Developing Country Debt and Economic Performance, Volume 2: Country Studies--Argentina, Bolivia, Brazil, Mexico 9780226733203

Citation preview

Developing Country Debt and Economic Performance

A National Bureau of Economic Research Project Report

Developing Country Debt and lxonornic Performance Volume

2

Edited by

Country StudiesArgentina, Bolivia, Brazil, Mexico

Jeffrey D. Sachs

The University of Chicago Press Chicago and London

JEFFREY D. SACHS is the Galen L. Stone Professor of International Trade at Harvard University and a research associate of the National Bureau of Economic Research.

The University of Chicago Press, Chicago 60637 The University of Chicago Press, Ltd., London 0 1990 by the National Bureau of Economic Research All rights reserved. Published 1990 Printed in the United States of America

99 98 91 96 95 94 93 92 91 90 5 4 3 2

1

@The paper used in this publication meets the minimum requirements of the American National Standard for Information Sciences-Permanence of Paper for Printed Library Materials, ANSI 239.48-1984

Library of Congress Cataloging-in-F’ublication Data (Revised for vol. 2) Developing country debt and economic performance. (A National Bureau of Economic Research project report) Papers presented at a conference held in Washington, D.C. on Sept. 21-23, 1987. Includes bibliographies and indexes. Contents: v. 1. The international financial system-v. 2. Country studies, Argentina, Bolivia, Brazil, Mexico. 1. Debts, External-Developing countries-Congresses. 2. Developing countries-Economic conditions-Congresses. 3. International finance-Congresses. I. Sachs, Jeffrey. HJ8899.D4815 1989 336.3’435’091724 88-20866 ISBN 0-226-73332-7 (v. 1 : alk. paper) ISBN 0-226-73333-5 (v. 2 : alk. paper)

National Bureau of Economic Research Officers Richard N. Rosett, chairman George T. Conklin, Jr., vice chairman Martin Feldstein, president and chief execurive officer

Geoffrey Carliner, executive director Charles A. Walworth, treasurer Sam Parker, director offinance and administration

Directors at Large John H. Biggs Andrew Brimmer Carl F. Christ George T. Conklin, Jr. Kathleen B . Cooper Jean A. Crockett George C. Eads Morton Ehrlich

Martin Feldstein George Hatsopoulos Franklin A. Lindsay Paul W. McCracken Leo Melamed Geoffrey H. Moore Michael H. Moskow James J. O’Leary

Robert T. Pany Peter G . Peterson Robert V. Roosa Richard N. Rosett Bert Seidman Eli Shapiro Harold Shapiro Donald S . Wasserman

Directors by University Appointment Charles H. Berry, Princeton Bill Brainard, Yale James Duesenberry, Harvard Ann F. Friedlaender, Massachusetts Institute of Technology Jonathan Hughes, Northwestern Saul Hymans, Michigan Marjorie McElroy, Duke

James L. Pierce, California, Berkeley Andrew Postlewaite, Pennsylvania Nathan Rosenberg, Stanford James Simler, Minnesota William S . Vickrey, Columbia Burton A. Weisbrod, Wisconsin Arnold Zellner, Chicago

Directors by Appointment of Other Organizations Richard Easterlin, Economic History Association Bruce Gardner, American Agricultural Economics Association Robert S . Hamada, American Finance Association Robert C. Holland, Committee for Economic Development David Kendrick, American Economic Association Eric Kruger, The Conference Board

Ben Laden, National Association of Business Economists Rudolph A. Oswald, American Federation of Labor and Congress of Industrial Organizations Douglas D. Purvis, Canadian Economics Association Dudley Wallace, American Statistical Association Charles A. Walworth, American Institute of Certified Public Accountants

Directors Emeriti Moses Abramovitz Emilio G. Collado Solomon Fabricant

Frank W. Fetter Thomas D. Flynn Gottfried Haberler

George B. Roberts Willard L. Thorp

Relation of the Directors to the Work and Publications of the National Bureau of Economic Research 1. The object of the National Bureau of Economic Research is to ascertain and to present to the public important economic facts and their interpretation in a scientific and impartial manner. The Board of Directors is charged with the responsibility of ensuring that the work of the National Bureau is carried on in strict conformity with this object. 2. The President of the National Bureau shall submit to the Board of Directors, or to its Executive Committee, for their formal adoption all specific proposals for research to be instituted. 3 . No research report shall be published by the National Bureau until the President has sent each member of the Board a notice that a manuscript is recommended for publication and that in the President’s opinion it is suitable for publication in accordance with the principles of the National Bureau. Such notification will include an abstract or summary of the manuscript’s content and a response form for use by those Directors who desire a copy of the manuscript for review. Each manuscript shall contain a summary drawing attention to the nature and treatment of the problem studied, the character of the data and their utilization in the report, and the main conclusions reached. 4. For each manuscript so submitted, a special committee of the Directors (including Directors Emeriti) shall be appointed by majority agreement of the President and Vice Presidents (or by the Executive Committee in case of inability to decide on the part of the President and Vice Presidents), consisting of three Directors selected as nearly as may be one from each general division of the Board. The names of the special manuscript committee shall be stated to each Director when notice of the proposed publication is submitted to him. It shall be the duty of each member of the special manuscript committee to read the manuscript. If each member of the manuscript committee signifies his approval within thirty days of the transmittal of the manuscript, the report may be published. If at the end of that period any member of the manuscript committee withholds his approval, the President shall then notify each member of the Board, requesting approval or disapproval of publication, and thirty days additional shall be granted for this purpose. The manuscript shall then not be published unless at least a majority of the entire Board who shall have voted on the proposal within the time fixed for the receipt of votes shall have approved. 5 . No manuscript may be published, though approved by each member of the special manuscript committee, until forty-five days have elapsed from the transmittal of the report in manuscript form. The interval is allowed for the receipt of any memorandum of dissent or reservation, together with a brief statement of his reasons, that any member may wish to express; and such memorandum of dissent or reservation shall be published with the manuscript if he so desires. Publication does not, however, imply that each member of the Board has read the manuscript, or that either members of the Board in general or the special committee have passed on its validity in every detail. 6. Publications of the National Bureau issued for informational purposes concerning the work of the Bureau and its staff, or issued to inform the public of activities of Bureau staff, and volumes issued as a result of various conferences involving the National Bureau shall contain a specific disclaimer noting that such publication has not passed through the normal review procedures required in this resolution. The Executive Committee of the Board is charged with review of all such publications from time to time to ensure that they do not take on the character of formal research reports of the National Bureau, requiring formal Board approval. 7. Unless otherwise determined by the Board or exempted by the terms of paragraph 6 , a copy of this resolution shall be printed in each National Bureau publication.

(Resolution adopted October 25, 1926, as revised through September 30, 1974)

Contents

Preface

X

Introduction Jeffrey D.Sachs

1

Book I

Debt and Macroeconomic Instability in Argentina Rudiger Dornbusch and Juan Carlos de Pablo

39

2. Some Debt History 3. From Martinez de Hoz to Alfonsin 4. The Process of High Inflation 5. The Austral Plan 6. Avenues and Obstacles to Growth Appendix A. Price Dynamics Under a Tablitu Regime Appendix B. The Budget and Inflation Appendix C. Statistical Data Notes References

41 54 64 77 91 115 131 133 134 147 150

Book I1

157

1. An Overview of Debt and Macroeconomic Problems

Bolivia’s Economic Crisis Juan Antonio Morales and Jeffrey D . Sachs

1. An Overview of Macroeconomic Performance 2. Political Economy and Macroeconomic Policymaking, 1952-87 vii

159 175

viii

Contents

3. State Capitalism and the Operation of the Public Sector 4. Trade Policies, 1970-85 5 . Aspects of Foreign Debt Accumulation, 1952-85 6. The Emergence of Hyperinflation, 1982-85 7. Ending the Hyperinflation, 1985-88 8. Bolivian Debt Management, 1985-88 9. Beyond Stabilization to Economic Growth and Development Notes References

Book I11

The Macroeconomics of the Brazilian External Debt Eliana A . Cardoso and Albert Fishlow

189 202 214 225 238 252 256 260 266

269

1. Introduction

27 1

2. Adjustment to the First Oil Shock: From Import Substitution to Macroeconomic Restraint

28 1

3. Adjustment in the 1980s: From International Monetarism to the Plano Cruzado

290

4. Stopping Inflation 5 . External Debt, Budget Deficits, and Inflation

305

6. Trade Policies and Consequences

335

7. Epilogue: Debt and Development Appendix. Brazilian Statistics

35 1 359

Notes

387

References

388

Book IV

Economic Policy and Foreign Debt in Mexico Edward F. Buf$e

318

393

1. Introduction

395

2. The Record of Stabilizing Development 3. Shared Development and the Echeverria Administration

398

4. The Lopez Portillo Administration 5 . The De La Madrid Administration and the Present Crisis

6. Import Compression, Real Wages, and Underemployment

417 428 447 465

7. Fiscal and Monetary Policy, Financial Intermediation, Inflation, and Growth 486

ix

Contents

8. Debt Management and Negotiations with Allen Sangines 9. Future Prospects: Is There A Way Out? Notes References

528

Biographies

553

List of Contributors

554

Name Index

555

Subject Index

559

517

536 547

Preface

This volume contains four country studies that were prepared as part of a research project by the National Bureau of Economic Research on developing country debt. Studies of Indonesia, Korea, the Philippines, and Turkey have been collected in a separate volume. In addition to the eight country studies, this project includes eight papers that examine other debt crises that occurred before World War 11, the role of the banks during the current crisis, the effect of developed country economies on the debtors, as well as possible solutions to the debt crisis. These papers have also been published in a separate volume. A fourth book contains shorter and less technical summaries of all sixteen papers. The findings of NBER’s debt project were presented at a conference for government officials of lending and debtor countries, economists at international organizations, and representatives of banks and other private firms with interests in the debtor countries. The conference was held in Washington, D.C. on 21-23 September 1987. We would like to thank the Agency for International Development, the Ford Foundation, Mr. David Rockefeller, the Rockefeller Brothers Fund, the Stan: Foundation, and the Tinker Foundation for financial support of this work. The success of the project also depended on the efforts of Deborah Mankiw, Yasuko MacDougall, Kirsten Foss Davis, Ilana Hardesty, Robert Allison, and Mark Fitz-Patrick.

Jeffrey D.Sachs

X

Introduction Jeffrey D. Sachs

1. Aims of the NBER Project on Developing Country Debt Latin America and Africa have suffered a collapse of living standards during the 1980s that in many countries rivals, and in some countries exceeds, the declines that were suffered during the Great Depression.’ The 1980s is widely regarded as “the Lost Decade” of economic development for large parts of the world, and even this appellation is too optimistic for many countries whose living standards have fallen back to the levels of the 1950s and 1 9 6 0 The ~ ~ collapse of living standards is intimately related to the external debt crisis that hit most of the countries of Latin America and Africa at the beginning of the 1 9 8 0 ~ As . ~ shown in table 1, economic performance-in terms of per capita growth, inflation, and the rate of capital formation in total output-has been particularly disastrous in those developing countries that experienced debt-servicing difficulties at the beginning of the decade. Few countries that fell into debt-servicing difficulties in the early 1980s have yet been able to extricate themselves from the financial crisis. Remarkably, at the end of 1988, after more than six years of the global debt crisis, not a single country in Latin America had regained normal access to loans from the private international financial market^,^ and countries such as Argentina, Brazil, and Peru were still caught in a dramatic process of collapsing incomes and exploding inflation. The NBER Project on Developing Country Debt was initiated in 1986 to improve our understanding of four fundamental issues concerning the debt crisis. First, what were the forces, both within the debtor countries and in the international financial system, that contributed to the onset of the debt crisis in so many developing countries? Second, what were the mechanisms by which the debt crisis contributed to the decline of living standards in the 1

2

Jeffrey D. Sachs

Table 1

Eeonomic Performance of Developing Countries with and without Debt-Servicing Difficulties

Per capita GDP growth 1970-79 1980-87 Average inflation rate 1970-79 1980-87 Gross capital formation 1980- 8 1 1982-87

With Debt-Service Problems

Without DebtService Problems

- .6

2.8

3.8 38

24.4 64.6

11.9 12.1

25.4 19.5

21.7 26.7

Source. IMF, World Economic Outlook, April 1988. GDP per capita, table A6, inflation, table A l I ; gross

capital formation, table A7.

Nores: Growth data are compound annual rates between the dates shown; inflation is measured as the nominal GDP-weighted average of the compound inflation rates of the countries in the category; gross capital formation is measured as the arithmetic average of the country ratios of gross capital formation of GDP.

debtor countries in the 1980s? Third, what are the reasons why recovery from the crisis has proved to be so difficult in many countries? And fourth, what are the lessons of the debt crisis more generally for the proper role of external borrowing in the long-term growth strategy of a developing country? To address these questions, the NBER project was divided into two parts: (1) a series of detailed country monographs, and ( 2 ) a series of analytical studies on several aspects of the international financial system. This volume (vol. 2 ) and a companion volume (vol. 3) present the monographs on the country studies. The analytical studies are contained in volume 1. A summary of the entire project may be found in Developing Country Debt and the World Economy. The eight countries examined in the NBER project (Argentina, Bolivia, Brazil, Indonesia, Korea, Mexico, the Philippines, and Turkey) were selected in order to shed maximal light on the comparative question of why some countries succumbed to financial crisis in the 1980s while others escaped the crisis. To provide a reasonable basis for comparison, all of the countries chosen are middle-income, capital-importing countries that enjoyed access to international commercial bank loans in the 1 9 7 0 ~Some .~ of these countries escaped the worst of the crisis of the 1980s, while others have virtually collapsed in response to the financial crisis. An overview of the very different economic performances of the eight countries can be found in table 2 . While a debt crisis is a multifaceted phenomenon, we will say that a country experienced a debt crisis if (1) the government was forced to reschedule the public or publicly guaranteed debts on an extraordinary basis; or ( 2 ) if the government went into substantial arrears on existing debt

3 Table 2

Introduction Economic Rrformance of the Eight NBER Countries Economic Variables by Country

1960-80

1980-85

2.2 2.1 5.1 2.6

-3.7 -4.8 - 1.1 - .9

Real per capita GDP growth:

Latin America Argentina Bolivia Brazil Mexico Rest of world Indonesia Philippines South Korea Turkey

3.4 2.8 7.0 3.6

2.3 - 3.4

6.0 2.5

1965-80

1980- 85

78.5 15.7 31.6 13.2

342.8 569.1 147.7 62.2

34.3 5.8 18.7 20.8

10.7 7.8 6.0 37.1

Inflation rate:

Latin America Argentina Bolivia Brazil Mexico Rest of world Indonesia Philippines South Korea Turkey

Sources: GNP growth rates, 1960-80, from the World Bank’s World Development Report, 1982; for 1980-85, from IhW, International Financial Statistics. Inflation rate is average annual rate of inflation from the World Bank’s World Development Report, 1987.

payments for more than ninety days. In all cases in which (1) or (2) occurred, the government also lost access to normal lending from the international capital markets. New lending, to the extent that it could be achieved at all, came only in the context of so-called concerted lending programs, in which existing creditors agree as a group to make new loans in proportion to their existing exposure. Table 3 shows the timing of debt-servicing difficulties and the dates of the first formal debt-rescheduling agreements. As shown in the table, Indonesia and Korea escaped the debt crisis entirely according to these criteria; Turkey fell into crisis in the late 1970s, bur overcame it by 1983, when the country no longer needed additional debt reschedulings and gained renewed access to international borrowing; and Argentina, Bolivia, Brazil, Mexico, and the Philippines fell into crisis in the early 1980s and had not recovered by the end of 1988. A useful measure of the depth of the debt crisis in each of the countries is shown in the last column of table 3, which records the secondary market value of the commercial bank debt of the eight countries, as of June 1988. The secondary market price is a simple index of a country’s creditworthiness, since it reflects the market’s expectation of the proportion of the

4

Jeffrey D. Sachs

Table 3

Debtor Government Relations with Commercial Bank Creditors for the Eight NBER Countries, 1W7-88 Date of First Debt-Service Difficulties

Date of First Rescheduling Agreement

1982 1979 1983 nonea 1982 1984 none 1978

1985 1980 1983 none 1983 1986 none 1979

Argentina Bolivia Brazil Indonesia Mexico Philippines South Korea Turkey

Date of Renewed Market Access

Nob No No na No No na 1983

Price of Debt, June 1988 27 13 53 na 53 55 na

99

Sources: The dates of the first debt-service difficulties are taken from the individual country monographs (the dates refer to the interruption of debt servicing on commercial bank debt). The timing of the debt rescheduling agreements may be found in the World Bank, World Debr Tables, 1987-88 edition, table IV-3, pp. xxxvi-xlii. Only ’hrkey has regained access to the private capital markets, as of 1983. See the Turkey monograph by Celasun and Rcdrik (vol. 3) for details. The price of debt refers to the asking price per $100 of par value of commercial bank debt, from Salomon Brothers, “Indicative Prices for Less Developed Country Bank Loans,” 9 June 1988. Note that there are no price quotations for Indonesia and South Korea.

“Indonesia experienced a limited external debt crisis in 1975, with foreign borrowing by the state petroleum company, Pertamina. This was not a generalized external debt crisis, and it was quickly resolved by 1977. See Woo and Nasution’s monograph on Indonesia (vol. 3). bNot yet renewed access.

na = not applicable

outstanding debt that a country will service in the long run. The most creditworthy nation, Korea, has bank debt that trades at par, reflecting very little fear in the international capital markets of a future default by these two countries. For this reason, Korea’s debt is not generally listed in quotations of secondary market prices for sovereign debt. Turkey’s secondary market price is also nearly at par, reflecting Turkey’s recovery of creditworthiness since the mid-1980s. The Latin American debt, by contrast, traded in June 1988 at about 50 percent of face value, and Bolivia traded at only 13 percent of par value, reflecting the fact that Bolivia experienced the deepest crisis in Latin America during the first half of the 1980~.~ Table 4 records the structure of the medium- and long-term external debt of the eight countries as of the end of 1986, both according to kind of creditor (i.e., whether the debt is owed to official creditors or to private markets, mainly the banks) and to the kind of debtor (i.e., whether the debt is a liability of the private or of the public sector). The creditor class of public and publicly guaranteed debt is available from the World Bank’s Debt Tables, but the creditor class of the private sector, nonguaranteed debt is not published. To construct the first two columns of table 4, therefore, it was necessary to assume that all debt of the private sector is owed to foreign private creditors. Note that in all of the countries, the bulk of the debt is owed by the public sector or is publicly guaranteed debt of the private ~ e c t o r This . ~ is a very

5

Introduction

Table 4

Structure of Medium- and Long-Term Debt by Category of Creditor and Debtor (billions of U.S. $, end of 1986) Creditor

Country

Private

Argentina Bolivia Brazil Indonesia

Debtor Official

Private

Official

38.4 (89) 3.5

(85) 82.5 (85) 31.9 (89)

Mexico

75.0

Philippines

19.8 (92)

South Korea

29.1 (85) 23.3

(82)

Turkey

(98)

Source: World Bank, World Debt Tubles, 1987-88 edition. To construct the table, it was assumed that all private nonguaranteed debt is owed to private creditors. The designation “official” in the debtor classification refers to debt of the public sector plus private sector debt with a guarantee by the public sector of the debtor country. Note: Proportion of total debt is in parentheses.

significant fact, with two profound implications. First, the debt signifies not only a burden on the country’s exports but also a burden on the fiscal resources, since government revenues must be used to service the public debt. Second, whereas an overhang of debt owed by private sector firms can be settled through bankruptcy and debt-to-equity conversions, writedowns, and buybacks arranged on an ad hoc basis, the public sector debt is not so easily discharged in the same manner. The available evidence suggests that, as a result, the private sector debt has been gradually extinguished through a number of ad hoc arrangements, while the public sector debt has been growing over time in most countries.8 The proportion of the total debt owed to official versus private foreign creditors differs significantly across countries. In Argentina, Brazil, Korea, Mexico, and the Philippines, most of the debt is owed to private financial markets (mostly banks), while in Bolivia, Indonesia, and Turkey, a higher proportion of the debt is owed to official creditors. The high proportion of official debt in these countries mainly reflects their lower per capita income, which made them less attractive to the banks for lending in the 1970s and made them more eligible and attractive for various forms of concessional official credits.’ Table 5 records the dramatic reversal of the net resource transfer from the private international capital markets (mainly banks) to the public sector of

6

Jeffrey D. Sachs

Table 5

Net Flows of Public and Publicly Guaranteed Debt, 1980-86 (as percentage of GDP, average per year) Country Argentina Net flows Net resource transfers Bolivia Net flows Net resource transfers Brazil Net flows Net resource transfers Indonesia Net flows Net resource transfers Mexico Net flows Net resource transfers Philippines Net flows Net resource transfers lbrkey Net flows Net resource transfers

1980-8 1

1982-84

1985-86

.3

2.7 .7

- 1.8

4.7 1 .o

- .3 -3.4

-.I - .5

1.9 .2

2.3 .2

- .3 -2.2

1 .o .4

I .5

3.0

1.1

2. I -1.8

1.7

1.9

.8

1 .o

.6

- .2

.7

.8

.4

- .7

2.4

.7 - .2

-.I -4.3

1.5

- .4 .7

- .4

Source: World Bank, World Debt Tables, 1987-88 edition

most countries after the onset of the debt crisis. The financial flows of public and publicly guaranteed debt from the private markets are measured in two ways in the table (in both cases as a percentage of the debtor country’s GDP). “Net flows” are loan disbursements minus loan amortizations during the year. The “net resource transfers” are new loan disbursements minus total debt servicing (interest plus amortizations). Thus, the net resource transfer is equal to net loans minus interest payments. After 1982, the Latin American countries were able to obtain new bank loans only as part of so-called involuntary or concerted lending packages (Bolivia received no funds in that form). In general, these concerted lending packages were more extensive during 1982-84 than later, so that the net resource transfers became more negative during 1985- 86 than earlier. Explanations of the Debt Crisis In examining why some of the countries succumbed to a debt crisis while others did not, an explanation can be attempted on several different levels. At the most superficial level, it is clear that the ratio of a country’s debt relative to its export earnings as of 1981 is a good, though by no means perfect, predictor of whether it fell into debt crisis during the 1980s. As shown in table 6, the countries with the highest ratios tended to be those which fell into debt crisis, while the two countries with the lowest ratios (Indonesia and South Korea) were those that escaped the crisis without any debt restructuring.

7 Table 6

Introduction Debt Ratios on the Eve of the Debt Crisis, 1981

Country Argentina Bolivia Brazil Indonesia Mexico Philippines South Korea 'hrkey ~~~~

Debt-GDP Ratio

Debt-Export Ratio

65.6 104.1 30.3 25.4 34.0 54.2 50.4 33.4

301.6 305.5 296.3 91.2 257.5 242.8 122.1 326.8

~

Source: All data are from the World Bank, WorldDebt Tables, 1987-88 editlon. Note: The data refer to total external debt (public plus private medium- and long-term debt, plus short-term debt).

Unfortunately, this finding (which has been reported in many econometric studies) doesn't really take us very far. Why did the debt-export ratio grow so rapidly in Brazil, but not in Korea? Interestingly, as shown in the table, Korea had as much debt as Brazil relative to GNP and yet maintained much lower debt relative to exports because of the high and rising share of exports in GNP. At the very least, we would like to know how the economic policies of the different governments affected the evolution of export promotion and debt accumulation. Even more deeply, we would like to understand how various underlying social and political factors (e.g., the distribution of political influence among competing social groups, the structure of government, the nature of electoral competition, etc.) have been important determinants of the economic policy choices of each of the governments. The NBER project seeks to shed light on the causes of the debt crisis at each of these levels of explanation, both the economic and the political. The attention to political factors might seem out of place to some readers. In the course of our research, however, it became very clear that the debt crisis is in large part a crisis of failed policy choices in the debtor countries, and that the policy choices themselves are best explained by appeal to important political as well as economic characteristics of the countries under study. In addition to focussing on the origins of the debt crisis, the country monographs examine the mechanics of a debt crisis once it is underway. They also reveal in considerable detail the profound economic and political difficulties that are encountered by countries attempting to recover from a large debt and a sudden cutoff of foreign lending. In the rest of this essay, I summarize the findings of the country monographs on these issues. Section 2 is a discussion of the origins of the debt crisis in a comparative perspective. In section 3 I describe the macrodynamics of the debt crisis in the countries worst hit by the external shocks in the late 1980s. In section 4 I use the country experiences to examine the pressing policy problem of how to recover from a debt crisis and to explore why recovery has proved so elusive in most of the Latin American countries. Section 5 provides a summary of the findings.

8

Jeffrey D. Sachs

2. Comparative Evidence on the Origins of the Debt Crisis The country studies volumes and the analytical studies in volume 1 shed considerable light on the origins of the debt crisis. As described in the International Financial System volume in chapters by Barry Eichengreen, and Peter H. Lindert and Peter J. Morton, debt crises have been episodic features of the world economy for the past 175 years. The wave of new lending to the LDCs that began in the early 1970s looks very much like the burst of new lending to the LDCs in the 1820s, 1870s, 1890s, and 1920s. In each episode, a wave of new lending was soon followed by a sharp retrenchment of lending and a widespread debt crisis, requiring substantial renegotiations of debt contracts and often involving widespread defaults. There is little doubt that major unforeseen external shocks at the end of the 1970s and early 1980s, especially the rise in world interest rates and the decline in the relative price of primary commodity exports of the debtor countries, played a major role in sparking the developing country debt crisis, as did similar macroeconomic shocks in the earlier historical episodes. At the same time, the NBER country monographs clearly reveal two additional facts about the crisis: (1) the effects of the external shocks on economic performance in a particular debtor country depended importantly on the quality of economic management in that country during the years when the borrowing was underway; and (2) even after the external shocks hit, in 1979 and 1980, there was time for policymakers to make adjustments in national policy to meliorate the shocks. Important adjustments were indeed made in some countries (e.g., Korea and Indonesia, and Turkey after 1979), but not in the Latin American countries or the Philippines. The external shocks were particularly important for the following reason. During the heady days of the 1970s when commodity export prices rose at annual rates that were greater than the interest rate on new loans, countries and their banks had the illusion of an unending Ponzi game (see my introduction to the summary volume for a further discussion of this point). Countries that borrowed all that they needed to repay interest and principal on past loans had a debt that grew at the rate of interest. As long as nominal export earnings were growing even faster than this (which was the case because of the steady rise of export prices), it was possible to borrow all of the funds needed to service past debt and at the same time enjoy a falling debt-to-export ratio. It seemed too good to be true: loans could be paid back out of new loans without provoking a rising debt-export ratio. And indeed, as shown in table 7, despite the heavy borrowing of the 1970s, the developing countries as a whole had a debt-export ratio at the end of 1980 that was about the same or even lower than in 1973! To the countries and to the banks, a money machine seemed to be at hand." Once interest rates rose and export prices started to fall, the debt-export ratios shot up dramatically. As shown in table 7, the increase between 1979

9

Introduction Debt-Export Ratios of Developing Countries, 1973-87

Table 7 Country Grouping

1973

1980

1982

1987

All nonoil developing countries Western Hemisphere Asia

115.4 176.2 92.9

112.9 178.4 68.2

148.3 271.8 87.1

155.1 346.3 89.1

Source: For 1973 and 1980 IMF, World Economic Outlook, 1983, table 33. For 1982 and 1987: World Economic Outlook, 1989, table A48. Note that for the category “nonoil developing countries” for 1982 and 1987, I referred to the category “nonfuel exporters’’ in table A48.

and 1982 was dramatic, particularly in Africa and Latin America. Suddenly, the old Ponzi scheme was no longer working. Unless countries actually began to service the debts out of their own resources, i.e., by running trade surpluses, the debt-export ratios were bound to explode, and indeed they started to. In a sense, it was not until the rise of interest rates and fall of export prices at the end of the 1970s that the most basic question on the debt was posed for each country: Could and would the country repay its debts out of national income, rather than out of new borrowing? It is easy to show that size of the external shocks on a country-by-countrybasis was not decisive in answering this question.” Consider in table 8, for example, the size of the terms-of-trade shocks in each of the eight NBER countries. The terms-of-trade shock is measured as the change in each country’s terms of trade (comparing the average for 1980-82 with the average for 1977-79) multiplied by the 1980 share of total imports in GNF? This measure roughly indicates the income loss as a percentage of GNP that resulted directly from the fall in each country’s export prices relative to import prices. Five of the eight countries experienced a terms-of-trade deterioration if 1980-82 is compared with 1977-79, but there is no relation between the size of the shock and the depth of the subsequent crisis. Bolivia and Mexico, for example, fell into a deep debt crisis despite an improvement in the terms Table 8

Terms-of-Trade Shocks, 1979-83 Terms of Trade

Country

1979 (1)

1983 (2)

Argentina BoI iv i a Brazil Indonesia Mexico Philippines South Korea Turkey

102 77 114 73 77 I12 127 125

96 99 87 97 98 99 101

94

(3)

Share of Exports inGNP, 1981 (4)

.94 I .29 .76 1.32 1.27 .88 30 .75

13 17 9 31 14 20 37 7

(2)

t

(1)

Size of Shock ( 5 ) = [(3) - (1)1 x (4)

- .8 4.9 -2.2 9.9 3.8 -2.4 -7.4 - 1.8

Sources: Terms-of-trade data from the World Bank’s Work Development Report, 1984 edition for 1979 data and 1987 edition for 1983 data. Share of exports in GDP from table 9 in this chapter.

10

Jeffrey D. Sachs

of trade. It is true that by 1982, export prices were beginning to fall in these two countries, but judged over a ten-year perspective, export prices were high, not low. On the other hand, Korea suffered a sharp terms-of-trade decline after 1979, but it largely escaped the debt crisis. It is apparent that other country-specific factors must have played a large role in determining the effects of the external shocks on the individual economies. The NBER country monographs suggest that the following country-specific factors were most important: 1. The shocks were less harmful in countries which had previously adopted a successful long-term growth strategy, based on an outward-oriented trade regime, a competitive exchange rate, and prudent $fiscal policies. 2 . The shocks were less harmful when foreign borrowing had contributed to a higher rate of capital accumulation in the 1970s, as opposed to a higher level of consumption spending or greater capital flight. 3. The shocks were less harmful when the government had adjusted rapidly to the external shocks in the early 1980s, rather than postponing adjustment measures. I now examine each of these points in detail.

The Long-Term Growth Strategy Trade and Exchange Rates The benefits of outward-oriented trade policies have now been firmly established by the economics literature, and the monographs in the NBER project lend further support to the existence of long-term benefits from an outward-oriented trade strategy.l 2 Outward orientation refers to the balance of incentives given by the trade regime to export sectors versus importcompeting sectors. Outward-oriented regimes are typically defined as regimes that are neutral between the sectors or on balance more favorable to exports. Importantly, the authors of the monographs for Korea and Turkey stress that in practice, outward-orientation can be very different from laissez-faire. Korea, for example, had a very interventionist trade policy with heavy import protection, but it has balanced the anti-export biases of the import restrictions with highly favorable fiscal incentives for exports. Table 9 provides some indicators of the trade regime of the various countries in the NBER project.I3 The table reports a World Bank index of trade orientation for 1973-77 and 1978-85, as well as the export-GNP ratios for the economies for 1960, 1980, and 1985. Korea has the highest index of outward orientation of the economies and, consistent with that, has had the most marked increase in the export-GNP ratio of all eight economies. Turkey moved from a posture of moderate inward orientation in the 1970s to one of moderate outward orientation in the 1980s, according to the World Bank index, and the export-intensity of the economy increased

11

Introduction Indicators of Trade Structure and Policy

Table 9

Index of Trade Regime Country Argentina Bolivia Brazil lndonesia Mexico Philippines South Korea Turkey

Share of Primary Exports in Total

Export Share of GNP

1963-73

1973-85

1981

1960

1980

1985

1 2 3 3 2

1 1 3 2 2 2 4 3

80 97” 59

10 13 5 13

13“ 17 9 31 14 20 37 7

15

2 4 1

96 73b 55

10 63

10 I1 3 3

18 14 23 16 22 36 19

Sources. The trade orientation index varies from 1 to 4, with 1 being the most inward oriented and 4 being the most outward oriented. The index is taken from the WorldDevelopment Report (1987, 83) of the World Bank. The share of primary exports in total exports is from the World Developmenr Report, table 10, various years. The export-GDP ratios measure the exports of goods and nonfactor services as a percentage of domestic GNP. They are from various editions of the World Developmenr Report: 1960 data from 1981 edition; 1980 data from 1982 edition; 1985 data from 1987 edition.

“1979. b1983

sharply after 1980, apparently in line with the shift in trade regime. Interestingly, Turkey is also the clearest case of recovery from a debt crisis, a recovery that Merih Celdsun and Dani Rodrik (vol. 3) attribute importantly to the boom in export earnings after 1980. All of the Latin American countries are judged to be moderately or extremely inward oriented, except for Brazil during the 1970s, which was designated by the World Bank as moderately outward oriented. No Latin American country showed an important rise in the share of manufacturing exports in GNP during the 1970s; the rise in Mexico’s export-GNP ratio between 1970 and 1980 is more than accounted for by the increase in oil exports as a percentage of GNP. The benefits of an outward-oriented trade strategy, especially for a country engaged in extensive foreign borrowing, come through very clearly in the country monographs. With an inward-oriented trade strategy, foreign borrowing is directed to privately profitable but socially unprofitable sectors, leading to a buildup of inefficient import-competing industries that eventually prove to be unable to generate the foreign exchange necessary to service the accumulated foreign debt. This point, which was spelled out in simple theoretical terms by Brecher and Diaz Alejandro (1977), is amply demonstrated by the NBER monographs. l4 Once debt repayments became necessary in Latin America, the foreign exchange earnings could be generated only with tremendous political and economic difficulty since the manufacturing sector could only begin exporting with a large cut in real wages.15 The protected manufacturing sectors in Latin America rely heavily on imported inputs in the production process. When the terms of trade

12

Jeffrey D. Sachs

deteriorated in the early 1980s, and when new lending stopped from international capital markets, the imported inputs also dried up (or rose sharply in price relative to nontraded goods), forcing much of the manufacturing sector into a sharp contraction.16 These firms could stay in operation only with very substantial cuts in real wages. The outward-oriented regime in Korea, by contrast, avoided many of the static inefficiencies of inward orientation, reduced Korea's vulnerability to external shocks, and also offered Korea various dynamic benefits, including the opportunity for export firms to exploit economies of scale and opportunities for learning-by-doing by producing for a large world market. The outward-oriented trade strategy, with its focus on spurring export growth, seems to offer the additional benefit of forcing policymakers to pay careful attention to exchange rate management. In Korea, as well as in Indonesia, and in Turkey after 1980, policymakers have devalued the currency in a timely fashion in order to avoid a costly loss of international competitiveness for the export firms. In Latin America, by contrast, devaluations are typically avoided until the last possible moment. As long as the central bank has a minimal level of reserves or access to international loans, nominal exchange rates are held fixed despite internal cost i n f l a t i ~ n . ' ~The result is that exporters are squeezed by an overvalued real exchange rate. This kind of chronic overvaluation adds a further barrier (in addition to the trade regime) to the development of a dynamic nontraditional export base. The Latin American tendency toward overvaluation on the official exchange rate at the end of the 1970s and the early 1980s (after the global shocks) is illustrated in table 10, where I show a measure of the real exchange rate vis-a-vis the United States for the eight NBER countries, comparing 1979-82 with a base of 1978. We can see that in Argentina, Bolivia, Mexico, and the Philippines, the real exchange rate appreciated in real terms, while it was stable or depreciating in Brazil, Indonesia, South Korea, and Turkey.

''

Table 10 Country

The Real Exchange Rate, 1978-82 (local currency vis-a-vis the U.S., 1978 = 100) 1978

1979

1982

Average (1980- 8 I )

138

59

I34

107 77 80 85 101 89

159 123 78 81 122 I07 95

60

78

1981

1980 ~

Argentina Bolivia Brazil Indonesia Mexico Philippines South Korea Turkey

100 100 I00 100 100 100 100 100

141 103 92 78 106 105 106 111

179 111 76

80 81 I27

81

117 108 96 84

I05 94 71 ~

~~

Source: IMF, International Financial Statistics.

Note: The real exchange rate v i s - h i s the U.S. is calculated as P/EP*, where P is the CPI of the country, E is the exchange rate in units of domestic currency per U.S.$, and P* is the U.S. CPI. A rise in the index signifies a currency appreciation.

13

Introduction

The inward-oriented trade structure in Latin America bolsters, in a political sense, the tendency toward an overvalued exchange rate. Since exporters who favor devaluations are limited mainly to the primary commodity sectors, devaluations in Latin America are typically opposed by most of the influential economic actors. There is no countervailing pressure for timely devaluations from a large manufacturing export sector, as there is in Korea and, increasingly, in Indonesia and Turkey.” Urban workers and firms operating in the sheltered manufacturing sector view devaluations as serving mainly to raise domestic costs and reduce real wages and to increase the rents earned by primary commodity exporters, whose supplies are deemed (incorrectly) to be inelastic in any event. The policy of maintaining fixed nominal exchange rates in Latin American countries, despite an overvaluation of the currency, was an important reason for foreign borrowing by the public sector. In Argentina, Bolivia, and Mexico, in particular, the central banks borrowed heavily in the late 1970s and early 1980s to gain foreign reserves in order to support the exchange rate in the face of massive private capital flight, which in turn was prompted by fears of an impending devaluation (as well as more generalized fears of political and economic instability).’O In the period of heaviest foreign borrowing, 1976-85, it has been estimated that about two-thirds of the increase in gross external debt in Argentina and Mexico went to finance private capital flight, as is shown in table 11.21 The bulk of this capital flight was concentrated before 1983, during this period in which the monetary authorities were holding fixed an overvalued exchange rate. In Brazil, where capital control restrictions were in place and where the exchange rate did not become so overvalued, private capital outflows were significantly smaller than in Argentina and Mexico. In Indonesia and Korea, capital flight was a much smaller fraction of total borrowing, largely because real exchange rates did not become highly overvalued in the critical period, 1979-82.” In Indonesia, ’Lgble 11

Estimates of Capital Flight, 1976-84 Change in Gross Debt

Country Argentina Boliviaa Brazil Indonesia” Mexico Philippines South Korea

(1)

36.3 3.0 79.3 27.0 79.4 19.4 33.2

Estimated Capital Flight (2)

Ratio (2)Nl)

25.0

.69 .33

17.3 5.0 53.4 3.7 3.5

.22

1 .o

.I9 .67 .I9 .I1

Sources: For Argentina, Brazil, Mexico, the Philippines, and South Korea, from R . Cumby and R . M. Levich, On the Definition and Magnitude of Recent Capital Flight, NBER Working Paper no. 2275 (June 1987), tables 1-5. The “Morgan” definition of capital flight is the one repeated here. For Bolivia and Indonesia, from Morgan Guaranty, World Financial Markets (March 1986). table 10, p. 13. Note that these sources did not contain estimates for Turkey. ‘For Bolivia and Indonesia, data are for 1976-85.

14

Jeffrey D. Sachs

for example, there was little capital flight between 1976 and 1985, despite the fact that the country has completely free international capital mobility, a point noted by Wing Thye Woo and Anwar Nasution (vol. 3).23 Another, and more subtle link between the exchange rate and foreign borrowing results from the fact that in several Latin American economies, the public sector is a large net exporter, while the private sector is a large net importer. In Bolivia and Mexico, for example, the major export earnings are garnered by state enterprises that produce primary commodity exports. In this circumstance, an overvaluation of the exchange rate can lead directly to a worsening of the budget deficit and thereby to increased foreign borrowing by the public sector.24 The Fiscal Balance

More generally, the management of fiscal policy, along with trade policy, was the second main determinant of which countries succumbed to a debt crisis. The importance of fiscal policy results from the fact that most of the foreign borrowing in the 1970s was undertaken by the public sector or by the private sector with public guarantees. We saw in table 4 that more than 80 percent of the external debt at the end of 1986 was held by the public sector.= The heavy indebtedness of the public sector poses special problems for debt servicing since the public sector has to generate resources in order to service the debt. The foreign debt servicing in this case therefore requires two resource transfers: first, from the private sector to the public sector, and then from the public sector to the rest of the world. The first kind of transfer requires a coherent set of fiscal policies; the second, an outward-oriented trade strategy and realistic exchange rates. The centrality of fiscal policy in the debt crisis is emphasized in all of the country monographs. In Latin America, Turkey, and the Philippines, governments ran chronically large budget deficits in the years leading up to the debt crisis, and often relied on foreign borrowing to finance current spending as well as capital expenditures. Moreover, the deterioration of the terms of trade in the 1980s caused a large loss of revenues, while higher interest rates raised the cost of debt servicing, so that the external shocks pushed the economies toward large deficits. Budget deficits were large at the end of the 1970s in all of the Latin American countries and rose very sharply in the early 1980s, after the rise in world interest rates and the fall in export prices. As I will discuss later, the failure of the Latin American countries to close the gaping budget deficits after 1982, combined with the inability to finance the deficits with foreign borrowing, is the explosive mix that has fueled the high inflations throughout the region in recent years. The Uses of Foreign Borrowing The trade regime and fiscal policy management played a large role in determining the ways that foreign borrowing was utilized. In Korea, for

15

Introduction

example, foreign borrowing largely financed investments in heavy industry that provided the basis for the export surge in the mid-1980s. It is safe to conclude that the foreign borrowing financed investment spending rather than consumption, since domestic saving rates rose throughout the whole period of heavy foreign borrowing in the 1970s. In the inward-oriented regimes in Latin America, foreign borrowing was much more likely to support consumption, capital flight, and investment in nontradables, than it was to augment the manufacturing export base. As we saw in table 1 1 , foreign borrowing was associated with heavy capital flight in three of the Latin American countries: Argentina, Bolivia, and Mexico. In Argentina, it is hard to identify any broad sector of the economy in which investment expenditure was spurred as a result of foreign borrowing, except for a binge of consumer durables purchases on the eve of the exchange rate crisis of 1981. In Mexico it appears that the small share of foreign borrowing that did not finance capital flight was used to increase government transfers and government investment in the oil sector, much of which proved to be unprofitable when oil prices started to fall after 1981. In Bolivia, the foreign borrowing financed extensive capital flight, some worthwhile development of the natural gas sector, and several major “white elephant” projects which have since been abandoned. In Brazil, domestic savings fell so sharply in the 1970s that despite the heavy foreign borrowing in that decade, national investment rates were lower in the late seventies (during the borrowing boom) than in the early seventies. Most of this decline was due to a fall in public sector savings as a percentage of GNP (i.e., a larger public deficit on the current account of the budget), suggesting that the borrowing really served to finance current government expenditures, such as subsidies on energy consumption. In Turkey, much of the surge in foreign borrowing was associated with the convertible Turkish lira deposit accounts (CTLDs) discussed by Cellsun and Rodrik. This financial mechanism effectively gave large subsidies to investment projects financed with foreign loans. Thus, Turkey sustained a large investment boom as the counterpart of the foreign borrowing. It is evident that much of the investment was rather inefficient, but at the same time, Cellsun and Rodrik suggest that the increased investment played an important role in generating manufacturing capacity for Turkey’s export boom in the 1980s. In the Philippines, heavy foreign borrowing was used to finance a significant increase in public sector investment, but much of this expenditure proved to be highly inefficient. As Robert S . Dohner and Ponciano Intal, Jr. stress: The expansion of investment had a particularly large construction component, much of it in public or quasi-public facilities: luxury hotels, cultural centers, and some of the notorious projects of Imelda Marcos such as the villa built entirely of coconut produc or the University of Life.

16

Jeffrey D. Sachs

Other investments were hurt by changes in world demand and prices. These include major investments in copper refining, sugar mills, and arguably, the nuclear power plant that the Philippines built. (1989, 176) Favoritism in allocating loans, kickbacks on government procurement, and the general use of public investment funds to enrich the Marcos “cronies” all contributed to giving the Philippines the highest ICOR (incremental capital output ratio) in Southeast Asia. Finally, with respect to Indonesia, Woo and Nasution stress that foreign borrowing was kept at a moderate pace (indeed the 1981 debt-GNP ratio for Indonesia was the lowest for the eight NBER countries), was tightly screened, and was used in a balanced manner to support export-oriented agriculture as well as industry for the domestic market. Capital flight was moderate despite the complete absence of capital controls. Woo and Nasution stress that political considerations under Soeharto have led to a consistent emphasis on support for the rural agricultural sector,26 which in turn has led the government to avoid an overvalued exchange rate which would tend to squeeze the agricultural sector. Policy Adjustments to External Shocks, 1979-82 The rise in world interest rates and the fall in commodities prices in 1979 and after caught most policymakers throughout the world off guard. Real interest rates had been negative between 1973 and 1978, and they were widely forecasted in 1979 to remain low or negative for many years into the future. Similarly, commodities prices boomed again in 1979 as they had in 1974, and there was little basis for believing that they would soon collapse to levels in real terms that had not prevailed since the Great Depression. Nonetheless, already in late 1978 there were enough clouds on the economic horizon to cause prudent policymakers in developing countries to pause from an aggressive strategy of foreign borrowing. Upon Paul Volcker’s accession to power as chairman of the Federal Reserve Board at the end of 1978, short-term dollar-denominated interest rates (e.g., LIBOR) began to rise sharply-part of a strategy to stabilize the dollar-reaching what was then a postwar high of fifteen percentage points on six-month loans in June 1979. Real interest rates (the interest rate minus the inflation rate of the past twelve months) also reached the highest levels of the past several years by mid-1979. Moreover, several important borrowing countries had already begun to experience debt difficulties: Peru, as of 1976; Turkey, as of 1977; and Jamaica, as of 1978.” It is fascinating to compare the policy reactions of the various countries to the shifting economic environment in 1979-80. These reactions were very important in affecting the later development of the crisis, since in Latin America in particular, a remarkably large proportion of the total debt as of 1982 had been incurred in just two years, 1980 and 1981. We also saw earlier that it was in these few years that the debt-export ratios jumped to their dangerous levels. As pointed out in Sachs (1987), the net debt of

17

Introduction

Argentina, Brazil, and Mexico to BIS-area banks nearly doubled in the two years between December 1979 and December 1981.” On the whole, Korea, Indonesia, and Turkey responded appropriately to the deteriorating international situation, while the Latin American countries and the Philippines either ignored the ominous signs from abroad or even accelerated their foreign borrowing in a misguided attempt to counteract the recessionary effects of the external shocks. It is instructive to give a schematic description of the policy choices undertaken in various countries during this crucial period. Characteristically, Korea proved to be the most prudent of the countries in the study. As early as spring 1979, Korea embarked on a Comprehensive Stabilization Program (CSP) and a retrenchment from the high rates of borrowing of the mid-to-late 1970s. The CSP included a turn to tighter monetary and fiscal policy, as well as a reallocation of investment from heavy industry to other less capitalintensive sectors. Nineteen eighty proved to be a deep crisis year for the economy (as a result of internal political unrest, a failed harvest, external shocks, and the internal adjustment effort). Nonetheless, the stabilization program was maintained and strengthened, in part through a significant depreciation of the exchange rate and a further tightening of fiscal policies. Indonesia was also relatively well positioned in the crucial period of 1979-82. In 1978, Indonesia devalued the rupiah significantly in order to spur nontraditional exports. Moreover, Indonesia enjoyed a terms-of-trade boom in 1979-80, in line with the second OPEC oil shock. Because of moderate fiscal policies during 1979- 82, Indonesia actually decreased its external debt net of foreign exchange reserves in that period. Turkey was “lucky” in having experienced its debt crisis as early as 1977-78. By January 1980, the government had decided to launch a major policy initiative to open the Turkish economy and expand manufacturing exports. The real exchange rate was significantly devalued, which greatly enhanced the export profitability for Turkish manufacturing firms. A military government seized office at the end of 1980 and continued the policy adjustments. Labor unrest was repressed, thereby preventing wage increases from undoing the real exchange rate depreciation. At the same time, Turkey was fortunate to receive substantial financial support from the official creditor community. In the Philippines, the needed adjustments were not put in place. As Dohner and Intal indicate, the Marcos government attempted to counter the recessionary effects of the second oil price increase by increasing public expenditures. The government actually embarked on a new strategy of Major Industrial Projects in 1980, causing a significant rise in the public sector deficit and a significant acceleration of foreign borrowing. As in Latin America, the foreign debt nearly doubled between 1979 and 1982. In the four Latin American countries, there is not a single case of significant retrenchment during 1979-82. Bolivia was in no condition politically to embark on a stabilization effort. In the tortuous path from

18

Jeffrey D. Sachs

military government in 1978 to democratic government in 1982, there were no less than thirteen heads of state! No government lasted long enough to implement a stabilization program, and only one really tried (for a short period in 1980). Bolivia fell into arrears with the commercial banks as early as 1980. In Argentina, the political situation was also difficult. The fixed exchange rate policy of Martinez de Hoz, in place since late 1978, was still being defended in 1979-80 as the best hope for fighting inflation. By 1980, the sharply overvalued exchange rate was provoking enormous capital flight, but the exchange rate was not devalued until February 1981. The period between March 1981 and the advent of democratic rule at the end of 1983 (with the accession to power of President Radl Alfonsin) was primarily one of political and economic disarray, capped by Argentina’s defeat in the Malvinas War in the spring of 1982. The policies in Brazil stand in interesting contrast to those in Korea. In spring 1979, at about the same time that Korea embarked on its stabilization program, Brazil was undergoing a presidential succession, from Ernest0 Geisel to Jog0 Baptista Figueiredo. The new planning minister, Mhrio Simonsen, embarked on a strategy of slower growth, reduced budget deficits, and less reliance on foreign borrowing. But instead of persevering as in Korea, these austerity measures were quickly abandoned, in part paradoxically as a response to the rise of oil prices in mid-1979. Rather than viewing the oil prices as yet another reason for retrenchment (as Brazil was a heavy oil exporter), the oil price increases were taken by many as a sign that new “stimulative” measures were necessary. In the event, Simonsen was replaced by Ant8nio Delfim Neto, who quickly ushered in a policy of accelerated foreign borrowing. Mexico provides a particularly remarkable case of failure to adjust. Between 1979 and 1981, the Mexican public sector increased expenditures so dramatically that Mexico ended up in a foreign debt crisis in 1982 despite the historically unprecedented boom in oil export earnings during 1980 and 198 1 . Government spending rose from 32.2 percent of GDP in 1979 to 46.4 percent in 1981! This increase of 14 percent of GDP was accompanied by a rise in revenues of only 4 percent of GDP (from 26.2 to 30.1 percent). It is interesting to speculate on the deeper factors that might explain the Latin American (and Philippine) failure to pursue more realistic fiscal policies. In another study, Andrew Berg and I (1988) surmised that the deep inequalities of income in Latin America and the Philippines have probably been an important systematic factor in undermining a sound fiscal policy. We hypothesized that higher income inequality would have the following effects, all making budgetary control more difficult:

. .

raise the pressure for redistributive policies toward the poor and working class; enhance the power of economic elites to resist taxation;

19

Introduction

decrease the political legitimacy of governments that defend the existing distribution of income; contribute to direct labor militancy; more generally, impede the development of a social consensus around policies that promote development in the long term, but which might impose costs on some social groups in the short term. Berg and I offered some circumstantial evidence in support of our hypothesis by showing that the extent of income inequality in a sample of thirty-four middle-income developing countries was a good predictor of which countries in the sample had succumbed to a debt crisis and which ones had not.

3. Macroeconomic Adjustments After the Onset of the Debt Crisis The previous section addressed the question of why some countries fell into crisis and others did not. We saw that the differences among the countries related to several factors, including the trade regime of each country, the exchange rate policies, the fiscal management, and the reaction of the various countries to the external shocks of 1979-82. In this section, I review some of the lessons in the country monographs concerning the macrodynamics of a debt crisis, starting from the moment that foreign lending is cut off to the overly indebted country. My focus here is on the experience of the Latin American countries, which suffered the deepest debt crises, though I also mention the cases of Turkey and the Philippines. I also explore the reasons that recovery from the crisis has proved to be so difficult. In order to understand the economics of a debt crisis, it is important to review the basic theory of macroeconomic adjustment to external shocks. Consider, for purposes of discussion, a simple economy with three productive sectors: a primary resource sector (e.g., oil, grains, primary metals), a tradables manufacturing sector, and a nontradables sector (which includes services, construction, and manufacturing that operates behind a high tariff barrier). The sectors will be denoted R, T, and N, respectively. The economy imports final consumption goods, final investment goods, and imported intermediate goods used in domestic production of R, T, and N. The relative importance of the three sectors in a particular economy will depend not only on the resource endowments of the economy, but also on the trade policies that have been pursued in the country over the preceding ten or twenty years. Outward-oriented trade policies will be more likely to result in a large tradables manufacturing sector, with significant levels of manufacturing exports, as in Korea. Inward-looking policies will tend to generate a manufacturing sector that is almost completely protected from world markets and is therefore really part of the nontradables sector, as in Argentina. Of course, the size of the manufacturing export sector will also depend on factor endowments. Resource rich economies will tend, ceteris paribus, to have a

20

Jeffrey D. Sachs

smaller manufacturing export sector since the export of primary commodities will maintain a strong exchange rate, which will crowd out the manufacturing export sector.29 I will suppose that in the short run of the production of the primary commodity R is inelastically supplied, while production in the other two sectors is determined by production functions that use sector-specific capital, imported intermediate inputs, and labor. Labor can move between the sectors in the short run (though perhaps with some adjustment lag). An appreciation of the real exchange rate (defined as a rise in the price of nontradables N relative to tradables 7') causes N firms to bid up wages and attract labor from the T sector. A rise in domestic aggregate demand (e.g., a fiscal expansion that is financed by foreign borrowing) pushes up demand for nontradables and tradables. Since the supply and demand for nontradables must be equal, the relative price of nontradables to tradables must rise (i.e., there must be a real appreciation), and labor is drawn from the T sector to the N sector. The urban real wage will tend to rise with an exchange rate appreciation (i.e., a rise in the relative price on N) and to fall with a depreciation. This will be true, for example, when the nontradables sector is relatively labor-intensive compared with the resource sector and the T sector is small (the typical pattern in the Latin American countries). If there is a large labor-intensive tradables sector (as in Korea, for example), the negative effect of a real depreciation on the real wage will tend to be smaller than in the case of Latin America, where the effect can be quite extreme.30 Also, the responsiveness of exports to a given percentage fall in real wages will tend to be lower in Latin America than in Asia because in Latin America the main export sector is the resource sector, which has a very low short-run supply inelasticity, while the more responsive T sector is much smaller in Latin America than in Asia as a percentage of total output. For this reason, the extent to which real wages must decline in order to increase exports by a given percentage of GDP will tend to be much larger in Latin America than in Asia. Exogenous shocks which reduce current aggregate expenditures on nontradables will tend to cause a depreciation of the real exchange rate and consequently a decline in real wages. For example, a fall in the world price of the R sector or a rise in interest costs for a net debtor both lead to a negative income effect and a decline in demand for nontradables. Similarly, in a case where foreign borrowing from international capital markets is supporting domestic spending above domestic production, a cutoff of the foreign lending will cause a drop of absorption, with a consequent fall in the price of nontradables relative to tradables and a decline in the real wage. Analogously, a country that is financing a trade deficit by running down foreign exchange reserves will experience a collapse of absorption and a real exchange rate depreciation at the moment that the reserves are depleted

21

Introduction

(assuming that there is no recourse to other forms of international capital inflow). This discussion highlights one central theme. The external shocks that hit the NBER countries (a fall in the terms of trade, a rise in world interest rates, and for five of the countries, a collapse of international lending) were of the sort that: (1) reduced domestic absorption; (2) provoked a real exchange rate depreciation; and (3) thereby provoked a decline in real wages.31 In economies with a very small T sector (such as Argentina, Bolivia, Mexico, and Turkey), the reduction of the real exchange rate and of the real wage had to be very large in order to induce a significant increase in the supply of new nontraditional exports to compensate for the terms-oftrade decline in the R sector and for the higher interest rates and the cutoff of new funds from abroad. This discussion of adjustments to external shocks, while conventional, does not do justice to the real-world drama that lies behind the adjustment to external shocks. In practice, the adjustment to adverse external shocks is likely to be politically destabilizing, both because the shocks tend to require reductions in real wages that can provoke social unrest and because the shocks tend to require governments to adopt fiscal austerity measures that are politically unpopular and sometimes even threatening to the survival of the government. In practice, governments may have very grave difficulties in convincing the general public that the austerity is the result of external forces rather than domestic mismanagement. In some countries, particularly Argentina and Bolivia, the real wage declines that followed the external shocks were vigorously resisted by powerful labor groups and by political groups acting on the behalf of urban labor interests. In these countries, the external shocks have provoked enormous labor unrest, including several general strikes since the early 1980s, as well as the periodic application of wage regulations (such as indexation agreements) that attempt to reduce or eliminate the downward movement of the real wage. The floor on real wage adjustments can of course contribute to open unemployment or more likely to the growth of an informal labor market, where the regulations are not honored and where real wages fall sharply. In other countries, especially in Brazil, Turkey, and Korea, which were all under military rule in the period 1980-83, real wage resistance was more readily countered by government repression of labor unions and pro-labor political parties. 32 In Indonesia, the Philippines, and Mexico, union activity is traditionally weak because of a long history of government repression or co-optation. The second kind of real-world adjustment difficulty arises from the fact that the external shocks directly and adversely affect government finances, thereby requiring politically painful fiscal adjustments by the governments. This is true of all three kinds of shocks that I have been discussing: terms of

22

Jeffrey D. Sachs

trade declines, higher world interest rates, and a cutoff in new lending. Governments of most countries in the NBER project rely on primary commodity exports as a major source of government revenues,33 so that a sharp drop in export earnings will cause a significant drop in government Even where taxes are levied on imports rather than exports, the decline of export earnings following a terms-of-trade deterioration can provoke a sharp drop in imports and thereby a sharp drop in government revenues. 35 Similarly, higher world interest rates directly affected the government budget since much of the foreign debt was an obligation of the public sector: higher world interest rates coupled with the heavy stock of foreign debt therefore contributed directly to an increase of the budget deficit. For the same reason, a cutoff in new loans also affected the government directly, since the lending cutoff generally affected loans to the public sector and therefore required that the government reduce a deficit that it had been financing with foreign loans or instead shift to other forms of finance (often with dire consequences). The practical implications of the link between government revenues and the external shocks are enormous. When the terms of trade deteriorates, or interest rates rise, or lending is cut off, there is not an automatic decline in domestic spending to the extent that these shocks hit the public sector. Rather, the government faces an explicit policy choice of absorbing the shock through a combination of reduced expenditures, higher taxes, or a larger budget deficit. The effects on the economy, and on the political fortunes of the government, are closely linked to the choices that it makes. The “textbook” approach to a permanent adverse shock, such as a lower terms of trade or higher interest costs, is to cut spending and raise taxes sufficiently to absorb the shortfall in government revenues. The “optimal” response to a cutoff in the availability of foreign credits is in general more complicated, though it will typically involve a reduction in overall borrowing, as well as a partial shift in financing from foreign sources to domestic sources. Typically, the necessary austerity measures are very difficult to carry out politically. When the government imposes austerity measures after a terms-of-trade shock or an interest rate increase, political opponents of the government will try to blame the economic hardships on government mismanagement rather than on the exogenous shocks themselves, and they will often enjoy significant political success with this gambit. This possibility alone will lead governments to try to postpone and disguise the necessary adjustments as long as possible. Moreover, many governments simply do not have the power to implement an austerity program, even if they want to. Large parts of the public sector (e.g., state enterprises, local and regional governments) may be outside of the effective control of the finance minister. And even if the executive branch formulates a realistic adjustment effort, it may be frustrated by a

23

Introduction

recalcitrant legislature (especially if controlled by an opposing political party) which refuses to support the executive's measures.36 The government may therefore choose, or be forced, to absorb much of the shock through increased borrowing. 37 Eventually, the creditworthiness for borrowing will erode, finally forcing the government into inflationary finance if it still cannot engineer a cut in spending or a rise in taxes. If we examine the actual path of adjustments taken in response to external shocks, certain salient trends are evident. In Latin America, where the foreign credit squeeze was most severe, governments responded to the external shocks with a combination of spending cuts and increased domestic borrowing. In cutting spending, public investment projects were the first to go, public sector real wages the second, and public sector employment a distant third. The decline in foreign loans and the large budget deficits after 1981 led to a shift toward domestic financing. That financing was in turn divided between financing from the central bank (i.e., seignorage financing) and financing from the domestic capital markets. Of course, increased domestic borrowing contributed to a crowding out of private i n v e ~ t m e n t .Since ~~ private investment was also crowded out by the increased public sector deficits, the combination of a lower public sector investment rate and a lower private sector investment rate led to a sharp drop in total investment rate in the economy, as a percentage of GDP, as shown in table 12. Between 1982 and 1986, the governments in Argentina, Brazil, and Mexico borrowed heavily in the domestic capital markets, thereby raising sharply the stock of domestic public debt as a share of GNP. By 1988, the interest burden on the internal debt rivalled the interest burden on the external debt in Brazil and Mexico. In Bolivia, the government lacked even the creditworthiness to borrow from the internal capital markets, so that the decline in foreign lending was matched, almost one-for-one, by an increase Table 12

Rate of Gross Capital Formation as Percentage of GDP, 1978-86 Country Latin America Argentina Bolivia Brazil Mexico Rest of world Indonesia Philippines South Korea Turkey

Source: IMF, International Financial Statistics

"1982-83. b1982-85. na = not available.

1978-81

1982-86

.23 .17 .22 .25

.18"

.24 .25 .31 na

.28 .20 .31 na

.12b

.I9 .21

24

Jeffrey D. Sachs

in central bank credits to the government. As pointed out by Juan Antonio Morales and myself (in this vol.), this shift from foreign financing of the government deficit before 1980 to central bank financing after 1982 was the proximate cause of the Bolivian hyperinflation during 1984-85. The large internal borrowing in Argentina, Brazil, and Mexico, with the attendant upward pressure on domestic real interest rates, led to increasing reliance over time on central bank financing of the budget deficits. Eventually, the central bank financing resulted in a sharp increase in domestic inflation. To summarize, large budget deficits during the pre-1982 period had modest effects on inflation since they were financed heavily by foreign borrowing, with a pegged exchange rate that kept domestic inflation low. After the foreign credits dried up and the budget deficits worsened after 1982, inflation began to increase. The increasing recourse to central bank financing in Argentina, Bolivia, Brazil, and Mexico between 1982 and 1985 then contributed to the outbreak of very high inflation, reaching a true hyperinflation in Bolivia during 1984-85 and more than 1,000 percent annual inflation in Brazil in 1988. 4.

The Difficulty of Economic Stabilization

Six of the countries in the NBER project experienced a serious debt crisis: Argentina, Bolivia, Brazil, Mexico, the Philippines, and Turkey. Some evidence on the successes and failures of stabilization after the debt crisis are given in table 13. For each country, I summarize the growth and inflation performance of the economy in the years following the onset of the debt crisis. It is clear that only Turkey has been able to resume sustained growth after the crisis, though as of 1988, the Philippines was also showing signs of a possible return to sustained growth with low inflation. All of the countries except the Philippines have had great difficulties in controlling inflation. Turkey’s inflation has been in the high doubledigit rates, while inflation in Latin America has been at triple-digit rates and higher. As of the end of 1988, it appeared that Bolivia and Mexico had been able to end their high inflations through programs of fiscal austerity, while Argentina and Brazil appeared to be on the verge of hyperinflation. One of the deepest puzzles in the debt crisis has been the difficulty of sustaining economic stabilization and growth, with the partial exception of Turkey. The Turkish recovery in growth in the 1980s is explained by Cel2sun and Rodrik by several factors: (1) the generous foreign support received by Turkey during 1979-82 from official creditors; (2) the shift toward export promotion in 1980, backed up by trade reforms and a significant real depreciation with a real wage cut; (3) special factors, like the boom in exports to the Middle East following the outbreak of the Iran-Iraq war; and (4) the fact that Turkey’s external debt burden was never as great, as a percentage of GNP, as in the Latin American countries.

25

Introduction Recent Economic Performance: GDP Growth and Inflation

Table 13

GDP Growth Country

1983

Latin America Argentina Bolivia Brazil Mexico Rest of world Indonesia Philippines South Korea Turkey

3 .O -6.6 -3.5 -5.3

4.2 1.1 10.9 3.4

1984

1985

2.6 .3 5. I 3.7

-4.5 - .2 8.3 2.7

6.0 -7.1 8.6 5.9

2.5 -4.1 5.4 5.1

-

1986

1987

I983-87

2.0

1.6 -1.6 4.5 - .7=

5.5 2.9 7.6 - 3.7

3.6 na

4.0 1.9 11.7 7.8

3.6 5.9 11.1 na

-

2.1

4.1 - .6

9.5 5.5"

Inflation Latin America Argentina Bolivia Brazil Mexico Rest of world Indonesia Philippines South Korea Turkey ~

~~~~

350.0 275.6 150.0 101.6

622.2 1,282.4 210.0 65.5

669.2 12,783.0 222.6 57.7

90.1 276.3 145.0 86.2

131.3 14.6 229.8 131.8

305.7 680.0 189.1 86.8

11.8 10.0 3.4 31.4

10.4 50.4 2.3 48.4

4.7 23.2 2.5 45.0

5.8 .8 2.8 34.6

9.3 3.8 3.0 38.8

8.4 16.3 2.8 39.5

~

Source: IMF, Internafional Financial Stafisfics. Note: Inflation is calculated using the consumer price index. The average for the period is a geometric

average. '1983-86.

I have already indicated some of the reasons for the great difficulties in recovery in the Latin American countries, but they bear reiterating: 1. The external shocks required a sharp reduction in real wage levels, which proved to be politically destabilizing in several countries; 2. The external shocks provoked a significant worsening of the budget deficit, as well as a reduced ability to finance the deficit from abroad. The fiscal crisis in turn prompted sharp cuts in public investment and increased domestic borrowing by the central government. The domestic financing of the deficit has led to a partial crowding out of private investment and a sharp increase in inflationary finance by the central bank; 3. The sharp reduction in aggregate investment rates has led to a significant reduction in the potential growth rates of the economy; 4. The sharp rise in inflation has undermined political and economic stability and has undermined private sector confidence to invest in new export sectors.

26

Jeffrey D. Sachs

The experiences of Argentina and Brazil in attempting to use wage-price controls as part of an anti-inflation effort are illuminating. The so-called heterodox shock programs, known as the Austral Plan in Argentina and the Cruzado Plan in Brazil, are discussed, respectively, in detail in the monographs by Rudiger Dornbusch and Juan Carlos de Pablo and by Eliana Cardoso and Albert Fishlow (in this vol.). The theory of the heterodox shock was the same in both countries: since inflation has an important inertial component, it is important to complement the orthodox measures of tight monetary and fiscal policy with an incomes policy designed to break the wage-price inertia.39 In the event, both countries carried out the heterodox elements but fell short on the more orthodox elements, i.e., the budget deficits remained excessively large. After a short period of price stabilization, high inflation resumed in both countries. It is important to underline three additional features of the economic environment that have hindered a strong recovery in Latin America. I put these factors under the heading of the “political economy of the Latin American debt overhang.” The first factor is the political weakness of the Latin American governments. In Argentina, Bolivia, and Brazil, the debt crisis coincided with the transition of democratic rule after almost two decades of military government. The new democracies were fragile, untested, and inexperienced, and subject to the pent-up social demands of a population that had been politically repressed for years. Populist sentiments to increase wages and social spending took hold just at the moment that budget austerity and real wage cuts were needed. Even in Mexico, where the PRI (Institutional Revolutionary Party) has dominated the political scene for decades, the political control of the government has been waning for years, thus limiting the possibilities for strong austerity rnea~ures.~’ The second factor is the particular problem of implementing austerity and reform programs in response to an external debt crisis. I have noted that interest servicing of the foreign debt now constitutes a large fraction of public expenditure in all of the Latin American countries (as well as in Indonesia and the Philippines). Budget austerity measures and other kinds of reform measures (e.g., trade liberalization measures) are viewed by the citizens of these countries as a squeeze on living standards that is undertakenfor the sake of foreign creditors. Since the foreign debt has little public legitimacy in the first place (particularly in South America where it was accumulated during the 1970s, mostly under repressive military regimes), there is little popularity in belt-tightening for the sake of repaying the foreign debt. Increasingly during 1982-88, the issue of whether to suspend debt payments has become a central topic of political debate. Governments that urge austerity programs are increasingly subject to rejection by a frustrated electorate that is attracted to the option of suspending foreign debt payment^.^' The third factor is the set of economic inefficiencies that arise from a large overhang of debt.42 Countries with a large overhang of debt face the

27

Introduction

following kinds of difficulties: sanctions from disgruntled creditors for incomplete payments on old debts (e.g., withdrawals of trade credits); high bargaining costs in renegotiating the old debt, including repeated breakdowns in financial relations between the government and its creditors; an inability to borrow for new investment projects, even when those projects have a high rate of return; and, as already mentioned, adverse incentives for carrying out austerity and reform programs since the costs of reform are borne by the debtor country while the benefits of reform will be appropriated by the foreign creditors. Debt Renegotiations After the Onset of the Debt Crisis In each of the six countries that succumbed to the debt crisis (Argentina, Bolivia, Brazil, Mexico, the Philippines, and Turkey), the country abruptly lost access to international lending after the international banks deemed the country to be uncreditworthy. In all cases, the contractual schedule of debt-servicing obligations faced by the country after the onset of the crisis could no longer be met, in part because the country had been relying on new lending to help finance the debt-service payments that were coming due. This was especially true for amortization payments, for which the normal expectation was that they would be refinanced through new loans. Once the debt crisis hit, even the refinancing of amortization payments became highly problematic. As a result, there has been a sustained period of renegotiation of the debt since the onset of the crisis. Only Turkey has “graduated” from the process of renegotiation to renewed access to market borrowing. For the other five countries, the renegotiation process has been nearly continuous for five years, with little prospect in sight for a resumption of normal market conditions for these countries. As I noted in table 3, the secondary market value of the commercial bank debt of the five countries remains far below par, indicating a lack of faith that the old loans will be fully repaid. Thus, the banks have little desire to resume normal market lending to these countries. The nature of the debt renegotiations has been widely discussed and analyzed, and a summary of the process may be found in my introduction to volume 1.43The basic structure of negotiations was designed by the creditor governments, mainly the U.S., working in conjunction with the commercial banks and the international financial institutions. The Mexican renegotiations of 1982-83 set the basic pattern for almost all of the negotiations that followed. In a nutshell, the negotiating process has contained the following elements: (1) a long-term rescheduling of all principal due on medium- and long-term debt owed to commercial banks; (2) a commitment by debtor countries to remain current on interest payments due on the debt; (3) in some years, a partial refinancing of interest payments through the mechanism of a “concerted loan,” in which the bank creditors agree jointly to re-lend a

28

Jeffrey D. Sachs

portion of the interest that is due; (4) a long-term rescheduling of bilateral debt-service payments, both principal and interest, in the Paris Club forum; (5) an IMF standby program, with conditionality focussing on budget austerity and export promotion via exchange rate depreciation; and (6) new lending from official sources, including the IMF, the World Bank, the multilateral development banks, and the export credit agencies of the industrial countries. Because of the creditor governments’ concerns with the financial situation of the commercial banks that were heavily exposed in the debtor countries, the official policy has been to press for continued interest servicing of the debt and to oppose any significant element of debt reduction in the course of the negotiations between the debtors and the banks. Thus, even as evidence has grown after 1982 that much of the debt may not be repayable, at least if economic and political stability is to be maintained in the debtor countries, the negotiations have not led to any significant long-term reduction in the debt burden, with the notable exception of Bolivia. Unlike Turkey, which received substantial credits after 1979 (enough to eliminate the net resource transfer from the Turkish government to its creditors), none of the Latin governments has been afforded a comparable financial “time-out” on debt servicing. Every year, the Latin governments have made large net resource transfers to the foreign creditors (equal to several percent of GNP each year). Only Bolivia has been able to reverse the negative resource transfers after 1985, through its policy of suspending debt payments to the commercial banks while at the same time receiving new credits from the official creditor community. The NBER monographs make clear the high costs of a strategy that has continued to insist on full interest servicing of the bank debt. Relations between the creditor banks and the debtor governments have become increasingly turbulent and unstable since 1982. Negotiations between the governments and the banks are drawn-out cliffhangers, and the hardline positions of the creditors and the debtors have often pushed financial relations to the verge of rupture or beyond (as with the moratorium of debt-service payments by Brazil during 1987, the suspension of payments by Bolivia after 1984, and the deep arrears of Argentina in 1988). Since the IMF has designed standby programs based on the presumption that the interest on the debt can and will be fully serviced, the conditions of these programs have repeatedly failed to be met by the Latin American countries. As Cardoso and Fishlow recall, “successive letters of intent under IMF programs had no sooner been dispatched than they were made obsolete by accelerating inflation that did violence to the monetary targets.’ ’ Brazil signed and then subsequently failed to meet the conditions of no less than eight standby arrangements during 1983-85. Similarly, Argentina entered and then failed to comply with several IMF programs from 1986 to 1988. The failure to the economies in the debtor countries to recover has led more and more market participants to expect that the debts of the Latin

29

Introduction

American countries will eventually be reduced, either through a negotiated settlement or unilateral action. (As Lindert and Morton, Sachs, and Eichengreen stress in volume 1, debt reductions have been the normal mode of resolving debt crisis in past historical episodes.) The unusual case of Bolivia provides some support for the argument that debt reduction may bring benefits both to the creditor and debtor.44 For a variety of economic and foreign policy reasons, the U.S. government and the rest of the official creditor community began after 1985 to support the concept of debt reduction for Bolivia. Morales and Sachs argue that this more lenient attitude on the debt was crucial in allowing Bolivia (alone of the high-inflation countries in the region) to end the hyperinflation of 1984-85 and to restore stability and renewed economic growth. During 1986-88, the official community supported Bolivia with new credits despite Bolivia’s policy of nonpayment of interest to the banks.45 At the same time, the official community supported Bolivia’s negotiations with the banks aimed at allowing Bolivia to use donated funds from friendly governments to repurchase part of its debt at the deeply discounted secondary market price (of 11 cents per dollar of face value). In 1988, Bolivia repurchased approximately one-half its debt from the banks. Subsequently, Bolivia and the banks have entered into a new round of negotiations to reduce the burden of the remaining debt. It would appear that the creditor banks have benefitted from the buyback along with Bolivia. Before the buyback, Bolivia was in complete suspension of debt payments and the secondary market price of Bolivia’s debt had fallen to 5 cents on the dollar. The buyback raised the secondary market value to 1 1 percent and allowed the banks to sell half of the debt for cash at this price. Bolivia surely benefitted from the lenient treatment on the debt, not only in the trivial sense of saving money on future debt servicing, but also in the sense of bolstering the political case for a strong domestic austerity program, which could no longer be attacked as a measure undertaken for the sake of the foreign creditors.

5. Summary and Conclusions The NBER monographs were prepared to shed light on several important questions regarding foreign borrowing and the developing country debt crisis. Why did the crisis occur? Why did it hit some countries deeply and not others? What are the economic mechanisms involved in an external debt crisis, and why is recovery so difficult? How might the process of debt negotiations be altered to enhance the recovery of the debtor countries? The monographs are very rich in insights on these points, and this essay could do no more than touch the surface of the individual country experiences. Nonetheless, the country experiences do lend themselves to certain general conclusions, and it is worthwhile to restate these at this point.

30

Jeffrey D. Sachs

First, while the debt crisis was provoked by important shocks in the international economy (mainly the rise in interest rates and the fall in primary commodities prices), the effects of these shocks on the individual debtor economies depended importantly on the quality of economic policies that were in place in the debtor economies. Countries with a sound long-term growth strategy, based on an outward-oriented trade regime, realistic exchange rates, and prudent fiscal policies, were well equipped to cope with the external shocks. Moreover, even after the shocks hit at the end of the 1970s, there was still time for a reorientation of policies by the debtor countries. Most countries (and especially those in Latin America) lost this opportunity to get policies under control; many other countries, such as Indonesia, Korea, and Turkey, underwent significant adjustment efforts, which helped them to avoid the worst of the crisis. The debt crisis, when it hit, showed a similar mechanism in the countries that fell into crisis. In the cases of Argentina, Bolivia, Brazil, Mexico, the Philippines, and Turkey, the onset of the debt crisis is marked by an abrupt withdrawal of new credits form the world capital markets and an inability to roll over existing debts on a normal basis. This shock is typically combined with (and indeed precipitated by) a terms-of-trade deterioration or a rise in interest rates that puts additional burdens on the debtor country. Proper adjustments to these shocks typically require a real exchange rate depreciation (i.e., a rise in the price of tradables to nontradables, a reduction of the urban real wage, a shrinkage of the nontradables sector and an expansion of the tradables sector, and a program of fiscal austerity. These adjustments can well be traumatic, both economically and politically, especially in economies that are particularly inflexible because of a past history of inward-oriented trade policies. If the nontraditional export sector is especially small, then a very large real exchange rate depreciation and a large fall in the real wage might be needed in order to restore external equilibrium. The process of reducing the real wage can provoke social unrest. It is notable that this process led to more open conflict in countries with militant trade unions (e.g., Argentina and Bolivia), than in countries that relied on repression of trade unions to restrict labor demands (e.g., Korea and Turkey in the early 1980s). The reduction of the budget deficit after the onset of the debt crisis also proved to be difficult and politically destabilizing in many countries. In part this is because governments simply wanted to avoid the opprobrium from pursuing an unpopular austerity program. In other cases, the government faced opposition in the legislature to austerity measures (such as tax increases or layoffs in the public sector) and thus was blocked from implementing an austerity plan. Also, the executive branch typically lacks the power to make a comprehensive fiscal adjustment since many of the fiscal problems are centered in powerful and independent state enterprises,

31

Introduction

and in regional and local governments, which are beyond the reach of the finance minister. These units often have independent access to central bank financing that is outside of the budget. Moreover, these entities often have a powerful political voice, which allows them to resist calls for austerity. Fiscal adjustments in the response to a debt crisis are particularly tricky from a political point of view, since the public comes to view the austerity measures as being put in place for the sake of foreign creditors, rather than domestic residents. In the Latin American countries and in the Philippines, there is a growing political current which therefore rejects the austerity measures in favor of a unilateral suspension of payments to the foreign creditors. The country monographs reveal only spotty success in restoring stability and growth after the onset of a debt crisis. Turkey has been the most successful case, for it has not only restored economic growth but has also restored its access to the international capital markets. But even in this most successful case, the inflation rate has remained extremely high (nearly 100 percent per year in 1988), so that macroeconomic stability remains in question. In the Latin American countries, there has been no successful revival of growth, and as of 1988, Argentina and Brazil continued to face the possibility of hyperinflation. Mexico reached triple-digit inflation in 1987, but apparently reduced it in 1988 at the cost of another year of output decline. Bolivia was able to end a hyperinflation in 1986 and to resume slight positive growth in 1987, but only in the context of a program that included a complete suspension of payments on commercial bank debt. The longer term prospects remain grim for many of the countries. Following the crisis, aggregate investment rates are down sharply in the four Latin American countries, as well as in the Philippines, so that the long-term growth potential has been diminished in recent years. Moreover, these countries seem to have little prospect of a quick return to the international capital markets in view of the market assessments that much of the existing debt will not be serviced in the long run.46 I noted several possible inefficiencies that may result from the large overhang of bad debt. These inefficiencies include the costs of repeated breakdowns in debtor-creditor relations; the withdrawal of trade credits and new lending to the countries, even when highly profitable new investment opportunities exist; and the disincentive for the debtor to undertake costly reform efforts, if the benefits of those reforms are likely to flow to the creditors in the form of increased debt-service payments. For these reasons, many of the authors speculated about the possible need to reduce the debt burden in the future and to abandon the presumption that all of the debt should eventually be repaid. The studies provide only one illustration of actual debt reduction, the case of Bolivia, but the evidence suggests that in that case, debt reduction has in fact worked to the mutual benefit of the creditors and the debtor country.

32

Jeffrey D. Sachs

Notes 1. For several studies of Latin America in the 1930s, see Thorp (1984). A particularly good survey of the period is provided by Carlos F. Diaz Alejandro in the same volume. These essays show that despite the enormous collapse of commodities prices in the Great Depression, many primary commodities producers experienced a vigorous recovery, particularly in agriculture and import-competing manufacturing. 2. In Bolivia, for example, real per capita GDP in 1988 was at the same level as in 1955. In Peru, the real per capita GDP in 1987 was as the same level as in 1960. 3. The debt crisis is generally dated to August 1982, when Mexico announced that it would be unable to meet its debt-servicing obligations. Soon thereafter, there was a cascade of similar announcements. It is true, however, that several countries fell into external crisis before Mexico. Of the countries in the NBER study, Turkey’s crisis arrived at the end of the 1970s, Bolivia fell into debt crisis in 1981, and Argentina fell into crisis in the spring of 1982. 4. Even Colombia, which did not reschedule its debts after 1982, found it to be impossible to get normal market loans to roll over principal falling due in 1988. Chile, which is widely considered by the financial markets to have been the best-managed of the South American countries, is still without access to spontaneous new lending from the banking world. 5. There are differences in the nature of the debt crisis in the poor versus middle-income countries because of the much less important role of bank debt in the case of the low-income countries. Thus, while this study inevitably sheds some light on the poorer countries in Sub-Saharan Africa, the relative importance of commercial bank debt in Latin America leads to special characteristics of the adjustment process and debt-renegotiation patterns in Latin America as compared with Africa. One major difference is that official creditors did not withdraw net lending as abruptly as did the commercial banks, so that there was not a decisive shift from net resource inflows to net resource outflows in the low-income countries of Sub-Saharan Africa as there was in the middle-income countries of Latin America. The second difference has been the greater effective debt relief offered by the official creditors than by the commercial banks. The Paris Club of government creditors, for example, routinely reschedules nearly 100 percent of interest payments due, while the commercial banks have never rescheduled interest payments. 6. Bolivia experienced the sharpest measured decline in per capita income during 1980-85 (-28 percent), as well as the worst hyperinflation in Latin American history (24,000 percent from August 1984 to August 1985). Peru, with hyperinflation beginning in 1988, appears to be the leading candidate for the worst crisis during the second half of the 1980s. 7. In practice, most of the debt is in fact the liability of a public sector entity. 8. Between 1980 and 1986, the proportion of total external debt owed by the public sector tended to rise significantly in the heavily indebted countries. According to the World Bank’s Debt Tables (1988-89 edition), the overall 1980 long-term debt of the heavily indebted countries was $205 billion, of which $148 billion was public or publicly guaranteed debt. Thus, in 1980, about 72 percent of the debt was owed by the public sector. By 1986, the total debt had risen to $421 billion, of which $360 billion, or 86 percent, was now owed by the public sector. One reason for the rise in

33

Introduction

this ratio has already been mentioned in the text: private sector firms made individual arrangements with their creditors to reduce the debt through various kinds of workouts. This is generally easier than with public debt, for several reasons: (1) the private debtors usually have only a small number of foreign creditors; (2) the private debt can be negotiated outside of the glare of publicity; deals can be made without fear that they will act as precedents for other debtors; (3) concerted lending packages have been available mainly to public sector debtors; and (4)some parts of the private sector debt have been absorbed by the public sector, through various kinds of bailout programs for private sector firms. 9. Also, the availability of official credits will depend on various noneconomic factors, such as the military/strategic importance of the country. As Merih CelPsun and Dani Rodrik point out in volume 3, the strategic importance of Turkey as a frontline member of NATO contributed to the large inflows of official credits to Turkey after the onset of Turkey’s financial crisis toward the end of the 1970s. 10. It may also seem to impute too much irrationality to market participants to suggest that they believed in a Ponzi scheme during the 1970s. Didn’t they know that such a good thing was bound to end once real interest rates rose back to more normal levels? The answer seems to be that many market participants and observers simply extrapolated the heady successes of the 1970s well into the future. And for many market participants, the incentives were such that there were few concerns about the far distant future. Within many debtor countries, the expected tenure in office of a particular government was too short to worry about the long-run constraints on borrowing. Even more surprising, the same was true within many large banks. As described in a recent history of Bank of America (Hector 1988), the lending at the bank in the mid- and late 1970s was judged according to the growth of earnings, rather than on the quality of the loan portfolio. Of course, the earnings on a loan in a particular year depended only on whether the loan was being serviced that year (even if out of borrowed money) and not on an assessment of the true long-term value of the loan. Hector describes the frenetic atmosphere at the bank and at competitor banks as follows: By 1977, as Clausen [the bank president] strove to keep quarterly profits climbing, the 10 percent a year growth limit fell to demands from the head office. Tom Clausen wanted to restore Bank of America to its rightful place as the nation’s most profitable bank. He pushed lenders to increase loans rapidly, even though Bank of America was running at the limits of its ability to control its growth. . . . (88) Managers learned quickly that headquarters always wanted more growth than they could supply . . . [Lloan officers realized that they were better off adding a pile of new loans of dubious quality than adding too few loans. The head office didn’t differentiate between the good and bad loans of its budget reviews, and a loan officer’s salary increases and promotions depended on only two variables: the size of his loan portfolio and the number of people working for his unit. (89)

1 1 . For the conclusion that the size of external shocks (both on commodities prices and interest rates) cannot account for cross-country differences in performance in the 1980s, see also my papers, Sachs 1985, and Berg and Sachs (1988). 12. For an authoritative statement in support of outward-oriented trade regimes, see Krueger (1978). See also various studies by Bela Balassa, such as Balassa and Associates (1982).

34

Jeffrey D. Sachs

13. Detailed descriptions of the trade strategies may be found in each of the country monographs. 14. Brecher and Diaz Alejandro proved that foreign borrowing at market interest rates must be welfare-worsening for a competitive economy operating with a tariff bamer for the import-competing sector. The tariff results in an excessive allocation of resources to the protected sector. The shadow value of capital in this sector is necessarily less than the world cost of capital, meaning that national income is reduced overall by the foreign borrowing once repayment on the foreign capital is taken into account. This misallocation caused by a high tariff barrier can in fact be so severe that national income is reduced by an inflow of foreign capital even if the foreign capital is a gift rather than a loan, a point first made by Harry Johnson (1967). The inflow of foreign capital reduces income by further encouraging the buildup of the high-cost protected sector and reducing the production of the more efficient export sector. 15. Among the Latin American countries, Brazil has shown the greatest capacity to spur new nontraditional exports. This is in line with Brazil’s export-promotion policies during much of the 1970s. According to many observers, including Eliana Cardoso and Albert Fishlow (in this vol.), the Brazilian trading system became increasingly inward-oriented in the mid- 1970s and afterward. Following substantial trade liberalization after 1985, Mexico has also enjoyed a mini-boom in manufacturing exports. The export growth has been very rapid, but starting from a very small base. 16. In some cases, where foreign exchange was rationed after the terms-of-trade decline and the onset of the debt crisis, the manufacturing firms simply could not obtain the necessary inputs, or only on the black market with a very depreciated exchange rate. In other cases, the exchange rate was devalued sharply so that the cost of the inputs rose sharply relative to wages and relative to nontradables in general. 17. In Argentina, Bolivia, and Brazil, the exchange rate has sometimes been maintained, even after reserves dry up, through the use of foreign exchange rationing and exchange controls. A black market exchange rate inevitably develops, which is sharply depreciated relative to the official rate. As is well known, the spread between the official rate, at which exporters must surrender foreign exchange, and the black market rate, which determines the marginal cost of imports, acts like an implicit tax on the export sector. The spread between the black market exchange rate and the official rate reached several hundred percent in Bolivia during 1984 and 1985, and has frequently been on the order of 30-50 percent in Argentina and Brazil. For further evidence on the existence of higher black market premia in Latin America than in East Asia, see Sachs (1985, 541, table 6). 18. Note that as with trade policy, Brazil comes across as somewhat more outward-oriented than do the other Latin American countries. 19. Interestingly, observers in Mexico have recently noted a growing constituency for devaluation among the new manufacturing exporters who have arisen with the liberalization-with-real-exchange-rate-depreciationsince 1985. 20. Argentina and Mexico were particularly susceptible to capital flight because of the absence of capital controls (Argentina, unwisely, had eliminated most capital controls during 1977-78). 21. The definition and measurement of capital flight are fraught with conceptual and data problems. What should be counted as capital flight? All private capital flows abroad? Only those that are unreported? Only those that are prompted by a

35

Introduction

specific motivation, such as tax evasion or to avoid political unrest? The definition used in table 1 1 is one that was proposed by Morgan Guaranty. It defines capital flight as all private sector capital outflows except those that are intermediated by the domestic banking system and except those that are recorded as foreign direct investment. Other definitions, equally plausible, also show the same pattern as in the table. 22. Apparently, ’hrkey also avoided extensive capital flight, though the studies reported in table 11 did not include Turkey in the countries examined. 23. Despite perfectly open capital markets, Indonesia has apparently been able to avoid rampant capital flight through the maintenance of a realistic exchange rate and prudent fiscal policy. 24. In some cases, however, the government is a net importer of goods and services (e.g., in Brazil, where the government services foreign debt and imports oil). Then, as Cardoso and Fishlow stress, a real appreciation can actually reduce a budget deficit and a real depreciation can worsen a deficit. In Brazil, for example, a real depreciation raises the domestic resource cost of servicing the foreign debt, which is not offset by higher earnings of public sector export firms. 25. Actually, the share of debt in the public sector increased after the onset of the debt crisis since much of the private sector debt was taken over by the government in the course of various bailout operations for private sector firms. Even before the debt crisis hit, however, the share of public sector debt in the total debt has always been more than half in all of the countries in the NBER project. 26. Among other factors, the regime is interested in reducing discontent in the heavily agricultural Outer Islands. 27. Of course, Mexico had already experienced one near-disaster with foreign borrowing in 1976, and Indonesia had been shaken by the Pertamina crisis of 1975, but those cases were viewed as special, isolated problems. The Mexican crisis was viewed more as a problem of exchange rate mismanagement (pegging to an unrealistic value of the peso), rather than as a foreign debt crisis. 28. The net debt here is defined as the total liabilities owed by residents of the country (public plus private sector) to commercial banks in the BIS reporting area, minus claims of residents on BIS-area banks. In Argentina, the net debt went from $5.3 billion at the end of 1979 to $16.3 billion at the end of 1981; in Brazil, the debt rose from $28.8 billion to 44.8 billion; and in Mexico, the debt increased from $22.5 billion to $43.4 billion. For the three countries, the net debt increased by 85 percent in just two years. 29. This is the phenomenon known as the “Dutch disease”: a strong resource sector crowds out other tradables sectors and promotes the expansion of the nontradables sector. 30. The real wage could actually rise with an exchange rate depreciation in the unlikely case of a very large T sector that is labor intensive relative to the N sector. 31. Actually, in Korea, where the upward trend growth of real wages was quite strong, the shocks only slowed the growth of real wages relative to the earlier trend, but did not reduce real wages absolutely, except in 1980. 32. In Brazil, the labor movement has historically been rather weak and much less organized and militant than in neighboring Argentina and Bolivia. 33. Either directly, through sales of a state trading company, as in Bolivia and Mexico, or indirectly, through trade taxes and royalty payments on commodity production.

36

Jeffrey D. Sachs

34. The revenues may be generated directly by state enterprises that own and produce the primary commodities (as in Mexico, where the state enterprise PEMEX is responsible for all oil exports, or in Bolivia, where the state tin company, COMIBOL, and the state oil company, YPFB, generate a very large share of total exports), or through export taxes (as in Argentina, which relies heavily on taxes on grain and meat exports). 35. Also, in Brazil and Turkey, where oil imports are undertaken by a state petroleum company, the rise in world oil prices can cause a significant budget deficit if the state firm, under domestic political pressures, is prevented from fully passing along to consumers the higher oil costs. Subsidies on oil imports was a large source of Brazilian budget deficits after 1973 and again after 1979. 36. In Bolivia, for example, during the 1982-85 administration of President Hernan Siles Suazo, the Bolivian Congress refused to support various tax measures proposed by the government to bring the deficit under control, thereby increasing the reliance on central bank financing and contributing to the eventual hyperinflation during 1984-85. 37. Of course, the recourse to borrowing in the face of a permanent external shock is feasible only in the short run; eventually the budget will have to adjust not only to the terms-of-trade decline, but also to the debt-service burden on the accumulated debt that results from the borrowing strategy. Even so, postponement of adjustment by the current government may make a lot of political sense, especially in a country with a rapid turnover of governments, where it can be expected that the political costs of adjustment will be borne by a future government of a different political party. In any event, a government may have no other option than borrowing, if it is unable to win legislative support for an austerity program. 38. Remember that this crowding out can be avoided if consumption (public and private) is reduced via higher taxes and lower government spending. 39. The highly successful Israeli stabilization program was based on the same principle. But in contrast to the Latin American cases, the Israeli program actually implemented a large measure of budgetary restraint. 40. Nonetheless, the Mexican government was still able during 1982-88 to impose harsh fiscal austerity measures and a sharp reduction of the real wage. These efforts were culminated at the end of 1987 in a new stabilization program christened the “Emergency Social Pact.” The program was successful during 1988 in reducing the inflation rate, but contributed to significant electoral setbacks for the PRI in the elections of 1988. The long-term success of the pact was still in question as of the end of 1988. 41. As I note below, the case of Bolivia is illuminating in this regard. Morales and Sachs stress that it was the Bolivian government’s suspension of interest payments to the commercial banks that gave the government the “political space” to undertake the sharp austerity measures at home. 42. See Sachs (1988) for a concise statement of the theory of adverse effects from a debt overhang, as well as references to further discussion in the economics literature. 43. See also the discussion and critique of the debt negotiation process in Sachs (1986, 1987, 1988). 44. The Bolivian debt arrangements are described at length in the monograph of Morales and Sachs, and in a separate article by Sachs (1988).

37

Introduction

45. In the other cases of nonpayment to the banks, e.g., Brazil’s debt suspension during 1987, the official community has refused to extend new official loans. 46. That is, the market believes that the expected discounted value of future net transfers from the debtor to the creditors is significantly less than the face value of the debt.

References Balassa, B., and Associates. 1982. Development strategies in semi-industrialized economies. Baltimore: Johns Hopkins University Press. Berg, A., and J. Sachs. 1988. The developing country debt crisis: Some structural explanations. Journal of Development Economics 29(3):27 1-306. Brecher, R., and C. F. Diaz Alejandro. 1977. Tariffs, foreign capital, and immiserizing growth. Journal ofInternationa1 Economics 7(4):317-22. Diaz Alejandro, C. F. 1984. Latin America in the 1930s. In Latin America in rhe 1930s, ed. R. Thorp. New York: St. Martin’s Press. Dohner, R. S . , and P. Intal, Jr. 1989. Debt crisis and adjustment in the Philippines. In Developing country debt and the world economy, ed. J. D. Sachs. Chicago: University of Chicago Press. Hector, G . 1988. Breaking the bunk: The decline of Bunk of America. Boston: Little, Brown and Company. Johnson, H. 1967. The possibility of income losses from increased efficiency of factor accumulation in the presence of tariffs. Economic Journal 77( 1):151-54. Krueger, A. 0. 1978. Liberalization attempts and consequences. New York: National Bureau of Economic Research. Sachs, J . 1985. External debt and macroeconomic performance in Latin America and East Asia. Brookings Papers on Economic Activity, no. 1523-73. . 1986. Managing the LDC debt crisis. Brookings Papers on Economic Activity, no. 2:397-432. . 1987. International policy coordination: The case of developing country debt. NBER Working Paper no. 2287. Cambridge, MA: National Bureau of Economic Research. . 1988. Comprehensive debt retirement: The case of Bolivia. Brookings Papers on Economic Activity, no. 2:705- 15. Thorp, R., ed. 1984. Latin America in the 1930s. New York: St. Martin’s Press.

This Page Intentionally Left Blank

1

An Overview of Debt and Macroeconomic Problems

In 1985, after forty years of financial instability, Argentina reached once again near-hyperinflation conditions. Budget deficits were the immediate cause, but the deeper roots can be traced to the ill-fated policy experiments of the 1970s. The destructive pendulum between populists and marketoriented reformists has meant that much of national wealth is held abroad. Taxes are paid by only a few, and the general atmosphere is one of skepticism about everything Argentine. Mallon and Sourrouille (1975, 11) draw attention to this steady conflict when they write Decision makers in Argentina have quite consistently attempted to adopt policy positions that seemed designed to tear society apart rather than to forge new coalitions. . . . Major policy disagreements in modem Argentine history have their main roots in the conflict between two divergent streams of thought: liberalism of the British Manchester School variety and what can be called national populism. . . . In general, the liberals have stood for the virtues of a society open to international opportunities and influences, whereas, the national populists have emphasized indigenous, autonomous development.

In our study we investigate the interaction between domestic macroeconomic instability and external constraints. We study these relations by focusing primarily on the past decade in which four very different periods can be distinguished. 1. The Martinez de Hoz period of the 1970s when external debts were accumulated in the context of an incompatible mix of policies: large and The authors are indebted to Eliana A. Cardoso, Richard D. Mallon, and F. Desmond McCarthy for their helpful suggestions. Katarina Nelson greatly improved the manuscript with her editorial assistance.

41

42

Rudiger Dornbusch and Juan Carlos de Pablo

persistent deficits, a strongly overvalued currency, and liberalization of capital flows. 2 . The period from the end of the 1970s to the hyperinflation, when debt and foreign exchange problems, war, and domestic politics were the reasons for an inflation explosion. 3. The Austral stabilization plan. 4. The post-Austral quest for a resumption of growth.

1.

A Long-Run Perspective

Although we focus only on the past ten years, we place our analysis in a long-run context. This is appropriate because debt problems and financial crises are at least one hundred years old in Argentina. One hundred years ago Argentina’s inability to service the foreign debt nearly brought down the City of London in the famous Baring panic of 1890; the Tornquist monetary reform dates back to 1899. It is important to view developments in this long-term perspective because it highlights how Argentina has steadily lost its position in the world economy during this century.’ Carlos Diaz Alejandro (1970, 1) reminds us of this decline: It is common nowadays to lump the Argentine economy in the same category with the economies of other Latin American nations. Some opinion even puts it among such less developed nations as India and Nigeria. Yet most economists writing during the first three decades of this century would have placed Argentina among the most advanced countries -with Western Europe, the United States, Canada and Australia. To have called Argentina “underdeveloped” in the sense that word has today would have been considered laughable. If Argentina in 1900 had a standard of living like that of the U.S., then the decline has been long and deep. Summers and Heston (1984) estimate that by 1950 Argentina had only 41 percent of the U.S. standard of living (against 80 percent of that in Australia and Canada). By 1985 the standard of living had slipped to only 30 percent of the U.S. level. Figure 1.1 shows the level of per capita real income in Argentina over the past forty-five years. There is a striking difference between the steady expansion in the thirty years before 1975 and the stagnation and decline that have occurred since then. The contrast could not be stronger: from 1945 to 1975, per capita income grew at an annual rate of 1.7 percent. From 1975 to 1985, it fell at an annual rate of 1.7 percent. The second dimension in which Argentine performance has shown a dramatic deterioration is inflation and fiscal stability. Of course, there have been frequent precedents for massive inflation and depreciation. But the experience of the past decade, with two near hyperinflations, stands out. In

43

ArgentindChapter 1 2.62 2.6 2.58 2.56 2.54

-

-

2.52 2.5 2.48

-

2.48 2.44 2.42 2.4

-

-

2.36

-

2.34

-/

2.38

2.32

I I I I I I I

40

45

I I

50

I t

55

I

I

I

60

I I 1 1 I I 4

I I 1~~~

65

70

1

I I I ! I I I I 1 1 I

7s

60

85

1899 Banker’s Magazine already reported of South Americans and their currency: [They] are always in trouble about their currency. Either it is too good for home use, or as frequently happens, it is too bad for foreign exchange. Generally, they have too much of it, but their own idea is that they never have enough. . . . the Argentines alter their currency almost as frequently as they change presidents. . . . No people in the world take a keener interest in currency experiments than the Argentines. The Argentine experience with the destruction of the financial system in the past fifteen years has certainly reinforced that keen interest and expertise. Figure 1.2 shows the monthly rate of inflation since 1970. In interpreting the graph, one should bear in mind that a monthly rate of inflation of 6 percent corresponds to 100 percent per year and 22 percent per month yields an annual rate of 2,500 percent. Inflation passed 1,000 percent in the Peronist period of 1975-76 and again in the pre-Austral period of early 1985. At no time in the past ten years did it fall below 100 percent for any length of time. The third broad feature to which we want to draw attention concerns the real exchange rate. This is a key price in any economy and even more so in the case of Argentina. Figure 1.3 shows the real exchange rate over the past thirty years measured as the ratio of domestic wholesale prices to the wholesale prices of import^.^ The extraordinary variations in Argentina’s external competitiveness are closely tied to macroeconomic policy mistakes, capital flight induced by these mistakes, and the present debt crisis. The outstanding episodes, clearly apparent in figure 1.3, are the real depreciation prior to 1976 and the

44

Rudiger Dornbusch and Juan Carlos de Pablo 30 20 26 24 22 20 10 16 14 12 10 0 6

4 2

0 -2

......................................................................

10

11

12

13

14

15

16

11

10

19

00

01

82

83 04

05

06

87

88

Fig. 1.2 Inflation rate of the CPI (quarterly average of monthly rates) 140

,

5 6 5 1 50 59 60 61 62 63 6 4 6 5 6 6 6 1 6 0 6 9 1 0 7 1 7 2 1 3 7 4 7 5 1 6 1 7 1 0 7 9 0 0 0 T 01 0 3 0 4 0 5 0 6 0 1

Fig. 1.3 The real exchange rate (WPI domestic/WPI imports, 1960 = 100)

appreciation of 1979-81. For the period 1970-78 the real exchange rate averaged 73; it increased to 108 over the next three years before declining back to an average of 75 in 1982-86. By March 1987 it had fallen to less than half of its peak value. The swings in the real exchange rate capture best the seesaw nature of Argentine policies. In some periods unimaginable damage is done to the productive and financial structures, and then a period of repair follows in which austerity and real depreciation restore the base for yet another political, fiscal, or foreign exchange adventure.

45

ArgentindChapter 1

Table 1.1 shows the debt accumulation over the past fifteen years. There is considerable uncertainty about the size of the external debt prior to the late 1970s, and available estimates from various official sources vary widely. Estimates of the Banco Central de la Repfiblica Argentina (BCRA) show that debt varied between $2.5 billion and $3 billion in the 1960s, ending at about the same level as it started. From 1970 on, external debt steadily increased for both the private and the public sector. Between 1970 and 1977 the external debt rose by $6 billion, and in the next four years it rose by more than $30 billion. We now turn to a review of the principal episodes. We will describe and explain the relevance to these episodes of the debt problem and the role of external debt in creating domestic macroeconomic difficulties. A brief chronology of dates and important facts helps place the events in context. 1.1.1 The Martinez de Hoz Period (3/1976-3/1981) At the time Jose Martinez de Hoz assumed power as finance minister of the military government, consumer prices had increased in the previous month at an annual rate of 5,000 percent and output had declined sharply. The black market premium for foreign exchange exceeded 200 p e r ~ e n tThe .~ new program was to stabilize the macroeconomy as a first priority, and then renovate industry and financial markets. Macroeconomic stabilization got under way quite rapidly so that inflation soon fell to less than 200 percent. A financial reform was implemented to liberalize capital markets and link Argentina more effectively with the world capital market. By late 1976, foreign exchange transactions were completely liberalized on capital account, and this was done so effectively that the black market premium was zero for the next four years. Figure 1.4 shows the black market premium and brings out the striking interlude of free capital mobility between the Peronist period and the aftermath of the collapse of Martinez de Hoz’s policies. Inflation failed to decline further once it was down to the 150 percent range. To make further inroads, policymakers opted for what Fernandez (1985) has called an “expectations management approach.” Beginning in 1979, they pre-fixed the rate of exchange depreciation with a tublitu, that is, Table 1.1

Argentina’s External Debt (in billions of $U.S. and percentages) 1975

Total external debt ($) Public ($1 Reserves ($) Net debVexports (%) DebVGDP (%) Interest payments/GDP (%)

7.9 4.0 0.6 260 18.6 0.7

1978

12.5 8.4 5.8 110 23.9 1.4

Sources: World Bank, BCRA, and Morgan Guaranty.

1979 19.0 10.0 10.1 120

30.2 I .4

1982

1985

I987

43.6 23.6

48.3 40.0 6.0 520

56.2

64.5

69.6 5.1

3.0 540 60.3 2.4

5.7

46

Rudiger Dornbusch and Juan Carlos de Pablo 280 260 240

I

220 200 180 160 140 120 100

80 60

40 20

0

Fig. 1.4 The exchange rate gap (parallel rate as percent of official, Australs/$U.S.)

they announced ahead of time a series of gradually declining rates of depreciation. These announcements were repeated on a rolling (though shortening) basis, so as to create an environment where economic agents could discern that a government commitment to disinflation was embodied in the timetable for reducing rates of exchange depreciation. This policy was expected to lower inflation through three separate channels. First, reduced rates of depreciation would directly lower the rate of import price inflation. Second, reduced depreciation would serve as a discipline on domestic price setters. Third, in an environment where inflation to a large extent depended on expectations, the rule or precommitment introduced a fixed point around which expectations could rally. Needless to say, the intellectual underpinnings of such a program relied on a belief that the Chicago School’s “law of one price” would be operative. Inflation responded to this policy, falling throughout 1980 to reach a bottom well below 100 percent. But gradually, during 1978 and 1979, the real exchange rate appreciated because inflation consistently outpaced the rate of depreciation. We saw in figure 1.3 that the cumulative overvaluation reached 50 or even 60 percent. But while the overvaluation ultimately led to capital flight and the collapse of the financial system, in the early stages there was quite the opposite trend. The high interest rates-relative to world rates and the preannounced rate of depreciation-gave rise to an (almost) risk-free speculation in favor of Argentine assets. As a result, private sector borrowing abroad increased to take advantage of the relatively low foreign interest rates, and a massive capital inflow developed. This is shown in table 1.1 in the large increase in Central Bank reserves between 1978 and 1979 and the matching increase in private external borrowing.

47

ArgentindChapter 1

The trade and employment effects of the overvaluation were slow to come. Diaz Alejandro (1964) has shown that the real income effects of a real depreciation tend to be dominant in the early stages, before substitution effects take over. For the real appreciation of 1977-80 the reverse applied: the increase in real income created an expansion in demand and thus seemed to validate the Martinez de Hoz approach by creating inflation reduction with rising real income. This was reinforced by the fact that trade protection, even with liberalization measures, kept the economy relatively closed, dampening the disinflation effects of the tubliru as well as the employment effects in the real sector. By 1979-80 the overvaluation had become so extreme that in financial markets there was the view that depreciation was inevitable. Even though the government asserted that the policy would be continued and could be financed, speculation increasingly went in the direction of dollar purchases. The regime of unrestricted capital mobility introduced in late 1976 maximally facilitated this capital flight. Hence, in 1979-80 the Central Bank and public sector enterprises were forced to borrow massively abroad to obtain the foreign exchange which was then sold in support of the exchange rate policy. Private speculators in turn bought the dollars and invested them abroad, With the round trip complete, commercial banks in New York, Zurich, and Tokyo had lent to the government the resources to finance capital flight which returned to the same banks as deposits. Of course, capital flight was not limited to dollar deposits-investments in financial markets were important as was real estate abroad. A variety of estimates is available on the accumulation of external assets by Argentines during this period. These estimates are typically residuals from debt and balance of payments data. They are obtained by deducting from the recorded increase in gross external debt the current account and recorded capital flows in the form of direct investment and changes in reserves. Dornbusch (1985a), for example, calculates that capital flight in 1978-82 amounted to $23.4 billion. In a review of various estimates, the IMF (Watson et al. 1986, 142) reports that capital flight amounted cumulatively to about $15 billion in 1979-8 1. Rodriguez (1987) estimates that between 1979 and 1982 Argentina’s private external assets increased from $10 billion to $34 billion. These estimates would have to be revised upward to reflect the extent to which the underinvoicing of exports and overinvoicing of imports was a significant channel of capital flight in this period. Both the fact of and the motivation for the wave of capital flight in the late 1970s are very clear. Unlike in other debtor countries (for example, Brazil or Chile), mismanagement of the exchange rate combined with an opening of the capital account is almost the full explanation for the massive debt accumulation. The particular background must be understood to appreciate that in Argentina’s case the government has an external debt but the private

48

Rudiger Dornbusch and Juan Carlos de Pablo

sector has matching external assets. Moreover, that process was carried further in the next few years as the government gradually took over all external debt in the course of sustaining failing financial institutions. In 1980 about half of the external debt was owed by the public sector; by 1985 that share had increased to 82 percent. 1.1.2 From Martinez de Hoz to Alfonsin (3/1981- 12/1983) The end to the military government did not come easily. The Martinez de Hoz overvaluation sowed the seeds of financial destruction, but the actual unraveling came only over the next four years. The world economy contributed to the problems induced by the debt crisis: sharply declining commodity prices and much higher interest rates brought with them difficulties in servicing the external debt. But domestic events were certainly the dominant factor. First came the reversal of the overvaluation. This process started with the change of presidents: the incoming president, months before taking office, declined to comment on his exchange rate policy. This served as an obvious indication that a devaluation was certain to occur; as a result, capital flight became massive. Central Bank reserves declined by more than $5 billion, and public external debt increased sharply. Finally Martinez de Hoz was forced by his successor, who was not yet in office, to bring his own expectations management and credibility approach to an end by devaluing the currency. Over the next three years exchange depreciation and inflation flourished, with inflation rising from less than 100 percent to 600 percent by the time Raul Alfonsin took office as president. Changes in public finance and financial markets were particularly important in this period. Exchange controls were instituted once again and the black market premium reemerged (see figure 1.4). The Central Bank, in an effort to ensure continuing trade flows, started exchange rate guarantee programs only to find that it could never hold on to the guaranteed exchange rates. As a result of the bank losing a string of bets in the foreign exchange market, the budget deteriorated dramatically. The deterioration was reinforced by financial failures that turned up in the public sector, by the burden of external interest payments, and by deteriorating terms of trade. The Malvinas conflict added to the loss of confidence and to the devastation of public finance. The economics of this period of deterioration can be expressed in terms of a simple model of deficit finance and financial markets. Suppose the budget deficit represents a fraction (g) of national income and that the velocity of high-powered money is an increasing function of the rate of inflation. Suppose the deficit is financed by money creation. Then it can be shown (see chap. 4) that the rate of inflation ( n )will be a steeply increasing function of the deficit but will also depend on financial institutions.

(1)

7r =

b g - Y)/(l

-

Pg)

49

ArgentindChapter 1

The higher the level of noninflationary velocity (a)and the more responsive velocity is to inflation as measured by the parameter (p), the more dramatic is the inflation impact of budget deficits. This framework helps to identify the interaction of deficits, external debt service, real depreciation, and financial markets in generating the inflation explosion of 1981-84. The growing burden of debt service, because of higher interest rates and real depreciation, increased the budget deficit ratio (g) and hence raised money creation and inflation. The institutional response of financial markets to higher inflation, namely a flight from money, aggravated this impact. The reduction in money holdings was facilitated by a growing range of interest-bearing substitutes. As these substitutes became more important, velocity sharply increased (a and p in eq. [l] increased), which meant that the inflation rate associated with a given deficit ratio also increased. The 1981-84 period thus represents an unraveling of what had been merely an artificial stability during the late 1970s. Several events, each in itself extraordinary, combined to make the crisis large: the initial overvaluation had been extreme, the financial sector had been allowed to become overexposed in speculation, private capital flight had been massive, and the world economy turned unfavorable at just the wrong time. Each of these factors caused the budget to deteriorate and hence reinforced inflation. 1.1.3 Alfonsin (1/1984-1989) These difficulties carried over to the beginning of the Alfonsin administration. Large real wage increases in 1983-84 created problems for the budget and for the external balance. Inflation rapidly escalated and negotiations with creditors and the IMF did not bring a solution. The inflation issue soon became the single most pressing problem. In early 1985 annualized monthly rates of inflation rose 1,500 percent and more. A hyperinflation was an entirely realistic possibility because the inflation process itself eroded the real value of tax collection and promoted a financial system which provided money substitutes, so that continual money creation was in order to finance an ever-widening deficit. Because IMF programs seemed unable to cope with the inflation problem in a timely and politically acceptable fashion, and because the sheer pace of disintegration was so rapid, the government considered extreme measures. The monetary reform known as the Austral Plan was just such a measure: an all-out attempt to stop hyperinflation. The details of the June 1985 Plan of Economic Reform, which is now called the Austral Plan, were as follows: 9

A real depreciation and a sharp increase in real public sector prices; an export and import tax, a forced saving scheme, and accelerated tax collection.

50

Rudiger Dornbusch and Juan Carlos de Pablo

A wage-price-exchange rate freeze. A new money, the Austral, and a promise not to create money to finance the budget. A conversion scale for existing contracts that would adjust them to keep real burdens unchanged in the face of an unanticipated reduction in inflation. An IMF agreement and a rescheduling agreement with creditors. The stabilization reduced inflation levels to only 1-2 percent per month. The immediate decline in inflation and the fiscal measures brought about a rapid and major shift in the budget. High real interest rates and the budget improvement created an atmosphere of an at least temporary stabilization. The black market premium vanished. For a country that had been on the verge of hyperinflation, the stabilization created an immense relief, but it also left considerable skepticism as to the possibility of stopping inflation by edict. The skepticism particularly revolved around the government’s ability to achieve sufficient budget control to reduce permanently the need for inflationary money creation. But even if skepticism existed, the stabilization proved to be an important political move and as such was a stepping stone toward a more deeply rooted stabilization. According to a public opinion survey (reviewed below in chap. 5 ) , the initial response was overwhelmingly positive. This was not the first time Argentina had used wage-price controls to try to stop inflation. Indeed, this was attempted in 1975-76, and the experience ended in an outburst of repressed inflation. The Austral Plan has not brought price stability-inflation by 1987-88 was back to more than 200 percent. The important achievements have been that inflation was brought down from more than 2,000 percent and that this was accomplished without either a major decline in economic activity, a rise in unemployment, or a reduction in the purchasing power of wages. In 1985 and early 1986 there was little risk that the stabilization would collapse in the near future. This gave rise to a confidence that could have made it both possible and fruitful for the government to concentrate on two key issues: how to achieve further budget improvement so as to bring inflation down to below 20 percent per year, and how to restore investment and growth. Unfortunately that opportunity was missed. A substantial fiscal deterioration took place and reignited high inflation. Thus the problem of growth without inflation remains unsolved. But what are the ingredients of a program of growth without inflation? We turn now to this question and to the implications for external debt and debt service in this context.

1.2 Investment, Debt, and the Budget The budget influences inflation as well as investment and growth because it influences the distribution of resources in the economy. If the government

51

ArgentindChapter 1

commands a large share of the resources, less is left for the private sector. The government may use its resources to service the external debt via noninterest external surpluses or to support consumption, or it may make them available for investment. Table 1.2 shows the budget of the consolidated government. Two points concerning the effect of the budget on the economy must be distinguished. The first is the way in which the government finances its outlays, i.e., by regular taxes, by borrowing, or by the inflation tax. The second, which may be related, is how the tax system determines the allocation of resources among sectors. As an illustration, the government can replace the inflation tax with outright taxes, and there will be little effect in the aggregate except that inflation will decline. If, however, the inflation tax declines without an offsetting increase in outright taxes, an offsetting reduction in absorption needs to occur: the government must either cut its spending or reduce its debt service. For the country as a whole there is a tradeoff among consumption, investment, and net resource transfers abroad. This tradeoff can be recognized from the GDP identity:

(2)

Output

=

Net Resource Transfer Abroad

+ Investment + Consumption,

where consumption refers to both private and public sector consumption, and investment similarly includes both the private and public sector. With a fixed amount of resources or output available (because the economy is already at full employment), the budget and the external debt strategy now determine inflation and future output potential. To show the range of options we can look at two particular scenarios. One possibility is to keep budget adjustments to a minimum so as not to interfere with consumption and yet foster growth via increased investment. This strategy requires, as (2) shows, that resource transfers abroad be stopped or even reversed. In a second scenario the government seeks both investment and continued, partial debt service. In this case, the resource shortage calls for crowding out of consumption by outright taxation or by the inflation tax. lsble 1.2

The Government Budget Year

9% of GDP ~

1980 1981 1982 1983

7.5 13.3 15.1 14.4

Year

W of GDP

~~

1984 1985 1986 1987

11.0 5.6 4.6 6.3

Source: BCRA Note: IMF definition for 1980-82 is on a budget basis; for 1983-86, on a cash basis. Data since 1983 include operating losses of the Central Bank.

52

Rudiger Dornbusch and Juan Carlos de Pablo

Over the past few years crowding out of investment, rather than of consumption, has been the rule. By maintaining relatively tight money and a strongly competitive exchange rate, the government has crowded out private investment, with consumption and transfers abroad absorbing the available resources. The adverse effect of positive and often high real interest rates on investment is all the more punishing in that uncertainty about future budget trends and debt service, hence interest rates, makes it unwise to repatriate capital or risk borrowing. Figure 1.5 shows the extraordinarily low rate of investment (as a percentage of GNP) in Argentina. Net investment in fact is zero or negative. With productive capacity stagnant, there is no source of growth in the standard of living. Hence the question is how much longer the current policy mix can be sustained without doing irreparable damage to the economy’s productive system and thus to the long-run viability of the economy. The flourishing of the underground economy is certainly a warning signal of a very undesirable trend. IMF programs for Argentina anticipate that in the absence of an official change in the debt strategy, the current account deficit (as a percentage of GDP) will gradually decline and ultimately turn toward surplus. The 1986 program, for example, anticipates that Argentina’s current account will reach a modest surplus of 0.2 percent of GDP by 1990. That means, of course, net resource transfers in the full amount of interest liabilities. This strategy, if it is to be consistent with even moderate growth of the economy’s supply side, requires a major shift in the budget so as to contain consumption. This shift can take the form of a much higher inflation tax or a much higher outright form of taxation. 25

,

70

71

12

73 .

74

15

76

77

78

79

80

Fig. 1.5 Gross investment (percentage of GDP)

81

82

03

84

85

86

81

53

ArgentindChapter 1

Latin American leaders advocate a different scenario. They argue that net resource flows need to be reversed and that the noninterest surpluses must come down. Resources need to be transferred inward again, they argue, to supplement scarce domestic saving in financing domestic investment. Such a reversal of resource flows compounds the problem of creditworthiness. If debtor countries like Argentina are experiencing difficulties in servicing the debt now, is it plausible that still more debt should be added? Feldstein (1987) has argued that some countries, in particular Brazil, can both borrow and grow without risking the buildup of an unsustainable debt. It is difficult to see that possibility in Argentina, however, except in the context of a major restructuring of the public sector. But if reliance on external resources is increased by reducing net transfers abroad, one must ask the question of how the extra space thus gained should be used. Once again a fiscal reform could translate these resources into growth of productive capacity. Using resources for consumption would simply reduce creditworthiness and thus presage yet another financial crisis. Argentina faces a crucial juncture with respect to fiscal policy. Fiscal choices today are critical because they affect inflation and growth and because there is little room left for mistakes. The external debt service is a key variable because it currently absorbs resources that could be available for growth. But resource savings due to reduced external debt service (assuming there is no debt forgiveness) can only be used productively if fiscal reform translates these savings into sharply higher growth. The decisions required to make that possible have as yet not been made. Moreover, if capital markets are unwilling to lend on a major scale, then most of the investment must be financed by reduced consumption. The policy mistakes of the 1970s directly translate into a growth crisis for the 1980s. The implications of the present effort to stabilize the budget and hence bring about growth and financial stability go far beyond the economic sphere. Political and institutional instability in Argentina resemble that of the Weimar Republic and Central Europe in the 1920s or the Fifth Republic in France. The political instability in turn influences economics because it stands in the way of continuity and farsightedness in private investment decisions. If, as has been the case in Argentina, the average tenure of a central bank president is less than a year, the situation is certainly not conducive to a long view. (No doubt, that is the reason for the clock in the antechamber of the office of the president of Argentina’s Central Bank. The inscription on the face of the clock reads “tempus fugit”.) The attempt at reconstruction underway today is thus of extraordinary significance. This also implies that increased flexibility of the external constraints associated with debt service is of particular importance.

54

Rudiger Dornbusch and Juan Carlos de Pablo

2

Some Debt History

Prior to the Great Depression, Argentina was one of the ten wealthiest countries in the world.' In this chapter we will review, from the perspective of external debt, the growth in the late nineteenth and early twentieth century and the subsequent decline that begins with the depression and World War I1 and extends to the present. We will only discuss the latter period in broad terms, although it must be understood that the Peronist experience of 1946-55 and the Frondizi years of 1959-62 have fundamental importance for the structure and performance of Argentina even today. Episodes such as the attempt at modernization under Frondizi or the Krieger-Vasena stabilization of 1967-69 remain important events in Argentine economic history.

2.1 Debt and Long-Run Growth In general terms, when did the economy of Argentina move ahead or fall behind? Table 2.1 shows a comparison of average growth rates of per capita income for several broad periods. Argentina became a significant country in world trade around the turn of the century. By 1912 Argentina accounted for 5.5 percent of British imports from the main trading countries, 4.2 percent of German imports, and 4 percent of French imports. The strong growth in real per capita income and in exports was rooted in the development of an export-based agricultural economy. By 1913 exports accounted for 40 percent of GDP.2 The period of decline clearly began in the 1930s. The collapse of commodity prices and the trade restrictions implied stagnation in per capita income. In 1939 per capita GDP was still at the level of 1928, and the same was true, after intermittent fluctuations, in 1945. The period from 1945 to 1974 had growth rates of per capita income equal on average to 1.9 percent. Since then, per capita income has been falling at 1.4 percent per year. Thus, even though inflation and external balance performance has often been troublesome and growth performance falls far short of that in Korea or Brazil, the extreme difficulties stem from events of the past fifteen years. Table 2.1 Period 1870- I9 I3 1900/04-1925/29 1925129- 1935139 1935139- 1980184

Average Rer Capita Growth (percent per year) Argentina

Australia

Brazil

Canada

U.S.

1.9 1.8 -0.3 1.3

0.6 0.8 0.3 2.6

n.a. 1.2 1.o 3.3

2.0 1.2 -0.9 3.2

2.0 1.3 -0.2 2.3

Sources: Cavallo (1986), Maddison (1982), and Elias (1987) n.a.

=

not available.

Argentidchapter 2

55

2.2 The Late Nineteenth Century Argentine economic development in the nineteenth century was financed by borrowing in the world capital market. Peters (1934) reviews the history of Argentine debt in this period and documents the waves of lending and subsequent defaults. His estimates place debt in 1879-80 at 300 percent of GDP and at 360 percent in 1890-91. Speaking of the 1890s, Peters notes, “Within a decade exports had doubled, imports had tripled, and debt had quadrupled. In no single year was there a budget surplus” (35). Hyndman ([1892] 1967, 153-54) writes of this period: Buenos Ayres surpassed every other city in its luxury, extravagance, and wholesale squandering of wealth. There was literally no limit to the excesses of the wealthier classes. While money, luxuries, and material poured in on the one hand, crowds of immigrants from Italy and other countries flocked in to perpetuate the prosperity of the new Eldorado of the South. Railways, docks, tramways, water-works, gas-works, public building, mansions, all were being carried on at once in hot haste. Debt service difficulties developed into a full-fledged debt crisis in 1890, spreading far beyond Argentina. Table 2.2 shows data on trade, the budget, and borrowing. Until 1890 large rates of capital inflow financed trade deficits and debt service. Ford ([1962] 1983, 140-41) observes: By 1890 this borrowing of 708 million gold pesos between 1885 and 1890 had increased the annual debt service charges to 60 million gold pesos (or 60 percent of export proceeds in 1890-a very heavy charge) of which the public sector’s share amounted to 28 million gold pesos. . . . It is important to note that because of the nature of these loans the annual Table 2.2

Argentine Trade and Finance, 1885-93 (in millions of gold pesos)

Year

Exports

1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 I896 1897

84 70 84 100 123 101 103 113 94 102 120 1I7 101

Trade Deficit

Debt Service

Foreign Borrowing

Budget Surplus

23 27 31 50

39 68 154 248 154 45 8 n.a. n.a. n.a. 17 37 38

- 13.9

8 25 33 28 42 41 36 22

2 -9 - 25

-5 -4

60 60 32 n.a. n.a. n.a. 38 40 44

Sources: Ford ([I9621 1983) and Schaefer (1922). Note: Foreign borrowing is calculated by date of debt issue. n.a. = not available.

-8.8 - 10.0 - 16.7 - 17.6

-9.0 - 14.1

-6.1 0.3 -5.9 - 10.3 - 36.0 -9.9

56

Rudiger Dornbusch and Juan Carlos de Pablo

service charges incurred had a hard core contractually fixed and payable in gold or sterling. Once the flow of loans ceased, thereby diminishing foreign currency receipts, . . . these service charges would remain. . . . Thus as the boom progressed service charge payments formed a growing portion of foreign currency payments and remained more intractable, once loans diminished and incomes fell, than imports. It is not surprising then to find the public debt service repudiated in 1891, and a moratorium arranged until 1987, for the government would have had to add to the depressive forces to raise the necessary funds-an action politically impossible. In fact, Argentina’s failure to service its debt brought down Baring Brothers, the English banking house.3 The initial debt rescheduling with the receivers of Baring Brothers (called the Rothschild Committee) was negotiated in 1890. Williams (1920, 125-28) notes two points of the deliberations that are highly relevant to present-day discussions. One is the insistence of the Argentine representative that “if the government had to buy bills of exchange in Europe, the premium on gold would probably go up so very considerably, which would make living standards unbearable, except for the richer classes, and might even cause revolution.” The other notable point concerns differences of opinion within the committee. The German and French members wanted to make a temporary loan to Argentina, believing that the country might shortly be in a position to resume payments, “but the English members refused to accede to this, for they thought it probable that at the end of six months the Argentine government would be in exactly the same position as at present. The foreign representatives then withdrew from the committee.” The initial agreement reached provided for a new bond issue to provide the means for debt service over the following three years. In return for this loan the government undertook to cancel (through budget surpluses) at least 15 million pesos each year. But, again quoting from Williams (1920, 127): The authors of it [the restructuring] had underestimated the gravity of the crisis, supposing that the temporary relief afforded by the loan would be sufficient to enable the government to assume the full burden of its foreign liabilities in 1894-a burden that would be augmented by the addition of the interest on the funding bonds to the previous obligations. The Funding Loan was at best a palliative, conceived in the spirit of the old policy which had brought Argentina to financial collapse, the policy of paying the interest on old loans with new ones. Further debt service difficulties, however, led to a large rescheduling in 1893. Interest on the three principal long-term loans was reduced for a five-year period from 5-6 percent to only 1 percent. The bond holders committee also agreed, in the so-called Romero Settlement, to a simultaneous reduction of interest payments to only 60 percent of the stated coupon rate for the same five-year period on all other debts. Full debt service was to resume in defined stages between 1897 and 1901. As Peters (1934, 47)

57

ArgentindChapter 2

notes, “the arrangement was highly successful; interest payments were resumed in full a year before the stipulated time, and Argentine bonds rose to high quotations almost immediately, indicating small loss of prestige.” The very rapid pace of immigration and the accompanying large inflow of capital stabilized debts per capita and strengthened the balance of payments. By 1899 the currency was stabilized in the Tomquist reform with a return to the gold standard. Argentina entered the twentieth century with a clean slate.

2.3 The Great Depression The sharp decline in commodity prices in 1921 brought renewed debt service problems for Argentina. Wheat prices fell by nearly 50 percent. The terms of trade which had been exceptionally favorable during World War I deteriorated by nearly 40 percent during 1925-31. The debt service problems did not last long. Recovery of the terms of trade and a wave of new lending, this time from the United States, masked the payments problems. The price of Argentine bonds which traded in the 90s at the end of the war fell to only 70 in 1920-21, but by 1923 a recovery of confidence helped push up the price. By 1926-28 full confidence was restored and Argentina was borrowing extensively. Peters (1934, 167) reports that net capital inflows from all sources averaged more than 100 million dollars per year between 1923 and 1930, or about 10 percent of the public external debt. Figure 2.1 shows net exports (in the GDP accounts) as a fraction of GDP and figure 2.2 shows the ratio of external public debt to GDP. Wartime surpluses and inflation helped reduce the debt ratio. Even with postwar

-”

-20

I

‘,,,,,,,,,,,,I 15

20

I

, , , , , , , , , , , , , , , , , , , I , , , , , , , , , , , , , , , , , , , , I , , , , I , , , , , , , , , , , , I

25

30

S5

40

45

50

55

60

Fig. 2.1 Net exports: 1913-86 (percentage of GDP)

65

70

75

80

85

58

Rudiger Dornbusch and Juan Carlos de F'ablo

n

7 220 210

1%

I\

4

./\

iy

120

I \

I I

80 70

A I

30 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1915

1920

1925

1930

1935

1940

1945

1 9 M 1955

l9sO 1965

1970

1975

1980

Fig. 2.2 Public external debt (percentage of exports)

deflation and the deficits, the 1930 ratio of debt to GDP was no higher than that prevailing on the eve of World War I.4 The 1928-33 period for Argentina represented an entirely traditional debt crisis: an adverse external shock coupled with a suspension of capital flows and trade restrictions. Kindleberger (1984, 317) notes that Latin America was hit-especially Argentina, Brazil and Colombia-by the abrupt halt in foreign lending in June 1928 when the New York market started its meteoric rise and interest rates tightened on the call money market. On this score a number of Latin American countries date the start of the Depression from the second half of 1928. . . . Whatever the merits of U.S. bankers in pushing foreign lending from 1925 to 1928, they were surely at fault in cutting it off abruptly in June of the latter year. Deflation is imposed by the Center on the Periphery, whenever the former suddenly stops lending, as in 1825, 1857, 1866, 1873, 1890 and 1907. First halting lending and then cutting way down on imports is a recipe for disaster. Table 2.3 shows key macroeconomic variables during the late 1920s and the depression years.5 Note first that the terms of trade and exports peak in 1928 and then suffer a dramatic decline. Real GDP reaches its lowest level in 1932, with a 13.5 percent decline. Just like Brazil, Argentina came out of the crisis rapidly and by 1938 had an output level 23.6 percent above the previous peak. Maddison (1985) has compared the performance of various countries during the Great Depression. Some of his conclusions are summarized in table 2.4. First, note that the experience of Argentina is not very different from that of other Latin American countries with respect to the decline in output, the export loss, and the decline in the purchasing power of exports. In order to complete the comparison, the change in the real value of debt

59

ArgentindChapter 2 Argentina During the Great Depression

Table 2.3 Year

GDP“

Agricultural GDPb

Terms of Trade‘

Price Leveld

Public Debte

Exports‘

1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938

100 107 106 119 114 106 103 I07 I I6 121 122 131

100

100 98 136 123 89 93 88

100 98 102 100 96 82 76 80

108

81

108

131

104

83 91 100 95

848 940 943 926 913 90 1 888 859 966 965 94 I 818 845

792 1,009 1,054 934 614 646 567 493 633 69 I 729 1,017 616

103 105 104 93 96

119

101 101

106 121 110 I I4

131 151 138

Sources: Vazquez-Presedo (1971) and €studios (no. 39, July-September 1986).

”Index, real GDP. bIndex, real agricultural GDP. ‘Index of the GDP deflator. dThe series was estimated from data in €studios (1986). ePublic external debt, in millions of $U.S. ‘Merchandise exports, in millions of $U.S.

Table 2.4

Comparative Performance in the Great Depression Pea!-to-Trough Decline (%)

Argentina Latin America Asia Industrial countries

GDP

Export Volume

Import Volume

Purchasing Power of Exports

-13.8 -17.3 -6.4 - 13.0

-35.8 -40.0 -22.4 -37.5

-53.2 -64.8 -30.0 -27.1

-41.9 -53.7 -31.8 - 30. I

External Shock”

1929-32

1932-37

-9.3

18.1 11.7 5.6

-11.1

0.1

Source: Maddison (1985, tables 4, 6, and 7).

“The external shock is defined as the real income loss from export volume and terms of trade changes measured as a percentage of base year real GDP.

service would have to be added. Since Argentina had a very high ratio of debt to GDP, this factor might have made Argentina’s experience appear to be particularly striking. The external shock measure represents a calculation of the real income as opposed to the GDP cost of the external shock. An adjustment is made for the reduced purchasing power of exports as well as their volume decline. The interesting point here is that while the initial loss was extremely large, so was the recovery. The base years for the decline and recovery are not the same, but an adjustment for this point would reduce the gain in 1932-37 to 15.7 percent. Thus on balance there was an exceptionally strong recovery.

60

Rudiger Dornbusch and Juan Carlos de Pablo

The reason for this is that partly as a result of several years of drought during the 1930s in the United States, Argentina experienced a major terms of trade improvement after 1933. Argentina’s response to the crisis, the terms of trade improvement notwithstanding, differs from that of other Latin American countries, notably Brazil. The Brazilian decline in real income in 1929-32 averaged only 0.4 percent versus 4.8 for Argentina; Brazil’s recovery in 1932-37 averaged 7.5 percent per year versus 5.2 in Argentina. Active domestic industrialization policy using import substitution explains Brazil’s superior performance because on the external side Brazil fared worse in 1932-37. Whereas Argentina in 1932-37 gained 18.2 percent in real income as a result of external forces, Brazil gained only 3.1 percent. However, there was another important difference: Argentina faithfully continued debt service throughout the period of adverse shock; Brazil, by contrast, declared a moratorium and ultimately wrote down its obligations in 1943. Figure 2.3 shows the price of Argentine debt in the New York market. Until 1930 there was no sign of the impending crisis. Then, with the collapse of exports and widespread default by European and Latin American countries, the price of the debt fell to less than $50 per $100 face value. But with the sharp rise in wheat prices-they doubled between 1931 and 1937-debt returned to near par and held steady until the outbreak of World war 11. Unlike virtually all other Latin American countries, Argentina did not suspend debt service in the 1 9 3 0 ~Countries .~ like Brazil relied on domestic industrialization and sacrificed debt service to gain the necessary room in the external balance. Or else, where industrialization was not the avowed policy,

100

60

50

192526 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 U 451946

Fig. 2.3 The Argentine bond price in New York (price per $100 face value)

61

ArgentindChapter 2

at least there was a vigorous attempt to stem the depression by fiscal policy. This was not the case in Argentina. Carlos Diaz Alejandro (1984, 33) writes explicitly about the tradeoff in Argentina between debt service and fiscal policies to fight recession: The counter-cyclical potency of Argentine fiscal policy during the 1930s was reduced by the increased share in total expenditures of debt-service payments, largely made to foreigners. All payments on the public debt reached 29% of expenditures in 1932; this may be compared with a meagre 5% devoted to public works. In Argentina a conservative government favored the interests of the cattle industry. The government won the continuation of access to the British market by granting privileged exchange rate and tariff treatment of British goods and by pledging continued debt service. Abreu (1984, 158) summarizes the policy of the 1930s, embodied in the Roca-Runciman Agreement of 1933, in the following terms:

Argentina’s foreign economic policy in the 1930s was defined under the heavy constraints placed by British bilateralism. Given the political basis of the Concordancia, Argentine concessions tended to assume a shape which distinctly favored cattle interests to the detriment of the national interest. This policy had costs in the long run in terms of a slower growth of the economy-and particularly of industry-than would have been the case had it made less concessions towards British interests. O’Connell (1984, 204) comments in particular on the conservative influence on policy. He writes that internationalism was a reflection of the favorable experience and the prosperity won in the past. That favorable experience in turn justified maintenance of that posture during a period of adversity: Insistence on maintaining the service of the external debt even when confronted with outright suggestions of suspension by the British authorities-in the context of the Anglo-Argentine Agreement negotiations of 1933-looks like a clear indication that such a set of ideas went beyond any external determination.

2.4 The Postwar Period through 1976 From figure 2.3 it was clear that Argentine bonds were trading near par by World War 11. The improvement in the terms of trade at the end of the 1930s and the large trade surplus accumulated during the war, which is strikingly apparent from figure 2.1, gave Argentina the means to retire its external debt and to buy from an impoverished Britain its direct investment in Argentina, in particular the railroads. Blocked sterling deposits of the BCRA with the Bank of England were the means with which the purchases were financed.

62

Rudiger Dornbusch and Juan Carlos de Pablo

Thus, by the late 1940s Argentina’s bonded public external debt had been very substantially reduced. The pace of redemption and the purchase of the railroads was too fast, however, and debt problems were again apparent by 1950. This time they concerned short-term credits. A refunding loan of $125 million from the Export-Import (Ex-Im) Bank helped pay off the arrears due American creditors. Most external borrowing in the 195Os, apart from short-term trade credit, took the form of lending by the Ex-Im Bank and by the World Bank. By 1955 the bonded public external debt of Argentina stood at $575 million. The next refunding operation occurred in 1956, following the overthrow of the Peronist regime. At this time, Argentina joined the IMF and had its first Paris Club rescheduling of official debts. By 1962 debt had risen to $2.6 billion with an additional $1 billion in unguaranteed private debt.’ Table 2.5 shows the very low ratio of public external debt service in the early 1950s and the rapid buildup, particularly in the form of amortization, toward the late 1950s. Further rescheduling of debts occurred in 1962-63 and again in 1965. The 1962-63 rescheduling involved public sector principal payments, with a stretching of maturities into the late 1960s. But the rescheduling was not sufficient, so that in 1965 principal payments due in the five-year period following 1968 had to be refunded. In the late 1960s, following the refunding agreement of 1965, there was a shift in financing from suppliers’ credits and official agencies to the international capital market. Borrowing in the world capital market allowed the automatic rolling over of maturing principal payments. Funding crises of the 1950s and early 1960s had arisen from an inability to pay principal, which created a need to reschedule these payments. The world bond market made it possible to roll over these payments apparently indefinitely. Thus any increase in the debthncome ratio could be kept to low levels, as shown in figure 2.4. The external surpluses shown earlier in figure 1.4 on average paid at least part of the interest. From the vantage point of 1970, this is how Bittermann (1973, 117-21) viewed the Argentine debt situation:

Table 2.5

Ratio of Public Debt Service to Exports (in percentages)

Period

Ratio

Period

Ratio

Period

Ratio

1926-29 1930-33 1953-56

9.3 24.6 2.0

1957 1958 1959

7.4 12.5 16.0

I960 1961 1962

23.5 26.1 22.8

Source: Avramovic (1966, 46)

63

ArgentindChapter 2 70

63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 7 8 79 80 81 82 83 84 85 86

Fig. 2.4 The ratio of debt to GDP (percent)

The outstanding public and private external debt, requiring heavy amortization in 1970, will constitute a crucial problem for the next few years. Obligations under the earlier refundings have been met and are scheduled to be paid off by 1972. . . . Without a renewal of bank loans and other credits, the contractual payment of the Argentine economy would be over a billion in 1970, compared with exports of $1.8 billion. In sum, the solvency of the Argentine economy depends on its ability to renew banking credits and issue securities. The formal refunding credits will have been paid off in 1972, but there will be continued service on the debt which has replaced them. Why was there no debt crisis until the early 1980s? The primary explanation is that automatic rollover meant that as long as creditors were confident, the external debt basically had no set maturity and no limit! From 1970 to 1975 public and private external debt each doubled. New loans financed old loans. One might have thought that the Peronist experience of 1973-76 would have made operation in international capital markets difficult. But reserves were high and initially the world commodity boom made for very large trade surpluses. A massive current account deficit did emerge in 1975, but that was also the last year of the Peronist regime. The policies under Martinez de Hoz, which we will discuss in the next chapter, brought the economy rapidly around to muster unprecedented rates of capital inflow. By the late 1970s the vulnerability of the economy to external payments difficulties and, especially, to a halt of foreign lending, was far removed from daily experience and memory. The stage was thus set for a major debt crisis.

64

Rudiger Dornbusch and Juan Carlos de Pablo

3

From Martinez de Hoz to Alfonsin

In this chapter, we review the central phase of Argentine debt accumulation. Between 1976 and 1981, the gross external debt increased by $27 billion. The debthncome ratio increased from less than 20 percent to nearly 50 percent. Near the end of the 1970s, the growth of debt was primarily the counterpart of capital flight. In the 1980s it resulted from the mechanics of debt accumulation-interest on debt rolled over (at high real interest rates) into an ever-growing indebtedness. Table 3.1 lists the presidents and economics ministers of the various administrations in this period. Politically, these years can be divided into the Martinez de Hoz phase (1976-81), the transition from military to democratic regime, and the Alfonsin regime. In the Martinez de Hoz period, it is common to distinguish the years through January 1979 from the rest which are characterized by the administration of an exchange rate tablita. In the transition period, the main events are the Malvinas War and the great debt liquidation under Economics Minister J. M. Dagnino Pastore and Central Bank President Domingo Cavallo in August 1982. Finally, in the Alfonsin administration, three phases are apparent: an initial mismanagement under Economics Minister Grinspun, the Austral Plan, and the subsequent drift toward renewed inflation and loss of control. In this chapter we will discuss the period through the Austral Plan, and will highlight thc chicf macroeconomic events and their reflection in the external buildup of debt. We focus on the Martinez de Hoz period because it is central to the buildup of external debt via capital flight. 3.1 Martinez de Hoz When the military coup overthrew the Peronist regime in March 1976, the country was on the verge of hyperinflation. Consumer price inflation declined Table 3.1

Presidents and Economics Ministers, 1976-88

President

Period

Videla* Viola* Galtien* Bignone*

311976-311981 31198 1 - I21198 I 1211981 -611982 711982- 1211983

Alfonsin

1211983-

Note: An asterisk (*) denotes a military government

Economics Minister Martinez de Hoz Sigaut Alemann Dagnino Pastore Wehbe Grinspun Sourrouille

Period 311976-31198 I 311981 - 1211981 1211981-61 I982 711982- 81 1982 811982- 1211983 1211983-211985 211985-611989

65

ArgentindChapter 3

from nearly 40 percent per month in March 1976 to less than 3 percent by June, but by the end of the year inflation had risen to 150-200 percent per year (table 3.2). Inflation fighting would thus become a constant preoccupation of the Martinez de Hoz administration. Among the policies of the Martinez de Hoz administration, two measures stand out. One is the financial market reform, initiated in June 1977; the other is the exchange rate tablita, which was introduced in January 1979 and lasted until early 1981. 3.1.1

Financial Reform

In the Peronist period interest rates had been regulated and, by the end, were significantly negative. Financial repression had been an important means of financing large budget deficits. Under the new rules, the budget deficit was to be financed in the capital market rather than by money creation. Banks were free to offer interest rates and deposits and to charge for services. In order to make the transition from the repressed system to the new, free market approach, extensive banking regulation was required. Specifically, reserve requirements had to be set high enough to ensure that the banking system would be required to hold the stock of government debt already in its portfolio. This was accomplished by setting high reserve requirements, initially at 45 percent. But these requirements implied a large spread between active and passive rates. To circumvent this problem, the Central Bank introduced a system called the Cuenta de Regulacidn Monefaria, which compensated banks for reserves. There was also a charge levied on that part of demand deposits not covered by reserve requirements.’ A number of factors combined to make the cuenfu run deficits: high market interest rates and hence a high ratio of M2 to M1, and a charge on demand deposits based on an estimate of expected inflation. As a result, the deficit of the cuentu averaged 1 percent of GDP during 1977-81 and reached 2.8 percent of GDP in 1978. The effect of financial liberalization on real, active interest rates can be seen in figure 3.1. Until 1977 real interest rates were overwhelmingly negative. Following the financial reform, real rates became more nearly positive, particularly in 1980. Calvo (1987) has argued that the financial reform was an important reason for the real exchange rate appreciation seen earlier in figure 1.3. Restraint of Economic Change in the Martinez de Hoz Period (percent per year)

Table 3.2

1976:I

Inflation Growth

706 -1.1

Source: Carta Econdmico

197631

1977

1978

1979

1980

I82 2.0

176 6.3

I76 -3.4

160 6.6

101 1.1

66

Rudiger Dornbusch and Juan Carlos de F'ablo

.........................................................................

-35

75.1

75.7

76.1

76.7

77.1

77.7

78.1

76.7

79.1

79.7

80.1

80.7

81.1

Fig. 3.1 Active real interest rates (percent per month)

domestic credit creation, combined with a deficit in the budget, which was no longer financed by the Central Bank, implied that both the private and public sector were borrowing abroad. Calvo interpreted this as the country borrowing abroad to put foreign exchange in the Central Bank, which then by monetization of the reserve inflow would provide domestic money. Thus the policy amounted to a strict application of the monetary approach to the balance of payments. Table 3.3 shows the balance of payments statistics. The Calvo interpretation certainly correctly represents the period up to 1979-80. Following financial reform, and given high average real interest rates, firms preferred to borrow abroad rather than in the home market. In this period, the increase in the gross external debt has a counterpart in capital imports by firms and reserve gains by the Central Bank. This pattern is particularly clear in 1979.

Table 3.3

Cument account Noninterest Capital inflows

Firms Banks Government Balance of payments Gross external debt

The Balance of Payments and Gross External Debt (in billions of $ U S ) 1977

1978

1979

1980

1981

1982

I .3

1 .8 2.6 1.3 0.7 0.1 0.5 3.2 12.5

-0.5 0.6 4.1 4.2 -0.0 0.5 4.4 19.0

-4.1 -2.6 2.6 2.0 -0.4 0.9 -2.5 27.2

-4.7 -0.9 I .5 -1.1

- 2.4

1.8 1.3 1.1 0.2 -0.0 2.5 9.7

Source: Indicudores de Coyunruru, various issues

0.0 2.5 -3.4 35.7

2.6 -2.3 -3.1 1.1 -0.4 -5.1 43.6

67

ArgentindChapter 3

Financial reform was not the only reason for capital inflows-the certainly was part of the attraction.

tablita

3.1.2 The Tablita The disappointing progress toward disinflation led the government to announce a disinflation strategy based on reduced, preannounced rates of disinflation in December 1978. The theory behind this disinflation policy relied on the Chicago School “law of one price,” popularized by Harry Johnson and Robert Mundell. This view held that prices in any country were strongly linked to the world price level. Given this link, a policy of depreciation inevitably meant inflation, thereby requiring further depreciation which, in turn, would lead to continuing inflation. In order to break this vicious cycle of inflation and depreciation, policymakers opted to set a path of reduced exchange depreciation, hoping that it would feed through into reduced inflation. At the same time, targets were set for public sector wages and prices as well as for the expansion of domestic credit. The Plan of December 20th had these specific announcements: Public sector wages and prices were to rise at a monthly rate of 4 percent over the following eight months. Domestic credit was to grow at 4 percent per month during the first semester of 1979. The exchange depreciation was to be 5.4 percent for January 1979, and would decline from that level by 0.2 percent per month until August

1981. Three linkages from depreciation to inflation were anticipated. First, expectations of depreciation were built automatically into pricing. Reducing the rate of depreciation would lead firms to reduce their inflation forecasts and hence their price increases. Second, the rate of increase of import prices would be dampened by reduced depreciation. Third, firms competing with importables would be forced into price discipline. As can be seen, therefore, reduced depreciation was thought to bring down inflation through various channels. Particular emphasis was placed on the expectations channel. To reinforce this effect, the future path of the exchange rate was preannounced. Thus in December 1978, the authorities announced that depreciation rates would fall from 5.2 to 4 percent per month over the next eight months. This procedure was followed until early 1981 when the system broke down as a result of the progressive overvaluation shown in figure 3.2. The law of one price was also expected to apply to interest rates. Hence pre-fixing the rate of depreciation was thought to rapidly reduce domestic interest rates to the level of those prevailing abroad, adjusted of course for the rate of depreciation. With an interest rate (including spread) in New York

68

Rudiger Dornbusch and Juan Carlos de h b l o

77.1

78.1

79.1

80.1

81.1

Fig. 3.2 The real exchange rate (index 1980-82 = 100)

of 0.15 percent per month, for example, and a depreciation rate of 4 percent per month, the Argentine interest rate should be 5.6 percent. Table 3.4 shows the quarterly monthly rates of depreciation, the CPI inflation rate, the nominal interest rate, and the real exchange rate. There are two margins to consider. One is between the nominal interest rate and the rate of depreciation. This margin, if announcements are believed, directs the choice between borrowing in Argentina or abroad. When the Argentine interest rate exceeds the announced rate of depreciation, there should be borrowing in the world market. The relation between the nominal interest rate and the rate of inflation tells us about the profitability of borrowing or the accumulation of real debts. Nineteen seventy-nine is a typical year for the combination of policies to attract capital inflows. Interest rates are very high relative to the rate of depreciation. The return from borrowing in New York and lending in Buenos Aires in December 1979, for example, is above 50 percent per year! Moreover, because of high (and growing) reserves, the policy is believable. The moderate current account deficit supports the idea that the exchange rate policy is sustainable. As a result, capital inflows are exceptionally high throughout the year. The increase in external debt exceeds the capital flows, but not by a wide margin. In 1980 real interest rates turned sharply positive, which suggests that financial market integration is as imperfect as goods market integration since the government could implement a tight domestic money strategy even under a pre-fixed exchange rate regime. 3.1.3

The Collapse

The large discrepancy between depreciation and inflation meant gradual appreciation of the real exchange rate. Between March 1976 and December

69

Argentidchapter 3 The Tublita Period (quarterly percentage rates, except as noted)

Table 3.4

Interest

Real Exchange Rate

Depreciation

Inflation

Passive

Active

Morgan Guaranty

PD/PM

21.4 11.5 8.3 15.5

28.1 26.3 19.4 27.8

21.7 22.3 20.7 21.5

37.8 26.9 25.0 24.6

82 87 93 100

71 80 89

I

15.5

I1 I11 IV

14.2 12.2 10.2

31.5 27.0 27.3 11.0

20.7 20.9 23.2 20.6

23.3 23.0 25.7 23.5

I20 I30 133

I09 I09 I17 I15

14.2

19.4 18.1 19.9 18.8

134 137 141 156

I10 I14 I16 114

26.7

157

I07

1978

I 11

Ill IV

LOO

1979

110

1980

I I1 111

IV 1981 1

8.3 6.5 4.5 3. I

9.7 14.2

16.6 15.0 16.2 15.1

17.6

14.0

21.7

15.5

Sources: Fernandez (1985) and Morgan Guaranty.

Nure: The inflation rate refers to nonagricultural wholesale prices. The real exchange rate is an index with base 1978: IV = 100. The measure of Morgan Guaranty compares nonfood wholesale prices in Argentina to the trade weighted average abroad. The second measure of the real exchange rate refers to the relative wholesale price of domestic goods and imports, P,/P,+,.

1978, the real exchange rate (as shown in fig. 3.2) had already appreciated by 20 percent. During the first year of the tublitu, real appreciation amounted to 33 percent and was an additional 17 percent in 1980. Only in 1980 did it become widely apparent that the disinflation strategy might fail. Inflation had declined, but not by an amount sufficient to justify a belief that exchange rate-based disinflation could be pursued long enough to bring inflation down to tolerable levels. The increasingly apparent overvaluation was not the only reason for unrest. The failure of major banks, resulting from dishonesty in management, created financial uncertainty and forced the government to take over these institutions. Due to very high real interest rates that persisted for an extended period, the government had started accumulating internal debts. This combination of factors brought about a wave of capital outflows. Beginning in the second quarter of 1980, the Central Bank experienced reserve losses. For the year, these losses came to nearly $3 billion, as shown in figure 3.3. In the first quarter of 1981 reserve losses were $3 billion. The current account deteriorated in a major way beginning in 1980. The main reason for the worsening was a large increase in imports. The huge real appreciation, along with some trade liberalization, led to an increase in imports of nearly $4 billion. Interest payments rose, but the rise in exports

70

Rudiger Dornbusch and Juan Carlos de Pablo

10

-

9 -

-

8 -

0 0

3 0

E "

7 -

6 -

5 -

Fig. 3.3 International reserves (in millions of $U.S.)

still more than compensated. As 1981 began, the growth in world interest rates started to have a strong effect on the current account. These actual and prospective reserve losses reinforced the belief that the exchange rate policy could not be sustained, and thus led to even more capital flight. Another factor was important, namely, the change from one president to another. In October 1980 the presidency of Roberto Viola was announced, and, from that time on, Martinez de Hoz was a lame duck economics minister. The new president was unwilling to commit himself, ahead of time, to Martinez de Hoz-style policy. As a result, there was an overt ambiguity about continuity in the domestic economy. The public drew the obvious lesson and shifted their monies into foreign assets. The large capital flight, only incompletely offset by public sector borrowing in world markets, implied unsustainable reserve losses. By late 1980 the exchange rate announcements ceased to be believed, and in February 1981 Martinez de Hoz was forced to violate the precommitted exchange rate targets by a 10 percent devaluation on top of the 2 percent preannounced depreciation. It is interesting to ask why the program failed. For some, the answer is as obvious as why a car on square wheels cannot move. But at a deeper level, the program was supported by at least a tendency toward the law of one price and a tendency for interest arbitrage to occur. A model suggested by Rodriguez and Sjaastad (1979) and Dornbusch (1982) gives the essential elements. We focus the discussion here using a diagram, while in appendix A we develop the necessary equations. Two variables are at the center of attention, the real exchange rate (R = P,/e) and the rate of price inflation of nontraded goods (IT). The vertical schedule in figure 3.4 shows a pre-fixed (constant) rate of depreciation (Ao). The rate of traded goods price inflation is

71

Argentidchapter 3 R ID

Fig. 3.4 The adjustment process under a tublita

assumed to be equal to the rate of depreciation. The real exchange rate is appreciating to the right of the vertical schedule and depreciating to the left. We assume that the rate of home goods price inflation increases whenever the real exchange rate depreciates. We also assume that the level of the real exchange rate and of the real interest rate affects home goods price inflation. A high relative price of home goods slows down inflation because of excess supply, as does a high real interest rate. In figure 3.4 we see the process of adjustment to a reduction in the rate of depreciation. We start at point A and the rate of depreciation is permanently reduced from A, to A , . The immediate effect of reduced depreciation is to slow the rate of inflation of traded goods. Home goods price inflation is affected through two separate channels: on one hand, there is a reduced rate of inflation via expectations effects, and on the other, the decline in the nominal interest rate resulting from international interest arbitrage would increase demand. We assume the former effect dominates so that inflation immediately starts slowing down.3 But, as is apparent from the figure, the slowing of inflation is only gradual, while the reduction in depreciation has moved ahead. Hence the real exchange rate is appreciating. Over time, inflation and the real exchange rate follow the path indicated by the arrows. Inflation does fall, but the real exchange rate keeps appreciating. The model suggests that given enough time the economy will converge to a point where the real exchange rate has declined to the initial level. Note, however, that at the very moment when inflation is first reduced to the target rate, at point D,victory on the inflation front comes at a price. Now the real exchange rate is highly overvalued, and the disinflation process must continue. Inflation has to fall below the rate of depreciation (via recession in the economy) until competitiveness is restored. Only at

72

Rudiger Dornbusch and Juan Carlos de Pablo

that point (at B ) will inflation settle down to the lower rate of depreciation and real interest rates return to the initial level. The problem is that on the way, at a point such as D , the large real appreciation poses credibility problems. Reserve losses can emerge because of fear that the authorities will not sit out the full adjustment. This is, in fact, what occurred in late 1980 and early 1981. The difficulty with what Fernandez (1985) has called the “expectations management approach” is primarily that the adjustment is not rapid and hence may not survive changes of regime. There is also, of course, the difficulty of accumulating current account deficits, which affects the long-run sustainable real exchange rate. The data in table 3.4 confirm the pattern of real appreciation, but do not support the decline in real interest rates predicted by this model. Thus, incomplete integration of the capital market is as much of a difficulty as the poor integration of goods markets. Of course, if capital markets were more integrated, the sharp reduction in the real interest rate could make disinflation more difficult. The slow disinflation is explained here in terms of inflation inertia, but we must also mention separately the issue of the budget. Budget deficits continued to be large, and hence on the fiscal side there was certainly no support to disinflation. Moreover, a point Carlos Diaz Alejandro (1964) makes comes into play. He argues that in a situation of real depreciation, the income effects and income redistribution appear rapidly, and the substitution effects work only gradually. He concludes that, as a result, a real depreciation tends to have recessionary effects. The converse, of course, applies in the situation of real appreciation which we are discussing here. The rise in the standard of living that comes from real depreciation sustains prosperity (la plutu d u k e ) for a while before substitution effects take over and create unemployment and trade problems.

3.2 Estimates of Capital Flight The extent of private capital outflows in this episode cannot be measured unambiguously. The balance of payments statistics and the debt are imperfect. Moreover, the current account statistics may misrepresent true trade transactions because of misinvoicing (for purposes of tax evasion or capital flight), and they certainly understate military imports. A possible method to calculate the increase in Argentine private assets abroad would use the debt and balance of payments statistics as fol10ws.~ The increase in assets has to be financed. The sources for this financing are the increase in debt plus the change in reserves minus the current account deficit and foreign direct investment in Argentina. Using equation ( l ) , which shows the counterpart of the increase in debt, we calculate the increase in private assets abroad (other private capital outflows) as a residual item:

ArgentinaChapter 3

73

Increase in debt = Current account deficit - Direct and long-term capital inflows Official reserve increases Other private capital outflows.

(1)

+ +

Table 3.5 gives the calculation for the period 1978-82 and shows that interest payments were financed almost entirely by the noninterest surplus. Thus virtually the entire increase in external debt has as a counterpart private accumulation of assets abroad, i.e., capital flight. Various estimates of capital flight from Argentina are reported in Rodriguez (1987) and Cumby and Levich (1987). These estimates, although the authors use different methodologies, broadly support the same conclusion: Argentina external debt has as a counterpart substantial private assets accumulated abroad. Moreover, when these asset estimates are combined with accumulated interest, they come close in magnitude to the entire external debt. Figure 2.4 above already showed the debVincome ratio of Argentina since 1970. The striking fact of the period under discussion is the dramatic rise in debt due not to conditions in the world economy which raised debt service or deteriorated commodity prices-that came later-but rather to capital flight arising from domestic macroeconomic and political experiments which backfired. It would be wrong to argue that Martinez de Hoz’s experiments are exclusively responsible for Argentina’s debt problems today, but his policies certainly reinforced the precarious financial position inherited from the Peronist period. They also left a severe mortgage for subsequent administrations. In retrospect, the Martinez de Hoz program is so striking because it reveals an arrogance of power of which only totalitarian regimes are capable.

3.3 The Transition The period between the collapse of the Martinez de Hoz program in March 1981 and the advent of democratic government in December 1983 is primarily one of disarray. Rising inflation and intensifying debt problems, the Malvinas War, and the internal debt crisis characterize this period. Table 3.5

Estimate of Capital Flight, 1978-82 (in billions of $U.S.) Current Account

Increase in Debt 26.8

Source: Dornbusch (1985a).

Interest

Noninterest

Increase in External Assets

-9.3

6.8

23.4

74

Rudiger Dornbusch and Juan Carlos de Pablo

The years from 1981 to 1985 can be best understood by considering the constraints faced by policymakers. First, there was an overriding external constraint. The inability to rollover debt automatically to finance current account deficits in the world market, following the 1982 Malvinas War and the Mexican crisis, meant that there was a permanent foreign exchange crisis. The foreign exchange crisis required high interest rates to prevent capital flight and to keep down the black market rate. It also required a competitive real exchange rate, but such a rate necessitated a reduction in real wages unless public sector prices were reduced to maintain the standard of living. This tradeoff between real wages and competitiveness was a growing source of inflation and budget difficulties in the coming years. Finally, growth could not be neglected. Policy at this time bounced around between objectives of growth and low inflation and the limits provided by the external constraint. The existing private debt played an important part in the attempt to manage the external constraint. The authorities encountered great difficulties in making private debtors maintain their external indebtedness in the face of real depreciation. Firms with dollar-denominated debts were facing the prospect of large depreciation and were consequently tempted to borrow domestically to pay off the external debt. To avoid the resulting reserve drain, two measures were taken: interest rates were raised dramatically, and exchange rate guarantees were offered. The high real interest rates led to bankruptcy problems, which we will discuss shortly. The exchange rate guarantees proved exceedingly costly later when they had to be met following a significant depreciation. They resulted in huge financing requirements as the government bought foreign exchange at a high dollar rate to give away cheap to those firms who had accepted the guarantee. Since the dollar purchases were financed by money creation, they provided the fuel for rising inflation. Another development of this period is the growing nationalization of the external debt. An alternative to exchange rate guarantees was for the government to assume directly external debts, for example, by taking over failing financial institutions. From the end of 1980 to the end of 1983, the external debt increased by $26 billion. The share of the public sector in that debt rose from 52 to 71.8 percent. The large increase in external debt meant increasing burdens on the budget. The policy of high real interest rates that was used in 1979-82, first as part of the disinflation program and then increasingly to stop capital outflows, led to a sharp rise in private debt. Figure 3.5 shows the hypothetical accumulation of real debt for someone who borrowed 1 peso in July 1977 at the outset of financial liberalization and rolled over the debt continuously at the unregulated active rate. Note, in particular, the extreme increase in indebtedness between late 1979 and mid-1981-a 60 percent rise in the real value of the debt. Such an increase in only two

75

ArgentindChapter 3

0.9 77.7

78.1

78.7

79.1

79.7

80.1

80.7

81.1

81.7

82.1

82.7

83.1

83.7

84.1

Fig. 3.5 The real value of debts (1977:6 = 1)

years outpaces any possible real return on investment. As a result, firms who had borrowed from banks were progressively moving toward bankruptcy. To cope with the internal debt problem, Economics Minister Dagnino Pastore and Domingo Cavallo, president of the Central Bank, liquidated debts in July and August 1982. The means used for debt liquidation was the fixing of nominal interest rates far below the rate of inflation. But, as is clear from figure 3.5, the measure provided only temporary relief. Debt accumulation began once again after the new economics team decided that their priority was to use tight money to fight rising inflation and continuing balance of payments deficits. Significant steps to address the problems using fiscal policy never occurred. In fact, the inflation tax was used increasingly to pay exchange rate guarantees, external debt service, and rising real wages in the public sector. We consider that process in more detail in the following chapter.

3.4 Alfonsin When the Alfonsin administration took office in December 1983, the economy was already in terrible shape. Inflation had risen from only 100 percent in 1980 to more than 400 percent. Real wages had been increased sharply in the final phase of the military government, and the real money stock-the base for the inflation tax-had been progressively eroded. M1 had declined to only 4.1 percent of GDP, down from 7.9 percent in 1980. The ratio of M4 to GDP had fallen from 27.8 percent in 1980 to only 12.4 percent.

76

Rudiger Dornbusch and Juan Carlos de Pahlo

External debt had risen steadily over the period since 1980. Table 3.6 shows the rising debt as well as the sharp increase in the interest bill. Since 1982 half of the interest had been paid using noninterest surpluses. The difficulty of achieving these surpluses was aggravated by sharply declining terms of trade beginning in 1981. In terms of resource transfers, the country had brought about a massive shift in the noninterest current account. Resource transfers that had been inward at the rate of more than 1 percent now moved outward. The shift abroad of resource transfers of 4-6 percent of GDP showed up throughout the macroeconomy in the form of high real interest rates and inflation, which were the means to crowd out enough private spending to free resources for debt service. Initial policies of the Alfonsin administration failed to effect fundamental changes. Real wages were increased, even as an IMF program was attempted. On the external debt side, the new economics minister, Grinspun, initiated a strong rhetoric that was endorsed by Alfonsin. But rhetoric notwithstanding, resource transfers remained enormous, inflation soared, and any semblance of control disintegrated when in early 1985 the country went to the brink of hyperinflation. When the Alfonsin administration came into office, expectations and hopes were limitless. By early 1985 output was declining and inflation, which on a December-to-December basis had been nearly 700 percent in 1984, accelerated sharply toward 3,000 percent. The budget deficit financed internally had risen from 6-7 percent in 1981-82 to 12 percent in 1983-84. In the second quarter of 1985 it rose to 23.7 p e r ~ e n t . ~ The government saw no option except to attack the inflation problem head on. But how? Creating a deep recession was politically excluded and economically unpromising. Relying on controls by themselves would not do much, as many previous attempts in Argentina had demonstrated. Hence the search began for a program combining orthodoxy with respect to monetary and fiscal adjustments with heterodoxy in avoiding the recessionary effect of macroeconomic restraint. In chapter 5 we study the Austral Plan, which was supposed to be the answer to the stabilization dilemma; but before doing so, we take a closer look in chapter 4 at the economics of hyperinflation. Table 3.6

External Debt and Resource Transfers (in billions of $U.S., except as noted)

External debt Interest due Current account Resource transfer (% of GDP) Source: World Bank

1980

I981

1982

1983

1984

1985

21.2 I .o -4.8 - 3.3

35.7 3.0 -4.7 - 1.2

43.6 4.4 -2.4 3.0

44.8 5.0 -2.5 3.9

47.8 5.3 -2.4

48.3 4.9

4.5

-1.0

6.0

77

ArgentindChapter 4

The Process of High Inflation

4

In this chapter we investigate the economics of high and explosive inflation.’ We develop a model of the main determinants of the inflation process and make the point that explosive inflation arises from the disintegration or “melting” of several institutions. Such an analysis shows how Argentina, from 1975 to 1980, went from about 100 percent inflation to a condition of near hyperinflation. The analysis also explains why successful stabilization is much less likely in the aftermath of a hyperinflation. Table 4.1 shows averages of monthly inflation rates for countries that recently undertook stabilization-Israel in June 1985, Argentina in July 1985, Bolivia and Peru in August 1985, Brazil in February 1986, and Mexico at the end of 1987. The table shows the stabilization following after double-digit (monthly) inflation rates and then, except in Israel, the gradual return of high inflation. For Mexico the stabilization is too recent to judge the longer run success. We develop here a model of the inflation process that highlights the roles of budget finance, tax and financial institutions, and contracts in creating a high-inflation scenario. The framework not only identifies the determinants of inflation, but has as its principal objective to explain the mechanics of the very sharp acceleration that occurred on several occasions. In the first Recent High-Inflation Experiences (annual and quarterly averages of monthly inflation rates)

Table 4.1

Period

Argentina

Bolivia

Brazil

Mexico

Peru

Israel

I980 1981 1982 I983 1984 1985:l 1985:2 1985:3 1985:4 1986: 1 1986:2 1986:3 1986:4 1987:l 1987:2 1987:3 1987:4 1988:l 1988:2

6.0 6.2 8.5 13.2 18.0 24. I 28.4 3.6 2.5 3.1 4.4 7.6 5.4 7.4 5.2 11.9

3.3 2.2 7.4 11.5 24.4 92.4 42.0 63. I 6.0 13.7 2.9 I .6 0.3 I .6 0.6 0.5 0.8 0.5 2.6

5.0 6.2

4.0 4.7 4.1 6.4 6.4

1.2 6.6 6.8 7.8 13.8 10.3 13.7

11.5

2.0 2.0 3.9 6.0 4.3 5.I 2.1 3.9

12.3

5.1

10. I

6.0

11.1

11.0 17.0

6.0

7.6 9.5 12.0 8.2

0.8 0.8 3.5 14.1 24.6 5.1

15.0 20.0 21.1

5.1 6.3 6.8 7.3 7.8 7.6 10.0 9.5 1.6

10.5

11.6 8.2 2.8 4.9 3.7 4.0 4.0 5.8 5.7 7.0 8.0 15.4 14.5

11.5

2. I 0.6 2.2 I .o 2.2 1.5 1.3 I .o I .5

I .8 1.6

Rudiger Dornbusch and Juan Carlos de Pablo

78

section we discuss the role of the budget in the inflation process. Then, we deal with the link among the real exchange rate, real wages, and inflation. In the third section we discuss the role of changing the frequency of contracting in accelerating inflation, and finally conclude with some remarks on why stabilization is so difficult. 4.1

Deficit Finance

There is considerable controversy in Argentina about the exact or even approximate size of budget deficits. Reliable public data, covering an extended period of time in a comparable fashion, are simply unavailable. Various series differ in their coverage of the public sector, in the distinction between budget and cash bases, and in the inclusion of certain expenditure items, especially with respect to the quasi-fiscal deficit of the Central Bank. Table 4.2 shows four different series for the government budget deficit. There are only two comprehensive, long-run time series available. One was prepared by Cavallo and Pena (1983), who measure deficits by summing government money and debt financing. For the period since 1960, FIEL (1987) has put together detailed public sector statistics. Among the attractions of these two series are the uniform methodology of construction for a long time period and their comprehensiveness in measuring the entire operation of the public sector. We also show a series for the deficit of the nonfinancial public sector published by the World Bank that is only available for the more recent years. This latter series is less comprehensive in that it does not include the quasi-fiscal deficit from Central Bank losses. Finally, we report the cash basis deficit of the nonfinancial public sector, including losses of the Central Bank, which is calculated by the Central Bank using IMF methodology. It is apparent that where the series overlap, they show widely diverging deficit estimates. Moreover, these differences become more important in the 1980s Table 4.2 Period 1960-64 1965-69 1970-74 1975-79 1980-84 1983 1984 19x5 1986 1987

Alternative Measures of the Budget Deficit (as a percentage of GDP) Cavallo-Pena

5.1 3.1 5.3 11.6 14.4 15.7 17.2 12.3 8.5 10.8

FIEL 44 2 1 24 76 16 4 15 6 25 2 96

World Bank

IMFiBCRA

5.5 8.9 12.9 16 1 12.6 6. I 4.3 8.0

14.4 II 0 5.6 46 6.3

Sources: Cavallo and Pena (1983) and update kindly providcd by A. Pena: FIEL (1987): World Bank; and BCRA.

79

Argentidchapter 4

when losses from exchange rate guarantees, for example, become an important part of government outlays. Table 4.3 shows in more detail the Cavallo-Pena measures. The table illustrates the financing of the deficit-internally by debt and money, and by external debt. The table also shows the debt service component, which includes payments related to exchange rate guarantees as well as indexation. 4.1.1

Passive Money, Deficit Finance, and Inflation

The most common view, certainly held by a group far more broad than that of the monetarists, asserts that high inflation is the result of budget deficits. If the government spends more than it receives in tax collection, the remainder is financed by creating money. That means more money, too much money, chasing too few goods, with the predictable outcome of inflation. This view, while basically correct, needs considerable refinement to be entirely correct. Three kinds of correction are essential. First, there is some room for noninflationary deficit finance. Second, deficits can also be financed by debt. Third, there is a channel of causation that runs from inflation to deficits, as well as the other way around. In this section we elaborate a model of these important qualifications. The deficit can be financed in one of three ways: with high-powered money, with domestic debt, or with foreign debt.

MIP

+ BIP + B*elP

=

gY,

Cavallo-Rna Measures of the Budget Deficit and Financing (as a percentage of GDP)

Table 4.3

Period

Total

1940-44 1945-49 1950-54 1955-59 1960-64 1965-69 1910-14 1915-79 1980-84 19x1 1982 1983 I984 1985 19x6 1981

2.1 8. I 6.6 6.0 5.1 3. I 5.3 11.6 14.4 15.2 12.6 15.7 11.2 12.3 8.5 10.8

Financed Internally

1.4 10.1 6.1 5.1 5.3 3.4 5.3 14.4 9.0 8.2 6.1 12.9 11.2 9.9 5.6 1.9

Financed Externally

Debt Service"

-4 1

09

18

2.0

0.8

-0.1 0.3 0.4 -0.3 0.0 -2.8 5.4 1 0 5.9 2.8 6.0 2.4 2.9 2.9

I .0 1.3 1.1

1.3 5.6 4.1 4.6 2.0 4.2 5.7 6.1 6.X 4.9 1.6 1.0 5.3 1.3 3.2

-

1.1 1.1

5.9 1.7 8.4 1.1 8. I 10.2 1.o 1.2 1.6

Deficit Excluding Debt Service

Source: Cavallo and Pena (1983) and update kindly provided by A. Pena.

"Includes all payments related to debt, specifically exchange rate guarantees and indexation payments.

80

Rudiger Dornbusch and Juan Carlos de Pablo

where g is the deficit ratio; Y is real GDP; M is the domestic base money; B and B* are domestic and foreign debt; and e is the exchange rate (Austral/$). A dot over a variable denotes the rate of change. It is immediately clear that deficits can be financed by borrowing from abroad or at home, entirely avoiding an increase in the money stock, at least for the time being. But it is important to stress the “for the time being” qualification. As pointed out by Dornbusch (1978) and Sargent and Wallace (198 l), a shift away from money finance may well be inflationary because the buildup in the real value of debt has no counterpart in extra taxation when the real interest rate exceeds the growth rate of output. Accordingly, there must eventually be an inflationary liquidation of debt or else a sharp increase in the inflation tax. But for the present we will not look at debt finance. Instead we will focus on the situation in which the entire deficit is financed by money. We are interested in how much inflation is generated by such a system. InJarionary Finance Financing deficits by money creation means that any money that is not in demand at the current level of prices must be forced on the public by inflation. In a growing economy some extra real balances are demanded in order to finance the growing level of transactions. But beyond that, the demand for nominal money expands only to the extent that inflation erodes the purchasing power of existing real balances. To restore their real balances (at least partially), the public has to add to nominal money holdings. Thus inflationary finance automatically creates a demand for the money issue which finances the deficit. Keynes ([1923] 1971), in his splendid description of the inflation tax, noted the potential for inflationary finance even in a country with the poorest economic and political conditions: So long as the public use money at all, the government can continue to raise resources by inflation. . . , A government can get resources by a continuous practice of inflation, even when this is foreseen by the public generally, unless the sums they seek to raise in this way are very grossly excessive. . . . . . . What is raised by printing notes is just as much taken from the public as is a beer duty or an income tax. What a government spends the public pays for. There is no such thing as an uncovered deficit. (28-29)

But, as Keynes has observed, the difficulty in this procedure of taxation is that major inflation or high erosion of the real value of money reduces the amount of money people choose to hold because they will substitute toward assets that are more inflation proof. Thus, just as high taxation erodes the tax base, high inflation leads to a reduction in real balances and hence to an increase in the rate of inflation necessary to finance a given deficit. Moreover, there may be a maximum amount of resources the government can extract.

81

ArgentindChapter 4

In appendix B we derive the relation between the budget deficit and the rate of inflation which results from financing that deficit by money creation. Equation ( 2 ) (which is derived in the appendix) shows this long-run relation between deficits and inflation:

(2)

IT

= (ag

-

YMl

-

Pg);

1>

pg,

where T and y are the rate of inflation and the growth rate of real GDP, respectively. The term a represents the noninflationary level of velocity, and f3 is the responsiveness of velocity to the rate of inflation. This equation shows that because the deficit is financed by money creation there is inflation. But it also shows that the inflationary impact of a given deficit can differ widely, depending on the financial structure and the growth rate of output. The key points of this relation are the following: The inflation rate is lower the higher the growth rate of output because when output grows strongly, so does real money demand. Accordingly, there is room for some extra money to be issued without introducing the risk of inflation. Inflation is higher the larger the budget deficit. Moreover, as will be shown in figure 4.1, this relation is very nonlinear. As the government tries to finance a larger deficit, the required rate of inflation increases steeply. Depending on the particular form of the money demand equation, there may even be a maximum deficit that can be financed by money. Going beyond that range implies hyperinflation. The inflation rate also depends on the parameters of the velocity equation, a and p. The higher the level of noninflationary velocity (i.e., because of dollarization as we argue below), the higher the rate of inflation associated with any given deficit. The parameter p, which captures the reaction of velocity to inflation, plays the same role. A high degree of responsiveness implies a larger rate of inflation. The increase in inflation brought about by a one percentage point increase in the deficit is given by the expression (a @)/(1 - bg). The higher the increase, the higher are the inflation and budget deficit from which one starts. This highlights the fact that inflationary finance exerts a very powerful impact on inflation if it is used in large doses or in an environment where a high level of velocity-and a strong responsiveness of velocity to inflation-leaves little scope for an inflation tax. Figure 4.1 represents how inflation is determined. Given the deficit ratio, say go, the corresponding inflation rate is no.As the schedule IT depends on the parameters of the velocity equation, we can immediately ask some questions. What happens if, at a point like E , a government authorizes or no longer impedes dollarization? Or what happens if the growth rate of income declines? The latter question is particularly important. As Melnick and Sokoler (1984) have shown, a decline in growth reduces seigniorage and

+

82

Rudiger Dornbusch and Juan Carlos de Pablo

2

0

l-

a

40 BUDGET S H A R E

Fig. 4.1 Budget deficits and inflation

hence raises inflation. This is shown as a shift of the TT schedule up and to the left to give T'T'. The new inflation equilibrium, given the deficit ratio, is at point E' where inflation has increased. The point of this exercise is that we must look not only at the deficit when asking how high is the inflation, but equally important are the determinants of money demand and output growth. Of course, without deficits there would be no inflation. But how much inflation there is depends critically on money demand and growth and also, as we will now show, on the tax structure. 4.1.2

The Olivera-Tanzi Effect

One of the striking effects of inflation is the erosion of the real value of taxation. The point is quite obvious: if there is any delay between accrual and payment of taxes, the inflation in the interim will mean that the real value of what is paid is lower the higher the rate of inflation. With moderate inflation it makes no difference that 1987 taxes are paid in 1988. But when inflation is high, this effect wreaks havoc with the real value of tax collection. Keynes, commenting on the impact of inflation on the budget, noted this point as did Bresciani-Turroni (1937) and Graham (1928). Harberger (19641, Olivera (1967), and Tanzi (1977, 1978) have recognized this effect in the specific context of Latin American inflation. The empirical importance of this effect, which we will refer to as the Olivera-Tanzi effect, is large whenever inflation is high and tax collection lags are long and when there is no provision for tax indexation. We can integrate into our model this dependence of the real value of tax collection on the rate of inflation by adding an equation for the determination of the real deficit ratio. (3)

g = IT)

+ ib + i*b*,

83

ArgentinafChapter 4

where u is the noninterest budget deficit, which increases with the rate of inflation as a result of the Olivera-Tanzi effect. We have also added as components of the budget deficit the domestic and external interest payments on debt. The terms b and b* denote the ratios of domestic and foreign debt to GDP. The debt terms are introduced to make explicit the possibility of shocks to the budget associated with debt service, with erosion of the real value of domestic debt, or with a real depreciation, which raises the real value of external debt in terms of GDP.* In figure 4.2 this equation is shown as the schedule gg. This schedule is drawn for a given debt ratio and capitalization coefficient. It is upward sloping because of the Olivera-Tanzi effect: the higher the rate of inflation, the larger the deficit ratio. We also give the TT schedule, which shows the dependence of inflation on the financing requirement and on the given financial structure. The diagram brings out the interdependence of the budget and inflation. 4.1.3

External Shocks and Inflation

Consider now the working of this model. Suppose, as was the case in Argentina, that the public sector has a large external debt when external debt shock occurs. Specifically, assume that prior to the disturbance, any existing external debt was rolled over with interest fully capitalized through the automatic “new money.” Suppose there is no domestic debt, and let 8 be the fraction of external debt service which is automatically financed by capitalization, so that: g =

(4)

u(7T)

+ (1

- O)i*b* T

BUDGET SHARE

Fig. 4.2

A budget shock

84

Rudiger Dornbusch and Juan Carlos de Pablo

Initially the economy will be at a point like E in figure 4.2, with inflation depending solely on the domestic deficit. Now suppose external creditors suspend voluntary lending so that the capitalization coefficient falls to zero, as occurred in 1982. It is immediately apparent that reduced access to automatic capitalization of interest payments increases the portion of the deficit that must be financed by money creation. Now the country has to earn the resources for external finance, or else finance the purchase of foreign exchange by money creation. The gg schedule shifts outward and to the right for two r e a ~ o n s First, . ~ the government will issue more money to finance the purchase of foreign exchange for interest payments (assuming, of course, that there are no expenditure cuts or tax increases). Second, there will typically be a real depreciation in order to improve the external balance. The new equilibrium at point E‘ shows a sharply higher rate of inflation. The increase in inflation is greater the larger the debt service shock and the real depreciation, but it also depends on the responsiveness of velocity to inflation and on the degree to which higher inflation erodes real tax collection. The important point to recognize is that each of these factors will increase significantly the inflationary impact of the debt shock. During high inflation episodes there will invariably be a discussion between two schools of thought. The “balance of payments school” will argue that external balance problems and the resulting depreciation of the exchange rate are the primary cause of the deficit. In the German hyperinflation of 1923, for example, this school considered reparation payments as the reason for exchange depreciation and the resulting high i n f l a t i ~ n .By ~ contrast, the “quantity theory school” will point to budget deficits financed by money creation as the reason for inflation. From figure 4.2 it is evident that the distinction between the two schools is much less clear-cut than the labels suggest. In fact, money is endogenous, and an external shock is a very plausible source of an inflationary spiral. Passive money is the essential ingredient in reconciling the quantity school and the balance of payments doctrine. Not surprisingly, suspension of reparation payments in Germany and of debt service in Bolivia in 1985 were essential steps in the stabilization of inflation. In Argentina, involuntary external debt service (after 1982) became an important source of inflation in exactly the manner the balance of payments school emphasizes. Terms of trade deterioration further aggravated the external debt shock by forcing real depreciation and hence an increase in the real value (in terms of GDP or the tax base) of the existing external debt service. 4.1.4

Endogenous Financial Innovation and Liberalization

When an inflationary process develops, we often observe an endogenous financial adaptation. The interest that traditional depository institutions can pay is typically controlled. There may be an outright limitation on interest

85

ArgentinaiChapter 4

rates, or else institutions may be required to hold reserves or government debt at controlled rates. These restrictions make institutions unable to compete in financial markets in which nominal interest rates more nearly reflect the ongoing inflation. New, unregulated financial institutions that offer depositors higher interest rates spring up and thus draw customers away from traditional depository institutions. In terms of the velocity equation, there is a fall in the ratio of conventional money to GDP. The government loses part of its inflation tax base and hence equilibrium inflation increases. In terms of equation (2) above, we have an increase in the parameters a and/or p. The government may aggravate matters if it responds to the increasing inflation by raising reserve requirements or forcing holding of government debt by traditional banks. This merely further restricts the ability of the banks to compete and accelerates disintermediation. Governments often actively (and ignorantly) promote this process, most obviously under the guise of financial liberalization. Since inflation is a tax on money (or commercial bank noninterest-bearing reserves), financial liberalization, not surprisingly, means that the public can avoid the tax on money. As a result of these incentives, financial liberalization will be pressed on the government. Velocity will rise and so will the inflation rate associated with the financing of a given deficit by money creation. It is quite clear that from an inflation point of view financial repression, not liberalization, is appropriate. Financial liberalization requires that extra tax revenue be available to avoid the inflationary impact of a reduction in the captive inflation tax base. Governments that condone dollarization, likewise, promote inflation. Dollarization is captured in equation ( 2 ) by the coefficients (Y and p. The shift from the domestic monetary base (Ml) into dollars reduces the base for the inflation tax and hence must increase inflation. Figure 4.3 shows the behavior of the ratio of M1 to GDP over the past twenty-five years. Growth in real output is hardly an explanation for the behavior of this ratio, as growth was very small over the period. The main explanation is the response of real money demand to inflation and the introduction of financial intermediaries and of dollarization in response to the potential demand for inflation hedges. It is tempting to try to fully explain inflation in terms of dollarization and new financial intermediaries in the sense that countries with stronger dollarization have higher inflation. But that would go too far since dollarization is also a response to inflation. Our point is that a government which experiences some inflation and makes dollarization easier will experience even more inflation. In terms of figure 4.4, dollarization would shift the T T schedule up and to the left. Accordingly, the rate of inflation rises for each level of the deficit. We look next at the role of contracts in the inflation process.

86

Rudiger Dornbusch and Juan Carlos de Pablo

60 61 62 63 64 65 66 67 6a 69 70 71 72 73 74 75 7.5 J J 7a 79 a0 01 a2 03 a4 85 86 a7

Fig. 4.3 The ratio of M1 to GDP (percent)

BUDGET S H A R E

Fig. 4.4 The effects of financial liberalization

4.2

Contracts

So far we have only looked at the interaction of financing requirements and the financial structure. We were able to do so by assuming full wage-price flexibility. But that, of course, is not realistic. On the contrary, an essential feature of the inflation process is a melting of contract structure. As inflation accelerates, contracts shorten, and that shortening of contracts is itself a factor that causes inflation to accelerate.

87

ArgentindChapter 4

4.2.1 The Pazos-Simonsen Mechanism Institutional wage-setting mechanisms often rely on a fixed contract length, with wage adjustments occumng at specified intervals. The adjustments are based on the cumulated increase in prices since the last adjustment. For example, earners might receive full compensation for past actual price increases at regular intervals, say yearly. Now suppose there is a shift to six-month intervals. Two interesting questions arise. The first concerns the dynamics of shifting to shorter contracts. What is the threshold for the inflationary erosion of wages that causes the shift, and what makes it economywide rather than just in a particular firm? The other interesting question is what happens when the frequency of adjustment increases still further. This point has been developed particularly by Simonsen (1983) and Pazos (1972). It is of interest here because contract deterioration is one of the important characteristics of an accelerating inflation and because exchange depreciation often plays an important role in triggering the process. If nominal wages are adjusted only periodically, the real wage follows a saw-toothed pattern. On each adjustment date, the nominal wage is increased by the inflation cumulated since the preceding adjustment, say by 50 percent. Until the next adjustment date, the real wage declines as the ongoing inflation erodes the purchasing power of the constant nominal payments. By the end of the adjustment interval, the real wage has declined below its period average. The higher the rate of inflation, moreover, the lower the average real wage, given the interval of adjustment. In a system of full, but lagged, indexation, the real wage can be lowered only by moving to a higher rate of inflation. Thus, once and for all, depreciation of the currency immediately raises the rate of inflation and erodes existing contracts, but the catch-up through indexation ensures that inflation must be pushed to an even higher rate, so that there is always some group of wage earners whose wages are still lagging the increasing rates of price increases. The same principle applies to the removal of subsidies established to correct the budget. Measures undertaken to correct competitiveness or the budget can be effective only if they achieve a reduction in the real wage, but because of full indexation that reduction can take place only if inflation is allowed to run at a higher rate. This mechanism often sets the stage for inflation explosions. Consider a country that requires adjustments in the budget and in external competitiveness. Suppose that the government lacks the political force to suspend full indexation, so that the removal of subsidies or a real exchange depreciation will speed up the inflation rate. Workers in the middle of their contracts, for example, will find that their real wages fall below what they consider a minimum standard of living. They cannot borrow, even in perfect capital markets. Hence they will call for a shorter interval between wage adjustments in order to recover the real

88

Rudiger Dornbusch and Juan Carlos de Pablo

wage losses imposed by inflation. They will ask for an advance of what they think is due. If the economy does shift from, say, six-month to threemonth indexation intervals, the inflation rate will simply d ~ u b l e But . ~ once the contract structure has moved to a three-month scheme, two facts are clear. First, it is exceptionally unlikely that the indexation structure will return spontaneously to a longer interval, even if shocks are favorable. Second, there is nothing to make the three-month interval more stable than the six-month interval that was just abandoned. New shocks will shift the economy to even more frequent adjustments and hence to correspondingly higher rates of inflation. At this stage, the exchange rate becomes critical. In his seminal study of the inflation process in Latin America, Pazos (1972, 92-93) has described the dynamics as follows: When the rate of inflation approaches the limit of tolerance, a growing number of trade unions ask for raises before their contracts become due. And management grants them. These wage increases give an additional push to inflation and bring about a further reduction of the adjustment interval. Probably the interval is initially shortened to six months, and then, successively, to three months, one month, one week, and one day. At first the readjustment is based on the cost-of-living index; but since there is a delay of one or two months or more in the publication of this index, it must soon be replaced by another. The best-known and more up-to-date of the possible indicators in Latin America is the quotation of a foreign currency, generally the U.S. dollar. This description of the inflation process makes it clear that the dramatic escalation of inflation, seemingly out of proportion to the disturbances, arises from the endogeneity of the adjustment interval. This is not really due to the direct impact on inflation of corrective exchange rate or price policies. Rather, it occurs because an increase in inflation-which may be minor but highly visible, such as a 10 percent devaluation over and above a purchasing power parity (PPP) rule, or a removal of bread subsidies-is the straw that breaks the camel’s back. It leads to an increase in the frequency of wage adjustments, which brings on a much higher inflation rate. The endogeneity of adjustment intervals is the mechanism that connects small inflation disturbances with a shift from 50 to 100 percent inflation, or beyond to hyperinflation (see table 4.1). The exact mode of the shift to increased frequency of adjustment will differ from one experience to another: the government may cave in under the impact of a strike, business may find it easier to give an “advance” on the real wage adjustment rather than risk labor unrest in the middle of a recovery or boom, or a planning minister may seek the popularity that comes from a wage policy which appears to favor labor. One way or another, the frequency will increase, and once it increases in a large part of the economy, it cannot fail to become generalized. It is interesting to note that there are no models

89

ArgentindChapter 4

to explain the actual process of shortening for the indexation period available. Cardoso and Dornbusch (1987) have drawn attention to this fact and speculate that perhaps Schelling’s (1982, chap. 3) model of group choice might be applicable. We have left unanswered the very interesting issue of the exact model of socioeconomic interaction that brings about contract shortening. Clearly inflation is the basic reason. But why are there discrete steps only: one year, six month, three month, one month, the dollar, and why does it require such large steps in inflation before the economy shifts to the next level? Perhaps the answers might come from the Schelling model, but it is certainly clear that this is a critically important research area for a better understanding of inflation dynamics. It is immediately clear from the Pazos-Simonsen mechanism that the optimal incomes policy designed to avoid creating the context for an inflationary explosion is one that monitors above all the frequency of adjustments. An entirely different view emerges with respect to exchange rate and budget policy. As long as full indexation remains, even seemingly small corrections are a dramatic threat to the stability of the inflation rate and hence may not be worth undertaking. 4.3

Dynamics

So far we have looked at the inflationary aspects of deficit finance and at the contracting process separately. The actual dynamics of the economy, however, emerge from the interaction of these processes. A stable equilibrium, as shown in figures 4.1, 4.2, and 4.4, may not actually exist. The dynamics of the interaction among deficit finance, institutional innovation in financial markets, dollarization, and shortening contracts may not lead us there, even if a stable equilibrium does exist. A more appropriate view of the inflation process is that when inflation gets to be high, it can only go higher. Moderate inflation has some inertia because there are many institutions that do not give way at the slightest sign of an inflationary shock. But when inflation rises significantly and permanently, institutions adapt. In doing so, they themselves feed the process of increasing inflation. Extreme inflation, in this view, stems from a radical melting of all institutions: near abandonment of domestic money, meaning the government must inflate at infinite rates to get any seigniorage, and contracts of short duration, namely dollar-based contracts. The most striking fact is the sharp asymmetry between low and high inflation experience. At moderate rates of inflation there is virtually no response of institutional behavior: contracts remain annual even when there is 50 or 100 percent inflation, and people hold domestic money even though they lose in real terms. But institutional adaptation accelerates when inflation gets high. The institutional meltdown takes only a few months, as recorded

90

Rudiger Dornbusch and Juan Carlos de Pablo

in table 4.4. The table shows how little time it takes to move from 15 percent per month to 50 percent per month, which qualifies for Cagan’s (1956) definition of hyperinflation. When interpreting the historical record of inflation such as that shown in tables 4.1 and 4.4, it is common to assume the presence of adaptive inflationary expectations, as did Cagan (1956). The reason is that there appears to be a significant sluggishness in the initial phases and a subsequent acceleration, which is suggestive of exactly such an expectations mechanism. Adaptive inflationary expectations are often the model device key to slowing the impact of money on inflation. An alternative and perhaps more accurate model has a stronger focus on the dynamics of deterioration of contracts, both in the goods and labor markets, and on the inflationary adaptation of financial institutions. The dynamics of institutions-their inertia at low rates of inflation and their gradual and eventually complete melting-seem to offer a more suitable framework for the study of high inflation. In a model that emphasizes the melting of institutions, the inflation process is quite naturally explosive, and there is also an accurate description of the fact that events play themselves out in shorter and shorter time intervals. As Allais (1966) has argued, the economic time interval shrinks along with contracts and maturities of financial assets until, when the economy has converged to a spot market with dollar pricing, the budget or external balance deficit leads to hyperinflation. Hyperinflation is Table 4.4

Month

The Process of Hyperinflation (monthly rate of inflation) Germany

1

2 3 4 5 6 7 8 9 10 I1 12 13 14 15 16 17 18 19 20 Source:

Austria

17 2 9 43

15 7 4

91

15 - 10 32 33 64 78 46 43 14 3 16 41 33 92 134 82

49 97 104 28 43 I45 - 13 4 43 139 205 1,216 4,126 3,773 35,879

1

Hungary

Poland

35 23 24 24

I1

-1

2 15 9 58 26 13 54 98 62 20 6 8 25 29 79

16 34 12 32 37 26 58 58 15 7 6 67 63 72 38 275 148 109 70

Greece 33 6 11 34 14 36 47 90 35 I45 152 90 100 158

60 305 349 1,909 8,894 85,507,000

Pazos (1972).

Nore: Germany, April 1922-November 1923; Austria, February 1921-September 1923; Hungary, July

1922-Febmary 1924; Poland, June 1922-January 1924; Greece, April 1943-November 1945.

91

ArgentindChapter 5

inevitable because the inflation tax, with sufficient financial adaptation, can be almost entirely evaded and hence the budget deficit cannot be financed. The Olivera-Tanzi effect, the shortening of contracts, and financial adaptation all react in a perverse way (from the point of view of stabilization) in that they widen the deficit and accelerate explosively the inflation process. 4.4

Concluding Remarks: Why Stabilization Is Difficult

The preceding discussion has an immediate bearing on the fact that stabilization is difficult and, more often than not, takes more than one attempt to succeed. In the process of high inflation all institutions melt. When stabilization is undertaken, there is neither immediate, spontaneous resumption of longer adjustment periods for wages and prices, nor an instant increase of real money demand to noninflationary levels. As a result, more sizable adjustments in the budget are required and more dramatic measures are necessary to create the confidence that stabilization will last. Because the fiscal measures have to be extremely large, they are also extremely difficult and hence often cannot be sustained. When they fail, inflation returns instantly at an exceptionally high level because institutional inertia has not recovered. We will see in the next chapter how incomes policy-freezing exchange rates, wages, and prices-can be an effective supplement to the inevitable budget cut. It makes up for institutional inertia and, to that extent, gives a government a better chance to start stabilization. But, as is clear from the experiences of Argentina, Brazil, and Peru, failure to correct the budget means that high inflation will soon return. The decline in the ratio of M1 to GDP that occurred in 1980-85, shown in figure 4.3, was not fully reversed in the initial stabilization. As a result, financing even a moderate deficit is much more inflationary than it was prior to the experience of extremely high inflation. This hysteresis effect of high inflation (similarly apparent in contracts, pricing, and tax collection) sharply reduces the chances of stopping inflation with anything short of a dramatic budget cut.

5

The Austral Plan

In early 1985 Argentina moved to the very brink of economic disintegration. The rate of price increase accelerated by the month, reaching an annual rate of 6,000, and was still rising. At this stage the government made a decisive move: recognizing both the need for austerity measures and the political and economic obstacles to a stabilization that relied only on the demand side, the

91

ArgentindChapter 5

inevitable because the inflation tax, with sufficient financial adaptation, can be almost entirely evaded and hence the budget deficit cannot be financed. The Olivera-Tanzi effect, the shortening of contracts, and financial adaptation all react in a perverse way (from the point of view of stabilization) in that they widen the deficit and accelerate explosively the inflation process. 4.4

Concluding Remarks: Why Stabilization Is Difficult

The preceding discussion has an immediate bearing on the fact that stabilization is difficult and, more often than not, takes more than one attempt to succeed. In the process of high inflation all institutions melt. When stabilization is undertaken, there is neither immediate, spontaneous resumption of longer adjustment periods for wages and prices, nor an instant increase of real money demand to noninflationary levels. As a result, more sizable adjustments in the budget are required and more dramatic measures are necessary to create the confidence that stabilization will last. Because the fiscal measures have to be extremely large, they are also extremely difficult and hence often cannot be sustained. When they fail, inflation returns instantly at an exceptionally high level because institutional inertia has not recovered. We will see in the next chapter how incomes policy-freezing exchange rates, wages, and prices-can be an effective supplement to the inevitable budget cut. It makes up for institutional inertia and, to that extent, gives a government a better chance to start stabilization. But, as is clear from the experiences of Argentina, Brazil, and Peru, failure to correct the budget means that high inflation will soon return. The decline in the ratio of M1 to GDP that occurred in 1980-85, shown in figure 4.3, was not fully reversed in the initial stabilization. As a result, financing even a moderate deficit is much more inflationary than it was prior to the experience of extremely high inflation. This hysteresis effect of high inflation (similarly apparent in contracts, pricing, and tax collection) sharply reduces the chances of stopping inflation with anything short of a dramatic budget cut.

5

The Austral Plan

In early 1985 Argentina moved to the very brink of economic disintegration. The rate of price increase accelerated by the month, reaching an annual rate of 6,000, and was still rising. At this stage the government made a decisive move: recognizing both the need for austerity measures and the political and economic obstacles to a stabilization that relied only on the demand side, the

92

Rudiger Dornbusch and Juan Carlos de Pablo

heterodox Austral Plan was conceived, combining traditional monetary reform with incomes policy. It is certainly not an exaggeration to claim, with two years’ hindsight, that this plan was dramatically successful in arresting the disintegration process and providing breathing space in which to address the deep-seated fiscal and growth problems. In this chapter we discuss the first year of the Austral Plan,’ and then turn in the following chapter to the central challenge of fiscal reconstruction and growth.

5.1

General Remarks

Among the key features of the stabilization program were the use of wage-price controls, a fixed exchange rate, fiscal correction, and a significant expansion in the nominal quantity of money. A conversion table was introduced to adjust existing contracts to the disappearance of inflation. The combination of fiscal correction and incomes policy has come to be known as “heterodox” stabilization policy, thus opposing it to the conventional IMF programs, which emphasize tight monetary and fiscal policies as the exclusive instrument of stabilization. A main point of this chapter is to isolate the precise role played by the heterodox features of the program. Our point is that these features provide an immensely valuable breurhing spell during which price stability can be established without creating recession. The resulting strong political support for the program and for the policymaker yields a platform from which can be made the inevitable adjustments in the budget that are the pillars of stabilization. We note that mistaking the breathing spell for success and the failure to use the political support at its height for the difficult task of fiscal correction will mean that the program must soon slip. And when it does slip, it often does so irrecoverably. The focus then is on how these programs, unlike traditional programs, provide an immediate, temporary opportunity for basic policy reform. Even if they do not afford “magical” relief from the necessity of fiscal correction, they surely represent a significant opportunity to try to achieve policy reform in cases where before it had always proved impossible or was overly delayed because of the perception that there would be enormous political costs associated with high unemployment and slow disinflation. Stabilization has two dimensions: the program may or may not include fiscal correction or austerity, and it may or may not have an incomes policy (wage, price, public utility, and exchange rate freezing, together with remonetization as explained below). The standard program is the IMF approach in which there is fiscal austerity but incomes policy is not a key instrument. It is true that orthodox programs favor wage restraint, but price controls are never an item on the conditionality list. On the contrary, price liberalization tends to be an element of these programs. Heterodox programs, by contrast, combine an incomes policy with fiscal austerity.

93

ArgentindChapter 5

Stabilization without austerity-that is, relying merely on incomes policy-is not a viable approach. The most common form of this approach is to attempt stabilization by controls only, without paying attention to the sine qua non of fiscal correction. History is crowded with some thousands of years’ worth of examples of failed experiments, from Emperor Diocletian to the modem day experiments of the Peronists in 1973-74, of U.S. President Richard Nixon, of Israel in early 1985, and of the Brazilian cruzado programs. Inevitably these programs end after a more or less brief period of effectiveness, failing when shortages have become a significant political headache. Eventually the “patient,” often too late, is rushed to the IMF. There is another dimension along which programs might be distinguished. Programs can be orthodox or heterodox, and they can involve gradualism or shock treatment. In this classification the present programs are “heterodox shocks” or treatments. An example of a heterodox-gradualist program would be the Brazilian Campos-Bulhoes stabilization of 1964-67. An orthodoxgradualist program might be the case of Chile. It is more difficult to cite an orthodox-shock example, perhaps because of a lack of survivors to recount the episode. The chapter is organized as follows: In sections 5.2 to 5.4 we review conceptual issues related to the use of incomes policy in stabilization when inertia is a central feature. The discussion includes the relation between deficits and inflation studied in the preceding chapter, inertial inflation, and the basics of monetary reform. In sections 5.5 and 5.6 we review the actual stabilization experience in Argentina. The chapter concludes with a discussion of the political dimension of stabilization, showing the extraordinary political popularity of the new programs.

5.2

Incomes Policy and Stabilization

In this section we review the analytic case for wage-price-exchange rate controls as an essential, but transitory, complement to the fiscal stabilization which, as already noted, is the sine qua non of successful anti-inflation policy. Economists and well-advised policymakers have long known that aggregate demand discipline is a necessary condition for sustained price stability. Yet it may not be sufficient to stop inflation, or it may fail to work under conditions of tolerable unemployment. This is demonstrated by the failure of a number of IMF-supported programs which ignored the problem posed by inflationary inertia. The result of this disregard is that IMF programs often lead to dismal stagflation and eventually to a resumption of expansion and little or no success at permanently reducing inflation. Not surprisingly, countries such as Argentina, Israel, and Brazil recently decided to focus on the supply side of inflation, attempting to stabilize prices by combining incomes policy with “monetary reforms.” Whether or not these experiments will yield success stories depends on a number of factors

94

Rudiger Dornbusch and Juan Carlos de Pablo

including, importantly, aggregate demand management. The ovemding lesson from the ongoing experiments is the need for sound respect for fiscal discipline and the need to recognize that a good dose of initial overkill may be a necessary ingredient for success. One interesting point is that such experiments were inspired by a sound game theory approach to inflation. It can be argued, as we do below, that an incomes policy is needed in order to coordinate individual behavior in a way not recognized by oversimplified versions of a rational expectations economy. The central issue is to understand what causes inertial inflation and how incomes policy can break the dependence of the inflation rate on its past behavior. 5.2.1

Inertial Inflation

A starting point for a discussion is to recognize that a large part of high inflation is essentially inertial. This observation applies equally well to the United States, Europe, or Latin America. Inertial inflation means that inflation today is approximately what it was yesterday. Let IT be current inflation, I T - , be last year’s inflation, and Gap denote the economy’s cyclical position. The actual rate of inflation would then be: (1)

IT

= IT-,

+ ~ G U +P V ,

where v denotes current period supply shocks. The essential point of modem inflation theory is that inflation is linked to the past through a variety of channels. It is not only “too much money chasing too few goods,” or supply shocks such as oil or agricultural price increases, or real depreciation, but also that inflation yesterday means inflation today. The reason for this persistence of inertia is primarily formal or informal indexation interacting with staggered wage setting. This may take the form of a legally imposed wage rule, or it may also be that much more informal wage bargaining leads to the same result. The same mechanism also works by means of expectations. In setting their prices, firms will have to estimate their own cost increases and the price increases of competing firms. Their best guess is that, cyclical and supply shock factors aside, inflation today will be approximately what it was yesterday. Because everybody shares this best guess, the public acts on these expectations and sets prices accordingly, without hesitating to give matching wage concessions. It is much easier to give wage increases in line with expected inflation than to run the risk of a strike. If everybody acts in this manner, then expected inflation turns out to be actual inflation; and if yesterday’s inflation is the benchmark, then today’s inflation will be much the same as it was in the past. Cyclical factors and supply shocks, including a need to depreciate exchange rates to cope with a debt crisis, are the chief reasons that inflation has exploded in many countries. The inertial part of inflation, other things

95

ArgentinaChapter 5

equal, would tend to result in rather stable inflation, but the additional elements can cause inflation to move and often to move sharply. The cyclical factor is quite obvious; it is simply demand inflation or the cooling down of inflation due to slack in activity and employment. But it is worthwhile to recognize an asymmetry: there is no upper limit to firms’ price increases in response to excess demand, but in reverse the argument does not apply. Stopping an inflation of, say, 400 percent by using slack is very difficult. Even as restrictive policy reduces nominal spending, firms are forced to make wage concessions based on past inflation. Their cost increases thus might be on the order of 400 percent, and it is quite inconceivable that they would be able to reduce inflation significantly simply by reducing profit margins. In the same way, reducing wage settlements below the prevailing rate of inflation will not make much of a difference to high inflation. Giving a wage increase of 360 instead of 400 percent would mean a very large cut in the real wage (8 percent!) but a very minor reduction in inflation. The idea of fighting inflation by using slack thus applies only to an economy where wage reductions of 2 or 3 percent, or cuts on that order in profit margins, mean cutting inflation in half. When inflation is very high and very inertial, then demand policies have a hard time making a rapid and large impact. Because such an impact is the only politically acceptable method, governments of high-inflation countries have little hope except to try to stem further inflation deterioration; they cannot see any opportunity to actually end inflation. In the new stabilization programs this problem is recognized in a way not done in the IMF programs. The need is seen for an incomes policy to break the inertial forces and thus instantly shift the economy from a state of high inflation to low. This incomes policy should be understood as follows: To stop inflation, someone must start offering reductions in profit margins or real wages in order to introduce disinflation. The initial disinflation can then be passed along by indexation onto a path of gradual, additional disinflation. Realistically, there will be no volunteers for such an approach. Everybody would like to leap together to a low inflation state, but nobody will make the leap unless others will join in. That means everybody wants to see zero inflation before each person will set his own price or wage increases at zero. But if everybody adopts a “wait and see” attitude then, of course, inflation will continue. An attempt to restrict demand would translate almost entirely into reduced employment and practically not at all into lower inflation. The dismal performance of the economy and the lack of success in fighting inflation would make any such campaign short-lived. 5.2.2

Game Theory and Incomes Policy

The scenario just described puts fighting inflation squarely in the domain of game theory. When economic agents interact strategically in the fashion related above, coordination becomes essential to achieve good results. A

96

Rudiger Dornbusch and Juan Carlos de Pablo

system of temporary controls of wages, prices, and public utility and exchange rates is the coordinating device that establishes the fact of price stability, which the economy left to itself could not establish quickly except at extreme cost. It might be argued that if the government undertakes to produce the right kind of monetary and fiscal policy, then the public cannot escape the conclusion that inflation has been left “dead in its tracks.” Everybody will respond accordingly, with the result that inflation will be dead. But there are two separate and crucial holes in this argument. One concerns the government’s inability to precommit credibly to future policies. The other, which is more novel, concerns the problem of coordination in a world of price setters. We review these in turn. The government cannot commit itself definitely, credibly, and beyond doubt. The institutional setting for such a precommitment does not exist (one thinks of constitutional amendments, the gold standard, and whatnot). Because the government cannot lock in its policies beyond doubt, the public recognizes that there is always some possibility that policy will not change to a noninflationary stance. Specifically, if the average agent does not quite believe that policy will change, then all agents will behave somewhat defensively, resulting in some wage and price increases, which force the government to suspend its policy. The expectation that this will indeed be the future course persuades the average agent to disbelieve the possibility of an instantaneous end to inflation. These ideas can be interpreted in a game theory perspective, which assigns the government a double task: to ensure credibility of an aggregate demand policy consistent with disinflation, and to coordinate the expectations and actions of individual wage and price setters. Assume that after a prolonged inflation, the Central Bank announces that it will stop printing money and the Treasury announces that the budget deficit will be eliminated as a result of increased taxes or expenditure cuts. Even if the general perception is that nominal GNP will be stabilized immediately, prudent price setters would not take the lead in stopping sectoral price increases as long as they consider that further price increases in other sectors are possible. In a noncooperative or noncoordinated game with many players, each individual player has little information concerning other players’ payoffs. As a result, there is no reason to believe that all players will reach the zero inflation-full employment equilibrium in the first move of wage and price setting. Uncertainty about the behavior of other players persuades the individual price setter to adopt a very cautious pricing p01icy.~How many moves are required or how long it takes for the economy to reach the equilibrium is an open question, depending in part on the learning mechanism used by individual agents. The speed of convergence may be painfully slow after a prolonged period of high inflation. The more prolonged the experience, the higher the

97

ArgentindChapter 5

unemployment rate that results when excessive prices are confronted by a given nominal income target that is sustained by the stabilization policy. The higher and more persistent the unemployment, the more agents will be inclined to believe that the authorities’ determination will falter. Accordingly, rather than speeding up their price responses, they may persist even longer in their overly prudent disbelief. After all, the history of previous stabilization attempts has taught them a lesson. The foregoing discussion provides the rationale for an incomes policy: if governments transitorily play the role of the Walrasian auctioneer, they can speed up finding a zero-inflation full employment equilibrium. From this point of view, incomes policy may be required to make economic agents behave in line with rational expectations models. It should be stressed that the central function of controls is not to constrain individual decisionmaking, but to tell each agent how other actors will play, clearing potential externalities in an imperfect information game. This role of controls, incidentally, eliminates a traditional argument against incomes policy, namely that governments are not better equipped than the private sector to discover the equilibrium. In fact, the central problem is not to identify the equilibrium but to orchestrate the simultaneous playing of wage and price setters so as to reach the equilibrium. A more fundamental contention is that the temporary success of an incomes policy may lead policymakers to forget that price stability can only be sustained with aggregate demand discipline. The temptation is to misread the price stability and produce a boom. The misleading signals are a true risk, as can be seen from countless examples in history. Yet the converse is also true. Trying to fight a high inflation from the demand side may only lead to stagflation so dismal that policymakers may conclude that life with inflation is less uncomfortable than life with an IMF-supported program. Worse yet, they often reach that conclusion only after a prolonged period of recession. The lesson then is that economists should advocate the superior recipe of heterodox programs and, with as much emphasis and urgency, note its limitations and temptations. Of course, the chances of hitting a zero-inflation, zero-unemployment, zero-shortage equilibrium instantly using an incomes policy are remote. Wage-price controls will almost inevitably lead to some shortages unless there is a generalized recession that lowers demand. The central question is, what is worse in terms of social welfare, a few product shortages that may eventually be overcome by imports or a generalized shortage of jobs? From this point of view, objections to an incomes policy should be balanced against the extreme costs of recession and unemployment. This is especially true when the problem is to fight a high inflation that has strong inertial roots. One reason to prefer an incomes policy is that it can be managed with appropriate flexibility, across sectors and across time, moving gradually from price freezes to price administration.

98

Rudiger Dornbusch and Juan Carlos de Pablo

The case for a coordinating role for an incomes policy owing to information externalities arises only when macroeconomic noise and uncertainty are large relative to the microeconomic uncertainties in each individual market. This explains why, in a second stage of the stabilization program, removing wage-price controls gradually, in successive sectoral steps, will result in less uncertainty and lower subsequent inflation than removing controls in one shot. The one-shot approach would simply bring back the uncertainty of individual players as to what every other player will do. Then, at the stage of liberalization there would be defensively large price increases which might well wreck the inflation stabilization. Achieving consistency during the removal of controls is thus an important and difficult task. 5.2.3

Incomes Policy Matching

We conclude this section by noting that the various instruments of an incomes policy-exchange rates, wages, public and private sector prices, and the nominal money stock-must be carefully matched. Failure to align these policy instruments can easily lead to dramatically poor performance. The clearest example of a poorly aligned policy may well be the Chilean stabilization of the late 1970s. The budget has not only been balanced, but indeed there was a surplus. Money was under tight control and inflation was gradually declining, although very slowly. To speed up disinflation, the government opted to stop the exchange rate depreciation that was being used to avoid loss of competitiveness in the face of continuing inflation. But the government failed to recognize that wage indexation, which was geared to the past inflation, meant cost increases for firms without an offsetting relief on prices. The exchange rate soon became grossly overvalued, leading ultimately to the worst kind of speculation and financial in~tability.~ The need to match instruments also applies to the money stock. AS we will discuss in the next section, successful disinflation requires determined (though careful and limited) monetization of the economy.

5.3 The Budget and Inflation The common perception is that inflation is caused by budget deficits. We draw attention here to two important channels that were studied in the previous chapter to show how high budget deficits can result from inflation. This more unusual direction is important in assessing public finance in inflationary episodes and in developing a judgment about the fiscal policy changes required to implement stabilization. The first reason inflation causes deficits is the Olivera-Tanzi effect (see sec. 4.1.2). Inflation, combined with lags in tax collection, means that the real value of the tax collection arriving in the hands of the government is lower the higher the rate of inflation. If there is 100 percent inflation and no indexation of liabilities, last year’s income taxes which are paid this year

99

ArgentinaChapter 5

represent only half of the real value of taxes the government would receive without inflation or without lags or if full indexation existed. Of course, in any high-inflation episode the lags in tax collection are sharply reduced and some form of indexation will appear, but it turns out to be impossible to make the entire tax collection fully inflation proof. Even during the Argentine hyperinflation, value-added tax (VAT) collection was monthly and income tax collection was only quarterly. Thus, if inflation rates reach 20 or 30 percent per month, even very short lags have a major effect on the real value of tax collection. This effect is reinforced from the public’s active interest in postponing tax payments in order to effect a large reduction in real payments. Charges for late payments and indexation help mitigate these problems, but they do not eliminate them. This is particularly true if the charges are not indexed nor geared to real market interest rates. The actual relevance of the Olivera-Tanzi effect depends on the tax structure. The second interaction between inflation and the deficit stems from the inflation component of debt service. Part of government outlays will be for service of internal and external debt. Interest rates will reflect expected inflation and depreciation. Because of this link between inflation and nominal debt service, two different measures of the deficit should be distinguished: the actual deficit and the inflation-adjusted or operational deficit. The former calculates the deficit taking full debt service as the measure of debt service, iB = (r T)B, while the latter only includes real interest payments (re)and excludes the inflationary component of interest

+

(W.

The rationale for this distinction is that corresponding to the inflation component of interest payments is inflationary erosion of the principal. To measure the increase in the real value of government indebtedness, the inflationary erosion of the debt should be subtracted from the addition to the stock of outstanding government liabilities, which is done by calculating the inflation-adjusted or operational deficit. The importance of inflation adjustments in the budget is apparent when we think of a government with a debthcome ratio of, say, 20 percent of GNP, and an inflation rate of 200 percent. The inflationary component of debt service is equal to 40 percent of GNP. The actual deficit therefore vastly overstates the increase in the government’s real indebtedness. If inflation were brought to zero, through whatever means, the budget deficit would be reduced correspondingly.

5.3.1 Money Illusion If the public has no money illusion, then its extra saving in the presence of inflation will be just enough to absorb the inflationary component of debt service. Otherwise, how would it be possible to finance budget deficits of 25 percent of GDP? There is, indeed, a serious theoretical flaw in using unadjusted budget figures to measure supply-demand gaps. As long as an

100

Rudiger Dornbusch and Juan Carlos de Pablo

increase in the budget deficit is matched by an equal increase in voluntary private savings, aggregate demand will not be affected at all. The assumption that deficits create voluntary savings in order to finance them looks rather strange, as far as the adjusted budget is concerned. Yet, it becomes highly plausible when the problem is to finance the inflation adjustment of the public sector debt. In the case of the foreign debt, this is obvious. If the public has no money illusion, the same principle applies to domestic public sector debt. Bond holders will understand that the inflation adjustment of their credits should not lead to additional consumption. Here again, inflation automatically creates the nominal savings needed to finance the inflation adjustment of the government debt. Of course, the no-money-illusion assumption can be questioned, especially in the absence of formally indexed public sector bonds. Even in an economy with widespread indexation, widows and retired workers might use part of the inflation adjustment of their savings accounts to finance consumption. Yet, two essential points should be stressed. First, money illusion is not likely to be a major problem, otherwise it would be impossible to finance deficits of 25 percent of GDP. Second, even if money illusion is important enough to prevent the inflation adjustment of domestic public debt from being financed entirely by voluntary savings, it brings us back to the Olivera-Tanzi effect. Once again, there is a circular cause-and-effect relation between aggregate demand and inflation. Stopping inflation will lead to an automatic reduction in aggregate demand. These distinctions between adjusted and unadjusted deficits play an important part in understanding the rationale for the new stabilization programs. If a government perceives that most of the deficit comes from the inflationary component of debt service, and the noninterest budget is in sufficient surplus to pay the real interest on the debt, stabilization becomes more plausible. All that is needed is a jump from high inflation to no inflation. Failure to understand the role of inflation adjustment could make it seem that an extraordinary reduction of the deficit through spending cuts or tax increases is required before stabilization can be considered. In this sense the famous IMF dictum, “a deficit is a deficit, is a deficit” is a poor starting point for analyzing the fiscal fundamentals required for a successful stabilization.

5.4

Monetary Reform and Monetization

Stabilization of high inflation often has monetary reform as an important and highly visible component. We briefly review the essential features of such a reform. The first, which is crucial, is to shift contracts from those appropriate to an inflationary economy to those appropriate in a zeroinflation environment. The second and less significant component is the

101

ArgentindChapter 5

introduction of a new monetary unit, the main purpose of which, other than ‘‘eliminating zeros ,’’ is to increase confidence and consolidate expectations.

5.4.1

Contracts

In an inflationary environment with high uncertainty, contracts will have a very short maturity unless, like rent contracts, they are indexed. Long-term capital markets will dry up, but even so, contracts for a month or even six months will still exist. These contracts specify nominal interest rates or implicit prices that are a reflection of the inflationary expectations prevailing at the time the contracts are written. For example, with an inflation rate of 10 percent per month, a one-month loan contract will carry an interest rate of at least 10 percent per month. Rent contracts entered into at any time will involve nominal payments over, say, six months that reflect the assumption of increasing prices. Wage contracts will be indexed in a formal or informal manner, so that they are adjusted for past or for future inflation when they come up for readjustment. A major problem for inflation stabilization is to take into account the presence of these contracts and institutions that are in force at the time of stabilization. If the economy were to move instantly from high inflation to zero inflation, outstanding contracts and institutions would cause major problems. It is immediately obvious that debtors can service loan contracts involving very high nominal interest rates only if the inflation expected at the time the contract was signed actually materializes. A six-month loan concluded with an inflation expectation of 10 percent per month would carry a 77 percent interest rate for the six months. If inflation were to disappear, the nominal interest rate of 77 percent would become the real interest rate and, hence, the debt service burden would be extraordinarily large. An adjustment in all loan contracts is required to avoid a massive, unintended, and unfair redistribution from debtors to creditors and the attendant risk of pervasive bankruptcy and financial instability. For wage contracts the problem is perhaps even more complicated. Suppose, as would be realistically the case, that wages are readjusted every three or six months. Every time a contract comes up for renewal, the money wage for the next three months is adjusted upward for the inflation which actually occurred over the past three months. With such a pattern of wage contracting, an instant end to inflation is nearly impossible. Just when the government seeks to impose zero inflation, some wage contract is coming up for renewal and workers will ask to be compensated for past inflation. This wage increase in turn creates cost increases and, inevitably, inflationary pressure. Accordingly, the transition to zero inflation needs to be accompanied by some restructuring of wage contracting to avoid this inertia effect. At the time of stabilization, some wage earners will have just received their adjustment and hence will find themselves in the high real wage position of their three-month cycle, while others will be almost at the bottom. Freezing

102

Rudiger Dornbusch and Juan Carlos de Pablo

wages in this situation would be perceived as extraordinarily inequitable and hence would serve as an impediment to stabilization. Monetary reform is the broad term that characterizes the rewriting of contracts and the reform of institutions to make them compatible with the zero-inflation target. In the case of wages, monetary reform requires that those who had recent increases, and hence have high real wages, must have their wages rolled back, while those who are in the low real wage phase need upward adjustments. This reform could, in principle, be achieved by money wage adjustments using the old ~ u r r e n c y .The ~ confusion of a new money may, however, help to achieve the transition in a simplified manner. Note, too, that a new money provides an important instrument to avoid legal complications, uncertainty, and challenges to the restructuring of contracts. Similarly, for contracts involving future nominal payments, a new money is a means of aligning the real value of payments with the expectations implicit at the time contracts were concluded. A conversion scale which sets the terms for translating the old money into the new according to a set timetable of depreciation is the practical means of achieving this end. Monetary reform can also involve a capital levy on the public in the form of a write-down of monetary assets. Interestingly, in none of the recent stabilizations has this been attempted. This fact is all the more noteworthy in that, unlike in the 1920s, government debts remained very large and hence made using a capital levy to balance the budget particularly attractive. In Brazil, where domestic public debt accounts for nearly 20 percent of GNP, the idea of a capital levy is often suggested. In fact, the Brazilian government used negative real interest rates to reduce the public debt. 5.4.2 Monetization Monetary reform also often includes a change in monetary institutions. Along with a change in the monetary unit, new institutional arrangements are meant to dramatize the end of inflationary finance. The traditional way to signal the new rules of the game is to announce the independence of the Central Bank and the end of automatic financing of the budget by the printing press. But it is important to read the fine print. In the 1920s, the stabilizations did involve institutional changes and limitations on the access of the government to the printing press, but that did not result in an end to money creation for two reasons. One was that in some cases the transition was characterized by a large, once-and-for-all issue of money. In Germany, for example, in 1923 the ceiling was set at 500 percent of the existing money stock.6 But beyond this once-and-for-all fiduciary issue, there was the second reason: the possibility of the money stock increasing in the course of domestic private credit expansion or the monetization of reserve inflows. The experience from the classical stabilizations was one of extremely large increases in nominal money-several hundred percentconsistent with price stabilization.

103

Argentidchapter 5

The explanation for this large, noninflationary money creation is obvious. During the high-inflation period, the cost of holding noninterest-bearing money is extremely high. As a result, real balances decline or the velocity of circulation increases. This is the famous “flight from money.” The financial system accommodates the flight from money by creating highly liquid interest-bearing or indexed liabilities-the “overnights,” which practically serve as money. Surely, a person could not pay for a taxi ride with an indexed bond, but they could transfer funds on any day from an indexed cash reserve account to a checking account by a simple telephone call to the bank before 10 A . M . In fact, living without MI was a fashionable and profitable exercise in survival. The fact that M I survived at all, even at an extremely reduced level, can only be explained by transactions costs. In the course of stabilization, the reverse occurs. The disappearance of inflation raises the demand for M 1, or transactions-based real balances. Unless deflation increases the real money supply, it is necessary to increase the nominal money supply by one means or another to avoid extravagantly high nominal and real interest rates. A noninflationary expansion in the money stock is needed to meet the additional demand for money, the well-known reliquification or velocity problem. M I can expand to replace other financial assets without risk of renewed inflation. The problem is to fine-tune this expansion and to identify which aggregates must expand. Failure to expand M I or too gradual an increase means that the economy will slide into a recession because of a liquidity crunch. But too rapid or too large an expansion leads to a loss of credibility and the reigniting of inflation. The safest solution might well be to set growth targets for total financial assets held by the public, allowing M1 to expand as long as other financial assets decline. Thus, if the government can accommodate the shift in the composition of the financial portfolio without expanding the size of the total liquid asset portion, then both recession and renewed inflation can be avoided. What are the implications of such a policy for broad monetary aggregates, such as, for example, M4? Here one would expect no major change (relative to GNP) unless there is a significant shift from dollar assets to domestic currency securities.

5.4.3 Financial Institutions

A major issue in the sudden end to high inflation is the fate of financial institutions and, specifically, that of commercial banks and financial intermediaries. During inflation the public seeks to avoid holding money because of the depreciation of the purchasing power of money. The higher the rate of inflation, the larger the implicit tax on money, and hence the greater the resources people are willing to devote to avoiding this tax. One natural consequence is the emergence of an industry that makes it possible to live with a minimum of real balances or, equivalently, to speed up the circulation of money.

104

Rudiger Dornbusch and Juan Carlos de Pablo

Commercial banks and other financial intermediaries are the natural agents to promote the moneyless economy, and credit cards are among the vehicles. Financial institutions will establish branches in order to surround potential customers and will attempt to attract deposits by paying some interest, thus helping their customers avoid a complete loss of the purchasing power of their monetary assets. The proceeds of deposits are in turn re-lent at the high nominal interest rates commensurate with the prevailing rate of inflation. Deposit and lending rate differentials leave ample room for the costs associated with an expansion of the banking system. The common observation, then, is that during inflation, bank branches and bank employment mushroom. In Argentina this process was clearly evident. The number of branches of financial institutions increased by 42 percent between 1977 and 1985. Commercial bank branches increased by 48 percent, and branches of finance companies nearly tripled! When stabilization occurs, inflation disappears and so does the absolute size of the deposit-loan rate spread. There is then an extraordinary profit squeeze, which forces banks to close branches and sharply curtail employment. The effect is totally predictable, and it is very serious, not only from the point of view of labor relations but even more importantly from the perspective of financial stability. A successful monetary reform must take into account the fact that the financial industry is hurt by the end of inflation. Mergers, employment cuts, and a shift to fee-for-service banking will generally occur in the immediate aftermath of the stabilization.’ We turn next to an analysis of the Argentine experience under the Austral Plan.

5.5 The Austral Plan Argentina’s extraordinary macroeconomic history reflects an interplay of adverse shocks-domestic and external, economic and political. But the basic difficulties are two: the unions are inside the country, and the money is outside; the unions are British, and public finance is Italian. In June 1985 Argentina implemented the Austral Plan, the first stabilization program to combine incomes policy with an initial dose of fiscal austerity. Figure 5.1 shows the recent inflation history of Argentina. Prior to stabilization, inflation had increased from less than 100 percent per annum toward the end of the destructive stabilization attempt under Martinez de Hoz, to nearly 1,000 percent; since early 1985 there had been a sharp acceleration. In June 1985 inflation, at an annual rate, exceeded 6,000 percent!

5.5.1

History

Table 5.1 gives details on the inflation rate, real wages, the real exchange rate, industrial production, and the budget deficit.

105

ArgentindChapter 5

0 81

82

83

84

85

86

I

87

Fig. 5.1 Monthly inflation (CPI, percent per month) Table 5.1

Inflation (% per year) Budget deficit' Real wageb Real exchange rateb Industrial productionb

Key Macroeconomic Indicators 1981

1982

1983

1984

1985:l

105 13.3 96 151 94

165 15.1 82 108 91

343 16.8

627 12.8 123 112 103

918 12.0 107 107 97

100 100

100

Sources: Carta Econdmico. Coyuntura Economica, Morgan Guaranty, and BCRA.

Tonsolidated cash deficit of the public sector, including operations of the Central Bank, % of GDP. bIndex 1983 = 100.

There is no natural beginning to the recent inflation bout. The last large inflation had occurred toward the end of the Peronist administration, in mid-1975 and early 1976. At its peak, that inflation reached 35 to 50 percent per month.8 During 1976-81, the military government achieved a reduction of inflation to an annual rate of less than 100 percent, but that reduction was bought at the price of a huge overvaluation of the exchange rate, which ultimately precipitated massive capital flight and accumulation of external debt.' Successors to Finance Minister Martinez de Hoz failed to contain inflation at this level, being handicapped by the preceding debt accumulation, the Malvinas War and the accompanying credit rationing in world markets, and a terms-of-trade deterioration. A notable event in the inflation history was the deliberate attempt to reduce substantially the real value of internal debts, public and private, by the Dagnino Pastore-Cavallo team in June 1982. For the remainder of the military government, until late 1983, a holding action contained inflation

106

Rudiger Dornbusch and Juan Carlos de Pablo

below 400 percent while achieving some recovery of economic activity. Budget performance throughout was very poor, a large part of the deficit representing losses on exchange rate guarantees granted by the Central Bank in the aftermath of the Martinez de Hoz overvaluation. 5.5.2

Stabilization

The Alfonsin government came into power in December 1983. Initially, the government attempted to cope with the problem of inflation by gradualist policies. From September 1984 on, an IMF program was in force. Not much was achieved, however, in great part because large real wage increases caused competitiveness and budget performance to deteriorate, forcing repeated major devaluations. The steady worsening of inflation, even in the face of halfhearted attempts at gradualist stabilization, ultimately forced the government to explore a different direction. The shift in policies in June 1985 was triggered by two facts. The first was that the economy was well on the way to hyperinflation. Inflation had risen from only 18 percent per month at the time of elections to the 25-30 percent range. The second was that the government faced elections in November 1985 and hence could afford neither a German-style hyperinflation nor an IMF-style depression at that time. Perhaps because the possibility of a hyperinflation was so actively on the minds of the economic team, the idea of monetary reform as a comprehensive framework for stabilization moved to the center of attention. Because of the team's structuralist persuasion, an incomes policy was thought to be an indispensable part of the stabilization. At the same time, the economic team had become distinctly more orthodox since the replacement of Economics Minister Grinspun by Juan Sourrouille, and it recognized that any attempt at stabilization without budget consolidation and correction of relative prices would inevitably fail. The loss of confidence in gradualist policy and an unwillingness to accept IMF austerity per se led to the conception of the Austral Plan of June 1985. The plan struck a balance between the fundamentals-monetary and fiscal austerity-and pragmatism (or "good theory") residing in the adoption of wage-price-exchange rate controls as the central feature of the disinflation program. The key features of the plan were the following:" (a) The government increased public sector prices, depreciated the exchange rate, imposed import duties and export tariffs as well as a forced saving scheme. Some tax rates were raised and tax collection was sped up. These measures were designed to improve the budget situation and to align key relative prices prior to the freeze. l 1 (b) A wage-price freeze and a fixed exchange rate based on the U.S. dollar went into effect until further notice. The wage freeze involved a cut in the real wage because there was to be no catch-up provision for the increases in public sector prices and in the exchange rate. At the same

ArgentindChapter 5

107

time, the purchasing power of wages increased as a result of the reduced inflationary erosion of wages during the payment period. A scale of conversion to adjust outstanding contracts for the immediate and unanticipated end of inflation was put into place. A new monetary unit was introduced, the Austral. The old money was allowed to continue to be used and circulate at par. The program was accepted by the IMF and served as a basis for debt rescheduling, with new money financing a significant part of arrears and current debt service. l 2

5.5.3 Two Years Later Where has the Austral Plan taken the Argentine economy during the first two years? Table 5.2 shows the performance prior to the plan and over the subsequent four semesters. The first point to note is that inflation has been cut dramatically, but that it has not disappeared. On the contrary, by mid-1986 it is at a rate of more than 100 percent per year and rising. The second point is that economic activity, as measured by industrial production, has picked up sharply and is back to the previous peak levels of 1979-80, before the Martinez de Hoz experiment crashed. The Austral Plan thus has two faces. It has clearly failed to eradicate inflation once and for all. But it has brought inflation down from the brink of hyperinflation and has achieved its results under conditions of sharp recovery. There is no comparable experience under an IMF program. In our judgment, the Argentine experience makes the case for programs that Table 5.2

Economic krformance and Two Years of the Austral Plan 1985:I

Inflation" CPI WPI Nominal interest rate' Money growthb MI M4 Budget deficit' Real waged Real exchange rated Industrial production"

198531

1986:I

1986:U

1987:I

26.2 28.5 27.7

I .9 0.9 6.9

3.8 2.1 6.0

6.5 5.7 7.5

6.3 5.6 7.9

23.4 24.3 12.0 98 I07 97

8.5 1.9 2.5 105 93 99

4.5 6.4 3.1 106 87 Ill

4.6 6.2

4.4

5.0 109

6.1 101 76 124

83 121

5.5

Note: For 1985:11, the date refers to August-December.

"Percent per month. interest rate is the active money market rate, percent per month. 'IMF measure, including losses of the Central Bank, % of GNP. dIndex 1983 = LOO. The wage refers to the effective real wage, adjusting for purchasing power effects of inflation. The series is published by FIEL with base of December 1983= 100. %e

'Industrial production is seasonally adjusted, and the series is published by Carta Economics. Sources: Curtu Econdmico. Morgan Guaranty, and BCRA.

108

Rudiger Dornbusch and Juan Carlos de Pablo

combine fiscal consolidation with an incomes policy. We return below to the insufficient use of fiscal restraint. For the pessimist, the Austral Plan has at least given Argentina a temporary respite from a 6,000+ percent inflation. For the optimist, it has provided a breathing spell during which to consolidate public finances and restore conditions of growth under moderate (albeit high by the standards of industrial countries) inflation. Whether moderate inflation can be restored and sustained is, of course, an open question. What went right is obvious: there was a significant correction in the budget. Part of this correction took the form of a reduced domestic currency cost of debt service, while much of the correction came from a sharp increase in the real value of tax collection. This increase reflected an.end to the inflationary erosion of revenues, and the introduction of new taxes and a scheme of forced saving. Even with these corrections, the deficit has remained very large and is growing, as shown in table 5.3. Moreover, there is some question about various government contingent liabilities which do not find a place in the flow measure of the deficit, but which represent a potentially very serious threat to fiscal stability. At the same time, there remain several points of concern with respect to taxation. For those few sectors or households that actually pay taxes, rates are probably too high. In addition, some taxes-for example, export retentions on agricultural products-are directly distortionary and have ease of collection as their only justification. Thus although some fiscal consolidation is taking place, public finance remains a disaster area. Until April 1986 the government maintained a fixed exchange rate based on the U.S. dollar. Since then, a policy of mini-devaluations to avoid a deterioration of external competitiveness has been followed. The ability to hold the real exchange rate relatively constant despite domestic inflation is explained by the large depreciation of the U.S. dollar in the world currency markets. A fixed peg of the Austral to the depreciating dollar meant that Table 5.3

Period

The IMF Measure of the Total Budget Deficit (cash basis, as a percentage of GDP) Nonfinancial Public Sector

Central Bank"

Combined Deficit

13.3 8.3 3.4 7.7 1.5 3.6 5.4

1.1 2.7 2.2 4.4 0.9 1.o 0.9

14.4 11.0 5.6 12.1 2.4 4.6 6.3

1983 1984 1985 1st half 2d half 1986 1987 Source BCRA

"Losses of the Central Bank

109

ArgentindChapter 5

there was an automatic partial offset against the real appreciation caused by domestic wage and price inflation. The dollar depreciation thus helped carry the fixed exchange rate policy for a few months. The government at no time committed itself to a zero-inflation target. This is an important difference with Brazil. In Argentina the government promised fiscal stabilization and no money creation to finance the budget deficit, while in Brazil the promise was for inJucuo zero. The Argentine government was wise not to commit to zero inflation. The economy is basically closed, and the lack of competition makes it possible for oligopolistic firms and unusually aggressive unions to interact in sectoral games to raise their relative income shares. The political force of the unions means that the government cannot avoid sanctioning most of the wage increases and, for employment reasons, most of the price increases as well. The Argentine program, after instantly shifting to zero inflation for a brief moment, moved immediately into a second phase of administered inflation. An economy run with significantly more slack would perhaps have avoided the inflationary pressure, but that might have made for much worse politics. There is an interesting question as to whence came the expansion. One might, perhaps most readily, point to the large and sustained rates of money growth, but several factors point in the opposite direction. A contraction rather than expansion should have occurred when one looks at the deficit, which declined, and at real interest rates, which were extraordinarily high-70 or 80 percent per year on an inflation-adjusted basis in the second half of 1985. Among the plausible explanations for the expansion is the restoration of credit, which resulted from an increased time horizon of economic actors. In the period of intense inflation, consumer credit was unavailable so that consumer durable sales slumped. With the end of extreme inflation, consumer credit reappeared, though at extraordinary real rates, which helped promote demand and production. The restoration of credit is a counterpart to a significant remonetization of the economy. Figure 5.2 shows the real value of M1 from 1983 to 1987. It is apparent that real M1 doubled in the aftermath of the Austral program. The second explanation relates to the purchasing power of wages. The depreciation and public sector price increases on the eve of the reform reduced real wages. At the same time, however, the purchasing power of incomes may have been increased to some extent by the fact that inflation came to a halt. When inflation slows, the part of income not spent at the beginning of the month more nearly preserves its purchasing power. Thus, the halt in inflation is equivalent to a shift in real income toward labor which, in turn, has a high spending propensity. The third explanation concerns the budget. During the high inflation the deficit was financed by an inflation tax. The fiscal correction shifted the burden from money holders to those who pay outright taxes (agriculture and large firms, for example). This redistribution may also have been a source of

110

Rudiger Dornbusch and Juan Carlos de Pablo 210

,

200 190 180 170 160 150 140 130 120 110 100 90

.................................................. -

seas. unadjusted

Fig. 5.2 Real M1 (index 1985= 100)

increased purchasing power for groups which have high spending propensities. But even with these factors in mind, the expansion in demand and output may not be fully explained.

5.5.4 What Next? At this stage two major concerns about the stabilization program are quite apparent. The first is whether the government can achieve further budget correction and thus provide assurance that inflation will remain moderate. Nobody expects zero inflation, but the issue is whether inflation can be reduced to, and held, around 50 or 100 percent. Of course, there is the question of whether triple-digit inflation can be stable. But that is more a question of future shocks than of the current problems of fiscal consolidation. The second problem is the poor growth performance viewed from a trend perspective. Per capita output today is more than 10 percent lower than it was fifteen years ago, and what is worse, the decline in per capita output is bound to accelerate.13 Net investment has been zero or negative for several years in a row, and there is no prospect of a change. These facts bring out the link between stabilization and growth. Stabilization and recovery involve primarily the demand side; but at some stage in the recovery, considerations of growth and the supply side must enter. The budget correction is required not only to contain excess demand and inflationary pressure but also to promote investment. The link between the budget and investment occurs through two channels. The first is credibility effects. If the private sector anticipates deficits and hence a worsening of

111

Argentidchapter 5

inflation, they will expect the government to use tight money to make up for a lack of fiscal consolidation. Tight money presents the firm with the risk of being caught in investments with a high cost of debt service and with no customers. Hence firms are reluctant to invest, instead responding to demand using price increases and overtime rather than durable expansion in capacity and employment. The other link between the budget and investment comes from the side of resource constraints. In a fully employed economy, resources for investment can only come from reduced consumption, reduced government spending, or increased net imports. To cut government spending is the popular option, but it has proved to be politically impossible, at least in the short term. Reduced consumption can, indeed, be achieved by a fiscal tightening. This is the correct option in a country where only 4 percent of total tax collection comes from income taxes. Finally, resources can come from abroad. That would pose the problem of increased external borrowing, perhaps in the form of forced lending by the creditor banks or reduced spreads. It also brings out the need to liberalize restrictions on imports so that potential investors actually have access to foreign goods. Increased foreign borrowing should certainly be part of an investment campaign in that it bridges the short-run political problems and provides a disinflationary effect. It is doubtful that the investment and growth problem can be solved by the budget alone. The budget can, at best, help provide a favorable context in which other forces can promote investment. The most favorable sign at this time is the prospect of an Argentina-Brazil common market. Such a development would give weight to the productive forces of the economy and might well be the decisive event that restores investment and growth. With a return of growth, public finance problems and distributive quarrels might well retreat to the background. In the next chapter we turn to these questions by looking closely at the links between growth and the budget. In that context, of course, the issue of external debt service takes a prominent place. But before we move to these topics, we briefly look at the issue of the political popularity of the Austral Plan.

5.6 Politics A very interesting and important aspect of the new stabilization programs is their political impact. They were initiated by governments already sharply weakened by their failure to stabilize. But, while stabilization is perceived ordinarily as politically difficult and harmful, these new programs have catapulted the politicians and technicians who initiated them to near immortality, at least initially. Alfonsin, who was the first president to attempt a full-fledged stabilization with incomes policy support, must be given credit for his courage and confidence.

112

Rudiger Dornbusch and Juan Carlos de Pablo

Public opinion surveys provide a ready means of checking on the political success of the stabilization plans. We therefore review some of that evidence here to reinforce the dramatic contrast between the food riots that occasionally emerge in the course of IMF programs and the extraordinarily positive response in these cases.

5.6.1 Public Opinion Data from a continuing opinion survey, SOCMERC, allows an assessment over time of the public's response to policy and management.I4 The data reported in table 5.4 show the fraction of the sample assessing the performance as positive. The notable point in this table is the immensely positive response to the Austral Plan and the sharp improvement in government popularity on all counts between May and August 1985. Moreover, the positive response has been quite persistent. Even after the initial enthusiasm wore off, there was still very substantial support, as is evident from the December 1985 rating. In addition, a public opinion poll at the end of 1985 showed that 35 percent of the sample felt that the Austral Plan had helped them and a further 42 percent estimated that it had not affected them significantly, while only 9 percent felt that they had been strongly hurt. Only 18 percent of the sample felt that the plan should be abandoned.I5

5.6.2

Explanations

Why do these new programs bring forth such strong public support, whereas traditional programs are viewed as a political liability? We see two immediate explanations. Perhaps the most important aspect of these new stabilization programs is that they occur in a context of sharply accelerating and extreme inflation. It is well known that the public views inflation as utterly threatening. If this view is accurate for the moderate inflation experienced in the industrial countries, then it is quite likely even more accurate for the extreme experiences reviewed here. The strong public reaction observed in this case far exceeds what economists can explain in terms of the economics of inflation.'6 By the same token, price controls are invariably a popular political move any time inflation is perceived as a problem. Using controls to stop accelerating inflation thus amounts to a well-known recipe in political Table 5.4

Austral Plan Economic management Government in general President Alfonsin

Percentage of SOCMERC Sample Responding Positively to the Government December 1984

May 1985

August 1985

December 1985

April 1986

19

46

10 35

64

68 35 52 71

52 19 36

12

74 40 57 74

Source: La Nacion, various issues.

64

113

ArgentindChapter 5

economy. More in-depth research in the psychology of inflation and controls is therefore certainly warranted. It is striking to note that the public is actually willing to accept the need for sacrifices in order to secure stabilization. A survey in Argentina found that 46 percent of the respondents were willing to make personal sacrifices to help the success of the Austral Plan, while another 46 percent said they could not reduce their standard of living further. Only 8 percent were unwilling to do so. Fiscal stabilization is also popular in that it does not involve a reduction in aggregate real absorption of goods and services in the private sector. The adjustment only involves a change in the incidence and distribution of taxation. An inflation tax finances the government during the high inflation period, whereas outright taxation finances basically the same level of real government spending after stabilization. Since no current account improvement is required for stabilization, the whole program amounts for the most part to a reshuffling, within the country, of the burden of financing a given real level of government spending. There may also be redistributions between sectors. For example, industries that collect sales taxes (say, the tobacco industry in Argentina) lose out when the float disappears, and workers gain from the stability of real wages. These redistributive features differ from the case of a country in which stabilization requires an improvement in the current account and hence a net reduction in aggregate absorption. Of course, with rising inflation and increasing focus on fiscal problems, and a resulting emphasis on tax collection, the Austral Plan was losing popularity during 1986-87. Table 5.5 shows the evaluation from a survey for both Alfonsin and the Austral Plan. The data reported in this survey offer a preview to the poor electoral performance of the radical party in the September 1987 election.

5.7 Concluding Remarks The new stabilization programs in force in Argentina (and in Israel and Brazil) represent a critically important, viable alternative to traditional, Table 5.5

Recent Evaluation of Alfonsin and the Austral Plan (percentage of respondents) Austral Plan

Alfonsin’s Management

July 1986 December 1986 July 1987

Positive

Negative

Positive

Negative

51.0 47.5 37.0

19.0 26.5 36.5

32.0 21.5 14.5

29.5 42.0 59.0

Source: El Cronista Comercial (24 August 1987).

114

Rudiger Dornbusch and Juan Carlos de Pablo

orthodox, IMF-style programs. They grasp a central economic fact-the need for coordination rather than sheer slack-as an essential part of stabilization. From a political point of view they are dramatically successful, at least in the initial stage, and as such they are feasible. Of course, without coming to grips with the fundamental problem of the budget, this program, like the Gelbard stabilization shown in table 5.6, had to fail. Programs using incomes policy as an explicit part of stabilization represent an important advance in macroeconomic policy, but even so, they do not afford miracles. There is no substitute for a correction of fiscal disorder, the orthodox part of stabilization. Perhaps surprisingly, governments seem to be unwilling to use the very strong improvement in their political standing to follow up on the initial stabilization with a program of enduring, substantial improvement in public finance. This unwillingness is very shortsighted because any program will ultimately buckle under as a boom and shortages force the return of inflation. There may be no outright collapse, at least for a long time, but the program tends to melt away gradually for lack of sustainability, credibility, and confidence. As a result, political support inevitably falls off. And with the loss of support, the chances of achieving important changes in public finance fall. The chance of turning from stabilization to growth is thus missed. A return to sustainable, long-term growth continues to be a pressing issue. In the three years prior to the Austral Plan, investment had declined significantly, and the supply potential has expanded little since then. The important changes in public finance which are required to promote long-term financial stability have not taken place. The lesson is that the incomes policy approach to stabilization does not dispense with the need for orthodoxy, but it does provide a rare political opportunity in the form of a brief breathing spell and the momentum of popular support for the hard task of fundamental policy reform.

Table 5.6

Four Stabilization Programs (monthly Rate of Inflation)

Previous three months Month of stabilization After stabilization 1st year 2d year 3d year Source: Carta Econdmico.

Krieger-Vasena November 1967

Gelbard May 1913

Martinez de Hoz March 1977

Sourrouille June 1985

1.2 2.4

6.9 3.5

8.8 3.9

28.3 25.1

0.7 0.7 I .5

I .o 4.8 19.6

8.7 8.5 6.8

6.5 11.9

3.4

115

Argentidchapter 6

Avenues and Obstacles to Growth

6

In late 1987 the government introduced a new stabilization program that was aimed at shoring up what was left of the 1985 Austral Plan. As shown in figure 6.1, inflation in the post-stabilization period had passed from a comfortable 3-4 percent per month to an uncomfortable 6-7 percent, before moving in October 1987 into the exotic range: 20 percent inflation for consumer prices and 30 percent for wholesale prices. As were previous stabilizations, this one was doomed because it did not come to grips with fiscal necessities and the need to meet fundamental conditions for sustainable growth. The government’s dramatic loss in the congressional elections of 1987 showed the strength of the new Argentine democracy, but it also left the government critically weakened. The hope for resumption of sustained growth, with moderate inflation, was yet one more step further away. In this chapter we investigate which policies and what environment would be conducive to growth. Our conclusions, not surprisingly, are fundamentally pessimistic. The external environment continues to be unfavorable. The debt overhang requires large resource transfers abroad, and these transfers are made more difficult by very unfavorable terms of trade. The domestic economic outlook is no more encouraging: the deficit continues to be too large, taxation remains inefficient, political and economic stability are lacking, and, as a result, capital formation is altogether

32

I

85

I

86

87

Fig. 6.1 Inflation since 1985 (CPI, percent per month)

116

Rudiger Dornbusch and Juan Carlos de Pablo

inadequate. The hyperinflation of 1985 shocked Argentineans into accepting a policy directed toward national economic reconstruction. The tragedy today is that the steady decline of living standards and erosion of economic stability which Argentina has experienced for more than a decade continues unabated. Gradualism may not offer an escape from the predicament. Argentina is learning slowly and painfully to introduce realism into public sector decisions within a democratic context, an experience that is entirely new.

6.1

Conditions for Economic Progress

In 1929 Argentina was among the seven richest countries in the world. Today, the living standard has fallen far below that in advanced countries and, unlike in most other countries, is still declining. The extent of decline revealed by estimates of official GDP may overstate the actual decline because of a large and growing underground economy. But the fact of continuing and significant decline is unquestionable. The main requirements for a resumption of sustainable growth are threefold: increased and sustained stability (institutional, financial, and political); increased investment, both in the private and public sectors; and improved resource utilization and increased efficiency throughout the economy. We now develop each of these requirements in more detail. But it is essential to point out at the beginning that their separation is artificial, since the various elements are significantly interdependent. To give an example, budget problems are responsible for financial instability, and financial instability is one of the explanations for low investment. But low investment also contributes to budget instability because it implies low growth.

6.2 Investment, Saving, and Debt A vast literature on growth, in advanced and in developing countries, supports the idea that capital accumulation is one of the most important sources of growth. Capital accumulation increases potential output directly, but it also tends to be a vehicle for technical progress. Capital accumulation is important in both the private and public sectors. Public sector enterprises need to expand capacity not only with respect to infrastructure (for example, in telecommunications), but also in more traditional industries which happen to be located in the public sector. 6.2.1

A Model

We can consider a stylized model of the growth process in an open economy in order to highlight the critical restrictions and tradeoffs.' We start by

117

Argentidchapter 6

fixing the target rate of output growth. Next we use the link between growth and investment shown by the ICOR (incremental capital output ratio). The ICOR gives us the net investment required (as a fraction of GDP) to support the target growth in output. In addition, there is the investment required to offset capital depreciation. The gross investment requirement (0) is:

IIY

(1)

=

0 = ay

+ 6,

where I and Y denote real gross investment and real GDP, respectively, y is the growth rate of output, and 6 is the fraction of GDP that needs to be invested to offset capital depreciation. The ICOR is denoted by the coefficient a. Next we consider the financing of growth. Net external liabilities will increase (as a fraction of GDP) due to debt service plus the noninterest current account deficit which is equal to the ratio of consumption and gross investment to GDP less unity:

b,

(2)

=

a b,-l

+ (0 + c -11,

where b and c are the ratio of debt and consumption to GDP, and the term @ = (1 r)/(l y ) , representing the ratio of the interest factor (1 r) to the growth factor ( 1 y). Equation ( 2 ) highlights an important point. Taking the investment requirement as given, growth and an improving debt ratio may not be compatible. Specifically, suppose, as do Selowsky and van der Tak (1986), that the level of consumption grows more slowly than GDP so that the consumption ratio declines and the saving ratio increases over time. Thus even if the real interest rate exceeds the growth rate so that @ > 1, the rising saving ratio can dominate and the debthcome ratio will eventually stabilize and then decline. Such is the force of compound interest that as the saving rate rises, it first is able to finance all of the required investment and then interest payments, and ultimately an ever-increasing trade surplus, which finances debt amortization. The interesting question concerns the time path of the saving rate, the current account, and the debVGDP ratio when a country embarks on an investment strategy that yields sustained growth. Table 6.1 reports simulations of this model with parameters adapted to the case of Argentina.

+

+

+

+

Sustained Growth Scenarios

Table 6.1

Scenario B

Scenario A Year

S

FinancingGDP

DebVGDP

S

FinancingiGDP

DebVGDP

1986 1987 1990 1995 2MM

15.6 16.8 20.4 26.0 31.3

0.5 9.4 6.9 1.9 -4.4

70.0 77.2 92.7 91.7 74.8

15.6 16.4 18.8 22.7 26.4

0.5 9.8 8.6 6.1 2.8

70.0 77.6 96.9 116.4 119.0

118

Rudiger Dornbusch and Juan Carlos de Pablo

In the simulations, the initial debt/GDP ratio is set at b = 0.7, and the consumption ratio is c = 0.844, as it was in 1986. The world real interest rate is r = 0.06. The simulations assume a target output growth rate of 3 percent which implies a gross investment requirement of 22 percent. This is arrived at by assuming depreciation of 13 percent and an ICOR of 3.’ TO place these numbers in perspective, we note that growth in Argentina in 1979-80 averaged 2.3 percent, while the ratio of gross investment to GDP averaged 21.8 percent. Thus a gross investment requirement of 22 percent is certainly not unrealistically high. We now explore the time path of the debt/GDP ratio under two alternative consumption scenarios: Scenario A: Per capita consumption is maintained constant. With population growth assumed to be 1.5 percent per year, consumption thus grows at 1.5 percent while GDP grows at 3 percent. Scenario B: Per capita consumption grows at 0.5 percent per year. Even in this case, consumption growth falls short of the growth of GDP so that the saving rate rises over time. Table 6.1 summarizes the results of simulations under the two scenarios. The column labelled s shows the ratio of national saving to GDP. The financing/GDP ratio represents the current account deficit. Both scenarios show that growth is compatible with a reduction in the ratio of debt to income, but the reduction occurs only very far in the future, beyond the year 2000.3 What is the recipe for growth with debt reduction? The model highlights three critical elements:

1. There is an increase in the investment ratio which supports the growth process. The model has no room for investment plans, so the critical question of why firms would want to invest goes unanswered. With respect to public sector investment, there is the unanswered question of financing the increase in budget deficits. 2. The increase in investment (without a compensating increase in saving) initially generates an increase in the current account deficit. Borrowing finances not only interest payments but also a trade deficit. The availability of external finance to cover an increase in the trade deficit is thus a critical feature of the early stage of accelerated growth. This goes far beyond the “new money” in present concerted lending programs which only covers part of the interest bill. 3. The saving rate rises throughout and reaches Korean levels in fifteen years. The increasing saving ratio assures that the current account deficit shrinks to zero and eventually becomes a surplus that helps reduce the debt/GDP ratio. Figure 6.2 shows the simulation of the debt/GDP ratio for the two scenarios.

119

ArgentindChapter 6

1.2

-

Debt Rot io

0.8

0.7

I

2

3

4

5

6

7

8

9

10 I I

1 2 1 3 1 4 15

Time

Fig. 6.2 The debt/GDP ratio: Alternative scenarios

There is no doubt that such a development path is conceivable. Korea, in fact, followed exactly this path.4 In the case of Argentina, however, there is considerable skepticism regarding each of the three points. An increase in investment is unlikely unless a much more stable business environment develops. But even if private or public investment increased spontaneously to start the growth process, there is absolutely no reason to believe that the external financing required for growth would become available. But without external financing of the large trade deficits that come at the beginning of the growth process, growth cannot get underway. Thus the external constraint is binding and bars growth from the outset. Finally, unless the saving rate increases over time and eventually exceeds the investment rate, the development of trade surpluses to pay interest and even amortization is impossible. If external financing is the constraint, the model suggests an alternative strategy: start off the growth process with an increase in the domestic saving rate, i.e., with a reduction in per capita consumption. The increased availability of resources now finances the required increase in investment without causing the trade balance to deteriorate. But this recipe brings into play the following difficulty: Can the cut in consumption be readily translated into an external surplus? This may not be possible if investment has a significant import content. If this is impossible, there will be a recession and a foreign exchange shortage. Suppose that the problem of resource flexibility does not arise. What might a growth program look like? If we assume that the initial reduction in the consumption ratio is sufficient to finance the increased investment and that the investment requirement is 22 percent, or twice the 1986 level, a reduction in the consumption ratio by 10 percentage points or some

120

Rudiger Dornbusch and Juan Carlos de Pablo

reduction in the consumption ratio along with a suspension of debt service is required. Suspension of debt service would imply that the noninterest current account is balanced (1 = c O ) , so extra resources would become available for investment. In this case, the consumption ratio would need to fall by only 6.4 percentage points. But, as a counterpart, the debtiincome ratio would be rising. The evolution of the debt ratio over time would depend on how the consumption ratio behaves. We can distinguish two possibilities. In one case, consumption and output grow at the same rate, so the consumption ratio is constant. The noninterest current account never improves and, as a result, the debt/GDP ratio steadily increases. Effectively, the country has repudiated its debts and used the flexibility provided by abandoning debt service, along with an initial cut in consumption, to embark on growth. The alternative is, once again, a path where consumption grows less rapidly than output. Hence the saving rate increases over time, the noninterest surplus improves, and, ultimately, the debt/GDP ratio stabilizes and starts declining. Compared to the scenarios in table 6.1, the strategy of an initial reduction in consumption involves a smaller financing requirement and hence, throughout, a lower ratio of debt to GDP.

+

6.2.2

Creditors’ Preferences

We have looked so far at alternative growth strategies defined in terms of a desirable path of consumption. Another perspective is that of the creditors, who are concerned with creditworthiness as measured by the debt/GDP ratio and naturally want to minimize new money commitments. Creditors would certainly think that home-financed growth is an excellent idea, but they would also believe that the increase in the debtiincome ratio should be kept to a minimum. Table 6.2 shows the recent debt facts. The rising debt/GDP ratio and the rising level of debt would certainly lead creditors to emphasize adjustment paths with debt reduction. Creditors might, for example, recommend a focus on the debt ratio rather than on the growth of that ratio. According to their type of plan, the consumption ratio would be allowed to increase only if the debt/GDP ratio Table 6.2

The External Debt (in billions of $US. and as a percentage of GDP)

Total external debt Debt of the public sector (%) Debt to official creditors (%) DebVGDP ratio InterestiGDP ratio Resource gap (% of GDP) InvestmenVGDP ratio Source: World Bank and IMF.

1981

1982

1983

1984

1985

1986

35.8

43.6 65.6 6.4 75.9 8.6 -3.4 16.4

45.1 70.3 15.5 73.5

46.6 77.0 11.2 71.1 8.5 -5.0 12.4

48. I 85.2 16.8 74.4 7.9 -6.8 10.6

51.7 85.5 20.5 73.7 6.6 -2.6 11.9

55.8 7.9 63.1 6.8

I .5 19.5

8.8 -4.4 14.2

121

ArgentindChapter 6

fell below the target level. In this scenario, much of the growth translates immediately into trade surpluses and thus would serve debt reduction. Such an adjustment process is clearly implicit in the IMF Staff Report of July 1987 on Argentina. This report envisages a reduction in the ratio of debt to GDP from 70 to 55 percent during 1987-92. The current account deficit is expected to fall from 3.5 percent of GDP in 1986 to just above 1 percent by 1992. Essential to this outlook, as always, is the assumption that budget balancing provides the increase in saving and that an investment and an export boom are the engines of growth. A more plausible route is suggested by table 6.3. Argentina in 1984-86 was running large trade surpluses that were sufficient to pay up to half of its interest liabilities. This suggests the possibility of abandoning the policy of resource transfers, with the increase in investment financed by suspending the trade surpluses rather than by an initial increase in saving.

6.2.3 Debt Service Alternatives There are two different ways in which a suspension of the resource transfer might work. One is an outright repudiation or suspension of debt service. The alternative is a moratorium or debt write-down such as has been urged by Rodriguez (1986a, 1986b, 1986c), among others. In this case, resources in the budget that can be used for home investment become available. The suspension of payments abroad dispenses with the need for a large trade surplus and hence provides the resources from the supply side to step up growth. A more attractive opportunity seeks a compromise between the interests of creditors and the requirement that resources be available for investment and growth. Such a program might take the following form. The government engages in a comprehensive fiscal reform that entails an improvement in public sector spending programs, a major tax reform to generate revenue, and a restructuring of the external debt service process. The domestic effort centers on eliminating the budget deficit and raising national saving. External creditors would receive half the interest payments in Australs, having the right to use them to make any investment in the economy, with the sole restriction that the resources cannot be repatriated. The remaining half of the interest payments would be suspended pending reconstruction of the economy. Table 6.3

Investment and Resource Transfers (as a percentage of GDP)

Period 1970-81 1982-86 Source: World Bank

Investment

Noninterest Current Account Surplus

21.2 11.9

1.6 4.7

122

Rudiger Dornbusch and Juan Carlos de Pablo

The part of the debt under suspension would accumulate real interest at a rate equal to the growth rate of per capita real income, i.e., per capita real GDP growth adjusted for changes in the terms of trade. Thus external debt service is, in part, recycled into the economy, suspending the external transfer problem. The rest is suspended to free budget resources for investment and to dispense with inflationary finance. If the program succeeds in reconstructing the economy, creditors fully benefit from the increased ability to pay. Creditors have an interest in such a program because the current arrangement offers a lesser chance of ever recovering the credits. The reason is that the current process of debt collection destroys financial stability and bars investment. As a result, there is no mechanism to enhance the ability (hence, willingness) to pay. A focus on the transfer problem helps in understanding the links among the budget, debt service, inflation, and the lack of investment. A domestic budget effort is, of course, indispensable. Much of the inflation problem is one of fiscal laxity and the inefficiency of government. The budget problem is not exclusively a reflection of the external debt. But a restructuring of the external debt can and certainly should be part of a reconstruction effort. Reconstruction has three features. First, there is a need for real resources for investment. These become available by budget correction and by recycling external interest payments into the economy. Second, there is a need to stop inflationary finance. The suspension of trade surpluses helps because it means a reduction in the real exchange rate and hence relief in the budget. Most of the correction, though, must come from an improvement in the budget such that foreign creditors are paid with the revenue from outright taxation (or increased public sector efficiency and expenditure cuts) rather than with the revenue from the inflation tax. Third, the credibility of the program is enhanced by a formula for capitalization of external interest geared to an ability to pay rather than to an exogenous and possibly excessively high world real interest rate. We conclude this section by summarizing the main message: there is a tight link between growth and external debt service. Unless the domestic saving rate can be raised, more debt service means less growth. As we shall see presently, more debt service also means less financial, economic, and political stability.

6.3 Efficiency of Resource Allocation Increased efficiency in resource allocation is an important complement to higher rates of investment in raising the sustainable growth of output. More efficient resource utilization simply means that a given factor input produces a higher level of output. In Argentina the experience with productivity growth is exceptionally poor. This is brought out in table 6.4, which shows

123

ArgentindChapter 6 Sources of Growth (average annual percentage rate)

Table 6.4 Period _

-

_

_

GDP

Capital

Labor

Residual

4.6 3.5 3.8 1.8 -0.3

3.5 4.3 4.5

1.6 3.5 3.6

2.0 -0.8 -0.2 -1.0 -3.1

~

1960-65 1965-70 1970-75 1975-80 1980-86

4. I

I .8

1.1

1.6

Note: The estimates for 1960-80 come from Elias (1982), where capital and labor growth are quality adjusted. The 1980-86 data are reported in McCarthy (1987). The residual is obtained by weighting the factor inputs by their distributive shares. McCarthy assumes a labor share of 47 percent.

estimates of the sources of growth in real GDP for the past twenty-five years. The striking fact in this table is the extraordinarily low contribution of growth to total factor productivity. By comparison, Korea, certainly one of the strong performers among developing countries, achieved an average GDP growth rate of 8.2 percent in 1963-83. Of that growth, nearly half was due to growth in total factor prod~ctivity.~ Only in the early 1960s did Argentina come close to that performance. What reasons for this poor performance on the productivity front can be given? One reason for poor productivity growth is the inefficient trade regime, which leads to a continuing high level of protection and to variability in key relative prices. A more trade-oriented development strategy would draw resources into more efficient uses. Sturzenegger (1986) has studied the antitrade bias of protection and the implicit import tax. Table 6.5 reports various measures of distortions in relative prices. The table brings out the strong bias against trade implicit in the protectionist structure: the domestic relative price of exportables in terms of importables is about one-half the world price ratio. There is thus a powerful incentive operating against exporting and in favor of import substitution, which in itself represents highly inefficient resource allocation. The inefficiency is, of course, compounded by the limited scale of the market, Table 6.5

Antitrade Bias and Implicit Import Tax, 1%0-84

Export taxa Anti-export biasb Implicit import taxC Source: Sturzenegger (1986)

"Export tax revenue as a fraction of exports. bRatio of domestic to external terms of trade 'Ratio of import to world prices

1960-69

1970-79

11.9 48.2 63.7

14.1 56.9

51.5

1980-84

10.0 44.4 96.6

124

Rudiger Dornbusch and Juan Carlos de Pablo

which prevents important scale economies and limits competitive pressures on pricing. The inward-looking production structure and lack of competition likewise stand in the way of exposure to and adoption of foreign technology. The systematic antitrade bias is only one of the respects in which resources are inefficiently utilized. The high variability of relative prices is another example. Between 1976 and 1980 the average tariff rate on imports declined from 55 percent to only 22 percent. By 1982 it had increased to 27 percent, and in December 1985 it stood at 37 percent. Half the tariff positions had duties in excess of 30 percent. In 1982 licensing applied to only 13.5 percent of imports. By 1985 it reached 46.6 percent of all the tariff positions.6 Another example of variability in policies is the extreme variation in the real exchange rate documented in earlier chapters. For instance, the real exchange rate for “traditional” (agro-based) manufactured exports declined from an index level of 100 in 1976 to 54 by 1980, and was back up to 99 by 1983; for “nontraditional” manufactured exports the real exchange rate fell to 50 in 1980 and returned to 87 by 1983. The same variability can be found in real interest rates. Figure 6 . 3 shows the monthly real interest rate for the past eight years. Episodes with extremely high real rates, such as in 1980-81 and in 1985-86, alternate with periods of extremely low real rates, as in mid-1982. In other periods real rates are simply variable, fluctuating between positive and negative, as in 1984-85. Such a situation inevitably leads to short horizons and speculation on near-term relative prices rather than to long-run efficient use of resources.

,

10

A

8-1 6

4

2 0 -2 -4

-6 -8 -10 -12

-

-14

-

-16

79.1

80.1

81.1

82.1

83.1

84.1

85.1

86.1

Fig. 6.3 The active real interest rate (percent per month)

87.1

88.1

125

ArgentinaiChapter 6

A discussion of efficiency in resource utilization cannot avoid mention of the public sector, where resources are being used in a highly inefficient manner. In part this is the result of militant unions, in part the legacy of military patronage, and in part it is endemic to public sector enterprises. Whatever the primary explanation for the current state of affairs, Argentina has much to gain from a trimmer and more efficient public sector.

6.4 Budget Deficits and Financial Instability The third problem area is stability of institutions and policies. The lack of stability can be measured by any number of variables: the wide swings in economic philosophy of those in power, ranging from Peronism to Alsogoray; the short average tenure of finance ministers; and the variability in the level of output (as shown in figure 6.4, for example) for the manufacturing sector. However, the most immediate indicator of underlying instability is the inflation rate. During the past twenty years, Argentine inflation has always been high, and most of the time a sharp acceleration is just around the comer. As a result, economic actors are always concerned with the possibility of inflation and with the shift in politics and economic policies that will occur when inflation gets out of hand. In practice, inflation uncertainty means short horizons for production decisions and a concentration on liquidity and inflation hedging. The economic structure that results emphasizes finance at the expense of production. The inflation problem is, of course, a direct outgrowth of persistent budget deficits. The budget deficit today is substantially financed (except for the

73

74

75

76

77

7a

79

ao

ai

a2

a3

a4

a5

Fig. 6.4 Real GDP in manufacturing (logarithm, 1970 Australs)

a6

a?

126

Rudiger Dornbusch and Juan Carlos de Pablo

new money portion coming from foreign creditors) by printing money. Occasionally, deficits are financed by domestic debt, now denominated in dollars. But the very high interest rates-well above 20 percent on dollar bonds in late 1987-discourage this channel of financing. Thus the inflation tax, rather than ordinary taxes, finances the government’s outlays. This raises two questions. First, why is it so difficult to shift from an inflation tax to ordinary taxation? After all, the inflation tax is highly regressive, so changing the tax structure should be politically attractive. Second, why is it so difficult to contain government outlays? Table 5.3 above has already shown the budget deficit calculated using IMF/BCRA measures. Other measures, such as those prepared by FIEL, are more comprehensive and show even larger deficits. We have already noted the existence of a wide range of deficit measures and the fact that public discussion has not settled on any particular measurement. The difficulties are chiefly coverage (central government, all levels of government, etc.), the inclusion of particular accounts (treatment of public sector enterprise investment, Central Bank losses, Central Bank discounts), and the distinction between budget and cash bases. Even so, various sources do portray roughly equivalent fluctuations in the budget, and hence we can focus on any one measure to make the point that there are large and persistent deficits. These deficits require financing, and that financing, as other sources have now dried up, increasingly takes the form of printing money. We concentrate here on two central points in the budget situation: the appalling record on enforcement of tax collection, and the outlays in the budget that result from financial instability or policy mistakes. Tax fraud is so common in Argentina that it is almost considered a citizen’s right. Income tax collection in 1985 accounted for less than 1 percent of GDP! Some idea of the extraordinary degree of evasion is given by a report of the Ministry of the Economy: in a country of more than 30 million inhabitants, only one and a half million were registered taxpayers. Of those registered, less than one-third filed a return and less than 1 percent of those registered declared any tax liabilities due. Worse, 84 percent of the total revenue from income and value-added taxes was paid by 6 percent of the registered taxpayers. Thus one must ask why those 29,081 Argentine households and firms do pay when nobody else does.’ In 1985-86 the government made considerable progress in collecting taxes. Unfortunately, most of the additional tax collection was of an emergency variety. There were new taxes on imports and exports, increases in real public utility rates, and increased taxation of those who had already paid. The role of public utility rates as a source of revenue is particularly problematic. The political and inflationary effect of changes in utility rates makes them targets for politicization. As a consequence, rates tend to fall behind in real terms, giving rise to deficits and hence inflationary deficit

127

ArgentinaiChapter 6

finance. When the time comes to restore the budget and the real value of utility rates, this results in a major inflationary shock. Essential in public finance is to index rates once and for all, and to put rate decisions at arm’s length from the budget and political process. Establishment and enforcement of a broad-based, efficient tax system is the most urgent task. With a broad-based tax system applying low rates to a large body of taxpayers, the extreme inequity and inefficiency of the present system can be reduced. At the same time, there would be significant room for collecting more tax revenue without adverse supply-side effects. With regard to outlays, we want to emphasize two categories directly related to past policy mistakes. One is the large external interest bill which reflects in part the fact that the government increasingly “nationalized” the external debt from 1981 on. This occurred partly as a means of stopping capital outflows when the government took over from the private sector external loans that had not yet matured but which private borrowers wanted to pay off in anticipation of depreciation. In part the debt was nationalized in the course of financial market problems. By now more than 85 percent of the external debt is a burden on the budget, but there is no significant offset in terms of revenues. As a result of this debt nationalization, there are extra outlays amounting to 5-6 percent of GDP which the government must finance. Moreover, every time the real exchange rate is depreciated, the ratio of debt service to GDP increases and so does the budget deficit. A 15 percent real depreciation, for example, raises the budget deficit by 1 percent of GDP.* Thus, once more, external debt and adjustments to ensure continued debt service play an important part in keeping macroeconomic instability alive. The other element in the budget that is directly due to past policy mistakes appears in the quasi-fiscal deficit. Specifically, in 1981-84 in the aftermath of the Martinez de Hoz overvaluation, the government had to pay out extraordinary sums as the result of exchange rate guarantees. On average, 10 percent of GDP was paid out in those years to cover bets lost by the public sector! Needless to say, financing of these losses is an important element in the buildup of inflation after 1982. The correction of budget deficits is essential because the current situation has two clear implications. First, high and rising inflation is always just around the comer. That possibility discourages any long-term investment and assures the public that holding assets abroad remains the best investment. The resulting speculation against the currency puts pressure on the exchange rate, forcing the monetary authorities to produce high real interest rates. The high real rates in turn discourage productive investment, aggravate bankruptcy problems, and worsen the budget by their adverse effects on activity. When the unreasonable fiscal position is routinely compensated for by tight money, normalcy is simply impossible. Exactly the opposite pattern

128

Rudiger Dornbusch and Juan Carlos de Pablo

is required: tight fiscal policy to raise saving, and easy money to finance domestic investment and reduce the risk of business bankruptcies.

6.5 External Constraints to Growth Most of Argentina’s economic problems are homemade, but not all. The external environment has been exceptionally unfavorable in recent years. These poor external conditions have reduced the flexibility for policymakers, making adjustment more inflationary. The U.S. policy mix of 1981-85, with record real interest rates, a massive dollar appreciation, and a sharp decline in commodity prices, was an important part of the Argentine debt crisis. The magnitude and persistence of these very poor policies were especially damaging to countries such as Argentina, where debts are in dollars and mostly geared to short-term market rates and where there is a significant concentration on commodities on the export side. Without this external shock, there certainly would have been a debt crisis, but the increase in indebtedness would have been far less, as we already saw in chapter 3. The outstanding feature of the poor external environment was, of course, the credit rationing and the short-leash approach that was introduced in the aftermath of the Mexican debt crisis of 1982. The concerted-lending approach not only proved cumbersome, but it also focused attention on financial aspects at the expense of long-run growth considerations. Debt negotiations kept policymakers in a perpetual state of emergency. IMF conditionality, oriented to quarterly targets, distracted all attention away from long-run policy reform toward balance sheet tricks and emergency taxation. The environment has been unfavorable in other respects, too. The terms of trade have deteriorated sharply since the early 1980s. As a result, there has been a need for more depreciation of the currency, which in turn has meant a worsening of the budget, more inflation, and a reduced standard of living. The terms of trade deterioration shown in table 6.6 and figure 6.5 reflects in part the decline in world agricultural prices, especially for those products of interest to Argentina. Wheat prices in dollars today, for example, are 30 percent below their 1981-85 levels, and the same is true for an index of food products in world trade. The dramatic decline in the real price of food Table 6.6

Terms of Trade (index 1970 = 100) 1970-74 1975-79 1980-84 1984

Source: World Bank.

119 94 112 110

1985 1986 1987

95 81 78

129

ArgentindChapter 6 150

140

130

120

110

1W

90 00

I

70 70

Fig. 6.5

71

72

73

74

75

76

77

70

79

00

01

02

03

04

05

06

07

The terms of trade: 1970-87 (Index 1970 = 100)

reflects the rapid growth in agricultural productivity and the increasing selfsufficiency of traditional importing countries. But it also reflects the agricultural support policies of industrial countries. The policies of the European Common Market, in particular, are costly for Argentina. As a case in point, when Spain joined the common market, the diversion of trade caused Argentina to lose a traditional buyer of Argentine wheat. The U.S. drought of 1988 gave a reprieve by sharply raising export prices for Argentine agricultural products. It remains to be seen whether policymakers can appropriate some of this windfall for budget correction rather than dissipate it in further macroeconomic instability and premature debt service. For manufactured exports, the news is no better. Although the world economy grew at a significant pace over the past five years, competition among newly industrialized countries has been intense. Morever, protection in the industrialized countries’ markets has been increasing. The growth of protection, in turn, has served to discourage investment in nontraditional exports.

6.6 Concluding Remarks In July 1986 Brazil and Argentina signed agreements to initiate a limited free trade regime. This is the only bright spot on the external side, but it is also an event with the potential to overshadow even the debt problem in its scope and promise. For the time being, the agreement covers only specific sectors. But there is no question that if it is expanded to unrestricted free trade, the implications

130

Rudiger Dornbusch and Juan Carlos de Pablo

for Argentina would be dramatically positive. The size of the market would increase fivefold.’ Firms could specialize in products for which they have the greatest comparative advantage, and thus reap major scale economies. The increase in competition in the home market would give consumers a major increase in real income. Free trade with Brazil (unlike unilateral trade liberalization) offers a plausible strategy for reaping massive gains from trade without the immediate risk of a serious destruction of industry. An opening to Brazil might be one of the very few means available to Argentina for reversing the steadily declining standard of living. We have argued in this chapter that poor economic performance in Argentina is the result of a number of interdependent factors: debt, inefficiency, poor public finance, and an adverse external environment. There is little scope for bootstrapping the country out of this predicament. Even a moratorium is unlikely to set off a process of sustained growth. The hyperinflation woke up the national conscience and gave the government a brief moment in which to achieve some consolidation. But the change in policies and opportunities was not sufficiently large and durable to reverse the drift back into poor performance. The Austral Plan of 1985 gave a powerful boost to confidence in economic management, as shown in figure 6.6. The gain in approval after stabilization in May 1985 exceeds 30 percent! But since then, public approval of economic management has declined below the prestabilization levels. Thus, political support for change has been dissipated. It is clear today that a turnaround in public finance, financial stability, and investment simply will not happen except as a byproduct of an extraordinary event. The world macroeconomy is unlikely to offer much good news. That

50

40

30

20

10

0 1/84

1o/84

5/85

5/85

10/85

3/86

7/86

12/86

5/87

8/81

10/88

3/88

Fig. 6.6 Public opinion poll rating approval of economic management Source; Gallup Poll.

131

ArgentindAppendix A

leaves the free trade option with Brazil as one of the few cards policymakers have left to play. A major fiscal reform, widening the tax base and achieving public sector efficiency, restructuring of the external debt service, and a new initiative in trade policies form a program that can reconstruct Argentina. Such a program now requires urgent attention. In concluding we draw on Carlos Diaz Alejandro’s (1988) words to balance the widespread skepticism about Argentina’s economic future: Paradoxically, the troubles of the 1970s in Argentina may have set the bases for political conditions allowing steadier and more efficient economic policies. . . . Perhaps swords and furnaces will be put away, and quieter hours may come. Under conditions of reasonable political stability, the unsinkable Argentine economy could recover from the catastrophes of the late 1970s and the early 1980s, including a high external debt for which so little growth can be shown. Abundant food stuffs and energy resources, plus an industry which, whatever its past costs, has shown itself capable of exporting, provide solid foundations for a growth which may or may not keep up with those of Australia and Brazil, but which assure a good life to all Argentines.

Appendixes Appendix A Price Dynamics Under a Tablita Regime In this appendix, we sketch a model of inflation dynamics under a Martinez de Hoz-style program. The essential linkages are purchasing power parity for traded goods and integration of the capital market. The former implies that prices of traded goods are determined by the exchange rate: (A. 1)

P, = e

and that interest rates are set by the world market with an adjustment for the rate of depreciation: (A.2)

i= i* + A

where i* is the international interest rate and A is the rate of depreciation. The real exchange rate is defined as R = P,/e, which is the ratio of home goods to traded goods prices. Let n denote the rate of inflation of home goods prices and A the rate of traded goods price inflation. From the

131

ArgentindAppendix A

leaves the free trade option with Brazil as one of the few cards policymakers have left to play. A major fiscal reform, widening the tax base and achieving public sector efficiency, restructuring of the external debt service, and a new initiative in trade policies form a program that can reconstruct Argentina. Such a program now requires urgent attention. In concluding we draw on Carlos Diaz Alejandro’s (1988) words to balance the widespread skepticism about Argentina’s economic future: Paradoxically, the troubles of the 1970s in Argentina may have set the bases for political conditions allowing steadier and more efficient economic policies. . . . Perhaps swords and furnaces will be put away, and quieter hours may come. Under conditions of reasonable political stability, the unsinkable Argentine economy could recover from the catastrophes of the late 1970s and the early 1980s, including a high external debt for which so little growth can be shown. Abundant food stuffs and energy resources, plus an industry which, whatever its past costs, has shown itself capable of exporting, provide solid foundations for a growth which may or may not keep up with those of Australia and Brazil, but which assure a good life to all Argentines.

Appendixes Appendix A Price Dynamics Under a Tablita Regime In this appendix, we sketch a model of inflation dynamics under a Martinez de Hoz-style program. The essential linkages are purchasing power parity for traded goods and integration of the capital market. The former implies that prices of traded goods are determined by the exchange rate: (A. 1)

P, = e

and that interest rates are set by the world market with an adjustment for the rate of depreciation: (A.2)

i= i* + A

where i* is the international interest rate and A is the rate of depreciation. The real exchange rate is defined as R = P,/e, which is the ratio of home goods to traded goods prices. Let n denote the rate of inflation of home goods prices and A the rate of traded goods price inflation. From the

132

Rudiger Dornbusch and Juan Carlos de Pablo

definition of the real exchange rate (remembering that traded goods price inflation equals the rate of depreciation), it follows that the real exchange rate appreciates at a rate equal to the difference between the inflation rates of home goods and traded goods: RIR

(A.3)

= T -

A

The model is completed by the dynamics of home goods price inflation. We assume that the rate of inflation of nontraded goods rises whenever the real exchange rate depreciates or when there is high demand relative to output. Excess demand depends on the real exchange rate, R , and on the real interest rate in terms of home goods. A high real exchange rate or a high real interest rate slows down demand.

(A.4)

ir = J ( A -

T ,R,

i*

+A

-T),

J , > 0; J 2 , J 3 < 0

+

We assume that J , J3 ; > 0 so that expectation effects dominate the real interest rate effect due to a more rapid rate of depreciation. The dynamics of the real exchange rate and the rate of inflation are shown in figure A . 1 for a given rate of depreciation, A,. When the rate of inflation exceeds the rate of depreciation, the real exchange rate appreciates to the right of R = 0, and depreciates to the left of the vertical line. Along ir = 0, the inflation rate is constant. To the right, the real exchange rate andlor the inflation rate are too high and hence inflation is falling. To the left and above ir = 0, the real exchange rate is too low and depreciation exceeds inflation by too much; hence, inflation is rising. From an initial position to the northeast of A the dynamics are indicated by the path to the steady state at point A. Note that in the steady state, the Chicago model fully holds: inflation and depreciation proceed at the same rate.

R = PN/e

Fig. A.l

The dynamics of inflation adjustment with inertia

133

ArgentindAppendix B

Appendix B The Budget and Injution In this appendix we briefly show the derivation of the key equation representing the interaction of inflation and financial structure. We start with the central equation which shows that the deficit is financed by money creation:

(B.1)

MIP = gY

or, using the definition of money growth, S

(B.2)

=

MIM:

6(MIP) = gY

Next we impose the assumption of a linear velocity equation and of monetary equilibrium. This implies equality of actual and planned velocity: PYIM =

(B.3)

ci

+ p9-r

The parameter P thus represents the responsiveness of velocity to the cost of holding money. We do not make any allowance for lags in the adjustment of velocity to the cost of holding money. Substituting from (B.3) into (B.2) we have:

Next we use the steady-state relation between money growth and inflation, having implicitly assumed a unit income elasticity of money demand:

(B.5)

.rr=S-y

This equation states that the inflation rate is equal to the growth rate of money less the growth rate of real output. Combining (B.5) and (B.4) yields the key equation used in the text:

03.6)

9-r

= (ag -

YMl - Pg); Pg < 1

To study the impact of external shocks on the budget deficit, we now decompose the actual deficit ratio g into the interest and noninterest budget. The noninterest budget, which is a function of the rate of inflation via the Olivera-Tanzi effect, is denoted by u. 03.7)

+ ib + i*b*,

g = u(n)

where ib and i*b* denote domestic and external debt service, b = BIPY and b* = eB*lPY. The last term depends on the ratio elP and hence will be increased by a real depreciation.

134

Rudiger Dornbusch and Juan Carlos de Pablo

A realistic setting is one where there is no domestic debt and where a fraction, of the external debt is automatically capitalized. The deficit ratio requiring internal financing by money creation is then:

+,

g

(B.8)

= a(7r)

+

(1 - 0)b*

Appendix C Statistical Data Table C . l

K e y Macroeconomic Variables

Year

GDP Growth

Inflation (CPI)

1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 I980 1981 1982 1983 1984 1985 1986 1987

5.4 3.7 I.9 3.5 5.7 -0.4 -0.5 6.4 - 3.4 6.7 0.7 -6.2 -4.6 2.8 2.6 -4.5 5.4 I.6

21.7 39.1 64.1 43.8 40.2 335.1 347.5 160.4 169.8 139.7 87.6 131.3 209.7 433.7 688.0 385.0 81.9 174.9

Table C.2

1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980

Real Exchange Rate

Terms of Trade

0.3 -0.6 0.2

56.7 49.7 52.9 58.6 71.2 36.9 46.4 50.7 64.7 83.4 100.0 69.5 48.9 58.8 58.4 48.9 45.2 42.6

106.6 116.4 125.7 134.6 117.8 111.5 95.8 92.7 83.5 88.4 100.0 107.0 89.0 86.0 93.0 81.0 69.0 65.0

I .5 1.3 -2.2 2.9 3.9 4.9 -0.1 -6.1 -2.7 3.3 3.9 4.2 6.2 I.9 0. I

The External Sector Current Account

-

Noninterest Current AccounVGDP

$ - 159 - 389

- 223

721 127 - 1,284 665 1.290 1,833 - 537 -4,767

W of GDP -0.8 - 1.8 - 1.0 2.7 0.4 -3.5 1.7 3.0 4.0 - 1.0 -1.6

Noninterest Current Account 64 - I33

50 1,038 425 - 824 1,130 1,660 2,238 -44 -3,824

Terms of Trade

DebV GDP(%)

Real Exchange Rate

Exchange Rate Gap ("7)

107 I I6 126 135 118 I12 96 93 84 88 100

16.7 18.2 21.8 20.0 20.4 18.6 18.6 19.2 23.9 30.2 37.3

56.7 49.7 52.9 58.6 71.2 36.9 46.4 50.7 64.7 83.4 100.0

2 27 43 32 85 I83 52

n

0 0 0

134

Rudiger Dornbusch and Juan Carlos de Pablo

A realistic setting is one where there is no domestic debt and where a fraction, of the external debt is automatically capitalized. The deficit ratio requiring internal financing by money creation is then:

+,

g

(B.8)

= a(7r)

+

(1 - 0)b*

Appendix C Statistical Data Table C . l

K e y Macroeconomic Variables

Year

GDP Growth

Inflation (CPI)

1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 I980 1981 1982 1983 1984 1985 1986 1987

5.4 3.7 I.9 3.5 5.7 -0.4 -0.5 6.4 - 3.4 6.7 0.7 -6.2 -4.6 2.8 2.6 -4.5 5.4 I.6

21.7 39.1 64.1 43.8 40.2 335.1 347.5 160.4 169.8 139.7 87.6 131.3 209.7 433.7 688.0 385.0 81.9 174.9

Table C.2

1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980

Real Exchange Rate

Terms of Trade

0.3 -0.6 0.2

56.7 49.7 52.9 58.6 71.2 36.9 46.4 50.7 64.7 83.4 100.0 69.5 48.9 58.8 58.4 48.9 45.2 42.6

106.6 116.4 125.7 134.6 117.8 111.5 95.8 92.7 83.5 88.4 100.0 107.0 89.0 86.0 93.0 81.0 69.0 65.0

I .5 1.3 -2.2 2.9 3.9 4.9 -0.1 -6.1 -2.7 3.3 3.9 4.2 6.2 I.9 0. I

The External Sector Current Account

-

Noninterest Current AccounVGDP

$ - 159 - 389

- 223

721 127 - 1,284 665 1.290 1,833 - 537 -4,767

W of GDP -0.8 - 1.8 - 1.0 2.7 0.4 -3.5 1.7 3.0 4.0 - 1.0 -1.6

Noninterest Current Account 64 - I33

50 1,038 425 - 824 1,130 1,660 2,238 -44 -3,824

Terms of Trade

DebV GDP(%)

Real Exchange Rate

Exchange Rate Gap ("7)

107 I I6 126 135 118 I12 96 93 84 88 100

16.7 18.2 21.8 20.0 20.4 18.6 18.6 19.2 23.9 30.2 37.3

56.7 49.7 52.9 58.6 71.2 36.9 46.4 50.7 64.7 83.4 100.0

2 27 43 32 85 I83 52

n

0 0 0

135

ArgentindAppendix C

Table C.2 (continued) Current Account

1981

1982 1983 1984 1985 1986

$

% of GDP

Noninterest Current Account

-4,714 -2,357 2,461 -2,391 - 953 -2,645

- 7.4

- 1,749

-3.8 -3.8 -3.5 - 1.5 -4.0

2,046 2,522 2,888 3,926 1,289

Terms of Trade

Debt/ GDP(%)

Real Exchange Rate

Exchange Rate Gap (%)

I07 89 86 93 81 69

48. I 60.3 59.5 60.5 64.5 65.3

69.5 48.9 58.8 58.4 48.9 45.2

0 55 39 38 19 12

Key Relative Prices (index 1980 = 100)

Table C.3

Agricultural/ Nonagricultural (WPI) 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986

I I9 I32 I52 I42 124 98 I05 112 I10 Ill 100 90

-

107 I I6 126 I35 118 112 96 93 84 88

57 50 53 59 71 37 46 51 65 83

100

100

100

100

I16

106 140 153 143 110 I10

107 89 86 93 81 69

70 49 59 58

I32 147 160 I48

49 45

Inflation (annual variations, average to average, in percentages) Consumer

Wholesale

1914 1915 1916 1917 1918 1919

-0.5 7.6 7.5 17.2 25.9 -5.8

0.8 6.7 14.0 23.9 9.4 3.2

I920 1921 1922 1923 1924

17.1

4.8 -20.5 -9.5 4.1 7.3

(continued)

Real Exchange Rate

263 216 268 264 218 124 83 80 71

107 86 I10

Terms of Trade

106

Ill

101

Real Wages

200 98 108 152 159 132 130 I34 I05

104

Table C.4 Year

Real Public Sector Prices

-11.1

- 15.9 -2.0 2. I

Year 1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935

Consumer -2.7 -3.1 - 1.1

1.1 - 1.1

1.1 - 13.9

- 10.4 13.0

- 11.5 6.0

Wholesale 1.6 - 10.3

1.7 0.6 -2.9 -

-4.2 -3.1 1.3 -4.4 13.9 - 1.2

136

Rudiger Dornbusch and Juan Carlos de Pablo

Table C.4 (continued) ~~~

~

Year

Consumer

Wholesale

Year

1936 1937 1938 1939

8.7 2.4 -0.4 1.6

1.8 14.5 -6.1 1.6

I940 1941 1942 I943 I944 1945 I946 1947 1948 I949

I .9 2.7 5.9 I .0 -0.3 19.8 17.7 13.5 13.0 31.1

13.8 2.8 25.8 9.4 8.9 8.5 16.1 3.6 15.6 23.0

1962 1963 1964 I965 1966 I967 I968 1969

26.1 26.0 22.1 28.6 31.9 29.2 16.2 7.6

30.4 28.8 26.3 23.8 19.9 25.7 9.4 6. I

1950 1951 1952 1953 1954 1955 1956 1957 1958 1959

25.6 36.9 38.7 4.0 3.7

13.6 34.7 58.5 60.3 24.3 182.8 444.0 176.0 175.5 159.5

14. I 39.6 76.6 50.4 20.0 192.5 499.0 149.4 146.0 149.3

13.4 24.7 31.6 111.6

19.8 49.3 31.1 11.6 3.2 8.9 26.0 24.3 30.8 133.9

1970 1971 1972 1973 1974 1975 1976 1977 1978 1979

I960 1961

31.1 13.7

15.5 8.2

1980 1981 1982 1983 1984 1985 19x6

100.8 104.5 164.8 343.8 626.7 672.2 90. I

75.4 109.6 256.2 360.9 575. I 662.9 63.9

12.3

Consumer

Wholesale

Source: Carta Economico

Table C.5

Foreign Debt (in millions of $ U S ,end of period)

Year

Public

Private

Total

Net DebVExports (Years)

DebliGDP (%I

1963 I964 1965 1966 1967 1968 1969

2,106 1,789 1,687 1,769 1,818 1,754 1,996

1,285 1,127 963 894 826 1,051 1,234

3,391 2,916 2,650 2,663 2,644 2,805 3,230

2.3 2.0 1.6 1.6 1.3 1.6 1.7

21.6 16.9 14.0 14.0 13.6 13.8 14.6

I970 1971 1972 1973

2,143 2,527 3,089 3,426

1,733 1,998 2,699 2,807

3,876 4,525 5,788 6,233

1.9 2.5 2.8 1.6

16.7 18.2 21.8 20.0

1974 1975 1976 1977 1978 1979

4,558 4,021 5,189 6,044 8,357 9,960

3,410 3,854 3,090 3,635 4,139 9,074

7,968 7,875 8.279 9,679 I 2,496 19,034

1.7 2.6 I .7

20.4 18.6 18.6 19.2 23.9 30.2

1.1 1.1

1.2

137

ArgentinaiAppendix C

Table C.5 (continued)

Year

Public

Private

Total

12,703 15,647 15,018 12,585 10,193 8,444 6,700

27,162 35,671 43,634 44,781 47,821 48,312 5 1.422

Net DebVExports (Years)

DebVGDP (%)

~~

1980 1981 1982 1983 1984 1985 1986

14,459 20,024 28,616 32,196 37,628 39,868 44,772

2.5 3.5 5.4 5.4 5.6 5.2

37.3 48.1 60.3 59.5 60.5 64.5 65.1

Real GDP and the Composition of Demand (in 1970 Australs)

Table C.6 Year

GDP

1950 1951 1952 1953 I954 1955 1956 1957 1958 1959

4,286.8 4,453.4 4,229.3 4,453.4 4,631.2 4,964.8 5,102.7 5,367.1 5,694.6 5,326.8

412.8 326.1 239.3 366.4 392.4 353.4 405.4 428.1 441.1 461.0

3,860.9 3,934.5 3,783.1 3,714.3 4.028.1 4,424.8 4,486.7 4,692.2 4,953.9 4,587.7

632.8 779.1 693.1 685.0 657.8 763. I 812. I 890.4 970.5 767.9

502.8 565.1 416.5 341.3 447.7 538.3 481.4 537.0 560.2 496.1

1960 1961 1962 1963 1964 1965 1966 1967 1968 1969

5,746.3 6,154.3 6,056.6 5,913.0 6,522.0 7,119.7 7,165.6 7,355.3 7,671.3 8,326.4

463.8 428.6 579.8 591.4 553.3 607.6 661.4 659.4 650.5 754.9

4,734.4 5,205.6 4,989.5 4,888.3 5,389.0 5,832.9 5,880.6 6,029.3 6,262.2 6,640.7

1,160.7 1,353.0 1,232.2 1,158.6 1,210.0 1,249.4 1,311.6 1,474.5 1,769.3

611.6 127.2 698.5 543.7 631.2 624.5 601.8 603.4 641.3 791.8

1970 1971 1972 1973 1974 1975 1976 1977 1978 1979

8,714.6 9,101.5 9,277.5 9,597.5 10,144.3 10,106.1 10,066.0 10,702.6 10,343.4 11,022.4

809.6 729.2 741 .O 829.3 831.3 737.4 970.5 1,227.7 1,322.5 1,294.8

6,893.8 7,179.3 7,268. I 7,532.8 8,155.8 8,234.1 7,596.7 7,748.3 7,604. I 8,539.6

1,861.2 2,004.0 2,015.2 1,874.6 1,961.5 1,963.4 2,159.2 2,577.0 2,267.9 2,368.4

789.4 862.1 797.1 738.8 849.7 907.1 674.5 868.2 783.7 1,185.2

1980 1981 1982 1983 1984 1985 1986

11,142.8 10,423.3 9,882.3 10,213.4 10.459.0 10.004.6 10,574.9

1,242.1 1,340.2 1,359.9 1,473.7 1,463.5 1,642.1 1,473.0

9,042.3 8,754.0 7,810.2 8,124.6 8,640.2 8 ,075.8 8,778.4

2,452.5 2,029.9 1,531.4 1,503.3 1,330.1 1,256.0 1,418.3

1,692.9 1,628. I 942.9 898.5 955.5 845.6 999.5

Exports

Consumption

Investment

1,044.1

Imports

138

ArgentindAppendix C Total and Per Capita Real Gross Domestic Product

Table C.7

Indexes (1986 = 100)

Annual Variations (%)

Year

Total G D P

Per Capita GDP"

Total GDP

Per Capita GDP

I900 1901 1902 1903 1904 1905

1909

686.7 744.8 729.8 834.2 923.1 1,045.7 1,098.3 1,121.4 1,231.3 1,292. I

151.50 159.70 152.19 169.79 183.58 201.66 203.58 198.13 206.27 205.95

6.5 7.0 6.9 7.9 8.7 9.9 10.4 10.6 11.6 12.2

44.3 46.7 44.5 49.7 53.7 59.0 59.6 58.0 60.4 60.3

8.5 -2.0 14.3 10.7 13.3 5.0 2. I 9.8 4.9

2.9 2.8 2.5 2.4 3.1 4.0 4.9 5.5 5. I

1910 1911 1912 1913 1914 1915 1916 1917 1918 1919

1,386.1 I ,411 .O 1,526.3 1,542.1 1,382.4 1.389.7 1,349.7 1,240.4 1,467.7 1,521.7

210.01 203.92 210.45 201.97 175.72 172.56 164.45 148.46 172.70 175.88

13.1 13.3 14.4 14.6 13.1 13.1 12.8 11.7 13.9 14.4

61.5 59.7 61.6 59. I 51.4 50.5 48.1 43.4 50.5 51.5

7.3 1.8 8.2 1.0 10.4 0.5 -2.9 -8.1 18.3 3.7

5.2 4.8 4.8 5.3 3.0 2.4 1.9 1.8 1.7 1.8

3.2 -4.0 - 13.0 - 1.8 -4.7 -9.1 16.3 1.8

1920 1921 1922 1923 I924 1925 1926 1927 1928 I929

1,632.5 1.674.2 1,808.1 2,007.3 2,163.9 2,154.9 2,258.8 2,419.0 2,568.8 2,687.2

184.65 184.56 193.45 207.26 215.72 208.51 212.53 221.11 228.21 232.34

15.4 15.8 17.1 19.0 20.5 20.4 21.4 22.9 24.3 25.4

54.0 54.0 56.6 60.7 63.1 61.0 62.2 64.1 66.8 68.0

7.3 2.6 8.0 7.8 -0.4 4.8 7.1 6.2 4.6

2.2 2.6 3.0 3.6 3.6 3.0 2.8 2.9 2.9 2.7

5.0 0.0 4.8 7.1 4.1 -3.3 1.9 4.0 3.2 1.8

1930 1931 1932 1933 1934 1935 1936 1937 1938 1939

2,576.0 2,397.2 2,317.8 2.426.7 2,618.3 2,732. I 2,766.9 2,987.4 3,028.0 3,138.2

217.04 191.41 187.31 192.68 204.47 209.93 209. I4 221.95 221.12 225.50

24.4 22.7 21.9 22.9 24.8 25.8 26.2 28.2 28.6 29.7

63.5 57.8 54.8 56.4 59.8 61.4 61.2 65.0 64.7 66.0

-4.1 -6.9 -3.3 4.7 7.9 4.3 1.3 8.0 I .4 3.6

2.6 2.3 I.9 1.8 1.7 1.6 1.7 1.7 1.7 1.6

-6.6 -9.0 -5.1 2.9 6.1 2.7 -0.4 6. I -0.4 2.0

I940 1941 1942 1943 1944

3,074.4 3,225.2 3,376.0 3,399.2 3,627.3

217.47 224.45 231.16 229.01 240.29

29. I 30.5 31.9 32. I 34.3

63.6 65.7 67.6 67.0 70.3

-2.0 4.9 4.1 0.7 6.7

1.6 1.6 1.6 1.6 1.7

-3.6 3.2 3.0 -0.9 4.9

1906

1901 I908

Total GDP

~

11.0

Population

Per Capita GDP

5.4 -4.7 11.6 8. I 9.9 1 .o

-2.1 4. I -0.2 2.0

- 2.9

139

Rudiger Dornbusch and Juan Carlos de Pablo

Table C.7 (continued) Annual Variations (B)

lndexes (1986 = 100)

Year

Total GDPB

Per Capita GDP"

Total GDP

Per Capita GDP

Total GDP

Population

1945 1946 1947 1948 1949

3,550.0 3,845.9 4,373.8 4,426.0 4,222.9

231.20 246.24 274.98 272.03 252.88

33.6 36.4 41.4 41.9 39.9

67.7 72. I 80.5 79.6 74.0

-2.1 8.3 13.7 I .2 -4.6

1.7 1.7 1.8 2.3 2.6

1950 1951 1952 1953 1954 1955 1956 1957 1958 1959

4,286.7 4,453.4 4,229.3 4,453.4 4,637.3 4,964.8 5,102.7 5,367.0 5,694.6 5,326.8

249.96 254.84 237.44 245.32 250.75 263.64 266.39 275.43 287.55 264.81

40.5 42.1 40.0 42. I 43.9 46.9 48.3 50.8 53.9 50.4

73.1 74.6 69.5 71.8 73.4 77.2 78.0 80.6 84.2 77.5

1.5 3.9 -5.0 5.3 4. I 7.1 2.8 5.2 6.1 - 6.5

2.7 1 .9 I .9 I .9 1 .9 1.8 I .7 1.7 1.6 1.6

I960 1961 1962 I963 I964 1965 1966 1967 1968 1969

5,746.3 6,154.3 6,056.6 5,912.9 6,522.1 7,119.7 7,165.6 7,355.3 7,671.3 8,326.4

281.36 296.95 288.10 277.36 301.78 325.01 322.80 327.00 336.62 360.65

54.3 58.2 57.3 55.9 61.7 67.3 67.8 69.6 72.5 78.7

82.3

86.9 84.3 81.2 88.3 95.1 94.5 95.7 98.5 105.5

7.9 7.1 -1.6 -2.4 10.3 9.2 0.6 2.6 4.3 8.5

1.5 1.4 1.4 I .4 1.4 1.3 1.3 1.3 1.3

6.3 5.5 -3.0 -3.7 8.8 7.7 -0.7 1.3 2.9 7.1

1970 1971 1972 1973 1974 1975 1976 1977 1978 1979

8,774.6 9,101.5 9,277.5 9,597.5 10,144.3 10,106 I 10,066.0 10,702.6 10,343.4 11,022.4

375.14 382.37 383.00 389.34 404.38 395.87 387.46 404.82 384.45 402.58

83.0 86. I 87.7 90.8 95.9 95 6 95.2 101.2 97.8 104.2

109.8 111.9 112.1 113.9 118.3 115 9 113.4 118.5 112.5 117.8

5.4 3.7 1.9 3.4 5.7 -0 4 -0.4 6.3 - 3.4 6.6

1.3 1.8 1.8 1.8 1.8 18 1.8 1.8 1.8 1.8

4.0 1.9 0.2 I .7 3.9 -2 1 -2.1 4.5 -5.0 4.7

1980 1981 1982 1983 1984 1985 1986

11,142.8 10,423.3 9,882.3 10,213.4 10.459.0 10,004.6 10,574.9

399.91 367.60 342.48 347.81 350.00 328.98 341.70

105.4 98.6 93.5 96.6 98.9 94.6 100.0

117.0 107.6 100.2 101.8 102.4 96.3 100.0

1.1

-6.5 -5.2 3.4 2.4 - 4.3 5.7

1.8 1.8 1.8 1.8 1.8 1.8 1.8

-0.7 -8.1 -6.8 I .6 0.6 -6.0 3.9

I .5

Per Capita GDP -3.8

6.5 11.7 -1.1 - 7.0 - 1.2

2.0 - 6.8

3.3 2.2 5.1 1.0 3.4 4.4 -7.9

Sources: CEPAL (1958); BCRA, Producto e ingreso de la Argentina. (1975) and national accounts department publications. INDEC, population census.

"In Australs, at 1970 prices.

Table C.8

Balance of Payments (in millions of $ U S ) Current Account Merchandise

Year

Exports

1951

1,169.4 687.8 1,125.1 1,026.6 928.6 943.8 974.8 993.9 1.0OO.6

1,480.2 1,179.3 795.1 979.O I , 172.6 1,127.6 1,310.4 1.232.6 983.6

-310.8 -491.5 330.0 47.6 -244.0 - 183.8 - 335.6 -238.7 17.0

1,079.2 964.I 1,216.0 1,365.5 1,410.5 1,488.0

1,249.3 1,460.3 1,356.5 980.7 1,077.4 1,195.0

- 170.I

1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 I964

1965

Imports

Services

Difference

-496.2 - 140.5 384.8 333.I 293.0

Financial -

19.4 10.0

0.8 -8.9 - 15.8 9.4 11.8 -28.9 - 18.3

-41.1 -93.9 - 136.2 - 161.9 -262.8 - 111.I

Real & Unilateral Transfers

-6.8 16.9 -6.7 18.8 18.3 43.3 21.1

8.3 13.3 12.9 5.4 3.9 10.9 -36.7 0.4

Autonomous Capital Movements Total - 337.0

-464.6 342.1 57.5 -241.5 -131.1 -302.7 -259.3 12.0 198.3 -584.7 -272.8 233.8 33.6 182.3

-

Private

78.3 108.I 68.7 43.6 42.4 110.6 53.8 60.4 72.0 355.I 158.9 297.9

160.1 10.4 15.7

Public

~

102 8 -8.6 -2.5 -9.7 - 13.8 -77.1 118.5 49.3 -11.8 138.9 77.0 -74.8 -20.7 -31.5 17.1

International Payments, Errors & Omissions - 155.9 - 365.I

390.3 91.4 -212.9 -97.6 -367.4 - 149.6 72.2 295.7 - 348.8

-49.7 373.2 12.5 215.I

1966 1967 1968 1969

1,593.2 1,464.5 1,367.9 1,612.1

1,124.3 1,095.5 I , 169.2 1,576.1

I970 1971 1972 1973 1974 1975 1976 1977 1978 1979

1,773.2 1,740.4 1,941. I 3,266.0 3,930.7 2,961.3 3,916.1 5,651.8 6,399.5 7,809.9

1,694. I 1,868. I 1,904.7 2,235.3 3,634.9 3,946.5 3,033.0 4,161.5 3,833.7 6,700.0

1980 1981 1982 1983 1984 1985 1986

8,021.4 9,143.0 7,623.7 7,836.1 8,107.0 8,396.0 6,849.0

10,450.6 9,430.0 5,336.9 4,505.0 4,584.0 3,814.0 4,700.0

Source: BCRA

468.9 369.0 198.7 36.0

-151.1 - 119.6 -205.1 -219.1

-65.3 -75.2 -42.2 -43.2

252.5 174.2 -48.6 -226.3

-215.7 244.7 149.6 11.6

5.3 -0.9 64.7 106.0

79.1

-222.5 -255.9 -333.6 - 394.4 -333.3 -429.6 -492.5 -578.5 -680.8 -920.0

- 15.5

- 158.9

-5.1 74.3 78.6 164.4 130.2 259.0 378.1 -51.4 - 726.3

- 388.7

-222.9 714.9 126.9 - 1,284.6 649.6 1,289.9 1,833.6 -536.4

330.9 -317.9 12.0 18.8 - 159.2 95.0 -219.5 1,309.9 808.9 4,158.3

82.8 128.4 -86.5 -98.0 105.5 96.1 - 113.2 -23.4 525.0 528.6

-297.4 635.7 73.5 -1,093.5 316.9 2,576.4 3.167.5 4,150.5

- 1,531.4 -3,699.7 -4,718.5 -5,408.0 -5,712.0 -5,303.8 -4,415.0

-717.2 -727.3 74.0 - 384. I - 201.9 -231.0 - 555.0

-4,767.8 -4,714.0 -2,357.7 - 2,461 .O -2,390.9 -952.8 -2,821.0

1,617.9

933.9 2,537.4 -381.0 29.7 2,152.3 964.7

I -3,224.3 -4,681.2 -3,880.0 - 1,810.7 54.6

- 127.7

36.4 1,030.7 295.8 -985.2 883.1 1,490.3 2,565.8 1,109.9 - 2,519.2

-287.0 2,286.8 3,331.1 3,523.0 4.582.0 2,149.0

- 1,047.7 -

1,942.5

- 1,448.7

-1,572.1 42.7

42. I 418.0 165.7 - 107.9 254.8 - 578.2

- 2,216.

142 Table C.9

ArgentindAppendix C Real Wages and the Purchasing Power of Wages (index 1980 = 100) Purchasing Power

Real Wages

Year

Collective Bargaining

Minimum Wage

Collective Bargaining

1960 1961 I962 1963 1964 1965 1966 1967 I968 I969

242.0 264.3 262.6 262.1 280.5 297.6 300.3 302.3 272.0 277.7

224.8 226.3 185.8 158.9 186.9

247.4 269.2 266.4 266.1 286.2 301.0 305.0 306.7 278.1 284.2

227.4 230.2 188.6 162.5 191.3

I970 1971 1972 1973 1974 I975 1976 1977 1978 1979

288.5 294.6 273.7 296.8 313.5 310.1 188.0 136.2 87.5 84.3

181.7 193.8 172.8 200.3 246.1 207.3 93.6 90.9 72.5 84.5

293.0 297.4 275.0 301.1 315.8 293.8 177.8 133.8 86.0 83.4

184.6 195.7 173.7 203.6 247.9 197.2 89.9 44.5 71.2 83.7

1980 1981 1982 1983 1984 1985 1986

100.0 114.2 119.3 184.7 195.8 146.8 138.3

100

100.0 112.6 115.9 175.5 180.8 140. I 138.0

100

95.3 96.7 146.3 172.7 115.7 107.4

Sources: INDEC, FIEL, and Carta Econdmico

Minimum Wage

94 94. I 139

159.6 110.7 107.2

143

Rudiger Dornbusch and Juan Carlos de Pablo

Table C.10

Public Utility Rates, Including Taxes (deflated by consumer prices, 1960= 100)

Transportation & Communications

Industry & Services

72.4 68.6 71.1 69.6 77.0 61.9 67.3 74.8 76.5 64.3

114.5 115.4 95.6 106.9 124.1 95.1

32.4 45.5 30.3 46.0 46.3 38.9

60.8 70.4 61.3 69.7 62.8 64.2 60.9

92.7 103.1 87.0 83.9 82.3 88.5 81.5

43.3 45.0 33.3 26.9 20.8 18.6 24.0

Year

Overall

Gasoline

Electricity

1958 1959

92.9 114.0

99.3 128.7

89.0 108.0

1960 1961 1962 1963 1964 1965 I966 1967 I968 1969

100.0 89.9 87.6 95.4 79.7 82.3 83.4 86.7 89.8 85.9

100.0 85.6 84.3 93.2 77.4 77.2 79.6 76.7 84.2 80.6

100.0

92.6 86.8 92.7 78.0 85.9 93.0 loo.I 92. I 83.0

1970 1971 1972 1973 1974 1975 1976 1977 1978 1979

81.0 76.3 74.4 82.0 115.6 120.8 100.3 99.1 101.8 79.7

75.1 71.5 68. I 79.9 128.5 140.5 113. I 103.3 99.7 77.7

1980 1981 1982 1983 I984 1985 1986

76. I 88.3 84.8 100.2 112.2 121.9 108.7

73.9 88. I 92.3 118.9 142.9 157.0 138.6

Source: S E E P

144

ArgentindAppendix C Interest Rates (monthly equivalent, in percentages)

Table C . l l

Passive

Active

Nominal

Real Consurner Prices

Regulated

Nominal

Real Whole-

Year

Regulated

1959

0.6

6.1

-5.0

1960 1961 1962 1963 1964 1965 1966 1967 1968 1969

0.6 0.6 0.9 1 .0 1 .0 1.1 1.0 1 .0 1.2 1.2

1.5 1.3 2.3 1.8 1.4 2.8 2.2 2.1 0.8 0.6

-0.8 -0.6 - 1.3 -0.8 -0.4 - 1.6 -1.1 -1.0 0.5 0.7

1970 1971 1972 1973 1974 1975 1976 1977 1978 1979

1.2 1.3 1.8 1.7 1.5 2.1

1.7 2.8 4.2 3.1 2.9 13.4 13.8 8.3 8.6 7.6

-0.4 - 1.4 -2.3 - 1.3 - 1.2 -9.4

1.2 1.4 1.9

2.0

1.8

1.9

1.9 2.8 4.5

2.1

1980 1981 1982 1983 1984 1985 1986

Free

7.2 6.7 5.0 8.1

11.6 14.0 12.2 4.1

7.1 11.2 19.4 14.5 5.7

5.4 1.3 10.0 15.0 18.8 14.7 5.1

-2.9 -4.0 - 1.9 -1.0

Sources: BCRA. INDEC. and El Cronista Comercial.

Free

Regulated

sale Prices

Regulated

0.9

6.0

-4.5

0.9 0.9 1.0

0. I 1.2 3.0 1.8 1.4 2. I 1.7 1.6 0.3 0.6

0.8 -0.3 - I.8 -0.5 .0 -0.8 -0.4 -0.2 1 .0 0.7

Free

1.3 I.3 1.3

1.3 1.3 1.3

1.2

1.7 1.3 1.6

-1.3 -0.8

8.7 7.4

2.0 3.4 4.9 2.3 2.6 13.8 14.9 7.9 7.7 7.2

-0.4 0.8 -5.6 -1.6 0.5 0.1 0.6

5.9 9.8 10.0 18.7 19.5 16.3 6.7

3.9 9.0 12.7 14.7 18.0 14.5 3.9

12.6 15.4 14.0 5.4

1.5 1

.n

-0.7 - 1.8

-2.7 -0 4 -0.7 -8.9 -7.8

Free

0. I 0.9 1 .0 -0.5 - 1.4 -2.6 -0 3 -0.4

I.0 0.3

1.7 -2.1 0.2 I .5 -

I.9 0.8 -2.1 3.6 1.3

2.2 2.8

145

ArgentindAppendix C

Table C.12

Exchange Rate (Australs/$U.S.) and Prices (index 1980= 100) Exchange Rate

Year

Official

1959

.000000019

.0000000x0

,004284

,007294

1960 1961 1962 I963 1964 1965 1966 1967 1968 I969

.000000083 .000000083 ,000000114 .000000139 ,000000 I40 .000000 169 .000000208 . m 3 31 .000000350 ,000000350

,000000083 .000000083 ,000000114 .00oooO139 ,000000243 ,000000243 ,000000338 .000000350 .000000350

,005615 ,006386 ,008054 .010151 ,012394 .015938 .02IO16 .027156 .a31564 ,033956

,008427 ,0091I9 .01I891 .015310 ,019336 ,023945 .02871X ,0361 10 ,039515 ,041915

1970 1971 1912 1973 I974 1975 1976 1977 I978 1979

,000000377 ,000000466 .000000806 ,000000869 ,000000869 ,000002600 .oooO18327 .000040989 .oooO79881 .ooOI31 988

,000000384 .00m603 .00ooOII54 .00ooO1151 .00O001608 .00OOO7262 .000025107 .00o040989 .000079881 .00013I988

,038566 ,051958 ,082330 ,131978 .164021 ,464 2.52 6.97 19. I9 49.8 I

,047817 ,066144 . I 17855 ,177216 .212681 ,622 3.73 9.29 22.87 57.0

1980 1981 1982 1983 1984 1985 1986

.O00184079 .OOO442917 ,002I94 167 ,010564833 .067888333 .601144167 ,944275ooO

.000184079 ,000442917 ,003586417 .O 14758333 ,089016667 ,702058333 1.062191667

Parallel

.00oooO155

CPI

100 204 54 1 2,402 17,462 134,836 256.314

WPI

100 209.6 746.6 3,441.3 23,233.3 111,249.4 290,448.3

MI Level (Australs) Year

Average

Annual variation (%)

December

Average

December

Level (Australs) Average

Annual variation (76)

December

Average

December

54.1

44.0

102.3 151.0

126.6 179.0

47.6

41.4

32.6 18.3 12.1 16.2 37.3 29.8 28.2 32..3 30.0 20.1

25.6 15.0 7.0 28.8 39.9 25.8 35.0 29.8 26.8 11.1

201.9 244.8 280.2 345.5 495.5 657.I 844.8 I , 127.I 1,508.4 1,898.9

228.6 269.7 300.1 407.1 589.0 743.4 996.9 1,321.5 1,740.6 2,030.5

33.7 21.3 14.4 2.3 43.4 32.6 28.6 33.4 33.8 25.9

27.7 18.0 11.3 35.7 44.7 26.2 34.1 32.6 31.7 16.7

1,467.5 13.2 2.003.1 25.2 2,881.8 39.4 5,620.7 76.9 8,885.0 69.4 103.I 26,020.0 92,630.0 302.8 151.6 208,440.0 563,340.0 138.1 1,382,900.0 154.7

19.6 36.5 43.9 95.0 58.I 192.9 256.0 125.0 170.3 145.5

2,219.6 2,836.8 4,373.4 7,775.9 13.434.8 24,572.5 117,560.0 365,985.O 1,024.211.7 2,917,795.8

2,477.5 3,558.3 5,501.9 10.614.7 16,766.5 41,990.0 208,160.0 577,180.0 1,550,450.0 4,599,850.0

16.9 27.8 54.2 77.8 72.8 82.9 378.4 211.3 179.9 184.9

22.0 43.6 54.6 92.9 58.0 150.4 395.7 177.3 168.6 196.7

1.851,8OO.O 2,735,000.0 118.5 51.9 2,812,558.3 2,609,500.0 174.6 7,724,483.3 14,864,300.0 .O 30,073,275 69,913,000.0 289.3 195,426,666.7 434,804,000.0 549.8 1,519,607,380.7 2,805,772,973.0 677.6 3,749,678,561.0 4,815,167,644.6 146.8

97.8 68.5 222.5 370.3 521.9 545.3 71.6

6,480,918.3 8,781,590.0 18,187,000.0 12,162,392.5 43,270,500.0 27,943,508.3 91,374,058.3 195,553,000.0 1,320,000,000.0 651,666,666.7 5,241,459,489.2 9,069,766,521.6 13,706,302,624.4 18,745,450,639.1

122.I 67.7 129.8 227.0 613.2 704.3 161.5

90.9 107.I 137.9 351.9 575.0 587.1 106.7

1958 1959

77.2 118.9

98.7 142.1

1960 1961 1962 1963 1964 1965 1966 1967 1968 1969

157.6 186.5 209.1 243.0 333.7 433.0 555.2 734.5 955.0 1,147.3

178.5 205.3 219.6 282.8 395.6 497.5 671.5 871.5 1,104.8 1,227.0

1970 1971 1972 1973 1974 1975 I976 1977 1978 1979

1,299.2 1,626.3 2,267.2 4,010.4 6,791.9 13,791.7 55,550.8 139,778.3 332,817.5 847,540.8

1980 1981 1982 1983 1984 1985 1986

M4

147

ArgentindNotes

Table C.14

(1985 = 100)

Money-Real

MI

M4

Year

Real terms

8 of GDP

Real terms

b of GDP

1958 1959

288.6 190.1

20.8 14.6

110.9 70.0

27.5 18.6

1960 1961 1962 1963 1964 1965 1966 1967 1968 1969

218.1 238.5 205.1 185.2 201.3 210.9 225.5 237.3 281.9 319.3

15.6 15.9 13.9 12.8 12.7 12.1 12.9 13.2 15.1 15.7

81.0 90.8 79.7 76.3 86.6 92.8 99.5 105.6 129.I 153.2

19.9 20.9 18.6 18.3 18.8 18.4 19.6 20.3 23.8 26.0

1970 1971 1972 1973 1974 1975 1976 1977 1978 1979

316.9 284.2 224.4 264.0 372.5 258.6 173.9 175.4 169.8 173.4

14.8 12.8 9.9 11.3 15.1 10.5 7.1 6.7 6.7 6.5

157.0 143.7 125.5 148.4 213.6 133.6 106.7 133.1 151.5 173.1

25.3 22.3 19.1 21.9 29.8 15.0 17.6 20.7 22.2

1980 1981 1982 1983 1984 1985 1986

216.0 156.5 120.7 101.9 98.1 100.0 150.6

7.9 6.2 5.0 4. I 3.8 4.1 5.8

219.2 196.2 126.6 89.8 94.9 100.0 159.6

27.8 26.6 18.1 12.4 12.8 14.1 21.3

18.7

~~

Source: BCRA.

Notes Chapter 1 1. See Ford (1983), Williams (1920), and Diaz Alejandro (1970) for Argentine economic history prior to World War 11. 2. Quoted by Cardoso (1987) and Ford ([1962] 1983, 92). 3. In chapter 3, where we discuss the Martinez de Hoz period, we also give the real exchange rate based on a comparison of manufacturing prices in Argentina and abroad. 4. On the Peronist experience, see, in particular, di Tella (1983) and de Pablo (1980a, 1984).

147

ArgentindNotes

Table C.14

(1985 = 100)

Money-Real

MI

M4

Year

Real terms

8 of GDP

Real terms

b of GDP

1958 1959

288.6 190.1

20.8 14.6

110.9 70.0

27.5 18.6

1960 1961 1962 1963 1964 1965 1966 1967 1968 1969

218.1 238.5 205.1 185.2 201.3 210.9 225.5 237.3 281.9 319.3

15.6 15.9 13.9 12.8 12.7 12.1 12.9 13.2 15.1 15.7

81.0 90.8 79.7 76.3 86.6 92.8 99.5 105.6 129.I 153.2

19.9 20.9 18.6 18.3 18.8 18.4 19.6 20.3 23.8 26.0

1970 1971 1972 1973 1974 1975 1976 1977 1978 1979

316.9 284.2 224.4 264.0 372.5 258.6 173.9 175.4 169.8 173.4

14.8 12.8 9.9 11.3 15.1 10.5 7.1 6.7 6.7 6.5

157.0 143.7 125.5 148.4 213.6 133.6 106.7 133.1 151.5 173.1

25.3 22.3 19.1 21.9 29.8 15.0 17.6 20.7 22.2

1980 1981 1982 1983 1984 1985 1986

216.0 156.5 120.7 101.9 98.1 100.0 150.6

7.9 6.2 5.0 4. I 3.8 4.1 5.8

219.2 196.2 126.6 89.8 94.9 100.0 159.6

27.8 26.6 18.1 12.4 12.8 14.1 21.3

18.7

~~

Source: BCRA.

Notes Chapter 1 1. See Ford (1983), Williams (1920), and Diaz Alejandro (1970) for Argentine economic history prior to World War 11. 2. Quoted by Cardoso (1987) and Ford ([1962] 1983, 92). 3. In chapter 3, where we discuss the Martinez de Hoz period, we also give the real exchange rate based on a comparison of manufacturing prices in Argentina and abroad. 4. On the Peronist experience, see, in particular, di Tella (1983) and de Pablo (1980a, 1984).

148

Rudiger Dornbusch and Juan Carlos de Pablo

Chapter 2 1 . See U.S. International Trade Commission (1986, 76). 2 . Historical annual data on the principal macroeconomic data for the period 1913-84 are reported in a special issue of Estudios (no. 39, July-September 1986). 3. Williams (1920, 125) quotes from the report of the Rothschild Committee to the Bank of England: “the result of the liquidation of Messrs. Baring’s affairs is dependent in a very great measure upon the future value of Argentine securities and obligations; besides which a much larger amount of British capital is engaged in Argentina, the value of which depends greatly on the rehabilitation of the exchanges.” See, too, Hyndman ([1892] 1967) and Fishlow (1985). 4. The debt ratio is calculated by assuming a 1914 debt of $784 billion, as reported by the Economic Commission for Latin America in 1964, and using the data for changes in public external debt in Estudios (no. 39, July-September 1986) to derive the cumulative external debt series. 5. On the Great Depression in Latin America, see Kindleberger (1984), Abreu (1984), O’Connell (1984), and especially Diaz Alejandro (1970, 1983, 1984). 6. The province of Buenos Aires did suspend debt service. 7. See Bittermann (1973) and Avramovic (1966) for some data on postwar external debt. Chapter 3 1. The deficit of the cuenta (per peso of deposits) is given by the combination of reserve requirements (a),the share of time deposits in total deposits (p), and the compensation on time deposit reserves versus the charge on demand deposits, i, and iD, respectively. DEFiD =

( 1 - a)( 1-p)iD

2. See Calvo (1986a, 1987), Dagnino Pastore (1983), Dornbusch (1982), Dornbusch and Fischer (1986). Fernandez (1985), Rodriguez and Sjaastad (1979), and Rodriguez ( 1983). 3. If the real interest rate effect dominates, there will be an initial phase of real appreciation with rising rather than falling inflation. 4. See Dornbusch (1985a) for a comparison of the experience of Argentina, Chile, and Brazil. 5 . See Fundaci6n Mediterranea Newsletter (December 1987).

Chapter 4 I . This chapter owes much of its inspiration to Mundell (1971) and Cardoso and Reis (1986). 2. With B and eB* the domestic currency values of internal and external debt and P the price level, these ratios are defined as b = B/PY and b* = eB*/PY. Note that b* depends on the real exchange rate. Given a debt in dollars, B*, and domestic real GDP, a real depreciation will increase b* and hence worsen the budget deficit. It now takes a larger number of “tax dollars” to pay a “debt service dollar.” 3. The real depreciation will increase the term b*, which for a given dollar value of debts depends on the real exchange rate.

149

ArgentindNotes

4. See Rist (1966) and Joan Robinson’s (1938) review of Bresciani-Turroni (1937) for an exposition of this view. 5. See the discussion in Simonsen (1983) on this point.

Chapter 5 1. This chapter draws on Dornbusch and Simonsen (1987) and de Pablo (1986, 1987). 2. How do we know that the economy cannot? This, too, is an issue in game theory. Schelling (1982) has written extensively on how to make threats stick. Ideas involve poison-pill strategies. Governments instinctively shy away from poison-pill strategies of no return. As a result, the public does not believe the fierce anti-inflation rhetoric and, therefore, the government in turn cannot afford to implement it. The late William Fellner (1976) developed these themes with great authority. See, too, Anti-inflation (1982). 3. See Simonsen (1986a, 1986b) for a game theory formulation where it is shown how players’ use of maximin pricing leads to an equilibrium different from the Nash equilibrium that is assumed in full information, rational expectations models. 4. See Dornbusch and Fischer (1986). 5. The Krieger-Vasena reform provided for initial wage adjustments that differed by sector, recognizing that across sector there were differences in the time since the last adjustment. See de Pablo (1972). 6. See Sargent (1982), Dornbusch (1985c), and Dornbusch and Fischer (1986). 7 . There is a difference in the required adjustments depending on the presence or absence of fees. In a system without fees, the banking system needs to achieve larger reductions in costs and hence in employment than under a fee-based banking service. 8. See Arnaudo (1979), di Tella (1983), and de Pablo (1980a) on the Peronist inflation and the failed attempt to stabilize it with controls. 9. On the Martinez de Hoz debacle, see Corbo and de Melo (1985), Diaz Alejandro (1982, 1988), Dornbusch (1982, 1985a), and Fernandez and Rodriguez (1982). 10. The most comprehensive treatment of the Austral Plan may be found in Heyman ( 1986). 11. Broda (1986) in particular emphasizes the importance of these early measures in the initial success of the Austral Plan. 12. Paul Volcker, as chairman of the Federal Reserve Board, is said to have played a significant role in achieving IMF and bank acceptance of the plan. He subsequently visited Argentina in November of 1985. 13. There is a need to make some allowance for growth of the informal and black market economy. Estimates are notoriously difficult, but there is a pervasive belief in Argentina that (just as in Italy) the underground economy has experienced much higher growth than the formal economy. A study by Guissari (1987), for example, reports growth rates for the formal (or registered) economy of 0.7 percent per year in 1974-83 and 1.3 for the entire economy, including the informal sector. For the early 1980s, this study finds that the share of the informal sector is 30 percent. No doubt, just as in other countries, the underground is an important and growing feature but, just as elsewhere, estimates of the size must remain tentative. 14. SOCMERC is a public opinion survey data base belonging to the consulting firm Aftalion, Mora y Araujo, Nogueira. 15. Reported in La Nacidn (2 January 1986). 16. See Fischer and Modigliani (1978).

150

Rudiger Dornbusch and Juan Carlos de Pablo

Chapter 6 1. We are indebted to Marcello Selowsky for suggesting this framework. The basic analysis is set out in Selowsky and van der Tak (1986). An alternative framework based on intertemporal optimization is reported in Blanchard (1983). 2. McCarthy (1987) reports a capital output ratio for Argentina of 4.35 in 1986. Assuming a depreciation rate of 3 percent per year, this yields an investment requirement of 13 percent just to keep the capital stock unchanged. It is apparent that the present level of investment implies that net investment has been negative. 3. In scenario B, the debtlincome ratio by the year 2010 (not shown in the table) reaches b = 66.7. 4. See Dombusch and Park (1987) for an analysis of Korean development. 5. See Dombusch and Park (1987). 6. See Carta Econdmico (April 1986), 75-83. For an extensive discussion of distortions in the Argentine economy, see Nogues (1985, 1986, 1987). 7. See Ministerio de Economia (1986). 8. With an external debt ratio of 64 percent and an interest rate of 10 percent, the interest payments amount to 6.4 percent of GDP. A 15 percent real depreciation thus raises interest rates by 1 percent of GDP. 9. For a discussion of the economics of a free trade agreement between Argentina and Brazil, see Dornbusch (1981).

References Abreu, M. de Paiva. 1984. Argentina and Brazil during the 1930s: The impact of British and American international economic policies. In Latin America in the 1 9 3 0 ~ed. ~ R. Thorp. London: Macmillan. Allais, M. 1966. A restatement of the quantity theory of money. American Economic Review 56, no. 5 (December): 1123-57. Anti-inflation policies and the problem of credibility: Symposium. 1982. American Economic Review 72, no. 2 (May): 77-91. Amaudo, A. 1979. El Programa antiinflacionario de 1973. Desarrollo Economico 19, no. 73 (April-June): 25-52. Avramovic, D. 1966. Economic growth and external debt. Baltimore: Johns Hopkins University Press. Banco Central de la Republica Argentina (BCRA). 1972. La CreaciBn del Banco Central. 2 vols. Buenos Aires: BCRA. Barletta, N. A., M. I. Blejer, and L. Landau, eds. 1984. Economic liberalization and stabilization policies in Argentina, Chile, and Uruguay. Washington, D.C.: The World Bank. BCRA. See Banco Central de la Repdblica Argentina. Bittermann, H. 1973. The refunding of international debt. Durham, N.C.: Duke University Press. Blanchard, 0. 1983. Debt and the current account deficit in Brazil. In Financial policies and the world capital market: The problem of Latin American countries,

150

Rudiger Dornbusch and Juan Carlos de Pablo

Chapter 6 1. We are indebted to Marcello Selowsky for suggesting this framework. The basic analysis is set out in Selowsky and van der Tak (1986). An alternative framework based on intertemporal optimization is reported in Blanchard (1983). 2. McCarthy (1987) reports a capital output ratio for Argentina of 4.35 in 1986. Assuming a depreciation rate of 3 percent per year, this yields an investment requirement of 13 percent just to keep the capital stock unchanged. It is apparent that the present level of investment implies that net investment has been negative. 3. In scenario B, the debtlincome ratio by the year 2010 (not shown in the table) reaches b = 66.7. 4. See Dombusch and Park (1987) for an analysis of Korean development. 5. See Dombusch and Park (1987). 6. See Carta Econdmico (April 1986), 75-83. For an extensive discussion of distortions in the Argentine economy, see Nogues (1985, 1986, 1987). 7. See Ministerio de Economia (1986). 8. With an external debt ratio of 64 percent and an interest rate of 10 percent, the interest payments amount to 6.4 percent of GDP. A 15 percent real depreciation thus raises interest rates by 1 percent of GDP. 9. For a discussion of the economics of a free trade agreement between Argentina and Brazil, see Dornbusch (1981).

References Abreu, M. de Paiva. 1984. Argentina and Brazil during the 1930s: The impact of British and American international economic policies. In Latin America in the 1 9 3 0 ~ed. ~ R. Thorp. London: Macmillan. Allais, M. 1966. A restatement of the quantity theory of money. American Economic Review 56, no. 5 (December): 1123-57. Anti-inflation policies and the problem of credibility: Symposium. 1982. American Economic Review 72, no. 2 (May): 77-91. Amaudo, A. 1979. El Programa antiinflacionario de 1973. Desarrollo Economico 19, no. 73 (April-June): 25-52. Avramovic, D. 1966. Economic growth and external debt. Baltimore: Johns Hopkins University Press. Banco Central de la Republica Argentina (BCRA). 1972. La CreaciBn del Banco Central. 2 vols. Buenos Aires: BCRA. Barletta, N. A., M. I. Blejer, and L. Landau, eds. 1984. Economic liberalization and stabilization policies in Argentina, Chile, and Uruguay. Washington, D.C.: The World Bank. BCRA. See Banco Central de la Repdblica Argentina. Bittermann, H. 1973. The refunding of international debt. Durham, N.C.: Duke University Press. Blanchard, 0. 1983. Debt and the current account deficit in Brazil. In Financial policies and the world capital market: The problem of Latin American countries,

151

ArgentindReferences

ed. P. Aspe Armella, R. Dornbusch, and M. Obstfeld, 187-97. Chicago: University of Chicago Press. Bresciani-Turroni, C. 1937. The economics of injlation. London: Allen & Unwin. Broda, M. 1986. The Austral Plan. World Bank Working Paper. Washington, D.C.: The World Bank. Bruno, M., G. di Tella, R. Dornbusch, and S. Fischer, eds. 1988. Stopping high inflation. Cambridge, Mass.: MIT Press. Cagan, P. 1956. The monetary dynamics of hyperinflation. In Studies in the quantity theory of money, ed. M. Friedman, 25-117. Chicago: University of Chicago Press. Calvo, G. 1986a. Fractured liberalism: Argentina under Martinez de Hoz. Economic Development and Cultural Change 34, no. 3 (April): 511-33. . 1986b. Incredible reforms. Department of Economics, University of Pennsylvania, mimeo. . 1987. Inflation and financial reform. Department of Economics, University of Pennsylvania, mimeo. Carballo de Cilly, M. 1987. Que pensamos 10s Argentinos. Buenos Aires: Ediciones el cronista comercial. Cardoso, E. 1986. What policymakers can learn from Brazil and Mexico. Challenge 29, no. 4 (September-October): 19-28. . 1987a. Latin America’s debt: Which way now? Challenge 30, no. 2 (May-June): 11- 17. Cardoso, E., and R. Dornbusch. 1987. Brazil’s tropical plan. American Economic Review 77, no. 2 (May): 288-92. Cardoso, E., and E. Reis. 1986. Debts, deficits and inflation (In Portuguese). Pesquisa e Planejamento Economico 4 (December): 575-98. Cavallo, D. 1982. La Reforma Jinanciera. Buenos Aires: Banco Central de la Republica Argentina. ~. 1984. Volver a crecer. Buenos Aires: Sudamericana-Planeta. . 1986. Long-term growth in the light of external balance policies. The case of Argentina. In The open economy, ed. R . Dornbusch and L. Helmers. Oxford: Oxford University Press. Cavallo, D., and A. Pena. 1983. Deficit fiscal, endeudamiento del Gobierno y tasa de inflacion: Argentina, 1940- 1982. Estudios 6, no. 26 (April-June). CEPAL. See Cornision economica para la America Latina. Cornision economica para la America Latina (CEPAL). 1958. El desarrollo economico en la Argentina. Santiago: CEPAL. Corbo, V., and J. de Melo, eds. 1985. Liberalization with stabilization in the Southern Cone of Latin America. World Development, special issue, 13, no. 8 (August): 863- 1016. Cumby, R., and R. Levich. 1987. On the definition and magnitude of recent capital flight. In Capital jlight and third world debt ed. D. Lessard and J. Williamson. Washington, D.C.: Institute of International Economics. Dagnino Pastore, J. M. 1983. Progress and prospects for the adjustment in Argentina. In Prospects for adjustment, ed. J. Williamson. Cambridge, Mass.: MIT Press, for the Institute for International Economics. de Pablo, J. C. 1972. Politica antiinjlacionaria en la Argentina, 1967-70. Buenos Aires: Amorrortu editores. -. 1980a. Politica economica del peronismo. Buenos Aires: El Cid Editor. . 1980b. La economia que yo hice. Buenos Aires: Ediciones el cronista comercial. -. 1984. Politica economica argentina: Materiales para el desarrollo del tema segun el metodo de 10s casos. Buenos Aires: Edicicines Macchi.

152

-.

Rudiger Dornbusch and Juan Carlos de h b l o

1986. La economia que yo hice. Vol. 2. Buenos Aires: Ediciones el cronista comercial. -. 1987. Proceso a la economia del proceso. Segundas jornadas anuales de economia. Montevideo: Banco Central del Uruguay. Diaz Alejandro, C. 1964. Exchange devaluation in a semi-industrialized country. Cambridge, Mass.: MIT Press. -. 1970. Essays on the economic history of the Argentine Republic. New Haven: Yale University Press. -, 1982. Southern Cone stabilization plans. In Economic stabilization in developing countries, ed. W. Cline and S. Weintraub. Washington, D.C.: The Brookings Institution. -. 1983. Stones of the 1930s for the 1980s. In Financialpolicies and the world capital market: The problem of Latin American countries, ed. P. Aspe Armella, R. Dornbusch, and M. Obstfeld, 5-40. Chicago: University of Chicago Press. . 1984. Latin America in the 1930s. In Latin America in the 1930s, ed. R. Thorp. London: Macmillan. -. 1988. No less than one hundred years of Argentine economic history plus some comparisons. In Trade, development and the world economy. Selected essays of Carlos Diaz Alejandro, ed. A. Velasco. Oxford: Basil Blackwell. di Tella, G. 1983. Argentina under Peron, 1973-76: The nation’s experience with a labor-based government. New York: St. Martin’s Press. di Tella, G., and R. Dornbusch, eds. 1988. Argentine economic policy 1Y4.5-83. Pittsburgh: University of Pittsburgh Press. di Tella, G . , and M. Zymelman, 1967. Las etapas del desarrollo economico argentino. Buenos Aires: Eudeba. Dornbusch, R. 1978. Inflation, capital and deficit finance. Journal of Money, Credit and Banking 1 (February): 141-50. -. 1981. Los Costos y beneficios de la integracion economica regional. Estudios 4 (July-September). Reprinted. integracion Latinoamericana 1 1 (June 1986). -. 1982. Stabilization policy in developing countries: What lessons have we learnt? World Development, special issue, 10, no. 9 (September): 701-8. -. 1985a. External debt, budget deficits and disequilibrium exchange rates. In International debt and the developing countries, ed. G. Smith and J. Cuddington. Washington, D.C.: The World Bank. -. 1985b. The Larida proposal: Comment. In Inflation and indexation, ed. J . Williamson. Washington, D.C.: Institute for International Economics. -. 198%. Stopping hyperinflation: Lessons from the German experience in the 1920s. In Macroeconomics andfinance: Essays in honor of Franco Modigliani, ed. R. Dornbusch, S. Fischer, and J. Bossons. Cambridge, Mass.: MIT Press. Dornbusch, R., and S. Fischer. 1986. Stopping hyperinflation: Past and present. WeltwirtschaftlichesArchiv 122, no. 1 (April): 1-47. Dornbusch, R., and Y. C. Park. 1987. Korean growth policy. Brookings Papers on Economic Activity 2: 389-444. Dornbusch, R., and M. Simonsen. 1987. Inflation stabilization with incomes policy support. New York: Group of Thirty. Elias, V. 1982. El crecimiento economico argentino y sus determinantes: 1970-80. Ensayos Economicos 21 (March). -. 1987. El crecirniento economico argentino a largo prazo. El Cronista Comercial (23 November). Feldstein, M. 1987. Muddling through is just fine. The Economist (London), 27 June. 21-25.

153

ArgentindReferences

Fellner, W. 1976. Toward a reconstruction of macroeconomics. Washington: American Enterprise Institute. Fernandez, R. B. 1985. The expectations management approach to stabilization in Argentina, 1976-82. World Development 13, no. 8 (August): 871-92. . 1987. Crecer en libertad. Buenos Aires: Ediciones el cronista comercial. Femandez, R. B., and C. A. Rodriguez, eds. 1982. Inflacion y estabilidad. Buenos Aires: Edicidnes Macchi. FIEL. See Fundaci6n de investigaciones economicas latinoamericanas. Fischer, B., U. Hiemenz, and P. Trapp. 1985. Argentina. The economic crisis in the 1980s. Tubingen: J. C. B. Mohr. Fischer, S . , and F. Modigliani. 1978. Towards an understanding of the real effects and costs of inflation. Weltwirtschaftliches Archiv 114 (4): 810-32. Reprinted in Indexing, inflation and economic policy, S . Fischer. Cambridge, Mass.: MIT Press, 1986. Fishlow, A . 1985. Lessons from the past: Capital markets during the 19th century and the interwar period. International Organization 39, no. 3 (Summer): 383-439. Ford, A. G. 1962. The gold standard, 1880-1914: Great Britain and Argentina. Oxford: Oxford University Press. Reprinted. New York: Garland Publishing Co., 1983. Fundacion de investigaciones economicas latinoamencanas (FIEL). 1987. El gasto public0 en la Argentina, 1960-1985. Buenos Aires. Graham, F. 1928. Exchange, prices and production in hyperinflation Germany. Princeton: Princeton University Press. Guissari, A. 1987. Economia informal. Documento de Trabajo no. 4. Buenos Aires: IDEC . Harberger, A. 1964. The inflation problem in Latin America. Paper prepared for the meeting of the Alliance for Progress, Buenos Aires. Heyman, D. 1986. Tres ensayos sobre inflacion y politicas de estabilizacion. Doc. no. 18. Buenos Aires: CEPAL. Hyndman, H. 1892. Commercial crises of the nineteenth century. Reprinted. New York: Augustus Kelley, 1967. Instituto de Estudios Economicos sobre la Realidad Argentina y Latinoamericana. Estatisticas de la evolution economica de Argentina. 1913- 1984. Estudios 9, no. 39 (July-September): Cordoba. Keynes, J. M. 1923. A tract on monetary reform. Reprinted. London: The Royal Economic Society, 1971. Kindleberger, C. 1984. The 1929 world depression in Latin America-From the outside. In Latin America in the 1930s, ed. R. Thorp. London: Macmillan. Lopez Murphy, R. 1984. Aspectos fiscales de la cuenta regulation monetaria. Skptimo jornadas de economia monetaria y sector externo. Buenos Aires: Banco Central de la Repdblica Argentina, May. McCarthy, D. 1987. Argentina towards the year 2000. The World Bank, Washington, D.C., mimeo. McCarthy, D., and A. Thorne. 1987. Argentina: Problems for achieving macro stability. The World Bank, Washington, D.C., mimeo. Maddison, A. 1982. Phases of capitalist development. Oxford: Oxford University Press. . 1985. Two crises: Latin America and Asia, 1929-38 and 1973-83. Paris: OECD. Mallon, R. 1987. Uncertainty and policy flexibility in Argentina. Harvard University, Cambridge, Mass., mimeo.

154

Rudiger Dornbusch and Juan Carlos de Pablo

Mallon, R., and J. Sourrouille. 1975. Economic policy making in a conflict sociery. Cambridge, Mass.: Harvard University Press. Melnick, R., and M. Sokoler. 1984. The government revenue from money creation and the inflationary effects of a decline in the rate of growth of GNP. Journal of Monetary Economics 13, no. 2: 225-36. Ministerio de Economia. 1986. Batalla contra la evasion fiscal. Buenos Aires, August. Mundell, R. A. 1971. Monetary theory. Pacific Palisades: Goodyear. Nogues, J. 1985. Distortions, factor proportions and efficiency losses: Argentina in the Latin American Scenario. Weltwirtschaftliches Archiv (June). -. 1986. The nature of Argentina’s policy reforms during 1976-81. Staff Working Paper no. 765. Washington, D.C.: The World Bank. -. 1987. Discriminatory economic policies as determinants of Argentina’s trade and employment prospects. The World Bank, Washington, D.C., mimeo. O’Connell, A. 1984. Argentina in the depression: Problems of an open economy. In Latin America in the 1930s, ed. R. Thorpe. London: Macmillan. Olivera, J. 1967. Money, prices and fiscal lags: A note on the dynamics of inflation. Banca Nazionale del Lavoro, Quarterly Review, no. 26. Pazos, F. 1972. Chronic inflation in Latin America. New York: Praeger Publishers. Peters, H. E. 1934. The foreign debt of the Argentine Republic. Baltimore: Johns Hopkins University Press. Polak, J. 1943. European exchange depreciation in the early twenties. Econornetrica 11, no. 2: 151-62. Ramos, J. 1986. Neoconservative economics in the Southern Cone of Latin America, 1973-83. Baltimore: Johns Hopkins University Press. Rist, C. 1966. History of monetary and credit theory. New York: Augustus Kelley. Robinson, J. 1938. The economics of inflation. Economic Journal 48, no. 3 (September): 507- 13. Rodriguez, C. A. 1982. The Argentine stabilization plan of December 20th. World Development, special issue, 10, no. 9 (September): 801 - 11. -. 1983. Politicas de estabilizacion en la economia Argentina, 1978- 1982. Cuadernos de Economia 20, no. 59 (April): 21 -42. -. 1986a. Un analisis estilizado de la reforma financiera de julio de 1982. Centro de Estudios Macroeconomicas de Argentina, mimeo, July. -. 1986b. El Problema de la deuda externa. La Nacidn (29 December). -. 1986c. La Argentina debe proponer, unilateralmente, el concurso de acredores. Interview. El Cronista Comercial (4 December). . 1987. The foreign debt of Argentina. Centro de Estudios Monetarios de Argentina (CEMA), Buenos Aires, mimeo. Rodriguez, C. A., and L. A. Sjaastad. 1979. El atraso cambiario en la Argentina, mito o realidad? Working Paper no. 2. Buenos Aires: CEMA, June 1. Sargent, T. 1982. The ends of four big inflations. In Inflation: Causes and effects, ed. R. Hall, 41-97. Chicago: University of Chicago Press. Sargent, T., and N. Wallace. 1981. Some unpleasant monetarist arithmetic. Federal Reserve Bank of Minneapolis, Quarterly Review 5, no. 4 (Fall): I - 17. Schaefer, C. A. 1922. Klassische Valuta-Stabilisierungen und ihre Lehren fur die Markstabilisierung. Hamburg: Verlag von C. Boysen. Schelling, T. 1982. Thermostats, lemons and other families of models. In Micromotives and macrobehavior. New York: Norton. Selowsky, M., and H. van der Tak. 1986. The debt problem and growth. World Development 14, no. 9 (September): 1107-24. Simonsen, M. 1983. Indexation: Current theory and the experience in Brazil. In

155

ArgentindReferences

Inflation, debt and indexation, ed. R. Dombusch and M. H. Simonsen. Cambridge, Mass.: MIT Press. . 1986a. Incomes policy as game theory. Rio de Janeiro: FundaGiio Getulio Vargas. . 1986b. Keynes versus expectativas racionais. Rio de Janeiro: FundaGiio Getulio Vargas, mimeo. Sjaastad, L. A. 1983. Failure of economic liberalism in the Cone of Latin America. The World Economy 6, no. 1 (March): 5-26. Sourrouille, J. 1983. Politica economica y procesos de desarrollo. La experiencia Argentina entre 1976 y 1981. Estudios e informes de la CEPAL, no. 27 (June). Sturzenegger, A. 1986. Los Efectos de la politica comercial sobre el tipo de cambio. Estudios 9, no. 37 (January-March): 37-46. Summers, R., and A. Heston. 1984. Improved international comparisons of real product and its components: 1950-80. Review of Income and Wealth 30, no. 2 (June): 207-62. Tanzi, V. 1977. Inflation, lags in collection and the real value of tax revenue. IMF StafSPapers 24, no. 1 (March): 154-67. . 1978. Inflation, real tax revenue and the case for inflationary finance: Theory with an application to Argentina. IMF Staff Papers 25 (September): 417-51. U.S. International Trade Commission. 1986. The effects of developing country debt-servicing problems on 1J.S. trade. USITC Publication 1950. Washington, D.C. Vazquez-Presedo, V. 1971. Estadisticas historicas argentinas (comparadas). 2 vols. Buenos Aires: Edici6nes Macchi. Watson, M., R. Kincaid, C. Atkinson, E. Kalter, and D. Folkerts-Landau. 1986. International capital markets. Washington, D.C.: International Monetary Fund. Williams, J. H. 1920. Argentine international trade under inconvertible paper money. 18XO-1900. Reprinted. New York: AMS Press, 1971. World Bank. 1980. Economic memorandum on Argentina. Washington, D.C. -. 1985a. Argentina. Economic memorandum. Washington, D.C. -. 1985b. Argentina. Srrategies toward industrial and export development. Washington, D.C. -. 1987. Argentina. Economic recovery and growth. Report no. 6467-AR. Washington, D.C.

This Page Intentionally Left Blank

1

An Overview of Macroeconomic Performance

In the first half of the 1980s, Bolivia experienced an economic crisis of extraordinary proportions. As seen in table 1 . 1 , Bolivia’s economic debacle in this period was striking even in comparison with the poor performance of Bolivia’s neighbors. Like its neighbors, Bolivia suffered from major external shocks, including the rise in world interest rates in the early 1980s, the cutoff in lending from the international capital markets, and the decline in world prices of Bolivia’s commodity exports. But the extent of economic collapse in the face of these shocks suggests that internal factors as well have been critical to Bolivia’s economic performance. The Bolivian hyperinflation of 1984-85, for example, was one of the most dramatic inflations in world history and one of the only hyperinflations that did not result from the dislocations of war or revolution. In some ways, Bolivia’s economic recovery since 1985 is as remarkable as the economic crisis during the first half of the decade. In the fall of 1985, the new government of President Victor Paz Estenssoro succeeded in eliminating Bolivia’s hyperinflation. The inflation was cut from 25,000 percent per year during the hyperinflation to a remarkable 10.7 percent per year in 1987.’ As shown in figure 1.1, the disinflation was almost instantaneous: by early 1986 virtual price stability had been restored. The stabilization is all the more amazing in view of the extremely large and adverse shocks which hit the Bolivian economy after the start of the stabilization program. As we discuss below, the prices of all of Bolivia’s major exports (even the illicit exports such as coca leaf) plummeted in late 1985 and after. The terms-of-trade deterioration further squeezed real incomes, worsening the decline of over 25 percent in per capita income levels that had occurred in the first half of the 1980s. The terms-of-trade decline further embittered the political conflicts in Bolivia, which have always been acute, and also directly reduced government revenues, thus threatening the macroeconomic stabilization program. 159

160

Juan Antonio Morales and Jeffrey D. Sachs

Table 1.1

Economic F'erformance in the 198Os, Selected Countries

Country

GDP Growth (annual rate),

Inflation (annual rate),

1980-85

Bolivia Latin America Argentina Brazil Mexico East Asia Malaysia Indonesia Korea

Terms of Trade, 1985

1980-85

Debt-GDP Ratio,a 1985

(1980= 100)

-4.5

569.1

136.8

86

-1.4 1.3

56.4 43.8 52.8

88 87

.8

342.8 147.7 62.2

98

5.5 3.5 7.9

3.1 10.7 6.0

62.0 36.6 43.0

85 97 105

Source: World Bank (1988).

"Medium- and long-term debt, public and private.

In 1987 Bolivia experienced its first positive GNP growth after seven years of decline, with an increase in GNP of 2.3 percent. Growth increased to about 3.0 percent in 1988. These positive growth rates during 1987-88 averaged about the same as population growth (2.7 percent per year), so that per capita income remained stagnant after eight years of sharp decline. The greatest disappointment of the stabilization experience after 1985 indeed has been the failure of Bolivia to resume solid and

JFMAMJJASONDJFMAMJJASONDJFMAMJJASONDJFMAMJJASONDJFMAM

1984 0

1985

1986

1987

First month of implementation of program

Fig. 1.1 Monthly inflation rates, 1984-88 Source; Institute of National Statistics, Bolivia.

1988

161

BolividChapter 1

positive per capita GNP growth. In our view, the continuing economic stagnation is not a result of the stabilization effort per se, but is rather a result of the recent external shocks together with internal disarray that preceded the stabilization program. This monograph focuses on Bolivia’s economic crisis and recovery in the 1980s, with a special emphasis on the role of external debt in the economic dynamics. The bulk of our attention is on the period from 1952 to 1985, from the Bolivian Revolution to the end of the hyperinflation. We do consider more briefly the period of economic recovery, 1986-88, but in view of the lack of perspective on recent events, we are not as comprehensive in our coverage of the years after 1985. We view the hyperinflation as the culmination of deep trends in the Bolivian economy and society, rather than the result of short-run forces in the early 1980s. One pervasive theme of our work is that the recent economic crisis is a reflection of political and social conflicts in Bolivian society that have undermined the development process for decades. We thus lay great stress on the failures of politics in accounting for the failures of economic performance.

1.1 Some Basic Characteristics of the Bolivian Economy Bolivia is a landlocked country with the second lowest per capita income in Latin America (U.S.$470 in 1985). All indicators of economic and social development in Bolivia lie well below the Latin American averages. The economy has been heavily dependent on two exports, tin and natural gas, for the past thirty-five years. Not surprisingly, the cycles of expansion and contraction of the Bolivian economy have followed quite closely the movements in the world market prices of tin and petroleum. Moreover, as we shall see, the accumulation of foreign debt up until 1980 took place in a procyclical pattern, in that Bolivia became able to attract foreign loans precisely during the phase of high international prices for its exports. At the same time, in the mid-l970s, a thriving underground economy developed based on coca-related exports, a development that has deep implications for the functioning of the economy and the country’s political institutions. Bolivia’s long-term growth problems begin with the inherent difficulties of growth in a landlocked mining economy centered in breathtaking, yet forbidding, terrain more than 14,000 feet above sea level. Bolivia is a very large and sparsely populated country. Its area is greater than the combined areas of the United Kingdom, West Germany, and France, with a population of only six million. The country is divided geographically between the Andean highlands, where the mining industry is located, and the lowlands to the east, where petroleum products and commercial agriculture are situated. From the beginning of colonial development, Bolivia’s political and economic

162

Juan Antonio Morales and Jeffrey D. Sachs

center of gravity has been in the highlands. The secular decline of the mining sector has led in recent decades to a shift in economic and political activity to the east. This shift, which is the cause of significant political conflict, was accelerated by the sharp crash of tin prices in October 1985. Bolivia’s economy since the colonial period has been based on mining, first silver from the colonial period until the late nineteenth century, and then tin in the twentieth century. With the population centered in the highlands and neighboring valleys, transport costs are extremely high. Therefore, while it has been profitable for Bolivia to export minerals such as silver and tin that have a high value added per unit weight, and therefore a low share of transport costs per unit value, Bolivia has been unable to overcome the transport costs for almost any kind of manufactured product. The problem of transport costs was greatly exacerbated by Bolivia’s loss of its littoral on the Pacific Coast in the traumatic War of the Sea against Chile and Peru in 1879. The requirement of high value added per unit weight in Bolivian exports helps to explain Bolivia’s only real diversification of exports in recent years: coca leaf derivatives (the precursors of cocaine) and petroleum products. Petroleum products, particularly natural gas exported to Argentina, can be carried out of the country by pipeline, while coca leaf can be profitably transported even if carried on a peasant’s back. Table 1.2 shows Bolivia’s concentration of non-coca exports in recent years. Among Bolivia’s measured exports (i.e., excluding coca), tin and petroleum products usually accounted for much more than half of total exports during the 1960s, 1970s, and 1980s. Coca exports in the 1980s are generally estimated to equal the sum of tin and natural gas exports. Share of Tin and Natural Gas in Total Exports of Bolivia, 1952-84

Table 1.2 Year

As % of Merchandise Exports

1952 1953 1954 1955 I956 1957 1958 1959 1960 1961 1962 1963 I964 1965 1966 1967 1968

59.9 64.2 55.2 56.0 55.1 58.9 56.1 68.2 71.0 87.3 89.7 80.6 80.8 80.5 70.2 58.5 58.9

Year

As % of Merchandise Exports

1969 1970 1971 1972 1973 1974 1975 I976 1977 1978 1979 1980 1981 1982 1983 1984

57.3 52.0 58.4 61.3 57.2 46.6 48. I 47.9 62.4 72.0 65.9 63.6 68.3 73.5 77.6 86. I

Source: Central Bank of Bolivia, Boletin Estodistico (La Paz), various issues

163

BolividChapter 1

Many of Bolivia’s problems can be seen as the tragic playing out of the secular decline in the mining sector, a process that has been underway for at least half a ~ e n t u r y .Actually, ~ Bolivia’s first deep crisis came with the depletion of silver deposits and the fall of world silver prices at the end of the last century. Good fortune, however, pulled Bolivia out of crisis when a boom in world demand for tin followed the development of the modem canning process. The apogee of modem Bolivian economic development relative to the neighboring countries was reached in the first two decades of the twentieth century, when tin was intensively developed and when export earnings provided the basis for a major extension of the country’s infrastructure, including roads and the railway system. Tin lodes were increasingly exhausted in the 1920s, and Bolivia began to lose export competitiveness with other producers. When tin prices collapsed at the start of the Great Depression, Bolivia became the first country in that crisis to default on its sovereign foreign debt, in January 1931. Bolivia was also the last country in Latin America to settle these defaulted debts after World War 11. The collapse of tin undermined both the economy and the political system, and eventually ushered in the Revolution of 1952. 1.1.1 Aggregate Growth Table 1.3 shows some indicators of long-term g r ~ w t hReal . ~ GDP grew between 1952 and 1985 at the average annual rate of 2.4 percent. If the crisis years of 1982-85 are left out, the average annual rate is 3.2 percent (or 0.4 percent in per capita terms). The growth, of course, was not smooth and was characterized by well-defined cycles. The years 1952-57, in the aftermath of the Revolution, showed negative rates of growth as important social changes prompted major dislocations of production. In particular, agricultural output in cash crops suffered a substantial reduction after the first year of an ambitious agrarian reform. The high inflation of those years, coupled with misguided policy measures that attempted to suppress the symptoms of inflation, also had important negative real effects. After the inflation stabilization of late 1957 until the early 1960s, growth resumed at a moderate pace. In 1959 the negative rate of growth is explained by a sharp drop in tin prices brought about by the mild world recession of 1958-59 and the sales of tin in international markets by the Soviet Union. From 1962 to 1972, very healthy rates of growth are observed, higher on average than the growth rates of the “export boom years” between 1972 and 1977. The growth in the 1960s is due importantly to a favorable international environment, internal political stability under President Barrientos, and the payoffs on earlier investments in petroleum and commercial agriculture in eastern Bolivia. Appropriate use of foreign aid funds, entrepreneurship provided by the educated Bolivians who had fled from the Revolution and had returned in the mid- 196Os, and the substantial investment expenditures made by the Bolivian Gulf Oil Corporation were also important sources of growth.

164

Juan Antonio Morales and Jeffrey D. Sachs

Table 1.3

Real GDP, Levels and Rates of Growth, 1952-88

Year

Real GDP

1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 I984 1985 19x6 1987 1988

7.600 6.000 7.023 7.395 6.957 6.726 6.887 6.864 7.159 7.309 7.717 8.213 8.608 9.202 9.847 10.470 11.222 11.757 12.370 12.976 13.729 14.646 15.400 16.417 17.418 18.151 18.761 19.104 18.990 19.064 18.528 17.314 17.160 16.660 16.375 16.752 17.255

Growth Rate (%)

-9.47 2.00 5.30 -5.92 -3.32 2.39

Five-year Average (%)

-2.27

- .33

4.30 2.10 5.58 6.43 4.81 6.90 7.01 6.33 7.10 4.77 5.21 4.90 5.80 6.68 5.15 6.60 6. I0 4.21 3.36 1.83 - .55 .35 2.81 - 6.55 - .87 - 1.75 -2.00 2.30 3.00

2.81

6.29

5.57

5.75

.43

-2.00

Source: Central Bank of Bolivia, Cuentas Nacionales no. 4 , Departamento Cuentas Nacionales (La Paz: 1983). and for 1984-88, IMF, Internarional Financial Statisrics. Notes: 1988, estimated. Five-year averages are calculated as geometric averages over the period.

Growth remained relatively strong during the Banzer regime, 1971-78, but the bases of the observed prosperity became more tenuous. Bolivian growth was based importantly on heavy foreign borrowing from private banks, as well as a temporary commodity price boom which favored the country. The pace of growth started to sharply decline from 1978 on. The high inflation of 1982-85, to which we devote substantial attention in chapter 5 , was accompanied by negative rates of growth. In regard to the composition of output, shown in table 1.4, primary production, including agriculture, accounted for 47 percent of GDP in 1952;

165

BolividChapter 1 Sectoral Composition of GDP (percentages)

Table 1.4 Sector

1952

1957

1962

1967

1972

1977

1982

198Y

Primary Manufacturing Nontradables I Nontradables I1 Indirect taxes GDP

46.51

44.84 12.34 4.60 38.23

39.94 14.05 5.42 40.60

38.40 15.28 6.30 40.02

29.08 14.60 5.35 50.97

25.75 15.86 5.93 52.46

34.21 11.96 4.04 46.15

32.23 9.83 4.15 50.69 3.10 100

14.15 4.61 34.73 b

b

b

b

b

b

100

100

100

100

100

100

2.64

100

Sources: Ministry of Planning, Revisfa de Planificacion y Desarrollo (La Paz: 1970); Central Bank of Bolivia, Boletin Esradistico (La Paz), various issues. Norest Percentages are computed from basic data in real terms. Nontradables I = energy and constmction.

Nontradables I1

=

commerce, transportation, and other services.

”Estimated figures. bIndirect taxes are included in each sector.

the corresponding percentage was 32 in 1985. The share of manufacturing in GDP in 1952 was 14 percent, compared with a low 9 percent in 1985. Construction and energy accounted for almost 5 percent of GDP in 1952 and for 4 percent in 1985. Commerce, transportation, and other services (denoted “nontradables 11” in the table) constituted 35 percent of GDP in 1952, with that share rising to 51 percent in 1985. This latter group of industries, which experienced such a dramatic increase in share, are generally nontradables. Between 1967 and 1977, services increased its share in the economy from 40 percent to more than 50 percent of GDP. This change in the composition of GDP, with a higher share of nontradables (services plus a significant but unknown percentage of agricultural and manufacturing output), became in the 1980s an important explanatory factor of the difficulties of servicing the accumulated foreign debt.

1.1.2 Foreign Trade Bolivia’s growth and development have been very strongly tied to foreign trade. With the exception of a few years in the period 1952-82, generally associated with strong political disruptions, exports amounted to more than 20 percent of GDP in current prices (table 1.5). The dependency on imports is obviously very high with import-GDP ratios rarely below 20 percent (in some years, the ratio is over 25 percent). As we noted in table 1.2, Bolivia’s legal exports during 1952-85 were highly concentrated in two major exports, tin and natural gas, with natural gas exports beginning in 1972. In some years, more than 70 percent of measured exports are in these two commodities. Export diversification was one of the principal economic policy goals of the revolutionary period of 1952-56. However, after the stabilization program of December 1956, the policy was largely abandoned and the export concentration on tin increased substantially. After 1966 the share of tin in total exports declined because of the growing importance of petroleum and agricultural exports. The cycle of

166 Table 1.5

Juan Antonio Morales and Jeffrey D. Sachs Trade as a Percentage of Bolivia’s GDP, 1956-85’ Period 1956-60 1961-65 I966 -70 I971-75 1976-80 1981-85

Exports

Imports

Total

17.7 18.5 20.7 26.8 20.7 15.5

24.6 25.0 24.0 22.5 23.4 13.7

42.3 43.4 44.7 49.3 44.1

29.2

Sources: For 1956-80, Ceneal Bank of Bolivia, Boletin Estadistico (La Paz), various issues; for 1981-85, unpublished data from the World Bank.

“Annual averages. Percentages are derived from National Accounts data in current prices.

decreasing dependency on tin and natural gas ended around 1977. During the crisis years of 1982-85, natural gas became the most important export. In 1985 natural gas and tin constituted almost 90 percent of all exports. The performance of exports was rather dismal from 1952 to 1985, as can be observed in table 1.6, but again very definite cycles of expansion and contraction can be distinguished. Between 1952 and 1961, the average annual rate of export growth was negative. In the quinquennium 1962-67 there was a healthy recovery that then slowed down between 1967 and 1971. The big boost in the dollar value of exports occurred between 1972 and 1977 because of higher prices, but in volume terms the growth of exports was negative during 1970-75 and very low during 1976-80. It is notable that despite the uniformly good real prices for exports during most of the 1970s, the output response was weak. The year 1974 has been called by some observers a “miraculous” one for Bolivian exports. The index of export prices almost doubled from its level of the previous year. In the export basket, the price of tin increased 70 percent in nominal terms and 40 percent in real terms. Tin prices would remain high with a slight downward trend until the collapse of the market in 1985 (figure 1.2). Imports generally followed the growth path of exports, though in the 1980s, imports fell sharply as export earnings were devoted increasingly to debt servicing. Transportation and insurance costs loom large in the CIF value of imports, given the geography of Bolivia and its rudimentary transportation network. Transportation costs are also a very important determinant of the structure of Bolivian exports. As mentioned earlier, because of the high costs of transportation, exports have to be of high value added per unit of weight.

1.1.3 Capital Formation and the Sources of Funds The share of fixed investment in GDP varied widely during 1952-85, as is shown in table 1.7. The long-run value of this rate seems to be, however, above 15 percent. In the two years following the Revolution of 1952, the investment rates fell below this long-run value, but there was a substantial recovery in 1955 and 1956 despite extremely high inflation. In fact, contrary to what occurred during the 1980s, forced savings (i.e., the inflation tax) helped to sustain a high investment rate.

167

BolividChapter 1

Tnble 1.6

Annual Rates of Growth of Main Export (volumes), 1950-85 Period

Tin

Natural Gas"

1950-55

-2.3

-

1955-60 1960-65 1965-70

-7.0 4.2 2.8 - 2.2 - 2.0 -6.4

1970-75

1975-80 1980-85

Total Exports - .3

-

-6.1 3.0 9.4 -5.5

5.6 1.7

-2.9

~

-

1.5

Source: Data are export volumes reported in the IMF, Inrernarional Financial Staristics (Washington, DC), various issues.

Nore; Growth rates calculated as geometric averages of period shown. aNatural gas exports began in 1972.

In the 1960s, investment rates were very high, comparable to those of the high-growth period of Brazil. However, these rates were sustained to a very significant extent by the contribution of foreign savings. Between 1975 and 1978, there was another upsurge in the investment rate, sustained similarly by foreign savings, leading to increased indebtedness. The investment rate fell in 1981 in line with the increasing difficulties in obtaining foreign financing. From 1982 onward, the investment rate fell precipitously as disbursements of foreign credits almost disappeared. As shown in table 1.8, the contribution of domestic savings to gross capital formation has been substantial and should not be underestimated.

65

I

I

I

I

I

1

68

71

74

77

80

83

REAL TIN PRICE

( 0

1

Fig. 1.2 Relative price of tin in world markets (1980= 1.0)

b

86

168 Table 1.7

Juan Antonio Morales and Jeffrey D. Sachs Investment Rates (percentage of GDP), 1952-85 Year

Investment Rate

Year

Investment Rate

1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968

16.0 10.0 13.0 20.0 20.0 19.0 15.0 15.0

1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985

20.0 14.0 15.0 16.0 15.0 15.0 17.0 17.0

17.0 14.0 21.0 20.0 20.0 20.0 17.0 18.0 25.0

17.0 18.0 17.0 14.5 12.4 9.8 6.6 8.8 6.5

Sources: Ministry of Planning, Revisru de Plunificucion y Desurrollo (La Paz: 1970); Central Bank of Bolivia,

Boletin Esrudistico (La Paz), various issues; and unpublished World Bank data. Note: Investment rates are computed as real investment divided by real GDP.

A high share of foreign savings in total investment was concentrated in a few subperiods: the 1960s, with foreign aid provided by the Alliance for Progress and with important foreign direct investment in the petroleum sector; and the late 1970s, with credits from commercial banks and suppliers. Gross national savings (i.e., domestic savings less net factor income plus current transfers) as a percentage of GNP experienced large fluctuations over the years 1958-85. During most of the 1970s, the saving rate was over 15 percent, and in 1974, it rose to more than 20 percent. From 1978 on, however, the rate began to fall, eventually to very low levels in the years of the economic crisis. A picture emerges in which two features stand out: first, the Bolivian economy can generate respectable national saving rates and, second, these high rates are generally the product of high growth in the economy supported by a favorable terms of trade or by access to foreign loans. The public enterprises played an important role in generating domestic savings during these high-growth years. The domestic financial system, constituted mainly by commercial banks, has played an important role in the formation of private savings. The data in table 1.9 indicate that financial deepening reached its maximum in the mid-1960s. It is interesting to note that the real value of monetary balances (measured as M2) in 1965 was the highest in the entire 1952-85 period and that this peak was obtained without the dollar-indexed savings deposits that were the most common form of deposit after 1975. The evidence seems to indicate that the most important factor that induced savings in the banking system in the 1960s was the low and stable inflation rates of those years.

169

BolividChapter I Sources of Financing for Gross Capital Formation (as a proportion of total),

lsble 1.8

1952-85

Year 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985

Domestic Savings 94.4 79.4 98.9 99.1 98.9 93.2 78.2 83.4 76.0 81.9 75.3 73.4 105.3 78.7 80.7 77.5 66.4 64.2 87.8 79.7 82.5 92.3 132.1 72.1 84.3 76.8 56.3 50.7 67.4 94.7 128.5 84.7 100.3 63.0

Foreign Savings 5.6 20.6 1.1 .9 1.1 6.8 21.8 16.6 24.0 18.1 24.7 26.6 -5.3

21.3 19.3 22.5 33.6 35.8 12.2 20.3 17.5 7.7 - 32.1 27.9 15.7 23.2 43.7 49.3 32.6 5.3 -28.5 15.3 - .3 37.0

Sources: For 1952-79, Central Bank of Bolivia, Cuentas NacioMles no. 4 ; for 1980-82, Departamento de Cuentas Nacionales; for 1983-85, elaborated with data from the World Banks.

1.1.4 Inflation The annual inflation rates during 1952-88 are shown in table 1.10. It will be observed in that table that Bolivia had a bout with high inflation in the wake of the Revolution of 1952, that lasted until the stabilization of 1957. At that time, Bolivian inflation rates were the highest ever known in Latin America. Inflation ended with an IMF-sponsored stabilization plan and with loans and subsidies from the U.S. government to support the budget. Details on that inflationary episode and how it was ended are given by Eder (1968). The 1960s and 1970s were characterized by low inflation, except after the devaluations of late 1972 (and the protracted relative price adjustments that lasted until the first months of 1974) and of late 1979. From 1980 onward

170

Juan Antonio Morales and Jeffrey D. Sachs

Table 1.9

Monetary Variables as a Percentage of GDP, 1961-85 (selected years) Year

Money Base

MI

M2

1961 1965 1969 1973 1977 1981 1985

8.0 10.0 10.6 8.7 10.4 8.6 2.3

8.6 11.8 11.2 9.4 10.4 8.9 2.3

13.7 17.6 14.0 12.5 14.5 15.5 3.5

Source: Elaborated with data from Central Bank of Bolivia, Memorias Anuales (La Paz), various issues. Note: Monetary indicators are annual averages.

lbble 1.10

AMUI Inflatiun Rates, 1954-88 Year

Inflation Rate

Year

Inflation Rate

1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 I966 I967 1968 1969 1970 1971

124.8 77.9 181.2

1972 I973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988

6.5 31.4 62.8 8.0 4.4 8.1 10.4 19.7 47.2 32.1 123.5 275.6 1,282.4 11,857.1 276.3 14.6 16.0

115.5

3.0 20.3 11.5 7.5 5.8 - .7 10.1 2.9 6.9 11.1 5.4 2.2 3.8 3.6

Sources: For 1953-70, Central Bank of Bolivia, Memorim Anuales (1.a Pa-r), varinus issues; for 1971-8I. Instituto Nacional de Estadistica, Resumen Esradistico (La Paz: 1982); for 1982-85, Unidad de Analisis de Politica Economica (UDAPE), Anexo (La Paz); for 1986-88, IMF, international Financial Starisrics. Note: Rates are calculated as year-to-year changes of annual average price levels.

until the fall of 1985, inflation rates were very high, crossing the 100 percent barrier in 1982. Inflation reached hyperinflationary levels in 1984 (i.e., monthly price changes in excess of 50 percent), with the hyperinflation lasting from May 1984 to September 1985. The inflation ended dramatically with the New Economic Policy of the Paz Estenssoro government, which was launched on 29 August 1985. The Bolivian inflations of 1953-57 and 1982-85 had their roots in weak fiscal budgets. Importantly, Bolivia (unlike Brazil, for example) did not have the institutional mechanisms that produced inertial inflation. Wage indexation was implemented only in 1983, when high inflation was well on its way, and

171

BolividChapter 1

lasted until August 1985. The labor unions have had little experience in collective bargaining, and there have been very few formal labor contracts with escalator clauses or even with well-defined periods of validity. The practice in the labor unions has frequently been to ask for wage increases whenever there has been a significant erosion in real wages and to go to the Ministry of Labor for arbitration if the unions do not reach a settlement with the employers. During the last months of the hyperinflation, wages in many private sector firms, especially for qualified labor and in the service sector, were linked to the dollar in informal arrangements that were not legally binding. In the high-inflation episode of 1953-57, the fiscal troubles were the result of deliberate policies to enhance production diversification projects in the eastern region of Santa Cruz with deficit financing, that is, by printing money. Inflation was used to generate forced savings to finance capital expenditures in a growing public sector. In the more recent experience of the early eighties, the replacement of foreign financing by internal financing to cover prolonged fiscal deficits was the most important explanatory factor of high inflation. 1.1.5 The Expansion of the Coca Economy An important question that arises when studying the Bolivian economy of the late 1970s and early 1980s concerns the role played by coca production and exports of processed coca leaves in the form of cocaine paste or refined cocaine. After 1978, illegal cocaine exports became a very important source of foreign exchange. A booming underground economy developed around the coca trade, with far-reaching consequences for the legal sectors of the Bolivian economy and for the international relations of Bolivia. Until around 1975, coca leaves were marketed essentially for legal domestic uses.5 The sharp increase in demand for cocaine (and other drugs) in the United States that started in the 1970s and emerged later in other industrial countries strongly increased the demand for coca leaves. Production responded rapidly to the demand shift. Coca paste was transported from Bolivia to Colombia, refined there, and marketed in the United States by Colombians. By 1980 some Bolivians started to refine the paste and market it without intermediaries in the U.S. market. Bolivia’s involvement in the cocaine business increased rapidly and took several forms: production of coca leaves, coca paste, and refined cocaine. It is very hard to come up with precise figures concerning Bolivia’s production and exports of coca-based products given the illegality of the activity, but preliminary estimates of the FOB value of cocaine chlorohydrate have been computed by Dona Medina (1986) using data on acreage prepared by the Earth Satellite Corporation and information on the technology of deriving cocaine from coca leaves. The figures, in 1980 dollars, are 412.2 million for 1980, 563.9 for 1981, and 758.5 for 1982. In 1980 this represented 40 percent of recorded exports of goods and services; in 1981, 56 percent; and in 1982, 83 percent. These estimates are fairly conservative.

172

Juan Antonio Morales and Jeffrey D. Sachs

If export values are difficult to estimate, gross value added in the cocaine industry (and related activities) are even harder to come by. Under the rather arbitrary assumption that material inputs other than coca leaves are 20 percent of gross output in cocaine chlorohydrates, the gross value added estimates would represent 5.7 percent of GDP in 1980; 7.5 percent of GDP in 1981; and 10.8 percent in 1982. If one considers cocaine and coca production as vertically integrated industries, these percentages increase to 7.3, 8.8, and 12.7, respectively. Undoubtedly, these figures are very high, but significantly less than the wild estimates sometimes given in the press.

1.2 Summary of the Monograph With this basic introduction to the Bolivian economy, we now turn to a summary of the remainder of the monograph. The outline of the work is as follows. Chapter 2 provides an overview of the political economy of macroeconomic policymaking in Bolivia since the 1952 Revolution. Great stress is put on the weakness of fiscal institutions in the face of heavy social and sectoral demands. The chronic fiscal weakness led to a remarkably overextended public sector, one that reached effective bankruptcy by 1985. During the period of hyperinflation, 1984-85, the government paid its way almost entirely through money printing. Chapter 3 highlights some of the main directions of development policy during 1952- 85, especially involving public investment spending and trade policy. We show that most of the period is characterized by a development strategy of “state capitalism,” in which the public sector is called upon to be the leading engine of growth through public sector investment spending and overall guidance of the private sector. The paradoxical result, however, is that as the government tried to control too much, it ended up controlling very little indeed. The model of state capitalism was particularly undermined by the chronic instability in policymaking that resulted from the rapid turnover of governments. In chapter 4 we consider important characteristics of Bolivia’s international trade, focusing bo&hon structural features (e.g., the heavy dependence on a small number of primary commodities), as well as policy choices. Our analysis of trade policies since 1952 makes clear that Bolivian governments continued to focus on primary commodity exports as the key to Bolivian growth, despite the secular decline in tin and the mixed fortunes regarding petroleum exports.6 Bolivian governments typically spent more time and energy on attempting to improve the terms of trade for tin (mainly through the International Tin Agreement, which collapsed in 1985) than on a diversification of the export base into manufacturing. Import policies have generally been ill-focused, leading to an enormous range of effective protection across sectors, with little discernible long-term

173

BolividChapter 1

strategy in the distribution of incentives and with little attention to the indirect effects of import policies on the profitability of exports. Similarly, exchange rate policies were often pursued for fiscal purposes (e.g., overvalued exchange rates allowed the central government to appropriate some of the rents from exportoriented state enterprises that were otherwise hard to tax), rather than for the purpose of stimulating nontraditional exports. Chapter 5 describes the process of foreign debt accumulation, which was the counterpart of the large budget deficits of the public sector in the 1970s and early 1980s. Bolivia’s external debt, particularly from the private financial markets, rose dramatically in the 1970s, mostly during the Banzer era. Bolivia’s rapid accumulation of external debt in the 1970s reflected three forces at work. First, part of the foreign borrowing financed a plausible attempt to generate a more diversified export base through various investment projects. Some of these projects were successful, but more often the investments proved to be expensive failures. Second, the foreign borrowing also reflected an attempt to finesse internal distributional conflicts by using borrowed money to pay for social demands for spending. Third, some of the foreign borrowing had the purpose and effect of enriching a narrow set of private interests through the public sector’s access to foreign loans. Similarly, an overvalued exchange rate, maintained through foreign borrowing, became a channel for capital flight for wealthier individuals. By the late 1970s, the precariousness of Bolivia’s financial situation was clear. Banks pulled back from new lending, and in 1980 Bolivia began to miss payments due to the banks. Note that this occurred almost two years ahead of the global debt crisis, usually dated to Mexico’s suspension of principal payments in August 1982. Bolivia signed a rescheduling agreement with the banks in April 1981, which fell apart when Bolivia missed payments due in September 1982. Payments were restarted in 1983, but then suspended again in mid-1984. Since then, Bolivia has made no interest payments to the commercial banks. Chapter 6 lays out the dynamics of the hyperinflation during 1982-85, focusing on the complex causal links among the budget deficit, the money supply, the exchange rate, and the price level. President Hernhn Siles Suazo therefore inherited an extraordinarily difficult financial position in October 1982. In the event, the Siles government was too weak and often too confused in its policies to restore fiscal order. Siles tried several times to implement a comprehensive stabilization program (during November 1982, November 1983, April 1984, August 1984, November 1984, and February 1985), but in each of these cases, some combination of Siles’ political opponents in Congress and his ostensible “allies” in organized labor torpedoed the adjustment efforts. As the inflation accelerated, tax collections fell sharply in real terms, thereby exacerbating the fiscal crisis. The inflation under Siles went from annual rates of several hundred percent in 1982 and

174

Juan Antonio Morales and Jeffrey D. Sachs

1983 to several thousand percent in 1984 and 1985. The hyperinflation began in May 1984 and ended in September 1985, upon the accession to power of the Paz Estenssoro government. In chapter 7 we detail the process of stabilization since 1985 and discuss some of the general lessons about ending high inflation that might be applied to other economies in the region. The successful stabilization program was carried out by the newly elected centrist government of Paz Estenssoro. The economic program, which was unveiled just three weeks after the government came to office, was remarkably wide ranging and indeed radical, encompassing not only plans for macroeconomic stabilization but also for trade liberalization, administrative and tax reform, and deregulation and privatization in the domestic market. The so-called New Economic Policy constituted nothing less than a program to dismantle the system of state capitalism that had prevailed over the previous thirty years. Chapter 8 describes the novel arrangements that Bolivia has negotiated in order to escape the severe overhang of external debt. Bolivia took one “heterodox” step in its stabilization program: it maintained a total suspension of interest payments on its commercial bank debt, pending a longer-term settlement on the debt. For this reason, the Bolivian program might be characterized as “internal orthodoxy, external heterodoxy.” The debt suspension was an integral piece of the political and economic strategy of the government. The Bolivian debt strategy paid off. After years of negotiations, Bolivia’s creditor banks ultimately accepted the argument that the debt should be reduced. Bolivia won the right to buy back part of its debt at the secondary market price, under the condition that it use money donated by donor governments to pay for the buyback. In May 1988, Bolivia repurchased $308 million in principal (approximately 47 percent of the outstanding debt) from the commercial bank creditors, at a price of about 11 cents on the dollar. Bolivia continued to repurchase debt at 11 percent of par and by mid-1989 had repurchased in total about 70 percent of the outstanding debt. In the concluding chapter 9, we discuss briefly the challenges facing Bolivia in the future, once stabilization has been accomplished. The key fact that we focus on is the weakness of Bolivia’s traditional export base and the consequent need for the development of new sources of export earnings. Bolivia will have to diversify beyond mineral exports to new agricultural exports and light manufacturing. The latter could be spurred importantly by an opening of the Brazilian market to Bolivian light manufacturing exports. We stress that accomplishing this transition will require careful and sensitive attention to regional and social conflicts which remain powerful in the country. The government will probably have to contemplate specific measures to address these conflicts so that the structural transformation can proceed with internal political stability.

175

2

BolividChapter 2

Political Economy and Macroeconomic Policymaking, 1952-87

In this chapter we review the major trends in Bolivian growth and development between 1952 and 1987. We have chosen 1952, the year of the Bolivian Revolution, as the relevant starting year for the analysis because the changes wrought by the Revolution shaped Bolivian economic and political development for the next thirty-five years. The review in this chapter will help to establish the links between foreign borrowing and the main macroeconomic developments. The developments in the export markets (and the attendant changes in foreign indebtedness) could have been used to define the main temporal phases of the economy. We have chosen instead to divide the 1952-87 period using the criteria of political economy, according to the degree and character of state intervention in the economy. Given that this period encompasses several subperiods of intense political turmoil that seriously affected economic policymaking, the issues of state intervention in the economy and political instability receive more attention in our analysis than is perhaps typical of a country study. In view of this emphasis on the political economy of Bolivian economic development, we begin this chapter with a detailed account of the major political developments since the Revolution of 1952, with a stress on their effects on the macroeconomy. Then, we turn to a summary of the main dimensions of economic performance, including growth, inflation, trade, and debt management, in the context of the political-economic history.

2.1 Bolivian Political Economy, 1952-87 This period witnessed a variety of policy phases that clearly deserve separate treatment. While the fundamental strategy of development set after the Revolution of 1952 did not change substantially until the end of 1985, the degree and the nature of state intervention in the economy differed significantly among the phases. Economic policy was shaped as much by political considerations as by macroeconomic objectives. Noneconomic objectives were very much present, and political turmoil itself left a deep mark on the period. Table 2.1 gives a brief political chronology of Bolivia since the Revolution of 1952. The most striking aspect of the chronology is the instability of constitutional rule in the country. The Revolution was led by the National Revolutionary Movement (the MNR), which governed Bolivia

176 Table 2.1

Juan Antonio Morales and Jeffrey D. Sachs Political Chronology of Bolivia, 1952-89

Year 1952

Political Event Bolivian Revolution, camied out by Nationalist Revolutionary Movement (MNR), under the leadership of Dr. Victor Paz Estenssoro.

1952-56 Presidency of Paz Estenssoro; important reforms, but sharp rise in inflation, reaching 178.8% in 1956. 1956-60 Presidency of Hernan Siles Suazo (MNR); economic stabilization under U.S. and IMF supervision and finance. 1960-64 Second presidency of Paz Estenssoro. 1964

Third presidency of Paz Estenssoro; Paz deposed in military coup led by General Rene Banientos Ortuno.

1964-66 Co-presidency of Alfred0 Ovando Candia and Banientos Ortuno. I966 -69 Civilian presidency of Banientos (dies in helicopter crash, April 1969). 1969

Vice President Luis Siles Salinas becomes president; Siles Salinas deposed in coup by Ovando Candia.

1970

Ovando Candia deposed by General Miranda; Miranda deposed by General Juan JosC Torres.

1971

Torres rules left-wing radical government; Torres deposed in coup, jointly sponsored by the military party Bolivian Socialist Phalange (FSB) and the MNR.

1971-73 General Hugo Banzer Suarez rules with MNR and FSB support. 1974-78 Banzer presidency under military rule; MNR and FSB withdrawal from government in 1974. 1979

General Juan Pereda becomes president in election marked by accusations of fraud; Pereda deposed by General David Padilla, who calls for 1979 election.

1919

Election results in stalemate (no majority); Senate President Walter Guevara Arze serves as interim president; Guevara deposed by Colonel Natusch Busch; Natusch Busch resigns in 15 days; president of Chamber of Deputies, Lidia Gueiler, becomes interim president.

1980

Electoral stalemate; Gueiler deposed in coup by General Luis Garcia Meza.

1981

Garcia Meza forced to resign in favor of General Bemal; Bemal resigns in favor of General Celso Torrelio.

1982

Torrelio deposed in coup by General Guido Vildoso; Congress reconvenes and names Siles Suazo as president.

1985

Siles Sumo announces early elections; Paz Estenssoro becomes president; new economic policy declared on 29 August 1985.

1989

Paz Zamora of the Movimiento Izquierda Revolucionario (MIR) elected president; governs in coalition with the Acci6n Democratica Nacionalista (ADN) party led by General Banzer.

for twelve years, winning elections in 1956 and 1960. The military toppled the civilian regime in 1964 and ruled without interruption until 1978. A period of political chaos followed during 1978-82, with a rapid alternation of military and civilian rule. Civilian rule was restored in 1982, with the accession to power of Siles Suazo, and then was continued with the presidency of Dr. Victor Paz Estenssoro, which began in August 1985. As in other Latin American nations, intense conflicts over income distribution contributed to a chronic alternation of power between populist

177

BolividChapter 2

and anti-populist politics. In Bolivia, however, this alternation has been particularly sharp. Generally, the military rule was anti-populist and especially anti-labor. There was, however, a brief populist phase of military rule under General Ovando and then General Torres in 1969-71, which coincided with the leftist military government in Peru under Juan Velasco. The civilian governments have all drawn their leaders from the MNR, though these governments have varied widely in their policies, with populist phases during 1952-56 and 1982-85, and conservative phases during 1956-64 and 1985-89. The leaders of the Revolution of 1952 drew inspiration from the Mexican Revolution and from Mexico’s Institutional Revolutionary Party (the PRI) in the formation and policies of the MNR. The experience under the Rosca (the disparaging name given to the pre-Revolution oligarchy) had thoroughly discredited private wealthholders as a class capable of leading national development. The leaders of the Revolution looked to the public sector as an engine of growth that would be more broadly based and equitable. They put in place an economic system which can broadly be called state capitalism, which assigned the bulk of capital formation to the public sector, both for infrastructure and for industrial production in state enterprises. The leading state enterprises were COMIBOL, the national mining company, and YPFB, the state petroleum company. The whole concept of more equitable growth through a large state sector collapsed in a mass of inconsistencies over the thirty years between the Revolution and the onset of the Bolivian hyperinflation. During the post-Revolution period, leftist leaders were always too weak politically to satisfy their distributional aims. To the extent, for instance, that they aimed to raise public sector salaries or to increase public sector investment and employment, they lacked the capacity to tax income and wealth, which was essential to finance the larger state sector. If necessary, the Bolivian army was prepared to intervene to forestall populist or redistributionist actions, as it did in 1971. For these reasons, leftist or populist leaders have been constantly forced to rely on inflationary finance or foreign aid and foreign borrowing to carry out their distributional and developmental goals. The first high inflation in Bolivia came in the wake of the Revolution, and the second came with the left-wing government of Tiles Suazo in 1982. Leaders on the right, such as Barrientos and Banzer, were not interested to limit the power of the state, but rather to use the state to satisfy their own agenda. While governments on the left sought redistribution through higher wages and a larger role for public sector workers, governments on the right often sought to bolster favored segments of the private sector through generous government subsidies. Governments of the left have generally paid for higher public salaries through printing money (i.e., the inflation tax) or through foreign borrowing, since they have been forestalled from raising taxes. Governments of the right, on the other hand, have rejected higher

178

Juan Antonio Morales and Jeffrey D. Sachs

taxes outright and have instead sought to finance the government through a reduction of public sector wages (often with overt repression of labor) and also through foreign borrowing. Note one common theme of both types of governments: let the foreigners pay! We now turn to a more detailed look at several key subperiods in Bolivia’s political history since the Revolution. In doing so, we use a taxonomy that borrows heavily upon Malloy and Borzutsky (1982) and Malloy and Gamarra ( 1988).

2.2 The Revolutionary Period, 1952-56 The Revolution of 9 April 1952 that swept the Movimiento Nacionalista Revolucionario (MNR) into power, with Dr. Victor Paz Estenssoro at the head, brought profound changes to the Bolivian economy and political structure.’ Four important economic changes should be highlighted: (1) the three largest mining concerns were nationalized; (2) a very broad land reform program was initiated; (3) import substitution and export diversification became official policies; (4) a march to the hitherto underdeveloped eastern lowlands was begun through various government-sponsored colonization projects.* The nationalization of mines was a natural outcome of a process of greater taxation and control of the private mining enterprises that had been underway for decades. Before nationalization, taxation occurred directly but also indirectly, through manipulations of the export exchange rate. With increased taxation and controls, the mining enterprises sought ways, legal and illegal, to evade them. As the government exhausted its taxing power, it was natural for the issue of nationalization to arise. An occupation of the mines by their workers in 1952 pushed the MNR leadership to undertake the nationalization. The MNR leaders, who were by and large moderate reformers with middle-class roots rather than extreme revolutionaries, thought that government control of the mines, as urged by the miners, was better than workers’ controversy. Land reform also went beyond the intentions of most MNR leaders and again was a policy imposed on them by the squatting by landless peasants. Export diversification, import substitution, and the march to the East were, however, deliberate development policies. Two instruments had an important role in those policies. First, through overtaxation of the mining sector with differential rates of exchange, resources were transferred to the new activities, especially in the East. Second, many public sector investment projects were financed with forced savings, that is, the inflation tax. A point of some importance about the Revolution should be noted. In spite of the official rhetoric of “anti-imperialism,” the government was generally supportive of foreign capital, as long as it was not from the Tin Barons. After 1955, following changes in the legislation governing the

179

BolividChapter 2

petroleum sector, Bolivia received significant amounts of foreign direct investment in petroleum. Substantial American mining interests belonging to W. Grace Co. were never threatened with nationalization. It should also be mentioned that, to the surprise of the neighboring countries, a close cooperation with the United States developed during the revolutionary period. At the same time that the MNR was undertaking development projects, it tried to satisfy the demand of its support groups for increased consumption. There was a clear inconsistency between these two claims that exacerbated the distributive problems. Malloy and Borzutsky (1982, 46) correctly point

out: The Bolivian revolution quickly demonstrated some inescapable facts of the reality of the situation of economic backwardness and structural dependency, namely that (a) a relatively backward country cannot follow a simultaneous policy of economic development and popular consumption and (b) any process of restructuring for development demands that some social groups pay the “cost of the new course.” 2.3 Stabilization and the Beginnings of State Capitalism, 1957-64

The revolutionary process entered into economic chaos by mid-1956. The period was marked by high inflation rates, foreign exchange rationing, a rapidly rising black-market premium for foreign exchange, shortages of food and other essentials, stagnant output in the mines and in traditional agriculture, and a decline in tin prices following the end of the Korean War boom. The deteriorating situation clearly required a drastic stabilization program. Such a program was announced on 15 December 1956. The program shares many elements with the one enacted almost thirty years later in August 1985: exchange rate unification, a sharp reduction in public expenditures (around 40 percent in real terms), complete price liberalization and elimination of all subsidies, and a wage freeze after an initial limited rise in compensation. All policies directed toward diversification in production and exports came to a halt with the anti-inflationary package of 1956. A shortage of funds in the public sector and the strictures of an IMF agreement forced a sharp cut in government investment. A planning board for state sector investment was created, but its role was very modest. The stabilization effort was largely sustained with donations from the United States, using food aid (Public Law 480 funds) and direct budgetary support. The world recession of 1957 and a substantial fall in tin prices in that year made things even more complicated and increased dependency on foreign aid. In exchange for U.S. support, the government of Bolivia committed to resume payments on foreign bonds that had been defaulted on in 1931 (see Ugarteche 1986; Baptista 1985). Notwithstanding the commitment, Bolivia defaulted again on this debt in 1960.

180

Juan Antonio Morales and Jeffrey D. Sachs

In regard to development policies, the government tried to rehabilitate COMIBOL, the state mining enterprise. The other two principal state enterprises, YPFB (oil) and the Development Corporation of Bolivia (CBF), were at that time relatively well-run enterprises that could stand on their own. COMIBOL, on the other hand, had been left completely decapitalized by the government management during 1952-56 and suffered from many ailments ranging from inadequate machinery to labor indiscipline. In 1961 the Triangular Plan was established with funds from the U.S. government, the newly created Inter-American Development Bank (IDB), and a consortium of West German mining firms. The plan called for extensive mechanization of the mines, the dismissal of several thousand workers, and the elimination of worker control of the activities of COMIBOL. Not surprisingly, the plan met strong labor opposition. It is interesting to note a significant change in declared export policy in this second phase, compared with the policies immediately after the Revolution. Mining again received high priority, and hopes for export diversification were reduced to the undertakings of the foreign oil companies. Active policies of import substitution were also largely forsaken. The second phase coincided with the beginning of the Alliance for Progress, a development program for Latin America created by President John Kennedy. The Bolivian government grasped right away the opportunities offered by this new program and became one of the main beneficiaries. The government under took extensive infrastructure investment, financed mostly with Alliance for Progress funds. By 1964 the remnants of the Revolution, including heightened political mobilization, worker participation in the management of public enterprises, and redistributive policies, had been all but abandoned. By the early 1960s, a form of state capitalism developed, controlled and exploited by various competing groups of the middle classes. As we shall note, the state enterprises became a source of enrichment for these private factions, some civilian and some military.

2.4 The Resurgence of the Private Sector, 1964-65 In 1964 President Paz Estenssoro, who was at the beginning of his third term, was deposed in a coup d’Ctat staged by his vice president, General RenC Barrientos Ortuno. The coup itself reflects the collapse of unity of the governing MNR, which had ruled since the Revolution. The party had by 1964 splintered into a myriad of factions and personalistic cliques, all organized to extract patronage and other spoils from the Treasury and from the state enterprises. When Paz attempted to maintain power for a third administration, many of the splinter groups of the MNR, which felt deprived of their share of the spoils, coalesced in support of the coup.

181

BolividChapter 2

General Barrientos continued the big push to modernize the public sector which had begun in Paz’s second term (1960-64). Significant effort and money were spent on the public sector, much of it financed by the Alliance for Progress. The domestic private sector, which had played a limited role in the economic developments of the previous twelve years, resurged as a result of two strong impulses. First, fresh investment funds were made available, at subsidized interest rates and with a high grant element, to private producers. A new group of producers formed a nascent entrepreneurial class that began to play an important political role in the late 1970s and early 1980s. The investment funds came mostly from international aid, with IDB and the World Bank being the principal sources. Second, many managers and middle-class professionals who had left the country during the MNR reign returned and engaged themselves in industrial and mining activities. As a result of this surge in public and private activity, economic growth was very high, reaching 6.6 percent during 1965-69. Foreign direct investment flowed into the mining and petroleum sectors. The expansion of petroleum production and exports was largely due to the investments of the Bolivian Gulf Oil Corporation. Very important natural gas deposits were discovered, and the construction of a gas pipeline to Argentina was started. At the same time, investments made in Santa Cruz, in the eastern portion of the country, during the MNR period (1952-64) started to pay off during the Barrientos era. Commercial agriculture increased in output very significantly. More importantly, economic development meant a larger political influence for an emerging class of Santa Cruz businessmen. It should be underlined that while foreign borrowing was high during the sixties, investment rates were also very high. Unlike during the 1970s, foreign borrowing indeed translated into higher domestic investment, and without a subsequent debt crisis. The nature of the credit sources in the 1960s might help to explain this fact. Official creditors then were both more careful in screening their loans and more generous in their terms than the banks proved to be in the 1970s. Bolivia joined the Latin American Free Trade Association (LATFA) in 1966 and in 1969 signed the Cartagena Agreement, which created the economic integration scheme of the Andean Pact (to be discussed in detail in chap. 4). The Bolivian government at that time held considerable expectations of the benefits of this association, expectations that failed to materialize.

2.5 The Populist Interlude of 1969-71 Barrientos was killed in a helicopter crash in April 1969, and the civilian vice president, Luis Adolfo Siles Salinas, assumed the presidency. In September 1969 he was overthrown in a coup led by General Alfred0 Ovando Candia, who in turn was overthrown by General Juan JosC Torres in 1970. Ovando and Torres represented the phenomenon of a left-wing,

182

Juan Antonio Morales and Jeffrey D. Sachs

populist military regime, which mirrored the political developments in Peru at the same time (with the ascension to power of General Juan Velasco). During the presidency of Ovando Candia, three important economic policy decisions were made: (1) the Bolivian Gulf Oil holdings were nationalized; (2) very costly, and ultimately uneconomical, tin smelters were built; and (3) the very influential Socioeconomic Strategy for National Development (SSND) was prepared. Even today, the reasons for the nationalization of Bolivian Gulf Oil are not very clear, except as a populist measure to gamer political support. On the other hand, the construction of the tin smelters was a response to a long-held Bolivian aspiration: the exportation of metallic tin instead of ores. This could indeed have been profitable to the country by saving heavy costs of transportation and charges for processing in foreign smelters, but unfortunately the tin smelters were built very ineffi~iently.~ The SSND will be discussed in chapter 3.

2.6 The Banzer Era, 1971-78 A coup d’Ctat in August 1971 installed General Hugo Banzer in the presidency. He was supported in the coup by civilians and members of the military discontent with the left-wing course of the Ovando and Torres regimes and afraid of emerging guerrilla threat^.^ Banzer’s supporting civilian force reunited the MNR of former president Paz Estenssoro with the far-right party Bolivian Socialist Phalange (FSB). In a turnaround, not uncommon in Bolivia, some party members of the Marxist-Leninist Party of the Revolutionary Left (PIR) entered government. In the military, commanders of the anti-guerrilla units had acquired the upper hand. Banzer’s government lasted seven years. From the end of 1971 until the first half of 1974, he governed with the political parties mentioned above; thereafter he relied essentially on the military and some technocrats and bu~inessmen.~ During his administration, Banzer faced strong opposition from organized labor, the leftist parties, and the intellectual elites; but he enjoyed, owing in part to great improvements in the economy, a high degree of popularity among the middle classes, small businessmen, and even in some favored “unions,” such as that of the railroad workers. Some peasant organizations also gave support to the regime, though on the whole, the peasant community was not favored by the Banzer regime. The emergent Confederation of Private Enterprises (CEPB) generally supported Banzer’s economic policies. Prominent businessmen held the key position of Minister of Finance during most of his term. It is important to realize that support and opposition to Banzer were established also along regional lines. Santa Cruz and, more generally, the eastern part of the country, constituted his strongest political base. Santa Cruz was not only Banzer’s home province, but the region also greatly

183

BolividChapter 2

benefited from his economic policies, including various rent-grabbing opportunities made available by the government. Economic growth had the highest priority in Banzer’s economic policies. Real GDP grew at an average 5.4 percent during his term. But the apparent prosperity had several profound economic weaknesses, a point that was to become painfully evident in the 1980s. First, much of the prosperity in the 1970s was based on a temporary commodity boom and heavy capital inflows. Once the inflows slowed, as was inevitable, the economy faced a sustained period of austerity. Moreover, a persistently overvalued exchange rate during the 1970s prompted investment in the nontradable sectors of the economy and in capital flight, so that there was little basis laid for a more dynamic export sector in the 1980s that would be necessary to service the debts accumulated during the Banzer years. Second, Banzer pushed growth in part by granting large subsidies to the vested interest groups that supported the regime. Friends of the government, particularly in the military and among the private business community, were frequently favored with property rights over hitherto public lands, with mining concessions, and, most importantly, with highly subsidized credits. The rationale of these policies was that above-normal profits, if saved and reinvested in the country, could be conducive to the desired high rates of growth.

2.7 The Years of Political Instability, 1978-82 There was another profound weakness of the Banzer years: the absence of a legitimate succession. In 1977 and 1978, General Banzer faced growing pressure from the Carter administration for a return to democracy. An election was called, and Banzer’s handpicked successor, General Juan Pereda, was named president in an election marked by massive fraud. When the results of the election were challenged, Banzer decided to remain in office, but was quickly ousted by Pereda himself. There ensued a period of intense political instability that brought to power no less than eleven heads of state between 1978 and 1982. There were several stalemated elections, which produced interim civilian presidents, and several coups. Just a listing of the heads of state captures the political instability of the period: 1979: Pereda (military), Padilla (military), Guevara (civilian), Natusch (military), Gueiler (civilian); 1980: Gueiler (civilian), Garcia Meza (military); 1981: Garcia Meza (military), Bernal (military), Torrelio (military); 1982: Torrelio (military), Vildoso (military), Siles Suazo (civilian). The events of 1979 deserve some notice. After two brief military governments, new elections were held in the summer of 1979. These elections ended in stalemate, with no candidate receiving 50 percent of the

184

Juan Antonio Morales and Jeffrey D. Sachs

popular vote and no candidate being able to command a majority in Congress. Walter Guevara, then president of the Senate, was chosen as interim president. He was quickly overthrown in a violent coup by Colonel Natusch Busch, who in turn was forced to resign shortly after. Another interim president came to power, Lidia Gueiler, who was then president of the Chamber of Deputies. She in turn was overthrown the following year, but not before she had launched a tough, coherent stabilization plan that, alas, did not have time to take effect. An attempt to elect a president failed in 1980 when the elections again ended in stalemate. At that point, the country reached its political nadir, with General Luis Garcia Meza deposing President Gueiler. The Garcia Meza regime was deeply implicated in the burgeoning cocaine industry in the country and, therefore, never received U. S. diplomatic recognition. The regime was nearly universally condemned for massive corruption and violence and was internationally isolated, except from the Argentinian military government which had helped in the coup and provided some financing, leaving behind debts that haunted Bolivia until the late 1980s. Capital flight reached new heights in the period, with errors and omissions in the balance of payments of 1980 and 1981 totalling $590 million, or about 10 percent of 1980 GNP, a remarkably high amount that probably understates the full extent of capital flight. The international commercial banks stopped all lending and negotiated an emergency rescheduling agreement in 1980, upon which the regime soon defaulted. The diplomatic isolation also had important financial consequences: the multilateral creditor organizations, including the IMF, the World Bank, and the Inter-American Development Bank, withdrew their support and new lending, leaving the country without any effective access to world capital markets. Bolivia started in 1980 to build up arrears with the commercial banks, skipping amortization payments. At the same time, the economy was feeling the stress of higher interest rates on the foreign debt. The need to normalize relations with the international commercial banks led to the debt rescheduling agreement of April 1981, but even with debt rescheduling, Bolivia again fell into arrears in 1982. Short-term Argentinian loans and some swaps provided by Latin American central banks allowed the Garcia Meza government to overcome the most pressing liquidity problems and provided some transitory alleviation of the financial squeeze. The government was also helped by a short-lived export boom in 1980 with important price increases for many nontraditional exports. In addition, Bolivia was able to negotiate better prices on its sales of natural gas to Argentina. Garcia Meza met enormous difficulties in his exercise of government and faced strong opposition even among members of the military who had conspired with him. His brutal ways alienated him from the business organizations and middle-class constituencies that had originally supported

185

BolividChapter 2

his coup. By mid-1981, it became clear that he had to step down. In August 1981, he resigned to a junta of commanders of the Armed Forces and, after a short interval, was replaced by General Celso Torrelio. Torrelio’s presidency coincided with the onset of the world recession and of the Bolivian economic crisis. Unable to control the economy, he was deposed in favor of General Guido Vildoso in the second half of 1982. Vildoso lasted only a few weeks, but in the meanwhile he prepared a stabilization plan, trying to make it palatable both to the IMF and to the public. Political fatigue owing to the deterioration of the economy and, especially, the instability of the military governments forced him to reconvene the Congress elected in 1980, which, in turn, acting as an electoral college, named the civilian Heman Siles Suazo as president. The political chaos between 1978 and the end of 1982 had a paralyzing effect on the economy. The uncertainties that arose from this situation delayed recognition of the external disturbances that the national economy faced and obstructed the process of decisionmaking needed to take appropriate action. Political antagonists attributed the effects of external shocks to their political foes, instead of looking to the true causes. Quite unfortunately for Bolivia, the internal political chaos overlapped with the onset of high world interest rates and world recession during 1980-82. Had opportune action been taken to redress the external imbalances, it is likely that the toll exacted on Bolivia by the world crisis would have been lighter. Bolivia remained largely isolated from the ongoing discussion in international academic and official circles about the way to cope with the crisis. Some of the macroeconomic mistakes that were made can be attributed to this isolation. The courageous attempt at stabilization of President Lidia Gueiler in November 1979 stands out as an exception. Her stabilization plan contained the following elements: (1) a 20 percent devaluation of the peso; (2) increases in prices of publicly provided goods and services (e.g., a 36 percent rise in gasoline prices, and a 30 percent increase in electricity rates and railroad fares); (3) a reduction in subsidies on imported staples; (4) a reduction of export taxes that had been created in 1972; and (5) strong regulations on dollar-denominated time and savings deposits in the domestic banking system. The package was completed with a small wage increase. In January 1980 an Extended Fund Facility loan was agreed upon with the IMF for U.S. $66.4 million, and in June of that year a Structural Adjustment Loan was obtained from the World Bank for U.S. $50 million. Both of these programs subsequently fell into abeyance, after the coup which brought Garcia Meza to power. It is worth noting that the agreement with the IMF was reached very rapidly after Gueiler’s stabilization plan was announced. This contrasts sharply with what happened in the ensuing years, when negotiations with the Fund would be protracted for several months.

186

Juan Antonio Morales and Jeffrey D. Sachs

It is difficult to assess the effects of the stabilization plan and the IMF and World Bank loans of those years. The Garcia Meza coup clouded the picture. Had the democratic process continued and the new government followed the economic policies set forth by President Gueiler, results probably would have been better than those observed. 2.8 The Return to Democracy and the Outbreak of Hyperinflation, 1982-85 As in other Latin American countries in the early 1980s, economic deterioration prompted the military to retreat to the barracks. In 1982, Herniin Siles Suazo, who had been the president to preside over the IMF stabilization program of 1957 and who had received the plurality of votes in the stalemated 1980 elections, was called on by a newly reconvened Congress to accept a four-year term of office. It is important to appreciate the political implications of Siles’ accession to power. The new administration represented the first elected government in eighteen years, so that pent-up social and economic aspirations were sure to boil over early in the term. Moreover, Siles represented the left wing of the MNR and governed in a coalition that included the Communist party, the left-wing MIR (Movimiento de Izquierda Revolucionario), and for part of the time, the centrist Christian Democrats. In the early phase of the administration, the union movement gave strong support to the new government, but demanded large real wage increases and other political concessions in return. Ultimately, when successive wage increases did not keep ahead of the accelerating inflation and when various political concessions were not granted, the unions turned sharply against the government. Indeed, labor unrest in 1984 and 1985 killed the final two stabilization attempts of the Siles administration. It is not unusual for inflation to accelerate sharply upon the accession of leftist governments, particularly if the left has been denied power for many years. The same phenomenon occurred, to a lesser extent, in Bolivia in the 1950s, in Chile under Allende (1970-73), and in Portugal in 1974 after the collapse of the right-wing regime in that year. The price explosion is typically due to a rise in government spending in favor of the new government’s key constituencies (e.g., real wage increases for trade union members or increases in social spending) and to the inability of the government to achieve a national consensus on raising taxes to cover such spending (with the right wing blocking tax increases). In the case of Bolivia, though, it was not only that the new government forced through large increases of public spending, but also that it was unable to reduce the huge deficit that it had inherited from earlier governments and was unable to attract noninflationary forms of finance. The coalition members of the Siles government were never able to agree on policies to restrain spending (the

187

BolividChapter 2

Communists, for example, rejected any kind of policy that implied a drop in public sector employment or real wages), while the government’s opponents in the Bolivian Congress opposed all tax reform proposals to stabilize or enlarge the shrinking tax base. Siles was unable to stop the economic decline, and indeed many of his policies (and sometimes his lack of action) aggravated the situation. The inflation rate jumped in the second quarter of 1984 to hyperinflationary levels. The detailed economic analysis of these events is postponed to chapter 4. A complete understanding of what precluded the necessary policy adjustments would take us deeply into the Bolivian political system and, therefore, is beyond the scope of this study. An excellent political analysis of the period is available, in any event, in Malloy and Gamma (1988). A few general points can be raised. First, a key condition for the eventual outbreak of hyperinflation was the cutoff in access to foreign borrowing in 1982: only a credit-constrained government would choose to finance current expenditures with a hyperinflation rather than with more foreign borrowing. Second, interest group politics were crucial in the process. At a fundamental level, an inflation tax is a highly regressive tax that affects a general and poorly organized part of the population, while cuts in government expenditures or increases in other kinds of taxes often affect better organized or more influential interest groups. Third, many wellconnected, rent-seeking individuals made considerable fortunes in the course of the hyperinflation. Anybody with access to official foreign exchange from the Central Bank could become wealthy almost instantly during the period by purchasing cheap dollars at the Central Bank and selling them at a several hundred percent profit in the black market. Similarly, commercial bankers, who took deposits at zero interest rates and lent at high nominal rates, shared in the government’s seignorage gains during the hyperinflation. Moreover, the government extended a large number of low interest loans during the period, which effectively became grants as a result of the inflation. Price controls on public sector goods, such as on petroleum and even flour, generated opportunities for lucrative smuggling operations. All of these opportunities for gain from the price distortions provided a natural constituency of powerful individuals who wanted to see the hyperinflation process continue. Fourth, there was a continuing and sometimes profound misunderstanding of the costs of stabilization. One of the most important checks on the stabilization process was the government’s political inability to impose real wage “decreases,” however transitory, on the union sector. The unions resisted any cuts in measured real wages as part of a stabilization program, despite the fact that the gains in reducing the inflation tax for the workers could have exceeded any direct real income losses from a reduction in public sector pay. The problem is that real wages are measured as a nominal wage,

188

Juan Antonio Morales and Jeffrey D. Sachs

W, deflated by a consumer price index, P, that does not include the price (i.e., opportunity cost) of holding money. Thus, while a rise in oil prices that allows the government to eliminate inflationary finance could well leave real wages net of the inflation tax higher than under the hyperinflation, the policy will almost surely reduce measured W/P, and therefore would typically be resisted by the unions. It should also be noted that much of the union movement’s opposition to the Siles government’s policies was not limited to narrow calculations about short-run movements in real wages, but rather to a broader agenda to push the country to the extreme left. The main union organization, the COB (Central Obrero Boliviana), pursued a strategy of insisting on co-government, with a goal of pushing the left-wing coalition toward a model of state socialism (see Malloy and Gamma 1988 for an extensive discussion of this point). This was much farther than most of the other coalition partners, and the president, were willing to go. Therefore, after failing to convince the government from the inside, the COB took to active and vocal external opposition. 2.9 Stabilization and the End of the State-Capitalism Model,

1985-88

By November 1984, President Siles admitted defeat in his government’s attempts to control inflation, and he called for new elections to be held in July 1985, one year before his term was due to expire. During the last year of the Siles government, inflation reached 24,000 percent (August 1985 compared with August 1984). Former presidents Hugo Banzer and Victor Paz Estenssoro were the front-runners in the national elections of 1985, but neither of them received a majority of votes. While Banzer had a plurality of votes, Paz Estenssoro was voted in by the newly elected Congress on 4 August 1985, with the support of left-wing representatives who opposed Banzer’s attempt to return to power. Three weeks after Paz’s inauguration, on 29 August 1985, the center-right government of Paz Estenssoro unveiled a wide-ranging program, known by its decree number, 21060, which encompassed not only plans for macroeconomic stabilization but also for trade liberalization, administrative and tax reform, and deregulation of domestic markets. The scope of the program is remarkable in that it attempted to address deep and politically sensitive structural issues in the economy at the same time that it battled against the hyperinflation. Hyperinflation was ended shortly after the program was announced. Details of how this was accomplished will be considered in chapter 5. In the set of structural reforms, announced in the program of 29 August 1985 and implemented during the following year, the most important were: (1) a tax reform that simplified a complicated structure, replacing it with one where valued added and wealth taxes loom large; (2) reform, decentraliza-

189

BolividChapter 3

tion, and expenditure cutting in the state enterprises, especially in the mining company COMIBOL; (3) reform in the trade regime, aiming at the elimination of most barriers to international trade and the implementation of a flat tariff schedule; and (4) normalization of Bolivia’s access to international financial flows from the official creditor community (i.e., the IMF, the World Bank, the IDB, and bilateral creditors). The political base for implementing this enormous program was established in a crucial Pact for Democracy signed by Paz’s MNR and Banzer’s party, Acci6n Democratica Nacionalista (ADN). This agreement established a congressional majority sufficient to put through the Congress the key features of Supreme Decree 21060. The pact allowed the decree to take effect in the context of democracy, a remarkable accomplishment in view of Bolivia’s volatile political institutions and its vulnerable position amongst politically unstable neighbors in the region. With the pact, the Paz government reserved important positions in state enterprises for members of Banzer’s ADN party. In addition to the political base of the MNR and ADN, Paz also had strong support in the business community. Several prominent businessmen, some who had been head of the CEPB, were given top Cabinet posts. The structural reforms did away with the state-capitalism model, with its several variants, that had been followed since the Revolution of 1952. The reforms were carried out with a remarkable smoothness. Of course, there was some bitter opposition, particularly in the labor unions and in the parties of the left, as well as lively criticism in the press, but the changes were peaceful and in the framework of democratic institutions. At no point during Paz’s administration was the process of reform imperiled.

3

State Capitalism and the Operation of the Public Sector

This chapter focuses on the strategy of state capitalism as the model of development in Bolivia during 1952-85 and particularly on the role of public finances, public investment spending, and trade policy in that period. We have already stressed that the Bolivian debt crisis in the 1980s, and the economic crisis more generally, can be traced largely to insufficient domestic savings and to a very weak public sector, despite the central role assigned to the public sector in the development strategy. The main point that we emphasize is that while the government faced large and politically urgent demands for spending-on, for example, the wages of public sector workers, social objectives (e.g., health and edu-

189

BolividChapter 3

tion, and expenditure cutting in the state enterprises, especially in the mining company COMIBOL; (3) reform in the trade regime, aiming at the elimination of most barriers to international trade and the implementation of a flat tariff schedule; and (4) normalization of Bolivia’s access to international financial flows from the official creditor community (i.e., the IMF, the World Bank, the IDB, and bilateral creditors). The political base for implementing this enormous program was established in a crucial Pact for Democracy signed by Paz’s MNR and Banzer’s party, Acci6n Democratica Nacionalista (ADN). This agreement established a congressional majority sufficient to put through the Congress the key features of Supreme Decree 21060. The pact allowed the decree to take effect in the context of democracy, a remarkable accomplishment in view of Bolivia’s volatile political institutions and its vulnerable position amongst politically unstable neighbors in the region. With the pact, the Paz government reserved important positions in state enterprises for members of Banzer’s ADN party. In addition to the political base of the MNR and ADN, Paz also had strong support in the business community. Several prominent businessmen, some who had been head of the CEPB, were given top Cabinet posts. The structural reforms did away with the state-capitalism model, with its several variants, that had been followed since the Revolution of 1952. The reforms were carried out with a remarkable smoothness. Of course, there was some bitter opposition, particularly in the labor unions and in the parties of the left, as well as lively criticism in the press, but the changes were peaceful and in the framework of democratic institutions. At no point during Paz’s administration was the process of reform imperiled.

3

State Capitalism and the Operation of the Public Sector

This chapter focuses on the strategy of state capitalism as the model of development in Bolivia during 1952-85 and particularly on the role of public finances, public investment spending, and trade policy in that period. We have already stressed that the Bolivian debt crisis in the 1980s, and the economic crisis more generally, can be traced largely to insufficient domestic savings and to a very weak public sector, despite the central role assigned to the public sector in the development strategy. The main point that we emphasize is that while the government faced large and politically urgent demands for spending-on, for example, the wages of public sector workers, social objectives (e.g., health and edu-

190

Juan Antonio Morales and Jeffrey D. Sachs

cation), the military, and foreign debt servicing-the tax base for financing these expenditures had always been extremely fragile. Domestic revenues were generally adequate only during periods of high foreign prices of raw materials. When export prices fell, governments generally tried to maintain spending with funds borrowed from abroad. The real crises arose when neither high export prices nor foreign loans could sustain the level of government spending. Thus, after 1980 and before the stabilization program of 1985, the government was forced to rely heavily on the inflation tax. As a general point regarding the structure of government revenues, the important thing to notice is that revenues were collected primarily through production taxes and foreign trade taxes, collected from a small number of activities undertaken by a handful of enterprises. Many, but not all, of those enterprises were publicly owned. The tax structure can be seen in table 3.1, from Breuer (1988). In 1976, for example, taxes from trade, state enterprises, and private mining (which i s a production tax on an export good) accounted for 49.7 percent of the overall tax revenues. These three sources proved to be highly unstable during the 1980s as real tax revenues fell sharply. Trade taxes fell as a rising share of trade went through the black market; state enterprise taxes fell as profitability of the state sector declined, partly as the result of unrealistic pricing; and private mining profits fell, partly due to terms-of-trade decline and partly due to a failure of collections. With a tax structure as in Bolivia, the search for tax revenue was easily confounded with the search for investment opportunities for the public sector that could yield those revenues.' Nobody was better equipped, it was asserted, than the public sector to seize large investment opportunities, and the public sector needed the revenues in order to finance the balance of its expenditures. In this setting, the heavy emphasis on public investment projects combined with the allocation of public sector revenues, necessitated a development plan. From 1956 until 1985, the governments emphasized economic development based on public sector planning. The design and implementation of Table 3.1

Direct taxes State enterprises Private mining Other Indirect taxes Trade Excise Sales Other Total

Tax Structure in Bolivia, 1972-80 1972

1974

1976

1978

I980

25.3 8.9

43.4 28.7 7.4 7.4 56.6 18.9 4.9 3.3 29.5 100.0

46.6 24.8 3.8 18.0 53.4 21.1 4.5 3.8 24.1 100.0

45.7 22.5 4.7 18.6 54.3 23.3 7.0 5.4 18.6 100.0

45.5 23.2 4.0 18.2 54.5 20.2 6.1 6.1 22.2 100.0

.o

16.5 74.7 31.6 8.9 6.3 27.8 100.0

Sources: Breuer (1988). table S.I(B), which is based on estimates supplied by UDAPE in the Ministry of Planning, Bolivia.

191

BolividChapter 3

development plans were considered essential for rapid economic growth in the productive and distributive spheres. The common view on the topic over a long period was that centralized planning should be compulsory for the public sector and indicative for the private sector. In this chapter substantial space is devoted to the public sector investment projects as they were framed in development plans, the main conceptions of which shed some light on the projects themselves. Indeed, the policy of investing public income and publicly guaranteed loans in huge projects of resource-based industrialization cannot be fully understood without appealing to the main conceptions on development as reflected in the development plans. Unfortunately, failed public investments, together with the narrowness and vulnerability of the tax base, are the main sources of the eventual collapse of the public sector.

3.1 The Mature State-Capitalism Model, 1969-84 The most important feature of the state-capitalism model is the degree of intervention of the government in the economy. In 1975, for example, the GDP generated in the public sector accounted for 22 percent of total GDP, while in 1982 this ratio reached a high of 28 percent. The data in table 3.2 are even more telling, as they show the very high share of public sector investment in total investment spending. During 1965-69, the public investment share in total investment was 51 percent; by 1975, the share decreased slightly to 48 percent; during 1976-80, it increased again, reaching an average of 60.4 percent. From 1981 to 1985, the ratio fell sharply because of the deep cuts in public investment spending during the financial crisis. It may be noted that from 1970 on, the years of high private investment rates are also associated with strong investment rates of the public sector. This is in part a manifestation of state capitalism, in that Public and Private Participation in Gross Fixed Capital Formation (in percentages)

Table 3.2

Year

Public

Private

Year

Public

Private

1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969

42 40 44 45 54 62 60 43 50

58 60 56 55 46 38 40 57 50 50 48 41

1970 1971 1972 1973 1974 I975 1976 1977 1978 1979 1980

56 60 59 37 34 41

44 40 41 63 66 59 44 36 34 42 42

50 52 59

Sources: Ministry of Planning and Central Bank of Bolivia.

56

64 66 58 58

192

Juan Antonio Morales and Jeffrey D. Sachs

private investment spending was often closely tied to specific public sector projects. It is worth looking at some explicit statements of the state-capitalism model contained in several development plans formulated by various governments. Though the model has roots that go back to the Revolution and even before, we will focus our analysis on the 1969-84 period. In 1969 the left-wing populist government of General Ovando presented to the nation a Socioeconomic Strategy for National Development (SSND). The SSND was more “statist” and nationalistic than the ten-year plan prepared by the MNR regime in 1962.* The SSND clearly defined a long-run vision of state-led growth, with many “strategic sectors” of production that were reserved to the public enterprises. The SSND targeted an increase in domestic saving based on the retained surpluses of the enterprises in the public sector; as a counterpart, and consistent with a nationalistic ideology, it called for a reduction of foreign savings through cuts in foreign direct investment and public external borrowing. A very important feature of the SSND was the emphasis given to resource-based industrialization. A key theme was that big operating surpluses can be generated in the state enterprises in the resource sectors, which in turn can be reinvested in other sectors. In addition, the SSND, in a clear case of misjudgment, relied quite heavily for the industrialization program on the market prospects offered by the economic integration scheme of the Andean Group Pact of 1969, formed by Bolivia, Colombia, Ecuador, Peru, and Venezuela (Chile belonged originally, but quit in 1976). With regard to agriculture, the SSND advocated a more communitarian organization of production, with strong emphasis on cooperatives and technological improvements. The SSND was short-lived, abandoned by President Banzer in 1972. However, the SSND exerted a powerful influence on the formation of ideas in Bolivia that went well beyond 1972. The strongly nationalistic nature of the SSND appealed to the left in the Bolivian political spectrum and, ironically, to many high-ranking officers in the military. In fact, the statecapitalism model reached its peak during the government of General Banzer and, although we can trace the origins of Banzer’s model to the Revolution of 1952, many of the most important concepts came from the SSND. The expansion of the public sector, especially as a producer of goods and services that could, under appropriate policies, be provided by private enterprises, indeed constitutes the landmark of the Banzer years. But this did not mean a confrontation with the private sector; instead, the policies of the Banzer regime aimed at a symbiotic relationship between the public and private sectors. Moreover, and this needs to be underscored, the public enterprises served frequently as a mechanism to transfer state-owned (or state-guaranteed) resources to privileged groups in the private sector. Access

193

BolividChapter 3

to government officials and government contracts were considered the most important asset from the viewpoint of many private sector businessmen. On the basis of research in progress, Ugarteche (1986) and Rivas (1986) claim that the main beneficiaries of the public sector external debt accumulated in the Banzer period were private sector supporters of the Banzer regime. Also, public employment was increased substantially between 1971 and 1978: many new public enterprises were created, and the government asserted its monopoly over the developments in the new mining sectors (especially iron), as well as in the commercialization of tin and metals. In 1975 Banzer announced a five-year development plan. In the summary of the plan given by Malloy and Borzutszky (1982), the following points stand out: (1) the plan’s aim was export-led growth accompanied by selective import substitution; (2) it gave the preponderant role to the public sector, targeting the public sector for 71 percent of total investment spending; (3) it allocated public investment primarily to the production of hydrocarbons; (4) it divided the nation into three investment zones: a primary zone of La Paz-Cochabamba-Santa Cruz, a secondary zone of Oruro-Potosi-Chuquisaca, and a tertiary zone of Beni and Pando, with the share in total investment of the primary zone to be 63 percent; and (5) it paid relatively scant attention to the peasantry. Most of Banzer’s five-year plan was never put in place. Two factors accounted for the failure to implement the plan. First, the financing of the plan was heavily dependent on Bolivian petroleum exports which failed to materialize. Second, the public administration did not adhere to the guidelines for the allocation of investment, so that many public investment projects were undertaken without due consideration for the investment allocations outlined in the plan. As this practice became common, the plan became wholly unenforceable. In the SSND and the five-year plan (as well as in the MNR’s 1962 ten-year plan), traditional exports occupy the center stage. In all of these plans, the expansion of traditional exports was deemed to be the “locomotive” of development, with the rest of the sectors of the economy following along. It has to be emphasized that nowhere was the conception of Bolivia as an export economy abandoned in favor of, for instance, import-substituting activities on a large scale. Planners and government bureaucrats did look on an expansion of a protected light manufacturing sector with some sympathy, but they did not think, given the smallness of the domestic market, that this sector could be the engine of rapid economic growth in the near future. Hopes were always placed on growth led by exports of primary commodities. The model of export-led growth embodied in the development plans that we have discussed differed significantly from the traditional pattern of growth, as well as from the export-led models in the newly industrialized countries (NICs). In contrast to the traditional pattern, based on the export of

194

Juan Antonio Morales and Jeffrey D. Sachs

a few primary products, the development plans aimed at a more diversified export basket, with the production of commodity-based goods having more domestic value added (e.g., the export of metals instead of mineral ores, and refined sugars and oils instead of the crude varieties). Unlike in the strategy of the NICs, there was little emphasis on the export of manufactures, except for some belated attempts in the context of the Andean Pact market. With the emphasis on petroleum and mineral exports, policymakers placed little importance on exchange rate management to promote new exports or on the deepening of domestic financial markets, though exchange rate and financial policies had played a crucial role in the successful takeoff of the NICs (cf. Tsiang 1986). The military successors of Banzer, by and large, tried to continue with the state-capitalism model. The civilian governments of Walter Guevara Arze and Lidia Gueiler attempted to follow a somewhat less “statist” course, but their efforts were curtailed by strong political adversities. In 1984 one more attempt was made to develop a plan, with the “National Development Strategy and Four-Year Plan.” Three different versions of the plan were offered to the public. Because of the internal squabbles of the coalition partners in the Union Democratica Popular (UDP) governments in 1984-85 and the high priority given to the inflation problem, no real importance was attached to any of the versions by the government or the public. 3.1.1

Public Investment Spending in the State-Capitalism Model

The deficiencies of planning were nowhere more revealed than in the execution of public investment projects. A full evaluation of investments in the very capital-intensive projects has yet to be made. Ideally, information should be gathered for an ex ante and ex post evaluation of each major project undertaken in the 1970s and up until 1985. Short of that, we may note the following points. First, many large projects responded more to noneconomic factors (prestige, national security, etc.) than to profitability, measured either in private or social terms. Second, grave mistakes were made in evaluating the endowments of natural resources. For example, overoptimistic assessments of oil and mineral reserves led to overinvestment in these industries. Third, the cost-benefit analyses performed before undertaking the projects were either incomplete or were ignored in the implementation phase. Fourth, the large projects were typically financed with expensive suppliers’ credits and foreign bank loans. The conditions of repayment were, from the start, likely to create problems in the cash-flow stream. Fifth, when the projects were financed with official loans, delays in disbursements often disrupted the investment schedules. Sixth, the execution of the projects was extremely poor, with frequent but avoidable long delays in deliveries and construction. Contrary to what occurred in other debtor countries, public investment was largely directed to sectors producing tradables or to physical infrastructure,

195

BolividChapter 3

and even in the last case, infrastructure served primarily for the production of tradables. The problem in Bolivia was not an excessive allocation of investment to nontradables, but rather the highly inefficient choice and execution of investment projects within the tradables sector. Political and regional pressures, as well as specific vested interests, held great sway over the selection of investment projects and the procurement of construction and supply contracts. Graft, of course, was a very important factor-kickbacks and unwarranted cost overruns were the order of the day. Several examples illustrate the assertion of investment mismanagement: the uneconomical expansion of the tin smelters of the ENAF company in Vinto; the huge Karachipampa polymetallurgic smelter in Oruro for which there were no minerals; the expansions in oil-refining capacity in Santa Cruz and Cochabamba when oil production was declining at a rapid rate; the refinery for soya oil built close to the Paraguayan border, so far away from producing and marketing areas that transportation costs made it completely unprofitable; the very modern international airport of Vim-Viru in Santa Cruz, with a traffic load of no more than fifteen planes a day in 1987; the high-cost military-run truck assembly plant of Cochabamba; and so on. Even when some projects could have been reasonably profitable under the right circumstances, cost overruns and technical misjudgments impaired profitability. For instance, the COMIBOL plant for the treatment of low-grade ores in La Palca was planned to be built in two years, yet after ten years it was still unfinished. Similarly, there is the sugar refining mill of Guabira, the abnormally high costs of which were essentially due to poor financial management that systematically delayed the sugar harvest by its suppliers, therefore causing a strong fall in the saccharose content of the inputs. It is interesting to note that private investment projects financed with foreign resources and public guarantees also showed poor performance, and substantial arrears were quickly built up after the loans were made. Loans to private firms that were guaranteed by the public sector institution, Banco del Estado, were used in white elephant projects and had a very low probability of being repaid. Similarly, the foreign loans channelled through the state-owned agricultural bank, Banco Agricola Boliviano, to large farmers in the eastern lowlands suffered extensive defaults. 3.1.2

State Enterprises in the State-Capitalism Model

The public enterprises were, during 1952-85, the main hope for economic development but also the Achilles’ heel of the public sector. As was shown before, the size of the Bolivian public sector relative to GDP has been considerable, far above the average in Latin America. Up to 1983, most of the larger public enterprises were able to generate savings in spite of adverse macroeconomic policies (mainly overvalued exchange rates), poor management, and labor indiscipline. Two exceptions were COMIBOL,

196

Juan Antonio Morales and Jeffrey D. Sachs

which already showed current account losses in 1980 (and chronic overall deficits, when account is taken of investment spending) and the smelting company, ENAF, which had troubles since its inception. For the other firms, however, the savings were usually insufficient to finance capital investment and debt servicing. As a result, the public enterprises were among the main borrowers internationally, as an increasing share of their capital outlays in the 1970s were financed by foreign savings. For several reasons, however, it is very difficult to evaluate with any accuracy the performance of individual enterprises. There are simply too many distortions in financing and pricing to make such a retrospective analysis feasible. Complex cross-subsidies among the state-owned enterprises distort the measured profitability of individual firms. A case in point is that of ENAF, which was heavily subsidized by COMIBOL. Also, the National Enterprise of Electricity (ENDE) subsidized all the other public companies through low electricity tariffs. A further complexity in evaluating the state enterprises is the need to place the enterprises in a macroeconomic context, while also taking careful account of trends in the international economy. The net incomes of the state enterprises are much more sensitive to macroeconomic policies than are the profits of private sector enterprises. The temptation to suppress the symptoms of inflation by manipulating the prices of the public enterprises or by delaying a devaluation of the official exchange rate, for example, is a constant danger. Lags during 1982-85 in raising public sector prices (especially petroleum products) and in devaluing the official exchange rate brought heavy losses to the public enterprises. As we shall note in chapter 5, the complete mismanagement of prices of publicly provided goods was an essential feature of Bolivian policymaking in 1982-85. To help understand the behavioral choices of the state enterprises in Bolivia, it is relevant to introduce here the concept of the soft budget constraint, as introduced by Kornai (1986) to explain the behavior of public sector firms in socialist economies. The essential idea is that normal budgetary constraints on a firm’s behavior do not hold in socialist economies since the firm’s managers know that in case of trouble, the firm will be bailed out by subsidies or credits from the central government. As Kornai points out, this leads to a nearly insatiable demand for public sector credit, since “sooner or later it can expect to be able to cover its costs on input, and if its proceeds on sales of output are insufficient, it will be able to cover costs from an external financial source” (44). Also, state enterprises facing a soft budget constraint lack incentives and sanctions to stick to their production goals and to limits on use of resources. This situation characterized most state enterprises in Bolivia during the period up to 1985 (and to some extent afterward). During the 1970s, individual firms could borrow internationally on the full faith and credit of the central government. There were few bureaucratic constraints that re-

197

BolividChapter 3

strained them from doing so, with the result, as we saw in the previous section, of a series of disastrously inefficient and uneconomical investment projects. The problem of soft budget constraints was especially severe during the high-inflation years of 1982-85. It was relatively easy to find ways to escape the budget constraints, for instance by delaying tax payments of the enterprises to the central government, accumulating liabilities and then shifting them to the Treasury, gaining access to heavily subsidized Central Bank loans, or contracting foreign debt beyond their needs when they could. Many state enterprises also suffered from a confusion of objectives: they were asked to generate profits to be reinvested, as well as to be shifted to other sectors; to generate employment, frequently as a disguised scheme of unemployment compensation; to provide education and health services to workers plus their families; and to act as a retirement fund and as a means to channel subsidies to the general consumer. This multiplicity of objectives impaired their normal functioning, as might be suspected. Most perniciously, state enterprises were used as a form of political patronage and spoils to bolster the political power base of the government. For this reason, employment in state enterprises increased remarkably rapidly in the 1970s, far outstripping the growth of population and the economy in general. Between 1970 and 1982, state enterprise employment rose at an annual rate of 4.6 percent. Political pressures played an important role in the expansion of jobs during these years. In fact, a nonnegligible part of the support for the Banzer government and the succeeding military regimes was the willingness to create employment in the public sector. The return to democracy in 1982 was also accompanied by a big spurt in the expansion of jobs in the most important public enterprises, particularly in COMIBOL. The distribution of sectoral ministries and public enterprises among the parties in the UDP coalition led to an unrestrained race of political patronage. The patronage of the Communist Party, through the Ministry of Mines that it controlled, was especially notorious. In the same vein, Rivas (1986), Ugarteche (1986), and many other Bolivian analysts also claim that it was through the mismanagement of public investment projects that the public foreign debt was diverted to private beneficiaries, often with political goals in mind. In this direct way, the public debt financed consumption and capital flight. While the government of Bolivia retained the commitments, the benefits were privatized. 3.1.3 The Public Sector as “Shock Absorber” for External Shocks

The public sector was frequently used to absorb, at least temporarily, external shocks hitting the economy. For instance, while YPFB reaped some important benefits and the Treasury substantial taxes from the oil price increases of the 1970s, each of these could have been greater had the government not pursued a policy of heavily subsidizing domestic oil

198

Juan Antonio Morales and Jeffrey D. Sachs

consumption. The protection of some consumers from the oil shocks was made at the expense of taxes and profits for the oil enterprises. The opportunity costs of this policy were high. Moreover, as some oil products were smuggled out, the subsidies gave much to the contrabandistas (smugglers) in addition to the general consumer. In some cases, not only opportunity costs were incurred but also heavy Treasury disbursements were spent in s ~ b s i d i e s For . ~ instance, when world prices of wheat and other grains climbed during 1974-75 and in 1979, domestic prices of wheat were kept constant, with the Treasury making up the difference between import prices and domestic prices. Moreover, the government established de facto price support schemes for some agricultural staples, like sugar and rice. These price support schemes were very costly in years of low international prices and good harvests. The essential point is that the fiscal sector was used to isolate the economy from external fluctuations. When the shocks were transitory, this policy did not have important consequences. However, as adverse shocks frequently lasted longer than expected, significant debt accumulation ensued. The domestic price repercussions of those deficits went unnoticed as long as it was possible to finance them by borrowing from abroad. The government deficits that resulted from subsidies was one of the channels that led to the financing of consumption with foreign debt. In some cases, the “shock absorber” function was used for the protection of a very narrow part of the elite. A good illustration is the case of cotton growers in the Santa Cruz region during the Banzer period. Tens of millions of dollars were loaned by the government’s agricultural bank (Banco Agricola Boliviano) to a small number of large landowners during an incipient cotton boom. The Bolivian producers sold their crops forward on world markets, but then tried to renegotiate when the spot-market prices rose above the forward contract price. They failed in the renegotiations and, in the end, much of the cotton remained undelivered. The growers quietly defaulted on the agricultural bank loans and the government absorbed the losses with no attempt to collect or to foreclose on any property. To this day, the bad cotton loans have undermined the solvency of Banco Agricola.

3.2 The Chronic Weakness of Public Sector Revenues In Bolivia, the sources of income for the public sector have suffered from extreme fragmentation since the beginning of the Republic. Before the tax reform measures of 1986, there were many central government taxes with as many collecting agencies; local taxes; and taxes earmarked for regional development corporations, the state universities, and for the social security system. In part because of this extreme fragmentation and in part because of notoriously bad record keeping for tax collections, it is hard to make a precise quantitative assessment of the public revenue system.

199

BolividChapter 3

The combined accounts of the central administration, the social security system, and the so-called decentralized agencies form the central government accounts. If one adds to the account of the central government the accounts of the local governments, this gives the general government accounts. The general government accounts together with the accounts of the nonfinancial state enterprises form the consolidated nonfinancial public sector. Note that transfers between parts of the system are very important for the consolidated nonfinancial public sector. Public enterprises pay taxes to (and on occasion receive subsidies from) the central administration and some local governments. Central administration transfers to the decentralized agencies and local governments are very substantial. Finally, the quasi-fiscal activities of the financial public sector (i.e., the Central Bank of Bolivia, together with the state banking system) should be consolidated with the nonfinancial public sector accounts to get a comprehensive view of the fiscal activities of the overall public sector. Because of data limitations, this complete consolidation would be virtually impossible for historical data. The best recorded data for the public sector are found for the account of the General Treasury of the Nation, TGN (Tesoro General Nacional), which handles the accounts for the central administration and which has traditionally been under the jurisdiction of the Ministry of Finance. We now turn to a look at revenues of the TGN.

3.2.1 The Revenues of the General Administration From 1961 to 1970, central administration revenues amounted to 8.0 percent of GDP on the average; in the next decade this ratio increased to 10.2 percent, with a slight rise in the second half of the decade compared with the first half. The maximum ratio of revenues to GDP of the TGN occurred in 1976 (12 percent of GDP). Data on revenues, expenditures, and the deficit, from Breuer (1988), are shown in table 3.3 for the years 1970 to 1984. The rise in revenues as a percentage of GNP from the 1960s to the 1970s was due to a determined effort to improve tax administration and to an increase in the tax base because of the expansion in the value of exports during the 1970s. However, after 1979, inflation started to take its toll on the real value of tax collections, in line with the so-called Olivera-Tanzi effect (see Tanzi 1977), which predicts a negative relation between inflation rates and real tax collections. During the high inflation years, the TGN revenues were as low as 2.8 percent of GDP in 1984 (and fell to only about 1 percent of GDP in the first quarter of 1985!). We shall describe in chapter 5 more precisely how the high inflation made a shambles of the Bolivian tax system. Changes in the structure of the TGN revenues should also be noted.5 In the 1970s, taxes collected by the internal revenue service amounted to 34 percent of the TGN revenues, while taxes on the foreign sector, mainly

200

Juan Antonio Morales and Jeffrey D. Sachs

Table 3.3

Budget Revenues, Expenditures, and the Deficit, 1970-84 (as percentage of GNP) Year

Revenues

Expenditures

Deficit

1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984

8.4 7.9 7.6 9. I 11.5 11.3 12.0

9.8 10.0 10.0 10.6 11.9 12.6 14.0 13.8 14.0 14.3 16.0 15. I 26.9 20. I 33.2

1.4 2.1 2.4 1.5 .4 1.3 2. I

11.5 11.4 9.4 9.6 9.4 4.6 2.6 2.6

2.2 2.7 4.9 6.4 5.6 22.3 17.5 30.6

Sources: Breuer (1988), table 5.5, which is based on estimates supplied by UDAPE in the Ministry of Planning, Bolivia.

import tariffs, represented 31 percent on the average. The remaining sources were production taxes on the mining and petroleum sectors, which averaged 31 percent. Property income and services provided by the central government yielded 4 percent of the TGN revenue. By the end of the hyperinflation, however, tax collections on internal income and on international trade had been sharply eroded. Only the taxes on petroleum production could be easily collected in the short term, even after the hyperinflation stopped. Therefore, in 1985, the year that stabilization began, taxes on petroleum represented no less than 64 percent of TGN revenue. We return to this point in chapter 5. In regard to the composition of internal taxes (more precisely, taxes collected by the internal revenue service and excise taxes on petroleum products), the most important component during the normal years 1975-78 was taxes on consumption and sales. The share of personal income taxes in total internal taxes has never been above 25 percent. The contribution of corporate income taxes was likewise modest, never going beyond 23 percent of internal taxes. Until the tax reform of 1986, property and road taxes were local taxes collected and earmarked for the municipalities and did not enter the TGN coffers. It is very hard to locate data of good quality for local governments, so only gross trends can be suggested. As happened with the TGN revenues, there was a slight improvement in tax effort during the 1970s, but in the 1980s, property taxes and vehicle taxes suffered very much with high inflation. Not only did lags in payments affect the real value of collections, but property assessments (which were undervalued to begin with) lagged far

201

BolividChapter 3

behind inflation, so that real property values for tax purposes fell sharply during the high-inflation period. Some municipalities in dire financial straits asked for, and obtained, frequent direct transfers from the TGN. The financing problems of the regional development corporations, the universities, and the social security system were also acute during the high inflation period.

3.3 Management of Public Expenditures The chronic weakness of public sector revenues and the low profitability of public investment spending explain a large part of the overall fiscal disaster in Bolivia in the 1980s. Also important were the dynamics of public sector spending on the current account of the budget (i.e., other than for capital expenditures). The key point here is that debt-servicing costs rose sharply in the 1980s just as the availability of new financing dried up. The government was squeezed by, falling tax revenues (due to rising inflation), higher expenditure charges (due to rising interest payments), and an inability to finance the resulting gap (as well as the deficit levels inherited from the 1970s) through foreign borrowing. The result was a turn to the inflation tax. Unfortunately, there is a profound difficulty in tracing interest payments carefully through the budget since the interest obligations were at various points charged to the state enterprises, the central administration, and the Central Bank. We will be able to review in chapter 5 some data on overall interest payments made to foreign creditors, but without precisely specifying how the spending was allocated among various parts of the public sector. As in the preceding section, a detailed analysis of expenditures must focus mainly on the TGN because of a lack of information regarding other parts of the state sector. Precisely because of the shifting of various financial burdens among different parts of the state sector, a focus on the TGN is not only incomplete, but may sometimes even be misleading. TGN expenditures as a percentage of GDP were 9.6 percent on average in the decade of the sixties, 12.2 percent in the seventies, and 16.7 percent from 1981 to 1985. The evolution of overall expenditures was shown in table 3.3. The large increase in overall spending after 1981 is mostly accounted for by a rise in debt-service expenditures and in transfers to financially strapped state enterprises. The composition of the TGN expenditures for selected years is shown in table 3.4. Note that in the 1970s, personnel expenses were generally about half of the total. In 1982, however, interest costs on the foreign and internal debt rose sharply, and payments to the Central Bank, also apparently related to foreign debt, accounted for about two-thirds of overall expenditures. In 1984, interest costs fell sharply as Bolivia suspended payments on the commercial bank debt. Was the central government profligate between 1970 and 1985? A key indicator of profligacy is the increase in employment. Employment in the

202 Table 3.4

Juan Antonio Morales and Jeffrey D. Sachs Structure of Expenditures in the Central Government, 1970-84 (as percentage of total expenditure)

Personnel Materials Transfers Debt service Payments to Central Bank Other Total

1970

1976

1980

1982

1984

54. I I .4

42.4 6.3 16.6 2.1 .0 22.7 100.0

46.4 7.3 11.6 16.3 .0 18.4 100.0

20.8 3.0 10.0 26.8 40.7 4.1 100.0

59.7 5.5 19.8 7.6 .0 7.6 100.0

16.3 13.6

.o

14.6 100.0

Sources: Data refer to central government (TGN) and are based on the data given in table 5.4 of Breuer (1988), which in turn were provided by UDAPE in the Ministry of Planning, Bolivia.

central administration increased by 92.4 percent between 1970 annd 1982, yielding an average annual rate of growth of 5.6 percent. This rate of growth was well above that of the urban population and of GDP, with the main increases occumng between 1970 and 1976. We know from the political analysis that succeeding administrations used patronage as a way to cement patron-client relations, and thereby build a political base of support. It does indeed seem that the result was a profligate and inefficient overextension of public sector employment. Expenditures on investment that were on average around 2 percent of GDP during 1976-79 fell to around 0.3 percent during the crisis years of 1981-85. Since central government investment expenditure is mainly on social overhead, the impact of its substantial reduction has important repercussions on income distribution and on growth. The fall in this particular type of investment will have long-lasting effects, the magnitude of which has not yet been fully appreciated.

4

Trade Policies, 1970- 85

It should be recalled from our overview in chapter 1 that the long-run growth of the Bolivian economy has been critically determined by the exports of primary commodities, mainly tin and natural gas. Bolivia’s economy depends crucially on the performance of the export sector. In turn, shifts in indebtedness have coincided, procyclically, with the export cycle. Bolivia’s dollar export earnings during 1970-88 are shown in table 4.1. Export earnings and, by extension, the domestic economy have been greatly affected by the instability of Bolivia’s export prices. As a result, policymakers have focused on measures to stabilize and improve Bolivia’s

202 Table 3.4

Juan Antonio Morales and Jeffrey D. Sachs Structure of Expenditures in the Central Government, 1970-84 (as percentage of total expenditure)

Personnel Materials Transfers Debt service Payments to Central Bank Other Total

1970

1976

1980

1982

1984

54. I I .4

42.4 6.3 16.6 2.1 .0 22.7 100.0

46.4 7.3 11.6 16.3 .0 18.4 100.0

20.8 3.0 10.0 26.8 40.7 4.1 100.0

59.7 5.5 19.8 7.6 .0 7.6 100.0

16.3 13.6

.o

14.6 100.0

Sources: Data refer to central government (TGN) and are based on the data given in table 5.4 of Breuer (1988), which in turn were provided by UDAPE in the Ministry of Planning, Bolivia.

central administration increased by 92.4 percent between 1970 annd 1982, yielding an average annual rate of growth of 5.6 percent. This rate of growth was well above that of the urban population and of GDP, with the main increases occumng between 1970 and 1976. We know from the political analysis that succeeding administrations used patronage as a way to cement patron-client relations, and thereby build a political base of support. It does indeed seem that the result was a profligate and inefficient overextension of public sector employment. Expenditures on investment that were on average around 2 percent of GDP during 1976-79 fell to around 0.3 percent during the crisis years of 1981-85. Since central government investment expenditure is mainly on social overhead, the impact of its substantial reduction has important repercussions on income distribution and on growth. The fall in this particular type of investment will have long-lasting effects, the magnitude of which has not yet been fully appreciated.

4

Trade Policies, 1970- 85

It should be recalled from our overview in chapter 1 that the long-run growth of the Bolivian economy has been critically determined by the exports of primary commodities, mainly tin and natural gas. Bolivia’s economy depends crucially on the performance of the export sector. In turn, shifts in indebtedness have coincided, procyclically, with the export cycle. Bolivia’s dollar export earnings during 1970-88 are shown in table 4.1. Export earnings and, by extension, the domestic economy have been greatly affected by the instability of Bolivia’s export prices. As a result, policymakers have focused on measures to stabilize and improve Bolivia’s

BolividChapter 4

203 Table 4.1

Export Earnings, 1970-88 (in millions of U.S. dollars) 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979

190.4 181.6 201.3 260.8 556.5 444.7 563.0 634.3 627.3 759.8

1980 1981 1982 1983 1984 1985 1986 1987 1988

942.2 912.4 827.7 755.1 724.5 623.4 545.5 518.7 542.5

Source: IMF, International Financial Statistics.

terms of trade, particularly with respect to tin. These measures have been pursued mainly by participation in international stabilization agreements on tin and by lobbying to forestall sales of this commodity by the industrial countries. In regard to natural gas, the trade policy has been much more passive. Many domestic economic policies have affected the development of Bolivia’s foreign trade performance in recent years. Some of the policies were particularly harmful and played an important role in the severity of the crisis in the 1980s. In part because of adverse trade policies and in part because of adverse terms-of-trade shocks that were out of Bolivia’s control, Bolivia suffered one of the sharpest declines in Latin America in the purchasing power of exports (PPX) between 1981 and 1988, as shown in table 4.2.’ In this chapter, we review the main trade policies and their effects on trade performance. Particular attention is paid to exchange rate management. The structure of tariffs and the taxation of natural resources are also examined. The important question that underlies the whole chapter is why, despite a high dependency on exports, Bolivia’s long-run export performance has been so poor. 4.1

Export Policies

Bolivia’s export policies during 1961- 8 1 were primarily aimed at strengthening or at least stabilizing Bolivia’s terms of trade in the major commodity markets.’ From 1982 to 1985, little attention was paid to trade policies given the overwhelming problems of internal stabilization. By far the most important scheme of price stabilization was provided by Bolivia’s participation in the International Tin Agreements (ITA). Five agreements were signed-1956, 1961, 1966, 1971, and 1976-but Bolivia did not join in signing the last one in 1976, in protest against price targets that it regarded as too low. In the 1970s, agreements of lesser scope were also signed for tungsten and antimony, other important mineral exports of Bolivia.

204

Juan Antonio Morales and Jeffrey D. Sachs

Table 4.2

The Purchasing Power of Exports (PPX) in Bolivia and Selected Countries, 1988 (1980 = 100 for all indices)

Country

PPX, 1988

Export Volume

Terms of Trade

Bolivia

57

69

89

I02 167 156 152 103

131

79 86

Argentina Brazil Chile Colombia Ecuador Mexico Peru Uruguay Venezuela

139

75 148 56

197 I58 174 159 228 71 I30 I06

101

90 67 62 108 I16 55

Source: Economic Commission for Latin America and the Caribbean, United Nations (ECLAC), “Preliminary Overview of the Latin American Economy, 1988” (3 January 1989): table 8, export volumes; table 10, terms of trade; and table 12, purchasing power of exports. As explained in endnote 1 to this chapter, the PPX should equal the product of the export volume index and the terms-of-trade index. This is only approximately true for the data shown, apparently because of the differing coverage of goods in the three indices reported by ECLAC.

The ITAs were agreed upon by the main tin producing and consuming countries, with the exception of the United States. The governing body of the ITA is the International Tin Council (ITC). The main, but not the only, instrument for achieving the price stabilization objective was a buffer stock of tin metal financed by the producing members. In negotiations for the five agreements, Bolivia, which had the highest production costs among the producing countries, lobbied systematically for higher floor and ceiling prices than those set by the ITC. Other producers did not follow Bolivia, feeling that a long-run policy of high prices would backfire on them. Time proved them right. There is considerable controversy over the workings of the ITC and of the buffer stock. For example, there was a problem with the small size of the agreed-upon stock. In fact, the buffer stock became irrelevant in the booming market of the 1970s. Moreover, the buffer stock could hardly cope with the most important destabilizing factor in the tin market, namely, the huge strategic stockpile of tin held by the U.S. General Service Administration (GSA). In the 1980s, the ITC held prices that were much too high instead of allowing a smooth accommodation to the weaker market conditions. High prices induced the entry of new producers in the market and hastened the process of technological substitution with other metals and materials. In addition, the financing of the buffer stock became a problem. This conjuction of an excessively high price with financing problems led to the collapse of the tin agreement in October 1985 and the collapse of tin prices from $5.60 per pound on the eve of the collapse to $2.55 per pound in July 1986, nine months later. The October collapse had a stunning effect: the buffer stock declared insolvency and the London Metal Exchange ceased trading in this metal. The

205

BolividChapter 4

evolution of the real price of tin (relative to the unit value of imports of the developing countries) was shown in figure 1.2 in chapter 1, in which the collapse in October 1985 is plainly evident. Besides the problem of price stabilization, the production and export activities of the mining sector during the 1970s were adversely affected by onerous tax legislation, which was somewhat eased after 1979. The mining sector was subjected to two main types of taxes: a regalia, which was initiated in 1965, and an export tax imposed with the devaluation of 1972. The regalia is a tax on presumed income, given that the nominal base of the tax results from the difference between world mineral prices and a presumed cost set by the Bolivian Ministry of Mines. Since presumptive costs changed infrequently, the regalia functioned in fact as a tax on the gross value of output. The regalia overtaxed the mining sector, and particularly the weakest enterprises, in years of low mineral prices, whereas it failed to fiscally appropriate the rents that were generated in years of rising mineral prices (Gillis 1978). Moreover, tax codes did not encourage investment in mineral exploration and development. Petroleum was a major export in Bolivia. But after 1973, with the rise in domestic consumption and the progressive depletion of reserves, the amount left for exports decreased substantially, and Bolivia ceased to be a net petroleum exporter in 1977. The systematic domestic underpricing of petroleum products encouraged the demand for both domestic consumption and for contraband exports, which hindered a sensible development of petroleum exports and appropriate tax revenues. In addition, petroleum production and exports have been subjected to punitive taxation and this, too, has had long-run costs in discouraging supply. Bolivia has important deposits of natural gas. In fact, the export prospects for energy lie mainly in natural gas. Exports of gas to Argentina have been a very important source of foreign exchange. In the 1970s, gas exports were already marred by controversies about price, and these controversies have gained in intensity in the last years. Unfortunately, pricing principles were not clearly established when the gas pipeline to Argentina was put into operation in 1972. A negotiation during the 1970s between Bolivia and Brazil to export natural gas to Brazil did not succeed because of domestic political opposition to sales of the “national patrimony” to Brazil. These negotiations have been resumed under the New Economic Policy begun in 1985. In 1977 the Bolivian government decided to subsidize nontraditional (or minor) exports, including selected agricultural products and manufactures. The Law of Fiscal Incentives of 1977, and its reform in 1982, for nontraditional exports included exemptions from all export taxes as well as from import duties for inputs into exports and a tax rebate certificate granted to the exporters. The certificate, which amounted to between 10 and 25 percent of the FOB value of exports, could be used to pay taxes on income,

206

Juan Antonio Morales and Jeffrey D. Sachs

sales, or imports. It could also be sold freely for use by other exporters. The tax certificate was a direct subsidy that partially compensated an increasing overvaluation of the peso. An ex post evaluation of export policies demonstrates that these policies were not always clearly stated, nor were their effects fully appraised. It is clear that with respect to traditional exports, fiscal measures were generally inimical to a long-term increase in supply. The fiscal system focused on expropriating economic rents-a legitimate objective, of course-more than on encouraging the opening of new mines or the drilling of new wells. In regard to the promotion of nontraditional exports, it is possible to make two appraisals. First, the scheme of 1977 (and the reform of 1982) was subject to considerable abuse, without really leading to increased incentives for more exports. Second, the exportable products that benefited from this export promotion policy constituted less than 5 percent of the value of all exports. In fact, the emphasis on fiscal measures obscured the fact that domestic firms and industrialists first had to learn how to improve their production and merchandising methods. Thus, it appears that more effective forms of encouraging nontraditonal exports could have been sought. 4.2

Import Policies

Major import tariff reforms took place in 1967, 1973, 1982, 1985, and 1986. Before the changes of 1985 and 1986, the most important was the reform in 1973, which had been distorted with piecemeal changes in the tax rates but which nonetheless affected the schedule in significant ways. A very important feature of the tariff structure in place until 1985 was the existence of preferential tariff provisions for (1) commodities, according to the final use to which they were put; (2) goods used in the northwestern regions of the country; and (3) goods coming from countries with which Bolivia had (and has) economic agreements for bilateral reductions in tariffs. Examples of preferential tariffs of the first type were the special provisions for imports for the mining and petroleum sectors and the exemptions accorded by the Investment Laws of 1972 and 1981. Preferences of the third type included the Bolivian Lists of Tariff Concessions to the member countries of the Latin American Free Trade Association and the Andean Group. These provisions for preferential tariffs affected an important proportion of Bolivian imports. Depending on the year, the value of imports subject to the preferential rates ranged between 25 and 35 percent of total imports. Considerations of government revenue and exigencies of the balance of payments (i.e., the need to constrain the fall of reserves in the context of a pegged exchange rate) prevailed over the view of using tariffs (and other import policies) as effective tools for guiding industrial policy. The piecemeal changes eroded the original intentions of coherent and limited protection in

207

BolividChapter 4

the 1973 reform and in subsequent tariff changes. While one could find economic reasons to justify the distinct tariff rates in the reforms on protection and revenue grounds, the piecemeal changes introduced a high degree of dispersion of the tariffs, reflecting ad hoc considerations with little economic justification. Frequent changes were often brought about by the pressures of special interest groups of industrialists and importers. Before 1986, tariff duties, as is to be expected,were high on luxuries and competitive consumer goods and exhibited significant variation. On the other hand, tariff rates for capital goods were very low. Duties on raw materials and intermediate goods, which are necessary for domestic manufacturing and hence could be treated in a manner like that for capital goods, were, however, quite variable. The effective rate of protection is better than the nominal tariff rate as an indicator of the extent to which a particular set of tariffs protects domestic producers. Table 4.3 shows the effective rates for selected products prevailing in the second half of the 1970s. It is clear that there is considerable variation among the effective rates. Note that effective rates have also been computed for imports subject to quantitative restrictions by finding the implicit tariffs involved, which were calculated as the relative difference between international and domestic prices. This procedure was used as well in the case of prohibited imports. More specific conclusions can also be drawn from the data in table 4.3. First, the high protection provided by the import bans stands out. Apart from the case of import prohibitions, the most important characteristic that appears in the structure of effective protection is the high effective rates for goods considered luxuries. The effective rates are considerably higher than the already high nominal tariff rates. Second, it is clear that there is high effective protection for domestic production. In the cases of goods subject to import bans there is complete protection, but this is also true in many cases which are only subject to tariffs. Third, most intermediate products for industrial usage have low (or even negative) effective rates, which are generally very close to the nominal rates. Fourth, the effective rates for capital goods are close to the nominal rates; however, in many cases the effective rates are negative. Quantitative restrictions, including prohibitions, were used along with tariffs to limit imports during 1970-82, but their scope was reduced during the decade. In 1978 less than 2 percent of the Brussels Trade Nomenclature was subject to prohibitions. During the crisis years of 1982-85, many luxury and competitive imports were banned for balance-of-payments purposes (around 10 percent of the items of the Brussels Trade Nomenclature). Smuggling has greatly limited the application of tariff and quota policies and has substantially hurt government revenues. Once again, the expansion of smuggling was a symptom of the increasing administrative weakness of

208 Table 4.3

Juan Antonio Morales and Jeffrey D. Sachs Bolivian Nominal Tariffs and Effective Rates of Protection by Industry (31 December 1977)

Mean

Standard Deviation

38.9% 74.4%

28.0% 97.5%

Nominal Tariff

Effective Rate of Protection

A . Summary statistics for a list of337 groups of commodities

Nominal tariff Effective rate of protection Simple correlation between nominal and effective rates = 0.88 Rank order correlation between nominal and effective rates = 0.88

B . Indices for selected items within this list

Livestock products Chemical & fertilizer mineral products Butter Cheese Canned fruits & vegetables Flour mill products Bakery products Processed tobacco carpets Lace products Jersey fabric Carpeting products Premanufactured wooden structures Papers for sanitary use Pharmaceutical preparations Paints, inks, & dyes Leather Soles and other shoe components Mining machinery Steel structures Hand twls Farm machinery, except tractors Textile industry machinery Industrial furnaces Business & office machines Domestic kitchen appliances Washers Fans & other domestic appliances Domestic refrigerators Trucks Household radio & TV sets Motorcycles, bicycles, & parts Wood furniture for homes Source: Morales, Ulloa, and Jimenez (1978), table 7

.I7 .27 .81 .67 .62 .29 .42 .97 1.18 .91

.I1 1.13 1 12

.67 .I7 .36 .65 .97 .07 .32 .24 .I1 .09 .I0 .38 .48 .79 .76 .37 .93 .53 .43 .94

.14

.38 4.83 2.33 1.81 .77 .58 4.29 5.83 1.88 2.15 2.33 2.36 1.67 .19 .65 1.72 1.89 .05 .49 .30 .09 .01

.05 .53 1.03 2.54 2.24 .61 2.82 .91 1.30 1.60

209

BoliviaKhapter 4

the public sector. Although there are no good data for this illegal activity, a fair guess for the late 1970s was that contraband imports constituted around 20 percent of legal imports. The expansion in contraband imports after 1978 is also related to the laundering of dollars earned in the drug trade.3 4.3

Economic Integration

Bolivia has adhered to two main economic integration schemes, the Latin American Free Trade Association (LAFTA), which later became the Latin American Integration Association, and the Andean Group, as well as to a host of other organizations of economic cooperation with less ambitious aims. Bolivia joined LAFTA in 1966 and was given a relatively less developed country status with preferential treatment that consisted essentially in postponing dates for the implementation of tariff reductions and dismantling nontariff barriers. The direct benefits of Bolivia’s association with LAFTA were virtually unnoticeable. Bolivian exports to the countries consisted mainly of petroleum, natural gas, and minerals. These exports, however important, would have taken place anyway, with or without LAFTA membership. Bolivian imports of manufactures from LAFTA grew at a very fast pace, but this expansion can hardly be attributed to its participation in the organization. The apparent failure of LAFTA, at least from the viewpoint of the relatively more poor Andean member countries, led to the formation of the Andean Group with the signing of the Cartagena Agreement in May 1969. The Andean Group integration scheme had two main instruments: (1) a customs union, and (2) a joint mechanism of investment programming for a list of goods for the Sectorial Industrial Development Program (SIDP). In addition, in order to counteract the adverse effects that these instruments might unintentionally provoke, the Andean Group countries agreed upon a set of measures to harmonize other policies that affected trade and agreed to set common policies for the treatment of foreign private investment. Bolivia was again given a relatively less developed country status in the group, along with Ecuador, and both were accorded preferential treatment for the two main instruments and subordinate policies. Economic integration within the Andean Group created considerable hope among Bolivian policymakers, who thought that it would provide the big push necessary for Bolivian industrial development with the incentive of a large market for manufactures. Bolivia, therefore, enthusiastically supported the Andean Group at the outset. By 1978 there was considerable disillusionment with the workings of the Andean Group among government officials and industrialist organizations in Bolivia. From their point of view, the benefits of integration seemed rather scant and the costs were presumed to be high. The fact that the whole

210

Juan Antonio Morales and Jeffrey D. Sachs

Andean Group entered into a state of crisis contributed to the problem. Chile, with the most healthy of the Andean Group economies, in fact withdrew from the group in the mid-1970s under the policy of the Pinochet regime. The Andean Group has continued in prolonged crisis, a crisis which deepened markedly with the international economic turmoil of the 1980s. In the aftermath of the hyperinflation, with public policies dominated by the need to consolidate the stabilization, Bolivia’s participation in the Andean Group and in all the other integration schemes is almost dead. Notwithstanding this, the collapse of the markets for traditional exports may inspire Bolivian policymakers to seek some fresh approaches to economic integration, especially with Brazil and Argentina, which represent large potential markets for light industrial exports from Bolivia.

4.4 Exchange Rate Policies Between 1957 and 1982, Bolivia followed a regime of unified pegged official exchange rates. The abundance of credits from 1957 until the late 1970s allowed the government to maintain a fixed exchange rate without the need to resort to explicit foreign exchange rationing, and thus prevented the development of a parallel market with significant premiums. Between 1957 and 1979, the exchange rate showed a surprising stability: only once, in October 1972, was the peso devalued. After the drying up of foreign inflows in the early 1980s and with the resistance of the government to undertake timely devaluations, the economy operated with what was in effect a dual exchange rate: an overvalued and rationed official rate and a floating, parallel (sometimes illegal) rate. After 1985 the exchange rate was again unified and operated as a managed float. On some occasions during the 1960s and 1970s, foreign exchange reserves fell significantly, prompting policy measures to avoid an outright devaluation through hidden or explicit rationing of foreign exchange. Various temporary trade policy instruments were used for this purpose. On at least two occasions, a uniform increase in import tariff rates was used as a substitute for devaluation from the import side: in 1969, an almost uniform surtax of 10 percent was levied on all imports; in 1975, another surtax of 3 percent was created. Export subsidies for minor exports were also used to compensate for overvaluation in 1977. However, the percentage of trade that benefited from those subsidies was very small. As mentioned in section 4.2, quantitative restrictions were also used for balance-of-payments purposes. For instance, in 1969 the restoration of external equilibrium was obtained with temporary prohibitions on the imports of automobiles and of luxuries. A new tool in the kit of import controls was introduced in 1976 in the form of prior import deposit^.^ It is important to note that these deposits were both a monetary measure and a tariff-like regulation raising the cost of imports. Because of both features,

211

BoliviaChapter 4

they were initially very effective in curtailing imports. However, to the extent that importers could roll over their deposits, the monetary contraction aspect was lost, except when there were increases in the level of imports. Thus fiscal and, to a lesser extent, monetary measures were used to avoid open devaluations of the peso in the 1960s and the 1970s. In accordance with the spirit of the times, devaluation was viewed as a declaration of failure in economic policymaking. General Banzer, w h o had to go through a devaluation in 1972, paid the costs, political and otherwise, of very painful adjustments in the economy distributed over more than a year after the devaluation. Although never publicly declared, a widely held opinion in government circles at the time was that the boom in export prices in 1973-74 saved Bolivia from a string of further devaluations. The devaluation of 1972 deserves some additional attention. Since the end of 1969 when the assets of the Bolivian Gulf Oil Corporation were nationalized, pressures on the peso had been building up. In 1970 the government decided to impose some mild administrative regulations on the convertibility of the peso; for instance, requiring a full registration in the Central Bank of demanders of foreign e ~ c h a n g eThese . ~ regulations were not sufficient to avoid the drain on foreign exchange reserves of the Central Bank. By the end of 1972, it became clear that a devaluation was unavoidable. The IMF was called for consultations, and Bolivia applied for a standby loan. The peso was devalued by 40 percent, and some public sector prices, as well as interest rates on savings deposits, were increased. Workers obtained a uniform compensation of $b 135, equivalent to U.S. $7 (1972 dollars), at the new rate of exchange. After the devaluation, many prices were subject to controls and fixed at their pre-devaluation levels. Some of the prices were revised upward only in October 1973 and the rest in January 1974. Strong excess demand conditions made the revisions unavoidable. In table 4.4 we show how the peso incurred a significant real appreciation vis-B-vis the U.S. dollar during 1973-84. The relatively long period of overvaluation had important implications for resource allocation. In the mining sector, the combination of overvaluation plus punitive taxation shifted resources from there to the nontradable manufacturing sector and the service sector. Overvaluation also encouraged the expansion of the very capital-intensive activities of tin smelting and oil refining. Traditional exports and nontraditional ones, such as commercial agriculture, suffered. If overvaluation hurt exports, one may wonder why the issue was not debated more fully at the time or why there was not a significant lobby to push for a devaluation. The following reasons may be hypothesized. First, oil and mineral exporters can usually live with overvalued exchange rates until the rates are severely misaligned. Given their cost structure, exporters usually place more emphasis on lessening the weight of direct taxation than on the exchange rate to maintain their after-tax profitability. Second, the high

212 Table 4.4

Juan Antonio Morales and Jeffrey D. Sachs The Real Exchange Rate in Bolivia, 1970-84 Year

Index (1970-100)

1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984

100.0

95.9 88.3 75.7 110.3 110.6 109.2 109.0 113.0 121.2 125.6 146.9 125.7 125.0 164.6

Source: IMF data. Note: The index is constructed as PIEP, where P is the Bolivian consumer price index, E is the exchange rate (pesos per dollar), and P is the U.S. CPI. For each year, annual averages are used for the three indexes. Note that E is the official exchange rate; in the 1980s there was a large and persistent gap between the official exchange rate and the parallel market exchange rate.

prices for the main exports, well above previous trend, concealed the need to correct the exchange rates. Although profit margins in the exporting sector decreased with overvaluation, they were still very high in mineral, oil, and gas exports. It was not fully realized that overvaluation hindered the expansion of potential exports. Since no significant actual exports were greatly damaged by overvaluation, no political lobby was established to gain a better price for the dollar earned in the exporting activities. Also, hopes for exports of manufactures were riding on the Andean Group market, and little attention was paid to the development of other markets. Markets in the Andean Group were protected by a relatively high common external tariff, while trade liberalization within the group benefited mainly noncompetitive imports from the partner countries. In those circumstances, overvaluation, if not severe, was not the major hindrance for export promotion of manufactures to the protected market. In the event, however, that market turned out to be much too limited to support much manufacturing export activity in Bolivia. The hypothesis that overvaluation constituted a fiscal measure to extract resources from the hard-to-tax public enterprises also has to be taken into account. The weakness of the central government vis-u-vis the public enterprises, and especially the inability of the central government to tax the state enterprises directly, may explain one attraction to overvalued exchange rates. Such rates permitted the transfer of resources from the exporting sector, formed mainly by public enterprises, to the nonexporting public sector, formed mainly by the central government.6 The abrupt reduction in net foreign reserve flows in 1982, combined with the underlying budgetary disequilibrium, at first caused a rapid loss of

213

BolividChapter 4

reserves and a collapse of the fixed exchange rate regime in March of that year. The collapse was followed by a dual market with a fixed official rate for a handful of transactions and a floating rate for all other transactions, either of current account or capital account. Unexpectedly for the public authorities, the exchange rate depreciated very rapidly in the parallel market, causing an upsurge of inflation. The difficulties of managing the exchange rate during the high-inflation period and the unification of rates at a realistic level with the stabilization program of August 1985 are discussed more completely in chapter 5 .

4.5 Capital Flight The overvalued exchange rate and lax management of the public sector contributed to widespread capital flight in the 1970s and 1980s. Ugarteche (1986) and the World Bank (1985) give some estimates of capital flight based on the “errors and omissions” account in the balance of payment^.^ The average annual capital flight is estimated to have been as follows (in millions of U.S. dollars): 1971-75, $77.3 (4 percent of the 1975 GDP); 1976-81, $216.9, (6 percent of the 1981 GDP); and 1982-83, $106.2 (3 percent of the 1983 GDP). Bank deposits held by Bolivians in banks in the United States were estimated to be on the order of $400 million in 1985, amounting to around 10 percent of GDP. This is an extremely conservative estimate of offshore bank accounts, especially in view of the fact that it is easy to hide foreign ownership of bank accounts and since many accounts are held in non-U. S . banks. What were the forces behind capital flight? We have already mentioned that overvaluation coupled with expectations of devaluation is an important explanatory factor. In addition, three other factors deserve to be mentioned. First, illegal transfers to private individuals resulting from the mismanagement of public sector investments were likely to be exported to safe havens abroad. Similarly, subsidized loans, diverted from their intended uses, were placed in assets abroad where they could not be seized by the debt collectors. Second, fears of expropriation or of controls on the free movements of capital have motivated a substantial portion of capital flight. In this regard, one of the most negative effects of the dedollarization measure of 1982 was its impact on capital flight, since individual savers were left with an unsatisfied demand for deposits in the domestic banking system and had to look abroad for a safe vehicle for their savings.’ Third, earnings from the coca trade have surely generated extensive capital flight, largely for nonmacroeconomic reasons.

4.6 Conclusions on Poor Export Performance It is clear from our survey of trade policies in Bolivia that relatively little careful policy attention was given to the promotion of Bolivia’s export

214

Juan Antonio Morales and Jeffrey D. Sachs

potential. Traditional exports were seen as offering rents that could be distributed to other parts of the economy. Nontraditional exports were hindered severely by an inadequate exchange rate policy and by a range of fiscal incentives which really did not have much effect on the margin in the incentives to produce nontraditional exportables, Public investments in the tradable sector, as discussed in the previous chapter, generally were unprofitable and socially costly. They were motivated more by political considerations and easy foreign credit, rather than by a careful cost-benefit analysis. Finally, unwarranted policy hopes were held for export promotion within the context of regional integration schemes, particularly the Andean Pact. These regional schemes proved to be superfluous for Bolivia, not only because the target market remained very small even after integration, but also because the Andean countries almost all descended into deep crisis in the 1980s.

5

Aspects of Foreign Debt Accumulation, 1952-85

As was shown in table 1.8, Bolivia has depended significantly on foreign savings to finance gross capital formation since the late 1950s. The bulk of that foreign financing has come in the form of medium- and long-term (MLT) loans to the public sector, which is the category of capital inflow that we will examine in this chapter. Unfortunately, it is difficult to study the foreign debt of the Bolivian private sector because of a lack of adequate data, though available information suggests that the debt of the public sector is indeed by far the dominant form of external indebtedness.' It should be mentioned, however, that private nonguaranteed debt increased very rapidly in the crucial subperiod 1978-82, just preceding the extreme macroeconomic crisis. The measured short-term debt remained fairly constant over time, but the quality of the data on this type of debt prevents us from drawing any firm conclusions. The frequent shifts in the classification of the debt because of reschedulings, arrears, and the assumption of the debt of one sector by another during the past several years makes the analysis even more difficult. An historical view of Bolivia's borrowing can help to discriminate among the different factors responsible for the debt crisis. Bolivia had access to loans from official multilateral sources and from governments since the final years of the 1950s. These credits had a concessional element, the size of which decreased significantly over time. Already by the first half of the

214

Juan Antonio Morales and Jeffrey D. Sachs

potential. Traditional exports were seen as offering rents that could be distributed to other parts of the economy. Nontraditional exports were hindered severely by an inadequate exchange rate policy and by a range of fiscal incentives which really did not have much effect on the margin in the incentives to produce nontraditional exportables, Public investments in the tradable sector, as discussed in the previous chapter, generally were unprofitable and socially costly. They were motivated more by political considerations and easy foreign credit, rather than by a careful cost-benefit analysis. Finally, unwarranted policy hopes were held for export promotion within the context of regional integration schemes, particularly the Andean Pact. These regional schemes proved to be superfluous for Bolivia, not only because the target market remained very small even after integration, but also because the Andean countries almost all descended into deep crisis in the 1980s.

5

Aspects of Foreign Debt Accumulation, 1952-85

As was shown in table 1.8, Bolivia has depended significantly on foreign savings to finance gross capital formation since the late 1950s. The bulk of that foreign financing has come in the form of medium- and long-term (MLT) loans to the public sector, which is the category of capital inflow that we will examine in this chapter. Unfortunately, it is difficult to study the foreign debt of the Bolivian private sector because of a lack of adequate data, though available information suggests that the debt of the public sector is indeed by far the dominant form of external indebtedness.' It should be mentioned, however, that private nonguaranteed debt increased very rapidly in the crucial subperiod 1978-82, just preceding the extreme macroeconomic crisis. The measured short-term debt remained fairly constant over time, but the quality of the data on this type of debt prevents us from drawing any firm conclusions. The frequent shifts in the classification of the debt because of reschedulings, arrears, and the assumption of the debt of one sector by another during the past several years makes the analysis even more difficult. An historical view of Bolivia's borrowing can help to discriminate among the different factors responsible for the debt crisis. Bolivia had access to loans from official multilateral sources and from governments since the final years of the 1950s. These credits had a concessional element, the size of which decreased significantly over time. Already by the first half of the

215

BolividChapter 5

1960s, Bolivia had a debt-GNP ratio in excess of 30 percent and by 1970, a debt-GNP ratio of over 40 percent, as we see in table 5.1. The innovation in the 1970s was Bolivia’s renewed access to financial markets from which it had been cut off after defaulting on its public sector bond debt in 1931. This led to a marked shift in the structure of the debt, as is shown in table 5.2, from official sources to private sources, particularly to banks. The share of bank debt soared from 2.3 percent of the total to 38.9 percent in 1981. The first incursion of the commercial banks into lending to Bolivia in recent history occurred in 1972 when the government received credits from Citibank, Swiss Bank Corporation, and the Bank of America to compensate foreign owners of those firms nationalized by President Torres (see Ugarteche 1986, 150, and the references therein). Other credits with very expensive borrowing terms followed. There was also a significant rise in suppliers’ credits. In 1977 Bolivia was still in good standing with the international banks, but the spillover effects of problems elsewhere in the developing world had negative repercussions on borrowing terms, resulting in shorter-term loans and higher-risk premiums. By 1980 Bolivia faced a severe debt problem that had not yet been resolved by the end of the 1980s. The problems with the commercial banks are the gravest, but are not the only critical aspect of the debt crisis. In the early 1980s, the military regimes in Argentina and Brazil gave short-term financing to the generals in Bolivia. In 1983 this short-term debt was refinanced into a longer maturity, thereby sharply increasing the MLT public debt. This explains why, while net capital flows were negative in 1983, there was a sharp increase in measured MLT debt. If one looks at the conventional measures of overall indebtedness (see table 5.1), MLT public debt relative to GDP was already high in the 1960s. Indeed, the ratio of MLT public debt to exports in the 1960s and early 1970s was actually above the average ratio between 1974 and 1977. But, as we should like to stress, this observation neglects the fact that the nature of Table 5.1

Debt Indicators, 1970-87 (public and publicly guaranteed debt) Year

Debt-GNP Ratio

Debt-Export Ratio

1970 1975 1980 1981 1982 1983 1984 1985 1986 1987

48.2 48.6 77.0 88.0 93.8 111.8 111.9 107.7 105.1 110.8

231.9 166.6 214.5 272.1 313.4 367.0 401.8 478.7 596.1 743.4

Source: World Bank Debt Tables, (New Yo&: Oxford University Press), 1988-89 edition.

Juan Antonio Morales and Jeffrey D. Sachs

216 Table 5.2

Structure of Medium- and Long-Term Debt, 1970-85 (proportion of total debt outstanding and disbursed)

Official creditors of which: Multilateral Bilateral Private creditors of which: Commercial hanks“ Suppliers Other Total

1970

1975

1980

1985

56.7

60.0

51.5

59.9

5.4 51.3 43.3

13.8

22.1

21.7

46. I 40.0

29.4 49.5

38.2 40. I

25.6

42.1

31.3

8.4

6.4

2.8

6.0 100.0

100.0

100.0

17.3 9.8 16.2 100.0

1 .o

.o

Source: World Bank Debt Tables, 1988- 89 edition. ”This category is “private financial markets,” and includes a small amount of bond debt not held by commercial hanks.

Bolivia’s debt had changed decisively, with a sharp increase in the share of debt owed to foreign commercial banks, a point that can be seen clearly in table 5.2. In part because of this change, the indicators of debt servicing (as opposed to indicators based on the stock of debt) suffered a persistent deterioration over time, as can be observed in table 5.3. It should be noted that the figures on interest service understate the contractual debt burden because they are based on actual payments and not on payments due, and since the portion of interest in arrears is very important, especially in the 1982-85 subperiod. The increasing debt-service ratios reflect the change in the nature of indebtedness and, more precisely, the shift from debt with a large concessional element toward debt on market terms, in a situation in which market interest rates were rising sharply. The debt-service indicators relative to export of goods and services worsened progressively in the 1970s, compared to the values at the very beginning of the decade. In sum, Bolivia’s creditworthiness improved very significantly in the first half of the 1970s, if creditworthiness is defined as access to market lending. The extraordinary upsurge of exports (and its effects on real GDP) in the 1970s created an illusion in regard to the long-run economic prospects of a country that had been, before this event, very dependent on foreign aid and loans with highly concessional terms. By 1980, however, that illusion had been shattered. Bolivia’s creditworthiness disappeared once again, and Bolivia found itself in a debt-rescheduling exercise with the banks, two years before the outbreak of the global debt crisis. 5.1 The Nature of Borrowing by the Public Sector in the 1970s

The big push for debt accumulation appeared between 1975 and 1980, as can be seen in table 5.4.Who were the beneficiaries of the growing external

217

BolividChapter 5 Debt-Service Indicators, 1970-87 (public and publicly guaranteed debt)

Table 5.3 Year

Debt Service-Export Ratio

Debt ServiceGNP Ratio

1970 1975 1980 1981 1982 1983 1984 1985 1986 1987

11.3 15.3 27.9 27.7 31.5 32.2 36.4 34.2 23.1 22.1

2.3 4.5 10.0 8.9 9.4 9.8 10.1 7.7 4.2 3.3

Source: World Bank Debt Tables, 1988-89 edition. Note: The data refer to total debt servicing on medium- and long-term public and publicly guaranteed debt.

Table 5.4

Debt Outstanding and Disbursed, Medium and Long Term, 1970-85 (in millions of U.S. dollars) 1970 1975 1980

1981 1982 1983 1984 1985

480 824 2,228 2,765 2,861 3,279 3,386 3,484

Source: World Bank Debt Tables, 1988-89 edition

debt? Among the public enterprises, the state oil company, YPFB, was probably the major recipient, followed by the smelting company, CMK (Complejo Minero Karachipampa), and the state mining company, COMIBOL. Loans contracted by the specialized state banks, with public sector guarantees, were channelled to the private sector producers. The stock of debt owed by the central government also grew very rapidly between 1975 and 1979. The big increase between 1979 and 1981 was caused essentially by a transfer of a debt from COMIBOL (and other less important enterprises) to the TGN. The important question, of course, is why the Bolivian government increased its indebtedness so rapidly at the end of the 1970s. As was discussed in earlier chapters, the data suggest that most of the debt was related to the rapid growth of public investment projects, which in turn were linked to a complex of political and economic factors. We have identified several of those factors at length in earlier chapters, including (1) the very short time horizon of Bolivian governments; (2) the use of state enterprises as a vehicle for political control; (3) the use of state enterprises as a conduit for channelling public money to favored parts of the private sectors; (4) the soft budget constraint of the state enterprises, which reduced the incentives

218

Juan Antonio Morales and Jeffrey D. Sachs

to monitor investment projects; (5) the overvaluation of the exchange rate, which led to a misallocation of investment spending into highly capitalintensive projects and which increased the budget deficits of the public enterprises; (6) the use of state enterprises as buffers for macroeconomic shocks; and so forth. And as we have seen, the mega-investment projects of the public sector in the end failed to pay the necessary returns. In addition, there was certainly a misjudgment about the country’s true macroeconomic situation, as well as a failure to predict (along with the rest of the world!) the sharp swings that were to take place in the international economy at the end of the 1970s and the early 1980s. One part of the misjudgment came from the fact that Bolivia’s strong economic performance in the 1970s reflected a temporary terms-of-trade improvement and the effects of the foreign loans themselves, rather than a true underlying improvement in the economy. The annual average terms of trade of 1976-80 was 22 percent higher than the average in 1971-75. This meant a positive real income effect of terms-of-trade change of 6.2 percent of GDP for the second half relative to the first one. This improvement turned out to be temporary, though the borrowing behavior implicitly assumed that it was permanent. With respect to the international environment, the low real interest rates on international loans were perceived to be permanent, when of course they turned out to be temporary. As stressed by Morales and Sachs (1989), this change in the interest rate environment helps to account for the fact that overborrowing (and overlending by the banks) was a common feature of the entire world at the end of the 1970s.

5.2 Private Nonguaranteed Debt In 1985 the private nonguaranteed debt was 8 percent of the total external debt ($314 million, or approximately 8 percent of GDP). While the amount is modest, it grew very rapidly in the late 1970s and early 1980s. Several channels were used by the private sector to contract nonguaranteed debt. Credits were given directly by the international commercial banks to the debtors, or they were intermediated by their local subsidiaries or by the domestic banks. A significant fraction of the nonguaranteed debt is actually suppliers’ credit from foreign manufacturers to their authorized domestic agents in Bolivia. The participation of the domestic banks in the marketing of foreign credits increased, somewhat surprisingly, when a hardening in the borrowing terms occurred and the international banks and their subsidiaries were reducing their direct exposure. The implicit assumption on the part of the Bolivian banks may have been that there were de facto public sector guarantees on the private sector debt. Unfortunately, there is no available information on the final users of the loans granted by the banks. We can, however, make a strong presumption

219

BolividChapter 5

that a high percentage of the loans were used in the service sector and in other nontradable sectors. The presumption is based on the fact that when depreciation of the peso accelerated, the delinquency rates in the banks increased considerably. In October 1982 it was reported that 40 percent of the banks’ portfolios were technically in default (World Bank 1985, 528). In fact, the debt situation is at least partially behind the measure of “dedollarization,” to be mentioned in chapter 6, which allowed the private sector to repay at a highly favorable interest rate the dollar-denominated debts incurred by the domestic banking system. Dedollarization and exchange controls also erased the effective distinction between private and public debt after 1982, since the foreign debts of the private banks were to be repaid in pesos to the Central Bank, with the Central Bank then required to honor the international obligations. The foreign debt of the domestic private banks with the international banks continues to be a thorny issue because of several legal and financial disputes between the banks and the government. No information is available on the service burden of the private nonguaranteed debt. 5.3

Short-Term Debt

The data on short-term debt are particularly poor. With the scant information available, we can obtain only a rough picture of what has happened since 1978. In view of the increasing difficulties of the economy after 1978 and the hardening of terms on long-term bank lending, the governments have increasingly resorted to short-term loans from other central banks in the region and from commercial banks. Balance-ofpayments problems (i.e., dwindling reserves as a result of fixed exchange rates, domestic inflation, and large budget deficits) prompted the appeal for swap credits from foreign central banks, while the state-owned Banco del Estado contracted commercial debts to channel them to the private sector. The worsening of the situation in 1980 caused a delay in payments of short-term commercial credits. This led to a refinancing agreement in April 1981, to which we refer below, that converted about one-third of the public sector’s short-term loans into medium-term loans. Unfortunately, the refinancing did not substantially reduce the outstanding short-term debt, since the government made appeals to other sources for more short-term loans: Central Bank swaps and reciprocal trade credits in the context of the economic integration scheme of ALADI. In 1983 a large part of short-term loans owed to Argentina and Brazil were converted to medium-term credits. In spite of the conversions to MLT debts, the stock of short-term debt continued increasing on account of arrearages. In 1980, short-term debt was 1 1 . 1 percent of the total external debt; in 1982, it reached a low of 5.8 percent; and in 1985, the percentage was 8.7 percent. Short-term debt was $347 million in 1985, or almost 9 percent of GDP.

220

Juan Antonio Morales and Jeffrey D. Sachs

5.4 The Hidden Costs of Easy Credit Availability Since the early 1960s, the prime focus of official policy had been to increase economic growth. The emphasis on this objective grew even more from 1973 to 1978, both as a product of design and as a result of very favorable external circumstances. The greater availability of foreign credits coincided with, and was in addition spurred by, a substantial increase in export earnings. Any source of financing for capital expansion was welcomed at that point. The government had long tried to cater to foreign direct investors with generous fiscal incentives, but the response was weak outside of the petroleum sector. Thus, as a substitute, it started to court foreign lenders, who obliged as part of the world credit expansion of the 1970s. Banzer’s government assured political and monetary stability, and the government offered, after some hesitation, exchange rate guarantees. This, and an ample supply of development projects, however poor in design or implementation, sufficed to induce a very significant flow of external resources. The increase in the contribution of foreign loans to financing domestic investment, had, however, the cost of further impairing the administration of the public sector. The easy recourse to indebtedness weakened the budget constraints and indeed allowed the government more leniency in fiscal policies and on the exchange rate than there would have been otherwise. Moreover, the undemanding fiscal attitude was aggravated when access was gained to commercial bank credits that were not tied to specific projects, in contrast to the case of official loans which almost always were based on specific projects. The access to loans on relatively easy terms also impeded the design of needed reforms in the financial sector, particularly in the banks. The intermediation of foreign loans, contracted by the government or with its guarantee, was a more important source of profits to the private banks than the lending out of deposits made by domestic wealthholders. The neglect by the financial intermediaries of the local demand for deposits in the banking system likely had a negative longer-term effect on the mobilization of domestic savings. The foreign loans were channelled to public investment projects and to the private sector through “refinancing” mechanisms. We have noted at length in the previous chapter that poor project design and, especially, poor implementation, led in many cases to results incommensurate with the resources that had been put out. A significant share of the increase in public indebtedness was due to these factors. Loans that were channelled to the private sector did not fare better in regard to results, as they were diverted to speculative uses and, frequently, never repaid. In many notorious cases, private lenders simply defaulted to the state banks that had channelled the external credits, and the government made no attempt to collect on the bad debts. This increased the demand for loans and, ironically, there was a ready supply to match.

221

BolividChapter 5

A special mention needs to be made regarding exchange rate policies. Overvaluation coupled with punitive taxation in the natural resources sector hampered investment in minerals, hence reducing the future supply of foreign exchange to honor the foreign debt obligations. In addition, overvaluation led to an expansion of imports of consumer durables that was again financed directly and indirectly with foreign debt. Lastly, as the reserves of foreign exchange became precariously low, partly because of the currency overvaluation and the looming budget deficit, there was an increase in capital flight. Because of the absence of sound macroeconomic policies and the poor administration of the loans intended for investment, many loans simply became consumption loans in 1978-8 1. Their servicing would later require a drop in consumption, as indeed happened. In the transition, however, more indebtedness was accumulated to refinance old loans and their interest charges. By 1980 Bolivia was already a highly indebted country, as indicated, for instance, by a debt-to-GDP ratio of 76 percent. It was then subjected to the sharp international interest rate shock. Arrears on amortizations of loans granted by private creditors started to build up. In spite of a debt rescheduling in 1981, the debt situation became aggravated. As seen in table 5.5, from 1982 on, the net foreign resource transfers (net new lending minus interest payments) turned negative and a shift from external sources of finance to internal sources occurred, throwing the country onto the path of hyperinflation.

5.5 Debt Management, 1970-85 In this section, some of the more important institutional features of the debt management are presented. One interesting feature is Article 56 of the Net Foreign Transfers on Medium- and Long-Term Public and Publicly Guaranteed Debt, 1970-87

Table 5.5

Year I970 1975 1980 1981 1982 1983 1984 I985 1986 1987

Net Resource Transfers ($ million)

NUT as Percentage

31 90 178 92 -61 - I35 - 102

3. I 5.3 6.2 2.9 -2.0 -4.6 - 3.3 -3.6 4.8 1.7

-116

184 72

of GNP

Source; World Bank Debt Tables, 1988-89 edition. Note; Net resource transfers are defined as new lending minus total debt servicing (amortization plus interest). All are measured on a cash-flow basis.

222

Juan Antonio Morales and Jeffrey D. Sachs

Constitution of the Republic of Bolivia, which requires congressional approval of all loans contracted by the government of Bolivia, the public enterprises, and all public institutions, or on their behalf. Strictly speaking, the loans contracted by unconstitutional governments are also unconstitutional. It should be recalled that during most of the period 1964-82, Bolivia had only de facto government^.^ Until 1974, government agencies and public enterprises negotiated foreign credits, which were furnished principally by official lenders and by suppliers of capital equipment. Central government agencies, municipalities, regional development corporations, and committees for public works themselves engaged in the search for foreign credits. The grant element in the official foreign loans was usually so important that the higher levels of the government, cognizant of this, approved and included them in the fiscal budget. In contrast, suppliers’ credits were systematically seen as too expensive and frowned upon, but the purchasers of goods with suppliers’ credits were usually in a position to overcome the objections of the Ministry of Finance and the Central Bank. In fact, the growth of suppliers’ credits led to increasing regulation in the mid-seventies. The haphazard way in which the borrowing took place before 1974 was such that there are no good records on the state of the public debt and, a fortiori, there was no policy on indebtedness, except the weak objections to suppliers’ credits, mainly under the advice of foreign agencies of international cooperation. This situation, and the large expansion in international credit between 1973 and 1978, led to two important regulatory measures. First, the National Institute for External Financing (INDEF) was created in 1974 as a decentralized government agency to keep track of all indebtedness incurred in the public sector through its financial and nonfinancial institutions, to negotiate new loans, and in specific cases to refinance old loans. INDEF was to be especially active in obtaining general balance-of-payments support loans. In 1974 a decree was passed establishing the National System of Projects. The objective was to create a set of agencies, in a hierarchical structure, to help in the generation of investment projects, to perform social cost-benefit analysis of them, and to oversee their implementation. The search for sources of financing of the projects was also included among the functions of the system. Large investment projects and their financing required final approval by the National Council of Economic Planning (CONEPLAN), composed of several cabinet members and undersecretaries. Before projects arrived at CONEPLAN, they were to be screened by the National Committee of Projects and its technical secretariat^.^ Very few investment projects followed the steps spelled out in the 1974 decree. The process of approval was slow and cumbersome, and the technical secretariats lacked competent personnel. Managers of government

223

BolividChapter 5

agencies frequently ignored the guidelines of the National System of Projects. More importantly, vested interests were able to circumvent the procedures and get their projects to CONEPLAN directly. Several reforms of the National System of Projects were proposed between 1976 and 1986, but to no important effect. INDEF ceased to exist in 1979, and its functions were transferred to the Central Bank, where a division on external finance (FINEX) was created. The general objectives of FINEX were very similar to those of INDEF. FINEX had the advantage of being part of the Central Bank where it could enjoy better information support. Unfortunately, FINEX objectives were not met, as it suffered from acute political interference. The Ministries of Finance and Planning frequently intervened in FINEX negotiations. Very often, debt negotiating committees were formed by making appeals to private Bolivian bankers with international connections and neglecting FINEX (a case in point is the debt rescheduling of April 1981, discussed in sec. 5.6). The situation worsened during Siles Suazo’s administration, when almost everybody in the Cabinet felt obliged to intervene in debt negotiations. For instance, crucial debt reschedulings with Argentina and Brazil were carried out by the Ministers of Defense and Foreign Affairs. A very personalistic and ad hoc style of debt negotiations developed, with a significant loss of institutional memory of past negotiations, which was important since many agreements were verbal. Bolivia, as well as its partners, suffered considerably from the rapid turnover in negotiators. The need to redress the confusion in the external debt accounts of Bolivia, as well as the need to have updated information on debt negotiations carried out by other countries in the same predicament, called for foreign expertise. Several debt consulting firms and personal consultants were hired.’ The consultants were, however, less helpful than expected, although it must be said in their behalf that they were contacted at a point when no solution was really in sight. If the determination of the correct size of the public external debt was a formidable task, keeping track of the foreign debt of the Bolivian private sector was even more difficult. It is worth underscoring that a good knowledge of the size and structure of the private debt was essential for the design of policies aimed at the restoration of external equilibrium. Moreover, although this was not presumed at that time, most of the private debt eventually became “nationalized” with the dedollarization measure of November 1982. In 1979, ceilings on the stock of short-term debts, both private and public, were set. The obligation to register the debts of the private sector in the Central Bank was also established, but unfortunately this obligation was not tightly fulfilled. The lack of adequate information on the debt of the private sector created needless problems after dedollarization.

224

Juan Antonio Morales and Jeffrey D. Sachs

5.6 Debt Reschedulings, 1980-85 The debt crisis of the 1980s commenced with a small crisis in 1980, during the government of Luis Garcia Meza. Arrears in amortizations to the commercial banks eventually led to a restructuring of the debt in April 1981, which was crucial for triggering further developments. The eventful years of 1980-81 had also led to an abnormal growth of short-term credit to the generals in Bolivia from the generals in Brazil and especially Argentina. Arrears on those debts were also refinanced in 1983 with important implications. The chronology of events that led to the April 1981 agreement has been described by Baptista (1985) and Rivas (1986). Bolivia had ceased to make amortization payments after the Garcia Meza coup in July 1980. Because of this, the creditor banks gathered in Caracas in August, formed a consortium, and elected the Bank of America as the leader of a Coordinating Committee formed by the Bank of America, Bankers Trust, Deutsche Sudamerikanische Bank, American Express, Crocker National Bank, Libra Bank, Manufacturers Hanover Trust, Texas Commerce Bank, Bank of Nova Scotia, and Irving Trust. The Coordinating Committee represented 128 creditor banks. In the same Caracas meeting, an agreement was reached to reschedule debts with repayments due between 28 August 1980 and 5 January 1981. In January 1981, the government of Bolivia was to meet again with the Coordinating Committee. No agreement was reached in January except on the need to meet again in April 1981. Notwithstanding, Bolivia made some bona fide payments. After all these postponements, the April 1981 meeting took place in New York. Debts to the consortium amounted to U.S.$722 million (or 19 percent of GDP), of which U.S.$457.3 million were to be rescheduled in four tranches. The April agreement consisted essentially of a conversion of short-term loans in arrears to a medium-term loan and a reprogramming of medium-term l o a m 6 The rescheduling called for a 10 percent down payment of the refinanced loans according to the original schedule of maturities. The April 29th refinancing has been very much criticized within Bolivia. The thrust of the argument is that Bolivia was overcharged, with its costs and conditions well above what other countries obtained at that moment or shortly after. A more telling point is that it was extremely unrealistic that Bolivia would be able to come close to achieving the terms of the agreement. It should also be noted that the agreement called for Bolivia to sign an IMF program which never occurred. The Coordinating Committee routinely asked for the fulfillment of this clause and, as routinely, waived it. Bolivia was unable to meet the terms of the April 1981 agreement and fell in arrears by September 1982. Several meetings took place to normalize the situation, but to no avail. A semblance of normality prevailed, however, during the term of Minister of Finance Flavio Machicado in 1983, when debt

225

BolividChapter 6

servicing was resumed. Bolivia finally declared a moratorium to the commercial banks in mid-1984. Bolivia was able to get some debt alleviation more easily with two bilateral creditors: Brazil and Argentina. A total of U.S.$716 million of short-term debt and principal on medium-term debt was refinanced with those countries in 1983 on relatively easy terms: a fixed interest rate of 8 percent, maturities between 8 and 10 years, and grace periods of 3 years. This scheme of refinancing was found acceptable by the Bolivian public and did not provoke the kind of criticism that the rescheduling with the commercial banks had received. The international organizations, on the contrary, objected to the status of “preferred creditor” that Argentina received because it was able to use the natural gas exports of Bolivia as collateral.

5.7 Developments After 1985 In chapter 8 we provide a detailed analysis of Bolivia’s debt renegotiations after 1985. The main point that can be mentioned here is that the government under President Paz took a very different approach to negotiations. Having inherited a unilateral suspension of payments on the bank debt from the Siles government, the Paz administration maintained the suspension and began negotiations with the banks addressed to a long-term solution to the crisis instead of another rescheduling. Some results were achieved in 1988 when Bolivia was able to retire approximately one-half of its commercial bank debt at a price of 11 cents per dollar. In 1989, more debt was retired and negotiations continued on the remaining debt that had not yet been repurchased.

6

The Emergence of Hyperinflation, 1982-85

The inflation in Bolivia in 1984 and 1985 was the most rapid in Latin American history up to that date and one of the highest in world history.’ During the first half of 1985, the inflation surged to an annual rate of about 26,000 percent (approximately 60 percent per month), and it reached an annual rate of 60,000 percent during May-August 1985. As shown in table 6.1, the inflation was brought under control in the second half of 1985, and then after a sharp jump in prices in January 1986, inflation was kept at low

225

BolividChapter 6

servicing was resumed. Bolivia finally declared a moratorium to the commercial banks in mid-1984. Bolivia was able to get some debt alleviation more easily with two bilateral creditors: Brazil and Argentina. A total of U.S.$716 million of short-term debt and principal on medium-term debt was refinanced with those countries in 1983 on relatively easy terms: a fixed interest rate of 8 percent, maturities between 8 and 10 years, and grace periods of 3 years. This scheme of refinancing was found acceptable by the Bolivian public and did not provoke the kind of criticism that the rescheduling with the commercial banks had received. The international organizations, on the contrary, objected to the status of “preferred creditor” that Argentina received because it was able to use the natural gas exports of Bolivia as collateral.

5.7 Developments After 1985 In chapter 8 we provide a detailed analysis of Bolivia’s debt renegotiations after 1985. The main point that can be mentioned here is that the government under President Paz took a very different approach to negotiations. Having inherited a unilateral suspension of payments on the bank debt from the Siles government, the Paz administration maintained the suspension and began negotiations with the banks addressed to a long-term solution to the crisis instead of another rescheduling. Some results were achieved in 1988 when Bolivia was able to retire approximately one-half of its commercial bank debt at a price of 11 cents per dollar. In 1989, more debt was retired and negotiations continued on the remaining debt that had not yet been repurchased.

6

The Emergence of Hyperinflation, 1982-85

The inflation in Bolivia in 1984 and 1985 was the most rapid in Latin American history up to that date and one of the highest in world history.’ During the first half of 1985, the inflation surged to an annual rate of about 26,000 percent (approximately 60 percent per month), and it reached an annual rate of 60,000 percent during May-August 1985. As shown in table 6.1, the inflation was brought under control in the second half of 1985, and then after a sharp jump in prices in January 1986, inflation was kept at low

226

Juan Antonio Morales and Jeffrey D. Sachs

Table 6.1

Monthly Inflation Rates, 1984-87

January February March April May June July August September October November December

1984

1985

1986

1987

9.6 23.0 21.2 63.0

68.9 182.8 24.9

11.8

33.0 8.0 .I 3.6

2.5 1.2 .7 1.6

47.2

35.7 78.5 66.3 16.5 56.5”

4.3 1.8 .7 2.3

4.1 5.2 15.0 37.4 59.3 31.6 61.1

- 1.9 3.2 16.8

1 .o

.4 ~

.2 .0

1.o

.6

.6

2.1

-.l .6

- .3

.8

Sources: 1/84-9/86, from national statistics; 10/86- 12/87, from IMF, International Financial Statistics.

‘Note that prices actually stop rising by September 9, though because of rapid inflation in August and the first week of September, September’s average price level is 56.5 percent higher than in August.

double-digit rates for the next three years. Since the Bolivian inflation is one of the only cases since 1950 of a “true” hyperinflation (by Cagan’s classic definition of monthly inflation exceeding 50 percent), it provides an important testing ground for alternative views of the dynamics of hyperinflation and of the design of anti-inflation programs.’

6.1 Origins of the Hyperinflation The proximate cause of the hyperinflation is the government’s loss of international creditworthiness in the early 1980s. We have noted that during the period 1975-81, various Bolivian governments relied heavily on foreign borrowing to finance government expenditures. The combination of a large buildup of international debt with domestic political instability, poor macroeconomic management, a weak tax system, and poor export prospects, precluded the Bolivian government from obtaining new international loans after 1981. When foreign capital inflows dried up in early 1982, the government did not raise taxes or cut expenditures, but rather substituted domestic credit expansion for capital inflows as the source of finance for the government. The rapid expansion of the money supply then set off the inflationary process. Thus the balance of payments played a critical role in the origin of the hyperinflation, but only insofar as it affected the rate of money creation. More evidence in support of this position is that after the beginning of stabilization, a massive deterioration of the Bolivian international terms of trade in 1986 did not reignite the hyperinflation since under the new policy rules, the terms-of-trade decline did not feed into the rate of domestic credit expansion to the public sector. The substitution of domestic credit expansion for capital inflows and the jump in real seignorage collection took place in the first half of 1982, almost

227

BolividChapter 6

nine months before the Siles government took office. But there is no evidence that the intens$cation of the inflation process involved a rise in seignorage collection after the beginning of 1982. (The simple correlation of inflation with the level of seignorage as a percentage of GDP is r2 = -0.001 for the quarterly interval 1982:l to 19853.) The time path of seignorage relative to GDP is shown in figure 6.1. The figure records the quarterly seignorage collection, measured relative to annual GDP. Therefore, to obtain the annual seignorage collection relative to GDP it is necessary to add the quarterly rates for a year. Seignorage is measured for each month as (M,- M,-,), where M is base money, and is then divided by an estimate of nominal GDP for the month. This ratio is then added for the months of the quarter, with the results shown in the figure. Surprisingly, it is difficult even many years later to uncover precisely the causes for the jump in money creation in early 1982, though the main culprit is almost surely a shift in the environment for foreign borrowing. The problem with nailing down a culprit lies with the disarray of Bolivian fiscal data during this period, a problem we have already encountered in chapters 4 and 5. We noted three kinds of problems inhibiting a clear assessment of the fiscal situation. First, most of the available data cover only the central government (the TGN) and not the consolidated public sector, including state enterprises, regional corporations, local government, state development banks, etc. This limited coverage is especially problematic for the

C

-

78

79

80

8.1

82

83

84

Fig. 6.1 Seignorage (per quarter, as percentage of annual GDP) Note: Annual seignorage earnings are the sum of quarterly earnings. Source: IMF, International Financial Statistics.

85

228

Juan Antonio Morales and Jeffrey D. Sachs

hyperinflation period, since there were large ad hoc transfers among the various parts of the government at that time. Second, there is great difficulty in interpreting on an ex post basis the accounting of arrears on debt repayments during this period. Third, there were (and remain) significant disagreements among the various parts of the government about the responsibility for portions of the foreign debt. At some points, for example, the Central Bank repaid foreign dcbts on behalf of the TGN, which the TGN refused to recognize. With these limitations in mind, it is still possible to piece together a rough interpretation of the rise in seignorage in 1982. Key fiscal data for the central administration for 1981-84 are shown in table 6.2. Notice the important fact that the TGN deficit rises from an estimated 5.7 percent of GNP in 1981 to an estimated 22.3 percent of GNP in 1982. This increase just barely exceeds in magnitude the rise in expenditures on “internal and external debt.” Most or all of the “internal debt” category in this period reflects Central Bank repayments of foreign debt on behalf of the TGN, so that the jump in debt repayments is almost exclusively related to foreign debt. At the same time that debt servicing jumps up, inflation starts to accelerate, with the result that real tax collections fall sharply, from approximately 9.4 percent of GNP in 1981 to 4.6 percent of GNP in 1982. This revenue shortfall is partially balanced by a simultaneous cut in the current and capital expenditures of the TGN. Notice that the combined expenditures on personnel, other services, materials, and fixed assets fall by 3 . 0 percent of GNP between 1981 and 1982. The importance of enlarged net debt-service payments for the burgeoning deficit in 1982 is also evident from the balance-of-payments data, which Table 6.2

The Evolution of Revenues and Expenditures, 1981-84 (as percentage of GNP)

Revenues Internal taxes Taxes on petroleum Other Expenditures Personnel Other services Materials Fixed assets Internal debt External debt Others Deficit

1981

‘1982

1983

1984

9.4 3.0 2.1 2.6 15.1 1.2 .7

4.6 1.8 .9

2.6 I .4 .4 .3 20. I 5.2 .3 .l .4 .3 10.8 2.4 17.5

2.6

I .5 I .0 1.8 1.3 1.6 5.7

Source: Ministry of Planning, Bolivia. ”Mainly central government transfers to state enterprises

.8 26.9 5.6 .4 1 .4 10.4 7.1 1.4 22.3

.o

.7 .6 1.1 33.2 8.0

.5

.8 .5 .4 2.2 20.8“ 30.6

229

BolividChapter 6

unfortunately cannot be matched exactly with the budget data. We saw in table 5.5 for 1980 and 1981 that net resource transfers to Bolivia were positive, meaning that net new borrowing by the public sector exceeded the level of interest payments on the public debt. In 1982, net new lending plummeted, so that the resource transfer to Bolivia (new loans minus interest payments) turned negative. As a percentage of GNP, net resource transfers toward Bolivia shifted from 6.2 percent in 1980 to 2.9 percent in 1981, -2.0 percent in 1982, and -4.6 percent in 1983. Were we to include the data on short-term debt, which is not collected by the World Bank in the case of Bolivia, the shift in net transfers would be even more dramatic, since an inflow of short-term credits in 1980 and 1981 dried up in 1982. As it turns out, real seignorage collection from 1982 to mid-1985 remains roughly constant, averaging about 12 percent of GNP each year. This does not reflect a stable path of government spending, taxes, and monetary emission during this period. Instead, the roughly constant seignorage collections hide a process of adjustment in which tax revenue collections all but collapse, while government spending is cut back sharply over time in the vain attempt to compensate for the falling tax collections. Once again, data problems at this point preclude a comprehensive account of this process. As we have mentioned, while there are reasonably acceptable accounts for the central government, data on the consolidated public sector is sparse. Nonetheless, relying on the central administration accounts, we can see much of the process at work. The government depended on three main forms of taxes during the period: internal taxes (mainly sales and income taxes), taxes on petroleum, and taxes on trade. Each of these taxes fell sharply in real terms during the period. Figure 6.2 records the collections from taxes during 1981-84 as a percentage of GNP (the category “other” includes trade taxes). Overall, revenues of the central administration fell from more than 9 percent of GNP in 1979 and 1980 to just 1.3 percent of GNP in the first nine months of 1985, before the new stabilization program went into effect (the preprogram period in 1985 is indicated by 85.1 in the figure). Upon institution of the stabilization program, tax revenues of the central administration jumped almost immediately to more than 10 percent of GNP (shown as period 85.2 in the figure), mainly through an increase in tax payments by the state oil company, YPFB. In view of the sharp decline in tax revenues, an increasing proportion of the central administration deficit was financed through fiscal credits from the Central Bank. The proportion of TGN expenditures covered by TGN revenues fell from 65.6 percent in 1979 to just 6.9 percent in the first nine months of 1985: 1979 1980 1981 1982

66% 60% 62% 17%

1983 1984 1985.1 1985.2

13% 8% 7% 134%

230

Juan Antonio Morales and Jeffrey D. Sachs

79 0

80 All

81 0 Income

82

83 0

Tariffs

84

85,l

85,2

A Petroleum

Fig. 6.2 Tax revenues as a percentage of GNP (by category of tax)

Note that the steep drop in revenues relative to expenditures in 1982 reflects the rise in expenditures on debt servicing in that year as well as the falling revenue collection. The reasons differ for the declines in the various kinds of taxes, and these reasons are worth stressing, since they help to explain the nature of the stabilization policies finally undertaken in August 1985. The drop in internal taxes (mostly income and sales taxes) is most readily explained. Indexation of the income tax system was not even attempted until an administrative change in March 1985, so that lags in collection combined with high inflation reduced the real value of tax collections substantially, in the way familiar from Tanzi (1977) and others. Many sales taxes were set at specific, rather than ad valorem rates, and the specific rates were adjusted with very long lags, if at all, and certainly not fast enough to keep up with accelerating prices under a hyperinflation. The decline in tariff revenues introduces some less conventional effects. Throughout the entire hyperinflation period, the official exchange rate was pegged by the Central Bank and was adjusted in steps with long lags to the underlying inflation. As the government resorted increasingly to fiscal credit creation by the Central Bank in order to fund expenditures, there was a constant pressure on foreign exchange reserves at the official exchange rate. Rather than maintaining a steadily depreciating but unified price of foreign exchange, the government maintained the official parity for long intervals and rationed foreign exchange. The persistent excess demand for foreign exchange at the official exchange rate of course resulted in an enormous premium for foreign exchange in an unofficial, illegal, but semi-tolerated black market. Table 6.3 shows the average premium on the

231 Table 6.3

BolividChapter 6 Percentage Gap Between Official and Black-Market Exchange Rate, 1980-86. In (quarterly average) 1980: 1 I1 111

IV

1981: I I1

III IV 1982: I

U UI IV

1983: I II 111

IV

.0

.o .o .o .0 .0 45.5 68.3 57.0 3.5 4.8 22.0 116.3

1984: I I1 111 IV 1985: I Il I11 IV 1986: 1 I1 III

327.4 69.8 309.7 302.7 330.9 321.0 798.1 8.0 5.8 2.1 .5

105.1

241.8 225.4

Source: Ministry of Planning, Bolivia.

black market relative to the official rate on a quarterly basis for 1980:l through 1986:3. In the month before the stabilization program, the premium was on the order of 1,400 percent! Since the stabilization program began, the exchange rate has been unified, with a small and declining deviation (which remains because of minor legal and administrative factors) between the official rate and the now-legal parallel market rate. With foreign exchange rationing at the Central Bank, progressively fewer import transactions went through legal channels and more imports came into the country via smuggling, outside of tariff control altogether. Under Bolivian law, all exporters are required to sell the foreign exchange from export earnings to the Central Bank at the official exchange rate. In sectors where smuggling was possible, the sharp divergence between the official effective rate for exporters and the black-market exchange rate made smuggling a highly attractive option (the smuggler converts the foreign exchange earnings from exports in the black market, where the number of pesos per dollar is much greater than at the official effective rate). A trenchant illustration is provided by the example of Peru, which, despite the absence of tin mines, became a tin exporter in 1983-85 on the basis of Bolivia’s smuggled tin. As dollar export earnings sold to the Central Bank at the official rate shrank markedly, the Central Bank had a diminished supply of dollars to sell to importers, who progressively turned to the black market in order to pay for their imports. The overall result was that a very high and apparently growing proportion of imports in the economy went unrecorded, and therefore largely untaxed, during the hyperinflation period. The story of YPFB tax collections is related. Just as the official price of foreign exchange lagged the domestic price inflation, so too the government altered the domestic price of petroleum products with a significant lag.

232

Juan Antonio Morales and Jeffrey D. Sachs

Indeed, petroleum prices were typically changed at the same time that the official exchange rate was devalued. As an example, consider the domestic price of a liter of gasoline, converted to U.S. dollars using the black-market exchange rate (the rate earned, incidentally, by smugglers who carried cheap petroleum from Bolivia to Peru across Lake Titicaca). During this period, the world market price was between $0.25 and $0.28 per liter, but the domestic price was as shown in table 6.4. Thus, on the eve of the stabilization program (which raised the internal petroleum prices to world levels), the internal price of petroleum was about one-sixth to one-seventh of the world market price. YPFB was obligated throughout the hyperinflation period to pay taxes to the central government on a percentage of its revenues from internal sales and external sales. With internal prices so depressed, the taxes on internal sales were also severely depressed. Moreover, because its profits on internal sales were so limited by the price ceilings on its output, YPFB also refused to pay taxes on its external sales, arguing that the revenues were necessary to meet operating expenses. It is much more difficult to offer a comprehensive account of expenditure behavior during the hyperinflation, for reasons to which we have already alluded. As a rough estimate, however, it seems that as government revenues diminished, real expenditures fell more or less in tandem to preserve a large but fairly constant deficit, since we know that the inflation tax, which was financing almost 100 percent of the cash-flow deficit, did not rise markedly above 12 percent of GDP from 1982 to mid-1985. It appears that the main casualties of the reduction in real spending were capital investment by the public sector and current central government expenditures on personnel and nonpersonnel services and materials. After mid- 1984, interest payments on the foreign debt also diminished sharply as the government built up large interest arrears. Public investment spending fell, according to the IMF, in the manner shown in table 6.5. Table 6.4

Domestic Price of Gasoline ($/liter), March-December 1986 March April May June July August September October November December

$.09 .07

.06 .05

.03 .04 .28 .21 .23 .23

Source: UDAPE,Bolivian Ministry of Planning. Nore: The price is calculated by dividing the peso price per liter by the black-market exchange rate (pesos/$).

233

BolividChapter 6

'IBble 6.5

Central administration State enterprises Total

Public Investment (percentage of GNP), 1980-85 1980

1981

1982

1983

1984

1985

2.9

3.1

I .9 4.6 6.5

1.8 3.0 4.8

1.5 2.6 4.1

1.8 1.4

3.2

Source: Unpublished IMF memorandum on Bolivia, 1985.

Current spending by the TGN evolved in a similar way, with overall spending on services and materials falling from 9.4 percent of GNP in 1980 to 6.8 percent in the first part of 1985.

6.2 The Dynamics of the Hyperinflation This section is devoted to a closer look at the dynamics of the hyperinflation. To organize the discussion, we begin with the simple and classic Cagan (1956) model. Real money balances (M,/P,) are written as an exponential function of expected inflation d,and inflation expectations adjust adaptively:

The government's reliance on seignorage taxation is denoted by s 7MIP, which may also be written asm n m ( n e ) ,with m = MIP, and r = P/P. In the steady state, s = n m ( n ) , which as is well known can be satisfied by rwo inflation rates, a low inflation rate nmin= nrnin($) and a high inflation rate 7~~~~ = nmax(s) for all s S Pax, the maximal seignorage rate. The maximal rate is achieved at the inflatipn rate n = l/b, with smax= m(l/b)/b. Given the relations in (1) and that M / P = s G s max , an economy starting at t = 0 will converge to mmIn(s) assuming that ~'(0)S T ~ ~ ~and ( sthat ) 1 > bX. In the important case in which s > smaxand in which the stability condition 1 > bX is satisfied, then starting from a finite expected inflation rate, actual inflation will rise steadily without bound. At any given moment, the actual seignorage s can be financed through a surprise inflation, in which actual inflation exceeds the expected inflation rate. But this continuing gap between actual and expected inflation continues to drive up the expected inflation rate, which in turn increases the actual inflation rate necessary to achieve the given level of seignorage. Inflation proceeds to increase explosively and without bound. One key property of the Cagan model should be stressed: starting from a steady state with no = nmin(so), a permanent rise from so to s1 will lead to a jump in n on impact, and then to a steadily rising inflation rate. If s1 is less

+

234

Juan Antonio Morales and Jeffrey D. Sachs

than smm,then a new higher steady state, IT"""(s,), is reached. If instead s1 is greater than Pax, then inflation jumps up on impact and continues to rise without bound. In either case, a period in which inflation is rising need not signify a rising financing need (i.e., a continuing rise in s), but rather a lagged adjustment of the inflation rate to an earlier once-and-for-all increase in s. This essential aspect of the model characterizes the Bolivian experience, as we have already seen in figure 6.1. Aside from seasonal spikes in seignorage in the fourth quarter of most of the years (for budgetary reasons associated with a Christmas wage bonus), there is basically one permanent step increase in the seignorage collection in 1982, which persists until the third quarter of 1985. This model can easily be extended to the open economy by postulating a law-of-one-price (or purchasing-power-parity) relation between domestic prices, foreign prices, and the exchange rate. Specifically, we can assume P, = P T - E,, where P* is foreign (dollar) prices and E is the exchange rate in units of pesos per dollar. Given the path of foreign prices and seignorage demands, we can solve for domestic prices and the nominal exchange rate. This simple extension model of monetary and price dynamics introduced earlier must be modified in one important respect in the case of Bolivia. We have seen that throughout the hyperinflation period, foreign exchange was rationed at the official exchange rate, so that a large black-market premium was present throughout the period. In order to understand the dynamics of prices with a black-market exchange rate, it is necessary to alter the purchasing-power-parity relation to reflect the fact that the black-market exchange rate, rather than the official rate, best reflects the marginal cost of foreign exchanges for most imports during the period. For a limited subset of "necessary" imports, official foreign exchange was available from the Central Bank without rationing. The official exchange rate applied, for most of the period, to newsprint and pharmaceutical imports, for example. For all other goods, limited amounts of imports were purchased at the official exchange rate, but on the margin the cost of foreign exchange was dictated by the black-market price rather than by the official price. As a result, we should expect that the price level would be a weighted average of the official exchange rate ( E ) and the black market exchange rate (l?) price of foreign exchange, so that As before, monetary equilibrium requires MIP = m(rr"). Using the convenient Cagan form for money demand used earlier, assuming adaptive expectations for inflation in a discrete-time setting, IT:= IT;-, + (1) and A(n, - IT;-^), and taking the logs of ( 2 ) , we can write:

235

(3)

BolividChapter 6 pt - ye, - (1 - y)Zt = Xu - bhT,

+

(1 - A ) (Pr-1 - pr-11,

where a = log A , p = log M ,e = log E , p = log P,and log P* = 0. By rearranging, we get an estimable equation for the black-market premium (2 - e ) as an increasing function of (p - e ) , as well as a rising function of inflation and a negative function of (log) real balances lagged one period: (4)

( E , - e,)

=

-ha/(l - y ) + [ l / ( l - y)](p, - e,) + [bX/(1 - y ) I ~ r - “1 - X)/(1 - Y ) I ( P t - I - Pt-1)

An estimate of equation (4) using monthly data for January 1982 to September 1985 is strongly supportive of this monetary interpretation of the black-market premium. The estimated equation i s 3 (5)

(a, -

+

e,) = 3.03 l.lO(p - e ) (0.24) (0.08)

+ 0.57 T , (0.17) R2 = 0.84; D.W.

=

-

0 . 5 8 ( ~ --~p p 1 )

(0.10)

1.26; s.e. = 0.27.

The point estimate on (p - e ) of 1.10 suggests that the black-market exchange rate gets a weight y of 0.9 in determining the price level. This high weight is broadly consistent with direct estimates for prices, to which we now turn.

6.2.1 Price and Exchange Rate Dynamics The microdynamics of wages, prices, and exchange rates changed markedly over the course of the inflation. During the early 1980s, nominal wages were changed only a few times a year, and the presence of staggered nominal wage settlements gave inertia to the wage-price process. As inflation accelerated, wage contracts were renegotiated on a more frequent basis until, by the end of the hyperinflation, nominal wages were being reset on a weekly or biweekly schedule. As for prices, an increasing proportion of transactions became dollar linked in the sense that traders quoted prices in dollars, converting into pesos (at the parallel exchange rate) at the time of the transactions. It was illegal during this period to use dollars for transactions or even to quote items in dollars, so that in most parts of the country (with the important exception of the Santa Cruz region), dollarization stopped short of actual transactions in U.S. dollars. A precise specification of price-exchange rate linkages would open up a number of econometric and conceptual issues better left to another study.

236

Juan Antonio Morales and Jeffrey D. Sachs

Here, the process of encroaching “dollarization” is shown more simply in an equation linking monthly inflation to a one-month lag of inflation, the change in the black-market exchange rate, and the change in the official exchange rate. The equation to be estimated is: (6)

(P, -

&I)

= a0

+

+

a,(p,-,

a3kr

-

- PI-2)

(?,-,I

+

a,(& -

Zr-1)

All variables are expressed as monthly averages, where the monthly average exchange rates for month t are proxied by the geometric weighted averages of end-of-month rates for months t - 1 and t. There are two maintained hypotheses: (1) as inflation accelerates, the weight of the exchange rates in the price equations increases relative to the weight of the previous month’s inflation; and (2) as inflation accelerates and as the black-market premium increases (see table 6.3), the black-market exchange rate increases in importance relative to the official exchange rate. Thus, we expect that a,/(a, +a,+a,) and a,l(a, +a,) will both fall as the hyperinflation intensifies. Estimates of (6) are shown in table 6.6, first under an unrestricted estimation of the parameters, and then under the restriction a , + a2 +a, = 1. The hypotheses are supported in each case. The average monthly inflation rates for the three intervals are: 1983, 13.2; 1984, 31.4; 1985, 66.9. We see clearly that as we move from 1983 to 1985 the weight of lagged inflation (even at a one-month lag!) disappears, while the weight of the black-market exchange rate grows in importance. In the final eight months of the hyperinflation, price change is basically equiproportional to change of the black-market exchange rate. The combination of equations (5) and (6) illustrates the utter futility of the government’s policy of maintaining an overvalued official exchange rate under conditions of rapid money creation. The government’s resistance to devaluation was ostensibly owing to the fear of provoking even higher inflation, but it is clear that money creation fed into prices even in the absence of devaluation, through the mechanism of depreciation of the black-market exchange rate. The maintenance of an overvalued official exchange rate did nothing to stop inflation. On the contrary, it not only created large distortions in the economy (particularly by acting as a tax on legal exports), but almost surely raised inflation by increasing the government deficit via its negative effects on tariff collections and other revenues.

6.3 Failed Attempts to End the Hyperinflation As with many other hyperinflations, the end did not come on the first try, but only after several attempts at stabilization had failed. Stabilization programs were launched in November 1982, November 1983, April 1984,

237

BolividChapter 6

Table 6.6

Price Inflation Equations

1 . Unconstrained estimation (monthly) l983:l- 1983:12

.33 (.27)

.18 (.11j

R 2 = .42 D.W. .13

1984:l- 1984:12

(S2)

R2 1985:l-1985:9

=

(.17)

=

.32 (.II)

.58 D.W.

-.I8

.20 (.I3

.47

.13

.38

- .17

.14

.57

.37

.16

.37

- .20

.14

1.76 .52 (.19)

=

.46

2.80

.I7

1.04

(.13j

(.16)

R 2 = .87 D.W. = 2.20

2. Constrained Estimation l983:l- 1983:12

1984:l- 1984:12

.57

.I6 (.11)

R 2 = .42

D.W.

.I6

.31 (.lo)

R 2 = .71

1985:1 - 1985:8

-.20

.27 (.13j =

1.97 .53 (.11)

D.W. = 2.83

.17 (.11)

1.03 (.lo)

R 2 = .94 D.W. = 2.18

August 1984, November 1984, and February 1985. The most ambitious programs were those of November 1982 and April 1984. Notably, the April 1984 package represented a fairly orthodox approach to stabilization that included some of the measures that later proved successful in August 1985. The package included a very large devaluation of the official exchange rate, an announcement of tax reform, and a major increase in public sector debt. In the event, the trade unions exploded in furious opposition to the program, and a month later, the government acceded to demands for a large wage increase to compensate for the devaluation. At the same time, the government reversed its debt policy under the pressure of the trade unions and entered into a unilateral moratorium on further servicing of the external commercial banks debt. The Siles government gave up further attempts to negotiate a package of debt rescheduling. The complete breakdown of the April 1984 package eliminated any remaining hopes of the public that the Siles government would prove able to stabilize the economy. It should be recalled that in November 1984,

238

Juan Antonio Morales and Jeffrey D. Sachs

President Siles was compelled to announce elections for July 1985, one year ahead of schedule. A final attempt at stabilization came in February 1985, with a program that prompted a domestic march to La Paz and a month-long sit-in by 10,000 miners. Again the government capitulated to popular demands, after which it gave up even the pretense of attempting to stabilize before the arrival of a new government in August.

7

Ending the Hyperinflation, 1985-88

The end of the hyperinflation came swiftly, just three weeks after the new government of Victor Paz Estenssoro took office. A comprehensive stabilization program, labelled the New Economic Policy, was unveiled on 29 August 1985, and within days, the hyperinflation ended. Later in 1985, after several weeks of low inflation, there was another sharp run-up in prices associated with a large emission of money at the end of the year, but this blip in inflation was quickly brought under control in mid-January 1986.

7.1 The New Economic Policy The main features of the New Economic Policy are shown in table 7.1. Note the program embraced widespread liberalization of trade and finance, as well as fiscal austerity. In principle, the stabilization measures are “shortterm” measures, while the liberalization measures are oriented toward “long-run” growth. In fact, as we discuss later, the Bolivian government believes, with considerable reason, that the liberalization measures played a key role in permitting the stabilization policies to take hold. The key stabilization measures (putting aside, for the moment, the question of liberalization policies) had four basic elements, two of which were to be implemented immediately and two of which were slated for implementation in the months following the beginning of the program. First, the government committed itself to a policy of a unified exchange rate, without capital controls or exchange controls. The exchange rate was initially floated, though with a maximum value of the peso beyond which peso appreciation would not be permitted, and then was managed in a dirty float during the following year. Second, the fiscal deficit was immediately reduced through a combination of (a) public sector price increases, especially for petroleum products; (b) a public sector wage freeze; (c) further

238

Juan Antonio Morales and Jeffrey D. Sachs

President Siles was compelled to announce elections for July 1985, one year ahead of schedule. A final attempt at stabilization came in February 1985, with a program that prompted a domestic march to La Paz and a month-long sit-in by 10,000 miners. Again the government capitulated to popular demands, after which it gave up even the pretense of attempting to stabilize before the arrival of a new government in August.

7

Ending the Hyperinflation, 1985-88

The end of the hyperinflation came swiftly, just three weeks after the new government of Victor Paz Estenssoro took office. A comprehensive stabilization program, labelled the New Economic Policy, was unveiled on 29 August 1985, and within days, the hyperinflation ended. Later in 1985, after several weeks of low inflation, there was another sharp run-up in prices associated with a large emission of money at the end of the year, but this blip in inflation was quickly brought under control in mid-January 1986.

7.1 The New Economic Policy The main features of the New Economic Policy are shown in table 7.1. Note the program embraced widespread liberalization of trade and finance, as well as fiscal austerity. In principle, the stabilization measures are “shortterm” measures, while the liberalization measures are oriented toward “long-run” growth. In fact, as we discuss later, the Bolivian government believes, with considerable reason, that the liberalization measures played a key role in permitting the stabilization policies to take hold. The key stabilization measures (putting aside, for the moment, the question of liberalization policies) had four basic elements, two of which were to be implemented immediately and two of which were slated for implementation in the months following the beginning of the program. First, the government committed itself to a policy of a unified exchange rate, without capital controls or exchange controls. The exchange rate was initially floated, though with a maximum value of the peso beyond which peso appreciation would not be permitted, and then was managed in a dirty float during the following year. Second, the fiscal deficit was immediately reduced through a combination of (a) public sector price increases, especially for petroleum products; (b) a public sector wage freeze; (c) further

239

BolividChapter 7

Table 7.1

Outline of Major Policy Initiatives in First Year of New Economic Policy

Policy Area

Key Policy Initiatives

Implementation

Exchange rate

Unification of exchange rate on current and capital account; free convertibility of foreign exchange

Completed 1985

Public sector pricing

Public sector prices (most importantly, energy and food) raised to world levels

Completed 1985

Consolidated public sector budget

Target deficit of 6.3 percent GNP, of which 5.3 percent to be externally financed

Budget approved by Congress, May 1986

Import regulations

Unification of tariffs to flat 20 percent rate; elimination of trade quotas and nontariff barriers

Quotas eliminated immediately; tariff reform, August 1986

Private sector salaries and employment

Elimination of government restrictions covering private wages, except for national minimum wage; removal of restrictions on hiring and firing

Completed 1985

Private sector pricing

Elimination of all price controls, except for public transportation and public utilities; elimination of previous monopolies in intercity transport

Completed 1985

Public sector salaries and employment

Successive wage freezes (with one-step increases between freeze periods) during August-December 1985, December-June 1986, June-December 1986

Implemented

Employment reductions in state enterprises and central administration Decentralization of major state enterprises

Partial implementation

Public enterprises

Most actions not taken

Taxation

Increases in taxes paid by YPFB

In effect Approved by Congress, May 1986

International financial organization

Major consolidation and reform of internal taxes; establishment of VAT, patrimony taxes, and uniform income taxes Negotiation of IMF standby agreement

Foreign creditors

Approved by IMF, June 1986

Reestablishment of creditworthiness with World Bank and IDB Negotiation of Paris Club rescheduling

Terms of agreement approved, June 1986

Commercial bank debt

Negotiations extended beyond

Elimination of restrictions on commercial bank interest rates

Completed 1985

Current on all obligations

1986

Interest rates

240

Juan Antonio Morales and Jeffrey D. Sachs

reduction in spending on public investment; and (d) budget austerities in other areas, such as subsidized credits to the private sector. Third, the program called for a tax reform to reestablish and broaden the tax base. The reform was enacted by Congress nine months after the start of the program, and implementation began one year after. Fourth, the government announced its intentions of reestablishing the country’s creditworthiness with international financial institutions (official and private), especially the IMF, the World Bank, and the Inter-American Development Bank, while maintaining a suspension of payments to the commercial banks. Renewed creditworthiness required the government to eliminate arrears on its debt-servicing obligations to the international institutions and to negotiate an IMF standby agreement, which it did by June 1986. In June 1986, a Paris Club rescheduling was obtained. As described in chapter 8, Bolivia entered into a novel debt repurchase agreement with its commercial bank creditors in 1988. To appreciate the drama of the end of the hyperinflation, consider the situation on the eve of the new program. During August 1985, the official exchange rate stood at 67,000 pesos per dollar, while the black-market exchange rate was approximately 1.1 million pesos per dollar, for a percentage gap of some 1,600 percent (in fact, after averaging 1.1 million for most of the month, the black-market rate shot up to 1.5 million pesos on the eve of the program announcement). The internal price of petroleum was about 3 cents per liter, compared with a world price of 28 cents per liter. Inflation in the weeks leading up to the program was (at weekly rates): August August August August

5-11: 12-18: 19-25 26September 1:

18.4 8.6 6.1 19.9

As the program began, the exchange rate was allowed to depreciate freely, before a new de facto peg was set. The market set the new unified rate at an opening level of about 1.1 million, giving a one-day dcprcciation of 1,600 percent relative to the previous official rate. Internal oil prices were immediately raised to 28 cents per liter, or by 833 percent. All other price controls in the economy were lifted by the decree of 29 August 1985. These enormous changes took only ten days to work through the price system. Prices jumped immediately following the devaluation and increase in internal energy prices, but then inflation abruptly ended: September 2-8: 36.8 September 9- 15: -4.6 September 16-22: - .8 September 23-29: - 2.5 September 30-October 6: .7 These data are reproduced in figure 7.1, where the remarkable break in the inflation from the second week of the program onward is clearly evident.

241

BolividChapter 7

I

-6

I

-4

I

I

-2

I

I

0

I

2

4

6

WEEK OF PROGRAM Fig. 7.1

Weekly inflation rates (week 0 = 8/26-9/1)

The inflation rate remained at relatively low levels during the following year, with the exception of a jump in prices during December 1985 and January 1986, following policy mistakes in monetary management at the end of 1985 (which resulted in a cabinet reshuffling and the reinforcement of fiscal austerity measures). The leading trade union organization, the COB, which had opposed and mobilized against similar-indeed less dramatic-policy packages in the past, called for a general strike in opposition to the program. But after three years of accelerating inflation and the near chaos of early 1985, the new government clearly had the upper hand. A state of siege was declared (as a temporary and constitutional emergency procedure), and the strike was quickly disbanded, after which the state of siege was cancelled. The stabilization of the exchange rate at the new depreciated level of 1.1 million pesos per dollar caused a gradual rebuilding of the public's real money balances. One of the central policy issues at the end of a hyperinflation is how that increase in real money demand should be accommodated-through domestic credit expansion to the public sector or to the private sector (e.g., via rediscounting of private paper), or through foreign reserve inflows through the balance of payments (i.e., Central Bank purchases of foreign exchange at a pegged rate or in a dirty float). The Bolivian government chose the third, and clearly most conservative, strategy of relying on the balance of payments. It was felt that with large Bolivian hoards of U.S. dollars, both in Bolivia and abroad, there was sufficient availability of foreign exchange holdings in the private sector to provide the

242

Juan Antonio Morales and Jeffrey D. Sachs

basis for the needed expansion of the domestic money supply. Moreover, with confidence in the peso likely to be a rising function of the extent of foreign reserve backing of the domestic money base, the conservative approach was felt to be the most consistent with a restoration of confidence in the new exchange rate and monetary regime. Net reserves climbed from approximately $63 million at the end of August 1985 to about $150 million by the end of December, exceeding the rise in the money base from $64 million in August to $1 12 million at the end of the year. We have already noted the sharp rise in government revenue collection as soon as the program started (see fig. 6.1). The rise in domestic oil prices plus the renewed YPFB tax payments on export earnings alone raised the rate of central government tax collection by about 7 percent of GNP. Adding in a small recovery of internal taxes and tariff collections, government revenues rose from about 1.3 percent of GNP during January-September 1985 to 10.3 percent of GNP during the remainder of the year. Overall, the central administration moved into a cash-flow surplus during OctoberDecember 1985. With a virtual elimination of domestic credit expansion to the fiscal sector, seignorage dropped off sharply after the program began (as was evident in fig. 6.1). Expressed as a percentage of GNP, the evolution of seignorage on a quarterly basis is:

1.9% .l% .8% Shortly after the stabilization program was announced, Bolivia faced the strong challenge of shocks to the external sector. Bolivian exports had depended heavily on three items in recent years. Tin and natural gas constituted about 90 percent of Bolivia’s measured and legal exports, while processed coca leaf, to become cocaine, represented the third (illegal and unrecorded) export. The rough estimates for Bolivian exports in 1985 are: 1985:l 1985:2 1985:3

2.9% 3.6% 3.4%

Tin Other metals Natural gas Others Coca based Total

1985:4 19861 1986:2

$ 150 million $ 81 million $ 377 million $ 22 million $ 600 million $1,230 million (approx. 31% of GNP)

Once the stabilization program got started, there were profound disruptions in all three of the major markets. The price of tin collapsed by 60 percent at the end of October 1985, as the worldwide tin cartel crumbled in financial distress. Three months later, hydrocarbon prices collapsed, forcing a renegotiation of Bolivia’s natural gas contracts with Argentina and resulting in a loss of dollar earnings on the order of $75- 100 million (the exact value is difficult

243

BoliviaKhapter 7

to compute because of a complicated barter arrangement between the two countries as part of the gas agreement). Finally, the Bolivian government began an interdiction effort against narcotics trafficking in the summer of 1986, disrupting the smuggling of coca-based products from the country; and, at the same time, the world price for coca paste fell sharply. A conservative estimate would put the terms-of-trade loss on the order of 10 percent of GDP. The shortfall in export earnings threatened the stabilization program in several ways. Most importantly, the export shortfall resulted in a sharp contraction in living standards, which is attributed to a significant extent within Bolivia to the effects of the stabilization program itself. With great weakness in the domestic demand for nontradable goods as a result of the decline in national income, there was enormous political pressure to revive government spending, increase public wages, and ‘‘reactivate” the economy. Second, each episode of export shortfall provoked public expectations of a devaluation and helped to maintain high short-term nominal peso interest rates as a result of continued speculation against the exchange rate. After the tin price collapse in October 1985, the exchange rate was, in fact, allowed to depreciate by almost 50 percent over a period of two months. After the collapse of world petroleum prices, the authorities pegged the peso exchange rate, despite skepticism of the public, who speculated heavily on a depreciation. Similarly, after the start of the drug interdiction effort in the summer of 1986, the public again speculated heavily against the peso, pushing up short-term peso interest rates once more and converting domestic currency holdings into dollars at the rate pegged by the Central Bank. Once again, the authorities resisted a currency devaluation, but at the cost of further months of high ex post real interest rates.

7.2 Interpreting the Rapid Success at Price Stabilization In his rightly celebrated and influential analysis of the ends of four hyperinflations, Sargent (1982) stressed that hyperinflations end when governments “change the rules of the game” regarding budget deficits and money creation. Sargent argues:

In each case that we have studied, once it became widely understood that the government would not rely on the central bank for its finances, the inflation terminated and the exchanges stabilized. . . . [tlhe changes that ended the hyperinflations were not isolated restrictive actions within a given set of rules of the game or general policy. Earlier attempts to stabilize the exchanges in Hungary under Hegedus, and also in Germany, failed precisely because they did not change the rules of the game under which fiscal policy had to be conducted. (89-90) How does the Bolivian experience fit into this schema?

244

Juan Antonio Morales and Jeffrey D. Sachs

At a trivial level, Sargent is certainly correct that without a change in budget practices, no noninflationary exchange rate management would have been successful for long in Bolivia. Sargent’s real assertion is much deeper, however, in suggesting that the end of the hyperinflation occurs suddenly not just because the inflationary budget or monetary practices change, but also because the public expects and understands that these changes will persist. In other words, the rapid end of the hyperinflation is a signal of a restoration of confidence. At this level, we believe that the Bolivian experience is not strongly supportive of Sargent’s view, in that the evidence suggests that stabilization proceeded well ahead of a complete restoration of public confidence. In our interpretation, fixing the exchange rate was nearly a sufficient condition for a short-run stabilization because so much of pricing was tied to the nominal exchange rate. Moreover, it was possible to fix the exchange rate in the short run despite the fact that the public lacked confidence in the ability of the government to Jix the rate in the long run. A temporary stabilization of the exchange rate was feasible because the Central Bank was willing to commit at least some international reserves (e.g., Bolivia’s gold reserves) to pegging the exchange rate in the short term and because the emergency measures allowed the government to temporarily not have to resort to money creation. Almost at the moment of stabilization, Central Bank credits to the public sector stopped rising, as revenues jumped in response to the higher domestic prices of petroleum products. During October and November 1985, increases in the money supply were due to increases in Central Bank holdings of foreign exchange reserves. The exchange rate stabilized at 1.1 million pesos per dollar, leading almost immediately to price stabilization. (Actually, the government did not peg the exchange rate, but rather intervened to prevent the rate from appreciating above the 1.1 million level. Given the pressures toward appreciation as households started to rebuild real money balances, the one-sided peg of the exchange rate worked as if the government were actually pegging the rate.) Before showing the evidence that inflationary expectations were slow to decline, let us consider some reasons why we should not be surprised that expectations did not shift dramatically upon the announcement of the program. First, the stabilization program was the seventh attempt in four years and was being carried out by a president without a majority in Congress and with a direct electoral mandate of less than 30 percent of the voters. Second, the program promised many reforms that would take a considerable amount of time to negotiate and implement. Third, the program reflected a radical change from the past policies of the president’s own party (which, as the vanguard of the Revolution of 1952, did not have a tradition of laissez-faire economics), and so could not be counted on to win the support of the governing party itself.

245

BolividChapter 7

Fourth, there were profound unknowns at the time of stabilization. The overhang of foreign debt remained crushing, and the stabilization in the intermediate run would remain possible only with new foreign funds from the official creditors. But would the IMF consent to a program without the immediate resumption of interest payments to the commercial banks? Would the Paris Club respond with generous debt rescheduling terms? Internally, the situation was as murky. Under Siles, the central administration had reneged on many financial commitments to regional governments. Would those claims by the regional entities now be honored, thereby threatening the solvency of the central government? Would tax reform measures be implemented? Would the unions be able to crush this stabilization program as they had, in part, the early ones? As of 9 September 1985, there were no budget accounts, reliable tax forecasts, or even statistics on Bolivia’s international reserve holdings (the data were in a mass of confusion because of several complications with bilateral payments arrangements with Argentina). There was simply no basis for an informed opinion about the longer-term prospects of stabilization. 7.2.1

Inflation Expectations in the Wake of the Stabilization

We have suggested that the hyperinflation ended immediately because of the stability and unification of the exchange rate, even though inflationary expectations, at least regarding the intermediate run, did not drop instantaneously. In other words, the program was successful even though the measures were not immediately credible. The key to this seemingly paradoxical view, we have suggested, is the fact that it is possible to peg an exchange rate in the short run, even if there are widespread expectations that the exchange rate will collapse in the future. There is one excellent source of evidence on exchange rate expectations, and that is from nominal interest rates on peso- and dollar-denominated assets in La Paz in the aftermath of the stabilization. Nominal peso interest rates remained extraordinarily high during much of the year after the beginning of stabilization. Nominal interest rates fell after the stabilization program began, but only gradually and with a long lag in response to the end of currency depreciation. Some data on nominal lending rates are shown in table 7.2. Under the deregulated financial system in place since the stabilization program began, loans could be contracted in either dollars or pesos, at unregulated interest rates. The spread between peso- and dollar-denominated borrowing reflects mainly the expected rate of depreciation of the peso vis-A-vis the U.S. dollar. Note that despite the sudden cessation of price increases in September 1985, nominal interest rates in pesos remained on the order of 20 percent per month from October 1985 to March 1986, compared with dollar-denominated interest rates of 1 % to 2 percent per month. The expectations of continued currency depreciation proved to be appropriate during December 1985 and January 1986, when policy mistakes led to a

246

Juan Antonio Morales and Jeffrey D. Sachs

Table 7.2

& and Dollar-Denominated Thirty-dayLoan Rates, April 1985-0ctober 1986 (beginning of month)

Peso

Dollar

Ex Post Dollar Rate on Peso Loan

Excess cost of Peso Loan

34 35 44 50 45 45 31 22 21

1.4 1.6

10.2 -6.7

8.8 -8.3

1.6 1.7 1.7 1.7 1.8 1.9 1.9 1.9 1.9 1.8 1.8

-8.1 34.5 16.8 17.5 13.0 8.4 8.0 7.5 6.8 6.8

-9.7 32.8 15.1 15.8 11.2 6.5 6.1 5.6 4.9 5.0

1985

April May June July August September (stabilization begins) October November December 1986

January February March April May June July August September October November a

19

20 20 19 13 8 8 8 7 7 3

+

Note: The ex post dollar rate on a peso loan is calculated at (1 i,)&/E,+, , where E, is the beginning-of-month exchange rate (pesos per dollar) and i, is the beginning-of-monthf interest rate. "Last week of October.

sharp depreciation of the peso. However, from February to September 1986, the interest rates remained extraordinarily high (see table 7.2), despite the absence of any further currency depreciation. Thus, it is clear from the data that there was not a decisive restoration of confidence in the peso at the beginning of the program, despite the fact that ex post, the stabilization succeeded. Similar evidence against an instantaneous recovery of confidence is offered by the time path of real money balances in the wake of the stabilization program. Holdings of real balances increased only gradually during 1986 and remained significantly below historical levels throughout the year.

7.3 Was the Stabilization Recessionary? The evidence on the recessionary effect of the program is ambiguous, since during the year of the stabilization program the economy was hit remarkably hard by a series of external shocks, whose overall effect was a loss of export earnings on the order of 10 percent of GNP. Such an enonnous

247

BolividChapter 7

loss of export earnings would normally be enough to create a deep recession, even in an economy starting from a position of macroeconomic stability. Moreover, during 1981-85, measured real GDP had declined every year, by 2.9 percent on average, so that the continuing economic stagnation and falling GNP in 1986 could not in any event be attributed easily to the stabilization program. In the event, GDP fell by 2.9 percent in 1986, but then began rising in 1987 and 1988 (by an estimated 2.3 and 3.0 percent, respectively), despite the sharp fall in the terms of trade. There is absolutely no evidence, therefore, that the effects of stabilization were in any way contractionary-if at all-for more than a very brief period. It should be stressed in any event that the “austerity” fiscal actions undertaken in ending a hyperinflation are not necessarily contractionary on balance, since the contractionary effects of a rise in tax revenues are balanced by the expansionary effects of the elimination of the inflation tax. Basically, the stabilization program does not involve a rise in overall taxes, but rather a shift from one tax (the inflation tax) to other, more efficient taxes. In addition, the reduction in uncertainty and the elimination of price distortions brought about by the stabilization program should have important supply-side effects. Moreover, the stabilization program made possible a restoration of net inflows of financial assistance from the rest of the world. Thus, the stabilization measures significantly eased the national budget constraint, if account is taken of the hundreds of millions of dollars of financial assistance that were made available by the IMF, the World Bank, the Inter-American Development Bank, and the bilateral creditors as a result of the success of the stabilization program. We have noted in chapter 1 that as a result of stabilization, the net resource transfer shifted from a significant negative amount (approximately 6 percent of GNP net outflow in 1983) to a large positive amount in 1987 (approximately 5 percent of GNP net inflow). One possible source of contractionary impulses from the stabilization program has been the high real interest rates in Bolivia since the beginning of the stabilization. As noted by Dornbusch and Fischer (1986), high ex post real rates have been present in the aftermath of most hyperinflations because of a scarcity of money as the demand for money rises sharply following stabilization. We have suggested that the high interest rates reflect lingering inflationary expectations, and so do not necessarily signal high ex ante real interest rates. 7.3 Further Lessons from the Hyperinflation

We conclude our discussion of the hyperinflation by discussing some of the central issues that have been debated in the design of stabilization programs and in the interpretation of their success or failure. We turn to two

248

Juan Antonio Morales and Jeffrey D. Sachs

topics in particular: the case for an “orthodox” versus “heterodox” shock program, and the linkages of short-run macroeconomic policy and long-run structural policies. 7.3.1

Orthodox versus Heterodox Anti-Inflation Programs

Much of the debate in Latin America in recent years has centered on the question of the appropriateness of “orthodox” policies in ending a high inflation. Advocates of the “heterodox” shock approach have argued that tight monetary and fiscal policies alone are insufficient to end a high inflation. They suggest that incomes policies, wage and price controls, and even monetary reform should play a central role in a stabilization program. Extreme advocates of heterodoxy go even further: they suggest that orthodox policies have no role to play in inflation stabilization. In this view, inflation is an “inertial” process, in which prices and wages rise because of selffulfilling expectations. The goal of policy, in that view, is to break those expectations through wage and price controls. The Bolivian experience sheds some light on this debate, although the lessons from Bolivia must be read with care. The most obvious point is that Bolivia was successful in stabilizing inflation without recourse to any elements of heterodox wage or price policies. There were no wage or price controls, and no monetary reform. Indeed, at the beginning of the program, virtually all existing price controls were removed, with the goal of liberalizing the price system. In this sense, it is clear that orthodox policies were sufficient to end the hyperinflation. And as already indicated, the end of the inflation was virtually instantaneous, with the runaway inflation being eliminated in a matter of a couple of weeks and remaining low for several years. This success stands in sharp contrast to the failed “heterodox” plans in Argentina, Brazil, and Peru. In those countries, inflation was held down by the stabilization programs for a few months, only to explode at very high rates after a short period of time. The long-term success of the Bolivian program compared with the failures elsewhere is clearly related to the greater fiscal stabilization achieved in Bolivia, and thereby in the lesser recourse to money financing of budget deficits. The links between the budget deficits and inflation in all of the countries mentioned is clear enough. Large deficits are financed by money printing, which in turn causes a downward pressure on the foreign exchange rate. If the exchange rate is allowed to depreciate, then the domestic prices of tradable goods (importables and exportables) rise in line with the exchange rate depreciation. If the exchange rate is pegged in order to avoid the inflationary consequences of depreciation, the central government loses reserves over time, which eventually forces it to devalue the exchange rate, but at a later date.2

’

249

BolividChapter 7

Typically, as happened in Argentina, Brazil, Peru, and Bolivia during 1982-85, the government tries to avoid outright depreciation or devaluation of the exchange rate, even as it runs out of reserves. Rather than devalue the official rate, the government rations foreign exchange, thereby creating a spread between the official exchange rate and the “market” (black or parallel) rate. In the end, the inflationary consequences are the same as if an official devaluation had occurred. On the margin, imports are purchased at the highly depreciated black-market rate, so that the internal prices of importables rise in step with the depreciated black-market rate, rather than the official exchange rate. Eventually, the gap between the official and the depreciated black-market rate becomes so large and produces so much cheating and corruption (in the form of overinvoicing of imports and underinvoicing of exports, bribes to get Central Bank allocations of foreign exchange, etc.) that the official exchange rate is grudgingly d e ~ a l u e d . ~ Of all of the high-inflation countries in Latin America in the 1980s, only Bolivia after 1985 was able to break this circle of budget deficits, depreciating currency, and high inflation. The orthodox tight monetary policies in Bolivia, made possible by a major reduction in the budget deficit, was the key that allowed Bolivia to maintain a stable and unified exchange rate vis-8-vis the dollar. In other countries, the large budget deficits that continued after the stabilization programs were put in place prevented the establishment of a stable and unified exchange rate for any period longer than a few months. While the Bolivian case illustrates powerfully the need for significant budgetary adjustment, it does not really settle the case for or against wage and price policies in addition to the more orthodox budgetary policies. In Bolivia, inflation stopped immediately without any additional “heterodox” pricing policies, but we should not make overly strong inferences from this. Part of the Bolivian experience reflects the special characteristics of a hyperinflation rather than a mere high inflation. In Bolivia, the inflation ended as soon as the exchange rate was stabilized because the preceding hyperinflation had eliminated all long-term price contracts from the Bolivian economy. By August 1985, most prices in Bolivia were set on a spot-market basis. Often, the prices were quoted in dollars, with the Bolivian peso price being determined by the dollar price multiplied by the black-market exchange rate at the moment. In these circumstances, inflation is basically equal to the rate of depreciation of the black-market exchange rate. If the exchange rate can be stabilized, inflation will immediately revert to the rate of increase of dollar prices. Indeed, in early September 1985, the exchange rate stopped depreciating and the Bolivian inflation was over. In other countries, however, it may happen that the exchange rate is stabilized but that domestic prices continue to rise. Such was the case of

250

Juan Antonio Morales and Jeffrey D. Sachs

Chile during 1979-81.4 The difference between Bolivia and these other cases seems to lie in the nature of hyperinflation, in which all vestiges of long-term pricing in the domestic currency disappear. For less rapid inflations, staggered price contracts denominated in the domestic currency, or backward-looking indexation to the domestic CPI, can break the tight link between inflation and the exchange rate. Thus, countries with high inflations but not hyperinflations might be wise to combine exchange rate stabilization with other pricing measures in order to achieve a rapid return to price stability. The 1985 Israeli stabilization is a clear example of the combination of orthodox stabilization policies combined with elements of “heterodoxy” in order to achieve a rapid disinflation. As in Bolivia, the key to Israel’s stabilization program was a dramatic and sudden cut in the budget deficit. And as in Bolivia, the stabilization of the exchange rate provided the “nominal anchor” in the stabilization, that is, the key price whose fixity was central to price stability throughout the economy. But in Israel, as opposed to Bolivia, the government also negotiated an agreement with the Histadrut, the national trade union movement, which called for a short-term freeze of wages and prices. Notably, the union agreed to this freeze conditional on the continued stability of the nominal exchange rate. Thus, as in Bolivia, the Israeli stabilization was, at the core, built upon nominal exchange rate stability. 7.3.2

Stabilization and Structural Adjustment Policies

One of the continuing debates among economists and policymakers is the appropriate mix of “stabilization” policies (budget cutting, tight monetary policy, etc.) with “structural” policies (trade liberalization, privatization, etc.). Some argue that stabilization should precede structural adjustment policies, while others say that the stabilization and structural adjustment policies should be pursued in parallel. Once again, the Bolivian case sheds some light on these arguments. The Bolivian stabilization program began with the remarkable presidential Supreme Decree 2 1060, which instituted a wide-ranging liberalization program in conjunction with the short-run stabilization measures. It is important to understand the motivations for this joint action. To a significant extent, to be sure, the designers of the stabilization program were aware that Bolivia needed to undergo major structural adjustments, especially to develop new export markets, in order to compensate for the long-term secular decline of Bolivia’s major export, tin. But the motivations for the liberalization went beyond the textbook arguments for outward trade orientation and an efficient allocation of resources. The designers of the stabilization program were aware of the enormous administrative overload that Bolivia’s model of state capitalism had created.

251

BolividChapter 7

The state sector was not only financially bankrupt but administratively bankrupt as well. There was little administrative capacity to carry out the most basic functions of government, much less to implement a refined program of industrial policy based on an elaborate scheme of price controls, transfers, and subsidies. Price controls that were in place were haphazardly enforced, subject to enormous corruption, and often had bizarre and unintended consequences. Grain subsidies, for example, resulted in enormous smuggling of flour to neighboring countries and the enrichment of millers, rather than any significant benefit to the Significant amounts of cheap gasoline were being smuggled in tanker trucks across the border with Peru. Cheap credits to agriculture were simply fueling the black market, rather than creating any additional agricultural output. Under these circumstances, much of the economy was already de facto liberalized, with many prices-even those under formal controls-effectively being determined by world market conditions. Price controls often had bizarre and regressive effects, with powerful or clever middlemen able to arbitrage the gaps between official prices and world prices. For these reasons, the liberalization of prices was viewed as a basic part of the short-term stabilization strategy, as well as a part of the long-term development strategy. By freeing up prices and withdrawing from the business of setting a wide range of quotas and prices, the government therefore hoped to accomplish many things: to free top government officials to focus on the crucial fiscal measures in need of immediate implementation; to eliminate many of the costly and distributionally perverse income transfers implicit in the existing system of price regulation; to reduce the need to monitor government policies, so that the scarce administrative machinery could focus on tax collection and budget cutting rather than on the administration of a complicated system of wage and price controls; and finally, to send the right price signals needed to encourage the movement of resources into new tradables sectors. As discussed below, the long-term structural adjustment to the liberalization policies is still in the early phases, despite several years of stabilization. While there has been some rise in nontraditional exports, there have still not developed major new export sectors that can compensate for the collapse of tin exports. Bolivia has still not generated the needed internal savings to finance the development of a major new export sector, and external savings are unlikely to make up the difference in view of Bolivia’s enormously high country risk. At the same time, Bolivia remains stuck without its most obvious new market: Brazil. It is only very recently that Brazil has shown an interest in easing its heavily protectionist policies to allow an increase in imports from Bolivia.

252

8

Juan Antonio Morales and Jeffrey D. Sachs

Bolivian Debt Management, 1985-88

We saw in the previous chapter that the cutoff in foreign lending to Bolivia in the early 1980s, combined with heavy debt-service payments during 1982-84, was a key factor in provoking the Bolivian hyperinflation. The commercial bank debt payments were unilaterally suspended in May 1984 by the Siles government at the insistence of the COB, the trade union organization. When President Paz came to office, the key intention with respect to the foreign debt was to avoid a resurgence of inflation that could be caused by a return to heavy debt-service payments. The Paz government began negotiations with its various creditors on a differentiated package of debt relief. With respect to the IMF and the World Bank, there was no prospect of debt relief via reschedulings, since those institutions do not reschedule their debts. Thus, the government attempted to arrange new credits that would offset the debt servicing, thereby leading to a net resource inflow from these institutions. With respect to the bilateral creditors, the intention was to get a fully negotiated postponement of interest and principal payments through a settlement in the Paris Club and to arrange for net new credits from friendly governments. Finally, with respect to the commercial creditors, the goal was essentially to get a new kind of settlement on the debt that would eliminate Bolivia’s debt overhang and obviate the need to makc cconomically and politically destabilizing net transfers to the bank creditors. In the event, Bolivia has maintained a suspension of interest servicing to the bank creditors since 1985 and negotiated a novel “debt buyback,” which eliminated approximately one-half of Bolivia’s commercial bank debt by the end of 1988. In this chapter we analyze both the conceptual underpinnings and the nature of negotiations that led to the Bolivian buyback, as well as discuss briefly the nature of Bolivia’s settlement with the other creditors. The chapter continues in three sections. In the next section, we explain why a comprehensive debt reduction mechanism, such as a debt buyback, can be highly desirable for the creditors as well as the debtor. Then, we describe the negotiations and implementation of the Bolivian buyback and argue that, indeed, the arrangement has been of benefit to the creditor banks as well as to Bolivia. Finally, we briefly describe the favorable debt relief that has been achieved by Bolivia in negotiations with its other creditors.

8.1 The General Theory of Debt Reduction Operations If a country owes $1 billion on which it can only pay an expected $50 million, the country may suffer enormous costs from being unable to pay the

253

BolividChapter 8

full amount due. It will face a great difficulty in new borrowing, even for highly productive investments. It will face high bargaining costs in handling the $1 billion of bad debt.* And, it will face sanctions from disgruntled creditors (e.g., a withdrawal of trade credits) that will hinder its future economic perf~rmance.~ In addition, there will be a major internal disincentive to economic reforms which increase the debt-service capacity of the country, since the costs of reform are borne by the country, while many of the benefits of reform will be appropriated by the creditors (who receive higher repayments in the event of r e f ~ r m ) . ~ For these reasons, it may well be beneficial for the country to pay an amount even more than the $50 million (in present value terms), in order to cancel the overhang of $1 billion of mostly bad debt. It will also be generally advantageous for the creditors to accept a partial payment on the debt, as long as it is in excess of the $50 million expected payments. The partial payments could come in the form of a direct cash buyback (especially if the country can borrow the funds for the buyback from friendly governments) or some other arrangement where future debt payments of over $50 million are guaranteed by the debtor country. A cash-starved country would obviously prefer to find ways to make the present value of payments in the future, rather than with current cash.5 By eliminating the overhang, the country would avoid the costs of default and regain the incentives for internal reform. In practice, even mutually advantageous debt reduction schemes (in which the debtor clears the debt overhang and the creditors raise the total value of payments that they receive) are hard to negotiate under the current debt management strategy of the IMF and the creditor governments. There are several decisive reasons why even mutually beneficial deals have not taken place. First, the few very heavily exposed banks have an inherent incentive to reject buyback deals, even when they are efficient from the point of view of the banks as a whole (i.e., when they raise the market value of overall debt repayments).6 Second, the U.S. government is the main arbiter of the kind of deals that take place, and it has vetoed almost all comprehensive debt reduction schemes, on behalf of the most-heavily exposed money center banks. Third, for the smaller countries, the debt negotiations are guided by the creditors’ concerns over precedent for the large debtors, rather than for the efficiency of the outcome for the small debtor. It is generally best to “strangle” a little country, even at the expense of the country’s debt servicing, if it sends a convincing signal to Brazil and Mexico to keep paying their debt.

8.2 The Bolivian Buyback The Bolivian buyback must be understood against the following background. Bolivia was the only case up to 1988 in which the U.S. government actually supported a policy of debt relief, though it came to that

254

Juan Antonio .Morales and Jeffrey D. Sachs

position only after long and difficult negotiations with the Bolivian government.’ The strategy has been highly beneficial for all parties, according to the theory of the “debt overhang” outlined earlier. Bolivia, alone of the high-inflation countries in the Southern Cone, has been able to stabilize and to resume growth because it has not been trapped by excessive debt repayments. There has also been a restoration of political stability in the country after the chaos and virtual anarchy of hyperinflation during 1984-85. After the Paz government came to power in mid-1985 and undertook remarkable stabilization efforts to halt the hyperinflation, it remained official U.S. and IMF policy in the spring of 1986 that Bolivia should resume interest payments on its foreign bank debt. Indeed, in March 1986, only two months after price stability had been restored to the country, the IMF was urging a large devaluation in Bolivia in order to facilitate increased interest payments to the commercial banks. The Bolivian government was convinced that such a move, in addition to destroying the economic and political basis of the stabilization program itself, would cause a collapse of the government. Instead, the Bolivian government urged a different approach in discussions with the U.S. government and the IMF.8 Ultimately, the IMF agreed to treat the Bolivian case on its own merits and acknowledged that Bolivia’s foreign bank debt could not be paid (at least, not without undermining economic and political stability in the country).’ The IMF also agreed to grant Bolivia a program based on its successful stabilization efforts, despite the fact that the Bolivian government had not reached any understanding with the commercial bank creditors. This was the first time that the IMF loaned money to a debtor country that did not plan to make interest payments to the commercial banks (or even to clear the arrears on back payments). In late 1986, the banks began to discuss with Bolivia a longer-term solution to Bolivia’s debt overhang, once they saw that the U.S. government and the IMF were not going to defend the banks’ position vis-a-vis Bolivia. Moreover, for several years, the U.S. regulators had been forcing writedowns of Bolivian debt in the banks’ books, thereby eliminating any important incentives that the banks might have had to hold on to the debt. After two years of complicated discussions and legal work, the buyback was arranged. Note that during the entire period of discussions, Bolivia did not pay any interest to the commercial banks. At the same time, Bolivia received large net resource transfers (on the order of about 5 percent of GNP per year) from the official creditors. As mentioned earlier, Bolivia repurchased with the buyback about one-half of its debt at 11 cents per dollar of face value. The money used for this purpose was donated from foreign governments. While some of the money might otherwise have come to Bolivia as foreign aid in other forms, much of

255

BoliviaiChapter 8

it would not (of the $34 million spent on the buyback, Bolivia might have been able to get $15-20 million of the money in other forms of aid). It might appear that the buyback could not have had much of a beneficial efficiency effect on Bolivia, since the country only repurchased one-half of the debt and the remainder still sells for 11 cents (indicating that most of the remaining debt also will not be paid, thus leaving Bolivia in a situation of default). But this abstract analysis misses the real point of Bolivia’s situation post-buyback. Under current U.S. and IMF policy, Bolivia is not being pressed on the remaining part of the debt, except to settle that remainder on a similar basis to the buyback.” In effect, the official community is recommending a gradual process in which Bolivia will clear all of its commercial bank debts at a price of about 11 percent of face value, and the process seems to be occurring: after the buyback, Bolivia has continued to repurchase debt at the buyback price, through individual deals with creditor banks. Meanwhile, as this process goes forward, the official community has agreed not to impose sanctions on Bolivia for nonpayment on the remaining bank debt. Was the debt strategy of the IMF and U.S. government successful in the case of Bolivia? The answer is a resounding yes, for all of the parties concerned. In effect, in 1986 the official community recognized the futility of trying to press Bolivia to pay unpayable debt. As a result, the Bolivian government got the time and international support to put in place a remarkably strong and effective stabilization program that has ended a hyperinflation and restored growth to the country for the first time in almost a decade. Bolivia’s political stability has been enhanced, as have its democratic institutions. The creditors as a whole benefited as well, as shown by the fact that Bolivia’s debt rose in value from 5 to 11 cents per dollar. This increase in the price of debt was not a giveaway by Bolivia. I It reflects, instead, the creditors’ share of the remarkable turnaround of the Bolivian economy, from the worst in the world during the early 1980s (with the world’s highest inflation in forty years) to one of stability and incipient recovery in 1988. Bolivia’s success story depended strongly on the supportive actions of the U.S. government and the IMF in providing a framework in which Bolivia could successfully negotiate with its bank creditors. Effective progress for other debtor countries will require similar oficial forbearance. As the Bolivian case has demonstrated, the debtor as well as the creditors (at least taken as a group) can benefit strongly from a realistic approach to comprehensive debt reduction.

8.3 Bolivia’s Relations with the Official Creditors The key to Bolivia’s debt strategy has been to maintain good financial relations with official creditors while at the same time pursuing debt reduction with the banks. To this end, Bolivia has kept current in payments

256

Juan Antonio Morales and Jeffrey D. Sachs

to the IMF, World Bank, and Inter-American Development Bank; negotiated two standby programs with the IMF for 1986-87 and 1987-88, and a three-year Structural Adjustment Facility with the IMF for 1988-90; negotiated several loans with the World Bank on concessional terms (from the International Development Association fund); and renegotiated the debt with the Paris Club on highly favorable terms in 1986 and 1988. The result has been a positive net resource transfer from the official creditor community at the same time that Bolivia has had essentially a zero net resource transfer to the banks. The relevant data are shown in table 8.1, where we see that after the stabilization program went into effect in 1985, Bolivia succeeded in reversing the overall net resource transfer by an elimination of net outflows to the banks and a reversal from outflows to inflows from the official community. Table 8.1

Net Resource Transfers on Medium- and Long-Term Debt, 1982-87 ($ million)

Type of Creditor

1982

Official Multilateral

Bilateral Private Suppliers Financial markets Total Net transfer as % of GNP

71 56 15 - 132 -7 - I25 -61 2.0

1983 14

1984

1985

1986

1987

53

- 86

1

195 142 54 - I1 -3 -8 I84 4.8

81 49 33 -9 -I -8 72 1.7

-

45

54

-41 - 45

- 149

- 49

- 30

- I7

-2 - 47 - 102

- 26

-31 - I32 - 135

-4.6

-

-3.4

-4

-116 -3.6

Source: World Bank Debt Tables, 1988-89 edition

9

Beyond Stabilization to Economic Growth and Development

The Bolivian stabilization has eliminated much of the panic conditions that surrounded the hyperinflation in 1984 and 1985. Also, significant progress has been made in easing the external debt overhang. Virtually complete price stability has been reestablished in Bolivia during 1987 and 1988. It is evident, however, that many of the deeper problems in the Bolivian economy and society that helped to cause the hyperinflation remain in place, and in

256

Juan Antonio Morales and Jeffrey D. Sachs

to the IMF, World Bank, and Inter-American Development Bank; negotiated two standby programs with the IMF for 1986-87 and 1987-88, and a three-year Structural Adjustment Facility with the IMF for 1988-90; negotiated several loans with the World Bank on concessional terms (from the International Development Association fund); and renegotiated the debt with the Paris Club on highly favorable terms in 1986 and 1988. The result has been a positive net resource transfer from the official creditor community at the same time that Bolivia has had essentially a zero net resource transfer to the banks. The relevant data are shown in table 8.1, where we see that after the stabilization program went into effect in 1985, Bolivia succeeded in reversing the overall net resource transfer by an elimination of net outflows to the banks and a reversal from outflows to inflows from the official community. Table 8.1

Net Resource Transfers on Medium- and Long-Term Debt, 1982-87 ($ million)

Type of Creditor

1982

Official Multilateral

Bilateral Private Suppliers Financial markets Total Net transfer as % of GNP

71 56 15 - 132 -7 - I25 -61 2.0

1983 14

1984

1985

1986

1987

53

- 86

1

195 142 54 - I1 -3 -8 I84 4.8

81 49 33 -9 -I -8 72 1.7

-

45

54

-41 - 45

- 149

- 49

- 30

- I7

-2 - 47 - 102

- 26

-31 - I32 - 135

-4.6

-

-3.4

-4

-116 -3.6

Source: World Bank Debt Tables, 1988-89 edition

9

Beyond Stabilization to Economic Growth and Development

The Bolivian stabilization has eliminated much of the panic conditions that surrounded the hyperinflation in 1984 and 1985. Also, significant progress has been made in easing the external debt overhang. Virtually complete price stability has been reestablished in Bolivia during 1987 and 1988. It is evident, however, that many of the deeper problems in the Bolivian economy and society that helped to cause the hyperinflation remain in place, and in

257

BolividChapter 9

some cases have deepened. We now mention some of the challenges that remain in converting the current stabilization period into the first phase of sustained economic development. The end of the hyperinflation did not bring a sudden rejuvenation of the economy. Real growth during 1987 and 1988, while positive, was not even enough to maintain real living standards. Indeed, it is fair to say that once the hyperinflation was lifted, Bolivia’s desperate underlying situation became even more apparent. We noted in the introduction that Bolivia has survived for hundreds of years on the exports of a few commodities: silver until the nineteenth century, tin during the twentieth century, and natural gas and (illicit) coca paste in the 1980s. None of these commodities can continue to act as the engine of growth of the Bolivian economy. Silver and tin deposits have been heavily mined, and remaining exploitation of these minerals will have to be on a smaller and much more technologically advanced basis. Tin mining in Bolivia had already become unprofitable at $5.60 per pound, the price that prevailed before the collapse of the world tin market in October 1985. At the post-collapse price of about $3.00 per pound, the Bolivian government was forced to lay off most of the tin miners, and Bolivia now stands little chance of maintaining significant amounts of tin exports. Natural gas is almost as problematical. The price on Bolivia’s natural gas exports to Argentina was slashed sharply after the collapse of world petroleum prices in early 1986. What is more, Bolivia’s export contract with Argentina expires in 1992. In view of Argentina’s recent large discoveries of natural gas deposits, it is quite possible that the gas contract with Bolivia will be suspended after 1992 or at least renegotiated on a much smaller scale. Coca paste is the most problematical and ironical case of all. As mentioned in chapter 1, starting in the early 1980s, when U . S . demand for cocaine soared, Bolivia was pulled into the extensive cultivation of coca leaf, the raw material input of cocaine. The coca leaf is processed into coca paste, a precursor of pure cocaine, and is then smuggled to Colombia. In the mid-l980s, it was estimated by the U . S . Drug Enforcement Authority (DEA) that Bolivian foreign earnings on coca paste exports roughly matched the sum total of all legal Bolivian exports. Bolivia has demonstrated a natural comparative advantage in coca cultivation. The climate of the Yungas and Chapare regions of the country are well suited to coca cultivation, and coca leaf has played an important role in the Andean culture for centuries. Illegal narcotics traffickers have demonstrated an enormous entrepreneurial activity in mobilizing resources into the sector, developing transportation and communications lines, etc. And yet, of course, the industry has been a disaster for the country from almost all points of view. It has encouraged the development of an internal mafia, linked with an international mafia of traffickers. This internal mafia

258

Juan Antonio Morales and Jeffrey D. Sachs

poses life-and-death risks for Bolivia’s democratic institutions and civil society. Coca cultivation has jeopardized Bolivia’s foreign relations with the United States and other industrialized countries. It has thereby created enormous uncertainties, as the U.S. government has repeatedly threatened to suspend foreign aid and international support to Bolivia. Coca production has also drained economic vitality from other economic sectors B la the “Dutch disease.”2 It has debilitated the tax system, since a “leading sector” is outside of normal tax c ~ l l e c t i o n . ~ Hundreds of thousands of well-organized (and in many cases well-armed) peasants now derive their meager livelihood from the cultivation of coca leaf, meaning that any plan for limiting coca cultivation must confront an enormous political and economic challenge from a large part of the population. Despite all of this, the Bolivian government has devoted a large share of its scarce resources and political capital to reducing coca cultivation and trafficking. For these reasons, Bolivia must now develop a new export base, indeed a completely new economic orientation, one that is a radical departure from its entire past history. As the planning minister, Gonzalo Sanchez de Losada, has said many times, “Bolivia must reinvent itself.” The basic strategy is to follow the lead of the outward-oriented developing countries: make the environment fertile for new exports, and then wait to see which industries respond to the incentives. Few observers could foresee that Chile’s brisk growth in recent years would be based on kiwi exports or that Korea’s export surge in the 1960s would be initiated with the export of wigs to the U.S. market! The key policies for promoting the new export base in Bolivia are a realistic exchange rate, an open trading system, and-budget permittingfiscal incentives for nontraditional exports. Bolivian export potential seems to be greatest in three broad areas. First, there is obvious potential in agricultural exports. The Bolivian lowlands in the East offer a vast and fertile area for grains (e.g., soybeans), tropical fruits, cut flowers, timber, etc. Bolivia has already begun to make soybean exports to world markets since the mid-1980s. Second, there is the potential for light manufacturing (e.g., yarns, textiles, furniture), especially for export to the Brazilian market. In July 1988, Brazil and Bolivia signed a new trade accord to open up some of the Brazilian market to new Bolivian goods. Third, there is the remaining potential in the mining and petroleum sectors. With new technologies for secondary recovery, some old silver and tin mines might once again become profitable. There also remains the potential, long under discussion, for a natural gas pipeline to Brazil. All of these new industries will require time, learning, and, above all, heavy investment expenditure. In turn, the investments must be predicated on a long period of social peace and political stability. Whether that stability can be achieved is certainly Bolivia’s most important question. There remain

259

BolividChapter 9

three sociopolitical cleavages that are serious obstacles to economic stability and long-term growth: income distribution, ideology, and region. The income distributional cleavage remains profound and is the source of considerable political conflict. The key political problem is to moderate the nearly continuous confrontation between powerful social groups, such as organized labor and private capital, and the various regional forces. The state-capitalism model attempted to finesse the income distribution problem through a combination of the inflation tax, heavy foreign borrowing, or internal repression of the lower classes. None of these alternatives is workable for a long-term development strategy. A key to a more equitable distribution of income in Bolivia is an increased tax burden on the higher income individuals. Rather than balancing the budget by eliminating the basic services of the state, such as health and education, considerations of equity and stability require that the government make increased efforts to secure an adequate tax base on the higher incomes. This might include a tax on land holdings and higher taxes on luxury consumption goods. A second key to a more equitable distribution of income would be greater public spending on education in the rural sector, where most of Bolivia’s poorest citizens live. Investment in the human capital of the rural peasantry is essential for long-term economic development. The second division to overcome is ideological, involving competing conceptions of the role of government. With the evident failures of state capitalism in the past two decades, there is a temptation on one side for a strict laissez-faire economic approach and, on the other side, for a fortified socialism. A more modulated approach is more likely to succeed. Such an approach would recognize the government’s responsibilities for infrastructure and social investments in health and education, but also recognize the limitations to the role of the state in the productive sector. Part of the push toward laissez faire in Bolivia is a frank acknowledgement of the limited capacity for honest, capable public administration in the country. But this limitation could be lessened by a concerted effort to raise the standards and capacity of the state bureaucracy. A determined effort at improved training of civil servants is vital in this regard. The third division is sectoral and regional. As we have noted, export diversification will require a change in emphasis to agriculture and light manufacturing, which in turn will surely entail some geographical shift in the locus of economic activity from the highlands to the lowlands. This kind of shift can be politically bruising and destructive if not handled with foresight and planning. The government will have to tread carefully between goals of allocating investment expenditures heavily toward the new sectors and regions, and the need to distribute the burdens and benefits of public spending in an equitable manner.

260

Juan Antonio Morales and Jeffrey D. Sachs

Notes Chapter 1 1. The hyperinflations in Peru (beginning in 1988) and in Argentina (1989) are the only other hyperinflations that did not occur following a war or revolution. 2. The inflation rate for the hyperinflation is measured as the twelve-month change for August 1985 over August 1984. The 1986 measure is December 1986 over December 1985. 3. See Malloy 1970 for an authoritative account of twentieth century Bolivian history before the Revolution, with an emphasis on the secular decline of the tin sector starting in the late 1920s. 4. There are several questions about the quality of the Bolivian National Accounts. Reported figures by the Central Bank of Bolivia differ, substantially in some cases, from those published by the IMF and the World Bank. Even more surprising, there are significant differences between the IMF and World Bank data. In the book, National Accounts data up to 1979 are sourced in Central Bank data because they are more complete. The discrepancies with other sources are relatively minor. We have preferred, in general, World Bank data for the period after 1980. It should be mentioned that the Central Bank substantially changed its methodology of national accounting after 1980. The new data seem to invite a thorough revision. 5. Coca production and consumption by Indian peasants in the Bolivian highlands have a secular tradition. Coca is used to fight hunger and the effects of physical stress in populations with very low standards of nutrition. Curative properties of coca leaves have also been known in the urban centers of Bolivia for centuries: brewed coca leaves are used regularly to cope with minor ailments as diverse as high-altitude sickness and toothaches. Many Bolivians, even in the cities, consider coca a miraculous plant. It must also be added that coca plants have a natural habitat in altitudes ranging between 1,000 and 2,000 meters. Bolivia has great extensions of land on those altitudes. Productivity there is very high, and little capital investment and labor are needed to grow the coca shrubs. Cropping and marketing are, however, very labor intensive, and most of the costs are incurred in those phases. High prices for the end-product coca usually mean high revenues for everybody in the chain, from middlemen to hitherto impoverished peasants. This partially explains why eradication is so unpopular, especially with poor peasants. 6. Bolivia briefly exported oil in the 1970s, but rising internal demand and falling reserves eventually led to an elimination of oil exports. Bolivia has continued to be a natural gas exporter to Argentina, but financial relations with Argentina have been unsettled since the mid-l980s, with continuing disputes between the countries over prices, volumes, and repayments schedules. In 1987 and 1988, Argentina went into arrears on payments to Bolivia, resulting in a serious loss of public sector revenues for Bolivia. Chapter 2 1. The Bolivian Revolution is a complex event that defies easy summary. It has been the object of excellent studies both in the United States (e.g., Malloy 1970; Malloy and Borzutsky 1982) and Bolivia. 2. In addition, the key political development of universal suffrage was also established.

261

BoliviaiNotes

3. From the very beginning there was technical and managerial trouble, adding a heavy burden to the fiscal budget and penalizing the mining sector, which was obliged to sell all of its ores to ENAF, the state enterprise that controlled the tin smelters. It should be noted that the construction of the tin smelters had been envisaged for decades, but private cost-benefit analysis had suggested the unprofitability of the operation. In the long run, it is possible that improvements in technology and especially in management could lower the costs of operation sufficiently to justify the project, but this has yet to happen. 4. In 1967, guerrilla units commanded by Ernest0 “Che” Guevara were defeated by the Bolivian Armed Forces, supported by U.S. intelligence and combat units. In the following years, there were several attempts, most of them by university students, to reorganize the guerrillas. At the time of the coup d’Ctat of August 1971, there were rumors of a strong Brazilian influence and even intervention among the military. This in itself would not be surprising in view of the place of Brazil in the geopolitics of the region. What is more important is the fact that the Brazilian model of development was regarded as the paradigm in economic matters by the new authorities. 5 . Paz Estenssoro broke away from the government in 1973 because of the small (and diminishing) role given to him and his party in the bureaucracy and in policymaking. FSB was a junior party in the coalition and it was dropped noiselessly after one of its leaders tried to stage a coup d’ttat in the eastern city of Santa Cruz. In spite of the rupture with the MNR and the FSB, Banzer retained in his bureaucracy some of the technocrats of both parties. Many of them would later join Banzer’s new party, the Acci6n Democratica Nacionalista (ADN).

Chapter 3 1. A good rationalization of this point can be found in the Estrategia Nacional de Desarrollo y Plan Cuatrienal (Ministry of Planning, Bolivia, 1984), 45-52. 2. Planning and a high degree of government intervention in the economy have a relatively long history in Bolivia. The extreme dependency on tin and the vagaries of that market in the 1930s led already in 1942 (in the so-called Bohan Plan) to a recommendation for govemment-induced production diversification and the opening of the eastern lowlands to modem agriculture and oil exploration. A United Nations mission in 1950 similarly recommended more planning and a more active governmental role in the economy and in the provision of social services. The MNR tried to implement this and Bohan’s advice after the Revolution of 1952. When President Kennedy created the Alliance for Progress program, one of the conditions to have access to the development funds provided therein was to have a national development plan. In 1962 the Paz Estenssoro administration announced a ten-year plan of economic development. At the core of this plan, there was a set of policies to promote an orderly recovery of tin exports to their pre-1952 levels. Some lip service was also paid to import substitution. One of the most important features of the ten-year plan resides in its investment program. Investment was to be oriented to the provision of social overhead to support production. With the exception of the needed funds for COMIBOL, little else was allocated for the direct involvement of government in production. The plan was also innovative in its basic needs approach to alleviate extreme poverty. In regard to the sources of financing, the plan emphasized the need to increase domestic savings. However, no specific measures

262

Juan Antonio Morales and Jeffrey D. Sachs

were included. Also in the plan there is an underlying two-gap model with a clear recognition of foreign exchange bottlenecks. 3. In contrast, the 1962 ten-year plan insisted more on social overhead investments for the peasantry holding individual plots of land. 4. In some other cases, such as the one with gasoline and other fuels, the public enterprises directly provided the subsidies to the consumer public. Those subsidies were not, however, deducted from the tax liabilities to the Treasury. 5. This discussion is based on F. Machicado (1987). 6. This discussion is also based on F. Machicado (1987).

Chapter 4 1. PPX measures the real volume of imports that can be purchased using the proceeds of export sales. The index is constructed as follows: PPX, = [P,,

* Xr/P,,] / [P,* x

P,lO

*

100,

where year zero is the baseline year (1980) in table 4.1, P, is the price of exports, P, is the price of imports, and X is the volume of exports. Note that movements in PPX can be decomposed into movements in the terms of trade and movements in the volume of exports: PPX, = [(P,t/P,,)/(P,o/P,o)l* Kr/xo). Bolivia in fact suffered large declines in the volume of exports (which fell by 31 percent between 1980 and 1988) and in the terns of trade (which fell by 11 percent between 1980 and 1988), according to the 1988 data of ECLAC, tables 8 and 10. 2. The foreign trade policies between 1964 and 1978 are discussed in detail in Morales ( 1982). 3. Contraband color television sets, personal computers, stereos, and even automobiles have been sold in the open market in the most important Bolivian cities at below international prices since the mid-1970s. Allegedly, this is possible because those contraband imports serve to launder drug money (or “coca-dollars”). 4. The system of prior import deposits consisted of a 120-day deposit in Bolivian pesos in the domestic banking system to which the private importers were liable. The deposit had to be equal to a given percentage, with a ceiling of 25 percent, of the CIF value of the imported goods. These deposits did not earn interest and were subject to a 100 percent reserve requirement. 5. The government tried, in doing so, to avoid a confrontation with the IMF, since most forms of administrative allocation of foreign exchange could have been judged a violation of Article VIII of the Articles of Agreement of the IMF. Registration of the demands for foreign exchange had a mild deterrent effect on capital flight. 6. The use of the foreign exchange system for several functions other than equilibrium in the foreign exchange market has been very clearly stated by Cooper (1971, 10-13). 7. Errors and omissions picks up capital flight that corresponds to measured exports and imports (i.e., when measured exports do not contribute to a rise in foreign exchange reserves or other measured foreign assets, the discrepancy appears in errors and omissions). However, underinvoiced exports or wholly uncounted exports (e.g., cocaine) will contribute to capital flight that is unmeasured by errors

263

BolividNotes

and omissions. It is true that understatement of imports may lead to an overstatement of capital flight, but in the Bolivian context, the understatement of exports (particularly drug-related ones) is likely to exceed the understatement of imports. 8. When the UDP government took office in the fall of 1982, one of the first measures taken was to eliminate the use of dollar-denominated contracts in the financial system. All deposits and liabilities that had been dollar-denominated were obliged to be converted into peso-denominated assets at a stated exchange rate. This led to a massive disintermediation of the banking system and to a flight for safer assets, particularly foreign exchange and other foreign assets.

Chapter 5 1. In 1982, according to the World Bank Debt Tables, the MLT public debt constituted 87 percent of the total external debt of Bolivia. The same ratio in 1985 was 82 percent. These data suggest that the MLT public debt is by far the most important component of total external debt. 2. To some extent, the effective subsidy to debtors was paid for by an effective tax on depositors. Specifically, depositors with dollar-denominated claims on the banks had their claims converted into pesos at a disadvantageously low exchange rate. This helped the banks to absorb the costs of converting their claims on private sector debtors from dollars to pesos at the same disadvantageous exchange rate. 3. This legal argument has frequently been employed in the internal political discussions on the debt, particularly by advocates of unilateral debt suspension. 4. A complete evaluation of the National System of Projects appears in JUNAC (1985). 5. Among them was the investment banking firm of Salomon Brothers. 6. The four tranches and the terms of the refinancing were as follows: Tranche IA: Short-term debts past due between August 1980 and April 1981. Maturity of 3% years and a grace period of 1 year. Interest charges at LIBOR 2 percent. Tranche IB: Medium- and long-term debts having reached maturity between April 1980 and April 1981. Maturity of 6 years and a grace period of 2 years. 2% percent. Interest charges at LIBOR Tranche 11: Medium- and long-term debts reaching maturity between April 1981 and April 1982. Maturity of 5 years and a grace period of 4 years. 2% percent. Interest charges at LIBOR Tranche 111: Medium- and long-run debts reaching maturity between April 1982 and May 1983. No agreement was reached on maturity or grace period. Interest charges at LIBOR 2% percent.

+

+

+ +

In addition to these terms, Bolivia was charged a myriad of fees and organization costs: a once-and-for-all flat fee of 1 percent of the rescheduled debt for the coordinator bank; an annual fee of U.S.$49,000for the coordinator bank; a fee of 0.375 percent on tranche IA and 1.125 percent on tranches IB and 11; and all organizational and legal expenses.

Chapter 6 1 . If hyperinflations are ranked by their maximal inflation rates over half-year calendar intervals (January-June, July-December), the Bolivian hyperinflation ranks

264

Juan Antonio Morales and Jeffrey D. Sachs

seventh in the twentieth century, following: Hungary (1946: l), Greece (1944:2), Germany (1923:2), China (1949:1), the Soviet Union (1922:1), and Poland (1923:l). 2. Cagan defined the beginning of hyperinflation as the month in which inflation exceeds 50 percent, and the end of hyperinflation as the month in which inflation last exceeds 50 percent and is then followed by twelve months of less than 50 percent inflation. 3. In this case, m, e , t? are measured at the end of the month; T,is the change in the monthly price level. Ideally, p t should be an end-of-month price, with T,= P, - P,-,. Since prices are available only on a monthly average basis, end-of-month prices for month b - 1 are proxied by a geometric average of averageperiod t and t - 1 prices.

Chapter 7 1. Monetary reform refers to the switch from one currency to another as part of the stabilization program. Argentina switched from the peso to the austral in June 1985 under the Austral Plan; Brazil switched from the cruzeiro to the cruzado in March 1986; and Peru adopted the inti in place of the peso in September 1985. The precise economic content of such a shift in currency depends on several factors, including: (1) whether there is a tax on old money balances as part of the conversion process, and (2) the terms under which long-term contracts in the old currency are to be settled in the new currency. 2. This paragraph, of course, describes a highly schematic view of the linkages between the budget and inflation. The point is not to be comprehensive, but to emphasize the crucial link in the Latin American context between budget deficits and the exchange rate, and between the exchange rate and the domestic price level. 3. As a formal matter, an overvalued exchange rate combined with a black-market exchange rate acts as a tariff. Marginal imports are priced according to the depreciated black market rate, while exports (to the extent that they are not smuggled) receive the lower official exchange rate. Thus, the gap between the official and the black-market rate acts as a tax on exportables by reducing their price relative to importables. As the gap widens, the resource misallocations caused by the overvaluation of the official rate eventually become so enormous as to force an official devaluation. 4. In a famous policy debacle, Chile fixed the peso-dollar exchange rate in the midst of a 35 percent annual inflation during 1979. While tradable goods prices soon stopped rising in Chile, nontradables prices continued to rise for two more years, in part because of a formal system of backward-looking wage indexation. The exchange rate fixing, combined with a huge inflow of short-term capital upon a liberalization of the financial system, led to an enormous squeeze of the tradable goods sectors in the country. Eventually the exchange rate had to be sharply devalued. The inadvisable fixing of the exchange rate is probably one factor that contributed to the severe recession in Chile in 1982 (the other factors include a reversal of the short-term capital inflow, a terms-of-trade decline, and the rise in world interest rates). 5. At the time of the stabilization program, official bread prices were set at a tiny fraction of world market prices. There was, however, very little bread available at the ridiculously low official prices. Rather, flour and bread were smuggled to Peru, effectively setting the internal bread price in Bolivia at the level of the border price, net of the costs of transport and smuggling. Most bread available on the streets of La

265

BolividNotes

Paz was sold at many times the official price. The middlemen, rather than the consumers of bread, were apparently the largest beneficiary of the subsidized grain prices.

Chapter 8 1. No bank will lend to the Argentine government, for example, even for a highly profitable public investment, for fear that the loan will simply become part of Argentina’s overall bad debt. It is dangerous to lend even if the individual project has a good return. 2. Negotiations may break down repeatedly, at high cost (e.g., with disruptions of normal trade financing), because the various parties have a continuing incentive to posture, to act tough, etc. For theoretical analyses, see Fernandez and Kaaret (1988) and Rotemberg (1988). 3. Even if the banks know that the debt cannot be paid, they may still impose sanctions for nonpayment, in order to impress other debtors with whom they are negotiating. 4. Consider a case in which a wonderful economic reform is identified, which would cost $100 million of current consumption in order to raise the debtor’s future income and its debt-servicing capacity by $200 million in present value. Suppose also that all of this incremental debt-servicing capacity would be squeezed out of the country by the foreign creditors in the course of future negotiations. The debtor has no incentive to undertake the reform, even though it has a very high return, because the benefits accrue to the foreign creditors. However, if the country first entered into a buyback in which it paid $60 million for the $1 billion in debt, thereby cancelling the debt overhang, the debtor would then be free to undertake the investment and to reap the large returns. Note that this incentive effect could work by incentives on a given government (by leading the government officials to a rejection of specific public investments or public sector reforms) or through the electoral process (by contributing to the election of governments that oppose the reform efforts). 5. As an example, the country could negotiate with creditors to use the receipts of future export earnings as collateral for future debt-service payments, in cases where it would be administratively possible to arrange for future export earnings to accumulate in an escrow account out of reach of the country. 6. There are probably only four or five such banks in the United States. They are, however, among the biggest banks. Examples include Citicorp, Bank of America, Chemical Bank, Chase Manhattan, and Manufacturers Hanover. 7. I have been closely involved in the Bolivian debt negotiations as the government’s main outside economic advisor. 8. There were no working meetings with the banks during this crucial period, as it was the official negotiations with the IMF and discussions with the U.S. government that were of most importance. 9. There are several complex reasons for this change of position. Most important was the ferocity of the crisis in Bolivia (with real per capita GNP having declined by almost 30 percent, with the terms of trade having collapsed, and with a virulent hyperinflation during 1984-85), combined with the strength of Bolivia’s adjustment program (which eliminated tens of thousands of jobs in state enterprises and closed the budget deficit by more than 10 percent of GNP almost overnight). Also, the

266

Juan Antonio Morales and Jeffrey D. Sachs

United States had important foreign policy interests in stabilizing democracy in the country, since Bolivia borders most of the large countries of South America and has often been feared as a focal point of unrest (Che Guevara died in the Bolivian jungles in 1967). Moreover, the United States was interested in pursuing an anti-cocaine policy in the region, which could only be accomplished with a friendly, stable government. 10. In 1988, Bolivia signed a three-year Enhanced Structural Adjustment Facility with the IMF, based on a program of balance of payments that presumes that Bolivia’s remaining debt will be settled on terms similar to the buyback. 11. The financial “costs” to Bolivia of the debt strategy have been minimal. If we judge the net cash costs of the buyback to Bolivia at $20 million, the country has paid in total over three years less than 1 percent of one year’s GNP ($20 million/ $3 billion) to its commercial bank creditors. At the same time, Bolivia has received large net resource inflows from the official creditor community. Thus, in contrast to Bolivia’s net resource inflow, all of the other countries in the region have been making large net resource transfers to the foreign creditors (as Bolivia did during 1982-84).

Chapter 9 1. Bolivia has already experienced the disaster of governance in conjunction with the coca mafia under the dictatorship of General Luis Garcia Meza. Garcia Meza’s brutal role led to the worldwide diplomatic isolation of Bolivia, which in turn contributed importantly to Bolivia’s financial rupture with the rest of the world. 2. In other words, the export of coca paste has caused a secular appreciation of the Bolivian real exchange rate, thereby squeezing the profitability of legal export sectors. 3. Taxation of coca production and exports is virtually impossible from an administrative point of view and is strenuously opposed by the international community. It is judged that normal taxation of the sector would effectively lead to its legalization. 4. In 1986, for example, the Bolivian government undertook joint military operations with the U S . government to interdict coca paste exports and to drive down the internal price of coca leaf. The President faced impeachment proceedings for his actions at the hands of nationalist elements in the Bolivian Congress. The action provoked a run on the peso and a nationwide banking crisis, as individuals sought to remove local deposits in anticipation of a currency devaluation. In the end, the peso was successfully defended.

References Baptista, Fernando. 1985. Estrategia Nacional Para La Deudo Externa. La Paz: Editora Siglo. Bolivia, Ministerio de Finanzas. 1978. Memorandum: Sobre El Financiamiento Externo de Bolivia. La Paz: Instituto Nacional de Financiamiento. Externo (INDEF).

266

Juan Antonio Morales and Jeffrey D. Sachs

United States had important foreign policy interests in stabilizing democracy in the country, since Bolivia borders most of the large countries of South America and has often been feared as a focal point of unrest (Che Guevara died in the Bolivian jungles in 1967). Moreover, the United States was interested in pursuing an anti-cocaine policy in the region, which could only be accomplished with a friendly, stable government. 10. In 1988, Bolivia signed a three-year Enhanced Structural Adjustment Facility with the IMF, based on a program of balance of payments that presumes that Bolivia’s remaining debt will be settled on terms similar to the buyback. 11. The financial “costs” to Bolivia of the debt strategy have been minimal. If we judge the net cash costs of the buyback to Bolivia at $20 million, the country has paid in total over three years less than 1 percent of one year’s GNP ($20 million/ $3 billion) to its commercial bank creditors. At the same time, Bolivia has received large net resource inflows from the official creditor community. Thus, in contrast to Bolivia’s net resource inflow, all of the other countries in the region have been making large net resource transfers to the foreign creditors (as Bolivia did during 1982-84).

Chapter 9 1. Bolivia has already experienced the disaster of governance in conjunction with the coca mafia under the dictatorship of General Luis Garcia Meza. Garcia Meza’s brutal role led to the worldwide diplomatic isolation of Bolivia, which in turn contributed importantly to Bolivia’s financial rupture with the rest of the world. 2. In other words, the export of coca paste has caused a secular appreciation of the Bolivian real exchange rate, thereby squeezing the profitability of legal export sectors. 3. Taxation of coca production and exports is virtually impossible from an administrative point of view and is strenuously opposed by the international community. It is judged that normal taxation of the sector would effectively lead to its legalization. 4. In 1986, for example, the Bolivian government undertook joint military operations with the U S . government to interdict coca paste exports and to drive down the internal price of coca leaf. The President faced impeachment proceedings for his actions at the hands of nationalist elements in the Bolivian Congress. The action provoked a run on the peso and a nationwide banking crisis, as individuals sought to remove local deposits in anticipation of a currency devaluation. In the end, the peso was successfully defended.

References Baptista, Fernando. 1985. Estrategia Nacional Para La Deudo Externa. La Paz: Editora Siglo. Bolivia, Ministerio de Finanzas. 1978. Memorandum: Sobre El Financiamiento Externo de Bolivia. La Paz: Instituto Nacional de Financiamiento. Externo (INDEF).

267

BolividReferences

Bolivia, Unidad de Analisis de Politica Economica (UDAPE). 1985. Analisis de la Gestion Economica en Bolivia: 1982- 1985. La Paz: October, mimeo. . 1985. Diagnostico de la Situacion del Sistema Fiscal en Bolivia. La Paz: December, mimeo. . 1985. Las Estadisticas Fiscales en Bolivia: 1970- 1985. La Paz:July, mimeo. . 1986-87. Analisis de Coyuntura. La Paz, mimeo (monthly). . 1987. Bolivia: Informacion y Estadistica 1986. La Paz, mirneo. Breuer, Luis E. 1988. Hyperinflation and stabilization: The case of Bolivia 1984-86. Ph.D. diss., University of Illinois at Urbana-Champaign. Cagan, Philip. 1956. The monetary dynamics of hyperinflation. In Studies in the quantity theory of money, ed. M. Friedman. Chicago: University of Chicago Press. Cooper, Richard. 1971. Currency devaluation in developing countries. Essays in International Finance, no. 86. Princeton University. Dona Medina, Samuel. 1986. La economia informal en Bolivia. La Paz: Editorial Offset Boliviana. Dornbusch, Rudiger, and Stanley Fischer. 1986. Stopping hyperinflation: Past and present. NBER Working Paper no. 1810. Cambridge, Mass.: NBER, January. Eder, George. 1968. Infation and stabilization in Latin America: A case history of injution and stabilization in Bolivia. Ann Arbor: University of Michigan Press. Fernandez, R., and D. Kaaret. 1988. Bank size, reputation, and debt renegotiation. NBER Working Paper no. 2704. Cambridge, Mass.: NBER, September. Gillis, Malcom. 1978. The structure of Bolivian mining taxes. In Taration and mining. Nonfuel minerals in Bolivia and other countries, ed. M. Gillis et al. Cambridge, Mass.: Ballinger. International Monetary Fund (IMF). 1986. Bolivia: Recent economic developments. Washington, DC: IMF, mimeo, December. Junta del Acuerdo de Cartagena (JUNAC). 1985. Estudio de Un Sistema Nacional de Proyectos para Bolivia. Lima: JUNAC, mimeo. Komai, Janos. 1986. Contradictions and dilemmas: Studies on the socialist economy and society. Cambridge, Mass.: MIT Press. Machicado, Flavio. 1987. Las Finanzas Publicas y la Inversion. IDLIS, Sene Estudio, Diagnostico, Debate. La Paz: ILDIS. Malloy, James. 1970. Bolivia: The uncompleted revolution. Pittsburgh: University of Pittsburgh Press. Malloy, James M., and Silvia Borzutsky. 1982. The praetorianization of the Revolution: 1964-1967. In Modern day Bolivia, ed. J. Ladman. Tempe, AZ: Arizona State University. Malloy, James M., and Eduardo Gamarra. 1987. The transition to democracy in Bolivia. In Authoritarians and Democrats: Regime transition in Latin America, ed. James Malloy and Mitchel Seligson. Pittsburgh: University of Pittsburgh Press. . 1988. Revolution and reaction. New Brunswick: Transaction Books. Morales, Juan Antonio. 1982. The Bolivian external sector after 1964. In Modern day Bolivia, ed. J. Ladman. Tempe, AZ: Arizona State University. Morales, Juan Antonio, and Jeffrey Sachs. 1989. Bolivia’s economic crisis. In Developing country debt and the world economy, ed. J. Sachs. Chicago: University of Chicago Press. Morales, Juan Antonio, Rodolfo Ulloa, and Gloria Jimenez. 1978. Analisis economico de las implicaciones para Bolivia del Arancel Externo Comun del Crupo Andino. La Paz: Informe a1 Ministerio de Finanzas de la Republica de Bolivia, November. Rivas, Hugo. 1986. El Modelo de Ecumulacion y el Endeudamiento Externo en Bolivia: 1972- 1983. Tesis de Licenciatura, Lima: Pontificia Universidad Cat6lica del Peru.

268

Juan Antonio Morales and Jeffrey D. Sachs

Rotemberg, Julio. 1988. Sovereign debr buybacks can lower bargaining costs. MIT, mimeo, October. Sachs, Jeffrey. 1986. The Bolivian hyperinflation and stabilization. NBER Working Paper no. 2073. Cambridge, Mass.: NBER, November. . 1987. The Bolivian hyperinflation and stabilization. AEA Papers and Proceedings, 77, no. 2 (May). Sachs, Jeffrey, and Juan Antonio Morales. 1988. Bolivia, 1952-1986. International Center for Economic Growth, Country Studies no. 6. San Francisco, CA: ICS Press. Sargent, Thomas J. 1982. The ends of four big inflations. In Injation: Causes and effects, ed. R. E. Hall. Chicago: University of Chicago Press. Tanzi, Vito. 1977. Inflation, lags in collection and the real value of tax revenue. IMF Stuff Papers, vol. 24. Tsiang, S. 1986. Reasons for the successful economic takeoff of Taiwan. Bank of China Economic Review, July-August. Ugarteche, Oscar. 1986. El Estado deudor, economia politica de la deuda: Peru y Bolivia 1968-1984. Lima: Instituto de Estudios Peruanos. World Bank. 1985. World Development Report, 1985. New York: Oxford University Press. . 1988. World Development Report, 1987. New York: Oxford University Press.

1

Introduction

1.1 Pre-World War I1 Experience Historical perspective shows that external debt is nothing new to Brazil. The accumulation of debt began shortly after the proclamation of independence in the 1820s. During the nineteenth century the government continuously borrowed abroad, mainly through the intermediation of British investment banking houses. The proceeds were on the whole applied to the construction of needed public works; once, however, in 1865, debt was contracted explicitly to support the war then underway with Paraguay. As an ever more active participant in the international economy after 1850, led by its increasing exports of coffee, Brazil stood second only to Argentina in Latin America as a recipient of foreign capital in the nineteenth century. As a consequence, Brazil benefitted from a growing railway network which supported its international trade, but also was subject to variable conditions of external capital supply. When English capital flows to Argentina came to an abrupt halt in 1890 in the aftermath of the Baring crisis and railroad overinvestment, the effect spilled over to the valuation of all South American securities and hence Brazilian creditworthiness. In the Brazilian case, the much increased domestic money supply-a product of new economic policies during the first years of the Republic-played a more important role than diminished foreign investment in the massive depreciation of the exchange rate during the decade. Brazil did not experience a debt crisis until the declining price of coffee decisively began to erode export receipts after 1895. Eventually, in 1898, it was necessary to contract a funding loan to capitalize prospective external interest payments for the next three years. As coupons on the prior debt came due, they would be redeemed by the new issue. In addition, amortization on much of the debt was postponed for thirteen years. The loan, however, was conditional on domestic stabilization. 271

272

Eliana A. Cardoso and Albert Fishlow

Finance Minister Joaquim Murtinho presided over a deflationary shock treatment. Paper currency was destroyed in proportion to drawings on the funding loan. Government income was increased largely through the imposition of new internal consumption taxes and later through a rising gold quota on import duties. Adjustment took place through a classical curtailment of income, imports, and investment. With recovery of the balance of payments, borrowing from abroad was renewed during the next cycle of international credit expansion at the beginning of the twentieth century. Foreign debt was now contracted to finance Brazilian stockpiling of coffee, which took place in order to help sustain its price. Other issues were applied more conventionally to infrastructure projects. So long as export earnings continued to grow, there was no problem sustaining increasing debt service. But by 1913, the cumulative effects of a weakened external performance amid continuing government deficits and mounting uncertainty about the capacity to sustain convertibility made new arrangements inevitable. In 1914, soon after the start of the First World War, a second funding loan became necessary. The 1920s saw another period of expansion, this time largely financed from expanding New York capital markets and extensively contracted by state and local governments. Other Latin American countries did the same: between 1926 and 1928, almost one-third of all foreign bonds issued in New York were of Latin American origin. Creditworthiness of developing countries was aided by the wartime inflation that reduced the burden of previous debt and by high commodity prices at the beginning of the decade. By 1929 commodity prices were already weakening. Soon thereafter, prices collapsed and so did the capital market. With the Great Depression, there were no additional capital outflows and the inevitable, and generalized, debt moratorium soon followed. Latin American countries were prominent among those declaring a moratorium. With Bolivia leading the way, Brazil, Peru, and Uruguay also stopped servicing their debts in 1931; they were joined in the following year by Argentina (albeit only on its dollar debts) and by Cuba. Brazil resumed partial payments in 1934, but because of a reduced trade balance in 1937 it ceased service again. This default was ranked by The Economist “among the most cynical that the London market remembers.” Such offense taken at the Brazilian decision was exaggerated during a decade in which international obligations were more honored in the breach than the observance.2

’

1.2 The Postwar Period In 1943, aided by wartime trade surpluses, Brazil rescheduled its debt on favorable terms, reducing the outstanding balance considerably. Bondholders also gained from the subsequent appreciation in market prices of Brazilian bonds. Brazil’s external debt contracted from its peak level of U.S. $1,300

273

BrazilKhapter 1

million in 1931 to some $600 million in 1946. Further real reduction was achieved through a new bout of wartime inflation. As a result, Brazil entered the postwar period with a virtually clean slate. It was not long before new entry of capital became necessary to finance an ambitious industrialization drive. Brazil, disappointed by the absence of expected large U.S. public lending in the aftermath of close wartime collaboration, had to turn to private sources instead. Total capital inflows, both from direct investment and loans, increased sharply after 1955, especially in the form of medium-term suppliers credits. Inflation and external debt again became part of the Brazilian scene, as budget deficits increased and balance of payments problems became more severe. President Juscelino Kubitschek, rather than jeopardize rapid growth and his industrialization Target Program, broke with the IMF and found finance elsewhere through reductions in reserves, short-term swap loans, and other compensatory operations. At the end of 1960, the external debt stood at double its 1955 level and the new president, Janio Quadros, faced a serious debt and accelerating inflation problem. In an effort to resolve these financial difficulties, ties with the IMF were restored and more orthodox policies followed. But that effort at stabilization was ended prematurely with Quadros’s resignation in August 1961. To economic disequilibrium was added intense political agitation. Quadros’s successor, Jog0 Goulart, was viewed with suspicion by conservative forces in society as a populist, at best, and perhaps more radically inclined. The withdrawal of the IMF mission in 1962, after the stabilization program had been irreparably weakened by the concession of large public sector wage increases, complicated access to needed new external resources. Despite the new funds potentially available under the aegis of the Alliance for Progress, the United States was increasingly reluctant to support Goulart’s reform efforts. Finally, as the economic situation deteriorated, a program loan was negotiated with the United States in 1963, but the conditions for its disbursement were never fulfilled. During this turbulent period, capital inflow had virtually ceased. The World Bank, previously an important source of official resources for Brazil, did not authorize a single loan to the country between 1960 and 1964. When the military seized power in April 1964, inflation had mounted to almost 100 percent on an annual basis and housewives were protesting in the streets. The middle class welcomed the demise of a Goulart government that was both leftist and ineffectual. The new government predictably found a more favorable international reception. Debt rescheduling was facilitated, and new loans from the Agency for International Development as well as new IMF credits were made available. Debt rescheduling and new credits did not prevent a stabilization-induced recession that was the result of contractionary monetary and fiscal policies. Import demand fell because of the reduction in aggregate demand; combined

274

Eliana A. Cardoso and Albert Fishlow

with a recovery of exports, this produced trade surpluses in 1964-66 and the elimination of arrears and other short-term debts. The supportive international environment, however, did help prevent an even steeper decline in income and allowed scope for the resumption of import growth in the early stages of recovery in 1967.

1.3 Integration into World Capital Markets and the Oil Shocks As Brazilian growth accelerated after 1967, the government consciously embarked on a policy of tapping private capital markets to underwrite rapid expansion. Brazil was one of the first countries to respond when in the late 1960s the Eurodollar market first welcomed and then actively sought developing countries to which to lend. Brazil was also the developing country that relied most extensively on this external capital market, implementing a debt-led model of development to help finance the mushrooming capital and intermediate goods imports associated with the “miracle” of rates of growth averaging close to 10 percent between 1968 and 1973. Another component of the strategy was explicit adaptation to inflation through indexation, not only of wages, rents, and financial assets (as introduced in 1964) but also of the exchange rate in 1968. When the oil shock struck in late 1973, Brazil was the largest oil importer among the developing countries. Already faced with the prospect of weakening growth as the boom matured, the government was not inclined to risk a significant decline in real income from the sudden adverse terms of trade. Increased external debt was an attractive alternative. Borrowing could postpone the contractionq effects of the petroleum “tax” and permit domestic expansion to proceed. After a surge of imports and an increased current account deficit, Brazil opted for adjustment through an ambitious program of generalized import substitution rather than export promotion or domestic recession. An elastic supply of debt that responded to Brazilian requirements and brief episodes of slowed domestic activity were adequate to keep the balance of payments under control after 1973. Higher coffee and other commodity prices also contributed by reversing the terms of trade decline. Brazilian economic performance after the oil price shock remained above its trend level of growth (of 7 percent per year), and the results evoked widespread admiration for the success of the growth-led debt model. Now growth requirements determined the need for debt, rather than available capital permitting higher rates of growth. Petro-dollar recycling apparently had worked to transfer considerable resources to worthy developing countries, sustaining high rates of investment and economic growth. What had been only expedient to offset a disequilibrium in the world balance of payments, turned into significant development finance.

275

BraziKhapter 1

On the eve of the second oil shock, the Brazilian debt was the largest in the world. As such, it would be especially vulnerable to the reversal in interest rates that was in the offing. Indeed, the debt had become so great that already in the late 1970s, before the sudden surge in interest rates, the reverse flow of service payments was beginning to cancel the new inflows. Whereas the first stage of debt accumulation saw a large transfer of real resources, in later stages more and more borrowing was going simply to cover interest obligations on earlier loans. The dynamics of debt-led debt had become part of the Brazilian story, a process that rising interest rates would magnify with a vengeance. Not surprisingly then, in 1980, after oil prices had risen and interest rates were rising, Brazil was one of the first countries to test the response of the private capital market to a large debtor encountering a balance of payments crisis. Limited additional finance was available and only on more expensive terms as spreads widened. Oil producers were the favored borrowers, not oil importers. Well before the generalized debt crisis in 1982, Brazil was forced to introduce a more austere set of policies and domestic adjustment in 1981. For the first time in the postwar period, income declined. Brazilian discipline was rewarded with new capital flows. Markets seemed able to respond to debt problems with the right dosage of conditional liquidity, and countries seemed able to implement the right amount of belt-tightening. When Mexico declared its virtual default in August 1982, Brazil, itself subject to the adverse balance of payments consequences of sharply higher interest rates, replayed its earlier approach to private creditors. The Brazilian government insisted that their situation was distinct from that of Mexico and capable of simple remedy. They claimed that Brazil continued to be creditworthy owing to its own domestic policies and that it had only minimal need for new resources. With a congressional election in November threatening the government’s control, politics precluded any official appeal to the IMF until after the votes had been counted. Then the inevitable acceptance of IMF supervision occurred, and Brazil joined the rapidly lengthening queue of problem cases. But Brazil did so with the disadvantage of having its own, and inadequate, stabilization program on the table beforehand, which only served to reduce the size of the finance made available. Brazil was therefore impelled to a stronger adjustment of its external accounts. What differentiated Brazil from other large Latin American debtors was a greater export recovery in 1984. Between 1982 and 1984, export receipts rose by 35 percent; with a decline in imports of almost 30 percent, the current account was quickly restored to balance and the foreign exchange constraint became less pressing. Output in 1984 was already on the way up, led by industrial exports. While the results were worse than the Great Depression for Brazil, the intervening decline in per capita income

276

Eliana A. Cardoso and Albert Fishlow

was smaller than for other problem debtors. Brazil’s more diversified industrial base allowed for both export increases and import substitution. In this sense, Brazilian adjustment was more successful than elsewhere. In another sense, it was less successful-progress on the external account was not matched internally. Successive letters of intent under IMF programs had no sooner been dispatched than they were made obsolete by accelerating inflation that did violence to the monetary targets. That experience led to the development of new deficit concepts, adjusted for the widespread indexation of government debt, which are now widely applied to other countries. The Brazilian experience also led many, but not all, to the understanding that improvement of the balance of payments and domestic equilibrium were not tightly linked. Indeed, trade surpluses might themselves create new problems for macroeconomic policy.

1.4 The Lingering Debt-Shock Aftermath A new civilian government took office in March 1985, and the rules of the stabilization game profoundly changed. In the hands of these leaders, Brazil’s large export surplus, only recently established, became a potent instrument of independence. Relations with the IMF deteriorated, and previous plans for a multiyear rescheduling agreement with the banks were scrapped. Although interest was fully paid, there were no inflows of new capital. Recovery was now based on internal demand, with limited import increase. Inflation began to accelerate from already record levels of more than 200 percent. Despite strong output growth, there were additional worrying signs: high real interest rates and increasing government internal debt service, low investment rates and stagnant exports. Bold measures were called for. The Pluno Cruzado was implemented in February 1986 as a substitute for a conventional recession-based strategy of stabilization. The similar Argentine Austral Plan had been put into place seven months earlier. Both were based upon the premise that high rates of inflation were driven by the inertial, self-replicating force of indexation. Accordingly, the Cruzado Plan forced a sophisticated, short-term standstill that maintained real income positions and abolished future indexation. Henceforth, inflation would be zero. In the words of the then finance minister, Dilson Funaro, Brazil would be a country of ‘‘Swiss inflation and Japanese growth.” For a few months it seemed true, and there was generalized euphoria. But signs of disequilibrium from excess demand mounted without inducing adequate compensatory response. Another election loomed, and in the best Brazilian political tradition, corrective actions were placed on hold. Indirect tax increases, announced immediately after the election, proved much too late and much too little. Because their immediate impact was to raise prices, they were ineffective. The dam of controls had broken, and

277

BrazilKhapter 1

there was no restoring an orderly process of readjustment of prices and wages. Events rapidly moved out of control as inflation rates mounted. There seemed to be policy impotence. The deterioration in the balance of payments was as significant as the mounting internal problem. Exports declined sharply in the last quarter of 1986 as imports continued their modest rise, and the trade surplus rapidly eroded. Suddenly, Brazil’s comfortable cushion of reserves had disappeared. In February 1983 Brazil declared a moratorium on the largest part of its commercial bank debt. In part, it was a strategy intended to appeal to internal nationalist sentiments and to strengthen support for a beleaguered president; in part, it was (as it turned out) a misguided effort to obtain broad international support and creditor-government intervention in behalf of a new political solution to the debt problem; and in part, it was a final and unavoidable act reflecting errors in macroeconomic policy that contributed to a new liquidity crisis. Thus Brazil had come from being the epitome of the successful and well-behaved debtor to being the challenger of the entire debt regime, unwilling to accept the burden of external adjustment at the expense of continuing economic growth. In the end, and not very long after the event, Brazil was forced to yield and to devise an alternative strategy that incorporated negotiations with the banks. There appears to be little sympathy in the industrialized countries for public intervention or broader solutions. The closest anyone has come to bold innovation is Citibank’s dramatic upward revision of loan-loss reserves, now followed by all other major banks and reaching $20 billion in total. That accounting change creates new possibilities for writing down debts on the bank books, but there is still a need to translate it into equivalent benefits for the debtor countries. The Brazilian renegotiation in 1988 once again moved the Brazilian debt to center stage, but with disappointing results. Not until the spring of 1989 would official policy move to the side of debt reduction. When it did, Mexico, and not Brazil, was the first beneficiary.

1.5 Overview This capsule history highlights several issues that we will analyze in the following chapters in more detail. One theme brought out by the Brazilian debt experience is the importance of the political economy of the domestic response. Borrowing was frequently used as a conscious policy instrument to achieve immediate gains. That is the sometimes fatal attraction of debt: it holds out the prospect of something for nothing, at least for the time being. But the piper must be paid. Brazilian policymakers incorporated reliance upon external finance into their adjustment strategy after 1973. When the country was unexpectedly buffeted by a more adverse international economy, their successors were faced with the

278

Eliana A. Cardoso and Albert Fishlow

consequences. The vaunted Brazilian technocratic capacity more than met its match in the demands placed upon policy by the oil and interest rate shocks. Ministers did not hesitate to improvise. Authoritarian though the government was, it still could not risk the consequences of an orthodox response to external disequilibrium. Policies of debt management were not always optimal. Sometimes they were flawed by inappropriate assumptions and incomplete models. Almost always they bowed to broader political objectives and considerations. For the most part in this period, a controlled transition toward greater civilian participation was an important goal. In chapter 2 we trace the origins of the debt strategy in the 1970s and its execution from 1973 to 1979. Another side of the blend of politics and economics was the requirements of adjustment to the changed international environment. In chapter 3 we examine the period of pre-IMF adjustment to the oil shock, the decision to seek a Fund agreement in 1982, the sequence of letters of intent and IMF-sponsored stabilization, and the conjunction of accelerating growth and inflation under the New Republic. Three essential elements emerge. One is the error of the heterodox initiatives adopted by the government in 1980 in a vain effort to sustain rapid economic growth in the face of the deteriorating international economy. Second is the inadequacy of the Fund stabilization program in its initial finance, in its nominal monetary and fiscal targets inadequate to a highly inflation-indexed economy, and in the striking divergence between successful achievement of trade surpluses and increasing domestic disequilibrium. Third is the commitment of the new civilian government to design its own policies and to seek its own solutions, of which the Cruzado Plan was the eventual outcome. The focus in chapter 4 is on the design of efficient stabilization programs: how to stop inflation without inducing large output costs. The basis of the Cruzado Plan was recognition of the central role of indexing in projecting price increases into the future. To stop inflation effectively under such conditions requires a coordinated standstill of wages and prices. In order to work, the plan required the elimination of other sources of inflationary pressure, especially fiscal deficits. In the event, fiscal policy was not neutral and wage policy encouraged rather than restrained the increases in real wages that had already been in motion as a result of the 1985 expansion. A comparison of the failure of the Cruzado Plan with the success in 1964-67 in decelerating inflation suggests three lessons. One is the greater ease of gaining control over the fiscal deficit when public debt is small and there is ready foreign aid. The second is the excessive boldness of the Cruzado Plan in aiming for zero inflation and abolishing indexation. While indexation contributes to the propagation of inflation, it is also protection against the kind of volatile inflationary acceleration that occurred at the end

279

BrazilKhapter 1

of 1986 and beginning of 1987. The third is the difficulty of managing neutral disinflation. In 1964 wage compression permitted continued progress in reducing the rate of inflation. In 1986 real wages were increasing at the expense of profit margins, helping to provoke shortages, disorderly growth, and the creation of black markets. This central question of the relation between budget deficits, external finance, and inflation is the focus of chapter 5. Our contribution is twofold. The first is the clarification and measurement of these deficits in an indexed setting. The second is presentation of a simple, two-equation model that incorporates debt finance in an open economy and specifies inflation dynamics in response to excess demand and indexing. It is designed to illuminate the linkage between the growing inability to externally finance the public sector deficit after 1979 and the acceleration of inflation. The debt problem here is shown to be much deeper than its balance of payments consequences. The domestic macroeconomic structure is also profoundly influenced by the suddenly limited access to external finance. In chapter 6 we turn to the trade component of the debt problem. The responsiveness of imports and exports to devaluation is key to the relative weights of expenditure reduction and expenditure switching in adjustment. While Brazilian trade policy has long been protectionist and oriented to import substitution, compensating export subsidies dating from the late 1960s have contributed to a rising trend of industrial and nontraditional agricultural exports. Exchange rate indexation from 1968 on, largely prevented the extremes of overvaluation that earlier periods had seen. Nineteen eighty was a prominent and costly exception, when Brazil dabbled in ‘‘global monetarism.” Fortunately, the experience was of less duration than in the Southern Cone. Unfortunately, the experiment occurred just when the second oil shock hit; a less propitious moment for overvaluation could hardly have been picked. Brazil, in common with other indebted countries, was forced into a massive reduction of imports after 1981. However, Brazil achieved a large rise in exports in 1984 and surpluses of more than $10 billion annually. That is what made continuing interest payments feasible and also led some to the conclusion that foreign exchange constraints to growth were no longer a central concern. Closer analysis of Brazilian import demand and export supply casts doubt upon that interpretation. Import elasticities in the 1980s show no decisive structural break with the 1970s. Higher rates of growth thus imply more than proportional increases in imported inputs, especially of capital goods. Relative price elasticities remain low, limiting the impact of real devaluation. For exports, such limited response is also the rule for primary commodities, which still make up one-third of total export receipts and are also less

280

Eliana A. Cardoso and Albert Fishlow

favorably influenced by expansion in industrial country income. A key point is the influence of capacity utilization upon both import demand and export supply. Prosperity threatens to rapidly diminish large trade surpluses and to present the need for increased external finance. Chapter 7 is both conclusion and epilogue. Our ultimate concern in examining the debt problem is its effect on growth and development. A direct measure of the gravity of the problem is the continuing low ratio of capital formation to total income. Brazilian investment is inadequate to support high rates of substained growth, let alone a competitive capability to ensure continuing performance in the export of manufactured products. The position of the public sector is badly compromised by the need to extract resources from the private sector for debt service. Uncertainty and inflation encourage speculation in real assets and in the black market and detract from productive capital accumulation. Highly variable real wages provoke strong defensive reactions from organized labor eager to protect its real income and hence unwilling to desist from nominal wage demands. Domestic adjustment can only go so far. Real resource transfers of 4 to 5 percent of product cannot continue indefinitely. That is the lesson of the February 1987 moratorium. Analyses that focus on projection of the balance of payments miss this point. Massive trade surpluses are incompatible with high and sustainable rates of Brazilian growth. There are two ways out: more lending or reduced debt service. Brazil has the capacity to absorb more debt productively. One problem, however, is the unreliable supply of additional flows. The Baker Plan formulated in 1985 has failed to meet its minimal, and inadequate, targets. The other is the great uncertainty of the international environment. Starting from a debvexport ratio that is close to 5 in 1987, there is no margin for error. Brazil is vulnerable to any and all adverse changes in interest rates, terms of trade, and industrialized country growth, not to mention domestic policy errors. The case for blending new flows with some relief is given by market prices that value Brazilian debt at a considerable discount. This tells us that full debt service is improbable and that creditors could do just as well by substituting a smaller debt with more secure payment. There is clear scope for a mutual gain through effective official intermediation in the market. It was only a matter of time, and of other adverse shocks, until this mechanism for relief came to be accepted in the Brady Plan in 1989. Now it remains to be seen when Brazil will qualify. The basic lesson of this study of Brazilian external debt is how quickly debt can turn from being part of the solution to becoming a central part of the problem. We may be approaching the point where it is again possible to unleash the productive forces in Brazil and in other developing countries, and to turn them to making up for a lost decade of development.

281

2

BraziliChapter 2

Adjustment to the First Oil Shock: From Import Substitution to Macroeconomic Restraint

In March 1974, when Emesto Geisel assumed the Brazilian presidency, the euphoria of the economic “miracle” of the previous five years still reigned. Dissident voices were few and discredited. Despite the sharp rise in petroleum prices a few months earlier, prospects for continuing prosperity were bright. Spectacular growth at rates in excess of 10 percent aroused visions of grundezu, the attainment of Brazilian destiny on a world stage. Geisel himself was the best prepared among the generals: a proven technocrat not only capable of managing the economy, but also persuaded of the need for a process of guided political liberalization to assure lasting social tranquility. In August 1979, only months after taking office, President Jog0 Batista Figueiredo recalled AntBnio Delfim Net0 to serve again as economic czar, replacing Mirio Simonsen. It was the first change of economic leadership during a presidential term since the military took power in 1964. Such an unprecedented decision signalled the increasing anxiety about the adequacy of Brazilian adjustment to the first oil shock; accelerating inflation and slowing growth threatened. Delfim Neto, the author of the earlier “miracle,” promised a second. In this chapter, we consider the accomplishments and deficiencies of the adjustment strategy of the 1970s, leaving for the next chapter the second oil shock and the evolution of altered policies in the 1980s. 2.1

Origins of the Debt Problem

Brazil faced the sharp rise in oil prices in October 1973 in apparently the best of circumstances. Growth had been extraordinary in the preceding years (figure 2.1), current annual inflation was officially recorded at only 15 percent, and impressive foreign exchange reserves were a defense against shocks in the international economy. Unfortunately, the complete picture was less encouraging. The main weaknesses of the economy were fourfold: (1) a low savinghncome ratio, ( 2 ) irregular export behavior, (3) an exchange rate that was becoming overvalued, and (4) resurgent inflation. They exposed Brazil to a special vulnerability to the direct and indirect effects of the increased price of oil. Economic growth during the ‘‘miracle” had benefitted from accumulated excess capacity. Relatively low fixed investment rates and domestic savings

282

Eliana A. Cardoso and Albert Fishlow I " " I " " ' I ' ' " I " * ' l '

12 a -

-

0

>

-

-

-2

1965

-

-

1970

1975

19801982

1986

Fig. 2.1 Growth rate of real output: Brazil and other large Latin debtors

had therefore been required. Compared to an incremental capital-output ratio of 2.67 in 1965-70, the average from 1971 to 1973 had been only 1.75.' Such good fortune could not be again anticipated. To sustain a continuing annual growth rate of 10 percent implied an increase in the ratio of investment to income of about 4 percentage points, even without considering the implications of the oil price increase. Such greater saving requirements conflicted with the prevailing incentives toward consumer durables as a motivating force in industrial development. Nor were the impressive international reserves at the end of 1973 an accurate measure of the strength of the balance of payments. Real import elasticities had been on the order of two since the late 1960s: growth at a rate of 10 percent a year implied expansion of real imports at 20 percent. Real export growth in the same interval was only about half as fast and only especially rapid in 1972 and 1973. The difference was made up principally by improved terms of trade. Between 1969 and 1972, Brazilian export prices rose 23 percent; import prices, 3 percent.' These favorable circumstances kept demands for external financing within reasonable bounds. The current account deficit amounted to about only 2 percent of total product. Larger quantities of capital actually flowed in-this was the first show of interest in Brazil by the Eurocurrency market-and reserves were accumulated. This access to loans helped to sustain an exchange rate that had become overvalued because changes in the crawling peg were determined by an underestimate of domestic inflation. Inflation was in fact on the rise. After several years of decline, to which high rates of productivity increase and lower real wage increases had

283

BraziUChapter 2

contributed, Brazil was subject to a reversal of the trend. Demand was strong, fueled by real liquidity increases. An expansionary policy was now confronting bottlenecks, and the response was increased reliance on administrative controls to keep the index at the levels forecast by the departing administration. Responding with varying intensity to these initial economic conditions and to the political objective of restoring law in addition to enforcing order, the Geisel economic policy unfolded in three distinct phases. First came the effort to cool down the overheated economy through the application of orthodox monetary and fiscal policy in 1974. These good intentions gave way by 1975-411 part because they were only marginally effective-to a more aggressive medium-term development strategy designed to accomplish the dual objective of sustaining high rates of growth while promoting adjustment to the oil shock. This bold strategy was soon modified in the third phase after mid-1976. A short-term, stop-and-go macroeconomic policy designed to keep inflation and external disequilibrium within bounds was superimposed on medium-term import substitution. Increasing capital inflows became the only means of reconciling the growth and stabilization objectives. This uneasy combination of government investment, monetary stringency, and larger external indebtedness characterized the remainder of the period to 1979.

2.2 From Orthodoxy to Foreign Borrowing The initial attempt at orthodoxy in 1974 was condemned to failure on two counts, one economic, the other political. Fiscal and monetary restraint was unable to make much headway in the short term against accelerating inflation, but did provoke a slowdown in industrial activity as well as lead to a major financial failure (of the Halles Group). Given a commitment to “corrective inflation” that deregulated administered prices, inflation in 1974 could hardly have failed to accelerate. Generalized increases in dollar import prices, part of the oil price explosion, added to the pressure. Even where moderated by subsidies, import prices were passed along to domestic prices. Efforts to sustain relative prices did the rest. At the same time that economic policy made few gains in the fight against inflation in the absence of further reductions of growth, politics called for bold expansion. The ruling government party’s unanticipated defeat at the polls in November 1974 did not provide scope for unpopular austerity measures. By early 1975 the die was cast for resumed expansion. Brazil was to be an “isle of tranquillity” in the midst of international economic turbulence. The choice was feasible for two reasons. First, higher inflation was tolerable domestically because of widespread indexing. And, second, a much weaker external payments position because of increased oil imports

284

Eliana A. Cardoso and Albert Fishlow

did not need to deter Brazilian growth; more liberal international finance was then available owing to petro-dollar recycling. Brazil had already discovered during 1974 that the old, preeoil crisis financial rules no longer held. A trade deficit of $6.2 billion was financed with the use of less than $1 billion in reserves. The sudden increase in capital inflows brought the gross debt level to more than $17 billion, with exports less than half as large. The debt strategy, although initially unplanned, had begun in earnest. Few voiced concern. Instead, foreign financing increasingly became an instrument of choice satisfying three objectives. First, it dampened domestic inflationary pressures inherited from 1973. It did so not only by augmenting the supply of imported inputs in 1974 but also by allowing the government the luxury of not raising domestic prices of imports, and especially of fuel, to the full extent that might have been anticipated. There was no imperative to cut back on consumption of energy or on other imports if the greater cost could be covered by borrowing. Only gasoline prices went up substantially, and even this rise was well below the percentage increase in the world price of oil. Second, abundant and relatively cheap imports helped to sustain high rates of fixed investment by providing access to needed equipment and intermediate inputs. Although the rate of industrial growth declined throughout the calendar year, the principal sectors affected were automobiles and other consumer goods. The producer goods sectors retained their dynamism, averting more serious deceleration in aggregate industrial growth and assuring a basis for the eventual resumption of rapid expansion. Third, access to foreign saving resolved the dilemma of inadequate finance for high growth rates, now that excess capacity had been used up. Ambitious growth targets could be compatible with continuing increases in consumption and did not require large increases in domestic saving. On the contrary, thought could be given to correcting the lagging consumption standards of the poor by implementing a more liberal wage policy.

2.3 External Debt As Solution A full commitment to external debt as a way to facilitate balance of payments adjustment and to finance growth progressively became the basis of the Geisel development strategy. This reliance was enshrined in the implementation, if not the formulation, of the Second National Development Plan. The strategy was based on stimulating public investment and import substitution while relying on debt to avert the need for changes in relative prices or reductions in aggregate expenditures. The deterioration in the terms of trade that threatened ambitious Brazilian growth targets would be neutralized. An orthodox response to the rise in oil prices through changed relative prices and reduced real income via real devaluation, was resisted for several reasons. For one, reliance on devaluation presumed significant price respon-

285

BrazilKhapter 2

siveness of imports and exports. Many Brazilian policymakers had limited confidence in the absorptive capacity of the world market, particularly during international recession; after all, even during the “miracle” years, exports had not performed all that well. At the time it was also believed that imports, particularly oil, had very low elasticities of substitution. In addition, it was certain that indexing would convert these ineffective price changes into generalized inflation through widely ramified nominal rcsponscs of wages and other inputs. Other parts of a conventional adjustment package, higher interest rates and rationed credit, would merely penalize production and investment without eliciting more saving. There was some truth to such a view. Yet the “heterodox” Brazilian response to the oil crisis was itself flawed in two important dimensions. Ironically, one weakness was the apparent coherence of the importsubstitution strategy, which attempted to resolve the short-term balance of payments problem at the same time that it deepened and extended the Brazilian industrial structure. Import substitution was too import intensive to work in the short run as an effective policy to ameliorate the balance of payments problem. As formulated in the development plan, moreover, with very large-scale projects requiring significant initial investment, this generic defect of the import-substitution approach was further exaggerated. Import substitution can only alleviate balance of payments problems in the short run when there is significant excess capacity to be exploited. Brazilian reality was otherwise. Brazil entered 1974 with the highest level of capacity utilization in the entire postwar period. In addition, the development plan called for moving into entirely new sectors, not just expanding domestic share in established industries. Therefore, as was also true in the 1950s, import substitution could only be pursued with the aid of external capital. There was a second contradiction in the strategy. It anticipated both a strong public sector and constructive relations with the private national sector, neither of which was borne out in reality. Public sector expansion entailed increasing deficit finance and came to rely on external resources to fill the gap. The state became larger but also economically weaker. At the same time, state enterprises increasingly impinged upon private terrain, requiring larger subsidies to national entrepreneurs to mollify their opposition. These transfers came at the expense of further imbalance. Thus, facilitated by privileged access to large capital inflows from abroad, government deficits-fiscal and monetary-steadily increased. In sum, Brazilian adjustment came to depend upon external indebtedness as the only means of reconciling its inconsistencies. But there were bounds. As the current account threatened to get out of control in 1976, when net interest payments already amounted to one-third of capital inflows, some domestic restraint became inevitable in the final years of the Geisel administration. “Stop-and-go,’’ as it was practiced, was an inelegant and ultimately ineffective solution to the problem of incomplete adjustment. Its

286

Eliana A. Cardoso and Albert Fishlow

limits became apparent in 1978 when inflation returned to its higher 1976 level despite only moderate GDP growth. Conventional restraint was powerless against adverse agricultural harvests and rising food prices which other sectors quickly felt through the transmission mechanism of higher wage demands. The Brazilian system of indexing was now operating to the disadvantage of policymakers. It had helped in promoting deceleration of inflation so long as it had been manipulated to reduce real wages and so long as supply-side shocks exerted a positive force. Now, as civil society was increasingly voicing its demands, and shocks were adverse, real wages could no longer be determined as the residual income share. Instead, controls proliferated at the expense of priorities. As sectors and interests defended their entrenched positions, the absolute level of subsidies rose without changing relative incentives. The net tax burden declined even while there was resistance to public sector encroachment. Private investment decisions were increasingly conditioned on government policy. The role of pure market forces in determining the allocation of resources progressively receded.

2.4 The External Situation External policy did not perform much more satisfactorily than the domestic instruments. The exchange rate depreciated only in line with Brazilian relative inflation and not much more. Larger devaluations were ruled out by the expectation that they would soon be passed along in prices and wages. Between 1974 and 1977, the real exchange rate moved within a narrow range. Its limited devaluation in 1978 derives primarily from the dollar’s decline relative to other currencies. Subsidies to manufactured products, excluding indirect tax exemptions permitted under GATT, did rise from about 20 to 40 percent of their value over this period; their effect, however, was partially offset by downward pressures on international prices of Brazilian exports. There was thus no mechanism to assure increased incentives to penetrate external markets. While exports continued to grow during 1974-78, as they had done earlier, and their composition diversified to include more manufactured goods, a decomposition of the sources of growth shows a marked difference relative to the preceding period. In 1971-74 increased Brazilian competitiveness accounted for close to half of the rise in exports; in 1974-78, less than 20 percent. Indeed, Brazilian participation in world trade remained approximately constant. For manufactured products exclusively, the conclusion is again telling: 71 percent is explained by competitive improvement in 1971-74, but only 43 percent in 1974-78. A vigorous export expansion was a necessary corollary of the debt adjustment strategy that had been followed. Because the economy was already so closed, additional import restrictions and substitution could make

287

BraziUChapter 2

available only limited foreign exchange to meet the increasing debt service. Export growth, however, did not play its role. Debt increased on the order of 28 percent per year between 1973 and 1978 and faster than exports. Ever larger debuexport ratios signalled the continuing postponement of the inevitable balance of payments realignment. At the same time, the real resource contribution of the debt was progressively diminishing. Each dollar of new borrowing brought less foreign saving that could be applied to real investment. Growing return interest payments reduced the margin by which imports could exceed exports and placed more of a burden on domestic saving to sustain the high rate of investment-about 25 percent of product-that had been attained. Brazil was clearly becoming more vulnerable after 1973, as its integration into the world economy was progressively more asymmetric: its share of debt far exceeded its share of trade. Were export growth not to manage to keep up, or the rigid control over imports to be breached, or the debt to give rise to much more costly interest remittances, the balance of payments constraint was waiting like the sword of Damocles. Even without deterioration of the external environment, there was a potential debt problem in Brazil's future. From table 2.1 the precarious nature of the external adjustment after the first oil shock is apparent. Reduction in the current account deficit, although Table 2.1

The Balance of Payments and the First Oil Shock (in billions of U.S. dollars) 1973

Trade balance Net interest Current account

0.0

-0.5 - 1.7

External effects resulting from:" Higher oil prices Reduced export volume Change in coffee price Actual medium- and long-term net debt Policy-adjusted net debt assuming:b Export increase Import limits Slower growth

1974

1975

1976

I977

-4.7 -0.7 -7.1

-3.5 - 1.5 -6.8

-2.3 - 1.8 -6.1

0.1 -2.1 -4.0

-2.3 1.5

- 1.4

-2.2 -0.6

-2.9

-3.1

- 1.9

0. I

1.4

2.0

11.9

17.2

19.5

24.7

11.5 9.5 10.3

16.3 13.3 14.5

17.4

21.8 20.7 23.2

-0.1 6.2

~

14.1

16.3

*External effects were calculated by comparing actual data with results obtained by assuming that: oil prices were fixed at 1973 value; deviation in estimated noncoffee export volume was caused by slower OECD growth relative to 1968-73 average (deviation calculated from regression for period 1969-82 relating Brazilian export growth to OECD growth rate, yielding an elasticity of 2); and coffee price was fixed at 1973 nominal value. 'Export increase: consequences for noncoffee export growth of a 10 percent real exchange rate devaluation, using a medium-term elasticity of 0.6. The latter is a composite of the medium-term value of 0.75 for manufactured products and 0.5 for nonmanufactures. See Instituto de Pesquisa Economic6 Aplicada do Ministerio de Planejamento (IPEA), Perspectivas de Longo Pram du Economia Brasileira (Rio de Janeiro 1985) and Musalem (1984). Import limits: consequences of a unitary real import elasticity, post-1973, with actual growth sales. Slower growth: consequences of 5 percent annual growth, 1974-77, with actual import elasticity in 1974 and 1 thereafter.

288

Eliana A. Cardoso and Albert Fishlow

continuous after 1974, was very gradual; in 1977 it remained $4 billion compared to the peak $7.1 billion in 1974. This style of accommodation translated into a rapid intervening accumulation of debt: in net terms after subtraction of reserves, the external debt stood in 1977 at almost $25 billion. It became an independent force. The current account deficit was large in 1977 even though the trade balance, helped by coffee and other commodity prices, had barely moved into surplus. An important reason was the mounting interest cost on the debt that already explains about one-third of the increase in the net debt between 1974 and 1977. An adverse international economy contributed to the deficit in two ways: higher oil prices and slower trade growth. Of these two factors, the first was far more important. Had oil prices remained at their 1973 level, Brazil’s actual imports would have cost $10.5 billion less between 1974 and 1977; the increasing effect of higher oil prices reflects the larger import volumes that Brazilian policy did little to deter. Slower OECD growth in 1974-77 than in 1968-73, by contrast, penalized export receipts by a cumulative $5.4 billion, or less than half as much as higher oil prices cost. Against these negatives, rising coffee prices provided an increasing offset by 1976 and 1977, virtually eliminating the OECD effect. Indeed, overall Brazilian terms of trade were more favorable in 1977 than they had been in 1973. On balance, then, a little more than half of the addition to the net debt can be directly traced to external circumstances. With the additional contribution of interest service, that leaves only a small residual in the increased debt to be explained, once Brazil had opted for its fuel-intensive, high-importrequirement model of adjustment. We can get at this issue another way, by asking what effect other potential Brazilian policies would have had on the net debt. Three such policies are shown in table 2.1: the effects on export growth of a real devaluation of 10 percent; less import intensity, such that real imports only increased proportionally to product growth; and slower product and import growth, reflecting a more conservative response to the oil shock, but taking effect only from 1975 on. Export gains would have had a modest, but increasing, impact in substituting for debt. Cumulative increases in exports would also have improved the debt/export ratio. A less import-intensive strategy, through immediately higher costs of fuel and other inputs, and a less ambitious industrialization program would have reduced debt more, particularly between 1974 and 1976 when the import elasticity remained high. The third option, slowing growth to 5 percent in 1974 but lowering the import elasticity only thereafter, shows the limitation of relying on income effects as long as imports remained relatively cheap in 1974, as they did. It is therefore not surprising that this initial policy was soon abandoned by the government. Import restrictions were increasingly imposed so that by 1977 the actual import elasticity fell below unity; that is why the hypothetical import-limit policies in table 2.1 would have been less effective than the actual trade barriers were in curtailing debt growth.

289

BraziUChapter 2

These calculations make clear that an early and steady implementation of a combination of policies, rather than a single measure, was necessary to prevent the oil shock from unduly interrupting growth or saddling the country with an extensive debt. Instead, Brazil had been late to react to a flood of imports in 1974 and had then gambled on an ambitious plan of import substitution and limited relative price change. Muddling through after 1976 was able to keep the resulting disequilibrium within bounds. So long as disaster was averted, as it was, and the economy continued to grow, as it did, attention could continue to be directed to the delicate task of commanding the pace of popular participation, which the Geisel administration did with skill. The new Figueiredo administration was left to deal with the incomplete economic adjustment problem.

2.5 New Initiatives The presidential succession in March 1979 defined a new approach to economic policy. It went beyond validation of just another episode of ‘‘stop-and-go’ ’ to slow down accelerating inflation and defend the fragile balance of payments. The guiding principles were two in number. First, the role of market signals, including the exchange rate, would be strengthened and controls diminished. Second, the effectiveness of fiscal policy would be enhanced by increasing the transparency of expenditure and reducing reliance upon the monetary budget to finance deficits. There was little attempt to conceal the probability that lower growth might be the price for such a substantial reordering of the Brazilian economy. This reformist vision of Mirio Simonsen, moved from his post as Geisel’s finance minister to Figueiredo’s planning minister and economic czar, was relabeled “recessionist” by its Brazilian critics, who were numerous. Private entrepreneurs challenged the utility of looming recession when their profits were already under pressure. Workers were experiencing the erosion of real wages from acelerating inflation because indexed adjustments were only made once a year; they did not look forward to weaker demand in addition. Private banks did not welcome the Bank of Brazil competing directly for prime clients rather than distributing credit subsidized by the Central Bank to priority sectors. Within the government, the new ministers were eager to spend, not to reduce, their expenditures and their power. There was no lack of worthy causes. Poor agricultural harvests reinforced the case for abundant and cheap credit to the sector. The alcohol program was in need of new investment for a second, hydrated phase based on alcohol-powered cars. State enterprises had critical projects underway; Petrobras, the state oil company, in particular, wished to increase substantially its expenditures for exploration. President Figueiredo, beset by the rise in OPEC prices in June 1979 and the need for a “war economy,” soon yielded. In August, Simonsen was dismissed in favor of Antijnio Delfim Neto. Delfim promised a supply-side

290

Eliana A. Cardoso and Albert Fishlow

approach to mounting internal and external disequilibrium that would make demand restraint unnecessary. That was sweet music to the ears of a president elected because of his pledge to wider popular political participation, who was persuaded that prosperity was a necessary condition to its success. Unfortunately, as the next chapter shows, the realization was much different than the expectation.

3

Adjustment in the 1980s: From International Monetarism to the Plano Cruzado

As our discussion has shown, Brazilian adjustment policies, even before the second oil shock, were in need of a midterm correction. Simonsen’s orthodox approach was rejected barely after its announcement and before it could be implemented. Delfim’s more optimistic heterodoxy was much more congenial. But it too proved inadequate, both because of its own limitations and the deteriorating international economic environment. Brazil by 198 1 was in the midst of a harsh austerity program designed to compensate for its mounting external disequilibrium. It could hope to succeed only if external credit were restored. The Mexican debt crisis in August 1982 dashed that hope and soon sent Brazil scurrying to the International Monetary Fund for assistance. The experience with the IMF was tumultuous and marked by repeated letters of intent and waivers for nonfulfilled targets. Improved external performance came partially at the expense of domestic inflation and investment objectives. Still, with the large increase in exports of manufactures in 1984, the economy began to show signs of recovery and resumed growth. The new civilian government that took office in 1985 soon defined itself as committed to expansion rather than macroeconomic restraint. Ample reserves made it possible to delay any long-term agreement with private creditors and to allow the extended program with the Fund to lapse. Accelerating growth in 1985 was accompanied by accelerating inflation that threatened the transition to sustained growth and provoked popular discontent and political dissatisfaction. The Pluno Cruzado, in February 1986, was the heterodox response. In the mold of similar programs previously launched in Argentina and Israel, it identified inertial inflation as the source of inflationary rates that had already exceeded annual rates of 300 percent. It was a bold, and temporarily successful, way to devise a recession-free domestic adjustment to match that of the external accounts.

290

Eliana A. Cardoso and Albert Fishlow

approach to mounting internal and external disequilibrium that would make demand restraint unnecessary. That was sweet music to the ears of a president elected because of his pledge to wider popular political participation, who was persuaded that prosperity was a necessary condition to its success. Unfortunately, as the next chapter shows, the realization was much different than the expectation.

3

Adjustment in the 1980s: From International Monetarism to the Plano Cruzado

As our discussion has shown, Brazilian adjustment policies, even before the second oil shock, were in need of a midterm correction. Simonsen’s orthodox approach was rejected barely after its announcement and before it could be implemented. Delfim’s more optimistic heterodoxy was much more congenial. But it too proved inadequate, both because of its own limitations and the deteriorating international economic environment. Brazil by 198 1 was in the midst of a harsh austerity program designed to compensate for its mounting external disequilibrium. It could hope to succeed only if external credit were restored. The Mexican debt crisis in August 1982 dashed that hope and soon sent Brazil scurrying to the International Monetary Fund for assistance. The experience with the IMF was tumultuous and marked by repeated letters of intent and waivers for nonfulfilled targets. Improved external performance came partially at the expense of domestic inflation and investment objectives. Still, with the large increase in exports of manufactures in 1984, the economy began to show signs of recovery and resumed growth. The new civilian government that took office in 1985 soon defined itself as committed to expansion rather than macroeconomic restraint. Ample reserves made it possible to delay any long-term agreement with private creditors and to allow the extended program with the Fund to lapse. Accelerating growth in 1985 was accompanied by accelerating inflation that threatened the transition to sustained growth and provoked popular discontent and political dissatisfaction. The Pluno Cruzado, in February 1986, was the heterodox response. In the mold of similar programs previously launched in Argentina and Israel, it identified inertial inflation as the source of inflationary rates that had already exceeded annual rates of 300 percent. It was a bold, and temporarily successful, way to devise a recession-free domestic adjustment to match that of the external accounts.

291

BmzWChapter 3

But it is necessary to begin the story of adjustment in the 1980s, that would eventually lead to the Cruzado Plan, with Delfim’s earlier heterodox experiment.

3.1

Supply-side International Monetarism

Delfim’s program to reduce inflation while abetting growth had four components. In sectoral terms, first priority would go to agriculture and energy. The former bore much of the weight, and hopes, of the policy. It was felt that rapid growth of apculture would do the following: end the relative food price shocks that had been so troublesome in recent years; provide the exports to assure continued service of the debt; permit energy substitution through the alcohol program; and facilitate more equal income distribution. There was also then, more than ever, a self-evident need to give equal importance to increased supplies of energy, whether from domestic production of crude or oil substitutes. Both sectors, agriculture and energy, were thus assured all the subsidized credit they needed or wanted. Macroeconomic policy was based on a theory of cost-push inflation, but was not always consistent with it. On the one hand, Delfim set out to undo the previous high real interest rates through strict controls in September 1979 that brought nominal rates sharply down. Yet at the same time, many administered prices were freed during the fall. In November a new wage law provided for more frequent semiannual inflation adjustments as well as relative gains for lower wage workers. While these actions accelerated inflationary pressures, the price increases were blamed upon the previous administration. On the positive side, the fiscal deficit was somewhat reduced and a potential labor conflict averted. The price increases set up what was hoped would be a substantial deceleration beginning in 1980 that could be claimed as a policy success. On the external side, Delfim decreed a maxi-devaluation of 30 percent in December, the first large devaluation in more than ten years of experience with the crawling peg. Export subsidies and prior deposits on imports were removed as redundant after the realignment of prices. In addition, alert to the deterioration in the balance of payments, Delfim actively took new measures to encourage private foreign borrowing to rebuild reserves. The final element was the true novelty in the program. Delfim preannounced the rates of both monetary correction and devaluation that would prevail during 1980, the former at 45 percent, the latter at 40 percent. Credit was to be limited accordingly. This move was intended to change inflationary expectations; if only everyone believed that inflation would be 45 percent in 1980, then it could be. Heavy doses of controls reinforced the message. The Delfim strategy, as it thus took form, was a mixture of three approaches: the standard IMF formula of devaluation to stimulate exports and import substitutes; Southern Cone international monetarism, predicated

292

Eliana A. Cardoso and Albert Fishlow

on a close relation between changes in domestic and international prices; and traditional Brazilian interventionism to induce desired supply response through subsidies. The program created confusion in the international financial community. At first, when the thrust was toward freeing markets in the Christmas package of December 1979, foreign bankers applauded; as disequilibrium became rampant in 1980, they rebelled. The bankers had the final word. Their refusal to roll over the debt without a more conventional stabilization package led to a new approach in November 1980. The results of Delfim’s heterodox approach had by that time fallen far short of its objectives. Although economic growth in 1980 exceeded 7 percent, it was fueled by consumer demand. The ratio of investment to GDP declined. Financial assets, now yielding much less than the inflation rate, were abandoned in favor of the speculative acquisition of consumer durables and real estate. Meanwhile, inflation soared and crossed the three-digit threshold for the first time in Brazilian history. And the current account deficit in 1980, under the impact of additional increases in the oil price, attained a record $12.4 billion and required massive finance. The net debt stood at almost $60 billion, three times the level of exports, compared to a 1977 debuexport ratio of little more than two. Delfim’s policy failed in 1979/80 for four reasons. First, it did not confront the excess demand under which the economy was laboring. The public sector deficit in 1980, excluding monetary correction, is variously estimated as between 5 and 7 percent of GDP. Although possibly smaller than the 1979 level, the deficit remained high and could not be voluntarily funded in the controlled financial markets of 1980; instead it had to be monetized. Unlike the “miracle” years, domestic supply was not elastic enough to satisfy demand. Capacity was nearly fully utilized, especially in the rapidly growing sectors. A second factor was the wage law of November 1979 that conceded semiannual rather than annual inflation adjustments. Increasing labor unrest in 1979, as accelerating inflation eroded real wages, had put pressure on the government to devise a new scheme. Delfim, in a bid to secure order, accepted not only more frequent adjustment but also a law that favored the lowest paid. The adjustment of their wages was to be greater than the inflation index. These two concessions, in theory, would have led to large increases in real wages. Many have therefore singled out the law as a principal determinant of the doubled inflation rate. The independent effect of the law is not so readily established. The World Bank mission to Brazil in 1982 concluded: “A simple examination of trends in total and per unit labor costs in industry from November 1979 to May 1982 suggests that the formula was not a major contributing factor to inflation” (World Bank 1984, 108). Among the reasons are high turnover at the bottom of the wage hierarchy, the lag in the new official consumer price index, the INPC, (to which the wages were linked) behind general inflation,

293

BrazilKhapter 3

and the more generous settlements and readjustments that already prevailed in the private sector. But what the law did do, because wage correction was based exclusively on past trends, was to make impossible significant deceleration in inflation without a large real wage increase. The much lower inflation target of 45 percent for 1980 was thus doomed from the start. The third reason for the failure of Delfim’s strategy was the absence of competitive imports to discipline domestic prices, as international monetarist theory required. Brazil was in the midst of a balance of payments crisis in 1980, despite rapidly rising exports, as a result of higher oil prices and increased interest rates. Imports remained controlled, as they had been for several years. It was an inopportune moment to experiment with this new approach, as Argentina and Chile were also to discover. Finally, this was not the moment to reverse inflationary expectations. Rising import costs, fears of oil shortages, and a demonstrable commitment to expansionary policies all negated the rhetoric of pre-fixed monetary correction and exchange rage devaluation. Rather, as the disparity between reality and the government forecast widened, the only uncertainty was when the policy would change. Expectations, and attendant financial speculation and holding back of exports, focused on the timing of devaluation and not the announced inflation target for the year. Delfim has been rightly critized for the errors of this aberrant policy. Bolivar Lamounier and Alkimar Moura (1986, 173-74) are especially harsh: The monumental failure of that heterodox experiment of economic policy can, in part, be explained by the attempt to implement a strategy of economic growth without consideration for the accentuated deterioration in the conditions of the international economy in 1979 and 1980. . . .It cannot be said, however, that there had been a generalized inability, among the government technocrats, to interpret the unequivocal signals of economic difficulty arising from the international economy. The predominant attitude was to try to exorcize such ghosts with the optimistic rhetoric inherited from the years of the Brazilian miracle.

3.2

The Second Oil Shock

Exorcising ghosts was not enough in the face of the new strong external shocks Brazil experienced after 1979. Table 3.1 presents the balance of payments effects of the second oil shock, the interest rate shock, and the attendant international recession. These show how Brazil was overwhelmed by the adverse turn of the external environment. More than half of the $4 billion deterioration in the current account in 1979 is explained by rising oil prices and higher interest rates. A still larger part of the $1.5 billion increase in net interest payments, $1.2 billion, is accounted for by the interest due on the change in accumulated debt (and smaller receipts on reduced reserves) during the year.

294

Eliana A. Cardoso and Albert Fishlow

Table 3.1

The Balance of Payments and the Second Oil Shock (in billions of U.S. dollars)

Trade balance Net interest Current account

1978

1979

1980

1981

1982

- 1.0 - 2.7

-2.8 -4.2

-6.0

- 10.0

-2.8 -6.3 - 12.4

1.2 -9.2 -11.0

0.8 -11.4 - 16.3

-5.7

-7.1 - 2.4 -2.5

-6.1

External effects resulting from? Higher oil price Reduced export volume Higher interest rate Actual net debtb Policy-adjusted net debt Export increase Import limits Slower growth

-1.8 -0.6 -0.3 36.2

- 1.4 -1.1

-5.9

46.4

57.7

68.0

83.5

45.6 44.8 45.8

55.8 54.0 55.7

64.5 64.9 65.9

78.3 82 5 81.46

"External effects calculated assuming that: oil price was fixed at 1978 nominal value; effect of reduced export volume was as in table 2. I , using deviation from 1974-78 average growth of 3.1 percent; and interest rate effect was based on constant 1978 average real interest rate (with respect to U.S. GNP deflator). bNet debt inclusive of short-term debt, from F'aulo Nogueira Batista (1985). 'Export increase: export response to 10 percent real devaluation. Import limits: real imports held constant at 1978 levels. Slower growth: Unitary elasticity effect on imports of product growth at 3 percent.

Conversely, an array of feasible alternative policy efforts would have provided only modest offsets to the balance of payments deterioration in 1979. Neither export promotion, strict import limits, nor slowed growth taken individually would have been equal to the effect of higher oil prices. It would have taken all together to begin to match the adverse turn in the international environment and the inertial effect of rising debt. In 1980 even higher oil prices and interest rates made matters considerably worse. The actual worsening of the balance of payments was smaller than would have been anticipated owing to the 58 percent increase in the value of exports since 1978. So substantial was the gain in earnings that in the absence of the rise in oil prices, Brazil would have enjoyed a healthy trade surplus in 1980. But oil prices did rise and there was little room for maneuver. Delfim inherited the problem of inadequate adjustment; it was not simply of his own making. Starting from its much higher debt in 1978 and the continuing large volume of oil imports, Brazil had much less flexibility in dealing with the second oil shock than the first. That said, however, there was a strong argument for conserving what few degrees of freedom remained. A more cautious policy would have marginally improved balance of payments performance, prevented spreads-and hence interest costs-from rising, and retained domestic credibility. A more conservative policy would have also avoided exchange rate appreciation that would later have to be undone. The beneficial effects of the December 1979 devaluation were more than wiped out by inadequate correction that failed to keep pace with domestic inflation.

295

3.3

BraziUChapter 3

he-IMF Orthodox Stabilization

Expectations did turn out to be rational. The virtually universal disbelief in the adequacy of the initial heterodox policy was confirmed by its abandonment in November and December of 1980 under increasing pressure from foreign creditors. Unlike some of his fellow policymakers in the Southern Cone, who believed both in their policies and that sustaining them even when they were not working was the only way to make them work, Delfim was more pragmatic. Yet it is a measure of the gap between ongoing political liberalization and the technocratic monopoly on economic policy formulation that he yielded to external influence rather than domestic critics. More orthodox policies of restraint became the order of the day. Capital expendituresof state enterprises were a principal target, both to reduce the budget deficit and to limit imports. Total loans to the private sector were limited to no more than a 50 percent increase over their December 1980 nominal value. Controls were removed from all loan rates, except for credit to agriculture and to exports. Monetary expansion was severely limited, provoking a liquidity shortage. Real interest rates rose from large negative to positive levels. Firms cut back on production and tied to work down their bloated, and increasingly expensive, inventories. Private investment declined. These deflationary impulses produced a decline in gross output of 1.6 percent between 1980 and 1981 and an even larger drop in industrial production. Urban unemployment became overt. Brazil had entered a period of falling income that was to prove more severe than the setback of the Great Depression. The strategy followed in 1981 was to conform to the shortage of foreign exchange by reducing absorption. Devaluation was ruled out by the failed attempt of late 1979 and the prevailing belief in limited opportunities for export in the midst of a world recession. The immediate gains from the new policy were relatively modest. Inflation decelerated from 121 percent in 1980 to 94 percent in 1981. The trade balance moved into modest surplus. The primary cffcct of the recession was to unloose a new flow of capital from commercial banks, placing Brazil further in debt. Instead of conceding the need for more fundamental changes and then implementing such changes, Delfim’s primary stabilization objective was to retain international creditworthiness and avert a liquidity crisis. In other words, it was a poor recession, just as the preceding period had been a false prosperity. To avoid going to the IMF, Brazil undertook an even more severe short-term stabilization to persuade international creditors of its sincerity. But in so doing, Brazil lacked a program of real adjustment or coherent strategy for coping with its expanding, and increasingly short-term and interbank, debt. The balance of payments problem thus remained. The real exchange rate, after an acceleration of the mini-devaluations in the second half of 1981, was not quite back to where it had been in early 1979. Now, with the dollar-to which the cruzeiro was linked-appreciating , more aggressive policies to

296

Eliana A. Cardoso and Albert Fishlow

improve the trade balance were necessary. Export growth, aided by restoration of subsidies to manufactured products, was respectable in 1981 but not spectacular. Almost as much of the trade improvement was achieved by more rigorous import controls. During 1982, exchange rate devaluation again fell behind Brazilian relative inflation, causing resumed appreciation. On the internal side, the tight monetary policy and fiscal restraint were not sustained. The deficit of the consolidated federal public sector actually rose, and in 1982 it reached 8.5 percent of GDP (see chap. 5). While controls over the money supply were apparently effective, with large real declines in liquidity as measured by narrow M1 and M2 aggregates, internal debt was used instead to finance the deficit. In 1981 and 1982, the augmented series of money and quasi money (M4) exceeded inflation. Progressively greater reliance on internal debt, which was to increase from 15 to 30 percent of GDP between 1980 and 1984, was a very dangerous course. (See figure 3.1 for the recent evolution of the debt of the public sector.) Since interest rates were much higher than the growth of revenues, today’s finance was converted into tomorrow’s larger deficit. In general, a potentially destabilizing situation results, in which explosive growth of the debt can crowd out real capital formation and lead to ever higher interest rates. In Brazilian circumstances, it also meant fewer degrees of freedom with respect to policy as government bonds had to be guaranteed against changes in the exchange rate and not only against domestic inflation. The hope was that a short, albeit severe, recession would permit Brazil to resume its access to external finance and economic growth. The crucial November election loomed in 1982, and government hopes for controlling the selection of the next president, turned on a respectable showing. That objective helps to explain why, despite the gathering clouds in financial

1981

82

83

84

1985

Fig. 3.1 Public sector debt (year-end, as a share of GDP)

297

BrazWChapter 3

markets in early 1982, domestic restraint eased and industrial decline was moderated. The political motivation was reinforced by the recalcitrance of inflation; it stopped falling in 1982, making the tradeoff with excess capacity much more unfavorable than in the previous year. The government was later to blame the decline in exports in 1982 on spreading international recession, the increase in net interest payments on the high international interest rates, and the closure of financial markets on the Malvinas/Falklands war and the Mexican default. As was shown in table 3.1, Brazil by 1982 was indeed laboring under very adverse external conditions. But the recession of 1981-82 was also poorly managed.

3.4 Going to the Fund Brazil waited too long to approach the Fund formally, only going after the election although contacts had been established earlier. Already by March 1982, the net reserves of the Central Bank were negative. Until virtually the very end, however, the technocrats insisted that they were capable and that Brazil was different from its profligate neighbors. Indeed, before going to the IMF, Brazil put together in October its own spartan plan for presentation to the private banks, a plan calling for minimal finance and exuding confidence: “It is precisely this blending of short- and long-term adjustment which will create the preconditions for the Brazilian economy to find a path of relatively more stable economic growth with smaller imbalances and being threatened neither by growing inflationary pressures nor by the unforecastability of external factors” (Conselho Monetario Nacional 1982, 10). And it was precisely this blending that had been absent in the previous three years. Policy had been very much oriented toward short-term goals and was frequently altered. Ad hocery was rampant. Solutions were designed for immediate problems, but frequently introduced new distortions that later would inhibit effective policy. The government failed to clear out the baggage of credit subsidies and tax incentives inherited from the past and establish meaningful priorities. Domestic political support was irrelevant. The judge and jury were the external creditors. Planning and finance ministers undertook well-orchestrated forays to the exterior to assure and reassure that overly optimistic targets were securely within reach. Meanwhile, domestic credibility dissipated. Delfim remained in office because there was not even enough governmental capability to define an alternative strategy. When agreement was reached in December 1982 with the international banks to reschedule principal and provide new finance, Delfim announced that a letter of intent was being presented to the IMF. The Fund board approved it with special speed to reassure the financial community that the debt crisis was under control. The Fund program incorporated the limited relief that had been previously asked of the private banks. Table 3.2 provides details of the quantitative internal and external performance criteria Brazil was obligated to fulfill in return.

IMF Agreements: Quantitative Performance Criteria

Table 3.2

Predicted Values January 1983 (letter # I )

February 1983 (letter #2)

September 1983 (letter #3)

November 1983 (letter #4)

March 1984 (letter #5)

November 1984 (letter #6)

December 1984 (letter #7)

Actual Values

Borrowing Requirement of the Public Sector (billions of cruzeiros) 1983

March 1,200 June 3,200 September 5,000 December 7,000 1984 March June September December 1985 March June Operational Deficit (billions of cruzeiros)**

2,800 5,000 6,600 8,800

14,900 19,350

December March June September December 1985 March June September Monetary Authorities' Net Domestic Assets (billions of cruzeiros) March June September December

4,050 4,650 5,150 5,800

11,750 23,750 35,500

44,500 67,800 35,500 70,000

1983 1984

1983

24,600 I 1,750

6,150 6,950 7,550 8,300

5,600 3,540

3,600 1,300

1,300 300 600

3.629 -638' - 1,692* 587* 6,169

1,100 - 2,100 -

3,540

3,578 8,334 13,263 23,891 9,686* 23,648' 45,466' 84,371 48,536 121,125

5,300 13,000

6,696 20.051 7,415 9,895 5,417 6,685

March 3,300 June September December March 1985 June Changes in Net International Reserves, relative to December of the previous year (US$ billion)

1984

March -1.5 June -1.5 September -0.7 December 0.0 March 1984 June September December March (no targets) 1985 Increase in Net External Indebtedness (US$ billion) 1983 March 2.2 3.0 June 3.0 4.5 September 4.0 5.3 December 6.0 6.0 March 1984 June September December March 1985 June September

5,350 4,550 2,800

1,600 - 50 - 5,ooo - 7,750

1983

4,365 3,105 223 - 2,064 -6,659 - 14,257 -

1.6

- 1.8

-3.1 0.0

- 2.8

0.1

5.5 9.0 2.5

1.7 2.7 3.7 4.3

3.9 6.8 9.I

Source: Banco Central do Brasil, Brasil: Program Economiro, various issues.

*As estimated by the Central Bank In early 1985.

**Operational deficit is equal to the BRPS minus payments of monetary correction on government bonds.

5.1 5.7

-3.3 2.7 4.2 5.8 7.0

8.6 10.8

1.3 2.5 2.7 3.5 3.1 5.1 7.0 8.4 0.8 1.7 -0.007

300

Eliana A. Cardoso and Albert Fishlow

Central to the initial letter of intent and its modifications was a targeted $6 billion trade surplus in 1983. Exchange rate devaluation of one percentage point per month in excess of inflation was deemed adequate. External debt was to increase modestly. Limited public sector borrowing and drastic restraint on net domestic credit (to the extent of a decline by one-half in its estimated previous year value) were the operative elements on the domestic side. Government expenditures were to be reduced to conform to the reduction in borrowing requirements. By the end of the year, an inflation rate of 78 percent was to be attained as a result of such policies. Barely had the first letter been issued than it was necessary to modify the Public Sector Borrowing Requirement (PSBR) and domestic credit limits. The February devaluation of 30 percent following a wave of speculation in the black market made the earlier nominal values meaningless. Larger credit allocation was allowed for export promotion and import substitution to help achieve the external goals. Monetary correction of government indexed bonds to match the devaluation of exchange rate altered borrowing requirements. The next series of letters of intent was provoked by the continuing acceleration of inflation. Only three months after formal approval of the stabilization program, the IMF suspended the scheduled $2 billion disbursement because the government failed to reduce its nominal deficits. This was no surprise, since any shortfall in attaining the inflation target translated into far different nominal results for equivalent real magnitudes. Two major supply shocks-the 30 percent February devaluation and the upward trend of agricultural prices-combined with indexation to prevent contractionary monetary policies from reducing inflation. In fact, inflation reached the 200 percent level during 1983. Two modifications were introduced. One reduced the extent of wage indexation. The IMF insisted upon limiting inflation adjustment of wages to 80 percent of any rise in the new consumer price index, despite the tendency of the INPC index already to understate true inflation. Thus there was explicit recognition of the force of inertial inflation. After considerable reluctance and the imposition of executive pressure, the Congress acceded. The other change was also a response to recalcitrant inflation. After prolonged discussion, a new measure of the deficit, the “operational deficit,” was introduced. In this concept, the payment of monetary correction on the public debt was subtracted from the global PSBR (see chap. 5, app. 2 , for more detail). The allowance for the effects of passive response to inflation provided a better measure of feasible fiscal restraint. These changes did not obviate the need for more new letters of intent. After the four in 1983, there were an additional three in 1984. They arose, as can be seen from table 3.2, largely as a result of the violation of targeted nominal public sector borrowing requirements. Domestic inflation remained immune to reduction in 1984, despite controls over increases in net domestic assets and favorable performance of the operational deficit. External

301

BrazilKhapter 3

objectives were overfulfilled, on the whole. After the sharp decline in output in 1983, steeper than that of 1981, there was even a modest recovery in 1984 to which rapid growth in industrial exports contributed.

3.5

Evaluating the IMF Program

The Fund program of 1983- 84 provoked increasing criticism within Brazil for being an inadequate response to its difficulties. Much of the opposition was directed to the continuation of onerous external interest payments. In magnitude these payments came to rival the entire import bill. Or to put it another way, debt service, even with rescheduled principal, came to about half of export earnings. Creditors were reluctant to concede a multiyear rescheduling along the lines of the Mexican agreement. That would have postponed a larger part of the principal that was coming due in the future, as well as reduced spreads. The longer that the agreement was delayed, the less sense it started to make, as it became clear that the soon-to-be successor government was likely to be led by the opposition. The Fund itself was always unhappy with Brazil’s performance and lack of compliance with policy targets. The series of revised letters of intent from Brazil offset the favorable impact of its excellent trade performance. The case of Brazil epitomized the limits of the IMF approach. Despite favorable management of the trade account, internal stabilization and a sound basis for renewed economic growth did not follow. Rather than decelerating, inflation more than doubled. Meanwhile, high real interest rates (figure 3.2), the counterpart of tighter money and large government sales of debt, discouraged private investment. Together with controls over

Fig. 3.2 Real interest rate nets of taxes (CDB: 30 days, average monthly yield, by quarter)

Eliana A. Cardoso and Albert Fishlow

302

public investment, this led to a decline in the ratio of gross capital formation to only 16 percent in 1984, just about its lowest level of the postwar period. Since the inception of the program, the public sector deficit was continuously pressured by the rapid growth of internal and external interest payments. Figure 3.3 shows that the share in GDP of the current account net of factor payments has turned positive since 1983. To critics of the IMF stabilization approach, the stark asymmetry of the results obtained on the balance of payments and domestic stability came as no great surprise. Contrary to the IMF’s implicit monetarist model, the Brazilian experience confirms a very different interpretation. Improving the external accounts has become an important source of internal disequilibrium. The very policies required to permit large trade surpluses and payment of external interest add to inflation and subtract from investment. Thus aggressive devaluation of exchange rates reflects itself sooner or later, and mostly sooner, in domestic inflation because of the ubiquity of indexing. In addition, the public sector must attract ever larger resources from the private sector in order to service the now largely public external debt. When it does so on a voluntary basis, interest rates are high and become a source of higher costs that are passed along to prices. In addition, government deficits, whether financed by money or internal debt, then maintain nominal demand and sustain the inflation. The state is too weak to raise taxes and accomplish the large transfer needed in a noninflationary way. The extensive resources that have been transferred externally, amounting to some 5 percent of gross product in 1983 and 1984 and reducing national

6 4

a

n

2

W

0

-2

-4

-6

1970

1974

Fig. 3.3 Nonfactor current account in Brazil, 1970-86

1983

1986

303

BrazilKhapter 3

income proportionally, have come primarily at the expense of investment. Consumption outlays have resisted further compression. Even with the changes in the wage-indexing arrangements required by the IMF, there were limits to further declines in standards of living. Saving has not been responsive despite the continuation of high real interest rates; bank certificates of deposit net of taxes were highly positive in 1984. These economic circumstances, despite attempts to show that Brazilian performance was better than elsewhere in Latin America, and despite a modest recovery led by manufactured exports in 1984, contributed to the Figueiredo government’s loss of political control. The rules of succession, which had been thought to assure not only continued dominance of the government party but also one more military president to guide transition, proved unable to withstand the clear lack of popular acquiescence. Tancredo Neves was selected by the electoral college in January 1985. With his untimely death before he assumed office, the New Republic’s leadership fell to JosC Sarney. 3.6

After 1985

During much of 1985 and 1986 the debt issue and the problem of the balance of payments took a back seat to domestic economic policies in the new government. Economic recovery quickened, but was accompanied by accelerating inflation. Brazil’s trade balance remained positive and adequate to service its interest payments up to mid-1986. Falling oil prices and interest rates, and sharply increased terms of trade in 1986, added up to significantly improved external conditions. From a high of 62 percent in 1982, the ratio of interest payments to exports fell to 44 percent in 1985 and 45 percent in 1986. The absence of additional capital flows meant real reductions in the level of the debt after the end of 1984. In these circumstances it is not surprising to see the recent external crisis blamed exclusively on poor domestic policy and the failure of the Cruzado Plan. Roberto Campos, a former planning minister in the first post-1964 government and now a senator from the opposition party, prominently did so in the national and international press. He was joined by Luis Inficio da Silva, better known as Lula, the leader of the Labor party on the other side of the political spectrum. Both have been joined by countless external observers who emphasize that there is no guarantee that exaggerated real domestic targets will not confront balance of payments constraints. Wishing to grow at high rates, and even needing such growth for laudable political ends, does not make it come true. The large debt burden of Brazil, however, was an important contributing factor to the failure of domestic policy that was a proximate cause of the precipitate decline in the trade surplus and the inability to continue normal interest payments. Three interactions are worth emphasizing: the lack of new

304

Eliana A. Cardoso and Albert Fishlow

capital flows to finance larger imports, the effect of the drain of external interest payments on domestic capital formation, and the immediate vulnerability of the balance of payments to even temporary shocks. First, it was unrealistic to expect Brazilian imports to remain at low levels despite accelerating growth and to project large continuing surpluses. Import elasticity is especially high when increased investment becomes necessary after a first phase of using up excess capacity. A popular view in Brazil was that import substitution had proceeded so far and so successfully that requirements were permanently reduced, and that made the debt, and the required surpluses, less central to domestic strategy than it ought to have been. Especially in the context of the Cruzado Plan, capital inflows would have been of considerable assistance in permitting import liberalization and thus a market check to inflation, rather than relying on price controls alone. In the second instance, the large resource transfer, amounting to more than $30 billion in three years, made it difficult to sustain domestic capital accumulation. Consumption could not be compressed to make the transfer; indeed, pent-up demand further reduced domestic saving, thereby aggravating the domestic imbalance. In addition, transfers of that magnitudeamounting to some 4 or 5 percent of product-complicated fiscal policy. The public sector was the principal remitter; it had to compete for the surplus of the private sector. A bias toward government deficits was a consequent, and ultimately disastrous, result. Thirdly, the fact of a large foreign debt gave little scope for error. Even a few months of diminished international performance were sufficient to erode reserves and provoke strong policy reactions. Those measures in November 1986 were badly received and contributed to the lack of credibility and a generalized sense of deterioration that quickly became self-fulfilling as the year ended and 1987 began. There was little time to regroup or to formulate new approaches. A moratorium on interest payments became inevitable. A large debt creates its own special problems for the formulation and implementation of policy, as the United States too is beginning to appreciate. This is not to exculpate Brazil’s responsibility in not sustaining its export growth. While everyone was busy citing Brazil as an example of successful adjustment, Brazilian export volume, after an increase of about 20 percent in 1984, grew by less than 2 percent in 1985 and declined by 8 percent in 1986. Irregular export performance has plagued Brazil even when its exchange rate policies have avoided extreme overvaluation. But that is precisely why the debt problem is so serious and so damaging to Brazilian developmental prospects. Ignoring the difficulty by projecting high compound rates of export growth is only more wishful thinking. Any serious analysis of the sad denouement of late 1986 and early 1987 is thus incomplete without reference to the larger story of the risky, and ultimately mismanaged, Brazilian adjustment policy via external debt since the early 1970s.

305

4

BrazWChapter 4

Stopping Inflation

Brazilian inflation is a problem that predates the current debt overhang. Figure 4.1 provides some recent history. In the five years between 1959 and 1964, increasingly populist administrations carried inflation from 10 to 100 percent. By 1968 it was down to 20 percent, a level that persisted until the first oil shock, when inflation jumped to 40 percent. There it remained until 1979. In the five years between 1980 and 1985, the government's failure to absorb the debt and oil shocks in a noninflationary manner pushed inflation from 50 to 220 percent. On 28 February 1986, with inflation at 400 percent per year, Brazil embarked on its second major stabilization effort in twenty-five years. In this chapter we highlight the main features of the Brazilian inflation process and contrast the two stabilization efforts. The dynamics of Brazilian inflation is best understood by recognizing the interaction of supply shocks and indexation as the main elements in generating acceleration. The two stabilization efforts demonstrate that an incomes policy is an essential ingredient to nonrecessionary stabilization. But they also show that demand restraint is indispensable if disinflation is to be viable and that external conditions make a difference. The 1964 program was gradualist and relied on wage repression to prevent cost forces from pressing on prices. Stabilization was non-neutral in its distributional impact. It was also conducted without concern for debt or the balance of payments. The 1986

1952

I

I

1964

1974

Wlc W W

FE co) o c

g: Fig. 4.1 Monthly inflation rate (annualized moving average)

306

Eliana A. Cardoso and Albert Fishlow

plan was a heterodox shock treatment centered around a continuing price freeze and paid insufficient attention to excess demand and the need for fiscal restraint. Supply was constrained by continuing large resource transfers abroad. Our closer analysis of the two stabilization programs is preceded by the presentation of a basic model of inertial inflation in the next section. 4.1

Inertial Inflation

Money expansion and velocity behavior are not enough to explain inflation dynamics in Brazil. There is little doubt that, from the demand side, large budget deficits in 1959-64 and 1979-85 supported the inflation process. But assigning them more than an accommodating role would mean neglecting the important contribution of the supply side to inflation and the mechanics of its propagation. Fischer (1977) and Taylor (1979) have drawn attention to the persistence of price disturbances in a setting of overlapping, long-term wage contracts even under forward-looking, rational expectations behavior and a wellunderstood program of monetary control. In the Brazilian setting, institutional factors have taken, to a large extent, the place of the relative wage and expectations mechanisms that characterize Fischer-Taylor contracts. From 1968 on, mandatory indexation has been backward looking, periodically readjusting wages and other contracts on the basis of recent experience. Figure 4.2 illustrates how the Brazilian indexation system worked between 1968 and 1986: while the wage legislation corrected wages for past inflation, the exchange rate was readjusted in short intervals by mini-devaluations, and interest rates, bonds, and rents were corrected for price increases through a system of indexation called monetary correction. Assuming that prices are determined by costs, current inflation depends on past inflation via indexation of wages, the exchange rate, and public sector prices. The output gap or unemployment affects current inflation because it

-

Public-sector prices

of living

Industrial prices

Wages Imported Inputs

Fig. 4.2 The Brazilian indexation system

Woge indexation Minidevoluot ions Monetory correction

307

BrazillChapter 4

influences the marginal costs of firms to the extent that the turnover of the labor force can be used to offset wage adjustment. Independent internal and external factors such as harvests, terms of trade, etc., also have an influence. Combining the effects of indexation, unemployment, and supply shocks, we write the inflation process as: (4.1)

pi

=

a pi-1

+ b GAP, + e,

where p is the rate of inflation, GAP is a measure of excess demand, and e is an indicator of supply shocks. Equation (4.1) shows that indexing plays a central role in projecting past price increases into the future. Where such a process has long been operative, the role of nominal demand in stopping inflation is weak relative to the replicative effects of formal indexation. Moreover, any escalation of prices from supply shocks gets permanently embedded in the inflation rate. The implications of this inflation process are worthwhile emphasizing. First, current supply shocks are automatically transmitted to future periods. A real depreciation, elimination of public sector subsidies, or increases in the price of oil or in indirect taxes or in the real price of agricultural goods raise the current rate of inflation and are transmitted via indexation into increased inflation in subsequent periods. In fact, in order to raise real prices or cut real wages in the presence of full indexation, the frequency of adjustment of exchange rates and public sector prices has to be greater than the frequency of wage adjustments. Only then is it possible to defeat the indexation, cutting the average real wage during the adjustment period by stepping up the rate of inflation. Indexation of the financial system, the tax structure, and the public debt means that changes in the inflation rate are automatically and fully accommodated. Second, a slowdown in the growth rate of nominal spending cannot eliminate inflation from one day to the next. Demand-side policies do not eliminate the cost-inflation captured in the lagged inflation term of equation (4.1). The neoclassical answer, which is to instantaneously recontract the labor force with reduced wages in the face of a shift to a noninflationary monetary regime, is implausible. The presence of inertia is thus a central reason for incomes policy in a stabilization program. Otherwise, only through significant recession can inflation be affected. Statistical estimates of the tradeoff in the Brazilian case show large output costs for small reductions in inflation. Third, any escalation of inflation initiated by some supply shock is supported by further accommodating endogenous elements that feed the inflation process. One element is the increase in the velocity of money in response to increases in inflation; another is the inflationary erosion of tax revenue. The larger fiscal deficit then implies increased rates of monetary expansion.

308

Eliana A. Cardoso and Albert Fishlow

Fourth, the inertia of the inflation process implies that contracts that are not explicitly indexed in a backward-looking way, such as short-term loans in the financial system, cany forward-looking inflation adjustments. At any point in time, a given stock of such contracts are outstanding. Their maturity may run as long as six months or one year. Sudden disinflation, by contradicting expectations, would result in an arbitrary redistribution between debtors and creditors. Finally, an important contributing factor to acceleration of inflation is the mounting pressure to shorten the interval for inflation adjustment of wages, public sector prices, and the exchange rate. In high-inflation economies, institutional arrangements provide for a periodical resetting of real wages to a peak. The peak real wage occurs at the date of the contract immediately after the nominal wage increase. Subsequently, up to the next adjustment, the real wage is eroded by inflation. Figure 4.3 shows the actual real minimum wage in Brazil over the period 1976-86. At fixed intervals, the real wage increases to a peak and then is eroded over the interval between adjustments, reaching a trough just prior to the next adjustment, a year or six months later. Because the average real wage declines with higher rates of inflation, increases in inflation provoke demands for shorter periods of adjustment to prevent adverse distributional effects. This shortening of the adjustment interval in turn increases inflation: in a context of overlapping contracts, the shortening of the intervals means that a larger number of contracts are revised on the same date, pushing up costs and hence inflation. Escalation of inflation to three or four digits invariably involves a shortening of adjustment periods for wage and price setting. In 1979 the annual adjustment of wages in Brazil was moved to a twice-yearly base. This translated into a supply shock for firms, as discussed in chapter 3. Given the large devaluation, discrete adjustments in public

76.5

Fig. 4.3 Minimum real wage (monthly data, 1976.5-1986.6)

6

309

BraziKhapter 4

prices, and the rise in real prices of agricultural products, all at the same time, it is not surprising that inflation rates doubled. Higher inflation in turn further shortened intervals, thereby pushing inflation up once more. By the end of 1985, firms and workers were beginning to move toward cycles of only three months. The government, keenly aware that the transition to ever shorter periods would have hyperinflationary consequences, tried to prevent the generalized legal requirement of three-month intervals. Eventually more dramatic action was needed and occurred in the form of the wage-price freeze of March 1986. Before we discuss the heterodox shock, however, it is useful to consider the stabilization program of 1965- 68. 4.2

Inflation and Stabilization in the Mid-Sixties

Our story begins in the early sixties. The main reason for the sharp increase in inflation between 1959 and 1962 was a rapid increase in demand. Brazil was undergoing a surge in growth driven by the internal market. Between 1957 and 1962, industrial output grew at 11 percent per year. The share of the central government budget deficit in output increased from 2.8 percent in 1960 to 4.3 in 1963, while the share of seignorage in GDP to finance it widened from 3.6 percent in 1959 to 5.7 in 1962. The combination of a 30 percent deterioration of the terms of trade, the lack of external financing, a bad coffee crop in 1963, and adverse climatic conditions leading to an agriculture disaster in 1964 contributed to the inflation problem from the side of supply. The economic crisis was the vehicle for a military takeover in March 1964. The Progruma de Ac&o EconBmicu do Govern0 (PAEG, 1964166) detailed a plan to reduce inflation gradually in three years using fiscal and monetary restraint along with incomes policy. The following were the main aspects of the program: (a) Sounder fiscal policies led to a gradual reduction in the deficit from 4.2 percent of GDP in 1963 to only 1 percent in 1966. The main instruments of this budget balancing were increases in public sector prices, cuts in subsidies, increased tax collections obtained through an increase in indirect taxes, and better administration to avoid tax evasion. Despite the initial increase in wages for the military and civil servants and an increase in investment expenditures after 1965, the deficit was reduced. Later on, the reduction of real wages also helped the budget. (b) The exchange rate was devalued by 70 percent at the outset of the program to deal with the external imbalance. Outstanding debt was rescheduled and a large program of external assistance was begun. (c) Monetary policy was erratic. An initial moderate expansion in 1964 was followed by a tightening in 1965-66. Indexation in financial markets

310

Eliana A. Cardoso and Albert Fishlow

was used to mobilize domestic saving and to create a market for public sector debt. The black market premium that had reached 60 percent in the last quarter of 1963 was already down to 4 percent by the end of 1964 and oscillated below 1 percent in 1965 and 1966. (d) Incomes policy took the form of granting wage increases, not in line with past inflation but rather geared to “expected inflation” and past average, not peak, wages. In terms of the inflation model described in equation (4.1), disinflation was achieved by breaking the link between current and past inflation. Wage adjustments became forward looking and limited to an officially imposed, and optimistic, inflation forecast. But the reduction in price inflation fell short, by a wide margin, of the anticipated decline built into wage agreements. The effect was to reduce real wages, although the original claim had been that real wages would remain constant; that is, smaller nominal adjustment would be compensated by lower inflation. The real minimum wage fell by 15 percent between 1964 and 1967. The wage cut made room both for budget balancing and for improved external competitiveness, while at the same time bankrolling a cut in the rate of inflation. (e) On the price side, the government introduced a program of fiscal, credit, and other incentives to firms that agreed not to raise prices by more than a stated percentage. Allowable cost increases excluded wage awards in excess of those contemplated in the government wage formula. In 1966 firms were promised a 20 percent reduction in excise taxes if they carried out wage increases in accordance with the government wage formula. From 1967 on, price guidelines became more pervasive, falling on most large industrial firms. This program was successful in reducing inflation without generating a dramatic decline in economic activity. Inflation declined from 144 percent in the first quarter of 1964 to 57 percent in 1965, and then to 38 percent in 1966. Industrial production declined in the first year of stabilization by 5 percent, but then showed a rapid recovery. By 1966 it was already 6 percent above the pre-crisis level. In 1968 a new strategy of stimulating demand and living with modest inflation was adopted. First, a crawling peg exchange rate policy was introduced to maintain stable real exchange rates and avoid the variability consequent upon discrete devaluations, thus extending indexing. Second, credit became more abundant. Third, in response to public reaction against the wage squeeze, the government revised the wage adjustment rule. The new formula corrected half of any underestimation of future inflation. Real minimum wages continued to fall until 1970, when inflation rates had stabilized at a lower level. A new formula was introduced in 1974 after inflation rates started up again. Wage readjustment then returned to the pattern of backward-looking, catch-up indexation.

311

Brazil/Chapter 4

The reforms of 1965-68 were the basis for an extended period of strong growth with stable inflation. Between 1968 and 1973, real growth averaged 10 percent per year and inflation declined to only 20 percent. After 1974 and the joint effect of external shocks and domestic overheating, inflation started to increase again, as discussed in chapter 2. 4.3

Inflation and Stabilization in the Mid-Eighties

The Cruzado Plan stopped inflation in March 1986. Less than a year later, in January 1987, interest rates were soaring and the economy was in disarray. Uncertainty was pervasive, fueled not only by the return of the three-digit inflation rate but also by the feeling that the government had lost control over the economy. In this section we describe the Cruzado Plan,* and in the next one we examine the factors that contributed to its failure, despite its clever design and appropriateness. In February 1986 monthly inflation had reached an annual rate of 400 percent. Inflation escalation had become severe with the second oil shock and the shortening of the intervals for wage setting, and was further accelerated in 1983 with the combination of a large real devaluation, an agricultural disaster, and increased prices in the public sector together with reduced subsidies. It was held in place at the expense of a major recession in 1983, as discussed in chapter 3. After a modest recovery in 1984, the new democratic government embarked on a program of expansion in 1985 that carried real growth to 8 percent. Another bad crop introduced supply shocks which, combined with the increase in economic activity, led to rising inflation. As noted earlier, workers’ concern over the erosion of real wages created new demand for shortening the interval for inflation adjustment to only three months. Some private firms acceded. Mindful of the dangers of inflation acceleration, and with elections coming later in the year, the government had little choice but to embark on a bold program of heterodox stabilization via incomes policy rather than recession. The decision did not come as a surprise. Plans for inflation stabilization which tackled the critical difficulty of inertia had been widely discussed in Brazil for more than a year. The Argentine and Israeli precedents were evidence that such plans were practical. Policymakers believed that past inflationary shocks were being perpetuated in a vicious circle created by indexation. Freezing prices, exchange rates, and wages would create a rupture with the past, thus permitting the economy to rid itself of inertial inflation. The main obstacle to the price and wage freeze was the absence of synchronization in price readjustments. In 1986 the typical wage contract ran for six months, with different groups of wage earners receiving readjustments at different points in time. Rent contracts ran for either one year or six months. Government bonds were readjusted monthly. Simply freezing wages

312

Eliana A. Cardoso and Albert Fishlow

and prices on a given day would greatly favor wage earners and entrepreneurs who had readjusted their prices immediately before the freeze. In a like manner, such a step would drastically punish those who were to have received their new settlements the following day. Others would be located along a spectrum whose midpoint would correspond to average real wages and prices; their position would depend upon how recently their nominal wage had been adjusted. The arbitrary choice of any particular day for a freeze would establish an unsustainable structure of prices and wages. This situation is illustrated in figure 4.4,which shows the main features of wage settlements in Brazil: the semiannual periodicity of readjustments, the staggered dates of readjustments, and the institutionalization of previous wage peaks. Recontracting with full past indexation restores the previous peak of the real wage at the beginning of each period. As prices increase, the real wage is reduced during the next six months until it reaches bottom on the day before the new contract takes effect. The figure shows the behavior of the real wage for six different groups of workers. The first group contracts in January and July, the second in February and August, and so forth. One can see that in each month three groups have their real wages above the mean and three others have theirs below it. Clearly, some workers can enjoy for sometime a real wage above the mean only because the wages of others are below it. Freezing wages at any one point in time would benefit workers whose wages were above the mean and hurt those with wages below. To avoid this problem, the freeze has to be done in such a way that all wages are restored to their average real levels. The new currency facilitated differential conversion of current wages to their average levels of the last six months. as

wage

I Peak Average Bottom

Jan.

Feb.

Mar.

Apr.

May

June

Fig. 4.4 Real wages for six groups of workers with staggered contracts (six-month indexation period)

313

BrazilKhapter 4

well as conversion of other contracts. Cruzeiro amounts were converted to cruzados (a new currency, with 1,000 cruzeiros per cruzado) according to specific tables. This translation, along with the coordination of all income recipients, were the central points of the new program. The key steps of the Cruzado Plan were the following: Wages were readjusted and frozen. Contracts with several months to go before the regular readjustment were augmented, and contracts that had experienced a recent readjustment were rolled back. As a sweetener, the minimum wage was increased by 15 percent over its past real average, while workers who were paid over the minimum received an 8 percent bonus. Rents and installment payments were converted into cruzados using their average real value during the previous twelve months and frozen at that level for one year. All prices and the exchange rate were frozen until further notice. A tablita was devised to compensate for the expected inflation built into extant contracts and thus avoid arbitrary redistribution between debtors and creditors. The new currency, the cruzado, was introduced to help facilitate the readjustment. Starting on 3 March 1986, the cruzeiro would depreciate at the rate of 0.45 percent per day relative to the cruzado. Indexation, which was central to the process of inertial inflation was virtually eliminated. An escalu move1 (sliding scale), with a 20 percent threshold, was substituted for wages. In financial markets, indexation was maintained only for instruments of more than one-year maturity. There was a sharp initial monetization of the economy to avoid the Argentine problem of exceptionally high real interest rates. In the first three months following stabilization, the monetary base doubled. Monetary demand was accommodated with the expectation that it was for the purpose of restructuring the composition of financial assets. On the fiscal side, the tax reform of December 1985 was thought to have laid much of the groundwork for stabilization. The expected revenue gains were calculated to close a budget deficit of 6 percent of GDP. The operational deficit in 1986 was projected to be zero. External factors favored the program in three respects. The decline in world interest rates reduced the burden of the debt service in the government budget and in the balance of payments. Sharply lower world oil prices made a major contribution in the same direction since oil made up 45 percent of imports. Dollar depreciation in the early part of the program helped provide a gain in Brazilian competitiveness despite a fixed exchange rate, Thus, the program was well timed from the vantage point of the international economy.

Eliana A. Cardoso and Albert Fishlow

314

4.4

What Went Wrong?

Between February and June, cumulative inflation was zero. Industrial production increased by 12 percent in the first half-year of the program relative to the same period a year before. Between June and November, however, the program took on a life of its own. Fuelled by strong popular support for the price freeze, Finance Minister Dilson Funaro elevated controls to a fetish. The budget was allowed to deteriorate dramatically, the trade surplus disappeared, shortages and black markets became pervasive. But “zero inflation” remained the ministerial obsession. Only one intervening and inadequate adjustment was attempted in July to raise indirect tax receipts. Among the factors leading to the failure of the Cruzado Plan, the most prominent was the overheating of the economy as a result of expansive wage, monetary, and fiscal policies. Money growth had already started to accelerate ahead of inflation in August 1985, as shown in table 4.1. The government turned to expansion of the monetary base rather than increases in internal debt to finance its deficit. As real interest rates net of taxes turned negative, economic agents rearranged their portfolios in favor of consumer durables, stimulating output growth, and in favor of real estate, pushing up the price of land (figure 4.5). Money growth further accelerated after March 1986, as policymakers recognized the need to remonetize the economy following disinflation. Price stability, along with an end to the inflation tax, suddenly increases the demand for real balances. Real interest rates turn sharply positive unless the government engages in a significant monetization. But it is difficult to know how much is enough. One criterion is the level of the real interest rate, the other, the behavior of monetary aggregates. It is difficult to judge the Real Money Stock and Real Interest Rates, Quarterly Average, 1984-86

Table 4.1

Real Money Stock (millions of March 1986 cruzados)

Period

M1

M4

1984 I

1.12 1.09 1.05 1.09 0.99 1.05 1.16 1.34 1.52 2.90 3.44

6.95 7.22 7.69 8.41 8.67 9.60 10.11 10.20 10.07 11.07 10.08

Il III IV 1985 I

II Ill IV 1986 I Il I11

Source: Conjuntura Econ6mica and Banco Icatu.

Real Interest Rate Net of Taxes (percent per year on CDs of one-month maturity) 0.37 5.76 5.25 0.78 8.10

54.36 -9.35 -11.15 - 14.32 14.49

315

BraziKhapter 4

4

400

300

200

120

Jon.1977 1978

1979 1980 1981

1982 1983

1984 1985 1986

Fig. 4.5 Average real estate prices in SPo Wulo, 1977-87 (in real dollars per M2)

appropriate level of real balances because financial adaptation to inflation partially destroys traditional linkages between interest rates and real balances. The stability of demand for money under such discontinuous change is dubious. But being too conservative is problematic because high real interest rates in the presence of a large public debt create a fiscal problem. With hindsight one can argue that monetary growth proceeded too fast to assure price stability. Part of the problem was that low interest rates permitted firms lacking confidence in the program to build up speculative stocks in anticipation of removal of the price freeze. The signs of excess demand and repressed inflation started to pile up, but the policymakerspreferred to deny the evidence rather than to give up the fetish of zero inflation for more realistic policies. The next point concerns fiscal policy. Tax revenues rose disappointingly little due to two features of the tax legislation approved in December 1985: a lowered income-tax-withholding schedule and an increased reliance on taxation of financial assets no longer widely held. Revenues of state-owned companies were hurt by the price freeze, spending ran higher than anticipated, and subsidies that were cut during 1983-84 have since surged back. The public sector wage bill increased in line with the economywide trend. Must the budget be strictly zero, or even in surplus, for monetary reform to succeed? Or is it possible to finance a small deficit in a noninflationary manner? To a large extent this depends on the growth rate of output, the prospective path of real tax revenues, and the real rate of interest. If output growth is high and the real rate of interest is negligible, there is room for small deficits. But the aftermath of the Cruzado Plan indicates that

316

Eliana A. Cardoso and Albert Fishlow

public budget deficits in Brazil were too large to be sustainable without accelerating prices. The inflation tax was still necessary to finance planned outlays. The third issue relates to wage policy. In terms of equation (4. l ) , the key point of the program was to eliminate catch-up inflation in prices and wages. This was done by the offsetting influences of rolling some contracts back and others up. Thus inertial inflation was simply suspended. But the real wage was increased, the cost of which was borne by a reduction in the profit margins of price-controlled firms. The increase in real wages promoted by the Cruzado Plan and ratified by the fast-growing economy sustained a consumer boom. During the first six months of 1986, retail sales increased 20 percent compared to the first six months of 1985, a year during which real GDP had increased by 8 percent. By July it was already obvious that the economy was overheated. Acute shortages, especially of meat and milk, commanded the headlines of every newspaper. Black markets for all sorts of goods started to flourish. Supply restrictions, to force price changes, aggravated the situation. Minor adjustments in the program were made in late July through increases in very special excise taxes. They were claimed to be large enough to solve the budget problem, yet small enough to be eliminated from the official price index to keep from influencing the wage trigger. Credibility of the plan started to dwindle. The government promise to maintain the price freeze contributed to their landslide election victory in November. It was immediately followed, however, by a second round of much larger excise tax increases, whose objective was to raise 4 percent of GDP in revenue. The package included price increases of 100 percent on beer and cigarettes, and an 80 percent price increase on cars and car parts. This latter increase, on top of the 30 percent increase in July, turned Brazilian cars from the cheapest into the most expensive in the world. Once again the government tried to eliminate these excise tax increases from the price index on the grounds that the products did not enter low-income expenditure baskets, but this time they encountered strong resistance. The government was implicitly trying to reenact the 1964 program of real wage cuts to restore the external balance and the budget. But the absence of fiscal austerity and the constraints of a democratic regime put severe limits on such an exercise. The November reforms also reintroduced partial indexation, creating a new system of daily corrections of the exchange rate and tying interest payments on savings accounts and other financial instruments to the short-term interest rate. It was a harbinger of what was to come. In February 1987 the prohibition of indexation for contracts of maturity shorter than one year was abolished. Brazil once again had to live with inflation, but under less favorable circumstances.

317

4.5

BrazillChapter 4

Comparing Two Experiments

During the 1964 stabilization the preceding high inflation with no indexation had reduced the real value of the public debt to less than 4 percent of GDP (figure 4.6). In the 1986 stabilization, by contrast, prevailing indexation and high real interest rates had left a significant handicap of a combined debvincome ratio of foreign and domestic debt of 50 percent. The large debt and insufficient budget improvement led to expectations of a need for inflationary finance. Rising nominal interest rates and a widening black market premium registered signals of doubt about the viability of the program. By January 1987, inflationary expectations had become extreme. The removal of long adjustment lags in wages, which previously had been an element of short-term stability in anchoring price increases, meant that inflation could accelerate dramatically. This possibility was reinforced by the escala move1 which potentially could put wage adjustments into the express lane. Under such an arrangement, the frequency of adjustment becomes very sensitive to the inflation rate and feeds back upon it. Self-fulfilling prophecies did not take long to verify themselves. In February the freeze was formally removed and prices exploded. Although the exchange rate was being devalued on a daily basis by then, the black market for dollars stood at a premium of more than 100 percent. The short-term interest rate reached 33 percent per month. Seeing the two Brazilian stabilization programs together teaches three key lessons. First, incomes policy is a valuable means for achieving disinflation.

I

Fig. 4.6 Net debt of the public sector as a percentage of GDP Notes: 1945-64: Total debt (external and domestic) of central and local governments (Goldsmith 1986). 1982-86: Net debt of the public sector including public enterprises (Banco Central do Brasil, Brasil: Program Economico, 1986).

318

Eliana A. Cardoso and Albert Fishlow

It helps avoid dramatic unemployment. But incomes policy by itself is not enough. Without careful fiscal policy, the disinflation is not viable. With a boom, price stability is very temporary. Moreover, incomes policy is difficult to implement in a neutral fashion. In 1964 wage repression was part of the price for disinflation. In 1986 there was a redistribution from firms to workers implicit in the rise in real wages. But firms were able to react to defend themselves and, in the process, made the price freeze an increasingly costly option. The second lesson concerns indexation. Indexation in the presence of supply shocks is a source of inflation propagation. But it also dampens the shocks. An economy with long adjustment periods has an inflation process that is protected against rapid acceleration. Indexation of assets reinforces the element of stability. In the 1964 episode, indexation was reinforced and broadened. In 1986, by contrast, it was eliminated altogether and replaced by a wage-adjustment trigger without a cap. The threat of setting off the ocala movel led to efforts to purge the price index at a significant cost to credibility. Financial assets were indexed to the short-term interest rate in November because expectations had turned adverse. Such setting led to a highly volatile atmospherein which inflationary expectations easily became the driving mechanism for actual inflation. The sharp deceleration of inflation in mid-1986 was thus replaced by an explosion of prices at the beginning of 1987. The third lesson concerns the debt overhang. In 1964 it was irrelevant. In 1986 it was prominent. There was much less margin for maneuver in order to finance government deficits in a noninflationary fashion. There was less import capacity to make up for domestic shortages or to make long-term investment and technological updating attractive. There was no abilityeven with initially abundant international reserves-to use the international accounts to compensate for internal excess demand. The death knell of the Cruzado Plan was, not surprisingly, sounded by the moratorium on external interest payments.

5

External Debt, Budget Deficits, and Inflation

In January 1987 Brazil faced an external debt of $103 billion, amounting to more than one-third of GDP. Debt service requirements remained onerous, and a precarious trade balance was on the verge of provoking a moratorium. The inflation front did not look any better. Table 5.1 shows the numbers for

318

Eliana A. Cardoso and Albert Fishlow

It helps avoid dramatic unemployment. But incomes policy by itself is not enough. Without careful fiscal policy, the disinflation is not viable. With a boom, price stability is very temporary. Moreover, incomes policy is difficult to implement in a neutral fashion. In 1964 wage repression was part of the price for disinflation. In 1986 there was a redistribution from firms to workers implicit in the rise in real wages. But firms were able to react to defend themselves and, in the process, made the price freeze an increasingly costly option. The second lesson concerns indexation. Indexation in the presence of supply shocks is a source of inflation propagation. But it also dampens the shocks. An economy with long adjustment periods has an inflation process that is protected against rapid acceleration. Indexation of assets reinforces the element of stability. In the 1964 episode, indexation was reinforced and broadened. In 1986, by contrast, it was eliminated altogether and replaced by a wage-adjustment trigger without a cap. The threat of setting off the ocala movel led to efforts to purge the price index at a significant cost to credibility. Financial assets were indexed to the short-term interest rate in November because expectations had turned adverse. Such setting led to a highly volatile atmospherein which inflationary expectations easily became the driving mechanism for actual inflation. The sharp deceleration of inflation in mid-1986 was thus replaced by an explosion of prices at the beginning of 1987. The third lesson concerns the debt overhang. In 1964 it was irrelevant. In 1986 it was prominent. There was much less margin for maneuver in order to finance government deficits in a noninflationary fashion. There was less import capacity to make up for domestic shortages or to make long-term investment and technological updating attractive. There was no abilityeven with initially abundant international reserves-to use the international accounts to compensate for internal excess demand. The death knell of the Cruzado Plan was, not surprisingly, sounded by the moratorium on external interest payments.

5

External Debt, Budget Deficits, and Inflation

In January 1987 Brazil faced an external debt of $103 billion, amounting to more than one-third of GDP. Debt service requirements remained onerous, and a precarious trade balance was on the verge of provoking a moratorium. The inflation front did not look any better. Table 5.1 shows the numbers for

319

BraziVChapter 5

Table 5.1

The Brazilian External Debt, 1984-86 (m billions of U.S. dollars and percentages)

Total Registered Nonregistered Medium- and long-term debt with: Foreign commercial banks Brazilian banks abroad Others External debt/GDP ratio (5%)

1984

1985

1986P

102

11

105 96 9

I03 95 6

48

60 8 28 48

59 7 28 40

91

Source: Banco Central do Brasil, B r a d Program Economico, November 1986 PPreliminary.

the external debt, and figure 5.1 indicates recent behavior of the inflation rate. As we discussed in chapter 4,the Cruzado Plan had stopped inflation cold in March 1986, and during the next six months inflation remained low. But less than a year later, inflation exploded again. In June 1987 Brazil faced an annualized inflation rate of 800 percent, twice as large an inflation as when the plan was implemented. Policymakers had emphasized inertia in contracts and expectations as the most important component of inflation. They had thus chosen a shock treatment centered around a rigid price freeze, while paying insufficient attention to the need for fiscal restraint. Their failure forces us to look more closely into fundamentals. In this chapter we argue that foreign debt and inflation in Brazil are related problems originating from the same source. We analyze the relations among the budget deficit, interest rates, domestic and foreign debt, and inflation. We start with a brief description of the Brazilian financial market, and then look at alternative measures of the budget deficit and discuss whether classic seignorage models can fit the Brazilian inflationary process. Having examined the nature of these large budget deficits, and having argued, in the subsequent section, that classic money-goods models do not explain inflation in Brazil, we establish the importance of the financial market in the analysis of inflation in Brazil. We then develop a seignorage model for an open economy with a financial market. We show that the enforced switch from external to domestic deficit finance has pushed both real interest rates and inflation upward.

5.1 The Financial Market Brazil has a complex financial system. Financial reform was a key element in the stabilization of the mid-sixties. In the 1950s and early 1960s,

320

Eliana A. Cardoso and Albert Fishlow

1979 1980

1981

1982

1983 1984

1985 1986

1987

Fig. 5.1 Monthly inflation rate in Brazil, January 1979-June 1987 (wholesale prices, internal supply)

price controls on public services, sectoral subsidies, and inadequate tax collection contributed to a growing budget deficit, almost exclusively financed by monetary expansion. Interest rate restrictions, combined with inflation, limited financial savings and made interest-bearing money substitutes scarce. Long-term financing was available only in limited quantities from government banks at negative real interest rates. The 1965 stabilization program initiated reforms which shaped the financial system during the 1970s. Those reforms introduced assets subject to monetary correction (indexation of the principal) and compulsory savings. The government made monetary policy the responsibility of the newly created National Monetary Council, started the Central Bank, and established the Housing Finance System headed by the National Housing Bank. Monetary authority was divided between the Bank of Brazil and the Central Bank. Their combined responsibilities still extend far beyond conventional central banking: the Central Bank manages development funds and programs, while the Bank of Brazil is the largest commercial bank in the country and the main supplier of rural credit. Following the above reforms, extensive financial deepening occurred. Financial assets rose from 23 percent of GDP in 1965 to 60 in 1985. The share of indexed assets, particularly savings accounts and compulsory savings, increased continuously, while external borrowing came to play an important role. After the first oil shock, capital inflows were encouraged, and credit subsidies and federal debt issue increased dramatically. The Central Bank acted as a financial intermediary with a negative spread,

321

BrazilKhapter 5

lending cheaply and borrowing at high rates. It seems likely that some portion of the considerable part of the agricultural credit known to be diverted from its intended applications found its way, directly and indirectly, through financial institutions into the holding of government debt. The years from 1975 through 1983 were characterized by heavy reliance on foreign borrowing, proliferation of subsidized credit lines, and increasing dispersion of interest rates. Rising economic stress after 1975 changed the size and composition of the financial system, although its broad structure, which dated from the reforms of the second half of the 1960s, persisted until the reforms of 1986. Figure 5.2 shows the share of main financial assets in M4. The Cruzado Plan substantially deindexed the financial system while retaining the indexation of savings deposits by a new correction index. But the OTN (Treasury Bill), which was frozen on February 1986 for one year, was revalued on March 1987 and subsequently, more frequently. Minidevaluations were also reintroduced at the end of 1986, and the consumer price index is to be used to revalue balance sheet assets and liabilities. To a substantial extent, Brazil is back to the pre-Cruzado indexed economy, with even shorter adjustment periods. Table 5.2 shows the main financial assets, yields, and taxation prevailing in March 1987. Brazilian financial instruments fall into four categories: those indexed by the consumer price index or by the new public debt instrument, the LBC (Central Bank notes), introduced in June 1986. These indexed instruments are mainly savings deposits and, increasingly, time deposits; assets with pre-fixed nominal yields, mostly sixty- and ninety-day time deposits; outstanding OTN-indexed bonds;

0.4 0.2 -r

Fig. 5.2 Main financial assets share in M4

322

Eliana A. Cardoso and Albert Fishlow

Table 5.2

Main Financial Assets, Yields, and Taxation at the Beginning of 1987

Asset

Yield

Taxation

Letra do Banco Central (LBC)

Rate of return set by government at “ I N K inflation” with 15-day lag.

No tax on interest, 40% tax on capital gains at source.

Cademetas de Poupanca

Principal corrected at LBC 0.5% per month interest rate. Minimal holding periods: individuals, I month; companies, 3 months.

No tax for individual holders. Some fiscal benefits. Companies taxed via profits.

Certificado de Deposito Bancario (CDB) with post-fixed yield

Corrected by LBC plus market interest rate. Minimum term 60 days.

Income tax on interest: 35% if holder identified, 45% otherwise. Forty percent tax on capital gains.

Certificado de Deposito Bancario (CDB) with pre-fixed yield

Market-determined nominal rate (60or 90 days).

“Reference” rate sets correction for tax purposes, tax rates same as yield for “post-fixed’’ CDB.

Obrigaco do Tesouro Nacional (OTN

OTN correction (now same as LBC) plus fixed interest of 4-8%. depending on term.

Income tax 40% on interest, 35% on capital gains above LBC correction.

Letra do Tesoum Nacional (LTN)

Market-determined discount.

Same as for CDB with pre-fixed yield.

+

Letras de Cambio

Market-determined interest rate.

Same as for LTN

Short-term assets held for less than 28 days

Market yield.

Forty percent income tax on total yield. No extra taxes on capital gains.

FGTS’

Corrected by LBC plus fixed interest rate.

No tax.

PIS/PASEP**

Corrected by LBC plus interest, depending on profitability of investment.

No tax

*Fundo de Garantia por Tempo de Servico. **Programs de lntegracao Social.

(4) dollar-indexed instruments. These are either five-year OTNs with an exchange correction, or dollar-indexed deposits unofficially offered by some commercial banks, or dollars traded on the parallel (black) market. Federal bonds and bills, which represented more than 30 percent of public sector grew to more than government relied more and more deficit.

were practically nonexistent in 1965, financial assets by 1985. The debt of the 50 percent of GDP in 1985, as the on debt creation to finance the budget

5.2 The Budget Deficit Not only was fiscal consolidation during the 1986 stabilization program difficult because of the size of the debt, but policymakers also made no

323

BraziUChapter 5

honest effort to correct the deficit. Some among them argued that Brazilian inflation was different from that elsewhere and that the budget deficit did not have a role in it; inflation was purely inertial and all that was needed to stop it was a price freeze and some formulae to recalculate wages, rents, and future installments. Others denied the existence of the problem by putting numbers together that would show a negligible budget deficit. Table 5.3 shows alternative measures for the budget.’ When inflation rises or abruptly falls, different budget concepts are strongly affected. The borrowing requirement of the public sector (PSBR), which moves dramatically with inflation because of the inflationary component of interest payments, is a faulty indicator. Rather, a preferable measure is the size of the budget deficit corrected for inflation. In 1982 the budget deficit, corrected for inflation of the consolidated public sector, exceeded 8 percent of the domestic product. An agreement with the IMF (whose staff calculated the PSBR as 15.8 percent of domestic product in 1982) was reached in December of that year. The following year inflation, rather than decelerating, more than doubled. The public sector deficit exceeded its targets regularly, not merely because it was hard to control expenditures and increase tax receipts but also owing to rapidly growing internal and external interest payments. There was a large increase in the budget deficit as a percentage of GDP in 1983 for two major reasons. First, in contrest with historically large and positive growth rates, output fell during 1983 by 3.2 percent. Second, interest paid on government bonds included compensation for the 30 percent devaluation of February because the return on these bonds had been linked to the rate of exchange depreciation. As shown in table 5.4, the domestic cost, corrected for inflation, of servicing the external debt greatly increased during 1983. Figures for the budget deficit are available up to 1985, and all of them, except perhaps for those under the ‘‘operational deficit” column (see table Different Measures of the Budget Deficit as a Share of GDP

’Igble 5.3

Year

Total DebUGDP

Increase in

Deficit Corrected for InflatiodGDP

FGV measurea/GDP

P.SBRb/GDP

1982 1983 1984 1985 1986

25.9 60.5 60.9 64.6 22.9

8.4 15.2 4.6 6.1 3.5

3.7 4.1 4.7 n.a. n.a.

15.8 19.9 23.3 27.8 11.2

Operational DeficitC/GDP

~

6.6 3.0 2.7 4.3 3.6

Source: Cardoso and Reis (1986) and Banco Central do Brasil, Brasil: Program Economico, February 1987.

Talculated on a cash-flow basis, excluding the monetary authorities’ deficit. bPublic Sector Bomwing Requirement, calculated on accrual basis, excluding the monetary authorities’ deficit. 5ubmcts monetary correction from PSBR.

324 Table 5.4

Year

I982 1983

1984 1985

Eliana A. Cardoso and Albert Fishlow Domestic Cost Rate of the External Debt Service in Brazil, 1982-85 (in percentages)

Interest Rate Paid on the External Debt

Inflation Rate During the Year

Devaluation Rate During the Year

13.02 10.04 10.20 10.00

95 155 22 1

93 22 1 220 236

226

Real Rate of the Domestic Cost of the External Debt Service

11.9 38.5

9.9 13.4

Sources: Conjuntura EconGmica and Bolenm do Bunco Central

5.3, col. 5 ) , indicate the existence of large deficits. In particular, nominal borrowing requirements for the public sector continuously increase until 1985. The question to be taken up in the next section concerns whether increasing inflation in Brazil can be explained by these growing budget deficits. 5.3 Seignorage Models of Inflation The observation that high and lasting inflation rates always involve monetary expansion has led to the study of situations that give rise to monetary expansions. The most common argument links money printing to the financing of government deficits. The link may be obvious, such as money issued to finance a war, or more roundabout, for example, involving an exchange rate collapse. A story that could be told for Brazil in the 1980s would start with a government with a large external debt. When foreign capital inflows suddenly cease, this government is forced to extract from the private sector the foreign exchange resources it needs, and does so either by money creation or increased domestic debt. On the other hand, to force the private sector to produce the trade surplus and the needed foreign exchange, the exchange rate needs to be greatly depreciated. The devaluation further increases the cost of the debt service in domestic currency, causing additional increases in government expenditure and money creation. Inflationary deficit finance inevitably leads to two types of vicious circles. First, if government prices are adjusted with delays and income taxes are collected on the basis of incomes earned one year before, higher inflation itself increases the budget deficit, inducing even larger increases in money. Second, the share of the inflation tax in output is inversely related to velocity. Since velocity increases with inflation, increasing budget deficits will require further increases in money creation once velocity responds to increasing inflation rates. Unfortunately, financing government expenditure through debt creation seems equally unattractive, especially when real interest rates are high and exceed the domestic growth rate. The rapidly growing stock of debt becomes

325

BraziVChapter 5

a major source of expenditure and makes deficit reduction more difficult in future years. External debt brings with it the additional burden of debt service in foreign currency. Table 5.5 presents the size of the deficit and the source of finance. Until 1984, and particularly in 1983, the increase in the budget deficit was financed primarily by an increase in the total debt, both domestic and external. As shown in figure 5.3, money growth in 1983 was approximately the same as in previous years, while inflation jumped to twice its previous level. The Brazilian monetary experience of 1983 provides one of those classic counterexamples to Milton Friedman’s claim that every inflation acceleration in history has been preceded by a monetary expansion. Money growth lagged behind inflation until the last quarter of 1984. Inflation seemed to enjoy a life of its own. A committed monetarist might be willing to argue that this was due to velocity adjustments: inflation accelerated due to the expectation of faster money growth in the future, even though current money growth was slow relative to the inflation rate. These expectations might have been fueled by the sight of an enlarged debt and the prospect of increasing interest payments. More eclectic economists would argue that the monetarist explanation is not convincing. To account for the dynamics of inflation, we must take into consideration the pervasive indexation schemes that have existed in Brazil and their effect on inflation inertia. This was the subject of the last chapter. Here we simply mention important supply shocks that took place during the period under scrutiny. Recall that a large devaluation in February 1983 followed the 30 percent devaluation of December 1979, that prices and interest rates administered by the government were corrected for past inflation, and that subsidies to oil and wheat consumption were cut at the same time as the mini-devaluations were accelerated. All these inflationary pressures were worsened by a crop failure in 1983. Add to that the exchange and bond indexation scheme and one is not surprised how easily the inflationary shocks spread, leading to inflation rates Budget Deficit Share in GDP and Budget Finance in Brazil, 1982-85 (in percentages)

Table 5.5

Year

Change in the Real Base/GDP

Inflation Tax on the Real Base/GDP

1982 1983 1984 1985 1986

-0.4 -1.3 0.3 -0.4 0.9

2.3 2.8 2.0 2.1 2.7

SeignorageGDP

Increase in Real Net Debt (net of the base)/ GDP

Budget Deficit Corrected for InflationiGDP

1.9 1.5 2.3 1.7 3.6

6.5 13.7 2.3 4.4 -0.1

8.4 15.2 4.6 6.1 3.5

Sources: Banco Central do Brasil, Brad: Progrumo Economico, August 1986, and Conjunruru Econdmico. Note: For the calculation of the budget deficit share corrected for inflation, see appendix 2 to this chapter.

326

Eliana A. Cardoso and Albert Fishlow

L

0

Seasonally adi. M I

500

m 3 1

L

a, Q

OLl

JAN 1981

I

JAN 1982

I

JAN 1983

I

JAN 1984

I

JAN 1985

I )

JAN 1986

Fig. 5.3 Inflation and money growth (three-month moving average)

of 200 percent. The restrictive monetary policy of 1982 and 1983 resulted in high interest rates, recession, and unemployment, but left the inflation rate unchanged. By 1984 the government had gone back to its traditional accommodating monetary policy. As the money growth rate converged on the inflation rate, economic activity began to recover, in part helped by export growth. Up to mid-1985, the fact that inflation was holding at the 1984 rates was interpreted as a confirmation of the theory that, in the absence of shocks and in a context of thorough indexation, inflation sustains itself through inertia. As a matter of fact, there were shocks during 1985 but these were counterbalanced by price controls and by public sector price increases that were lower than the general inflation rate. The inflation explosion in August, November, and December of 1985 aroused the suspicion that something was very wrong with the inertialists' explanation.

5.4 Can Money-Financed Budget Deficits Explain Inflation in Brazil? Seignorage models consist of a combination of two equations. The first one shows the seignorage share in output (or, in other words, the share in output of the budget deficit financed by money creation) as equal to money growth divided by velocity, which is assumed to be a positive function of the inflation rate. The second one makes inflation equal to money growth. If increasing money-financed budget deficits are to explain the ever increasing inflation between 1979 and 1985, one would expect seignorage as a share of GDP to rise. This did not happen. Between 1970 and 1985, seignorage as a share of GDP is fairly constant at around 2 percent. Seignorage models as an

327

BraziKhapter 5

explanation for inflation in Brazil can thus be dismissed on the grounds that seignorage as a share of GDP shows absolutely no correlation with inflation (figure 5.4). The money-goods model of monetarism is inappropriate to the Brazilian economy because it fails to account for changes in deficits not financed by money creation. The model predicts that seignorage requirements drive the system. But the Brazilian experience has to be interpreted in the light of the institutional reality of financial markets and growing external debt. There is yet another reason why a more complete model is necessary to account for inflation in Brazil. This concerns the dynamics of inflation. Fully flexible prices permit equality, at all times, between seignorage adjusted for growth and the inflation tax on the monetary base.' But the Brazilian data rule out this possibility. Figure 5.5 shows that an increase in inflation increases the inflation tax but also increases velocity and reduces seignorage. Ruling out full price flexibility, different assumptions about price dynamics yield unattractive models of seignorage. The reason is that money holders are assumed to acquire disequilibrium levels of real balances to satisfy needs of the monetary authorities. A more appropriate solution is to introduce financial markets into the model. In the next section we sketch a seignorage model that does exactly that.

5.5 A Seignorage Model for the Open Economy with a Financial Market Consider an economy in which the current account is financed either by commercial loans or by changes in foreign reserves. All external borrowing

0.03

a

D

-

l

'

l

'

l

~

l

-

'

l

-

0

-8 2 0.020 c

.-0, 0)

0.015

o

Ql

0

-

0 O , "

0

0

-

-

o.ol.-op

Fig.

l

0.025-

P

v,

'

-

0

'

1

'

I

'

1

'

I

"

328

Eliana A. Cardoso and Albert Fishlow

L l

Seignorage i

1

Inflation tL\

4

Fig. 5.5 Seignorage and inflation tax in Brazil, 1973-86

is done by the public sector. The government finances the budget deficit by borrowing abroad and by creating both money and domestic debt. We can combine the government budget constraint and the balance of payments equation to obtain an equation for the growth rate of the real money base: (5.1)

=

a/h

-

n

where h stands for the real monetary base, I*. is its growth rate, II stands for inflation, and a represents the sum of the domestic component of the budget deficit financed by money creation plus the noninterest current account. Equation (5.1) is derived in appendix 1 at the end of this chapter. The next question concerns the inflation dynamics. The nominal interest rate adjusts to clear the money market at all times. We also assume that there is inflation inertia: inflation increases whenever the level of activity exceeds full employment, that is, whenever the actual real interest rate, i - II,defined by goods and money market equilibrium, is below the full-employment real interest rate, r:

(5.2)

fI

=

o[r(G,TB) - ( i - I I ]

where G and TB represent, respectively, permanent government expenditure and the trade surplus. The model described by equations (5.1) and (5.2) is represented graphically in figure 5.6, where we also show the adjustment path for an increase in government expenditure financed by money creation. A larger budget deficit financed by money creation shifts the schedule (k = 0) to the right. Increased government expenditure requires a higher full-employment real interest rate, thus shifting (fI = 0) to the left. The economy moves with

329

BrazilKhapter 5

p=O h Fig. 5.6

A money-financed increase in government expenditures

oscillations from the initial low-inflation equilibrium to the new equilibrium with a higher inflation rate and smaller real money balances. As money increases, the nominal interest rate falls and so does the real interest rate, stimulating activity and pushing up the inflation rate. Gradually, inflation catches up with money growth and then exceeds it, reducing real cash balances and increasing the real interest rate. This story seems appropriate to the trajectory of the post-Cruzado Plan data, but it certainly does not fit the period of increasing inflation between 1979 and 1985, which requires a different explanation. Figure 5.7 shows the adjustment of inflation and real balances within the same basic model, but now with unchanged seignorage and rising equilibrium real interest rates. Seignorage is unchanged as long as the increased sum of budget deficits and noninterest current account surpluses does not get monetized but rather is financed by larger domestic debt. The increase in the equilibrium real interest rate can be attributed to crowding out, either as a result of growing government expenditure financed by debt or due to increased trade surpluses required to finance interest payments on the foreign debt. Consider a situation where a balance of payments crisis, such as the halting of capital inflows at the end of 1982, requires a real devaluation, which induces a growing trade surplus. Monetary targets imposed by the IMF program bring about a change in domestic government finance from money to debt. The sum of the budget deficit and noninterest current account

330

Eliana A. Cardoso and Albert Fishlow

L

h

Fig. 5.7 Successive increases in the full-employment real interest rate

financed by money creation is thus left unchanged. The higher real exchange rate and trade surplus are counterbalanced by a higher equilibrium real interest rate. How does the system move from one low-inflation equilibrium to another with higher inflation? The higher real exchange rate brings about increased activity in the tradable goods sector, thus raising demand and inflation. As inflation increases and money growth lags behind, the real interest rate increases. The economy adjusts in a cyclical fashion. Figure 5.7 shows a leftward-looping pattern of adjustment for inflation and real balances, induced by successive increases of the equilibrium real interest rate, This matches the Brazilian data in figure 5.8 extremely well. Figure 3.2 (chap. 3) confirms the upward trend of real interest rates in the years between 1979 and 1985. 5.6

Concluding Remarks

Between 1979 and 1985, Brazilian inflation doubled three different times. From 45 percent in mid-1979, it jumped to 100 percent in 1980-82, and then to 200 percent in 1983-84. At the end of 1985 and the beginning of 1986, the annualized inflation rate grew to 400 percent. The Brazilian inflationary process cannot be explained, as we argued above, simply by reference to increasing budget deficits financed by money creation. This does not mean that Brazilian budget deficits were not large or that they did not have an important role in sustaining inflation. But such a process has to be understood in the light of changing sources for financing the budget and the economy. The inflation acceleration between 1979 and 1985 is linked to the switch from external to domestic finance and to the progressively larger trade surpluses that pushed up interest rates and inflation.

331

BraziVChapter 5

Real monetary base Fig. 5.8 Inflation and real monetary base (quarterly average)

The Cruzado Plan failed to pay attention to this aspect of the debt problem. Larger deficits were experienced and were financed by monetary expansion. At the same time, excess demand required a higher real interest rate to compensate. This combination led to a classic inflationary finance situation which could have been avoided by reducing the excess demand (by increased taxation) and by improved conditions of finance. Debt relief would have permitted smaller trade surpluses for interest service and would have provided a buffer against resumption of higher inflation rates.

Appendix 1: A Model of Seignorage The government budget constraint is defined as: (GI - T I )

(A.1)

(K, -

+ i f B , - l + i ; E f D t p 1=

K-11

+ (4- &,I

+

E,P, - Dt-1)

where: the nonfinancial component of the budget deficit; = interest payments on domestic debt; = interest payments on the external debt; E is the exchange rate; 8, - B l p 1 = domestic borrowing; D, - D I P ] = external borrowing; K, - K I P = domestic credit creation, equal to the change in the monetary base, H, minus the change in foreign reserves, F: G, - TI il4-l i:EfDf-l

=

Eliana A. Cardoso and Albert Fishlow

332

(A.2)

K, - K,-I = ( H , -

Ht-l) -

E,(F, - F , - l ) .

Consider now the balance of payments under the assumption that only the government borrows abroad:

('4.3)

+

F, - F t p 1 = N X , - iTD, -

(D, - D , - I )

where N X , = the noninterest current account. We also define:

n, = (P,/P,-I )

- 1 as the inflation rate;

j , = (E,/EfI ) - 1 as the devaluation rate; 1 it = ( 1 r , ) ( l II,) as the relation between nominal and real domestic interest rates.

+

+

+

We substitute (A.2) and (A.3) in (A.l) and use the definitions above to obtain (A.4), where R = the real exchange rate:

64.4)

(Gt-Tt) + r t ( B t - i / f ' - 1 ) domestic deficit corrected for inflation

+

R, N X , noninterest current account

= ( H , - H,-l)IP, + [(BtlP,) - (B,-l/Pt-I)l real seignorage increase in the real domestic debt = creation of real paper at home

+

Observe that as long as the current account does not deteriorate with a real devaluation, a real devaluation must be financed by a reduction in the domestic deficit, an increase in domestic debt, or by money creation. We can rewrite (A.4) as: ('4.5)

a, = ( H , - H , - 1)/P,

where: a, = (G, - T,)

+ r t ( B f - l / P , - l )- [(B,/P,) - (Bt-l/Pt-l)] + R,NX,.

We multiply the right-hand side of ( A S ) by H , H , - I / H , H , - l , thus obtaining:

(A.6)

a, = h,.Q,,

+

where h, = H , / P , , and a, = z,/(l z,), having defined z as the growth rate of the monetary base. We also define the growth rate of the real monetary base as: c~ = Q, - II, obtaining:

(5.1)

= dh -

rI

Appendix 2: Calculating the Budget Deficit Recall the budget constraint defined in appendix 1 as (A. 1):

333

Brazil/Chapter 5

(G,-T,)

+ i , B f P l + iTE,D,-l + E, (Dt - D f - 1 )

= (K, -

K f p 1+) (B, - B , - l )

Deflating ( A . l ) by P , and using the definitions of II,j , and r (app. l), we can write the budget deficit corrected for inflation as:

(Gr - Tr)/Pr + rr(Br- 1 IPr-

(A.7)

1)

+

{ [ (1 + i X 1 + j M 1 + ~ , ~ l - ~ H ~ , - l ~ ~ ~ =budget deficit corrected for inflation = BD, =[(K- G l ) ~ ~+f [l( B P f )- ( ~ f - l ~ p f -+l )[l( R f D , )- (R,-lDt-I)l.

- l ~

Using (A.2) and observing that: (Hf - H f - I ~ ~ = P , [ H f / P , )-

+

~ ~ , - l ~ ~ , - [l (~Hl ,

-l~P,-,)c~,/1+~,)1,

we can write the budget deficit corrected for inflation: (A.8)

BD,

=

b,

+ d, + h, + infltax,

where: =

bf

4

=

= hr infltux, =

B, IP, - Btp1l P f - l = increase in the real domestic debt; R,D, - R,- I D,-l - R, (F, - F,- = increase in the real external debt net of foreign reserves; H , / P , - H , - / P , - = increase in the real monetary base; (II,/l + II,)(H,~l/P,~l).

Table 5.5 in the text shows the share of the budget deficit, corrected for inflation, in GDP and its financing for the last four years. We next describe how those numbers were calculated. We use the information given in table 5A. 1. Net Debt of the Public Sector (in millions of cruzados)

Table 5A.1

I98 I

1982

1983

1984

1. Federal government and Central Bank Domestic debt (including the base) External debt 2. State and local government Domestic debt External debt 3. State enterprises Domestic debt External debt

2,324 919 1,405 1,325 1,052 273 4,875 1,824 3,051

5,903 1,961 3,942 3,570 2,815 755 12,206 4,832 7,374

35,060 8,387 26,673 1 1,530 8,579 2,951 47,903 16,839 3 1,064

129,541 47,781 81,760 38,655 27,602 11,053 161,230 54,335 106,895

446,319 176,518 269.80 1 152,789 103.4 I8 49,371 625,280 214,550 410,730

Domestic debt External debt

3,795 4,729

9,608 12,071

33,805 60,688

129,718 199,708

494,486 729,902

8,524

21,679

94,493

329,426

1,224,388

Total debt

=

(1)

+ (2) + (3)

Source: Banco Central do Brasil, B r a d : Program Economico, August 1986, pp. 34-35

1985

334

Eliana A. Cardoso and Albert Fishlow

To obtain the fraction of the budget financed by the increase in the real debt of the public sector, we first deflate the consolidated net debt of the public sector at the end of the year by the general price index at the end of the year. The difference between the real debt in two consecutive years divided by the real GDP is equal to the sum of columns 1 and 4 of table 5.5. Observe in table 5.5 that the numbers for the inflation tax are different from the numbers one can calculate for seignorage, since seignorage also includes the changes in the real monetary base. Adding columns 1, 2, and 4 of table 5.5, we obtain our measure of the share of the budget deficit, corrected for inflation, in GDP. We observe that our measure is different from other measures available for the share of the budget deficit in GDP, as shown in table 5.3. We next explain why our measure in column 2 of table 5.3 is different from the “operational deficit” (OD) published by the Central Bank (col. 5, table 5.3). The OD is calculated by eliminating from the deficit defined as “borrowing requirement of the public sector” (PSBR) the actual payments of monetary correction. Even if the PSBR were an appropriate measure of the total borrowing requirements of the public sector in Brazil, the operational deficit would not be a proper measure of the budget deficit corrected for inflation for the following reason. The monetary correction index is not always equal to the inflation rate, and the difference between the two represents capital gains or losses for the public sector. Consider, for instance, a year such as 1983, during which the government sold domestic debt with a clause for monetary correction equal to the devaluation rate. There was a 30 percent devaluation in February, and the inflation rate that year was well below the monetary correction paid on the domestic debt. By excluding monetary corrections from interest payments rather than inflation, the Central Bank is leaving aside capital losses actually incurred by the public sector. The next problem concerns the exclusion of the deficit of the monetary authorities from the PSBR. In Brazil, most subsidies are paid directly by the monetary authorities. Therefore, a concept that excludes the deficit of the monetary authorities underestimates the actual borrowing by the public sector. The Central Bank only started publishing data for the total debt of the public sector in January 1986, and a series for the period 1982-85 was then made available. The data for 1985, for instance, shows that the share of the increase in the total debt in GDP was 65 percent. In that same year, the data published for the PSBR share in GDP was 27 percent. The large difference comes from the fact that the concept of PSBR does not include the deficit of the monetary authorities. Our measure, by contrast, includes all expenditures by the public sector and takes into consideration the capital losses that the OD excludes.

335

6

BraziKhapter 6

Trade Policies and Consequences

Trade performance is central to Brazilian adjustment to the debt problem. With or without debt relief, net export earnings must be larger than past earnings to pay interest, let alone principal. Such interest claims, even under favorable projections, come to more than one-third of export receipts over the next several years. Brazil is exceptional among the Latin American countries in its successful combination of import substitution and growth, as well as diversification of exports from the late 1960s on. Three factors have been responsible. One is the size and efficiency of the industrial sector initially established behind protectionist barriers and its continuing extension to intermediate and capital goods. Second is a crawling peg exchange rate policy that since 1968 has led to repeated small devaluations as a response to differential internal and external inflation, thus avoiding long periods of overvaluation. Third is an active promotion of manufactured exports through subsidies that offset the negative effects of import tariffs and controls. Together, these circumstances explain the growth of Brazilian merchandise exports from 1965 to 1980 of 9.4 percent a year, placing the country ahead of Singapore and narrowly behind Hong Kong. Such a performance permitted Brazil to follow a strategy of indebtedness in adjusting to the oil shock of 1973. The continuation of this performance, although attenuated since 1980 after the reemergence of the oil shock, has been central to coping with the debt. At the same time, Brazil has been able to restrain imports as a result of continuing expansion of alternative domestic supply sources. The 1980-84 level of real imports, excluding wheat and oil, is lower than in 1970-74, despite an increase in output of almost 80 percent. Taken together, this combination of export growth and import substitution have produced large export surpluses in recent years. Table 6.1 provides the numbers for the trade balance and figure 6.1, an illustration. The trade surpluses of the 1980s have underwritten interest payments of $10 billion a year, along with domestic recovery since 1984. These interest payments and recovery, however, have not been without the attendant and cumulative internal problems that we have stressed in previous chapters. In this chapter we examine more closely the policies that have made such an outcome possible. We begin with a discussion of the importsubstitution industrialization of the 1950s and the liberalization of the 1960s. We then examine in more detail exchange rate policy since 1964 and import and export responsiveness. We conclude with some thoughts on future prospects.

336

Eliana A. Cardoso and Albert Fishlow

Table 6.1

Trade Balance (annual averages, in millions of U.S. dollars) Period

Exports

Imports

1950-59 1960- 69 1970-73 1974-76 1977-78 1979-80 1981-82 1983-85 1986

1,447 1,590 3,958 8,917 12,390 17,688 21,734 24,848 22,393

1,276 1,380

4,045 12,376 12,853 20,520 20,743 14,171 12,866

Trade Balance 171 210 - 87 - 3,459 464 -2,832 99 I 10,677 9,527

Source: Boletim do Banco Central.

6.1 From Import-substitution Industrialization to Trade Liberalization Brazilian trade policies after the Second World War were profoundly shaped by the experience of the Great Depression of the 1930s. The significant decline in import capacity after 1929 as a result of falling prices of coffee and disruption of international capital markets created an immediate balance of payments problem. Brazil’s response was to cease debt service, limit imports, and expand internal demand. Expansionary fiscal and monetary policies coupled with real devaluation and exchange licensing created favorable conditions for industrial development in the 1930s. Although one can trace the beginnings of industrialization from an earlier period, the impetus of the Depression was decisive in emphasizing how reduced reliance on international trade was not only compatible with, but also favorable to, domestic growth. By 1939 imports of consumer nondurables accounted for less than one-tenth of total supply. Production of many other goods was making rapid headway. The continuing constraints on imports from the supply side during World War I1 reinforced the impulse to industrialization from the Depression decade. There was a difference. Brazil emerged from the war with a trade surplus as a result of expanded exports. The accumulation of reserves and favorable commodity prices, especially the surge during the Korean War, provided enhanced import capacity. But liberation was temporary. After 1947, exchange rate overvaluation and import licensing were used to support domestic industrialization and discriminate against primary exports. More sophisticated rationing devices were progressively introduced. In 1953 an auction system replaced the earlier quantitative controls. The auctions functioned as a variable tariff system whose level for different categories depended upon the amount of exchange made available. At one extreme stood wheat, petroleum, and certain essential capital goods whose imports were facilitated at a special, below-auction rate; at the other extreme were finished consumer goods on which a premium rate of more than 500

337

BrazilKhapter 6 14

12

--

I

I

1965

1975

I

n C 0

v)

3

P

-2L

-4

1955

1985

Fig. 6.1 Trade balance (in millions of U.S. dollars)

percent was charged at the time the system was abolished in 1957. Such a system enabled the government to expropriate the surplus derived from quantitative controls and to turn some of it to subsidizing nontraditional exports. In 1957 the multiple exchange rate system was simplified as a result of a tariff refom. Three basic tariff levels were established: 0-10 percent, 10-60 percent, and 60-150 percent. In order to establish surcharges, the purchase of many items required exchange procured at special auctions and at a large premium relative to the general rate. A new administrative structure was set up to grant discretionary adjustments to the stipulated tariffs, and the Law of Similars was reactivated, limiting importation of products already or potentially domestically produced. Until the end of the 1950s, the policy of favoring domestic industry worked reasonably well. Overvalucd exchange rates taxed primary exporters and converted the gains to the direct advantage of priority importers. It was a second-best policy for reallocating resources when the government itself could not tax directly. As long as exports were inelastic in their supply (or their international prices enjoyed a boom), such a strategy was feasible. Over time, however, as export elasticity increased, the balance of trade deteriorated as export volume stagnated. Foreign direct investment, lured by protected access to the Brazilian market and granted favorable exchange rates, partially compensated for the deterioration of the trade balance and kept the balance of payments from similar erosion. But that source, too, became smaller after the initial investment in new sectors was undertaken. As the balance of payments became more problematic, so too did the fiscal position of the government. As long as import substitution could be subsidized by indirect taxes, deficits were avoided. Increasingly, as exchange rates had to be revalued and bonuses provided for nontraditional exports,

338

Eliana A. Cardoso and Albert Fishlow

that source disappeared. At the same time, as industrialization proceeded, demands upon the government for investment in infrastructure simultaneously increased. The inflation tax had to substitute for other sources of revenue, as well as meet rising expenditures. It could do so only at the cost of accelerating inflation rates that reached annual levels of 30 percent and threatened to go out of control. A classic devaluation with stabilization was attempted in 1961 during the short-lived presidency of Janio Quadros. Its failure was followed by accelerating deterioration in both of the external accounts, and internal growth and inflation. The coup in March 1964 brought a military government with a new, more market-oriented agenda and greater authority with which it could be implemented. The 1964-67 period saw a more thorough effort at liberalization than had been attempted earlier. Serious attention was paid to the exchange rate and incentives for exports. Import liberalization was much less profound. While tariffs were lowered in 1967 and real prices of imports fell, an equal determinant of the price decline was real appreciation of the cruzeiro after its initial devaluation in 1964. Unadjusted measures of nominal and effective protection that fail to allow for the redundancy of much of the legal provisions-because no imports were realized and domestic competition determined price-exaggerate the changes after 1964. Perhaps the best measure of that redundancy is the limited impact of the large tariff reform in 1967 on domestic industry or on the composition of imports. Brazilian industry was more efficient than simple calculations of effective protection suggest. Some of those estimates are presented in the statistical appendix, but must be interpreted with caution. On the export side, however, more decisive changes were introduced after the change in government. Export incentives for manufactures began in 1964 with measures reducing the anti-export bias of value-added taxes. Gradually these policies shifted emphasis to credit subsidies that became important in the 1970s. Cardoso (1980) computes a time series estimate of the combined effect of fiscal and tax incentives on prices received by exporters. Musalem (1981) broadens the estimates by the inclusion of financial subsidies. I The major incentives and when they were introduced were: 1. June 1964: Effective implementation of the drawback system (import duty exemptions for inputs used in manufactured exports). 2. April 1965: Exemption of the IPI, value-added tax paid into manufactured exports and export sales. 3. February 1967: Exemption of the ICM, value-added tax paid on manufactured exports and export sales. 4. February 1967: Introduction of an income tax credit. A tax refund was granted to exporters of manufactured products proportional to the share of total output exported.

339

BraziUChapter 6

5. July 1968: Creation of an IPI credit consisting of a subsidy for manufactured exports based on the product’s IPI rate. 6. September 1971: Creation of an ICM subsidy similar to that for the IPI.

This system of subsidies was a source of increasing conflict with the United States in the 1970s. At the end of the decade it was agreed to phase out the tax subsidy-as opposed to exemption-components, and initial steps were taken. The seriousness of the debt crisis in the early 1980s led to the reinstatement of subsidy payments. A second major decision was the adoption of the crawling peg exchange rate regime in August 1968. It was designed to avoid the discontinuity of the large readjustments to nominal rates necessitated by higher rates of Brazilian inflation. Discrete devaluations permitted an intervening appreciation of the real rate that discouraged exports and encouraged self-fulfilling speculation. Small and frequent mini-devaluations were linked to domestic and external inflation, particularly of the United States, and significantly reduced the variance of the real exchange rate. This system, analyzed more closely in the next section, underlies the better Brazilian export performance in the 1970s and 1980s.

6.2 Nominal and Real Exchange Rates As noted above, the adoption of a crawling peg in 1968 offered an escape from the problem of overvaluation resulting from the higher rates of price increase in Brazil compared to its trade partners. The existence of an exchange rate policy designed to stabilize real exchange rates is evident from the comparison between inflation, nominal devaluations, and the real exchange rate behavior shown in figure 6.2. Real exchange rates measure competitiveness, and they behave quite differently than nominal exchange rates. The analysis of trade issues, such as the performance of the export sector and import growth, requires measures of the domestic and international competitiveness of the traded goods sector. In this section we look at different measures of competitiveness in Brazil. What matters for competitiveness is prices relative to costs, and prices relative to the prices offered by an alternative competing supplier. Before proceeding further, a word about terminology is warranted. All cruzado prices are denoted by P and p (the former are foreign, the latter are domestic), all foreign currency prices by (*). The cruzado price of foreign exchange is represented by E. As a convention we conduct all price comparisons in cruzados and thus translate all foreign currency prices into cruzados at the relevant exchange rate. The different measures of the real exchange rate are denoted by e and a subscript. An increase in the real exchange rate denotes a real appreciation. Conversely, its reduction denotes a real depreciation and an increase in competitiveness.

340

Eliana A. Cardoso and Albert Fishlow 140

i0

I

I

I

I

I

I

Fig. 6.2 Real exchange rate (appreciation up)

The first measure to consider is a broad measure of cost levels in Brazil compared with costs in the rest of the world. As measures of costs we utilize economywide wholesale prices. Brazilian competitiveness is enhanced if foreign wholesale prices, in cruzados, rise relative to Brazilian wholesale prices. Denoting home and foreign wholesale prices by P and EP*, respectively, we have: (1)

e, = PIEP*

In order to examine the performance of the manufactures export sector, we need to develop price measures that indicate whether that sector is gaining competitiveness. It is natural to think in terms of demand and supply. We first look at the foreign demand side for Brazilian exports by comparing prices of Brazilian exports with those of manufactures of foreign competitors. The relevant price comparison is the price of Brazilian manufactured exports, Px, compared with the export price of manufactures of competitors, EPZ:

(2)

ed = P x / E P :

On the supply side, an increase in competitiveness or profitability occurs if export prices rise relative to the prices of manufactures in the home market. If so, we would expect increasing amounts of manufactures production to be diverted toward exports and away from home sales. A natural measure of competitiveness on the supply side, then, is the ratio of domestic manufactures price, Px, to the export prices of Brazilian manufactures:

(3)

es =

Px

lP,

341

BraziKhapter 6

Since there are advantages to sales in the export market relative to the domestic market because of the subsidy programs, we include another index taking into account the fiscal and credit subsidies: (3')

ess =

Px

IPXS

Further information about real exchange rates comes from the relative price of home goods and traded goods. This measures the incentive to reallocate resources to the production of tradables. We calculate the ratio of the index of construction costs, pn, which we take as a proxy for the behavior of prices of home goods, to the domestic price of industrial goods, PA-

(4)

en, = P n

JP~

There exists one more index which is readily available and helps clarify further the change in trade performance. That is the index of terms of trade, which is simply the ratio of the price index of total imports, P,, to the price index of total exports, P , ; (5)

4 = p,

IPtx

Table 6.2 and figure 6.3 show the different measures of the real exchange rate we have introduced above and the terms of trade between 1970 and 1986. The ratio of economywide wholesale prices, e,, shows relative stability from 1970 until 1977. Thereafter, because of depreciation of the dollar, Brazil gained against other currencies. The devaluation in 1979 extended this pattern, but it was reversed in 1981 as a result of the limited devaluation of the cruzeiro during the course of 1980. The appreciation of the dollar undermined the Brazilian competitive position, a deterioration that was finally remedied by the maxi-devaluation in early 1983. The ratio of home to traded goods, en,, confirms a generally similar pattern, with a strong tendency in favor of tradables after 1979. A principal conclusion, then, is that Brazil's aggregate real exchange rate showed modest depreciation, but in the face of the decline in the terms of trade, q, after 1977, that may not have been adequate to maintain a strong balance of payments. Debt, to some extent, served as a substitute source of foreign exchange. A more favorable picture emerges when manufacturing competitiveness is considered. Brazilian export prices fell relative to those of competitors throughout the period. Such a movement, at the beginning of the period and again after 1982, possibly reflects the impact of Brazilian subsidies. Exporters could gain larger market share by passing along some of the benefits in the form of final prices.

342

Eliana A. Cardoso and Albert Fishlow Different Measures of the Real Exchange Rate and the Terms of Trade in Brazil, 1970-86 (1977 = 1)

Table 6.2

Year

ec

ed

e,

e**

e.,

4

1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985

1.059 1.059 1.027 0.952 0.929 0.925 0.969 I .ooo 0.896 0.819 0.766 0.929 0.981 0.812 0.849 0.837

1.218 1.168 1.126 1.092 1.067 1.034 1.025 I .ooo 0.916 0.815 0.714 0.866 0.941 0.723 0.723 0.714

1.620 1.479 1.380 I .085 0.901 0.986 1.056

I .95 1 1.732 1.610

0.936 0.932 0.942 0.977 0.978 0.955 1.101 1.ooo I . 105 1.153 0.951 0.909 0.896 0.811 0.738 0.803

0.89 0.82 0.87 0.95 0.78 0.76 0.85 I .oo 0.86 0.79 0.65 0.55 0.54 0.53 0.58 0.58

1.000

1.056 0.986 0.944 1.127 1.296 1.056 1.056 1.040

1.244

1.024 1.049 1.049 1.ooo 1.049 1.024 1.024 1.195 1.341 1.122 n.a n.a

Source: The statistical appendix, tables A.14 and A.18. Nore: A description

statistical appendix.

of the procedures used in the calculations of the real exchange rates is found in the

e, = ratio of wholesale prices of Brazil and U.S.,both in cruzados. e, = effective real exchange rate published by Morgan Guaranty. e, = ratio of domestic manufactures price and export price of manufactures. e, = same as above, including fiscal and credit subsidies to exporters. e,, = ratio of construction costs and domestic price of industrial goods. q = terms of trade.

Such a pattern on the demand side is echoed on the supply side, especially when the effect of subsidies is included. Domestic industrial prices did not rise as rapidly as prices of manufactured exports up to 1977. Thereafter, however, the incentive to export was attenuated and began to reverse. By 1982 there was a significantly greater attraction to domestic sales rather than exports, as measured by prices. A fall in the real volume of industrial exports between 1981 and 1982 triggered the devaluation in 1983. Real depreciation became inevitable as the interest rate shock was added to the rise in oil prices, as well as to the slowdown in the growth of the industrial economies. It added to the real income losses associated with the continuing decline in the terms of trade and loss of purchasing power over imports. One needs only to look at the measures of gross product in dollars presented in table A.2 of the statistical appendix to grasp the extent of the loss. On a per capita basis, income in 1985 (again, taking into account the terms of trade loss of 5 percent) was less than in 1977! The brunt of the fall occurs after 1982 and is also registered in the decline of real industrial wages by more than 20 percent between 1982 and 1984.

343

Brazil/Chapter 6

g 0

pc

0, x

Q)

U

C -

Fig. 6.3 Different measures of the real exchange rate

6.3 Import Performance As noted earlier, liberalization efforts after 1964 yielded some advance in reducing the relative costs of imports. In conjunction with export subsidies, the strong bias in favor of import substitution was reduced. Although tariffs rose modestly in 1969, they were reduced again in 1972. Between 1964 and 1973 the import ratio actually rose, reversing a long downward tendency. It reached its high point in 1974 with large increases in oil imports as well as of other speculative purchases. Thereafter, under the pressure of persistent imbalance in the external accounts, imports have been limited. The ratio of imports to product in 1985, at less than 7 percent, is back to its levels of the mid-1960s. It did not take long for protection to increase after the oil shock. In June 1974 all forms of import financing for products subject to nominal tariffs of 55 percent or higher were suspended; tariff rates for nearly 900 products were doubled, and import duty reductions on capital goods were suspended. In 1975 tariffs on intermediate goods were raised. Imports subjected to tariff rates equal to or higher than 37 percent were made subject to prior payment. At a time when inflation was 30 percent, import legislation imposed a 100 percent prior deposit rate, without interest, refundable six months after the transaction. In 1976 the economic authorities imposed a complete ban on automobiles, pleasure crafts, toys, and other goods judged to be luxuries. Thereafter, levels of protection have not been modified much. Donald Coes (1986) has recently constructed an index of Brazilian trade liberalization that is graphed in figure 6.4. It traces the dismantling of earlier liberalization after 1973. By the early 1980s the degree of restriction was comparable to the last severe balance of payments crisis of the early 1960s.

344

Eliana A. Cardoso and Albert Fishlow

I

I

1

I

I

I

I

1947 1951 1955 1959 1963 1967 1971 1975 1979

Fig. 6.4

Index of trade liberalization in Brazil, 1947-82

Again, import protection depends for its effectiveness on utilization of nontariff barriers. Table 6.3 shows the legal and implemented average tariff rates by broad categories for 1982. The former are considerable, averaging 47 percent excluding the nondutiable fuel and wheat imports. The reason is straightforward: if imports are permitted at all, they are almost always exempt from tariffs or receive large reductions. It is the license that has again become the binding constraint rather than the price. In table 6.4 these 1982 data are classified in another way that further illuminates this point. Only 11.3 percent of imports pay the legally stipulated rate. Others enter as privileged products, such as wheat, or by virtue of special regional trade agreements involving preferential access, and still others enter by virtue of their priority in connection with exports or domestic development projects. There is a large degree of administrative discretion. Import reduction, as we have seen, was a central part of the Brazilian balance of payments adjustment after 1980. The relative success of the policy depended not merely upon import restriction but also on an active policy of investment directed to import substitution. Despite its limitations, discussed in chapter 3, the Brazilian strategy of subsidizing the intermediate and capital goods sectors in the 1970s did make possible a decline in imported inputs without disrupting domestic production. Table 6.5 provides evidence on the sectoral import coefficients. Paper and paper products are an example of not only a more than proportional decline in imports, but even of conversion to exports. Similar transitions have occurred in machinery, metals, vehicles, and electrical products. But it is also important to point out the role played by the increase in domestic production of crude oil. It is the largest source of import saving after 1979. These supply adjustments did not permit the Brazilian economy to escape unscathed. A severe recession was also necessary in 1982-83 to help restrict demand for imports. Declines in investment reduced imports of intermediate

345

BrazilKhapter 6

Table 6.3

Nominal and Actual Tariff Rates in Brazil, 1982 (weighted averages, in percentages) Category

Nominal

Actual

Durable consumer goods Nondurable consumer goods Construction materials Capital goods for industrial usc Capital goods for agricultural use Transport equipment Inputs for industrial use Inputs for agricultural use Fuels Other imports

139.4 57.5 54.5 55.6 32.6 51.8 40.9 9.4 0.7 2.8

18.0 9.2 4.0 15.7 22.5 4.3 13.5 I .5 0.0 1.2

Total imports except fuels and wheat Total imports

41.5 22.4

13.7 5.9

Source: Moreira and Araujo (1984).

Table 6.4

Nominal and Actual Tariff Rates for Imports Classified by Tax Regime in Brazil, 1982 (weighted averages, in percentages)

Tax Regime Oil Wheat Related to exports Development programs International agreements Products without fiscal exemption Others Total

Share in Total Imports

Nominal Rate

Actual Rate

46.7 4.0 9.4 5.1 5.3 11.3 18.3

0.0 45.0 57.4 62.3 42.4 44.5 27.1

0.0

0.3 0.8 7.0 44.6 0.4

1m.o

22.4

5.9

0.0

Source: Moreira and Araujo (1984).

and capital goods more than proportionally. A review of recent research on import demand confirms its sensitivity to economic conditions. In table 6.6 we present the findings of three recent studies. All the equations show large income elasticities ranging at the low end from 2 . 3 in Abreu (1987) to 2.8 in Moraes (1985), and finally to 3.4 in Dib (1981). It is no accident, moreover, given the continuing low level of imports in recent years, that the most contemporary study yields a lower income elasticity. Price elasticities tend to be less statistically significant and uniformly lower than unity in the most recent studies. Moraes takes into account the possibility of bias due to the existence of quantitative restrictions. He assumes that the government looks at the ratio between reserves and debt, X . It restricts imports whenever the ratio is low and liberates imports whenever this ratio grows above a desired target. The coefficient of X appears with the expected sign and is significant, but its inclusion does not affect the other estimated elasticities.

346

Eliana A. Cardoso and Albert Fishlow Import Coefficients. Share of Imported Inputs in the Total Sales of Different Industrial Sectors in Brazil, 1974-75 and 1982-83 (in percentages)

a b l e 6.5

Sector Nonmetallic minerals Metals Mechanical instruments Items for electrical use Material for transport Paper and cardboard Rubber Chemical products Artificial plastics Total

1974-75

1982-83

3.0 7.8 9.9 16.6 6.8 7.3 6.4 15.9 7.1

0.7 3.0 3.6 7.7 4.0 1.3 2.7 9.0 I .7

7.5

3.8

Source: Kume 1985.

Bble 6.6

Import Equations

Dib (1981): period 1960-79; total imports except oil and wheat; real import price includes tariffs; OLS. Income

R2 = 0.97

Relative Price -1.11

3.42 (5.83)

Coefficients t-statistics

SER

D.W. = 1.85

=

(-

Time Trend

-0.16 3.49)

3.5)

(-

0.117

Moraes (1985): period 1976-84; quarterly data; imports of manufactured goods; tariffs not included in import prices; OLS. Income I .39

Coefficients t-statistics

R2 = 0.94

Relative Price

Lugged Imports

Time Trend

-0.30 (-2.34)

0.52 (5.59)

- 0.02 (-5.59)

(5.68) Q(18) = 10.55

SER

=

0.08

h = 0.284

Abrcu (1987): period 1976-85; quarterly data; total imports except oil and wheat; MQC.

Ccefficients t-statistics R2 = 0.96

Income

Relative Price

Capacity Utilization

Time Trend

Lagged Imports

1.58 (7.49)

- 0.33

1.03 (3.55)

-0.018 (-6.01)

0.32 (4.03)

SER

=

0.06

(-

2.93)

h

=

-0.296

Abreu introduces capacity utilization as a causal variable as well. The response of imports is positive as utilization increases. That may measure not only lack of domestic supply response but also the strictness of nontariff limitations. Licenses are more readily conceded when domestic activity is close to full employment. Import substitution, as measured by the trend variable, is significant in all three versions and has almost identical estimates in the quarterly Moraes and Abreu equations. Interestingly, it is highest in the Dib's study for 1960-79

347

BrazilKhapter 6

rather than for the more recent period. Thus the emphasis given to the structural change in import substitution in the early 1980s, which was a result of the previous investment in the 1970s, must be qualified. Abreu’s explicit tests for change in the income elasticities in the recent period yielded mixed results. In the annual data there was no evidence of a shift. For the quarterly data, however, one could accept the hypothesis of a difference in 1984 and 1985 for capital goods and competitive industrial imports. 6.4

Export Performance

Brazilian export performance was central to establishing creditworthiness in the 1970s and remains critical for satisfactory debt service in the 1990s. Brazil was an attractive debtor not only because of record industrial growth from 1968 to 1974 but also because of improved export results during the same period. After stagnation in the 1950s and early 1960s, export receipts experienced large increases in the beginning of the 1970s. Despite borrowings from the Eurocurrency market, the current account deficit and the debuexport ratio remained moderate, confirming Brazil’s creditworthiness. Throughout the 1970s Brazilian exports continued to expand, encouraged again by favorable prices after 1975. They were the basis for the importsubstitution adjustment strategy of that decade. Unlike the surge of industrialization in the 1950s, exports were given close attention, and the subsidies for exports of manufactures (described above) substantially countered the effects of greater protection and exchange rate overvaluation. A World Bank (1983) study found that for the manufacturing sector as a whole in 1980, the anti-export bias was only 2 percent. Some sectors, however, most prominently transport equipment, actually received greater rewards from exports. Favorable prices for agricultural commodities, along with a shift in the composition of exports toward fast-growing industrial products, were important factors in the improved export performance of the 1970s. From 1970 to 1979, total receipts increased from $2.7 billion to $15.2 billion. Price increases in primary exports accounted for almost 40 percent of this rise in value; real growth in industrial exports accounted for almost another 20 percent. Table 6.7 shows this change in the structure of exports from 1965 to 1985. Coffee alone accounted for almost half of export earnings in 1965, while manufactured goods accounted for less than 10 percent. Twenty years later the proportions were reversed. Within the category of manufactured exports, as table 6.8 shows, the transformation was of three kinds. First, there was a large increase in more traditional exports, footwear in particular. This was especially true after 1975. Second, there was an emergence of new sectors after 1975, such as paper and chemical products, that had previously been exclusively import substituting. Third, and not made explicit in the table,

348

Eliana A. Cardoso and Albert Fishlow

Table 6.7

Composiion of Exports (as a percentage of total exports)

coffee beans Other primary goods Semimanufactured goods Manufactured gwds Spccial transactions

1965

1970

1975

1980

1985

44.3 37.3 9.6 8.1 0.7

34.3 40.8 9.1 14.9 0.9

9.9 48. I 9.8 29.8 2.4

12.4 29.8 11.7 44.8 1.3

9.1 24.2 10.8 54.9 1 .O

Source: Boletim do Banco Central.

%ble 6.8

Composition of Manufactured Exports (as a percentage of total manufactured exports)

Footwear and textiles Industrialized agricultural products Chemical products Metal and mechanical products Paper and cardboard Other

1965

1970

1975

1980

1985

6.7 23.0 0.0 27.6 0.1 42.6

7.8 24.1 2.9 32.4 0.1 32.7

17.9 17.0 2.6 26.2 1.1 35.2

11.1 15.5 7.8 29.0 2.1 34.5

11.7 13.6 11.9 26.3 2.3 34.2

Source: Boletim do Banco Central.

there was a shift over time within categories to more sophisticated industrial exports. For example, the export of entire vehicles replaced auto parts, and the export of steel replaced pig iron. In addition, entirely new products emerged in the export market, such as airplanes. Despite the relatively strong performance of exports throughout the 1970s, especially considering the emphasis given to import substitution from 1975 on, there is a significant change in the rhythm of export performance after the middle of the decade. A recent decomposition of the sources of Brazilian export growth brings out the difference. The results are presented in table 6.9. Note that in the 1974-78 period, Brazilian exports are more influenced by world trade growth than by increased market share. The participation of total Brazilian exports in world trade (excluding fuels), after rising sharply from the late 1960s to 1977, falls off thereafter; the 1979 level is lower than Table 6.9

Source of Growth of Exports (as a percentage of total change) Total (excluding fuel)

Growth-of-world-tradeeffect Composition-of-exportseffect Destination-of-exportseffect Competitivenesseffect Source: Horta 1985.

Manufactured

1971178

1971174

1974178

1971178

1971174

1974178

71.4 - 9.0

64.8 -0.1

- 1.5

- 13.6

39.1

48.9

100.9 -20.0 1.0 18.1

30.2 -0.1 -3.9 73.8

33.1 -0.2 -4.5 70.6

57.2 -0.1 -0.2 42.7

349

BrazilKhapter 6

in 1973. Gains in manufactured products continued, but at a slower rate. The consequence is a notable slowing in the real growth of exports after 1974, both the total and that of manufactured products alone. In the 1980s, after the second oil shock, the deterioration of primary product prices, along with slowing world trade growth, hit Brazil hard. Export receipts followed a more irregular pattern than before, falling from 1981 to 1982, and after a sharp rise in 1984, again declining. In 1987 exports were only 10 percent greater than in 1980. The value of world nonoil trade grew by more than 30 percent in the same interval, hence the Brazilian share fell, but only after 1984. The volume of Brazilian exports grew at a rate of about 5.6 percent a year from 1980 to 1986, while prices have declined by 3.5 percent. Brazil’s performance is inferior to that of nonfuel developing country exporters, whose volume has expanded at a rate of 6.5 percent. In addition Brazil has experienced a larger decline in export prices, almost 1 percent a year. The most dramatic difference, however, emerges in a comparison with the performance of the principal developing country exporters of manufactures, whose volume expanded by more than 8 percent a year and whose terms of trade actually improved over this period. Just when the close of capital markets forced the debt service burden more substantially upon export receipts, Brazilian Performance became more erratic and less competitive despite real devaluation. Part of the reason for this is the continued, although diminished, importance of primary exports subject to a pronounced fall in prices. Another part is the strong recovery in internal demand after 1984. Table 6.10 presents some estimates of the response of Brazilian manufactured exports to relative prices and domestic capacity utilization. Although it is clear that Brazilian exports do adjust to relative prices with elasticities considerably larger than unity, and hence are sensitive to exchange rate and subsidy policies, it is also the case that strong internal demand typically subtracts from export supply in a significant way. The decline after 1984 in the volume of industrialized exports owes itself to this factor. Brazilian exports of manufactured products are still relatively small compared to the size of the internal market. Overall, the ratio is approximately 10 percent compared with values of roughly 50 percent for Korea and Taiwan. This difference explains much about post-1984 export performance in these countries. For the Asian nations, exports and production are virtually synonymous; trade penetration and industrialization are interlinked.

6.5 Conclusions A central element in the Brazilian capacity to service its debt is its trade performance. Large trade surpluses since 1984 have been achieved primarily as a result of continued import compression rather than regular growth of

Eliana A. Cardoso and Albert Fishlow

350

Table 6.10

Export Equations

Coes (1986): period 1959-77; OLS; all variables in logs; r-statistics in parentheses x = ratio of manufactured exports to total industrial production.

p = world price of manufactures adjusted by the Musalem incentive series and home industrial prices. b = measure of excess capacity. x = -3.78

+

(-2.82)

R2 = .I4

D.W.

=

1 . 6 7 ~- 0.018b (6.17) (-2.88)

1.92

Braga and Markwald (1983): period 1959-81; 3SLS; all variables in logs; r-statistics in parentheses.

X

=

e

= world price of Brazilian manufactured exports adjusted by the Cardoso incentive series and home

B

= measure of excess capacity. = world price of Brazilian manufactures. = world income.

quantum of manufactured exports. industrial prices.

P, P P*

=

price of competing manufactures.

X = - 1.9

-k 2.5e -

1.5B

(11.7)(-2.1)

P, = -1.4 R2 = 0.97

-

0.1X

+ 0.5Y' + P*

MSE = 1.7

exports accompanied by expanding import capacity. This makes an important difference in developmental consequences, as we elaborate in the next chapter. The initial conditions of a large trade surplus tend to create an illusory optimism about the ease with which large external transfers of resources can be realized. Applying even differential compound rates of growth to exports and imports leaves a scope for larger trade surpluses in the future. Such exercises are applications of arithmetic rather than economic analysis. That analysis suggests a more cautious view, One question is the degree of Brazilian competitiveness required for large and regular growth of manufactured exports. To the extent that competitiveness can only be achieved in the short term through real devaluation and hence declining real wages, political feasibility also becomes a major factor. At recent low rates of investment and imports, Brazilian technological capacity is handicapped vis-A-vis strengthened Asian NIC competitors. A second question deals with the prices of primary exports. An end to the abrupt decline experienced since 1979, the most serious such decline during the postwar period, would be an important boost to Brazilian prospects. Despite dollar devaluation, primary export prices have not risen steadily, while those of manufactured products have done so. Current projections forecast some modest relief in the next few years, but they have been consistently wrong in the last several.

351

BrazilKhapter 7

Third, the continued imbalance in the payments of the industrial countries threatens expansion of world trade in two ways. One is a recession induced by restrictive macroeconomic policies in the United States unmatched by expansion in Germany and Japan; the other is rising protectionism and increased attention to bilateral trade balance that limit the scope for debtor country import penetration. Fourth, on the import side, there may be an overestimate of the structural reduction in Brazilian income elasticities owing to the large program of import substitution. Artificially low investment, despite income recovery in 1985 and 1986, has helped to keep import demand in check. But such a pattern cannot be projected, nor is it consistent with sustained development. Export growth itself becomes dependent upon access to imports. Brazilian ability to generate large trade surpluses due to previous investment should not be exaggerated. Taken together with recent experience, these considerations suggest that Brazilian trade performance and policy are not settled issues. It will require a sustained effort to assure that export growth occurs in a more regular fashion. It will also require careful attention to assure that import limitations and emphasis upon the internal market do not, in the end, worsen rather than improve the balance of payments. In the last analysis, meaningful trade projections cannot be made within the context of debt service requirements alone. A broader perspective, incorporating the debt as an integral part of the Brazilian development problem, is necessary.

7

Epilogue: Debt and Development

Thus far we have stressed the problems of macroeconomic adjustment to the large Brazilian external debt. Debt service has imposed new demands upon trade performance and upon the public sector. The former has responded better than the latter. Since 1984, record export surpluses have been attained. Fiscal deficits and inflationary pressures have persisted, however, and the Cruzado Plan, despite its promising start, has failed. Once again, while trade surpluses have recovered to the $10 billion level, inflation exceeded 400 percent in 1987. Instead of external and internal equilibrium being joint, the strong balance of trade has been gained, to some extent, at the cost of domestic imbalance. But the debt problem goes even deeper. It is rapidly evolving into the central problem of economic development in the 1990s. After almost a decade of slower growth to facilitate adjustment through limited imports, future prospects are dimmed by low rates of investment since 1982.

351

BrazilKhapter 7

Third, the continued imbalance in the payments of the industrial countries threatens expansion of world trade in two ways. One is a recession induced by restrictive macroeconomic policies in the United States unmatched by expansion in Germany and Japan; the other is rising protectionism and increased attention to bilateral trade balance that limit the scope for debtor country import penetration. Fourth, on the import side, there may be an overestimate of the structural reduction in Brazilian income elasticities owing to the large program of import substitution. Artificially low investment, despite income recovery in 1985 and 1986, has helped to keep import demand in check. But such a pattern cannot be projected, nor is it consistent with sustained development. Export growth itself becomes dependent upon access to imports. Brazilian ability to generate large trade surpluses due to previous investment should not be exaggerated. Taken together with recent experience, these considerations suggest that Brazilian trade performance and policy are not settled issues. It will require a sustained effort to assure that export growth occurs in a more regular fashion. It will also require careful attention to assure that import limitations and emphasis upon the internal market do not, in the end, worsen rather than improve the balance of payments. In the last analysis, meaningful trade projections cannot be made within the context of debt service requirements alone. A broader perspective, incorporating the debt as an integral part of the Brazilian development problem, is necessary.

7

Epilogue: Debt and Development

Thus far we have stressed the problems of macroeconomic adjustment to the large Brazilian external debt. Debt service has imposed new demands upon trade performance and upon the public sector. The former has responded better than the latter. Since 1984, record export surpluses have been attained. Fiscal deficits and inflationary pressures have persisted, however, and the Cruzado Plan, despite its promising start, has failed. Once again, while trade surpluses have recovered to the $10 billion level, inflation exceeded 400 percent in 1987. Instead of external and internal equilibrium being joint, the strong balance of trade has been gained, to some extent, at the cost of domestic imbalance. But the debt problem goes even deeper. It is rapidly evolving into the central problem of economic development in the 1990s. After almost a decade of slower growth to facilitate adjustment through limited imports, future prospects are dimmed by low rates of investment since 1982.

Eliana A. Cardoso and Albert Fishlow

352

Investment averaged 23 percent of gross product in 1975-79 and 21.6 percent in 1980-82, but only 17.7 percent in 1984-86. Figure 7.1 illustrates the behavior of investment and savings shares in GDP. Recent investment rates are well below the requirements necessary for sustained rates of growth adequate to absorb a growing potential labor force and capable of underwriting a continuing transition to political democracy. In this concluding chapter, we explore both this development problem and the adequacy of debt policy. The moratorium on interest has come to an end, and a conventional new money package was granted in 1988. That agreement, unfortunately, did not provide a happy ending to this story of Brazilian indebtedness. Now, a debt reduction agreement awaits the inauguration of a new president in 1990.

7.1 Crowding Out Investment At the heart of the Latin American adjustment to the requirements for continuing interest payments after 1982 has been a large export surplus and transfer of resources to creditors. Although principal has been rescheduled, interest payments have not. Voluntary lending has been minimal, and concerted arrangements have been limited. As a consequence, between 1983 and 1987, Latin America has paid some $120 billion more in interest than it has financed from new borrowing. The experience in Brazil has been typical for the region, as shown in figure 7.2. Indeed, its resource transfer between 1983 and 1987 amounted to about $45 billion, more than its share of regional debt. The export surpluses

24 -

n

20

a

-

2 16Y-

O

2

0

12 -

8-

4-

-

O-,

1970

1972

1974

Domestic savings

1976

1978

1980

1982

1984 1986

0Foreign savings

Fig. 7.1 Gross investment and savings in Brazil, 1970-86

353

Brazil/Chapter 7

needed to effect the transfer have averaged more than 5 percent of gross product since 1984. The counterpart.domestic resources have come primarily at the expense of investment rather than consumption. Instead of reduced consumption paying the lion’s share of external interest in response to reduced national income, and thus a rise in the national saving rate, consumption has been sustained. The consequence has been high real interest rates that crowd out private capital formation in order to satisfy the equality between uses and applications of resources. More austere fiscal policy and greater saving would avert such an implication. But political realities and economic behavior prevent such a solution. Brazil, despite relatively high levels of taxation, has not found it easy to increase taxes enough to raise the cash surplus required to meet its internal and external interest obligations while simultaneously paying for its purchases of goods and services. The lesson from the failure of the Cruzado Plan was the difficulty of augmenting public sector financial capacity. The government eventually resorted to increasing indirect taxes, with only modest effect. Larger collections and reduced outlays are a policy imperative, but they may be easier to achieve while the economy is expanding, rather than during stagnation. Taxation has adverse effects on incentives. These effects, however, can be offset by the larger incomes in a growing economy. High interest rates have made it possible for the public sector to sell the internal debt needed to finance its borrowing requirements. But they have done so by portfolio substitution rather than by inducing an outpouring of new saving. Theory is ambiguous about the relative weight of income and substitution effects of interest rates. Careful empirical analysis suggests that

6 4

r % 2

W

re

0

- 0 c

a

2 a a

-2

-4

-6

Fig. 7.2 Real resource transfers (as a share of GDP)

354

Eliana A. Cardoso and Albert Fishlow

there is virtually no systematic net effect. On the other hand, high interest rates tend to increase the large component of government expenditures dedicated to debt service and thereby to build in continuing high deficits. If consumption is to be restrained, other instruments will be required. Whether by reason of pent-up demands after earlier austerity or uncertainty about the future, Brazilian consumption has been quite strong and unresponsive to larger foreign transfers. Saving is too small, particularly in a world in which foreign saving is likely to remain limited over the next decade. Domestic saving will have to increase, but the more promising route is through higher rates of growth rather than austerity. There is a medium-term vicious circle. Investment is too limited to sustain adequate rates of growth while underwriting expanding exports and permitting the transfer of technology required to remain competitive in world markets. Growth is driven by consumption demand and public sector deficits that negatively affect the balance of payments, inflation, and real interest rates. Structural reforms have limited priority when immediate macroeconomic disequilibrium occupies center stage. Soon, the only option is another dose of stabilization and austerity. This encounters political resistance and fails to yield durable consequences. At the heart of this sequence is the adverse effect of large transfers of resources to service the external debt. Note that the effect stressed here is not the additional short-term problem of domestic resource reallocation to tradables associated with making the export surplus possible. Because much of that burden falls upon expenditure reduction and import contraction, stabilization is usually costly. For both reasons, analyses of capacity to pay debt service based on projections of imports and exports are inherently excessively optimistic. They ignore short-term inflexibilities and pay no attention to where the counterpart real domestic resources are coming from in the medium term. Large trade surpluses are not feasible when there is a continued reduction in capital formation. There will be neither growth nor assured debt service. 7.2

The Debt/Export Ratio

Figure 7.3 plots the debuexport ratio from 1929 to 1986. The debvexport ratio today is almost as high as it was in the early 1930s. A solution to the debt crisis and the restoration of creditworthiness would be translated into a reduction of the debvexport ratio. The basic debt dynamics model focuses on this ratio, which evolves according to the familiar equation: (1)

d,= b d,- I - n,-ft;

+

+

b = ( 1 i* )/(1 x)

where: d is the ratio of debt to exports; i* is the average interest rate on the external debt; x is the rate of growth of exports; n is the noninterest current

355

BrazilKhapter 7

tI

4

Fig. 7.3

Debt/export ratio, 1929-86

account surplus as a ratio of exports; and f is the non-debt-creating capital inflows as a ratio of exports. The equation highlights three main determinants of the behavior of the debt to export ratio. The first one captures the automatic element of debt accumulation. As long as the international interest rate exceeds the rate of export growth, the current debvexport ratio exceeds the past ratio. The next element is the noninterest current account surplus. Other things being equal, the higher the noninterest current account surplus, the lower the debvexport ratio. The noninterest current account depends in part on domestic policies and in part on the external environment. The ease with which Brazil will be able to generate a noninterest current account surplus-consistent with domestic growth, moderate inflation, and high investment-is the key issue in the solution of its debt problem. Finally, non-debt-creating capital inflows can help reduce the rate of debt accumulation. Direct investment would be beneficial, while repatriation of foreign investment and capital flight would create extra problems. Figure 7.4 shows the historical series for the Brazilian debuexport ratio, as well as three scenarios for the 1987-98 period. The projections start from the initial ratio of 4.9 in 1986. Scenario 1. This scenario is a pessimistic one. We assume that the noninterest current account surpluses and non-debt-creating capital inflows cancel each other out to equal zero throughout the period (n + f = 0). The international interest rate is assumed to be 9 percent throughout the period.

356

Eliana A. Cardoso and Albert Fishlow

Scenario

2.0

75

80

85

90

95

Fig. 7.4 Debuexport ratio, 1975-98

The export growth rate is assumed to behave as erratically as in the past-its behavior between 1975 and 1986 is repeated year after year between 1987 and 1998. The result is dismal. By 1998 the debvexport ratio would be equal to 4.9, as bad as the starting point in 1986. Scenario I f . This scenario is more optimistic than the previous one. The noninterest current account surplus and the non-debt-creating capital inflows are assumed to be as large as 10 percent of exports. We also assume that the export growth rate is kept constant and falls short of the international interest rate by 1 percent. Under these circumstances we would observe a small reduction of the debvexport ratio to 4.3 in 1998. Scenario IZZ. This scenario is overly optimistic. We assume that every year the growth rate of exports would match the international interest rate. The noninterest current account surplus and the non-debt-creating capital inflows are assumed to add up to 20 percent of exports. We observe that 20 percent was the average ratio of noninterest surpluses to exports between 1982 and 1985, when Brazil achieved its highest trade surpluses ever. That means that in any year between 1987 and 1998, Brazilian imports would not exceed 80 percent of the revenue from exports and direct investment. Under these circumstances, the debvexport ratio would be reduced to 2.5, just below its level in 1975. These simulations give some idea of how important noninterest surpluses and nondebt capital inflows are in order to compensate for the effects of large interest payments. The key problem is how to achieve those surpluses without sacrificing imports necessary for investment. Here, finding a reliable mechanism to finance that portion of interest payments not supported by exports will play a central role in solving the debt problem.

357

BraziVChapter 7

7.3 Debt Policy There is scope for both international and internal policy to improve present prospects. Reduced resource transfers require new understandings with creditors; they will not come about through spontaneous new credit flows nor lower international interest rates. The conventional settlement of the moratorium did not resolve the issue. It provided finance for 1987 interest arrears, as well as limited new loans. In 1989 these resources have not even been available because of a lack of an IMF agreement. These packages have been the norm for other debtors, and with the same negative implications for investment and growth. The capital flows required to significantly reduce the resource transfers as a percentage of income are not in prospect. That was a major defect of the initial Baker Plan that called for annual flows of less than $13 billion. There is a broad consensus that such a level was too small by half, and such flows were never realized. Banks refused to make new commitments and continue to remain reluctant. Martin Feldstein’s (1987) optimistic assessment of the debt problem in general, and the Brazilian situation in particular, is flawed for this reason. His calculations, presented in The Economist, for net external borrowing requirements based on a resource transfer of 2.5 percent of product were seen as eminently reasonable. Yet the $4 billion that would be needed annually for the next few years is much larger than capital inflows obtained in any year since 1982. If there had been access to such lending, there would have been no moratorium, a much stronger basis for increased investment, and, very probably, a much more successful outcome to the Cruzado Plan. Brazil is apparently one of the countries capable of absorbing new loans as a means of adjusting to its debt overhang. Up until recently, it has not been afflicted by a large capital flight, as have other debtors, so there is a counterpart stock of real assets corresponding to external liabilities. Moreover, much of the investment was productively applied. As long as international interest rates are not larger than the growth rates of exports and of product, even modest trade surpluses will permit continuing reductions in debvexport and debvproduct ratios. Those more modest surpluses would be possible because of new credits and increases in the debt. In terms of figure 7.4, this would yield a favorable outcome intermediate between scenario I1 and 111. But one should also not ignore the peril of such a strategy. The problem is that new loans are added on top of a 1987 debt export ratio that has already reached a level much higher than projected or desired. Brazil begins with a ratio of 4.9. That translates into considerable short-term vulnerability. Increases in interest rates, higher oil prices, slower industrialized country economic growth, and domestic policy mistakes would all mean larger borrowing requirements than initially projected. But there is a limited willingness, given the level of exposure, to accept the risk of financing such

358

Eliana A. Cardoso and Albert Fishlow

additional needs. Domestic growth rates then suffer. They are not the target to which capital flows accommodate, but rather the principal variable that must adjust. In the last analysis, a policy of growing into debt by new borrowing works when the supply of new capital is highly elastic and the environment relatively stable. Neither accurately describes the present situation. Efforts to introduce greater certainty through interest rate capitalization or by establishing a new interest rate compensatory fund at the IMF have made very little headway. There is continued resistance to assured interest capitalization. Muddling through works by avoiding a crisis at the expense of developing countries' growth. It has involved too little additional credit expansion to permit countries to adjust, but too much for the taste of banks. The implications of another decade of muddling through are nicely captured by figure 7.4's scenario I of temporary, but unsustainable, improvement in the debvexport ratio. The case for debt reduction derives from the possibility of benefitting debtors and creditors. For a long time, developing country loans have been quoted at discounts of more than 50 percent (table 7.1). Banks do not expect the loans to be serviced fully, which is why yields are so high and why they have added to their loan loss reserves. Creditors could do just as well by substituting a smaller debt with more secure payment. It makes little sense for countries to contradict the market expectation when there is no reward. If full service translated into increased capital flows, there would be that incentive. Now, there is none. There is an open invitation to find reasons to reduce debt service. Brazil dared a unilateral moratorium on interest payments in February 1987. It failed because it did not evoke broad political support and because it was seen to emanate from domestic policy errors rather than an incapacity to pay. Conversely, the Brazilian decision to suspend the moratorium and to adopt a less confrontational posture should not be seen as a mistake. In a world where only conventional solutions were possible, it is rational to borrow as much as one can at the lowest spread in order to alleviate the resource transfer. Large short-term credits and interbank deposits are hostages to fortune that limit Brazilian negotiating strength. In 1989 the Brady Plan finally accepted the reality of debt reduction as an integral part of debt renegotiation. Mexico concluded an agreement in which 'zgble 7.1

Discount

The Discount in the Secondary Market for Brazilian Debt (centsper dollar, selling price) July 1985

January 1986

July 1986

January 1987

May 1981

July 1987

September 1987

75

75

73

74

64

51

39

Source: Salomon Brothers.

359

BrazilIAppendix

discounts of 35 percent of principal or comparable reductions of interest rates were accepted. Equally important, the new money option extends over a four-year period. Brazil needs no less. Two issues are central. One is the acceptance by the IMF, the World Bank, and the creditor banks of the simultaneity of a domestic stabilization package and a meaningful debt renegotiation. Our basic message is that the success of an adjustment policy is linked to reduction of the external resource transfer. The second requirement is the availability of adequate industrial country support to underwrite an appropriate package. Like the Baker Plan, the Brady Plan is substantially underfunded. Better Brazilian economic management will be necessary, not only to qualify for debt reduction, but to permit resumption of sustained economic development. Domestic saving will have to increase. Exports will have to receive continued priority. The public sector deficit will have to be controlled. Debt will continue to severely limit the range of the possible, but this is no excuse for not trying to make the best of a difficult situation.

Appendix: Brazilian Statistics List of Tables A. 1 Real Per Capita GDP A.2 Nominal GDP and Per Capita GDP in Dollars A.3 Sectoral Composition of GDP A.4 Output Deviations From Trend A S Investment and Savings Shares in GDP A.6 Population, Labor Force, and Its Sectoral Distribution A.7 Income Distribution A.8 Wholesale Price Index, Inflation, and Monetary Correction A. 9 Nominal Wages A. 10 Real Wages A. 11 Price Indexes A. 12 Export Prices A.13 Import Prices A. 14 Terms of Trade A. 15 Indexes of Subsidies to Exports A. 16 Effective Rates of Protection A. 17 Nominal Exchange Rates A. 18 Real Exchange Rates A. 19 Balance of Payments A.20 Composition of Exports A.21 Real Exports

359

BrazilIAppendix

discounts of 35 percent of principal or comparable reductions of interest rates were accepted. Equally important, the new money option extends over a four-year period. Brazil needs no less. Two issues are central. One is the acceptance by the IMF, the World Bank, and the creditor banks of the simultaneity of a domestic stabilization package and a meaningful debt renegotiation. Our basic message is that the success of an adjustment policy is linked to reduction of the external resource transfer. The second requirement is the availability of adequate industrial country support to underwrite an appropriate package. Like the Baker Plan, the Brady Plan is substantially underfunded. Better Brazilian economic management will be necessary, not only to qualify for debt reduction, but to permit resumption of sustained economic development. Domestic saving will have to increase. Exports will have to receive continued priority. The public sector deficit will have to be controlled. Debt will continue to severely limit the range of the possible, but this is no excuse for not trying to make the best of a difficult situation.

Appendix: Brazilian Statistics List of Tables A. 1 Real Per Capita GDP A.2 Nominal GDP and Per Capita GDP in Dollars A.3 Sectoral Composition of GDP A.4 Output Deviations From Trend A S Investment and Savings Shares in GDP A.6 Population, Labor Force, and Its Sectoral Distribution A.7 Income Distribution A.8 Wholesale Price Index, Inflation, and Monetary Correction A. 9 Nominal Wages A. 10 Real Wages A. 11 Price Indexes A. 12 Export Prices A.13 Import Prices A. 14 Terms of Trade A. 15 Indexes of Subsidies to Exports A. 16 Effective Rates of Protection A. 17 Nominal Exchange Rates A. 18 Real Exchange Rates A. 19 Balance of Payments A.20 Composition of Exports A.21 Real Exports

360

Eliana A. Cardoso and Albert Fishlow

A.22 Composition of Imports A.23 Real Imports A.24 Domestic Production of Oil A.25 Industrial Product and Employment A.26 Monetary Aggregates A.27 Interest Rates A.28 Public Sector Budget Deficit A.29 Taxes and Transfers A.30 Public Sector Debt A.31 External Debt and Reserves A.32 Net Capital Flows A.33 Average Cost of the External Debt A.34 Foreign Debt Indicators A.35 Currency Breakdown of the External Debt Table A . l

Year 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 I965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980

Real and Per Capita GDP, 1950-80 Real GDP (Index 1950 = I )

Population (millions)

Growth Rate GDP

1 .Ooo 1.059

51.9 53.5 55.2 56.9 58.6 60.4 62.2 64.I 66.1 68.1 70.2 72.2 74.3 76.4 78.6 80.9 83.2 85.6 88.0 90.5 93.0 95.1 97.4 99.8 102.3 104.9 107.5 110. I 112.8 115.6 118.5

6.5 5.9 8.7 2.6 10.0 6.8 3.2 8.2 7.1 5.6 9.7 10.3 5.3 1.5 2.9 2.7

1.151

1.181 1.299 1.388 1.432 1.549 1.669 I .762 1.933 2.132 2.245 2.279 2.345 2.408 2.531 2.652 2.896 3.157 3.419 3.806 4.266 4.863 5.301 5.577 6.123 6.405 6.712 7. I96 7.850

5.1

4.8 9.2 9.0 8.3 11.3 12. I 14.0 9.0 5.2 9.8 4.6 4.8

7.2 9. I

Growth Rate GDPKapita 4.2 2.8 5.4 - .5

6.8 3.7 .2 5.0 4.4 2.5 6.4 7.3 2.3 - 1.3

.o - .3

2.2 1.9 6.3 6.0 6.2 8.6 9.3 11.2 6.4 2.7 7.1 2. I 2.3 4.6 6.5

361

BraziUAppendix

Table A . l (continued)

Year

Real GDP (Index 1950 = 1)

Population (millions)

Growth Rate GDP

Growth Rate GDP/Capita

1981 1982 1983 1984 1985 1986

7.583 7.652 7.460 7.886 8.540 9.240

121.5 124.5 127.6 130.7 134.0 137.3

3.4 .9 -2.5 5.7 8.3 8.2

-5.7 -1.5 -4.9 3.1 5.6 5.6

-

Source: Conjuntura EconBmica and Banco Central do Brasil. Nominal and Rer Capita GDP, 1970-85

Table A.2 Year

GDP at Market Prices (in millions of cruzados)

GDP/Capita (in U.S.dollars)

1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986

196.1 261.1 345.0 483.3 708.0 1,009.7 1,625.1 2,486.8 3,621.0 6,041.0 12,639.0 24,737.0 48,148.0 118,195.0 387,968.0 1,406,077.0 3,687,473.0

453 517 594 787 1,015 1,180 1,411 1,591 1,786 1,959 2,028 2,192 2,155 1,604 1,619 1,694 1,977

Source: Banco Central do Brasil.

Sectoral Performance of the Brazilian Economy: Indexes of Real Output

Table A.3

(1949= 100)

Year

Total

Agriculture

Industry

Commerce

Transport & Communication

1950 1951 1952 1953 1954 1955 1956 1957

106.5 112.8 122.6 125.7 138.4 147.9 152.6 165.0

101.5 102.2 111.5 111.7 120.5 129.8 126.7 138.5

111.3 118.4 124.3 135.1 146.8 162.4 173.6 183.5

107.1 117.5 121.8 119.3 132.6 137.9 140.1 153.6

109.5 121.3 130.0 143.3 155.3 161.3 169.5 182.7

(continued)

362

Eliana A. Cardoso and Albert Fishlow

Table A.3 (continued)

Year

Total

Agriculture

Industry

commerce

Transport 8 Communication

1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 I968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985

177.7 187.7 205.9 227.1 238.9 242.7 249.7 256.4 266.1 278.9 310.1 341.1 369.4 413.7 459.6 522.2 572.8 603.7 662.3 700.1 735.1 782.1 838.4 825.0 832.4 805.8 842.1 912.0

141.3 148.8 156.1 167.9 177.1 178.9 181.3 206.3 199.8 211.1 214.0 226.8 230.2 256.2 266.7 276.3 299.0 313.3 322.4 360.4 351.1 368.6 391.8 416.9 406.5 415.4 432.8 470.9

213.2 238.5 261.4 289.2 311.8 312.4 328.5 313.0 349.6 360.0 415.8 460.5 508.9 569.9 644.0 749.0 817.9 866.1 973.5 1,011.5 1,084.3 1,153.7 1,244.9 1,176.4 1,183.5 110.3 1,168.1 1,273.2

164.3 179.7 190.3 203.6 215.5 215.6 218.0 221.6 238.6 249.1 278.3 302.7 327.8 370.8 413.0 466.3 511.6 525.4 563.7 590.2 615.0 648.2 692.3 672.9 678.9 655.2 674.9 733.6

193.9 212.5 249.3 257.6 279.2 301 .0 305.8 311.3 331.9 357.8 389.4 434.8 462.6 509.4 578.1 699.5 821.9 934.5 1,070.0 1,170.6 1,283.0 1,461.4 1,608.9 1,612.2 1,702.4 1,704.2 1,815.0 1,960.2

Source: Conjunhm Ecodmica.

Output Deviations from Trend

Table A.4

Year

Total

Agriculture

Industry

1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965

1.77 .93 2.68 -1.41 1.64 1.69 - 1.76 - .53 .30 - .81 1.87 5.09 3.57 -1.44 -5.17 -9.11

1.20 -2.56 1.70 -2.56 .57 3.56 -3.31 1.15 - 1.29 - .51 - .23 2.61 3.50 .07 -3.05 5.42

-2.31 -3.69 -6.39 -5.62 -4.88 -2.35 -3.25 -5.26 2.17 5.82 7.42 9.97 9.92 2.55 .01 - 12.40

363

BrazilIAppendix

Table A.4 (continued) Year

Total

Agriculture

1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985

- 12.00 - 13.90

-2.23 -1.17 -4.25 -2.89

-9.84 -6.89 -5.51 - .76 3.18 9.36 12.00 10.70 13.40 12.40

Industry -8.89 - 13.50

-6.68 -4.04 - 1.61 2.15 6.81 14.30 15.60 13.70 17.90 14.10

-5.85

.40 - .03 - .94

2.51 2.73 1.15 7.85

10.60

.78

13.50

10.30 10.60 2.44 -3.25 - 13.10 - 15.30 - 13.90

1.20 2.86 4.62 -2.35 -4.63 -4.98 - .99

12.50 12.20 - 1.03 -7.99 -22.60 -24.40 -23.40

Source: Projecoes Economicas, July 1986. Note: The positive sign indicates production above the normal level; the negative sign indicates production below the normal level. Investment and Savings Shares in GDP

Table A.5

Year

Gross Investmenta/ GDP

Nonfactor Current AccountEDP

1970 1971 1972 1973 1914 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986

21.5 21.9 22.1 22.4 23.9 25.0 23.3 21.3 22.6 22.5 22.4 22.3 20.2 15.6 16.5 17.3 19.2

-1.7 -1.7 - 1.2 -5.9 -4.0 -2.4 - .7 -1.2 -1.9 -2.1 - .3 - .7 2.5 5.6 5.3 2.3

- .5

Net Factor PaymenWGDP

Net Foreign SavingdGDP

Domestic SavingsGDP

1.3 2.6 2.6 2.1 6.8 5.4 4.0 2.3 3.5 4.6 5.3 4.4 6.1 3.3

20.2 19.3 19.5 20.3 17.1 19.6 19.3 19.0 19.1 17.9 17.1 17.9 14.1 12.3 16.5 17.2 17.2

- .8

- .9 - .9 - .9 - .9 - 1.4 - 1.6

-1.6 -2.3 -2.7 -3.2 -4.1 - 5.4 -5.8 -5.6 -5.4 -4.3

.o

.I 2.0 -

Source Conjuntura EconBmtca

"Except for 1981, 1982, and 1983, gross investment includes changes in stocks

364

Eliana A. Cardoso and Albert Fishlow Population and Labor Force

Table A.6

~~

Total population (millions) Labor force (millions) Growth rate population Growth rate labor force

~~~~

~

I940

1950

1960

1970

1980

41.24 14.76 1.5

51.94 17.12 2.3 I .5

70.19 22.75 3.1 2.9

93.14 29.56 2.9 2.7

119.07 43.80 2.5

-

Sectoral Distnbution of the Labor Force (%) Agriculture Industry Service

1960

I970

1980

54.0 12.9 33.1

44.3 17.9 37.8

29.9 24.4 45.7

Source: Fun&qao IBGE.

Table A.7

Income Distribution, 1970 and 1980 Percent of Total Income Percent of Population Poorest 10 percent 20 30

40 50 60 Richest 30 20 10 5 1

I980

.9 2.7

.8 2.4

5.5

5.0

9.3 14.1 20.3 71.7 61.0 45.2 32.9 15.2

13.1 19.0 73.3 62.7 47.0 34.6 16.4

8.5

Income Classes

Percent of Labor Force

up to !A Minimum Wage >%to1 > I to2

28.4 32.2 21.7 12.7 3.3 1.3 .4

>2to5

> 5 to 10 > 10 to 20 more than 20 Source: Rossi (1986).

1970

12.5 20.8 31.1 23.6 7.2 3.2 1.6

4.0

BraziVAppendix

365 Table A.8

Wholesale Price Index, Inflation, and Monetary Correction, 1960-85 Wholesale Price Index (1977 = 100, year average)

Year

1.117 1.959 3.732 5.852 8.070 10.360 12.870 15.550 18.630 22.610 26.620 30.700 39.640

1962 1963 1964

I965 1966 I967 I968 I969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986

ORTN’ IChange (average over average of previous year)

Inflation During the Year (I) 51.6 75.4 90.5

56.8 37.9 28.4 24.2 20.8 19.8 21.4 17.7 15.5 35.4 27.8 40.4 40.6 38.9 55.4

50.660 71.130

100.OOO 138.900 215.800 451.500 961.900 1,866.100 4,942.900 16,453.500 53,202.400 128,909.415

63.0 39.2 23.2 25.0 18.5 19.6 22.7 19.8 13.7 21.3 29.7 30.1 36.1 32.0 40.1 50.0 73.3 93.1

109.2

113.0 94.0 164.9 232.9 223.3 142.3

131.1

~-

194.4 220.4 130.0 ~

Source Conjuntura EconBmrca

Wbngaqks Reajustaveis do Tesouro Nacional

rable A.9

Nominal Wages (cruzaddmonth, and indexes)

Minimum Wage

Industry Median (2)

Industry Average (3)

,08400

,01670 ,02700 ,04720 .07910 . I lo00

.I0500

.I5800

,12900 ,15600 .I8720 .22560 ,25440 ,29760

,17200 .21m .26200 ,31700 ,46900 ,65700

,01620 ,03440 ,06580 ,09970 ,13820 ,18490 ,23710 .3oooo .37800 .46740 ,58980 ,72690

Year

(1)

1962 1963 I964 1965 1966 1967 1968 1969 1970 1971 1972 1973

,01344 .02100 ,04200

(continue4

.06600

Industry Average Index (4)

Rural SBo hula

(3

Rural Wage CEA Index (6)

Industry ADBIB (7)

16.1 22.5

,00135 ,00167

Industry

,00669 ,01086 ,02292 ,04110 ,05340 ,07470 ,09870 .I 1910 ,15420 ,19350 ,25140 ,34050

RESP Index (8)

366

Eliana A. Cardoso and Albert Fishlow

Table A.9 (continued)

Year

Minimum Wage (1)

Industry Median (2)

Industry Average (3)

1974 1975 I976 1977 1918 1979 1980 1981 1982 1983 1984 1985

,35840 ,39360 ,68960 ,99360 1.40880 2.14280 4.01720 8.15000 16.20800 34.76400 95.38800 322.08000

,87900 1.16300

,95040 1.35240

Industry Average Index (4)

4.5 6.6 10.0

Rural Sio Paul0 (5)

Rural Wage CEA Index (6)

,47550 ,61950 ,83550

36.3 49.1 68.3 100.0 137.0 217.0 445.0 902.0

15.1

23.8 45.6 100.0 215.6 478.3 1,384.4 4,800.0

1644.0

3707.0 11201.0

Indusq HESP Index (8)

Industry ADBIB (7)

.00217 ,00307 ,00461 ,00689 .01005 ,01558

,02994 ,06679 ,14820 .33230 .91440 3.22210

27.1 43. 64.1

loo.( 161.' 314.: 620.: 1,248.' 2,534.1 7,371.1 27,205.1

Sources: Columns ( I ) , (6), (7), and (8). Conjunturu EconBrnicu. Column (2). Bacha (1981). Column (41,

Fundaqao IBGE. Column

( 3 ,Bacha (1981).

Column ( I ) , weighted average during the year of the highest minimum wages. Lcgislation gradually united minimum wages paid in the different regions of the country.

Nores.

Table A.10

Real Wages

Year

Minimum Wage

1962 1963 1964 1965 I966 1967 I968 1969 1970 1971 1972 1973 I974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985

154.87 139.97 144.86 145.17 133.97 130.45 129.60 129.13 129.33 128.42 123.00 124.76 116.37 100.00 124.79 127.89 130.56 127.82 114.52 109.06 111.79 90.53 74.62 77.93

Industry Median

65.1 60.0 55.1 58.9 59.4 66.4 58.2 60.5

61.3 61.1 76.8 93.3 96.6 100.0

Industry Average

54.33 65.77 66.04 63.82 64. 15 66.85 69.01 72.27 76.00 77.43 82.99 88.69 89.81 100.00

Source: Table A.9 deflated by wholesale prices. "FundaGHo Instituto Brasileiro de Geografia e Estatistica

Industry FIBGE"

Rural SBo Paul0

Rural (CEA)

Industry (F'IESP)

48.97 45.33 50.22 57.43 54.11 58.96 62.71 62.63

61.78

100.00 104.49 112.36 122.47 123.59 113.48 116.85 130.34 108.99 94.38 101.12

69.98 77.23 90.69 98.09 l00.00 96.05

60.59 73.13 91.57 96.96 96.06 100.00 98.63 100.56 48.56 93.77 88.98 75.00 68.08

72.60 83.70 89.10 100.00 107.70 116.70 122.20 125.10 109.70 112.90 138.80

367

BrazilIAppendix

lsble A . l l

Price Indexes Construction Costs

Year

CPI (R.J.) (1977= 100)

1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 I972 1973 1974 1975 1976 I977 1978 1979 1980 198I 1982 1983 1984 1985

.3816 ,4935 ,6575 ,9962 1.7010 3.2560 5.4010 7.6300 9.9600 12.2000 15.4000 18.8800 22.6900 26.4500 29.8000 38.0300 49.0400 69.5700 100.oooO 138.7000 211.8000 387.2000 795.9000 1,575.7000 3,812.9000 11,314.5000 36,994.7000

Industrial Goods (1977= 100)

Agricultural Goods (1977=100)

.4799 S908 ,8389 1.2200 2.2300 4.0940 6.6040 8.7400 10.9780 14.3150 17.2600 20.1650 23.7000 27.5000 31.6100 40.m 52.8300 71.8500

,477 .643 1.029 1.702 3.397 4.837 6.862 8.544 10.024 12.180 14.390 18.000 22.020 26.260 33.890 42.060 66.890 100.000 142.400 222.500 472.700 966.400 1,691.000 5,602.600 20,070.200 67,693.700

100.oooO 135.3000 210.5000 428.8000 898.6000 1,804.2000 4,379.4000 14,196.5000 46,169.4OOO

R.J. (Mar.86 = 100)

,0186 ,0217 ,0254 ,0298 ,0355 ,0460 ,0580 ,0910 . I150 ,1810 ,2950

.4690 ,9390 1.8600 4.0830 12.0500 42.6500

Brazil (SINAPI) (Dec.73 = 100)

100.0 149.6 189.5 278.2 373.6 504.4 832.5 1,818.7 3,102.4 6,550.5 10,421.8 29,062.3 95,626.9

Source: Conjuntura E c o n h i c a .

Indexes of Export Prices (dollar prices, 1977 = 100)

lsble A.12

Total Year

Total

Except Coffee

Coffee

Minerals

Industrialized Goods

Primary Goods

1960 1961 1962 1963 1964 I965 1966 1967 1968 1969 1970 1971 1972 1973 1974

31.04 32.60 28.44 28.34 33.93 34.26 32.83 32.76 32.33 33.27 37.59 36.28 40.98 56.39 71.05

37.84 41.96 33.64 34.50 36.30 35.18 37.39 40.09 38.59 39.71 40.82 43.67 47.76 66.94 86.12

15.65 15.46 14.49 14.17 18.80 19.31 16.76 15.56 15.49 15.94 21.69 16.59 21.02 25.81 27.98

68.5 68.0 65.4 58.8 50.4 51.6 52.1 49.1 45.6 45.1 48.1 49.0 47.8 51.0 61.1

28.2 29.1 29.5 33.3 36.6 32.9 34.3 33.6 34.6 36.3 38.3 42.5 47.5 67.1 94.3

23.3 23.2 21.6 21.5 26.2 27.2 24.9 23.2 23.8 25.8 30.5 26.3 31.4 43.0 46.3

(continued)

368

EIiana A. Cardoso and Albert Fishlow

Table A.12 (continued)

Year

Total

Total Except Coffee

Coffee

Minerals

1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985

71.05 81.95 100.00 92.00 101.00 107.00 100.70 94.60 89.50 91.30 86.10

87.76 87.76 100.00 98.00 107.80 116.60 115.10 105.90 98.90 100.00 94.50

24.30 60.09 100.00 69.70 74.80 70.40 40.80 46.50 49.70 55.20 50.90

83.2 96.2 100.0 91.2 95.0 109.3 112.3

121.0 112.9 100.0 99.8

Industrialized Goods

Primary Goods

93.5 89.3 100.0 100.0 112.0 122.1 120.2 109.2 101.4 103.1 98.0

42.7 69.5 100.0 75.0 78.9 72. I 49.2 50.4 52.9 60.4 53.2

Source: Conjuntura EconGmica.

Table A.13

Indexes of Import Prices (dollar prices, 1977 = 100)

Year

Total

Oil

Wheat

1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1971 1978 1979 1980 1981 1982 1983 1984 1985

38.26 38.91 39.23 40.16 38.91 39.03 40.28 41.17 42.38 41.57 42.37 44.07 47.03 58.90 90.68 93.64 96.19 100.00 107.00 128.00 163.90 182.10 176.10 166.80 158.40 149.20

6.67 8.86 11.69 12.18 12.70 12.03 13.32 12.39 14.70 15.71 16.08 20.26 22.19 27.81 92.60 93.57 96.14 100.00 101.00 135.20 226.30 270.00 260.50 235.30 229.40 222.40

63.22 66.21 66.21 68.16 72.05 65.56 63.48 66.21 62.57 61.99 58.48 65.50 70.76 114.60 196.50 152.60 143.30 100.00 124.60 154.80 198.60 198.30 181.20 172.90 155.20 146.90

Total Imports Except Oil & Wheat

52.2116 5 1.4963 5 1.9082 53.6835 54.3170 54.4690 5 1.6203 51.4912 52.1658 53.5533 5 1.6897 53.9123 62.9 156 83.6086 87.7054 94.6253 100.oooO 108.9700 112.4820 133.6160 139.8900 136.2900 136.9700

Source: Conjunrura EconGmica. Column (4): FundaGBo Getljlio Vargas, unpublished data

369

BrazillAppendix

Table A.14

Terms of Trade, 1960-85

Year 1960 1961 1962 1963 1964

1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985

Terms of Trade 81 84 72

70 87 88 81 79 83 80 89 82 87 95 78 76 85 100 86 79 65 55 54 53 58 57

Source: Banco Central do Brasil.

Indexes of Subsidies Given to Exports

lsble A.15

(continued)

Year

Fiscal Subsidies

Fiscal & Credit Subsidies

1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977

1 .Ooo 1.004

I .Ooo 1.004 1.005 1.050 1.213 1.262 1.340 1.438 1.475 1.485 1.505 1.525 1.623 1.740 1.715

1.050 1.050 1.219 1.265 1.316 1.389 1.413 1.423 1.430 1.470 1.491 1.493 1.506

370

Eliana A. Cardoso and Albert Fishlow

Table A.15 (continued) ~

~

Year

Fiscal Subsidies

Fiscal & Credit Subsidies

1978 1979 1980

1.487 1.464 1.381 1.442 1.400 I .400

1.730 1.670 1.585 1.655 1.666 1.628

1981 1982 1983

Sources: Cardoso (1980), Musalem (1981). and Braga and Markwald (1983).

Table A.16a

Effective Rates of Protection by Sector: 1958-67 1958

Agriculture Vegetable products Animal products Mining Nonmetallic minerals Metals Machinery Electrical & communication equipment Transportation equipment Wood products Furniture Paper products Rubber products Leather products Chemicals Pharmaceuticals Perfumery & cosmetics Plastic products Textiles Clothing & footwear Food products Beverages Tobacco products Printing & publishing Miscellaneous industry All sectors (wtd. by 1959 value added)

1963

1966

1967=

1967b

- 47

- 15

- 13

- 14

- 14

24 -5 73 61 22 83 82 I38 22 I 86 139 248 56 17 279 281 239 264 502 171

12 34 130 124 68 169 147 176 367 I69 22 I 405

16 24 72 63 30 112 103 120 25 1 91 158 174

N.A. 9 48 33 31 57 81

146

56

60

10 28 1 332 232 32 1 423 183 299 142 95

18 13 45 35 32 61 84 81 90 43 126 127 29 10 121 133 162 107 252 104 114 4 47

273 139 88

453 489 298 481 6,718 243 469 305 175

44 92 42 I82 84 20 10 74 I I7 88 I54 71 76 79 8 45

29.7

75.2

43.7

24.2

13.9

Industry

106.1

183.5

108.0

63.3

47.8

Consumer goods

242.0

359.9

230.1

122.2

65.7

64.9 53.0

130.6 112.5

68.0 69.1

40.3 55.7

38.5 52.4

Intermediates Capital goods Source: Fishlow (1975). "Based on 1959 input-output table. bBased on 1971 input-output table.

371

BrazillAppendix

Table A.16b

Effective Rates of Protection by Sector: 1%-73 June 1966

Agriculture Vegetable products Animal products Mining Nonmetallic minerals Metals Machinery Electrical & communication equipment Transportation equipment wood products Furniture Paper products Rubber products Leather products Chemicals Pharmaceuticals Perfumety & cosmetics Plastic products Textiles Clothing & footwear Food products Beverages Tobacco products Printing & publishing Miscellaneous industry Average industry

April 1967

November 1973 25

8 17 13 39 36 32 97 75 25 124 59 1 I6 85 42

35 164 25 86 58 41 215 151

45 239 1 I8 136 1 I7 59

39

35

8,480 183 379 337 87 447 313 142 128

3,670 58 162 142

181

76

14 46 35 32 61 34 68 74 50 66

81 19 17 46 41 I18 293 83 114 83

40 173 124 67 72

30 37 47

Sources: 1966 and 1967 estimates from Bergsman (1970). 1973 estimates from Tyler (1976)

Table A.16c

Effective Proteetion Estimates a t the Two-Digit Level, 1980-81 (in percentages)

Industry Mining Nonmetallic minerals Metals Machinery Electrical & communication equipment Transportation equipment Lumber & wood products Furniture Paper products Rubber products Leather products Chemicals Pharmaceuticals Perfumery & cosmetics Plastic products Textiles Clothing & footwear (continued)

Effective Protection

Net Effective Protection

-4.3

- 18.0

- 19.6

-31.1 15.0

34.2 77.0 111.9 -9.6 17.7 52.7 - 18.5 -21.4 13.9 86.4 116.3 91.6 28.3 36.7 46.7

51.7 81.6 -22.5 0.9 30.8 - 30.2 -32.6 - 2.4 59.7 85.3 64.2 9.9 17.1 25.7

372

Eliana A. Cardoso and Albert Fishlow

Table A.16c (continued) Industry

Effective Protection

Food products Beverages Tobacco Printing & publishing Miscellaneous industry

26.1 -1.1 5.7 31.9 171.7

Averages" Primary agricultureb Manufacturing Capital goods Intermediate goods Consumer goods

-8.2 43.6 59.6 42.0 35.7

Net Effective Protection 8.1

- 15.3 -9.4 13.0 132.8 -21.3 23.1 36.8 21.7 16.3

Source: World Bank (1983). "Value-added weights of 1979 are used for aggregating from the four-digit to the two-digit level and for computing the more aggregated means. bIncludes forestry & fishing, agriculture, and livestock & poultry.

Table A.17

Nominal Exchange Rates, Cruzeiros/U.S. Dollars (yearly average, sale rate) Year

Official Rate

Black Rate

~~

1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1917 1978 1979 1980 1981 1982 1983 1984 1985 1986 June (cruzados)

,190 ,270 ,390 ,580 1.250 1.900 2.220 2.660 3.396 4.074 5.594 5.288 5.934 6.126 6.790 8.126 10.670 14. I28 18.063 26.870 52.700 93.100 179.500 577.040 1,848.030 6,205.250 13.840

,520 ,890 1.530 1.930

2.220 2.880 3.710 4.440 5.010 5.880 6.570 6.550 7.450 9.375 13.525 16.917 21.717 32.592 58.417 108.675 27 1.250 926.250 2,214.200

8,455.000

Sources: Conjuntura EconBmica, Pechman (1984), and Projecoes Economicas.

21.000

373

BrazilIAppendix Real Exchange Rates (1977 = 1)

Table A.18 Year

(1)

1965 1966 1967 1968 I969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985

,898 1.034 1.124

(2)

(3)

(4)

1.083 1.050 1.059 1.059 I .027 .952 ,929 ,925 ,969 I .Ooo ,896 ,819 ,766 ,929 ,981 .812 ,849 337

1.218 1.168 1.126 1.092 1.067 1.034 1.025 1.000 .916 ,815 ,714 ,866 ,941 ,723 ,723 ,714

.937 ,936 ,932 ,942 ,977 ,978 ,955 1.101 I .ooo 1.105 1.153 ,951 ,909 .896 ,811

,738 303

1.328 1.311 1.227 1.177 1.176 1.180 1.199 1.183

.ooo

1

1.105 1.153 ,863 ,845 .956 ,634 ,522 ,549

(5)

(6)

1.423 1.620 1.732 1.718

2.561 2.659 2.463 2.366 2.122 1.95 1 1.732 1.610 1.244 1.024 1.049 1.049 I .Ooo 1.049 1.024 1.024 1.195 1.341 1.122 N.A. N.A.

1.648

1.620 1.479 1.380 1.085 ,901 ,986 1.056 I .Ooo 1.056 ,986 ,944 1.127 1.296 1.056 1.056 1.040

Sources: Column (1): domestic wholesale pricedweigbted average of wholesale prices of main trade partners (IMF, International Financial Statistics). Column (2): effective exchange rate (Morgan Guaranty TNS~Co.). Column (3): construction prices/industrial prices (Conjuntura Econbmica). Column (4): construction prices/agricultural prices (Conjuntura Econbmica). Column ( 5 ) : industrial pricesiprices of manufactured exports in dollars multiplied by the exchange rate (Conjunruru EconBmica). Column (6): same as (3,but including subsidies.

Table A.19

Balance of Payments (in millions of U.S. dollars) 1950

Trade balance Exports Imports Net services Capital services Other Unilateral transfers Current account Capital account Net foreign direct investment Loans and financial flows Amortization Other Errors and omissions Balance of payments

1951

1952

425 68 1,359 1.771 -934 -1,703 - 283 - 469 - 74 -90 - 209 - 379 __ -2 __ -2 140 - 403 - 65 - I1 3

-4

28

38

- 85

- 27

-11 __

- 23 -

52

- 18 __

123 -291

1953

- 286

1.416 -1,702 - 336 - 36 - 300 -2 - 624 35 -

424 1.540

-1,116 - 355 - 127 - 228 - 14

55 59 -

9 35

24 __

--26 -615

148 1,558 -1,410 -338 - 97 -241 -5 - 195 - 18 -

-

22

- 33 3

1955

1954

11

44

109

-46

- 134

9

- 98 -

16

-4 -

1

0

- 203

320 1,419 -1,099 - 308 - 78 - 230 - 10 2 3 -

1956 437 1.483 -1,046 - 369 -91 - 278 -11 51 151 -

43

89 23 I

84

- 140 16 __

12 17

- 187

18 -14

194

1957 107 1,392 -1,285 - 358 - 93 - 265 - 13 - 264 255 143

1958 65 1,244 -1,179 - 309 - 89 - 220 __ -4 - 248 I84 110

319

373

- 242

- 324

35 -

25 - 189 -

- 171 __

- 180

- 253

1959 12 1,282 -1,210 - 373 -116 - 257 - 10 -311 182 124 439 -377 -4 - 25 - 154

-

1960 - 23 1,270 -1,293 -459 - 155 - 304 4 - 478 58 -

99 348 -417 28 10 -410

1961 113 1,405 -1,292 - 350 - 145 - 205

15

- 222 288 -

I08 579 - 327 - 72 -

49 1 I5

1962 Trade balance Exports Imports Net services Capital services Other Unilateral transfers Current account Capital account Net foreign direct investment Loans and financial

flows Amortization Other Errors and omissions Balance of payments (continued)

-

89 1,215 -1,304 - 339 - 136 - 203 39 - 389 181

69 325 -310 97 - 138 - 346

-

1963

1964

1965

1966

I967

1968

I969

1970

1971

112 1,406 -1,294 269 - 87 182 43 -114

344 1,430 -1,086 -359 -131 -228 55

655 1,596 -941 -362 -174 -188 75 368 -6

438 1,741 -1,303 - 463 - 197 - 266 79 54 124

213 1,654 -1,441 - 527 - 251 - 270 71 - 237 27 -

26 1,881 -1,855 - 556 - 228 -328 22 - 508 54 I

318 2,311 -1,993 - 630 -261 - 369 31 -281 871

232 2,739 -2,507 -815 - 353 - 462 21 - 562 1,015 -

-341 2,904 -3,245 -980 - 420 - 560 14 - 1,307 1,846

3,991 -4,235 -1,250 - 520 -730 5 - 1,489 3,942

168

318

-

30 250

- 364

30 - 76 -

- 244

40 82

28

22 1 -217 110 -218 - 96

-

70 363 -304 -135 -31 33 1

-

-

-

74

76

508

530 -444 - 135 - 35 - 245

- 350

61 583

177 1201

132 1,510

- 484

- 493

- 672

32

549

545

- 244

- - - 2,523 -850 5 -9 530

- 108 381 - 14 45 - - 25 -1 -41 92 - -

153

1972

4,320

1973 7 6,199 -6,192 -1,722 - 712 -1,010 21 - 1,688 3,512

940 4,299

- 1,202

- 1,673

2,439

2,179

56 - 54 436 355 -

Table A.19 (continued) 1974

1975

1976

1977

1978

1979

1980

1981

1982 ~~

Trade balance Exports Imports Net services Capital services Other Unilateral transfers Current account Capital account Net foreign direct investment Loans and financial flows Amortization Other Errors and omissions Balance of payments

- 4,690 ~

-3,540 8,670 - 12,210 - 3,162 - 1,733 - 1,429

-2,147 10,130 - 12,277 -3,919 -2,422 - 1,497

- 1,024

L

A

- 6,700

- 6,062

6,189

6,867

97 12,120 - 12,023 -4,134 - 2.559 - 1,575 0 - 4,037 5,269

887

892

1,010

810

2,047

7,355

7,483 -2,172 - 14 - 439 - 950

7,921 -2,888 824 387 1,192

8,727 -4,060 - 208 - 602 630

15,388 -5,324 - 220 - 639 4,262

7,951 12,641 - 2,433 -901 - 1,532 1 -7,122 6,254

- 1,920

- 68 - 68

- 936

Source: Banco Central do Brasil

-

1983

1984

1985

~~

12,659 - 13,683 -6.037 -4,232 - 1.805 71 -6,990 11,891

-2,840 15,244 18,084 - 7,920 - 5,542 -2,378 18 - 10,742 7,717

-2,823 20,132 - 22,955 -10,152 -7,032 -3,120 168 - 12,807 9,679

1,202 23,293 -22,091 - 13,135 - 10,272 -2,863 199 -11,734 12,773

780 20,175 - 19,395 - 17,083 - 13,494 - 3,589 -8 -16,311 7,851

6,470 21,899 - 15,429 - 13,415 -11,008 - 2,407 108 - 6,837 2,103

13,090 27,005 - 13,915 - 13,215 -11,471 - 1,744 171 45 253

12,471 25,639 - 13,168 - 12,893 -11,191 - 1,702 155 - 268 -2,729

2,212

1.532

2,326

2,547

1.359

1,549

1,253

11,288

13,315 -5,010 - 158 - 343 -3,471

16,783 -6,242

12,451 -6,952 - 195 - 368 -8,828

7,778 -6,863 - 171 - 670 - 5,405

8,768

5,614 -8,890 - 706 - 529 - 3,526

~

- 6,385

602 - 130

-3,155

- 94

-414 625

- 6,468 - 3,596

403 700

377

BrazilIAppendix Composition of Exports, 1960-85 (value of exports in millions of U.S. dollars)

Table A.20

Year

Total Exports

1960 1961 1962 1963 I964 1965 1966 1967 I968 1969 1970 1971 1972 1973 1974 I975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985

1,270 1,405 1,215 1,406 1,430 1,596 1,741 1,654 1,881 2,311 2,739 2,904 3,991 6,199 7,951 8,670 10,130 12,120 12,651 15,244 20,132 23,293 20,175 2 1,899 27,005 25,639

Industrialized Products”

Coffee

Minerals

Primary Goodsb

209.0 293.7 287.4 319.5 366.8 482.3 647. I 865.7 1,198.7 2,002.8 2,974.9 3,232.9 3,619.6 4,649.4 6,179.0 8,166.4 10,964.0 13,646.0 11,361.1 13,415.6 18,215.1 17,101. I

712.7 710.4 642.7 748.5 759.9 707.4 773.5 733.0 797.3 845.7 981.8 822.2 1,057.I 1,344.2 980.4 934.2 2,398.2 2,625.0 2,294.7 2,326.2 2,773.0 1,761.0 2,113.1 2,347.0 2,856.0 2,632.5

86.94 94.57 99.23 97.45 101.62 135.66 133.09 123.55 136.16 177.55 262.10 295.26 287.81 411.86 665.92 ,088.61 .I0934 971.00 ,086.72 ,367.77 ,701.42 ,980.87 2,005.98 1,681.76 1,674.40 1,795.70

470.36 600.03 473.07 560.05 359.48 459.24 547.01 477.95 580.74 805.45 848.00 920.84 1,447.39 2,440.14 3,329.78 3,4 14.29 3,002.36 3,874.60 3,090.58 3,383.63 4,963.58 5,905.13 4,694.82 4,454.64 4,259.50 4,110.70

“Includes special transactions. 1960, 1961, 1962, and 1963, includes manufactures

Table A.21

Year 1960 1961 1962 1963 1964 1965 I966 I967 1968 1969 1970 1971 1972 1973 1974 1975 (conrinued,

Real Exports (in 1977 U.S. dollars) Industrial Goods

Coffee

Minerals

Primary Goods

Total

571.04 892.71 837.90 950.89 1,060.12 1,328.65 1,689.56 2,036.94 2,523.58 2,984.80 3,154.72 3,457.65

4,553.99 4,695.08 4,435.47 5,282.29 4,042.02 3,663.39 4,615.16 4,710.80 5,147.19 5,305.52 4,526.51 4,956.00 5,029.02 5,208.06 6,503.93 3,844.44

126.92 139.07 151.73 165.73 201.63 262.91 255.45 25 I .63 298.60 393.68 544.91 602.57 602.11 807.57 1,089.89 1,308.43

2,018.71 2,586.34 2,190.14 2,604.88 1,372.06 688.38 2,196.83 2,060. I3 2,440.08 3,121.90 2,780.33 3 5 0 I .29 4,609.52 5,674.74 7,191.75 7,996.00

6,699.62 7,420.49 6,777.34 8,052.90 6,186.75 6,507.39 7,905.34 7,973.45 8,945.99 10,149.75 9,541.31 11,096.80 12,764.23 14,675.17 14,940.29 16,606.52

378

Eliana A. Cardoso and Albert Fishlow

Table A 2 1 (continued) ~

~~

~

Year

lndustnal G w d s

Coffee

Minerals

Pnrnary Goods

Total

1976 1977 1978 1979 1980 1981 1982 1983 1984

4,053.30 4,649.40 6,179.00 7,291.43 8,758.39 11,352.75 10,403.94 13,230.37 17,667.41

3,991.01 2,625.00 3,292.25 3,109.89 3,938.92 4,316.18 4,544.30 4,722.33 5,173.91

1,153.68 971 .OO 1,191.58 1,439.76 1,556.65 1,763.91 1,657.83 1,489.60 1,674.40

4,319.94 3,874.60 4,120.77 4,288.50 6,884.30 12,002.30 9,315.12 8,420.87 7,052.15

13,517.93 12,120.00 14,783.60 16,129.58 21,138.26 29.435. I4 25,921. I9 27,863.17 31,567.87

Source: Banco Central do B r a d and Conjuntura Econdmica

Table A.22

Composition of Imports, 1960-85 (import values in millions of US dollars, f.0.b.) Year

Total Imports

1960 1961 1962 1963 1964 1965 1966 I967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 I979 1980 1981 1982 1983 1984 1985

1,293 1,292 1,304 1,294 1.086 94 I 1,303 1,441 1,855 1,993 2,507 3,247 4,232 6,192 12,642 12,210 12,383 12,023 13,683 18,084 22,955 22,091 19,395 15,429 13.916 13.168

Source: Banco Central do Brasil.

Oil & Derivatives

Wheat

194.6 190.6 191.9 192.8 176.6 154.0 165.8 153.6 204.0 203.8 236.1 326.9 409.2

122 117 139 139 176 1 I4 142 153 168 135

710.8

334 468 325

2,840.1 2,875.4 3,612.5 3,813.9 4,195.8 6,434.4 9,844.3 I I ,005.8 10,120.2 8,179.0 6,866.6 5,693.6

104

107

122

504 260 54 1 545 890 832 762 727 755 762

379

BraziUAppendix

Table A.23

Real Imports (in 1977 U.S. dollars)

Year

Oil

Wheat

Others

Total

1961 1962 1963 1964 I965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985

2,151.2 1,641.6 1,582.9 1,390.6 1,280. I 1,244.7 1,239.7 1,387.8 1,297.3 I ,468.3 1,613.5 1,844.1 2,555.9 3,067.1 3,073.0 3,757 .5 4,081.0 4,438.6 5,009.6 4,507.3 4,199.6 4,014.2 3,657.9 3,213.2 2,777.0

176.7 209.9 203.9 244.3 173.9 223.7 231.1 268.5 217.8 177.8 163.4 172.4 291.4 238.2 213.0 351.7 279.0 563.4 635.0 624.9 542.6 468.0 523.4 538.0 497.6

1,885.40 1,889.65 1,853.66 1,366.16 1,239.02 1,827.09 2,197.59 2,880.10 3,171.04 4,046.25 5,442.28 6,864.48 8,181.12 I 1,163.80 10,272.57 8,736.04 7,949.10 8,209.78 9,872.34 9,146.14 7,329.47 6,246.09 4,762.36 n.a. n.a.

4,2 13.30 3,741.15 3,640.46 3,001.06 2,693.02 3,295.49 3,668.39 4,536.40 4,686.14 5,692.35 7,219.18 8,880.98 11,028.42 14,469.10 13,558.57 12,845.24 12,023.0 12,787.9 14,128.1 14.005.5 12.128.0 11,014.2 9,249.4 8,798.6 8,839.8

Source: Conjuntura EconBmica n.a. = not available.

Table A.24

Domestic Production of Oil (in thousands of tons) Year

Oil

Year

Oil

1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971

3,181 3,991 4,692 4,511 4,891 4,602 4,756 5,827 7,315 8,062 8,621 8.209 8,389

1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984

8,233 8,372 8,727 8,459 8,225 8,166 8,166 8,412 9,199 10,822 12,998 16,676 23,195

Source: Banco Central do Brasil.

Table A.25

Industrial Product and Employment Real Industrial Product

Year

1965 1966 I967 I968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980

Industrial Employment (FIESP)

Hours Worked (FIESP)

63.0 69.2 71.3 80.8 90.6 100.0

112.0 126.5 147.0 160.6 170.1 191.2 198.7 213.1 226.7 244.6 231.2 232.5 216.8 229.6 250.3

1981

1982 1983 1984 1985

88.4 94.6 93.9 100.0 103.6 108.3 95.0 89.3 80.4 83.6 95.5

89.7 95.5

97.3 100.0 103.5 107.3 99.9 95.0 87.6 87.4 98.4

Source: Conjunrura EconBrnica.

Table A 2 6

Monetary Aggregates (in millions of cruzados) Yea$

I960 1961 1962 1963 1964 1965 1966 1967 1968 I969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985

Base

,361 ,579 .952 1.619 2.890 4.990 6.142 8.035 11.410 14.690 17.160 23.390 27.700 40.800 54.200 73.900 110.800 166.900 241.800 445.900 699.500 1,188.100 2,225.400 4,367.400 15,015.100 45,468.800

Source: Banco Central do Brasil aEnd of year.

MI ,6519 .9940 1.6310 2.6850 4.8750 8.7500 9.9590 14.5100 20.1700 26.7400 33.6400 44.5000 61.6000 90.5000 120.8000 172.4000 236.5000 325.2000 462.7000 803.1000 I ,367.oooO 2,558.5000 4,222.oooO 8,231.9000 24,985.oooO 106,975.oooO

M4 .74 1.10

1.78 2.91 5.27 9.78 12.13 18.56 26.32 36.83 52.00 73.00 113.00 167.00 235.00 370.00 575.00 857.00 1,312.00 2,121.OO 3,662.00 8,820.00 1 8,362.OO 45,998.00 182.040.00 732,442.00

Brazil/Appendix

381

Table A.27

Interest Rates (CDB: one month, yield per year) Gross Nominal

Year

(1)

1974

,1610 ,1685 ,1693 ,1829 ,2103 ,1971 ,2209 ,2901 ,3092 ,3607 ,4144 ,4410 ,4618 ,3409 ,4242 ,3622 ,4261 ,4412 ,3868 ,5223 ,5096 ,4098 ,3529 ,3255 ,3783 ,2669 ,4338 ,6297 ,7845 ,7750 ,8910 1.0220 1.0140 1.0109 1.1814 1.4755 1.4408 2.3881 2.0649 1.9029 2.5779 2.1133 2.6356 2.7063 3.3391 2.9073 2.0640 2.7062 3.4295 ,1656

1975

1976

1977

1978

1979

1980

1981

1982

1983

1984

1985

1986

Source: Banco Icatu.

Net of Taxes Nominal (2) ,1610 ,1685 ,1693 ,1829 ,2103 ,1971 .2209 ,2901 .3092 ,3607 .4144 .4410

,4618 ,3409 ,4242 ,3622 ,4261 ,3804 ,4276 ,4874 ,5096 ,4098 ,3529 ,3255 ,3783 ,2669 .4338 .6297 .7845 ,7750 3910 1.0220 1.0140 1.0109 1.1814 1.4755 1.4408 2.3881 1.9270 1.7867 2.4078 1.9809 2.4613 2.5261 3.1052 2.7099 1.9353 2.5249 3.1647 ,1585

Real Net of Taxes (3) -.1814 - ,2179 - ,0127 - ,0406 - ,0515 .0556 - ,0814 - .01 I5 - ,1458 - ,0740 - .0944 ,1134 - ,0519 .0713 ,1626 .0163 ~

~

- .0105

- .0756

,0380 .1308 .0928 - ,0337 - ,3247 - ,3426 - ,2864 - ,3688 - ,3453 - ,2701 - ,2445 - ,0522 - ,0142 ,2008 ~

- ,0840 - ,0481

,1977 ,3297 - .0839 .1947 - .2454 - .0542 .m37 .0576 .0525 ,0078 .Oslo ,5436 - ,0935 -.1115 - ,1432 ,1449

382

Eliana A. Cardoso and Albert Fishlow Different Measures of the Budget Deficit as a Share of GDP

Table A.28

Year

Increase in Total DebVGDP

Deficit Corrected for InflationGDP ~~~~~~~

1982 1983 1984 1985

25.9 60.5 60.7 65.6

~

FGV MeasureYGDP ~~

8.4 15.2 4.6 6.1

~~~

PSBRb/GDP

Operational Deficit'GDP

15.8 19.9 23.3 27.8

6.6 3.0 2.7 4.3

~~

3.7 4.1 4.7 N.A.

Source: Cardoso and Reis (1986) and Banco Central do Brasil, Brasil: Programa Economico. February 1987.

Talculated on cash-flow basis, excludes the monetary authorities' deficit. bPublic Sector Borrowing Requirements: calculated on accrual basis, excludes the monetary authorities' deficit. 'Subtracts monetary correction from PSBR Share of Taxes in GDP (in percentages)

Table A.29

Transfers

Direct Taxes Year

(1)

1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980

9.1 9.5 10.5 10.9 10.7 11.0 11.7 12.2

1981

1982 1983 1984

Government Interest Payments (5)

Indirect Taxes (2)

Total (3)

Total (4)

16.6 15.5 15.6 15.6 15.5 14.4 13.6 13.4 12.7 11.5 12.8 12.4 12.5 12.5 10.4

25.7 25.8 26.1 26.5 26.4 26.2 25.3 25.6 24.7 23.3 23.2 23.6 25.1 24.4 21.5

9.1 8.2 8.6 8.3 7.5 8.2 8.6 9.2

I .2 2.3 1.2 I .2

9.0

2.0

1 .o

9.0 9.0 10.0 11.3 12.3 13.4

2.0

1 .0 3.5 2.6 2.5 2.3 1.5

11.0 11.0 10.1 11.2 12.6 11.9 11.1

Source: Conjuntura EconBmica.

1.1

1.2 1.1 I .9

1 .o

2.2 .0 4.3 5.0

Subsidies (6)

.o .0 .7 1.2 2.0 2.0 1.4 1.5

Net Taxes (7) 15.5 16.0 16.0 17.0 16.6 15.2

15. I 14.9 13.I 12.2 10.7 11.0 10.8 9.0 6.4

Table A.30

Net Debt of the Public Sector (in millions of cruzados) External

lntemal Federal Government and Central Bank Year December December December December December December

1981 1982 1983 1984 1985 1986

(1)

State and Local Governments (2)

Public Enterprises (3)

Total (1) (2) (3) (4)

919 1,961 8,387 47,255 176,724 25 1,143

1,052 2,879 8,831 29,315 110,120 192,357

1,824 4,832 16,839 54,335 214,568 379,618

3,795 9,672 34,057 130,905 501,412 823,118

Source: Banco Central do Brasil

+ +

Federal Government and Central Bank

Total

(5)

State and Local Governments (6)

Public Enterprises (7)

Total (5) + (6) + (7) (8)

1,405 3,942 26,673 81,760 269,801 554,836

273 755 2,951 1 1.053 49,371 77,105

3,051 7,374 31,064 106,895 410,730 557,928

4.729 12,071 60,688 199,708 729,902 1,189,869

(4)

+ (8) (9)

8,524 21,743 94,745 330,613 1,231,314 2,012,987

384

Eliana A. Cardoso and Albert Fishlow

Table A.31

Year

Brazilian Foreign Debt and Reserves (in billions of U S . dollars)

Medium- and Long-Term (I)

1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986p

3.644 3.666 3.281 3.780 4.403 5.295 6.622 9.521 12.572 17.166 21.171 25.985 32.037 43.511 49.904 53.848 61.411 70.198 81.319 91.091 94.857 101.540

Short-Term (2)

3.700 4.600 6.200 10.800 14.100 13.067 10.313 10.948 9.269 9.032

NB NB NB NB NB CB CB CB CB CB

Total Debt (I) (2)

+

Gross Reserves

3.644 3.666 3.281 3.780 4.403 5.295 6.622 9.521 12.572 17. I66 21.171 25.985 35.737 48.111 56.104 64.648 75.511 83.265 91.632 102.039 105.126 110.572

,483 ,421 ,198 .257 ,657 1.187 1.723 4. I83 6.416 5.269 4.041 6.544 7.256 11.895 9.689 6.913 7.507 3.994 4.563 11.995 11.608 6.760

Reserves Net of Liabilities of Monetary Authorities - .777

,412 ,617 - ,488 ,143 ,786 1.413 3.706 5.994 4.870 3.668 5.657 6.216 11.535 9.443 5.163 5.695 -2.332 ,086 5.096 4.400 ,771 -

Sources: Banco Central do Brasil and Nogueira Batista (1980) Notes: (1): Medium- and long-term debt outstanding at the end of the year. (2): Nonregistered debt. NB: Nogueira Batista’s estimates. CB: Central Bank figures. PF’reliminary estimate.

Table A.32

Net Capital Flows (in millions of U.S. dollars) Net Interest and Profits

Year

(1)

Amortization (2)

1975 1976 1971 1978 1979 1980 1981 1982 1983 1984 1985

1,804 2,039 2,462 3,342 5,348 7,457 10,305 12,551 10,263 11,449 11,092

2,172 2,888 4,060 5,324 6,385 5,010 6,242 6,952 6,863 6,468 8,890

Source: Banco Central do Brasil.

Capital Service (1) + (2) - (3)

+

(3)

Net Loans Financial Flows (4)

3,916 4,927 6,522 8,666 11,733 12,467 16,547 19,503 17,126 17,917 19,982

7,483 7,921 8,727 15,388 11,288 13,315 16,783 12,451 7,778 8,768 5,614

(4) - (3)

=

(5)

~

3,507 2,994 2,205 6,722 - 445 848 236 - 7.052 -9,348 -9,149 14,368

(5)

385

BraziVAppendix Average Cost of Brazil’s Foreign Debt

Table A.33

Yeas

1968 1969 1970 1971 1972 1973 1974 1975 I976 1977

1978 1979 1980 1981 1982 1983 1984 1985 1986

Net Interest Payments ($ millions) (1)

Lagged Net Debt ($ millions) (2)

Average Interest Rate (%) (3)

144 182 234 302 359 514 652 1,498 1,810 2,104 2,696 4,186 6,311 9,161 11,353 9,555 10,203 9,659 9,093

3,083 3,523 3,747 4.108 4,899 5,338 6,156 1 1,897 17,310 19,441 24,781 31,616 40,125 46,935 53,904 66,204 76,756 79,096 84,249

4.7 5.2 6.2 7.4 7.3 9.6 10.6 12.6 10.6 10.8 10.9 13.2 15.7 19.5 21.1 14.4 13.3 12.2 10.8

Source: Banco Central do Brasil

U.S. Inflation (%) (4)

Real Interest Rate (4%) (5)

4.4 5.1

.287

5.4

.759 2.286 2.975 3.592 1.654 3.019 5.133 4.726 3.259 4.236 5.952 9.232 14.245 10.137 9.251 8.522 7.985

5.0 4.2 5.8 8.8 9.3 5.2 5.8 7.4 8.6 9.2 9.4 6.0 3.9 3.1 3.4 2.6

,095

Table A.34

Foreign Debt Indicators (in millions of U.S. dollars) Amortization" (2)

Interest Payments (3)

Registered Debt (4)

Reserves (5)

Net Debt (4) - ( 5 ) (6)

Short-Run Debt (7)

Total Debt (4) + (7) (8)

Interest PaymentsiExports (9)

DebUGDP

Year

Debt Service (2) + (3) (1)

1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986

4.023 5,089 6,585 8,770 I 1,875 14,146 17,801 20,766 15,557 14,135 13,999 13,255

2,160 2,998 4,123 5,426 6,527 6,689 7,496 8,215 5,294 2,866 2,907 3,200

1,863 2,091 2,462 3,344 5,348 7,457 10,305 12,551 10,263 11,449 I 1.239 10,055

21,171 25,985 32,037 43,511 49,904 53,847 61.411 70,198 81,319 91,091 95,857 94,938

4,040 6,544 7,256 11,895 9,689 6,913 7,507 3,994 4,563 11,995 11,608 6,760

17,131 19,441 24,781 31,616 40,215 46,934 53,904 66,204 76,756 79,096 84,249 88,178

3,944 6,160 5,914 8,676 5,899 10,397 12,552 15,166 12,237 10,948 9,269 9,032

25,115 32,145 37,95 I 52,187 55,803 64,244 73,963 85,364 93,556 102,039 105,126 110,572

21.5 20.6 20.3 26.4 35. I 37.0 43.5 62. I 46.9 42.4 43.8 44.9

20 21 22 25 24 26 27 30 45 48 47 42

Source; Banco Central do Brasil. "Amortization up to 1983 is on an accrual basis

(10)

387

BraziVNotes Currency Breakdown of Brazilian Foreign Debt,' 1973-85, Selected Years (in percentages)

Table A.35

US. Year

Dollar

1973 1976 1979 1982 1985

79.9 84.9 75.8 82.9 74.8

German Mark

Japanese Yen

French Franc

British Pound

Swiss

Franc

Other

4.0 4.1 9.0

2.5 3.3 7.2 5.7 6.8

2.9

6.3

1.7

2.2

2.2 1.7 1.8

2.2 .9 .7

.7 .5 1.6 .9 2.0

3.6 3.3 2.0 2. I 8.4

5.8

4.4

Source: World Bank. "Public debt outstanding and disbursed at the end of the year.

Notes Chapter 1 1. See Albert Fishlow (1988a) and Eliana Cardoso (1988). 2. See Eliana Cardoso and Rudiger Dornbusch (1988).

Chapter 2 1. See World Bank (1984). 2. The trade price and quantity indexes are those of Conjuntura Econornica.

Chapter 4 1. The Brazilian indexation experience is described and analyzed in Rudiger Dombusch and M&io Henrique Simonsen (1986). 2. There are many studies of the Cruzado Plan, several of which are given in the reference list. See, for instance, Arida and Lara-Resende (1985), Cardoso and Dombusch (1987), Lopes (1986), and Modiano (1986).

Chapter 5 1. The different concepts of the budget deficit of the public sector are explained in appendix 2 in this chapter. 2. When prices are fully flexible and the budget deficit increases, the economy can jump from one equilibrium characterized by low inflation and high real cash balances to another with higher inflation and smaller real cash balances. The price level jumps, adjusting velocity. The growth rate of the monetary base, p, then, is all the time equal to the inflation rate, II, plus the rate of growth of income, Y, times the elasticity of the demand for the real monetary base in relation to income, p. That means that under full price flexibility, seignorage is equal to:

(AH)/P = p ( H / P ) =

(n + Y P ) ( H / P ) .

387

BraziVNotes Currency Breakdown of Brazilian Foreign Debt,' 1973-85, Selected Years (in percentages)

Table A.35

US. Year

Dollar

1973 1976 1979 1982 1985

79.9 84.9 75.8 82.9 74.8

German Mark

Japanese Yen

French Franc

British Pound

Swiss

Franc

Other

4.0 4.1 9.0

2.5 3.3 7.2 5.7 6.8

2.9

6.3

1.7

2.2

2.2 1.7 1.8

2.2 .9 .7

.7 .5 1.6 .9 2.0

3.6 3.3 2.0 2. I 8.4

5.8

4.4

Source: World Bank. "Public debt outstanding and disbursed at the end of the year.

Notes Chapter 1 1. See Albert Fishlow (1988a) and Eliana Cardoso (1988). 2. See Eliana Cardoso and Rudiger Dornbusch (1988).

Chapter 2 1. See World Bank (1984). 2. The trade price and quantity indexes are those of Conjuntura Econornica.

Chapter 4 1. The Brazilian indexation experience is described and analyzed in Rudiger Dombusch and M&io Henrique Simonsen (1986). 2. There are many studies of the Cruzado Plan, several of which are given in the reference list. See, for instance, Arida and Lara-Resende (1985), Cardoso and Dombusch (1987), Lopes (1986), and Modiano (1986).

Chapter 5 1. The different concepts of the budget deficit of the public sector are explained in appendix 2 in this chapter. 2. When prices are fully flexible and the budget deficit increases, the economy can jump from one equilibrium characterized by low inflation and high real cash balances to another with higher inflation and smaller real cash balances. The price level jumps, adjusting velocity. The growth rate of the monetary base, p, then, is all the time equal to the inflation rate, II, plus the rate of growth of income, Y, times the elasticity of the demand for the real monetary base in relation to income, p. That means that under full price flexibility, seignorage is equal to:

(AH)/P = p ( H / P ) =

(n + Y P ) ( H / P ) .

388

Eliana A. Cardoso and Albert Fishlow

Chapter 6 1. The two series are reproduced in the statistical appendix. Chapter 7 1. To obtain equation (l), we start from the equation for the balance of payments.

D, - D,-1= I*D,-l- NICA,

-

NDCF,

where D = external debt, NICA = noninterest current account, and NDCF = nondebt-creating capital flows. We divide both sides of this equation by current exports and use the definition of the growth rate of exports to obtain (1).

References Abreu, M. de P. 1978. Brazilian public foreign debt policy: 1931-1945. Brazilian Economic Studies, no. 4. Rio de Janeiro: Instituto de Pesquisa Economica Aplicada. -. 1987. Equacoes de demanda de importacoes revisitadas: Brasil, 19601985. Texto para Discussao, no. 148. Rio de Janeiro: Pontificia Universidade Catolica. Arida, P., and A. Lara-Resende. 1985. Inertial inflation and monetary reform. In Inpation and indexation: Argentina, Brazil and Israel, ed. John Williamson. Cambridge: MIT Press. Bacha, E. 1981. Crescimento economico e salarios: Precos e distribuicao na industria de transformacao. Pesquisa e Planejamento Economico, 2 , no. 4, (December). Bacha, E., and L. Taylor. 1978. Brazilian income distribution in the 1960s: Facts, model, results and the controversy. Journal of Development Studies, 14. Baer, W. 1979. The Brazilian economy: Its growth and development. Columbus: Grid Publishing. Batista, Paul0 Nogueira. 1983. Mito e realidade da di'vida externa brasileira. Rio de Janeiro: Paz e Terra. -. 1985. International financial flows to Brazil since the late 1960's. Rio de Janeiro: Fundaqb Getulio Vargas. Mimeo, May. -. ed. 1988. Novos ensaios sobre o setor externo brasileiro. Rio de Janeiro: FundaGBo Get6lio Vargas. Bergsman, J. 1970. Brazil: Industrialization and trade policies. London: Oxford University Press. Bittermann, H. 1973. The refunding of international debt. Durham: Duke University Press. Braga, H . , and R. Markwald. 1983. Exportacoes brasileiras de manufaturados. Pesquisa e Planejamento Economico 4 (3). Bresser Pereira, L. 1984. Development and crisis in Brazil: 1930-1983. Boulder: Westview. Cardoso, E. 1980. Incentivos As Exportaqoes de Manufaturas: Strie Historica. Revista Brasileira de Economia 34 ( 2 ) .

388

Eliana A. Cardoso and Albert Fishlow

Chapter 6 1. The two series are reproduced in the statistical appendix. Chapter 7 1. To obtain equation (l), we start from the equation for the balance of payments.

D, - D,-1= I*D,-l- NICA,

-

NDCF,

where D = external debt, NICA = noninterest current account, and NDCF = nondebt-creating capital flows. We divide both sides of this equation by current exports and use the definition of the growth rate of exports to obtain (1).

References Abreu, M. de P. 1978. Brazilian public foreign debt policy: 1931-1945. Brazilian Economic Studies, no. 4. Rio de Janeiro: Instituto de Pesquisa Economica Aplicada. -. 1987. Equacoes de demanda de importacoes revisitadas: Brasil, 19601985. Texto para Discussao, no. 148. Rio de Janeiro: Pontificia Universidade Catolica. Arida, P., and A. Lara-Resende. 1985. Inertial inflation and monetary reform. In Inpation and indexation: Argentina, Brazil and Israel, ed. John Williamson. Cambridge: MIT Press. Bacha, E. 1981. Crescimento economico e salarios: Precos e distribuicao na industria de transformacao. Pesquisa e Planejamento Economico, 2 , no. 4, (December). Bacha, E., and L. Taylor. 1978. Brazilian income distribution in the 1960s: Facts, model, results and the controversy. Journal of Development Studies, 14. Baer, W. 1979. The Brazilian economy: Its growth and development. Columbus: Grid Publishing. Batista, Paul0 Nogueira. 1983. Mito e realidade da di'vida externa brasileira. Rio de Janeiro: Paz e Terra. -. 1985. International financial flows to Brazil since the late 1960's. Rio de Janeiro: Fundaqb Getulio Vargas. Mimeo, May. -. ed. 1988. Novos ensaios sobre o setor externo brasileiro. Rio de Janeiro: FundaGBo Get6lio Vargas. Bergsman, J. 1970. Brazil: Industrialization and trade policies. London: Oxford University Press. Bittermann, H. 1973. The refunding of international debt. Durham: Duke University Press. Braga, H . , and R. Markwald. 1983. Exportacoes brasileiras de manufaturados. Pesquisa e Planejamento Economico 4 (3). Bresser Pereira, L. 1984. Development and crisis in Brazil: 1930-1983. Boulder: Westview. Cardoso, E. 1980. Incentivos As Exportaqoes de Manufaturas: Strie Historica. Revista Brasileira de Economia 34 ( 2 ) .

389

-.

BraziUReferences

1986. What policymakers can learn from Brazil and Mexico. Challenge 29, no. 4 (September/October): 19-28. -. 1987a. Inflation, growth and the real exchange rate: Brazil and Latin America, 1850-1983. New York: Garland Publishing. -. 1987b. Latin America’s debt: Which way now? Challenge 30, no. 2 (May/June): 11-17. -. 1988. Lessons of the 1890s for the 1980s: Comments. In Debt, stabilization and development, ed. R. Findlay. Oxford: Basil Backwell. Cardoso, E., and R. Dombusch. 1987. Brazil’s tropical plan. American Economic Review 77 (May): 288-92. -. 1988. Brazilian debt: Past and present experiences. Cambridge: Massachusetts Institute of Technology, mimeo. Cardoso, E., and E. Reis. 1986. Deficits, dividas e inflacao no Brasil. Pesquisa e Planejamento Economico 7, no. 4 (December): 575-98. Coes, D. 1986. Brazilian trade policy: Inferences for sequencing of liberalization policies. Washington, D.C.: World Bank, mimeo. Conselho Monetario Nacional. 1982. Foreign sector programme in 1983, Brasilia. CNZ, P. D. 1984. Divida externa e politica economica: A experiencia brasileira nos anos sextenta. SLo Paulo: Brasiliense. Diaz Alejandro, C. 1983. Some aspects of the 1982-1983 Brazilian payments crisis. Brookings Papers in Economic Activity 1515-52. Dib, M. 1981. Equacoes de demanda de importacoes. Revista de Economia Brasileira 35 (2). Dombusch, R. 1985. The Larida proposal: Comments. In Inflation and indexation: Argentina, Brazil and Israel, ed. John Williamson. Cambridge: MIT Press. . 1987. The world debt problem. New York: The Twentieth Century Fund. Dombusch, R., and M. Simonsen, eds. 1986. Injlation, debt and indexation. Cambridge: MIT Press. -. 1987. Injlation stabilization with incomes policy support. New York: Group of Thirty. Evans, P. 1979. Dependent development: The alliance of multinational, state, and local capital in Brazil. Princeton: Princeton University Press. Feldstein, M. 1987. Latin America’s debt. The Economist (June 27). Fischer, Stanley. 1977. Long term contracts, rational expectations, and the optimal money supply rule. Journal of Political Economy 85:191-205. Fishlow, A. 1972. Some reflections on post-1964 economic policy. In Authoritarian Brazil, ed. A. Stepan. New Haven: Yale University Press. -. 1975. Foreign trade regimes and economic development: Brazil. University of California, Berkeley, mimeo. . 1976. Brazilian size distribution of income. In Income distribution in Latin America, ed. A. Foxley. Cambridge: Cambridge University Press. . 1984. The debt crisis: Round two ahead? In Adjustment crisis in the third world, ed. R. Feinberg and V. Kallab. Washington, D.C.: Overseas Development Council. -. 1985a. The debt crisis: A longer perspective. Journal of Development Planning, no. 16:83-104. -. 1985b. Lessons from the past: Capital markets during the 19th century and the interwar period. International Organization 39:383-439. -. 1988a. Lessons of the 1890s for the 1980s. In Debt, stabilization and development, ed. R. Findlay. Oxford: Basil Backwell. -. 1988b. Tale of two presidents. In Democratizing Brazil: Problems of transition and consolidation, ed. A. Stepan. Oxford: Oxford University Press.

390

Eliana A. Cardoso and Albert Fishlow

Furtado, C. 1971. The economic growth of Brazil. Berkeley and Los Angeles: University of California Press. Goldsmith, R. 1986. Desenvolvirnento Jinanceiro sob um seculo de inflacao: Brasil, 1850-1984. SLo Paulo: Harper & Row do Brasil. Horta, Maria Helena T. T. 1985. Sources of Brazilian export growth in the 70s. Brazilian Economic Studies 9: 164-65. Knight, P. 1985. Economic stabilization and medium-term development strategy in Brazil. Washington, D.C.: World Bank, mimeo. Knight, P., and R. Moran. 1981. Brazil: Poverty and basic needs. Washington, D.C.: World Bank. Kume, Honorio. 1985. Uma nota sobre a retomada do crescimento economico e a necessidade de importacoes. Revista Brasileira de Comercio Exterior, November. Lamounier, B., and A. Moura. 1986. Economic policy and political opening in Brazil. In Latin Americnn political economy, ed. J. Hartlyn and S. Morley. Boulder: Westview. Leff, N. 1982. Underdevelopment and development in Brazil. London: Allen and Unwin. Lopes, F. 1986. 0 choque heterodoxo. Rio de Janeiro: Editora Campus. Macedo, R. 1977. A critical review of the relation between post-I964 wage policy and the worsening of Brazil’s income distribution in the 1960’s. Explorations in Economic Research 4:117-40. Malan, P., et al. 1977. Politica economica externa e industrializacao no Brasil. Rio de Janeiro: IPEA, Relatorios de Pesquisa. Modiano, E. 1986. Da injlacao ao cruzado. Rio de Janeiro: Editora Campus. Moraes, P. B. 1985. Equacoes de demanda de importacoes. Text0 para discussao. Rio de Janeiro: Pontificia Universidade Catdica. Moreira, Heloisa Correia, and Aloisio de Araujo. 1984. Politica brasileira de importacoes: Uma descricao. Rio de Janeiro: Instituto Brasileiro de Pesquisa Economica e Social, mimeo. Morley, S. 1982. Labor markets and inequitable growth: The case of inequitable growth in Brazil. Cambridge: Cambridge University Press. Musalem, Albert0 Roque. 1981. Politica de Subisidios e ExportaGbes de Manufaturados no Brasil. Revista Brasileira de Ecomonia 35 (1). Neuhaus, P., coordenador. 1980. Economia brasileira: U r n visao historica. Rio de Janeiro: Editora Campus. Pechman, C. 1984. 0 Dolar paralelo no Brasil. Rio de Janeiro: Paz e Terra. Pelaez, C., and W. Suzigan. 1976. Historia monetaria do Brasil: Analise da politica, comportamento e instituicoes monetarias. Sene Monogrhfica. Rio de Janeiro: Instituto de Pesquisa Economica Aplicada. Pfeffermann, G. 1985. The social cost of recession in Brazil. Washington, D.C.: World Bank, mimeo. Rossi, J. W. 1986. Distribuicao da renda. Rio de Janeiro: INPES, mimeo. Salazar-Carillo, J., and R. Fendt, eds. 1985. The Brazilian economy in the eighties. New York: Pergamon Press. Simonsen, M. H. 1972. B r a d 2002. Rio de Janeiro: APEC-Bloch. . 1983. Indexation: Current theory and the Brazilian experience. In Inflation, debt and indexation, ed. R. Dornbusch and M. Simonsen. Cambridge: MIT Press. -. 1985. The developing country debt problem. In International debt and developing countries, ed. G . Smith and J. Cuddington. Washington, D.C.: World Bank. . 1986. Brazil. In The open economy, ed. R. Dornbusch and F. L. Helmers. Washington, D.C.: World Bank.

391

BraziVReferences

Skidmore, T. 1967. Politics in Brazil, 1930-1964: An experiment in democracy. Oxford: Oxford University Press. -. 1988. The politics of military rule in Brazil: 1964-1985. New York: Oxford University Press. Stepan, A., ed. 1973. Authoritarian Brazil. New Haven: Yale University Press. Taylor, John. 1979. Staggered wage setting in a macro model. American Economic Review (papers and proceedings) 69(May): 108- 13. Tyler, W. G. 1976. Manufactured export expansion and industrialization in Brazil. Kiel: Kierlen Studien. Versiani, F., and J. R. Mendonca de Barros, organizadores. 1978. Formacao economica do Brasil. Slo Paulo: Edicao Saraiva. Villela, A., and W. Suzigan. 1973. Politica economica e crescimento da economia brasileira: 1889-1945. Sene Monografica. Rio de Janeiro: Instituto de Pesquisa Economica Aplicada. Wells, J. 1978. Brazil and the post-1973 crisis in the international economy. In Inflation and stabilization in Latin America, ed. R. Thorp and L. Whitehead. New York: Holmes & Meier. Werneck, R. 1983. Estrangulamento externo e investimento publico. In Divida externa, recessao e ajuste estrutural-0 Brasil diante da crise, P. Arida, organizador. Rio de Janeiro: Paz e Terra. World Bank. 1983. Brazil: Industrial policies and manufactured exports. A World Bank Country Study, Washington, D.C. . 1984. Brazil: Economic memorandum. Washington, D.C. . 1985. Brazil: Financial systems review. A World Bank Country Study, Washington, D .C.

This Page Intentionally Left Blank

~

1

Introduction

The purpose of this work is to critically examine macroeconomic policy in Mexico from 1958 to 1986. This period is of special interest as it embodies two distinct, sharply contrasting phases. Between 1958 and 1972, the inflation rate never exceeded 6 percent, while annual output growth averaged 6.7 percent. The period since 1972, by contrast, has been marked by a succession of increasingly severe macroeconomic crises. Figures 1.1 - 1.3 depict how

8.0. 7.0

’

6 .O 5.0 4.O 3.0.

2 .o I.o

0 -1.0

-2.0 -3.0

Fig. 1.1 Growth rates of total and per capita real output, 1950-86 Source: Raw data are from Indicadores Econornicos (Bankof Mexico). 395

396

Edward F. Buffie

real output, the foreign debt, and inflation have fluctuated. Inflation and the foreign debt began rising rapidly after 1972. During the Echeverria administration the foreign debt increased fourfold to reach $28 billion, and in the last quarter of 1976 a major balance of payments crisis erupted. Just six years later and notwithstanding an enormous increase in oil revenues, the foreign debt had climbed to $88 billion and the country found itself in the grips of a much more threatening crisis: in 1982 the inflation rate jumped to 98 percent, the fiscal deficit was an astounding 17.6 percent of GDP, and neither the public nor the private sector was capable of meeting its scheduled debt service payments. The adjustment following the 1982 crisis has been and continues to be traumatic. In a country that experienced continuously positive growth from 1933 to 1981, real per capita income has fallen 13.4 percent (as of 1986) and real wages have declined to levels not seen since the early sixties. At the end of 1986, inflation was running well over 100 percent, the investment rate

loo

1

?

1 '

6oi --Total

Foreign Debt

-

40-

3020-

1950

1955

1960

1965

1970

1975

1980

1985

Fig. 1.2 Total foreign debt and public sector foreign debt (billion $)

Source: Raw data are from Perspectivas Economicas de Mexico (CIEMEXWHARTON), various issues.

397

Mexico/Chapter 1

I00 90

80 70

60 50 40

58 60626466687072 74 76 78 80 82 8486

Fig. 1.3 The domestic inflation rate, 1958-86 Source: Raw data are from Indicadores Economicos (Bank of Mexico).

was at an historical low, and the foreign debt was $10 billion larger than in 1982. Relations with foreign creditors remain tense; to date there have been three major debt reschedulings. In chapters 2 through 5 I analyze economic performance during the era of Stabilizing Development and the Echevenia, Lopez Portillo, and De La Madrid administrations. These chapters explore various aspects of recent macroeconomic history. They track the evolution of economic policy and the major macroeconomic variables of interest such as real wages, the foreign debt, capital flight, the fiscal deficit (broken down according to the deficits of the parastatal sector, of financial intermediaries, and of the rest of the government), and different measures of financial intermediation. In the next two chapters I develop formal economic models in an attempt to gain some insight into the factors responsible for the dismal post-1982 record of falling per capita incomes, low real wages, high inflation, and widespread underemployment. Chapter 6 is an analysis of the repercussions

398

Edward F. Buffie

of import compression on real wages and underemployment, while in chapter 7 I investigate the links between capital accumulation, inflation, fiscal deficits, and financial intermediation. Chapter 8 covers various topics relating to the evolution of the foreign debt, with a detailed discussion of the different debt reschedulings undertaken since 1982 and institutional aspects of the debt management process. The final chapter briefly examines the economy’s future prospects and summarizes the main policy implications of the study.

2

The Record of Stabilizing Development

After the devaluation of the peso in 1954, the Mexican economy entered a phase of high growth and low inflation that would last until the end of the sixties. This period has since come to be known as the era of Stabilizing Development (SD). Though it is difficult to pinpoint its exact starting date, there is general agreement that the SD period covered at least the years 1958-70; that is, mainly the administrations of Presidents Adolfo L6pez Mateos (1959-64) and Gustavo Diaz Ordaz (1965-70). As stated by the then Minister of Finance, Antonio Ortiz Mena, the main objectives of economic policy during SD were to increase private sector savings and capital accumulation, maintain price stability and a fixed parity with the dollar, and increase real wages (Ortiz Mena 1970). These goals were largely achieved (tables 2. l a and 2. lb), leading observers to speak of a “Mexican miracle.” The exchange rate was kept fixed at 12.5 pesos per dollar, and the annual inflation rate averaged 3.8 percent. Real output grew at an average rate of 6.7 percent, and the share of gross fixed investment in GDP rose (at 1960 prices) from 16.2 to 20.8 percent. The real industrial sector wage inclusive of fringe benefits grew at an annual average rate of roughly 4 percent. Workers in the urban informal and agricultural sectors also appear to have experienced large real wage gains. (The data bearing on real wages in the latter two sectors will be discussed in section 2.3.2.) In the next two sections I discuss in detail the macroeconomic, trade, and industrial policies that constituted the SD program.’ Section 2.3 is a critical examination of the conventional view that the SD strategy was responsible for a severe worsening in underemployment and the distribution of income and that by 1970 it could no longer deliver sustainable, high rates of growth.

398

Edward F. Buffie

of import compression on real wages and underemployment, while in chapter 7 I investigate the links between capital accumulation, inflation, fiscal deficits, and financial intermediation. Chapter 8 covers various topics relating to the evolution of the foreign debt, with a detailed discussion of the different debt reschedulings undertaken since 1982 and institutional aspects of the debt management process. The final chapter briefly examines the economy’s future prospects and summarizes the main policy implications of the study.

2

The Record of Stabilizing Development

After the devaluation of the peso in 1954, the Mexican economy entered a phase of high growth and low inflation that would last until the end of the sixties. This period has since come to be known as the era of Stabilizing Development (SD). Though it is difficult to pinpoint its exact starting date, there is general agreement that the SD period covered at least the years 1958-70; that is, mainly the administrations of Presidents Adolfo L6pez Mateos (1959-64) and Gustavo Diaz Ordaz (1965-70). As stated by the then Minister of Finance, Antonio Ortiz Mena, the main objectives of economic policy during SD were to increase private sector savings and capital accumulation, maintain price stability and a fixed parity with the dollar, and increase real wages (Ortiz Mena 1970). These goals were largely achieved (tables 2. l a and 2. lb), leading observers to speak of a “Mexican miracle.” The exchange rate was kept fixed at 12.5 pesos per dollar, and the annual inflation rate averaged 3.8 percent. Real output grew at an average rate of 6.7 percent, and the share of gross fixed investment in GDP rose (at 1960 prices) from 16.2 to 20.8 percent. The real industrial sector wage inclusive of fringe benefits grew at an annual average rate of roughly 4 percent. Workers in the urban informal and agricultural sectors also appear to have experienced large real wage gains. (The data bearing on real wages in the latter two sectors will be discussed in section 2.3.2.) In the next two sections I discuss in detail the macroeconomic, trade, and industrial policies that constituted the SD program.’ Section 2.3 is a critical examination of the conventional view that the SD strategy was responsible for a severe worsening in underemployment and the distribution of income and that by 1970 it could no longer deliver sustainable, high rates of growth.

Table 2.la

MacroeconomicAggregates (% change)" 1958

1959

1960

1961

1962

1963

1964

1965

1966

1967 ~~~

Real GDP Manufacturing Agriculture, forestry, and fisheries Inflationb Gross fixed capital formation Real exchange rate'

Table 2.lb

1968

1969

1970

~

5.3 5.4

3.0 9.0

8.1 8.5

4.9 5.6

4.7 4.9

8.0 9.2

11.7 17.5

6.5 10.0

6.9 9.6

6.3 7.1

8.1 10.5

6.3 8.4

6.9 8.7

6.9 .5

-3.1 3.7

5.2 5.0

1.9 3.4

3.8 2.9

5.2 3.2

7.5 5.9

5.4 2.2

1.7 4.0

2.7 2.8

3.1 2.4

1.1 4.0

4.9 4.8

-5.9 108.7

1.3 104.8

14.9

100.0

.8 94.7

5.4 92.2

11.5 89.1

21.8 84.7

6.0 84.2

8.9 83.6

13.7 82.3

9.6 82.6

7.4 82.0

8.3 81.6

Composition of Output (% of G D P ) ~

Private consumption Government consumption Gross fixed capital formation Change in inventories Exports Imports

79.6 6.7

79.1 6.3

76.2 6.3

75.3 6.5

75.1 7.1

73.6 7.4

72.6 7.4

71.8 7.2

72.1 7.3

72.5 7.4

73.2 7.5

71.9 7.4

71.9 7.5

16.2 .6 11.5 14.7

15.9 .4 11.4 13.2

16.9 2.7 10.3 12.6

16.3 3.1 10.7 11.8

16.4 2.0 10.8 11.5

16.9 3.1 10.5 11.6

18.5 3.2 10.5 12.1

18.4 4.3 10.1 11.8

18.7 3.2 9.9 11.3

20.0 2.5 9.1 11.4

20.3 1.6 9.1 11.8

20.5 1.9 10.0 11.7

20.8 3.0 8.7 11.9

Sources: All national income accounts data is from lndicodores Economicos (Bank of Mexico). Wage data is from the Bank of Mexico's survey of large-scale manufacturing firms (Encuesta Industrial Mensun. "Real variables are expressed in terms of 1960 prices. bDecember-to-December change in the CPI. 'Calculated as the period average exchange rate multiplied by the ratio of the U S . wholesale price index (now called the producer price index) to the Mexican GDP deflator.

dOutput shares at 1960 prices.

Edward F. Buffie

400

2.1 Macroeconomic Policy Macroeconomic policy was geared toward promoting capital accumulation and industrialization while preserving price and exchange rate stability. A variety of tax and expenditure measures were employed to raise the return on domestic investment. The 1955 industrial promotion law (Ley de Industrias Nuevas y Necesarias) provided an extensive set of tax subsidies to new and “necessary” industries (defined to be industries in which the market share of domestic firms was less than 80 percent). Firms in such industries received rebates covering 40 percent of the corporate income tax and 100 percent of stamp and sales taxes and all duties on imported machinery, equipment, and raw materials (Solis 1981, 6). In 1961 the corporate income tax was amended to allow for accelerated depreciation allowances. Dividends and interest income were taxed at low, flat rates and accumulated to other income sources in calculating the taxable income base. To promote reinvestment of profits, neither capital gains nor retained profits were taxed after 1965. And lastly, high levels of evasion of the corporate income tax were tolerated. The statutory rate of 42 percent was not the effective rate for most firms. Public sector investment in projects complementary to private sector capital and low prices for publicly provided inputs also enhanced the profitability of private investment. Most public sector prices, especially energy prices, increased more slowly than the inflation rate. According to an index constructed by Clavijo (1980), the real price of goods and services provided by the public sector fell 12.5 percent between 1961 and 1970 (table 2.2). Public investment favored the industrial sector to a greater extent than in earlier periods. Table 2.3 shows how the composition of public sector capital outlays shifted over the 1954-70 period. The share of the industrial sector in total investment climbed from 35.4 percent in 1954-58 to 40.1 percent during the Diaz Ordaz administration, while the shares of agriculture and communications and transportation declined. In real terms (deflating by the GDP deflator) public sector industrial investment rose 204 percent during Table 2.2

Source:

Real Prices of Public Sector Goods and Services” Year

Price

Year

Price

1961 1962 1963 1964 1965

101.7 99.7 98.6 94.4 97.1

1966 1967 1968 1969 1970

96.9 95.9 96.2 92.9 89.0

Clavijo (1980).

“Period average price deflated by the period CPI.

401

Mexico/Chapter 2 Composition of Public Investment

Table 2.3

Communications Period

1954-58 1959-64 1965-70 1954-70

Agriculture

lndustry

& Transportation

Social Welfare

Administration & Defense

13.30 10.60 10.96

35.41 37.49 40.06 38.76

33.55 24.86 21.83 24.08

15.13 24.22 25.20 23.76

2.61 2.83 1.95 2.29

11.11

Total 100 100

I00 100

Source: Estadisticas Historicas de MCxico (MCxico, D.F.: INEGI [Instituto Nacional de Estadistica Geogrofia

e Informatica], 1985).

SD, with much of the increased expenditures going to expand the supply of electricity and oil. Monetary policy played an important, complementary role in stimulating capital accumulation. Real interest rates on bank deposits were at positive levels, several points above those prevailing in the United States. The high real rates along with the stable exchange rate succeeded in attracting a much larger fraction of private sector savings into the banking system, producing what some have termed a “financial miracle” (table 2.4). Led by an enormous increase in the demand for interest-bearing, high-liquidity deposits-the real growth of bonos Jinancieros averaged 17.5%-the supply of bank funds and private sector credit expanded rapidly. While tax, expenditure, and monetary policies were all enlisted to stimulate investment and accelerate the pace of industrialization, this effort was combined with a commitment to prudent macroeconomic management. The “rules of the game” were well defined and called for fiscal and monetary policies to be coordinated in a fashion consistent with the goals of price and exchange rate stability. The growth rate of the monetary base was closely monitored, and it was well understood that if the fiscal deficit exceeded the level consistent with the planned rate of monetary emission, expenditures were to be lowered until the gap was eliminated. Institutional arrangements were of crucial importance in this respect. The Ministry of Finance (Secretaria de Hacienda y Credit0 Publico, or SHCP) was responsible for controlling both revenue collection and public expenditures. This made Hacienda the main economic authority; the Central Bank was in charge of the less important tasks of setting interest rates and regulating the financial system. Reinforcing the centralization of economic power in Hacienda was the immense personal prestige of Ortiz Mena, who headed the Ministry of Finance from 1958 to 1970. Even in periods when relations with the president were strained, Ortiz Mena’s authority in financial matters was regarded as indisputable. Fiscal deficits were generally small but by no means trivial during SD, ranging from 1 to 4 percent of GDP during the sixties.2 Deficits of this magnitude, however, were not highly inflationary. The main source of

Table 2.4

Monetary Aggregates and Real Interest Rates

Real growth ratesa Monetary base M2 M3 M4 Total stock of bank fundsb Total credit of the banking system' Percentage of GDP Monetary base M2 M3 M4 Total stock of bank funds Total credit of the banking system Interest rate on finance bonds Nominal rate Real rate'

1958

1959

1960

1961

1962

1963

1964

1965

1966

1967

1968

1969

1970

-

-

-

8.7 7.0

6.4 7.7

2.6 8.4

4.1 1.4 7.2 9.0

12.3 9.8 12.9 14.7

15.4 12.1 17.3 16.2

20.3 9.9 15.0 13.0

9.8 3.9 13.6 12.8

7.9 6.7 14.1 16.0

13.3 4.7 12.7 15.4

13.0 10.1 13.5 14.2

8.4 6.3 8.2 14.6

5.6 5.4 5.7 12.9

11.7

17.4

18.1

14.1

15.6

18.8

17.6

14.8

16.4

14.1

20.1

12.3

12.8

10.5

16.3

15.1

13.6

13.0

11.8

15.8

11.0

8.6 10.8 14.5 17.9

8.9 10.9 15.3 19.1

9.4 11.3 16.4 20.5

9.8 11.1 16.8 20.8

10.7 11.3 18.3 22.3

10.8 11.0 19.4 23.8

11.3 11.0 20.8 26.0

11.9 10.9 21.7 27.6

12.2 11.0 22.5 29.5

12.2 10.9 22.4 31.3

13.2 24.7

14.5 26.6

15.8 27.5

16.2 27.1

17.8 30.5

19.4 32.4

21.6 34.6

23.2 36.1

25. I 38.4

27.0 40.6

9.0 5.6

9.0 6.1

9.0 5.8

9.0 3.1

9.0 6.8

9.0 5.0

9.0 6.2

9.0 6.6

18.8

13.3

-

-

-

11.2 12.8

11.5 14.1

11.1 14.0

-

-

-

Source: All raw financial data are from Indicudores Economicos (Bank of Mexico).

Notes: M2 = Currency held by the public + peso- and foreign-currency-denominated demand deposits. M3 nonliquid (i.e., fixed-term) savings accounts.

aReal monetary aggregates are calculated as the end-of-year balance deflated by the end-of-year CPI. bM4 less currency held by the public. 'Credit of the Central Bank, the development banks, and commercial banks. dAverage of the end- and beginning-of-year monetary aggregate relative to GDP 'Nominal rate less the percentage change in the end-of-year CPI.

=

M2

9.37 5.4

+ liquid savings accounts. M4 = M3 +

9.37 4.6

403

Mexicotchapter 2

funds for financing the fiscal deficit was not the printing press but rather forced “loans” extracted from the commercial banking system through the imposition of high reserve ratios (= 34 percent). Since bank deposits grew at a rapid pace, this provided a considerable margin for noninflationary financing of the fiscal deficit. In most years, the government was able to extract seignorage in excess of 1 percent of GDP (table 2.51, even though inflation remained very low. The modest fiscal deficits and brisk growth in tax revenues supported a considerable increase in total public sector expenditures. Real tax revenues rose at a pace of 8.6 percent per annum. As this was well above the growth rate of real output, the share of public sector revenues in GDP increased by almost two percentage points between 1960 and 1970. From table 2.6 it can be seen that direct taxes were the main source of revenue growth. Indirect taxes grew at the sluggish rate of 4 percent per annum and in 1970 supplied only 57 percent of total tax revenues, a rather low figure for a less developed country. The bulk of the growth in direct taxes came from taxation of wages and salaries, which were taxed at increasing marginal rates. Concern about inequities in the tax system and the desire to finance a more ambitious public investment program led to an attempt at tax reform in 1964-65. I discuss the failure of this attempted tax reform at length not because it was, as is often claimed, responsible for mounting fiscal deficits toward the end of the SD era. Tax revenues grew rapidly despite the failure to achieve tax reform and, as is shown later in section 2.3.3, after taking account of the normal workings of the political business cycle, there is no evidence that fiscal discipline deteriorated during the Diaz Ordaz administration. The failed campaign for tax reform in 1964-65 is significant instead because it foreshadowed failures in the following Echevema, Lopez Portillo, and De La Madrid administrations, when expansion of the tax base would be essential for averting a loss of fiscal control. 2.1.1 The Attempt at Tax Reform in 1964-65 In 1963 Ortiz Mena invited Nicholas Kaldor to prepare a report on restructuring the tax ~ y s t e r nKaldor .~ proposed that the fractionalized system of reporting income be replaced by a global income tax. The exemption level Table 2.5

Seignorage (96 of GDP)

1961 1962 1963 1964 1965

.67 1.30 1.73 1.68 .75

1966 1967 1968 1969 1970

1.17 .84 1.35 1.14 1.12

Table 2.6

Real growth rates" Total taxes Direct Indirect Share of GDP Total taxes Direct Indirect Share of direct taxes in total tax revenue

Tax Revenue Performance 1958

1959

1960

1961

1962

1963

1964

1965

1966

1967

1968

1969

1970

8.6 -4.6 15.4

2.8 5.4 1.7

6.8 13.4 3.9

2.7 7.9 .2

8.7 12.6 6.6

11.1 12.3 10.4

15.0 25.6 9.1

-1.4 -16.4 8.3

15.5 36.8 4.9

5.6 13.6 .4

15.5 15.3 15.6

7.4 11.1

4.7

7.5 5.3 9.2

7.1 2.1 5.0

7.1 2.2 4.9

7.0 2.3 4.7

6.9 2.3 4.5

7.1 2.5 4.6

7.3 2.6 4.7

7.6 3.0 4.6

7.0 2.3 4.7

7.6 3.0 4.6

7.5 3.2 4.3

8.0 3.4 4.6

8.1 3.5 4.6

8.1 3.5 4.6

30.0

30.7

32.6

34.3

35.5

35.9

39.2

33.2

39.4

42.3

42.3

43.7

42.8

Sources: All data for 1965-70 are from Esfadisticas Hacendarias del Sector Publico: Cifras Anuales, 1965-1982 (SHCP). Data for 1958-64 are from Estadisricas Hisroricas de M h i c o (MCxico, D.F.: INEGI, 1985): 632. The series for total taxes from this source differs significantly from that found in Esradisticas Hacendarias. Our series for total taxes over 1958-64 is constructed by splicing the Esfadisticas Hacendarias series to the Esradisricas Historicas series using the 1965 overlap.

"Nominal revenues deflated by the GDP deflator.

405

MexicoJChapter 2

was $l,OOO, and the taxable base was to be calculated by summing all income regardless of its source. Progressive rates would be applied against the global base, with the maximum rate being 40 percent. Very modest wealth and inheritance taxes were to supplement the global income tax. The proposed wealth tax required full disclosure of assets and would be levied against both tangible and nontangible wealth (net of liabilities) valued at acquisition prices. The exemption level was set at $40,000; wealth holdings exceeding this level were to be taxed at a rate of 0.25 percent, increasing in equal 0.25 percent intervals with each additional $SO,OOO until the rate reached a ceiling of 1 percent. By and large, the proposed tax reform was rejected. A few piecemeal changes were introduced (interest payments from fixed rate securities and housing rents became subject to taxation), but efforts to globalize the income tax and institute a wealth tax foundered on two contentious issues that have ever since undermined attempts at substantive tax reform: 1. The anonymity of wealth. The wealth tax required bonds and stocks to be

nominative and registered. The private sector (and many important politicians) opposed this measure, which would disclose the amount and possibly the origin of their wealth. 2. The equal treatment of property and labor income. This was and continues to be the major obstacle to reaching an agreement on the definition of a broad tax base. In August 1966 Ortiz Mena (1973, 46-47) stated publicly that tax reform was more a “process” than a “radical change” and suggested leaving the date “adequate for its implementation to more favorable circumstances.” After this, the drive for a major tax reform was abandoned.

2.2 Trade and Industrial Policies The manufacturing sector was the engine of growth during SD. Real manufacturing growth was consistently high, averaging 9.0 percent during the terms of both L6pez Mateos and Diaz Ordaz. As a fraction of GDP, manufacturing output increased from 23.3 percent in 1960 to 27.9 percent in 1970. Manufacturing growth was fostered by an import-substituting trade strategy involving an escalated structure of protection. Tariff rates were 5- 15 percent on raw materials and intermediate products, 20-25 percent on machinery and tools, 25-35 percent on other manufactured goods, and 100 percent on automobiles (Solis 1981, 6). The tariff structure, however, may not accurately reflect the actual pattern and degree of protection in view of the fact that quantitative restrictions came into widespread use in the sixties. While 35.1 percent of imports (in value terms) were subject to licenses in

406

Edward F. Buffie

1957, by 1970 this figure had increased to 68.3 percent (Gil Diaz 1984b, table A-7). Nonetheless, most studies concur that, by LDC standards, the trade regime was modestly prote~tionist.~ Besides the trade regime, public sector pricing policy, interest rates, and tax credits affected the structure of relative factor prices. In the industrial sector, the user cost of capital and the real price of electricity exhibited sharp declines, while real raw material prices increased slightly. The real (product) wage, by contrast, grew very strongly after 1961, ending up 50 percent higher in 1970 than at the beginning of the decade. It is difficult to ascertain how the mix of trade and industrial policies affected employment growth in different sectors. There are numerous serious problems with the employment data in the 1960 and 1970 population censuses. The original 1960 census was marred by gross processing errors, and the corrected version still appears to overenumerate greatly the size of the agricultural labor force. Classification schemes also differ as between the two censuses, and in the 1970 census a large number of labor force participants were not assigned to any category. (This was also a problem, but to a lesser extent, in the 1960 c e n ~ u s . )Estimates ~ of employment growth differ widely depending on the nature of the adjustments made to “correct” these and other flaws in the data. Although the quality of the data is problematic, the weight of the evidence favors the conclusion that, despite large, sustained increases in the real wage, employment growth in the industrial sector considerably outstripped the growth rate of the labor force. Table 2.7 presents the estimates made by Unikel (1978) and Altimir (1974) for the sixties. According to Altimir’s Table 2.7

Employment Growth (average annual rate) 1950-60

1960-69

.4 3.9 4.2

2.0

.5 5.2 3.9 2.7

1960-65

1966-70

-1.36

- 1.92

Altimir

Primarya Secondaryb Tertiary“ Total

Unikel Agriculture Industry Services Total

3.11

4.88 1.40

2.80 4.05 1.33

Sources: Unikel (1978); and Altimir (1974), cited in Gregory (1986, 30)

aAgriculture, livestock, forestry, and fishing. bMining, petroleum, manufacturing, construction, and electric power generation ‘Commerce, finance, transportation, communications, government, and other services.

407

Mexico/Chapter 2

estimates, employment growth in the high-wage industrial sector accelerated in the sixties and, at an annual average rate of 5.2 percent, was the highest of the sectoral figures. Unikel’s estimate of employment growth is much lower, but is still well above the growth rate of the labor force.

2.3 Another Look at the Record of Stabilizing Development In the initial, quick examination in section 2.1 of the macroeconomic data for 1958-70, I observed that the record of SD with respect to growth of aggregate output, growth of real industrial sector wages, investment, and inflation was impressive. Many students of Mexican economic history, however, are of the view that a more detailed investigation reveals that the SD program was inherently flawed. Conventional wisdom holds that starting sometime around the mid-sixties the Mexican economy was beset by a host of intractable problems: 1 Inadequate employment growth. Underemployment is alleged to have worsened as a result of policies aimed at stimulating investment, which made capital relatively cheap and encouraged firms to use less labor-intensive technologies, and the protectionist trade regime, which promoted the capital-intensive, import-substituting industrial sector at the expense of the labor-intensive agricultural sector. 2. A worsening distribution of income.’ Neglect of agriculture and inadequate employment growth meant that the poorest groups gained little in the growth achieved under SD. 3. Progressive loss of Jiscal control.* Concern about the deteriorating distribution of income created pressure to increase social welfare expenditures, leading to a sharp increase in overall public sector spending in the last half of the sixties. Due to an earlier failure to achieve any significant tax reform, revenue growth could not keep pace and the fiscal deficit started rising, climbing from 0.9 percent of GDP in 1965 to 3.8 percent in 1970 (tables 2.8 and 2.9). The larger fiscal deficits, in turn, caused the payments balance to deteriorate, and by 1970 the current account deficit had reached the unprecedented figure of $1.19 billion. 4. Diminishing growth p ~ t e n t i a lIt. ~is often claimed that the economy began to lose steam after 1965 when growth in agricultural output declined steeply and the opportunities for “easy” and efficient import-substitution had been largely exhausted. I am unable to find much support for this critique. Most of the critique, if not incorrect, rests on very shaky foundations. 2.3.1

The Distribution of Income

Utilizing data from various household-expenditure surveys dating back to 1950, numerous studies have been made of how the distribution of income

408 Table 2.8

Edward E Buffie Public Sector Revenues and Expenditures (or0 of GNP)

Expenditure Current Interest on foreign debt” Other Capital Revenues Economic deficit Deficit on financial intermediationb Monetary deficit

I965

1966

1967

1968

I969

1970

18.8 15.1 .4 14.6 3.7 18.0 .8 .1 .9

18.4 14.9 .5 14.4 3.5 17.3 1.1 .1 1.2

19.7 14.7 .6 14.1 5.0 17.5 2.2 .2 2.4

19.6 14.5 .7 13.8 5. I 17.7 1.9 .3 2.2

20.0 14.6 .7 13.9 5.4 18.1 1.9 .3 2.2

22.3 15.3 .8 14.5 7.0 18.9 3.4 .3 3.8

Source: Estadisticas Hacendarias del Sector Publico: Cifras Anuales, 1965- 1982 (SHCP). “Estimated by multiplying public sector interest payments on the foreign debt by the period average controlled exchange rate. bDeficit of La Banca de Desarrollo Table 2.9

Breakdown of the Fiscal Deficit (9% of GDP)

PEMEX Expenditure Cument Capital Revenues” Deficit Non-PEMEX parastatalsb Expenditure Current Capital Revenues” Deficit Othef Expenditure Current Capital Revenuesa Deficit

1965

1966

1967

1968

1969

1970

3.0 2.6 .4 3.2 - .3

3.1 2.7 .4 3.2

2.9 2.1 3.5 - .5

2.9 2.3 .5 3.4 - .5

2.8 2.4 .4 3.4 - .6

2.7 2.3 .4 3.3 - .6

6.0 .8

6.8 6.2 .6 5.5 1.3

7.2 6.2 1.o 6.0 1.2

7.0 5.8 1.2 5.8 1.2

7.2 6.0 1.2 5.9 1.3

9.9 6.3 3.6 6.8 3. I

9.1 6.2 2.9 8.8 .3

8.5 6.0 2.5 8.6 p.1

9.5 6.3 3.2 8.0 1.5

9.8 6.4 3.4 8.5 1.3

10.0 6.2 3.8 8.8 I .2

9.7 6.7 3.0 8.8 .9

6.8 6.3 .5

0

.a

Source: Estadisticas Hacendarias del Sector Publico: Cifras Anuales, 1965- 1982 (SHCP)

’Sum of revenues and taxes paid. bBudget- and nonbudget-controlled parastatal enterprises ‘Includes DDF (Department of the Federal District).

evolved during SD. The general conclusion reached by these studies is that the distribution of income worsened significantly. Tables 2.10 and 2.11 show how the distribution of income by decile and several standard distributional measures varied from 1950 to 1977. All of the distributional measures (except possibly Atkinson’s Coefficient) suggest a marked increase in inequality between 1958 and 1970. According to table 2.10, the middle classes and the very rich were the main beneficiaries of growth. The income

409

Mexico/Chapter 2 Distribution of Income by Deciles

Table 2.10 Decile

I

II 111

IV V

VI VII VIII

Ix

X

Total

1950 2.43 3.17 3.18 4.29 4.93 5.96 7.04 9.63 13.89 45.48 100.00

1958 2.32 3.21 4.06 4.98 6.02

7.49 8.29 10.73 17.20 35.70 100.00

1963

1968"

1970

1975

1977

1.69 1.97 3.42 3.42 5.14 6.08 7.85 12.73 16.45 41.60 100.00

1.21 2.21 3.04 4.23 5.07 6.46 8.28 11.39 16.06 42.05 100.00

I .42 2.34 3.49 4.54 5.46 8.24 8.24 10.44 16.61 39.21 100.00

0.69 1.28 2.68 3.80 5.25 6.89 8.56 8.71 17.12 45.02 100.00

1.08 2.21 3.23 4.42 5.73 7.15 9.11 11.98 17.09 37.99 100.00

Source: Hernandez and Cordova (1979, 443), cited in Gallardo (1983, 2240).

"Revised data from the Bank of Mexico.

share of the sixth, seventh, and top deciles rose, while the share of all remaining deciles fell. Notice, however, that any worsening that may have occurred in the distribution of income took place entirely in the very short period from 1958 to 1963. Between 1963 and 1970, the various income distribution measures either remain stable or improve. The sharp reversal after 1963 raises doubts about how much of the overall deterioration in the distribution of income over 1958-70 can be attributed to the policies of SD. g o n e accepts that there was a strong causal link, one can say from the data either that the distribution of income had ceased to deteriorate by 1963 or that the initial deterioration caused by SD policies was temporary and after 1963 was in the process of being reversed. The conclusion that the distribution of income deteriorated between 1958 and 1970 is also open to question. Apart from problems in the quality of the data across surveys, the summary income distribution measures may be biased. It is well known, for example, that when individuals change their position in the income distribution ranking, the Gini coefficient and the income share of the poorest may suggest that the distribution of income has worsened when in fact it has unambiguously improved.lo This would appear to be a potentially serious problem in the Mexican case, for, as noted earlier, employment growth in the industrial sector was well above the growth rate of the labor force during SD. The resulting substantial transfer of labor from low- to high-wage sectors of the economy probably enabled many of the (formerly) poor to move up the income distribution ladder. The latter observation suggests one final point. Regardless of how the overall distribution of income may have changed, it seems the poor benefitted substantially in absolute terms from the high rates of growth achieved under SD. From the data in table 2.10 one can infer that real incomes of the poorest 40 percent of the population grew at an annual

Table 2.11

Inequality Indices Atkinson's Coefficient of Inequality

1950 1958 1963 1968 1970 1975" 1977

E

Gini

Theil

E = .5

S16 ,450 ,527 ,526 .496 370 ,496

.748 ,406 ,494 ,488 ,498 ,556 ,426

-

-

-

.20 .I6

.45 .42

.62 .62

~

-

.I7

=

1.5

-

-

.44

E

=

3.0

~

-

.78

Richest 20%

Middle 30%

Poorest 40%

Poorest 40%

Ab

B'

4.5 3.6 5.5 5.4 4.7 7.3 5.0

1.37 1.49 1.81 1.85 1.86 2.45 2.01

60 45 35 30d

-

% of poar families

-

-

-

63.2 48.6 49.5 -

Sources: Measure A of the percentage of poor families is from van Ginneken (1980,19). Measure B i s based on estimates made by a World Bank mission and is from Solis (1981,147). The estimates of Atkinson's Coefficient are from Aspe and Beristain (1984b,45).The other income distribution measures are from Gallardo (1983,2241).

'For the per capita income distribution. bPoverty line is 10,000 pesos per year. +Poverty line is the 1975 minimum wage. dThe figure is for 1969. T h e data in the 1975 Income-Expenditure Survey are known to be unreliable.

411

Mexico/Chapter 2

average rate of 5.8 percent during the SD era.” According to a study by van Ginneken (1980), the percentage of families living in poverty declined from 45 percent in 1958 to 30 percent in 1969.’*

2.3.2

Underemployment

The claim that underemployment worsened during SD strikes me as particularly weak. First, the assertion that decreases in real energy prices and in the user cost of capital slowed employment growth is theoretically dubious. Lower prices for energy and capital might induce firms to adopt less labor-intensive technology, but they also raise the profit-maximizing level of output. From production theory and most empirical estimates one can argue that normally the favorable output effect on labor demand will dominate the adverse substitution effect. Factors of production, in other words, tend to be gross complements so that reductions in the cost of capital and energy would be expected to raise, not lower, the rate of employment growth. Observations that the capital and energy intensity of production increased are beside the point; it is precisely the greater utilization of cooperating, complementary factors that enhances labor productivity and expands labor demand. The thrust of my analysis so far has been that the SD policies promoted employment growth in the high-wage industrial sector. This is not, of course, sufficient to rule out the possibility that underemployment worsened during SD. Indeed, there is a sizable school of thought which contends that while industrial sector employment growth was respectable, it was achieved at the cost of generally stagnant employment. Shrinking employment opportunities in agriculture, it is asserted, caused a large increase in migration out of rural areas. Only a small fraction of the rural migrants could be absorbed by expansion in the capital-intensive industrial sector; the remainder spilled over into the low-productivity informal sector. The employment data, unfortunately, are inconclusive on this point. Different stories emerge from the different methods various authors use to adjust the employment data in the 1960 and 1970 population censuses. Referring back to table 2.7, Unikel’s estimates show aggregate employment growth lagging behind the growth rate of the labor force. Altimir’s estimates, on the other hand, show not only much greater rates of employment growth in agriculture and industry but also a doubling in the growth rate of labor productivity in the tertiary sector from the fifties to the sixties. On balance, it seems the evidence lends greater support to the view that the SD policies succeeded in greatly reducing the extent of underemployment. Altimir’s estimates are corroborated by a number of other findings which suggest that labor demand grew very strongly throughout the sixties. Gregory (1986) reviews the data on wages and productivity in the informal sector and concludes that they strongly contradict the hypotheses that (in the sixties): (a) low wages and low productivity

412

Edward F. Buffie

generally characterize the informal sector and (b) the large shift of labor out of agriculture depressed informal sector incomes. Labor productivity increased strongly in the service sector, and in many branches wages exceeded the minimum wage in 1970 (34-49). Table 2.12, which is constructed from Gregory’s tables 7.4 and 7.5, shows that real wages and labor productivity increased substantially across all size establishments in industry, commerce, and services. Migration studies provide additional evidence of improving employment opportunities.l 3 According to anthropological studies and sample surveys of migrants in Mexico City and Monterrey, most migrants found employment very quickly and viewed migration as having substantially improved their standard of living. In addition, the studies do not confirm the notion that migrants flooded into the informal sector. The share of migrants taking their first job in the tertiary sector declined in each succeeding decade from 1930 Table 2.12

Average Annual Growth Rates of Total Real Remunerations and Net Value-Added per Employee, 1960-70s

Sector and Size of Establishment

Average Remuneration per Employee

Net Valuc-Added per Employee

No paid employees Industry Services Commerce Establishments with paid employees Small Industry, 1-5 workers Services, 1-2 workersb Commerce, 1-2 workers Medium Industry, 6-25 workers Services, 3-8 workers‘ Commerce, 3-8 workers Large Industry, 26- 100 workers Industry, 101-500 workers Industry, >500 workers Services, >9 workersC Commerce, >9 workers

5.0

-

-

3.3 6.1 2.4

4.3

2.6 2.9

6.7

1.5

-

-

1.2 3.4 4.4 6.1 1.5

Sources: The growth rates of real remunerations are calculated from Gregory (1986), table 7-3 (232). Net value-added per employee is calculated by deflating Gregory’s estimates of nominal value-added in table 7-5 (240) by a price deflator for industrial sector value-added. The deflator was constructed by forming a weighted average of the price deflators for the manufacturing, construction, mining, and electricity sectors, where the respective weights were given by the sector’s share in total industrial value-added in 1960. aTotal remunerations are wages and salaries plus fringe benefits and payroll taxes. Nominal remunerations are deflated by the CPI for Mexico City. b l - 3 workers in 1960. ‘4-10 workers in 1960 d l 1 or more workers in 1960.

413

Mexico/Chapter 2

to 1970; in the sixties, 56 percent of migrant unskilled workers went into the industrial sector. 2.3.3

Fiscal Discipline

The claim that fiscal discipline began to break down during the Diaz Ordaz administration in the wake of political unrest and pressures to increase social welfare expenditure does not appear to be any better founded than the claims that SD had adverse repercussions on the distribution of income and employment growth. The share of public investment devoted to social welfare did increase during the Diaz Ordaz administration, but the increase was far smaller than in the preceding L6pez Mateos administration. Moreover, while 1965-70 was a time of considerable social and political tension, so also was the 1958-64 period. If Diaz Ordaz had to contend with a students’ strike in 1965 and student riots in 1968, L6pez Mateos faced the railroad strike and the teachers’ strike in 1959, a rural guerilla campaign in 1962, and a physicians’ strike in 1964 (which almost brought down the government). If there was no weakening of fiscal discipline, what then accounts for the steady increase in the fiscal deficit from 0.9 percent in 1965 to 3.8 percent in 1970? A quite plausible answer is that the growth of the deficit reflected nothing more than the normal workings of a well-defined political expenditure cycle. l4 Table 2.13 displays the results of regressing the detrended values of current, capital, and total government expenditure for 1965-85 on six dummy variables (01-06) corresponding to the six years making up the presidential term. Serial correlation was tested using the limits for the Durbin-Watson statistic developed by Farebrother (1980) for regression Table 2.13

The Political Expenditure Cycle Total Public Sector Expenditure

Current Expenditure

Capital Expenditure

DI

- .076

- ,026

03

-.I51

04

i.46) - ,016

(.63) - ,067 (1.61) - ,038 i.92) - .002 i.04) ,082 (1.62) .091 (2.18) .70 .47 1.47

-.I9 (4.24)

D2

(2.32) - ,063 (1.93)

D5 06 R2

R Z

Durbin-Watson statistic

i. 15)

.I2 (2.98) ,079 (2.40) .71 .56 1.22

- .04

(.W) ,043

(. 96)

,059 (1.32) 218 (3.96)

.80 .65 1.81

414

Edward F. Buffie

equations without a constant term. In those cases where the Durbin-Watson value fell in the indeterminant range, Bartlett’s (1946) test was then applied as a second check for serial correlation. In none of the regressions was there evidence of first-order serial correlation. It is clear from table 2.13 that fiscal policy follows a very distinct cycle. For total government expenditure, the dummy variables are negative and significant in the first two years and positive and significant for the last two years. Capital expenditures exhibit a significant decrease in the first year and a significant increase in the fifth year, while current expenditures show a nearly significant decrease in the second year and a significant increase in the final year. The expenditure cycle seems to stem from both the perceived political advantages of increasing expenditures shortly before elections and the incongruity between the natural gestation period of investment projects and the fixed, six-year term (sexenio) of each administration (reelection is not allowed). Fiscal expansion invariably occurs in the two years preceding the upcoming election. Capital spending first increases strongly in the fifth year in the rush to complete investment projects before the term of the existing administration expires. In the following year, spending surges again as current expenditures rise in the campaign to strengthen political support just before the election. Immediately after the election, spending falls sharply as capital expenditures temporarily decline while a new set of investment projects are being designed and the new administration strives to reduce the fiscal deficit. Fiscal control then prevails until the fifth year when the cycle starts to repeat itself. Returning to the issue of fiscal discipline in the latter part of the SD period, since 1965 was the first year of the Diaz Ordaz sexenio, the increase in the fiscal deficit between 1965 and 1970 was not at all out of the ordinary. The relevant comparison is between the fiscal deficits of 1964 and 1970. This comparison does not support the notion of mounting fiscal problems. In both years, the deficit was approximately 4 percent of GDP. 2.3.4 . Diminishing Growth Momentum? Finally, I also disagree with the claim that the economy’s growth momentum began to decline after the mid-sixties. Much has been made of the drop in the growth rate to 4.2 percent in 1971.I5 But this drop is readily explained by the fiscal retrenchment that occurs in the first year of the political expenditure cycle. Table 2.14 confirms the expectation that the expenditure cycle is associated with a similar cycle for real GDP.I6 The difference between the actual growth rate in 1971 and the fitted value of the model is only 0.0012 and is not statistically significant (the SEE is 0.0328). Thus, the 1971 slowdown was hardly unusual. Concerning the pattern of agricultural output, the high rates of growth between 1945 and 1965 were based on the development of large-scale

415 lsble 2.14

MexicoIChapter 2 The Political Expenditure Cycle and Real Output, 1940-85’ Real GDP DI

- .02

02

(1.85) - .01

03

- .006

04

(.57) ,015 (1.30)

(.89)

D5

.021

06

(1.75) .005

R2 = -

.52

(.W

R2 = .45 DW = 1.06 %statistics are in parentheses.

irrigation schemes in the northwest that improved existing lands or brought vast amounts of new land under cultivation. By 1965 this source of growth had been largely exhausted.17 Agricultural growth fell off sharply after 1965 because of political constraints on land redistribution that prevented investment to develop the more populous, rainfed agricultural areas, not because SD entailed “neglect” of the agricultural sector. Furthermore, despite the deceleration in agricultural growth, overall growth remained satisfactory owing to the strong performance of the industrial sector. Industrial sector productivity continued to grow at an impressive rate, the investment share in GDP increased 2.4 percentage points (measured at 1960 prices), and total output growth averaged 6.8 percent during the Diaz Ordaz administration. The continuing dynamism of the industrial sector would seem to belie the claim that the process of import substitution was encountering increasing difficulties. The argument that much of this growth was inefficient because it was achieved under a protectionist trade regime is also quite dubious. Free trade is not necessarily optimal if distortions are present and cannot be remedied by the imposition of appropriate lump-sum taxes and subsidies. Recent theoretical work, in fact, suggests that an escalated structure of protection is an appropriate (second-best) policy when either the level of private investment falls short of the socially desired level or the labor market is distorted by wage rigidity in the industrial sector.’* Given the sizable gap between wages in the industrial sector and the informal and agricultural sectors, the record of sustained growth in industrial productivity, and the moderate nature of Mexican protection, it is difficult to construct a strong case for the view that import-substituting industrialization was inefficient during SD. The one obvious flaw in the trade regime was the relatively high

416

Edward F. Buffie

degree of protection granted to the domestic capital goods sector. A trade regime that did not protect this sector would have been more effective in stimulating economywide capital accumulation.

2.4

Concluding Observations

Thc SD era was one of the most successful periods of Mexican economic development. A remarkable degree of macroeconomic stability prevailed at the same time that annual GDP growth averaged 6.7 percent. In my view, there is no firm evidence that the accomplishments of high growth and macroeconomic stability were tainted by a worsening in either underemployment or the distribution of income. In fact, some evidence points to the opposite conclusion or, at the very least, to the conclusion that the record of SD was adequate on these two counts. Labor demand grew strongly in the high-wage industrial sector. Even Unikel's pessimistic estimates show industrial sector employment expanding at an annual clip of 3 percent despite real wage growth averaging over 4 percent. And while employment studies yield disparate conclusions about the growth of aggregate employment, wage and productivity data suggest that substantial progress was achieved in reducing the extent of underemployment. Distributional studies are plagued by problems in the comparability of data at different points in time and likely biases in the summary measures of the income distribution. Putting these reservations aside, the data, such as they are, show that inequality increased between 1958 and 1970. But the worsening in the distribution of income occurred entirely in the 1958-63 subperiod; after 1963, the distribution of income improved. This peculiar pattern, coupled with the substantial transfer of labor from low- to high-wage activities, makes one suspicious of the claim that a heightening of income inequality was inherent in SD policies. Furthermore, though the distribution of income may have deteriorated and SD policies may have been partially to blame, it also seems that in absolute terms the poor reaped substantial gains. Average real income of the poorest 40 percent of the population increased 97 percent, and the percentage of families living in poverty greatly declined. l9 The SD period was marred by numerous outbreaks of social unrest. It is hardly clear, however, that these outbreaks had much, if anything, to do with the economic policies of SD. The growth of social discontent reflected principally the dissatisfaction of the middle classes at being excluded from the political process. In earlier years, political hegemony had been maintained by co-opting the growing middle class into either the government or party bureaucracy. By 1960 the middle class was simply too large to be placated in this fashion and the political consensus began to unravel.20 Some authors (e.g. Tell0 1979) contend that the social unrest of the sixties can be traced to increased underemployment and a deterioration in the distribution of income which adversely affected the welfare of the middle

417

Mexico/Chapter 3

classes. But the claim that underemployment worsened is difficult to substantiate, and the data in the income-expenditure surveys contradict the notion that economic factors underlay middle class dissatisfaction. The income share of the middle classes increased in each succeeding survey (1958, 1963, 1968, 1970) and rose far more over the 1958-70 period than that of any other group. In vigorously defending the record of SD, I am not saying that policy mistakes were not made. Tax reform, less rapid real wage growth in the industrial sector, and greater efforts at promoting agricultural development would, I believe, have led to greater reductions in underemployment and a more equitable distribution of income. Overall, however, SD worked and worked well.

3

Shared Development and the Echeverria Administration

The presidential campaign of Luis Echevem’a generated great enthusiasm and high hopes among the general population. Echevem’a crisscrossed the country, exhibiting a level of political energy not seen since the days of Lkzaro C6rdenas in the thirties. He repeatedly stressed two basic themes in his campaign: prevention of another social conflict like that of 1968 and preservation of the fixed exchange rate of 12.5 pesos per dollar. The first objective reflected Echevem’a’s intention to achieve a reconciliation with the young and the middle class. The second signalled a commitment to perpetuate the successful financial system inherited from SD. Although the Ministry of Finance and Bank of Mexico were placed under the direction of professionals who had served the two preceding administrations, the economic program of SD was rejected as having done too little to reduce underemployment and improve the distribution of income. It was announced that henceforth the government would take a more active role in ameliorating social ills-that is, in promoting “Shared Development.” The initial economic program proposed six measures to foster Shared Development and reduce the large current account deficit of 1970:’ 1 . Increase the supply of credit to, and government investment in, the agricultural sector. 2. Replace licenses by tariffs, eliminate tax rebates given to the industrial sector, and redirect trade policy toward export promotion. 3. Increase government revenues by raising public sector prices, by tax reform, and by a reduction of tax evasion.

417

Mexico/Chapter 3

classes. But the claim that underemployment worsened is difficult to substantiate, and the data in the income-expenditure surveys contradict the notion that economic factors underlay middle class dissatisfaction. The income share of the middle classes increased in each succeeding survey (1958, 1963, 1968, 1970) and rose far more over the 1958-70 period than that of any other group. In vigorously defending the record of SD, I am not saying that policy mistakes were not made. Tax reform, less rapid real wage growth in the industrial sector, and greater efforts at promoting agricultural development would, I believe, have led to greater reductions in underemployment and a more equitable distribution of income. Overall, however, SD worked and worked well.

3

Shared Development and the Echeverria Administration

The presidential campaign of Luis Echevem’a generated great enthusiasm and high hopes among the general population. Echevem’a crisscrossed the country, exhibiting a level of political energy not seen since the days of Lkzaro C6rdenas in the thirties. He repeatedly stressed two basic themes in his campaign: prevention of another social conflict like that of 1968 and preservation of the fixed exchange rate of 12.5 pesos per dollar. The first objective reflected Echevem’a’s intention to achieve a reconciliation with the young and the middle class. The second signalled a commitment to perpetuate the successful financial system inherited from SD. Although the Ministry of Finance and Bank of Mexico were placed under the direction of professionals who had served the two preceding administrations, the economic program of SD was rejected as having done too little to reduce underemployment and improve the distribution of income. It was announced that henceforth the government would take a more active role in ameliorating social ills-that is, in promoting “Shared Development.” The initial economic program proposed six measures to foster Shared Development and reduce the large current account deficit of 1970:’ 1 . Increase the supply of credit to, and government investment in, the agricultural sector. 2. Replace licenses by tariffs, eliminate tax rebates given to the industrial sector, and redirect trade policy toward export promotion. 3. Increase government revenues by raising public sector prices, by tax reform, and by a reduction of tax evasion.

418

Edward F. Buffie

4. Introduce tax incentives to encourage employment growth and greater dispersion of industrial activity. 5. Develop new tourist sites as an additional source of foreign exchange earnings. 6. Improve the “efficiency” of government current expenditures. This program was supported by the new Minister of Finance, Hugo Margain, but quickly came under fire from other cabinet members who felt that the pursuit of Shared Development required a more aggressive approach. Elements of the program would be resurrected and then abandoned again in 1977 and 1983 in the face of opposition from large segments of the bureaucracy, labor, and industrialists operating in the protected industrial sector. Shared Development got off to a slow start in 1971. Contractionary fiscal and monetary measures were adopted since restoration of external balance was the paramount concern, Capital expenditures of the federal government were cut 12 percent in nominal terms, the domestic price of sugar was raised 48 percent, and a surtax on luxury goods was imposed. The nominal monetary base increased 18.8 percent, but this reflected the desired improvement in the balance of payments; Central Bank credit to the public sector actually decreased. Contractionary policy produced a mild recession. GDP growth slowed to 4.2 percent, and excess bank reserves rose to 2.6 billion pesos. Though the recession was not particularly severe and was predictable in light of the political business cycle, it considerably strengthened the hand of those militating for a much more expansionary fiscal program. The Ministries of the Presidency and of National Patrimony vigorously attacked the restrictive monetary policy of the Central Bank and the low-level budget proposed for 1972 by the Ministry of Finance. A large increase in public investment and easier monetary policy were essential, they argued, to pull the economy out of the recession and make progress toward the goals of Shared Development. President Echevem’a concurred, and starting in 1972 economic policy changed radically.

3.1 1972-76: Public-Expenditure-Led Growth and the First Debt Crisis The Minister of Finance counseled President Echevem’a that a tax reform increasing revenue growth was imperative if the ambitious investment program planned by the government was to be financed soundly. In 1972 the Undersecretary of Revenue, Gustavo Petricioli, put forward a program involving reform in three key areas: 1. Reduction of tax evasion, especially evasion of the corporation income tax. If evasion could be lessened, a cut in rates would be compatible with an increase in total revenue.

419

MexicolChapter 3

Definition of a broad taxable base without preferential treatment for any group or individual. The tax base would include all income regardless of its source. To solve the problem of anonymity, securities issued to the bearer would pay the maximum marginal rate. Otherwise, interest and dividend income would have to be declared and accumulated to other income. Greater revenue generation by state-owned enterprises (SOEs). Many SOEs had not adjusted their prices in the last ten years. Revenue stagnation was a major problem for PEMEX (oil), CFE (electricity) and DDF (Department of the Federal District). The underlying issues were the same as those Ortiz Mena had wrestled with in 1964-65. The private sector staunchly opposed the reform, greatly fearing the loss of anonymity and the possibility of a wealth tax. In policy circles there was apprehension that any reform would provoke capital flight and a severe devaluation and, in particular, that the elimination of bearer securities would cause financial panic and a collapse of the banking system.2 After many heated debates, President Echevem’a decided on 26 December 1972 to kill the tax reform bill. Shortly afterward, the Minister of Finance, Margain, and his Undersecretary, Petricioli, were fired. The revenue-raising effort in 1973 was limited to an increase in the excise tax rate from 2.8 to 4 percent and the imposition of a 15 percent surcharge on luxury goods. During the same period in which it suffered defeat on tax reform, the Ministry of Finance lost a second important battle. Since the late nineteenth century, the Ministry of Finance had been responsible for monitoring both government expenditure and revenue collection. Beginning in 1973, the Ministry’s responsibility for controlling outlays was limited to current expenditures. Control of public investment expenditures, and hence control over the bulk of discretionary spending, was shifted to the Ministry of the Presidency. Increasingly, the Ministry of Finance was forced to relinquish control over the spending process as more and more programs were directly approved by the president himself. There was no pretense to the contrary. After firing Margain as Minister of Finance, President Echevema stated bluntly that “the national finances are handled from Los Pinos” (the presidential residence). Under this new institutional arrangement, the limited revenue base no longer deterred expenditure growth. At the same time, Echeverria came under intense political pressure to increase spending4 From the thirties onward, the presidential successor had emerged as the compromise choice of the various factions constituting the ruling political elite. By providing the new president with a solid base of political support at the beginning of his term, this consensual system made it much easier to conduct a responsible, clear-cut economic policy. The era of strong presidents came to an end in 1970. Diaz Ordaz’s brutal handling of the student riots in 1968 produced a deep political schism. For the first time since the selection of Plutarcho Calles in 1924, the political elites could not

420

Edward F. Buffie

agree on a compromise candidate, and Diaz Ordaz had to choose his successor unilaterally. Hence, in contrast to his predecessors, Echevem’a faced the difficult task of creating his own supporting coalition after assuming office. The simplest method of shoring up the weakening political consensus was to spend on everyone’s behalf dole out subsidies to education and agriculture, increase government jobs for the middle classes, grant large wage increases to mollify organized labor, etc. The inflationary repercussions of greater spending were not a grave concern; reestablishing the political consensus was. The alternative, in Echevem’a’s view, was “fascism” (Newell and Rubio 1984, 204). After retrenchment in 1971, succeeding years saw enormous fiscal expansion. Total public sector expenditure increased from 20.5 percent of GDP in 1971 to 32 percent in 1976. Much of the increased spending took place in the parastatal sector. Between 1971 and 1975, SOEs increased their real current expenditures at an annual average rate of 18 percent and their real capital expenditures at a rate of 29.3 percent.’ New, large-scale industrial projects accounted for most of the expansion in the parastatal sector. During this period, the steel mills of Lizaro Ckdenas-Las Truchas, several industrial seaports, and the petrochemical complex of La Cangrejera were constructed and many of the most important PEMEX investments were carried out. Expenditure increased strongly in other branches of the government as well. Outside the parastatal sector, expenditure growth reflected principally the undertaking of expensive investment programs and a general policy of expanding government employment. Between 1970 and 1976, the number of federal government employees doubled and the growth rate of general government employment averaged 10.8 percent. A series of large wage hikes after 1972 further inflated the government wage bill (table 3.1). The wage hikes combined with growth in new hirings resulted in a 101 percent increase in the real public sector wage bill during the Echevema sexenio. Much of the Table 3.1

Public Sector Employment and Wage Payments

General government employment W change Total real public sector wages and salaries‘ W change I of GDP

1970

1971

1972

1973

1974

1975

1976

100.0

109.5 9.5

123.2 12.5

134.7 9.3

149.4 10.9

168.4 12.7

185.0 9.9

100.0

111.6 11.5 6.1

120.3 7.9 6.0

143.1 18.9 6.5

146.2 2.2 6.3

176.1 20.4 7.2

200.7 14.0 7.6

-

5.7

Sources: For 1970-75, the data on general government employment is from the National Income Accounts (MCxico, D.F.: INEGI), vol. 6. The 1976 figure is from the National Income Accounts, Cuentas de Produccion del Sector Publico (INEGI). Public sector wages and salaries are from Estadisticas Hacedarias del Sector Publico: Cifras Anuales, 1965-1982 (SHCP).

‘Deflated by the period average CPI.

421

Mexico/Chapter 3

increase came in the administration’s last four years. Even in 1976 when efforts were made to restrain spending, real public sector wage payments rose 14 percent. The priority afforded to expanding government employment and to the investment projects of the SOEs came at the expense of investment in agriculture and social welfare. Table 3.2 shows how the sectoral composition of public investment varied over the Echeverria sexenio. Expenditure by CONASUPO (the agency in charge of price supports and subsidies for agriculture) and investment in agriculture increased strongly from 1973 to 1975, but when financial pressures became acute in 1976, agriculture’s share of public investment funds was slashed from 18.1 to 13.9 percent. Spending on social welfare programs increased until 1973 and then declined very rapidly as expansion in the parastatal sector gathered steam. By 1976 the share of public investment allocated to social welfare was barely one-half its 1970 level. In light of table 3.2 and his unwillingness to press energetically for tax reform in 1972, it is difficult to comprehend Echeverria’s reputation as a “populist.” The populist image apparently derives largely from Echeverria’s rhetoric and his lenient approach to public finance issues. An examination of the fiscal record reveals that Echevem’a was a statist, not a populist; his main priority was to increase the role of government in the economy, both as a source of employment and as a producer of goods and services.6 Despite the rejection of tax reform in 1972, public sector revenue growth outpaced GDP growth during the Echeverria term (table 3.3). Direct taxes increased rapidly as inflation pushed a larger portion of the population into higher tax brackets (the phenomenon of “bracket creep”). The increase in the excise rate to 4 percent and the introduction of a 15 percent surcharge levied on luxury goods raised the average real growth rate of indirect taxes to 12.3 percent after 1972. The high growth in revenues of the SOEs, however, is deceiving, reflecting as it does simply the enormous expansion in the parastatal sector. Until 1974, public sector price increases were kept very low as part of a deliberate policy to repress inflation. Sectoral Composition of Public Investment

Table 3.2

Year

1971 1972 1973 1974 1975 1976 1971-76

Agriculture

Industry

Communications Transport

Social Welfare

Other

Total

14.57 14.85 14.13 16.92 18.09 13.89 15.41

41.65 34.47 32.54 36.02 41.51 45.99 38.70

20.49 23.65 25.38 23.98 20.70 19.I7 22.23

21.67 23.07 25.75 20.75 16.47 14.51 20.37

1.62 3.96 2.28 2.33 3.23 6.44 3.29

100

Source: Estadisricas Historicar de Mdxico (Mtxico, D.F.: INEGI. 1985)

100 100 100

100 100

100

422 Table 3.3

Edward F. Buffie Public Sector Prices and Revenues.

Total public sector revenue General tax revenue Income taxes Personal Indirect taxesb Revenues of parastatal sector‘ Real public sector pricesd

1970

1971

1972

1973

1974

1975

1976

18.9 8.1

18.4 8.0

18.7 8.1

3.5

3.1

1.5 4.6 9.6

1.6 4.6 9.7 96.2

1.7 4.4 9.7 94.0

21.1 8.9 4.0 1.7 4.9 11.5 106.6

23.1 9.7

3.5

20.2 8.9 3.8 1.7 5.1 11.0 88.7

23.8 10.3 4.9 2.3 5.5 12.7 104.3

100.0

4.5 2.0 5.3 13.1 110.6

Sources: The real public sector price series is from Clavijo (1980). All other figures are generated from data in Estadisticas Hacendarias del Sector Publico: Cifras Anuales, 1965-1982 (SHCP).

“Revenues are expressed as a percentage of GDP. General tax revenues and parastatal revenues do not sum to total revenues because of other, unclassified revenues and because some tax receipts are tax payments made by the parastatal sector. bThe sum of value-added taxes, taxes on production and services, taxes on foreign trade, and “other” tax revenues. Does not include gasoline taxes (which I classify as revenues of PEMEX). ‘Sum of revenues (exclusive of any transfer payments received) plus taxes paid. dPeriod average price deflated by the period average CPI.

Relatively strong revenue growth was not sufficient to prevent the fiscal deficit from rising rapidly. Enormous increases in government spending drove the consolidated public sector deficit upward from 2.5 percent of GDP in 1971 to 10 percent in 1975. As can be seen from tables 3.4 and 3.5, the loss of fiscal control was widespread. The growth in the deficit can be attributed in almost equal parts to increased losses of the non-PEMEX SOEs and larger deficits run by the nonparastatal sector (“deficit on financial intermediation” and “other”). Unlike during SD, the fiscal deficits were financed in large measure by borrowing from the Central Bank. Table 3.6 contains information on how the main monetary aggregates behaved during the Echevem’a term. The growth rate of the monetary base accelerated from 19.6 percent in 1971 to 33.8 percent in 1975, and the share of seignorage in GDP rose to triple the Table 3.4

Public Sector Revenues and Expenditures (% of GNP)

Expenditure Current Interest on foreign debt Other Capital Revenues Economic deficit Deficit on financial intermediationa Monetary deficit

1970

1971

1972

1973

1974

1975

1976

22.3 15.3 .8 14.5 7.0 18.9 3.4 .3 3.8

20.5 15.3 .8 14.5 5.2 18.4 2.1

22.9 16.2 .7 15.5 6.8 18.7 4.2 .7 4.9

25.8 19.0

27.0 20.6 1.0 19.6 6.4 21.1 6.0 1.3 7.2

31.9 23.7 1.2 22.5 8.2 23.1 8.8 1.2 10.0

32.0 23.6 1.5 22.1 8.4 23.8 8.3 1.6 9.9

.4

2.5

.8 18.2 6.7 20.2 5.6 1.2 6.9

Source: Estadisticas Hacendarias del Sector Publico: Cifras Anuales, 1965-1982 (SHCP).

‘Deficit of La Banca de Desarrollo.

423 Table 3.5

MexicoIChapter 3 Breakdown of the Fiscal Deficit (96 of GDP)

PEMEX Expenditure Current Capital Revenues" Deficit Non-PEMEX parastatakb Expenditure Current Capital Revenues" Deficit Other' Expenditure Current Capital Revenues" Deficit

1970

1971

1972

1973

1974

1975

1976

2.7 2.3

2.7 2.0

4.1 2.9

.9

1.1

3. I -.4

3.0 2.1 .9 2.8 .2

3.2 2.3

3.3 -.6

3.0 2.2 .8 3.2 -.2

4.0 - .8

4.1 0

3.6 2.0 1.6 4.0 - .4

9.9 6.3 3.6 6.8 3.1

8.7 6.9 1.9 6.7 2.0

8.6 6.8 1.8 7.0 1.6

10.4 8.5 2.0 8.0 2.4

12.2 9.4 2.8 8.3 3.9

14.6 10.2 4.4 9.0 5.6

12.9 9.9 3.0 9.1 3.8

9.7 6.7 3.0 8.7 1.o

8.8 6.2 2.6 8.4 .4

11.6 7.3 4.3 8.6 3.0

12.4

11.6 9.0 2.6 8.8 2.8

13.3 10.6 2.6 10.0 3.3

15.5 11.6 3.8 10.7 4.8

.4

.7

8.5

3.9 9.3 3.1

Source: Estadisticas Hacendarias dei Sector Pubiico: Cifras Anuales, 1965- I982 (SHCP)

"Sum of revenues and taxes paid. bBudget- and nonbudget-controlled parastatal enterprises. 'Includes DDF (Department of the Federal District).

average level of the sixties. In the last three years (1974-76), foreign debt replaced domestic debt as the main method of financing the deficit (table 3.7), but the monetary base continued to grow strongly. A second important shift in monetary policy concerned the management of interest rates. Whereas real deposit rates were maintained at positive levels throughout SD, after 1972 this policy was allowed to lapse. Nominal interest rates were not adjusted upward in step with inflation, and as real rates turned negative, the financial miracle terminated abruptly. The total stock of real bank funds fell 13.3 percent from 1973 to 1976. Financial disintermediation, in turn, by reducing the growth of demand for bank reserves-much the largest component of the monetary base-made it far more difficult to prevent excessive growth in the high-powered money supply. The government reacted by raising the reserve ratio from 0.313 in 1970 to 0.511 in 1976. Nonetheless, real bank reserves grew at an annual rate of only 5.7 percent over 1973-76, a figure far below the 9.8 percent recorded during the preceding Diaz Ordaz administration. For a couple of years, expansionary demand policies were successful in stimulating strong output growth (tables 3.8a and 3.8b). However, problems soon began to appear. After 1972, when recovery from the 1971 recession was complete and excess capacity had largely disappeared, inflation accelerated, rising above 20 percent in 1973 and 1974. Furthermore, while aggregate growth was high from 1972 to 1974, much of the growth was concentrated in the public sector. Private investment weakened, dropping

424

Edward F. Buffie

Table 3.6

Monetary Aggregates and Real Interest Rates

Real growth rates’ Monetary base M2

M3 M4 Total stock of bank fundsb Credit to the business sectof Total credit of the banking systemd Percentage of GDP Monetary base M2

M3 M4 Total stock of bank funds Credit to the business sector Total credit of the banking system Real interest rates‘ Average real bank deposit rate8

1970

1971

1972

1973

1974

1975

1976

5.6 5.4 5.7 12.9 14.1

13.8 2.4 7.4 8.3 9.1

13.5 14.9 13.8 11.7 11.0

0.2 4.4 - 1.8 -5.9 -7.6

6.0

20.5 9.1 8.7 14.2 14.9

1.9 6.7 -9.8 - 10.2 - 15.6

10.0

13.1

9.8

- 14.8

.6

9.5

3.9

11.0

7.8

10.3

- 2.4

2.2

15.1

8.4

12.2 10.9 22.4 31.3 27.0

12.8

13.1 10.9 22.5 31.4 27.0

12.6 10.3 20.0 28.0 23.7

13.6 10.2 19. I 28. I 23.8

14.3

22.8 32.9 28.7

13.1 10.7 23.1 33.2 28.9

24.6

26.1

26.6

23.7

20.5

20.4

20.9

40.6

42.2

42.1

40.4

37.6

38.7

41.4

6.1

5.2

4.3

10.8

-11.3

- .7

-6.9 -2.4 -3.3

-9.2

.7

10.5

18.0 27.0

22.2

-

15.4

Sources: The bank deposit rate series is from CIEMEX-WHARTON, Mexican Economic Outlook (Philadelphia: Wharton Economehic Forecasting Associates, 1987). 18(4):185-86. All other data is from Indicadores Economicos (Bank of Mexico).

Notes: M2 = Currency held by the public + peso- and foreign-currency-denominated demand deposits. M3 = M2 + liquid savings accounts. M4 = M3 + nonliquid (i.e., fixed-tern) savings accounts.

‘Real monetary aggregates are calculated as the end-of-year balance deflated by the end-of-year CPI.

bM4 less currency held by the public. ‘Total credit of the commercial and development banks less credit extended to federal, state, and municipal governments. dCredit of the Central Bank, the development banks, and the commercial banks. =Average of the end- and beginning-of-year monetary aggregate relative to FDP ‘End-of-year interest rate (December value) less the December-to-December CPI inflation rate. gWeighted average of bank deposit interest rates (CPP, or costo prornedio porcentual).

Table 3.7

Financing the Fiscal Deficit (% of GDP) I97 1

1972

1973

1974 ~

Monetary deficit AIntemal debt AExternal debt

25 14 11

50 47 3

69 39 30

1975

1976

10 0 51 49

99 39 60

~~

72 32 40

425 Table 3.8a

Mexico/Chapter 3 Macroeconomic Aggregates (% 1971

1972

1973

1974

1975

1976

4.2 3.9 5.7 5.2 -1.7 8.9 -23.2

8.4 9.7 .8 5.5 12.2 2.5 40.2

8.4 10.5 4.0 21.3 14.7 3.0 39.6

6.1 6.3 2.5 20.9 7.9 11.4 2.4

5.6 5.0 2.0 11.0 9.3 2.0 21.6

4.2 1.o 27.2 .4 6.1 -7.6

70.3 8.2 20.6 12.5 8.1 1.9 8.7 9.7

69.7 8.2 20.2 13.2 7.0 3.9 8.2 11.0

69.7 8.9 21.7 12.7 9.0 3.0 7.1 10.4

69.9 9.0 20.9 12.9 8.0 2.3 7.9 10.1

~~

Real GDP Manufacturing Agnculture, forestry, and fisheries Inflationb Real gross fixed investment Private Public

’Igble 3.8b

5.0

Composition of Output (% of GDP)c

Private consumption Government consumption Gross fixed capital formation Private Public Change in inventories

Exports Imports

72.6 7.7 18.9 14.0 4.9 1.9 7.7 8.8

71.5 8.1 19.5 13.2 6.3 1.7 8.3 9.0

Source: Indicadores Economicos (Bankof Mexico)

‘Real variables are measured at 1970 prices. bDecember-to-December change in the CPI. ‘Output shares at 1970 prices.

from 14 percent of GDP (at 1970 prices) in 1971 to 12.7 percent in 1975. Government financial policies seem to have been at least partly responsible for the slump in private investment. Negative real deposit rates, slowing the growth of bank funds, and higher reserve requirements caused the supply of credit to the private sector to diminish sharply (see table 3.6). Lending recovered in 1975 when price controls suppressed inflation and bank deposits paid a small, positive return, but the recovery only partially offset the substantial financial disintermediation of the previous two years. At the end of the Echevema sexenio, the supply of commercial and development bank real credit to the “business” sector (i.e., the private sector plus the parastatal sector) had still not regained its 1972 level. More threatening than either the acceleration in inflation or the decline in private investment was the deterioration in the payments balance (table 3.9). As the nominal exchange rate remained pegged at 12.5 pesos per dollar, the real exchange rate fell steadily after 1970, despite rising world commodity prices. The decline in the real exchange rate and the large increase in public sector absorption set off a surge in import demand. From 1971 to 1975, the dollar value of private sector imports increased at an annual rate of 22.8 percent. In the public sector import growth was a phenomenal 53.8 percent, reflecting the high import intensity of many of the large industrial investment projects undertaken by SOEs in this period.

426 Table 3.9

Edward F. &fie External Accounts

Current account (billion $) Merchandise exports (billion $) Merchandise imports (billion $) Volume of merchandise imports (% change) Intermediate inputs Consumer g d s Capital goods Volume of merchandise exports (% change) Nonoil manufactures Agriculture, forestry, and fisheries Real exchange rate' Real price of total merchandise importsb Real price of total merchandise exportsb Nonoil manufacturesb

1971

1972

1973

1974

1975

1976

-.93 1.37 2.26 -6.4 -4.6 5.1 -13.2 2.4 6.3 -7.9 97.3 98.7 97.4 98.6

-1.01 1.67 2.76 10.3 5.0 27.1 14.7 13.3 15.6 14.3 94.4 98.9 98.8 99.4

-1.53 2.07 3.89 22.9 28.8 14.1 15.1 8.0 14.0 -.4 91.4 99.6 100.9 101.8

-3.23 2.85 6.15 25.1 39.6 - 12.2 9.0 3.6 - .3 -21.5 85.8 99.6 109.1 118.5

-4.44 3.06 6.70 -2.6 -9.7 - 13.0 20.3 - 10.0 - 15.1 - 8.0 82.2 96.0 105.3 101.6

-3.68 3.66 6.30 - 14.5 - 16.1 - 16.5 - 10.6 6.4 10.6 8.1 88.7 103.9 120.0 112.0

"1970 = 100, nominal exchange rate multiplied by the ratio of the U.S. wholesale price index (now called the producer price index) to the Mexican GDP deflator. b1970 = 100; deflated by the GDP deflator.

Export promotion efforts were confined to the creation of the National Institute for Foreign Trade (inspired by Japan's JETRO) and to the granting of rebates (the CEDIS) on all indirect taxes for exporters. In response to these measures and the strong growth in worldwide trade in the early seventies, real exports increased strongly in 1972 and 1973. But as the real exchange rate continued to fall, nonpetroleum exports first stagnated in 1974 and then declined 5 percent in dollar value in 1975. Even before the deceleration in export growth, the current account deficit worsened considerably. By 1975 the deficit had reached the alarming level of $4.4 billion, a figure equivalent to 5.1 percent of GDP. The reluctance to raise domestic interest rates in the face of higher inflation and a clearly overvalued peso caused the overall payments balance to deteriorate to an even greater extent. Capital flight withdrew approximately $5.3 billion from the country over 1974-76.' The large payments deficits were mirrored in a fast-mounting level of foreign indebtedness (table 3.10). From a figure of $6.6 billion at the start of 1971, the total foreign debt more than quadrupled to $27.9 billion by the end of 1976. Almost all of this debt was taken out by the public sector with the commercial banks. In 1970 the public sector external debt was $4.7 billion and 51.8 percent of the debt was held by private financial institutions. By 1976 these figures had changed to $21.6 billion and 82.7 percent. The administration stubbornly refused to acknowledge the accumulating evidence that some adjustment was required in its macroeconomic policies. Domestic economic difficulties were attributed instead to adverse external shocks. The upsurge in world prices in 1973 was cited as the main cause of higher domestic inflation. The world recession that followed in 1974-75

427 Table 3.10

MexicoIChapter3 Debt Burden Measures ~~

Total debt (billion $) Total debt/GDP" (%) Public sector debt serviceb (billion $) % of Merchandise exports Q of Current account income I of GDP" Total debt serviceC(billion $) % of Merchandise exports % of Current account income % of GDP" Total debt service #2' (billion $) I of Merchandise exports % of Current account income 90 of GDP

1970

1971

1972

1973

1974

1975

1976

6.6 18.4 .86 66.4 26.3 2.4 .98 76.2 30.2 2.8

7.1 18.2 .82 59.8 23.1 2.1 .95 69.8 27.0 2.4

8.3 18.3 .92 55.4 21.6 2.0 1.08 65.0 25.3 2.4 2.91 174.7 68.0 6.4

11.0 19.9 1.34

15.6 21.7 1.40 48.9 20.4 1.9 1.66 58.2 24.3 2.3 4.31 151.1 63.0 6.0

21.6 24.5 1.89 61.6 26.4 2.1 2.29 74.8 32.1 2.6 6.52 212.8 91.3 7.4

27.9 31.4 2.47 67.7 29.9 2.8 2.88 78.8 34.8 3.2 8.44 231.0 102.0 9.5

-

-

-

-

-

-

64.6 24.8 2.4 1.54 74.5 28.6 2.8 3.37 162.8 62.4 6.1

Sources: Mexican Economic Outlook (CIEMEX-WHARMN) for data on the total debt and short-term public and

private sector debt. All other data comes from Indicudores Economicos (Bankof Mexico).

"GDP measured in dollars was calculated by dividing nominal GDP by the period average controlled exchange rate. bPublic sector interest payments plus amortization of the medium- and long-term debt. 'Public sector debt service plus private sector interest payments. dThe sum of public and private sector interest payments, public sector amortization of the short-, medium-, and long-term debt, and amortization of the short-term private sector debt. Amortization of the short-term debt is assumed to equal the previous period's short-term debt.

was blamed for the slow growth in Mexican exports and used to justify the continuation of expansionary demand policies (needed to offset the weakening in external demand). Events proved all too soon, however, that something was amiss in domestic policy. In 1976, as the world economy began to recover, the economic program of the Echevem'a administration collapsed under extreme balance of payments pressures. Extensive import controls were imposed and parastatal expenditures were sharply curtailed, but little was done to check spending by other branches of the government or to curb monetary expansion. As a result, though the burden of debt service was not exceptionally high (35 percent of current account income if short-term debts could be rolled over), the current account deficit remained sizable ($3.7 billion for the year), capital flight persisted, and the Central Bank's stock of foreign exchange reserves became severely depleted. On August 3 1, the peso was devalued nearly 100 percent and the economy went into a severe tailspin. One month after the devaluation, compensating wage increases averaging 23.1 percent were decreed for industrial workers. During the last four months of the year, manufacturing sector employment declined 4.2 percent, the inflation rate surged to 60 percent, and there were frequent threats of bank runs. Shortly before Lopez Portillo's inauguration, negotiations began on the terms for a standby agreement with the IMF.

428

Edward F. Buffie

3.2 Concluding Observations The Echeverria economic program was a clear failure. For a couple of years following the 1971 recession, output grew strongly. The 1972-74 expansion was necessarily temporary, however, given the fundamental economic imbalances created by large fiscal deficits and mismanaged monetary policy. In the administration’s last two years, output and employment growth slowed considerably while inflationary and balance of payments pressures became extreme. Distributional considerations do not alter this assessment. None of the studies discussed in the previous chapter turn up any evidence that the overall disthbution of income improved.’ Nor does a less formal examination of Echeverria’s policies suggest that they benefitted either the urban or rural poor. The real blue-collar manufacturing wage grew at a slower pace than during Stabilizing Development. Public investment in agriculture increased initially but was later severely reduced when budgetary problems became acute in 1975 and 1976. Overall, the agricultural sector stagnated, experiencing average annual growth of only 2.6 percent. The Echevem’a administration failed politically as well as economically. The crisis of political legitimacy that came to the fore in 1968 was never squarely faced. Echeverria made no progress toward reconstructing a stable political consensus and, at the very end of his term, in an effort to revalidate his tarnished populist credentials, he introduced new political tensions through a heavy-handed land expropriation in the northwest. Once again, the political elites could not reach agreement on a presidential successor, and Echevem’a had to choose his successor unilaterally, just as he had been chosen unilaterally six years earlier by Diaz Ordaz.

4

The Lopez Portillo Administration

The Lopez Portillo administration began under difficult circumstances. In the last three years of the Echevem’a administration, the economy’s performance had deteriorated steadily. Real GDP growth fell for the third consecutive year in 1976, dropping to 4.2 percent, while the inflation rate rose to the relatively high level of 27 percent. Despite widespread imposition of import controls, the current account registered a deficit of $3.68 billion, and in October, after having been pegged at 12.5 pesos per dollar for twenty-two years, the currency was devalued to 23 pesos per dollar. The overall fiscal

428

Edward F. Buffie

3.2 Concluding Observations The Echeverria economic program was a clear failure. For a couple of years following the 1971 recession, output grew strongly. The 1972-74 expansion was necessarily temporary, however, given the fundamental economic imbalances created by large fiscal deficits and mismanaged monetary policy. In the administration’s last two years, output and employment growth slowed considerably while inflationary and balance of payments pressures became extreme. Distributional considerations do not alter this assessment. None of the studies discussed in the previous chapter turn up any evidence that the overall disthbution of income improved.’ Nor does a less formal examination of Echeverria’s policies suggest that they benefitted either the urban or rural poor. The real blue-collar manufacturing wage grew at a slower pace than during Stabilizing Development. Public investment in agriculture increased initially but was later severely reduced when budgetary problems became acute in 1975 and 1976. Overall, the agricultural sector stagnated, experiencing average annual growth of only 2.6 percent. The Echevem’a administration failed politically as well as economically. The crisis of political legitimacy that came to the fore in 1968 was never squarely faced. Echeverria made no progress toward reconstructing a stable political consensus and, at the very end of his term, in an effort to revalidate his tarnished populist credentials, he introduced new political tensions through a heavy-handed land expropriation in the northwest. Once again, the political elites could not reach agreement on a presidential successor, and Echevem’a had to choose his successor unilaterally, just as he had been chosen unilaterally six years earlier by Diaz Ordaz.

4

The Lopez Portillo Administration

The Lopez Portillo administration began under difficult circumstances. In the last three years of the Echevem’a administration, the economy’s performance had deteriorated steadily. Real GDP growth fell for the third consecutive year in 1976, dropping to 4.2 percent, while the inflation rate rose to the relatively high level of 27 percent. Despite widespread imposition of import controls, the current account registered a deficit of $3.68 billion, and in October, after having been pegged at 12.5 pesos per dollar for twenty-two years, the currency was devalued to 23 pesos per dollar. The overall fiscal

429

Mexico/Chapter 4

deficit increased slightly from the previous year and, at 9.9 percent of GDP, was obviously unsustainable. Shortly after the October devaluation, a Letter of Intent was submitted to the IMF outlining a stabilization program to be implemented in stages over the next three years. Table 4.1 lists the main targets of the Fund program. The program called for the standard mixture of trade liberalization and economic austerity. Public sector savings, international reserves, and net domestic assets of the Central Bank were to increase, and a $3 billion limit was imposed on additional foreign borrowing by the public sector. Wage restraint was to accompany monetary and fiscal restraint: annual wage increases of 10 percent, 12 percent, and 15 percent were planned for 1977-79.' The trade reforms were aimed at increasing the openness of the economy and rationalizing the system of incentives for exporting and import substitution. Import licenses were to be progressively replaced by tariffs granting, in most cases, a lesser degree of protection.2

Table 4.1

Macroeconomic Targets for IMF Stabilization Program of September 1976 (% of GDP) 1976 Projection

Balance of payments Change in international reserves Merchandise trade and services balance Interest payments on the external debt Transfers and other factor payments Direct investment Change in net indebtedness of the public sector Public sector operations Revenues Current expenditures Public sector savings Investment Deficit Net external financing Net domestic financing Savings and investment Investment Gross fixed investment of the public sector Gross fixed investment of the private sector private sector inventory accumulation Total savings National savings Public Private Foreign savings Accumulation of international reserves (total savings minus investment)

1977

1978

1979

1.o

-

1.6 3.0

.5 .1 -2.2 -1.0 1.6 2.0

26.4 25.9 .5 8.7 -8.2 5.7 2.5

27.7 25.4 2.3 8.3 -6.0 3.0 3.0

28.8 24.8 4.0 8.0 -4.0 2.0 2.0

25.0 8.7 14.3 2.0 25.0 20.1

26.0 8.3 15.7 2.0 27.0 22.4 2.3 20. I 4.6

27.0 8.0 17.0 2.0 27.5 23.9 4.0 19.9 3.6

1 .o

.5

-

-2.3 -1.7 - .9 1.5 5.7

.5 19.6 4.9

- .5 -2.1

-1.0

0

Source: Economic Memorandum, Government of Mexico (13 September 1976)

-

-2.2

-1.0 1.6

1.0 30.0 24.5

5.5 8.0 -2.5 1.o I .5 28.0 8.0 18.0 2.0 28.0 25.4 5.5

19.9 2.6 0

430

Edward F. Buffie

This stabilization program was fairly successful in its first year. Although the revenue share of the public sector increased only slightly, the consolidated fiscal deficit was still lowered from 9.9 to 6.7 percent of GDP through a sharp reduction in investment spending. Cutbacks in the parastatal sector accounted for a large fraction of the decrease in total real investment spending-nominal non-PEMEX parastatal investment expenditures actually declined by 1.11 billion pesos. Wage restraint and the reduction in public sector absorption exerted a favorable impact upon the price level and the payments balance. The inflation rate declined from 27.2 to 20.7 percent (December-to-December change), while the current account deficit fell by over $2 billion in response to sharp increases in real export and import prices. The improvement in the current account was matched by a similarly large improvement in the capital account as capital flight decreased following the stiff devaluation of the currency in the last quarter of 1976. The lower current and private sector capital account deficits together with an additional $2.7 billion of public sector foreign borrowing enabled Central Bank reserves to increase by $657 million after declining by over $1 billion the previous year. A certain measure of success was also achieved with respect to the targets for real economic activity. The economy went into a recession, but the general downturn was less severe than anticipated: real GDP, which had been forecasted to remain constant, grew 3.4 percent. The one area in which the economy’s performance was unsatisfactory was private sector investment spending. After registering modest growth in the preceding two years, real fixed capital formation in the private sector declined 6.7 percent in 1977.3 This sharp contraction was apparently induced by the series of currency devaluations starting in the last quarter of 1976, which strongly increased the real price of imported capital goods and diminished the profitability of new investment. In the face of a 19.4 percent increase in real import prices, the volume of capital good imports fell 27.6 percent in 1977.4 Exactly how the economy would have evolved in the last two years of the stabilization program must be left open to conjecture, for in the course of the year, policy perspectives changed radically as it became widely known that Mexico’s oil wealth was far greater than formerly thought. The 1975 figure for proven hydrocarbon reserves of 6.4 billion barrels was increased to 11.2 billion during 1976 and then raised further to 16 billion at the end of 1977 (Zedillo 1985, 304).5 This constituted a stupendous increase in national wealth; by the time oil prices reached $31.25 per barrel in 1980, oil wealth would measure 1,370 percent of GNP and 570 percent of the value of the aggregate capital stock (Rizzo 1984, 109). An immediate consequence of the discovery of enormous oil wealth was the virtual disappearance of any constraints on foreign borrowing. Fierce

431

Mexico/Chapter 4

competition arose among foreign banks to extend new loans to Mexico. Naturally, the Mexican government took advantage of its enhanced credit rating to improve the terms of its foreign debt. During 1978 and 1979, the average maturity on public sector debt was lengthened from a little less than five years to over eight years. In addition, the interest spread over LIBOR (London interbank offer rate for dollar deposits) was reduced from an average (for long-term credits) of 1.625 percentage points in the preceding three-year period to between 0.625 and 0.825 percentage points, a rate that compared favorably with that charged to prime customers in the OECD countries (Zedillo 1985, 308). A second and more important repercussion of oil wealth came in the policy sphere. Not surprisingly, Mexican officials felt that they now faced a less rigid set of constraints and that economic policy ought to be reoriented toward recovering the development momentum lost in preceding years. The Fund program, therefore, was dropped in favor of a “new,” more expansionary policy package. 4.1

1978-81: Public-Expenditure-Led Growth Once Again

The new development plan called for large, sustained increases in real government expenditures. In this respect, the plan appeared to continue the discredited public-expenditure-led-growth (PEW) strategy of the Echevema administration. It was argued, however, that an economic base expanded ant strengthened by oil wealth could support a much enlarged role for the public sector. Furthermore, strong fiscal stimulus was to be only one part of a comprehensive reform package that would avoid the main policy errors of the Echevema administration. The liberal wage increases granted in the early and mid-seventies were discontinued and replaced by a quite restrictive wage policy: even though the 1977 inflation rate was 20.7 percent, the wage increase announced for government employees was only 10 percent in 1978 and the contractual “guideline” for private sector wage increases was set at 12 percent. The exchange rate was to be managed more flexibly in order to avoid balance of payments crises and speculative runs against the peso. Nominal interest rates would also be more flexible and were to be set so as to maintain positive real interest rates and a high level of financial intermediation. To keep the fiscal deficit in check, the operations of the public enterprises would be rationalized and tax revenues would be increased by broadening the tax base and improving tax administration and taxpayer compliance. At the same time that the size of the fiscal deficit would be diminished by these measures, new debt instruments (government bonds known as CETES) would be introduced to reduce monetization of the deficit. Finally, public sector expansion was not to occur at the expense of productive capacity in the private sector. Capital goods were exempted from

432

Edward F. Buffie

the value-added tax (VAT), and a more favorable treatment of depreciation allowances was introduced in an effort to revive private investment. It is undeniable that between 1978 and 1981 the Mexican economy recorded some impressive accomplishments (tables 4.2a and 4.2b). Real GDP growth ranged between 8.0 and 9.1 percent, and employment growth in the high-wage manufacturing sector and the public sector increased 27.2 and 41.4 percent, respectively. Both private and public sector investment spending increased rapidly. The share of public sector investment in GDP rose (at 1970 prices) from 7.2 to 10.8 percent and that of the private sector increased from 11.7 to 14.1 percent. The inflation rate began creeping upward after 1978, but never exceeded 30 percent. For 1978 and perhaps part of 1979, it can be argued that the demand stimulus provided by higher public sector spending was an important element in the economic recovery. The source of rapid growth beyond early 1979, however, seems to have been strong supply-side expansion fueled by sharp decreases in the real prices of domestic and imported intermediate inputs (table 4.3). Price controls kept the internal price of energy growing at a pace barely one-third that of the GDP deflator. The relative price of imported intermediates also fell considerably as the “fixed but adjustable exchange rate” proved to be more fixed than adjustable. The nominal exchange rate rose at an annual average rate of 3.6 percent, far less than the spread between the U.S. and Mexican inflation rates. Consequently, the real exchange rate declined by a full 30 percent between 1977 and 1981, provoking a 128 percent increase in the volume of imported intermediate inputs. Since factors are normally gross complements (a decrease in the price of one input raises the demand for other inputs), the large decreases in intermediates prices would be expected to raise strongly the demand for labor and capital, stimulating growth in employment and investment. The elastic supply response also helps to explain why the huge growth in government spending did not prove highly inflationary until the 1982 crisis.6 Table 4.2a

Macroeconomic Aggregates (% change)”

Real GDP Manufacturing Agriculture, forestry, and fisheries hflationb Manufacturing sector employment‘ General government employment Public sector employment Real investment Private Public

1976

1977

I978

1979

1980

1981

1982

4.2 5.0 1.0 27.2

3.4 3.6 7.5 20.7 1.9 5.7 5.5 -6.7 -6.7 -6.7

8.3 9.8 6.0 16.2 7.9 7.5 7.3 15.2 5.1 31.6

9.1 10.6 -2.1 20.0 6.7 9.9 9.2 20.2 22.7 17.1

8.3 7.2 7.1 29.8 7.2 10.8 10.4 14.9 13.7 16.7

8.0 7.0 6.1 28.7 2.9 9.6 9.3 14.7 14.0 15.8

-.5 -2.9 -.6 98.9 -8.5 5.3 5.9 -15.9 -17.3 -14.2

-.06 9.8 9.1 .4 6.1 -7.6

433 Table 4.2b

MexicoKhapter 4 Composition of Output (% of G D P ) ~

Private consumption Government consumption Gross fixed capital formation Private Public Change in inventories Exports Imports

69.9 9.0 20.9 12.9 8.0 2.3 7.9 10.1

69.0 8.6 18.9 11.7 7.2 3.5 8.8

8.8

68.9 8.8 20.0 11.3 8.7 3.0 9.1 9.9

68.7 8.8 22.1 12.7 9.4 2.8 9.3 11.7

68.2 8.9 23.5 13.4 10.1 4.6 9.1 14.3

67.9 9.1 24.9 14.1 10.8 5.1

9.0 15.9

69.0 9.3 21.0 11.7 9.3 .5 10.2 10.1

Sources: National Income Accounts, Production Accounts of the Public Sector, 1975- 1983 (Mexico, D.F.: INEGI) for government employment data. The manufacturing sector employment series is from Indicudores Economicos (Bank of Mexico). All other data is from the National Income Accounts (INEGI). ”Real variables are expressed in terms of 1970 prices ”December-to-December change in the CPI. ‘December-to-December change. dOutput shares at 1970 prices. Table 4.3

Real Prices of Intermediate Inputs (1977 = 100)

Real exchange ratea Real domestic price of energy inputsb

1971

1978

1979

1980

1981

100 100

93.2 89.8

86.2 80.6

16.4 68.6

70.0 63.2

Source: The internal producer price index for energy inputs is from the series “Combustible y Energia” in

table 20.8, Esradisticas de Mkxico (Mexico, D.F.: INEGI, 1985): 753-57.

“Calculated as the period average official exchange rate multiplied by the ratio of the U S . wholesale price index (now called the producer price index) to the Mexican GDP deflator. bDeflated by the GDP deflator.

While the overall performance of the Mexican economy was impressive during 1978-81, there is considerable disagreement about the extent to which labor benefitted from this phase of historically high growth. Employment in the high-wage public and the manufacturing sectors increased considerably, and though reliable employment data does not exist for other sectors of the economy, it appears that the growth in aggregate labor demand was healthy as well. According to some accounts, labor shortages even began to appear toward the end of 1981 (Zedillo 1985, 305; Gregory 1986, 303). But if labor gained from better employment opportunities, the limited data available also suggests that real wage compression accompanied employment growth during this period. Various real wage indices are computed in table 4.4. Between 1977 and 1981, the average real minimum wage decreased 10.6 percent and the real public sector wage grew by only 4.4 percent. Of course, the information conveyed by these two wage indices is limited. Neither the minimum wage nor the government sector wage is necessarily an accurate index of private sector wage costs. This would seem to be especially true of the late seventies. Unions strongly resisted government wage guidelines implying real wage cuts, and to avoid or settle strikes many firms agreed to

434 Table 4.4

Edward F. Buffie Real Wages (1977 =

CPI deflator Average minimum wageb Public sector wage‘ Manufacturing sector Blue-collar wage White-collar wage Overall wage (inclusive of fringe benefits) Tornell contract wage GDP deflator Average minimum wage Public sector waged Manufacturing sector Blue-collar wage White-collar wage Overall wage (inclusive of fringe benefits) Tornell contract wage

1977

1978

1979

1980

1981

1982

100 100

96.6 100

94.6 100.4

87.9 97.6

89.4 104.4

79.0 80.7

100 100

95.7 96. I %.6 95.2

91.6 92.4 93.8 95.7

92.6 94.3 91.1 96.5

93.3 89.7 96.3

100

97.3 97.0 98.0 97.2

100 100

97.2 103.6

93.5 102.0

85.4 98.0

86.9 101.7

76.1 93.1

100 100 100 100

97.9 97.7 98.7 91.9

94.7 95.0 95.6 94.2

88.9 89.8 91.1 93.0

90.1 91.7 94.4 93.8

89.8 86.3 92.6

I00

Sources: Minimum wage data are from INEGI. The blue-collar, white-collar, and overall wage series for the manufacturing sector are from Encuesfa Industrid Mensual. as reported in Indices de Precios (February 1986). The Tornell contract wage is based on the contract wage series found in Tornell (1983). “The minimum wage index is a weighted average of minimum wages in different regions, where the weights are given by the region’s share of the total salaried population in the nation. In years in which there is more than one wage adjustment, the period average figure is generated by weighting the wage in each subperiod by the fraction of the year during which it prevailed. bPeriod average nominal wage deflated by either the period average CPI or the GDP dellator. ‘End-of-year wage deflated by the end-of-year CPI. dAverage of the beginning- and end-of-year wage deflated by the GDP dellator.

grant wage increases well in excess of the guidelines. Also, a shift toward greater fringe benefits (social security, vacation pay, year-end bonuses, and employer-subsidized housing, food, and transportation) pushed up labor costs more than is suggested by nominal wage settlements. Fortunately, for the manufacturing sector at least, a good deal is known about the nature of wage contracts during this period. The first three indices in table 4.4 were constructed using the data gathered from the Bank of Mexico’s survey of contractual wages in large manufacturing firms. The rows labeled “Tornell contract wage” are based on wage series computed in a careful study by Tornell (1983). Tornell used data from actual contracts to adjust nominal wages for all fringe benefits. A monthly manufacturing sector wage series was then constructed by weighting the wage in each contract by the fraction of the total labor force in the sample covered by that particular contract. The sample consisted of data from forty-one firms that produced more than half of total manufacturing sector output.’ The contractual wage series confirm the general picture of real wage restraint in the formal sectors of the economy. Regardless of whether the CPI or the GDP deflator is used, the real contract wage declined over 1977-81, though none of the indices decreased as much as the real minimum wage.

435

Mexico/Chapter 4

Despite real wage restraint and a substantial increase in the economy's investment rate, the acceleration in growth after the 1977 recession was not sustainable. In retrospect, it is clear that little, if any, policy reform took place and that the oil bonanza simply resulted in the policy mistakes of the Echevem'a administration being repeated on a larger scale. Both current and capital expenditures of the public sector grew more rapidly than projected and got completely out of hand after 1980 (table 4.5). Total real public sector expenditure increased by 97.7 percent in the space of four years (calculated by deflating by the period average CPI), climbing from 29.5 percent of GDP in 1977 to 41.3 percent in 1981, a figure some nine percentage points above the peak value recorded during the Echeverria administration. This massive increase in expenditure led to large fiscal deficits as it was not matched by a similar buildup in revenues. After declining to 6.7 percent of GNP in 1977, the consolidated public sector deficit grew steadily and then skyrocketed to 14.7 percent of GNP in 1981 when real public sector spending (net of interest payments on the foreign debt) rose an astounding 28.6 percent. The breakdown in the overall deficit is shown in table 4.6 and points to stagnation of nonoil revenues, in addition to rapid expenditure growth, as an important factor in the rising deficits. PEMEX initially registered a small surplus, but after 1978, when petroleum exports commenced on a large scale, the surplus rose rapidly, reaching 6.3 percent of GDP in 1980 and then falling back to 4.1 percent in 1981. This sizable revenue windfall was offset to a large extent by slow revenue growth elsewhere in the public sector. Between 1978 and 1981, the deficit of the non-PEMEX parastatal sector increased from 2.8 percent of GDP to 5.1 percent, with more than half of the increment owing to the decline in the sector's revenue share. The 1979 tax reform improved the efficiency and equity of the tax system, but did not succeed in increasing revenues significantly:8 the revenue share of the nonparastatal sector declined to an even greater extent than that of the (non-PEMEX) parastatal sector, dropping from 10.5 percent of GDP in 1978 Table 4.5

Public Seetor Expenditures and Revenues (9%of GDP)

Expenditure Current Interest payments on the foreign debt Other Capital Revenues Economic deficit Deficit on financial intermediation" Monetary deficit

1977

1978

1979

1980

1981

1982

29.5 22.0 1.9 20.1 7.5 24.2 5.4 1.4 6.7

31.0 22.3 2.0 20.3 8.7 25.5 5.5 1.2 6.7

32.2 22.6 2.1 20.5 9.6 26.2 6.0 1.4 7.4

34.6 24.9 2. I 22.8 9.7 27.8 6.8 1 .o 7.9

41.3 28.0 2.3 25.7 13.3 27.7 13.6 1.2 14.7

46.4 36.0 5.1 30.7 10.6 30.1 16.3 1.4 17.6

~~

Source: Estadisricas Hacendarias del Sector Publico: Cifras Anuales, 1965- 1982 (SHCP). 'Deficit of La Banca de Desarrollo.

436

Edward F. Buffie

Table 4.6

PEMEX Expenditure Currenta Capitalb RevenuesC Deficit Non-PEMEX parastatals Expenditure Current” Capitalb Revenues‘ Deficit OtheP Expenditure Current Capital Revenues Deficit

Breakdown of the Fiscal Deficit (% of GDP) 1977

1978

1979

1980

1981

1982

3.9 2.0 1.9 4.9 - 1.0

4.8 2.1 2.7

5.5 2.5

1.5

- 1.9

5.8 2.9 3.0 12. I -6.3

3.6 3.9 11.6 -4.1

1.5 4.5 3.0 15.8 -8.3

5.8

-1.0

3.0 7.4

12.0 9.8 2.3 9.0 3.0

12.0 9.5 2.5 9.2 2.8

11.5 8.9 2.7 8.7 2.8

12.0 9.0 3.1 8.2 3.9

13.0 9.9 3.1 7.9 5. I

12.5 9.9 2.6 8.1 4.4

13.6 10.3 3.3 10.2 3.3

14.2 10.7 3.5 10.5 3.1

15.2 11.2 4.0 10.1 4.9

16.7 13.0 3.7 7.5 9.2

20.8 14.5 6.3 8.3 12.5

26.5 21.5

5.0 6.2 20.2

Source: Estadisticus Hucendurius del Sector Publico: Cfrus Anuules, 1965-1982 (SHCP). “Gusto de operucion plus ajenas de gasto (operating expenditure plus “outside account” expenditure).

bPhysical investment only (excludes financial investment).

‘The sum of current income, capital income, taxes paid, and ujenas de ingreso (outside account income). dIncludes DDF (Department of the Federal Disbict).

to 8.3 percent in 1981. Moreover, part of current expenditures of the federal government probably reflects expenditures induced by revenue shortfalls in the nonparastatal sector. In the detailed fiscal accounts of Secretaria de Hacienda y Credito Publico (SHCP), it is not possible to trace the majority of transfer payments made by the federal government. These unaccounted for transfers are quite sizable and reflect mostly expenditures to cover the losses of various price support schemes, local “development institutions,” and firms in which the government has a minority interest (but which are not classified as state-owned enterprises).’ Such transfer payments increased steadily throughout the Lopez Portillo senenio and exceeded in each year the deficit of the non-PEMEX parastatal sector (table 4.7). If the unaccounted for transfer payments in the table are treated as a negative revenue item (i.e., “induced” subsidies), the revenue share in GDP of the non-PEMEX public Table 4.7

Hidden Transfer Payments (% of GDP)

Unaccounted for federal government transfers

1977

1978

1979

1980

1981

1982

3.6

3.4

3.9

4.6

5.7

8.4

Source: Estadisticus Hucendarius del Sector Publico: Cifras Anuales, 1965-1982, p. 22 (SHCP). Sum of unaccounted for current and capital transfers.

437

MexicolChapter 4

sector fell by 5.8 percentage points over 1978- 8 1, indicating, remarkably, a three-percentage-point decrease in the sum of non-PEMEX revenues and the PEMEX surplus. There can be little doubt that the large decrease in the share of nonoil revenues was due principally to a reluctance to raise public sector prices. Some evidence in support of this claim is presented in table 4.8.After 1977 the pace of public sector price increases was less than half that of the inflation rate, and sales of goods and services by the non-PEMEX parastatal sector declined relative to GNP. The share of general tax revenues in GDP, on the other hand, exhibited a modest rise before dropping sharply in 1982.” The failure to maintain real public sector prices not only slowed the growth of non-PEMEX revenues but also greatly diminished the size of the PEMEX surplus. Domestic energy prices changed very little as world petroleum prices shot upward after 1973, so that by 1980 the average internal price of petroleum products was less than one-quarter of the world market price (table 4.9). If the share of private sector investment and consumption in GDP is taken as a rough approximation of the private sector share in total energy consumption, the revenue loss from the implicit subsidy on domestic consumption of PEMEX products amounted to 6.2 percent of GDP in 1980, a figure almost as large as the entire public sector economic deficit that year. As occurred earlier in the Echevem’a administration, the large fiscal deficits gave rise to unsustainably large balance of payments deficits which ultimately proved to be the undoing of the PELG strategy. Consider the familiar decomposition of the current account deficit B: B = (S - I )

+ (R- G ) ,

where S is private sector savings, I is private sector investment, and R and G are total public sector revenues and expenditures (i.e., including those of the parastatal sector). Strictly speaking, it is improper to make conjectures about Table 4.8

Public Sector Prices and Revenues

Percentage increase in public sector pricesa Period average inflation rate Revenues from non-PEMEX parastatal sales of goods and services (% of GDP)b General tax revenues (% of GDP)C

1977

1978

1979

1980

I981

1982

28.4 28.9 6.3

6.4 17.5 6.5

8.3 18.2 6.1

-

26.3 5.5

24.7 28.0 5.3

72.4 58.9 5.3

10.5

10.7

10.9

11.2

10.8

9.0

Sources: Estadisticas Hacendarias del Sector Publico: Cijias Anuales, 1965- 1982 (SHCP) for non-PEMEX revenues from sales of goods and services. Clavijo (1980) for public sector price increases between 1977 and 1979. Indicadores Econornicos (Bank of Mexico) for public sector price increases between I980 and 1982. a

Period average increases. There is a series break in 1980. Budget- and nonbudget-coneolled parastatal firms.

Sum of direct taxes, taxes on production and services, value-added taxes, and “other” tax revenues. Does not include gas taxes (which I classify as part of PEMEX revenues).

438 Table 4.9

Year 1973 1974

1975 1976 1977 1978 1979 1980

Edward F. Buffie Subsidy on Domestic Energy Consumption International Price/ Domestic Pricea

Domestic Subsidy (billion pesos)

Gross Subsidy (% of GNP)

1.29 2.30 1.90 2.90 3.17 3.49 4.06 4.68

5.4 39.9 30.0 74.3 114.7 147.0 228.1 350.9

4.4 2.7 5.4 6.2 6.3 7.4 8.2

.8

Net Subsidyb .7 3.6 2.0 4.4 4.9 4.9 5.8 6.2

Source: The relative domestic price of PEMEX products and the implicit cost of the domestic subsidy are from Rizzo (1984). aAverage international price of PEMEX products relative to the average domestic price. bNet subsidy is obtained by multiplying the gross subsidy by the share of private sector consumption and investment in GNP at current prices. (The change in inventories is not included in private investment.)

the impact of the fiscal deficit on B without specifying a full model that takes into account induced changes in S and I. Nonetheless, the general picture is clear enough in the Mexican case. Private sector investment spending increased from 11.6 to 14.0 percent of GDP (at current prices) over 1978-81 in response to the tax breaks provided in the 1978 tax reform, the large decreases in the real prices of factors complementary to capital, and the fall in the real exchange rate, which effectively subsidized the purchase of imported machinery. The shift in the distribution of income away from labor and agriculture led to an even larger increase in the private sector saving rate (the share of private consumption in GDP declined from 66 to 61 percent at current prices), but the greater surplus of private sector savings over private investment was not nearly large enough to compensate for the massive decrease in public sector savings. The trend toward rising current account deficits and external indebtedness is spelled out in greater detail in tables 4.10 and 4.11. Trade liberalization combined with real exchange rate appreciation lowered the real price of imports (deflating by the GDP deflator) by approximately 28 percent from 1977 to 1981, provoking a stupendous, across-the-board increase in demand. Between 1978 and 1980, real imports of capital goods and intermediate inputs increased by more than 100 percent. As the relaxation of quotas favored consumption goods more than other types of imports, the volume of imported consumer goods increased even more strongly, rising by over 200 percent in the same three-year period." In 1981, fears about the growing payments deficit resulted in the reimposition of quotas on many items, particularly consumer and capital goods imports. Nevertheless, the overall import volume still rose 15.2 percent. On the export side, oil sales became very sizable after 1978. From 1978 to 1981, dollar earnings generated by petroleum exports increased 682 percent. Overall export earnings, however, rose at a considerably slower rate as

439

MexicoKhapter 4

Table 4.10

External Accounts ~~

Current account deficit (billion $) Merchandise exports (billion $) Merchandise imports (billion $) Real exchange rate' Real price of total merchandise exportsb Nonoil manufactures Real price of total merchandise importsb Volume of total merchandise exports (W change) Nonoil manufactures Volume of total merchandise imports (% change) Intermediate inputs Consumer goods Capital goods

~~~~

~

1977

1978

1979

1980

1981

1982

- 1.6

-2.7 6.1 7.9 98.8 142.2 119.9 118.6

-4.9 8.8 12.0 91.4 156.2 108.7 112.3

-7.2 15.1 18.8 81.0 191.1 95.2 99.7

- 12.5

4.6 5.7 106.0 155.7 141.2 124.1

19.4 23.9 74.2 188.1 82.4 89.4

-6.2 21.2 14.4 112.1 256.9 98.8 111.2

13.7 8.5

29.8 29.4

10.0 5.6

11.4 -2.8

7.1 -4.0

16.9 11.1

-6.1 6.4 - 18.1 -27.6

27.8 25.2 64.2 26.2

32.9 20.8 38.6 64.8

36.9 34.4 81.0 29.8

15.2 12.0 6.5 25.4

- 39.0 - 36.2 -46.3 -42.1

Source: National Income Accounts (INEGI) for traded goods price indices and the indices of import and export volumes. Indicadores Economicos (Bank of Mexico) for the current account deficit and dollar value of imports and exports. "1970 = 100. calculated as the period average official exchange rate multiplied by the ratio of the U.S. wholesale price index (now called the producer price index) to the Mexican GDP deflator. b1970 = LOO; deflated by the GDP deflator.

nonoil exports suffered from both an appreciating real exchange rate and the dismantling of the CEDIS system of subsidies. After jumping to a decade-level high in 1977, the real price of manufactured exports plummeted, declining more than 40 percent in the next four years. Predictably, the volume of manufactured exports slowed sharply in 1979 and then turned negative in 1980 and 1981.12 The financial counterpart to the large current account deficits was a fast-growing level of external indebtedness. The total foreign debt increased almost threefold to $81 billion at the end of 1981. This figure, however, considerably overstates the increase in net foreign debt. Table 4.12 gathers together various estimates of the magnitude of capital flight during this period. The wide variation in the estimates arises from different data bases. l3 According to the Cumby and Levich (1987) estimate (col. l), capital flight siphoned off roughly 46 percent of the extra debt accumulated between 1977 and 1981. A problem with their estimate is that the net inflow of external resources is calculated from World Bank data on the change in gross external indebtedness. But as Zedillo points out (1987, 175-76), this is not an accurate measure of net new indebtedness because in certain years some of the increment in the reported debt figures simply reflects more extensive coverage by the government's debt-reporting systems. Zedillo (col. 2) uses the Bank of Mexico's balance of payments data to measure the change in net indebtedness (a much more accurate measure), but also makes the odd adjustment of subtracting from the official current account data imputed

440

Edward F. Buffie

Table 4.11

Debt Burden Measures ~~~~

Total debt (billion $) Total debt/GDP Public sector debt serviceb (billion $) % of Merchandise exports % of Current account income % of G D P Total debt serviceC(billion $) % of Merchandise exports % of Current account income 96 of G D P Total debt service #2‘ (billion $) % of Merchandise exports % of Current account income % of GDPa Net debt‘ (billion $) Net debt service‘(biI1ion $) % of Merchandise exports % of Current account income % of GDP” Net debt service #2g (billion $) % of Merchandise exports % of Current account income % of G D P

~

1976

1977

1978

1979

1980

1981

1982

27.9 31.4 2.5 67.7 29.9 2.8 2.9 78.8 34.8 3.2

30.3 37.0 3.8 82.5 41.8 4.7 4.3 91.8 46.5 5.2 10.9 235.1 119.1 13.4 22.7 3.8 81.2 41.1 4.6 10.4 224.5 113.8 12.7

35.1 34.2 6.3 103.7 54.0 6.1 6.8 112.7 58.7 6.6 12.5 206.2 107.2 12.2 24.3 6.0 99.7 51.9 5.9 11.7 193.2 100.5 11.4

42.4 31.5 10.2 115.4 62.6 7.6 11.0 124.7 67.6 8.2 15.7 178.6 96.8 11.7 27.0 9.6 109.4 59.3 7.2 14.4 163.3 88.5 10.7

54.4 29.2 7.7 50.8 30.8 4.1 9.2 60.8 36.9 4.9 15.5 102.7 62.3 8.3 31.9 6.9 45.8 27.8 3.7 13.3 87.7 53.2 7. I

81.0 33.8 10.3 52.9 33.4 4.3 13.2 67.9 42.8 5.5 24.3 125.0 78.8 10.1 39. I 8.9 45.6 28.7 3.1 19.9 102.7 64.7 8.3

87.6 53.4 14.9 70.0 53.0 9.1 17.4 82.1 62.2 10.6 39.9 187.8 142.4 24.3 51.6 12.4 58.5 44.3 7.6 34.9 164.2 124.5 21.3

19.0 2.3 63.8 28.2 2.6

Sources: Mexican Economic Outlook (CIEMEX-WHARTON) for data on the total debt and short-term public and private sector debt. All other data come from Indicadores Economicos (Bank of Mexico).

“GDP measured in dollars was calculated by dividing nominal GDP by the period average controlled exchange rate. There is no correction for deviations of the actual exchange rate from the equilibrium exchange rate. bPuhlic sector interest payments and amortization of the medium- and long-term debt. ‘Public sector debt service plus private sector interest payments. dThe sum of public and private sector interest payments, public sector amortization of the short-, medium-, and long-term debt, and amortization of the short-term private sector debt. Amortization of the short-term debt is assumed to equal the previous period’s short-term debt. ‘Calculated as the cumulated value of official current account deficits starting from 1951 ‘Calculated by scaling total interest payments by the public and private sectors by the ratio of net debt to total debt. No attempt is made to adjust for the fact that the interest rate on private sector foreign assets differs from the rates charged for foreign loans to the public and private sectors. gCalculated as the sum of public sector amortization of the short-, medium-, and long-term debt, amortization of the short-term private sector debt, and net interest payments. Net interest payments are total interest payments by the public and private sectors scaled down by the ratio of net debt to total debt. Amortization of the short-term debt is assumed to equal the previous period’s short-term debt.

interest payments in identified Mexican deposits abroad. (For some reason, reinvested interest income from foreign assets is not treated as capital flight.) In the third column, labeled “Modified Zedillo,” I remove this latter adjustment. This gives a figure for capital flight that is $3.9 billion less than that of Cumby and Levich. Finally, in the fourth column, the previous three estimates are corrected using Gulati’s estimates (1987, 73) of net capital flight effected through trade-invoice faking. In Mexico, underinvoicing of imports exceeded underinvoicing of exports during this period, so that estimated capital flight is reduced. The Gulati adjustment suggests that

441

Mexico/Chapter 4 Capital Flight (billion $)”

Table 4.12

Gulati-Adjusted CLC

z*

.98

5.61

.60

1.17

1.06 3.89 14.03 9.03 3.39 3.67 3.75

1.52 3.72 6.15 6.60 6.88

-

1.31 -.52 p.62 -3.71 7.32 7.59 -2.41 -4.35 -.68

1.60 .01 .25 .86 11.62 8.39 -1.44 -3.01 1.15

24.77

11.37

22.69

-

10.83

26.29

Year

Cumby and Levich

-

1977 1978 1979 1980 1981 1982 1983 1984 1985

4.99 I .76 2.37 6.75 8.56 7.24

.69 .07 .23 - .68 9.73 8.23 2.42 2.33 I .92 18.27

29.59

24.94

40.40

11.71 6.02

Cumulative total 1977-82 Cumulative total 1977-85

31.67

~

Zedillo

Modified Zedillob

-.66

m

e

~

Sources: Cumby and Levich (1987, 58). Gulati (1987, 73), and Zedillo (1987, 177). aMorgan Guaranty definition of capital flight: the change in the foreign debt plus net foreign direct investment plus the current account surplus minus the change in short-term foreign assets of the banking system minus change in foreign exchange reserves. The Zedillo estimates also subtract the change in other official external assets. bEstimate obtained using Zedillo’s data and the official figures for the current account deficit. ‘Cumby-Levich estimate with current account data adjusted by Gulati’s estimate of net trade invoice faking. *Zedill0 estimate with current account data adjusted by Gulati’s estimate of net trade invoice faking. ‘Modified Zedillo estimate with current account data adjusted by Gulati’s estimate of net trade invoice faking.

approximately one-third of the increase in total gross debt may have ended up financing capital flight. l4 Nearly all of the new debt was contracted by the public sector; the private sector debt tripled during 1978-81, but still stood at only $21.9 billion in 1981. Most of the $53 billion of debt held by the public sector took the form of medium- or long-term commercial loans extended to different SOEs (PEMEX alone had contracted $15.7 billion of foreign debt by 1981), but the short-term debt also grew rapidly and by the end of 1981 accounted for 20.3 percent of the total public sector debt. Since over half of private sector borrowing was short term, for the aggregate debt the corresponding figure is a much higher 27.7 percent. By contrast, just three years earlier the share of short-term debt stood at only 13.5 percent.15 Despite the large windfall conferred by oil discoveries and the high rates of GNP growth achieved between 1978 and 1981, it seems safe to say that an increase of this magnitude in the foreign debt was excessive. The standard debt burden measures may be found in table 4.11. l6 The net debt service measures take account of the fact that the private and public sectors hold income-earning foreign assets as well as foreign debts. In these figures, the net debt is calculated, crudely, as the cumulative value of past official current account deficits starting from 1951.

442

Edward F. Buffie

While all of the debt burden measures decreased sharply after 1980 when oil exports increased by $6.5 billion, it is also evident that the heavy binge of short-term borrowing in the immediately preceding years had placed the country in a financially precarious position. Even in 1981, debt service inclusive of short-term amortization claimed nearly 80 percent of total current account income. If short-term amortization is excluded (which gives a better sense of the medium-run debt service profile), the debt service burden was not particularly onerous in 1980 or 1981, judged by the usual standards. Observe, however, that just one year later and notwithstanding a 13 percent increase in the dollar value of oil exports, a much less sanguine picture emerges. In 1982, debt service exclusive of short-term amortization absorbed 62.2 percent of total current account income and 10.6 percent of GNP.” The corresponding figures for net debt service are smaller but still quite large. Even though the burden of debt service eased considerably in 1980, the large current account deficit in that year was a clear warning signal that some adjustment in economic policy was needed. For reasons that are difficult to fathom, the signal was ignored. Apparently still in the grip of oil euphoria, the Lopez Portillo administration bulled ahead with yet more vigorous fiscal expansion, faithfully following the dictates of the political expenditure cycle.18 Although the budget approved by Congress called for an unchanged nominal fiscal deficit, in practice no serious attempt was made to achieve fiscal restraint. Just as Echevem’a had done earlier, Lopez Portillo overrode the congressional budget by routinely authorizing ampliaciones presupuestales (“out-of-budget” expenditures). Those likely to object to such practices were simply excluded from policymaking circles: since early 1980, the head of the Central Bank had not been invited to attend economic meetings of the cabinet (Zedillo 1985, 312). In mid-1981, developments in the oil market forced a reevaluation of economic policy. The 1981 budget had been predicated on the assumption that Mexico would be able to increase its oil exports by 75 percent at a price 10 percent above that prevailing at the end of 1980 (Zedillo 1985, 313). By June it had become clear that additional oil sales could not be made without accepting a cut in price. As the perception spread that a devaluation of the peso was on the horizon, Mexdollar accounts swelled and capital flight assumed tremendous proportions. Confronted with an impending financial crisis, the Lopez Portillo administration exhibited massive inertia. The daily crawl of the peso continued at an annual rate of just 9 percent. On the fiscal front, an across-the-board, 4 percent cut in expenditures was announced, but the proposed cuts were with respect to the higher than budgeted levels of expenditures and, in any case, were not actually carried out. In fact, spending increased apace so that by the end of the year the share of public expenditures (net of interest payments on the foreign debt) in GDP had increased by six percentage points and the deficit was more than double its originally projected value.

443

MexicoIChapter 4

Foreign lending was undoubtedly an important permissive factor in allowing necessary policy reforms to be postponed. The current account deficit in 1981 reached the staggering figure of $12.5 billion, and capital flight was of the same magnitude if not several billion dollars larger. To stem the potentially large drain on reserves, the public sector contracted an additional $20.3 billion of foreign debt. The new debt was obtained at a stiff price and bought the government only a very short breathing spell. Short-term loans constituted almost half of the total, and the favorable interest rate spreads of earlier years disappeared. This is certainly one of the most puzzling episodes in recent Mexican economic history. It is more than a little difficult to understand why commercial banks were willing to grant new loans of this magnitude in the absence of any evidence that the Lopez Portillo administration was committed to reversing the direction of economic policy. l 9 4.2

The 1982 Crisis

The inability to substantively alter the course of economic policy persisted through the first half of 1982. In mid-February, the Economic Adjustment Program was announced. The program entailed an immediate 40 percent devaluation of the peso and severe import restrictions, and called for an immediate increase in public sector prices and a substantial curtailment in public expenditures over the coming year. A number of these measures, however, were introduced only after a delay of several months or were not implemented at all. The first increases in controlled prices and the prices charged for public sector goods and services did not occur until early June. And while certain expenditure cuts took place, a host of other measures strongly boosted public sector spending: the large currency devaluation of February 17 was followed three weeks later by a compensating 30 percent wage increase; in many sectors, the pressure to finish projects started in earlier years overrode the intention to reduce expenditures; to aid firms holding dollar debt, it was announced that the state would absorb 42 percent of the capital losses due to changes in the exchange rate (Villareal 1983); an emergency scheme involving fiscal relief and outright subsidies was drawn up to support “productive firms”; and finally, to fortify labor demand, CEPROFIs (Fiscal Incentive Certificates) providing tax exemptions equalling 15 percent of the monthly payroll were granted to firms that could prove they had maintained employment levels .*O The net effect of these conflicting measures was further fiscal and monetary expansion. Real public sector investment spending contracted sharply, but the various aforementioned subsidy schemes apparently induced a large increase in current expenditure. (Unaccounted for federal government transfers jumped to 8.4 percent of GNP.) Higher interest charges on the public sector foreign debt were an important but secondary factor in the loss of fiscal control. Net of interest payments on the foreign debt, real current

444

Edward F. Buffie

expenditure increased by 20.5 percent and total public sector spending by 6.9 percent (calculated by deflating by the period average CPI). As a result, despite a 37.6 percent increase in real revenues provided by PEMEX, the consolidated public sector deficit rose to 17.6 percent of GNP.21 Not surprisingly, the Economic Adjustment Program did little to allay the fears of the private sector, and the flight out of peso-denominated assets and Mexdollars recommenced on a large scale beginning in mid-March.22 In the second quarter alone, capital flight totalled $3.65 billion.23 The government staved off another sizable devaluation of the peso by borrowing an additional $5.7 billion through three medium-term, syndicated loans. The difficulty in arranging the last and largest loan ($2.5 billion), in which the federal government was the debtor, underscored the recent, rapid deterioration in Mexico’s creditworthiness: despite being offered very attractive terms, 575 of 650 banks invited to form the syndicate initially refused to subscribe to the loan facility (Zedillo 1985, 316). Finally, in August, massive capital flight forced the government’s hand.24 Additional large price increases were announced for bread and tortillas (100 percent) and gasoline (50 percent), and a dual exchange rate system was established, involving a preferential rate set initially at 50 pesos and a free rate. The preferential exchange rate was adjusted by a daily crawl and applied to most merchandise trade and foreign debt payments. All other transactions (tourism, nonessential imports, capital account items) had to be financed through the free market. The dual exchange system proved incapable of containing capital flight and was followed in short order by two desperate measures. On August 13, dollar deposits at Mexican commercial banks were converted into pesos at an unfavorable exchange rate. Less than three weeks later, on September 1, the banking system was nationalized and comprehensive exchange controls were imposed. None of this brought Mexico much closer to being able to service its foreign debt. The public sector faced $14.3 billion of payments on principal coming due in 1983 and 1984.25 The private sector’s repayment schedule was even more onerous. Of $18 billion owed by the private sector to foreign commercial banks, two-thirds was to be repaid by the end of 1984. It was soon conceded that Mexico could not adhere to the existing repayment schedule and negotiations began to restructure the public sector’s foreign debt. In the last four months of the year, a de fact0 moratorium on debt service existed; all payments on the private sector debt ceased, as did most payments of principal on the public sector debt. Nineteen eighty-two came to a close with Mexico burdened not only by an immense foreign debt but also by severe stagflation and a depressed level of private investment spending. The strong growth in notional supply that had checked inflationary pressures in previous years was reversed as extremely restrictive quotas and a series of large real devaluations of the peso forced a

MexicoIChapter 4

445

36.2 percent reduction in imports of intermediate inputs. Contraction on the supply side coupled with expansionary fiscal policy sent the inflation rate soaring to 99 percent while, for the first time since 1932, real output fell. The large fiscal deficit led to a large real increase in the domestic component of the monetary base, but expectations that the peso would be devalued and the failure to adjust deposit rates in step with inflation caused a substantial decrease in the volume of real bank funds (table 4.13). Thus, at the same time that real credit to the public sector expanded 44.8 percent, real lending to the private sector contracted 23.8 percent. The credit crunch together with a 42 percent curtailment of capital goods imports provoked a 17.3 percent decrease in real private sector investment spending. The decline in investment spending and the cutback in imported intermediates (90 percent of which go to the manufacturing sector) hit the manufacturing sector hardest. After increasing strongly in the first quarter, manufacturing sector output and employment declined by 13.2 and 9.5 percent, respectively, in the succeeding three quarters. Table 4.13

Monetary Aggregates and Real Interest Rates

Real growth ratesa Monetary base M2 M3 M4 Total stock of bank fundsb Total credit of the financial system' Credit to the private sector krcentage of GDP Monetary base M2 M3 M4 Total stock of bank funds Total credit of the financial system Credit to the private sector Real interest rates' Average real hank deposit ratef Real effective bank loan rate

1977

1978

1979

1980

1981

1982

9.8 4.6 1.5 8.9 13.2 12.0

10.6 13.3 7.0 16.0 16.9 6.2 20.3

12.5 11.3 13.9 12.8 13.7 9.3 12.5

8.7 2.7 5.6 6.6 7.8 4.5 8.8

12.8 3.6 4.5 16.4 17.1 18.8 7.5

-4.3 -22.4 -10.5 - 14.6 -15.3 14.1 -23.8

13.9 10.2 15.9 24.7 20.2 41.9

14.3 10.4 15.5 26.1 21.7 42.6 17.6

14.4 10.5 15.5 26.9 22.6 41.4 19.4

14.3 10.1 15.2 26.4 22.4 39.7 20.0

15.0 9.8 15.0 27.9 23.8 42. I 20.3

16.1 8.8 14.9 28.2 24.0 51.9 19.3

-1.1 4.3

-3.6 2.8

-9.1 3.9

-.I 17.3

-

-

-7.8 -

-58.5 -26.5

Sources: The bank deposit rate series is from Mexican Economic Outlook (CIEMEX-WHARTON). All other data is from Indicadores Economicos (Bank of Mexico). Notes: M2 = Currency held by the public =

M2

+ liquid savings accounts. M4

=

+ peso- and foreign-currency-denominated demand deposits. M3

M3

+ nonliquid (i.e., fixed-tern) savings accounts.

'Real monetary aggregates are calculated as the end-of-year balance deflated by the end-of-year CPI. bM4 less currency held by the public. 'Credit of the Central Bank, the development banks, and the commercial banks. dAverage of the end- and beginning-of-year monetary aggregate relative to GDP. 'End-of-year interest rate (December value) less the December-to-December inflation rate. weighted average of bank deposit interest rates (CPP, or costo promedio porcenruaf).

446

Edward F. Buffie

4.3 Concluding Observations The 1982 debt crisis came, ironically, at the end of a period in which the Mexican economy had been presented an exceptional opportunity to embark upon an era of high and stable growth. During 1977-82, Mexico enjoyed very favorable terms of trade and was blessed by the discovery of enormous oil wealth.26 The Lopez Portillo administration simply squandered these windfalls in a sustained bout of extraordinary fiscal indiscipline. Real public expenditures (net of interest payments on the external debt) increased by 115.1 percent between 1977 and 1982. This huge increase in fiscal spending was not matched by a similar buildup of the public sector revenue base. Instead, nonoil revenues were allowed to decline to the point where they largely offset the fiscal surplus yielded by PEMEX, and the ensuing fiscal deficits were financed by taking out ever greater amounts of foreign debt. When the inevitable reversal in net foreign lending occurred in 1982, the inconsistencies in policy immediately drove the economy into deep stagflation. Oil wealth and favorable external conditions provided such a large margin for error that, despite the policy blunders, Mexico was not in serious difficulties until 1981. At that stage, a doubling of the gasoline tax, higher public sector prices, and modest restraint in the growth of real government expenditure could have lowered the fiscal deficit to 2-4 percent of GDP in a year or two. Confoundingly, fiscal policy became not more restrained, but rather more reckless. Political pressures reinforced by the naive belief that future oil sales would cure all macroeconomic imbalances appear responsible for the total collapse in fiscal control in 1981 and 1982. The political business cycle called for strong expenditure increases in the last two years; higher public sector prices would have hurt the middle and upper-middle classes disproportionately, alienating the main political base of the PRI (Institutional Revolutionary Party)? The overly rapid accumulation of foreign debt by the Lopez Portillo administration would not have inflicted lasting damage on the economy had the funds been used to finance efficient investment projects. Unfortunately, this did not happen. According to the various estimates I discussed earlier, between 50 percent and 83 percent of the debt financed capital flight. A large portion of the remainder financed higher public sector consumption and investment. It is difficult to believe that the increase in current expenditures did much to enhance the economy’s productive capacity, particularly as the share of human-capital-related expenditures remained small. And though little hard data exists on the productivity of state-owned enterprises, there is little doubt that many of the public sector investments undertaken in this period were fundamentally unsound and have subsequently yielded very low social returns.

447

MexicoKhapter 5

After making due allowance for capital flight, the splurge in government consumption, and inefficient investments by the parastatal sector, it is difficult to escape the conclusion that Mexico obtained remarkably little for the $59.7 billion of foreign loans taken out during the Lopez Portillo years. Perhaps the best evidence in support of this conclusion is provided by the extreme hardship the economy has subsequently suffered in servicing the debt. This is the topic of the next chapter.

5

The De La Madrid Administration and the Present Crisis

The De La Madrid administration began with a two-year respite from large-scale debt service payments. On 10 December 1982, an agreement was reached with the commercial banks to reschedule $23 billion of capital payments on the public sector debt coming due between 23 August 1982 and 31 December 1984. The maturity of the debt was extended to eight years and allowed for a four-year grace period. The price for lengthening the repayment period was a 1 percent restructuring fee and an increase in the interest rate of approximately one percentage point. Whereas the previously contracted debt involved spreads of 0.83 and 0.66 percentage points over the U.S. prime rate and LIBOR, respectively, the restructured debt gave lenders the option of a rate 1.75 percentage points over the prime rate or 1.875 percentage points over LIBOR. The new (public sector) debt service schedule involved minimal amortization until the end of 1984 and then called for $61.4 billion of capital payments from 1985 to 1990. The debt restructuring at the end of 1982 was followed in 1983 by two additional, smaller reschedulings. Private firms able to convert their short-term debt into long-term debt according to government guidelines became eligible for a program of insurance against exchange rate risk (covering both principal and interest) offered by FICORCA (see sec. 8.3 below for a detailed description of the program). By the end of October, some $12 billion of private sector liabilities were tentatively covered by the FICORCA facility; almost all of this debt was renegotiated to mature at eight or more years and included a four-year grace period. Earlier, in June, $2 billion in export credits had also been rescheduled. These reschedulings were supplemented by $5 billion in new loans to the public sector: The new loans carried even harder terms than the restructured

447

MexicoKhapter 5

After making due allowance for capital flight, the splurge in government consumption, and inefficient investments by the parastatal sector, it is difficult to escape the conclusion that Mexico obtained remarkably little for the $59.7 billion of foreign loans taken out during the Lopez Portillo years. Perhaps the best evidence in support of this conclusion is provided by the extreme hardship the economy has subsequently suffered in servicing the debt. This is the topic of the next chapter.

5

The De La Madrid Administration and the Present Crisis

The De La Madrid administration began with a two-year respite from large-scale debt service payments. On 10 December 1982, an agreement was reached with the commercial banks to reschedule $23 billion of capital payments on the public sector debt coming due between 23 August 1982 and 31 December 1984. The maturity of the debt was extended to eight years and allowed for a four-year grace period. The price for lengthening the repayment period was a 1 percent restructuring fee and an increase in the interest rate of approximately one percentage point. Whereas the previously contracted debt involved spreads of 0.83 and 0.66 percentage points over the U.S. prime rate and LIBOR, respectively, the restructured debt gave lenders the option of a rate 1.75 percentage points over the prime rate or 1.875 percentage points over LIBOR. The new (public sector) debt service schedule involved minimal amortization until the end of 1984 and then called for $61.4 billion of capital payments from 1985 to 1990. The debt restructuring at the end of 1982 was followed in 1983 by two additional, smaller reschedulings. Private firms able to convert their short-term debt into long-term debt according to government guidelines became eligible for a program of insurance against exchange rate risk (covering both principal and interest) offered by FICORCA (see sec. 8.3 below for a detailed description of the program). By the end of October, some $12 billion of private sector liabilities were tentatively covered by the FICORCA facility; almost all of this debt was renegotiated to mature at eight or more years and included a four-year grace period. Earlier, in June, $2 billion in export credits had also been rescheduled. These reschedulings were supplemented by $5 billion in new loans to the public sector: The new loans carried even harder terms than the restructured

448

Edward F. Buffie

debt: for a six-year loan with a three-year grace period, a 1.25 percent commitment fee was charged and the interest rate was fixed at either (according to the lender's choice) 2.125 points over the U.S. prime rate or 2.25 points over LIBOR. During the same period in which the restructuring of the external debt was negotiated, a wide-ranging stabilization-cum-structuralreform program was agreed upon with the IMF. Sizable increases in the value-added tax, upper-bracket income tax rates, and public enterprise prices were to be combined with large expenditure cuts in an ambitious attempt to lower the public sector budget deficit to 8.5 percent of GDP in 1983. As in 1970 and 1976, along with fiscal retrenchment came trade liberalization and a whole host of measures aimed at reform of the financial system. Growth in the monetary base was to be limited to the rate consistent with the target for the fiscal deficit, and nominal interest rates were increased sharply in an effort to raise real rates and promote financial intermediation. The prevailing system of exchange controls was replaced by a dual exchange system in which most merchandise imports and all merchandise exports and debt payments were carried out at a controlled rate, while all other transactions took place at a higher, free market rate. The controlled rate was adjusted by a daily crawl and initially set at a value above that judged to be the equilibrium purchasing power parity (PPP) rate. Trade liberalization consisted of replacing import licenses by tariffs, rationalization of the tariff structure, and a gradual reduction in the overall level of protection. Contrary to widespread expectations, fiscal discipline was rigidly enforced and the ambitious goal of halving the public sector deficit relative to GDP was attained (table 5.1). The consolidated deficit declined from 17.6 to 8.9 percent, with the public sector's expenditure share falling and its revenue 'hble 5.1

Public Sector Revenues and Expenditures (% of GNP)

Expenditure Current Interest on foreign debt Other Capital Revenues Economic deficit Deficit on financial intermediationb Monetary deficit

1982

1983

1984

1985

1986P

46.4 35.8 5.1 30.7 10.6 30.1 16.3 1.4 17.6

42.8 35.0 5.1 29.9 7.8 34.3 8.5

40.3 33.4 3 .O 30.4 6.9 33.0 7.3 1.4 8.7

40.9 34.6 2.6 32.0 6.3 32.6 8.4 1.6 10.0

45.7 40.1 4.4 35.7 5.6 31.0 15.2' 1.1 16.3

.5

8.9

Sources: Data for 1982-85 are from SHCP. The 1986 figure for capital expenditure is from Informe A n d . All other data for 1986 are from Indicudores Economicos (Bank of Mexico). Current expenditure is calculated residually by subtracting capital expenditure from total expenditure. PReliminary figures. There is an inexplicable discrepancy of 474.2 billion pesos between the revenue and expenditure calculation of the deficit and the sources of funds measure of the economic deficit. bDeficit of La Banca de Desarrollo.

449

MexicoIChapter 5

share increasing by four percentage points apiece. It is important to observe, however, that the adjustment in both revenues and expenditures was highly uneven. As is clear from the breakdown of the deficit in table 5.2 and from table 5.3, greater revenue generation of PEMEX through higher internal and external sales accounted for over 100 percent of the increase in total public sector income; the revenue share of the non-PEMEX public sector actually fell (relative to GDP) by 1.7 percentage points due to a 18.6 percent decline in real income taxes and a 8.7 percent decrease in the real revenue take of the non-PEMEX parastatals. Expenditure reductions were achieved mostly through huge cuts in real public sector wages and investment spending of PEMEX and the federal government. While total real capital expenditures were cut by a thud and the real wage bill was lowered by almost a quarter, real investment spending by the non-PEMEX parastatal sector increased slightly and public sector employment rose 7.5 percent. Stringent monetary policy accompanied fiscal austerity. The real monetary base fell 12.5 percent, and real credit extended to the public sector declined by 15.3 percent. Although the real return on longer term deposits increased a b l e 5.2

PEMEX Expenditure Current Capital Revenues Deficit Non-PEMEX parastatal; Expenditure Current Capital Revenues Deficit OtheP Expenditure Current Capital Revenues Deficit

Breakdown of the Fiscal Deficit (% of GDP) 1982

1983

1984

1985

I986P

7.5 4.5 3.0 15.8 -8.3

6.2 4.2 2.0 21.5 - 15.3

5.7 4.0 1.6 19.3 - 13.6

5.1 3.8 I .3 18. I -- 13.0

5.4 4.2 1.2 13.0 -1.6

12.5 9.9 2.7 8.1 4.4

14. I 11.2 2.9 7.8 6.3

13.8 10.9 2.9 8.5 5.3

14.4 11.7 2.7 9.1 s.3

13.6 11.0 2.6 9.4 4.2

26.5 21.5 5.0 6.2 20.2

22.5 19.6 2.9 5.0 17.5

20.8 18.4 2.4 5.2 15.6

21.4 19.1 2.3 5.4 16.1

26.8 25.0 1.8 8.6 18.2

Sources: Figures for 1982 and 1983-86 are not fully comparable. For 1982, data are from Esfadisticar Anuales, 1965-1982 (SHCP). For the parastatal sector, current expenditure is calculated as operating expenditures plus a j e m de garto (outside account expenditure), and total revenue is the sum of current income, capital income, taxes paid, and ajenas de ingreso (outside account income). Data for 1983-86 are from Indicadores Economicos (Bank of Mexico). Current expenditure is operating expenditure plus variacion de cuentar ajenas (change in outside accounts). Total revenue is income (net of transfen) plus taxes paid.

Hucenhrias del Sector Publico: CI@

‘Preliminary figures. ‘Includes expenditures and revenues of DDF (Department of the Federal District) after 1982. In 1982, DDF expenditures and revenues are in “other.” bDoes not include DDF expenditures and revenues after 1982. The “out-of-budget” deficit is treated as part of current expenditures.

450 a b l e 5.3

Edward F. Buffie Real Internal Energy Prices (1980 = 100) Year

Fm Input Rice”

1960

148.0 147.2 149.2 148.5 142.1 139.4 137.6 138.0 144.1 140.0 141.3 143.1 134.7 124.6 133.1 123.8 119.9 145.7 130.8 117.5 100.0 92.1 129.4 196.I 237.3

1961 1962 1963 1964 1965

1966 1%7 1968 I969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985

Consumer Priceb

128.5 123.0 119.5

113.8 159.1 179.6 165.2 164.9 142.2 121.5 100.0

86.8 128.7 175.8 164.8 161.1

Sources: Firm input price are from the series “Combustible y Energia” in table 20.8, Estodisticas Hisforicos de Mkxico (Mexico, D.F.: INEGI, 1985). Consumer prices are from the series “Rtroleo y Derivados” in the decomposition of the consumer price index found in lndicadores Economicos (Bank of Mexico).

‘Period average input price deflated by the GDP deflator. bPeriod average price deflated by the period average CPI.

sharply, the average real interest rate paid on bank deposits remained highly negative. The low real returns together with uncertainty about the exchange rate provoked further financial disintermediation. Capital flight diminished in 1983 but still totalled approximately $8 billion; and while there was a noticeable shift toward longer-term bank deposits, all measures of financial intermediation exhibited large real declines (table 5.4). Consequently, the private sector was subjected to a sharp reduction in credit, with real loans to agriculture and industry each declining by roughly 15 percent. With debt service claiming 50 percent of total current account income and over 10 percent of national income (table 5 . 3 , it was necessary to delay trade liberalization. Stringent import controls were employed to force a 43 percent reduction in the volume of total imports (table 5.6). The private sector bore the brunt of the adjustment: the ratio of private to public sector imports fell from 1.67 to 0.99, exceeding the previous post-World War I1 low (which occurred in 1981) by some sixty-six percentage points. Even after making allowance for the unusually high level of imports in 1981, this

451

MexicoKhapter 5 Monetary Aggregates md Real Interest Rates

Table 5.4

Real growth ratesa Monetary base

1982

I983

1984

-4.3

- 12.5

-3.2

1985

-28.3 ( - 11.2)b

M2 M3 M4 Total stock of bank funds' Total credit of the financial systemd Credit to the private sector' Percentage of GNP' Monetary base

- 22.4 - 10.6

14.6 - 15.3 14.1 -23.8

-21.8 18.0 -11.6 -9.1 - 14.6 - 12.9

2.0 46.2 6.0 4.9 -6.7 12.5

16.1

14.9

13.9

-

M2 M3 M4 Total stock of bank funds Total credit of the financial system Credit to the private sector Real interest ratesg Average real cost of bank funds Effective commercial bank loan rateh

8.8 14.9 28.2 24.0 51.9 16.0

-58.5 -26.5

~

7. I 13.0 25.3 21.9 50.3 13.6

24.9 21.6 45.3 13.9

-24.1 10.4

-8.1 4.5

6.5 15.5

-6.1 18.4 - 12.9 - 12.9 9.5 - .4 ~

11.6 (13.1)b 6.5 16.0 24.0 20.8 47.8 14.8

1986

-28.2 ( - ll.7)b

- 16.3 ~

21.2

- 10.9

- 10.4 4.8 -9.6 9.0

( 12.7)b

6.2 14.0 23.4 20.3 56.5 15.3

-7.6

Sources: The nonpreferential loan rate series is from Esfodisficus de Mixico (Mtxico, D.F.: INEGI, 1985):825. All other raw financial data are from Indicadores Economicos (Bank of Mexico). Nofes: M2 M3 = M2

=

Currency held by the public

+ liquid savings accounts. M4

+ peso- and foreign-currency-denominated checking accounts.

=

M3

+ nonliquid (i.e., fixed-term) savings accounts.

"Real monetary aggregates are calculated as the end-of-year balance deflated by the end-of-year CPI 1985 the reserve requirement was lowered from 48 tu 10 percent. Simullanruusly, banks were required to use 38 percent of their funds to purchase various government assets (see n. 13). The figures in parentheses show the change in the monetary base under the assumption that the ratio of bank reserves to M4 was the same in 1985 and 1986 as in 1984. 'M4 less currency held by the public. dCredit of the Central Bank, the development banks, and the commercial banks 'Credit extended to the private sector by the Central Bank (which is negligible), the development banks, and the commercial banks. 'Average of the end- and beginning-of-year monetary aggregate relative to GDP. gEnd-of-year interest rate (December value) less the December-to-December CPI inflation rate. hThe nonpreferentid loan rate.

represents an extraordinary degree of import compression; the cutbacks in 1982 and 1983 brought the private sector import volume 25 percent below its level in 1970.' Contractionary fiscal and monetary policies, tight import restrictions, and higher real export prices produced a dramatic $11.6 billion turnaround in the current account balance. The contribution on the export side came mostly from manufacturing exports, whose 33 percent increase more than compensated for falling oil prices. The overall export volume rose by 16 percent and dollar export earnings by 5.1 percent. As in so many Fund stabilization programs in Latin America, deepening stagflation was the price exacted for improvement in the external accounts

452 Table 5.5

Edward F. Buffie Debt Burden Measures

1983 Total debt (billion $) Total debVGDP" Public sector debt service" (billion $1 8 of Merchandise exports 8 of Current account income % of GDPb Total debt serviceC(billion $) % of Merchandise exports % of Current account income % of GDP Total debt service #2d (billion $) % of Merchandise exports % of Current account income 5% of GDPb Net debv (billion $) Net debt service'(bi1lion $) 8 of Merchandise exports % of Current account income 8 of GDPb Net debt service #28 (billion $) 8 of Merchandise exports 8 of Current account income 8 of CDPb

93.8 65.8 12.3 55.0 42.4 8.6 14.6 65.4 50.4 10.2 30.5 136.8 105.5 21.4 57.9

10.7 48.0 37.0 7.5 26.7 119.5 92.1 18.7

1984

1985

96.6 56.4

11.7 48.3 35.5 6.8 14.1 58.1 42.8 8.2 27.9 115.5 84.9 16.3 52.4 8.7 36.0 26.5 5.1 22.6 93.4 68.7 13.2

*

97.3 55.1 11.1 51.3 36.1 6.3 13.0 60.1 42.3 7.4 17.5 80.8 56.9 9.9 48.2 7.9 36.4 25.6 4.5 12.4 57. I 40.2 7.0

1986 98.3 77.3 9.6 59.1 39.5 1.5 11.1 69.5 45.9 8.8 15.0 93.4 61.7 11.8 47.0 6.8 42.3 28.0 5.3 10.6 66.2 43.8 8.3

Sources: Mexican Economic Outlook (CIEMEX-WHARTON) for data on the total debt and short-term public and private sector debt. All other data comes from fndicodores Economicos (Bank of Mexico). 'Public sector interest payments and amortization of the medium- and long-term debt. bGDP mcasured in dollars was calculated by dividing nominal GDP by the period average controlled exchange rate. There is no correction for deviations of the aCNd exchange rate from the equilibrium exchange rate. 'Public sector debt service plus private sector interest payments. dThe sum of public and private sector interest payments, public sector amortization of the short-, medium, and long-term debt, and amortization of the short-term private sector debt. Amortization of the short-term debt is assumed to equal the previous period's short-term debt. 'Calculated as the cumulated value of official current account deficits starting from 195 1. 'Calculated by scaling total interest payments by the public and private sectors by the ratio of net debt to total debt. No attempt is made to adjust for the fact that the interest rate on private sector foreign assets differs from the rates charged for foreign loans to the public and private sectors. Talculated as the sum of public sector amortization of the short-. medium-, and long-termdebt, amortization of the short-term private sector debt, and net interest payments. Net interest payments are total interest payments by the public and private sectors scaled down by the ratio of net debt to total debt. Amortization of the short-term debt is assumed to equal the previous period's short-term debt.

and the public sector finances (table 5.7). Notwithstanding stiff monetary and fiscal contraction and wage restraint sufficient to produce cuts of 17 percent in the real minimum wage and 21 percent in the overall manufacturing sector wage, the inflation rate declined but slightly from 98 to 81 percent. While inflation remained high, real GDP declined 5.9 percent and aggregate underemployment increased substantially. The greatest decline in economic activity occurred in the manufacturing sector, where output decreased 7.3 percent and employment fell 6 percent.

453 'IBble 5.6

MexicoKhapter 5 External Accounts 1982

1983

I984

1985

5.4

4.2

1.2

3.4 22.3 8.6 -43.3 -30.9 -68.5 -62.2 16.1 33.9 153.3 143.5 240.8 104.4

3.7 24.2 11.3 21.5 32.5 34.1 13.7 12.8 19.6 135.0 127.1 206.0

3.8 21.7 13.2 17.1 14.8 28.0 23.3 -1.5 1.8 135.6 132.1 197.1

1986

~

Current account (billion $) Capital flight (billion $)

Mz= Merchandise exports (billion $) Merchandise imports (billion $) Volume of merchandise imports (% change) Intermediate inputs Consumer goods Capital goods Volume of merchandise exports (% change) Nonoil manufacturers Real exchange rateb Real price of total merchandise imports' Real price of total merchandise exportsd Nonoil manufacturers'

-6.2 9.0 21.2 14.4 -39.0 -36.2 -46.3 -42.1 16.9 11.1 138.8 111.2 256.9 98.8

- 1.3 16.0 11.4 - 12.0 - 13.2 -20.2 -4.7

179.4 180.7 192.9

Sources: Cumby and Levich (1987, 58); and Zedillo (1987, 177). National Income Accounts (INEGI) for traded goods price indices and the indices of import and export volumes. Idicadores Economicos (Bankof Mexico) for the current account deficit and the dollar values of imports and exports. "Modified Zedillo estimate (see table 4.12). b1980 = 100; calculated as the period average controlled exchange rate multiplied by the ratio of the US. wholesale price index (now called tbe producer price index) to the Mexican GDP deflator. '1970 = 100; deflated by the GDP deflator. For 1984-86, spliced to CIEMEX-WHARTON series (Mexican Ecommic Outlook [June 1987]:181) for the internal price deflator for exports of gooh and services. d1970 = 100, deflated by the GDP deflator. Fbr 1984-86, spliced to CIEMEX series (ibid.) for internal price deflator for imports of goods and services. 'I970 = 100; deflated by the GDP deflator.

The second straight year of severe stagflation also brought a collapse in aggregate investment spending, jeopardizing the future growth prospects of the economy. Undoubtedly, a substantial reduction in investment spending of the state-owned enterprises was in order, but private sector investment also declined to an almost equal extent (down 24.2 percent). Given the high rate of inflation that prevailed throughout 1983, it is improbable that demand contraction induced the fall in private sector investment spending. Rather, the main explanatory factors appear to lie elsewhere. Financial disintermediation and the abrupt cutoff in foreign lending led to a sharp reduction in bank credit: at 13.6 percent of GNP, real lending to the private sector stood at its lowest level since (at least) 1978. Large upward jumps in the real prices for capital goods and complementary inputs sharply diminished profit margins, reinforcing the contractionary effect of the credit squeeze.' The real price of imports rose approximately 29 percent while the increase in the real domestic price of energy inputs was an even larger 52 percent. The high real import prices led to reductions of 62.2 and 31 percent, respectively, in the volumes of imported capital goods and intermediate inputs. The large increases in the real prices of domestic and imported intermediate inputs also appear to explain the puzzling coexistence of high

Edward F. Buffie

454 Table 5.7a

Macroeconomic Aggregates ( % change)'

Real GDP Manufacturing Agriculture, forestry, and fisheries Inflationb Manufacturing employment Real investment Private Public

Table 5.7b

1982

1983

1984

1985

1986P

- .5 -2.9 - .6 98.9 - 8.5 - 15.9 - 17.3 - 14.2

-5.3 -7.3 2.8 80.8 - 6.0 - 27.9 -24.2 -32.5

3.7 4.8 2.5 59.2 2.3 5.5 9.0 .6

2.8 5.8 3.8 63.7 .2 6.4 13.4 -4.4

-3.8 - 5.6 -2.1 105.7 -6.1 - 12.2 -9.8 - 16.5

Composition of Output (% of GDP)d

Private consumption Government consumption Gross fixed capital formation Private Public Change in inventories Exports Imports

69.0 9.3 21.0 11.7 9.3 .5 10.2 10. I

67.3 9.7 16.1 9.4 6.6

I .o 12.1 6.2

66.6

66.1

10.0 16.3 9.9 6.4 I .4 12.9 7.2

9.9 16.9 10.9 6.0 2.7 12.2 7.7

66.2 10.1 15.4 10.2 5.2 2.2 13.0 6.9

~

Sources: The manufacturing sector employment series is from lndicadores Economicos (Bankof Mexico). All other data is from the National Income Accounts (INEGI).

PPreliminary figures. "Real variables are measured at 1970 prices. bDecember-to-December change in the CPI. "December-to-December change. %up111

shares at 1970 prices.

inflation and rising underemployment, on the one hand, and strongly contractionary fiscal and monetary policy and substantial wage repression, on the other. The downward pressure exerted on prices by contractionary demand policies and real wage repression seems to have been neutralized in large measure by the reduction in notional supply induced by higher intermediates prices. Similarly, since decreased usage of complementary intermediate inputs tends to lower labor's marginal physical productivity, a decline in real wages is not sufficient to guarantee a rise in employment. In fact, given the magnitude of the relative price swings that occurred in 1983, it is quite likely that the adverse productivity effect would dominate. In the next chapter, I demonstrate for a wide range of plausible technologies that when import quotas are tightened, even substantial real wage cuts will often be insufficient to prevent the emergence of open unemployment.

5.1 1984-85: Modest Recovery At the start of 1984, the economy began to recover from the 1982-83 recession. Sluggish growth in the first two quarters was followed by strong

455

MexicoIChapter 5

growth in the last half of the year, led by a resurgence in private investment spending and purchases of consumer durables. Overall GDP growth for the year was 3.7 percent. In the manufacturing sector, output increased 4.8 percent, but employment increased only 2.3 percent despite an additional 7 percent decrease in the real wage. This very modest “recovery” (from deep recession to mild recession) reflected the stimulus of a number of reflationary demand and supply-side measures implemented during the course of the year. Real public sector investment declined another 10.4 percent, but real current expenditures net of interest payments on the foreign debt rose 5.1 percent and fiscal incentives were introduced to encourage private investment spending. Most importantly, the favorable payments balance recorded in 1983 allowed import controls to be greatly relaxed: imported intermediates rose by 32.5 percent and capital goods imports by 13.7 percent in 1984, with most of the extra imports going to the private ~ e c t o r . ~ Progress on the price front coincided with recovery from the trough of the 1982-83 recession. Although the target of a 40 percent inflation rate proved unattainable, the actual inflation rate fell by twenty-one percentage points to 59.2 percent. While wage restraint helped contain inflationary pressures, the most important deflationary factor at work was the expanded flow of imports at lower real prices. After rising nearly 30 percent the previous year, the real average import price fell 11.4 percent in 1984. The payments balance also evolved favorably in 1984. Dollar export earnings rose 8.4 percent, principally from a 19.6 percent increase in the volume of nonoil exports. As the growth in imports took place from an extremely depressed level (in 1983, the aggregate [public + private] import volume was only 14.6 percent above its 1970 level), a $12.9 billion trade surplus was recorded. This translated into a current account surplus of $4.2 billion which was used to prepay part of the foreign debt and accumulate an additional $3 billion of international reserves (exceeding the target figure of $2 billion). Capital flight, however, remained a problem: $3.7 billion, the equivalent of one-third of merchandise imports in 1984, left the country. The second consecutive favorable showing in the payments balance yielded an immediate dividend in enabling Mexico to restructure its foreign debt on much better terms. In the last quarter of the year, all public sector debt payments maturing between 1985 and 1990 were rescheduled. Almost all of the $48 billion to be paid over 1985-90 was renegotiated to mature over fourteen years. The interest rate on the restructured debt was cut roughly one percentage point, and LIBOR replaced the U.S. prime rate as the reference rate for most of the debt. The adjustment program drawn up at the beginning of the De La Madrid administration was formally ended early in 1985. On 24 March 1985, a new reform program was outlined in a Letter of Intent to the IMF.4 This latest reform package emphasized the need for an accelerated pace of trade

456

Edward F. Buffie

liberalization and further reductions in the fiscal deficit. The fiscal goal was to lower the consolidated public sector deficit from 6.2 percent of GDP in 1984 to 5.1 percent in 1985. The program for trade liberalization comprised a variety of measures aimed at promoting nonoil exports and reducing the level and degree of dispersion in the structure of p r ~ t e c t i o n . Later ~ adjustments called for the existing system of import licenses to be fully replaced by the end of 1988 by a compact schedule of five tariff rates ranging from zero to 30 percent (Informe Anual 1986, 117).6 The reduction in the fiscal deficit was to be accomplished exclusively through a reduction in expenditures; the share of public sector revenues in GDP was to be kept fixed at its 1984 level by offsetting an anticipated decline in oil revenues through enlargement of the tax base and improved collection procedures (but not higher tax rates). The intended expenditure cuts were concentrated primarily in current expenditures, but also included a freeze on some 100 billion pesos (equivalent to 4 percent of the 1984 deficit) of nonpriority investments. To curb the deficit on financial intermediation, increases in the interest rates charged on preferential loans were announced and a ceiling of 350 billion pesos was placed on lending by the development banks. Lastly, a number of administrative reforms were introduced in an attempt to gain better control over public sector expenditures, especially those of the state-owned enterprises (SOEs): 1. Monthly and quarterly schedules were drawn up for SOE revenues and

2.

3.

4.

5.

6.

expenditures, and intersecretarial commissions formed to monitor progress toward the deficit targets. Commercial banks (owned by the government since 1982) have been ordered not to extend credit to SOEs or other branches of the government without receiving prior authorization from SHCP. The banks are also to furnish regular reports to Hacienda detailing their credit transactions with the public sector. The Treasury is to make payments in connection with debt service of the SOEs directly in order to avoid diversion of the funds earmarked for this purpose into other channels. Only net transfers among SOEs are to be carried out. This measure was evidently necessary because certain SOEs obtain ‘‘unbudgeted financing” by being less than scrupulous about paying their bills with other SOEs. New disbursement procedures have been instituted so that fiscal transfers will not be effected until they actually become necessary. From the monthly allocation of funds to programs and projects slated for expenditure cuts, the Treasury is to withhold a sum equivalent to 15 percent of total fiscal transfers.

The rather bizarre nature of these reforms is revealing. It is not too difficult to discern that internal organizational problems have been, and probably still

457

MexicoKhapter 5

are, a severe impediment to efforts to constrain expenditure growth. Put more plainly, the SOEs are apparently loose cannon^.^ During the course of the year it became evident that, whatever intentions may have been, the fiscal deficit was once again assuming dangerous proportions. The deficit for 1984 turned out to be 8.7 percent of GDP, not 6.2 percent as stated in the Letter of Intent. Moreover, actual revenues and expenditures moved further away from their targetted levels in 1985, causing the deficit to climb to 10 percent of GDP. Declining oil prices lowered PEMEX’s surplus, but other factors contributed as well to the growth in the deficit. While higher public sector prices raised the income of the non-PEMEX parastatals, the deficit on financial intermediation worsened considerably and general tax revenues continued to stagnate (table 5.8). Clearly, despite avowals to the contrary, no substantive effort was made to enlarge the overall tax base. Remarkably, the share of income taxes was allowed to decline 1.6 percentage points over 1981-85, pulling down the share of general tax revenues by an almost equal amount. Only part of this decline can be attributed to the depressed level of corporate profits; since 1982, the lower yield from personal income taxes accounts for nearly all of the reduction in the income tax share.8 To a substantial extent, these large fiscal deficits reflect an inflated level of current expenditures associated with the inflationary component of interest payments on the internal debt. Table 5.9 shows two calculations of the inflation-adjusted deficit (IAD) obtained using the Central Bank’s estimates of the impact of inflation on the value of peso-denominated internal debt (Informe Anual 1986, 185). The first estimate subtracts from the unadjusted monetary deficit the entire reduction in the value of the internal peso debt caused by inflation even when the implied ex post real interest rate is negative.’ The second estimate is calculated using an inflation a b l e 5.8

Public Sector Prices and Revenues’

Real public sector pricesb Revenues of non-PEMEX parastataW General tax revenues Income taxes Personal Indirect taxesd Foreign trade

1981

1982

1983

1984

1985

100.0 7.9 10.8 5.8

108.5 8.1 9.0

131.3

139.1 8.5 9.0

133.1 9.1 9.3

2.6 5.0 1.1

4.9 2.6 4.1 .90

7.8 9.0

4.2 2.0 4.8 .50

4.2 1.9

4.8 .50

4.2 2.0 5.0 .67

1986

9.4 9.4 4.3 2.0 5.1

37

Source: Indicadores Econornicos (Bankof Mexico). ‘Revenues are expressed as a percentage of GDP. bPeriod average price deflated by the period average CPI. “Sum of revenues (exclusive of any transfer payments received) plus taxes paid. dSumof value-added taxes, taxes on production and services, taxes on foreign trade, and “other” taxes. Does not include gasoline taxes (which I classify as revenues of PEMEX).

458

Edward F. Buffie Estimates of the Inflation-Adjusted Deficit (IAD), 1970-86 (9%of GDP)

Table 5.9 Year

IAD I

IAD2"

Year

IADl

IAD2"

1970 1971 1972 1973 1974 1975 1976 1977 1978

3.15 1.80 4. I7 3.95 4.54 8.33 6.22 4.30 4.76

3.41 2.12 4.59 5.09 5.43 8.81 7.28 4.96 5.22

1979 1980 1981 1982 1983 1984 1985 1986

5.04 4.59 11.29 6.99 .86 1.04 1.58 2.88

5.64

5.34 11.63 9.28 .22 1.38 1.65 2.86

~

~

Sources: Data for the inHationary component of interest payments on the peso-denominated internal debt are from Informe Anual 1986, 185. The figures used for the unadjusted deficit are from EsfadisficasHarendarias del Secfor Publico. Cifras Anuales. 1965- 1982 (SHCP) and lndicadores Economicos (Bank of Mexico).

Talculated using an inHation rate consistent with a long-term real interest rate of 5 percent

rate consistent with a value of 5 percent for the long-run real CETES (government bonds) rate. According to both estimates, the IAD has been small in recent years. lo It is often claimed that a low value for the IAD indicates that inflation is mostly inertial and additional fiscal adjustment is unnecessary. But while the IAD is clearly a better measure than the actual budget deficit, the latter conclusions are not at all obvious. Consider the nature of the savings-investment constraint in an open economy experiencing ongoing inflation. In an inflationary context, the proper definition of disposable income incorporates anticipated capital gains or losses on different assets (see Turnovsky 1977, chap. 3). Thus, under the assumption of perfect foresight, the private sector budget constraint reads" C

(1)

+ S = Y + iB

- T - ITM,

where C is real consumption, Y is real income, B is the real value of government (peso-denominated) debt, i is the real interest rate on government debt, IT is the inflation rate, M is the real stock of high-powered money, T is real taxes, and S is real net savings (gross savings equal S + ITM). Equation ( 1 ) may be rewritten as

(2)

C

+S

=

Y

+ IAD - G -

DS

-

ITM,

where DS is real interest payments on the public sector external debt, G is other real government expenditure, and IAD = G DS iB - T . Substituting for Y from the goods market identity gives

+

(3)

S - I = (NX

-

DS) + I A D

+

- ITM

where I and NX denote, respectively, private investment and net exports. The first term on the right side is the current account surplus. Now suppose the IAD is independent of IT and consider the tradeoffs offered by a Cruzado or Austral plan that eliminates the inflation tax.I2 Then (3) implies that

459

MexicoIChapter 5

unless a reduction in the inflation tax increases real net savings one-for-one, either the current account balance will deteriorate or investment will decline (as happened in Argentina and Brazil). With less forced savings generated through the inflation tax, some other component of savings must rise to prevent a decrease in in~estment.'~ If the size of the current account surplus (foreign savings) is tightly constrained by the schedule for debt repayment, measures must be taken either to stimulate private savings or further lower the IAD. A zero or even negative value of the IAD does not eliminate the need for fiscal adjustment. In the Mexican case, the impact of the mounting fiscal deficit was felt most strongly in financial markets. To lessen inflationary pressures, strict control over the growth rate of the monetary base was maintained.14 Consequently, a large part of the deficit had to be financed by the sale of CETES to the banking system and the public (table 5.10). Early in the year the decision was made to place 250 billion pesos of CETES with Banca Multiple. After October, lending to the private sector was frozen and virtually all excess bank funds were diverted to purchases of various government-issued assets (CETES, petrobonds, etc.). l5 Predictably, financing the deficit in this fashion led to generally rising interest rates and a strong contraction in lending to the private sector. Whereas the inflation rate (December to December) rose slightly from 59.2 to 63.7 percent, the average cost of bank funds increased each month, rising from 47.5 percent in December 1984 to 65.7 percent in December 1985. During the same period, the yield on three-month CETES jumped from 49.2 to 74.1 percent. The increased interest rate spread between CETES and bank funds provoked a new wave of financial disintermediation, reversing the gains made in 1984. While real credit to the public sector increased 12.6 percent in 1985, the real stock of bank funds fell 12.9 percent and real credit (of the entire financial system) to the private sector contracted slightly. The severe credit squeeze imposed on the private sector, falling oil prices, and the catastrophic September earthquake in Mexico City pushed the economy back into recession. In the second half of 1985, real GDP growth turned negative as private investment spending and manufacturing sector output contracted sharply. Overall GDP growth for the year was just 2.7 Real Internal Debt, 1979-86

lsble 5.10

~

Total" %Change % of GDP

1979

1980

1981

1982

1983

1984

1985

1986

100.0

107.8 7.8 18.9

126.7 17.5 20.8

175.9 38.8 35.9

155.3 -11.7 31.4

137.9 -11.2 26.5

147.4 6.9 29.2

167.7 13.8 40.1

~~~

-

18.8

Source: Data on the total internal debt are. from CIEMEX-WHARTON, Mexican Economic Outlook, 19, no. 2 (1987): 188. 'End-of-year stock of debt deflated by the end-of-year CPI. Includes both pesodenominated and dollar-indexed debt.

460

Edward F. Buffie

percent. In the manufacturing sector, output expanded a respectable 5.8 percent, but employment failed to increase. As the economy slid back into recession in 1985, the large payments surpluses of the preceding two years gave way to an overall deficit. Real import payments rose by 19.9 percent, with public sector imports declining 9.1 percent and private sector imports increasing 41.2 percent.16 Total export earnings decreased $2.33 billion. Falling world market oil prices accounted for much of the decline, but nonoil export earnings also decreased $500 million. The steep decline in export earnings combined with the resurgence of private sector import demand to cut the current account surplus from $4.2 billion in 1984 to $1.24 billion in 1985. Unfortunately, the movement toward current account balance was not matched by a similar movement in the capital account. While the long-run capital account registered a small surplus, capital flight removed $3.8 billion from the country ($9.9 billion after adjusting for underinvoicing of exports), forcing the Central Bank to absorb a $2.3 billion decrease in its gross reserve holdings.”

5.2 The 1986 Oil Crisis Four months after the devastating earthquake, the Mexican economy was battered by a second severe shock. The government’s economic program for 1986 presumed that the average price for Mexican oil would drop 9 percent in world markets to $23 per barrel (Znfirme Anual 1986, 17). Early in the year, however, prices began plummeting, and by July Mexican crude was fetching only $8.45 per barrel. Prices recovered somewhat thereafter, but the average price for the year still came to only $1 1.82, 53 percent below the 1985 average of $25.35 (113, 178). The dollar value of oil exports declined $8.5 billion, a loss equivalent to 6.7 percent of the 1985 GDP (17). Following previous declines, this brought the country’s terms of trade to its lowest level in more than thirty years (table 5.11). Adjusted for changes in world market interest rates, Mexico’s terms of trade had deteriorated nearly 60 percent since 1981 and over 40 percent since 1970. The De La Madrid administration responded to the oil price shock by digging its heels in deeper. Essentially, the pre-shock policy course was continued, but with an extra measure of austerity. To blunt the impact on the trade balance, the rate of depreciation of the peso was accelerated strongly, producing, by the year’s end, a 32 percent increase in the real exchange rate. A real devaluation of this magnitude, it was conceded, would be strongly stagflationary in the short run. In view of the experience in 1982-83, however, the alternative of imposing tight import quotas was judged to be even worse (Informe Anual 1986, 22). Aggressive devaluation was supplemented by limited fiscal adjustment and extremely contractionary monetary policy. Some new expenditure cuts and tax increases were introduced, but these measures fell far short of neutralizing

461

Mexico/Chapter 5 Terms of Trade Indices

Table 5.11 Year

Unadjusted

Adjusted"

Year

Unadjusted

Adjusteda

1960 1961 I962 1963 1964 1965 1966 1967 1968 1969

87.8 89.6 83.9 89.1 85.7 84.1 85.2 83.9 89.0 88.0

87.2 88.3 82.0 87.3 84.8 83.5 85.1 84.1 90.4 92.2

1974 1975 1976 1977 1978 1979

100.1 97.8 113.0 113.0 113.I

116.4 96.1 119.2 121.9 115.6 110.4

1970 1971 1972 1973

96.7 100.0 103.3 115.2

100.0 100.0 100.7 121.2

127.6 124.3 108.2 98.8 97.1 91.9 65.6

123.5 127.4 94.5 77.4 66.5 71.5 54.6

1980 1981 1982 1983 1984 1985 1986

104. I

Source: lnforme Anual 1986. "Adjusted for changes in international interest rates

the impact of the oil price drop, and the consolidated public sector deficit soared to 16.3 percent of GNP. As in previous years, the fiscal deficit was financed largely by depriving the private sector of credit. The real monetary base fell sharply, while the real internal debt rose 13.8 percent. Even though nearly all marginal bank credit (77-92 percent) remained reserved for the public sector, the large increase in bond supply could not be absorbed without inducing a strong rise in real interest rates. The real (compounded, annual equivalent) interest rate paid by three-month CETES averaged 19.54 percent. Bank rates followed suit. The average cost of real bank funds was 6.3 percent, and the real nonpreferential loan rate fluctuated between 13 and 18.2 percent (lnforme Anual 1986, 27). Renewed austerity coming on top of the terms of trade loss brought the weak 1984-85 recovery to a grinding halt: real output declined 3.8 percent, real investment 12.2 percent, and manufacturing sector employment 6.7 percent, while the inflation rate jumped from 63.2 to 105.7 percent in 1986. Despite the introduction of quarterly wage adjustments, the real (minimum) wage fell (8.4 percent) for the fifth consecutive year. The sole consolation was an unexpected improvement in the overall payments balance. High domestic interest rates elicited a substantial capital inflow which, together with a two percentage point fall in the interest rate applicable to the foreign debt (lnforme Anual 1986, 25) (keeping the current account deficit to $1.3 billion), enabled Central Bank reserves to rise by $950 million. In an attempt to head off the growing crisis, a standby agreement was signed with the IMF and a large-scale restructuring of the debt negotiated in the last half of the year. This latest debt package provided $12 billion of new funds for 1987 and 1988 and restructured the old debt on very favorable

462

Edward F. Buffie

terms for Mexico. Capital payments coming due between 1985 and 1990 in the amount of $51.2 billion have been rescheduled to mature over twenty years with a seven-year grace period. Interest charges have been lowered substantially through a reduction in the LIBOR spread from 121 to 80 basis points and the replacement of the U.S. prime rate by LIBOR or an average of three-month CD rates quoted in various countries."

5.3 Post-1982 Economic Policy: An Evaluation Judged against almost any set of economic, criterion, the post-1982 adjustment record has been a dismal failure. At the end of 1986, real output stood slightly below its 1982 level and the inflation rate had accelerated to over 100 percent. In per capita terms, real income fell 11 percent during this four-year period, with labor bearing the brunt of the decline: since 1982, real wages have decreased approximately 30 percent (32 percent for the minimum wage), falling far below the levels that prevailed at the end of the Stabilizing Development (table 5.12). l9 Nor do the prospects for recovery look particularly promising. Both private and public sector investment remain heavily depressed, and while large current account surpluses were achieved in 1983 and 1984, by early 1986 balance of payments problems had emerged once again. Adverse external shocks and the burden of servicing the debt made some deterioration in the economy's performance inevitable. Over 1983- 86, Mexico's terms of trade (adjusted for changes in world market interest rates) declined 42.2 percent, the most severe blow coming in 1986 with the collapse of world market oil prices. The worsening terms of trade coupled with debt service claiming 40-50 percent of total current account income forced an extraordinary degree of import compression upon the private sector.*' Contrary to textbook models, import compression is almost certain to bc strongly stagflationary. Thc reason for this is simply that in Mexico, as in most LDCs, intermediate inputs and capital goods account for over 80 percent of total imports. On the normal assumption that factors of production are gross complements, a reduction in imports, whether imposed directly by import controls or induced by a real devaluation, exerts a powerful contractionary effect upon economic activity. Cutbacks in imported intermediates lower labor demand at a given real wage and discourage investment by reducing the productivity of capital. Restrictions on capital goods imports further depress investment by raising the supply price of capital. Since most imported machinery lacks close domestic substitutes, there is little, if any, demand stimulus created by expenditure switching; instead, the construction sector goes into a slump as investment orders fall off sharply. Figure 5.1 shows how the private sector import volume evolved from 1970 to 1986. Clearly, import compression went far beyond simply offsetting the rapid growth of the oil boom years. Between 1981 and 1983, real private

463

MexicoIChapter5

Table 5.12

Real Wages (1970 = 100)” Manufacturing Sector

Year

Average Minimum Wageb

1964 1965 1966 I967 I968 1969

76.4 74.8 85.7 83.3 94.7 91.0

1970 1971 1972 1973 1974 1975 1976 1977 1978 1979

100.0 95.0 107.6 101.1 110.8 112.1 124.7 123.9 119.7 117.2

1980 1981 1982 1983 1984 1985 1986

109.0 110.8 91.9 81.4 75.9 74.9 68.6

OverallC

Blue-Collar

97.9 99.5 100.0

101.2 106.7 104.3 104.5 107.8 116.7 118.5 116.2 114.5 111.2 115.1 114.1 90.3 84.0 85.2

100.0 100.7 106.6 103.9 106.6 110.8 123.2 125.3 121.9 119.9 114.8 116.1

116.9 87.0 83.5

Sources: Minimum wage data are from INEGI. The blue-collar and overall wage series for the manufacturing sector are from Encuesta Industrial Mensual, as reported in Indices de Precios (February 1986). “Period average wage index deflated by the period average CPI. %e minimum wage index is a weighted average of minimum wages in different regions, where the weights are given by the region’s share in the total salaried population of the nation. In years in which there was more than one wage adjustment, the period average figure is generated by weighting the wage in each subperiod by the fraction of the year during which it prevailed. ‘Composite index for manufacturing sector wages and salaries inclusive of fringe benefits.

sector imports were cut 73 percent; even after two years of “recovery,” the import volume in 1985 barely exceeded its 1978 level. But while the terms of trade shock and the burden of debt service made some contraction inevitable, errant policy must also shoulder a good portion of the blame for the post-1982 debacle. The excessive use of quantitative restrictions to regulate the current account caused import compression to be deeper and more prolonged than necessary. In view of the complementary nature of domestic factors and imports, policy should have been directed far more toward promoting manufacturing and agricultural exports so as to minimize the impact of debt service on import flows. Although some export promotion measures were introduced, others were withdrawn and nonoil exporters remain handicapped by controls aimed at ensuring “sufficient” supply for the domestic market. The CEDIS scheme of subsidies has been largely phased out and credit subsidies for exporters sharply curtailed.”’

464

Edward F. Buffie

270

-

250 -

230 210 190 -

170 150 -

130 110 90 -

70 ,

I

1970

Fig. 5.1

4

1

I

1

,

1975

1

,

,

1

1

1980

,

,

1

,

,

1985

The private sector import volume, 1970-86

Sources: Indicadores Economicos (Bank of Mexico) for dollar import values and the National Income Accounts (INEGI) for the aggregate import volume. Nore: The private sector import volume is estimated as the aggregate import volume multiplied by the ratio of the dollar value of private sector merchandise imports to the dollar value of total merchandise imports. The 1986 aggregate import volume is estimated by deflating the dollar value of total merchandise imports by the dollar price index for imports (the index given in Informe Anual I985 [201], updated to 1986 using the change in import prices in Indicadores Economicos [April 19871) and then splicing to the National Income Accounts index for the aggregate import volume.

Export controls have been loosened somewhat, but at the end of 1985 44.1 percent of nonoil exports were still subject to (domestic) quotas and nominal protection of the agricultural sector was negative (World Bank 1986, 12- 13). Fiscal and interest rate policy intensified the contractionary blow delivered against the private sector by import compression. The fiscal deficit diminished but remained large and was financed in good measure by imposing high marginal reserve requirements on the banking system. In addition, the deficit drove bond rates strongly upward which, together with the failure to raise deposit rates to positive levels, provoked a sharp decrease in the supply of bank funds. Financial disintermediation and the redirection of credit toward the public sector have depressed bank loans to the private sector to historically

465

Mexico/Chapter 6

low levels: total real credit to the private sector was 11.7 percent lower in 1986 than in 1978, and the share of M4 in GDP was at its lowest level since 1965. Finally, private investment was further depressed by the manner in which fiscal cuts were achieved. No doubt a substantial reduction in parastatal investment spending was necessary, but investment expenditures for infrastructure capital were also severely slashed. As many types of infrastructure capital enhance the productivity of private sector capital, the latter cutbacks, like import compression and the reduction in bank lending, lowered the profitability of private investment. Once investment declines, it is easy for the economy to slip into a downward spiral in which capital decumulation, rising inflation, and growing fiscal deficits become mutually reinforcing. As lower investment rates take their toll on the capital stock, output declines and inflation accelerates. For a given level of real government expenditures, the decline in real output widens the fiscal deficit by lowering real tax revenues. If the larger deficit is financed by printing money, inflation rises further (the budget is “balanced” by the inflation tax), triggering a new round of financial disintermediation and capital decumulation. If an attempt is made to contain inflationary pressures by covering the revenue shortfall through greater bond sales, the bond rate jumps upward and again the outcome is further financial disintermediation, capital decumulation, and intensified inflationary pressures. In the next two chapters, formal theoretical models are developed in an attempt to gain a fuller understanding of the factors that seem to be driving the Mexican economy into a low growth, low real wage, high inflation, high underemployment equilibrium. In chapter 6 I analyze the repercussions of import compression on real wages and underemployment, while chapter 7 is an investigation of the links between capital accumulation, inflation, fiscal deficits, and financial disintermediation.

6

Import Compression, Real Wages, and Underemployment

Perhaps the most striking aspect of the post-1982 adjustment process has been the imposition of an extreme and unprecedented degree of import compression upon the private sector. Highly restrictive quotas were placed on imports of all types between 1982 and 1984 as part of a comprehensive stabilization program aimed first and foremost at eliminating the current account deficit. On 25 July 1985, a large number of quotas were eliminated, but, until very recently, import controls (licenses, high “reference” prices

465

Mexico/Chapter 6

low levels: total real credit to the private sector was 11.7 percent lower in 1986 than in 1978, and the share of M4 in GDP was at its lowest level since 1965. Finally, private investment was further depressed by the manner in which fiscal cuts were achieved. No doubt a substantial reduction in parastatal investment spending was necessary, but investment expenditures for infrastructure capital were also severely slashed. As many types of infrastructure capital enhance the productivity of private sector capital, the latter cutbacks, like import compression and the reduction in bank lending, lowered the profitability of private investment. Once investment declines, it is easy for the economy to slip into a downward spiral in which capital decumulation, rising inflation, and growing fiscal deficits become mutually reinforcing. As lower investment rates take their toll on the capital stock, output declines and inflation accelerates. For a given level of real government expenditures, the decline in real output widens the fiscal deficit by lowering real tax revenues. If the larger deficit is financed by printing money, inflation rises further (the budget is “balanced” by the inflation tax), triggering a new round of financial disintermediation and capital decumulation. If an attempt is made to contain inflationary pressures by covering the revenue shortfall through greater bond sales, the bond rate jumps upward and again the outcome is further financial disintermediation, capital decumulation, and intensified inflationary pressures. In the next two chapters, formal theoretical models are developed in an attempt to gain a fuller understanding of the factors that seem to be driving the Mexican economy into a low growth, low real wage, high inflation, high underemployment equilibrium. In chapter 6 I analyze the repercussions of import compression on real wages and underemployment, while chapter 7 is an investigation of the links between capital accumulation, inflation, fiscal deficits, and financial disintermediation.

6

Import Compression, Real Wages, and Underemployment

Perhaps the most striking aspect of the post-1982 adjustment process has been the imposition of an extreme and unprecedented degree of import compression upon the private sector. Highly restrictive quotas were placed on imports of all types between 1982 and 1984 as part of a comprehensive stabilization program aimed first and foremost at eliminating the current account deficit. On 25 July 1985, a large number of quotas were eliminated, but, until very recently, import controls (licenses, high “reference” prices

466

Edward F. Buffie

and domestic content requirements) were widespread. Even now that liberalization is quite far along, a severely depreciated peso has kept the import volume heavily depressed. In this chapter, I construct a simple model showing how import quotas affect the price level, underemployment, and the payments balance. The model suggests that import compression has been one of the critical causal factors underlying the post-1982 slide into low growth, declining real wages, and worsening underemployment. Under very weak conditions, a tighter quota on imported inputs used by the protected (or nontradables) manufacturing sector produces open unemployment in the short run. Over time, open unemployment diminishes and is eventually eliminated altogether as the nominal (and real) wage outside the manufacturing sector declines. Even in the long run, however, the contractionary effects of import compression do not fully disappear; if the import restrictions are maintained, the economy settles into a new steady-state equilibrium characterized by permanently higher underemployment and lower real output. The economy is not necessarily recompensed for tolerating high underemployment and low real wages by a strong improvement in the trade balance. In fact, if quotas are highly restrictive (in the sense of generating large implicit tariffs) and nontradables and tradables are relatively good substitutes in consumption, the trade balance worsens. This is, perhaps, an extreme result, but it serves to underscore the point that the cumulative trade surplus generated by a prolonged recession may be disappointingly small. This seems to be roughly consistent with recent Mexican experience: sizable trade surpluses were achieved in 1983 and 1984, but payments difficulties resurfaced by the end of 1985, when the economy was still deeply mired in recession. In section 6.3 I demonstrate that higher internal energy prices are also likely to have contributed to the protracted post-1982 recession. An increase in the domestic price of oil exacerbates underemployment in much the same fashion as import compression. The overall impact on output and the payments balance may be either positive or negative depending on how the magnitude of the loss from a more severely distorted labor market compares to the direct efficiency gain brought about by reducing the gap between the domestic and world market price of oil. Given the difficulties associated with using import restrictions to redress payments deficits, it is natural to ask whether other policies offer more favorable macroeconomic tradeoffs. This issue is tackled in sections 6.4 and 6.5, where the effectiveness of devaluation and export subsidies is analyzed. Devaluation succeeds in improving the trade balance but is contractionary: real output always declines, and if tradables sector technology is relatively inflexible, the adjustment process entails open unemployment and declining real wages over the short and medium run. Export subsidies usually exert a more favorable effect upon labor demand and have the potential to simul-

467

Mexico/Chapter 6

taneously improve the payments balance and reduce underemployment provided they are financed by tax increases. 6.1

A Simple Macroeconomic Model With Import Quotas

The main message I wish to convey is that import compression gives rise to severe macroeconomic problems largely through adverse effects operating on the supply side. Accordingly, the model that follows has a very simple macroeconomic structure along the lines of models found in the monetary approach to the balance of payments (Dornbusch 1973, 1974; Frenkel and Johnson 1976). The economy produces two goods, a nontradedlquota-protected manufactured good and an export good. (An exogenous, oil-earning sector may be added to the model, in which case the economy may be a net importer of the so called “export” good.) Each good is produced by labor, a noncompetitive imported input, and fixed, sector-specific capital under conditions of constant returns to scale. Consumer imports have already been banned (or are negligible), so that when balance of payments problems arise, foreign exchange rationing necessarily entails a reduction in imports of intermediate inputs. (In 1986, consumer imports were 5.5 percent of total private sector imports.) There is no government sector, no banking system, and no capital accumulation. In addition, it is assumed that all wealth is held in the form of high-powered money; nothing substantive changes if domestic residents are also permitted to hold a foreign currency denominated asset. The economy is small in world markets, and units are chosen so that all world market prices equal unity. The peso price of the export good, P,, thus equals the exchange rate, e:

’

(1)

P, = e.

Firms in each sector are perfectly competitive. Perfect Competition together with constant returns to scale implies that the zero profit condition obtains: Pi = ~ ’ ( w ’g’, , r’), i = n,x,

where C’, wi, g’, and ri are, respectively, the unit cost function, the nominal wage, the price of the imported input, and the capital rental in sector i. Different g’ are specified in order to allow for the possibility that quotas will be imposed in one sector but not the other; in particular, many governments appear to find the direct but naive policy of rationing imported inputs to the nontradables sector-the sector that does not earn foreign exchange-to be an attractive method of treating a payments deficit. The wage also differs across sectors. This reflects the functioning of a dualistic labor market. In the protected manufacturing sector, the bargaining

468

Edward F. Buffie

power of unions or adjustments in a binding minimum wage ensure that the nominal wage responds at least partially to general increases in the price level: (3)

+" = a ( y n P n + y x P x ) , y x + y n = l ,

a =z 1,

where yi is the consumption share of good i; a is the wage indexation parameter, and a circumflex denotes the percentage change in a variable. I treat a as an exogenous, politically determined variable and characterize later results in terms of cuts in the real wage required to avoid a worsening in underemployment. It would be preferable, of course, to have a more complete model in which ci was determined endogenously. Unfortunately, empirical knowledge about the wage-setting process in Mexico is very limited and does not permit one to choose with much confidence any of the competing theories of wage rigidity (implicit contracts, the efficiency wage hypothesis, optimizing unions, etc.). Although wage rigidity is not confined exclusively to the protected manufacturing sector, it is clear that labor markets are considerably more competitive elsewhere in the economy. In 1977 average earnings of agricultural laborers were 34 percent of those in the industrial sector and 70.8 percent of the labor force in agriculture earned less than the minimum wage.' In view of these stylized facts, I assume both that wx is substantially below W" and that wx adjusts so as to clear the labor market.3 The degree of wage flexibility in the export sector, however, is asymmetric; wx rises instantaneously to eliminate any excess demand but is inflexible in the downward direction, declining slowly in response to the pressure created by open unemployment: (4)

W X I

L"

+ L" < L

= p(L"

+ L"

-

L),

p > 0,

where L' stands for employment in sector i, L is the fixed supply of labor, and an overdot signifies a time derivative. Employing Shephard's lemma, the sectoral labor demands may be stated as

where Qiis output in sector i. Similarly, the market-clearing conditions for the sector-specific capital stocks, K', and the sectoral allocation of imported inputs, z', are

(7)

I'

=

CiQ'.

Equations ( 1)-(7) characterize supply behavior. The remaining equations complete the model by specifying the demand side of the economy and the dynamics of wealth accumulation.

469

Mexico/Chapter6

References are homothetic, and demand depends upon prices and total consumption expenditure, E. P, fluctuates to continuously clear the market:

D"(P,, P,, E )

(8)

= Q"

.

Expenditure, by definition, is the difference between income, Y, and savings, S: (9)

E = Y - S ,

where income is the sum of private sector value-added and import premiums accruing to license holders: (10)

Y = R(P,, P,, gx, g", L", L") + (g" - e)Zx

+ (g" - e)Zn .

R is the private sector value-added function and possesses the properties: aRiapi = aRiagi = -zi, aRiaLi = wi. Finally, saving is motivated by the difference between the desired and the actual stock of wealth. In the current model, wealth is held entirely in the form of domestic money, M. Assuming desired wealth is a fixed multiple, h, of income, the savings function takes the form

ei,

(11)

S

=

@(hY - M), CD > 0.

The nominal money stock is fixed in the short run but changes over time through the trade balance, B. From (9) and (10): (12)

M = B = S .

Equation (12) is the classic specie-flow mechanism: the trade balance is the vehicle through which saving brings the actual stock of wealth into equality with its desired level.

6.2 Import Compression Suppose the government finds itself faced with increased debt service obligations and attempts to engineer a series of trade balance surpluses by restricting the supply of imported inputs to firms operating in the nontradables sector. The rationale typically put forward on behalf of this policy is that the imposition of quotas is less inflationary than a comparable devaluation of the currency (though this is far from obvious) and that the export sector should be exempt from restrictions since it processes imported inputs in order to e m foreign exchange. What, if any, are the merits of this argument? To figure out the repercussions on employment and prices, differentiate (5) and utilize Uzawa's result that uij = C,C/CiCj , where uij is the Allen partial elasticity of substitution between factors i and j. While it is not difficult to work out a set of fully general results, little of interest is lost if, for algebraic simplicity, production functions are assumed to be separable between primary factors and imported inputs, viz.: , a = uKI = avI, where uvIis the elasticity of substitution between value-added and imported

470

Edward F. Buffie

inputs. Letting 8; represent the cost share of factor j in sector i, in the appendix to this chapter I show that in the separable case u2g"

(13)

jn

(14)

P' = blp,, - b27',

= alpn -

where

As expected, a , , b, , and b, are all positive. The unconditional cross-price elasticity (-a,) is negative. A higher price for the imported input both contracts the profit-maximizing level of output and induces substitution toward labor. With a separable production function (or any production function for which [uu - uK,]is not extremely large), the output effect dominates, making L" and I" gross complements. In the export sector, the output price and the price of the imported intermediate are fixed at e , the nominal exchange rate. Employment, therefore, varies only insofar as the nominal wage changes:

i"

(15)

= -a3GX,

where a3 = ( 1 -

e;)

[ u ~ ~-( e;) i

+ u;,,e;l

> 0.

When labor demand in the nontradables sector increases, w xis bid upward to clear the labor market and L" declines. If L" falls, w" is temporarily rigid and L" is unchanged in the short run. The equilibrium value of P , is found from the market-clearing condition (8). Choosing units so that P , = 1 initially and noting that $" = 0;fi" + O;p, (8), (9), and (11) yield (16)

(0; - Q'W;u, - Q"e;b,)p, + Q78;u2

+ 8 , " b 2 ) r = ~ ,-(s1) d Y ,

where ci denotes the marginal propensity to consume good i, and s = @h, the short-run marginal propensity to save. Defining t = g" - 1 , the implicit tariff on I", from (10) and (13) we have (17)

d~

=

~ " ( -i

e;fa,)P,, - pe;a,g.

+ tdi"

in the case where L" decline^.^ Substituting for Y in (16) from (17), employing the Slutsky decomposition, and collecting terms yields

471

(18)

MexicoIChapter 6

fop, + fig"

= c,(l - s)tffZ",

where

and E is the compensated own-price elasticity of demand for nontradables and is defined to be positive. Equations (14) and (18) can be solved for P, and P,. Unsurprisingly, a more restrictive quota always raises both prices:

where

A = (E + C,4b2 + e;(cx + c,s)u;,(i

- y,a)[~:~(i - e;)

+ u;,e;pe;

> 0.

Turning back to (13) and substituting the above solutions for P, and g", we obtain, after simplification, the following critical condition:

(21)

i" - 5 0, as E + c,s i"

S u;,(l

-

y,a)

Figure 6.1 illustrates the conflicting factors influencing nontradables labor demand. The reduction in imported intermediates simultaneously shifts the marginal product schedule leftward and cuts the product wage W n by pushing up P,. If there is little substitutability between imported intermediates and domestic factors (avrsmall), employment is likely to fall as the labor demand curve is steeply sloped and the leftward shift of the schedule is large. The extent to which P, increases depends on the degree of substitutability between nontradables and exportables (E), the reduction in demand elicited by the fall in real wealth (c,s), and the contraction in demand stemming from the direct fall in real income caused by the tightening of the quota [c,(l - s)t(l t)-']. For any given decline in P,, the product wage falls by the amount (1 - y,a). Large values of E, c,, t, a, and 'y, thus limit the fall in the product wage and increase the probability of a decline in labor demand. Though there are circumstances in which employment increases, in my view (21) argues that a contractionary outcome is likely. On the right side, uv, is scaled down by the product of two fractions. Thus, employment

+

472

Edward F. Buffie

gnlMpL

I

Fig. 6.1 The impact of a tighter quota on labor demand in the nontradables sector

contracts unless the degree of substitutability is considerably greater on the supply side than on the demand side. Observe as well that deindexation of wages does not necessarily suffice to prevent a reduction in labor demand-a nominal wage cut may be required. The value of a that leaves employment unchanged is

a*

= y;l'

[

1-

E

+ c,s

[l - C"(1 - s)t/(l

+

I

+ t)]u$,'

which is negative, for example, when (E c,s) > uv,. Even when u$, is relatively large, the requisite value of a will often be extremely low, particularly if t is initially nontrivial in magnitude. Table 6.1 shows how a* varies with uv, , E, and t when c, = yn = 0.50 and s = 0.20. In all cases where uv, =S 2~ and t 3 0.20, a* is far below unity. Consider next the question of whether import restrictions are efficacious in strengthening the trade balance. From (11) and (12) it is seen that the trade balance improves when nominal income rises, inducing increased savings. The general solution, therefore, may be obtained by replacing P, and g" in (17) by the expressions in (19) and (20). However, given that my objective is only to show that there is no strong presumption in favor of an improvement in the trade balance, in the remainder of this section I shall derive results under the following two simplifying assumptions: (1) there is a rigid real wage in the protected manufacturing sector (a = l), and (2) the nontradables sector production function is fully separable (avr= ,a = u,). In this case, it turns out that the trade balance worsens if and only if

MexicoKhapter 6

413 Table 6.1

Critical Degree of Real Wage Rigidity (a*) t

0

.2

.4

.6

.8

1 .0

E

0 1 .0 >I >I >I

NV .93 >I >I >l

NV

NV .82

NV

.87 >1

>1

>I

>I >l

>I >I >1

.I0

>I

NV .75 >I >I >I

.2 .4 .6

NV 0

NV NV .43 .82

NV NV .33

1

NV NV .49 .87

NV NV .38

.8

NV NV .57 .93 >I

.75

21

>1

.78 >I

1 .oo

NV NV NV .39 .71

NV NV NV .31 .65

NV NV NV .24 .59

NV NV NV .I8 .54

NV NV NV .I2 .50

UVI

.2 .4 .6

.8 1 .0

.67

.oo

1 .0

>I

.2 .4

.6

NV NV 0

.8

.50

1 .0

.80

NV

=

.78

.30

SO

negative value.

The trade balance is likely to deteriorate when the impact upon the price level is comparatively weak and the reduction in real income due to lower employment and the worsening of the trade distortion is comparatively large. As expected in light of the previous analysis of the employment outcome, the rise in P, is small relative to the real income loss when E, t, and 0;: are large and anis small. Viewed from another angle, under these circumstances the trade balance worsens because the rise in P, induces an increase in consumption of the exportable that lowers export earnings by an amount exceeding the enforced reduction in intermediates imports. What is surprising about the condition stated in (22) is that, if existing quotas are somewhat restrictive, there is no general presumption that the trade balance will improve. For t = 0, (22) requires implausibly large values of E. But when t is on the order of 0.30 or more, the trade balance may worsen when E assumes believable values. For example, with yx = cx = 0.25, 6; = t = 0.40, and an = 0.50, trade deficits arise whenever E > 1.06. So far, I have dealt only with the short run. Beyond the short run, the payments balance alters the stock of money balances and wx begins to decline as workers laid off in the nontradables sector seek employment in the export sector. Can it be hoped that after these adjustments are complete the contractionary effect of import compression will be reversed or at least substantially blunted?

474

Edward F. Buffie

Figure 6.2 depicts the nature of the adjustment process. The M M and LL schedules reflect, respectively, the set of points for which the trade balance is zero and the labor market clears. These schedules are derived by substituting the reduced-form solutions for Y and L' into (4) and (1 1):

ni

(23) (24)

= B = @[hY(M, w") -

MI

P[L"(M,w") + L"(w") - L ] , LT > 0, Y 2 , L Y , L'f < 0, hY, - 1 < 0. W" =

A higher level of wealth reduces savings (hY, < l), thereby worsening the trade balance. The M M schedule, therefore, is negatively sloped: a fall in w" to encourage exportables production is needed when M rises if a trade balance deficit is to be averted. To the right of M M there is a payments deficit and M is falling, while to the left the payments balance registers a surplus and M is rising. The LL schedule is positively sloped as a larger money stock drives up P, and stimulates employment in the high-wage nontradables sector, thus drawing labor out of the export sector and bidding up w". Above LL, open unemployment exists and w' is falling. Points below the LL schedule are irrelevant since w' is upwardly flexible; once the economy arrives at a point on LL south of B, the adjustment path moves up the LL schedule until the long-run equilibrium is reached. The adjustment process is thus globally stable. Following the imposition of a tighter quota, the long-run value of the export sector wage is obtained by the requirement that the labor market clear when saving is zero. Since wx and L" move in the same direction in the long run., the qualitative outcome can be determined simply by setting s = 0 in (21). This yields dL"/dI", dw"ldI" < 0 if and only if

0

Fig. 6.2 Import compression in the nontradables sector

M

475

Mexicotchapter 6

Given that s is usually small, the prospects for reducing underemployment do not improve much in the long run. As with the short-run outcome, real income will be lower and underemployment greater in the new steady state unless the degree of substitutability on the supply side far exceeds that on the demand side. The cumulative impact upon the payments balance turns on whether nominal income rises after wx has adjusted to clear the labor market. From (10) and (13), the change in nominal income in the long run is (26)

dy = u n p e e ( i -

$1 {[(Y,

+ ~,ey)ie;tlS,- (ey/ert)p}+ td",

where J, = wX/w"< 1. Solving for the long-run changes in P, and g" and substituting these into (26) yields

as the necessary and sufficient condition for a cumulative payments surplus. The condition for a cumulative surplus is less stringent than that for a short-run surplus, reflecting the fact that, in the long run, those laid off in the high-wage manufacturing sector find employment in the export sector. The likelihood of an initial payments deficit being reversed in the long run is sensitive to the magnitude of the sectoral wage gap. Inspection of (27) suggests that, whatever may happen in the short run, if the labor market is not extremely distorted (IJJ > O S O ) , wage flexibility in the export sector eventually succeeds in generating a cumulative payments surplus. In subsequent analysis, I assume (27) is satisfied so that the nominal money stock rises across steady states. In figure 6.2, path ABC portrays the nature of the adjustment process when (25) holds and the payments balance improves in the short run. Point A is the initial equilibrium. When the import quota is reduced, the LL and M M schedules shift horizontally to the right and C becomes the new long-run equilibrium. A payments surplus emerges, but labor demand contracts in the nontradables sector and the economy immediately goes into a recession, experiencing open unemployment. Over phase AB, payments surpluses increase the money supply and the soft labor market depresses w". Higher employment in the export sector and expansion in the money supply both raise nontradables demand. Consequently, P,, continues to rise and L" begins to recover from its initial decline. By the time point B is reached, open unemployment has been eliminated but the trade balance still exhibits a surplus. As the money stock increases further, the path moves up the LL schedule. During this second phase, there is continuous full employment and P,, L", and wx all increase steadily. The payments surplus finally disappears

476

Edward F. Buffie

at point C and the economy then settles into a new steady-state equilibrium characterized by a lower nominal (and real) export sector wage and higher underemployment. It was demonstrated earlier that when import quotas are already in place and are generating sizable implicit tariffs, it is quite possible that further import compression will, paradoxically, worsen the payments balance in the short run. Path DEFC corresponds to this case. On impact, the MM schedule shifts downward and LL shifts horizontally to the right. The trade balance deteriorates at the same time as employment and real output contract, and the economy heads off on a southwesterly path with both wx and M declining. The payments deficits persist until the recession has extracted a sufficiently large decrease in the real export sector wage and the economy finds itself at point E on the M M schedule. Thereafter, further decreases in wx give rise to payments surpluses and the remainder of the path is similar to that of ABC. 6.3.1

Import Rationing in the Export Sector

In the preceding section I assumed that import controls were imposed only on firms operating in the protected manufacturing sector; export sector firms continued to have access to imported inputs at world market prices. In fact, it is improbable that exportables production will not be affected to some extent by import restrictions. When import compression assumes the massive proportions it did in Mexico in 1982-83, some export sector firms will inevitably be subject to controls. Furthermore, the government may not be able to completely segment the market, in which case higher prices for intermediates used by firms in the nontradables sector will spill over and raise input prices for firms engaged in export production. To see how this affects the macroeconomic repercussions of import restrictions, reverse the treatment of firms in the export and nontradables sectors: let import restrictions fall entirely upon export sector firms and allow nontradables sector firms unrestricted access to imports at world market prices. In terms of the model developed in section 6.1, set g" = e and treat I" and g" as endogenous variables and f x as an exogenous variable. Clearly, import compression is now an unmitigated disaster. Under the usual, weak assumption that intermediate inputs and labor are gross complements, employment in the export sector contracts. The reduction in export sector value-added lowers the demand for nontradables, depressing P, and causing employment in the nontradables sector to also contract. To make matters still worse, the payments deficit widens. (Nominal income falls, including a decrease in savings.) The new long-run equilibrium in figure 6.3 (0)thus lies southwest of the initial equilibrium (A), and after the import quota is tightened, both wx and M start decreasing. If the export wage does not exhibit substantial downward flexibility, wx and M overshoot

477

MexicoJChapter 6

L

0

M

Fig. 6.3 Import compression in the export sector

their new, lower steady-state values and the adjustment process entails a protracted recession with open unemployment prevailing over the entire ABC stretch of the transition path. In practice, of course, import restrictions are seldom confined to firms operating in one sector or the other. The main point of this analysis is that import compression is more likely to be contractionary and less likely to improve the payments balance the more import controls impinge upon export production. Clearly, one could argue on the basis of these results that the large trade surpluses Mexico achieved in 1983 and 1984 reflected primarily the impact of extremely contractionary monetary and fiscal policies and that the surpluses would have been larger had import controls been loosened instead of tightened. While a careful empirical study would be needed to evaluate this argument, at the very least the recent Mexican experience strongly suggests that point C in figure 6.2 is way south but not far to the east of point A; import compression purchases a small cumulative payments surplus at the very dear price of much greater underemployment and much lower real wages. 6.3 Raising Internal Energy Prices In recent years, the internal price of petroleum has been increased substantially in Mexico. The usual justification for this policy is that Mexican users should pay a price closer to (ideally, equal to) the world market price, which presumably reflects the opportunity cost of domestic oil consumption. By now it should be clear that this line of reasoning overlooks the possibility that higher energy prices will lessen overall efficiency by increasing the extent of underemployment. To avoid excessive taxonomy, in

478

Edward F. Buffie

this and the following two sections the analysis is limited to the case where the real wage is rigid in the manufacturing sector (a = 1). The model employed earlier is easily adapted to analyze the impact of higher domestic oil prices. Set g" = g" = g, where g is the common internal price of energy, and interpret I' in equation (7) as domestic consumption of oil in sector i. (I' is endogenously determined.) Also, let @ denote the fixed level of oil production and redefine R to be the gross output function R(Px, P,, I" I", L", L"). With these changes, (10) becomes

+

(10')

Y = R(P,, P,, I"

+ I", L", L") + e(Q' - I"

-

I").

Equation (10') defines national income as output of final goods and services plus oil exports. Differentiating with respect to g yields

The last term measures the direct efficiency gain from lowering the domestic subsidy. Due to this direct efficiency gain, the case for raising internal energy prices is inherently more favorable than that for imposing import controls (where the corresponding term involves an efficiency loss). The overall outcome, however, is highly uncertain as the direct efficiency gain has to be weighted against potential losses from decreased employment. Since labor and oil are gross complements, L" always declines. In the nontradables sector, employment rises or falls depending on whether

where T = e/g - 1 > 0. The right-side term picks up the impact on nontradables sector employment of the change in export sector value-added measured at world market prices [ e ( p Q' - Z")]. On the left side, the presence of T in the term multiplying a;; reflects the favorable impact on nontradables demand of the rise in income produced by the direct efficiency gain associated with lower usage of I". Given that the labor market cannot be reformed, it is unwise to completely eliminate the domestic subsidy. Trivially, for T = 0, an increase in the domestic price of oil has the same effect as tightening import quotas (when t = 0 and g" = g" = g). Export sector employment and value-added decline. Nontradables sector employment and value-added also decline unless, as usual, the degree of substitutability is much greater on the supply side than on the demand side. With the term on the right side in (29) negative, u ; , > (E c,p)/yx is a necessary condition for L" to increase.

+

+

479

MexicolChapter 6

While second-best considerations will normally call for a domestic price below the world market price, it should be stressed that there are distinct limits to the magnitude of the subsidy that can be justified. Note that when the domestic price is less than half the world price (7 > l), export sector value-added always increases and E

+ c,s

< U+f2YX

is a suficient condition for L" to increase. Thus, there appears to be a presumption that the very large subsidies of the Lopez Portillo years were excessive and that some increase in internal prices was warranted after 1982. Several remarks are in order about the proper interpretation of these results. First, the analysis applies only to energy consumption by firms; there is no justification for keeping the price of oil to consumers below the world price. Second, the size of the optimal, second-best energy subsidy might be considerably larger when the repercussions on capital accumulation are taken into account. If other policies are not manipulated to maintain the profitability of investment, higher internal energy prices will trigger capital decumulation (on the usual assumption that capital and oil are gross complements) and impose welfare losses in the future by worsening the extent of underemployment. (Even ignoring the adverse interaction with the labor market distortion, capital decumulation will give rise to welfare losses if the private discount rate exceeds the social discount rate.) Third, since the optimal policy entails balancing the gain from less underemployment against the direct efficiency loss created by driving a wedge between domestic and world prices, thc optimal subsidy will obviously be larger once distributional effects are incorporated into the welfare calculations. Indeed, given the current levels of underemployment and real wages in Mexico, distributional considerations may carry as much or more weight than concerns about efficiency. 6.4

Devaluation

Import controls, as we have seen, have little to recommend them. In exchange for suffering permanent decreases in real output and real wages and a permanent increase in underemployment, the economy obtains nothing more than a problematic impact on the payments balance-there is no guarantee that the payments balance will improve, or if it does, that the cumulative payments surplus will be sizable. (Of course, if it were known that a payments deficit would result, the appropriate policy would be to relax import quotas.) Needless to say, this is not exactly a bargain, and it is worthwhile to consider alternative policies in the hope of finding a better deal. In the remainder of this section, I analyze the merits of the IMF and World Bank's favorite remedy, devaluation of the currency. Section 6.5 is an investigation of the case for export subsidies.

Edward F. Buffie

480

Viewed over the long run, devaluation compares favorably with import rationing. With a rigid real wage in the nontradables sector, the real equilibrium is clearly not affected by equiproportionate increases in e, P,, wx, and M. Hence devaluation is neutral in the long run. This is shown in figure 6.4. The post-devaluation, long-run equilibrium lies at point C on the ray running from the origin through the initial equilibrium A. In the long run, all real variables are unchanged and the cumulative payments surplus as a fraction of the money supply equals the percentage devaluation. Unfortunately, though devaluation ultimately succeeds in generating a cumulative payments surplus without disturbing the real equilibrium of the economy, the adjustment process is painful and involves traversing a path in which real output is continuously below its previous level. In the short run, a nominal devaluation generates a real devaluation (i.e., elP, rises) by reducing the level of real money balances. The real devaluation, in turn, lowers L" both by raising the product wage ( p , < Gn < 2)and by increasing the real price of imported inputs to the nontradables ~ e c t o rMore . ~ formally, set e = gx = g" and treat I' as endogenous in the model of section 6.1. Grinding through the usual manipulations then yields6 ink?= - sal/Al < 0 ,

(30) where

A, = {E

+ C,S + O;bl + alOf[cx+ C,S + c,pJI(l - s)]}

>0

and p = -dLx/dL" s 1 is the fraction of workers laid off in the nontradables sector who gain employment in the export sector. Whether or not contraction in the nontradables sector leads to open unemployment depends in large measure on the flexibility of export sector

Fig. 6.4 The impact of a devaluation when export sector technology is relatively flexible

481

MexicolChapter 6

technology. The rise in P, increases labor demand in the export sector at a given nominal wage. If technology is relatively flexible, job creation in the export sector exceeds layoffs in the nontradables sector and, though real output declines (by the amount [w" - wX]dLn),open unemployment is averted. The flexibility condition that export sector technology must satisfy is (31)

ufK(i -

of) + u+Ief> o w , L w ( i

- of)^,,

where A, is evaluated at p = 1 . When (31) holds, the economy follows a path such as ABC in figure 6.4. Devaluation produces a strong increase in export sector labor demand that bids up wx by AB in the short run. The balance of payments shifts to a surplus, and as the money stock rises over time, P , and L" increase. The tightening of the labor market pushs wx up further as the economy moves toward the long-run equilibrium C along the LL schedule. Open unemployment never emerges, but the worsening in underemployment on the transition path results in real output being everywhere below its predevaluation level. Figure 6.5 applies when export sector technology is relatively inflexible and fails to satisfy (31). The LL schedule is steeper in slope than the ray OF, reflecting the fact that the export sector is incapable of absorbing all of the labor released by the nontradables sector without a decrease in the nominal wage. There is an initial phase AB of open unemployment and falling nominal (and real) wages in the export sector; over the medium run, devaluation works principally by reducing import demand and has effects similar to import controls. Once point B is reached, full employment prevails and on the remainder of the path w x rises steadily, eventually surpassing its previous steady-state level.

~

0

M

Fig. 6.5 The impact of a devaluation when export sector technology is relatively inflexible

482

Edward F. Buffie

6.5

Export Subsidies

The basic problem with using quotas or devaluation to remedy current account deficits is that the import volume contracts. Under weak conditions, the curtailment in the import volume reduces labor demand in the high-wage manufacturing sector, thereby exacerbating underemployment. Conversely, if underemployment is to be lessened, import flows must be increased, not reduced. This makes it natural to consider export subsidies as an alternative policy that might reconcile the potentially conflicting objectives of expanding import flows and improving the trade balance. When analyzing the repercussions of an export subsidy, some assumption must be made about how the subsidy is financed. Consider first the rather optimistic case in which the subsidy is financed by an increase in lump-sum taxes. Letting v denote the ad valorem subsidy, the basic model (without import restrictions) changes only in equations ( 1 ) and ( P,

(1') ( 10")

=

e(1

+ v)

Y = R(P,, P,, L", Lx) - evX,

where X = Q" - D", the export volume. Upon the introduction of a small export subsidy'

(32)

&IFx = u ~ A ~ ' [ EyX(a$/

+ ~/0;)] 3 0.

where a4

= {a;;[0p

+

(1 - 0;)0;]

+ a,n,efe;}/Op.

The impact on underemployment is uncertain, reflecting the conflicting effects on manufacturing sector labor demand of a higher product wage and a lower real price of imported intermediates (elP, 1).Unlike with a devaluation, L, may increase. Furthermore, when labor demand contracts, a comparison of (30) and (32) shows that the decline in L, is smaller under an export subsidy if and only if

(33)

E

+ ens > a$/?, .

This is precisely the condition for a tighter quota on intermediates utilized by the nontradables sector to lower L, (when the implicit tariff is initially zero). The condition reappears here because an export subsidy differs from a devaluation only in that it lowers instead of increases the real price of imports. A duality proposition thus links the employment effects of quotas, devaluation, and export subsidies: if a more restrictive quota on imported inputs used by the nontradables sector exacerbates underemployment, an export subsidy has a less adverse (and possibly favorable) impact on underemployment than a devaluation.'

483

MexicoIChapter6

Export subsidies also compare favorably with devaluation in being less likely to create open unemployment. This is not only because manufacturing sector employment is likely to decline less. Rather, in addition, an export subsidy provides greater stimulus to labor demand in the export sector by lowering the real price of imported intermediates, generating a favorable cross-price effect that is absent under a devaluation. The counterpart to (31) is ~ f K ( 1-

0f) + 0$,0f > @ u ~ [ E- yX(u$[ + dO;)]L"/LXA,,

which is less stringent than (31) even when (33) holds as an equality so that an export subsidy and a devaluation contract manufacturing sector employment to the same extent." The long-run outcome may be found by setting s = 0 in (32). This yields the conclusion that L" rises whenever (34)

e

'

U

a

X *

Again, this is the condition for a tighter quota on imports used by nontradables sector firms to lower L". Thus, in the long run, an export subsidy has qualitatively the same effect on underemployment as import liberalization in the nontradables sector (when s,t = 0 initially). Figure 6.6 extends the comparison of export promotion and devaluation to cover the entire adjustment process. The initial equilibrium is A, and ABC is the transition path on the assumption that export sector technology is flexible enough (i.e., [31] is satisfied) to prevent the emergence of open unemployment. Since devaluation is neutral in the long run, the new steady state lies on the ray OF. The M'W and L"L" schedules define the transition path following the introduction of an export subsidy. Extremely weak conditions suffice to

F

0

Fig. 6.6 Export promotion compared with devaluation

M

484

Edward F. Buffie

guarantee an improvement in the payments balance. It can be shown, however, that the cumulative payments surplus is always smaller than under a comparable devaluation. If (34) holds, underemployment diminishes and the real export sector wage increases unambiguously in the long run (kX> px > pn).The new long-run equilibrium, therefore, lies to the northwest of C . At all points on the transition path, the real wage in the export sector and real income are higher than under a devaluation. The cumulative payments surplus is smaller, but this merely reflects the fact that the price level rises less. The real money supply is higher at E than at C (and the initial steady state A ) .I ' It is more difficult to compare export subsidies and devaluation when (33) holds but not (34). Figure 6.7 depicts the time paths of real income Y' in this case. In the short run, a devaluation results in greater underemployment and lower real income than an export subsidy. At some point, however, this ranking is reversed. Over the long run, real income declines under an export subsidy but eventually returns to its previous steady-state level under a devaluation. I end on two cautionary notes. First, these results demonstrate only that under certain conditions an infinitesimally small export subsidy compares favorably with a devaluation. As such, they show that a plausible case can be made for the introduction of some export subsidy; they do not, however, indicate how large a subsidy can be justified (i.e., the size of the optimal export subsidy). Second, although export subsidies may expand the tax base by increasing real income, they normally will not be self-financing. Active measures to increase tax revenues or cut expenditures will be needed in order to avoid a

Y'

I 0

I

to

t

Fig. 6.7 The short- and long-run effects upon real income of an export subsidy and a devaluation

485

MexicoKhapter 6

deterioration in the fiscal deficit and a possible worsening in the payments balance. Suppose, for example, that the government expenditures are fixed in real terms and that revenues derive from a proportional income or value-added tax z.In this case, if the fiscal budget FB is initially in balance'*

(35)

.. = P,X - (1 - z) -

Y-ldFBIP,

Y

pne" Y e;(l -

+),

where q = L"/k,. The second term reflects the revenue gain from the reduction in underemployment. For believable values, this gain is very small compared to the direct worsening of the deficit caused by the introduction of the subsidy. For instance, if z = 0.20 and the share of nonoil exports in GNP is 0.05, then the direct revenue loss equals 0.04. Turning to the indirect gain, it can be checked that, extreme cases aside, the employment elasticity q will not exceed 0.20.13With this large value for q, 02 = = 0.50, and PnQ"/Y = 0.60, the indirect revenue gain is still only 0.006, leaving a large residual deficit.

+

6.6 Concluding Observations Since the end of 1981, employment growth has virtually ceased despite reductions in real wages in excess of 30 percent. In this chapter I have put forward the thesis that conventional microtheoretic factors can account for much of the extreme worsening in underemployment seen in the first five years of adjustment to the debt crisis. Tight import quotas (until July 1985), a heavily depreciated currency, and huge increases in internal energy prices have greatly reduced usage of intermediate inputs complementary to domestic labor. In short, as regards job creation, relative prices were wrong. More moderate increases in internal energy prices (for industrial users) and greater use of export subsidies instead of real devaluation to attain external balance would have allowed adjustment to proceed with at least a lesser increase in underemployment. Unfortunately, each of these policies would have conflicted with the task of reducing the fiscal deficit, a task already made very difficult by increased debt service and sharply declining terms of trade.

Appendix

To derive the expressions for L" and I" given in (13) and (14), first substitute for Q" in (5) from (6): (Al)

L" =

c;K"/c:.

Now differentiate (Al) and note that C,CIC,Cj = ujj,where u, is the Allen partial elasticity of substitution between factors i and j . This gives

486

Edward F. Buffie

Substituting for G" from (3) and utilizing the adding-up restrictions, -uuB, = uLK8, u,e, and -uKKeK= a,,€), + u,e,, (A2) becomes

+

(A3)

i"

- 0;)

= -ay,[uL;((l

+ ufIO,"]p,+ [u&(1 - 0;) + oj$~;]i-" + e;(u,; - u;,)g" .

In the case where the production function is separable between primary factors and imported inputs, , u =, u = uvI and (A3) simplifies to ('44)

jn = [ ~ : ~ ( i-

e;) + ~ ; , e ; ] ( i ~-

aynFj~).

From equations (2) and (3):

Substituting this expression into (A4) gives equation (13) in the text jn = u,Pn -

('45)

adg" ,

where al

= [i

u2

= e;[u:K(i

e;)][~:~(i - e;) + ~;~e;]/e; - e;) + ~;,e,"]/e; .

- ay,(i -

The expression for I" stated in (14) is obtained by the same procedure.

7

Fiscal and Monetary Policy, Financial Intermediation, Inflation, and Growth

In previous chapters I have often emphasized the self-reinforcing and stagflationary nature of the various macroeconomic mechanisms linking large fiscal deficits, high inflation, financial disintermediation, and slow growth. At present, Mexico, like so many other Latin American countries, seems to be trapped in a self-perpetuating spiral of this sort: high inflation provokes a flight of funds from the banking system; the low level of financial intermediation curtails the supply of bank loans for productive investment

486

Edward F. Buffie

Substituting for G" from (3) and utilizing the adding-up restrictions, -uuB, = uLK8, u,e, and -uKKeK= a,,€), + u,e,, (A2) becomes

+

(A3)

i"

- 0;)

= -ay,[uL;((l

+ ufIO,"]p,+ [u&(1 - 0;) + oj$~;]i-" + e;(u,; - u;,)g" .

In the case where the production function is separable between primary factors and imported inputs, , u =, u = uvI and (A3) simplifies to ('44)

jn = [ ~ : ~ ( i-

e;) + ~ ; , e ; ] ( i ~-

aynFj~).

From equations (2) and (3):

Substituting this expression into (A4) gives equation (13) in the text jn = u,Pn -

('45)

adg" ,

where al

= [i

u2

= e;[u:K(i

e;)][~:~(i - e;) + ~;~e;]/e; - e;) + ~;,e,"]/e; .

- ay,(i -

The expression for I" stated in (14) is obtained by the same procedure.

7

Fiscal and Monetary Policy, Financial Intermediation, Inflation, and Growth

In previous chapters I have often emphasized the self-reinforcing and stagflationary nature of the various macroeconomic mechanisms linking large fiscal deficits, high inflation, financial disintermediation, and slow growth. At present, Mexico, like so many other Latin American countries, seems to be trapped in a self-perpetuating spiral of this sort: high inflation provokes a flight of funds from the banking system; the low level of financial intermediation curtails the supply of bank loans for productive investment

487

MexicoIChapter I

and exacerbates short-run inflationary pressures by reducing the demand for the monetary base; as low investment rates translate into slower growth in productive capacity, a new wave of financial intermediation takes place, tax revenue declines, the fiscal deficit increases, and the cycle begins anew but at a higher rate of inflation than before. The Mexican government has attempted to break out of this destabilizing spiral by cutting fiscal spending, by setting high marginal reserve ratios (nearly 100 percent) to increase the demand for the monetary base, and by financing a greater proportion of the fiscal deficit through bond sales. This approach has clearly not worked well, and a large number of unresolved policy issues remain. Is it better to reduce the deficit by raising taxes or cutting expenditures? Will higher reserve ratios and regulations requiring banks to purchase more government bonds reduce the inflation rate and lower bond rates or, perversely, will they lead to greater financial disintermediation, capital decumulation, and higher inflation and larger fiscal deficits as the tax base shrinks? Similarly, will reduced monetization of the deficit lower the inflation rate, as claimed by simple monetarist models; or, instead, will the inflation rate increase either because interest payments on the government debt increase sharply or because higher bond rates strongly crowd out private investment, lowering future output and tax revenues? The aim of this chapter is to gain a deeper understanding of these policy issues by developing a model that captures various key elements making up the fiscal deficit-financial intermediation-growth-inflation nexus. In section 7.1 I develop the basic model, while in sections 7.2-7.7 I analyze the consequences of different policy packages designed to service the external debt. The results are used in the final section to evaluate the stabilization program followed by the De La Madrid administration.

7.1 The Model The standard analysis of inflation and fiscal deficits follows the early work of Cagan (1956) and Mundell (1965) and is based on a minimal model in which real output is exogenous and high-powered money is the sole financial asset in the economy. This framework might be adequate for analysis of hyperinflation, but it is too simple to yield any insight about the difficulties facing the Mexican economy, an economy that is stumbling deeper into stagflation but has not yet reached the dysfunctional point of hyperinflation. In what follows, a fairly rich model is developed that allows for inside and outside money, interest-bearing government debt, and endogenous capital accumulation. Due to the detailed specification of the financial sector, the dynamics for wealth accumulation are analytically intractable. The analysis, therefore, will be confined to a comparison of steady-state outcomes. This is an acknowledged shortcoming. Comparative steady-state analysis, however,

488

Edward F. Buffie

is one important element in judging the adequacy of different policy packages in that it defines the long-run economic tradeoffs that will ultimately have to be confronted. Such comparisons enable one to judge whether a stabilization program is inherently sound or instead a "quick fix" that will eventually create more severe macroeconomic problems. The central thesis of this chapter is that many of the recent Mexican stabilization measures have been of the quick-fix variety. The model is laid out in stages, beginning with the goods and labor markets.

7.1.1 Aggregate Supply and the Labor Market The economy is small and completely open, producing one importable manufactured good and two agricultural export goods. There are no trade taxes and the world market price of each good is fixed at unity. Consequently, all domestic prices are set by the exchange rate and all relative prices also equal unity (choose any good as the numeraire). Capital and labor are used in production of the manufactured good, and land and labor in production of the agricultural goods. Technology exhibits constant returns to scale in each sector, and the aggregate supplies of land and labor are perfectly inelastic. Land is intersectorally mobile. As in the model of the preceding chapter, the labor market is dualistic. A rigid real wage, wm, prevails in the manufacturing sector while a lower, market-clearing real wage, wx, is paid in the two agricultural sectors. Besides the private manufacturing sector, the public sector pays a relatively high wage. Parastatal firms employ N g workers at a fixed real wage, w g > wx. Both N g and w g are treated as government policy variables. The public sector capital stock is fixed, so parastatal output, Q g , varies only with N g : (1)

Q g = F(Lg), F'

> 0.

This specification for the real side of the economy implies that all real factor prices are constant as long as relative goods prices are constant.' Since real factor prices are fixed, employment in the high-wage manufacturing sector, N", varies proportionately with the capital stock, K :

(2)

dN"

=

edK,

where 4 = N"/K, the (constant) labor-capital ratio.' Using standard methods from duality theory, it is easy to verify that the value of private sector income, Y p , and total wage income, Y", may be represented by functions of the form

(3) (4)

Yp = Yp(K, N g , wg)

489

Mexico/Chapter 7

where

Y:,

= rK

y$ = y y y p -

3 -

+ (wm -

wx)e,

Y;

=

(Wm - w x ) e ,

= wg - wx,

yy = Ng ,

and rK is the competitive capital rental. When the capital stock rises, output increases directly by the amount r, and indirectly by ( w m - wx) as labor is reallocated from the low-wage agricultural sectors to the high-wage manufacturing sector. Similarly, an increase in public sector employment raises private income by the extent of the sectoral wage gap, w g - wx. The main reason for postulating this particular productive structure is that it permits a simple and clear delineation of a key general equilibrium relation, namely, the interdependence of capital accumulation and financial intermediation. An increase in the capital stock raises total wage income by expanding employment in the high-wage manufacturing sector. In general equilibrium, this triggers a mutually reinforcing, virtuous cycle of capital accumulation and financial deepening: as workers increase their holdings of demand deposits, banks extend more loans, which leads to further capital accumulation, further growth in wage income and bank lending, etc. The same feedback mechanism operates and broadly similar qualitative results emerge in any model in which capital accumulation increases labor income (for example, in models with integrated labor markets where a rise in K bids up the market-clearing wage).

e

7.1.2 The Banking System The government owns and regulates all banks. Bank interest rates are administratively determined and are set below their market-clearing levels. From the funds supplied by the public at the regulated deposit rate, banks either purchase government bonds or make loans to private sector firms. As the real loan rate is too low to clear the market, loan demand is r a t i ~ n e d . ~ Furthermore, in keeping with regulations that require banks to apportion a sizable fraction of their investible funds to the purchase of various government financial assets, the division of the bank portfolio between capital and bonds is treated as exogenous. Banks face a reserve requirement, k, and must channel the fraction kb of their deposits, D , into bond purchases. Captive bank bond demand, Bb, is thus

(5)

Bb = kbD

and (assuming excess reserves are zero) the quantity of loans, L , is determined residually as

(6)

L

=

xD.

490

Edward F. Buffie

1 - k - kb. wherex Though the “commercial” banks in Mexico are managed by and classified as part of the private sector, clearly their losses (or, improbably, their profits) must be absorbed by the public sector budget. It is useful at this juncture, therefore, to note from (5) and (6) that real bank profits, Z, are given by

z = (TLX

(7)

-k rbkb -

rd

- mk]D,

where m represents the inflation rate and r,, rb, and r, denote, respectively, the real loan rate, the real bond rate, and the real deposit rate. 7.1.3 Capital Accumulation and Private Sector Asset Demands

Any model of the financial sector in Mexico should faithfully reflect certain important stylized facts concerning asymmetries in the pattern of asset holdings across agents. First, whereas cash (or currency) accounts for a negligible fraction of total wealth of high-income groups, the poor hold virtually all of their wealth in the form of money broadly defined (currency demand deposit^).^ Second, even among those who are relatively well off, few have the opportunity to invest in capital assets due to the absence of an extensive and well-functioning equities market. With these stylized facts in mind, I assume the specialized pattern of asset holdings described in table 7.1 .5 This particular configuration of asset holdings can be plausibly explained by a mixture of institutional constraints and different patterns of asset dominance. Capital dominates bonds, but neither landowners nor workers can acquire capital as they are not directly involved in organization of the productive process in the manufacturing sector. In addition, workers do not hold bonds because either their time preference rate is too high or large fixed transactions costs render small bond purchases unprofitable. Bonds are an important saving instrument only for the landowners, who may be thought of in the current model as proxying for the upper middle class. And finally, while both currency and demand deposits serve as mediums of exchange and hence enter agents’ utility functions, the transactions needs of family firms and landowners are such that, for these two agents, demand deposits dominate currency. The main consequence of this latter assumption is that when the government pegs the

+

Table 7.1

The Pattern of Asset Holdings Agent Family firms Landownen

Workers

Assets

Liabilities

capital, demand deposits bonds, demand deposits demand deposits, currency

bank loans

none none

491

MexicoKhapter 7

real deposit and real loan rates, workers alone shoulder the burden of the inflation tax. Private sector asset demands are obtained by solving a set of explicitly specified intertemporal optimization problems. Each agent is infinitely lived and has a recursive utility functional of the type formulated by Uzawa (1968). This class of utility functionals has the desirable property of allowing local time preference to vary continuously as a function of current utility (Epstein and Hynes 1983; Epstein 1987). In steady-state equilibrium, utility remains at the level u* and the time preference rate becomes the constant p(u*). Following the usual practice in the literature, I assume (8)

(i) p' > 0 and (ii) p - p'u > 0.

The first condition reflects the notion of increasing marginal impatience, to use the terminology of Lucas and Stokey (1984). The second implies that in a comparison of two stationary consumption streams, the stream having higher instantaneous utility confers higher total utility and will be preferred. Succeeding sections are devoted to characterizing the nature of each agent's asset demands. The discussion in the text is largely informal. Derivations of the asset demands may be found in the appendix to this chapter. Notational conventions are as follows: all variables are expressed in real terms, and c, v, and w superscripts refer, respectively, to capitalist family firms, landowners, and workers; common, unsuperscripted symbols are used for the utility function and the time preference rate, but it is understood that these differ across agents. Family Firms The family firm holds its (gross) wealth in the form of two assets, physical capital and demand deposits. Deposits, unlike bonds, are not dominated by capital because they yield nonpecuniary services in facilitating transactions. These nonpecuniary services are indirectly accounted for by incorporating real deposits into the utility function. Current utility, u, is represented by the indirect utility function u = V(E', Dc),

where VE, VD > 0, VDE = VED> 0, VEE,VDD

0

and E' is real consumption expenditure and D ' is real demand deposits. Increases in E' and D ' raise utility but at a diminishing rate.6 V has positive cross partial derivatives, indicating that a higher level of real consumption raises the marginal utility of deposits.

492

Edward F. Buffie

Investment in physical capital and accumulation of real deposits are financed out of retained profits and bank loans, L. Since the firm is rationed in in its access to bank loans, retained profits are the marginal source of investment funds and time preference affects steady-state asset demands. In a stationary equilibrium where K , E", Dc, and L are all constant,

(9)

+ ~ ( 1-

7)

= p{V[+(K)(l -

7)

f rdDc

-

rLL,

o']}

+ ( K ) is the restricted profit function with the constant w m suppressed. the return on capital, equals the competitive capital rental, r,. T is a flat ad valorem tax that applies to all noninterest income (gross profits, wages, and land rents). In the present model, 7 is equivalent to a valueadded tax. Equations (9) and (10) state that capital will be accumulated until its return equals the time preference rate and that the marginal rate of substitution between consumption and dcposits will be equated to the opportunity cost of deposits, +,(1 - 7) - r,. Solving these two equations for K and Dc gives

+,

(12)

Dc = hC(7),

where

Under constant returns to scale, the return on capital is constant for a given value of w m .As the return on capital does not vary across steady states, the optimal long-run response to increases in L , rd, or r, is to keep real deposit holdings unchanged and to adjust K until real income (inclusive of net interest payments) returns to its original level, at which p = + K ( l - 7). The expressions for ,fi and f$are thus quite simple:

ft

fi

=

fi

= L/+,(l

-D/+,(l

-

7)

- 7).

Provided that r, is positive and L exceeds D', we have, as noted above in (9) and (lo), f t > 0 and (f: f;) > 0.

+

493

Mexico/Chapter 7

Landowners The nonfirm private sector does not receive loans and is restricted in its asset choices to bonds and money. For landowners, optimizing behavior requires that asset demands satisfy (13)

p{v[V(l -

7 )-t

rdD"

f

r$", D"]} = rb

where v stands for income from land rents. Equations (13) and (14) yield (15)

B" = f 2 [ r b , rd, ~ ( -l

T)]

Bonds play the same role in the landowner's portfolio that capital does in the family firm's portfolio. The real bond rate fixes the time preference rate so that, in the long run, increases in after-tax rental income induce bond decumulation but do not alter real deposit holdings. Note also that there is a positive relation between savings and the real bond rate, as reflected in the fact that steady-state wealth incrcascs with rb (f: h'; > 0). The impact of variations in real returns is largely as expected. An increase in the real deposit rate induces substitution away from bonds toward deposits. A higher real bond rate, however, may increase the demand for both assets. Real deposit holdings rise or fall with rb depending on whether

+

(17)

(a + y ) ( l - rd/rb) 3 q >

where u = V,,E"/V, > 0 , the elasticity of the marginal utility of deposits with respect to real consumption expenditure; y = - VEEE"/VE,the "partial" Arrow-Pratt measure of relative risk aversion (i.e., y is defined for a given value of D"); and q = (dp/dE)E/p is the elasticity of time preference with respect to real consumption expenditure.' In what follows, I assume that p is strictly concave in C so that q < 1. A rise in rb makes it more costly to hold money in financial terms but also leads to a higher level of steady-state consumption, which increases the marginal utility of the nonpecuniary benefits of money relative to the marginal utility of goods (VDE> 0, V,, < 0). If q is relatively small, a large increase in consumption is required to bring p up to the higher value of rb and the latter, positive effect is likely to dominate the former, negative effect. On the other

494

Edward F. Buffie

hand, when q is relatively large, steady-state consumption rises little and the larger spread in financial returns causes a reduction in deposit holdings. In general, there is no clear presumption as to the sign of hi'. Although q must be less than unity, relatively little is known about the magnitudes of u and y. Even if existing estimates of relative risk aversion are taken as evidence that (a y) will range between 1 and 3, the left side of (17) may be below unity when the initial interest rate spread is not too large. A feature of the asset demands that will figure importantly in future results is the relative magnitudes of the own- and cross-price effects. The own-price effect is larger than the cross-price effect for deposits, but the reverse is true for bonds. Overall, however, own-price effects dominate cross-price effects. It is readily demonstrated (see the app.) that

+

Workers Due to a high rate of time preference and/or fixed transactions costs that absorb most of the return on small bond purchases, workers allocate all of their wealth to money. Unlike capitalists and landowners, their transactions needs are such that they hold significant quantities of both deposits and currency, C. The representative worker's indirect utility function is of the form V ( E " , D", C) and has the properties: VE, VD,

v,,

-

V c , V E D , VEc > 0; V c D ,

v,, < 0; v,,

-

VEE,

v,, < 0 .

Vcc, V D D

< 0;

The negative sign for V,, signifies that the two monies are substitutes as mediums of exchange. We assume as well that increases in E do not alter the nonpecuniary benefits of deposits relative to those of currency: (18)

VC - V D

c),

= [(O?

50

> O,

[C

< O.

Workers' holdings of deposits and currency are determined by the conditions that

(19)

VDIYw(l v,r[Yw(l =

T)

7)

p{v[Y"(1

+ r,D"

-

7rC, D", C ]

+ rdDw - d,D", c] - 7)

+ rdDw - TC,D", c]}- rd

The marginal nonpecuniary benefits of deposits are equated to the net price of deposits, p - rd. Currency is accumulated up to the point at which its marginal nonpecuniary benefits exceed those of deposits by an amount equaling the nominal deposit rate.

495

MexicoIChapter 7

In the general solution for the asset demands, the signs of both cross-price terms (aclar, and aDwlan) are uncertain owing to conflicting income and substitution effects. Maintaining r, as the inflation rate rises, for example, may not increase deposit holdings; for while an increase in the nominal deposit rate induces substitution toward deposits, the higher inflation tax lowers real steady-state consumption, which exerts a countervailing contractionary effect. To abstract from the ambiguous cross-price effects, I work with special versions of the general solution in which the demand for each asset depends upon total after-tax wage income and its own return: (21)

Dw

(22)

c

=

hw[YW(l-

= g[Y"(l -

T ) , id]

71, T]

where hi", hY, gi > 0,

g2

< 0.

The critical implication of neglecting cross-price effects is that to promote financial intermediation, the government must offer the public a higher real deposit rate. Merely pegging the real deposit rate in the face of higher inflation will not suffice. 7.1.4 Aggregate Asset Demands and Asset-Market Equilibrium

The real bond rate adjusts to ensure that the stock of outstanding bonds supplied by the government is willingly held. Combining the demand of banks and landowners, we have the market-clearing condition

B = kJl

(23)

+ f 2 [ r b ,rd. v(1 - 711.

The real bond rate is the only market-determined interest rate in the economy. All other asset stocks reach their long-term equilibrium values through quantity variation; investment by family firms and financial savings of the public bring K , D, and C to their desired levels. Aggregating D", D" and D' and using (4), the general equilibrium solutions for D and C may be expressed as (24) (25)

D

=

D[Yw(K,Ng, wg)(l -

C

=

T),

C [ Y w ( K N, g , wg)(l -

rd, rb, T]

T),

TI,

where D I , C1, 0

2

> 0 , C2 < 0 ,

0 3

,D , S 0 .

Since workers' holdings of deposits increase with their income, capital accumulation and financial intermediation are interdependent, mutually rein-

496

Edward F. Buffie

forcing processes. Growth of the capital stock raises total wage income and real deposits which, in turn, expands the supply of bank loans and lowers rb. The expansion in bank lending stimulates further capital accumulation, which leads to further financial deepening, and so on. To ensure that capital accumulation, deposit accumulation, and fluctuations in the real bond rate do not feed back upon one another in an unstable manner, it is assumed that

e,, = 1 - f { ( l - k ) D , Y p ( l -

(26a)

= 1 -

T)

pd(e,m/e,m)

(1 - wx/wm)rL(l- k)Dw/YW(l-

T)

>0

where Oy denotes the cost share of factorj in the manufacturing sector and P d stands for the elasticity of D" with respect to after-tax wage income. The first restriction ensures that the capital accumulation multiplier is finite when the bond market does not exist (or when open-market operations peg the real bond rate). The second is necessary and sufficient to guarantee stability of the adjustment process when rb fluctuates to clear the bond market. This restriction requires that bonds and deposits not be overly strong substitutes. 7 . 1 . 5 The Government Budget Constraint The government makes outlays for employment, consumption, internal and external debt service, and to cover the losses of the banking system. External debt service net of long-term capital inflows and concessional aid equals S in real terms and is treated as strictly exogenous.' Public sector income consists of revenue from the value-added tax and sales of government-produced goods and service^.^ Any revenue shortfall is covered by printing money M or selling bonds:

M

+b =G -

-I-

WgNg 4-

TYP(K,Ng,

s

4- rbB -

wg)

z

- F(Ng) -

TM,

where an overdot signifies a time derivative. Since M and B are constant in a steady state, satisfaction of the budget constraint requires an inflation tax of

(27) mM = G

+ wgNg + S + r$

- 2 - TYP(K,N*, wg) - F(Ng).

As the asset preferences and saving behavior of the public determine only M and either B or r,, there are a number of different ways in which budget balance can be achieved. For the most part, I will treat B and (G + S) as

497

MexicoXhapter 7

exogenous and assume that the inflation rate is manipulated along with other policies (k, kb, etc.) to satisfy (27). While equation (27) is the standard representation of the government budget constraint, in the current context it conceals rather more than it reveals. When the government owns the banking system, net revenue from the inflation tax depends in part on how bank interest rates are managed. We assume that the government pegs the real deposit and real loan rates and treat any changes in rd and rL accompanying fluctuations in the inflation rate as separate, conscious policy decisions. The crucial point to note is that under this interest rate policy, the net revenue gain from raising T is limited to just the increase in revenue from the higher inflation tax on currency holdings; the higher inflation tax on bank reserves is nullified by greater losses on bank operations." To see this, substitute for Z in (27) from (7) and decompose M into bank reserves (kD) and currency. We then have

(28)

3TC = G -k

s -k

WgNg -k r$

- rYP(K, Ng, wg)

-

- [rLX

+ rbkb - rd]D

F(Ng).

The right side of (28) shall be referred to as the adjusted fiscal deficit. A second important and obvious link connecting financial policy and the fiscal deficit arises through net losses incurred by the banking system. We assume, generously, that the deficit on financial intermediation (Z) is initially zero. This implies that initially (ILX

+ rbkb

-

rd)D

=

TkD.

Thus, an asymmetry exists: although higher inflation taxes only C, it remains true that growth in bank reserves (kD) increases the revenue yield of the inflation tax. I

'

7 . 1 . 6 Characterizing the Steady-State Equilibrium

As a first step toward characterizing the steady-state equilibrium, replace D in (23) by the solution in (24) and solve for rb: (29)

(f: -k kbD3)dr-b = dB - kbDl(1 - r)(w" - w")k'dK - kbD1(l -

7 ) (W'

(f;-k k@*)drd

- Wx)dNg- kbDI(1 - T ) N ~ ~ W '

+ [ f : V + kb(D,Y" - D q ) ] d ~ .

The supply of deposits effectively determines bank bond demand. Since higher total wage income results in greater deposit holdings, increases in K , w g , and N g all lower r,. The inverse, general equilibrium relation between K and rb is represented by the negatively sloped RR schedule in the second quadrant in figure 7 . 1 .

498

Edward F. Buffie

Fig. 7.1 The steady-state equilibrium

The degree of financial intermediation influences capital accumulation as well as the real bond rate. Substituting for L in (1 1) from (6) and (24) gives (30) e&K

=

(ffxD2

+ f i ) d r , + f i d r , + [ f ; x ( D 4- D I Y w )+ f $ ~

+ f;xD1Ng(l - T)dWg+fixD1(Wg - wX)(1- 7)dh" - fiDdk fiDdk, + f:xDgdrb .

-

Equations (29) and (30) can be solved simultaneously for K and r, as a function of the policy variables r,, r,, k,, k, w g ,Ng, and 7.This enables the general solution for K to be written as (31)

K

=

H(rd, r,,

T,

k, kb, WP, N g ) .

With the solutions for K and r, in hand, the steady-state inflation rate can be pinned down through the government budget constraint. Let X = d, the inflation tax on currency holdings. Making use of (25), one obtains (32) dX = ( 1 - kC)Cd.rr+ d l ( l -

+ d l ( 1 - T)Ngdwg+ d

T)(w"'l ( 1

-

w")k'dK

~ ) (-d wX)dNg-

dlYwdT,

where k, = -.rrC,/C, the elasticity of currency demand with respect to the inflation rate. The ZZ schedule in the fourth quadrant of figure 7.1 is based on (32) and shows how X varies with r for given values of K, T , w g , and N g . Since I am not interested in the well-known stability problems which arise when higher inflation lowers seignorage, pc is restricted to be less than unity and IZ is positively sloped throughout.

499

MexicoIChapter 7

Exactly where the economy locates along the ZZ schedule depends on the state of the fiscal budget. Recall that revenue from the currency inflation tax must suffice to cover the adjusted budget deficit as given by the right side of (28). This provides a second relation involving X:

+

+

+

(D - PkDz)drd - xDdr, rLDdk (33) dX = d(G S) - (rb - rL)Ddkb r,dB - {w, ( w m - W*

+

+

+ ~ k D i ( 1- ?)]}dK - {Yp - TICIYw + k(D1Y” - D,)]}d? (B” - PkD3)drb

(7

+ [ W g - F‘

- (Wg - W”)

+ TkD,)]dW + (1 - wkD,)Ng(l - ?)dW8 .

Capital decumulation worsens the adjusted budget deficit both by lowering tax revenues and reducing real deposit holdings (which diminishes the base of the overall inflation tax). The X X schedule in the first quadrant is thus negatively sloped. X X is defined for given values of S, the entire set of policy variables, and r,. Shifts of or movements along the RR schedule will, therefore, produce shifts in X X . The ambiguous sign of the coefficient on r, is due to the fact that while a higher bond rate worsens the fiscal deficit by raising interest payments to the nonbank public, it also expands the monetary base if bonds and deposits are complements in landowners’ portfolios (D3 > 0). As it seems clear that higher real bond rates intensify inflationary pressures in Mexico, I ignore the opposite possibility by assuming D3 < B”/d-that is, bonds and deposits are not extremely strong complements. The graphical depiction of the steady-state equilibrium is completed by using the reduced form solution for K in (31) to fix the position of the KK schedule.12 Since X = PC and the steady-state capital stock is independent of both C and P when the government pegs r, and r,, KK is horizontal. Associated with a given KK schedule are a particular ZZ schedule in the fourth quadrant and a particular equilibrium bond rate, which fixes the position of the X X schedule. The intersection point of KK and X X determines the size of the adjusted fiscal deficit and hence the required currency inflation tax. Once X is known, the steady-state inflation rate is found by dropping a vertical line to the ZZ schedule. The corresponding mathematical procedure involves using (29) and (31) to eliminate K and r, in (32) and (33). The resulting two equations can then be solved to obtain the reduced-form solutions for X and P . 7.2 Fiscal Adjustment and Debt Service Suppose the government is faced with increased debt service obligations (or a reduction in oil revenues). If real bank interest rates (r, and rL) are

500

Edward F. Buffie

maintained and none of the instruments of fiscal policy (G, w g ,Ng, and T) is altered, the entire burden of adjustment falls upon the inflation tax and we have the outcome shown in figure 7.1. The shock simply shifts the X X schedule to the right. Neither K nor rb changes, and the inflation rate rises from T~ to IT^. Apart from its ugly distributional effects-workers alone pay the costs of adjustment-the main drawback of responding so passively to the external shock is, unsurprisingly, that the inflation rate increases very strongly. From (32) and (33): (34) Under a policy of pegging real bank rates, the base for marginal increases in the inflation tax is very small and the shock will prove highly inflationary even when p., is well below unity. In Mexico in 1986, currency held by the public was just 3.1 percent of GDP (calculated using the average of the beginning and end-of-year currency holdings). With this value for UGDP and kb = 0.33, an increase in debt service equalling only 2.5 percent of GDP implies a long-run increase in the inflation rate of 120 percentage points. In the foregoing I assumed that nominal bank interest rates were actively adjusted so as to keep real interest rates fixed. In fact, since 1972, the Mexican government has frequently allowed high or rising inflation to substantially reduce real deposit and real loan rates. By contrast, during the era of Stabilizing Development when single-digit inflation prevailed, real bank interest rates were consistently positive and above U.S. rates. It is not too difficult to see why policymakers are tempted to follow a more passive interest rate policy when the government owns the banking system. Note that in pegging the real deposit and loan rates, the deterioration in the actual budget deficit caused by the external shock is magnified by an increase in the deficit on financial intermediation of Id>d.rr (see [7]). Under pressure to contain the growing fiscal deficit, the government may abandon its commitment to maintain real bank rates. Another source of temptation may be the hope that a lower real deposit rate will indirectly ease budgetary difficulties by putting downward pressure on the real bond rate and reducing interest payments on the internal debt. To ascertain whether a more passive interest rate policy will help blunt inflationary pressures, coneider the consequences of equal decreases in the real deposit and real loan rates: dr, = dr, = -dg. (A pure nominal interest policy corresponds to -dg = - d n . ) Equation (33) gives

dX

=

-[Bb

+ kD(1 - nD,/D)]dg

for the direct impact on the adjusted budget deficit. Provided the semi-interest elasticity of deposit demand (D,/D)is not too large, the direct impact upon the budget is favorable.

501

Mexico/Chapter 7

But this is by no means the end of the story. The steady-state capital stock declines:

(fi +

f;)< 0 and the own-price effects are larger than the cross-price since effects (f:D2>f2D3).I3 The parallel decrease in deposit and loan rates directly induces capital decumulation by raising real income of capitalist family firms, who are net debtors of the banking system. In addition, as the supply of deposits shrinks, bank lending is curtailed. This gives rise to further capital decumulation as reflected in the first term involving f f. The decrease in the capital stock lowers real output and total wage income, causing tax revenues and real currency holdings to decline. Furthermore, the real bond rate, instead of falling together with rd , is almost certain to rise, adding a third source of countervailing inflationary pressure. Substituting the solution for K into (29) yields (36) drbldg

=

[bDi(fbD:!

+ fi

f

fi)

+ (f,"+

bD2)(1 - fkDi)I/ei

.

The first positive term captures the reduction in bank bond demand stemming from the reduction in wage income and deposit holdings occasioned by capital decumulation. In the second term, the sign of (f:+ k,D,) determines whether, in the aggregate, bonds and deposits are complements or substitutes. A decrease in rd induces substitution toward bonds on the part of landowners in the amount f& but banks, finding themselves with fewer funds, are forced to lower their bond purchases by k&. The latter effect is the stronger one when (37)

pd

>

-C@'/Bb

,

where pd and ad are, respectively, the elasticities of total deposit demand and private sector bond demand (i.e., landowners' demand) with respect to rd.I4 The above condition is virtually certain to hold; p d , an ownprice elasticity, will normally exceed (Yd a cross-price elasticity and, in the Mexican case, BvIBb is much smaller than unity (B"lBb = 0.096 in 1985). With the real bond rate higher and tax revenues and real currency holdings lower, it is not at all improbable that, after the dust settles, the inflation rate will increase. Figure 7.2 depicts one possible scenario. Capital decumulation shifts the KK schedule down to K'K' and I1 inward to 1'1'. Equation (37) is satisfied so that the RR schedule shifts to the left and the real bond rate rises from rg to rh. The increase in r, partially offsets the initial leftward shift of the X X schedule and, together with the decrease in tax revenues and deposit holdings, produces a larger adjusted fiscal deficit. The larger deficit, reinforced by the reduction in real currency holdings, drives the inflation rate up to TI.

502

Edward F. Buffie

Fig. 7.2 The effect of equal decreases in the real deposit and real loan rates

7.2.1 Tax Policy and Current Expenditure Cuts As figures 7.1 and 7 . 2 clearly demonstrate, an increase in debt service creates grave macroeconomic problems when fiscal policy remains passive. If the government weakens in its resolve to maintain real bank interest rates, the economy is thrown into a cumulatively reinforcing spiral of accelerating inflation, rising real interest rates on the internal debt, capital decumulation, and financial disintermediation. This vicious cycle can be averted by pegging real interest rates, but the inflation rate is still certain to increase strongly. The simplest and most effective way out of these difficulties is to cut government consumption expenditures by an amount equalling the increase in debt service. This leaves private disposable income unchanged and keeps the fiscal budget in balance, thereby preserving the initial equilibrium. Other types of fiscal adjustment are not adequate substitutes for reducing governmentconsumption spending. Consider firstthe repercussionsof cutting the real public sector wage. The government wage bill decreases and workers reduce their holdings of both currency and deposits, so the X X and IZ schedules shift leftward to X‘X’ and Z’Z’ in figure 7.3. From (32) and (33) it is seen that the overall direct effect is deflationary under the weak assumption that15

PA +

PdZZ


0 requires (A20)

p'virb

-

-

Jo[q(l

(l

-

qq)[VDD

+ r,)q

-

-

vED(rb

I] > 0.

-

rd)l

The first two terms are positive as is the third under the very weak restriction that qq < (1 + r b ) - l . Workers Workers hold both types of money, deposits, D", and currency, C,but do not purchase bonds. They solve the optimization problem:

(A231

A

= Sj,

+

where A = D" C; Y" denotes total wage income; and T is the inflation rate. The indirect utility function V(-) exhibits the following characteristics: VE, VD,

V c , V E D , V E c > 0; V c D , v.1~9 Vcc, V D D Vcc - V D C < 0; V D D - V c D < 0 .

< 0;

516

Edward F. Buffie

The negative sign for V,, reflects the notion that the two assets are substitutes as mediums of exchange. Optimizing behavior requires that in a stationary equilibrium the marginal rate of substitution between deposits and goods equals the time preference rate less the deposit rate:

and that the marginal nonpecuniary benefits of currency exceed those of deposits by an amount equalling the nominal deposit rate (the opportunity cost of currency):

where

Ew = Y w ( l -

T)

+ r,D"

-

TC