The Politics of Canadian Public Policy 9781442682061

Table of contents :
Contents
Acknowledgments
1. Strategies for Policy Analysis
2. A Comparative Study of Language Policy in Quebec: A Political Economy Approach
3. The Politics of Provincial Resource Policy
4. Determinants of Legal Aid in Canada: Actors, Policies, Programs, and Futures
5.Technological Innovation and Industrial Policy: Canada in an International Context
6. Institutions, Constitutions, and Public Policies: A Public-Choice Overview
7. The Management of a Common Property Resource: Fisheries Policy in Atlantic Canada
8. The Political Economy of Policy Instruments: Tax Expenditures and Subsidies in Canada
9. Crown Corporations in Canada: The Choice of Instrument
10.The Development of Health Policy in Canada
11.The Policy Consequences of Northern Development
12. Will Program Evaluation Be Used in Formulating Policy?
Note on Contributors

Citation preview

The Politics of Canadian Public Policy

Edited by Michael M. Atkinson and Marsha A. Chandler The essays especially commissioned for this volume document the response of the state in Canada to developments in several crucial policy spheres. These include both traditional problem areas such as health, language, and natural resources and some of emerging interest or deepening concern such as legal aid, northern development, and industrial strategy. In this one volume readers may sample a number of approaches to the study of public policy, including new work on the political economy of Canada and in the research area known as public choice. In their introductory essay the editors argue that research on public policy in Canada has focused on four fundamental themes: the question of what determines variation in patterns of public policy; the importance of the policy field itself for understanding state activity; the choice and consequences of alternative means to a policy goal; the question of what happens after a decision is taken and acted upon. Each of the essays in this volume addresses one or more of these themes: policy determinants, policy type, policy instruments, and impact. Together they support the editors' contention that there are several routes to an understanding of the policy process in Canada and, indeed, several policy processes. Finally, the essays bring us some distance closer to the goal of understanding how state and society are related in Canada. MICHAEL M. ATKINSON is a member of the Department of Political Science at McMaster University. MARSHA A. CHANDLER is a member of the Department of Political Science at the University of Toronto.

This page intentionally left blank

The Politics of Canadian Public Policy EDITED BY MICHAEL M. ATKINSON AND MARSHA A. CHANDLER

UNIVERSITY OF TORONTO PRESS Toronto Buffalo London

www.utppublishing.com © University of Toronto Press 1983 Toronto Buffalo London Printed in Canada Reprinted 1985 ISBN 0-8020-2485-8 (cloth) ISBN 0-8020-6517-1 (paper)

Canadian Cataloguing in Publication Data Main entry under title: The Politics of Canadian public policy (Studies in Canadian public policy) ISBN 0-8020-2485-8 (bound). - ISBN 0-8020-6517-1 (pbk. ) 1. Canada - Politics and government - Addresses, essays, lectures. 2. Canada Economic policy - Addresses, essays, lectures. 3. Canada - Social policy Addresses, essays, lectures. 4. Policy sciences - Addresses, essays, lectures. I. Atkinson, Michael M. II. Chandler, Marsha A., 1945- III. Series. JL65 1983. P64

320. 971

C83-094086-3

Contents

Acknowledgments / vii

1 Strategies for Policy Analysis / 3 MICHAEL M. ATKINSON AND MARSHA A. CHANDLER

2 A Comparative Study of Language Policy in Quebec: A Political Economy Approach 721 WILLIAM D. COLEMAN

3 The Politics of Provincial Resource Policy / 43 MARSHA A. CHANDLER

4 Determinants of Legal Aid in Canada: Actors, Policies, Programs, and Futures 769 DALE H. POEL

5 Technological Innovation and Industrial Policy: Canada in an International Context 7 93 ARPAD ABONYI AND MICHAEL M. ATKINSON

6

Institutions, Constitutions, and Public Policies: A Public-Choice Overview/127 MARK SPROULE-JONES

vi / Contents 7 The Management of a Common Property Resource: Fisheries Policy in Atlantic Canada/ 151 SUSAN MCCORQUODALE

8 The Political Economy of Policy Instruments: Tax Expenditures and Subsidies in Canada/ 173 KENNETH WOODSIDE

9

Crown Corporations in Canada: The Choice of Instrument / 199 J. ROBERT S. PRICHARD AND MICHAEL J. TREBILCOCK

10 The Development of Health Policy in Canada / 223 GEOFFREY R. WELLER AND PRANLAL MANGA

11 The Policy Consequences of Northern Development / 247 SIMON MCINNES 12 Will Program Evaluation Be Used in Formulating Policy? / 267 JOHN MAYNE AND ROBERT S. MAYNE

Note on Contributors / 285

Acknowledgments

We would like to acknowledge the generous assistance of the Social Sciences and Humanities Research Council. The council provided a conference grant that enabled the authors to meet and assess the state of public-policy research in Canada. At the conference, the first drafts of the essays were presented, and were substantially revised in light of the criticism and discussions that arose. To all of the contributors, we are grateful for their diligence in preparing their own essays and for the assistance each gave to the others. As a group we are much indebted to the three referees selected by the University of Toronto Press. Each referee provided us with insightful and constructive comments. They will certainly see how crucial their comments were for the final versions of the essays. Our editor, Virgil Duff, has been unflagging in his support of this volume, and has gone far beyond the call of duty in providing assistance and encouragement. Kathy Johnson, our copy-editor, is perhaps the one individual most aware of our failings. We appreciate her help and good humour. We would like to thank Richard Simeon, Bill Coleman, and Bill Chandler for their comments on the introduction and various other aspects of the book. We would also like to thank the Department of Political Economy and the Faculty of Law at the University of Toronto and the Political Science Department at McMaster University for having supported this project. Finally, to the many students to whom we have presented this material, we would like to dedicate this book. It is written with the hope that it will shed some light on the politics of Canadian public policy. M. M. A. M. A. C.

This page intentionally left blank

THE POLITICS OF C A N A D I A N PUBLIC POLICY

This page intentionally left blank

I/Strategies for Policy Analysis MICHAEL M. ATKINSON MARSHA A. CHANDLER

The study of public policy in political science is no longer restricted to describing what governments do. Analysis has come to centre on a search for patterns and relationships that explain as well as describe the actions of governments. Central to this concern has been the view that political authorities are not simply receptors of political demands and the state not simply a mechanism for sorting out and pronouncing on them. Rather, it is held that governments reach out into their environments, moulding and shaping demands and often relying on nongovernmental institutions and actors to administer decisions. Accompanying the more traditional concern of how issues reach the public agenda, there is a new interest in how the issues themselves are structured by government. Social scientists steeped in the analysis of observables are coming to realize that the area of 'non-decisions' - options or issues that are not part of the public agenda - is no longer something that can be left untouched. Traditional studies of policy determinants are being augmented by studies of policy instruments, the particular means chosen to achieve policy goals. Added to detailed analyses of the content of policy has been a growing concern with policy impacts, the consequences of policy for the broader political and social environment. Under these circumstances it seems particularly necessary that students of public policy in Canada and elsewhere develop a clear conception of the state. Who are the key actors in the state apparatus? To what or whom do they respond? How do they generate and structure demands? How well do policy outcomes reflect their intentions? Unfortunately, there are no unambiguous answers to these questions. Indeed, as the contributions to this volume indicate, the search continues; and as it does, our image of the state becomes richer and more complex. It is true that two general images of the state can be taken as starting-points in any analysis. First, in the neo-Marxist paradigm, the state cannot be understood

4/M. M. Atkinson and M. A. Chandler apart from the class struggle inherent in capitalist societies. Capitalists do not rule directly through the instruments of the state, but the state's primary role 'entails maintaining the social conditions for economic growth and the reproduction of classes in a way consistent with the dynamics of the capitalist economy. '1 There is little consensus in the neo-Marxist literature on precisely how to conceive of the state.2 In some versions the state is simply a reflection of the power relations that exist among the classes in society. 3 Since the state is capitalist, existing in a society dominated by the capitalist class, the functions of the state and the interests of this class naturally coincide. In other, instrumentalist, versions the state is viewed as the tool of a self-conscious capitalist class, which is the ultimate beneficiary of virtually all state actions, including those which on the surface appear to benefit labour. Finally, it has been suggested that there is a division of labour between those who manage the state apparatus and those who direct the economy; the state has a certain amount of autonomy,4 although it is not entirely clear what that autonomy entails, how much of it there is, and under what conditions it can be exercised. In the neo-Marxist view of the state, public policies are portrayed on a large canvas, and the key to understanding the actions of the state is to recognize that the capitalist class enjoys certain structural advantages. These advantages derive from the fact that in a capitalist system investment decisions (what is to be produced and how resources are to be allocated) are decisions made in large measure by business. Private sector decision-makers are entrusted not only with ensuring their own well-being, but also with decisions affecting the production of goods and services used by everyone else.5 In this kind of system politicians ignore the demands of business at their peril. It is the task of the state, after all, to maintain the social, economic, and political stability congruent with capital accumulation. Of course the state in Canada, in addition to being capitalist, is also liberal democratic. Neo-Marxist views of the state often underestimate the requirements imposed on the state by its liberal democratic form. Elections, independent political parties, and autonomous interest groups are all features of a liberal democracy with which politicians must come to terms. Equally important for public policy is the state's institutional form-in Canada's case, federal and parliamentary.6 Moreover, it would be hard to explain the actions of the state without some reference to the quality of the bureaucracy and the level and nature of bureaucratic rivalry. In short, an understanding of the constraints imposed, and opportunities offered, by the political system is as important to the study of public policy as an appreciation of social and economic factors. The second view of the state, the pluralist version, is constructed around the requirements of liberal democracy - chiefly that the regime be open to extensive

Strategies for Policy Analysis / 5

public contestation. 7 The actions of the state are, in this view, not a function of class relations but instead are the outcome of competition in the political marketplace. Political parties compete for electoral support and citizens trade their votes for a preferred package of goods and services. The state emerges as something close to a huge public utility enterprise providing services for the mutual benefit of all citizens.8 Unfortunately, this rather benign image of the state has limited the usefulness of the pluralist view. The image is premised on the belief that the competition implied in these institutional arrangements is genuine and that no interest enjoys an unassailable advantage. Furthermore, those working within this tradition -students of political parties, interest groups, bureaucracies, and legislatures -have often stressed the input functions of these structures, and, as Simeon points out, the tendency has been simply to assume that they are important in determining the course of policy.9 The evidence, however, is inconclusive, and it is unlikely that the assembly-line model frequently implied in these studies is entirely appropriate. We are becoming increasingly aware that political parties do not merely aggregate demands; they also mould opinion and socialize their members. Interest groups not only seek to influence policy, but may in fact also assume the responsibility for administering state policies and for controlling their membership. In short, while we must acknowledge the importance of liberal democratic institutions for public policy, many of the presumptions with which pluralism approaches its study need to be reconsidered. Insights from each of these perspectives on the state are crucial to the study of public policy, and no effort has been made to force either of these models of the state on the authors of these essays. The essays are similar in their general approach to the study of policy-making in a state that is both liberal democratic and capitalist. In each essay the state is considered an active intervener, shaping the environment as well as responding to it. Governments are shown to have their own interests and to be actively pursuing them. Whether it is more appropriate to consider these actions within the context of a neo-Marxist view of the state or a pluralist one depends in some measure on the questions being posed. In considering some of the larger questions of economic development, political change, and social inequality, the authors of these essays go beyond narrow legalistic views of the state. In exploring the nuances and subtleties of public policy the authors are reluctant to reduce all issues to matters of class relations, for example, and thereby forgo the opportunity of constructing an image of the state and a theory of public policy based on both economic and political variables. Many of the essays in this volume move beyond a narrowly defined view of the political process and seek either to set the actions of the state within an economic

6/MM. Atkinson and MA. Chandler

and social context or to link public policies to economic processes. Increasingly in public-policy analysis the traditional concerns of political science are married to ideas drawn from macro-economic theory, from micro-economics, and from Marxian economics and its derivatives. As Lindblom has argued, political science and economics have both been impoverished by the practice of pressing questions of governments and markets on one another when, in fact, it is in the joint consideration of these themes that public-policy studies should be conducted.10 Although each essay is devoted to some specific aspect of the policy process, an attempt is made to place the particular findings in a broader context. Several of the essays draw upon some of the theoretical frameworks that have been slowly taking shape in the public-policy field, especially public choice and political economy. These essays tend to be deductive in form. They begin with an explicit set of general propositions, draw inferences, and seek empirical support by examining actual government behaviour. Other essays compare policy developments across provinces and over time. Authors using this approach begin with a description of various policy patterns and then move inductively to propose factors that will account for them. Crucial to such an approach is the selection of cases that appear to differ in regard to variables the author argues are important, and yet are sufficiently similar that the cases can be compared. In each of the essays, whether an inductive or deductive approach is used, there is an attempt to move beyond the particular and to discuss public policy in more general terms suitable for the development of theory. The authors differ in the manner in which they define their particular policy problem and in their mode of analysis. Represented in this volume are four strategies of analysis, each of which proposes a separate route into the policy process by posing a distinctive initial problem. These strategies correspond to the major problems that have dominated policy research in Canada: How is policy determined? Do processes vary across policy fields? How do policy instruments differ? What are the consequences of government decisions? In outlining these problem areas, and the tradition of research within them, we are proposing that the manner in which problems are defined is a valuable guide to the current state of public policy research and a basis for building an understanding of policy-making in Canada. These strategies are not mutually exclusive; several problems may be addressed simultaneously. By treating them separately at this stage, we will be able to discuss some of the consequences and advantages of different modes of problem definition. Policy Determinants The first strategy for studying policy is to focus on the determinants of policymaking. Do political factors account for diversity in policy outputs? What are

Strategies for Policy Analysis / 7

the policy consequences of variations in political institutions? To what extent do the policy choices of governments issue from the nature of class relations? What are the policy consequences of a capitalist mode of production and distribution in a dependent economy? These kinds of questions address the problem of how policy is shaped. Although the bulk of research in political science has focused primarily on the relevance of political forces, the determinants strategy embraces the effects of social and economic forces as well. 11 In Canada, the tradition of political economy has consistently sought to link the role of the state to the structural conditions of the economy. 12 The nature of the class structure associated with changing modes of production and changing patterns of economic imperialism associated with American capital are among the forces thought to condition the action of the Canadian state. The political economy tradition places Canadian political experience within a broad and unique set of social and economic circumstances, emphasizing that the economy has operated as a staple-extraction hinterland largely under the domination of foreign capital. Other policy analysis undertaken to date in Canada has relied on a determinants strategy that emphasizes the direct role of political factors. The primary concern has been to specify the nature of political activity and to demonstrate how it shapes policy outputs. The term 'political activity' can refer either to behaviour itself (for example, class voting patterns) or to political institutions such as the electoral system, whose rules and incentive systems structure behaviour. Explicit in the political determinants strategy is the view that variations in political behaviour or adjustments to political institutions will have an effect on public policy. Political parties are a case in point. They are structures that link individuals and the state, and as such the policy impact of political parties is a key element in any assessment of policy determination. Can elections and party strengths account for differences in public policies? Does it matter which party comes to power? Do parties of the left fashion policies that differ from those of other parties? The analysis of policy determinants in Canada has documented a number of important linkages between policy outputs and politics. Dale Poel's study of policy diffusion, for example, identified electoral maldistribution as an important factor in governing the rate at which provinces adopted certain policies.13 William Mishler and David Campbell found that legislative development had a marked impact on the responsiveness of provincial governments to health care needs, 14 while William Chandler concluded that in some circumstances the strength of left-wing parties in opposition strongly influenced changes in the pattern of government expenditure.15 Pursuing a model of policy outputs based on electoral self-interest, Donald Blake found that in its early stages the process of distributing Local Initiatives Projects (LIP) grants was riddled with partisan

8/MM. Atkinson and M. A. Chandler

considerations, 16 and Bruce MacNaughton and Conrad Winn concluded that Liberal governments were seeking vote optimization in the distribution of Department of Regional Economic Expansion (DREE) grants. 17 In a similar vein, Duff Spafford found overwhelming evidence of an electoral cycle in which provincial highway expenditures were related to the timing of provincial elections. 18 These works, as well as political economy research, constitute a valuable core of policy studies. But they have by no means exhausted the range of factors that may influence government policy, nor have they been able to synthesize the many and sometimes contradictory findings. Generally speaking, when public policy is defined in broad terms such as expansion of the public sector, equally broad macro-political or socio-economic factors have been called upon to account for the shape of policy. This is why such variables as class structure, 19 party composition and ideology,20 and degree of political centralization21 are often treated as policy determinants. When students of public policy place emphasis on more proximate factors, such as legislative institutionalization22 and bureaucratic incentives, 23 they are frequently trying to account for narrower, more tightly defined actions or policies. Regardless of the choice of variables, there are three crucial steps in the analysis of the determinants of policy. First, it is necessary to identify as precisely as possible those government actions or policies that are to be explained. 24 Second, it is necessary to have adequate measures of political, social, and economic activity. Finally, the process that links these variables to one another should be explicit. It should be clear how the social and economic factors bear on political forces as well as how political factors themselves are related to policy.25 Several of the essays in this volume are primarily studies of policy determination. It is our belief that no one set of variables or single process applies across the entire range of policy-making; the essays in this volume concern a wide number of policy fields and focus on several potential explanations. In the context of language policy, William Coleman assesses the conditions under which parties with different ideologies also offer different policies. He argues that one can expect few differences among parties when their policy initiatives impinge on capital accumulation. In a study with similar objectives, Marsha Chandler suggests that a considerable amount of the variation in the extent and type of provincial intervention in the resource policy field can be attributed to differences in the ideologies of the parties in power, particularly since 1970. While these authors are both persuaded of the importance of party ideology to public policy, there remain differences between them on the conditions under which ideological factors are decisive. In their discussion of policies in the area of technological innovation, Arpad Abonyi and Michael Atkinson eschew ideology altogether, focusing instead on

Strategies for Policy Analysis /9

the organization of the bureaucracy and the absence of a constituency willing to support new initiatives. Dale Poel's analysis of the determinants of legal aid concentrates on proximate factors, such as the posture of the law societies in each province and the extent of provincial financial resources. He notes that the demands of clients play a relatively small role in determining the delivery of legal aid services. In a paper that outlines the conceptual and propositional foundations of public choice, Mark Sproule-Jones makes a more general case for the importance of institutional arrangements in determining how much of a public good will be produced, who will consume it, and who will pay for it. Each of these essays attempts to characterize the process of policy determination. What knits these studies together is a general interest in ascertaining not only which particular forces are important for the output of public policy, but also under what circumstances they are important. Policy Types Examinations of different fields of public policy sometimes yield diverse explanations of policy-making. The analysis of any one policy field usually leads to the question of whether the process might have varied if there had been another policy field under consideration. For example, the influence of pressure groups on federal-provincial relations appears to vary by policy type: Richard Schultz found groups to be important actors in transportation policy,26 but on such questions as tax reform Richard Simeon concluded that they had little impact.27 In fact, some studies of the determinants of policy have come to conclude that aggregating across policy fields obscures significant differences in the policymaking process.28 Even a division of policies into traditional substantive fields such as transportation, social welfare, and education reveals useful distinctions.29 The real significance of separating policy into substantive categories is that doing so is the beginning of the search for those underlying dimensions of policy that are linked to process. In essence, it is a way of treating aspects of policy as independent variables that have a part in shaping the political process. One of the earliest and most influential efforts in this regard was Theodore Lowi's classification of policy based on expected outcomes.30 In subsequent formulations Lowi argued that policy types could be distinguished by the degree of coercion applied. What has made Lowi's work so important is the explicit linkage he sought to draw between the type of policy under consideration and variations in the policy-making process that are related to policy types. Policy, it appeared to Lowi, could predict politics. 31 Since Lowi's seminal article, a new generation of analysts has suggested numerous criteria on the basis of which policies might be categorized and types identified and ultimately related to the policy-making process. A particularly

10/MM. Atkinson and M. A. Chandler useful discussion appears in Public Policy in Canada, where Peter Aucoin argues that the degree of coercion, the nature of distribution, and the extent of systemic impact constitute three central properties of policy that are associated with differences in decision-making processes. 32 A major contribution of the policy type strategy is that it underlines the fact that there are several pathways to policy, and that the path or process observed is, to some degree, contingent on the nature of the policy. In other words, no single pattern describes the policy process. Depending on the policy considered there may be differences in the location of decision-making, the participants, and the relationships among decision-makers. It is this knowledge that has forced researchers to outline in more detail the nature of their policy area and to isolate those features that condition the response of policy-makers. In this volume, Chandler draws attention to the fact that at least some natural resources are, by their nature, exhaustible, which means that the principal policy struggles will likely involve proprietary rights. Geoffrey Weller and Pran Manga point out that much of the health care policy field, involving as it does inter-governmental negotiations, is, by virtue of that fact, characterized by limited interest-group access. Coleman goes even farther in insisting on the importance of the policy field, suggesting that the relative autonomy of the state varies with policy type, and that the more distant the content of the policy from the process of capital accumulation, the more autonomy that exists. It is in the field of public choice, however, that most attention has been paid to the precise nature of the goods that governments seek to provide. Sproule-Jones suggests that categorizing public goods along two dimensions - their relative exclusivity and their relative availability - is a necessary step for an understanding of the varying nature of policy problems in different areas and for an appreciation of government response. The pattern of decision-making is affected by the nature of the public good, and the efficacy and appropriateness of public policy can be evaluated only if the nature of the good is first identified. In a paper influenced by this line of thought, Susan McCorquodale begins her examination of fisheries policy by pointing out that the fishery is a common property resource. The common-property nature of this good and the problem of establishing proprietary rights have in large measure influenced the behaviour of fishermen and conditioned the responses of succeeding governments. Similarly, as Abonyi and Atkinson point out, the difficulty that firms face in establishing proprietary rights to technology and their own response to this problem set the stage for government action in the area of technological innovation. These papers illustrate that attention to differences in the nature of a policy under consideration is the first step in acknowledging that there is more than one

Strategies for Policy Analysis /11

policy process. Distinguishing among policy fields will not, by itself, reveal how governments will behave. Analysis of policy types does, however, mean a recognition of differences in the structure of the policy-making process. To this should be added the observation that there is more than one way to define the problems that policies are intended to address. Poel's study, for example, illustrates several problems that are seen as the rationale for legal aid policies. The perceptions of decision-makers, as well as the structure of the policy field, can condition the response of governments. 33 Policy Instruments The third strategy used to explore the policy process is the comparison of policy instruments. This strategy is based on a recognition that policy-making does not end with the decision to allocate resources among competing objectives; the process of policy-making also entails the choice and deployment of policy instruments, the means employed to attain policy objectives. The study of policy instruments and instrument choice is a relatively new area of research. Until recently, policy analysis has concentrated on the choice of goals, and much less attention has been devoted to the various ways in which they might be attained. The growth of the public sector has been an issue interpreted largely in expenditure terms, creating the erroneous impression that governments seek their objectives primarily by spending money. Beginning with the study of regulation, a considerable effort has been made in recent years to consider other arenas of government action, including taxation and public enterprise.34 Governments themselves, under increasing criticism for uncontrolled spending, have sought to reform the budgetary process and to consider alternative instruments for the achievement of their objectives. More recently still, students of public policy have begun to ask what prompts governments to prefer some policy instruments over others. There is no consensus on the precise range of available policy instruments, but most lists would include the following: exhortation, subsidies, taxation, regulation, public ownership, government reorganization, and a variety of symbolic gestures. Given the objective of increasing the extent of Canadian ownership of natural resources, for example, a government may offer direct grants to subsidize the purchase by Canadians of foreign-controlled resource firms, or it may use the tax system to provide advantages to Canadian firms. Alternatively, the government may regulate the sale of shares in resources companies to nonCanadians, or it may turn to public ownership as a means of keeping resources Canadian-owned. In each case the objective is the same, but the consequences of choosing one means or another differ considerably.

12/M. M. Atkinson and M. A. Chandler

Let us consider briefly three types of consequences of instrument choice, remembering that these anticipated consequences are likely to have a considerable effect on the preferences of decision-makers for one instrument rather than another. First, the personal goals of the decision-maker will be affected by the instrument chosen. If the decision-maker is a politician interested in re-election he may prefer to achieve the goal of regional economic expansion by providing subsidies, since these are large, visible transfer payments that lend themselves to carefully orchestrated announcements and considerable political benefit. The bureaucrat, on the other hand, may prefer the creation of a public corporation for this purpose on the ground that the interests of his department will be undermined by anything less than an instrument that permits the close monitoring of public investment in the regions. In both cases the means preferred are strongly influenced by the ends to be achieved, and these will differ from actor to actor. Second, the choice of policy instruments has immediate consequences for the type of political process required to implement the choice. Each instrument is considered in a distinct policy arena with its own set of participants, rules, and relationships. The regulatory arena, for example, is often a narrow, policyspecific, quasi-judicial setting, whereas the expenditure budget is an arena in which both broad objectives and programmatic details are contested by competing bureaus, with all participants attuned to the norms imposed by a hierarchical structure. Like the policy type strategy, the analysis of policy instruments assumes that there are, in fact, several policy processes. The choice of an instrument is in effect the choice of a process, and subsequent developments in the policy field will depend in large measure on the procedures for monitoring and review that are in place in that particular policy arena. Finally, the choice of policy instruments has both equity and efficiency implications. Whether or not distributional impacts are intended, the means chosen to accomplish some end invariably has its own consequences. If the tax system is the chosen instrument for encouraging investment in research and development (R & D), for example, large firms will tend to benefit more than small ones since large firms are better placed to maintain their own R & D capacity. A system of subsidies can be targeted more carefully and can include, if desired, eligibility guidelines with respect to firm size. Similarly, the chosen instrument will have efficiency implications. As Robert Prichard and Michael Trebilcock point out in their essay, crown corporations are efficient instruments for some purposes; they suggest that such efficiency considerations weigh heavily in determining instrument choice. In one of the first attempts to identify a pattern of instrument choice, Bruce Doern and Vince Wilson argued that politicians tend to respond to issues by moving from the least coercive to the most coercive instrument. Exhortation or

Strategies for Policy Analysis /13

government reorganization is often followed by a distributive spending approach, until at last the government employs regulatory devices.35 It is unlikely, of course, that the application of coercion is the only factor governing instrument choice. The relative substitutability of instruments is a theme pursued by Prichard and Trebilcock, and these authors and their colleagues have also introduced a calculus of instrument choice based on politicians' electoral self-interest.36 Why some instruments should be preferred to others is the subject of Kenneth Woodside's comparison of tax expenditures and subsidies. Identifying the winners and losers of instrument choice, as Woodside does, is an essential requirement for a theory of instrument choice based on selfinterest. Several essays in this book deal with the comparison of policy instruments within specific policy fields. Distinguishing between direct and indirect instruments, Abonyi and Atkinson consider why successive Canadian governments have employed a limited range of indirect policy instruments in the area of technological innovation. Chandler compares the instruments used to collect resource rents in the provinces, and Poel considers the various methods of delivering legal aid. The intrinsic advantages and disadvantages of each instrument, as well as their appeal for particular actors and parties to the policy process, are themes developed in these and other essays. Policy Impact

Those who study policy impact begin with the question: what are the consequences of government decisions? From the impact perspective, the policy output is not the last stage in the policy process. On the contrary, analysis starts with the policy and proceeds by tracing its effect. Instead of attempting to ascertain the determinants of policy, policy itself is treated as the determinant and an effort is made to ascertain the role of policy in changing the political environment. Or, as John Grumm puts it, 'with policy-impact research we are given an explanation and are* asked to find some phenomenon which it explains. '37 This research strategy faces some serious definitional questions. In the first place, what is policy? Extremely broad definitions of policy - 'What governments choose to do or not to do, ' for example-tend to blur the distinction between policy and its impact because it is not always easy to decide what governments have done by choice. Is urban decay, for example, a government policy, namely something that has deliberately been permitted to happen, or does it show the impact of government policy combined with other social and economic factors? To avoid ambiguity, impact analysis must distinguish between the policy and its effects. By the same token, the concept of impact itself

14/M. M. Atkinson and M. A. Chandler

demands some refinement. Many conceivable impacts may result from a single government action. Some of these, first-order impacts, will occur immediately, others perhaps over an extended period of time. There are intended and unintended consequences, individual effects, group effects, and general societal effects. 38 Even more pressing is the general issue of just what kind of impacts we should be interested in. There is, of course, the issue of who profits from government action and who loses by it. Although the subject matter of political science has sometimes been summarized as the study of who gets what when and how, the distributional effect of government policies is an issue that, in Canada, has been addressed primarily by economists. Since government actions often create private burdens, but also bestow both public and private benefits, the task has often been one of determining just what effect government actions have on the distribution of income in a society.39 The measurement of certain public benefits - who benefits most, for example, from expenditures on highways? - is a serious, although not intractable, problem. Research in this area has concentrated on specific policy areas, efforts being made, for example, to determine who benefits from and who pays for public expenditures on post-secondary education.40 The impact of policies has also been studied under the general rubric of 'program evaluation. ' During the last decade, in particular, governments have shown increasing interest in whether programs with announced objectives are in any sense achieving them. In their paper for this volume, John Mayne and Robert Mayne outline some of the difficulties that beset program evaluation, particularly methods of evaluation based on quasi-experimental designs. They extend their analysis to the issue of how program evaluation can inform subsequent policy change. As they make clear, it is by no means easy to inject the results of program evaluation into the policy-making process. In this area the need to distinguish between policy and impact is particularly pressing, since programs need to be assessed against policy objectives. Yet the process of defining policy objectives is itself a frustrating one. In the study of policy impact, political scientists have shown most interest in the effects of policy on political behaviour and public institutions.41 It has long been recognized, for example, that in the creation of programs governments are also creating vested interests: sometimes represented by the bureaucracies charged with delivery of a service, sometimes by the clients who are on the receiving end. What effect do these whirlpools of complementary and competing interests have on future policy development? To what extent does the creation of a new department or agency provide a channel of access for previously unrepresented groups?42 By specifying the impact of policy on various components of the political system, the basis for subsequent policy action

Strategies for Policy Analysis /15

becomes more apparent. Heclo, for example, has demonstrated that the development of social welfare policy in Britain and Sweden can be traced to monitoring and the subsequent adjustments made by the bureaucracy rather than to new sets of demands.43 In his paper on northern development, Simon Mclnnes outlines the conflict between contending objectives in a single policy area, identifies a series of intended and unintended consequences, and comments on the political mobilization triggered by government policy. Recent interest in the study of policy impacts has been prompted, at least in part, by the feeling that the process of policy-making, particularly its shortcomings, cannot be fully appreciated without examining policy outcomes. The impact strategy brings a much-needed developmental perspective to the study of policy, emphasizing, as Weller and Manga do in their paper on health policy, that the output of one stage of the policy process can serve as an input for the next. Kenneth Bryden's study of Canadian pensions policy44 and Malcolm Taylor's work in the area of health45 are early examples of this type of thinking. The dynamics of policy-making, so hard to capture by studying institutions and personalities, yield much more readily if one can make sense of, and generalize about, policy development. A Composite View

The object of this collection of essays is to explore the roles of the state in a variety of policy settings, and to provide an overview of different approaches to the analysis of the policy process. To this end, these studies describe the substance and processes of policy-making as well as the forces that give rise to public policy. In the belief that there is no one path to policy development or any single compelling explanatory theory, we have chosen to survey a range of substantive fields and analytical questions.46 Although there is no commitment among the authors to a single paradigm or model, the essays do share a common approach to policy-making. Each study treats the state as an active agent in the policy process. Although the studies all are essentially concerned with political forces and the workings of the political system, this concern does not preclude consideration of social and economic settings. The questions of policy determination, type, instrument, and impact provide distinctive but complementary perspectives on the study of public policy. Both the issue of instrument choice and that of policy type, for example, are based on the hypothesis that there is more than one policy process; the study of policy determination and that of impact are both concerned with identifying those forces behind the development of policy. Although individual authors may have a preferred route into the study of policy, no general understanding of the policy

16/M. M. Atkinson and M. A. Chandler process can afford to neglect entirely the issues raised in each of the four strategies. While several of the essays deal with a particular policy field, each author uses his or her analysis to draw out generalizations that go beyond the specific topic. In this way the application of the four research strategies to various policy fields yields a broad composite picture of the politics of the policy process. What emerges is a multitude of processes and a number of significant political forces. A tidy, consistent view of the state and policy-making would be more satisfying, but these essays suggest that no such view is readily available. There are, instead, multiple avenues of analysis, and the challenge is to combine these for a wide and cumulative understanding of Canadian politics and policy-making. NOTES 1 Leo Panitch, 'Dependency and Class in Canadian Political Economy, ' Studies in Political Economy 6 (Autumn 1981), 24 2 For a discussion of different approaches to the state in the neo-Marxist literature see Bob Jessop, 'Some Remarks on Recent Theories of the State, ' Cambridge Journal of Economics 1 (Dec. 1977), 353-73, and Theda Skocpol, 'Political Response to Capitalist Crisis: Neo-Marxist Theories of the State and the Case of the New Deal, ' Politics and Society 10 (1980), 155-201. The discussion in this paragraph follows Skocpol. 3 Nicos Poulantzas, Political Power and Social Classes (London: NLB and Sheed & Ward 1973). For an adaptation of Poulantzas to the Canadian context see Rianne Mahon, 'Canadian Public Policy: The Unequal Structure of Representation' in The Canadian State: Political Economy and Political Power, ed. Leo Panitch (Toronto: University of Toronto Press 1977), 133-98. 4 Leo Panitch, 'The Role and Nature of the Canadian State, ' in The Canadian State, 3-27. 5 This point is made at some length by Charles Lindblom, Politics and Markets (New York: Basic Books 1977), esp. ch. 13, and is discussed approvingly in Colin Crouch, 'The State, Capital, and Liberal Democracy, ' in The State and Economy in Contemporary Capitalism, ed. Colin Crouch (New York: St Martins 1979). 6 A recent Canadian contribution that assesses the importance of federalism to the development of policy is Keith Banting, The Welfare State and Canadian Federalsim (Montreal: McGill-Queen's University Press 1982). 7 The term often used to refer to such regimes is polyarchy. See Robert Dahl, Polyarchy: Participation and Opposition (New Haven: Yale University Press 1971). 8 An outline and critical review of this version of the state is found in Charles

Strategies for Policy Analysis /17 Lindblom, 'Another State of Mind, ' American Political Science Review 76 (March 1982), 9-21. 9 Richard Simeon, 'Studying Public Policy, ' Canadian Journal of Political Science 9 (Dec. 1976), 552. For an example of one effort to correct this deficiency see Richard Simeon and Robert Miller, 'Regional Variations in Public Policy' in Small Worlds, ed. D. Elkins and R. Simeon (Toronto: Methuen 1980), 242-84. 10 Lindblom, Politics and Markets, ix. 11 A recent collection of essays that survey a broad set of variables is Contemporary Political Economy, ed. Douglas Hibbs and H. Fassbender (New York: Elsevier North-Holland 1981). 12 The literature here is too extensive to cite in any detail. However, some examples of old and new include Hugh Aitken, 'Defensive Expansionism: The State and Economic Growth in Canada' in Approaches to Canadian Economic History, ed. W. T. Easterbrook and M. H. Watkins (Toronto: McClelland and Stewart 1967), 183-221; Wallace Clement, Continental Corporate Power (Toronto: McClelland and Stewart 1977); The Canadian State, ed. Leo Panitch; Glen Williams, 'The National Policy Tariffs: Industrial Underdevelopment through Import Substitution, ' Canadian Journal of Political Science 12 (June 1979), 333-68; and John Richards and Larry Pratt, Prairie Capitalism: Power and Influence in the New West (Toronto: McClelland and Stewart 1979). 13 Dale Poel, 'The Diffusion of Legislation among the Canadian Provinces, ' Canadian Journal of Political Science 9 (December 1977), 605-25. 14 William Mishler and David B. Campbell, 'The Healthy State: Legislative Responsiveness to Public Health Care Needs in Canada, 1920-1970, ' Comparative Politics 10 (1978), 479-97. 15 William Chandler, 'Canadian Socialism and Policy Impact: Contagion from the Left?' Canadian Journal of Political Science 10 (Dec. 1977), 755-80. 16 Donald Blake, 'LIP and Partisanship: An Analysis of the Local Initiatives Program', Canadian Public Policy 1 (Winter 1976), 17-32. 17 B. D. MacNaughton and C. J. Winn, 'Economic Policy and Electoral Self-interest: The Allocations of the Department of Regional Economic Expansion, ' Canadian Public Policy 1 (Spring 1981), 318-27. 18 Duff Spafford 'Highway Employment and Provincial Elections, ' Canadian Journal of Political Science 14 (March 1981) 135-42. 19 For an example in the Canadian case see Pierre Fournier, The Quebec Establishment (Montreal: Black Rose Books 1976). 20 For example, David R. Cameron, 'The Expansion of the Public Economy: A Comparative Analysis,' American Political Science Review 12 (Dec. 1978), 1243-61, and Douglas Hibbs, 'Political Parties and Macro-economic Policies, ' American Political Science Review 11 (Dec. 1977), 1467-87.

18/M. M. Atkinson and M. A. Chandler 21 Christopher Leman, 'Patterns of Policy Development: Social Security in the United States and Canada, ' Public Policy 25 (Spring 1977), 261-91. 22 Mishler and Campbell, 'The Healthy State. ' 23 Douglas Hartle, A Theory of the Expenditure Budgetary Process (Toronto, Ontario Economic Council 1976). 24 On the 'problem of the dependent variable' see Simeon, 'Studying Public Policy, ' 552. 25 The need to explore these relationships is well illustrated in a recent study that attempts to specify the conditions under which party competition affects welfare policies by distinguishing between class-based and non-class-based competition: Edward T. Jennings, 'Competition, Constituencies and Welfare Policies in American States, ' American Political Science Review 73 (June 1979), 414-29. 26 Richard Schultz, Federalism, Bureaucracy and Public Policy (Montreal: McGillQueen's University Press 1980), chap. 8. 27 Richard Simeon, Federal-Provincial Diplomacy (Toronto: University of Toronto Press 1972), 280-3. 28 Richard Bird, The Growth of Government Spending in Canada (Toronto: Canadian Tax Foundation 1970), ch. 8; Issues in Canadian Public Policy, ed. G. Bruce Doern and V. S. Wilson (Toronto: Macmillan of Canada 1974), 5; and Simeon, 'Studying Public Policy,' 555. 29 Lewis Froman, 'The Categorization of Policy Contents' in Political Science and Public Policy, ed. Austin Ranney (Chicago: University of Chicago Press 1969). 30 Theodore Lowi, 'American Business, Public Policy, Case Studies and Political Theory, ' World Politics 16 (July 1964), 677-715; and Lowi, 'Four Systems of Politics, Policy and Choice, ' Public Administration Review (July/Aug. 1972). 31 For an application to Canada see Robert S. Best, 'Youth Policy, ' in Issues, ed. Doern and Wilson, 137-65; and John Shiry, 'Distributive and Regulative Policy in Ontario: A Test of Lowi's Arenas of Power Schema,' paper presented to the annual meeting of the Canadian Political Science Association (Montreal, May 1976). 32 Peter Aucoin, 'Public Policy Theory and Analysis,' in Public Policy in Canada, ed. G. Bruce Doern and Peter Aucoin (Toronto: Macmillan of Canada 1979), 1-26. 33 Charles W. Anderson, 'The Logic of Public Problems: Evaluation in Comparative Policy Research,' in Comparing Public Policies: New Concepts and Methods, ed. Douglas E. Ashford (Beverly Hills: Sage Publications 1978X 19-42; and, in Canada, Ronald Manzer, 'Public Policies in Canada: A Developmental Perspective,' paper presented to the Canadian Political Science Association (Edmonton, June 1975). 34 Bruce Doern has called regulation 'the "other" half of the reality of government. ' See his 'The Concept of Regulation and Regulatory Reform, ' in Issues, ed. Doern

Strategies for Policy Analysis /19

35 36 37

38 39

40 41

42 43 44 45 46

and Wilson, 8-35; and Douglas G. Hartle, Public Policy Decision Making and Regulation (Toronto: Butterworths 1979). Doern and Wilson, 'Conclusions and Observations, ' in Issues, ed. Doern and Wilson, 338-41. M. J. Trebilcock, D. G. Hartle, R. S. Prichard, and D. N. Dewees, The Choice of Governing Instrument (Ottawa: Economic Council of Canada 1982). John Grumm, 'The Analysis of Policy Impact, ' in The Handbook of Political Science vol. 6, ed. Fred Greerstein and Nelson Polsby (Toronto: Addison-Wesley 1975), 449. Grumm, 'Policy Impact, ' 443. See, for example, W. Irwin Gillespie, The Redistribution of Income in Canada (Toronto: Gage 1980); and Grant Reuber, 'The Impact of Government Policies on the Distribution of Income in Canada: A Review, ' Canadian Public Policy (Autumn 1978), 505-29. O. Mehmet, Who Benefits from the Ontario University System? (Toronto; Ontario Economic Council 1978). On the issue of just what political scientists should be concerned with, see Theodore Lowi, 'What Political Scientists Don't Need to Ask about Policy Analysis,' in Policy Studies and the Social Sciences, ed. S. S. Nagel (Lexington: Heath 1975). Richard Rose, 'Comparing Public Policy, ' European Journal of Political Research (April 1975), 67-93. Hugh Heclo, Modern Social Policies in Britain and Sweden (New Haven: Yale University Press 1974). Kenneth Bryden, Old Age Pensions and Policymaking in Canada (Montreal: McGill-Queen's University Press 1974). Malcolm Taylor, Health Insurance and Canadian Public Policy (Montreal: McGillQueen's University Press 1978). A similar attitude is readily detectable in Francis G. Castles, 'Introduction: Politics and Public Policy' in The Impact of Parties, ed. Francis G. Castles (Beverly Hills: Sage Publications 1982), 1-18, although here the focus is provided, as the title suggests, by concentrating on a single institution.

This page intentionally left blank

2 / A Comparative Study of Language Policy in Quebec: A Political Economy Approach WILLIAM D. COLEMAN

In the study of public policy in advanced capitalist states, it has become increasingly difficult to assess the policy impact of political parties and the specific ideologies they espouse. This difficulty arises from the fact that the usual lines of cleavage dividing parties along- socialist and non-socialist lines in capitalist democracies often become blurred when one turns to assess the performance of those parties in office. When those socialist parties that had committed themselves to achieving reform within the liberal democratic system first came to power in the late 1920s and early 1930s, they hardly fomented substantial change. 'Socialists behaved like all other parties: with some distributional bias toward their constituency but full of respect for the golden principles of the balanced budget, deflationary anti-crisis policies, gold standard, and so on. '1 In the late 1930s and early 1940s, these same parties adopted Keynesian economics as their central policy and affirmed their commitment to the maintenance of the capitalist mode of production. Such a commitment appears to have forced them to continue to behave like most bourgeois parties when in office. The question arises whether students of public policy should pay much attention to apparent differences in the ideologies and social bases of parties that have a common commitment to capitalism. The answer to this question, I shall argue, is yes, provided that one takes into account the specific relationship of the policy in question to the process of capital accumulation. By capital accumulation, I refer simply to the process whereby surplus monies are accumulated through profits in the capitalist economic system. The greater the distance between a given policy and the activities of capitalists appropriating surplus and reinvesting to maintain and expand profits, the more ideologies and differences in clientele among parties are likely to have an independent impact on policy. One way to examine this proposition in a systematic fashion would be to take a case where two political parties espousing different ideologies and drawing

22/W. D. Coleman

from different social bases have legislated in the same policy areas in the same system. The example would require that two policy areas differing in their distance from the process of capital accumulation be available for study. Ideally, the two parties should have proposed different approaches to the formation of policy in these areas. By focusing on the system, preferably within a relatively short period, it is likely that one could isolate the impact of party and ideology on the respective policies. In the predominantly French-speaking Canadian province of Quebec, a case with many of the above characteristics occurred between the spring of 1974 and the summer of 1977. During that period, two different political parties, the Parti liberal du Quebec (PLQ) and the Parti quebecois (PQ), formed successive governments. The PLQ, elected to power in 1970, lost the right to govern when the PQ gained a majority of seats in the November 1976 election. These two parties diverged both in social base and in ideology. The PQ drew its support largely from the French majority in the province; the PLQ had the virtually unanimous support of the province's ethnic minorities, particularly the largest of these -English speakers of British descent-as well as support from the French majority. Within the francophone community the PQ drew most of its support from professionals and semi-professionals, particularly those in the public sector; managers and'top administrators in the private sector supported the PLQ. Among workers, those who were in favour of social change favoured the PQ.2 Fournier's study of the corporate elite of the province did not find one of his subjects who supported the PQ.3 A comparison of candidates for the two parties in the 1970 and 1973 elections shows that 53. 2 per cent of the PQ candidates were drawn from a 'new middle class' (public servants, teachers, professors, journalists) compared with only 22. 9 per cent of the PLQ candidates. 4 These differences in social base were complemented by differences in ideology. The PQ was explicitly a social democratic party, calling for extensive direct public intervention in the economy and for the promotion of francophones to positions of influence. In contrast, the PLQ was in favour of giving incentives to the private sector to develop the economy and relying upon private firms to take the initiative in promoting francophones to positions of influence. The PQ favoured moving toward the political independence of Quebec; the PLQ wanted Quebec to remain part of Canada. Between the spring of 1974 and the summer of 1977, both parties sought to implement a policy on the French language. The PLQ introduced its Official Language Act (bill 22) to the National Assembly in the early spring of 1974; the bill was made law in late June 1974. The PQ published a white paper on language

Language Policy in Quebec / 23

in March 1977, introduced the first version of its Charte de la langue francaise in April (bill 1) and a second version (bill 101) in July 1977. The latter version was made law in August of the same year. Both parties developed policies in two spheres that clearly differed in their distance from the accumulation process and that were viewed differently by the two parties. The first of these spheres centred on the objectives of expanding the use of French as a language of work at middle and senior levels of management in the private sector (francization) and promoting more francophones to these same levels (francophonization). Policies designed to realize these objectives could conceivably impinge directly on the rights of capitalists to control production and investment. Restrictions were being placed on the personnel they could draw upon. The second sphere pertained to the objective of integrating the local public institutions of Quebec's non-francophone communities - municipalities, school-boards, hospitals, and social service agencies - more closely with those of the francophone majority. Policies affecting these kinds of institutions were considerably more distant from the process of accumulation of capital than those of francization and francophonization. This particular case allows us to examine the effect of party differences on the character of policy because it has the properties defined above. The case is suitable for two additional reasons. The policy developed by the PLQ was only beginning to be implemented when that party was defeated in the 1976 election, and did not significantly constrain the PQ when it came to formulate its own policy. What is more, no major social event occurred between 1974 and 1977 to change the conditions creating a need for a language policy or the environment in which such a policy was to take effect. In order to demonstrate the relative role of party and ideology in determining these policies, I shall begin by examining theoretically the particular dependence of the capitalist state on the process of capital accumulation. This discussion will lead to the presentation of a proposition concerning the place afforded to party differences in the development of policy. This proposition will then be studied by comparing the differences between the PLQ and the PQ in their respective formulations of policy for francization and francophonization in the private sector, and for integrating local public institutions serving the English-speaking community into the broader Quebec society. Theoretical Introduction The first and most important assumption underlying the approach to the study of public policy developed here is that policies must be assessed in terms of their closeness to or impact on the process of capital accumulation. This assumption

24/W. D. Coleman

follows logically from another, namely that the capitalist state is itself dependent upon the successful persistence of accumulation. What is meant by the use of the term 'dependent' in this instance? It is a basic fact that in any economic system, if long-term growth is to take place, a part of the product of that system must be withheld from immediate consumption and directed toward increasing productivity. A capitalist economic system is distinguished first of all by the fact 'that the part withheld from current consumption is derived to a great measure from the part withheld from the immediate producers, ' namely the workers. 5 Second, in such a system, the allocation of the product to be withheld is made on the basis of decisions taken by private entrepreneurs. As Offe writes, 'The basic prerogative of free enterprise is a negative one: the right not to produce unless production is accumulative, that is, not only production of useful things but simultaneously production of surplus value or profit. '6 These two distinguishing characteristics of the capitalist system have a number of important consequences. As long as control over accumulation remains in private hands, all of society, including the state, is dependent upon enterprises being profitable and capitalists being efficacious in their investments. It is in the interest of everyone that capitalists turn profits. The occupants of positions of political power can only appropriate the resources they need from taxation and from the capital market if wages are being earned and capital is being accumulated. 7 In such a system capitalists come to appear as the champions of universal interests because the realization of their material interests is a precondition for the realization of the material interests of any other group. The state and the parties controlling the state apparatus then have a double interest in supporting the private accumulation of capital. Unless accumulation takes place, it will be impossible to satisfy the material interests of the various groups in society and it will be difficult to sustain the continued existence of the state apparatus itself. Offe concludes: Thus, every interest the state (or the personnel of the state, its various branches and agencies) may have in their own stability and development can only be pursued if it is in accordance with the imperative of maintaining accumulation; this fundamental dependency upon accumulation functions as a selective principle upon state policies. '8 This dependency influences not only which policy areas the state selects for consideration but also the particular form the state gives to a policy. At the same time it must be noted that the capitalist state is by necessity active in determining what the general interest of capital is. It must put some distance between itself and individual capitalists who often are competing with each other, and who give precedence to the realization of their own short-term interests. The capitalist state develops a series of structures or selective mechanisms that allow it to distil the general interest of the capitalist class out of the

Language Policy in Quebec / 25

particular interests of individual capitalists and to fashion policies designed to preserve the overall process of capital accumulation.9 Although the need to preserve the accumulation process acts as a constraint on policy formation, it by no means determines exactly what a policy will be. Przeworski cautions correctly that 'Appropriation of profit by capitalists is a necessary but not a sufficient condition for the realization of interests of any group. '10 He adds that in capitalist democracies the outcomes of conflicts over policies are to a degree uncertain and that these outcomes are not uniquely determined by class positions. This is not to say that class is irrelevant. The very organization of politics in democracies is such that it favours groups that have access to particular resources and skills. 11 Furthermore, the more corporatist a system is (that is36.74 pt, the more the organizations representing class interests are formally integrated into the structure of the state), the less likely it is that there will be open politicking and room for manoeuvre. Bargains may be struck within the state itself and outside the party system. None the less, the system in Canada and Quebec is not strongly corporatist and outcomes are not at all predetermined in this sense. 12 For our purposes it is not necessary to discuss in detail the interface between the democratic form of government and the capitalist economic system. Suffice it to say that in a democracy citizens of all classes have the opportunity to pursue their material interests within certain limits. 13 Within those limits, what determines whose interests are attended to? Przeworski argues that the manner in which conflicts over goods are organized will be the determining factor. 14 In a democracy, these conflicts will often be organized differently depending on the party in power. Different parties may have support from different components of society and may succeed in mobilizing those components more or in giving them access to particular resources useful in certain struggles over goods. Similarly, the ideologies that guide their leaders may predispose those leaders to select particular groups or demands for special attention. In short, parties will draw upon their own resources, bases of support, and ideologies to structure conflicts in capitalist democracies and thereby influence the policies that emerge. Further, the more distant the content of the policy area is from the process of accumulation itself, the more frequently such explicitly political factors might be brought into play. The relative autonomy of the state varies with the policy being considered. In most cases, it will be a complex task to unravel the relationships among policy area, distance from the accumulation process, and the impact of party and ideology. By using the study design outlined above - selecting two different parties and examining their treatment of two policy matters that clearly differ in their relative impact on accumulation - some of these complexities can be

26/W. D. Coleman controlled. Furthermore, by demonstrating variations in the impact of a political party and its ideology on a policy, some credence may be given to that much-abused concept in the political economy literature, the 'relative autonomy of the state. ' Language Policies in Quebec

The proposal that the provincial government in Quebec should become active in promoting the use of the French language in that province was not new in the late 1960s and early 1970s: it had been put forth by the 'father of Quebec nationalism,' Abbe Lionel Groulx, in the 1930s, most eloquently at the Second French Language Congress held in 1937. The ideas were kept alive by the various nationalist societies and finally became a plank in the electoral platform of the PLQ in the 1960 election.15 In 1961, the Liberals followed through by creating the first Office de la langue francaise, which was placed under the stewardship of Guy Fregault and the new Department of Cultural Affairs. This first office has been largely unnoticed by social scientists despite its pathbreaking work in the development of French technical vocabularies appropriate for use in Quebec's anglicized industrial world. Language use became an issue evoking widespread public debate and unprecendented emotion in the late 1960s. With the decline of Catholicism as an integral component of French Canadian culture, the French language became more and more the means used by French Canadians to distinguish themselves from other nationalities. In Quebec novels and poetry of the 1960s (particularly those of the Parti pris group) one finds intensive reflection on language and its impact on culture. The educational reforms begun in 1962 with Gerin-Lajoie's Grande Charte started to furnish graduates educated for jobs in the manufacturing sector of the economy, a sector that worked largely in English. An increasingly better-educated working class was finding promotion to higher positions open only to those who knew English. A falling birth rate among francophones meant that they could no longer rely on the revanche des berceaux strategy to keep their relative demographic strength in Quebec at a constant level in the face of large-scale non-francophone immigration. The cries of alarm became even more intense when it was learned that 90 per cent of immigrants were educating their children in the English. This seemingly sudden endangerment of the French language led to increased pressure on the provincial government to develop a global language policy. The need for francization and francophonization in the private sector on the one hand and for defining local public institutions as unilingual on the other were two important components in this demand for a global policy. Both needs were addressed in the legislation of the PLQ in 1974 and of the PQ in 1977.

Language Policy in Quebec / 27 Language Use in the Private Sector

It is by now well known that the French-speaking community in Quebec, which forms the overwhelming majority of the province's population, occupies an inferior position in the provincial economy. A study done by Andre Raynauld for the Royal Commission on Bilingualism and Biculturalism showed that French speakers controlled little of the economic activity in Quebec. 16 Subsequent studies done for the Commission of Inquiry on the Position of the French Language and on Language Rights in Quebec (Gendron Commission) demonstrated that French was rarely used at middle and upper management levels of most major corporations in the province. 17 Sales, in a more recent study, verified many of these findings. 18 This situation was unsatisfactory to the governments which since the Quiet Revolution had endorsed industrial capitalism wholeheartedly and had engineered educational reforms designed to graduate the cadres that would give francophones a stronger voice in Quebec's economy. In formulating their policies on language both the PLQ and the PQ sought to develop a program that would increase the use of French at the management level in businesses. The objectives of francization and francophonization were by definition closely related to capital accumulation. Any attempt to increase the use of French at executive and middle management levels of private enterprise in Quebec and to hasten the promotion of francophones to management positions in numbers reflective of their demographic strength might impinge on the capitalist's right to appropriate and invest profits following his own dictates. Increasing use of French might weaken the efficiency of an enterprise and thus lower its profits. It might also be costly enough to deprive employers of the return on their investment they felt they needed to stay in business. Similarly, accelerated hiring of francophones at higher levels of management might weaken a firm either because of reduced morale or, more important, by depriving employers of the right to choose the personnel they felt were most valuable to the firm in its search for profits. The PLQ in its Official Language Act developed policies that allowed it to move toward its objectives while maintaining the support of the employer class. 19 Ideologically, the PQ was predisposed toward pushing much harder in order to achieve francization and francophonization. This predisposition is documented in bill 1, the first version of the party's charter, a version that was drawn up after minimal consultation outside the party membership. When parliamentary hearings were held, bill 1 was roundly attacked by virtually every group representing employers in the province, both francophone and anglophone. The law that emerged after this pressure from employers very much resembled bill 22, the law devised by the PLQ. I shall argue below that the party

28/W. D. Coleman retreated in face of arguments that it had gone too far in its intervention into the free-enterprise system. The basic framework devised for francization by the PLQ after extensive consultation with the business community was retained by the PQ. 20 Under both parties' policies, agencies were created (the Regie de la langue francaise or RLF under the Liberals and the Office de la langue francaise or OLF under the PQ) to administer francization. Firms were asked to initiate contact with the agency and were then given a provisional francization certificate. They were then asked to carry out a detailed analysis of the language used in their operations. These operations were to be divided into organizational groups such as accounting and sales, and virtually all possible transactions in each group, written and oral, were to be assessed in terms of language usage. All relevant documents, manuals, and working memoranda were to be listed and classified according to language. The detailed forms worked out by the RLF were retained virtually to the letter by the OLF except for the additions required by differences in the actual provisions of the two laws.21 Despite these overall similarities in the process of francization under the two laws, there were important differences between the Official Language Act of the PLQ and bill 1, the first version of the PQ law, relating to the way in which firms were to enter into this process.22 These differences were successively muted in the first-reading and the third-reading versions of bill 101, the successors to bill 1. In general, where bill 22 attempted to persuade enterprises to francize, bill 1 sought to coerce them. In being more coercive, bill 1, in the eyes of representatives of the employer class, came close to depriving that class of its rights under the free-enterprise capitalist system. Several differences between the two laws are salient here. First, bill 22 sought to persuade firms to obtain a francization certificate by making such a certificate a requirement for any firm that wished to receive governmental subsidies or to enter into any contract of purchase, service, public works, or the like. In contrast, bill 1 made francization compulsory for all firms employing fifty employees or more workers. Article 106 provided that firms without a certificate would not have the right to receive from the government subsidies, premiums, or even permits. (Permits might include licences to operate a business, to use the facilities of Hydro Quebec, and the like. )23 Furthermore, bill 1 provided for a more rigorous regulation of francization. It proposed the creation of a Commission de surveillance that had the power to investigate suspected violations of the law and to submit suitable cases to the attorneygeneral for prosecution. While bill 22 also had an inquiry mechanism, it was less elaborate in structure and more informal in operation. Offending parties could be asked to appear before the Regie, and if they could not be persuaded to change, the Regie could inform the cabinet or its designated minister of the problem.

Language Policy in Quebec / 29

Both the provisions in article 106 and the powers of the Commission de surveillance sparked a strong protest from both anglophone and francophone employers.24 The Canadian Manufacturers' Association summarized these sentiments best when it stated, in reference to article 106, Timplantation d'une politique linguistique ne devrait pas se realiser au sacrifice d'une liberte aussi fondamentale que celle de pouvoir exploiter une entreprise. II ne s'agit pas ici de semantique, mais bien d'une liberte essentielle sous un regime de libre entreprise. '25 Representatives of this class were unanimous in calling for a scalingdown of the Commission de surveillance and for the institution of some appeal procedure. When bill 101 was presented a month later, article 106 had disappeared completely. Provision was made for an appeal procedure in cases where a francization certificate had been revoked.26 The Commission de surveillance was required to inform a firm of a violation of the law and to give it a certain amount of time to conform. If it did not do so, only then could the commission submit the case to the courts. A second difference in the laws was the manner in which francization was to be implemented in companies. Under bill 22, the francization process was to be supervised by the management of the enterprise. Bill 1, in contrast, directed that francization be carried out by a committee. One-third of the members of the committee were to be named by the workers in the firm. Again, the employers were unanimous in rejecting this arrangement. French and English groups alike argued that overseeing francization was the right of management. In its brief, CIL wrote: 'II devrait etre reconnu que ceux qui n'ont pas la responsabilite ultime pour les resultats de 1'entreprise ne peuvent pretendre a partager 1'autorite avec les cadres et 1'administration a qui cette responsabilite incombe. '27 In the first version of bill 101, no response was made to these plaints. In the version tabled for third reading, however, article 149 had been changed: the francization committee was now to report to the management of the firm, which would in turn deal with the OLF. Management controlled two-thirds of the committee and now dealt with the office on its own, and the situation that had existed under bill 22 had been virtually restored. A third difference that generated discontent among employers was the structure of the OLF itself. The RLF had been composed of nine members, of which only the president was a full-time functionary. The other eight members were appointed by the cabinet, and gave direct representation to the business community and others.28 In contrast, the OLF had a purely bureaucratic structure, and was headed by a president appointed for a ten-year period. Employers had no direct representation in the OLF. Instead, the law provided for an advisory council, which included business representatives. 29 The francophone representatives of the employers, in particular the Chambre de commerce de la Province de

30/W. D. Coleman

Quebec (CCPQ) and the Chambre de commerce du District de Montreal (CCDM), objected strenuously to this arrangement.30 They argued that employers needed a stronger voice in the operations of the OLF. Again, their arguments had some effect. Bill 101 changed the structure of the office so that it was composed of a president and four others, a structure again similar to that of the RLF. A fourth important distinction among the three bills divides bill 22 and bill 1 on the one hand and bill 101 on the other, and again indicates the willingness of the PQ to deal with the employer class. Article 29 of bill 22 had stated that francization programs should be set up 'while taking account of the situation and structure of each firm, of its head office and of its subsidiaries... ' Article 113 of bill 1 had used the same language, stating that francization programs were to take account of firms with head offices in the province supervising operations that extended beyond the province's borders. In the briefs of the CCDM and the Positive Action Committee (a group supported mainly by large corporations), the government was asked to define head offices broadly enough to embrace research centres and regional head offices. 31 While again the first version of bill 101 remained unchanged from bill 1, the version submitted for third reading contained a novel element. Instead of following the normal francization procedure, head offices were given the option of signing particular agreements with the OLF that would allow them to francize to a degree compatible with their own situations. The regulation issued by the OLF defined head offices broadly; research and development operations as well as normal head office functions were included. 32 To summarize, the francization program developed in conjunction with the employer class under the Liberal regime was substantially changed when the PQ submitted bill 1. That bill was more coercive in approach, potentially impinging on the process of accumulation of capital within the firm. It also reflected the social democratic tendencies of the PQ by bringing workers into what was perceived to be a sphere of management. After strong representations by all parts of the employer class, the legislation was revised and made similar to the previous law of the PLQ, bill 22. In addition, the PQ excluded head offices and research and development operations, both of which are fundamental centres of economic power, from the regular francization program. It seemed the party was willing to promote the use of French at the middle management level but drew back from pushing its use at higher levels; there, the initiative was left with the enterprise itself. With respect to francophonization, the process of placing francophones at management levels of firms, the PQ went no further than the PLQ had gone. In bil 22, francization programs were supposed to relate to 'the francophone presence in management. ' Bill 1 had defined as a goal of francization an increase in the

Language Policy in Quebec / 31

numbers of 'Quebecois' at all levels of employment. Bill 101 removed the political content from this goal by calling simply for an increase in the number of persons 'ayant une bonne connaissance de la langue francaise, ' thereby divorcing francophonization from the explicit promotion of persons of the French Canadian nationality only. A related aspect of francophonization follows the pattern of francization, namely a tendency for bill 101 to converge with bill 22. Under the terms of the Official Language Act of 1974, those groups falling under Quebec's Code of Professions were required to designate themselves in French and to use French when communicating with the provincial government. Under Bill 22 members of the professions were required to have a working knowledge of French in order to obtain a licence to practise in the province. Article 23, however, exempted professionals working exclusively for one employer and not coming into contact with the general public. This exemption was an important one because it meant that large national and multinational corporations could bring professionals to the province to work in research facilities and head offices in English without penalty. Bill 1 followed the policy set out in bill 22 but did not provide for this latter exception. This decision led to a strong protest from the banks, the Positive Action Committee, and the Canadian Jewish Congress. 33 Yet again the government relented - bill 101 reintroduced the special exemption for professionals found in bill 22. In conclusion, it seems clear that the laws in question tended to converge. Despite the ideological differences between the two parties sponsoring the laws - differences well illustrated by the comparison of bill 22 with bill 1 -the PQ felt a need to revise its policy to a point where it closely resembled bill 22, which had gained the approval of Quebec's employer class. The one difference that distinguished the francization program outlined in bill 101 from that in bill 22 was that the former was coercive rather than persuasive. In the end, however, as Allaire and Miller point out, this difference also made bill 101 universal and thus more economically fair. 34 Firms doing business with the government of Quebec who were forced to francize if they wished to continue to do so were no longer placed at a competitive disadvantage with respect to those who did not deal with the government. In their view bill 101 was, at least in the areas dealing with francization and francophonization, a less ambiguous bill 22. 35 Language Use in Local Institutions

Until recently most descriptions of Quebec emphasized that the province was composed of two societies. Both the French and the English communities were geographically and institutionally segregated. Both had full control over their

32/W. D. Coleman

own school systems and universities; both had their own municipalities, hospitals, charities, and social service agencies. To a degree, both had different museums, library systems, and artists. Beginning in the late 1960s, the separation of the two communities was increasingly challenged by nationalists to the point where the languages used in municipal governments, school-boards, hospitals, and various local social agencies became a political issue. The existence of these local institutions was seen as an additional magnet attracting immigrants to the English community rather than the French. The questions raised were less directly related to the accumulation of capital than were those arising from francization and francophonization. Conceivably the rules governing language use in local institutions might determine for some whether they would live in the province, and might indirectly affect the efficiency of certain enterprises who relied on outsiders to manage the firm. Still, the connections to accumulation are more tenuous; according to our initial proposition, the issue of language use here should lend itself to more variation deriving from differences in the social bases and ideologies of the two political parties. The respective ideologies of the PQ and the PLQ certainly looked at these local institutions in different ways. The basic demand of the PQ was that Quebec become a sovereign nation-state. Its platform was built around this demand and called for changes to strengthen the state of Quebec both before and after independence. The pequiste ideology separated Quebec from the rest of Canada and assumed that Quebec society had the capacity to thrive politically in its own right. The province of Quebec, of course, is a political entity and its borders do not coincide exactly with the cultural border between the French Canadian nationality and other groups. Accordingly, for the PQ, Quebec society was pluraliste, embracing several cultural groups. The two largest groups were the French Canadians and the anglophones of British descent. The former were viewed as the majority culture and the latter as the largest of the society's minorities. In general, each of these groups had the right to its own culture. This right was qualified, however, in order to take account of the perceived precariousness of the French Canadian nationality in anglophone North America. In order to ensure that no threat to the existence of this nationality arose in Quebec itself, the PQ developed a strong policy on the French language. There appear to be two rules underlying this policy. First, one set of institutions should be shared by all cultural groups in the province. In deference to the particular situation of the majority francophone community, these institutions should function in the French language. This I shall call the unilingualism rule. Second, all cultural groups in Quebec should be represented in these institutions in proportion to their demographic strength. This rule, the proportionality rule, implied that francophones had a right to redress in economic institutions, and

Language Policy in Quebec / 33

that non-francophones had a similar right in provincial public administration, areas where each group was respectively underrepresented. In contrast to the PQ, supporters of the PLQ believed that Quebec should remain a part of the federal political system in Canada. Quebec was a distinctive province because it was the patrie of one of Canada's two founding nationalities. At the same time the border between Quebec and Canada was viewed as a provincial not a national border, and members of the party saw themselves to be Canadians or Canadians and Quebecois. 36 According to the Liberals, Quebec society was pluralist, but it was also part of the larger Canadian society, which was also pluralist. The anglophone minority in the province was viewed as a part of the anglophone majority in Canada and deserving of full linguistic rights appropriate to that status. Accordingly, while the PLQ endorsed the proportionality rule, it was not particularly supportive of the unilingualism rule. Rather, the party followed the recommendations of the Gendron Commission and acted upon a 'French priority' rule. This rule stated that all members of the French majority in the province should experience no difficulty in living, working, and purchasing goods and services in French. If French and another language were in conflict in a jurisdiction, then French should prevail. As long as this rule was respected, those belonging to the English community should be able to use their own language in their daily lives. Symbolically, the PLQ indicated its support for Canada by agreeing that the provincial Assemble nationale and the courts should be officially bilingual. The issue of the language used in local institutions was important because it related to the degree to which Quebec's English-speaking community might control the institutions it had erected. The policy formulated by the PLQ in its Official Language Act allowed those institutions to remain in the control of the English community. Under article 6 of the act, municipalities and school-boards that served communities that were at least 10 per cent English-speaking were required to draw up official texts and documents in French and English. Where French was the language of internal communication in the public administration (OLA, 12), French and English were designated as languages of internal communication for those municipal bodies and school-boards in which most of the persons served were English-speaking. Where French had to be used by the public administration to communicate with other governments in Canada and with moral persons in Quebec, those municipal and school bodies serving a majority of English speakers could use either French or English. In the public service in general, hirings, promotions, and transfers were to be contingent on the applicant's knowledge of the official language, but this condition was waived for municipal and school bodies in which the majority of clients were anglophones. Contracts were to be drawn up in French or in French and

34/W. D. Coleman

English by local bodies. The law placed no restrictions on the language to be used in posters, notices, and the like. The Official Language Act of 1974 did provide that health and social service institutions were to ensure that their services were offered to the public in the official language. They were also required to use the official language when communicating with the public administration. No restrictions were put on internal communications, nor were they required to stop offering their services in English. Because of its ideology and its tendency to treat Quebec as separate from Canada, the PQ constructed a policy based closely on its view ofpluralisme and the unilingualism rule. Unlike the PLQ, which treated health and social services separately, the PQ combined them with municipal bodies and school-boards. 37 All local institutions were required to draw up and publish texts and documents in French. Municipal bodies were to use French as the language for internal communication. School-boards and health and social service institutions were to use French for internal communications unless the majority of the people they served were anglophones. In that case, they were to use both French and English.38 All local institutions were to use only the official language when communicating with other governments, moral persons in Quebec, or other departments and agencies of the government of Quebec. Under the terms of bill 101, all hirings, promotions, and transfers were to be made only if the candidate possessed an appropriate knowledge of the official language. The criteria and modalities for determining what knowledge was 'appropriate' had to be approved by the Office de la langue francaise. All contracts drawn up within the province were to be in French. All notices, posters, bills, and the like were to be in the official language except for those produced by local institutions serving for the most part those who spoke 'une langue autre que francais. ' In that case, printed material could be in French and another language providing that French was more highly visible than the other language. All local governmental institutions were to develop a francization program similar to that required of private businesses. They were to apply for a provisional francization certificate, carry out an analysis of language use inside the institution, and devise a procedure to francize their operations. In short, under the terms of the Charte de la langue francaise, local institutions serving non-francophone communities were to gradually make themselves compatible linguistically with the institutions of the majority. If the provisions of the law were adhered to, no francophone should find difficulty in working in French in these institutions or in being served in French by them. The formerly separate system of local institutions serving Quebec's anglophone community was to be merged into the larger pluraliste whole.

Language Policy in Quebec / 35

In conclusion, the differences in the treatment of language use in local institutions by the two parties were quite consistent with their respective ideologies. These differences were evident from the outset with the publication of bill 1 and, if anything, widened in the first and the final versions of bill 101. 39 This point becomes interesting when it is noted that the response of the employer class to the policy of the PQ on local institutions was no different from its response to francization and francophonization. In both cases, the class was virtually unanimously opposed to the policy. The CCDM and the Centre des dirigeants d'entreprise (CDE) asked for a return to bill 22 and the exceptions for local institutions serving the English-speaking community. The Montreal Board of Trade and the Conseil du patronat felt that documents issued by the government at whatever level should be in both English and French. The Positive Action Committee, the Montreal Stock Exchange, the Canadian Jewish Congress, and the Federation des groupes ethniques du Quebec all spoke against the tendency to change the separate status of local anglophone bodies.40 Arguments based on sustaining the accumulation process were not brought forward to justify these positions; appeals were made to history, to tradition, or to individual rights. Such appeals failed to sway the leadership of the PQ, and its policy on language in local institutions was not amended significantly. It certainly was not adjusted to converge with the policy of the PLQ, a policy that had enjoyed the favour of the employer class. The policy sponsored by the PQ may be seen to have been in the economic interests of some of those who had come to support the party. In the short run, the policy opened the way for employment of francophones who were trained for work in the public sector. In the longer run, it ensured that francophones would have access to more institutions of Quebec society in their search for employment. The policy also meant that individuals working in other public bureaucracies in the province, individuals who were predominantly francophone and who were strong supporters of the PQ, would not need to use a language other than French in their dealings with these local institutions. Finally, by attempting to dissolve some of the barriers around the local institutions of Quebec's English-speaking community, the policy was seen as a challenge to the aloof and privileged position of the English-speaking community in the province. Conclusion This essay has sought to, examine the possible roles in policy-making of the particular political party to which decision-makers belong and of the ideology

36/W. D. Coleman

that party espouses. Questions about whether such factors have any role at all arise from the consideration of the performance of parties of the left once they become governing parties. It would appear that in many cases, despite obvious differences in social bases and ideologies, the policies designed by these parties once they are in office are hardly distinguishable from those of their more conservative opponents. One possible conclusion that might be drawn from this observation is that the constraints upon the state in the capitalist mode of production are sufficiently strong to negate differences in party and in ideology of policy-makers. Such a conclusion, however, would be premature if not naive. While the capitalist state is dependent on the reproduction of the process of capital accumulation, when that state is organized in a democratic form the degree to which particular material interests will be attended to is not determined. Operating within the constraint of generating profits for capitalists, the outcome of conflicts over the distribution of goods will depend very much on how the conflicts are organized. The less that conflicts are likely to impinge upon the process of accumulation, the greater the likelihood that differences in party and ideology will affect conflict organization and outcomes. These theses, which have been drawn largely from the work of Przeworski, have been confirmed in the analysis of language policy in Quebec. Under the particular conditions that prevailed in Quebec between 1974 and 1977, two parties developed policies in two areas. The first area, francization and francophonization in the private sector, involved several aspects that touched the rights of capitalists to realize profits and dispose of surplus. It was shown that despite obvious differences in ideology between leaders of the two parties (differences made clear by a comparison of bill 22 and bill 1) the policies decided upon by the parties were quite similar. The second area, language use in local public institutions, was more distant from the activity of accumulating capital. In that instance differences in the ideologies of the two parties had a clear impact upon the form of the policy that emerged. One might say that the state enjoyed more 'autonomy' in this instance than in the previous one. Before ending this discussion, we should consider another possible explanation for convergence in policy development. It might be argued that the convergence between the PQ and the PLQ on the language of business is a logical result of the working of a two-party system in Quebec. Downs has argued, following Hotelling, that two parties competing for the same electoral space are likely to change their programs in order to resemble each other.41 Sartori has refined Downs somewhat, arguing that convergence is likely if four conditions are satisfied: undecided or floating voters are moderates; one party is not predominant; the two parties compete in the same 'space' (which can be ranged along an

Language Policy in Quebec 737

Figure 1. Schema for the Quebec party system

ideological dimension); and one party at least is able to win a plurality. Since the 1970 provincial election and the entry of the Parti quebecois, the party system in Quebec has been moving more and more toward satisfaction of these conditions. Figure 1 describes the party system in 1977. Group A, which includes public-sector professionals and students, is solidly supportive of the PQ. Group C, which includes the francophone and anglophone business class and the nonfrancophone cultural communities, strongly supports the PLQ. Group B, which includes the francophone middle class and the francophone working class, is not firmly attached to either party and is the main object of party competition. Given this somewhat simplified understanding of Quebec's two-party system, one would expect the PQ and the PLQ to converge on policies when they are seeking to win over group B. Can the convergence on the language of business be seen in this light? The answer would appear to be negative for the following reasons. First, the changes from bill 1 to bill 101 were not changes that primarily benefited the classes in group B. The changes made the process of francization easier for the business class by softening the compulsory aspects of bill 1, and by creating exemptions for key personnel (head office staff and non-client-serving professionals). These changes narrowed the avenues of advancement for the

38/W. D. Coleman

francophone middle classes. Further, they weakened considerably the role played by workers in the francization process. In fact, all representatives from the francophone middle classes and all the trade union congresses had supported bill 1 as it was originally presented. The changes in bill 1 may have threatened the support enjoyed by the PQ from group B. One might counter that the PQ changed bill 1 in order to try to expand group B by bringing into it at least the francophone business class. The PQ might have felt its base in group B was secure enough to make such an overture. It is evident that the changes made in bill 101 appealed to the business class in general and particularly to francophones. Yet, if this were the case, why did the PQ not alter as well its policy on language use in local institutions? The francophone business class was as adamantly opposed to this policy as it was to the language of business policy. If the PQ were seeking to win over this segment of the business class, might it not have made concessions here as well? One might respond that the policy here was probably less important to the francophone business class because it was directed at institutions that served other groups. This response is not convincing, since the francization and francophonization programs were not primarily directed at the francophone business class; rather, they were created to transform language usage in corporations owned by non-francophones. Instead of focusing on direct benefits from the legislation, one might use another approach and see the language policy as mainly symbolic. The changes in bill 1 might be interpreted as being directed at group B in order to counter the negative image of the bill that had resulted from the opposition of business groups and non-francophone communities. In showing flexibility, the PQ might have been trying to solidify or expand its support from group B. (It is well known that direct attacks on non-francophone communities are not popular among francophones. ) If this were the case, however, one might have expected that the changes made to bill 1 would have been widely publicized. This did not happen. Bill 101 was presented early in July 1977, after the publicity from bill 1 had died down. Little attention was paid by the government or the media to the changes introduced. McRoberts and Posgate in their excellent analysis of bill 101 do not even mention the differences between bill 1 and bill 101.43 In short, there does not appear to be a strong basis for using the logic of partysystem functioning to explain the language policies compared here. The impact of the language-of-business provisions on the usual rights of capitalists to direct their enterprises remains a relevant factor in explaining the convergence of the two parties in this policy area. In making this case, the comparision with the policy on language use in local institutions has clearly been instrumental.44 The evidence presented in this article is drawn from a narrow context. The method used maximizes similarities in the systems compared by focusing on one

Language Policy in Quebec / 39 society at two different times. None the less, this strategy permits specific examination of the roles of party and ideology with minimal worries about confounding factors. The support given to my original proposition on the conditions under which political parties will have more influence on the shape of policy has been encouraging. I recommend that further research inspired by the theories discussed in this article be carried out. NOTES 1 Adam Przeworski, 'Social Democracy as a Historical Phenomenon, ' New Left Review 122(1980), 51. 2 Richard Hamilton and Maurice Pinard, 'The Bases of Parti Quebecois Support in Recent Quebec Elections, ' Canadian Journal of Political Science 9, no 1 (1976), 16. 3 Pierre Fournier, The Quebec Establishment (Montreal: Black Rose Books 1975) 45-6. 4 Figures were reported in Vera Murray, Le Parti quebecois: de lafondation a la prise dupouvoir (Montreal: Hurtubise HMH, 1976), 33-4. 5 Adam Przeworski, 'Material Bases of Consent: Economics and Politics in a Hegemonic System,' Political Power and Social Theory (forthcoming), ms. 8. 6 Claus Offe, 'The Theory of the Capitalist State and the Problem of Policy Formation' in Stress and Contradiction in Modern Capitalism: Public Policy and the Theory of the State, ed. Leon Lindberg, Robert Alford, Colin Crouch, and Claus Offe (Lexington, Mass.: Lexington Books 1975), 126 (emphasis in original). 7 Claus Offe and Volker Ronge, Theses on the Theory of the State, ' New German Critique (1975), 139. 8 Offe, 'Policy Formation, ' 126. 9 For this particular view of the capitalist state, see Claus Offe, 'Class Rule and the Political System. On the Selectiveness of Political Institutions, ' German Political Studies 1(1974), 31-57. 10 Przeworski, 'Material Bases of Consent, ' 10. 11 This is discussed in Claus Offe, 'Political Authority and Class Structures: An Analysis of Late Capitalist Societies, ' International Journal of Sociology 2 (1974). 12 For a preliminary historical discussion of corporatism in Canada, see Leo Panitch, 'Corporatism in Canada, ' Studies in Political Economy 1 (1979), 43-92. 13 These limits are elaborated upon in Przeworski, 'Material Bases of Consent, ' passim. 14 Ibid., 14. 15 For a discussion of the origins of language as a political issue in the period after 1945, see William D. Coleman, 'The Class Bases of Language Policy in Quebec, 1949-1975, ' Studies in Political Economy 3 (1980), 93-118.

40/W. D. Coleman 16 This study was subsequently published as Andre Raynauld, Lapropriete de entreprises an Quebec (Montreal: Les presses de 1'universite de Montreal 1974). 17 See Commission of Inquiry on the Position of the French Language and on Language Rights in Quebec, Report, vol. 1: 'Language of Work' (Quebec: L'Editeur officiel du Quebec 1973). 18 Arnaud Sales, La bourgeoisie industrielle au Quebec (Montreal: Les presses de 1'universite de Montreal 1979). 19 For documentation on this point, see Coleman, 'The Class Bases, ' 96-103. 20 For a description of his process of consultation under the RLF, see RLF, Rapport annuel 1976, app. 2. 21 This conclusion is based on a comparison of the form entitled Analyse linguistiqe, formulaire B, developed by the RLF, and Analyse linguistique de I'entreprise, published by the OLF. 22 References to the laws are the following: 'Official Language Act, ' SQ 1974, c. 6; Projet de loi no 1: Charte de la langue francaise au Quebec, published in Les dossiers du Devoir 3 (juin 1977), 22-35; 'Projet de loi no 101: Charte de la langue francaise, ' premiere lecture; and 'Charte de la langue francaise, ' SQ 1977, c. 5. 23 Although it is difficult to say for certain, this particular interpretation of article 106 appeared to catch Dr Camille Laurin, the sponsor of the bill, by surprise. See his reply to the brief of the Canadian Manufacturers' Association (Quebec Division) contained in Assemblee nationale du Quebec, Commission permanente de 1'education, des affaires culturelles, et des communications, Auditions des memoires sur le projet de loi no 1: La Charte de la langue francaise, 20 June 1977. 24 These can be found in the briefs of Le Conseil du Patronat du Quebec, La Chambre de Commerce du District de Montreal, La Chambre de Commerce de la Province de Quebec, Canadian Manufacturers' Association, Bank of Montreal, Montreal Board of Trade, Positive Action Committee, Royal Bank of Canada, Canadian Jewish Congress, Le Centre des dirigeants d'entreprise. All were submitted to the parliamentary committee that conducted hearings on bill 1. 25 CMA Brief to the parliamentary committee, 4. 26 For the modalities of this procedure, see Reglement regissant la Commission d'appel de francisation des entreprises, AC 465-79, La Gazette officielle du Quebec, 21 fevrier 1979, 1915-16; Reglement fixant les modalites d'un appel interjete aupres de la Commission d'appel de francisation des entreprises, AC 466-79, ibid., 21 fevrier 1979, 1917-19. 27 Brief of CIL Ltd. to the parliamentary committee, 6. 28 In 1976, the Regie included the director of communications from Quebec Cartier Mining Company and the vice-president-personnel of Pratt and Whitney Aircraft. 29 This body was called the Conseil consultantif de la langue francaise and was set

Language Policy in Quebec741

30 31

32

33 34

35 36

37

38 39

40 41 42 43

up to advise the minister on language issues and to receive information on the progress of francization from enterprises. See their briefs to the parliamentary committee. The Positive Action Committee had representatives from such firms as Celanese Canada, CIL, the Royal Bank of Canada, and Alcan. Its brief listed two lawyers and three academics as being active in the committee, but no other nonemployers. See Reglement de POffice de la langue francaise relatif a la definition de siege sociale et a la reconnaissance des sieges sociaux pouvant faire Pobjet d'ententes particulieres avec 1'office, AC 3646-78, La Gazette offtcielle, 30 novembre 1978, 7124-^. These objections may be found in the briefs of these groups submitted to the parliamentary committee studying bill 1. Yvan Allaire and Roger Miller, Canadian Business Response to the Legislation on Francization in the Workplace, Accent Quebec Series (Montreal: C.D. Howe Research Institute (1980), 37. Ibid., 38. A survey of voters in the Montreal area in 1970 showed that supporters of the PLQ saw themselves as mainly Canadian (54.4 per cent) or Canadian and Quebecois (43.4 per cent). In contrast only, 11 per cent of PQ supporters saw themselves as mainly Canadian, and 37.6 per cent saw themselves as mainly Quebecois. These figures are reported in Serge Carlos and Daniel Latouche, 'La composition de 1'electorate pequiste,' in Leprocesses electoral au Quebec, ed. Daniel Latouche, Guy Lord and Jean-Guy Vaillancourt (Montreal: Hurtubise HMH 1976), 201. this decision to combine under the same category municipal, school, and social institutions was taken by the PQ in the final version of the Charte. In bill 1 and bill 101 (first reading), the party treated social institutions separately, following the practice of the PLQ. This provision was in fact a concession. Originally in bill 1 such institutions were to use only French for internal communication. These changes are examined in William Coleman, 'From Bill 22 to Bill 101: The Politics of Language under the Parti Quebecois,' Canadian Journal of Political Science 14(1981). All of the positions described in this paragraph may be found in the briefs by these groups submitted to the parliamentary committee studying bill 1. Anthony Downs, An Economic Theory of Democracy (New York: Harper & Row 1957)114-22. Giovanni Sartori, Parties and Party Systems: A Framework for Analysis (Cambridge: Cambridge University Press 1976), 376. Kenneth McRoberts and Dale Posgate, Quebec: Social Change and Political Crisis, rev. ed. (Toronto: McClelland and Stewart 1980).

42/W.D. Coleman 44 One other possible hypothesis that might be explored has been suggested to me by an assessor of the manuscript: the PQ was more flexible about the francization proposals because it considered them less important. The party might even be ambivalent about proposals to make French the dominant language in the private sector since encouraging more francophones to seek managerial jobs in the private sector would undermine the support base of the PQ. Such an argument certainly is plausible in light of our breakdown of support for the two principal parties. Still, the original policy of the party in bill 1 was very tough for a party that is argued to be giving the policy less importance. Furthermore, the retreat made by the party hardly undermined creation of a francophone managerial class. Rather, it removed the state from directly legislating francophones into positions of senior management.

3/The Politics of Provincial Resource Policy MARSHA A. CHANDLER

The impact of political parties on public policy has been a long-standing controversy in political science. Does it matter which party is in office? Do the ideological characteristics of the party in power account for differences in policy-making? Although early empirical studies in the U.S. and Canada found little evidence of a relationship between party and policy,1 subsequent research has pointed to a number of serious methodological and conceptual flaws in the studies.2 As these problems have been dealt with and the political and policy variables have been refined, partisanship has been found to be of considerable utility in explaining policy choice. A growing body of comparative research shows that left-wing governments produce policies in keeping with their economic priorities. There is evidence that in contrast to more conservative governments, social democratic regimes are more likely to pursue macro-economic policies that counter unemployment rather than inflation, to expand the size of the tax sector, to provide more social welfare, and to reduce social inequality.3 Even though party may once again be considered a factor in some areas of policy-making, it has received little attention as a determinant of Canadian natural resource policy. Most explanations of provincial resource policy have relied on a convergence model, in which it is argued that regardless of party ideology, governments have behaved similarly in response to socio-economic conditions.4 The importance of resources as core elements in the economic development are the bases for provincial actions. Provincial policies are seen to be pragmatic efforts to provide conditions favourable to economic growth. Despite this prevailing view, there are important reasons to suspect that partisan political forces may be significant determinants of resource policy. Both the nature of the policy and the workings of the federal system provide a basis

44/M.A. Chandler for a linkage between party and resource policy. Although resource development questions may consist of a number of issues (including foreign ownership and control, federal-provincial relations, and environmental concerns), the fundamental issue rests on the relationship between the state and the private sector. The debate relates not only to the appropriate division of resource revenues between the public and private sectors but also to the extent of direct government involvement in resource management. Given these characteristics, natural resources may be hypothesized to be a policy area that divides left and right. Disagreement as to the role of the state is at the heart of the ideological cleavage between left and right: the left is associated with a greater willingness to create a large public sector and with intervention in the economy.5 In the case of natural resources there has been more opportunity to develop distinctive provincial policies than in the case of social welfare, another arguably ideological issue. Unlike social welfare, an area in which there has been pressure for uniform national standards, natural resource policy has not been subject to the homogenizing influence of federal cbnditional grants. Distinctive provincial resource policies are less likely to be obfuscated by inter-governmental arrangements. The object of this study of provincial resource policy is to compare the utility of a convergence model, which emphasizes the preponderance of environmental variables and tends to dismiss the effects of political system factors,6 with the party impact formulation, which focuses on the explanatory power of partisanship. Unfortunately, generalizing about resource policy is made difficult because there is no accepted yardstick for comparing policies. In order to deal with this problem, this essay first develops a framework for comparison through an examination of the policy instruments available to provincial governments. In the second section the models are assessed by evaluating the hypothesis that governments of the left adopt resource policies that differ from those of other governments. The Changing Resource Arena

The long history of state involvement in promoting the development of natural resources substantiates the fact that at the provincial as well as the federal level Canada has never been a laissez-faire state. Provincial mineral resource policies historically have been fairly straightforward. The overriding goal - to facilitate the exploitation of the resources - has always been considered essential to economic development. The exploitation and the development of resources have for the most part been left to the private sector, and the role of provincial governments has been to stimulate private activity through tax incentives, provision of ancillary services, infrastructure, and direct subsidies.

Provincial Resource Policy / 45

But natural resources have represented more than a vehicle of economic development. Through the tradition of crown lands, natural resources have been established as the property of the state with the benefits thereof belonging to the people.7 This collective element is an important component of provincial policy. It provides a basis for revenue collection beyond taxation of a profitable industry or recovery of the costs of infrastructure. As owner, the province has a standing right to the benefits of its resources. Although the bulk of resource policy-making takes place at the provincial level, as in most areas of public policy both Ottawa and the provinces have some basis for claiming authority in the field.8 The British North America Act provides for provincial ownership and control over natural resources in section 109 and parts of section 92, which gives the provinces powers of direct taxation (2), management and sale of public lands (5), local works and undertakings (10), property and civil rights (13), and all matters of a merely local or private nature (16). Provincial authority is based on more than ownership: since provinces are also permitted control over development, even where land and mineral rights have been alienated or given over to the private sector, the provincial state still has some basis for regulating resource development. Federal power in the resource area is derived mainly from the 'peace, order and good government' clause, authority over trade and commerce under section 91(2), any mode or system of taxation (91(3)), and subjects excepted from those assigned to the provinces (91(29)). Until 1970 provincial resource policies clearly reflected the primacy of the economic development objective: taxation rates were low and profits taxes and royalties were the main revenue instruments. Provincial policies resulted in the private sector retaining most of the proceeds of resource exploitation. Resources never constituted a significant source of provincial income; they rarely ever went above 10 per cent of the provinces' total revenues. The policies followed by provincial governments indicate little or no concern for the costs of their actions. Because it was believed the resources were inexhaustible, there was no concern for user costs and the related notion that each generation should compensate future generations for what would not be available to them. The benefits of economic development obscured the opportunity costs (or the forgone revenues) resulting from minimal royalties. Finally, the social costs were not understood; externalities such as pollution did not enter the decision-making calculus of provincial governments. The traditional arena that gave rise to these policies can best be characterized as narrow and clientele-based, a symbiotic relationship between the provincial government and private-sector interests.9 Organized along sectoral lines, government departments that dealt with mining interests did not have to compete among themselves; nor were there any significant challenges from outside (i.e.,

46/M.A. Chandler

non-producer) interests. Given this system of limited representation, it is easy to understand the pre-eminence of the economic development objective and the consequent favourable promotional policies. Garth Stevenson has aptly summed up the nature of this relationship between the public and private sectors: The role of the state in mineral resource management, while not lacking in importance, was largely entrusted to a small body of officials whose training and expertise were in the technical side of the mining industry, rather than in policies, economics or public administration. Between this group and the private sector no major difference of opinion existed concerning the proper objectives of governments in relation to mining or the means of arriving at these objectives.10

Once the western provinces were given jurisdiction over their resources by the BNA Act, 1930, federal-provincial relations in the resource arena were relatively free of conflict. From time to time, of course, issues did emerge.11 The proceedings of the Rowell-Sirois Commission provide evidence that resource taxation was a point of some debate.12 However, Premier Hepburn's strong criticism of federal actions in the resource sector was perhaps more indicative of the close relationship between the Ontario government and the mining industry than of the state of federal-provincial relations.13 The uses of federal taxing powers were generally consistent with the provinces' efforts to provide incentives for resource industry growth. By allowing the industry to deduct provincial royalties and taxes (from 1946 on), Ottawa permitted the bulk of any public resource revenues to remain with the provinces. The traditional arena of resource policy can be described as relatively harmonious, for prior to the 1970s neither the public-private split of resource revenues nor the federal-provincial division of the public share generated a great deal of conflict or any significant variations in provincial policy. In the decade that followed, both public-private and federal-provincial relationships became much more discordant. This change can be traced to two important shocks to the traditional provincial policy arena. The first was the large increase in resource prices - not just of oil and gas, but also of copper, potash, and other minerals. As prices rose, the economic rents associated with resource exploitation were no longer minimal. Obviously, revenue policies became more salient when there appeared to be large profits available for taxation. Perhaps most significantly, the increased value of resources led to the belief that mineral developers were reaping 'windfall profits.' In light of these seemingly large and undeserved profits, the notion that the land and its resources belonged to the people became a basis for action.14

Provincial Resource Policy 747

With increases in resource prices came the perception that minerals were both valuable and exhaustible. Once it was recognized that natural resources could not provide an unending stream of revenues, some efforts were made to control the rate of exploitation and to capture the present revenues in order to provide for the future, when resources would be depleted. There was increased popular support for the idea of the provinces gaining direct benefit from their resources. The provinces responded to price increases with altered revenue and management policies, which disturbed the close relationship between provincial government and producer interests. At the same time, the harmonious relationship between the federal and provincial governments were also changing radically. Provincial governments' awareness of their position as resource owners (and therefore of their entitlement to the economic rents) was heightened by federal policies to capture some of the rents. The second shock to the provincial political system was Ottawa's new, more active role in the resource field. The Royal Commission on Taxation (the Carter Commission) initially challenged the necessity for preferential tax treatment for the resource industry. It was unsuccessful because those provinces linked with the mining industry were able to bring about postponement or rejection of many of the proposed changes that would have increased the taxes on the industry and made its burden comparable to that of other sectors of the economy.15 However, rising prices and provincial efforts to capture the resource rents led to defensive action by the federal government. In the 1974 budget, Ottawa declared an immediate end to a number of tax advantages for the mining sector and, most significantly, held that provincial royalties and taxes would no longer be deductible. Resource taxation quickly became a zero-sum game. Ottawa's gains were the provinces' losses. The federal challenge, however, went beyond revenues to the provinces' right to manage their resources. Although that right seems straightforward, at some point management blurs into marketing. To the extent that provincial management affects extraprovincial trade, the province is acting beyond its authority and encroaching on federal territory. For example, in the Central Canada Potash case,17 Saskatchewan's prorationing scheme to manage potash production was declared ultra vires because it interfered with extraprovincial trade (and because it constituted an indirect tax).18 The federal government's policy aimed at promoting security of supply of fuels has also resulted in a challenge to provincial authority to control the rate of development of resources. In summary, the provincial policy arena that developed in the 1970s differed substantially from the traditional one that had persisted for nearly a century. First, a wedge came between the provincial government and the resource sector. No longer could it be assumed that public and private interests were identical. Second, resource issues became major points of controversy in federal-

48/M.A. Chandler

provincial relations. The federal government had become a key actor in the resource arena. Since 1970 the provinces with significant resource wealth have appeared to be seeking to increase the public share of resource revenues.19 All have broken away from their almost exclusive concern with promoting the industry. This uniform change of policy direction obscures a number of important differences among the provinces. A means of measuring and comparing provincial resource policies is now outlined. The instruments of policy Most systematic analyses of public policy have tended to operationalize the activities of the state in terms of expenditures. Absolute levels of spending have been used to indicate the extent of state involvement, and share-of-budget has been used to reflect government priorities. The problem is that expenditure measures are based on a limited notion of state involvement since they focus on essentially one instrument of policy - subsidies - and ignore other instruments such as taxation and regulation. 20 In those fields in which expenditures are the main policy instruments, budgetary analysis may be quite useful; in other areas, such as natural resources, expenditures are inadequate indicators of policy. For example, even though numerous incentives have been provided to the resource industry, these have come as much through the tax system as through direct and indirect subsidies. Minimal portions of provincial budgets have been spent on natural resources, and the level of spending has been relatively constant over time and across provinces. For example, in 1950 the provinces spent an average 6 per cent of their budgets on resources; and in 1975 the figure was 5 per cent. The dramatic policy changes that occurred in the 1970s cannot be identified by a survey of constant resource expenditures. If expenditures are inappropriate measures, what then is a good synoptic indicator of resource policy? Given the provinces' constitutional status as owners and managers of natural resources, indicators should reflect variations in both of these two policy functions. Measures based on the provinces' tax systems are one means of assessment.21 As owners, provincial governments have to deal with two often contradictory objectives - facilitating resource development and obtaining a public share of resource revenues. Provincial resource policies can be compared by the size of the revenues the province extracts from the industry. However, there are some difficulties in using tax burden as a measure of policy. It is almost impossible to arrive at a meaningful approximation of an average effective tax rate because the

Provincial Resource Policy / 49

rate varies greatly, depending on such factors as degree of processing, expenditures on exploration, and metal prices. This is equally true for gas and oil royalties, which depend on such factors as location, the age of the well, total production, and price.22 Rather than trying to estimate the tax burden associated with each instrument, resource revenues as a proportion of production value will be used as an indicator of the revenue collection capability of provincial policy. Resource taxation entails more than the appropriation of resource rents to the public sector; it also involves the provinces in resource management. For example, it is quite possible that a profits tax and a gross royalty may generate the same level of revenues while representing two different policy choices that have different effects on the industry. Although in general the provinces have left exploration and development to the private sector, most of the methods of collection are not neutral, since they affect the rate and level of exploration. In some cases the province has no intention of regulating;23 in others the provinces have sought explicitly to use the revenue system to affect allocation.24 The other important provincial policy-making role is that of. manager of a resource. There are a number of ways this role can be performed; revenueproducing instruments can be distinguished by the extent to which the state intervenes in the economy or the degree to which it acts as a manager of the resource sector. This dimension is implicit in the distinction posed by Norrie and MacMillan in their discussion of rentier versus province-building policy.25 Under the rentier strategy, the state views taxes from resource exploitation mainly as another source of revenue, and resource development decisions are left to market forces. The province-building strategy uses revenue collection deliberately to influence resource development and to shape the economy. The notion of policies varying by the degree of state involvement is implicit in a number of other analyses of policy-making. For example, Lowi and Doern and Wilson suggest policies may be differentiated by degree of coercion and that decision-making will generally move from less coercive to more coercive policies.26 In their study of the oil industry, Trebilcock et al. order revenue instruments by the share of risk borne by the state. They hypothesize that overtime, with greater experience with the industry, the state takes on more risk andis able to obtain more revenues.27 Throughout the 1970s a broad range of revenue instruments was adopted by the provinces. In addition to employing new methods, levels were increased on such existing mechanisms as royalties and mining profits taxes. Each instrument is described below in terms of its implications for public-sector involvement. Starting with profits tax and moving to gross royalties, super-royalties, and public ownership, the instruments can be ordered in terms of increasing government involvement in the private sector. Although the provinces can also capture

50/M.A. Chandler

rents through land and mineral rights acquisition payments and taxes on exploitation, I shall focus only on those revenues related to exploitation and development. Profits tax Prior to the 1970s the most popular tax on minerals was a net royalty or profits tax. Under this system a firm that makes no profits pays no tax. But if the costs of labour, capital, and other expenses are deducted and a profit remains, the excess is viewed as rent and is subject to taxation. If the normal return to capital has not been included on the costs side, then a profits tax may be based on more than rent. Thus, in order to separate rent it is necessary to approximate a normal return on capital. For example, in Ontario the Smith Committee on Taxation suggested a 15 per cent return to be deducted from income before applying a mineral profits tax. For firms that make a profit, a profits tax involves no interference with the industry - normal market incentives operate. For those firms that generate only normal returns without any excess or rent, a profits tax that does not provide for a return on capital is in fact a tax on the firm's capital. This tax on capital is peculiar to the mining industry.28 There is one situation in which the profits tax can become more intrusive. Because the tax is based on the exploitation and not on the processing of the mineral, it is often necessary to deduct processing costs to arrive at the taxable pit-head value of the mineral. Although the provinces have generally sought to estimate the real processing costs, variable processing allowances have also been used to discriminate in favour of domestic processing. For example, Ontario provided for much higher allowances (deductions) if the processing took place in Canada. Similarly, British Columbia permitted lower royalty rates for processing within the province. Gross royalties Royalties that take a fixed percentage of the value of production are the oldest mode of taxing the mineral industry. In modern times this has been the most popular means of obtaining revenues from gas and oil. As a tax on output, a simple royalty does not allow for differences in the quality of ore or the costs of production. However, distinctions between old and new oil and gas that allow for lower royalties on the new supplies, which are assumed to be more costly, reflect efforts to make the system more sensitive to variations in costs.29 A production royalty generates public revenue regardless of whether the firm makes a profit. For this reason it must be seen as a tax instrument that adds to the cost of production. The royalty represents more of an intrusion into the firm's decision-making than does a profits tax, which is imposed after costs have been deducted. Although it appears that a production royalty can easily be a tax

Provincial Resource Policy 751

on more than economic rent (especially for the marginal firm), the rationale for the royalty is that it is a user cost. The present exploiters are paying those who, in the future, will not have the use of that resource.30 Unlike the profits tax system, in which the state shares the risk with the private sector, royalties require the industry to carry more of the risk; royalties are therefore generally lower than profits taxes to compensate the industry for bearing that risk.31 Super-royalties Surcharges, super-royalties, or excess profit taxes represent direct efforts to capture economic rents. The super-royalty is a second tier of taxation specifically designed to tap the windfalls resulting from price increases. This scheme taxes at a very high rate the profits derived from price increases. For example, the Saskatchewan royalty surcharge taxed away all of the extra income generated by increases over the 1974 price of oil. Manitoba's Metallic Minerals Royalty Act (1975) taxed all mining profits above 18 per cent at the rate of 35 per cent, while profits below 18 per cent were taxed at the rate of 15 per cent. As in the case of gross royalties, super-royalties can be based on a split rate in which there is a higher levy on old oil (which is assumed to be lower in cost). The object is, of course, to make the taxes more attuned to rent and not to destroy incentives to produce higher-cost resources. Public ownership Until the 1970s, aside from the public hydro corporations there were no significant examples of public ownership in the resource sector. Today, state activities include a provincial trading agency (BC Petroleum Corporation), take-overs of existing firms (Potash Corporation of Saskatchewan), and joint ventures with the private sector (Syncrude). In each case the province shares directly in the risks and benefits of exploitation. As owner and developer, it can collect not only the rents but also the normal return on capital. Of all the revenue-producing instruments, public enterprise involves the greatest involvement in the economy; however, not all forms of public ownership are the same.32 Provincial governments have used public enterprise to intervene in the economy in order to foster economic development, to redistribute economic benefits and power, and to further nationalism or provincialism. Each objective entails a particular relationship between the private and public sectors and a distinctive state involvement in the economy. Public enterprises designed to foster economic growth are not necessarily challenges to the private sector. On the contrary, they use public funds to supplement and support the private sector.33 Redistributive crown corporations involve direct intervention in the economy to change some of the consequences of the market. The Potash Corporation of Saskatchewan is a redistributive crown corporation whose purpose is to extend

52/M.A. Chandler

the state's control over the economy. A public enterprise established to realize nationalistic objectives is used to enhance provincial interests at the expense of interests outside the province. Nationalistic enterprises are set up to assist indigenous economic interests by promoting economic growth, although not growth at any cost. It is a complex and restricted kind of incentive system and represents a greater interference by the state in the economy than the corporations designed simply to foster private-sector development. Nationalistic state enterprise has been especially important in Quebec. Enterprises such as SOQUIP (Societe quebecoise d'initiatives petrolieres) and SOQUEM (Societe quebecoise d'exploration miniere) have been initiated with the view of developing the Quebec economy by and for Quebecois. Nationalism has also been an important basis for public enterprise in western Canada. Certainly much activity in Alberta has centred on replacing eastern dominance with control by Albertans.34 Because public ownership may fulfil more than one function, it is important to draw distinctions within the broad category of state enterprise. We know that governments of diverse ideologies have all used crown corporations; the question is whether they used them for diverse objectives. Recognizing that there are variations within each instrument does not detract from the value of ranking the four instruments by degree of state intervention in the economy. The least intrusive is profits tax, followed by gross royalties, super-royalties, and crown corporations. (As we have noted, types of crown corporations could also be ranked by degree of state intervention, but there is little doubt that in general crown corporations represent greater public involvement than the three taxing instruments.) Taken together, the revenue and degree of involvement reflect the provinces' roles as owners and managers of natural resources. Using these indicators of resource policy, it is now possible to look for variation in provincial policies and to consider the utility of party politics in explaining policy differences. Parties and Policy

The proposition that variations in policy outputs may be explained by differences in the parties in power is predicated on the ideological distinctiveness among parties and differences in each party's constituent groups. Even when parties fail to articulate different policies, if their support comes from different segments of the population it is assumed that their actions as governments in power will reflect the differences in their support. Party ideology Provincial parties may be ideologically distinguished on a left-right continuum. The left connotes state intervention in the economy; the right is associated with

Provincial Resource Policy 753

the principle of free enterprise unfettered by public control. The left favours a positive role for the state; the right leaves the economy to be shaped by market forces. Their commitment to goals of economic and political equality makes social democratic parties critical of the distributive consequences of the market.35 They are willing to intervene and to use the power of the state to redistribute wealth and achieve social justice..Linking economic and political resources, those on the left argue that the deconcentration of economic power is necessary for the decentralization of political power. For the purposes of this discussion, the 'left' is composed of the New Democratic Party (NDP) and its predecessor the Co-operative Commonwealth Federation (CCF) and the Parti quebecois (PQ). The NDP has been consistently ranked to the left of other parties, and it is generally agreed that the PQ is a social democratic party that occupies the left in the spectrum of Quebec politics.36 Unlike most social democratic parties, including the NDP, the PQ has no formal links with the trade-union movement. Otherwise, the similarity of their positions on economic and social issues is well established.37 The category of 'non-left' is composed of other parties that compete in provincial elections (Liberal, Progressive Conservative, Union Nationale, and Social Credit). Although there are significant differences among provincial parties bearing the same label, on economic issues all have shared a commitment to a market economy and have accepted the perpetuation of an economic order based on private property ownership as a goal in and of itself. The question then is whether the policy outputs of governments reflect their ideological distinctiveness. If there is a causal relationship between party ideology and resource policy, then we can expect social democratic governments to appropriate for the public sector a larger share of resource revenues and to use those revenue instruments involving more interference in the private sector. Provincial policy-making This analysis focuses on the provinces that at some point had a left government -British Columbia, Saskatchewan, Manitoba, and Quebec-and the provinces that are major contributors to natural resource production - Alberta and Ontario. The policies of each of these six provinces are examined separately and policy variations corresponding to changes of party in power are noted. The policies of left governments from all provinces will be aggregated and compared with the aggregated policies of the non-left. This allows a comparison of how the two kinds of governments handle resource revenues and management. British Columbia The mining sector in British Columbia is a highly concentrated industry that exports primarily unprocessed minerals. The industry, second in importance

54/M.A. Chandler only to forestry in the provincial economy, has long enjoyed a clientele relationship with the Department of Mines. Although W.A.C. Bennett's government (1952-72) was interested in resource revenues and raised the mining profits tax from 4 per cent to 10 per cent in 1954 and from 10 per cent to 15 per cent in 1968, the Social Credit policies very much conformed to those of the other provinces. Revenues were not the primary consideration; the overriding objective was to stimulate economic development by encouraging and assisting private producers. When the NDP came to power in 1972, the mining industry was an obvious target. Large profits were being made by a few multinational firms exporting British Columbia's raw materials. The NDP government tried to move on three fronts: it sought a larger share of corporate resource revenues; its concern with economic development was manifested by efforts to encourage more domestic processing of BC minerals; and it sought more public control over the development of the mineral sector. In order to increase mineral revenues, Premier Barrett's government passed bill 31, which became the Mineral Royalties Act (1974).38 The act added to the existing 15 per cent mining tax a gross royalty of 5 percent of production value and an incremental royalty or super-royalty. The super-royalty provided for a 50 per cent tax when the price of a mineral was 120 per cent of the basic value (a price set by the minister based on the average price and adjusted to reflect increased costs). The purpose of the super-royalty was to give the state one-half of the 'windfall' gains derived from government increases in mineral prices. In order to facilitate further processing, bill 31 provided for a reduced gross royalty if a mineral was smelted or refined in the province. Revenues and domestic processing were not the only objectives of the government. An equally important (and in the eyes of the industry, an equally objectionable) element in the NDP's approach was section 64 of the Mineral Act, which required ministerial approval of industry production plans. As Payne points out, the 1973 amendments to the Mineral Act39 'centralized and focused the administration of most aspects of mining regulation in the hands of the mines ministry and gave the department increased powers in dealing with private developers.'40 The NDP was not only seeking to exercise its power as owner; it was also seeking a management role.41 Although the Cabinet Resource Committee did recommend that the government seek an equity share in new mining projects, the NDP never took any major initiatives in public ownership of minerals, the British Columbia Petroleum Corporation was established in 1974 to act as a monopsonistic trading agency. It buys all the natural gas produced in the province and acts as the sole agent of sale for export. Although the difference between purchase and sale prices provides

Provincial Resource Policy 755 TABLE 1 Provincial resource* revenuesf as a percentage of production value

British Columbia Alberta Saskatchewan Manitoba Ontario Quebec

1945

1950

1955

1960

1965

1970

1975

1977

4.5 6.0 8.0 4.8 2.3 4.8

2.0 36.0 7.4 3.7 2.0 2.8

8.1 41.1 12.9 3.4 2.3 3.1

11.6 28.3 7.6 3.6 2.2 2.6

14.7 33.8 10.8 2.0 2.3 1.9

10.8 16.9 8.3 1.6 1.7 2.9

25.4 30.9 35.7 3.1 2.8 3.7

17.3 37.8 30.6 2.6 0.9 2.6

SOURCE: Revenues - Provincial Government Finances Statistics Canada 68-207. Production General Review of the Mineral Industry Statistics Canada 26-201 * Includes only fuel and mining resources. t Does not include revenues from corporate income tax. TABLE 2 Revenue instruments used by provincial governments (since 1945) Profits tax British Columbia left non-left Alberta left non-left Saskatchewan left non-left Manitoba left non-left Ontario left non-left Quebec left non-left

X XX

Royalty X

Superroyalty

Public ownership

X

X

XX

XX

X

X

XX

XX

X XX X

X

X

XX

XX X

X

XX

XX

significant public revenue and has eliminated the need for royalties on gas, the main reason for the corporation's establishment was to increase the export price of BC gas.

56 /M.A. Chandler

On its return to power in December 1975, the Social Credit party made a number of changes to these policies. Royalties and super-royalties were replaced by a return to a slightly increased profits tax (increased from 15 per cent to 17'/2 per cent). Given the many deductions available, the effective rate was significantly lower than the stated amount. The 1977 amendments to the Mineral Act removed many of the regulations that had permitted the minister of mines to intervene in private-sector decision-making. There can be, no doubt that the Social Credit government consciously stepped back from an active management role. It did establish another state enterprise, the British Columbia Resources Investment Corporation (BCRic), in 1978. BCRIC'S stated purpose, however, was to manage and oversee the eventual dispersal of assets acquired by the previous NDP government. In 1979, with the province retaining 20 per cent ownership, five shares of BCRIC were given to every resident of British Columbia. Certainly in British Columbia there is evidence that an NDP government acted quite differently from a Social Credit administration. Tables 1 and 2 illustrate these differences: in terms of the size of the public share and the instruments employed, the NDP acted as a more interventionist government. Manitoba Until 1930, Manitoba's resources were under federal jurisdiction. From the transfer of jurisdiction in 1930 until 1969, successive Manitoba governments acted similarly to those in other provinces. Continued emphasis was placed on promoting private-sector exploration of the province's mineral wealth. Any and all initiatives in exploration and development came from the private sector. The province used a progressive mining profits tax that ranged from 6 per cent to 11 per cent. There is no evidence of attempts on the part of the province to play any managerial role. The concentrated mining sector maintained a close clientele relationship with the Department of Mines. When the NDP formed a government in 1969, it undertook a review of the mining industry. The mining projects tax was raised to 15 per cent, the level in most of the other provinces. In 1971 the legislature passed bill 17, which established Manitoba Mineral Resources Ltd. This crown corporation was created as a vehicle for provincial participation in mining ventures. Premier Schreyer's government also commissioned Eric Kierans to report on the proper role of the provincial government in managing valued resources. Produced in 1973, the Kierans Report made the case for greater provincial intervention in the resource sector. Most important, Kierans argued that the state must be ready to fill in the slack from any downturn in private exploitation resulting from increased taxation. Subsequent legislation increased the province's share of resource revenues and permitted provincial participation in mining

Provincial Resource Policy 757

activity. The Metallic Minerals Royalty Act (1975) provided for a two-tier system of a basic royalty and an incremental royalty, or super-royalty. Based on the size of the investment, profits under 18 per cent were taxed at 15 per cent, and profits over 18 per cent were taxed at 35 per cent. The Mineral Disposition Regulation (328 / 74), inter alia, gave the province the option of 50 per cent participation in new resource ventures, and required companies to obtain production leases before commencing production. An Explorations Operations Branch (EOB) was created in the Department of Mines, Resources and Environmental Management to administer mineral policy.42 The intention was to encourage private-sector development as well as to 'take up any exploration slack resulting from private sector decisions.' Manitoba Mineral Resources Ltd and the EOB gave the government opportunities to become involved in resource development either directly or by influencing private-sector decisions. Tables 1 and 2 show the increased revenue levels and involvement in the private sector. When the Progressive Conservative party led by Sterling Lyon came to power in 1977, the two-tiered royalty system was abolished and replaced by an 18 per cent profits tax. The Lyon government also took steps to withdraw from a managerial role in the resource industry. The administrative arm of the EOB was dismantled and requirements that industry submit exploration plans were abolished. No new participation agreements were entered into. Saskatchewan Saskatchewan has the distinction of having been governed by the left for a greater length of time than any other province. It also has the most extensive public involvement in the resource sector. Most of this involvement did not arise until the 1970s. When jurisdiction over mineral resources was transferred in 1930, Saskatchewan tried to encourage resource exploitation. In 1944, when the CCF came to power, the government's immediate concern as resource owner was not so much to raise revenues but rather to regain control over its mineral resources. Under the Mineral Taxation Act of 1944 many of the alienated mineral rights reverted to the crown. The policies promulgated by Premier Douglas's government were like those of the other provinces.43 Royalties and taxes were low, policy was directed toward encouraging development, and a close relationship existed between the Mineral Resources Department and private industry. The highly concentrated and predominantly foreign-owned producers had the technology, expertise, and capital so desperately needed by the province.44 The CCF government was in no better position to risk driving the companies out than were the more conservative governments of Duplessis, Manning, or Bennett. There can be little doubt that this dependence effectively

58/M. A. Chandler constrained any ideological predilections toward greater state involvement in the economy. The Liberal government elected in 1964 continued the promotional policies of the CCF. Like the CCF, it failed to assume a real managerial role, it sought outside private capital, and it kept resource taxes to a minimum. Pro-rationing, introduced in 1969 at the request of the producers, meant regulations on price, volume, and marketing of potash. These regulations served to shield the private producers from dropping prices because of an international glut of potash. When the NDP returned to power in 1971, the potash industry was flourishing under the pro-rationing regulations. The government tried to increase the public benefits, but the industry refused to renegotiate the royalty arrangements, which were to run for another ten years. The pro-rationing fee imposed in 1972 and the reserve tax imposed in 1974 were government efforts to obtain greater revenues from the potash industry. The boom in oil prices, like that in potash, was met by provincial efforts to secure for the public sector a large part of the windfall gains. The Oil and Gas Conservation, Stablization and Development Act represented an attempt to collect these rents. The act created a royalty surcharge that collected 100 per cent of the difference between prevailing prices and a basic well-head price, which corresponded to 1973 prices. This tax was found by the Supreme Court of Canada to be ultra vires because it was ruled an indirect tax and it interfered with interprovincial and international trade.45 Today the royalties on crown lands are based on old and new oil, an area adjustment factor, and production. Based on its oil royalties and taxes, uranium production royalty and profits tax, and potash taxes based on profits and production, Saskatchewan's share of resource revenues is one of the largest of all the provinces. Although the NDP has clearly opted to raise the provincial level of resource revenues through the tax system, it is through direct participation that its policy-making has been unique. Saskoil (1973), the Saskatchewan Mining Development Corporation (1975), and the Potash Corporation of Saskatchewan (1975) are crown corporations established to raise revenue and to increase provincial control over resources.46 Each enterprise is a vehicle for direct government involvement in the development of Saskatchewan's resource wealth. In addition, the compulsory joint venture program permits the province to participate to the extent of 50 per cent ownership in mining projects. Quebec Most of Quebec's resource policies have been directed toward promoting the mining industry. As in other provinces, the industry is predominantly foreignowned and based primarily on exporting unprocessed minerals. Until 1960 there

Provincial Resource Policy / 59

was very little benefit to the province from the exploitation of its resources. As owner, the province derived limited revenues from a mining profits tax which ranged from 4 per cent to 7 per cent. Quebec's resource revenues fell below even the minimal revenues of other provinces. Furthermore, the province played no managerial role. Policy during the Duplessis period consisted of a number of ad hoc arrangements to promote private-sector investment. The Department of Natural Resources, established in 1936, performed a major service role for its clientele in the industry. The Quiet Revolution marked a transition in many policy areas. In the field of natural resources, it brought about a close examination of the relationship between the resource industry and the state. Although the promotional objective was by no means rejected, several other goals received consideration. These reflected the general concerns of the Quiet Revolution for greater government control over the economy and greater participation by Quebeckers. Two important policy changes emerged. The first was structural: in 1961 there was a reorganization of the Department of Mines and within it the creation of a planning bureau and a committee for promoting mining activity. This provided the organizational capability for greater state involvement in the industry. The department was thus equipped to carry on much more than its previous service role. The second important change that occurred in the 1960s was the use of public enterprise as an instrument of policy. For example, in 1968 Siderurgie Quebecoise (SIDBEC) was created in an effort to involve Quebeckers in the ownership and management of an integrated iron and steel complex. Similarly, SOQUEM was established to carry out and encourage mining exploration and to participate in mineral development. The use of state enterprise was not meant to supplant the private sector; rather, it was to substitute for other instruments in promoting development, but in keeping with the nationalistic objectives of the government.47 Despite many public statements claiming that the public had a right to the benefits of its resources, little was done in the 1960s to collect a larger share of resource revenues. The tax changes introduced in 1965 brought Quebec in line with the favourable policies of the other provinces. The profits tax rate increased to a range of from 9 per cent to 15 per cent. A number of generous allowances brought the effective tax rates far below the declared levels. In the 1970s Quebec, like the other provinces, reacted to the growing value of its minerals and to Ottawa's efforts to increase its own resource revenues. Bill 89, which became the Mining Duties Act, raised the profits tax to a minimum of 15 per cent for profits above $3 million and a maximum of 30 per cent for profits exceeding $20 million. But, as in 1965, numerous allowances kept the effective rate well below those figures. Table 1 shows that there was little change in

60/M. A. Chandler

Quebec's share of resource revenues. What the figures do not reflect is that the 1960s did represent a break with the past; the new policies focused on increasing social gains from the resource sector by stimulating more integration of mining into the economy, by facilitating increased processing, and by greater participation by Quebeckers. The Parti quebecois, elected to office in 1976, continued the state's positive role in resource development begun by Rene Levesque's ministry in the Lesage government. The PQ's most visible action was the creation of the Societe nationale de 1'amiante in an effort to take over the asbestos industry. Since 1960, the Liberals, the Union Nationale, and the PQ have all used public ownership to further provincial autonomy. Unlike its predecessors, the PQ has tightened government control over the state corporations.48 There has been much more use of public enterprise specifically to carry out government policy. Given the degree of foreign ownership and the extent of federal power over the economy, much of the left's concerns with the skewed distribution of wealth and the concentration of economic power is directed toward external forces. Ontario Since 1907, Ontario's mineral resource revenues have depended on a mining profits tax. Ontario's consistent pattern of resource taxation is summed up by Ross Hynes: Both the form of Ontario's mining tax, as a levy on profits, and its basic justification, as the public's share in the 'Bounty of Nature' remain in effect today. Its execution, however, has gone through a considerable transition in recent years. Indeed in only the past seven years has the Bounty of Nature begun to resemble anything more than a hollow promise. For this objective has been overshadowed during the greatest part of its existence by the government's desire to promote the exploitation of the province's mineral resources.49

The revenues shown in table 1 indicate the favourable treatment of the industry. Until the 1970s Ontario was similar to the other provinces in its dependence on low profits taxes and extensive deductions. In 1946, when Ottawa made provincial taxes deductible, Ontario raised its profits tax to range from 6 per cent to 9 per cent. In 1957 the maximum was increased to 12 per cent, and generous exemptions were introduced. The recommendation of the Ontario Committee on Taxation (the Smith Committee) to change the tax to a flat 15 per cent on profits over $50,000 was implemented in 1969. In 1974, a steeply graduated profits tax was enacted. For profits over $100,000 the rate was to be 15 per cent; over $40 million, the rate was 40 per cent. (The top rate was

Provincial Resource Policy / 61

subsequently lowered to 30 per cent.) Since 1970, Ontario's share of resource revenues has been significantly below that of the other major mineral-producing provinces. The province has been consistent in not exercising many of its revenue options as a resource owner. In modern times Ontario has used only one instrument of policy-the least intrusive, the profits tax. Other than exercising its promotional role, the Ontario government has not played an active part as resource manager. The only direct attempts to intervene in private-sector decisions have been made in order to promote domestic processing. However, legislation regulating processing has often been emasculated through the granting of exemptions or through non-implementation. For example, although the profits tax was amended to prohibit deductions for foreign processing, pressures from the industry succeeded in watering down the legislation by gaining numerous exemptions. By 1978 the exemptions were made permanent, and the only section of the legislation still in force is a higher deduction for processing in northern Ontario.50 Ontario's policies contain few of the features we associated with left governments. Successive non-left governments in Ontario have dealt with natural resources in a manner consistent with a conservative ideology. Reliance on a profits tax has meant that only a small proportion of the resource industry's revenues are taken by the public sector. As will also be evident in the case of Alberta, conservative governments in Ontario have intervened in the economy solely to promote private-sector development - a purpose quite different from those of left governments. Alberta Alberta, at present the province with the most significant resource wealth, has employed a number of policy instruments. It has introduced gross royalties and super-royalties as well as some forms of public ownership. The federal measures taken to benefit from rising oil prices in the early 1970s were especially provocative to Alberta. The 1973 session of the legislature passed a number of bills to defend provincial jurisdiction against federal encroachment. The gross royalty rate on production was raised, and a supplementary royalty on old oil was put in place. As in 1954, when the Alberta Gas Trunk Line was created as a defence against gas-gathering and to pre-empt federal action, the establishment of the Alberta Energy Corporation (1973) as well as Pan Alberta (1972) must be seen as defensive public enterprise vis-a-vis Ottawa and a market that is believed to be dominated by central Canada. It might at first appear that the initiatives of the Conservative government of Peter Lougheed in Alberta run counter to my hypothesis, but a closer examination suggests another interpretation. First, the public enterprises created in Alberta are quite different in form and function from those observed, for

62 /M.A. Chandler

example, in the Potash Corporation of Saskatchewan. Alberta's corporations have not been created to supplant the private enterprise; in fact, the Alberta Energy Corporation is not permitted to compete in the development of conventional oil and gas. The public corporations have been used to promote a home-grown private sector. The Alberta government has intervened in the name of Alberta's private enterprise; this intervention has at times meant public action to counteract the pressures of the federal government and the national economic system. Alberta's government seems to be more concerned with changing the distributional and allocational consequences of the east-west economic relationship than with providing a more equitable distribution of economic benefits within the province or with using the power of the state to deconcentrate economic power. In Alberta, the use of public ownership was designed primarily to strengthen the province's hand vis-a-vis outsiders and to benefit the province's business community and urban middle class.51 Comparing Policy Models

From measures of revenues and state intervention, it is apparent that the long-standing policy uniformity in provincial resource policy was shattered in the 1970s. Much of the subsequent policy variation could be associated with the ideology of the party in power. Not only have social democratic governments consistently employed more interventionist instruments, they have for the most part also acted to increase the public portion of resource revenues. The fact that the resource industry tends to be highly concentrated and predominantly foreign-owned has perhaps made the redistributive policies of the left more palatable to the public. To some extent nationalistic sentiments may have supported redistributive policies. Where the NDP has held power and then lost it, its policies are brought into greater relief by the actions of those governments that have succeeded it. In the cases of Manitoba and British Columbia, the incoming non-left governments lost little time in changing NDP mineral policies. It might be argued that a returning NDP government may also have changed those policies, but the evidence remains that when the left had the opportunity, it moved in keeping with its ideology. The case of Alberta highlights another factor in provincial politics. Although Alberta's Progressive Conservative government has employed instruments which on their face are more associated with the left, in fact most of its interventions in the economy have not been for internally redistributive purposes. The Alberta government has been using resource taxation to promote

Provincial Resource Policy/63 industrial development, just as in earlier times all provinces employed tax policies to promote resource development. Alberta's actions in the name of nationalism represent a different kind of intrusion from the redistributive policies of the left. Similarly, the nationalistic policies of the Quebec governments that preceded the PQ were interventions not to supplant the private sector, but to promote a certain kind of development. Conclusion This paper has examined the linkage between party and policy. Measures of revenues and state intervention make it apparent that the uniformity of earlier provincial resource policies disappeared in the 1970s. The policies that were then adopted varied significantly. From the discussions relating to the individual provinces, a general picture of resource policy emerges that provides support for both the convergence and party models. For many years the provinces had similar policies, which appear to be related to two economic factors - limited resource rents (and hence limited resource revenues available to the provinces) and an undisputed need to attract venture capital. Although these conditions may determine the need for a policy response, they do not necessarily determine the nature of that response. For this we must look to the actors and institutions within each province. There we find a similar set of governments that were fairly weak, financially pressed, and unable to bear much of the risk of resource development. In addition, prior to 1969 only Saskatchewan had a left government, and even there it should be noted the CCF was concerned primarily with its health and social policy programs. Redistribution via the resource system had to take a back seat. The similar policies of the provincial governments during this period do not indicate that politics were unimportant, but rather that similar policy outputs came from similar actors. In the 1970s, Canada's economic and political setting changed dramatically. There was a huge increase in resource prices, and the federal government began to compete for the ensuing rents. These stimuli called forth new political responses. Unlike the earlier period in which social and economic conditions acted upon relatively similar sets of political forces, by the 1970s the actors and institutions had also changed. The province-building of the 1960s and 1970s had resulted in more capable provincial governments that were more willing to bear some of the risks of exploration.52 It is clear that the provinces have become much stronger in dealing with the resource industries. The provinces have been willing to use state enterprise as an instrument of industrial strategy. This has

64/M.A. Chandler meant not only a continuation of promoting the private sector, but also some use of more intrusive nationalistic instruments. Besides the general strengthening of provincial governments (which resulted in more nationalistic policies), a number of provinces were governed by social democratic parties that were predisposed to challenging industry. In conclusion, a convergence model that emphasizes the impact of environmental forces is not without utility here. Neither phase of resource policymaking can be understood without adequate reference to its social and economic setting. However, the basic premise of a convergence model - that external forces overwhelm internal political factors - is not supported. Resource policy has not been solely the product of external forces; it is also a function of internal political factors. Among those factors party ideology has proved to be an important determinant of resource policy. NOTES 1 For a summary of the U.S. material see T. Dye, Politics, Economics and the Public (Chicago: Rand McNally 1966). For the Canadian findings see D.J. Falcone and M.S. Whittington, 'Output Change in Canada: A Preliminary Attempt to Open the "Black Box",' paper presented to the Canadian Political Science Association, June 1972. See also Richard Bird, The Growth of Government Spending in Canada (Toronto: Canadian Tax Foundation 1970). 2 See, for example, S.R. Rakoff and G. Schaefer, 'Politics, Policy and Political Science; Theoretical Alternatives,' Politics and Society 1 (1970), 51-78, and Joyce Munns, 'The Environment, Politics and Policy Literature: A Critique and Reformulation,' The Western Political Quarterly 28 (1975) 646-67. See also E. Jennings, 'Competition, Constituencies and Welfare Policy' American Political Science Review 13 (1919), 414-29. 3 D. Hibbs, 'Political Parties and Macroeconomic Policies,' American Political Science Review 71 (1977), 467-87; David Cameron, 'Expansion of the Public Economy,' American Political Science Review 72 (1978), 1234-61; F. Castles and G. McKinley, 'Does Politics Matter: An Analysis of the Public Welfare Commitment in Advanced Democratic States,' European Journal of Political Research 7 (1979), 169-86; E. Tufte, 'Political Parties and Social Class and Economic Policy,' Government and Opposition 14 (1976), 18-36, and D. Hibbs and M. Fassbender, Contemporary Political Economy (New York: Elsevier 1981). 4 See C. Winn and J. McMenemy, Political Parties in Canada (Toronto: McGrawHill Ryerson 1977), ch. 8, and S.D. Berkowitz, 'Forms of State Economy and the Development of Western Canada,' Canadian Journal of Sociology 4 (1979), 287-312.

Provincial Resource Policy 765 5 See O. Kirschen et al., Economic Policy in Our Time (New York: North Holland Publishing 1964); S. Harris, The Economics of the Political Parties (New York: Macmillan 1962); E. Tufte, Political Control of the Economy (Princeton, NJ: Princeton University Press 1978); and The Impact of Parties, ed. F. Castles (Beverly Hills: Sage Publications 1981). 6 The term 'convergence' is used here to mean similar outputs despite differing political structures and ideologies. 'Convergence' has been used elsewhere to refer to the development of similar structures and consequent policy responses by divergent governments. See, for example, Harold Wilensky, The Welfare State and Equality (Berkeley: University of California Press 1975.) 7 H.V. Nelles, The Politics of Development (Toronto: Macmillan 1974), 87-103. 8 G. LaForest, Natural Resources and Public Property under the Canadian Constitution (Toronto: University of Toronto Press 1969). See also R. Burns, Conflict and Its Resolution in the Administration of Mineral Resources (Kingston: Centre for Resource Studies, Queen's University 1976), ch. 3, and W. MacDonald, Constitutional Change and the Mining Industry in Canada (Kingston: Centre for Resource Studies, Queen's University 1980). 9 The relationship between the state and resource sector is not specific only to Canada. See Leon Lindberg, The Energy Syndrome (Lexington, Mass.: Heath Books 1974). 10 G. Stevenson, 'The Process of Making Mineral Resource Policy in Canada,' in Natural Resources in U.S.-Canadian Relations, vol 1, ed. C. Beigie and Alfred Hero Jr (Boulder: Westview 1980). 11 Nelles, Politics of Development, 87-103. 12 Royal Commission on Dominion-Provincial Relations, Report, book II (Ottawa: Queen's Printer 1940), 14. 13 C. Alway, 'Hepburn, King and the Rowell-Sirois Commission,' Canadian Historical Review 48 (1967), 113-41. 14 D. Smiley, 'The Political Context of Resource Development in Canada' in Natural Resource Revenues: A Test of Federalism, ed. A. Scott (Vancouver: University of British Columbia Press 1976), 52-61. 15 Meyer Bucovetsky, 'The Mining Industry and the Great Tax Reform Debate,' in Pressure Group Behaviour in Canadian Politics, ed. P. Pross (Toronto: McGrawHill Ryerson 1975), 87-114. 16 MacDonald, Constitutional Change. 17 Central Canada Potash Co. Ltd. et al. v. Government of Saskatchewan et al. (1978) 88DLR(3d)609. 18 S.I. Bushnell, 'The Constitution and Natural Resources,' Canadian Public Policy 6 (Summer 1980).

66/M.A. Chandler 19 For a discussion of the initial changes see R.D. Brown, 'The Fight over Resource Profits,' Canadian Tax Journal (July-Aug. 1974), 315-37. 20 For a discussion of the characteristics of the range of instruments available to governments see M. Trebilcock et al., The Choice of Governing Instrument (Ottawa: Economic Council of Canada 1981). For comparisons of specific instruments see the essays in this volume by Woodside and Prichard and Trebilcock. 21 It should be noted that the use of regulatory instruments such as land regulation and pro-rationing (production restrictions) as well as the mechanisms used to oversee the regulatory system might help to further distinguish among provincial policies. 22 For a description of these formulae see 'Current Tax and Royalty Treatment of the Petroleum Industry' in 1978 Conference Report (Toronto: Canadian Tax Foundation 1978), 165-218. 23 See J.C. Leith, 'What Is Ontario's Mineral Policy?' Canadian Public Policy 4 (Fall 1978). 24 For a discussion of Alberta's and Saskatchewan's efforts, see John Richards and Larry Pratt, Prairie Capitalism (Toronto: McClelland and Stewart 1979). 25 M.L. McMillan and K. Norie, 'Province-Building vs. A Rentier Society,' Canadian Public Policy 6 (special supplement 1980), 213-20. 26 T. Lowi, 'Decision-making vs. Policy-making, An Antidote for Technocracy,' Public Administration Review (May-June 1970), 314-25; Issues in Canadian Public Policy, ed. G.B. Doern and S.V. Wilson (Toronto: Macmillan 1974). 27 'The Regulation and Taxation of the Oil Industry,' in The Choice of Governing Instrument: Some Applications (Ottawa: Economic Council of Canada 1981) 28 J.C. Leith, 'Exploitation of Ontario Mineral Resources: An Economic Policy Analysis,' Ontario Economic Council working paper (Dec. 1976). See also Ross Garnant and A. Clunies Ross, 'A New Tax for Natural Resource Projects' in Mineral Leasing as an Instrument of Public Policy, ed. M. Crommelin and A.R. Thompson (Vancouver: University of British Columbia Press 1977), 78-90. 29 For a critique of these and similar assumptions, see Mason Gaffney, 'Objectives of Government Policy in Leasing Mineral Lands' in Mineral Leasing, 3-26. 30 Richards and Pratt, Prairie Capitalism, 321. 31 M. Crommelin, P. Pearse, and A. Scott, 'Management of Oil and Gas Resources in Alberta: An Economic Evaluation of Public Policy,' discussion paper no. 76-19 (University of British Columbia, June 1976). 32 For an elaboration of this theme, see Marsha Chandler, 'State Enterprise and Provincial Politics,' Canadian Journal of Political Science 15 (Dec. 1982). 33 For a discussion of Quebec's facilitative activities see Silvana Guindon, Provincial Mineral Policies: Quebec 1945-15 (Kingston: Centre for Resource Studies, Queen's University 1977).

Provincial Resource Policy 767 34 Richards and Pratt, Prairie Capitalism, 234. 35 For interesting comparative analyses that link the ideology of the left with public policies see The Impact of Parties, ed. F. Castles. 36 See K. McRoberts and D. Posgate, Quebec: Social Change and Political Crisis, rev. ed. (Toronto: McClelland and Stewart 1980); J. Niosi, The New FrenchCanadian Bourgeosie,' Studies in Political Economy 1 (Fall 1979), 113-61; and Pierre Fournier, 'Les nouveaux parametres de la bourgeoisie quebecoise,' in Le capitalisme au Quebec, ed. P. Fournier (Montreal: Albert St. Martin 1978), 135-82. 37 For a comparison of the two parties see Ivan Auakumovic, Socialism in Canada (Toronto: McClelland and Stewart, 1978), 210-15. For an analysis of the program of the PQ see Vera Murray, Le Parti quebecois, de lafondation a la prise dupouvoir (Montreal: Hurtubise HMH Ltd. 1976), ch. 2. 38 SBC 1974, c. 54. 39 SBC 1973, c. 52. 40 Raymond Payne, 'Corporate Power, Interest Group Activity and Mining Policy in British Columbia,' paper presented at the 1980 Canadian Political Science Association meetings, 16. 41 See articles by J.L. McPherson and Dale Jordan in Crommelin and Thompson, Mineral Leasing, for critiques of the NDP'S attempts to manage the resource sector. 42 B. Owen and W. Kops, The Impact of Policy Change on Decisions in the Mineral Industry (Kingston: Centre for Resource Studies, Queen's University 1979). 43 Ron Murray, Provincial Mineral Policies: Saskatchewan, 1944-75 (Kingston: Centre for Resource Studies, Queen's University 1978). 44 Richards and Pratt, Prairie Capitalism, ch. 6. 45 For a critique of the court's action see Arne Paus-Jenssen, Resource Taxation and the Supreme Court of Canada: The Cigol Case,' Canadian Public Policy 5 (Winter 1979), 45-58. 46 Personal interviews with Hon. Allan Blakeney, premier of Saskatchewan, and Gary Beatty, president of Crown Investments Corporation, Saskatchewan. 47 Gerard Gaudet, 'Forces Underlying the Evolution of Natural Resource Policies in Quebec,' in Beigee and Hero, Natural Resources, 247-60. 48 See Pierre Fournier, Les societes d'etat et les objectifs economiques du Quebec: une evaluation preliminaire (Montreal: La Documentation Quebecoise 1979). 49 Ross Hynes, Provincial Mineral Policies: Ontario 1945-1975 (Kingston: Centre for Resource Studies, Queen's University 1978), 16. 50 Leith, 'What is Ontario's Mineral Resource Policy?' 51 See Larry Pratt, The State and Province-Building: Alberta's Development Strategy,' in The Canadian State, ed. Leo Panitch (Toronto: University of Toronto Press 1976). 52 On province-building see M. A. Chandler and W.M. Chandler, Public Policy and Provincial Politics (Toronto: McGraw-Hill Ryerson 1979), ch. 1.

This page intentionally left blank

4 / Determinants of Legal Aid in Canada: Actors, Policies, Programs, and Futures DALE H. POEL

Hattie got a divorce this summer. Most of the campers who bought blocks of ice at her small house on a rural road found it hard not to hear her story. Hattie is a middle-aged country woman, and her version of the story includes a no-good husband, hard drinking, abuse, and poverty. Friends kept telling her not to put up with her problems, and she finally did something. In her own words, she 'got a legal aid divorce.' Paul is 18. While driving around he and two friends decided to break into a small-town store. It seemed like a good idea; his two friends had done it the week before and knew where the money was kept. This time the owner was sleeping in the back room. Later that week Paul pleaded guilty to a charge of breaking and entering. A legal aid lawyer advised him on his plea and at trial a month later spoke to sentencing. Because of Paul's secondary role in the episode, as brought out by his lawyer, Paul was given only thirty days in jail. On a warm summer night George went to the county fair. After an hour he left the sound of the rides and the smell of popcorn. With him were two young girls, ten and twelve years old. One ran away; the other was found dead in a distant field. Some time later a legal aid staff lawyer stood beside George as he pleaded 'not guilty.'1 Of the examples given, the availability of legal aid services was possibly least important to George. Because he was charged with murder, he would have received a court-appointed lawyer if he so wished. In most provinces the private lawyer would have received some remuneration for his legal services. He may or may not have had the years of criminal courtroom experience of George's legal aid lawyer. In the absence of legal aid, Hattie and Paul might well have experienced quite different developments in their lives. Without a free 'legal aid divorce' and

70/D.H. Poel without funds to pay a private lawyer, Hattie would likely have continued life with an abusive and unfaithful husband. Given her meagre resources and limited social skills, finding other solutions to her problems would be unlikely. Neither Paul nor his family could afford the services of a private lawyer. Because he was caught in the act and charged with a common offence, Paul might have gone through the justice system without representation. Without a lawyer to emphasize several mitigating circumstances in his situation, and because of his previous record, Paul would have received a longer sentence and served it in a larger correctional centre where the chances of physical abuse would be great. In the Canadian provinces and territories, criminal and family law matters similar to those described in the above cases are the largest components of legal aid workloads. They do not exhaust the range of legal work being done under the banner of legal aid, however. This essay assumes that the creation and development of legal aid in Canada is a consequence of the interplay of organized groups who share overlapping interests - financial, professional, and ethical-in legal aid as a governmentsupported program. The political-professional interplay of these actors has brought legal aid to its current status in Canada. I will use some recently available program statistics to outline the structure, scope, and eligibility profiles of the provincial and territorial legal aid plans. I will then consider the determinants of legal aid per se and the determinants of the alternative program directions found in the different plans. The use of the terms 'actors' and 'determinants' suggests some explanation or model of decision-making with respect to legal aid and might suggest a conclusion, resolution, or final portrayal of the state of legal aid in Canada. But decision-making does not stop, the social and professional environment of any government program does not remain static, and any cluster of 'actors' will likely constitute in some way both winners and losers of past decisions. Inevitably, the actors will rejoin at some later date to continue the arguments and press their interests. The chapter concludes with a discussion of the future of legal aid and the currently pressing determinants of change in Canadian legal aid -changes in process, yet to be concluded. Issues and Directions in Canadian Legal Aid

Legal aid consists of the free delivery of legal services by lawyers or paralegals, supported by public funds, to all individuals for whom the fees of a private lawyer would be prohibitively expensive. Legal aid as a formal concept and a potential government service in Canada became recognized during the second

Legal Aid/71 TABLE 1 Total expenditures (in thousands of dollars) for legal aid by province/territory (1977-78 and 1979-80) and percentage changes over that period

Province/territory

1977-78

1979-80

Per cent change

Ontario Quebec British Columbia Alberta Saskatchewan Manitoba Nova Scotia New Brunswick Newfoundland Northwest Territories Prince Edward Island Yukon

$32,343 26,168 8,404 5,636 3,768 3,306 1,464 702 602 381 95 73

$36,653 35,995 11,611 4,555 3,927 3,512 1,993 745 699 453 87 151

+13 +38 +38 -19 +4 +6 +36 +6 + 16 +19 -8 +107

SOURCES: Legal Aid Services in Canada 1977-78 and Legal Aid Services in Canada 1979-80

World War. After the war, a gradual transition was made from individual lawyer- or law-society-organized volunteer services to formal structures and programs involving law societies, independent legal aid commissions, provincial and territorial governments, and the government of Canada.2 The 1950s and 1960s were years of experimentation by law societies and their respective provincial governments. The first cost-sharing agreement with the federal government for the provision of legal aid in criminal matters was signed with the Northwest Territories in 1971. In 1973 and 1974, all the remaining provinces signed similar agreements (the Yukon signed in 1977). Each of the twelve agreements contains similar provisions, but room is also left for provincial discretion in many areas.3 Table 1 suggests the range of absolute effort in support of legal aid programs for the budgetary years 1977-79 and 1979-80. This table shows the absolute contrasts without per capita adjustments, and permits comparisons between respective budgetary increases. The absolute contrasts are extreme. Prince Edward Island spent only $87,000 in 1979-80, while Ontario spent $36,653,000. The figures suggest three groupings among the plans. The largest and most complex plans are those of Ontario, Quebec, and British Columbia. Alberta,

72/D.H. Poel TABLE 2 Legal aid expenditures per capita and federal contribution as a percentage of total provincial/territorial expenditures (1977-78 and 1979-80), ranked Expenditures per capita NWT Yukon Quebec BC Ontario Saskatchewan Manitoba Nova Scotia Alberta Newfoundland New Brunswick PEI

1977-78

1979-80

Federal contribution

$8.73 3.41 4.17 3.35 3.85 4.00 3.21 1.75 2.94 1.06 1.02 0.78

$10.58 7.07 5.72 4.42 4.28 4.06 3.42 2.34 2.20 1.21 1.06 0.70

PEI New Brunswick Newfoundland Yukon Nova Scotia NWT Alberta Manitoba BC Saskatchewan Ontario Quebec

1977-78

1979

90% 80 64 49 49 34 28 26 25 21 21 20

95% 79 53 50 44 48 44 30 22 25 24 18

SOURCES: Legal Aid Services in Canada 1977-78 and Legal Aid Services in Canada 1979-80

Saskatchewan, Manitoba, and Nova Scotia fall in the middle range. New Brunswick, Newfoundland, the Northwest Territories, the Yukon, and Prince Edward Island have the smallest plans. The expenditure changes shown for this period also distinguish the plans. Saskatchewan, Manitoba, and New Brunswick show fairly small percentage increases which, when compared to rates of inflation, suggest a falling-back. Ontario, Newfoundland, and the territories did somewhat better, while Quebec, British Columbia, and Nova Scotia show the largest gains except for the Yukon. The Yukon's large increase reflects the start-up period of its first shared-cost agreement with the federal government. During this period most of the plans received minimal, no-growth funding; none was in a position to expand its services based on these increases. Alberta, but not Prince Edward Island, experienced a serious drop in legal aid expenditures. Table 2 shows that most of the provinces with small increases also showed an increased reliance upon federal government funding. Nova Scotia, British Columbia, and Quebec show modest decreases, while Newfoundland, an exception, shows the largest decrease in federal government support. The decrease is in proportion to federal government support and is not an absolute decrease. The left side of table 2 converts the initial expenditures shown in table 1 into a per capita figure. With the exception of the territories, Quebec has consistently shown the highest legal aid expenditure per capita. Observers have attributed this to the Quebec government's strong commitment to a comprehensive legal

Legal Aid 773 TABLE 3 Provincial rank scores on per cent of families designated as 'lowincome' and on legal aid clients per 1,000 population (1978-79) Rank on low-income families

Rank on legal aid clients

Difference in ranks

Newfoundland Prince Edward Island New Brunswick

1 3 4

8 9 10

-7 -6 -6

Manitoba Saskatchewan Alberta Nova Scotia

5 2 8 6

6 2 7 4

-1 0 4-1 +2

10 9 7

5 3 1

+5 +6 +6

Province

Ontario British Columbia Quebec

SOURCES: Statistics Canada, 1981 Census: Family Income; Levy, address to the Commonwealth Law Conference

aid program and its belief in governmental organization as a means of meeting social needs.4 The right side of table 2 shows that Prince Edward Island and New Brunswick rely almost entirely upon federal monies to support their limited legal aid plans. At the other extreme, Manitoba, British Columbia, Ontario, Saskatchewan, and, to the greatest extent, Quebec only have federal support for 18 to 30 per cent of their total budgets. The larger the provincial portion of a legal aid plan's budget, the more comprehensive are the plan's services. The provincial expenditures support legal services in areas other than criminal law (which is the focus of the existing federal-provincial agreements). Table 3 ranks the provinces on the basis of the proportion of low-income families in the total population and the number of legal aid clients served per 1,000 population. Newfoundland has the largest proportion of low-income families, Ontario the smallest. Proportionally, Quebec's legal aid plan serves the most clients, New Brunswick's the least. The client figures do not include 'duty counsel' service - minimal advice given spontaneously in a courtroom context. Duty counsel service constitutes a large part of New Brunswick's legal aid activity. From the differences between the two rank-order scores, we can conclude that Newfoundland, Prince Edward Island, and New Brunswick are providing less legal aid service than one would expect given their low-income populations.

74/D.H. Poel TABLE 4 Payments to members of the private bar as a percentage of total provincial/territorial legal aid expenditures, 1977-78 and 1979-80 Province/territory

Legal aid model

1977-78

1979-80

Yukon Northwest Territories Ontario Alberta New Brunswick

95 72 76 56 68

77 73 71 71 70

Judicare

Manitoba British Columbia Newfoundland Quebec

44 46 60 32

49 48 45 38

Mixed

Saskatchewan Prince Edward Island Nova Scotia

26 31 4

6 2 2

Staff lawyer

SOURCES: Legal Aid Services in Canada 1977-78 and Legal Aid Services in Canada 1979-80

Manitoba, Saskatchewan, Alberta, and Nova Scotia are providing services at approximately the level they should. Ontario, British Columbia, and Quebec each are providing more service than one would expect given the proportion of low-income families in their respective populations. Again, Quebec is making the greatest program effort of the three largest plans, considering the province's financial resources. Table 4 uses the fees paid by legal aid plans to participating members of the private bar to array the provinces and territories into three groupings that reflect the separate responses to Canada's judicare-staff lawyer debate5 as it was carried out between respective law societies and the provincial governments. The Yukon, Northwest Territories, Ontario, Alberta, and New Brunswick are operating legal aid plans in which the judicare model is operative. This model provides legal services through private lawyers who submit a bill to the plan for services rendered to legal aid plan clients. Ontario is on the verge of leaving this camp as its staff lawyer-clinic program continues to expand.6 Manitoba, British Columbia, Newfoundland, and Quebec have mixed-model plans with staff lawyers and judicare services. Newfoundland's figures suggest the most significant move toward a new, equal emphasis on the two components. Quebec's ratio shows a balance favouring the clinic-staff lawyer component. Prince Edward Island's choice of model is not yet clear. It has only recently ex-

Legal Aid 775

panded its services beyond the criminal law representation provided by two public defenders. Saskatchewan and Nova Scotia have clear preferences for the staff lawyer model. These structural differences are the result of major political struggles between governments and law societies. The determinants of these outcomes will be discussed later. The information reviewed in the tables shows that the plans are quite different in their comprehensiveness and organizational structure. These patterns reflect differences in values, program priorities, and resources. The policy decision to have legal aid services available across Canada seems to be settled, but a variety of program issues remain for debate. These issues are not always on the legal aid agenda, but come and go in response to pressures for change. Almost every term in the definition of legal aid given earlier contains the seed of an argument. Should legal aid services be completely free or should a plan collect some use or partial fee to generate revenue or discourage clients? Should the legal services provided be 'remedial' or 'pro-active'?7 The remedial approach accepts the existing law and deals with clients on an individual basis; the pro-active stance views legal aid as 'one weapon among others in the struggle against poverty.'8 Should legal services be provided only by lawyers or is there an important role for paraprofessionals? Who should receive legal aid services - only individuals with a specific legal need? Or should groups in society benefit from more broadly aimed advocacy? Can staff lawyers, on a government-paid salary, provide the quality and independence of counsel to which the legal profession aspires, or are these values better assured under the judicare model? The concluding section of this essay will consider the determinants of program outcomes that relate to most of these issues. The issues, however, do not rise or fall by themselves. Many individuals have interests in the expansion, maintenance, or reduction of legal aid plans. They differ in their resources, their priorities, and the amount of attention they pay to legal aid programs. The next section will briefly discuss the actors in legal aid policy and program decisionmaking. Actors in Legal Aid Policy and Program Decision-making

The major actors in Canadian legal aid decision-making include lawyers within their professional organizations, the judiciary, the provincial governments, the government of Canada, the taxpayer, legal aid clients, and legal aid lawyers, administrators, and commissioners.9 Each of these is a convenient, but possibly obscuring, shorthand label for legal aid actors. We may speak of Ontario or of the federal government, but we are really referring to individual ministers, ministers in cabinet, or civil servants who have values, attitudes, priorities, and

76/D.H. Poel resources with which they approach legal aid issues. The collective expression of these priorities or resources through cabinet documents, budgets, law society briefs, and federal-provincial agreements constitutes the determinants of legal aid policy and programs. These determinants will be considered in the next section. Here it is sufficient to remind ourselves that the values, attitudes, priorities, and resources are held by individuals making choices within political and professional structures. The legal profession All Canadian lawyers are members of provincial or territorial law societies, the organizations created by provincial statute that govern the profession. The Canadian bar has been one of the most significant actors in legal aid policy and program decision-making. We would be mistaken, however, to treat the 'profession' as a monolithic entity. One author writes that 'to speak of "the legal profession" is to ignore important internal conflicts which have arisen and been resolved in different ways over the years.'10 Little systematic research has been done on the distribution of lawyers' views in Canada, and none has focused specifically on the views held by individual lawyers concerning legal aid.'' One might expect views on legal aid to vary according to a lawyer's seniority in the bar, leadership role within the profession, or type of legal practice. Commentators have emphasized the differing opinions of those practising law relative to those teaching or studying in Canadian law schools.12 The profession as a whole and individual lawyers are clearly major actors in legal aid decision-making. They have much to gain or lose in maintaining professional principles, preserving economic rewards, and controlling occupational opportunities. The judiciary The judges in Canadian courts, especially in the lower courts, have an interest in the quality and efficiency of the legal action taking place before them. If a criminally charged defendant appears without legal counsel, a judge may have to fill the awkward dual role of both judge and counsel. Lawyers with little criminal trial experience may frustrate the judge's hopes for efficient and effective representation. If there are differences between private and legal aid lawyers, judges can be expected to develop favourable or unfavourable views toward legal aid. In Nova Scotia, for example, judges have spoken very favourably about the quality of legal aid criminal defence casework and, in some areas of

Legal Aid 777

the province, considerably less favourably about the performance of some private lawyers with no criminal trial experience.13 Provincial and territorial governments The provincial and territorial governments are responsible for the delivery of legal aid services. As with the legal profession, the provinces and territories cannot be taken as one actor. Each contains within it sub-actors with unique interests and perspectives. The cabinet, collectively, will be a powerful force in arbitrating competing interests and establishing government priorities. Within the cabinet, the individual ministers represent departments differently affected by legal aid and competing with legal aid for scarce financial resources. The provincial department of the attorney-general has the most immediate professional relationship with any legal aid plans, since crown attorneys prosecute persons charged in criminal matters. Despite this adversarial relationship defined by the justice system itself, many legal aid plans submit their budgetary proposals to the provincial treasury boards through the attorney-general's office. Provincial departments of social services will also have an interest in legal aid plans because many of their clients will be legal aid clients. They may see legal aid either as another 'helping agency' or as an adversary in a child-custody case. The Government of Canada Constitutional responsibility for criminal law and a concern for national standards in social programs have made the federal government a prime actor in decision-making and agenda-setting in legal aid. The Department of Justice has been the major means through which the government's interest has been expressed, but other departments or agencies are also interested in legal aid. The Department of Health and Welfare will provide cost-sharing monies to the provinces for civil legal aid (if the provincial legal aid administrators can generate the necessary background information on clients and cases).14 Health and Welfare has also competed in the past with Justice for control of legal aid as a federal program. The solicitor-general provides housing for the legal aid clients who are sentenced to a federal penitentiary, and shares an interest with legal aid lawyers in reducing Canada's current prison population. Statistics Canada has a general responsibility to compile justice statistics, and the Department of Indian and Northern Affairs is concerned with the legal needs of Canada's native populations. In addition to shared program interests, these and

78/D.H. Poel

other federal programs grouped in the social policy area of the federal budget find themselves competing with legal aid for financial support. The taxpayer The assessment of legal aid by citizens could be either financial or substantive or both ('We shouldn't spend money to defend that kind of person'). Most Canadians have little awareness of legal aid, and neither governments nor scholars have conducted systematic research on the public's opinion of legal aid plans. Without such research, the safest conclusion is that major legal aid participants - leading members of the bar, federal and provincial politicians, and program administrators - probably overestimate the amount of information about legal aid that citizens have. Informed opinion on legal aid is likely to be found only among those with the most direct contact with the service, either as clients, employees of a legal aid plan, professionals who interact with legal aid, or political decision-makers.15 Legal aid clients Legal aid clients are low-income individuals or groups who are in conflict with the law or who could benefit economically or socially from a positive application of or change in the law. They are potential actors in legal aid decision-making and agenda-setting; to date, they have been only passive objects of professional or personal concern on the part of others. Clients are the largest group of actors with a stake in the continued existence and quality of legal aid in Canada. They are least likely to have a significant direct impact on legal aid developments except to the extent that their legal needs and views are carried into the decision-making arena, with or without interpretation, by other actors such as lawyers or other professionals. The public hearings associated with major reviews of legal aid plans are likely to be dominated by participating professionals, not by bona fide clients.15 The past, present, and future clients of legal aid plans are likely to be politically unorganized and probably do not think of themselves as an interest group. Those receiving legal counsel in criminal cases do not anticipate a need for similar future services (even though many do return to legal aid offices).16 This would certainly be true of the two criminally charged individuals described at the beginning of this essay. Many family law matters are ongoing, but a client like Hattie would have no expectation of needing another legal aid divorce.

Legal Aid / 79 Legal aid lawyers, administrators, and commissioners Once legal aid plans were established, the professional participants became significant in decision-making and agenda-setting. They will try to protect what they have and make it grow if they can. The directors and other major administrators of the plans meet annually. They are a nationally organized interest group with personal and professional stakes in the continuation, expansion, and financial support of legal aid programs. The actors involved in legal aid have different resources and interests: lawyers have their tradition, professional principles, and political experience; federal government departments have a significant interest in legal aid and substantial financial resources; the provinces have the constitutional responsibility, some have the financial resources, and most have the programmatic commitment. Interaction between these groups creates the collective expression of priorities and commitments that become the determinants of legal aid policy and programs. The Determinants of Canadian Legal Aid Policy and Programs

The existing documentary record of legal aid in Canada does not permit the measurement of each factor that might be brought forward to 'explain' legal aid policy decisions. For example, the direction and strength of opinion held by lawyers seems an obviously important variable. The literature informs us about the views of lawyers in those provinces where a law society's impact on legal aid decision-making was obviously significant or where the views were strongly held and fought for. There is some assessment of the private bar's collective views on legal aid in Ontario and Quebec,17 but very little on positions taken by law societies of New Brunswick, the Yukon, or Manitoba. Systematic comparison would require the measurement of each factor for each of the twelve legal aid plans. References to ten provinces and two territories should not mislead the reader into accepting the previous discussion as 'comparative analysis.'18 The comparative analysis that follows, then, is an exploratory attempt to specify the key factors in an explanation of legal aid policy and program decision-making without necessarily having a firm measure of each variable for each unit of analysis (province or territory). The suggested relationships should be taken as hypotheses. To pursue this analysis objective, it is useful to distinguish five outcomes. The first is the decision to formalize and finance legal aid programs. This is a basic policy decision, one which was taken in most provinces and territories between

80/D.H. Poel

1971 and 1974 in the context of federal-provincial negotiations, although provincial initiatives had been made in the latter half of the 1960s.19 Once the decision to implement legal aid was taken, further decisions about the structure of the plan were necessary. Although the basic policy outcome was inevitably interlocked both in time and in substance with decisions about the nature of the program, separating policy and program decision-making is analytically useful and substantively necessary. Three program outcomes are considered - decisions concerning the legal aid program's structure, scope, and client eligibility. Our discussion of the last outcome, the future direction of the programs, must of necessity be incomplete since it will deal with pressures for and directions of possible change. That discussion will be a reminder that decisions taken in policy and program areas are not static conclusions to some debate, but are simply temporary points of reference in the ongoing interplay of interests and resources of the dominant actors. In trying to explain legal aid policy, we must remember that a single factor rarely accounts for anything adequately. None the less, the law of parsimony will lead us to seek the simplest 'stable' multi-determinant explanation of legal aid policy and program decision-making.20 Determinants of legal aid policy The division of responsibilities found in Canada's Constitution spells out the areas of jurisdiction in the administration of justice but has little to do with the choice of an institutionalized, government-sponsored legal aid plan. Commentators have suggested the federal shared-cost proposals,21 idealistic students and members of law faculties,22 the example of the United States,23 and the United Nations international agreement on civil and political rights24 as significant determinants of the implementation of legal aid in Canada. The various determinants can be assessed within a framework that not only orders the items by time but also suggests some priorities. In the broad context we must refer to the professional principles that emphasize an individual's right to legal counsel and define the adversarial legal process as one in which both parties should have representation. A general awareness of legal needs as expressed in international documents, Canada's status as a signatory of those documents, and the passage of Canada's own Bill of Rights are also determinants in the broad context. The judicare example of legal aid in Great Britain and the U.S. implementation of a staff lawyer system in response to Supreme Court decisions and the War on Poverty are probably the most important determinants at this level. They took the general principles of the legal

Legal Aid 781

profession and translated them into specific government programs dealing with recognized problems in the administration of justice. 25 Typically, Canada has drawn upon both models in its own system. The determinants in the intermediate context of legal aid decisions were a growing professional discontent with the purely voluntary legal aid efforts of the law societies and the general solvency and willingness of governments to make funds available for new social programs through the 1960s and early 1970s. Professional discontent was most visible in the areas of criminal law, and resulted from the failure of the various attempts in the provinces and territories to provide legal counsel on a voluntary basis for criminally charged individuals. Many defendants were going through the courts without representation26 and the work of representing those who appeared with counsel was falling disproportionately on the shoulders of those whose social conscience or lack of seniority in their firms did not allow them to avoid the professional obligations imposed upon them by the court or senior bar. The discontent was both pragmatic and principled. Lawyers, like most people not required to take a vow of poverty, disliked working for little or no pay and doing work that others in their profession were able to avoid. Some members of the legal profession, most often law professors and students, felt and articulated a principled response to the inadequate provision of legal services to the Canadian poor. The 1971 Conference on Law and Poverty brought together those concerned with the issues of poverty law.27 The intermediate context also contained general views and programmatic examples of the expanding role of governments in attacking social problems. Legal aid is a form of social service, and was subject to the same forces that affected other government-funded social programs. One author writes, 'Government phrases such as "the war on poverty" and "the just society" were consistent with public support for governments assuming... responsibility for society's disadvantaged groups...It is as a part of the larger movement to help the disadvantaged that legal aid in Canada became established as a government spending program.'28 Experiments with legal aid were tried at the provincial level during the 1960s in those provinces having the resources, commitment, or both. Ontario and Quebec are frequently cited as leaders in this regard, although less comprehensive pilot projects were attempted in almost every province.29 In the intermediate context, then, preliminary versions of legal aid existed in some provinces on a limited basis, and the primary emphasis was placed on legal aid in criminal matters. The immediate context of decisions leading to cross-national legal aid services consisted of two concrete inducements offered by separate departments of the

82/D.H. Poel federal government. The first initiative was the clinic-funding grants given by the Department of Health and Welfare to four clinics in Nova Scotia, Quebec, Ontario, and Saskatchewan. Garth remarks that these 'federal grants sought to foster [neighbourhood law firm] innovations, and they succeeded to a remarkable extent, providing powerful models of and advocates for provincial change.'30 The second federal initiative followed on the heels of the Health and Welfare grants, and was possibly the outcome of strong departmental competition for jurisdiction over the legal aid program area. The Department of Justice negotiated a shared-cost agreement with the provinces to provide federal monies for the support of legal aid in criminal matters.31 The legal aid proposals spelled out the areas of the Criminal Code and other federal statutes that were to have priority (for example, indictable offences), allowed the provinces discretion with respect to the eligibility requirements imposed upon applicants, and allowed the provinces to determine the structures of their respective legal aid plans. The signing of the individual agreements with the federal government represents the final determinant of the current legal aid policy in Canada: provincial willingness to accept federal monies (and priorities) for a program area of shared constitutional responsibility. Figure 1 summarizes this discussion of policy determinants. Only our earlier call for parsimony prevents us from either adding more determinants or drawing more arrows in figure 1. The following hypotheses might be suggested: 1 Discontent with the unmet legal needs of the poor and the quality of criminal justice led to federal proposals for a cost-shared, criminal legal aid, 2 which provincial governments were willing to accept as a new source of funds to address a recognized problem, 3 in a way that was politically feasible because the private bar's discontent with the quality of legal practice made it willing, collectively, to see governments assume responsibility for a matter traditionally within the domain of a 'selfregulated' profession. Determinants of legal aid structure, scope, and eligibility The signing of the federal-provincial agreements in the provinces and territories made available comprehensive criminal legal aid services to Canadians unable to afford the fees of a private lawyer. The agreements created the framework within which the provinces passed legislation establishing the structures of their plans, the scope of services to be offered, and the financial eligibility criteria by which applicants would be accepted as clients. The provincial legislation32 has the commonality necessary to honour the requirements of the federal-provincial

Legal Aid / 83

Figure 1. Canadian determinants of legal aid

agreements and the distinctiveness required to implement politically feasible plans. Except for the Health and Welfare clinic grants, federal government actors apparently remained neutral on the issue of the structure of the legal aid plans. Ontario had its judicare plan firmly in place by the time its federal-provincial agreement was signed. The determinants of legal aid structure are found in the characteristics, priorities, and resources of the law societies and the provincial governments. The structural issue, as noted in the discussion of table 4, is whether the plan will follow a judicare, mixed judicare / staff lawyer, or staff lawyer approach to the delivery of services. Provincial law societies have consistently voiced their preferences for the judicare model in which members of the private bar accept legal aid clients and bill the government according to an agreed-upon tariff. The ultimate resolution of this issue hinges on five determinants: 1 The strength and apparent consistency of private law society opinion in favour of the judicare option. 2 The law society's respective skills as an interest group within provincial politics (normally thought to be fairly high)

84/D.H. Poel

3 The level of provincial financial resources 4 The provincial government's cabinet-level assessment of the relative costs of the alternatives 5 Provincial responses to minority views in favour of poverty-law clinics The Yukon and Northwest Territories complicate the analysis here, because the federal government was responsible for the administration of justice prior to agreements with the two territories. The federal government worked within a judicare framework before the agreements and the pattern remained unchanged after the agreements were signed. Ontario, Alberta, and New Brunswick operate under the judicare model. In these provinces the law societies successfully pressed the case for judicare and the governments made provincial financial resources available to support the plan. New Brunswick's presence in this group is a result of a variation on the ability-to-pay calculation. Until recently, New Brunswick has been able to operate a judicare system with mostly federal monies by limiting its legal aid program to criminal legal aid and by placing greater emphasis on duty counsel services. Ontario's shift towards a mixed system is in part the result of a reassessment of its ability to pay the increasing costs of the more open-ended judicare program (in which the private lawyer has the ability to determine the services required and bill accordingly) and recent responses to minority view pressures for specialized clinic funding.33 Nova Scotia represents the extreme alternative structural outcome. Without strong views favouring the clinic approach to legal aid, the provincial government rejected a committee's recommendations for a predominantly judicare system on the ground that the province did not have the financial resources to implement a comprehensive judicare legal aid plan.34 The remaining provinces reflect a mixture of determinants. The judicare-staff lawyer debate has been largely settled in most provinces. The private bar's defence of the virtues of judicare remains successful only in New Brunswick and Alberta. The bar is losing ground in Ontario and is trying in the mixed system of British Columbia to preserve its hold on criminal legal aid.35 At the other end of the continuum, Nova Scotia may move a little toward a mixed delivery system in response to a challenge from the Nova Scotia Barristers' Society on matters of conflict - that is, cases in which legal aid already represents one party. Such a case could involve co-accused in a criminal matter, or equally eligible spouses in a family-law dispute. The challenge is especially problematic in family-law matters, and puts pressure on the legal aid plan to contract out more cases to lawyers in private practice. Over time the provincial plans may all consist of this mixture of judicare and staff lawyers, and the provincial law societies will have won a partial victory in their defence of the British judicare model.

Legal Aid 785

The scope of any of the twelve legal aid plans will be shaped primarily by the priorities established in the federal-provincial agreements, the provincial government's financial resources and commitment to a comprehensive (criminal and civil) legal aid plan, the presence or absence of a poverty-law perspective on the part of legal aid staff and administration, and the existence of caseload pressures. The private bar has not played a constructive role in defining the scope of legal aid, although it may have resisted provision of some legal aid services in areas of civil law. The starting-point is certainly found in the priorities expressed in the federalprovincial agreements. They read like a litany in most legal aid documents. The key conditions of these agreements include the requirements that: - Legal aid be made available to any eligible person charged with an offence against an act of Parliament punishable by way of indictment or subject to the proceedings under the Extradition Act or the Fugitive Offenders Act. - Legal aid be available at the discretion of the province concerned to any eligible person charged with a lesser offence against a federal act or regulation, punishable by summary conviction or an offence subject to proceedings under the Juvenile Delinquents Act... if conviction might result in imprisonment or loss of livelihood, or where in the opinion of the province special circumstances exist. - Legal aid be made available when appeals are taken to higher courts by the Crown in the cases outlined above, with the province being required...to ensure that the accused can obtain a lawyer as soon as he has been arrested or detained.36

These priorities derive from the federal government's constitutional jurisdiction in matters of criminal law. The scope of legal aid services is ultimately decided by the additional factors suggested above. The degree of reliance on federal monies for the financing of a legal aid plan reflects the extent to which the province or territory is committed to legal aid services beyond the requirements of the federal agreement, and whether it has the financial resources to back up that commitment. A reference to table 2 is instructive in this regard. The range of reliance on federal funds suggests that by 1979-80 Prince Edward Island and New Brunswick had made only the minimal commitment to legal aid as defined in their federal-provincial agreements. Newfoundland, the Yukon, Nova Scotia, the Northwest Territories, and Alberta have expanded services into areas of civil law, but not to the extent that the remaining five provinces have. Manitoba, British Columbia, Saskatchewan, Ontario, and Quebec each receive 30 per cent or less of their program expenditures from the federal government. A willingness to commit provincial resources and a poverty-law perspective on legal aid will lead to an expanded scope of services. The individual lawyer or office caseload within a legal aid system may restrict the scope of services

86/D.H. Poel regardless of what the official mandate or legal aid statute permits. Lazar notes that 'the caseload which all legal aid plans and specialized clinics face almost inevitably overwhelms the intention to do much more than merely provide reactive case-by-case service.'37 A large caseload may mean that applicants involved with civil law disputes or summary-conviction offences are denied legal aid certificates. The definition of'serious' legal need may tend to be restricted to an indictable offence under the Criminal Code or other statutes. Provincial definitions of legal aid eligibility have focused on the ability to pay for private legal services. Although legal aid plans are free to develop their own criteria, for many the starting-point for determining legal aid eligibility is eligibility for provincial or territorial social services. The emphasis on financial eligibility, however, can obscure the obvious fact that a poor person with a legal need in an area not covered by a provincial plan is no more eligible for legal aid than someone who has been disqualified for financial reasons. The pressures of large caseloads combined with the legal profession's definition of'serious' legal needs will interact with financial criteria as determinants of legal aid eligibility. Future directions of legal aid The views, priorities, and resources of actors in the federal and provincial governments and in the existing legal aid systems will be the primary determinants of the future direction of legal aid in Canada. The priorities and opinions expressed by or on behalf of the provincial law societies may push a judicare model toward a mixed system (for example, through pressures for increased judicare fees) or a staff lawyer system toward a mixed system (for example, Nova Scotia's challenge over conflicts), but the basic dimensions of legal aid policies and programs are in place and cannot be easily challenged in a society that is increasingly critical of the legal profession.38 Criticisms that 'legal aid bureaucrats have attempted to instruct lawyers on how to conduct their cases'39 are not likely to affect the priorities of governments and other legal aid actors or the actual delivery of services. The continued federal contribution to criminal legal aid cannot be taken for granted. The Department of Justice believes that the federal government does not get sufficient public recognition for its financial contribution to a 'provincial program.' This and the more significant threat of competition from other social policy programs mean that the federal contribution must be continually defended within the federal government itself. An outcome of these pressures was the cabinet-level decision in the spring of 1981 to renew the federalprovincial agreements for only a one-year period.40 The federal actors must decide whether legal aid is just another social program with which they may experiment, or whether it is an integral component of

Legal Aid/87 the Canadian system of justice. Cabinet social policy committee language seems to imply a long-range commitment to legal aid, and defines legal aid as a requisite of the justice system.41 The same priorities, resources, and interdepartmental competition that led to the initial federal initiatives are likely to sustain the federal role in the future. The federal initiative in the funding of civil legal aid comes from the povertylaw priorities of the Department of Health and Welfare rather than from the administration-of-justice concerns of the Department of Justice. Following its earlier clinic-funding grants, Health and Welfare has offered new support for civil legal aid. Civil legal aid became cost-shareable under the Canada Assistance Plan as of July 1980.42 British Columbia's legal aid plan was the first to claim this funding successfully, and Quebec's was a close second.43 The remaining provinces are busy sorting out procedures and strategies in the hope of following the first two. Scarce resources, a willingness to expand into civil legal aid, and a desire to collect federal money for existing services combine to suggest the further development of federal-provincial relations in the area of legal aid. The new initiatives for federal funding actually involved three federal departments. The Department of the Solicitor-General pressed the new Young Offenders Act through Parliament. The Department of Justice participated in federal-provincial discussions of this replacement of the Juvenile Delinquents Act, and Health and Welfare offered a trade - financial support for civil legal aid in exchange for provincial acceptance of the new act. Implementation of the new legislation will have a potentially major impact on the legal aid plans. Large numbers of young people currently not receiving legal counsel will require such counsel, and most will qualify for legal aid services. The strain on legal aid caseloads and budgets could be enormous. The provincial governments are themselves making significant contributions to legal aid funding. The competition for scarce provincial resources, however, may lead provincial plans to modify their eligibility criteria,44 or to introduce user fees or partial fees to discourage potential clients or generate revenues.45 Another future development in legal aid might be improved record-keeping and standardization of terms that would facilitate attempts at cross-national comparisons of legal aid services. A recent report suggests that the structure of the plan is a determinant of record-keeping behaviour: Those provinces which operate a judicare model of legal aid... possess reliable systems designed to control billings and pay the lawyers for services rendered. Since services ar provided under certificates issued on a case-by-case basis, aggregated data are frequently available. Provinces which depend more on salaried staff to provide services tend to have less complete and less reliable data systems since financial control systems focus to a lesser extent on the completion of individual case assignments. 46

88/D.H. Poel This excursion into economic 'determinism' might suggest that staff lawyers should have some portion of their salary associated with the completion of comprehensive case reports. Whatever the determinants, some consistency in terms used to describe legal aid services would greatly assist those who wish to compare the various plans. One last comment can be reserved for the clients of legal aid. The legal needs of the program clients might have been considered a determinant of legal aid policy and program, except that the needs are not, for the most part, being expressed by the clients themselves. The unmet legal needs that contributed to the professional discontent with the quality of justice in Canada were recognized and defined within the professional and political arenas. Legal aid clients are likely to be the least politicized citizens of Canadian society. NOTES 1 I have obtained information about these three cases solely from personal contacts with the individuals or from the public record. 2 Information about the development and history oflegal aid in Canada is available from a variety of sources, not all of which are easily found in university libraries. The two most recent reviews oflegal aid are Legal Aid Services in Canada 1979 / 80: A Report Prepared for the Implementation Work Group on Justice Information and Statistics (Ottawa: National Legal Aid Research Centre, April 1981), and Canadian Centre for Justice Statistics, Legal Aid, 1981 (Ottawa: Statistics Canada July 1981). An earlier piece frequently referred to is Ian B. Cowie, 'Legal Aid in Canada - 1975,' an address to the Second National Legal Aid Conference, Victoria, June 1975. Additional perspectives are provided by Frederick H. Zemans in 'Legal Aid and Legal Advice in Canada,' Osgoode Hall Law Journal 16, no. 3 (1978), 663-93, and James L. Wilkins, Legal Aid in the Criminal Courts (Toronto: University of Toronto Press 1975). 3 A useful listing of the provincial legal aid statutes and the federal-provincial agreements, with comment, is found in Legal Aid, 1981, 10-18. 4 Personal interviews with federal Department of Justice staff. 5 P.T. Burns, Evaluating the Delivery of Legal Services, a preliminary report to the federal Department of Justice (Vancouver 1978), 8. Burns provides a review of the various arguments for and against both the judicare and public defender options. The Burns report was part of a larger and ambitious proposal to compare systematically the two alternative approaches to criminal legal aid. That project is now reported in P.L. Brantingham and Peter Burns, The Burnaby, British Columbia, Experimental Public Defender Project: An Evaluation, 7 vols. (Ottawa: Department of Justice and Vancouver: Legal Services Society of British Columbia 1981).

Legal Aid/89 6 See the Report of the Commission on Clinic Funding (the Grange Report) (Toronto: Ontario Ministry of the Attorney General 1978) and Rudy Platiel, 'Storefront Clinics Bring Legal Aid to People,' The Globe and Mail, 2 July 1981. 7 See Bryant Garth, Neighborhood Law Firms: A Comparative Study of Recent Developments in Legal Aid and in the Legal Profession (Alphen aan den Rijn: Sijthoff & Noordhoff 1980), 3-13. 8 Commission des services juridiques, 6th Annual Report (Quebec 1978), 55. 9 Burns, Evaluating the Delivery of Legal Services, 19-20. 10 P. Dianne MacFarlane, 'The Legal Profession in Canada: A Research Perspective and Prospectus,' Chilly's Law Journal no. 1 (Feb. 1980), 50-9. 11 See H. Arthurs, J. Williams, and L. Taman, 'The Toronto Legal Profession: An Exploratory Survey,' University of Toronto Law Review 3 (1971), 498-528. 12 See MacFarlane, 'The Legal Profession in Canada,' 55, and Zemans, 'Legal Services in Canada,' 664. 13 Personal interviews with selected Nova Scotia judges, 1979-80. 14 Legal Aid, 1981,28. 15 This is certainly true of the Ontario Grange Commission. See Report of the Commission on Clinical Funding, app. c. 16 Dale H. Poel and Richard Conway, Legal Aid Applicants, Application Criteria and Application Outcomes (Halifax: Nova Scotia Legal Aid Commission and the Federal Department of Justice and Evaluation Research Associates, 1980), 22-5. Approximately 37 per cent of the NSLA applicants (in criminal, family, and civil matters) were previous NSLA clients. 17 See, for example, Zeman's, 'Legal Aid and Legal Advice in Canada.' 18 I prefer to reserve the term 'comparative' for research in which the covariation between factors or variables in some model of explanation is compared across selected units of analysis in such a way that the research can be systematically criticized and, in principle, replicated by other sharing the same substantive interest. Geographic place-names (for example, of the provinces and territories) may become last-resort, surrogate variables as a result of failure to measure more analytical features of the political system; they are not a first-choice measurement strategy. For a discussion of comparative research strategies see A. Przeworski and H. Teune, The Logic of Comparative Social Inquiry (New York: John Wiley and Sons 1970). 19 See Legal Aid, 81,9-14. 20 For an additional discussion of causality see H.M. Blalock Jr, Causal Inferences in Non-experimental Research (Chapel Hill: University of North Carolina Press 1964) especially ch. 3, 'Evaluating Causal Models.' 21 Cowie, 'Legal Aid in Canada- 1975,' 2. 22 Garth, Neighborhood Law Firms, 86.

Legal Aid 790 23 R. Penner, 'The Development of Community Legal Services in Canada,' prepared for Health and Welfare Canada (Montreal 1977), 14. 24 Legal Aid Services in Canada 1977-78: Comparison of Services (Ottawa: National Task Force on the Administration of Justice 1979), 4. 25 For a general discussion of the emulation of innovations, see E.M. Rogers, Diffusion of Innovations (Glencoe: Free Press 1962). 26 Harold J. Levy, address to the Sixth Commonwealth Law Conference, Lagos 1980, 4. 27 Cited in Garth, Neighborhood Law Firms for the Poor, 86. 28 A. Lazar, 'Legal Aid in the Age of Restraint' (Ottawa 1979), 3. 29 See overviews of the provincial plans in Legal Aid Services in Canada 1979/80, 7-86. 30 Garth, Neighborhood Law Firms for the Poor, 87. 31 A summary of the provisions of these agreements is found in Legal Aid Services in Canada 1979/80, 2. 32 See the summary of the provincial legislation in Legal Aid, 1981, 10-30. 33 Platiel, 'Storefront Clinics.' 34 The conclusion is based upon personal interviews and a review of Nova Scotia documents. See P. Aucoin, R. Conway, and D. Poel, Nova Scotia Legal Aid: A Review and Proposal for Evaluation (Halifax: Nova Scotia Legal Aid Commission and the Federal Department of Justice 1980). 35 The Canadian literature seems to assume that the judicare model is more expensive to implement than the staff lawyer alternative. Burns discusses the alternative findings in the American literature (see n. 5). The potential cost benefit of the staff lawyer system may depend upon the maintenance of the large caseloads criticized for limiting the scope of legal aid services. The B.C. public defender project should shed light on the issue of cost, and may contribute to the current arguments over preserving the province's judicare in criminal matters against pressures to implement a parallel public defender system: see Brantingham and Burns, n. 5. 36 Legal Aid, 1981, 16. 37 Lazar, 'Legal Aid in the Age of Restraint,' 11. 38 See comments in MacFarlane, 'The Legal Profession in Canada,' 50-1. 39 The comment was made at the 1981 Canadian Bar Association meeting in Vancouver and was carried in a Canadian Press report in Halifax Mail-Star, 2 September 1981,4. 40 This outcome was conveyed to the provincial plans in a letter to directors from the Department of Justice. 41 Legal Aid 1981 claims that 'Legal aid is now seen as a component of an effective judicial system rather than as a facet of social welfare' (9). 42 Legal Aid, 1981, 28.

Legal Aid 791 43 Discussions of federal support for civil legal aid are in process at the time of writing. 44 Introduction to the New Form of Application for Legal Aid (Toronto: Ontario Legal Aid Plan and the Ministry of Community and Social Services 1980). 45 Manitoba instituted such a user fee in April 1978 under pressure from the former Progressive Conservative government. It was abolished in May 1982 because of concern for the user fee's deterrent effect on potential applicants and because of administrative difficulties in collecting it: Ontario Legal Aid Plan, News Update 1982). 46 Legal Aid Services in Canada 1979-80, 4.

This page intentionally left blank

5/Technological Innovation and Industrial Policy: Canada in an International Context ARPAD ABONYI MICHAEL M. ATKINSON

The term 'industrial policy' covers a wide range of government measures aimed at altering the structure of an economy - that is, the manner in which resources are used and the resulting patterns of production and trade.1 A variety of discrete measures, including tariffs and competition legislation, contribute to an industrial policy, although these programs may not be comprehensive or coordinated to achieve designated objectives. A particularly important aspect of industrial policy is the programs and measures aimed at enhancing technological innovation. The contribution of innovation to economic growth has been appreciated for many years;2 this paper examines recent additions to our understanding of this phenomenon and considers the range of policy instruments open to governments seeking to enhance the innovative capacities of industry. The observations and conclusions of this paper are informed by the view that the emerging pattern of technological innovation in the world economy has important public-policy implications. The governments of industrialized countries and multinational corporations (MNCs) have sought increasingly to concentrate those activities associated with innovation-particularly research and development - in a very few countries. MNCs in particular have endeavoured to keep innovative activities in the parent firm while ensuring that the specialized production and assembly of component products is distributed on a world-wide basis. Behind these developments lies a desire to improve allocative and technical efficiency. The need to achieve economies of scale and reduce the fixed costs associated with research and development is an important factor prompting this global division of labour. In addition, cheaper labour costs in less-developed countries (LDCs) have increasingly led to the transfer of standardized production to industries located in those countries.

947 A. Abonyi and M.M. Atkinson

The policies of governments with respect to innovation need to be assessed in the context of this emerging global economy. The most pressing problem is that of creating an indigenous innovative capacity. The first section of this essay outlines the specific problems of innovation and assesses Canada's performance in relation to that of its major trading partners. Canadian industry, it is suggested, has chosen a defensive and adaptive approach to innovation which, in spite of certain economic advantages, may also bring about a gradual erosion of Canada's industrial base. The second section describes the policy instruments employed by the federal government to address the problem of innovation. A distinction is made between direct and indirect instruments, and Canadian policies are compared with those initiated by other countries that have endeavoured to create an indigenous capacity for innovation. Attention is drawn to Japan and to the countries of the Andean Pact and Eastern Europe that have sought explicitly to increase their capacity to adapt imported technology to their own ends. In addition to giving a critical assessment of the policy instruments employed by Canada, it is argued that successive governments at the federal level have limited themselves to a narrow range of instruments and have eschewed entirely the type of selective, sector-based approach to innovation that has been pioneered in other countries. The final section of the essay examines, in a preliminary fashion, some of the obstacles to the development of such an approach in Canada. Technological Innovation and Industrial Growth

The term 'technological innovation' is often confused with invention or technical change. Although there is no consensus on definitions,3 some distinctions can be made. For instance, inventions are new ideas for improving products or processes.4 Innovations embody inventions for commercial purposes. This includes the introduction of processes that result in a marked enhancement of efficiency or productivity. Innovations are related to technology in the sense that they usually evolve from systematic (scientific) organization of knowledge, relying on the steps from basic research to application that are associated primarily with the process of research and development (R & D).5 Technological change, which encompasses any change in a product or a production process (such as the adoption of an existing, but better, method of production), differs from technological innovation in that the latter embodies a new method or idea in the development of a product or process.6 For our purposes, then, 'technological innovation' means the creation of a new commercial product, process, system, or device through the application of scientific knowledge.

Technological Innovation and Industrial Policy 795

Technological innovation contributes to industrial development by introducing new products and by stimulating productivity, international trade, and economic growth.7 For new production processes to be efficient relative to existing techniques, they must either obtain greater output from the same factor inputs or raise productivity by reducing reliance on more expensive inputs, or do both.8 Given favourable circumstances, technological innovations tend to have a multiplier effect on the economy by inducing a series of consequential changes that raise the performance level and the growth potential of industries.9 At the company level, innovations help to produce better-quality products, enhancing the firm's competitive edge in international markets. The patented innovation is a particularly important source of market power, for it is unavailable to competitors except through licensing procedures that guarantee royalties for the innovating firm and control over the type of production to be undertaken by the licensee.10 Consequently, technologically innovative firms tend to develop greater independence with regard to what to produce, how to employ new products effectively, and how to extend the technology. Although not necessarily related to economic efficiency, this aspect of technological innovation provides a basis for avoiding possible dependence on the knowledge base of foreign competitors. Technological independence increases flexibility, which permits a wider range of development choices within an industry and reduces a firm's vulnerability to competition or changes in market demand. Problems in the generation and diffusion of technology Private firms and government enterprises may acquire technological innovations in two ways - by the generation of new products and processes from their own resources, or by the acquisition of products and processes from abroad. The major problem associated with the generation of technology is cost. Generation involves a significant expenditure and commitment to all aspects of inhouse and extramural R & D.'' Funds are channelled into pure scientific inquiry to produce inventions, into research to adapt inventions to commercially usable products and processes, and into development and design activities to exploit all production and commercial possibilities. To this end enterprises in knowledgeintensive industries, such as chemicals and electronics, employ a large scientific and technical infrastructure outside the firm to aid innovative efficiency.12 Approximately one-third to two-thirds of the knowledge input into innovation comes from outside the firm, mostly from universities and government-funded labs.13 Many enterprises, however, are unable or unwilling to pursue this type of offensive innovative strategy.14 They may lack sufficient capital, the scientific

967 A. Abonyi and M.M. Atkinson

personnel, or access to the necessary technical infrastructure. Many firms are reluctant to divert scarce resources from production activities for the purpose of 'reinventing' technologies available from other suppliers in the international economy, especially if they have traditionally found that the purchasing of technology, through licensing for example, is a convenient and profitable arrangement. Such acquisitions may result in the diffusion of technology, a process by which the use of an innovation spreads and becomes incorporated into production activities of firms other than the originators.15 This permits those firms to participate in the benefits of technical change without taking the risks associated with technological innovation. As companies assimilate borrowed technology, performance disparities between economic units may be reduced and technological gaps narrowed.16 Reliance on foreign technology is not a specific attribute of any particular group of countries: even those with high levels of R & D activity, such as the United States and Japan, import and adapt technology. Neither the generation nor the acquisition of new technology necessarily results in its effective absorption by the firm. Firms unable to pursue an offensive innovative strategy may also be incapable of assimilating new technology effectively. If they wish to share in the benefits of innovation, such firms must improve absorptive capacity by learning to use the new technology and developing the ability to improve on it. 17 The learning process involved in assimilating technology and converting it to innovative capacity is predicated on access to knowledge and on the availability of people who understand how to use it in the production process.18 When successful, technological apprenticeship or 'learning by doing'19 permits receiving firms to control the subsequent generation of technology and to ensure that it is appropriate to domestic needs, resources, and incomes. Several factors in the global economy complicate this diffusion process and make generation of new technology increasingly important. In the first place, industrial concentration in the world economy has grown.20 New technology, which is increasingly located in the corporation and which emerges through in-house R & D, is closely controlled by patents and other restrictive practices.21 This enables corporations, particularly MNCs which organize across national boundaries, to control the diffusion and transfer process and to limit market competition and dominate particular sectors of the economy.22 As a consequence, industrial R & D and innovative activity are now located in a smaller number of firms in industrialized countries.23 In addition to concentration, the contemporary world economy is distinguished by greater specialization. In order to utilize fully the market power derived from technological advantage and to cover the high costs of innovative activity, corporations, usually in the form of MNCs, are motivated to increase

Technological Innovation and Industrial Policy 797

world-wide sales, attain greater economies of scale, and increase specialization in production through the internationalization of operations. Internationalization, as Michalet defines it, 'consists of turning various subsidiaries of the MNC into production units specialized in a part of the finished product. The geographical distribution of the production process changes the affiliates into workshops and there is... strict coordination of the production flow of the various components in the industrial process.'24 Under these conditions the autonomy of subsidiaries is reduced in the sense that the nature of their output and the strict technical specifications on which that output is based are centrally determined within the corporation. This new internationalized specialization results in technical production standards being imposed by MNC headquarters on a worldwide scale. Given these aspects of the global economy, the successful extension of borrowed technology into a capacity for generation frequently does not occur. Critics have argued that the terms on which production is transferred abroad do not encourage the transfer of innovative capacity, that is, the ability to absorb imported technology. They point out that firms, especially MNCs, are preoccupied with preserving their monopoly advantages, not with transferring innovative capacity.26 These corporations are plainly not in the business of transferring technology except on terms advantageous to them. Only from this firm-centred perspective does it become clear why MNCs are insensitive to national goals and even government inducements in the area of technology transfer.27 In transfers not involving the subsidiaries of MNCs, importing firms face a series of obstacles. Licensing contracts frequently include stipulations regarding the actual use of technology. These can tie purchases of intermediate products, capital goods, and spare parts to the MNC and place restrictions on competition from importing firms.28 In addition, MNCs may not make available the most up-to-date technology,29 and the diffusion of that which is made available may be limited to importing sectors because it is incompatible with the technical levels of suppliers of domestic firms. Users also run the risk of adopting the specifications and standards of the exporting firm rather than developing their own which might be more appropriate to domestic markets and development goals.30 Consequently, even an investment in industrial R & D, if it is essentially product-oriented and adaptive in character, will not ensure that borrowers will avoid the trap of a never-ending cycle of adoption. The diffusion process is further complicated by the fact that the emerging structure of the global economy indicates the growing importance of organized R & D, and the generation rather than the acquisition of technology. The diffusion process, which has allowed some firms to acquire the benefits of R & D without investing in it directly, may no longer be sufficient to maintain a high level of

98/A. Abonyi and M.M. Atkinson

manufacturing productivity and a large share of manufacturing exports. With the advent of internationalization, homogeneous products, such as consumer durables and intermediate manufactured products produced by capitalembodied process innovation, have saturated Organization for Economic Cooperation and Development (OECD) markets. As a result OECD countries appear to be shifting production to technology-intensive goods, placing emphasis on the development of new products of higher quality and greater reliability and on exporting technology as well as skills.31 In order to be successful, however, this approach requires increased capability in R & D and product design. Countries and firms that previously relied on borrowed technology can no longer hope to realize increased manufacturing productivity and rising manufacturing exports shares without expanding indigenous innovative activity. Studies suggest that when domestic innovative capacities are insufficient, manufacturing productivity and export share will decline.32 If this is so, industrial development will have reached a new stage in which technology creation rather than aggregate demand is the most important source of economic growth. The problem in Canada The approach adopted by Canadian industry toward technological innovation has been described as defensive and adaptive.33 Instead of investing heavily in the generation of new products and processes, firms operating in Canada have tended to depend on the acquisition of technology from foreign sources. This phenomenon is manifested in two important ways. First, a great deal of technology is obtained from MNCs, primarily through intracorporate transfers but also by subsidiaries' purchases of technology-related services. Precise estimates of the value of these technology imports are difficult to make. Preliminary research indicates that American-based MNCs import more than three times as much R & D as they finance in Canada, and that for these firms the cost of this 'invisible R & D,' although declining, is more than twice their Canadian R & D spending.34 Not unexpectedly, the presence of a high foreign subsidiary share within an industry in Canada tends to lower its R & D intensity.35 Second, Canada's rate of investment in R & D is low by OECD standards and is declining. In 1972, when R & D spending comprised 1.12 per cent of GNP, the Senate Special Committee on Science Policy declared that a target of R & D spending amounting to 2.5 per cent of GNP was necessary to foster technological innovation.36 However, as shown in table 1, by 1979 total R & D spending had declined even further, to 0.94 per cent of GNP. Table 1 also indicates that Canada's present level of R & D funding as a percentage of GNP is among the

Technological Innovation and Industrial Policy / 99 TABLE 1 GERD as a percentage of GDP in selected OECD countries

United States Japan Germany France United Kingdom Italy Netherlands Switzerland Canada Sweden Belgium Australia Denmark Norway

1964

1970 1975 1976

3.14 1.47 1.41 1.84 2.32 0.65 2.03 2.21 1.07 1.20 0.97 n.a. n.a. n.a.

2.79 1.80 2.05 1.91 2.29 0.88 2.13 2.25 1.21 1.21 1.26 n.a. 0.95 1.10

2.44 1.94 2.22 1.80 2.12 0.93 2.12 2.40 0.97 1.75 1.33 n.a. n.a. 1.34

2.43 1.93 2.15 1.77 n.a. 0.86 2.07 2.45 0.92 n.a. n.a. 0.99 n.a. n.a.

1977 1978

1979

2.37 1.93 2.24 1.77 2.20 0.84 1.97 2.42 0.98 n.a. n.a. 0.90 n.a. 1.40

2.41 2.04 2.27 1.82 n.a. 0.84 1.98 2.45 0.9 1.89 1.40 n.a. 0.97 1.38

2.39 1.91 2.14 1.76 n.a. 0.89 1.99 2.29 0.93 1.87 1.37 n.a. n.a. 1.40

SOURCE: OECD, Science and Technology Indicators, Basic Statistical Series, vol. B: Gross National Expenditures on R&D 1963-1979 (Paris,OECD, DSTi/SPR/82.05, Jan. 1982) GERD = Gross domestic expenditure on R&D (natural sciences and engineering and social sciences and humanities) The countries are ranked in decreasing order according to the absolute level of domestic expenditures on R&D in 1979.

lowest of OECD countries. Federal government spending on scientific activities as a percentage of the budget has been decreasing since the early 1970s.37 This situation is aggravated by the fact that 50 per cent of the funds provided for R & D in Canada, a disproportionate amount by international standards, comes from government sources. Moreover, 80 per cent of the R & D funded by governments in Canada is also undertaken by governments. As a result, Canada does relatively more research and less development than other OECD countries.38 The portion of R & D spending financed by industry in Canada has increased marginally since 1963,39 but as table 2 shows, unlike that of Japan and Sweden, Canada's industry spending has not grown as fast as the industrial domestic product. Although the government has based its expectations for a higher national rate of R & D spending solidly on the private sector,40 there is little indication of how industry will be induced to increase its commitment. 41 The character of the R & D activity sustained by this defensive and adaptive posture has been criticized on the ground that it is truncated, failing to encompass a full range of functions - from the original research to all aspects of

100/A. Abonyi and M.M. Atkinson TABLE 2 R&D spending in OECD countries (as a percentage of GDP or GNP)

Canada Belgium France Germany Italy Japan Netherlands Sweden United Kingdom United States

1969

1971

1973 1975

1.2 1.1 2.0 1.7 0.8 1.4 2.0 1.3 2.3 2.8

1.2 1.3 1.9 2.1 0.9 1.6 2.0 1.5 2.1 2.6

1.0 1.3 1.8 2.0 0.9 1.7 1.8 1.6 2.5

1.0 1.2 1.8 2.1 0.9 1.7 1.9 1.8 2.1 2.4

1977 R&D in millions ofu.s.S

1977

1979

0.9 1.2

0.6

1,802

n.a.

1.8*

984 6,299 10,522 1,500 12,605 1,708* 1,500 44,744

2.0

1.4 1.9

0.9*

n.a.

1.7

1.4

1.9*

n.a.

1.9 2.4

2.0 1.9|

1.9

SOURCES: 1969-77- Statistics Canada Annual Review of Science Statistics 1980, (Ottawa Ministry of Supply and Services 1981, 23.) 1979-calculatedfrom OECD Bulletin, 1982. * 1976 f1978-79

marketing.42 Instead of exploiting opportunities for greater specialization and export, R & D in Canadian industry, particularly in foreign-controlled firms, has been primarily directed toward adapting imported technology for domestic use and achieving the goal of import substitution.43 Those who defend this pattern do so largely on the grounds of allocative efficiency. A small country like Canada, it is argued, derives greater advantage from appropriating the results of subsidized research performed elsewhere than it does from performing the research itself.44 Moreover, the difficulties associated with extracting rents from proprietary technology are best met by MNCs. Since their decisions with respect to the location of R & D will be governed by matters of proximity and cost, it is argued that Canadians will have to accept that 'nearly the only sort of R & D efficiently carried on in Canada is that aimed at adapting foreign technology to Canadian uses.'45 If adaptation is the most efficient policy, concern about the level of R & D spending is somewhat misplaced since the extent of industrial R & D effort is not necessarily an accurate measure of the ability to innovate. Innovations based on imported technology, particularly process innovations, are less expensive than those based on technology developed in-house.46 If, however, the capacity to

Technological Innovation and Industrial Policy/ 101

generate new technology becomes increasingly important to a country's competitiveness, reliance on adaptation may not be enough. A study prepared for the Economic Council of Canada, which examined 283 innovations, found both Canadian and foreign-controlled firms relying increasingly on technology developed in-house.47 Such a shift is significant since the performance of imported technology, in terms of spin-offs and subsequent export sales, was found to be distinctly inferior to that of technology developed in-house.48 While imported technology is an efficient means of imitating products and processes and of achieving import replacement, when technology is imported via MNCs (the manner in which most of it comes), it does not necessarily lead to anything except further importation. In particular, it does not enhance exports. This does not mean that foreign firms do not invest in R & D: overall, most of the R & D undertaken in Canada is done by multinational corporations.49 What is disturbing is that both Canadian and foreign-controlled firms, perhaps for different reasons, invest so little by OECD standards. Perhaps even more disturbing is the clear tendency for firms surveyed in the Economic Council study to rely increasingly on 'improvement innovations,' to be attracted to innovations with rapid pay-backs, and to avoid expenditures on major innovations.50 Whatever advantages might accrue from a trend to self-reliance in matters of innovation appear to be offset by a lack of acknowledgement on the part of firms in Canada of the trend's overall importance. It has been argued that the strategy of defensive adaptation, in which MNCs play a large role, has eroded Canada's competitiveness in manufactured products. The composition of Canada's merchandise trade balance has been well documented.51 Export surpluses in primary and partly manufactured products have traditionally offset deficits in the trade of highly manufactured endproducts, but since 1974 Canada has accumulated a current account deficit approaching $30 billion. This deficit shows no signs of disappearing in spite of productivity gains in Canadian industry.52 From 1955 to 1970 world demand sustained a tremendous increase in Canada's trade in manufactured goods, but by 1981 Canada's export share among industrialized market economies had dropped to 4.5 per cent.53 Countries such as Japan, which invested heavily in innovative activities during this period, continued to enjoy an expansion in their exports. Of particular importance is the fact that Canada's trade balance in technology-intensive manufactures deteriorated rapidly between 1965 and 1976; by 1976 Canadian imports of technology-intensive manufactures exceeded exports by approximately 70 per cent.54 If the adaptive strategy is working, it does not appear to be doing so in the context of manufacturing exports. While there are many other ways of evaluating the performance of Canadian industry,

1027 A. Abonyi and M.M. Atkinson

it is Canada's role in an emerging international division of labour that should command the most attention. There is more to Canada's present deficit in trade in manufactured products than an underinvestment in innovative activities. Increases in the cost of manufacturing, the decline of the dollar, and the long-term deleterious effects of tariffs and foreign investment have all made a contribution to both the trade picture and the scope and nature of the R & D performed in Canada. The debate over the relative importance of these factors and their relationship to one another continues.55 We have concentrated on the more immediate issue: the role of imported technology and a low level of industrial R & D with respect to Canada's trade position. The following section is devoted to government policy in these areas. Government Policies and Technological Innovation

The response of governments to problems in the generation and diffusion of technology depends in the first instance on assessments of the adequacy of the market as a means of allocating resources for innovation. Governments may adopt two distinct views. First, they may take the attitude that underinvestment in innovation is the result of market failure: a low level of innovative activity occurs because the value of technology to private producers is insufficient to generate a higher, socially desirable, level of investment. Governments respond to this problem not by usurping managerial discretion, but by attempting to induce firms to change their priorities. These efforts frequently take the form of subsidies intended to cover the difference between the income private producers can expect to obtain from R & D and its value to society. Since technological choices continue to be influenced by market forces, policy instruments premised on this approach will be referred to as 'indirect measures.' Governments that employ them are typically concerned with the overall level of research and development spending on the ground that the beneficial effects will eventually filter down to consumers. Alternatively, governments may determine that with respect to innovation the market itself is unresponsive to the goals of domestic development and that inducements are insufficient to ensure the level and type of innovation deemed desirable. Such a judgment is often based on perceived deficiencies in the existing industrial structure. A high degree of industrial concentration or foreign ownership, for example, may induce governments to intervene directly in the creation and utilization of technology within firms or across industrial sectors. Direct intervention may also result from scepticism about the efficacy of market

Technological Innovation and Industrial Policy / 103 forces in providing public goods. Enhancing manpower-training facilities or creating research centres is unlikely to come from normal market mechanisms. When direct intervention occurs, market forces are not suspended entirely, but policy instruments such as state enterprises and tripartite industrial boards may be used to change the market's complexion and inject a measure of political authority into the decision-making process.56 Whether policy instruments are being employed directly or indirectly is an empirical question and usually one of degree. Tax expenditures (provisions in the income tax system by which governments forgo tax revenues to encourage certain activities) and subsidies are normally employed indirectly, but these indirect steps may be marshalled in a larger effort, which includes state enterprise and rigorous regulation designed to transform the market structure. The taxonomy of policy instruments outlined below is offered with this caveat. The combination of instruments in a particular industrial sector is referred to as a 'sector strategy.' It will be impossible to survey all policy instruments. Attention is focused on the choice of measures whereby governments encourage the depackaging of technology, the improvement of engineering and design, increased R & D, and greater use of local technology. Direct and indirect instruments

Theoretically, governments can use either direct or indirect instruments, or both, to encourage the generation and absorption of technology. In the case of absorption, the most noteworthy feature of Canadian policy has been a heavy reliance on direct measures. Successful absorption requires firms to break down a particular technology into its core components, adapt it to local conditions, improve on it, and find local suppliers. Among indirect instruments designed to encourage absorption, loans and credits are the most common incentives. Peru, for example, passed the General Law of Industries in 1970 which established the Institute of Industrial Technological Research and Technical Standards (ITINTEC). The ITINTEC research fund subsidizes industry's attempts to adapt, modify, and absorb imported technology.57 Colombia has a similar arrangement under its Fund for Scientific Research and Special Projects (COLCIENCIAS).58 More ambitious mechanisms exist in Hungary, despite the fact that the country's economy is centrally administered. Loans are available from the national bank, and enterprises may obtain subsidies from the Ministry of Finance and the State Committee for Technical Development (SCTD).59 In Japan, firms receive loans from the National Technology Promotion Fund of the Japan Development Bank at favourable interest rates to facilitate application of new technologies.

104/A. Abonyi and M.M. Atkinson TABLE 3 Canadian federal government incentives for industrial R&D, all industries, 1981-82 (in millions of dollars) Program Indirect incentives (tax deductions) Special allowance Investment tax credit Direct financial and technological assistance Enterprise development program* (EDP) Special electronics fund (SEF) Microelectronics support program (MSP) Industrial research assistance program (IRAP) Technical information service (TIS) Mini-iRAP Science and engineering student program (SESP) Defence industry productivity programf (DIPP) Remaining programs Procurement (contracting) Contracting-out Remaining programs Departmental programs (An array of programs ranging from manpower training to development of energy conservation technology)

$ value 100.0

Per cent

14.0

61.1

8.6

14.4

2.0

37.8

5.3

131.6 24.2

18.4 3.4

160.0 25.0

22.4 3.5

159.7 713.8

22.4 100.0

SOURCE: Based on Federal Government Incentives for IndustrialR&D (Ottawa: MOSST Jan. 1982) * This program consists of two parts. The first is innovation assistance grants for selected projects concerned with development of new and improved products ($61.1 million in fiscal year 1981-82). The second part is support for companies to adjust to market changes. The fund also provides government guarantees of private loans for adjustment projects ($200 million in fiscal year 1981-82). t The grant (direct contribution) component in 1981-82 is $131.6 million. In addition, $20.0 million is budgeted for government loans.

Japan also employs depreciation measures for the purpose of promoting the utilization of new techniques.60 In all of these countries additional funds are available for firms involved in government-designated priority sectors. Of the many programs introduced by the Canadian government in the last two decades (see table 3 for the major categories), very few have been designed to promote indirectly the absorption of imported technology.61 Most have been aimed instead at the indigenous development of technology and the product design stage of the innovation process. An exception to this pattern has been the Defence Industries Production Program (DIPP) through which grants are made available to promote innovation in export-oriented military equipment and

Technological Innovation and Industrial Policy/ 105

which in 1981-82 accounted for 18.4 per cent of the total incentives offered by the federal government. DIPP permits the outright purchase of foreign technology with a gross sales criterion limiting the size of the grant. But this program is an exception. On the whole, the efficient and effective absorption of technology has not been a priority in spite of the fact that firms in Canada have generally assumed a defensive posture which calls for investment in innovation largely for purposes of acquisition from foreign sources. Prominent among the direct measures employed by most countries to facilitate the absorption of technology have been government laboratories and research institutes. As the experience of Colombia shows, training programs help strengthen the technical infrastructure needed to support absorption.62 In Peru, the government has established under the ITINTEC system research centres charged with aiding in applied research and finding solutions to technical problems of industry. The most serious difficulty associated with these efforts has been that of ensuring that the work of institutes and laboratories is co-ordinated with the needs of industry. In the Japanese case, government research centres have a thoroughgoing involvement with industry. In the early 1970s there were fifteen institutes and laboratories established under the Agency for Industrial Science and Technology (A 1ST) for research and experimentation in industry-related fields, eleven of which operated in fields specifically relevant to national interests.63 These research institutes are somewhat similar to the ones found in Eastern Europe, where substantial research activity exists outside firms along sectoral and functional lines. This means that several research institutes and laboratories operate under each ministry. For example, the Hungarian Ministry of Engineering and Metallurgy has over a dozen institutes dealing with such matters as transportation, metallurgy, and telecommunications.64 In Canada the transfer of technology from government laboratories and institutes to the private sector is important given the large amount of government R & D spending. Problems arise, however, largely because there is no natural synchronization between government efforts and industry needs. The typical sequence is one in which firms are presented with solutions for problems that they may have never considered.65 The lack of sensitivity to commercial application implied in this pattern is a primary reason for industries' often negative assessment of government efforts. Cordell and Gilmour add to this the fact that government laboratories have no mandate to work in support of the manufacturing sector.66 Consequently, few laboratories make a habit of promoting their capabilities or disseminating the results of their work. 67 The diffusion process is also hindered by the structure and performance of Canadian industry. With high levels of foreign ownership and generally low levels of R & D activity,

1067 A. Abonyi and M.M. Atkinson companies seem to be inadequately prepared to adopt the basic research of government laboratories, frustrating attempts at effective transfer. Where success has occurred, for instance in the telecommunication industry, it has been the result of recent efforts to satisfy specific government requirements and to transfer the technology to selected firms.68 Policies designed to stimulate the generation of technology are generally aimed at increasing the overall level of R & D performed in the economy. The primary direct instrument for these purposes is the complex of government research institutes and laboratories mentioned above; outlined below are the numerous indirect means. First, governments have attempted to increase the demand for industrial R & D by adopting contracting and purchasing policies that discriminate in favour of local suppliers. Brazil, Mexico, and Peru have pursued such policies with mixed success.69 The federal government's major thrust in this regard came in February 1972, when the government announced a 'Make-or-Buy' policy at the heart of which was the intention to contract out to industry much of the government's innovative needs.70 The program was considerably enhanced in 1974, when cabinet provided for the consideration and financing of unsolicited proposals, thus allowing industry a voice in the type of research that ought to be performed. Thousands of contracts have been awarded, and early assessments indicate that R & D performed under government contracts increased considerably between 1972 and 1976.7I None the less, inducing industry to undertake the R & D that would otherwise have been performed by government has done little to arrest the general decline in industrial R & D. This is partly because large firms are apparently unprepared to increase R & D resources or divert them from established commercial objectives for the sake of obtaining a government contract. The second indirect means of stimulating generation has been to offer firms various forms of credit, subsidies, loans, and tax incentives. Although the terms and amounts differ, all of the countries mentioned in connection with absorption programs also employ these measures. Canada has made a strong commitment to the support of generation.72 Traditionally, much of this support has come in the form of shared-cost programs administered by the Department of Industry, Trade and Commerce aimed at civil and defence research needs and at various aspects of product design and development. Although the range of programs was impressive, their purpose was never clearly established. Were they intended to support the ongoing research needs of industry, or were they short-term programs launched to entice firms into research and development in the expectation that a taste of the benefits would be sufficient to encourage a serious commitment of resources?73 Industry clearly preferred continuous support and objected to programs such as the Industrial Research and

Technological Innovation and Industrial Policy / 107

Development Incentives Act (IRDIA), which paid a portion of increases in R & D expenditures.74 In 1977 the Canadian government made several changes to these programs, not to reduce the level of funding, but to change the criteria for eligibility. Most of the original programs (DIPP being the major exception) were combined under the umbrella of the Enterprise Development Program (EDP) which, initially at least, had more funds at its disposal than the programs it replaced (see table 3). These funds are allocated under two general headings. The first, EDP adjustment, provides loans (at the prime rate plus 2 per cent) to help firms reorient technologically and to assist those that are losing their competitive position. It can be argued that this feature of the program represents a retreat from the original objectives of the former programs inasmuch as the loans provided are not necessarily destined to promote technological innovation, but rather to help declining enterprises 'develop.'75 EDP innovation (or 'positive adjustment') is the second category under which funds are distributed. This part of the program provides grant money up to 75 per cent for innovations. Firms must meet three criteria to qualify. First, EDP grants should be incremental, representing additional resources committed to the firm's existing strengths. Second, the project for which the EDP grants are awarded must place a 'significant burden' on the firm's resources: it should be a large undertaking relative to firm size. This means, in effect, that small and medium-sized Canadian firms are favoured, since MNCs normally do not experience problems meeting capital requirements. Third, the grant and the innovation must be used as instruments to build the company. This appears to be a non-specific, future-oriented criterion by which projects are chosen that are central to the growth plan of the firm. The most immediate effect of this criterion has been to force firms to develop a corporate plan. By 1978 changes had been made to the Income Tax Act. The benefits paid under the old shared-cost program, IRDIA, were brought into the act in the form of a tax allowance of 50 per cent of incremental R & D expenditures. A year earlier the investment tax credit had been extended to cover current and capital expenditures on R & D with a higher rate of credit for small businesses and firms in economically disadvantaged regions. These allowances and credits were added to an existing provision for the immediate write-off of 100 per cent of current and capital R & D expenses. The result of these measures has been a spectacular increase in tax expenditures devoted to research and development from $40 million in 1976 to $102 million in 1978, to an estimated $142 million in 1980.76 The costs involved underline the government's commitment to maintaining a high level of support for industrial R & D. But transfer through the tax system also illustrates the government's willingness to relinquish what little control it possessed under IRDIA over how some R & D funds would be spent.

108 / A. Abonyi and M.M. Atkinson

It is too early to assess the precise impact of these changes, but two general observations can be made. First, a firm's decision to spend or withhold R & D funds does not appear to be decisively influenced by the presence of shared-cost programs.77 Part of the reason may be that R & D financing in Canada has often been made available to firms regardless of the environment for innovation within the industry.78 It is not clear whether EDP will be used to meet an R & D spending target, in which case a broad and rather indiscriminate distribution of funds will continue, or whether the program's new discretionary features will channel funds into high-opportunity areas such as electronics where firms appear to be responsive to the funding programs.79 Second, the movement toward more elaborate criteria with respect to some EDP projects appears to be more than offset by the government's tax measures. On the surface at least there is no common framework informing the use of these policy instruments. The present system of tax expenditures is intended to meet the demands of industry for more funds and fewer controls. Unfortunately, the definition of scientific and technological activities for tax purposes encourages the inclusion of marginal expenditures under the general rubric. Moreover, these incentives are essentially windfall gains since there are no requirements attached to them. It should be added, in this regard, that the tax system can be employed as a direct instrument that requires R & D spending. Under the Peruvian ITINTEC system, for example, firms are obliged to devote 2 per cent of their before-tax income to R & D or else forfeit these funds to ITINTEC and the projects it sponsors.80 By comparison, the Canadian government's use of the tax system has not only been generous, but also benign. Toward sector strategies To this point the policy instruments designed to enhance the generation and absorption of technology have been discussed as discrete programs. These measures may be combined, however, in selective, sector-based strategies. The term 'sector strategy' refers to the co-ordination of direct and indirect policy instruments in a selected industry for the development of an indigenous technology base. This form of priority industrialization enables governments to establish, restructure, and promote further development of those industries that are responsive to domestic needs or export potential. The adoption of a sector strategy with respect to technology implies the existence of a more general objective: the achievement of industrial growth and efficiency by altering the industrial structure. The term 'industrial strategy' is frequently used to describe the co-ordination of various measures, including labour, foreign investment, and competition policies, to this end.81 Technology

Technological Innovation and Industrial Policy / 109 policy is thus made to serve more elaborate objectives. It is no longer sufficient merely to increase the level of industrial R & D; it must be demonstrated that the R & D performed is appropriate. The choice of sectors and the establishment of criteria defining what is 'appropriate' can be accomplished in several ways, but the pattern of decision-making with respect to R & D priorities is no longer left to the individual firm. As the Japanese experience reveals, the success of a sector strategy depends on the sector selected and the co-ordination imposed. In establishing the petrochemical industry, for example, the Japanese government was interested in securing a stable supply of intermediate raw materials, developing a substitute for increasing imports, and lowering the price of domestic petrochemical products.82 There are also other factors involved in selecting an industry, especially ones relating to the industry's spin-off potential. For example, the current Japanese sector strategy in computers has important technological benefits for both the telecommunications and electronics sectors.83 In the Japanese case, after an industry is selected, the Ministry for International Trade and Industry (MITI) is given responsibility for developing it. In co-ordinating the Petrochemical Development Program since 1955, MITI set out to establish an industry that would be capable of satisfying the entire domestic demand of the synthetic fibre and plastics sector.84 It first required the creation of new companies for the purpose of absorbing imported technology. Some companies merged in order to attain the necessary managerial and engineering skills. In addition, a state-owned synthetic rubber enterprise was established. The new firms were obliged to submit proposals detailing how they intended to depackage and absorb the new technology. In addition to forecasting sales for a five-year period, these companies were required to reveal the content of the technology package being imported, the terms of the contract (including the royalty payments involved), and the type of export restrictions, if any.85 So that firms could receive a larger share of the market and develop economies of scale, the Japanese government limited participation in the new petrochemical sector to four enterprises. As the first group became established and the sector developed, more firms were encouraged to enter the market. The government employed all the direct and indirect policy instruments already discussed to aid these firms in absorbing the new technology. In addition, it imposed a variety of import controls to protect the newly established industry and in 1964 set up an 'administrative guidance cartel' to regulate competition and co-ordinate investment. All of these initiatives, as Okita and Tamura point out, have had a favourable impact on absorption and have significantly decreased the price of domestically produced petrochemical products. It should be emphasized that this type of selective strategy, like attempts to promote

110 / A. Abonyi and M.M. Atkinson development of local technology in LDCs, is closely tied to government procurement policies, which play an important part in maintaining demand for goods.87 A selective strategy is unlikely to succeed if policy instruments are not coordinated with other measures within the framework of an industrial strategy. In the Japanese case this has been accomplished through administrative guidance - the government's authority to issue directives, warnings, and suggestions to enterprises and the general willingness of the latter to acknowledge that ministries like MITI speak for the national interest.88 The Japanese example is not unique. Other countries such as Hungary and those in the Andean Common Market have introduced sector strategies in bus manufacturing (Hungary) as well as machine tools and petrochemicals (Andean group).89 In contrast, the Canadian government has shown little inclination to experiment with sector-based strategies. In spite of the limited utility of each of the policy instruments discussed in the preceding section, the Canadian government has persisted in its efforts to enhance them as means of channelling aid to Canadian industry90 while avoiding the temptation to lead a restructuring of industry itself. A possible exception is the government's introduction of the National Energy Program (NEP) in 1980. However, the NEP falls short of the Japanese sector-based strategy because it is not co-ordinated with other policy instruments and is less concerned with technological innovation than with Canadianization and self-sufficiency. Similarly, the mega-projects strategy outlined in Economic Development for Canada in the 1980s lacks the selectivity and comprehensiveness implied in a sector strategy. In 1978 the federal government established twenty-three sector task forces, made up of business, government, and, to a more limited degree, labour representatives, to discuss the problems of specific industrial sectors and apparently to devise sector strategies.91 These sector studies were the first stage in a two-tier operation: the second-tier exercise was to draw sector recommendations together and offer a synthesis that cut across sectoral lines. These events marked a departure from existing government policy only in the sense that they represented a comprehensive consultative effort. The government did not signal its intention to transform the existing market structure in any particular way, although the sector profiles provided by ITC and the Department of Finance did contain implicit and explicit criticisms. The sector task forces responded with a barrage of criticisms of their own. Their recommendations were designed in almost every case to cater to the demands of specific industries. Tax credits, the elimination of regulation, and an enhanced procurement policy figured heavily in many of the task force recommendations. The second-tier exercise was handicapped by the absence of common points of reference in the first-tier

Technological Innovation and Industrial Policy /111

reports. Some uneasy compromises were reached, but no recommendations regarding industrial rationalization were made. The government, by providing a minimum of direction, bears a great deal of the responsibility for the outcomes. But the rudimentary state of official consultative mechanisms in Canada probably dictated the format and the results. The problem is aptly summarized in a recent study prepared for the Science Council: The dilemma is that a top-down process in which governments assert clear policy objectives and seek only advice on how to implement them risks being damned as illegitimate intervention in the private sector. But a bottom-up process, in which the private sector is simply invited to 'tell us what you want' is going to produce an uncoordinated and incoherent stream of proposals which provide something for everyone but avoid the hard choices.92 Obstacles to Sector-Based Strategies in Canada

Canada's approach to the stimulation of innovation has been predicated on direct mechanisms (with their acknowledged deficiencies) and on the avoidance of sector strategies. Why this should be so while the problems of innovation in Canada deepen is a complex political question. Neither those who would strip away the existing programs and seek rationalization through a reduction in tariffs93 nor those who advocate a policy of direct government intervention94 have been able to account for the Canadian government's persistent adherence to its present policy direction.95 In this section we discuss in preliminary fashion some of the factors that have induced the existing pattern of programs and in particular have discouraged in Canada the employment of sector strategies. We begin with some of the economic problems posed by an effort to enhance industrial innovation, and then turn to two important political factors: the manner in which the bureaucracy is presently organized to face these problems and the absence in Canada of a constituency prepared to support the initiatives implied in a sector-based strategy. The identification of these factors as important ones is not intended to rule out others. In particular, the traditional role of the state in economic development will condition to some degree the willingness of governments to intervene. In Japan, for example, formal and informal links between government and industry are the product of an institutional heritage originating in the feudal period.96 This is not meant to imply that in Japan conflict does not exist among industrial interests or between and within state agencies. Indeed, there are frequent differences with regard to what sectors and technologies should be promoted and the

112/A. Abonyi andM.M. Atkinson

appropriate policy instruments to employ./Nevertheless, there is widespread acceptance in industry that it is the guiding role of the Japanese state bureaucracy that secures long-term economic growth.97 Government-industry relations in Canada have not been based on the same understanding of the state's role. Examples of direct intervention are numerous, particularly in sectors such as transportation and energy, but the Canadian state has never become a legitimate director of private-sector investment decisions. Bearing this general observation in mind, this section concentrates on some of the more immediate obstacles to sector strategies. The general problem has been one of mobilizing support for a policy that could significantly disrupt existing relationships. Neither the bureaucracy nor the private sector in Canada is united behind a strategy of industrial development that might inform policies with respect to technological innovation. On the contrary, the existing panoply of programs has created a set of mutually reinforcing interests wedded to the status quo with minor variations. These problems are compounded by the circumstances of the Canadian economy. The scope of the problem One of the main objectives of a sector strategy is to change the existing industrial structure to enhance the prospect that technological innovation will take place. But the size of the domestic market is an important constraint. The Japanese strategy of fostering the growth of a domestic market before making an international challenge is difficult to duplicate in a country with one-fifth the population of Japan. The small Canadian market and the presence of unexploited economies of scale have been major objects of concern for almost all observers of secondary manufacturing in Canada.98 Critics argue that industry in Canada suffers from relatively small plant size and short production runs. A great diversity of products is manufactured, particularly by foreign firms, but in quantities too limited to realize economies of scale. The presence in Canada of MNCs and a high proportion of technology from foreign sources also influences how much and how effectively local technology can be developed. MNCs can fragment industrial strategies and hinder government efforts to integrate the various policy instruments designed to stimulate indigenous technology. The internationalization of technology within MNCs means that they will normally be unresponsive to national industrial priorities. In general, MNC operations are oriented toward neutralizing the competitive position of domestic firms, in some cases through outright purchase. To compensate for small internal markets and offset the effects of MNCs, some analysts have suggested that large 'world-scale corporations' are viable vehicles for the purpose of an industrial strategy.99 Others have countered that large

Technological Innovation and Industrial Policy 7113

firms may be inefficient and their structure and organization may impede innovation. Small, medium, and large firms all have a role to play in the innovative process: 'No single firm size is uniquely conducive to technological progress ... what we want therefore may be a diversity of sizes, each with its own special advantages and disadvantages.'100 However, if sector strategies are introduced in Canada, large firms will be a necessary ingredient not only to neutralize the market power and extensive resource base of MNCs, but also to stimulate specialization and economies of scale in production activity so that the firms can go beyond the domestic market and capture economic gains from exportoriented growth. Whether governments are at all capable of successfully moulding the present industrial structure is open to question on two fronts. First, sector strategies imply that choices must be made with respect to sectors, firms, and the desired industrial structure. Critics have warned that governments are entirely unsuited for this task. The capacity of economic analysts, let alone governments, to identify key sectors has been challenged on the ground that social returns on technological investments are extremely difficult to calculate.101 Moreover, the capacity of bureaucrats to select particular firms or sections for preferential treatment (to 'pick winners') is widely considered to be seriously deficient.102 Without contesting the general view that government bureaucracies are not ideally suited to make the decisions required, it must be acknowledged that the history of government enterprise in Canada is not an unbroken story of failure and defeat. Criticisms of government involvement are frequently premised on the view that bureaucracies operate entirely separate from industry, making decisions with little or no consultation. This is hardly ever the case. What is clear, however, is that success in coping with the problems outlined will depend to some degree on whether consultative mechanisms capable of drawing participants together and facilitating decision-making are in place and whether both bureaucratic and political resources are committed to the strategy. An even greater obstacle is the penchant of politicians for picking losers. This is done not out of ignorance but to preserve the employment and economic relationships that exist. These are much more tangible than the jobs and opportunities that may or may not result from a policy of leading the process of rationalization. Certain aspects of the Enterprise Development Program are clearly intended to guarantee economic security to the greatest degree possible. Unfortunately, as Thurow has recently pointed out, 'this guarantee locks us into current activities and makes it difficult to shift to the new products and processes that are at the heart of economic progress."103 Whether such actions are prompted by a crude electoral calculus, or whether they suggest the role of the state as the guarantor of a high rate of profitability for capital, the result is an uneven and contradictory approach to innovation.

114/A. Abonyi and M.M. Atkinson A fragmented bureaucracy At the federal level Canada does not have a department of government charged with drawing together science and technology policy on the one hand and economic policy on the other. Instead, several departments share these responsibilities, and each has developed a separate view of the contribution of innovation to economic performance and the best means of promoting industrial innovation. This division of responsibility, combined with departmental rivalries and unequal bureaucratic resources, ensures that obstacles will thwart all but the most innocuous efforts to adopt a strategic approach to innovation. Recent organizational changes - the creation of a new Department of Regional Industrial Expansion, the creation of the Ministry of State for Economic and Regional Development, and the radical restructuring of the Department of External Affairs-may eventually overcome some of these obstacles.104 An attitude of cautious pessimism is most appropriate, given the recent history of bureaucratic relations. The Ministry of State for Science and Technology (MOSST) was created in 1971 and was given responsibility for the formulation and co-ordination of science policy. But the absence of particular program or control responsibilities has forced the ministry out of central decision-making circles.105 Its major success has been Make-or-Buy, but as French and Aucoin have observed, this policy 'is based on only the vaguest vestige of a more comprehensive policy direction.'106 Without responsibility for industrial programs in the area of innovation, MOSST is left with the task of assessing scientific and technological possibilities and acting as a liaison between government and a disparate scientific community. The Department of Industry, Trade and Commerce (ITC) has held a highly unstable amalgam of industrial and trade responsibilities since its creation in 1968. ITC has undergone numerous organizational upheavals,107 and observers have concluded that the Industrial Development Branch of the department has never been able to challenge successfully the department's export promotion philosophy with which it has been forced to coexist.108 Line branches on the industrial side of the department have developed strong clientele relations with those sectors of industry whose incentive programs they administer. This community of interests has defied policy co-ordination in ITC just as it has in other departments responsible for industrial development.109 The new Department of Regional and Industrial Expansion is intended to strengthen the regional focus of industrial development programs and will do so without the presence of International Trade in the same departmental structure. It is by no means clear, however, that the problems of coherence and co-ordination that gave rise to a reorganization will be solved by combining regional development and industrial expansion in the same portfolio.

Technological Innovation and Industrial Policy 7115

Even if ITC had been committed to sector strategies, it is unlikely that it would have had much success in dislodging the Department of Finance as the government's primary articulator of economic policy. Finance has resisted the development of economic expertise elsewhere in the government and has successfully promoted the economically orthodox attitude of non-intervention in the economy. The creation of the Ministry of State for Economic Development (MSED) in 1978 did signal the government's desire to impose some order on the process of economic decision-making in Ottawa. But MSED did not become a planning centre. Instead, it assumed the responsibility for administering the economic development envelope and performing the function of a central agency attached to the cabinet committee on economic development. Policy formulation at MSED took the form of developing an assessment of the country's economic prospects (a 'medium-term track') based on both Finance and ITC economic models. Its successor, the Ministry of State for Economic and Regional Development (MSERD), promises to be a much more decentralized ministry charged with co-ordinating and reconciling national economic policy and regional aspirations. With the apparent demise of economic-development-by-megaproject,this may prove a rather difficult task. If MSERD is to realize its potential for directing industrial policy it will require an agency ideology more complete than the megaproject strategy and strong political support to compensate for an absence of programmatic responsibilities. The bureaucratic power relations of the late 1970s permitted only incremental, pragmatic adjustments to Canada's industrial policy.110 the new relationships of the 1980s are aimed at neutralizing some bureaucratic rivalries and making economic policy more sensitive to regional aspirations. It is unclear, however, whether they will constitute a break with the past as long as responsibility for the development of an industrial policy continues to be fragmented and divided among a host of government departments. The absence of a supportive constituency An approach to innovation based on sector strategies would represent a major departure in current government policy. This is unlikely to occur without the development outside of government of a strong constituency interested in demanding change and capable of assisting in the development of policy. Some economists have suggested that such a constituency is already present in the form of Canadian scientists, engineers, and the middle class at large, all of whom would profit from the development of an indigenous capacity for innovation. They suggest that the nationalism implicit in such an approach, and an emphasis on generation instead of acquisition, would detract from allocative efficiency.111 The ultimate losers, it is asserted, would be consumers and the working class.112

1167 A. Abonyi and M.M. Atkinson

In spite of the supposed distributional effects of nationalist policies, interest in them appears limited to those Canadian firms, such as SPAR Aerospace and Northern Telecom, already deeply committed to research and development. The community of Canadian scientists and engineers is a small and disparate one, insufficiently concentrated to carry much political weight.113 The 'middle class' at large has evinced no great interest in the topic. Of course, the benefits to scientists and members of the middle class may be entirely illusory. It would be exceedingly difficult to offer a definitive demonstration of the distributional impact of sector strategies.114 More important, it is misleading to speak of the middle class in Canada as a class sufficiently coherent and united to constitute a supportive constituency. Even a 'national economic elite,' only a slightly more precise concept, is unavailable for these purposes. There are at least three reasons why no such supportive constituency exists. First, the Canadian economic elite has never been a single unified body of finance and industrial capital. Although disputes over the degree of division continue, Canada's indigenous economic elite has concentrated on what Clement has called the 'circulation sector,' the financing of transportation and utilities systems to Canada and elsewhere. The bulk of the manufacturing sector has been left to foreign corporations. Canadians wishing to pursue careers in manufacturing are frequently absorbed into the Canadian subsidiary operations of MNCs, and constitute what Clement has called a 'comprador elite.'115 Their futures are played out within a continental context. This process has evoked no great dismay. As Rotstein has put it, 'No other business class in an advanced industrial society has presided so gracefully over its own liquidation.'116 The integration of the national economic elite into the larger MNC, which Stephen Hymer anticipated some time ago,117 has reached perhaps its highest level of development in Canada. Second, any rationalization of Canadian industry to realize expected gains in innovative capacity will almost certainly require centralized decison-making authority. The federal government is in the best position to act as this authority to the degree that economies of scale are deemed important, spillover effects are great, and there is a need to preserve existing locational advantages (in Central Canada) against the claims of the peripheries.118 It is likely that the restructuring of Canadian industry implicit in sector strategies would mean that all of these conditions would apply. To allow provincial governments to embark upon their own strategies in the face of this assessment would only mean prohibitive costs and the ever-present likelihood of competitive rather than collaborative strategies. For the provinces to accede to federal direction would require provincial governments to accept that decisions regarding the future of industries located in

Technological Innovation and Industrial Policy 7117

particular provinces would, at best, be shared with the federal government. The likelihood that provincial governments would find this acceptable has diminished considerably in recent years. Provincial governments and provincially owned utilities have adopted a series of policies designed to discriminate in favour of local (that is, provincial) producers. These policies, which include discriminatory government purchasing and employment programs, further fragment an already small domestic market by erecting non-tariff barriers to the movement of goods, capital, and labour across provincial boundaries.119 The provinces have added to this their own initiatives in the fields of microelectronics, computers, and biotechnology.120 Provincial governments may not have carefully considered industrial strategies, but the phenomenon of provincebuilding has reached the stage that provinces do feel free to alter market forces in the interests of provincial or regional objectives. Finally, to reduce the enormous information costs associated with a strategic approach to innovation and to ensure compliance, governments require the active collaboration of business and labour. It is necessary that a supportive constituency exist, and that this constituency actively participate in the required restructuring. Without prior commitments of this sort very few governments will be induced to absorb the short-term costs of intervention (including the inevitable dislocations) for elusive long-term gains. In Canada there is only the weakest of corporatist traditions on which to erect the required collaborative structures. Observers of the consultations between governments and the private sector confirm that discussions were limited largely to information exchange and produced no consensus on overriding goals.121 Participants were suspicious of being co-opted into support for government positions, and both labour and business expressed the view that responsibility rested with elected officials. In any event, neither business nor labour in Canada is sufficiently inclusive or authoritative to bind a large constituency to negotiated arrangements, even if this were their inclination. Finally, it is unlikely that the development of corporatist structures is merely a matter of business, labour, and government getting to know one another. There is nothing inevitable about the development of these structures and they are by no means easy to legitimate and maintain.122 Conclusion

We have argued in this essay that the contribution of technological innovation to industrial growth is likely to increase. In response to changes in the world economy, several countries have adopted a sector-based approach to innovation in which direct and indirect measures are brought to bear on particular sectors of

1187 A. Abonyi and M.M. Atkinson

the economy, consultative mechanisms are established to link business and government, and adjustments are made to the existing market structure to realize gains in innovation and to facilitate the effective absorption of technology. The Canadian government has not endorsed such an approach. A variety of policy instruments have been employed by the federal government, but not within the context of an overall industrial strategy. These instruments appear to have had little effect on the general deterioration of industrial innovation in Canada. Canadian industry has adopted a defensive and adaptive approach, but Canadian government policies do not complement this orientation. The policies that have been adopted have resulted in a considerable amount of R & D being undertaken by government, and very little change in industry's contribution. A principal objective of the essay has been to place the Canadian experience in a larger international context. Indeed, one of the initial requirements in this policy area is to establish just what the problem is. Canada's current level of innovative activity is a fact, not a problem. Before it can be treated as a problem, a set of propositions, both normative and empirical, is required, and the first part of the paper sought to offer an initial formulation. An underlying and unexamined reason for Canada's response may be the fact that the problem as we have outlined it is not the problem as defined by bureaucrats and politicians in Canada. Very little is known about the attitudes and beliefs of crucial actors in the process and much more information will be needed before firm conclusions can be drawn. In its absence we have suggested several factors that might have contributed to the rather narrow and conservative policies that have emerged. A clearer appreciation of their importance will come when Canada is examined in a comparative context and when efforts are made to map the attitudinal structures of senior officials and politicians. NOTES We would like to thank Liz Smythe, Steven Langdon, Bill Coleman, Jeanne Kirk Laux, and our fellow contributors to this volume for their helpful comments on an earlier draft of this paper. We would also like to acknowledge the assistance of officials in the Ministry of State for Economic Development, Industry, Trade and Commerce, the Science Council of Canada, and the Ministry of State for Science and Technology, all of whom gave us a considerable amount of their time. 1 John Diebold, Industrial Policy as an International Issue (New York: McGrawHill 1980), 6. 2 The work of Joseph Schumpeter was an outstanding contribution in this regard. See his Business Cycles, vol. 1 (New York: McGraw-Hill 1939). For a good

Technological Innovation and Industrial Policy / 119 review and analysis of the relationship of technological innovation to economic growth, see P.M. Scherer, Industrial Market Structure and Economic Performance (Chicago: Rand McNally Co. 1980), 407-57. 3 K. Norris and J. Vaizey, The Economics of Research and Technology (London: Allen and Unwin 1973), 11-18, 76. 4 Christopher Freeman, The Economics of Industrial Innovation (Harmondsworth: Penguin 1974), 22. 5 Norris and Vaizey, Research and Technology, 11-18; Freeman, Industrial Innovation, 14-32; and P. Bourgault, Innovation and the Structure of Canadian Industry (Ottawa: Science Council of Canada 1972), 23. 6 T.H. Hill and J.M. Utterback, Technological Innovation for a Dynamic Economy (New York: Pergamon 1979), 3. 7 Organization for Economic Co-operation and Development (OECD), Technical Change and Economic Policy (Paris: OECD 1980), passim; Scherer, Economic Performance, 407-38. 8 Hill and Utterback, Dynamic Economy, 1-12. 9 J. Schmookler, Invention and Economic Growth (Cambridge: Harvard University Press 1966). 10 C. Vaitsos, 'The Process of Commercialization of Technology in the Andean Pact,' in International Firms and Imperialism, ed. H. Radice (Harmondsworth: Penguin 1975). 11 Extramural R & D is carried on outside the firm, usually in private, government, or university laboratories and research centres. 12 Freeman, Industrial Innovation, 263-9. 13 K. Pavitt, 'Government Policies towards Industrial Innovation: A Review of Empirical Findings,' Omega 5, (1976), 539-58. 14 An offensive innovative strategy requires a very considerable investment in intramural R & D and an advanced capacity in product design and engineering. Firms pursuing such a strategy are typically linked to sources of basic research. For a more complete discussion of this strategy and others, including defensive, imitative, and dependent strategies, see Freeman, Industrial Innovation; and in the Canadian context, James Gilmour, 'Industrialization and Technological Backwardness: The Canadian Dilemma,' Canadian Public Policy 4 (Winter 1978), 20-33. 15 J.E.S. Parker, Economics of Innovation (London: Longmans 1974), 99-124. A recent example is Bombardier Inc., a Canadian company that has purchased licences for the production of mass transit equipment and is now selling its products in Mexico and the United States. See New York Times, 24 May 1982. 16 OECD, Technical Change, passim. 17 F.R. Sagasti, Technology, Planning, and Self Reliant Development (New York: Praeger 1979), 86.

1207 A. Abonyi and M.M. Atkinson 18 L.K. Mytelka, 'Technological Dependence in the Andean Group,' International Organization 32 (Winter 1978), 101-40; Janta del Acuerdo de Cartagena, Andean Pact Technology Policies (Ottawa: International Development Research Centre 1976) 7. 19 These concepts are advanced in the work of Kenneth Arrow, 'The Implications of Learning by Doing,' Review of Economic Studies 39 (June 1962); and E.M. Rogers, The Diffusion of Innovation (New York: The Free Press 1962). 20 United Nations (U.N.), Multinational Corporations and World Development: A Reexamination (New York: United Nations 1978), 34-5. 21 The greater role of the corporation in innovative activity is confirmed by data from the United States which show that the share of patents obtained by corporations increased from 19 per cent in 1908 to 70 per cent in 1973. See S.P. Magee, 'Multinational Corporations, the Industry Technology Cycle and Development,' Journal of World Trade Law 11 (July-Aug. 1977), 297-321. Similar data for Britain reveal that patent applications from corporations rose from 15 per cent in 1913 to 68 per cent in 1955. See Parker, Innovation, 27. Although inventors often remain outside the large - and frequently multinational - firm, R & D is still necessary to turn the inventions of these idea people into innovations. If innovations occur in smaller firms, they are often made possible by scientists and technicians with previous experience in large firms. See K. Pavitt, 'The Multinational Enterprise and Transfer of Technology,' in The Multinational Enterprise ed. J.H. Dunning (London: George Allen and Unwin 1971), 62. For a critical assessment of how this corporate innovative activity affects technology commercialization, see E.A. Nadal, 'Multinational Corporations in the Operation and Ideology of International Technology Transfer,' Studies in Comparative International Development 10 (Spring 1975), 11-29; and Vaitsos, 'Commercialization,' passim. 22 Nadal, 'Multinational Corporations,' passim; C.A. Michalet, 'The International Transfer of Technology and Multinational Corporations,' Development and Change 8 (1976), 154-74; and U.N., Measures Strengthening the Negotiation Capacity of Governments in Their Relations with Transnational Corporations: Technology Transfer through Transnational Corporations (New York: U.N. Centre for Transnational Corporations 1979), 16. 23 L.K. Mytelka, 'The Lome Convention and the New International Division of Labour,' Journal of European Integration 1 (1977), 63-76. Such concentration is especially the case in knowledge-based industries such as Pharmaceuticals, chemicals, engineering, and electronics, which are characterized by costly and complex technologies. For analysis see Freeman, Industrial Innovation, 161-71. 24 Michalet, 'International Transfer,' 170-1. 25 Ibid.; Nadal, 'Multinational Corporations,' 13; and Michalet, 'International Transfer,' 167-8.

Technological Innovation and Industrial Policy 7121 26 Mytelka, 'The Lome Convention,' passim. 27 This perspective also emerges from what has been called the theory of internalization. In Canada its chief proponent has been Alan M. Rugman, Multinationals in Canada (Boston: Martinus Nijhoff 1980), and 'Research and Development by Multinational and Domestic Firms in Canada,' Canadian Public Policy 1 (Autumn 1981)604-16. 28 See United Nations Conference on Trade and Development (UNCTAD), Towards the Technological Transformation of Developing Countries, (Manila: U.N. 1979); and UNCTAD, Major Issues Arising from the Transfer of Technology to Developing Countries (New York: U.N. 1975), 165-70. 29 J. Szita, Perspectives for All-European Economic Cooperation (Budapest: Akademiai Kiado 1977), 170; and Nadal, 'Multinational Corporations,' passim. 30 See G.K. Helleiner, 'The Role of the Multinational Corporations in Less Developed Countries, Trade and Technology,' World Development 3 (1975), 161-90; F. Stewart, Technology and Underdevelopment (London: Macmillan 1977). 31 K. Pavitt, 'Technical Innovation and Industrial Development: The New Causality,' Futures 11 (Dec. 1979), 458-70. 32 Ibid., 466. 33 Richard E. Caves, Michael E. Porter, A. Michael Spence, with John T. Scott, Competition in the Open Economy: A Model Applied to Canada (Cambridge: Harvard University Press 1980), 382. 34 Ministry of State for Science and Technology (MOSST), Importation of Invisible Research and Development (Ottawa: Supply and Services 1978), 5-6. 35 Caves et al., Competition in the Open Economy, 193. 36 Senate Special Committee on Science Policy, A Science Policy for Canada, vol. 4 (Ottawa: Queen's Printer 1970), 499. 37 See also the statistical series reproduced in Donald G. McFetridge, 'Research and Development Expenditures,' in How Ottawa Spends Your Tax Dollars, ed. G. Bruce Doern (Toronto: James Lorimer and Co. 1981), 287-94. 38 Bourgault, Innovation, 62. 39 MOSST, Overview of Research and Development in Canada (Ottawa: Supply and Services 1978), 7. 40 Financial Post, 25 October 1980. 41 McFetridge, 'Research and Development,' 263-4. 42 John N.H. Britton and James M. Gilmour, The Weakest Link: A Technological Perspective on Canadian Industrial Underdevelopment (Ottawa: Science Council of Canada 1978), 96. 43 MOSST, R&D in Canadian and Foreign-Controlled Manufacturing Firms (Ottawa: Supply and Services 1979), 39. For a detailed analysis of the balance between imported and in-house technology in foreign- and Canadian-controlled firms,

122/A. Abonyi and M.M. Atkinson see Dennis P. DeMelto, Kathryn E. McMullen, and Russel M. Wills, Preliminary Report: Innovation and Technological Change in Five Canadian Industries, (Economic Council of Canada, discussion paper no. 176, mimeo 1980), 105-22, 22560. The argument that the origins of technological dependence are to be found in the nineteenth century, particularly the 1872 Patent Act and the 1879 National Policy Tariffs, is made by Glen Williams, 'The National Policy Tariffs: Industrial Underdevelopment through Import Substitution,' Canadian Journal of Political Science 12 (June 1979), 333-68, esp. 346-54. 44 Steven Globerman, 'Canadian Science Policy and Technological Sovereignty,' Canadian Public Policy 4 (Winter 1978), 34-45; and D. J. Daly and Steven Globerman, Tariff and Science Policies: Applications of a Model of Nationalism (Toronto: Ontario Economic Council 1976). 45 Caves et al., Competition in the Open Economy, 176-381. 46 DeMelto et al., Innovation and Technological Change, 35. 47 Ibid., 262. 48 Ibid., 105-12. 49 Herman P. Bones, 'Are Foreign Subsidiaries More Innovative?' Foreign Investment Review 3 (Spring 1980), 20-3. DeMelto et al., Innovation and Technological Change, found that while the Canadian firms in their sample spent more on basic research for product innovations, overall the foreign-controlled firms spent significantly higher amounts on R & D: 114. Compared with their U.S. counterparts, however, subsidiaries spend less, suggesting to Caves et al., Competition in the Open Economy, that 'the foreign firm is an accomplished practitioner of R & D, but primarily for the adaptation of technology imported from abroad': 381. 50 DeMelto et al., Innovation and Technological Change, 226, 248-9. 51 For example, Economic Council of Canada, Looking Outward (Ottawa: Supply and Services 1975), 18-22. 52 Donald J. Daly, 'Weak Links in the "The Weakest Link,"' Canadian Public Policy 5 (Summer 1979), 307-17. 53 Fred Lazar, 'An Industrial Strategy within the Context of Canada-u.s. Trade Relations,' prepared for the Center of International Affairs, Harvard University, April 1982, 2. On the mid-1970s see Britton and Gilmour, Weakest Link, 41. 54 MOSST, Canadian Trade in Technology-Intensive Manufactures (Ottawa: Supply and Services 1978), 13-14. 55 Britton and Gilmour, Weakest Link; Daly, 'Weak Links'; A.E. Safarian, 'Foreign Ownership and Industrial Behaviour: A Comment on "The Weakest Link,"' Canadian Public Policy 5 (Summer 1979), 318-35. 56 For a discussion of exchange, authority, and persuasion, see Charles Lindblom, Politics and Markets (New York: Basic Books 1977). 57 Sagasti, Self Reliant Development, 124.

Technological Innovation and Industrial Policy / 123 58 A. Araoz, Science and Technology for Development: STPI Module 5 (Ottawa: International Development Research Centre 1980), 38-41. 59 A. Abonyi, 'Imported Technology, Hungarian Industrial Development and Factors Impeding the Emergence of Innovative Capability,' in Hungary: A Decade of Economic Reform, ed. H. Radice, P. Hare, and N. Swain (London: Allen and Unwin 1981), 150-1. 60 OECD, Industrial Policy in Japan (Paris: OECD 1972), 88. 61 See Daly and Globerman, Tariff and Science Policies. 62 Araoz, Science and Technology, 40. 63 OECD, Japan, 86. 64 See Abonyi, 'Imported Technology,' 148-9. 65 A.J. Cordell and J. Gilmour, The Role and Function of Government Laboratories and the Transfer of Technology to the Manufacturing Sector (Ottawa: Science Council of Canada 1976), 241. 66 Ibid., 240. 67 In June 1978 the government announced plans to expand the NRC'S Industries Laboratory Projects and Technical Information Services. McFetridge, 'Research and Development,' 269; MOSST, Measures to Strengthen Research and Development in Canada (Ottawa: Supply and Services 1978). 68 Bohumir Pazderka, 'Innovation in the Canadian Telecommunications Industry' (Kingston: Queen's University, mimeo 1982). 69 F.R. Sagasti, Science and Technology for Development: Main Comparative Report of the Science and Technology Instruments Project (Ottawa: International Development Research Centre 1978), 75. 70 Further measures to enhance procurement policies were announced in June 1978. See MOSST, Measures. 71 MOSST, The Make or Buy Policy (Ottawa: Supply and Services 1976), 16-17. 72 Bourgault, Innovation, has described the level of government support for commercially oriented R & D in Canada as 'unmatched in the Western World': 59. 73 A.H. Wilson, Governments and Innovation (Ottawa: Science Council of Canada 1973), 74. 74 Bourgault, Innovation, 59. 75 McFetridge, 'Research and Development,' 273. 76 Department of Finance, Government of Canada Tax Expenditure Account (Ottawa: Department of Supply and Services 1980), 14. 77 J.D. Howe and D.G. McFetridge, 'The Determinants of R & D Expenditures,' Canadian Journal of Economics 9 (Feb. 1976), 57-71. 78 Caves et al., Competition in the Open Economy, 187-8. 79 To overcome EDP requirements with respect to a firm's viability, a special electronics program was established to increase flexibility and thus meet the

1247 A. Abonyi and M.M. Atkinson

80 81 82

83 84 85 86 87 88 89

90 91

92 93 94 95

particular needs of that sector. The $29 million originally allocated to this fund was devoted to firms with exciting growth prospects and, after two disbursements, was nearly exhausted. In January 1982 the federal government added another $67 million: Globe and Mail, 26 January 1982. Sagasti, Self Reliant Development, 124. OECD, The Aims and Instruments of Industrial Policy: A Comparative Study (Paris: OECD 1975), 7-9. S. Okita and S. Tamura, 'Transfer of Technology and the Japanese Experience,' in Technology Transfer in Pacific Economic Development ed. K. Kojima and M.S. Wionczeck (Tokyo: Japanese Economic Research Centre 1975), 70. M. J. Peck and A. Goto, 'Technology and Economic Growth: The Case of Japan,' Research Policy 10 (July 1981), 202-21. Okita and Tamura, 'Technology Transfer,' 70. Ibid., 71. Ibid. Sagasti, Science and Technology, 99. Chalmers Johnson, MITI and the Japanese Miracle (Stanford: Stanford University Press 1982), esp. ch. 7. For an elaboration of this sector strategy in Hungary see A. Balassa, 'Central Development Programs in Hungary,' Acta Oeconomica 14 (1975), 91-108. The Andean Group was established in 1969 by Bolivia, Chile, Colombia, Ecuador, and Peru. Venezuela joined in 1973. For a review of the original objectives of the Common Market and aims of the sector strategies see Junta del Acuerdo de Cartagena, Andean Pact, 21-31, and Lynn K. Mytelka, Regional Development in a Global Economy (New Haven: Yale University Press 1979). MOSST, Measures, passim, and Economic Development for Canada in the 1980s (Ottawa: Department of Finance 1981), 15-16. D. Brown and J. Eastman, The Limits of Consultation: A Debate among Ottawa, the Provinces and the Private Sector on an Industrial Strategy. Prepared by the Institute of Intergovernmental Relations, Queen's University (Ottawa: Science Council of Canada 1981). Ibid., 177. For example, Daly and Globerman, Tariff and Science Policies. Science Council of Canada, Forging the Links: A Technology Policy for Canada (Ottawa: Supply and Services 1979). Daly and Globerman do imply that the discrimination in favour of nationals, which they identify in present policies, might be attributed to political gains made by scientists and professionals, but no effort is made to outline the political process associated with the creation of these policies.

Technological Innovation and Industrial Policy / 125 96 See Okita and Tamura, 'Technology Transfer,' passim; M. Saito, 'Introduction of Foreign Technology in the Industrialization Process: The Japanese Experience since the Meiji Resoration,' The Developing Economies 13 (Jan. 1975), and UNCTAD, Case Studies in the Transfer of Technology: Policies for Development of Technology in Pre-war Japan (New York: UNCTAD 1978). 97 Ira C. Magaziner and T.M. Hout, Japanese Industrial Policy (London: Policy Studies Institute 1980), 30-1, 43-4; and Johnson, MITI, passim. 98 R.J. Wonnacott, 'Industrial Strategy: A Canadian Substitute for Trade Liberalization,' Canadian Journal of Economics 8 (Nov. 1975), 536-47. 99 M.J. Gordon, 'A World Scale National Corporation Industrial Strategy,' Canadian Public Policy 4 (Winter 1978), 46-56, and Science Council, Forging the Links, 51. 100 Scherer, Economic Performance, 414-15. 101 Harry Johnson, Technology and Economic Interdependence (New York: St. Martins 1975), 21-3, 58. 102 Ibid., 43, 85; and Pavitt, 'Government Policies,' 550. 103 Lester C. Thurow, The Zero-Sum Society: Distribution and the Possibilities for Economic Change (Harmondsworth: Penguin 1981), 22. 104 Office of the Prime Minister, Reorganization for Economic Development, 12 January 1982 and Reorganization for an Economic Strategy, 27 April 1982. 105 Peter Aucoin and Richard French, Knowledge, Power and Public Policy (Ottawa: Science Council of Canada 1975), 77. 106 Ibid., 81. 107 Richard Phidd and G. Bruce Doern, The Politics and Management of Canadian Economic Policy (Toronto: Macmillan of Canada 1978), ch. 8. 108 Rianne Mahon, 'Canadian Public Policy: The Unequal Structure of Representation,' in The Canadian State: Political Economy and Political Power ed. Leo Panitch (Toronto: University of Toronto Press 1977), 180-2. 109 Richard French, How Ottawa Decides (Toronto: James Lorimer and Co. 1980), 110-11. 110 Ibid., 123,129-32. 111 Daly and Globerman, Tariff and Science Policies. 112 Globerman, 'Canadian Science Policy,' 42. 113 Aucoin and French, Knowledge, Power and Public Policy, 78-80. 114 Mel Watkins, 'The Economics of Nationalism and the Nationalism of Economics: A Critique of Neo-classical Theorizing,' Canadian Journal of Economics 9, supplement (Nov. 1978), s 109-11. 115 Wallace Clement, Continental Corporate Power (Toronto: McClelland and Stewart 1977).

126/A. Abonyi and M.M. Atkinson 116 Abraham Rotstein, The Precarious Homestead (Toronto: New Press 1973), 24. 117 Stephen Hymer, 'The Multinational Corporation and the Law of Uneven Development,' in International Firms and Modern Imperialism ed. Hugo Radice (Harmondsworth: Penguin 1975), 58. 118 Kenneth Norrie, 'Regional Economic Conflicts in Canada: Their Significance for an Industrial Strategy,' in The Politics of an Industrial Strategy: A Seminar (Ottawa: Science Council of Canada 1979), 74. 119 Garth Stevenson, Unfulfilled Union (Toronto: Gage 1982), ch. 5. 120 The strongest initiative in this regard seems to have been taken by the government of Ontario which announced in 1981 the creation of a Board of Industrial Leadership and Development (BILD) fund totalling $70 million. Initiatives and problems are discussed in 'Report on Ontario,' Globe and Mail, 26 April 1982. Other provinces are also taking steps to attract high-technology firms. On Saskatchewan's efforts in this regard see Globe and Mail, 13 July 1981. 121 Brown and Eastman, The Limits of Consultation, 175-9. 122 Leo Panitch, 'Trade Unions and the State,' New Left Review, no. 125 (Jan.-Feb. 1981), 21-43.

6 / Institutions, Constitutions, and Public Policies: A Public-Choice Overview MARK SPROULE-JONES

Public choice is a comparative approach to the study of public policies. It is centrally concerned with comparing different ways of providing public policies to individual citizens. It is also an economic approach to the study of public policies. Public choice uses the relatively recently devised concepts of public goods, externalities, and common pools to compare how and why public policies differ. One of the objectives of this essay is to outline and illustrate the public-choice approach to comparative policy analysis. One of the major concepts used in public-choice analysis is that of 'institutional arrangements,' a term used to describe the laws (statute and common), the government regulations, and the number and type of organizations (government and private) that may be involved in providing public policies. These institutional arrangements are ultimately established and maintained by 'constitutional arrangements' that determine who may exercise governmental authority, what limits (if any) are placed on this authority, and what kinds of agreements must exist between government decision-makers before laws, regulations, and organizations are changed. The constitutional arrangements may all be written down in a single document, but usually (as in Canada) they are contained within a number of documents (such as the British North America Act, the Treaty of Westminster, the treaties between native Indians and the British Crown before 1867). Public choice is interested in seeing what differences institutional and constitutional arrangements make in providing public policies. A second objective of this essay is to outline such differences. The public-choice approach is normative in cast, and is aimed at determining what kinds of institutional arrangements and constitutional arrangements work better for citizens. It argues for the reform of laws, regulations, and government and private organizations, as well as constitutional arrangements that do not

128/M. Sproule-Jones perform in the interests of citizens. The structure of government - the composition of Parliament, for example - should be modified in favour of a different structure if necessary. Existing institutional and constitutional arrangements are only of value if they work better than others in providing the policies that citizens prefer. The third objective of this essay is to outline the criteria used in public choice for assessing performance. The first section describes the essential elements of the public-choice approach. Three core concepts for comparative analysis are discussed: the nature of the policies, the nature of institutional arrangements, and the nature of citizens as they demand and respond to policies and institutions. A schematic representation is included for ease of interpretation. The second section describes the major 'analytical units' that have been found to be helpful in understanding the provision of public policies. These analytical units encapsulate the findings and research agenda for the approach to comparative policy analysis. The third section is explicitly concerned with constitutional arrangements. It contains a framework that classifies the major kinds of differences that constitutional arrangements can make in policy provision. The fourth section briefly describes the kinds of criteria used to assess the performance of the system in providing policies that promote human welfare. The concluding section will briefly discuss the origins and future directions of the approach, and will contrast the approach with some of its major competitors. The Core of Public Choice

The public-choice approach builds its analysis of public policy on three essential elements. These elements or concepts are then used to reason about and to derive testable hypotheses about differing public policies. The public-choice approach is as much about the logical reasons policies differ as it is about discovering what the policies are in the first place. The student of politics is therefore counselled to read the following descriptions of the essential elements with some patience. The first element is the nature,of policies or goods in question; the second is the institutional arrangements for their provision; the third is the nature of individuals. Each will be discussed in turn. The nature of policies or goods Public choice warns us against making generalizations from the provision of one policy to the provision of another. The characteristics of the policies may differ;

Institutions, Constitutions, and Public Policies / 129

for example, we cannot readily compare education policies with fishing policies, or energy policies with defence policies. Policies differ in two crucial ways that allow us to measure their differences and similarities. First, we must ask if it is technologically or legally easy to exclude some citizens from benefiting from a policy. It is easy, for example, to exclude some persons from having a polio shot or from using an automobile. In contrast, it is difficult to exclude some from enjoying the benefits of national defence. Think of the technical difficulties involved if the government of Canada were to provide defence to all residents of Ontario except those living in Toronto! Policies must be seen as varying in their degree of exclusivity. Second, some policies are not 'packageable' and available for one citizen or another to enjoy alone. Much of the good is left for a resident of Hamilton to enjoy after a resident of Vancouver has enjoyed the national defence provided by the Canadian government. Similarly, someone can use a park, and there will be still benefits left for the next citizen. In contrast, if one person consumes a loaf of bread, none of the loaf is then available to another consumer. All goods and policies vary in the degree to which the good is available for others to enjoy after one person has enjoyed the good. Table 1 illustrates these two dimensions of exclusivity and availability from which comparisons of public policies may be made. Policies or goods that would fall into box A are normally labelled 'private goods'; those that fall into box c are normally labelled 'public goods.' It is important to remember that these labels do not refer to the mode of provision- private goods can be provided by the public sector, and often are; public goods can be provided by either the private or public sector. The labels refer solely to the manner in which they are consumed or enjoyed by citizens. Goods that would fall in box D are sometimes labelled 'common pools'; that is, their availability is reduced by consumption but it is difficult or costly to exclude citizens from using them. A freeway from a downtown to a new suburb may be used by a commuter, for example; once his car is on the freeway less of the road is available for other commuters. While tolls are sometimes used to exclude certain car owners from using the freeway (those having a low income or who take public transport), less and less of the roadway is available to commuters as more and more cars find their way onto the road. The highway may eventually become completely congested. The same process may occur if a second highway is built. Many natural resources are common pools; fishermen may exhaust a fishing stock or oil companies an oil pool unless some form of exclusion (such as fish licences or production quotas) is established. Some form of exclusion must be practised in the management of natural resources to ensure that use does not exceed biological or geological renewal.

130/M.Sproule-Jones TABLE 1 The main characteristics of policies Low availability after consumption

High availability after consumption

Low costs of exclusion

A

B

polio shot (pure private goods)

park

High costs of exclusion

D

fisheries (common pools)

c national defence (pure public goods)

Some public policies have indirect consequences on other policies. Certain logging policies may have beneficial consequences for outdoor recreation in developing logging roads or increasing deer stocks. Other logging policies may have negative consequences on bear stocks, for example, or on soil erosion (without proper silvaculture practices). Beneficial interdependencies are called 'positive externalities' if they are not taken account of by the users of a good. Harmful interdependencies can yield 'negative externalities.' We should anticipate positive or negative externalities with the provision of any good. For example, education policies are interdependent with manpower policies, with immigration policies, with industrial policies, and so on.1 The nature of arrangements Public goods, private goods,common pools, and goods that create externalities are provided to citizens under a variety of institutional arrangements. Laws, regulations, and organizations are important in establishing who gets what, with what consequences, and for what reasons. Investigations into the nature of institutional arrangements fall into a sub-field of public choice called 'property rights analysis.' Institutional arrangements affect the 'demand side' for the provision of policies by government, by the private sector, or by both jointly. They also affect the 'supply side.' This division of concerns is a fundamental part of property rights analysis. Examples of demand and supply considerations will illustrate how public choice treats the concept of institutional arrangements or property rights. Consider first the problem of providing a pure public good to a large number of citizens, such as the population of a city. If all the government did (at local, provincial, or federal levels) was resolve disputes and maintain contracts, public goods would probably be insufficiently provided by the private sector. Citizens

Institutions, Constitutions, and Public Policies/ 131

who want police patrols, for example, would reason that this policy would also be available for their neighbours, and that their neighbours would not be excluded from enjoying the benefits of protection from such patrols. They might also reason that they need not 'buy' the police patrols from a private security force, because their neighbours (who might have more money) would purchase them and they could avail themselves of the good. If everyone thought the same way, no one would contribute toward the cost of the police patrol, and the private security firm would go out of business. A different set of institutional arrangements is necessary to see that citizens are provided with the benefits of a policy they want but will not contribute toward voluntarily. One change in the institutional arrangements would be for a government to impose taxes on citizens and force them to contribute money toward the cost of police patrols. This 'free-rider' problem is a classic example of the ways in which institutional arrangements can affect the provision of a public good.2 More precisely, the demands of citizens for a public good are not articulated (signalled) and aggregated under the existing system. Note that it is a demand-side problem, not a supply-side problem; the government can contract with the same private security firm to provide police patrols, and the government need not change arrangements on the supply side.3 A second example of demand-side considerations is now generally known by the term 'rent-seeking.'4 Consider a taxi driver who has invested in a suitable vehicle, and who finds that casual drivers or commuters will pick up fares at convenient times for a modest payment. He may band together with other taxi drivers to pressure city council to license all taxis and ensure a stable market. By so limiting entry into the market, he can insulate himself from the pressures of competition. He can also resell his licence to other prospective taxi drivers. If he happens to work in Vancouver, such a licence would sell for over $30,000.5 In other words, the cost to the taxi-takers of having a so-called stable market is $30,000 per licence. The real cost is even higher; the taxi owners will invest time and energy (opportunity or transaction costs) in banding together, lobbying the city council, and ensuring that limited entry is policed. There will also be direct administrative costs to the local government for enforcing the regulation. This example illustrates how certain institutional arrangements can be constructed to establish cartels and to transfer wealth from one part of society to another. It also illustrates how institutional arrangements that are responsive to such groups (or individuals) can skew and bias the demands of all of the public. There is much evidence to indicate that Canadian society is biased in favour of a myriad of rent-seeking groups and individuals.6 A third and final example of demand-side considerations, and of how differing institutional arrangements can affect the articulation and aggregation of preferences for public policies, is found in the analysis of voting rules. Different

132/M. Sproule-Jones voting rules - such as plurality rule with a limited franchise - can yield a different set of decisions from, say, simple majority rule with an extensive franchise. Public choice has analysed in detail a variety of voting arrangements. Take, for instance, the procedure by which bills may be voted on by a committee. If the committee votes on bills in pairs, and votes to prefer bill A over bill B, and then to prefer bill C over bill A, the decision may be taken that bill c is the winner. However, the committee never voted on the merits of bill B versus bill c, and bill B might be preferred in this pairwise comparison. One conclusion that might be drawn is that the group or political party or chairman of the committee that sets the agenda for voting can manipulate the outcome by altering the procedure by which bills are compared and voted upon. The generic point is that the voting rules can be altered to yield differing solutions or differing patterns of demand aggregation.7 Institutional arrangements can also affect the supply or production of governmental policies, just as they can affect the demands for differing policies. Government policies, whether produced directly by a government agency or contracted to an agency of another government or to a private sector organization, require personnel, capital, and land to be put into effect. One classic example of the way that institutional arrangements can affect the supply side is in the budgeting process.8 If elected officials are presented with a budget from an agency that has a monopoly in producing a service, it is likely that the budget request will be larger than necessary to provide the level of service that citizens demand. The Department of National Defence may request a budget to cover the operations of the military in Canada, in NATO, in NORAD, and so on. The department will ask for a budget to cover the total cost of its operations. While the Treasury Board, the cabinet, and Parliament may not meet this request, they are never supplied with the necessary information about the relative costs of different levels of service. They cannot make judgments about the value of adding or subtracting a different level of operations with the costs of different levels. By equating the total value of operations with the total costs, they are asked to treat equally additional levels of operations that are more costly to produce and have less value (for example, naval operations in the South Pacific) on the same basis as operations that have a high value and are relatively cheaper to produce (for example, domestic security). Comparing total benefits with total costs, rather than marginal benefits with marginal costs, leads to a greater level of the supply of the good. Different institutional arrangements may have to be adopted to alter the monopoly of information possessed, in this case, by the Department of National Defence. Another way in which institutional arrangements can affect the supply of services is through the discretion granted to employees in an agency to set

Institutions, Constitutions, and Public Policies/ 133

priorities. For example, traffic and road engineers typically have considerable discretion to set priorities for road-building, road-widening and road maintenance. The only check on this discretion may come from elected local officials responding to individual complaints about potholes or other hazards. Priorities may be set to stress maintenance and building of arterial highways, which move greater volumes of traffic per mile, leaving neighbourhood roads to be built and maintained only when funds permit. Such priorities may have more impact on neighbourhoods with old roads than on newer subdivisions with roads recently built by developers. In turn, the older neighbourhoods may be disporportionately populated by ethnic, linguistic, or class minorities. In other words, institutional arrangements affecting the supply of goods and services may ultimately affect the equality with which they are distributed. In the case of public goods and services that are not sold in the marketplace, there may be no ready test to see whether the public prefers or rejects the priorities established through bureaucratic service delivery rules.9 In sum, a variety of institutional arrangements affect the degree to which individual preferences are articulated and aggregated by public means (the demand side), as well as the degree to which public policies are produced by organizations in the private and public sectors (the supply side). Not only do institutional arrangements differ according to the particular policy under consideration, but they also operate differently depending on the nature of individuals. We now turn to the public-choice conceptions of individuals. The nature of individuals Public-choice analysis deals with the provision of policies that individual citizens demand, and that are supplied by one or more organizations in the private and public sectors. Public choice expects that citizen-consumers, elected and appointed officials, businessmen, and lobbyists will have their own conceptions of how and why policies should be provided. Two aspects of this concern with individuals are important in policy analysis. First, public choice emphasizes that policies are of value if they meet the demands of citizens as consumers, not of individuals who are employed in or by the private and public sectors. Health policies that satisfy the aspirations of doctors, nurses, and hospital administrators are of no value and may waste resources unless they are also valued by patients or future patients. Policies, and the institutional arrangements for their provision, are of instrumental value and have no intrinsic merit. Second, public choice emphasizes that the demands of individuals are made within and structured by institutional arrangements. A farmer may demand a

134/M. Sproule-Jones high degree of fire protection, and will normally satisfy that demand by installing fire extinguishers, buckets, and hoses. Institutional arrangements that would provide him with a nearby fire hall are unlikely to be implemented. Similarly, a bureaucrat working in the Department of Indian Affairs may have limited discretion in carrying out his preference for a devolution of power to Indian bands; he will be restricted by the Indian Act, its regulations, the budget available to him, and so on. Individuals in any position, in other words, will have constraints and opportunities within which they can attain their preferences. Individuals are thus not treated as synoptically rational persons. They act on the basis of differing unseen preferences within the incentive systems established by laws, organizations, and other features of institutional arrangements. Urban mobility provides an example of individuals receiving the benefits of public policies within an ongoing system of institutional arrangements. Individuals living in a particular urban neighbourhood may receive public policies and have access to private goods that are inferior to those that can be attained in another neighbourhood. The natural environment, housing, life-style, or even garbage-collection services may be improved by moving elsewhere. More important, such advantages may be attained more easily by moving than by voicing concerns to the local government, by challenging urban redevelopers, or by readjusting life-styles and preferences. In greater Vancouver, for example, 26 per cent of the residents report that they have moved within the urban area for housing reasons, 18 per cent for environmental reasons, and 11 per cent for both housing and environmental reasons.10 In short, many citizens find they can better attain their preferences by relocation, and many public policies do not match neighbourhood demands.11 The Units of Analysis

The three essential elements of public choice - the nature of the good, the nature of institutional arrangements, and the nature of individuals - are building blocks from which public-policy analysis proceeds. They generate the units of analysis that may be compared and treated empirically. The units of analysis called 'provision systems' and 'industries' are built in the following way.12 Consider a policy or good that individuals want or are being provided with. The policy might be fire protection, water-quality management, police protection, national defence, or garbage collection. Consider also how these goods are provided to a population living in a contiguous area, such as a neighbourhood, a water basin, a city, or a region. It is likely that private enterprises, public enterprises, groups of citizens, and individual citizens are all involved in providing this service. Two examples will illustrate this.

Institutions, Constitutions, and Public Policies/ 135

Fire protection is provided through the interaction of a number of organizations, mainly of a public-enterprise character. A municipal bureaucracy usually organizes and provides fire-fighting equipment and personnel. Other local government departments typically enforce construction standards and building codes. Urban renewal and housing departments (often at different levels of government) may select sites for clearance or establish building and fire codes. The water supply, which may be organized by a local government, co-operative local government agencies, or by public utilities regulated at higher levels of government, affects water pressure and hydrant location. Private households contribute to security through educating children, maintaining dwellings in good repair, and purchasing fire insurance. Insurance firms provide some inspection and investigating services in conjunction with fire departments. Associations of private firms, insurance underwriters, and fire-fighting personnel influence construction, inspection, fire department personnel practices, and technology. Police departments provide some surveillance, reporting, and occasional fire-fighting services. Private firms may use, often under public regulation, water supply and sprinkling systems, security guards, janitors, and supplementary fire-fighters. In other words, a multiplicity of public and private organizations contribute in interdependent ways to the provision of fire protection.13 The management of water quality is a second example. Typically, river basins are used for a wide variety of purposes-navigation, storage, fisheries, boating, water supply, waste disposal, industrial and residential sites, and the production of lower forms of biological life, to name a few. Some of these uses are for the production of private marketable goods, and one typically finds a large number of individuals and locally, nationally, and internationally owned private firms jointly involved in the production or consumption of water quality. A large number of local, regional, national, and international agencies of government are also involved, either because they participate in the production or consumption of private goods and services (such as fish) provided by the basin; because they are involved in the regulation of the production or consumption of marketable goods and services; or because they are involved in the production or consumption of public goods and services. Again, many organizations are typically involved in the provision of water-quality management.14 The network of organizations and individuals involved in providing a public policy is termed a 'provision system' in public-choice analysis. Provision systems for a particular policy can exist across many communities, in which case they are collectively called a 'public service industry.' These provision systems are the units of analysis normally taken as points of comparison in public-choice policy analysis. One compares, for example, the provision systems for solid waste collection in Vancouver, Toronto, Geneva, and Rome; or one compares the

136/M. Sproule-Jones public service industries for natural resources in Canada, Australia, and Sweden. Two important features of these units of analysis must be stressed. First, it is theoretically possible to find only a very simple organizational network in any provision system. It is possible to find, for example, a single organization with a monopoly in the provision of a policy, although this is unlikely given the nature of most goods and services. I will return to this point in the next section of the paper. Second, the units of analysis differ considerably from the units conventionally used in many public policy analyses. By beginning with the nature of the policy in question, one typically finds the institutional arrangements (including organizations) of more than one level of government involved in providing any policy. This approach implies that it is invalid to compare the policies of national governments, provincial governments, or local governments. Each of them may or may not be involved in articulating the demands or contributing to the supply of any public policy. Focusing on the level of government will take only those parts of a number of provision systems this level of government is involved with, and then attempt to re-aggregate them into a misleading collection of policies. In other words, different descriptions, analyses, and conclusions will be found depending on whether a 'provision system' or a 'level of government' is used as a basic unit of analysis. To illustrate the form and kind of conclusions reached in public-choice analysis of public policies, consider the provision of public safety in a residential community, such as the municipality of Oak Bay, British Columbia. Public safety, like any public policy, consists of a variety of activities, including crime prevention, responses to reported crimes (including investigation and apprehension), traffic regulation and control, and general public services (such as crowd control and escort duty). After identifying the activities that make up the public policy, public-choice analysts would then attempt to describe all of the means by which such activities are produced.15 It becomes apparent that these activities are not the exclusive prerogative of the municipal police department. Some are produced by the RCMP, some by private security forces, some by institutional security forces, and some by adjacent municipal police departments (because criminal activities may cross municipal boundaries); some are co-produced by the private citizen (especially in crime reporting and prevention).16 This list must be expanded to include those organizations that provide (often by contract) support services: for example, the British Columbia Police Academy, which trains police recruits, or the RCMP crime laboratory. Another group of producing organizations are those that operate at the 'street level' of the municipality

Institutions, Constitutions, and Public Policies/ 137

through the decisions made by the municipal police department; examples would include the British Columbia Co-ordinated Law Enforcement Unit (Joint Forces Operation), which combats organized crime, or the Emergency Response Team (the 'SWAT' team) organized by Oak Bay and two adjacent municipal police departments. On the production or supply side alone, it is possible to identify over twenty-five producing agencies (excluding citizens) that are directly or indirectly involved in producing public safety in a residential municipality of 18,000 persons. On the demand side, there is no single body that articulates citizen demands about the general level and quality of public security services or responds to complaints about the major producers and their employees. Key actors would include the municipal council (which funds the municipal police department), the police board (which oversees policy), the British Columbia police commission, the attorney-general, and a variety of elected officials at local, provincial, and federal levels who can initiate and participate in policy-making that affects suppliers paid by the level of government in question. Provision systems such as the one outlined above appear not only complex but confusing. Studies of provision systems find, however, that there are regular patterns of relationships and interactions between producers and between producers and demand-responding organizations. A multiplicity of organizations does not, in other words, imply unco-ordinated and unplanned provision of services.17 Provision systems may be described as possessing a variety of degrees of horizontal and vertical integration, depending on the activity in question. Activities that are easily measurable and are capital-intensive, and are in turn amenable to economics of scale in production, are normally vertically integrated (for example, crime laboratories in public security-provision systems). One also would expect to find both co-operation and competition, as well as collusion and coercion, operating within provision systems.18 The concepts of 'provisions systems' and 'public-service industries' were devised in order to make public-policy analysis better correspond with the complexity and multifaceted nature of policies. Given the differences in the nature of public policies (and activities), one would not expect to find similar institutional arrangements for all policies for any community. Given also the differences between individuals and their preferences, one would not expect to find policies provided in comparable levels and qualities from one community to the next. And given differences between institutional arrangements, one would not expect to see individuals articulating their preferences and responding to existing advantages and disadvantages in policy provision in identical ways from community to community. In other words, one should anticipate complexity,

138/M. Sproule-Jones variety, and overlapping authority and jurisdictions. The question of whether these kinds of provision systems are better than alternatives for providing public policies is a matter to be addressed later in this essay.19 Constitutional Arrangements We have seen that public-choice analysts treat the provision of public policies as necessarily complex and diverse, and that they compare public policies by identifying provision systems and public-service industries as units of analysis. Comparative studies using such concepts reveal that the institutional arrangements involved in providing public policies differ from place to place and from time to time.20 This forces one to ask 'Why?' At this point, public-choice analysis borrows extensively from one of its other sub-fields - the analysis of constitutions. It appears that the authority given to legislative bodies to pass laws, establish regulations and agencies, and set the framework within which the private sector operates rests on the nature and structure of constitutional arrangements. The role of constitutional arrangements in structuring the various provision systems for public policies may be summarized and classified in the following way. First, public institutions vary in the degree to which they share authority for the provision of goods with other public institutions and with individual and corporate persons in the private sector. At one extreme, one may think of public institutions as possessing concurrent authority with other public and private organizations to provide goods, and having limited entry costs for additional organizations entering the provision system in question. For example, in Canada, agricultural services are subject to the concurrent authority of both the provincial and federal governments; because of the international character of most agricultural markets, services are also influenced by the operations of non-domestic governments and corporate and individual persons resident abroad (the foreign private sector). At the other extreme, it is possible to conceive of the monopoly provision of goods, although real-world instances are difficult to find. Even the classic case of military services, which in many formal constitutional documents is assigned to a national government alone, is, on closer scrutiny, a severely constrained example of monopoly provision because of the international character of military markets and the relatively low entry costs for paramilitary and guerrilla movements (normally in the private sector). Nevertheless, it is possible to suggest that the two polar extremes for characterizing the 'product' market for provision systems are concurrency and monopoly. Second, governments may vary in the degree to which they possess concurrent authority to regulate and produce the inputs for provision systems. Provision systems require the supply and transformation of factors of production - labour,

Institutions, Constitutions, and Public Policies / 139 TABLE 2 Policy and provision system consequences of constitutional choice Monopoly provision

Concurrent provision

Monopoly procurement

A

B

Concurrent procurement

D

C

capital, and land. Labour markets in Western societies are regulated by statute or common-law provisions through courts and quasi-independent administrative tribunals that enjoy arm's-length relationships with the executive branches of governments. The only exceptions occur in the attenuated control of labour markets granted to one or more governments for purposes of military conscription and international migration. Similarly, capital markets are rarely subject to the exclusive jurisdiction of a single government, and capital is a highly mobile factor even across international boundaries. Land as a factor of production is more often subject to the exclusive monopoly control of a single government, but all three necessary factors for the production of the products of provision systems are rarely subject to the monopoly jurisdiction of one government. Governments find that they must come to terms with institutional constraints on their ability to provide policies on both the input and output sides of policy provision. These two considerations related to the degree of concurrency in the product and factor markets for the provision of goods and services allow us to draw a simple 2-x-2 matrix describing the polar extremes of the policy and provision system consequences of constitutional choice (see table 2). Different communities of individuals will be subject to goods and services provided under different conditions approximating any of the four boxes (A, B, C, and D) in the matrix of constitutional possibilities. Our previous discussion implied that for most goods and services, communities will find themselves in box C. No one government, at any level, will exercise monopoly control of either the products of a provision system or over the factors (land, labour, and capital) that are used in production by a provision system. It would be improbable to expect all goods and services to be constitutionally organized in the same way for all communities at all times. Military services, for example, are often provided in wartime under constitutional conditions approximating box A in the matrix; a single government is granted monopoly authority over the conduct of a war and the procurement and consumption of factors of production. Many small, geographically isolated communities often find that

140/M.Sproule-Jones their so-called local government services (such as police, garbage collection, or fire protection) approximate the conditions of box D; alternative producers of the service are not organized because of market size. Much of the concern expressed by democratic theorists about the consequences of public-sector unionization may be interpreted as a concern about the consequences of provision systems organized under conditions of box D; the monopoly bargaining power of unions may not be offset by the countervailing monopoly power of the government as employer. An important empirical question about constitutional arrangements is this: how far do the processes of constitutional change afford public (and private) enterprises the opportunities to prevent rivals or potential rivals from providing goods and services? To what degree is it possible for one government to reduce the rivalry in provision systems? An instructional example may be drawn from changes in the number of powers of local governments (both municipalities and special districts) as a result of laws passed by provincial governments in Canada, by equivalent regional governments in other countries, or by national governments in unitary political systems. When the secession and formation of new local governmental units from larger units has been prohibited, when small governments have been amalgamated, or when one local government has been permitted to annex surrounding areas, the degree of concurrency in the product markets has been reduced. As economic theory would predict, institutional changes that support monopoly provision and reduce rivalry bring increased costs to the provision of services without compensating increases (and often with decreases) in the levels and quality of services provided in the community.21 These changes would not have been possible had not constitutional authority been granted to a provincial or national government. Constitutions vary in the degree to which they facilitate the entry and exit of public institutions into different provision systems. The following three examples taken from Western societies indicate the types of control over entry and exit that are typical of constitutional authority. 1 Constitutional authority over provision systems, subject to a majority vote of elected representatives in a unitary government, as in the United Kingdom, with the important exceptions of non-resident public and private enterprises subject to willing consent with foreign governments (explained further below). 2 Constitutional authority over provision systems, typically subject to a majority vote of judges as arbitrators appointed by the national government, as in the United States, but again with the international exceptions noted in the previous example. 3 Constitutional authority over provision systems, typically subject to the willing consent of both regional and national governments, as in Australia and Canada (at least until 1982), again with the international exceptions.

Institutions, Constitutions, and Public Policies / 141 These three examples draw attention to two additional features of constitutional arrangements that must be considered for a full understanding of comparative policy analysis. First, constitutional arrangements for providing goods and services actually include multiple constitutions of different communities. Non-resident public and private enterprises, organized under non-domestic constitutions for any community, can enter and leave the product or factor markets for different provision systems. Foreign governments and multinational corporations are the names given to enterprises organized under constitutions abroad. But domestic communities within nations may also have constitutions and domestic enterprises that interact with non-resident enterprises (public and private) organized under different constitutions. Multiple constitutions imply inter-constitutional relations that must be analyzed for an accurate understanding of publicpolicy provision. Second, the arrangements for resolving differences between multiple constitutions deserve extended scrutiny and analysis. For example, the provision of goods and services in Australia and Canada cannot be understood outside the context of unanimous-consent decision rules of the seven Australian and eleven Canadian governments. In both cases, the governments operate like cartels to exclude rivals, and display the acknowledged instability of cartels as they bargain together, plan holdout strategies, and impose high transaction costs on other governments to secure unanimous-consent agreements.22 We are now in a position to summarize the major elements in a public-choice approach to comparative policy analysis. Public and private goods, common pools, and goods with externalities are provided to communities of individuals by public organizations, private organizations, groups, and individuals. For each kind of good there exists a provision system. Provision systems for any good make up a public-service industry. Provision systems vary in the degree to which any public or private institution is subject to the rivalry of other enterprises. This rivalry is a function of constitutional arrangements (including inter-constitutional relations). A discussion of how public-choice analysis evaluates the performance of provision systems follows. The Performance of Provision Systems The object of the public-choice approach to comparative policy analysis is to evaluate as well as to describe and explain public policies in different regions and countries. This concern stems from the public-choice interest in individual citizens. Institutional arrangements and the policies provided thereby are presumed to meet the demands of individuals, and if better institutional and constitutional arrangements can be designed, they deserve to replace the existing ones.

142/M. Sproule-Jones This normative posture does raise the question of the criteria to be used to judge whether provision systems, or indeed entire industry structures, are operating better than alternatives that may be established. Public-choice analysts have shifted their criteria over time in the light of newer theoretical investigations, it may be useful to review briefly the two kinds of criteria used in earlier studies (and which, incidentally, are used extensively in pure economics) before specifying the criterion of assessment given primacy in later works. Criteria based on individual utilities The policies provided by any provision system could be judged by whether they increase the satisfactions or preferences of individual citizens. The private sector may fail or be inadequate to provide preferred goods and services at prices that consumers are prepared to pay, and changes in institutional arrangements maybe necessary to provide public-type goods that are closer to the preferred quantities, qualities, and prices. Pollution may be generated, for example, as a negative externality from an industrial process, and under certain institutional arrangements, it may be difficult for certain citizens to band together to bargain effectively with the polluter. A change in institutional arrangements, such as making the polluter legally liable for damages in the face of class-action suits, may lead to greater 'global efficiency' or satisfy the preferences of those damaged by the externality. The problem with such a simple preference-based or utility-based criterion is that we assume that the value of damages to those experiencing pollution is commensurate with the benefits accruing to the polluter from an absence of pollution controls. We may be comparing apples and oranges. We may not be able to simply add up the benefits and costs of different ways of providing policies and assume that when the ratio of benefits to costs is greatest, then that particular policy output is the one to adopt. We would need to smuggle into our analysis a criterion besides that of preference satisfaction in order to make judgments. The list seems endless - should it be to satisfy a majority of citizens (but what if the policy included incarceration of a sizeable minority?)? Should it be to satisfy each individual (but what if the individual were a child-abuser?)? Should it be to satisfy those with intense preferences (to make war at any costs because some feel strongly about military adventures?)? Clearly, criteria based solely on citizen satisfaction are inadequate. The tradition still enjoys considerable popularity in Canada, however, particularly in governmental circles. Benefit-cost analyses, cost-effectiveness studies, and program-planning budget methods appear to be logical techniques by which governments can choose between a variety of programs, all of which have legitimate claims upon public treasuries.23 But the techniques all involve choices

Institutions, Constitutions, and Public Policies / 143

about what benefits and costs should be calculated, what effectiveness measures should be developed, what goals and objectives government agencies should pursue, and what trade-offs between competing programs should be made. Elected officials may have neither the information nor the self-interest to acquire the information about citizen preferences; the demand-side problems cannot be solved by such techniques. The criterion ofPareto Optimality Economists and public choice theorists have long recognized the problems of simple preference satisfaction, whether of consumers or officials, as a criterion to judge policy performance. One of the more popular substitutes was posed by the Italian economist, Vilfredo Pareto. Pareto formulated the criterion that if a policy change would make at least one person better off and no one worse off, then that change should be effected. The criterion was derived from examining how, in the private market, buyers and sellers would exchange goods in a voluntary manner to satisfy their preferences. If a seller or a buyer did not like the price at which a good was offered for sale, either of them could voluntarily reject the offer. This is, in effect, an 'any-person rule' criterion, as it requires the willing consent of all parties to an agreement prior to the agreement being implemented. In one sense, the criterion is a useful one. Take, for instance, the issue of whether a local government should not only provide but also produce services such as fire protection, policing, garbage collection, libraries, and recreation. The government can choose between tax-financed municipal production, taxfinanced franchised private production, tax-financed franchised public production by an agency of another government, user-charge-financed production under any of the three preceding modes of production, or even user-chargefinanced competitive private production (privatization). The municipality of Surrey, British Columbia, has chosen the second option, that of a tax-financed franchised private production, while Burnaby, a similar municipality in the same region, has chosen the first option, tax-financed municipal production. The cost per household of garbage collection and disposal in Surrey is approximately 82 per cent of that in Burnaby. This suggests that the residents of Burnaby would all be better off, and a Pareto Optimal change attained, if the Burnaby city council contracted out the service.24 Pareto's criterion is not very useful in most circumstances, however. When governments provide public goods by taxing citizens, and when these citizens have different demands for the level and quality of a service and the tax burden they are willing to bear, it is highly unlikely that everyone would be satisfied with the terms on which the good is provided. For example, some people may be

144/M. Sproule-Jones worse off if the government reduces the level of educational services; others may be worse off if the government increases the level of educational services (and no alternatives such as mobility to another community or private schools are feasible). In other words, the criterion cannot be applied to the provision of public goods unless one assumes that all of the beneficiaries and taxpayers have similar preferences. Criteria based on institutional fairness Public-choice theorists have shifted the focus of their attention from the benefits and costs incurred by individuals in the provision of policies to the institutional and constitutional arrangements that affect policy prformance.25 Their reasoning is that if most policies will make some people better off and other people worse off, then the rules or institutional arrangements for policy provision should form the locus of evaluative criteria. If individuals perceive the institutional and constitutional arrangements as being fair to them as well as to others, then the policy outputs or policy changes can be evaluated as fair. This approach is sometimes referred to as 'rule utilitarianism.' Most of the reasoning focuses on the equality of opportunity that each individual has to alter the terms and conditions of policy provision. Do individuals, for example, have equal access to tribunals or the courts to bargain with a producer of externalities? Are individuals treated equally when their votes for representatives are counted in constituting a legislative assembly? This notion of institutional fairness as equality in setting and amending institutional arrangements is not without its critics. It is usually defended as being the criterion that any one person would choose if he were uncertain about his position in society in the future. It would ensure that anyone would have the long-term capacity to be of influence. Critics point out that many persons would opt for their present positions even if they could not predict social and political conditions. They also point out that the criterion may conflict with the concept of personal liberty in that it may proscribe individual actions that seem necessary to accomplish ends that individuals feel should have top priority (for example, selling oneself into slavery to provide income for one's family). Despite such controversies, this newer thrust to policy evaluation is a radical departure from previous norms. Institutional and constitutional arrangements that exclude the public or skew decision-making to the advantage of those with the will and resources to win can now be considered as potentially inefficient in a broad global sense.26 The challenge is not only to explain who wins and who loses, but to develop new norms to assess the place of the individual in the organization of the state.

Institutions, Constitutions, and Public Policies/ 145 Conclusion

Public choice offers a different approach to policy analysis. It does not treat policy-making and decision-making in government as the same for all public policies; the nature of the good places constraints and opportunities on collective action. It does not treat institutional and constitutional arrangements as a given; those arrangements can have independent and important effects on policy provision and on citizen demands. It does not treat citizen or governmental official interests as the same; diversity in preferences among individuals is expected in any community and state. What public choice does do is fashion a new approach out of the nature of the good, the design of institutional and constitutional arrangements, and the preferences of individuals. This new approach centres on provision systems and public-service industries as basic units of analysis, and then describes and evaluates how provision systems and public-service industries perform in different communities. The approach is an economic one, and it shares many of the conclusions of Marxist and neo-Marxist political economy. The approach would agree, for instance, that many governmental officials possess preferences and ideologies in common with corporate and organized interests. But the reasons it advances for such conclusions and its prescriptions for reform differ. First, public choice emphasizes 'process-type' criteria such as institutional fairness in assessing the allocational and distributional activities of the state and the market. In contrast, Marxists and neo-Marxists emphasize 'end-state' criteria such as the distribution of wealth and capital in society in evaluating the relationships between state and society. Second, public choice emphasizes the importance of individuals rather than classes of individuals as basic elements of normative value in society. This leads, in turn, to priority being placed on the economic and civil liberties of individuals in the face of collective action, in contrast to the approach of many Marxists and neo-Marxists. Third, the public-choice approach does not attribute any 'inner rationality' to the evolution of economic relationships over time. Constitutional and institutional arrangements are as likely to produce regressive effects in human welfare as they are to produce revolutionary or mutually productive relationships within society and between society and the state. There are other differences and similarities between the public-choice and the Marxist and neo-Marxist political economies. A task facing students is to examine in detail the symbiotic relationship between government in its many forms and the economic power possessed by individuals sharing common preferences and aspirations. One challenge for students in Canada is to extend

146/M. Sproule-Jones the scope of empirical investigations on these issues. This is a time in which serious reconsiderations are being made of constitutional arrangements, of federal-provincial-local relationships, of the role of the public in resource and other policy arenas, of the place of organized communities of interest (indigenous peoples, women, French Canadians, linguistic minorities) in the structures of governments, and of the scope of the private and public sectors. Public choice, which is essentially liberalism with new economic tools, has much to offer in these debates. NOTES 1 There is a large literature on the characteristics of public goods, common pools, and externalities. Most has developed since Samuelson's formal analysis in the mid-1950s, and the basic analytics can now be found in most modern publicfinance texts. A clear non-technical introduction is found in Robert L. Bish, The Public Economy of Metropolitan Areas (Chicago: Markham Press 1971), ch. 2. 2 The classic work on the 'free-rider' problem in politics is Mancur Olson's The Logic of Collective Action (Cambridge: Harvard University Press 1965). Recent modifications of the theory have specified the conditions under which wealth, influence, and intensity of preferences as well as mere size of group can induce the form of strategic behaviour known as 'free-riding.' For an example of this recent work, see John B. Chamberlin, 'A Collective Goods Model of Pluralist Political Systems,' Public Choice 33 (1980), 97-113. 3 This analytical distinction between demand and supply considerations, and its importance for reassessing how public goods can and should be provided to heterogeneous communities within urban areas, was first discussed at length in Vincent Ostrom, Charles M. Tiebout, and Robert Warren, 'The Organization of Government in Metropolitan Areas,' American Political Science Review 55 (1961), 831-42. It should be noted that many traditional economics and political science models still assume that the government (or level of government) that aggregates the demands for goods and services also must produce or supply these goods themselves. For a critique of such models in understanding federalism, see M.H. Sproule-Jones, Public Choice and Federalism in Australia and Canada (Canberra: Australian National University Press 1975). 4 The term 'rent-seeking' was first coined by Anne O. Krueger in 'The Political Economy of the Rent-Seeking Society,' American Economic Review, 64 (1974), 291-303. Many leading and recent articles on the topic are collected in Toward a Theory of the Rent-Seeking Society, ed. James M. Buchanan, Robert D. Tollison, and Gorden Tullock (College Station: Texas A and M University Press 1980). 5 G.B. Reschenthaler, 'Regulatory Failure and Competition,' Canadian Public Administration 19 (1976), 466-86.

Institutions, Constitutions, and Public Policies/ 147 6 For an overview see Studies on Regulation in Canada, ed. W.T. Stanbury (Toronto: Butterworths 1978). For evidence on urban property markets and the climate of local government activity, see Michael Goldberg and Peter Harwood, Zoning (Vancouver: The Fraser Institute 1980) as well as the more radical-popular literature such as James Lorimer, The Real World of City Politics (Toronto: James Lewis and Samuel 1970); Donald Gutstein, Vancouver City (Toronto: James Lorimer and Co. 1975); and James Lorimer, The Developers (Toronto: James Lorimer and Co. 1978). 7 An excellent recent introduction to this sub-field, sometimes known as social choice theory, will be found in Robert Abrams, Foundations of Political Analysis (New York: Columbia University Press 1980). A more advanced overview is Charles R. Plott, 'Axiomatic Social Choice Theory: An Overview and Interpretation,' American Journal of Political Science 20 (1976) 511-96. Readers should also consult J. Nicholas Tideman and Gordon Tullock, 'A New and Superior Process for Making Social Choices,' Journal of Political Economy 84 (1976), 1145-59, for a concise introduction to the recent literature on the 'Clarke-tax' method of overcoming free-riding by large groups. 8 This example is known in the literature as the Niskanen hypothesis, after the thesis found in William A. Niskanen, Bureaucracy and Representative Government, (Chicago: Aldine-Atherton Press 1971). Modifications and elaborations of the thesis have been made since the publication of Niskanen's book, a survey of which will be found in William Orzechowski, 'Economic Models of Bureaucracy: Surveys, Extensions and Evidence,' in Budgets and Bureaucrats, ed. Thomas E. Borcherding (Durham: Duke University Press 1977). For work that integrates the Niskanen hypothesis with demand-side considerations, see especially Robert J. MacKay and Carolyn L. Weaver, 'On the Mutuality of Interests between Bureaucracies and High Demand Review Committees,' Public Choice 34 (1979), 481-91. For work that integrates the hypothesis with intergovernmental arrangements, see especially Roger B. Parks and Elinor Ostrom, 'Complex Models of Urban Service Systems' in Urban Policy Analysis: Directions for Future Research, ed. Terry N. Clark Urban Affairs Annual Reviews, vol. 21 (Beverly Hills: Sage Publications 1981), 171-99. 9 The literature on bureaucratic service delivery rules and their effects on the supply and distribution of services is both recent and large. An overview can be found in Bryan D. Jones, Suadia Greenberg, and Joseph Drew, Service Delivery in the City (New York: Longmans 1980). Other major works in the field include Robert Lineberry, Equality and Urban Policy (Beverly Hills: Sage Publications 1977), and Michael Lipsky, Street Level Bureaucracy (New York: Russell Sage Foundation 1980). Systematic studies in Canada are few, but useful insights can be obtained from Lorimer, The Real World of City Politics, especially chs. 5 and 6.

148/M. Sproule-Jones 10 Unpublished data as part of survey designed to measure citizen preferences for environmental intangibles. See M.H. Sproule-Jones, 'The Social Appropriateness of Water Quality Management,' Canadian Public Administration 21 (1978) 176-94. 11 The study of urban mobility by public-choice analysts dates from the seminal article by Charles M. Tiebout, 'A Pure Theory of Local Expenditures,' Journal of Political Economy 64 (1956), 416-24. The subsequent literature is extremely large, three of the more important studies being John M. Orbell and Toru Uno, 'A Theory of Neighbourhood Problem Solving,' American Political Science Review 66 (1972), 471-89; David F. Bradford and Wallace E. Gates, 'Suburban Exploitation of Central Cities and Governmental Structure,' in Redistribution Through Public Choice, ed. Harold M. Hochman and George E. Peterson (New York: Columbia University Press 1974), ch. 3; and Dennis Epple and Allan Zelenitz, 'The Implications of Competition Among Jurisdictions,' Journal of Political Economy 89 (1981), 1197-1217. A recent survey article is Richard J. Cebula, 'Voting with One's Feet,' Regional Science and Urban Economics 10 (1980), 91-107. A recent reassessment of the extent of urban mobility in Canadian cities is that of Barry Wellman and Barry Leighton, 'Networks, Neighbourhoods and Communities,' Urban Affairs Quarterly 14(1979), 363-90. 12 This re-conceptualization of the organizational arrangements for providing public policies was first suggested in Vincent Ostrom and Elinor Ostrom, 'A Behavioural Approach to the Study of Intergovernmental Relations,' Annals of the American Academy of Political and Social Science 359, no. 165, 137-46. Five recent books that have developed the concepts of'industry' and 'provision systems' (or comparable terms) are Sidney Somenblum, John J. Kirlin, and John C. Ries, How Cities Provide Services (Lexington: Ballinger Press 1977); Eugene S. Savas, The Organization and Efficiency of Solid Waste Collection (Toronto: D.C. Heath 1978); Elinor Ostrom, Roger B. Parks, and Gordon P. Whitaker, Patterns of Metropolitan Policing (Lexington: Ballinger Press 1978); Roger B. Parks, 'Assessing the Influence of Organization on Performance,' PHD thesis, Indiana University 1979; Mark Sproule-Jones, The Real World of Pollution Control (Vancouver: University of British Columbia, Westwater Research Centre 1981). 13 Phillip M. Gregg, 'Units and Levels of Analysis,' Publius 4 (1974), 59-86. 14 Mark Sproule-Jones, 'Co-ordination and the Management of Estuarine Water Quality,' Public Choice 33 (1978), 41-53. 15 Note that production (or supply) is not the same as provision (or demand aggregation). 16 Co-production of public policies by citizens is generally ignored both by administrative studies and by studies of citizen participation. See Gordon P. Whitaker, 'Co-production: Citizen Participation in Service Delivery,' Public Administration Review 40 (1980), 240-6; Mark Sproule-Jones, 'Urban Bads and the Structure of

Institutions, Constitutions, and Public Policies/ 149

17 18

19

20

21

Institutional Arrangements' in Managing Urban Settlements: Can our Governmental Structures Cope?, ed. K.G. Denike (Vancouver: University of British Columbia Centre for Human Settlements 1980), 2-10; Roger B. Parks, Paula C. Baker, Larry Kiser, Ronald Oakerson, Elinor Ostrom, Vincent Ostrom, Stephen L. Percy, Martha B. Vandivort, Gordon P. Whitaker, and Rick Wilson, 'Consumers as Co-producers of Public Services,' Policy Studies Journal 9 (1981), 1001-11. Mark Sproule-Jones, The Real World of Pollution Control. Robert L. Bish, 'Intergovernmental Relations in the United States,' in Interorganizational Policy Making, ed. Kenneth Hanf and Fritz W. Scharpf (Beverly Hills: Sage Publications 1978), 19-36. The diversity, variety, and complexity is best termed 'polycentricity' (see Ostrom, Tiebout, and Warren, 'The Organization of Government in Metropolitan Areas'). It should be distinguished from 'pluralism,' which is normally reserved for sociopolitical diversities, varieties, and complexities irrespective of institutional arrangements and the nature of policies. Apart from the citations in n. 12, three useful summaries of the findings on urban public policies are Robert L. Bish and Vincent Ostrom, Understanding Urban Government (Washington: American Enterprise Institute 1973); The Delivery of Urban Services, Urban Affairs Annual Review, vol. 10, ed. Elinor Ostrom (Beverly Hills: Sage Publications 1976); Comparing Urban Service Delivery Systems: Urban Affairs Annual Review, vol. 11, ed. Vincent Ostrom and Frances Pennell Bish (Beverly Hills: Sage Publications 1977). For a lucid, non-technical introduction to this argument, see Robert L. Bish and Robert Warren, 'Size and Monopoly Problems in Urban Government Services,' Urban Affairs Quarterly 8 (1972), 97-122. Some of the best evidence for the argument comes from the United States and Sweden. See especially Dolores T. Martin and Richard E. Wagner, 'The Institutional Framework for Municipal Incorporation,' Journal of Law and Economics 21 (1978), 409-26; Robert B. Hawkins, SelfGovernment by District (Palo Alto: Hoover Institute Press 1976); Jorgen Westerstahl, Decision-Making Systems in 36 Swedish Communes (Gothenburg: University of Gothenburg Press 1974); and Bengt Owe Birgersson, 'The Service Paradox: Citizen Assessment of Urban Services in 36 Swedish Communes,' in Comparing Urban Service Delivery Systems, ed. Ostrom and Bish 243-68. There are similar indications that in Canada and Britain the larger local governmental units have failed to articulate and aggregate adequately citizen preferences. For introductions, see respectively Lionel D. Feldman, Ontario 1945-73: The Municipal Dynamic (Toronto: Ontario Economic Council 1974); and L.J. Sharpe, 'The Failure of Local Government Modernization in Britain,' Canadian Public Administration 24 (1981), 92-115.

150 / M. Sproule-Jones 22 The situation will change in Canada under the new Canada Act. See M.H. Sproule-Jones, Public Choice and Federalism in Australia and Canada, for a description and explanation of federal-provincial bargaining and its public policy consequences under the pre-1982 constitutional arrangements. 23 See John W. Mayne and Robert S. Mayne, 'Will Program Evaluation Be Used in Formulating Policy?' in this volume. 24 This is preliminary data made available by Jim McDavid. It should be noted that public-choice analysis evaluates the issues of privatization and nationalization in an empirical fashion. Contracting out to private producers can, for instance, be more expensive in certain circumstances, particularly if the producers can 'rig' the market. For a thorough recent overview, see Michael Denning, 'The Public Ownership of Productive Resources,' paper delivered at the Western Political Science Association Conference San Diego, March 1982. It should also be reemphasized that user charges are inappropriate financing mechanisms for pure public goods. A useful introduction to their applicability is Richard M. Bird, Charging for Public Services (Toronto: Canadian Tax Foundation 1976). 25 The classics in the field include James M. Buchanan and Gordon Tullock, The Calculus of Consent (Ann Arbor: University of Michigan Press 1962; John A. Rawls, A Theory of Justice (Cambridge: Harvard University Press 1971); Vincent Ostrom, The Intellectual Crisis in American Public Administration (Birmingham: University of Alabama Press 1973); Robert Nozick, Anarchy, State and Utopia (New York: Basic Books, 1974); Frederick A. Hayek, Law, Legislation and Liberties, 3 vols. (Chicago: University of Chicago Press 1973, 1976, 1979). 26 Robert L. Bish, Governing Coastal Resources (Seattle: University of Washington Press 1982). This book also includes a number of criteria of institutional fairness that can be used to evaluate provision system performance.

7 / The Management of a Common Property Resource: Fisheries Policy in Atlantic Canada SUSAN MCCORQUODALE

The fishery is an activity that affects public policy outcomes and generates a distinctive type of politics. Fish in the sea are not owned: they are, like air, water, or wildlife, a common property resource. H. Scott Gordon and Anthony Scott were among the first to draw attention to the economic (and indeed biological) consequences of this.1 Government policy, which first sought to deal with the conservation consequences of common property, has of late turned attention to the economic consequences. Some of these developments will be outlined below. It was in the field of evolving international law that the absence of property rights (in this case national property rights) had its greatest domestic impact in recent years. International management of the common fish in the northwest Atlantic, one of the richest fishing grounds in the world, meant almost no management and in the end brought near disaster, in the form of lower catches, to the fishing industry along the whole Atlantic coast. In response, both Canada and the United States in 1977 extended to 200 miles the boundaries of their territorial waters. While this move limited or controlled foreign fishing, it did not settle all issues. The boundary between New England and Nova Scotia remains disputed, and the limited joint management of the rich scallop fishery of the George's Bank creates, as before, the danger of the depletion of a common property resource.2 Within Canada the right to regulate commercial fisheries fell exclusively to the federal government under the provisions of section 91(12) of the British North America Act 1867 ('sea coast and inland fisheries'). The provinces, upon losing significant court decisions in 1914 and 1930, were forced to leave the definition of policy to the federal authorities.3 The extension of jurisdiction, however, had the effect of 'nationalizing' a new potential source of wealth. When this was tied to a new federal statement of policy, the provinces, now better equipped administratively and financially, began to demand a share in the decision-making affecting

152/S. McCorquodale an important provincial natural resource. These disputes over conflicting definitions of the right to regulate climaxed at the constitutional conferences of 1980; this essay will trace some of the issues involved, and will also outline the politics associated with the management of a common property resource. In the first place, it seems clear that the resource must in fact be 'managed.' Given easy and relatively low-cost entry into the fishery, it is possible to deplete the resource, perhaps not to the point of catching the last fish in the sea, but certainly to the point of catching the last profitable fish. In doing so, resources of labour and capital have been wasted and misused. Second, only the state can effectively manage the resource. It is true that in some instances (for example, the management of a village common grazing land or a small area of seabed) tradition and social sanction can limit the use of a common property resource. But once the economic activity grows to the point that the politics of social interaction are no longer effective,4 only the state can enforce the necessary sanctions. Third, the level of state intervention and the public goals being sought have changed over time: the goals of policy have become more ambitious and the instruments of policy more coercive. The politics of Canadian fisheries policy has involved a legislative response to the business and biological cycles to which the industry is subject, bureaucratic and pressure-group politics, and, recently, the politics of inter-governmental conflict between federal and provincial governments. The Atlantic Fisheries Fishing and the fish-processing industry must be set in a regional context in order to assess their relative importance. In terms of a variety of possible indices - employment or contribution to the GNP - the industry is relatively insignificant in the Canadian context. But if these various units of measurement are placed in an Eastern Canada context and within that context are set against individual provinces, then fishing becomes tremendously important. In Newfoundland, for example, in 1976 nearly 15 per cent of the labour force was employed in fishing and processing. In Nova Scotia the industry accounts for 10 to 15 per cent of value added in all commodity production.5 There are strong links in both provinces with the shipbuilding and service industries, and it has been estimated that about 75 per cent of the communities in Atlantic Canada (Newfoundland, the Maritimes, and coastal areas of Quebec) take part in commerical fishing. Of these communities perhaps 20 per cent (roughly 250,000 people) have no other economic base.6 Over the period from 1955 to 1980 Atlantic Canada landed about 90 per cent of the total Canadian catch, although this amounted to only 65 per cent of the total landed values.7 This is because the principal species caught on the east coast

Fisheries Policy in Atlantic Canada/ 153 TABLE 1 Atlantic sea fisheries-catches (0 and values (V) by province 1955-80, expressed as percentages of the Atlantic total Nova Scotia

1955 1960 1965 1970 1975 1978 1980*

Q

V

33 24 40 27 44 38 39

46 43 50 42 48 47 45

New Brunswick

Q 11

15 15 18 15 13 9

PEI

Quebec

Newfoundland

V

Q

V

Q

V

Q

V

13 15 11 13 13 12 9

2 2 3 11 2 2 3

6 1 1 8 6 6 5

9 1 8 11 7 6 7

6 6 6 8 8 7 10

43 41 34 40 32 40 41

30 26 25 27 24 28 29

SOURCE: Annual Statistical Review of Canadian Fisheries, vols 9 and 11 Government of Canada, Fisheries and Oceans * Preliminary figures

is cod, while that of the Pacific is the much more valuable salmon. Within the Altantic region the relative provincial positions in terms of quantity and value are as given in table 1. Nova Scotia's greater share of value comes partly from the value of the species caught (scallops, lobster) while Newfoundland's industry rests where it has always rested-on the cod. Employment in the fishery differs between these two provinces (as table 2 shows) as does the composition of their fishing fleets. The Nova Scotia fleet is largely, although not exclusively, trawlerbased, while Newfoundland's fishermen use smaller boats based in numerous scattered 'outports.' The primary resource sector is of significantly different size in the five provinces, contributing 49.6 per cent, 38.6 per cent, 27.4 per cent, 27.2 per cent, and 15.8 per cent to Newfoundland, PEI, Nova Scotia, New Brunswick, and Quebec respectively.8 Fishing as a component of this sector ranged from 10 per cent of PEI'S economy to less than 1 per cent in Quebec.9 Figures that measure the relative strength of each sector of an economy in terms of value added to industrial output do not measure that other critical indicator, employment. In Newfoundland, for instance, mining appears to outweigh fishing as an industry; but mining is very capital-intensive and ranks behind the fishery in terms of employment, generating about 6,000 man-years of employment annually compared to perhaps 19,000 in fisheries.10 In 1981 the federal government estimated that there were almost 70,000 fishermen and plant workers employed in the commercial fisheries of Canada.11 However, the fulltime man-years involved were probably less than 35,000. In the majority of cases fishermen had other part-time jobs. Actual employment in the fisheries on a

154/S. McCorquodale TABLE 2 Estimated number of fishermen in Newfoundland and Nova Scotia, 1965-80 Year

1965 1966 1967 1968 1969 1970 1971 1972 1973 1974* 1975 1976 1977 1978 1979 1980

Atlantic total

Newfoundland

Nova Scotia

49,335 45,918 45,210 45,709 42,931 41,757 39,726 39,741 38,996 36,464 40,564 40,785 n/a 48,533 55,885 60,032

21,701 20,286 19,814 19,355 17,770 17,765 15,961 14,452 15,313 12,792 15,802 15,351 20,243 26,447 32,352 35,080

14,049 13,067 12,589 13,108 11,717 11,018 10,688 11,735 10,600 10,460 10,435 10,409 n/a 10,311 10,799 11,432

SOURCE: Annual Statistical Review of Canadian fisheries, and for Newfoundland, Fisheries and Oceans Canada, Economic Services Branch, St John's * After 1974, the numbers represent licensed fishermen only.

province-by-province basis has been difficult to estimate, owing in part to the difficulty of classifying individuals as full-time, part-time, or occasional fishermen. In general there are three times as many fishermen in the Atlantic region as there are in the Pacific; figures for Nova Scotia and Newfoundland are given in table 2. These numbers, while they must be used with caution, show an unmistakable trend. Between 1965 and 1974, the numbers decreased as conditions in the fishery deteriorated, and then the trend began to reverse when foreign fishing was reduced and both fish stocks and markets improved. It has been estimated that in Newfoundland, between 1974 and 1979, the numbers of inshore fishermen increased 170 per cent.12 During the same period a transition in the technology of fishing took place in Canada affecting both fishing methods and onshore employment in domestic processing plants. The traditional Atlantic fisherman worked either alone or as shareman in a small inshore boat and used not very elaborate equipment to earn

Fisheries Policy in Atlantic Canada/ 155

a not very large income from fishing. A considerable portion of his catch was salted and sold to the West Indies, South America, and Mediterranean countries. However, in the years after the Second World War modern developments in refrigeration and transportation meant a gradually increasing proportion of the catch was being sold to the United States either as fresh or frozen product. Starting in the 1960s (and encouraged by government subsidies), private investors turned to the modern trawler to fish the 'banks' off Newfoundland out of southern Newfoundland and the Atlantic coast of Nova Scotia. These boats were large (over 150 gross tons) and could fish in all seasons of the year for a variety of species, including newly developed fisheries in herring and squid. Beginning in 1962 catches rose in both quantity and value, reaching a peak in 1968. A parallel growth occurred in employment in the processing plants, reaching some 15,000 jobs in 1981.13 This growth was accompanied, however, by a decrease in the number of fishermen, principally because of a serious decline in the share of the fish available to inshore fishermen. As investment in fishing craft in Atlantic Canada rose from $32.6 million in 1955 to $495.5 million in 197814 the specialization in provinces increased. Over 50 per cent of the offshore fleet is stationed in Nova Scotia, and it is in the interest of this fleet to hunt for fish in all the waters of the region in order to keep onshore processing facilities open. As table 2 indicates, however, over 50 per cent of Atlantic Canada's fishermen are in Newfoundland; most of these are inshore fishermen who are often in direct competition with offshore fleets for the same stocks of fish. Jurisdiction

The central instrument of Canadian fisheries policy is the Fisheries Act and its amendments which now run to 62 pages of text and 851 pages of regulations.15 Today's legislation has evolved directly from a pre-Confederation statute of the legislature of the United Canadas.16 From the beginning, the principle that governed fishery regulation was conservation. The absence of property rights dictated that fisheries policy would differ from that of forestry, for instance. Timber was recognized by governments to be a valuable asset that could be exploited by private interests which in turn would bolster the province's revenues. Fishing rights could rarely be sold (except rights relating to private inland streams) and as a consequence legislation was aimed not at revenue but at enforcing regulations with respect to seasons, gear, and catch limits. After Confederation, however, it transpired that the fishery provisions of section 91 of the BNA Act conflicted with the property and civil rights provisions of section 92. In 1892 Ontario successfully argued before the Judicial Committee

156/S. McCorquodale of the Privy Council that proprietary rights to inland fisheries in public waters lie with the crown in the right of the provinces. In 1913 and 1920, when British Columbia and Quebec claimed proprietary rights in tidal waters, the Privy Council ruled that proprietary rights existed only in non-tidal waters, and therefore commercial fisheries were within the exclusive jurisdiction of Parliament.17 It was conceded that in the case of Quebec, certain proprietary rights could also exist in non-navigable tidal waters. While the federal government might delegate administrative responsibility for inland fisheries to the provinces and authority over aspects of marine fisheries to Quebec, overall it has maintained its dominant position in the regulation of the fisheries. Until quite recently, the most serious abridgment of the federal government's regulatory power resulted from forces external to the country.18 Under the conventions of international law the legal jurisdiction of coastal states generally was limited to no more than the breadth of the territorial sea, defined as the waters three to twelve miles from land. Most of the world's marine fish stocks were to be found outside these narrow territorial limits and were thus beyond the regulatory reach of coastal states. It was through international joint commissions that some satisfactory solutions to problems of overfishing were found, for the Pacific halibut in 1923 and for the Fraser River salmon in 1930-7. The most important of these commissions, the International Commission on the Northwest Atlantic Fishery (ICNAF), came to prominence after 1951. In the next twenty years a virtual fishing explosion took place in the northwest Atlantic. New technologies and new foreign, distant-water fleets combined to virtually destroy certain species offish. It was not until the 1970s that ICNAF was able to establish national allocations within overall quotas. But this reform came too late. In 1977, Canada, arguing the evolving notion of'customary law,' took over most of the area, and by bilateral agreement phased out most nations' fisheries. Canada's fishery policy, in so far as it has been placed in the hands of the federal government, has been directed at the conservation of the resource. This 'philosophy' has been at the core of policy since pre-Confederation times. In large part this goal has been set by the nature of the resource itself. Fisheries policy, however, is as much about men and communities as it is about fish. Politicians and fishery managers have become preoccupied with the implications of their management policies for the income of local fishermen, who often live in communities with no other economic opportunities. Beginning in the 1930s, a variety of subsidies and stimulative measures have been aimed at encouraging expansion in the fisheries. Conservation goals have been supplemented by distributional goals and policies aimed at increasing the earning power of fishermen.19

Fisheries Policy in Atlantic Canada/ 157 The Evolution of Fisheries Policy 1950-70 It is useful to think of the evolution of fisheries policy in terms of cycles. Christopher Hood has argued that the cycles that are important to the fishery are biological and economic; in fact, these are the chief factors responsible for change.20 Biological depletion of a stock leads to economic cycles, which present the greatest challenge to policy-makers. Depressions in the economic cycles, as measured by low landed values for fish, seem to raise immediate pressure to devise schemes of subsidization that will cater to the cyclical change without propping up inefficient operations. This cyclical process can frustrate long-term planning because the short-run policies adopted to cope with economic cycles result in counter-productive responses. We can see in the policies adopted by the federal government during the 1950s and 1960s some evidence of the adoption of a long-term strategy, frustrated by the need to respond to short-term input demands. In the years just after the Second World War the Department of Fisheries made certain attempts at new policy directions. Particularly after 1949, when Newfoundland joined Confederation, it became necessary to deal with what was for the poorest province its most important industry. As economic enterprises, fish-harvesting and fish-processing in all of Atlantic Canada were backward. There was a notable absence of technological development, and there had not been a serious move into the explanding North American markets for fresh and frozen fish. Generally the industry was exhibiting none of the characteristics of a modern food industry such as economies of scale and standardization of specialized products (for example, breaded fish sticks). The cure for low incomes at the fishermen's level, low productivity at the industry level, and general undercapitalization seemed to lie in massive infusions of public and private money, with investment aimed particularly at the offshore. It was hoped that the creation of new job opportunities in other sectors would naturally draw some of the excess labour away from the inshore sector, either onto the trawlers or into the processing plants. Legislation passed in 1955 gave the federal government broad powers to guarantee loans from private financial institutions for virtually any fishing operation. The Fisheries Development Act of 1967 authorized the minister of fisheries to make payments for the 'construction, modification, conversion and equipment of fishing vessels' and to grant subsidies to the secondary sector for cold storage, ice-making, and related facilities. The result was a rich variety of subsidy and other stimulative measures, augmented by generous provincial loan board programs and, after 1969, by the spending of the Department of Regional Economic Expansion (DREE).

158/S. McCorquodale The direct costs of these programs were (and remain) very high, and the results were not in the directions anticipated. On the matter of costs, two studies, one done in the 1960s and one for 1978, suggest the nature of the problem. In 1961 Copes related the costs of direct support to the Newfoundland fishing industry to the value of fish landed. He concluded that support equalled 93 per cent of the value of the fish landed, or 65 per cent of the total income generated in the combined fishing and fish-processing industry. For comparison, he calculated that the net general expenditure of all levels of government in Canada on natural resources and primary industries as a percentage of the net value of production in the corresponding primary (resource) industries varied between 6.0 per cent and 9.5 per cent over the years 1954-60.21 To Copes it was a clear case of 'fiscal favouritism.' In a more recent case study on Prince Edward Island, R.D.S. MacDonald of the federal Department of Fisheries has concluded that using the most conservative interpretations, in 1978 it cost $13 million in government funding (administrative costs and subsidies) to produce something less than $16 million in income to fishermen and processing workers. Large sums for support translate into low average incomes for inshore fishermen of between $5,000 and $5,500 and approximately $6,500 for offshore fishermen.22 Policy-makers gradually came to understand that government outlays far exceeded direct revenues to government from the commercial fishery.23 At the same time, the incomes of individuals working in the industry remained low. It seems clear in hindsight that several of the strategies adopted in practice worked against what had been the initial aim of policy, the modernization of the offshore sector. Throughout the period, politicians and on-the-spot fishery managers had responded to the need to maintain employment and income. Subsidies and income support formerly extended to the traditional inshore sector had worked to draw men to this less efficient sector, or at least to discourage them from leaving. This point hints at another factor that worked against policy: the general economic conditions of Atlantic Canada. In many parts of the area there simply were no employment alternatives. If the government subsidized an increase in catching capacity through improved gear on larger boats, it did so in the hope of producing higher incomes. But higher income simply encouraged overexpansion of the fishery, resulting in a more rapid depletion of the resource, which in turn lowered the catch per fisherman. Another strategy was the attempt to move people from outports in the hope that if the catch was divided over a smaller number of men, incomes would improve. The 1965 Fisheries Household Resettlement Program in Newfoundland and the 1969 Development Plan for Prince Edward Island are seen today as failures. Fishermen were not easy to move, and job opportunities in designated 'growth centres' did not materialize. Many men and families ended up on

Fisheries Policy in Atlantic Canada/ 159

welfare rolls or drifted back into inshore fishing. The social costs had been high and nothing much had been accomplished.24 The extension of unemployment insurance benefits to fishermen, which began in 1957 as another attempt to protect individual incomes, is also seen today as being counter-productive. The Economic Council of Canada in a recent report on the Newfoundland economy argues that the present-day the uic system is drawing fishermen into the industry and appears to 'directly offset conservation programmes by enticing too many fishermen into the industry, each of whom works too little.'25 To round out this rather bleak picture, by 1974 it was clear that the industry itself and the resource on which it was based were in serious trouble. In large measure, the problem was an increased fishing effort in the international waters of the northwest Atlantic. Canada's total share of the catch offshore had fallen from 33 per cent in 1955 to a low of just under 20 per cent in 1973. At the same time, the total groundfish catch had increased from 1,260,000 metric tons in 1951 to a peak of 2,829,000 metric tons in 1965, and had dropped again to 1,568,000 tons in 1975. The resource was being depleted. The impact was felt immediately in the processing plants. Having expanded during the 1960s to cope with then-rising volumes of fish, the plants now began to suffer the effects of excess capacity. Serious problems were delayed for a time by a steady rise in the real price of fish products. By by 1974-5 rapidly rising fuel costs combined with the impact of the world-wide decline in fish prices brought the entire Atlantic coast groundfish industry to a state of crisis. Without large federal government subsidies - $130 to $140 million over a three-year period - widespread bankruptcies in the industry would have occurred. Pressure thus mounted on the federal government to work through the Law of the Sea Conference for Canadian control of the fish resources off its coasts and to reconsider domestic policy. The result of the former was the 1977 extension of jurisdiction; the result of the latter was a 1976 policy document entitled Policy for Canada's Commercial Fisheries, which signalled that a new plan had been adopted as government policy, which was to become a guide for the rebuilding of Canada's commercial fisheries over the next ten years. New Policies and New Actors Since 1977 we have witnessed various attempts to come to terms with the domestic implications of the 200-mile limit. The history and background of this event are recounted elsewhere.26 It is important to note, however, that the difficulties and delays encountered along the way are today responsible for part of the lack of confidence of provincial authorities in their federal counterparts.27

160/S. McCorquodale To provincial governments facing a serious resource depletion and to representatives of a vitally affected industry it seemed that Canada's international negotiators behaved like 'boy scouts in a thieves' den' and were not aggressive enough in pushing for national self-interest. In 1972 Mitchell Sharp addressed the Newfoundland Board of Trade and cautioned that 'Canada must continue to respect international law. We will take unilateral action only where justified.'28 At a Resources of the Sea Conference in 1973 Dr Art May, a senior official of the federal Department of Fisheries, was quoted as saying: 'It would be politically unrealistic to put forth the idea [at the Law of the Sea Conference scheduled for 1974] that coastal states have absolute control over fisheries resources... Canada favours less than what Iceland is asking for. [It favours instead] a more moderate approach. This preference is justified by the costs of conservation efforts. Costs must be balanced by "authority to manage." '29 Memories of this period underlie Newfoundland's desire for an assured constitutional place in the management of the fishery. Once a major source of Canadian fishery problems, foreign fishing, was removed, it was found that beneath that surface issue there existed a host of new questions. In nineteenth-century Canada, Hodgetts has noted, there was an absence of concentrated organized interest groups in the fisheries policy area partly due, he suggests, to the open-access nature of the resource. This absence, coupled with the apathy of the legislature, left more than the usual freedom for permanent departmental officials to develop what seemed to them an appropriate view of the public interest.30 One hundred years later, such wide discretion is no longer possible. Those groups that attempt to influence policy are no longer few in number or unorganized. In particular, the new era has brought the provincial governments most affected into the policy arena. It may be, as Hood argues, that electoral cycles have little impact on fisheries policy because the issue is non-partisan.31 In a federal state, however, provincial governments often provide the opposition to federal proposals. The federal government In the 1976 Policy for Canada's Commercial Fisheries, the Department of Fisheries and Oceans made explicit its desire for an integrative approach to the management of fishery.32 The Strategies adopted reflect a fundamental redirection in the government's policy for fishery management and development. Although commercial fishing has long been a highly regulated activity in Canada, the object of regulation has, with rare exception,

Fisheries Policy in Atlantic Canada/ 161 been protection of the renewable resource. In other words, fishing has been regulated in the interest of the people who depend on the fishing industry. Implicit in the new orientation is more direct intervention by government in controlling the use of fishery resources, from the water to the table, and also more direct participation by the people affected in the formulation and implementation of fishery policy. [Emphasis added]

Federal policy-makers have spent the intervening years working out the detailed implications of this policy. Attempts have been made to evolve new decision-making structures and procedures, and policy in general has shown a strong bias toward the inshore sector, possibly as a counterweight to the more highly organized processing interests, possibly as an expression of the policy choice of the minister, Romeo LeBlanc. The first three years of extended jurisdiction meant the holding back on demands from large processors for an immediate development in the offshore fleet. To the anger of the industry and the chagrin of the provincial governments, that period also meant a deliberate strengthening of the links between the federal government and fishermen's organizations by such things as granting to the latter a licence to engage directly in 'over-the-side' sales to foreign buyers. The aim was to break the monopsony power of the industry. The 1976 document warned that the federal government felt it necessary to 'apply systems of entry control in all commercial fisheries,' the most coercive policy instrument a regulator can undertake. First introduced in 1974, implementation of a licensing scheme that would categorize inshore fishermen as 'full-time' or 'part-time' has been difficult and delayed, but in 1981 with the full support of the vitally important Newfoundland fishermen's union, the licensing of inshore groundfishermen was undertaken. Licensing carries with it the implication that sometime in the future only a limited number of full-time fishermen will be granted the right to fish commercially inshore. The union supports the move as a means of increasing the professional status and income of inshore fishermen. The administrative organization of the Department of Fisheries, highly centralized in the 1950s, has transformed itself into a geographically disbursed and decentralized organization with regional directors-general in St John's, Halifax, Moncton, Quebec City, Winnipeg, and Vancouver. The department, established in 1867, suffered something of a bureaucratic identity crises beginning in 1969: it was linked first with Forestry and then became part of an omnibus Department of the Environment in 1971, only to be separated again in 1979. Continuity has come in recent years from the minister, Romeo LeBlanc, appointed to cabinet in 1974, who thus has had by Canadian standards a remarkably long association with only one department. LeBlanc has not hesitated to defend his department's

162/S.McCorquodale jurisdiction even to the point of public disputes with a cabinet colleague. In a recent book, Savoie describes how LeBlanc prevented the signing of a fisheries agreement negotiated by DREE with New Brunswick officials because the minister thought the agreement extended provincial jurisdiction into a federal area.33 Thus a decentralized department, led by a minister with a strong sense of his jurisdiction, who proclaims himself the 'Minister of Fishermen,' may in fact be, at least to a reasonable degree, as 'local' as the provincial ministries. The industry Throughout this period, changes have been taking place within the private sector which have in turn affected other actors within the policy environment. In the early months of 1977 the H.B. Nickersons firm of Lunenberg, Nova Scotia, took over a larger company, National Sea Products. In Newfoundland something similar happened in 1980 when the Canadian Development Corporation bought a 40 per cent interest in the second-largest company operating in that province. Access to capital, managerial expertise, and market strength were now available through large, multinational companies operating in both provinces. The initial interest of 'Nick/Nat' in freezer trawlers and joint ventures with European capital brought a confrontation of major proportions within Newfoundland where the provincial government, acting through a joint proposal with Nova Scotia, advocated a $900 million development plan. This plan called for the renewal of the offshore fleet and the building of ice-strengthened freezer trawlers.34 Inshore fishermen, through their union, opposed these schemes, fearing a loss of employment and competition for the resource. The jointventure discussions also precipitated a less heated but important dispute within the processing industry when the Independent Fish Producers' Association urged that processing should be dispersed along the whole northeast coast and Labrador, rather than being centralized in only one area.35 Eventually all of these proposals, from both Nova Scotia and Newfoundland, failed, partly for economic reasons and partly because the federal minister rejected them as being ahead of the recovery of the stock and an unwelcome intrusion of foreign capital into Canada.36 Thwarted in one direction, the larger companies have adopted an alternative strategy to gain access to cod stocks. Investment would be not in freezer trawlers, but in Newfoundland-based processing plants. The irony is that private capital may be providing something of the modernization and offshore development once favoured by the federal government at a time when federal policy has changed.

Fisheries Policy in Atlantic Canada/ 163 The provincial governments For Newfoundland and Nova Scotia, policy has followed a somewhat erratic course and seems to have gone through three phases. In the first phase, roughly from 1977 to 1979, the two provinces forged an alliance with each other and with the industry that favoured, offshore expansion. In the second phase, which will be discussed briefly here (a fuller examination is available elsewhere37), the two provinces acted together at the federal-provincial conferences of November 1978 and February 1979. In the third phase, from late 1979 to the present, Nova Scotia and Newfoundland (and indeed the other Atlantic provinces) have pursued separate courses. In the months prior to the first ministers' conferences of 1978 anger and frustration led Nova Scotia to support the gradually evolving Newfoundland position that the best solution to problems of fisheries management was the amendment of the BNA Act. At root, of course, Nova Scotia's interests lay with the federal position that 'there are no provincial fish - there are only Canadian fish.' The province's fishing industry was dependent on an offshore fleet, whose access to traditional fishing areas was now restricted for conservation reasons. The Canadian cod in the northern waters off Newfoundland looked inviting. But Nova Scotia's dissatisfaction with federal policy led it to support the Newfoundland position. Newfoundland argued that ownership of resources by the provinces is a generally accepted Canadian concept and that fishing was too important locally to leave it solely in federal hands. In a series of meetings with officials the two provinces drew up a draft constitutional amendment, which would have divided authority as follows:38 Assigned to federal paramountcy: - international negotiations - surveillance - international enforcements - basic research - applied research (to help set global quotas) - licensing of foreign vessels (based on residual quotas) Assigned to provincial paramountcy: - determination of quotas after global quotas had been set - division of quotas - harvesting plans - licensing of local boats and those of other provinces Included was the provision that should provincial agreement on the division of quotas fail, the final determination would be left with a provincially constituted review board.

164/S. McCorquodale At the conferences the federal authorities were unyielding. The scheme was unworkable and would produce a Canadian ICNAF. Besides, the federal minister was as much a representative of fishermen as were some others at the table and the information coming from his sources, from both fishermen's organizations and the processing industry, either urged the federal government to move cautiously or specifically rejected provincial claims to paramountcy over quota division, harvesting, and licensing.39 Gradually it became clear that the provinces had been deserted by the constituency they claimed to represent, and the so-called Northern Cod War of November-December 1979 accelerated a deteriorating partnership. This incident had been forced by the newly elected premier of Newfoundland, Brian Peckford, the newly elected minister of fisheries, James McGrath, a fellow Newfoundlander and member of the Clark government (May 1979 to February 1980). Peckford wanted the stock of fish that swim in northern waters off Newfoundland 'reserved' for the province either to be exclusively caught by Newfoundland fishermen or to be landed for processing only in Newfoundland ports. No federal minister of fisheries could accede to such a demand. In an atmosphere of mounting crisis, Peckford put teeth in his demands in December 1979 by putting all 'mainland' (i.e., Nova Scotia) fish companies operating in Newfoundland on notice that if they landed 'Newfoundland northern cod' for processing on the mainland, they would become ineligible for provincial assistance and their processing licences might not be renewed. Peckford might have thought himself to be defending a stock offish on which the inshore fishermen of Newfoundland had depended for 400 years, a resource of great cultural and emotional significance to his province, but the effect on Nova Scotia was immediate. Joint dissatisfaction with the federal government could not outweigh divergent provincial economic interests. At the constitutional conference of September 1980, while Romeo LeBlanc recited the familiar federal claim to jurisdiction, the former provincial allies collided head-on and the Nova Scotia minister of fisheries said the Newfoundland plan was 'unworkable.' With the support of New Brunswick, Nova Scotia abandoned its aspirations to provincial management of one of its major industries.40 Since that time the provinces have pursued their separate courses. Nova Scotia seems generally in support of corporate planning and management as the best guide to the prosperity of the industry. Newfoundland continues to link ownership of the fisheries resources, control over the offshore oil and gas resources, and the transmission of power from Labrador through Quebec. All form part of a province-building model that aims at the expansion of income and employment within the province. Fish stocks adjacent to the coast of a

Fisheries Policy in Atlantic Canada/ 165

province are seen as belonging to that province, and the preservation of the inshore fishery as vital to the economic and social well-being of the province. Newfoundland and the Structure for Decision-Making

The province of Newfoundland is making claims that reach beyond those of the other provinces of Atlantic Canada. This is in line with the economic importance of the industry, and with the historic patterns established in Newfoundland in the 1930s (and earlier) and arguably not yet established within any of the other Atlantic provinces. During the 1930s the Newfoundland government created a regulatory regime that was, relative to its size and resources, the best fisheries service in North America, and far superior to anything available in Canada's east-coast industry.41 At Confederation this lead in policy direction was lost, partly because the federal government was indifferent to a positive state role and partly because the provincial government turned to a variety of other development schemes aimed at breaking an age-old dependency on the fishery. It is worth noting, however, that between the 1950s and the 1970s provincial spending on the fisheries was, proportionately, double that of the other maritime provinces, and approached the share of gross governmental expenditures (approximately 1 Vi to 2 per cent) spent by the federal government on fisheries.42 Since the mid-1970s the province has re-established fisheries at the centre of its development strategies and has, through a series of sophisticated studies and the publication of a five-year plan, put forward expansionary plans that rival those of the federal government.43 No other maritime province has made such a political and administrative commitment to fisheries development. The private sectors of the industry in Newfoundland have also in recent years consolidated their structures and set broad objectives. One big union, the Newfoundland Fish Food and Allied Workers (NFFAW), has won the right to collective bargaining on a province-wide basis. The members of the union are drawn from workers in processing plants, trawler crews, and inshore fishermen. In the rest of the maritimes the fishermen's organizations remain fragmented. One recent count identified thirty-six fishermen's unions in the Maritimes and Quebec, each of which is concerned with specific problems in particular fisheries and geographic areas.44 On the 'trade' side too the pattern is the same. The Fisheries Association of Newfoundland and Laborador, alone of similar associations in the region, has consolidated its members and developed a full-time professional organization. The association, as a body, today buys and sells fish, charters vessels, negotiates trawler wage contracts for the whole industry, and, with the NFFAW, sets prices for various species and qualities of fish.

166/S. McCorquodale This greater interest and development in the industry has as yet found no responsible means by which to express itself in the federal policy arena. What is true now for the 'lead' province is, moreover, likely to be true in the future (if it is not already) for other regional interests, both governmental and private. The federal government said in its 1976 policy document that it aims at the direct participation in the formulation and implementation of fisheries policy by those affected. What we have not seen so far, however, is the establishment of new structures and procedures for decision-making. Conferences, seminars, and consultation have not succeeded in altering the seeming arbitrariness of federal government policy. Throughout the region the debate about policy has spilled outside the traditional forums and onto the front pages of Maritime and Newfoundland newspapers. Clusters of interest groups, each with a plausible case for what is 'best' for the industry, for Canada, or for a provincial community, seek not peaceful coexistence but hegemony. Often the voices raised for a particular point of view have a messianic quality, as each spokesman purports to show the one true way to deliverance for the Atlantic fishery.45 There is a continuing degree of animosity and mistrust among the various participants that is disturbing to concerned observers. In addition to this we must consider the phase into which federal policy seems to be moving. I stated earlier that state intervention in fisheries policy appears to have moved along a continuum from less coercive to more coercive. Federal policy in the years prior to about 1965 might be characterized as passive. Decisions were taken in response to representations from the industry, provincial governments, and the scientific community.46 Hood's thesis is that governments act only in response to immediate crises in the biological or economic cycles to which fisheries as natural resources marketed world-wide are subject, and that this tends to deflect policy-makers from certain long-term planning goals. We have seen the development of subsidy programs and emergency help to the industry in much the same light, certainly in the years preceding 1976-7. In the 1976 policy statement (and in its 1981 updated restatement)47 the federal department embarked on a long-term strategy aimed at the economic health of competing sectors of a domestic resource industry. In the years since the Department of Fisheries and Oceans has moved to implement these long-term goals, using what it itself calls the 'powerful and controversial tool' of fisheries management, limited-entry licensing.48 A growing volume of regulations must be arbitrated and policed by federal officials who are basically allocating access to wealth. The obvious consequence is controversy and political conflict. We have been studying a policy field in transition, one in which the environment, the goals, and the outcomes of policy are being redefined. Domestic

Fisheries Policy in Atlantic Canada / 167

concerns now match international ones, and the goal of conservation has been matched with an equal concern for the economic and social well-being of fishermen. In short, the outcomes of policy are in transition from those that are characteristically distributional to those that are characteristically regulatory.49 What has not sufficiently shifted, it would appear, is the structure through which conflict is to be resolved. The federal department seems to prefer a decentralized administrative structure and extensive consultation. Others, particularly the provinces, would prefer new structures. Conclusions

In this essay I have argued that two characteristics of the policy area affect the outcome of fisheries policy. First, the common-property nature of the resource alters the familiar legal and economic regimes under which resource policy is generally made. The absence of ownership allows for greater conflict over rights and jurisdiction. This conflict is enhanced in a federal state where provincial interests, which themselves clash, in turn challenge exclusive federal jurisdiction. Second, the resource is subject to biological and economic cycles outside the control of policy-makers, cycles to which they must respond in a variety of short-term measures which often, as in the case of unemployment insurance, end up in conflict with the long-term goals of managers. In Canada, the conflict that almost seems inherent in the policy area has become intensified in recent years as the federal government has moved from a passive to an active role. It was one thing for policy to centre on conservation and modernization in an era when the resource seemed limitless. It is quite another to undertake management when necessity dictates a reduction in effort to allow the restoration of nearly depleted stocks. To achieve this goal and at the same time increase the economic benefits to fishermen, the policy instrument being used, limited-entry licensing, is both controversial and coercive. As the regulatory state widens, controversy and conflict are enhanced. Common property creates jurisdictional conflict, and active policy-making, which seeks to smooth out the cycles to which the industry is subject, has intensified that conflict. It would seem that there is at least a plausible case to be made that, in order to reduce the present administrative and political costs of determining fisheries policy in Canada, there should be a division of the function between the federal and provincial authorities. Perhaps we could build on the model of intergovernmental delegation that Quebec was granted in a 1922 order-in-council.50 Alternatively, there could be delegation to a representative regional council modelled on the United States regional fisheries councils.51 In a recent paper,

168/S. McCorquodale Professor Scott has suggested a division of the regulatory function, partly to the market and partly to government. He argues that oversight of the market function (dealing with a transferable individual catch-quota system) might best fall to the provincial governments. The federal government would be left the management of the fishery and its environment.52 Whatever the model, the vital decision would be whether to divide the function. As I have said above, the current level of mistrust and animosity is disturbing to all those concerned with fisheries policy in Canada. The provinces would seem to have a legitimate concern over policy formulation and implementation, and as sovereign bodies are not likely to remain content with consultation. NOTES 1 H. Scott Gordon, 'The Economic Theory of a Common Property Resource: The Fishery,' Journal of Political Economy 62 (April 1954), 124-42; and Anthony Scott, The Fishery: The Objectives of Sole Ownership,' Journal of Political Economy 63 (April 1955), 16-24. 2 Sang-Myon Rhee, 'The Application of Equitable Principles to Resolve the United States-Canada Dispute over East Coast Fishery Resources,' Harvard International Law Journal 2\, no. 3 (Fall 1980), 667-83. 3 For Privy Council decisions see, A.-G. B.C. v.A.G. Canada [1914] A.c. 153, 15 D.L.R. 308 (P.C.); and A. G. Canada \. A. G. B.C. (The Fish Canneries Case), [1930] A.C. 111. 4 M.L. Miller and John Van Maanen, 'Boats Don't Fish, People Do,' Human Organization 38, no. 4 (1979), 377-85; and Susan Peterson and Leah J. Smith, Small-scale Commercial Fishing in Southern New England (Woods Hole, Mass.: Woods Hole Oceanographic Institution 1981). 5 Environment Canada, Fisheries and Marine Services, Policy for Canada's Commercial Fisheries (1976), 12. 6 Ibid., 7. 7 Annual Statistical Review of Canadian Fisheries, vols. 9 and 11 (Ottawa: Department of Fisheries and Oceans). 8 Statistics Canada Report no. 61.202 (1978), table 1. 9 Ibid. 10 Government of Newfoundland and Labrador, Managing All Our Resources (St John's; Newfoundland Information Services 1980), 5. 11 Government of Canada, Policy for Canada's Atlantic Fisheries in the 1980s: A Discussion Paper (Ottawa: Department of Fisheries and Oceans 1981), 3. 12 Gordon Munro, A Promise of Abundance: Extended Fisheries Jurisdiction and the

Fisheries Policy in Atlantic Canada/ 169 Newfoundland Economy, a study prepared for the Economic Council of Canada (Ottawa: Supply and Services Canada 1980), 43. 13 Atlantic Development Board, Fisheries in the Atlantic Provinces (Ottawa: (ADB 1969), 61. 14 Annual Statistical Review of Canadian Fisheries, vols. 9 and 11. 15 Ibid., and Policy for Canada's Atlantic Fisheries in the 1980s (1981), 3. 16 The Fisheries Act, RSC 1970, c. F-14. Other statutes include the Department of Fisheries and Oceans Act, the Fisheries Development Act, the Fish Inspection Act, and the Fishing and Recreational Harbours Act. 17 J.E. Hodgetts, Pioneer Public Service (Toronto: University of Toronto Press 1955), 143-50. 18 See note 5 above. 19 Parzival Copes, 'The Evolution of Marine Fisheries Policy in Canada,' Journal of Business Administration ll,nos. 1 and 2, 125-63. 20 Christopher C. Hood, 'The Politics of the Biosphere: The Dynamics of Fishing Policy,' in The Dynamics of Public Policy: A Comparative Analysis, ed. Richard Rose (Beverly Hills: Sage Publications 1976). 21 Parzival Copes, Government Assistance, Productivity and Income in the Fishing Industry of Newfoundland (§\. John's: Memorial University 1974), 31-2, tables 27-29. 22 R.D.S. MacDonald, 'Fishermen's Income, and Inputs and Outputs in the Fisheries Sector: The P.E.I. Case,' Canadian Issues 3, no. 1 (Spring 1980), 29. 23 Policy for Canada's Commercial Fisheries, 19. 24 Copes, 'The Evolution of Marine Fisheries Policy in Canada,' 150. 25 Economic Council of Canada, Newfoundland: From Dependency to Self-reliance (Ottawa: Supply and Services Canada 1980), 88. 26 R.D.S. MacDonald, 'Inshore Fishing Interests on the Atlantic Coast,' Marine Policy 3, no. 3 (July 1979), 179. 27 Hon. Brian Peckford, speech to St John's Rotary Club, Evening Telegram, St John's, 26 March 1981. 28 Evening Telegram, St John's, 26 Feb. 1972. 29 Evening Telegram, St John's, 17 and 19 May 1973. 30 Hodgetts, Pioneer Public Service, 144. 31 Hood, 'The Dynamics of the Biosphere,' 58. 32 Policy for Canada's Commercial Fisheries, 5. 33 Donald J. Savoie, Federal-Provincial Collaboration (Montreal: McGill-Queen's University Press 1981), 85-8. 34 Evening Telegram, St John's, 31 Aug. 1977. A 1978 provincial white paper on fisheries policy summarized the findings of a feasibility study by Kellog

170/S. McCorquodale

35

36 37

38 39 40 41

42 43

44 45 46 47 48

Consultants. Included in the latter was a recommendation for a primary landing and distribution centre (PLDC) to be located at Harbour Grace, Newfoundland. Norco Report, It Were Well to Live Mainly Off Fish: The Place of the Northern Cod in Newfoundland's Development. Prepared for the government of Newfoundland and Labrador, February 1981 (St John's: Jesperson Printing Limited 1981), 55. Evening Telegram, St John's, 10 Feb. 1978. Sheilagh Dunn, S. McCorquodale, and A.P. Pross, 'East Coast Fisheries: Economic Resurgency and Constitutional Issues,' unpublished paper presented to the Atlantic Provinces Political Science Association Conference, College of Cape Breton, Oct. 1980. Ibid., 41. Ibid., 45. Ibid., 47-9. David Alexander, The Decay of Trade: An Economic History of the Newfoundland Saltfish Trade, 1935-65. Newfoundland Social and Economic Studies, no. 19 (St John's: Memorial University 1977), 37. Statistics Canada, 68-207 and 68-211, relevant years. (1) Setting a course: A Regional Strategy for Development of the Newfoundland Fishing Industry to 1985, Government of Newfoundland and Labrador, Department of Fisheries, Aug. 1978, 4 vols.; (2) The Position of the Government of Newfoundland and Labrador on the Harvesting of the 2J and 3KL Cod Stocks,' presented at the Government/Industry Seminar, Corner Brook Aug. 1979; (3) White Paper on Strategies and Programs for Fisheries Development to 1985, St. John's Nov. 1978; (4) Towards the Twenty-First Century-Together: The Position of the Government of Newfoundland Regarding Constitutional Change, St. John's Aug. 1980; (5) Managing All Our Resources: A Development Plan for Newfoundland and Labrador, 1980-1985, St. John's Oct. 1980. The government of Quebec has also produced a five-year plan; see Gouvernmen du Quebec, Ministere de 1'agriculture, des pecheries et de 1'alimentation, Direction generale des peches maritimes, Strategic Quinquennale d'allocation des ressources pour les peches maritimes de Quebec 1980-1984, reimpression juillet 1980. MacDonald, 'Fishermen's Income,' 179. W.J. McCarten, 'Economics of the Fisheries in the 1980s Maritimes and Quebec,' speech to the Fisheries Council of Canada annual convention, 13 May 1981, 3. Copes, 'The Evolution of Marine Fisheries Policy,' 133. Policy for Canada's Atlantic Fisheries in the 1980s: A Discussion Paper (Ottawa: Supply and Services Canada 1981). Ibid., 29.

Fisheries Policy in Atlantic Canada/ 171 49 Theodore J. Lowi, 'American Business, Public Policy, Case-Studies, and Political Theory,' World Politics 16 (1963-64), 677-719; and The Structures ofPolicymaking in Canada, ed. Bruce Doern and Peter Aucoin (Toronto: Macmillan of Canada 1971), 19-22.

50 P.c. no. 360, dated 13 Feb. 1922. 51 William R. Rogalski, 'The Unique Federalism of the Regional Councils under the Fishery Conservation and Management Act of 1976,' Environmental Affairs 9 (1980), 163-201. 52 Anthony Scott, 'Regulation and the Location of Jurisdictional Powers: The Fishery,' unpublished paper prepared for Liberty Fund Symposium on Regulation, Alton, Ontario, Oct. 1981.

This page intentionally left blank

8 / The Political Economy of Policy Instruments: Tax Expenditures and Subsidies in Canada KENNETH WOODSIDE

The rapid growth in the size of government since the Second World War has been accompanied by an explicit and widespread use of policy instruments to intervene in the operation and change the character of the economy. While intervention was already practised by Canadian governments prior to the war, the espousal of Keynesian economic doctrine by the political elites during and after the war resulted in an expanded use of interventionist techniques by politicians and officials. The growing intervention by governments has made analysts of all political persuasions aware of and interested in the actions of politicians and the institutions of government in attempting to influence the working of the economy. This relationship between political and economic factors is the focus of study in political economy. One approach involves an examination of the ways that different policy instruments used by government involve different political processes and are used to solve the problems of different social and economic groups. It can be argued that politicians and officials face a market in policy instruments, and that a range of factors will contribute to their choice.1 Governments may attempt to regulate certain activities in order to eliminate or minimize undesirable behaviour. Alternatively, governments may prefer to provide a service through a public or crown corporation either de novo or as a result of some method of nationalization. Governments may also attempt to change or influence behaviour through some form of fiscal activity such as monetary or taxation policy-making or through the spending of public funds. Each of these policy instruments may be used separately or in some combination as a means of satisfying a mix of policy goals. In this essay I will undertake a comparative analysis of the different political processes that encompass two of these policy instruments, the tax incentive and the expenditure subsidy.

174/K. Woodside A tax incentive, or tax expenditure as it has come to be described in recent years, is a provision within the tax statutes that affords special treatment for certain designated types of income or wealth. This special treatment takes the form of a reduction in tax liability; it is called a tax expenditure because the recipient benefits by the amount of the reduction in taxes. A subsidy is a direct payment made by government to individuals or organizations that satisfy certain conditions. Each of these two policy instruments can be used to achieve a diverse range of goals. Expenditure subsidies may be used to assist Canadian shipyards in building, repairing, or converting ships, or to encourage multinational automobile companies to set up parts production facilities in Canada rather than in the United States, Europe, or in Southeast Asia. Expenditure subsidies may also be used to supplement the incomes of low-income pensioners or to provide family allowances. The diversity of such expenditures is matched by tax incentive provisions. While taxes have traditionally been thought of as instruments to raise revenue, over the past few decades they have also been used to pursue a wide range of other goals. For instance, the traditional personal income tax system with its standard deductions related to family size and dependency has been further complicated by deductions aimed at promoting home ownership and establishment of retirement savings plans. Similarly, tax incentives have been introduced to encourage the installation of pollution abatement equipment by industry and to promote the location of industry in the less developed areas of Canada. In the next section I will discuss the general character of these two policy instruments and attempt to relate their characteristics to the management of conflict and other general goals of politicians. In the third section I will assess the extent to which the instruments are used. The fourth section will compare the different political processes within the government and Parliament that produce the two different types of policy output. In the fifth section this comparison will be extended to the political process outside the government and Parliament in order to assess differences in the responses of constituent groups to each policy instrument as well as the broader income distribution implications of each type of policy instrument. In the sixth section some of the administrative and efficiency considerations of any choice will be discussed. Finally, some proposed reforms of the tax policy process and the 1981 effort to eliminate some tax expenditures will be considered. The focus is largely on the federal government. This reflects the character of Canada's taxation system, where most tax expenditures have been introduced by the national government and provincial taxes are expressed as a function of the federal tax.

Tax Expenditures and Subsidies/ 175 Policy Characteristics: Taxes and Subsidies

In recent years the wide use of tax incentives has become a topic of growing controversy. Public-finance specialists, in increasing numbers, have begun to distinguish between the normal tax structure - those sections necessary to the revenue-raising function of the tax structure - and what they call tax expenditures.2 The purpose of the distinction has been to highlight the use of tax measures as a form of expenditure and to distinguish it from the usual revenueraising function. While there are difficulties associated with this distinction, tax expenditures may be defined as special provisions in the tax statutes, applicable to particular types of business or sources of income, which result in those designated types of income being taxed at a lower rate than would otherwise be levied.3 We will use the terms 'tax expenditure' and 'tax incentive' interchangeably. Tax incentives are viewed by most public-finance practitioners as being conceptually equivalent to expenditure subsidies, except that the former are administered through the tax system. Although some tax expenditures are difficult to reformulate in the terms of an expenditure subsidy, they do represent deviations from a normal rate of tax and thus involve a tax saving, which is passed on to those who are eligible to receive it. The net result of a tax expenditure is that tax revenue that ordinarily would have been collected by the government is not collected. This is a crucial difference between tax expenditures and expenditure subsidies. Whereas the subsidy involves money that comes to a recipient from the government in the form, for instance, of a grant - a transaction that involves government ownership and control over the money prior to receipt of the grant-the tax expenditure involves money that the government has not taken from the recipient, which suggests that the recipient has been and remains the owner of those funds. The tax expenditure is thus a less direct form of government action, involving the failure to do something (i.e., to tax income at the regular rate) that the government supposedly would otherwise have done. In a mixed economy with a strong private sector and supported by widely held beliefs in free enterprise, the directness or indirectness of governmental action may be an important factor in policy instrument choice. A second and related difference between tax expenditures and expenditure subsidies concerns their visibility. While expenditures will be funded through the Consolidated Revenue Fund, to which all taxpayers have contributed, and leave an audit trail that clearly ties them to government action, the tax expenditure involves forgone revenue, the cost of which appears less obviously to be a burden on other taxpayers.4 The visibility or relative invisibility of government

176/K. Woodside behaviour is important because, when government acts, its behaviour is discriminatory. Government interventions are bound to benefit one group more than another.5 They are both a response to and a source of conflict within society and, as a result, the extent to which the government is seen to be favouring or abusing any group or individual may well rebound to its disadvantage. The importance of the visibility or relative invisibility of the exercise of government power, therefore, is that this exercise of power entails winners and losers, and thus conflict. The extent to which the action is easily associated with the government can become an important variable in government decisionmaking. Incumbent politicians normally will seek to minimize the level of conflict focused on them in order to enhance their chances of staying in office. The desire to be re-elected, to promote one's career, and to reward one's allies usually leads to efforts by politicians to control and minimize the level of conflict created by the interventions of their governments.6 The two characteristic differences between tax expenditures and expenditure subsidies - their directness or indirectness and their visibility or relative invisibility-are obviously closely related. The more indirectly a government policy instrument operates, the less visible the government's actions will probably be. A policy instrument that is more direct in its application will make the source of the policy more visible. These underlying differences are reflected in and enhanced by the political processes that produce the different forms of intervention. The Canadian Experience

Before proceeding with a comparison of the political processes and characteristics that differentiate the use of tax expenditures and subsidies, a brief assessment of the extent of their respective use is necessary. First, evidence will be provided on the amounts of money being spent by Canadian governments, especially the federal government, through expenditure subsidies and tax expenditures. Second, we will examine the extent to which tax expenditures and subsidies have been growing in relative importance. It will be seen, first, that huge sums of public funds are being dispersed by both expenditure subsidies and tax expenditures and, second, that in recent years the growth in tax expenditures has substantially exceeded the growth in expenditure subsidies. The federal government spent just over $53 billion in fiscal year 1979-80. This money was spent on a wide range of objects from government salaries, police protection, and welfare payments to medical services and subsidies to corporations. While it is difficult to determine what proportion of this spending subsidized non-governmental activities, about $10 billion went for income

Tax Expenditures and Subsidies/ 177

maintenance activities and over $5 billion for economic development and support.7 When people talk about government spending, they are usually referring to this accounting for expenditures. Indeed, until very recently it was the only accounting available. However, as we have already argued, there is another type of expenditure accomplished through the tax system. In 1979 the federal government published its first description of these tax expenditures. According to the estimates provided, the federal corporate and personal income tax structures alone incorporated special provisions which reduced tax revenues by almost $26 billion, a sum equal to almost one-half of federal expenditures for that fiscal year.8 Furthermore, this Tax Expenditure Account did not include all tax incentives - many were omitted or not accounted for 'because of resource constraints and data limitations.'9 While previous estimates of the size and scope of tax expenditures were often controversial with respect to both their methodology and conclusions and while the federal government estimate is inconclusive in character, it is useful to review the findings of this literature. The first attempt to measure tax expenditures was made by the Royal Commission on Taxation, which estimated the revenue cost of a selected number of corporate income tax measures and several other personal income tax exemptions at 16 per cent of the total governmental revenue for 1964 derived from personal and corporate income taxes.10 A decade later, David Perry estimated that all corporate income tax expenditures for 1973 cost Canadian governments 67 per cent of the total federal and provincial corporate income tax revenues collected.11 In a subsequent study, Roger Smith estimated slightly higher ratios of 74.9 per cent and 68.7 per cent of corporate tax revenues for 1974 and 1975 respectively, using the same tax measures studied by Perry.12 A study done by the National Council of Welfare estimated that a sum equivalent to 62 per cent of federal and provincial personal income tax revenue in 1974 had been disbursed in the form of tax expenditures.13 In a subsequent study, the council estimated that twenty personal income tax expenditures alone cost the federal government $7.1 billion in 1976, a sum $800 million greater than the federal deficit for that year (whose size was, at the time, the subject of great controversy).14 The National Council of Welfare studies, however, included a number of exemptions that usually are considered to be part of the normal income tax structure, not tax expenditures. Allan Maslove estimated that 31 per cent of personal income tax revenue had been lost as tax expenditures in 1976, a sum over 50 per cent higher proportionately than that estimated by Jonathan Kesselman for 1973.15 Given the introduction of a large number of exemptions and deductions in the intervening years, these two estimates are roughly compatible.16 In assessing all the studies, the obvious conclusion is that while the federal

178/K. Woodside government has spent substantial sums as expenditure subsidies, it has disbursed at the same time large amounts as tax expenditures. Indeed, in areas such as economic development, tax spending has exceeded expenditure subsidies, at least in recent years.17 Not only do tax expenditures amount to substantial sums when compared to government expenditure subsidies, they have also been growing at a faster rate.18 Edward Tamargo estimated that tax expenditures had grown by 42 per cent from 1976 through 1979 as compared to a 30.4 per cent growth rate for expenditures.19 Roger Smith, in a projection of Kesselman's analysis for 1973 through 1976, found that personal income tax expenditures increased by 50 per cent in 1974, 18 per cent in 1975, and 34 per cent in 1976.20 Richard Bird, as reported by Smith, estimated federal and provincial corporate tax expenditures at 37 per cent of total corporate tax expenditures collected in 1964 by both the federal and provincial governments.21 In 1973 David Perry estimated federal and provincial corporate tax expenditures at about 67 per cent of total federal and provincial corporate taxes for that year; while Roger Smith's extension of Perry's analysis to 1975 set the value of corporate tax expenditures at almost 69 per cent of total federal and provincial corporate tax revenues.22 All of these studies indicate that, compared to expenditure subsidies, tax expenditures have been growing at a relatively rapid rate in recent years. This surge in tax spending has occurred at the same time that a heightened sensitivity to and gradual reduction of government expenditures has been developing. The Political Processes in Government and Parliament

In comparing the processes through which tax incentives and subsidies are devised within the government and passed by Parliament, the most important element may well be the range of interests and actors that play an influential role in each area of the policy-making process and must be accommodated. Taxation policy decision-making is guided by and carried on within the parliamentary tradition of budget secrecy. This doctrine constrains open discussion of proposed tax changes even within the cabinet, and therefore provides fewer opportunities for opponents of a measure to intervene before the government is committed to a position by the finance minister in his budget speech. As well, it reduces the ability of the government's tax officers to obtain outside advice on possible alternatives.24 The result is that tax policy decisions are made by a small number of officials within the Department of Finance and are less directly subject to the scrutiny of interested parties.25 In the case of subsidies, such wide-ranging scrutiny is the norm. All proposals are subject to review and analysis by the appropriate department, the policy-

Tax Expenditures and Subsidies/ 179

sector committee, the Treasury Board and its secretariat, and cabinet.26 Departmental proponents of a measure will have to confront and justify its costs to the Treasury Board Secretariat. Subsidies normally involve complicated interdepartmental negotiations, and consequently their approval is more awkward. The net result is that decision-making power over tax policy is highly concentrated, and that more actors and a wider range of interests are involved in the review of subsidy decisions. The new system of 'policy envelopes' introduced by the Conservative government in 1979 encompasses both tax expenditures and subsidies, so that tax incentives sought by departments other than Finance must undergo more rigorous review procedures involving the policy sector committee, the Treasury Board, Finance, and cabinet.27 However, Finance's ability to introduce independently new tax incentives and expand or reduce others has not been affected by this new system. While the authority of the Department of Finance in tax policy may have been somewhat eroded over the last decade, the department continues to operate within a policy process comparatively well insulated from many political pressures.28 A second factor differentiating the political processes of tax incentives and subsidies within the government is the implications of each for policy administration. All tax measures are administered by officials in National Revenue although in some cases the use of a tax incentive may have to be approved by the relevant department. Thus, while an accelerated capital cost allowance for pollution abatement equipment was introduced in 1965, it was not until 1970 that claimants needed certification by Environment Canada to establish eligibility.29 Subsidies are usually administered by the department normally responsible for the relevant area of activity. The result may be badly administered policy when tax incentives are used in place of subsidies because the government administrators may be more concerned with the integrity of a provision as a tax measure than they are with its cost-effectiveness or its implications for the policy area generally. Third, recent developments in the character and complexity of federalprovincial relations make tax expenditures more attractive as a policy instrument for the federal government. In most areas of government policy apart from taxation the federal and provincial levels of government pursue highly interdependent policy strategies, and unilateral action by one level of government can be a source of concern and ill will for the other level.30 Co-operation has been increasingly difficult to obtain because of the conflict between pressures of province-building and nation-building.31 However, while subsidies may be introduced only after lengthy negotiation between the federal and provincial governments, the federal government can still act unilaterally in the area of taxation. While the sales tax dispute of 1978 and the 1980 proposals for resource taxation

180/K. Woodside may be harbingers of a more complicated and interdependent taxation policy process in years to come, this area remains one in which federal autonomy, both de jure and de facto, is high.32 Since tax expenditures appear less intrusive as a result of their more indirect character, federal action through the tax structure is usually construed to involve less interference with provincial government responsibilities. The irony of this is that federal tax expenditures may not only involve intrusions into provincial areas of legislative competence but may also entail a loss of revenue to the province because of shared tax bases and the provincial tax levies being calculated as a percentage of the basic federal tax.33 A fourth area of comparison relates to the character of information available to the public concerning the use of tax expenditures and subsidies. Until 1979 the government published no data on the cost of tax measures beyond the occasional estimate of the cost of a new provision for the first year, which would be presented in the budget statement. Since 1979 the government has published an annual Tax Expenditure Account, but these documents leave much to be desired. The tax expenditure estimates are far from complete; they show no totals for even sub-categories of tax expenditures, and they are full of disclaimers as to the ultimate value of the exercise despite the large sums involved.34 The costs of subsidies, on the other hand, are detailed annually in the main or supplementary estimates and, unlike tax expenditures, are audited by the auditor-general. Further, descriptions of the purposes and goals of subsidies, while not above criticism themselves, tend to be better documented than tax expenditures.35 The relatively greater visibility of subsidies contributes to the apparently greater interest of MPs and even prime ministers in the examination of subsidies, and thus makes the introduction of new subsidies somewhat more subject to cgmment and evaluation, if not control, within Parliament.36 A fifth and final area of comparison with respect to the political processes involved in tax expenditures and subsidies relates to the political process within Parliament. At first glance the differences in process are substantial, although, as we shall see, there are grounds to question the extent to which one process is more effective as a means of control than the other. Changes in tax incentives are normally introduced into the House through the budget speech. Until that time the proposals will have been kept secret from all but a few officials within the Department of Finance and, as budget day approaches, the prime minister and finally cabinet.37 The proposed tax changes will not have been debated within the government party caucus. Still, once the changes are introduced through the budget, the government normally regards itself as committed to the stated position, and for this reason is quite inflexibile toward admendments.38 The budget proposals will receive detailed committee examination from the Committee of the Whole, in which the opposition is unable to call expert witnesses or

Tax Expenditures and Subsidies 7181

conduct hearings. The result of this procedure is that proposed tax expenditures are subject to relatively little scrutiny once they are introduced. Even where a tax measure includes a time limit providing for its expiration, the tax expenditure is regularly extended with little comment.39 Moreover, some tax changes can be introduced by regulations issued through order-in-council as a result of provisions in the Income Tax Act.40 In this way parliamentary review can be avoided almost entirely. Subsidy proposals will have been subject to much more open discussion and analysis within the government and within the caucus and, on introduction to the House, are referred to a standing committee for detailed examination after second reading. The standing committees can receive testimony and evidence from outside experts and can even meet if the House is not sitting.41 Once passed, the estimated cost of a subsidy will always be reflected in the estimates as long as it is in place and even though the measure itself does not receive specific examination frequently. These differences suggest that tax expenditures are subject to significantly less careful examination and control by Parliament than are subsidies. While in general it is true that the scrutiny of tax proposals is less far-reaching, many questions have been raised about the effectiveness of parliamentary control over the expenditure process. Various critics have noted a long list of problems with the House of Commons standing committees, including poor attendance by MPs, a lack of expertise and background knowledge which might make MPs' committee questioning effective, an excessively high turnover in committee membership that makes expertise difficult to acquire and does not allow good working relationships to develop, excessive control over committee work by the government, inadequacies in the format of the estimates as they are received by the House and the committees, and a lack of interest on the part of MPs in really controlling spending costs.42 By the time that the estimates reach the House, they have already gone through a long negotiation process and reflect a balance of interests reached within the government, thus making the government more resistant to adjustments in its plans.43 The Royal Commission on Financial Management and Accountability reported in 1979 that 'ministers and officials consider review of Estimates, for example, a waste of time.'44 Finally, many government expenditures, both statutory and non-statutory, are far from easily changed because they reflect a consensus of the federal and provincial governments at a particular point in time or are necessary to provide for the orderly operation of the country.45 Looking back over the argument so far, it is clear that tax expenditures are the product of a political process subject to a more concentrated and narrowly based range of interests, that they offer more freedom to the federal government within the context of inter-governmental relations, that their costs are somewhat

182/K. Woodside less visible, and that their passage through the House of Commons is easier and involves less careful scrutiny. These features make tax expenditures attractive to some elements of the government in that, depending on what the government is attempting to achieve, it can act more decisively and less conspicuously through the tax system. Among these factors the relatively lesser visibility of tax expenditures as compared to expenditure subsidies is of great importance. It perpetuates the existing pattern wherein a narrow range of interests are influential in tax policymaking and it generates less pressure to increase the scrutiny of tax expenditures by Parliament and to produce more information about tax preferences for public evaluation. The lack of visibility also generates less concern over conflicts and inconsistencies in departmental responsibilities and makes the provincial governments either less willing or less able to oppose tax reductions that interfere with their jurisdiction while benefiting their constituents. To the extent that visibility and an openness of the policy process to the broadest range of interests possible are positive features of a healthy and balanced policy, the characteristics of the process of legislating tax expenditures do not contribute to the healthy functioning of a democratic state. The Political Processes beyond Government and Parliament

Looking at the political process more broadly, a first area of comparison is the distributive and equity implications of each policy instrument. Can choice of policy instrument affect who gets what when? Distributive implications can be seen in three areas: with respect to individuals, with respect to corporations, and as between Canadians and non-residents. In the case of the distributive implications of each instrument for individuals, it must be cautioned that tax expenditures or incentives are not necessarily regressive, nor are expenditure subsidies necessarily progressive. Indeed, there are some measures of each kind in each category. However, as tax expenditures are usually structured, few can be described as progressive. This is because of the way that most of them are designed, and especially because of what is called the 'upside-down effect' of tax expenditures where those with the largest incomes receive the greatest absolute benefit.46 Tax expenditures are normally available through an exemption of income from taxation, a deduction from taxable income, a credit against taxes payable, or a reduction in the rates applicable to certain income. A common feature is that one must have taxable income in order to receive any benefit. Moreover, the more income one has, the better off one is likely to be.47 As a consequence, a tax expenditure with social-policy implications (such as a deduction for child care or health care expenses against taxable

Tax Expenditures and Subsidies/ 183

income) will provide the greatest benefit to those with the largest income because their high income makes them subject to higher rates of tax under the progressive income tax structure. Tax expenditures by their very nature, therefore, tend to be regressive in character.48 While many expenditure subsidies such as subsidies to shipbuilders or pulp and paper firms are likely to be equally advantageous to the well-to-do, a subsidy can be more directly targeted to benefit individuals equally despite variations in their incomes, or indeed especially to benefit those with the lowest incomes. It is largely through the use of subsidy instruments that most social security and welfare payments are implemented, although the 1978 introduction of a refundable child tax credit is a small step toward making tax expenditures more useful to those with lower incomes.49 With respect to corporations, tax expenditures tend to be of more financial benefit to large firms than to small firms. Taxes are not paid on tax expenditures (they are usually paid on subsidies or grants). A corporation must make a profit in order to be able to make use of the tax expenditure. The larger and more regular the profit, the greater will be the benefits of these tax expenditures. Thus, since small, private, Canadian-controlled corporations pay taxes at the rate of 25 per cent of taxable income they will benefit less from tax expenditures that reduce their taxable income than larger and more profitable corporations paying tax at the normal rate of 46 per cent. Further, tax expenditures that reduce the cost of capital or reduce the cost of borrowing money to take over another firm will benefit the larger firms that are better able consistently and effectively to use the tax saving potential that such actions generate.50 Subsidies can be made available to firms regardless of income and profitability, and the benefits available to the recipient company may even be geared to provide more benefits for the less well-off firm. Finally, there are the differential benefits available to residents and nonresidents as a result of tax expenditures and subsidies. Distinctions between the two groups, while mentioned on occasion in the tax code, have not been the rule, and indeed the most general and valuable tax expenditures (especially those provided for the resource industries) have tended to provide greater benefit to non-resident corporations and individuals. The Royal Commission on Taxation estimated that 78 per cent of the revenue derived from its recommended eliminations of tax expenditures would have been extracted from non-residents.51 The high degree of foreign ownership in the manufacturing sector also ensures that the generous tax expenditures available to that sector will be particularly useful and valuable to non-residents.52 Subsidies, such as those proposed in 1980 for the oil and gas sector by the federal government, can be more effectively and carefully targeted and designed so that foreign firms are not eligible or only eligible under special circumstances.53 However, in practice it is not clear that

184/K. Woodside subsidies to the corporate sector have been used much more frequently to aid resident over non-resident firms. The favouring of Canadian residents over non-residents may put Canada in conflict with its obligations under GATTnot to discriminate against foreign investors. In reviewing the distributive implications of tax expenditures and subsidies it is evident that the tax system as a means of program delivery works to the advantage of wealthier rather than poorer individuals and corporations. Although tax measures could be designed to reverse this impact, in practice this has not been the case. While subsidies have been used for the whole range of distributive programs, it is through them that governments have sought to assist the less fortunate individual and corporation.54 The character of tax expenditures is reflected in the political process that generates them. David Good has described the making of Canadian federal tax policy as the product of an informal process dominated by the interests of business groups and (to some degree) their interpretations of likely outcomes, all of which is balanced against the financial requirements of the government.55 The mutual trust and informality in the process would appear to work to the advantage of those who are well organized, able to maintain a continuing presence, and establish relations of 'mutual trust,' those who share a common perspective with Finance officials, and those who command resources that the government needs or wishes to influence. A second area of comparison closely related to the first is the regional impact of programs implemented through the two policy instruments. In Canada, the regional distribution of government spending has become an important factor in assessing the desirability of that spending. The government has established and developed programs such as those administered by the former Department of Regional Economic Expansion that are particularly intended to encourage economic development in areas of low growth. The expansion of regionallysensitive subsidy programs proceeds despite the considerable doubt that exists as to their effectiveness or desirability in strictly economic terms.56 While tax expenditures such as the investment tax credit work to the advantage of emerging firms or firms expanding in less economically developed regions, subsidies have the advantage of helping firms whether or not they are profitable. Subsidies may also be useful to small and new firms in a poor cash-flow situation. Tax expenditures can allow government to act with less concern for the regional impact of a policy. Since a subsidy is a more visible policy instrument than a tax incentive, by using the subsidy the government will not only be attempting to promote business in the less economically developed regions but will be seen to be doing so. The greater visibility of subsidies makes them-more symbolically satisfying for the federal government in a time of heightened regional sensitivity. Tax expenditures do not usually require that a firm be located in a particular part

Tax Expenditures and Subsidies/ 185

of the country; as a result, they allow the government to act with less concern for the regional impact of its policy. A third area of comparison flows directly out of the greater visibility of subsidies as government programs. The growth of anti-government sentiment in the 1970s and the concomitant emphasis on government expenditures as a measure of the weight of government activity has encouraged proponents of new or expanded programs to pursue their goals through other policy instruments. Since the costs of tax expenditures have not been as well documented, increased reliance on tax expenditures to implement government programs has been a less visible, less conflictual way of expanding government action. As a result, the appearance of government restraints reflected in expenditure cuts could coexist with actual program expansion through the use of new or broadened tax incentives.57 The pressures for government restraint also made subsidies a less dependable form of aid, as they were more readily subjected to pressures for delay as a measure of spending restraint and, unlike tax expenditures which are administered by National Revenue, they might require the creation of a new bureaucratic apparatus to administer the program.58 Because tax expenditures do not appear in the estimates, they are less subject to restraint and, indeed, governments have received little credit for restraint in this area. In fact, tax expenditures may appear costless to their proponents. Until the publication of the 1979 Tax Expenditure Account there was no official tabulation of the costs of tax expenditures, and there is evidence that the costs of tax incentives were not monitored with any care.59 Most tax expenditures are not subject to any limits, and even where there are limits (as with the investment tax credit) they apply only to the individual applicant and are not a global budgetary restraint for the program as a whole.60 Whether the new envelope system will alter the perception that tax spending is free and is somehow not tied up with the issue of 'big government' is difficult to say, but it seems unlikely that the proponents of budgetary austerity will override the defenders of political opportunism. Fourth, businessmen usually are more hostile to government subsidy programs. Subsidies are seen to involve excessive intervention and interference with management prerogatives and are 'tainted' by association with other government expenditure programs such as unemployment insurance, welfare, or financial aid to troubled companies. Subsidy programs generally involve procedures to establish eligibility and some degree of control over the actual spending of the subsidy. Substantial discretion is left in the hands of government administrators, and there exists a concomitant threat of interference in the management of the firm's subsidized activities. If governmental assistance is to be received, the preferred policy instrument is a tax expenditure. Tax incentives leave the initiative more clearly in the hands of the private sector beneficiary, especially where the tax expenditure is itself a rather broad and open-ended incentive, and there is

186/K. Woodside the appearance of less governmental interference with management decisions.61 The word 'appearance' is important in this context because the tax expenditure, if it is effective, obviously involves interference and intervention by government in that satisfying the conditions necessary to use it may have altered the course of action a corporation would otherwise have followed. The underlying desire, therefore, is to diminish the visibility and directness of government support. Thus, the federal government Ministry of State for Science and Technology in its January 1981 statement on research and development policies, planning, and programming explicitly recognizes the preference of businessmen for government assistance through the tax system rather than as a form of subsidy.62 In order to highlight the significance of the greater discretion afforded government officials in expenditure programs, it is useful to compare briefly the differential treatment of individuals charged with income tax and social security fraud. 63 The income tax structure and its administration has been subject to the rule of law since 1948; the statute and subsequent regulations are published, and interpretations and court rulings are public.64 Social security administration and adjudication is much more subject to discretion, is less open to outside assessment, and affords less protection of the rights of individuals. Rulings by social assistance review boards or by administrators are either unpublished or, if published, are so attenuated that it may be difficult to determine just what principles of law are being followed.65 The right of search is more extensive in areas of suspected social security abuse than in cases of income tax evasion, and the likelihood of being convicted and incarcerated is much higher. The fact that the income tax system is the main vehicle of assistance for the well-to-do while expenditure subsidies are used to implement social security policies affecting less fortunate individuals is no doubt a critical factor in the different rights that exist in each area of law. A fifth area of comparison relates to the government's wish to be more or less closely associated with some course of action. The greater visibility of government involvement entailed by subsidies as opposed to tax incentives may influence the choice of policy instrument. If a government has a public commitment to certain policies, it may not want excessive attention focused on deviations from that course. Similarly, the character of a government party's voting constituency may affect the degree of openness with which it will want to be seen courting the support of certain groups: depending on who is in receipt of the benefits, there may be ideological or practical reasons to choose one instrument over another. As we have argued, for instance, businessmen see tax incentives as more of a free-market solution than subsidies, while a new tax exemption will not be of much help to poor people who lack the requisite income to take full advantage of it.

Tax Expenditures and Subsidies/ 187 Administrative and Efficiency Considerations

Tax incentives take effective immediately upon their introduction into Parliament and thus, in theory, permit a government to respond quickly to a problem. Subsidies only come into effect after they have been passed into law. The speed of delivery of the measures, however, may not reflect these procedures. In the 1970s, for instance, a pattern developed whereby budgetary provisions might take up to a year to be finally passed into law, with the result that while the measure was itself effective immediately, the exact wording of the statute remained unclear. In consequence, the value of the tax expenditure was undercut because of uncertainty about the final wording. While subsidies are subject to a slower process of introduction, the discretion they normally give to administrators allows for a more flexible, less legally-bound approach. Further, while tax incentives can in theory be introduced quickly, evidence suggests that much of their impact is subject to considerable time lags and that as a result the cure may have its greatest impact after the problem has become less serious. Moreover, the tax system is structured to provide benefits on an annual basis, whereas subsidy payments can be scheduled with closer attention to the real needs of the recipient. The net result is that subsidies can be a more flexible and responsive policy instrument than tax incentives. A second difference between the two measures involves their relative efficiency. There is a growing literature that suggests that tax incentives are inefficient and overly costly to government.66 Subsidies are easier to target than tax expenditures, especially when the object is a narrow sector or group, and they are subject to much closer supervision.67 (While this closer supervision can make subsidies more cost-effective than tax expenditures, it is also a major source of irritation for many businessmen.) The relative inefficiency of tax expenditures is not just a result of lack of supervision and the excessively wide definition of many tax incentives. It is also a consequence of the increased complexity of the Income Tax Act, where incentives are piled on incentives and many tax expenditures may operate (or fail to operate) in unexpected ways because of unanticipated interaction with existing provisions. In addition, the dense language of the act makes it difficult to be aware of available opportunities. This is less of a problem for larger firms than it is for small private corporations, who may be unable to afford the legal or accounting expertise necessary to interpret the act. Clearly there are grounds to question the efficiency of tax expenditures and to advocate the use of subsidies as a more cost-effective alternative. This thinking is reflected in the new emphasis on investment grants in the federal Petroleum Incentives Program.68 In one important area, however, tax expenditures may be preferable to subsidies. Recently, countervailing duties have been levied against the exports of

188/K. Woodside several Canadian firms such as Michelin and Honeywell by the United States government as a result of subsidies paid to the companies. It may well be that the high visibility of subsidies has disadvantages in international trade in that the subsidies threaten to trigger countervailing duties from Canada's trading partners and especially from the United States. In conclusion, it is argued that tax expenditures do not have the administrative advantages over subsidies that are sometimes ascribed to them. Moreover, tax expenditures tend to be relatively costly and inefficient as instruments of policy. To some extent this higher cost results from government's attempt to create indirectly an environment favourable to certain undertakings. By choosing this oblique course of action, the government may unintentionally fund other unrelated activities through the tax system. Conclusions and Reconsiderations

This analysis has suggested that substantial differences exist between tax incentives and subsidies. On the one hand, tax expenditures are less visible and less subject to constraint, offer more freedom for a government to act decisively and inconspicuously, are a means of assistance for the well-to-do and for successful large corporations, appear to be without cost to the federal government, and are seen (by the corporate sector especially) not to involve intervention in the economy. On the other hand, subsidies are much more cost-efficient, can be subject to much closer controls on cost and conditions of use, go through a more broadly based and open process of policy formulation, and are commonly the means of redistributive social policies. These differences raise several points worthy of further discussion. First, a focus on policy instruments alone ignores the substance of a policy. What we want to consider, therefore, is the way in which these two components of the policy process-the substance and the instrument-interact. Second, how did these policy instrument differences evolve? Tax expenditures and subsidies do not necessarily have to take their present form. A tax incentive need not be subject to the 'upside-down' effect. The existing differences reflect to some degree the residue of policy development over the decades. The Income Tax Act, for instance, has evolved in response to a narrower range of interests than can be said of many subsidy policies. Third, what will be the impact of possible reforms of the tax policy process that have been gaining in popularity over the last decade? Finally, what can be learned from the 1981 budget and the tax reforms proposed in that document? The first question might be posed as follows: does the use to which each type of policy instrument is put produce the differences between them? Are tax incentives less visible because they are the preferred instrument for providing

Tax Expenditures and Subsidies / 189

social benefits to the well-to-do and to successful corporations? Are subsidies more visible and open to scrutiny because they are used to aid the poor and to subsidize less successful businesses? These questions mirror one of the perennial disputes in political analysis: how important are political institutions as determinants of public policy outputs?69 While it is not my purpose to come to grips with this broader question, it is worth noting that many subsidy programs, like tax incentives, are also consistently able to avoid serious political attention. On the one hand, grants to the pulp and paper and to the shipbuilding and repair industries are routinely subject to relatively little public intervention and outsider review. On the other hand, the greater scrutiny usually given to subsidies makes policies that benefit the poor more easily attacked and the benefits of these programs less secure. Tax policies beneficial to the poor, to the very limited extent to which they exist in the Income Tax Act, appear to be better shielded from opposition. In general, assistance to the well-to-do and to successful large corporations is relatively free from criticism and secure in the availability of benefits whether it takes the form of a subsidy or a tax expenditure. The subsidy policies that usually draw the most fire from critics are those such as welfare, unemployment insurance, and DREE grants that assist businesses in the poorer regions of the country. The choice of policy instrument may have an independent impact where policies that benefit the lower income categories are implemented through tax expenditures. Therefore, while the choice of instrument is an important factor in determining the character of the political process that evolves, the substance of the policy is also important. Many of the differences between tax expenditures and expenditure subsidies may have less tc do with the instrument used to implement the policy than with the substance of the policy being implemented. On these grounds the scope for substantial change in policy content as a result of procedural changes may be smaller than some analysts have suggested. The second question relates to the origins of differences between policy instruments. Policies pursued by social democratic parties in North America and Western Europe normally have been paternalistic in their solutions to social issues, focusing on correcting specific problems. The progressive income tax was initially seen by radical groups on the left as a means of destroying the capitalist state. The moderate social reformers, whose subsequent influence has been widespread, were less interested in destroying capitalism and preferred a more direct and paternalistic instrument than was available in the tax system. Moreover, at the time that these new social policies were gathering support, the income tax still was narrowly focused on those earning high incomes. Pressure for increased redistribution through taxation would have antagonized the upper classes and threatened the successful implementation of new expenditure programs. As a result, new redistributive policies introduced by left-wing parties

190/K. Woodside were subsidy policies. Since 1940, however, the redistributive character of the tax system has declined significantly. This occurred because the rapid spread of the income tax net during the Second World War to encompass most of the working population (who are now taxed at relatively high rates) was continued into the post-war period, and because of the sharply increased use of corporate and personal income tax expenditures over the same years. A lack of attention to tax policy developments by social democrats made it easier for the tax system to develop subject to a much narrower range of political interests, largely those of the affluent. This conjunction of historical and social factors served to sharpen and even increase the intrinsic differences between tax incentives and subsidies. Only over the last decade or so has there been any appreciable growth in awareness of what has happened in the area of tax policy. In recent years both social reformers and government officials have begun to take a fresh look at the wealth of tax expenditures that have taken root within the Income Tax Act. In the face of continuing and growing pressures to open up the tax policymaking process, it is interesting to speculate on the implications of the many recent proposals for reform. It appears that for many observers the closed and secretive character of federal tax policy-making undermines the legitimacy of the resulting decisions. Calls for reform have usually been heard in periods of dissatisfaction with federal economic leadership and in the face of controversial tax measures being proposed by the Department of Finance. The overall thrust of these reform proposals is to open the tax policy process to greater and more intense scrutiny at a stage when the government can back away safely from unpopular initiatives. Some possible reforms include relaxing the restraint of budget secrecy to allow more consultation with tax professionals outside government prior to delivery of the budget, special or permanent advisory committees to review possible budget alternatives, and changing procedures in the House of Commons to allow a committee either to consider budget proposals in the context of their broad policy implications or to examine them on a clause-byclause basis.71 The various proposals suggest different implications for the tax policy process. In general, if the impact of the reforms is to increase consultations with tax professionals, the result would reinforce the influence of the existing tax community. This would make subsequent tax increases on capital and high incomes more difficult to implement; it would increase the obstacles to eliminating or reducing the value of existing tax expenditures, and might increase pressure to expand them. However, reforms that bring into the tax policy-making process viewpoints that are usually excluded might produce policy outputs more representative of the full range of interests within Canada. Several other implications deserve notice. A more open policy-making process would create additional

Tax Expenditures and Subsidies/ 191 delays for the implementation of tax measures, but it seems likely that improvements in policy would offset any costs of delay. Further, the federal government might find it more difficult to resist pressures from the provincial governments to be more responsive to provincial concerns. A brief look at the Liberal budget of November 1981 and its aftermath is useful as a means of putting taxation politics into some perspective. In the 1981 budget, Finance Minister Allan MacEachen proposed the closing off and reduction of a number of tax expenditures in conjunction with a reduction in personal income tax rates. Among the proposed changes were a reduction in the first-year capital cost allowance for depreciable assets, taxation of a number of employee benefits that had been exempt, limitations on the deduction of interest costs, elimination of tax deferrals through income-averaging and capital gains reserves, and an end to the capital gains exemption for more than one principal residence per family. These changes were accompanied by reductions in the personal income tax rates that largely benefited those with high incomes. Almost immediately there were cries of outrage as the budget came under attack. It was criticized for its failure to include new tax expenditures (for example, a provision for the deductibility of mortgage interest). The removal of existing tax expenditures was described as an attack on investment incentives. The competence of officials in the Finance Department was impugned. Within months Finance had made significant concessions in withdrawing from many of its initial positions through the provision of transitional relief and 'grandfather' clauses. Further retractions or adjustments were expected in subsequent budgets. The struggle was particularly revealing of the character of tax politics. The distribution of benefits from tax expenditures favours those with high incomes, especially non-salaried incomes. The dominant actors in tax politics are the groups and individuals who receive these benefits. Finance recognized this factor in concentrating its rate reductions in the high-income categories. The implication of this struggle, as well as the earlier attempt at tax reform in the period from 1969 to 1971, is that if the government is serious about reducing tax expenditures it must either carry on with its reforms despite the attacks of opponents or it must mobilize new support to offset the opposition of aggrieved interests. Equally important, the government must show more care in authorizing future tax expenditures; once in place, they will be difficult to reduce or eliminate. NOTES 1 R. W. Phidd and G. Bruce Doern, The Politics and Management of Canadian Economic Policy (Toronto: Macmillan 1978), ch. 14. Roger Smith has argued that

192/K. Woodside governments are increasingly aware of the opportunity to choose one or the other as policy instruments. See Roger S. Smith, Tax Expenditures: An Examination of Tax Incentives and Tax Preferences in the Canadian Federal Income Tax System, Canadian Tax Papers no. 61 (Toronto: Canadian Tax Foundation 1979), 8. 2 Stanley S. Surrey, 'Tax Expenditure Analysis: The Concept and its Uses,' Canadian Taxation 1, no. 2 (Summer 1979), 3-14, and Stanley S. Surrey, Pathways to Tax Reform (Cambridge: Harvard University Press 1973). Also see Smith, Tax Expenditures. 3 The basic difficulty with tax expenditure analysis lies in trying to distinguish between those tax provisions that are subsidies or incentives intended to encourage some particular result and those that are integral to the revenue-raising function. The problem of revenue loss estimation for tax expenditures is one shared with all government programs. On this question, see the symposium on tax expenditures in Canadian Taxation 1, no. 2 (Summer 1979). Also see Smith, Tax Expenditures, 1-6, and Stanley S. Surrey and Paul R. McDaniel, The Tax Expenditure Concept and the Legislative Process,' in The Economics of Taxation ed. Henry J. Aaron and Michael J. Boskin (Washington: The Brookings Institution 1980), 123-44. Critics of the tax expenditure concept and its application argue that the tax expenditure was initially intended as an instrument of budget reform and that it has been perverted into a way of increasing tax progressiveness by emphasizing the costs of tax preferences of all kinds. On this argument see Seymour Fiekowsky, 'The Relationship of Tax Expenditures to the Distribution of the "Fiscal Burden",' Canadian Taxation 2, no. 4, (Winter 1980), 211-19. 4 Ibid.,215. 5 The different political relationships of government to the various groups and sectors of a society are at base, the critical feature distinguishing the liberaldemocratic, corporatist, and socialist state systems. 6 Smith, Tax Expenditures, 1. 7 Edward Tamargo, 'Comparing Direct Spending and Tax Spending,' Canadian Taxation 1, no. 4 (Winter 1979), 42-5. 8 Government of Canada Tax Expenditure Account (Ottawa: Department of Finance 1979), 33-46. 9 Ibid., 27. 10 Smith, Tax Expenditures, 16. 11 David B. Perry, 'Corporate Tax Expenditures,' Canadian Tax Journal 24, no. 5 (Sept.-Oct. 1976), 528-33. 12 Smith, Tax Expenditures, 18-19 (see tables 2a and 2b). 13 The Hidden Welfare System (Ottawa: National Council of Welfare 1976). 14 The Hidden Welfare System Revisited (Ottawa: National Council of Welfare 1979), 1-2.

Tax Expenditures and Subsidies/ 193 15 Allan M. Maslove, The Other Side of Public Spending: Tax Expenditures in Canada,' in The Public Evaluation of Government Spending ed. J. Bruce Doern and Allan M. Maslove (Toronto: Institute for Research on Public Policy 1979), 149-68, and Jonathan R. Kesselman, 'Non-Business Deductions and Tax Expenditures in Canada: Aggregates and Distributions,' Canadian Tax Journal'25, no. 2 (March-April 1977), 160-79. 16 Smith, Tax Expenditures, 15, 22. 17 Edward Tamargo, 'Comparing Direct Spending,' 44. 18 Tax Expenditure Account, 1979, 30, and Allan Maslove, 'The Other Side of Public Spending,' 157. 19 Edward Tamargo, 'Comparing Direct Spending,' 42-5. 20 Smith, Tax Expenditures, 22. 21 Ibid., 16. 22 Ibid., 18-19; and David B. Perry, 'Corporate Tax Expenditures,' 530-2. 23 W. Irwin Gillespie and Allan M. Maslove, 'The 1980-81 Estimates: Trends, Issues and Choices,' in Spending Tax Dollars: Federal Expenditures, 1980-81 ed. G. Bruce Doern (Ottawa: Carleton University, School of Public Administration 1980), 23-48. 24 Donald R. Huggett, 'The Budget Process and Income Tax Changes,' in Report of Proceedings of the Twenty-Ninth Tax Conference, 21-3 Nov. 1977, 20-40; and David A. Good, The Politics of Anticipation: Making Canadian Federal Tax Policy (Ottawa: Carleton University, School of Public Administration 1980). Huggett notes (22-3) that a new method of getting around the budget secrecy problem has evolved inadvertently. Increasingly in the 1970s the minister ended up submitting budgetary legislation and then resubmitting it later, thus allowing the government to respond to criticisms of the first effort. 25 Ibid., ch. 3 and 152-6. 26 Ibid.: also see Douglas G. Hartie, The Expenditure Budget Process in the Government of Canada, Canadian Tax Papers no. 60 (Toronto: Canadian Tax Foundation 1978), ch. 2. 27 The New Expenditure Management System (Ottawa: Department of Finance 1979), 10. 28 Government of Canada Tax Expenditure Account (Ottawa: Department of Finance 1980), 2-4. On the question of the decline in the authority of the Department of Finance, see Douglas G. Hartie, The Revenue Budget Process of the Government of Canada: Description, Appraisal and Proposals, Canadian Tax Paper no. 67 (Toronto: Canadian Tax Foundation 1982), 33. 29 Joan Allin, The Tax Subsidy for Pollution Abatement Equipment' Canadian Taxation 1, no. 2 (Summer 1979), 47-50. 30 Richard Simeon has presented a strong case for the need to 'disentangle' the two

194/K. Woodside levels of government. See 'The Federal-Provincial Decision-Making Process,' in Intergovernmental Relations (Toronto: Ontario Economic Council 1977), 25-38. 31 For an example see R. Brian Woodrow, Kenneth Woodside, Henry Wiseman, and John B. Black, Conflict over Communications Policy, Policy Commentary no. 1 (Montreal: C.D. Howe Institute 1980), 29-74. 32 See Thomas E. McDonnell, 'The Tax Reform Process: Comments on the Report of the Tax Legislative Process Committee,' Canadian Taxation 1, no. 4 (Winter 1979), 27-9. 33 According to the Fiscal Arrangements and Established Programs Financing Act, 1977, all provinces except Quebec levy their personal income tax as a percentage of the basic federal tax. In the case of the corporate income tax, all provinces except Ontario and Quebec (and soon Alberta) share the same corporate income tax base as that of the federal government. The federal government collects the revenue for both levels of government, and provincial corporate taxes are expressed as a percentage of the federal taxable income. See Robin W. Boadway and Harry M. Kitchen, Canadian Taxation Policy, Canadian Tax paper no. 63 (Toronto: Canadian Tax Foundation 1980), 116. A major exception to this lack of concern is in the area of resource taxation where the western provincial governments have been especially concerned that federal resource taxations interfere with and reduce their control over revenue from their resources. However, in areas of non-resource income and personal income taxation there is little evidence of continuing concern apart from the rare case such as the battle over tax reform from late 1969 through the end of 1971. 34 Government of Canada, Tax Expenditure Account, Dec. 1979, 27-28. It is noteworthy that the government has not committed itself to the publication of an annual tax expenditure account as is the case in the United States. 35 For a discussion of some of the criticisms see Report of the Royal Commission on Financial Management and Accountability (Ottawa: Ministry of Supply and Services 1979), chs. 6, 21, and 22. 36 See Huggett, 'The Budget Process,' 28; Good, The Politics of Anticipation, 134; Report of the Royal Commission on Financial Management and Accountability, 386-7, 399. 37 Good, 'The Politics of Anticipation,' 152-6; Hartle, The Expenditure Budget Process, 24-33. The finance ministers will, of course, receive many suggestions and requests from MPs and representatives of non-governmental groups but the Finance officials must play the role of good listeners and avoid giving hints about what the government might or might not do. 38 Good, The Politics of Anticipation, 121-61. As mentioned in n. 24 the trend in the 1970s was for many budget proposals to be resubmitted; in the process some

Tax Expenditures and Subsidies/ 195

39

40

41 42

43 44 45 46

47 48 49

controversial measures (such as the 1978 proposals to remove special tax provisions for self-employed professionals and athletes) were substantially revised. For instance, the two-year write-off for manufacturing and processing machinery and equipment introduced in 1972 was extended indefinitely in 1974 with relatively little comment. A general example of tax changes introduced by regulations is found in the capital cost allowance rates and the designation of qualified investments eligible for consideration under an existing tax provision. Report of the Royal Commission on Financial Management and A ccountability, 406. Ibid., chs. 6, 21, and 22; Robert J. Jackson and Michael M. Atkinson, The Canadian Legislative System, 2d rev. ed. (Toronto: Macmillan of Canada 1980), 99-102, 107-9; Paul G. Thomas, 'Parliament and the Purse Strings,' in Parliament, Policy and Representation, ed. H.D. Clarke, Colin Campbell, F.G. Quo, and Arthur Goddard (Toronto: Methuen 1980), 160-81. Report of the Royal Commission on Financial Management and Accountability, 384. Ibid., 399. Ibid., 100-3 and Hartle, The Expenditure Budget Process, 36-8. Neil Brooks, 'The Tax Expenditure Concept,' Canadian Taxation 1, no. 1 (Jan. 1979), 31-5, and Gerard M. Brannon, 'Tax Expenditures and Income Distribution: A Theoretical Analysis of the Upside-Down Subsidy Argument,' in The Economics of Taxation, ed. Henry J. Aaron and Michael J. Boskin (Washington: The Brookings Institution 1980), 87. It is true, however, that the smaller absolute gains for those in the lower income categories may in fact produce larger gains proportionately. Surrey, Pathways to Tax Reform, 134-8. A regressive measure is one that benefits high-income earners more or costs them less than it does low-income earners. The refundable child tax credit is unique because those below the tax threshold (i.e., those without taxable income) receive a credit for each child which comes in the form of a payment from the government. The value diminishes as income increases beyond a certain maximum. Thus the benefits tend to be concentrated on those with lower incomes, although Brigitte Kitchen has noted that families near the median income level benefit the most. See Brigitte Kitchen, 'The Refundable Child Tax Credit,' Canadian Taxation 1, no. 3 (Fall 1979), 44-51. An attempt to introduce a second refundable credit ended in December 1979 with the defeat of the Crosbie budget. This was the refundable energy tax credit. For discussion of its redistributive character see Andrew Doman, 'The Effect of Federal Budgetary Policies 1978-80 on the Distribution of Income in Canada,' Canadian Taxation 2, no. 2 (Summer 1980), 112-22.

196/K. Woodside 50 Richard M. Bird, Tax Incentives for Investment: The State of the Art (Toronto: Canadian Tax Foundation 1980), 52-3. Also see R.P. Simon, The Efficacy of Recent Corporate Income Tax Reductions for Manufacturing - An Unsimulated View,' Canadian Tax Journal (March-April 1974), 160-5. 51 Report of the Royal Commission on Taxation, vol. 6 (Ottawa: Queen's Printer 1966), 69. 52 Bird, Tax Incentives for Investment. In this regard Bird places particular weight on the heavy use of capital cost allowances, investment tax credits, and reduced rates of taxation. 53 The National Energy Program, 1980 (Ottawa: Department of Energy, Mines and Resources 1980), 38-41. 54 Many expenditure subsidies are not particularly redistributive because they are not sufficiently focused on the poor. Thus the unemployment insurance revisions of 1971 'provided virtually no fiscal benefits for the poorest families': W. Irwin Gillespie, In Search of Robin Hood (Montreal: C.D. Howe Research Institute 1978), 18, 25, 40-2. 55 David A. Good, The Politics of Anticipation, 11-78, 92, 108, 188. 56 For instance, see R.S. Woodward, 'The Effectiveness of DREE'S New Location Subsidies,' Canadian Public Policy (Spring 1975), 217-30; D. Usher, 'Some Questions about the Regional Development Incentives Act,' Canadian Public Policy (Autumn 1975), 557-75; and Industrial Incentives Programs in the Atlantic Region (Halifax: Atlantic Provinces Economic Council 1976). A more recent example of dissatisfaction with DREE programs can be found in David Crane, 'Ottawa wastes investment funds, economists say' Toronto Star, 25 Nov. 1980. 57 Gillespie and Maslove, 'The 1980-81 Estimates," 44; Ken Woodside, 'Tax Incentives vs. Subsidies: Political Considerations in Governmental Choice,' Canadian Public Policy (Spring 1979), 248-56. 58 Wolfe Goodman in 'The Small Business Credit: A Panel Discussion' Canadian Taxation 1, no. 2, 44-5. Attacks on 'big government' virtually never involve a reference to the number of tax expenditures or the amount of revenue forgone through the tax system. See, for example, the following instances: Toronto Star, 17 June 1976; 11 Nov. 1975, and the Globe and Mail, 20 Nov. 1975, 18 March 1976, 18 May 1976, and 14 Dec. 1976. 59 Good, The Politics of Anticipation, 82. Also see the Toronto Star, 15 Septemb 1979. On the reasons for the introduction of the tax expenditure account, see Hartle, The Expenditure Budget Process, 14-15. 60 Tax Expenditure Account, 1979, 31. Not all expenditure subsidies have limits, however. The 1980-81 plan to provide grants for exploration and development by the oil industry is essentially open-ended and therefore limited only by the degree to which the industry responds to the incentives.

Tax Expenditures and Subsidies/ 197 61 David Vogel, 'Why Businessmen Distrust Their State,' British Journal of Political Science (Jan. 1978), 45-78; and Leonard Silk and David Vogel, Ethics and Profits: The Crisis of Confidence in American Business (New York: Simon and Schuster 1976), ch. 2. 62 R & D Policies, Planning and Programming (Ottawa: Ministry of State, Science and Technology 1981), 24. 63 Reuben Hasson, 'Tax Evasion and Social Security Abuse - Some Tentative Observations,' Canadian Taxation 2, no. 2 (Summer 1980), 96-108. This excellent article is part of a useful symposium on the subject. 64 Ibid., 106-8; also W. Goodman, 'Panel Discussion: Tax and Social Security Abuse,' Canadian Taxation 2, no. 2 (Summer 1980), 109-11, and Robert W. Davis, Capital Cost Allowances, Studies of the Royal Commisson on Taxation, no. 21 (Ottawa: Queen's Printer 1967). 65 Hasson, 'Tax Evasion and Social Security Abuse.' 66 Bird, Tax Incentives for Investment, 32-45. 67 Tax Expenditure Account, 7979,31. 68 Petroleum Incentives Program: The Basic Rules-A Framework (Ottawa: Department of Energy, Mines and Resources 1980). 69 Two examples of a large literature on this question are Arthur Maass, Area and Power (Glencoe: The Free Press 1959) and J. Roland Pennock, 'Agricultural Subsidies in England and the United States,' American Political Science Review 61, no. 3 (Sept. 1962), 621-33. 70 For instance, see Paul Addison, The Road to 1945 (Quartet Books, London 1977), 117-18 and the interview with Stuart Holland in Bertram Silverman, 'The Crisis of the British Welfare State,' Challenge (Sept.-October 1980), 33-9. 71 A review of many of these proposals can be found in Hartle, The Expenditure Budget Process, 47-60. Also see the Tax Legislative Process Committee, The Tax Legislative Process,' Canadian Tax Journal 26, no. 2 (March-April 1978), 157-82, and John B. Stewart, The Canadian House of Commons: Procedure and Reform (Montreal: McGill-Queen's University Press 1977), 101-8, 277-81. Also see The Budget Process (Ottawa: Department of Finance 1982).

This page intentionally left blank

9 / Crown Corporations in Canada: The Choice of Instrument J. ROBERT S. PRICHARD MICHAEL J. TREBILCOCK

Introduction In this essay we will analyse public ownership as an instrument of intervention and examine the characteristics of public ownership that are likely to be influential in determining its substitutability for alternative policy instruments. In particular we will focus on the question of why a government might resort to public ownership instead of some other instrument-taxation, expenditure policy, regulation - to accomplish a particular interventionist objective.1 The puzzle of public enterprise is the patchwork pattern of its manifestations and substitutes across Canada. For example, in the last two decades most electric utilities have been taken over by provincial governments, but in Newfoundland, Prince Edward Island, and Alberta the electric utilities are still either wholly or partly privately owned. In provinces where the electric utilities are government-owned, natural gas distribution systems are typically privately owned and publicly regulated. In Newfoundland, Nova Scotia, Prince Edward Island, New Brunswick, Quebec, Ontario, and British Columbia, the telephone systems are privately owned and subject to government regulation. In Manitoba, Alberta, and Saskatchewan, the telephone systems are publicly owned, although in Manitoba and Alberta they are regulated by public utility boards and in Saskatchewan by the cabinet. A major national air carrier, Air Canada, a publicly owned enterprise, is in competition with another national carrier, CP Air, a privately owned enterprise, and both are subject to extensive regulation by the Canadian Transport Commission; in regional markets, both airlines compete with various regional and local carriers, some of which, such as Pacific Western Airlines, are owned by provincial governments. Canadian National Railways, which is publicly owned, coexists with the privately owned CP Rail.

2007 J.R.S. Prichard and M.J. Trebilcock

In the oil industry, a large private sector coexists with a publicly owned oil company (Petro-Canada). In the broadcasting field, the private networks and the Canadian Broadcasting Corporation, a publicly owned enterprise, are all subject to various forms of regulation by the Canadian Radio-Television and Telecommunications Commission. In some provinces, especially a number of the western provinces, automobile insurance is provided by publicly owned corporations, while in other provinces it is provided by privately owned enterprises subject to government regulation. The Post Office, run until recently as an executive department of government, competes with various privately owned parcel and courier services; its transformation into a crown corporation occurs at a time when the United States government is considering converting its postal service from a corporation to an executive department of government. In capital markets, crown corporations (such as the Canadian and Ontario Development Corporations) provide capital to private-sector firms, as do private-sector financial institutions; in similar contexts, lending functions are performed directly by departments of government (such as the Department of Regional Economic Expansion). In the face of these apparent contradictions, it is tempting to rely largely on random factors and ideology as explanations for the divergent patterns in the selection of public enterprise as an instrument of intervention. However, in the Canadian context, ideology alone is not an adequate explanation of the emergence of public enterprise, at least if party policies are treated as being in some respects ideological. As Vining and Botterell have shown,2 in the case of the provincial crown corporations all political parties at the provincial level have used public ownership as an instrument of intervention. While some recent work suggests that there may be some systematic differences between parties of the left and right with regard to the frequency of resort to public ownership in some sectors,3 it remains clear that ideology does not offer anything approaching a complete explanation. At the federal level, Canada has never had a social democratic government; nevertheless a large number of government-related corporations have been created over the years by both Liberal and Conservative governments. In the two provinces with arguably the strongest ideological commitment to free enterprise, we observe, first, that the Alberta government has purchased Canada's third-largest airline (Pacific Western) and, second, that the Ontario government owns Canada's largest crown corporation (Ontario Hydro). To resort to historical accidents as explanations for the emergence of public enterprise is to acknowledge that the factors that bear on instrument choice are incapable of specification, an acknowledgment we do not accept. The traditional reasons offered for the establishment and operation of public enterprises reflect a rich mixture of political, economic, cultural, historical, and

Crown Corporations 7201 ideological factors.4 The striking feature common to the rationales for the choice of public ownership as the instrument of intervention is their lack of explanatory power. It is true that each rationale sometimes leads to public ownership, but virtually none of them leads exclusively to public ownership as the instrument of response. For example, while it is widely accepted that public ownership is one way to regulate natural monopoly, it is also accepted that it is not the only way. It is possible (and indeed fairly common) to have a private firm regulated by a public regulatory board charged with overseeing the firm's behaviour so as to overcome the potential economic inefficiences in situations of natural monopoly. In Canada and elsewhere, both publicly owned firms and regulated private firms are engaged in similar activities in various areas. As a result, it is not enough to use a field of activity requiring government intervention, such as natural monopoly, as a satisfactory explanation of the existence of public ownership; such an approach fails to capture the diversity of alternative modes of government intervention. Rather, since the intervention can normally be achieved in more than one way, one must go further to identify those characteristics of public ownership that will be influential in determining the choice of instrument in a particular context. The approach taken in this essay focuses on the characteristics of public ownership as an instrument of intervention. We do not suggest that this approach will provide an explanation in every case, but it does offer substantial insights into why governments choose public ownership as the instrument of intervention in those circumstances where more than one instrument is available, and into the more general phenomenon of the substitutability of instruments.5 Our approach may appear highly abstract. However, if the reader applies our conceptual framework to any factual situation involving a decision to utilize public ownership, we are confident that the abstract considerations will take on a vitality not immediately apparent in the formal presentation. Indeed, some readers may find that attempting to apply our framework to a particular case study is a useful way to test both their understanding of the framework and its value in studying public ownership as a policy instrument. Public Enterprise: Definition and Characteristicssstics

The precise meaning of the term 'public enterprise' resists definition. At its broadest, public enterprise might encompass all functions of government, including those carried on by governmental departments, quasi-independent agencies, and crown corporations, and those accomplished by public support of and involvement in the private sectors. Similarly, the term 'crown corporation' resists exact definition in legal or functional terms. Legal considerations such as the form

2027 J.R.S. Prichard and M.J. Trebilcock

of organization, degree of ownership, and powers of appointment of directors are all relevant to the definitional problem.6 Here we will confine the use of 'public enterprise' to those situations in which the government is engaged in the provision of goods or services to the public on a commercial or quasi-commercial basis. Similarly, the term 'crown corporation' will refer to corporations in which the government has a de facto controlling interest and that provide goods or services to the public on a commercial or quasi-commercial basis.7 We will attempt to identify the major legal and institutional characteristics of public enterprise and crown corporations that are influential in determining instrument choice. Our reference point will be the federal crown corporations. Legal characteristics The essential legal characteristics of crown corporations may well be influential in government's decisions to choose public ownership instead of other available instruments. Below we summarize those characteristics as they relate to the creation, taxation, regulation of labour relations, accountability, and financing of crown corporations. At the federal level, all crown corporations are created by one of three methods; a special constituent act of parliament, letters patent (typically pursuant to the Canada Companies Act8), or articles of incorporation under the Canada Business Corporations Act.9 Under the current Canada Business Corporations Act, a minister can apparently incorporate any company without reference to the governor-in-council or Parliament. Similarly, an existing crown corporation may create a subsidiary under the Canada Business Corporations Act without prior government or parliamentary approval.10 With respect to taxation, section 149(l)(d) of the Federal Income Tax Act11 expressly exempts from tax 'a corporation, commission or association not less than 90 per cent of the shares of capital of which was owned by Her Majesty in right of Canada... or a wholly owned subsidiary to such a corporation, commission or association...' Department practice apparently extends this exemption to wholly owned sub-subsidiaries of crown corporations. Section 27 of the Income Tax Act negates section 149(l)(d) immunity for those corporations listed in schedule D to the federal Financial Adminisration Act, and expressly declares them to be subject to income tax. Moreover, the section provides that such corporations shall be deemed not to be private corporations and thus not to be entitled to any of the special benefits accruing to private corporations (such as the small business tax rate) provided in the act. It should be noted, however, that according to departmental practice, a wholly owned subsidiary of a schedule D corporation will be entitled to the section 149( l)(d) immunity, as it is not itself listed in

Crown Corporations / 203

schedule D. It should also be noted that these provisions apply to both federally imposed income tax and provincial corporate tax for the seven 'agreeing provinces' (i.e., not Ontario, Quebec, or Alberta) who collect their taxes through Ottawa. At the federal level, employer-employee relations in crown agencies are subject to the provisions of either the Public Service Employment Act12 or the Canada Labour Code.13 Broadly speaking, the Public Service Employment Act applies to departmental employees and employees of designated government agencies. Crown corporations, in the sense in which the term is used here with very few exceptions fall under the Canada Labour Code. There are several important differences in the two regimes. First, some matters that may be subject to collective bargaining under the Canada Labour Code are not subject to collective bargaining under the Public Service Employment Act or the Public Service Staff Relations Act;14 these include the setting of classification standards and of certain terms and conditions of employment. Second, under the publicsector legislation, certain designated employees are prohibited from striking; this includes those employees whose duties are duties the performance of which is or will be necessary in the interest of the safety or security of the public. At the federal level, only those crown corporations named in the schedule to the Financial Administration Act are subject to a generalized scheme of financial accountability to the government and to Parliament. For other crown corporations, financial accountability either is defined on an ad hoc statute-bystatute basis in the case of corporations set up under constituent acts, or follows from the shareholder - corporation relationship defined in the Canada Business Corporations Act in the case of crown corporations incorporated under or governed by that Act. Beyond the realm of financial accountability, in the case of most crown corporations the federal government has the power to approve by-laws and to appoint and remove directors, board chairmen, and chief executive officers, either pursuant to powers in the constituent act of a corporation or under applicable provisions of the Canada Business Corporations Act. Sometimes, constituent acts confer on the government the power to issue, through a minister, policy directives to a crown corporation. In the case of companies incorporated under the Canada Business Corporations Act or predecessor acts, the designated minister, as sole shareholder, would appear to have an equivalent power to issue policy directives in the form of unanimous shareholder agreements. With respect to financing, at the federal level crown corporations seeking access to government funds face the budget approval requirements pursuant to the Financial Administration Act. With respect to grants or loans from the government, no payments may be made out of the consolidated revenue fund

2047 J.R.S. Prichard and M.J. Trebilcock

without the authority of Parliament.15 This authority may be contained in a variety of sources, such as a special or general statute or an appropriation in votes on estimates. With respect to guarantees of private-sector borrowings by crown corporations, a government guarantee can only commit the consolidated revenue fund if it has been authorized to do so by Parliament.16 By virtue of his role as manager of the consolidated revenue fund,17 all guarantees both in law and in practice are given only by the minister of finance. Institutional characteristics In Hodgett's striking phrase, crown corporations are 'structural heretics.'18 In some ways, they resemble private-sector enterprises maximizing profits subject to the constraints of applicable regulation or direction. In other ways, they resemble bureaucracies executing public policies designed to promote nonmarket objectives. In examining why politicians choose crown corporations as a policy instrument in particular settings, it is important to attempt to identify the institutional characteristics that distinguish crown corporations from privatesector enterprises that are subject to government direction and influence through direct regulation, tax policy, subsidy policy, or procurement policy, and from bureaucracies where government takes over and directly performs given economic functions. This task poses two major problems: differentiating the characteristics of public ownership from private-sector regulation, and differentiating departmental bureaucracies from crown corporations. Viewing instrument choice from the perspective of the calculus that faces political decision-makers, a number of characteristics would seem influential in determining the location of these two boundaries. Public ownership versus private-sector regulation A number of institutional factors might suggest a policy preference for public ownership over private-sector regulation (compendiously defined). Many of them derive from the notion of monitoring and information costs first developed in the economic literature on the theory of the firm. We will begin with a discussion of this concept. The economic literature on the theory of the firm19 has sought to answer the question of why firms internalize the process of co-ordinating inputs rather than rely on the price system through independent contracting between entrepreneurs and input owners. Coase suggests that the main reason it is profitable to establish a firm is that there is a cost to using the price mechanism; the most obvious cost of organizing production through the price mechanism is that of discovering what the relevant prices are. He also argues that the cost of negotiating and concluding a separate contract for each exchange transaction that takes

Crown Corporations / 205 place in the market must be taken into account. When a firm internalizes factor co-ordination, the character of the contract into which the owner of a factor enters with the firm is such that for a certain remuneration the factor owner agrees to obey the directions of the owner of the firm (or his agents) within certain limits. Within those limits the owner directs the other factors of production. When the costs associated with this process of direction are less than the costs associated with ascertaining relevant factor prices in the market and the transaction costs associated with negotiating independent contracts with these factors, it will benefit a firm to internalize the co-ordination of its factors. McManus emphasizes the enforcement costs associated with using the price mechanism as a constraint on behaviour: resources must be expended in measuring the activity for which one is paying.20 In many contexts, pecuniary constraints on behaviour are not perfectly enforced, because some changes in the activity of an individual will not be detected to the mutual satisfaction of buyer and seller. Where the buyer cannot perfectly specify, or enforce, desired outputs from independent contracting, pecuniary incentives exist for the seller to shirk or cut corners to the point where it pays the buyer to specify the contractual constraints more clearly and enforce them more strictly, or to choose a different and less costly form of economic organization for co-ordinating these factors. In choosing a firm that either owns or employs many of the relevant factors, an entrepreneur may be able to reduce his monitoring costs by acquiring (at a price) the right to engage in continuous direction of the allocation of productive activities within the firm. Alchian and Demsetz suggest that in the classical private-sector firm, the entrepreneur becomes a specialized monitor in directing the allocation of resources, and has strong incentives to perform his role efficiently by virtue of his position as residual claimant to the income of the firm after payment of the factors.21 That is, since the entrepreneurial owner of the firm is entitled to the profit remaining after all factors of production have been paid, the owner has a strong and direct interest in the efficient management of the firm. A government contemplating regulating or otherwise influencing the behaviour of private-sector enterprises faces many of the same kinds of costs faced by an entrepreneur engaging in independent contracting for factors of production. For example, to the extent that a regulatory fiat imposes costs on a privatesector firm, it obviously has incentives to undertake less than complete compliance with the fiat. Similarly, where the government is contemplating provision of a subsidy to a private-sector firm, it faces the costs associated with obtaining and validating information from the firm about its real subsidy requirements, and specifying and enforcing the conditions governing use of the subsidy (and probably also the conditions determining qualifications for further subsidies).

2067 J.R.S. Prichard and M.J. Trebilcock

Suppose a government should decide that, in the interest of maintaining employment in a region of high unemployment, it is necessary to save a 'failing' private-sector enterprise. One option open to the government is to provide continuing subsidies. In order to determine appropriate subsidies, the government needs a great deal of information about the firm's costs, about conditions prevailing in the firm's output markets, about the potential for the firm acquiring over the long run more efficient technology, and about the likely effects of continuing subsidies on the incentives of the owners to improve the performance of the firm. In addition, if the purpose of the subsidies is to maintain jobs in the firm, presumably some conditions would need to be specified as to the type and level of employment that the firm would have to maintain in order to qualify for the subsidies. These conditions may be difficult to specify and difficult to enforce. Because of the firm's superior access to much of this information, the government faces the risk of strategic behaviour on the part of the firm in exaggerating the size of the subsidy required and of exploiting the threat to terminate activities in the region, and thus precipitating the politically costly elimination of jobs and other spillover activities in the area. These costs, all of which can be subsumed under the rubric of monitoring costs, are likely to be substantial; they create the same incentives for government to internalize factor co-ordination as those facing entrepreneurs in the private sector in choosing between factor co-ordination through the price system or through the firm. This concept of monitoring costs lies at the core of the explanation of the characteristics of public ownership that might induce a government to favour this instrument over private-sector regulation.22 The concept is reflected in a number of other characteristics identified below, which similarly affect the choice of instrument. These are policy co-ordination, industry structure, legal and functional limitations on substitute instruments, low-visibility taxation, symbolism and ideology, and national security and international relations. In situations involving multiple public programs or intervention objectives, the crown corporation may be a relatively more effective instrument than a regulated private firm for enhancing policy co-ordination. It may be the case that in situations where the degree of public support for the private enterprise is substantial and takes the form of a variety of different programs, the cost of co-ordinating these activities can be reduced by internalizing them to the public sector through public ownership. For example, Tupper suggests that the creation of the Cape Breton Development Corporation can be traced in part to the complexities caused by the array of support programs that were previously directed at the Cape Breton area in an attempt to moderate recurring economic dislocation.23 Similarly, where multiple public objectives are sought through a particular enterprise, it may be that the cost of co-ordinating and reconciling

Crown Corporations / 207

these different objectives can be reduced by internalizing them through the management and accountability mechanisms of crown corporations. This effect will be enhanced in situations where competing public policies are being advanced; a trade-off among the policies which may be difficult to achieve in explicit terms in an external forum can be internalized through public ownership. For example, in the case of Telesat, it was necessary to trade off the competing considerations of Canadian content and commercial and technical viability, a task that would have been extremely difficult if done externally.24 The structure and nature of the relevant industry will affect the relative desirability of public or private ownership. In particular, the absence or presence of a competitive market structure in the relevant part of the private sector will be influential, because it will affect the monitoring costs identified above. Where numerous private firms are available to perform a given function, and where there is competition among them to undertake this function, the industry structure itself will generate superior information for the government, thus reducing the costs of specifying and monitoring desired outputs. That is, the competition among the private firms will serve as a form of monitoring and information production, making reliance on the private sector relatively more attractive. However, in cases where only a single firm or a small number of firms are available, or where the existing firms are able through anti-competitive practices to behave as a single firm, the monitoring costs are increased by virtue of this market structure, and public ownership will become relatively more attractive. Another context in which market structure may favour public ownership is where the government is the primary purchaser of the product of the industry. In these cases, which may display certain attributes of bilateral monopoly, the strategic and gaming costs of contracting with private firms may be high and the attractions of vertical integration for the government may be increased. In the result, the government as public purchaser may wish to integrate by becoming a public producer in the form of a crown corporation to achieve the economies available through vertical integration. Certain of the wartime crown corporations might be explained in this way.25 Certain legal factors may also limit the effectiveness of the various substitute regulatory instruments as techniques for aligning private-sector activity with specified public objectives, thus favouring the choice of public ownership as the instrument of intervention. For example, in sectors of economic activity where the constitution allocates regulatory authority to the federal government, a provincial government is deprived of the ability to use direct regulation as a technique of intervention and may therefore choose public ownership as the only instrument available to it for participating in public decisions in those

2087 J.R.S. Prichard and M.J. Trebilcock

sectors. A case in point is the regulation of aeronautics, which is a federal responsibility under the division of powers in the constitution. Here, a provincial government's only opportunity to participate may be through public ownership of an airline. Indeed, Alberta's purchase of Pacific Western Airlines is consistent with this proposition. Similarly, regulatory constraints on institutional behaviour, which may in general be well justified in terms of the intended policy objectives, may create special situations where public ownership or at least public participation in business activity becomes advantageous.26 For example, limitations on the voting stock of companies that can be held by banks, life insurance companies, and trust and mortgage-loan companies, together with rate ceiling laws applicable in other cases, may create imperfections in capital markets that inhibit the supply of risk capital to certain kinds of ventures that the government wishes to see encouraged (such as small businesses). This may create a rationale for the government to provide capital itself, either in the form of loans or equity. The creation of the Federal Business Development Bank and the Ontario Development Corporation can be explained in part in these terms. In a functional sense, the very nature of direct regulation limits its effectiveness as a mechanism for aligning private-sector activity with desired public objectives. To the extent that these limitations are substantial in a given context, the crown corporation becomes relatively more attractive to decision-makers. Direct regulation relies primarily on explicit legal orders. These orders may appear in the form of decisions of regulatory tribunals, rules and regulations under statutes, terms and conditions under contracts, undertakings given in memoranda of understanding or agreement, and conditions and qualifications attached to the receipt of public assistance. However, regardless of the particular form in which they appear, legal orders require definition and specification of a private firm's future conduct. As a result, in situations where setting such definitions or specifications is difficult or impossible, direct regulation becomes less effective and public ownership relatively more effective. While crown corporations also require direction, that direction can be constantly evolving, communicated less formally and less openly, and stated with less precision; it need not anticipate accurately the financial consequences of the required conduct. The primary situation in which the limitations of legal orders as a regulatory device will become apparent is one where the regulatory objectives are evolving or uncertain. This will arise most commonly in situations of relative novelty, either because of new technologies or because of new 'environmental' considerations. In these cases, the costs of a proposed undertaking are likely to be highly uncertain, and the particular objectives the firm will be required to attain are

Crown Corporations / 209

likely to be extremely difficult to state. This uncertainty will be magnified in cases where the objectives can only be stated in a general way. For example, an objective for a particular project such as Telesat may be to have the greatest possible Canadian participation in the supply of materials, parts, and services. The objective 'as great as possible' cannot be reduced to a percentage or dollar amount in advance because of uncertainty about the availability of suppliers and the terms on which they will be able to supply. To a private firm faced with the necessity of forecasting the financial consequences in order to assure a profit, this uncertainty may result in unacceptable degrees of risk in its financial projections. A public firm assured of financial support as the actual degree of Canadian participation is clarified over time may not share the same concern over this uncertainty. These considerations may be particularly important in the regulation of enterprises such as railways, airlines, and electric utilities, which serve a central economic function such as providing an infrastructure service that the government wishes to use as a vehicle for supplying incentives or disincentives for other economic activities. In these cases, the policies of the firm must vary in response to the continually changing circumstances of the secondary industries, and policy judgments may be required on a virtually continuous basis. In a situation requiring frequent marginal adjustments in policy, it may be extremely difficult to issue a stream of legal orders to dictate the behaviour of the private firm, to calculate the financial implications of each order, and to compensate the firm appropriately. In such a case, a publicly owned firm, not constrained in the same way as a private firm by the financial implications of changes in policy, may be relatively more attractive. More generally, the ability of a government to reverse policy decisions effectuated through a crown corporation in a low-visibility, informal, incremental way minimizes the political costs associated with more public and deliberate acknowledgments of governmental error. A rather different characteristic of public ownership is that by combining in a crown corporation a set of profitable activities or objectives that are not financially self-sustaining, politicians may be able to realize political advantages through the imposition of a form of tax (cross-subsidization) that has low visibility for the bearers of it (it never appears on the government's books) while at the same time being raised relatively efficiently through 'business-like' management of the tax-bearing resources. This strategy will generally require the conferring of a government-sanctioned monopoly of the profitable activities on the crown corporation to prevent entry and the competitive erosion of the capacity to cross-subsidize. Even where a government provides direct subsidies to Crown corporations, whether in the form of capital grants, forgivable loans, loans at below-market

2107 J.R.S. Prichard and M.J. Trebilcock

interest rates, or guarantees, these may be perceived by the taxpayers as being within the normal investment functions of a shareholder and designed to produce a long-term return, rather than to provide a subsidy to employees or customers of the corporations (as may in fact be the case). As Borcherding has argued, public ownership compared to private-sector regulation as an instrument of redistribution is less open, more flexible, and more selective.27 These are important political properties, and to the extent they are systematically more available through crown corporations, they should play a substantial role in determining instrument choice. Public ownership is also a symbol of philosophical and ideological preferences. Its symbolic connotations are material in a number of contexts. For example, in situations where the government is providing substantial public support to a private firm, the politician may perceive a risk that, if the firm were at some future date to make a profit, the government would be accused of having used public funds to generate private profit. Therefore, in situations where the public support of the industry or firm is so substantial that the residual private contribution to the firm's potential success is minimal, the government may wish to own the enterprise in the form of a crown corporation so as to capture the benefits of the firm's success. To put it another way, in situations where the government assumes virtually the entire downside risk of an enterprise, it may also wish to capture upside risks so as to avoid potentially damaging political criticisms of 'corporate welfarism.' Simply undertaking to tax upside profits in the future may not be perceived by voters as an even trade. Similarly, issues of foreign ownership may have high symbolic content, and in those cases where there are no existing or potential Canadian enterprises to provide certain services or goods, the government may have no realistic alternative other than to create a crown corporation. The creation of Trans-Canada Airlines (the predecessor of Air Canada) may be partially explained in these terms.28 Public ownership may also be attractive as a way of symbolizing and dramatizing a government's commitment to a particular cause or set of values. In some situations, even if it is possible for the government to regulate private-sector activity so as to achieve its objectives, it may be too difficult to generate public confidence in and understanding of this reality, and an assertion of public ownership may be the only way to communicate clearly the government's commitment to a particular public objective. A number of other characteristics of public ownership may sway the decisionmaker's choice of instrument in particular circumstances. First, the opportunity for relatively secret decision-making may make public ownership particularly suitable in situations involving national security and substantial amounts of

Crown Corporations 7211

confidential information.29 Second, in situations where Canada's international relations will be affected by a firm's conduct and where Canada's international posture is intimately related to a range of other international objectives, public ownership may offer the only reasonable vehicle for operating in the international sector in order to internalize the decision-making and to permit Canada to speak with a single voice.30 While the discussion to this point has emphasized the cost of monitoring private-sector activity as a major factor favouring the utilization of public ownership, the concept of monitoring costs is also the source of major factors that militate against the choice of a public-ownership policy instrument in many circumstances. A sensitivity to these competing considerations is essential. First, as the literature on the theory of the firm points out, while a firm may be able to economize on some of the monitoring costs faced by an entrepreneur in purchasing factors through independent contracting, new and different kinds of monitoring costs are created within the firm by virtue of the attenuation of the relationship between a factor's income and his productive activities. There are always some opportunities in any organization to direct one's activities to non-pecuniary forms of consumption. Thus, there will be an increase in monitoring costs associated with specifying and enforcing non-pecuniary constraints.31 Relative monitoring costs within and outside a firm will largely determine the method of economic organization. At the public ownership-private-sector regulation boundary, this implies that while the government will face monitoring costs in attempting to regulate or otherwise influence private-sector activity, offsetting monitoring costs must be confronted in the event that these activities are internalized. A related factor militating against the choice of public ownership as a policy instrument derives from the concept of the residual interest maintained by the controlling owners in a private-sector firm. It will be recalled that Alchian and Demestz in part explain the emergence of firms on the basis that the firm structure creates incentives for residual claimants to firm income to develop specialized monitoring skills. In comparing a publicly owned enterprise to a privately owned enterprise, clearly these incentives are attenuated. When the firm owner is the government, there is no cohesive set of individuals who stand to be at a financial advantage because of more efficient rather than less efficient monitoring. The minister responsible for the oversight of publicly owned firms does not possess the financial incentives of owners of private-sector firms to ensure efficient monitoring. The incentive structures facing political overseers of publicly-owned activities should not be over-simplified; while there may be no economic returns for

2127 J.R.S. Prichard and M.J. Trebilcock

efficient management, there are presumably political returns. As Becker32 and Breton and Wintrobe33 have pointed out, politicians derive no returns from waste per se and indeed run political risks from acquiescing in it. However, it can probably be accepted that the ultimate shareholders in publicly owned firms - the voters - have sufficiently small stakes in whether such corporations are well or badly managed and face sufficiently high information costs in ascertaining which is the case that the penalties attaching to weak monitoring in the public sector are significantly less exacting than those that obtain in the private sector. This proposition will not always hold. The strength of the monitoring incentive created by the owner's residual interest in a private-sector firm is directly related to the extent to which the magnitude of his residual claim varies with his performance. Therefore, in situations in which government involvement in a private firm in the form of loans, subsidies, or grants is extensive, the residual incentive may be severely diminished by the magnitude and pervasiveness of the public involvement. In extreme cases, the private 'owner' may become little more than a manager with severely attenuated concerns for efficiency. In these cases, public ownership may be relatively more attractive as it then permits explicitly bureaucratic forms of incentives to be created for proper performance by the firm's managers. To this point we have considered only the institutional characteristics of public ownership relative to regulation of the private sector. We will now attempt to identify those characteristics that make a crown corporation a policy instrument preferable to direct ownership and administration of economic resources by an executive department of government. Crown corporations versus departmental bureaucracies There appear to be a number of characteristics of crown corporations that favour their utilization over departmental bureaucracies in the provision of goods or services to the public at a unit price in circumstances closely resembling those in which private-sector enterprises operate in the same or similar areas. As a number of writers, including Von Mises,34 Downs,35 and Niskanen36 have pointed out, bureaux of government typically face no economic markets on the output side. Therefore, they have no direct way of valuing their output in relation to the costs of the inputs used to make them.37 In the case of organizations facing markets for their output, the sale of outputs in market transactions provides an automatic evaluation of the work of the producer. If he can sell his outputs for more than the cost of his inputs (treating normal returns on capital and entrepreneurship as costs), then he knows his product is valuable to its buyers. If he fails to cover the costs of his inputs by selling his outputs, then he knows that his product is not valuable enough.38 Downs suggests that one of the

Crown Corporations / 213

main reasons extensive formal rules are necessary in bureaux is because there are no direct measures of the value of their output. 'In many cases members of private firms can shape their behaviour on an ad hoc basis because they do not need rules to indicate how they can make profits... But whenever there is no clear linkage between the nature of an action and its value or ultimate end, pressure arises for the development of formal rules to help individuals decide their behaviour.'39 Because the outcome desired by owners of private-sector enterprises can be reduced, for the most part, to a simple profit calculus, a ready measure of the performance of an organization and individuals within it is available. To that extent, dependence on extensive formal rules can be reduced. In the absence of such a measure, other rules must be substituted. A similar phenomenon can be observed in the regulation of markets for professional services, where clients or patients possess an imperfect ability to judge the quality of service outcomes. Given the highly particularized nature of professional services, which makes generalized regulation of outputs unfeasible, public policies are often reduced to adopting input regulation, such as regulation of the quality of entrants to a profession, on the assumption (often tenuous) that there is a high correlation between prescribed training inputs and desired service outcomes.40 However, there is a significant cost attached to the substitution of detailed formal rules, primarily focused on inputs, for measures of productivity focused on outputs, such as profits. By constraining and prescribing the nature and mix of inputs, such rules also constrain the potential for innovation and dynamism in the system. These costs may not be high in many bureaucratic contexts, at least in cases of relatively routinized functions, but in an entrepreneurial setting they are likely to be substantial. To the extent that crown corporations are providers of goods or services in market settings, there are obvious advantages to emphasizing output measures of productivity rather than input measures; thus the case for removing such activities from the normal departmental setting. However, one must assume that to a greater or lesser extent every crown corporation is intended to maximize some set of policy objectives in addition to, and indeed in opposition to, profits, if this were not so, it would be difficult to conceive of any reason for a crown corporation to exist. For these non-market objectives, as for bureaucratic objectives, output measures of effectiveness will be difficult to specify. To the extent that a crown corporation is expected to engage in any substantial balancing of market and non-market objectives, the intended joint output may be difficult to specify and measure. Thus, politically uneasy and conceptually untidy compromises between input and output measures of the value of a crown corporation's activities seem unavoidable. How these compromises are struck is presumably

2147 J.R.S. Prichard and M.J. Trebilcock

in large part a matter of the relative weights of pecuniary wealth and political arguments in a corporation's objective function. These weights are likely to be reflected in institutional variations within the crown corporation mode, such as the presence or absence of both government and private-sector shareholders maximizing competing objectives, and the choice of accountability regimes - whether to schedule a crown corporation under the Financial Administration Act or leave it unaccountable to government and Parliament in this respect. In addition, to the extent that a public enterprise carries on activities in competition with private-sector enterprises, it is likely to be important that it be able to compete in both input and output markets on similar terms to those applying to its competitors. For example, if specialized expertise is required in the management of a public enterprise for it to be competitive in managerial skills with private-sector enterprises in the industry, it will be necessary for the publicly owned enterprise to be able to offer executive remuneration and follow personnel and hiring policies that are competitive with those prevailing in the industry, rather than being constrained in these respects by public-service pay scales or personnel policies. This is also the case for procurement and advertising policies. While the potential for continuous marginal policy adjustments was earlier described as an advantage of crown corporations over private-sector regulation, 'distancing' crown corporations from government constrains, to a greater or lesser extent, what may be perceived by management as political and bureaucratic interferences with management prerogatives and with an objective of competitive performance. Excessive 'closeness' or 'intermeddling' may lead to demoralization. Somewhat related to the previous point is the notion that if publicly owned enterprises are carrying on market-type activities in competition with privatesector enterprises, they should be sufficiently removed from the political process to be free to compete without excessive political intrusions that may compromise this objective; but they should not be so heavily dependent on preferential treatment by the government that the other firms in the industry perceive themselves to be unfairly prejudiced in the terms on which they are able to compete. The distancing of a public enterprise from the executive arm of government may promote these considerations. In particular cases, politicians will find it advantageous to be able to claim credit for positive activities of government or government agencies, but at the same time will want to be able to establish some distance from these activities where there are zero or negative political returns. The crown corporation structure facilitates this distancing strategy to a greater extent than a departmental bureaucracy, where the principle of ministerial responsibility severely circumscribes its utilization. Furthermore, to the extent that a government wishes

Crown Corporations 7215

to espouse (or at least to appear to espouse) a political philosophy of 'keeping government out of business' or 'reducing the size of government,' the crown corporation mode of organization may offer some advantages over direct departmental responsibility for the same activities. The one major cost to adopting the crown corporation mode over a departmental bureaucracy for organizing economic activity is that to the extent that non-market objectives are assigned to crown corporations involving outputs that cannot be readily specified or measured, the greater distance between political decision-makers and a crown corporation may increase the monitoring costs faced by government in implementing its policies. Moreover, when competitive conditions compel remuneration arrangements for executives that are tied to market measures of effectiveness and productivity, and to the extent that executives perceive their future value elsewhere in the industry to be judged largely in these terms, incentive effects are set in motion to maximize market objectives over non-market objectives in much the same way as may have led politicians to reject private-sector regulation in the first instance. The presence of non-government shareholders in a crown corporation may exacerbate these effects. Substitutability

This analysis of the legal and institutional characteristics of crown corporations serves as a basis for identifying the critical factors determining the substitutability of crown corporations for other instruments. There are three dimensions to this substitutability. First, within the class of instruments defined as crown corporations there is considerable variation with regard to the legal attributes of the corporations. As the discussion of the definition of crown corporations indicated, there is no single legal model at the federal level for crown corporations. Although current reform proposals (referred to below) would reduce this variability in legal characteristics, there is no doubt that to date the existing variability permits substitution. This allows the decision-maker some range of choice even after a decision to utilize a crown corporation as the instrument of intervention in a particular situation has been taken. The second dimension of substitutability relates to alternative forms of public enterprise. Once a decision is taken to produce goods or services publicly, a second choice must be made between the corporate form and the departmental form of production, since the two are close substitutes. At this point the legal and institutional characteristics identified above that distinguish crown corporations from departmental bureaucracies become the critical factors influencing the choice calculus.

2167 J.R.S. Prichard and M.J. Trebilcock

A current example of this second dimension of substitutability relates to the continuing controversy over the appropriate form of organization for the Post Office. Recent legislation41 has effected a change from a department to a crown corporation, adopting in statutory form the apparently widespread belief that such a change would bring substantial gains to the public.42 In the United States, various legislative proposals have called for converting the U.S. Postal Service, a public corporation, into the Post Office Department.43 The impetus for the Canadian change can be traced to both the legal and institutional characteristics of crown corporations. With respect to legal aspects, the dominant consideration is that the shift from departmental to corporate status would result in the application of the Canada Labour Code to postal service labour relations in place of the existing Public Service Staff Relations Act.44 This would change the processes for the resolution of disputes, the determination of terms and conditions of employment, and numerous other aspects of labour relations. In light of the almost constant state of controversy concerning labour relations in the Post Office, it is not surprising that an organizational substitute that holds out the promise of a better era is seen as an attractive instrument. This legal consideration is complemented by numerous institutional characteristics. The Post Office produces a service that is substantially susceptible to output measurement; it operates in an increasingly competitive market; it is a source of constant political embarrassment, which might be reduced through the 'distance' inherent in the corporate form; and it is not so integrated with other government policies and programs that the monitoring and co-ordination costs are likely to become oppressive in the corporate form. The third dimension of substitutability, and the dimension most relevant to this essay, concerns the substitution of crown corporations for taxation, expenditure policy, and regulation. We deal with this below in somewhat greater detail. Substitutability between crown corporations and other instruments The large number of crown corporations at all levels of government in Canada makes it exceedingly difficult to categorize all the activities they are engaged in. However, it is fair to say that public ownership has become an important instrument of intervention in the following fields of activity: 1 Natural monopoly regulation, as instanced by provincial hydroelectric companies and some provincial telephone systems. 2 Nation-building and community development, as instanced by Canadian National Railways, BC Railway, Air Canada, and the Canadian Broadcastin Corporation.

Crown Corporations 7217

3 Moderating economic transitions and stabilizing income, as instanced by the Cape Breton Development Corporation, DeHavilland, Canadair, the Canadian Saltfish Corporation, and the Canadian Freshwater Fish Corporation. 4 The provision of capital funds, as instanced by the Canadian Development Corporation, the Ontario Development Corporation, the Federal Business Development Bank, the Farm Credit Corporation, and the Export Development Corporation. 5 The promotion of national security and security of supply, as instanced by Petro-Canada, Polymer, Telesat, and Eldorado Nuclear. 6 The creation of a yardstick competitor, as instanced by Petro-Canada. 7 The control of externalities, as instanced by the provincial liquor control boards and the federal and provincial lottery corporations. The utilization of public ownership as an instrument of intervention in these various fields of activity reveals certain patterns regarding the relative importance of the different characteristics of crown corporations.45 Certain of the characteristics dominate as choice variables, with monitoring and information costs (Cape Breton Development Corporation, DeHavilland, Canadian Saltfish Corporation, Petro-Canada), policy co-ordination costs (provincial hydroelectric companies, Cape Breton Development Corporation), structural limitations on substitute instruments (Pacific Western Airlines, provincial hydroelectric companies), functional limitations on substitute instruments (provincial hydroelectric companies, Telesat, provincial telephone systems, Air Canada, PetroCanada, lotteries, and liquor boards), and symbolic effects (Quebec Hydro, CBC, Air Canada, Cape Breton Development Corporation, Atomic Energy of Canada Ltd, lotteries, and liquor boards) being influential in many cases. The monitoring and information costs associated with substitute instruments often make public ownership an attractive alternative, particularly when the extent of public involvement in the industry is so substantial that the residual monitoring incentives inherent in private ownership have been seriously weakened. The possibility of improving policy co-ordination by internalizing all decisionmaking to the public sector also favours the utilization of crown corporations in situations where either multiple support programs are being used or the development of the industry is integrally linked to other government programs or plans. Structural limitations on substitute instruments have induced utilization of public ownership where this has enabled provincial governments to exempt local consumers from the effects of federal taxes (hydroelectric companies) or to influence an activity they were otherwise precluded from regulating (Pacific Western Airlines). The inherent functional limitations of substitute instruments that depend on the issuing of legal orders, broadly defined, also support the

2187 J.R.S. Prichard and M.J. Trebilcock utilization of public ownership as the limits of direct regulation become apparent in situations involving novelty, uncertainty, evolving policies, and continual marginal adjustments in policy. Finally, in a substantial number of cases the symbolic characteristics of crown corporations have played an important role in the calculus of choice. Undoubtedly a full understanding of the dynamics of instrument choice requires an analytical framework even more complex than that presented above. However, a systematic consideration of the relevant legal and institutional characteristics of crown corporations in particular and instruments of intervention in general46 can illuminate the process of choice and focus attention on the dimensions of institutional change and reform. Reform Proposals

Two reform thrusts have recently won wide currency in Canada with respect to crown corporations. First, it has been widely asserted that crown corporations should be made more closely accountable to the executive and legislative arms of government. Second, it has been asserted in some quarters that a number of existing crown corporations should be 'privatized' by selling their shares or assets to the private sector. The concept of instrument choice described in this paper raises some questions about the efficacy of either reform. Proposals requiring greater accountability of crown corporations to the executive and legislative arms of government imply closer assimilation of crown corporations with line departments of government. While greater accountability may have some benefits, it also has some costs. The factors described earlier in this paper that might lead rational government decision-makers to choose crown corporations over line departments of government for the management of particular economic activities still exist. Recent reform proposals would in effect shift the boundary between crown corporations and government bureaucracies and forgo some of the advantages of distancing that have hitherto been attainable through the use of the crown corporation. The reform proposals,47 in calling for tighter budgetary controls, more explicit commitment to long-term strategic plans, tighter control over the creation of crown corporations and subsidiaries, and wider ministerial directive powers, appear to have a highly homogenizing aspect in applying essentially the same accountability regimes to broad classes of crown corporations.48 But as we have sought to show, the crown corporation has in the past displayed wide institutional variations. The flexibility that has been possible in ownership interests, accountability regimes, creation, and financing has permitted a substantial degree of custom design, whereby a crown corporation can be fashioned to meet specific policy needs. To

Crown Corporations/219

the extent that proposed reforms reduce this degree of flexibility, costs in terms of instrumental effectiveness will be incurred. Whether, in the light of possible reductions in the instrumental effectiveness of crown corporations and the substitution effects that proposed reforms are likely to engender, a net increase in social welfare (however measured) can be expected is open to question. With respect to the second reform thrust - privatization42 - proposals to sell crown corporations to the private sector rarely say explicitly whether they are motivated by a change in policy objective whereby public involvement in the sector is seen to be no longer necessary or desirable, or whether a continuing public-sector involvement is thought to be justified but superior instruments have been identified for realizing policy objectives. In the latter case, a failure to be explicit about the substitute instruments that will be invoked makes evaluation of the merits of privatization proposals extremely difficult. Such an evaluation would focus on the question of what circumstances have changed so as to affect the relative strengths of policy instruments. Thus, when crown corporations are put up for sale, the terms of the sale must include not only the price of the equity or the assets but also the commitments made openly or privately with regard to future government support or regulatory programs bearing on the industry. Undoubtedly the potential purchasers will require such information regarding future support or regulatory impact; the price they will be willing to pay for the corporations will be directly related to their evaluation of the likely impact of these alternative instruments. It is too easy for a government to inflate the apparent gains from privatization if it proceeds without disclosure of the information to the public. A final point illustrated by proposed reforms of crown corporations is the substitutability of all governing instruments. Once it is understood that the government normally has available to it a number of instruments that might be utilized to accomplish a particular policy objective, one must not be unduly sanguine about the likely impact of the reform of any one instrument. The government's ability to choose an alternative instrument (unaffected by the reform) provides an opportunity for the potential benefits of the reform to be dissipated through the substitution process. Proposals for increased accountability and for privatization appear to be particularly susceptible to this phenomenon. NOTES 1 This essay is derived in part from a study by Michael J. Trebilcock and J. Robert S. Prichard, 'Crown Corporations: The Calculus of Instrument Choice,' in Crown Corporations: The Calculus of Choice, ed. J. Robert S. Prichard (Butterworths (Canada), forthcoming).

2207 J.R.S. Prichard and M.J. Trebilcock 2 An Overview of the Origins, Growth, Size and Functions of Provincial Crown Corporations (Vancouver: University of British Columbia, Faculty of Commerce and Business Administration, Policy Analysis 1979). 3 Marsha Chandler, 'The Politics of Public Enterprise,' in Crown Corporations, ed. Prichard. 4 See, for example, W.G. Shepherd, Public Enterprise: Economic Analysis of Theory and Practice (Lexington, Mass.: D.C. Heath and Co. 1976); F.L. Pryor, 'Public Ownership: Some Quantitative Dimensions,' in Public Enterprise, ed. Shepherd, 9; U.S. Congress, Senate Committee on Governmental Affairs, Study on Federal Regulation Prepared Pursuant to S. Res. 71 to Authorize a Study of the Purpose and Effectiveness of Certain Agencies, vol. 6 (Framework for Regulation) (Washington: U.S. Government Printing Office 1978), 102-24; Robert Sexty, Canadian Government Business Corporations: Definition and Characterization, working paper no. 78-7 (mimeograph 1978). 5 For a fuller treatment of this subject, see Michael J. Trebilcock, Douglas Hartle, J. Robert S. Prichard, and Donald Dewees, The Choice of Governing Instrument (Ottawa: Economic Council of Canada 1982). 6 This discussion merely hints at the complexity of the definitional problem. For a fuller treatment of the issue, see Trebilcock and Prichard, 'Crown Corporations,' section 2. The leading reference sources include the Royal Commission on Financial Management and Accountability, Final Report (the Lambert Report), part iv (Ottawa: Ministry of Supply and Services 1979); Privy Council Office, Crown Corporations: Direction, Control, Accountability (Ottawa: Ministry of Supply and Services 1977), hereinafter 'Blue Paper'; John W. Langford, 'The Identification and Classification of Federal Public Corporations: A Preface to Regime Building,' Journal of Canadian Public Administration 23 (1980), 76-104. 7 This definition departs from the legal definition contained in the Financial Administration Act, RSC 1970, c. F-10. 8 RSC 1970, c. C-32. 9 sc 1974-75-76, c. 33. 10 See generally Elaine Kirsch, 'The Legal Environment of Federal Crown Corporations'(mimeograph 1979), 19-21. 11 sc 1970-71-72, c. 63. 12 RSC 1970, c. P-32. 13 RSC 1970, c.L-1. 14 RSC 1970, c. P-35. 15 Section 19, Financial Administration Act (hereinafter FAA). 16 Section 22, FAA 17 Section 9, FAA 18 J.E. Hodgetts, The Canadian Public Service: A Physiology of Government 1867-1970 (Toronto: University of Toronto Press 1973), ch. 7.

Crown Corporations 7221 19 Especially Ronald Coase, 'The Nature of the Firm' Economica 4 (1937), 386-406; Armen Alchian and Harold Demsetz, 'Production, Information Costs, and Economic Organization' American Economic Review 62 (1972), 777-95; John McManus, 'The Costs of Alternative Economic Organization,' Canadian Journal of Economics 8 (1975), 333-50. 20 McManus, 'The Costs of Alternative Economic Organization'. 21 Alchian and Demsetz, 'Production, Information Costs, and Economic Organization.' 22 See Thomas E. Borcherding, 'A Positive Theory of Public Enterprise,' in Crown Corporations, ed. Prichard. 23 See Alan Tupper, The Nation's Business: Canadian Concepts of Public Enterprise,' PHD thesis, Queen's University, Department of Political Science 1977. 24 See Trebilcock and Prichard, 'Crown Corporations,' section 4. 25 See Sandford Borins, 'World War II Crown Corporations: Their Functions and Their Fate,' in Crown Corporations, ed. Prichard. 26 J.M. Mintz, 'Mixed Ownership', working paper, Department of Economics (Kingston: Queen's University 1979). 27 See Borcherding, 'A Positive Theory of Public Enterprise.' 28 See generally, C.A. Ashley and R.G.H. Smails, Canadian Crown Corporations (Toronto: Macmillan of Canada 1965). 29 For example, the expropriation of Eldorado Mining can be explained primarily in these terms. See Trebilcock and Prichard, 'Crown Corporations,' section 4. 30 For example, both Telesat and Petro-Canada are responsive to this consideration. 31 In this respect, the agency literature in finance theory is highly relevant: see, for example, M. Jensen and W.H. Meckling, 'Theory of the Firm: Managerial Behaviour, Agency Costs, Ownership Structure,' Journal of Financial Economics 3 (1976), 305-60; Eugene F. Fama, 'Agency Costs and the Theory of the Firm,' Journal of Political Economy 88 (1980), 288-307. 32 Gary S. Becker, 'Comment,' Journal of Law and Economics 19 (1976), 245-8. 33 Albert Breton and Ronald Wintrobe, 'An Economic Analysis of Bureaucratic Efficiency,' workshop paper, Law and Economics Workshop Series, University of Toronto, Faculty of Law, 1979. 34 Ludwig Von Mises, Bureaucracy (New Haven: Yale University Press 1946). 35 Anthony Downs, Inside Bureaucracy (Boston: Little, Brown and Co. 1967). 36 William A. Niskanen, Bureaucracy and Representative Government (Chicago: Aldine 1971). See also Niskanen, Bureaucracy: Servant or Master? (London: Institute of Economic Affairs 1973), 10. 37 Downs, Inside Bureaucracy, 30. 38 Ibid., 29. 39 Ibid., 59. 40 See Michael J. Trebilcock, Carolyn J. Tuohy, and Alan D. Wolfson, Professional Regulation: A Staff Study of Accountancy, Architecture, and Law in Ontario

222/J.R.S. Prichard and M. J. Trebilcock

41 42

43

44 45 46 47 48

49

prepared for the Professional Organizations Committee (Toronto: Ministry of the Attorney General 1979), ch. 4. Bill c-42, first reading, 17 July 1980. Canada Post, Considerations Which Affect the Choice of Organization for the Canada Post Office: Report of a Study Group to the Postmaster General (Hull: Ministry of Supply and Services 1978). American Enterprise Institute for Public Policy Research, Postal Service Legislative Proposals (Washington: American Enterprise Institute for Public Policy Research 1977). See supra, nn. 12-14. For an extended review of this proposition, see Trebilcock and Prichard, 'Crown Corporations,' section 4. The summary above is based on that essay. See Trebilcock et al., The Choice of Governing Instrument. For example, bill c-27, first reading, 26 Nov. 1979. See also Lambert Report. For a detailed and damaging critique of recent federal definition and classification proposals, see Langford, 'Identification and Classification of Federal Public Corporations.' Little has been written to date in Canada on privatization. See, however, Robert W. Sexty, 'The Profit Role in Crown Corporations,' Canadian Business Review 5, no. 3 (1978), 9-13; T.M. Ohashi, T.P. Roth, and Zane Spindler, Privatization Theory and Practice: Distributing Shares in Private and Public Enterprise: BCRIC, PETROCAN, ESOPS, GSOPS (Vancouver: The Fraser Institute 1980); and Sexty, 'Autonomy Strategies of Government-owned Business Corporations in Canada,' Strategic Management Journal (forthcoming).

10/The Development of Health Policy in Canada GEOFFREY R. WELLER PRANLAL MANGA

This essay compares the impact of differing health policies over time. A major theme is that the nature of federal-provincial relations has been one of the main determinants of the changing nature of Canadian health policy. Health policy outputs are the result of a tremendous variety of shifting forces;1 the nature of federal-provincial relations is one crucial factor, because health policy outputs and impacts are directly related to a staged chronology largely dictated by the oscillation of power between the two levels of government. A second major concern is the evaluation of the impacts of major policy outputs in each of the stages in the development of health policy. While there are so many difficulties inherent in the analysis of policy impacts that many commentators try to remain value-neutral, an attempt will be made in this chapter to distinguish between positive and negative impacts. Before proceeding with the discussion the stages will be delineated and the criteria for evaluation will be stated. In very broad terms, health policy outputs in Canada can be said to have developed in three stages. The first stage, from 1867 to 1945, could be described as one of benign neglect when health care was basically in the private realm, even though it was gradually becoming politicized as a result of some fundamental inadequacies. Since health care was, by and large, a matter of private initiative the fact that the BNA Act assigned responsibility for health matters to the provinces was of little consequence, and health was largely unaffected by the oscillation of power between the two levels of government.2 The second stage, from 1945 to 1977, could be called the era of the shared-cost agreements, for it was one in which several federal-provincial shared-cost programs were initiated by the federal government. In this period Canada moved in slow steps toward a system of national health insurance that was intended to overcome the inadequacies of the earlier stage. Health care ceased to be provided entirely by private

224/G.R. Wellerand P. Manga

institutions and became a publicly financed service, thereby becoming subject to the oscillation of power between the federal and provincial governments. Federal-provincial disputes over health matters greatly overshadowed the arguments between government and health professionals over substantive health issues. The third and present stage, in existence since 1977, could be called the era of the Established Programs Financing Act. In this stage the provinces began to exercise greater control over their health care systems as a consequence of the introduction of a new set of financial arrangements intended to overcome the problems associated with the shared-cost method.3 The analysis of policy impacts is not an easy task. Some policy impacts are intended, others are not. Some may be felt in the short run, others in the long run. Some may affect the general public, while others may only affect certain groups in society. Moreover, most policies are likely to have a distributive component in that their impact may be unevenly felt by different groups in society.4 The analysis becomes even more complicated if attempts at evaluation are made. However, evaluation is essential, for most policy outputs have explicit or implicit value assumptions that cannot, or should not, be ignored. In the health policy field, the major thrust of policy outputs over many decades in most Western nations has been to enhance the degree of equity in health services.5 In recent years some policy outputs have been clearly intended to slow or even reverse this trend.6 Thus, any discussion of intended and unintended or longand short-run impacts of health policies must deal with the fundamental normative criteria of equity embodied in the policy outputs. The most commonly accepted understanding of equity in health care is that of equal access to health services; this implies the elimination of barriers to access for either individuals or aggregate groupings within society - financial, educational, geographic, ethnic, or any other kind. We shall regard those impacts that enhance the achievement of the equity objective in health as positive and those that impede it as negative. We shall also regard those impacts that enhance the administrative efficiency and cost-effectiveness of health programs and that tend to lessen federal-provincial conflict over health matters as positive. Those that seem to do none of these things we shall describe as being neutral. In this essay we proceed by examining the policy objectives intended to be achieved by governmental outputs in each of the three periods, and then analysing the positive, negative, and neutral impacts of these outputs. We shall illustrate that the impact of policy in each of these periods was, to some degree, a determinant of the policies that were pursued in the subsequent period. We shall also attempt to discuss some of the other forces that acted as determinants of Canadian health policy.

Health Policy 7225 Health Services before 1945

Health care was not highly politicized before the end of the Second World War. Only one of the two branches of health care, namely public health services, involved any significant degree of government action. Personal health services were generally regarded as a private sphere of activity. Most governmental action emanated from the provinces, which had constitutional jurisdictional over health matters. Their activities were largely regulatory, and attracted little political attention. Not surprisingly, bureaucratic structures dealing with health care tended to develop quite late in the period and rather haphazardly. These structures, like their political architects, tended to concern themselves largely with regulatory concerns in the field of public health. In the field of personal health services, the Canadian health care system approximated the 'closed' or 'medical' model of health in which the dominant health professionals controlled the system.7 Both the federal and provincial governments permitted a maximum of self-regulation and control by the medical professionals.8 Governments in Canada had a lower profile in health care than those in the United Kingdom and the United States.9 The consequence of this stance was a situation in which the dominant professionals were able to penetrate10 the bureaucratic and political systems and entrench a restrictive perception of health care as largely the curing of sickness in individuals. The Canadian health care system in this era was characterized by the primacy of the physician in a structure that provided care largely in terms of services to individuals in a doctor-patient relationship. This model and the delivery system associated with it were consistent with the individualistic ethic of the times. Although health care remained largely outside the political system throughout this period, there were forces inescapably pulling it into the political arena. These arose from changes within the health care system and from other values being espoused within the political system. Most of the changes within the health care system can be seen as the result of the continual and rapid rise in the costs caused by the invention and application of new technology. The continual advance of medical technology changed the pattern of illness from the predominance of widespread infectious diseases that were relatively cheap to combat to social and degenerative diseases that were very expensive to combat. It also created specialists who needed expensive equipment that only hospitals could provide and a scientific base that only the research medical schools could provide. This led to a ballooning of medical costs, changed the internal structure of the health services, and created professional rivalries. The combination of rapidly rising costs and the economic crises

226/G.R. Weller and P. Manga

of the 1930s led in Canada, as it did in the United States, to the rapid growth of private 'voluntary' insurance schemes.11 To some degree the politicization of health care was delayed by the development of these private-sector mechanisms to cope with the problem of rising costs, but it was by no means stopped. In the 1930s provincial and municipal governments, especially in western Canada where private insurance companies never really established a foothold, became increasingly involved in such matters as hospital funding.12 Later, when it became evident that the private schemes were unable to control rising costs and that they provided coverage (which became rapidly more expensive) for only a relatively small percentage of the population (generally that least in need of it) the provincial and municipal governments, along with the federal government, came under pressure to enter the health insurance field. When federal action came in Canada, it occurred in stages in the post-1945 period and, as in other nations, was rarely more successful than the earlier private-sector schemes in holding down costs in the face of advancing technology. The changes in the political realm that led to the politicization of health care were largely the result of the Depression and the Second World War. Canadians have always been more sympathetic to government activity in many aspects of social and political life than have Americans; however, the Depression sparked a greater willingness and indeed a desire for governmental activity in a wide range of policy fields. This attitude, when coupled with the tendency toward the centralization of the political system generated by the war and the fascination with Keynesian economics within the federal government and the bureaucracy, led to a situation in which proposals for a scheme of national health insurance began to be taken seriously.13 British thinking about social reconstruction, and in particular the ideas contained in the Beveridge report,14 had quite an effect in Canada. Canada, in fact, had its own equivalent of the Beveridge Report in the Marsh Report15 and it is significant that Leonard Marsh had worked on the preparation of the Beveridge Report. In 1945 the federal government brought a proposal for a scheme of national health insurance to the federal-provincial conference on reconstruction. In doing so it politicized health care and exposed it to the oscillation of power between the federal and provincial governments. The Era of the Shared-Cost Programs

The federal government took the initiative in proposing new health policy departures in the immediate post-war period at a time when federal authority was pre-eminent. Proposals for a scheme of national health insurance were part of a whole package (the so-called Green Book proposals) dealing with

Health Policy 7227

constitutional, taxation, and social policy matters presented to the provinces at the reconstruction conference of 1945.16 The federal government's proposals were part of an overall plan to enshrine federal dominance. Even though they were in a relatively weak position, the provinces balked at accepting proposals that would assign them a subordinate role in the Canadian federal system. While the English-speaking provinces, such as Ontario, objected mainly on economic grounds, Quebec objected on both economic and cultural grounds. Quebec saw provincial constitutional jurisdiction over social policy areas as a necessary defence of the French Canadian culture in the wider English environment. Largely because the federal government refused to modify its proposals significantly, the conference adjourned without agreement being reached. After these initiatives the federal government remained relatively inactive on the health care front. Gradually, however, provincial pressure began to build for a national hospital insurance plan. In 1947 the Co-operative Commonwealth Federation (CCF) government of Saskatchewan instituted a compulsory hospital insurance scheme of its own. This was probably the most dramatic health policy innovation in North America to that date, and was especially significant because Saskatchewan was a poor province.17 In 1949 Alberta and British Columbia set up their own provincial schemes. When Newfoundland entered Confederation in 1949 it became the fourth province to have a health insurance plan. The four schemes varied considerably in their terms and conditions. In 1948 the federal government instituted a system of grants, which it billed as 'a fundamental prerequisite of a nationwide system of health insurance.' This National Health Program was a composite of ten separate but related grants for such things as health surveys, general public health care, tuberculosis control, mental health, venereal disease control, facilities for crippled children, cancer control, and hospital construction. Only a few of these were shared-cost programs. The hospital construction grant, quantitatively the most significant, was a shared-cost program designed to reduce what was felt to be an acute shortage of hospital beds.18 It was well received by the provinces and presaged the Hospital Insurance and Diagnostic Services Act (HIDSA) which was to be enacted ten years later.19 The introduction of HIDSA was not only the result of the pressure from the examples of provincial public hospital insurance plans and the popularity of the shared-cost agreement for hospital construction. The analysis of the Canada Sickness Survey supplied a basis for the justification of a public program providing universal hospital insurance coverage.20 Pressure for action mounted from various interest groups, including the labour movement, and it was evident such a plan would be politically popular with the electorate. Moreover, hospitals

228 /G.R. Weller and P. Manga

were obviously in dire financial straits. It is significant that private insurance companies (most of which had little health business at the time) and doctors did not oppose the idea very strenuously. In fact, some doctors perceived that they would gain significant financial advantages. The main objectors to a publicly financed scheme of national health insurance were many of the provincial governments, most notably Quebec and, initially, Ontario. Their reservations related mainly to constitutional matters; they saw such a plan as little more than a constitutional subterfuge that would enable the federal government to meddle in their affairs. The health care objectives of the federal government in introducing H IDS A are evident in the principles the act embodied. These principles are especially noteworthy because they exerted a tremendous influence on later federal health policy and, indeed, on the Royal Commission on Health Services (the Hall Commission). The major objectives were, first, universal coverage: everyone should be covered under uniform terms and conditions; second, accessibility: services should be provided in a manner that did not impede or preclude, whether by fees or other means, reasonable access to insured services; third, the portability of benefits to all Canadians; and fourth, comprehensive coverage: all approved medical services were to be covered. The provision of out-patient services was optional: a province could, if it wished, provide all, some, or none of the in-patient services on an out-patient basis as a shareable cost under the act. The primary political objectives of the federal government in introducing HIDSA were twofold. The first was the promotion of equity both in an interpersonal and inter-governmental sense. The act was meant to be, and was, redistributive. Second, the federal government was implementing its own priorities, which were regarded as of national benefit and thereby by implication of greater relative importance than the more parochial priorities of the various provinces.21 HIDSA received royal assent in 1957 under the Liberals, but was largely implemented in 1958 when the Progressive Conservatives were in power. By 1961 all of the provinces had entered into agreements with the federal government. The financial arrangements of this act were embodied in a formula by which the federal share of a province's hospital costs were calculated to be 25 per cent of the national per capita costs plus 25 per cent of the provincial per capita costs, all multiplied by the number of insured persons for the province. The formula made no explicit reference to provincial fiscal need and it implied an open-ended scheme with no limits on federal contributions to the provinces. Specifically excluded from the financial arrangements were tuberculosis hospitals and sanatoria, institutions for the mentally ill, and institutions providing custodial care, such as nursing homes. The method of financing and administering

Health Policy 7229

the program was left to provinces but with the proviso that they were not to impose obstacles to access. The act, in fact, defined shareable costs in such a way as to provide strong disincentives to the provinces to impose direct charges on patients for hospital care: such charges were treated as 'offset' revenue, which was not shareable. HIDSA had a number of positive policy impacts. First, it provided insurance coverage to all, rich and poor alike, and eliminated the financial risk of hospital costs. Second, it was a popular program with the public, the medical profession, and most provincial governments. Third, over the years it helped to reduce disparities in the availability of hospital care. A fourth positive impact (which is often overlooked) was that federal funds greatly increased the salaries of hospital workers, who had previously been very poorly paid.22 However, HIDSA also had a number of unintended negative policy impacts. Perhaps most important, it distorted the allocation of health care resources by favouring acute care treatment. The result was rapid inflation in the cost of health care in the 1960s. Second, the open-ended nature of the arrangements resulted in the rapid inflation of hospital costs. This open-endedness meant that the federal contributions became large and largely unpredictable, especially as the act did not contain any explicit incentives for cost containment. Third, both levels of government found the act inflexible and rigid, particularly in terms of the services that could be provided and the definition of shareable costs. This rigidity hampered the task of holding down costs. Fourth, federal and provincial bureaucracies tangled over the joint responsibility for the administration of the act. The provinces complained frequently and vigorously that the administrative arrangements undermined their autonomy in the health policy field. HIDSA also had unanticipated policy impacts. The most important of these related to the question of inter-governmental equity. The act was meant to favour the poorer provinces, but it produced anomalous results. The financial formula embodied in the act was specifically designed to allow the federal government to pay a higher proportion of the hospital care expenditures of the poorer provinces. Were it not for HIDSA, the expenditures of the maritime provinces on health services would have been lower and the gap between them and the wealthier provinces would have been wider. In a more meaningful sense, however, the formula was judged to be inequitable since the per capita transfer going to the richer provinces was usually greater than for the poorer provinces. This was because the formula favoured those provinces that could make relatively high per capita expenditures; these were usually the richer provinces. The next major development in Canadian health policy occurred in 1962 when the NDP government of Saskatchewan extended its own public insurance plan to cover the cost of physicians' services. The ensuing doctors' strike was a

230/G.R. Wellerand P. Manga clear indication of the lengths to which the medical profession and associated interests were prepared to go in resisting such plans.23 However, the Saskatchewan controversy revived public interest in the issue and sparked demands from groups such as trade unions for federal action. Moreover, following Saskatchewan's example, other provinces began to introduce public insurance schemes to cover physicians' services; none of them was universal in scope or comprehensive.24 The Progressive Conservative federal government responded to these demands and pressures by establishing the Royal Commission on Health Services chaired by Emmett Hall. The commission produced a report in 1964 that called for the gradual implementation of a health charter for Canadians, including a national system of public insurance to cover doctors' bills.25 The Commission's report is the most extensive Canadian statement of the insurance approach to tackling health problems. The report was heavily criticized by the medical profession and by the insurance companies, which had built up private schemes precisely to counter the demand for a public plan such as that proposed by the Hall Commission. In addition, provincial governments were by and large opposed to the idea. The federal government, now Liberal, did not move directly to implement the main recommendations of the Hall Commission. Instead, in 1966 it introduced another shared cost program, the Canada Assistance Plan. This plan allowed Ottawa to help provide, on a 50-50 basis, a wide range of health services, drugs, dental services, and prosthetic applicances to needy persons. The program served notice of the federal intention to act upon the recommendations of the Hall Commission concerning the health charter. There were a number of reasons the federal government eventually acted upon the Hall Commission's recommendation to provide public insurance for physicians' services. Among them was a desire on the part of the Liberal party to complete the last major part of a social platform first promised in 1919, and the belief that the implementation of a plan so popular with the public (although not with many provincial governments) might help to remedy the Liberal governments' minority status. In addition, the federal insistence on Medicare in the face of the economic difficulties of the late 1960s might be explained as part of an attempt to stem the increasing decentralization of the political system by initiating a scheme that, it was hoped, would promote uniformity in health standards across the nation, thus perhaps enhancing mobility and national unity. The policy objectives of Medicare were, in fact, much the same as they were for HIDSA. The intent was to establish a national program with public financing to promote interpersonal and inter-governmental equity and to provide free access to services.

Health Policy 7231

The Medical Care Act received royal assent in 1966 and was implemented in 1968.26 By 1972 all the provinces and territories had entered into agreements with the federal government. Under Medicare the federal government paid 50 per cent of the national per capita medical costs multiplied by a province's insured population. Unlike HIDSA, where joint administration had been undertaken largely to appease physicians, the administration of the Medical Care Act was almost exclusively provincial. The auditing requirements, for instance, were minimal when compared with those applied in HIDSA. It should be noted that the Medical Care Act made reference to the need for reasonable incomes for physicians and the freedom of doctors to choose their preferred method of remuneration. The act did not prohibit opting out or extra billing, although it was clear that resorting to user fees that inhibited access to physician services would be considered inimical to the intentions of the act. The act outlawed private insurance for basic physicians' services. The Medical Care Act had a number of positive policy impacts which, like the negative ones, are difficult to isolate in so far as they resulted from the combination of both insurance plans. The act established one public insurance scheme to replace the patchwork of private and public plans that had provided variable coverage at varying costs to members. The act extended this uniform system to all, rich and poor alike. Another positive policy impact was the partial reduction of the inefficiencies produced by having only a public insurance scheme; the incentives for doctors to admit patients for hospital care that had been engendered by HIDSA were reduced. A significant increase in medical manpower was brought about by the act; one consequence was a marginal improvement in accessibility to physician's services for those living in remote regions and earning low incomes. The act produced a more equitable interprovincial transfer of funds. The Medical Care Act also had a number of negative, and to some degree unexpected, policy impacts. Physicians' incomes were significantly increased largely because patients' demand for care increased and bad debts were eliminated. By 1971 the relative incomes of physicians were at their peak.27 The effect of Medicare was to redistribute income from the general taxpayers to physicians. In the long run, Medicare was also inflationary. It placed great pressure on provincial and federal public finances for spending in the health care field. The financial provisions of the Medical Care Act were, like HIDSA'S, open-ended, and this made the federal government's contributions unpredictable - a fact that the federal government regarded as a negative policy impact. Provinces objected to what they considered an extension of the unwarranted invasion of provincial jurisdiction begun by HIDSA. Medicare allowed the federal government to dictate social policy inter-sectorally (e.g., health versus housing) andintra-sectorally (e.g., physicians' services over other health services) within the provinces.

232/G.R. Wellerand P. Manga Another provincial objection related to the development of shared-cost programs in general. By the late 1960s, there was widespread recognition that all social services were closely related and that separate and largely uncoordinated programs that were poorly tailored to regional needs, that were often sprung upon the provinces without warning, and that often existed at the mercy of federal benificence, were perhaps not the most organized and efficient way of approaching the problems. The shared-cost agreements distorted the type of health services to which people had access. Access to expensive highly technological, curative, individually oriented medicine was improved, and these services were emphasized to the detriment of preventive and less costly forms of health care services. The Lalonde report28 was one of a number of studies that questioned the assumption, built into the insurance approach and the shared-cost agreements, that doctors plus hospitals equals health. The report showed that uncritical acceptance of this assumption could have extremely costly consequences. Because of the accumulation of the negative policy impacts of both HIDSA and Medicare, the federal and provincial governments began searching for an alternative to shared-cost financing, although not to the principles embodied in the programs. Essentially the shared-cost approach was regarded as having produced serious constitutional problems and problems of economic efficiency.29 From the federal point of view the main complaint was the open-endedness of the programs. Both levels of government were worried about the broader issues of accountability to their electorates for health care spending and the distortion of provincial priorities created by the programs. Both were beginning to be worried by distortions produced within health services as between curative and preventive services and were searching for ways to correct the problem30 and both levels of government saw that some of the anomalous interprovincial equity impacts of HIDSA had to be corrected. Health Services and the Established Programs Financing Act The federal government formally began the process of renegotiating the sharedcost agreements in November 1971. The negotiations lasted more than six years and were notably contentious. They culminated in the passage of the FederalProvincial Fiscal Arrangements and Established Programs Financing (EPF) Act on 1 April 1977.31 Like the act, the negotiations concentrated on the financial arrangements between the two levels of government and attempted to disentangle the provincial and federal administrative structures. There was little discussion of (and few changes in) the principles underlying the health programs. The

Health Policy 7233 idea of dropping the public insurance concept did not occur, and there was little discussion of the specifics of health care delivery programs. Several characteristics of the negotiations reveal a great deal about the process of health policy determination. There was little pressure group involvement. Some major groups expressed opinions now and then; the federal government received a large number of briefs, most of which supported its position, although some, like those from the Canadian Labour Congress and various voluntary health agencies, opposed it. Pressure groups did not, however, participate directly and really seemed only to keep a 'watching brief on the situation as it developed. Another characteristic of the negotiations was that they were never very much in the public eye. The media covered only the highlights, and then only in a perfunctory way. The complexity of the subject and the long, drawn-out negotiations also mitigated against continued public interest or concern. During the negotiations, the ministers of health took a back seat to the ministers of finance and the first ministers, especially at the critical junctures and toward the end. This was because substantive health matters were not under discussion, but simply the financing of established programs of which HIDSA and Medicare, though the most important, were only a part.32 Politically the objective of both the federal and provincial governments was to come to an arrangement that was constitutionally more appropriate than that embodied in the shared-cost agreements. In terms of finance, the main objective of the federal government was to create a closed-ended fiscal transfer that was reasonably predictable and stable. Another major federal objective was to create a more equitable outcome in transfers to the provinces. The government also wanted to promote cost-cutting and greater efficiency by instituting a system where any savings would accrue to the province, and by eliminating the distortion of provincial spending that had resulted from the distinction between insured and non-insured programs, such as home care and extended health care, that had been a part of the shared-cost programs.33 The EPF Act constituted a major change in the financing of post-secondary education, hospital care, medical care, and selected health services previously covered under the Canada Assistance Plan. The fiscal arrangements were complex, and essentially replaced the HIDSA and Medicare cost-sharing arrangements with a cash transfer coupled with surrendered federal tax room. The tax transfer under the EPF Act arrangements consisted of one corporate and 13.5 personal income tax points, equalized to the national average. The 'tax transfer' meant a reduction in federal tax accompanied by a corresponding increase in provincial income taxes. The cash payments consisted of 'basic cash' contributions, 'transitional adjustments,' and 'levelling' payments. The basic cash portion

234 /G.R. Weller and P. Manga

was calculated by taking 50 per cent of the federal contributions under the three established programs in 1975-6 and escalating by the rate of growth of the GNP. The basic cash contributions were intended to provide for stable, longterm, and fairly predictable funding and for a continued federal presence. The transitional adjustments and levelling payments topped up the value of the tax transfer to ensure that no province lost as a result of accepting part of the federal contribution in the form of a tax transfer. The EPF Act arrangements had a number of positive policy impacts. The provinces were much happier with the new constitutional situation, for the EPF Act essentially returned full responsibility for health policy determination to the provinces where, it is agreed, the constitutional responsibility resides. The provincial briefs to the Hall Review of 198034 were generally supportive of the EPF Act constitutionally, especially Ontario's. It should be emphasized that only the level of jurisdictional squabbling diminished; there was still a dispute over the level of financing, and now a controversy concerning the underlying principles of the program is developing. A positive political policy impact was the greater provincial satisfaction with the lower level of federal bureaucratic involvement in health programs. From the federal government's point of view, another positive impact was the supposedly more clearly defined financial obligations resulting from the closedendedness of the new arrangements. This is not to say that the federal government remained happy with the quantum of the transfer. In fact, the federal government felt that it had been too generous in the 1977 agreements, and pointed to figures showing the increasing federal proportion of total health expenditures. The federal share was roughly 50 per cent in 1977, and had risen to 56 per cent by 1979-80.35 Still another positive financial policy impact was the increased financial flexibility afforded the provinces by the new arrangements. This was certainly regarded as positive by all of the provinces except Saskatchewan, judging by their briefs to the Hall Review. Saskatchewan's objection was based on the argument that the flexibility came at the expense of mechanisms for enforcing the principles carried over from the shared-cost agreements. Saskatchewan argued that this implied that the federal government had opted out of its commitment to bringing about the full realization of the health charter for Canadians. There is evidence that the provinces did utilize their new-found flexibility, and did de-emphasize hospital care. It is important to note, however, that even before the incentives and flexibility allowed under the EPF Act came into being, the provinces were practising cost-containment policies. Because these policies began in the early 1970s, it is difficult to judge the EPF Act's contribution to holding down costs. Theoretically it should have had this effect.

Health Policy 7235

In terms of negative policy impacts, one must be careful to distinguish between those that may properly be ascribed to the EPF Act and those that are associated with it in the public mind but for which it may not be truly responsible. It is certain that the act enhanced trends toward the balkanization of the health care system and provincial efforts to privatize the costs and financing of hospital and medical care. Specific provincial cost-containment measures should not necessarily be directly attributed to the act, nor should the recent problems relating to opting out and extra billing. The most feared negative policy impact of the EPF Act is that it is balkanizing the Canadian health care system. This means that the design and operation of the various provincial health care plans have diverged to the degree that universal coverage under uniform terms and conditions and with full portability of benefits has been endangered. The federal government and certain powerful lobby groups, such as the Canadian Labour Congress, argue that the homogeneity of health programs achieved under HIDSA and Medicare, which gave the schemes a national character, is at risk because of the inability of the federal government to enforce the basic principles embodied in HIDSA and Medicare and still technically in place under the EPF Act. The federal government has only the 'all-or-nothing' sanction of withdrawing all cash transfers from a province that does not live up the those principles. This penalty is politically almost impossible to impose. The EPF Act created a situation in which provincial attempts to privatize the costs and financing of hospital and medical care were not clearly and strongly discouraged. 'Privatization' means the imposition of various types of direct charges to patients, such as hospital per diems, co-payment for physicians' visits, deterrent charges, and the like. Although the wording was vague, the HIDSA and medicare schemes did discourage user charges, and only minimal fees were in place. The potential privatization of health financing under the EPF Act is regarded as a negative policy impact because it is possible that the gains in equity and accessibility achieved under HIDSA and Medicare will be lost and portability endangered.36 Since the EPF Act came into force, the provinces have raised the level of existing authorized charges for hospital care and imposed new ones. Resort to such financing deters access to care by the poor, and will reintroduce the financial risk associated with illness - a risk that, under HIDSA and Medicare, was almost climated.37 These negative effects have so disturbed the critics of the EPF Act that they fear the eventual demise of public health insurance in Canada. Their fears are real, for these impacts do jeopardize the achievements of the federal government in the health care field.38 The EPF Act has also engendered what might be termed mixed policy impacts. This is clearly seen in the issues of the sharing of the burden of total health costs

236/G.R. Wellerand P. Manga between the federal and provincial governments and the equitability of federal transfers as between the provinces. The EPF Act has increased the percentage of total health care costs paid by the federal government. Monique Begin, the minister of national health and welfare, has claimed39 that some of the provinces have 'misappropriated' funds, and has threatened to stop all cash transfers to some provinces.40 The suspicion that there had been misappropriation of funds helped determine the federal government's position in the 1982 renegotiations. The minister of finance stated that the federal government was paying too high a proportion and wished to reduce it. The provinces contested the federal view; they argued that the immediate or short-run effect of Medicare was fully expected by all concerned to increase the federal proportion but that, because of the design of the EPF Act, its long-run impact will be to reduce the federal proportion of total costs, since provincial costs will rise faster than the federal contribution.40 It has been noted that one of the policy objectives of the EPF Act was to obtain a more equitable transfer of health care funding from the federal government to the various provincial governments. In one sense this was achieved, because the cash transfers were identical per capita and the burden of taxation was heavier on the richer provinces. This situation was certainly an improvement over that which had existed under HIDSA and Medicare.41 But the Maritime orovinces complained that since the value of tax points is equalized not to that of the richest province but to the average of all provinces, the richer provinces stood to gain more than the poorer provinces.42 These dual effects were the consequence of the EPF plan being a composite of cash and tax transfers. The cash component was an equalizing factor, but the tax transfer may have pulled in the opposite direction. The issue of fiscal redistribution has always been a contentious one in the designing of specific inter-governmental transfers. The poorer provinces have always insisted that the matching formula explicitly include fiscal capacity variables. The richer provinces have always resisted such an inclusion; they argue that there are other formal equalization grants that should deal with the problem. In the final analysis, the degree of equalization built into any intergovernmental agreement is a strictly political decision. There are a number of what can only be described as alleged policy impacts of the EPF Act. Numerous briefs submitted to the Hall review argued that the new arrangements generated incentives for provincial treasuries to cut, squeeze, or freeze hospital operating and capital budgets. The EPF Act may not, however, have been truly responsible for these phenomena because they were realized in a much more dramatic fashion long before the introduction of the Act.43 This is

Health Policy 7237

perhaps most obvious in the case of Ontario's attempts to hold down hospital costs and, indeed, close some hospitals long before the EPF Act came into force. It is sometimes argued that the act was responsible for the small increases in physicians' fee schedules made by the provincial ministries of health. The resulting opting out and extra billing that caused so much alarm nationally was a post-EPF Act phenomenon, which led many to believe there was a cause-andeffect relationship between them. This allegation, however, runs contrary to the available evidence. In fact, the relatively small fee schedule increases were a part of the 1971 6 period, during the pre-EPF Act years. Since the inception of the act the annual average fee schedule increases have been considerably higher than before in virtually every province.44 It should also be remembered that between 1975 and 1978 physicians' fee increases were subject to Anti-Inflation Board regulation, and it is possible that the sudden spate of troubles with opting out and extra billing may be the result of the end of the AIB.45 The Present Medicare Crisis

Governmental restraint on hospital expenditures, opting out, and extra billing are the principal issues that define the latest 'Medicare crisis.' There are, however, two other major ways in which the crisis is manifested. One is the imposition of new and additional user charges; the other is the dramatic increases in health care premiums that have occurred in the three provinces that levy them. Provincial restraint policies created a great deal of concern among those groups that had originally pressured for Medicare and among the general public. This was because these policies primarily involved cutting back resources, both physical and human, and did little or nothing to effect systemic changes such as increasing the use of physician substitutes or promoting the use of community health centres and preventive services. There were no attempts to change the style of doctors' practice in order to produce greater efficiency. Nevertheless, the provincial efforts to contain costs were remarkably successful. Health care costs as a percentage of the GNP were stable at about 7 per cent throughout the 1970s. This experience is unique in the Western world.46 It is also notable because it demonstrates that publicly financed universal health insurance schemes are not inherently inflationary, and that the Canadian health system is not incompatible with the containment of health care expenditures. The present medicare crisis also manifests itself in heightened conflict between the medical profession and the provincial governments. This conflict relates to such matters as the administration of the insurance plans, doctors' allegations of excessive governmental regulation, and a lack of meaningful consultation

238/G.R. Wellerand P. Manga

between the two groups. The most contentious issue, and the one that characterizes the relationship between the medical profession and the provincial governments, the issue of fee schedules, which, in effect, means incomes. While the provinces have successfully held down physicians' fee increases in response to escalating health care costs, doing so has had a number of negative consequences. It has encouraged 'revolving-door' practice as doctors respond to limited fee increases by increasing the volume of services supplied. It has encouraged extra billing and opting out. Moreover, it has magnified the division between surgeons and specialists on the one hand and general practitioners on the other. There have been attempts to improve the negotiating process in some provinces (such as Ontario), but the relationship between the two parties is generally still poor. It should be noted here that throughout this period physicians' incomes rose much faster than the nominal increase in fee schedules might lead some people to believe.47 The physicians' perceived decline in their relative incomes resulted in opting out, most notably in Ontario, and extra billing, again most notably in Ontario but also in Alberta and several other provinces. There were, in fact, threats of opting out and extra billing in virtually every province in Canada. Opting out involves leaving a plan and billing separately; extra billing means that a physician stays within a Medicare plan but levies additional charges. The effects of both these actions are identical. They threaten the survival of Medicare because they lead to a two-class system of health care by reducing the accessibility of health services to low-income patients and violating the principle of portability. They raise the issue of consumer risk-bearing relative to a system of universal first-dollar insurance coverage. They transfer the burden of health care costs from the general taxpayer to the sick, and are therefore inequitable. Neither extra billing nor opting out is likely to enhance cost control, reduce utilization levels, or increase the technical efficiency with which medical services are delivered. The only real rationale for opting out and extra billing is that they raise the level of physicians' incomes. The provincial governments did not react quickly to the development of these two phenomena, and at first expressed an amazing indifference to them. Even now very little has been done by the provincial governments to tackle the problem, other than agreeing to increase the fee schedules in attempts to mollify the physicians. The federal government, in contrast, quickly and unequivocally condemned both practices.48 The current Medicare crisis is also reflected in the imposition of new and additional direct charges to patients for hospital use. There have been dramatic increases in health care premiums in those provinces that levy them. These devices are generally regarded as a particularly regressive way of financing

Health Policy/239 health care. Direct charges in particular impose a direct financial burden on the sick and are thus especially inequitable. They have little or no effect on the utilization of hospital care, are an administrative burden and, most importantly, they shift the financing of hospitals from the public to the private sector.49 The effects of premiums have been fully described elsewhere.50 We simply note here that premiums have always been condemned as regressive and as adversely affecting the principles of universality and portability so fundamental to Canada's health programs. They have no substantive impact whatsoever on the actual operation of a health care system. It is true that the provinces that levy premiums do exempt the aged and those living below the poverty line, and provide partial assistance and subsidies to low-income families, but an Ontario study revealed that a substantial proportion of persons entitled to these subsidies (approximately 30 per cent) did not receive them.51 It might be noted also that the three provinces that impose premiums have relatively low rates of personal income tax and healthy corporate tax bases; they have the ability to raise taxes and do away with premiums as the other provinces have done since the introduction of HIDSA. The developments just described so alarmed those groups that had initially pressured for HIDSA and Medicare that they began to organize to attempt to preserve hard-won advances in social policy. In November 1979 the Canadian Labour Congress initiated and organized the SOS Medicare Conference. This conference brought together over one hundred concerned organizations, ranging from the United Church of Canada and the Canadian Public Health Association to the Inuit Tapirisat of Canada, to make common cause in the defence of Medicare. The objective was not only to save Medicare but to try and improve the program by making it more cost-effective. It was suggested that this could be achieved through measures such as a greater ratio of convalescent and chronic care beds and home care programs; the use of nurse-practitioners and paraprofessionals; the provision of community health centres and life-style health programs; and the development of an organization of the consumers of health services.52 Pressure was also brought upon both levels of government in many other ways. The New Democratic Party, for instance, made the preservation of Medicare a major issue. In Ontario, the party launched a major campaign on the issue, opposing such things as hospital cutbacks and extra billing. The party even presented the Ontario legislature with a petition signed by over 250,000 people opposed to extra billing.53 Partly as a result of these pressures, a newly elected progressive Conservative federal government decided to institute not only a review of the EPF arrangements but also health policy in general. Health Minister David Crombie appointed Emmett Hall to conduct the review. The review document was

240/ G.R. Weller and P. Manga

published in 1980 and turned out to be a disappointment. While Hall's mandate was significant in that it called for a review of the fundamentals of the health care system, the document54 was superficial and contradictory and gave no clear indication of where the future should lead. The review clearly shows that Hall was sympathetic to the doctors' view that they were underpaid, even though studies done specifically for the review showed that doctors' incomes were significantly higher than published figures indicated.55 According to Hall, the major problems confronting the health care system were doctors' low incomes and the imposition of various forms of direct charges. Hall's main recommendations were that extra billing should be outlawed because it was a threat to the system, and that binding arbitration should be the device used to resolve the conflict between the provincial governments and the medical profession over fee levels. The provincial governments liked the idea of binding arbitration even less than the medical profession, for they thought it would give the physicians greater influence at the bargaining table.56 The Hall group was the first body to have a mandate to review the fundamental principles of Medicare; unfortunately, the review represents a lost opportunity. Another response to the crisis in health care was the creation of the ParliamenForce on Federal-Provincial Fiscal Arrangements are bound to become points which reported in 1981,57 agreed with Hall that extra billing inhibited access to health care. It also stated that hospital and Medicare premiums were a regressive form of taxation. The task force agreed with Hall's suggestion that there should be resort to binding arbitration if negotiations between provincial governments and physicians failed. The task force recommended that clear program conditions should be established, supported by explicit criteria against which compliance with the conditions could be measured. If a province failed to meet the conditions, federal support for that province's plan was to be withheld. The original five-year funding period of the EPF Act expired on 31 March 1982. There were lengthy and contentious discussions before that date between the federal government and the provinces concerning the conditions and level of funding that should apply in the next period. In February 1982 the federal government declared that while negotiations would continue with respect to the clarification of national standards for health care, it was going to proceed unilaterally to amend the act's financial formula. Bill c-97 was duly introduced and received royal assent on 7 April 1982. The amendments revised the formula such that it more clearly tied federal contributions to the growth of the Canadian economy, with the likely effect that over the next five-year period the federal government will pay out about $5.7 billion less than it would have under the old formula.

Health Policy 7241 Conclusions: Future Policy Directions

All recent reviews of health care services58 have supported and defended the basic principles of the health insurance schemes. It is probable that the concepts of universal coverage essentially free of charge - that is, free access to care in a publicly administered program - will be maintained. There will, however, be a great deal of disagreement over precisely how the provinces can be made to uphold fully these principles. The suggestions made by the Parliamentary Task Force on Federal-Provincial Fiscal Arrangements are bound to become points of contention in the negotiations. One suggestion was that the conditions for the receipt of federal cash transfers be much more clearly defined. Another was that the new agreement incorporate financial penalties or disincentives to be applied against those provinces that do not meet the more clearly defined conditions. For instance, there could be a stipulated cash penalty for those provinces not covering at least 99.5 per cent of their residents; or the federal government could withhold a portion of the cash transfer to a province equal to the amount of extra billing in that province. Whatever the outcome of the discussions, it is clear that the long-range future of health policy in Canada will increasingly reside with the provincial governments. Although the federal government may seek to ensure the maintenance of basic principles, it will not have a major say on substantive health policy issues. This has a number of serious implications for both levels of government. The provinces will now have to determine policy on substantive health issues. While a number of federal and provincial reviews, such as the Lalonde and Hastings reports,59 have been concerned with substantive issues such as the balance between curative and preventive medicine, the need for community health centres, and physician substitutes, these issues have never been seriously addressed by either level of government because the constitutional, financial, and inter-governmental aspects of health policy have always overshadowed them. The provinces will have to live with the fact that there are going to be fewer and fewer chances to pass the buck to the federal government as a means of placating either the medical profession or the public. Experiencing direct fiscal and political accountability may not be pleasant for many provinces. As far as the federal government is concerned, the provincialization of health policy determination leaves them with at least one very important question to answer; namely, what will be the role of the Department of National Health and Welfare? In the 1940s, 1950s, and 1960s the department played a significant role in designing and implementing major health policy innovations. These included not only the HIDSA, CAP, and Medicare insurance schemes, but also the Medical

242/G.R. Weller and P. Manga Services Branch which delivers health services in the Canadian North, and the Health Protection Branch. The EPF Act has, however, dramatically changed the responsibilities and therefore the influence of the department. The responsibilities of the department may be further reduced by the eventual loss of the Medical Services Branch as a result of the drift toward provincial status of the Yukon and Northwest Territories and the transfer of responsibility for Indian health to the provinces.60 The federal government's response to these changes is not clear, but certainly one of the policy impacts of the EPF Act will eventually be major structural changes in the Department of National Health and Welfare. Two major conclusions concerning policy analysis can be drawn from the preceding discussion. The first is that the development of Canadian health policy is not the result of any single causal factor. If it had been, it would have presumably been much smoother and more homogeneous. Moreover, no one methodology fully explains what went on. While it is clear that the peculiar nature of Canadian federalism has had a considerable effect on the timing and nature of Canadian health policy, numerous other institutional, ideological, and technological factors also had major roles to play. Thus, a case has implicitly been made for eclecticism. The second conclusion is that the provincialization of health is going to affect markedly the nature of the health policy process in Canada, and thereby the nature of the analyses that will have to be made. The institutional factor of the nature of federal-provincial relations will recede in importance, and ideological and pressure-group explanations of the determination of policy within the provincial context will become more important. Life will be more complicated for health policy analysts, because they will find it increasingly important to understand the particular nature of the political cultures and policy processes of each province. The fact that health policy is increasingly being provincialized and becoming subject to ideological debate can be seen in the disputes over physicians' incomes and user fees. Both of these issues are being resolved largely at the provincial level, and both are markedly ideological in nature; the user-fee issue involves an attempt to reintroduce elements of the private sector into those parts of the health system brought into the public sector by the health insurance schemes. It would seem that the health policy field will continue to be as complicated and controversial as it has always been. NOTES 1 See Geoffrey R. Weller, 'The Determinants of Canadian Health Policy,' Journal of Health Politics, Policy and Law 5, no. 3 (Fall 1980), 405-18. 2 For a brief description of the alternation of power between the two levels of government see P. Fox, 'Regionalism and Confederation,' in Ontario Advisory

Health Policy 7243 Committee on Confederation, The Confederation Challenge, vol. 2 (Toronto: Queen's Printer 1970, 1-27. For a good survey of federal-provincial fiscal relations see A.M. Moore, J.H. Perry, and D.I. Beach, The Financing of Canadian Confederation: The First Hundred Years (Toronto: Canadian Tax Foundation 1966. For more recent developments see P. Lewis, 'The Tangled Tale of Taxes and Transfers' in The Fraser Institute, Confederation at the Crossroads (Vancouver: The Fraser Institute 1978), 39-105. 3 For general descriptions of the Canadian health care system see Maurice LeClair, 'The Canadian Health Care System,' in National Health Insurance: Can We Learn from Canada? ed. S. Andreopoulos (New York: Wiley 1975); M.G. Taylor, Health Insurance and Canadian Public Policy: The Seven Decisions That Created the Canadian Health Insurance System (Montreal: McGill-Queen's University Press 1978); L. Soderstrom, The Canadian Health System (London: Croom Helm 1978); C.H. Shillington, The Road to Medicare in Canada (Toronto: Dell Graphics 1972); D. Swartz, 'The Politics of Reform: Conflict and Accommodation in Canadian Health Policy,' in The Canadian State: Political Economy and Political power, ed. L. Panitch (Toronto: University of Toronto Press 1977), 311-43; Perspectives on Canadian Health and Social Services Policy: History and Emerging Trends, ed. C.A. Meilicke and J.L. Storch (Ann Arbor: Health Administration Press 1980); Health and Canadian Society: Sociological Perspectives, ed. D. Coburn, C. D'Arcy, P. New and G. Torrance (Toronto: Fitzhenry and Whiteside 1981). 4 For a discussion of these matters see J. Grumm, 'The Analysis of Policy Impact' in F.I. Greenstein and N.W. Polsby, Policies and Policy-Making (Reading, Mass.: Addison-Wesley 1975). 5 See Pranlal Manga and Geoffrey R. Weller, 'The Failure of the Equity Objective in Health: A Comparative Analysis of Canada, Britain and the United States,' Comparative Social Research 3 (1980) 229-67. 6 See Geoffrey R. Weller and Pranlal Manga, 'The Re-emergence of the Private Sector in Health: A Comparative Analysis of Canada, Britain and the United States,' paper prepared for presentation at the Twelfth International Political Science Association Congress, Rio de Janeiro, Aug. 1982. 7 See G.M. Torrance, 'Socio-historical Overview' in Health and Canadian Society, ed. Coburn etal., 13-19. 8 See G.R. Weller, 'Health Care and Medicare Policy in Ontario,' in Issues in Canadian Public Policy, ed. G.B. Doern and V.S. Wilson (Toronto: Macmillan of Canada 1974), 85-114. 9 See Torrance, 'Socio-historical Overview,' 19. 10 See M.G. Taylor, 'The Role of the Medical Profession in the Formulation and Execution of Public Policy,' Canadian Journal of Economics and Political Science 26, no. 1 (Feb. 1960), 108-27.

244/G.R. Wellerand P. Manga 11 See Shillington, 'The Road to Medicare.' 12 See Report of the Royal Commission on Health Services, 2 vols. (Ottawa: Queen's Printer 1964 and 1965). 13 These developments are detailed in D.V. Smiley, Constitutional Adaptation and Canadian Federalism Since 1945 (Ottawa: Queen's Printer 1970), ch. 11, 9-24, and P. Fox, 'Regionalism and Confederation.' 14 Sir William Beveridge, Social Insurance and Allied Services, Cmd. 6404 (London: HMSO 1942). 15 L. Marsh, Report on Social Security for Canada (Ottawa: King's Printer 1943). 16 Dominion and Provincial Submissions and Plenary Conference Discussions, Dominion Provincial Conference, 1945 (Ottawa: King's Printer 1946). 17 See Taylor, Health Insurance and Canadian Public Policy, ch. 2. 18 See D.M. Herron, 'National Health Grants,' Medical Services Journal (Oct. 1967), 1118-23. 19 RSC 1970, c. H-8. 20 See M.G. Taylor, 'Government Planning: The Federal-Provincial Health Survey Reports,' Canadian Journal of Economics and Political Science 19 (Nov. 1953), 501-10. 21 See G. Carter, Canadian Conditional Grants Since World War II (Toronto: Canadian Tax Foundation 1971). 22 See G.M. Torrance, 'Hospitals as Health Factories,' in Health and Canadian Society, ed. Coburn et al., 257. 23 See R.F. Badgley and S. Wolfe, Doctors' Strike: Medical Care and Conflict in Saskatchewan (Toronto: Macmillan of Canada 1967); E.A. Tollefson, Bitter Medicine (Saskatoon: Modern Press 1963); K. MacTaggart. The First Decade (Ottawa: Canadian Medical Association 1972). 24 See Shillington, 'The Road to Medicare.' 25 Report of the Royal Commission on Health Services. 26 RSC 1970, c. M-8. 27 A.D. Wolfson, R.G. Evans, and J. Lomas, Physicians' Incomes in Canada (unpublished background study commissioned by E. Hall for the 1980 Review). 28 M. Lalonde, A New Perspective on the Health of Canadians: A Working Document (Ottawa: Department of National Health and Welfare 1974). 29 T.J. Courchene, Refinancing the Canadian Confederation: A Survey of the 1977 Fiscal Arrangements Act (Montreal: C.D. Howe Research Institute 1979). 30 Lalonde, A New Perspective, and Hastings Committee, The Community Health Centre in Canada (Ottawa: Information Canada 1972). 31 sc 1976-77, c. 10.

Health Policy 7245 32 See R.J. Van Loon, 'From Shared Cost to Block Funding and Beyond: The Politics of Public Health Insurance in Canada,' Journal of Health Politics Policy and Law 2, no. 4 (Winter 1978), 454-78. 33 See G.R. Weller, 'Recent Intergovernmental Health Care Relations in Canada: The View from Ontario,' paper presented to the annual meetings of the Western Political Science Association, Los Angeles, 16-18 March 1978. 34 E.M. Hall, Canada's National-Provincial Health Program for the 1980s: A Commitment for Renewal (Ottawa: Department of National Health and Welfare 1980). 35 Ibid., 11. 36 For a discussion of the extent of the equity gains in a comparative context, see P. Manga and G.R. Weller, 'The Failure of the Equity Objective in Health: A Comparative Analysis of Canada, Britain and the United States,' Comparative Social Research 3 (1980), 229-67. 37 See R.G. Beck, 'The Effects of Co-payment on the Poor,' Journal of Human Resources 9,129-42; R.G. Beck, 'Economic Class and Access to Physicians' Services under Public Medical Care Insurance,' International Journal of Health Services 3, no. 3 (1973), 341-55; M.L. Barer, R.G. Evans, and G.L. Stoddart, Controlling Health Care Costs by Direct Charges to Patients: Snare or Delusion? (Toronto: Ontario Economic Council 1979). 38 M.C. Brown, 'The Implications of Established Programs Financing for National Health Insurance,' Canadian Public Policy, 6 no. 3 (Summer 1980), 521-32. 39 M. Begin, statement at the SOS Medicare Conference, Ottawa, 5 November 1979. 40 See P. Manga, 'Fiscal Aspects of the Hall Review,' Canadian Taxation 3 (Spring 1981), 45-53. 41 See Robin Boadway, Intergovernmental Transfers in Canada (Toronto: Canadian Tax Foundation 1980), ch. 4. 42 See Brown, 'The Implications of Established Programs Financing. 43 See C. Kapsalis, Block Funding and Provincial Spending on Social Programs (Ottawa: Department of National Health and Welfare 1980). 44 See Canada, Department of National Health and Welfare, Price Changes and Physicians' Services (Ottawa: Health Information Division, Department of National Health and Welfare 1980) and A.D. Wolfson and C.J. Tuohy, Opting Out ofOHIP, (Toronto: Ontario Economic Council 1980). 45 Kapsalis, 'Block Funding, 15. 46 Ibid. 47 R.G. Evans and A.D. Wolfson, 'Moving the Target to Hit the Bullet: Generation of Utilization by Physicians,' paper presented to the National Bureau of Economic Research Conference on the Economics of Physician and Patient Behaviour, Stanford, Calif., 27-28 Jan. 1978.

246/G.R. Weller and P. Manga 48 See P. Manga, 'Physician Extra Billing: What Can Governments Do?,' paper presented to the Canadian Health Forum, Ottawa, 21-23 June 1979. 49 See Barer et al., Controlling Health Care Costs. 50 Issues and Alternatives: Update 1979 (Toronto: Ontario Economic Council 1979), 23-40. 51 Report of the Select Committee on Health Care Financing and Costs (Toronto: Queen's Printer 1978). 52 See D. McDermott, Keynote Address to the SOS Medicare Conference (Ottawa: Canadian Labour Congress 1979) and Preliminary Report in the SOS Medicare Conference (Ottawa: Canadian Labour Congress 1979). 53 Globe and Mail, 21 Nov. 1979. 54 E.M. Hall, Canada's National-Provincial Health program. 55 A.D. Wolfson et al., Physicians' Incomes in Canada. 56 See P. Manga, 'Arbitration and the Medical Profession: A Comment on the Hall Report,' Canadian Public Policy 6, no. 4 (Autumn 1980), 670-7. 57 Parliamentary Task Force on Federal-Provincial Fiscal Arrangements, Fiscal Federalism in Canada (Ottawa: Department of Supply and Services 1981. 58 Illustrative of the sudden spate of reports issued in the early 1970s are Committee on the Costs of Health Services, Task Force Report on the Costs of Health Services in Canada, 3 vols. (Ottawa: Queen's Printer 1970); Report of the Committee on the Healing Arts, 3 vols. (Toronto: Queen's Printer 1970); Report of the Commission of Inquiry on Health and Social Welfare, 1 vols. (Quebec City: Quebec Official Publisher 1970; Manitoba Department of Health and Social Development White Paper on Health Policy (Winnipeg: Queen's Printer 1972); Health Security for British Columbians (Faulkes Report) (Victoria: Queen's Printer 1973); H.R. Robertson, Health Care in Canada: A Commentary (Ottawa: Information Canada 1973); Report of the Health Planning Task Force (Toronto: Queen's Printer 1974). 59 Lalonde, A New Perspective. 60 Geoffrey R. Weller, The Delivery of Health Services in the Canadian North,' Journal of Canadian Studies 16, no. 2 (Summer 1981), 69-80.

11 /The Policy Consequences of Northern Development SIMON MCINNES

This study focuses on Canada's policies regarding northern development in the Northwest Territories (NWT).1 It will be demonstrated that the dominant concern in northern development is resource exploration and exploitation as it benefits Canada as a whole. Northern development policy in this regard is fairly coherent and consistent. However, the government has two other objectives: enhancing Northerners' lives and protecting the environment. The consequence is that these two objectives are simply irreconcilable with the interests of Canadians as a whole. Although the Inuit Northerners have not been entirely successful in altering government and resource plans in the past decade, it will be argued that the growing maturity of northern political institutions and Ottawa's increased sensitivity to aboriginal concerns could bring about a reconciliation of Northerners' and national goals. This essay approaches the study of public policy by comparing policy impacts. My concern is to describe the consequences of general policies on political behaviour, public institutions, and specific policy decisions. In considering the main issues of policy impact studies, a number of assumptions have to be made. First, because the policy process is continuous, the analyst must decide where to begin his analysis. Here, the analysis will begin with the three policy packages that formed the framework for northern development in the 1970s. Second, the implementation of any policy will have both expected and unexpected effects, welcome and unwelcome results. No policy, however carefully designed, will perform exactly as its makers intended. Where several contradictory policies speak to the same policy area, the consequences of policy-making will be even more difficult to control, making the impact of a single policy far from predictable. The third assumption is that the impacts of policy vary in their

248/S. Mclnnes

Figure 1. Policy impact studies

importance. The policy analyst is more interested in the first-order intentional impacts (see figure 1) than in third-order unintended consequences. The three cases selected for study here are major examples of NWT resource activity. The first, offshore drilling for oil and gas in the Beaufort Sea, examines the activities of Dome Petroleum, which continue to be vitally important in light of Canada's need for energy and the National Energy Policy announced in 1980 by Marc Lalonde, the minister of energy, mines and resources.2 The second case is lead-zinc mining on north Baffin Island. Nanisivik Mines was the only operating mine north of the Arctic Circle in the 1970s. Its performance sets a precedent for other mines, such as the Cominco mine on Little Cornwallis Island, which commenced operations in 1982. The third case, uranium explora-

Northern Development / 249

tion in the Keewatin, was the subject of a significant court case brought by the Inuit against the federal government. Its outcome may have a first-order impact in setting the pace for exploration throughout the 1980s. All three cases examine the interface of government, industry, and the majority local population, the Inuit (some 17,500 in the N WT). This essay will examine Inuit response to northern development, and how it alters Inuit behaviour and northern political structures. The Major Governmental Actor: The Ottawa Hydra

The federal government has several departments and agencies involved in northern development policies. The major Ottawa actors include the Department of Indian Affairs and Northern Development (DIAND); the Department of Energy, Mines and Resources (EMR); the Department of the Environment (DOE); the Department of Fisheries and Oceans; External Affairs; the Ministry of Transport; the Department of Employment and Immigration; the Ministry of State for Economic and Regional Development; the Ministry of State for Social Development; the Department of Industry, Trade and Commerce (ITC); the National Energy Board; the Treasury Board; and the newest actor, the Canada Oil and Gas Lands Administration. Of these, DIAND remains Ottawa's most visible agent in the Northwest Territories. DIAND's chief responsibilities incompatibly include protecting aboriginal rights and ensuring adequate access to resources to meet Canada's needs. The bureaucratic politics of interdepartmental struggles for influence in northern policy development will not be examined in detail here. However, it must be remembered that there are functional overlaps among the mandates of each department and that the northern priorities of the weaker departments may be rearranged to suit those of the stronger.3 Northern policy planning in the 1970s was governed by three major policy packages, the most important being Canada's North: 1970-1980.4 Policy goals included enhancing the life of Northerners in a way compatible with their own preferences and aspirations, protecting the environment, and encouraging economic development to benefit Canadians generally. These goals may be viewed as potentially mutually conflicting. The second policy package governed oceans policy and the need to ensure Canadian sovereignty over the offshore waters of the Canadian Arctic archipelago.5 The third package was the federal government's commitment to double frontier exploration and development between 1976 and 1979 and to reduce Canadian dependency on foreign oil.6 While Canada's North: 1970-1980 was the most comprehensive of the three, each department or agency chose the one it preferred. The case studies show that the objective of economic development has been paramount.

250/S. Mclnnes Political Development in the Northwest Territories A knowledge of institutional developments in the Northwest Territories is essential to understanding the northern political economy. In the early 1960s the government of the Northwest Territories (GNWT) was a branch of the federal government. Territorial council members were appointed or elected, and the deputy minister of northern affairs and national resources (based in Ottawa) served as the commissioner, the chief territorial public servant. The 1966 Carrothers Report7 resulted in the establishment of the GNWT in Yellowknife in 1967, the appointment of a full-time resident commissioner, Stuart Hodgson, and the growth of territorial government departments. In 1977 Prime Minister Pierre Trudeau appointed C.M. Drury as a special representative to examine constitutional development in the Northwest Territories. Drury recommended regional governments as a means of bringing the territories a greater degree of responsible government.8 This recommendation conflicts with the aim of the Inuit to divide the NWT into eastern and western territories.9 The federal government had earlier approved the expansion of the territorial council to twenty-two members. Ten of the members elected in 1979 are from the Inuit areas in the Western, Central, and Eastern Arctic. Only four or five are from primarily non-aboriginal areas in the Mackenzie Valley. Stuart Hodgson was replaced by John Parker, who has given the executive and council far more say in running the government. Since 1979, the council has increasingly demonstrated the independence of a provincial assembly, which has provoked an increasing policy distance from Ottawa. For instance, the council supported holding a plebiscite in April 1982 on the proposed division of the Northwest Territories. Overall, 55 per cent supported division; in the eastern territories the figure was over 80 per cent. John Munro, minister of Indian affairs and northern development, declared that the plebiscite was merely advisory. However, the results pressured Ottawa to take northerners' political aspirations seriously. Second, the ninth council has actively supported settling aboriginal claims. This has resulted in vastly improved relations between the GNWT and the Inuit and Dene-Metis. However, if the federal government views the 1975 James Bay and Northern Quebec Agreement with the Cree and Inuit of Quebec as the paragon of claims settlements, the greater expectations of the aboriginal peoples of the Northwest Territories may not be met. The GNWT has disagreed with Ottawa on resource development; it considered challenging the National Energy Board's 1981 approval of the oil pipeline from Norman Wells down the Mackenzie Valley to Alberta.10 But because the GNWT lacks sovereignty, any legal action would, in effect, be the federal government opposing itself. In 1981 the GNWT also opposed bill C-48, which regulates oil and

Northern Development / 251

gas developments in the Northwest Territories and the Canadian offshore, stating that its land 'rights' were not taken into consideration, and that it supported Drury's recommendation that the GNWT be given ownership of lands and natural resources.11 However, the GNWT ultimately lacks the power to frustrate the plans of the federal government. Despite the importance of the GNWT today, the federal government presence is still significant. This situation will likely continue as long as the area north of 60° remains a territory. Beaufort Sea Offshore Drilling

The case of offshore drilling in the Beaufort sea illustrates the impact of oil on the permanent residents of the western Arctic. The part of the Beaufort Sea involved is that free of permanent ice from July to the end of September, lying north of the Mackenzie Delta and the north slope of the Yukon. The exact quantity of recoverable hydrocarbons is uncertain, but estimates of gas reserves range from 50 trillion to 112 trillion cubic feet, while estimates of oil reserves range from 5 billion to 38 billion barrels.12 The major problems are extracting the hydrocarbons without damaging the environment and shipping them to southern markets. The severity of Arctic conditions imposes limitations on traditional solutions. Drilling in the Beaufort Sea is difficult because of the short season and drifting ice. The damage caused by oil spills or blow-outs in the Arctic would last far longer than in warmer climates, and recovery might take as long as a decade.13 The low productivity of the tundra and the inland and offshore waters prevents large-scale commercial harvesting.14 Hydrocarbon accidents would severely hamper reproduction of marine life. The Inuvialuit of the western Arctic, some 2,500 Inuit living in six communities, have traditionally depended on the land and water for food and clothing. Fur-trading provides cash to buy nontraditional goods and services. While increasing numbers of Inuvialuit are wage-earners, most hunt part-time, supplementing their diet and income. Depriving them of the traditional products of land and water, the basis of their culture, would have incalculable consequences. The Inuvialuit view any resource activity with deep concern. Interest in the Beaufort area grew dramatically after the 1968 discovery of oil at Prudhoe Bay on the north slope of Alaska. The federal government formed a committee in February 1973 to consider the environmental aspects of offshore drilling. In September, Jean Chretien, minister of Indian affairs and northern development, approved in principle drilling by Hunt International commencing in the 1975 season. The Committee for Original Peoples' Entitlement (COPE), representing the Inuvialuit, called for full consultation before drilling.15 Ottawa

252/S. Mclnnes

announced in March 1974 a delay of drilling until 1976. As Chretien wrote to Sam Raddi, president of COPE, 'Any Drilling Authority issued will be conditional on a satisfactory demonstration that the risks associated with drilling a well from both a technical and environmental point of view are minimal.'16 In April 1974 Chretien and Jack Davis, minister of the environment, announced that the Arctic Petroleum Operators Association would put up $4.5 million for twenty-one studies, in addition to studies proposed by DOE and EMR. COPE contended that no studies had yet been completed on the consequences of a blow-out, and insufficient recognition had been given to the high-pressure zones under the sea bottom that increased the chances of shallow blow-out.17 A DOE study admitted that the ecological effects of an oil blow-out could be severe, and acknowledged that no technology existed for drilling a relief well in the event of a blow-out in the Beaufort Sea transition zone (except during the brief summer open-water period).18 In January 1976 COPE asked that the question of Beaufort drilling be submitted to an Environmental Assessment Review Panel hearing, with the results to be submitted to the Inuit communities for consideration, and that the government thoroughly brief the communities before seeking their opinion. In February the Arctic Waters Oil and Gas Advisory Committee (AWOGAC) made a hasty tour of the Beaufort communities to reveal government environmental plans. COPE complained that the 'AWOGAC oil spill contingency plan... affords virtually no protection for the marine and shore line environments.'19 Dome Petroleum argued in favour of proceeding with drilling in 1976. First, Canada as a net importer of oil needed new domestic supplies. Second, the lead time from drilling to marketing oil is seven to fifteen years, and if self-sufficiency was to be achieved by 1985, a start would have to be made in 1976. Third, Dome was prepared to invest $400 million over the next five years. Fourth, approval in principle had been granted in 1974, and Dome hinted that this obliged Ottawa to grant final approval. Fifth, according to Dome, the Canadian government had exaggerated the environmental impact of a blow-out. Sixth, without Dome's efforts, there would be increased tanker traffic bringing imported oil to the east and west coasts (a case of Dome using the danger of one environmental risk to sanction another). On balance, concluded Dome, the Beaufort Sea is 'no more hostile... than many other offshore areas of the world; notably the East Coast of Canada, the Labrador Coast or the North Sea.'20 On 15 April 1976 cabinet announced its approval of Beaufort drilling, saying that 'the risks are low enough to be acceptable and, balanced with the need to conform to Canada's energy resources, justify proceeding with drilling this year.'21 Dome was required to post a $50 million bond to cover oil-spill costs, and was restricted to drilling between 15 July and 15 September, allowing a month at the end of the season as 'blow-out' insurance. The decision was

Northern Development / 253

opposed by the Canadian Wildlife Federation and the Anglican Church of Canada because of concern for the environment and the traditional Inuit way of life.22 The national Inuit association, the Inuit Tapirisat of Canada (ITC), criticized the decision and the inadequate safety provisions. Judd Buchanan, now minister of Indian affairs and northern development, hastened to reassure ITC President James Arvaluk that 'while there may still be certain deficiencies, the most up-to-date technology has been applied...in the counter-measures.'23 The opening season was an ostensible success. Canmar, Dome's subsidiary, employed 127 northerners, of whom 113 were aboriginal, in a work force of 400.24 The major economic impact was felt in Tuktoyaktuk, where Canmar had its shore facilities and winter berths for its drillships. While no major ecological disaster occurred, the Canadian Arctic Resources Committee (CARC) accused DIAND and Dome of covering up the fact that the existing drilling technology was inadequate. CARC pointed out that none of the wells drilled had reached target depth, and that CARC had evidence of water and gas blow-outs encountered as drills hit unexpectedly high-pressure zones of sub-sea water and gas.25 Gordon Harrison, president of Canmar, countered that the quantity of gas involved had been very limited. He also downplayed the environmental concerns by revealing that 'ice itself... actually is an excellent containment device for oil.'26 Amid Dome's proposals to begin drilling eleven holes in 1977, a secret government report placed blow-out chances at one in 300, not one in 15,000 to 20,000 as the oil industry had calculated.27 In May 1977, Ottawa approved the resumption of drilling by Dome. In July DIAND announced a $7 million government program, matched by industry, to improve oil spill clean-up technology.28 Industry could well afford to pay its share. Of a yearly budget of $120 million for its Beaufort Sea activities, Dome's after-tax costs were calculated to be $12 million a year for the first three years, costs that markedly reduced the company's financial risk.29 At the end of the season Dome reported major gas finds30 and 1,150 barrels of oil per day at its Nektoralik well.31 The good showing of the 1977 season, owing to an absence of major technical problems, encouraged the federal government to extend the 1978 drilling season to mid-October, and the shallow-water drilling season until December.32 Throughout the 1970s, doubts had arisen regarding the federal enforcement of environmental regulations. In 1980, CARC concluded that there had been no improvement in the regulation of development activities in the Beaufort since 1974. Commenting on a Dome proposal for an oil-refuelling farm for its summer operations, CARC indicated that 'advice from government experts calling for full review of the environmental consequences of Dome's proposal was circumvented to allow Dome to proceed with all due speed.'33 Pressure grew to expand oil and gas exploration in the western Arctic because of Dome's successes; but in

254/S. Mclnnes May 1978 the federal government deferred the issuing of new exploration permits on crown reserves in the western Arctic for one year to aid the settlement of land claims being put forward by COPE. COPE was under tremendous pressure because of the intense Beaufort Sea activity, and was anxious to achieve a land-claims settlement before the Inuit way of life was seriously disrupted. In October 1978 COPE and Ottawa signed an agreement in principle, the first to be reached in the NWT.34 Under the agreement COPE would receive surface rights to 96,000 square kilometres, special hunting, fishing, and trapping rights, a $45 million cash settlement (in 1978 dollars), business development grants, and social development programs. The agreement in principle was roundly criticized by ITC for extinguishing all aboriginal rights.35 Others argued that the Inuvialuit were powerless to protect their way of life because no mechanisms were provided in the agreement to prevent them from being engulfed by the development industry.36 Two reasons were cited in 1979 by Jake Epp, minister of Indian affairs and northern development in the Clark government, for refusing to sign the final agreement.37 First, the Yukon government did not like the provisions giving the Inuvialuit hunting rights across the north slope of the territory; it was felt that these jeopardized the government's land rights.38 Second, the eastern boundary described in the COPE agreement overlapped the western boundary of the land traditionally used and occupied by the Inuit of the central Arctic. A further problem was that the Dene-Metis also used some of the Inuvialuit land. By 1980 two additional reasons for refusing to sign the COPE agreement arose. The Yukon objected to the COPE agreement because it could mean that oil and gas developments in the Beaufort Sea would benefit COPE if COPE was successful in reaching its aim of receiving 3 per cent royalties. In addition, the Inuit of the western Arctic would have first access to employment opportunities.39 Second, the federal government was itself having second thoughts. In a letter dated 24 December 1980, John Munro, Epp's successor, pointed out to Senator Davie Steuart, the federal negotiator, that cabinet would support COPE only if the Inuit agreed to the creation of a wilderness park on the north slope of the Yukon. The government also wished to establish a transportation corridor and onshore drilling facilities.40 Some progress was achieved with the passage of bill c-48 in 1981. The government agreed that the 25 per cent share reserved to the crown of oil and gas rights in Canada lands would be transferred to the Inuvialuit homelands designated in the final agreement.41 Further, Marc Lalonde, minister of energy, mines, and resources, vowed not to take any action that would prejudice land claims.42 By mid-1982, however, the impasse on the COPE agreement in principle remained. Industry will continue to drill actively, and oil production is likely to occur by the early 1990s. This leaves some time to settle the dispute between COPE and the federal government.

Northern Development / 255

Of the three policy packages defining the framework of northern policy planning, the energy package has had a first-order impact on the Beaufort and has been reinforced by the 1980 National Energy Program. The difficulties associated with east-coast drilling and with financing oil-sands plants puts even greater pressure on the Beaufort in the 1980s, to the advantage of companies involved there. However, an unintended consequence of northern policy is the threat imposed by industry ambitions on Inuvialuit interests. As CARC wrote, 'the highly competitive individualized, profit-motivated assumptions running through much of Dome's [proposals for development] may be the leading edge of exactly what will cut the native people off from whatever benefits southern [type] development may bring.'43 The impasse faced by COPE has not yet encouraged it to take drastic political action. COPE still hopes an agreement can be reached with Ottawa. However, a delay in signing does not prevent resource exploration from continuing and thus threatening the Inuvialuit aboriginal rights. As long as no agreement is reached, the federal government cannot be entirely sure that the courts will not find some government actions infringing those rights. This is perhaps the most important first-order impact that the recognition and affirmation of aboriginal and treaty rights contained in part II of the new Canadian constitution may have on northern policy. It is difficult to calculate whether delaying the signing of an agreement in principle is in the best interests of the Inuvialuit; as long as the residents of the western Arctic lack the means to have a first-order impact on the economic development taking place at their doorstep, the developers may continue to ignore their interests. Baffin Island Lead-Zinc Mining

The next example to be studied here is the second Arctic mining venture above the tree-line. As in the Beaufort Sea case, industrial and governmental interests governed decision-making in this instance. Nanisivik Mines on Strathcona Sound is located at the northern tip of Baffin Island. The project was announced in June 1974 by Jean Chretien. An earlier mine had operated from 1957 to 1962 in Rankin Inlet on the west coast of Hudson Bay, where climatic conditions and social and environmental concerns were fairly similar to those of Baffin Island. The mine's closing had been a disaster locally. Its demise enabled the Inuit to comprehend the pitfalls of trading a hunting life for short-term participation in a wage economy. The Texas Gulf Sulphur Company had been undertaking geological surveys of the area from the early 1960s. In 1972 the company decided to exchange its Strathcona leases with Calgary-based Mineral Resources International (MRI). A feasibility study showed that an ore body of 6,000,000 tons contained 14 per cent

256/S. Mclnnes

zinc, 1.4 per cent lead, and two ounces of silver per ton.44 The plan was to mine the ore, to ship lead, zinc, and silver concentrates to Germany during the ice-free months, and to dump the tailings in Strathcona Sound. The project would have major environmental consequences for five north Baffin communities (the nearest, Arctic Bay, was a scant seventeen miles away), but it could generate substantial local and national economic benefits. DIAND persuaded MRI to extend the life of the mine from the most economical eight years to the less economical but still feasible fifteen years in order to encourage more Inuit participation and to spread wage-earning over a longer period.45 The Council of Arctic Bay wanted Nanisivik to be a bunkhouse operation, to protect their community from the influence of a southern-dominated permanent settlement. They also wanted to protect local marine life, particularly seals, which were an important source of food. Assured that no townsite was planned, the council did not object to the mine at the outset.46 Arctic Bay workers could even commute to work daily on the road linking the community with the project. However, by 1974, the infrastructure package included an all-season class A airstrip, a wharf, and a permanent townsite. The cheapest way to handle toxic mine tailings was to dump them into the ocean. However, toxins could enter the food chain, increase in concentration, and affect the seals and narwhals that were an essential source of food for the Inuit. ITC argued that, in direct contravention of an internal government directive of 20 December 1973, no prior environmental assessment had been conducted. Further, ITC argued that the tailings would get into the food chain regardless of where they were dumped.47 Judd Buchanan, the minister of Indian affairs and northern development, pointed out that if MRI did not conduct any studies on the effect of ocean-dumped tailings, the tailings would have to be contained in tailings ponds on land.48 But as an ITC study suggested, 'it is a technological nightmare to get such ponds to work even during the operation of the mine, let alone to have them stay intact for infinite time afterwards... it sometimes makes sense to sacrifice one lake and fill it with tailings, if no better site is available.'49 In November 1975 the federal government agreed that ocean dumping was out of the question. Strathcona Minerals, the mine operator, agreed with ITC that the only feasible alternative to sea disposal was to use one of two nearby lakes, West Twin Lake.50 The Inuit thus were successful in persuading the mine to choose an environmentally safer (though more costly) method of tailings disposal. The project required considerable federal government support; $16.7 million was given in the form of grants and loans, and by June 1974, when Nanisivik Mines was established, Ottawa held an 18 per cent equity, in the project.51 The economic viability of the mine was cause for concern. MRI assumed that leadzinc prices would remain above 22.8 cents per pound, and that government

Northern Development / 257

would provide assistance to ensure the project's success.52 MRI concluded that the cheaper costs associated with local labour would, along with an absence of environmental problems, make the Nanisivik project economically viable. The Department of Indian Affairs and Northern Development approved the mine, although the two European partners of Nanisivik were to receive 80 per cent of the zinc and 100 per cent of the lead. The agreement between Ottawa and the Nanisivik consortium contained a weak clause specifying usage of Canadian shipping where feasible. EMR pointed out that the supposed benefits of the mine could not automatically be guaranteed.53 The costs of the project ($18.3 million in infrastructure assistance less $ 10.6 million recoverable, plus $700,000 yearly in training costs and worker relocation)54 were weighed against the benefits (wage employment, tax and royalty revenues, and experience for future ventures). But some costs were not measured in detail: the environmental consequences, effect on Arctic Bay, and the loss of jobs and taxes through the export of secondary processing. Rapid approval of the project was encouraged by the advent of 1974 election; the European partners wanted government assurances by the end of June, and Inuit opposition to the project was growing. Environmental concerns were not unfounded. In 1976 a federal report found that only 33 per cent of the conditions placed on the mine's construction had been met: 'The mine's attention to the conditions of the water licence and to general environmental matters was unacceptable.'55 J.B. Sprague Associates warned ITC that by 1982-3 the tailings disposal should be monitored very closely, as the condition of the retaining walls and overflow problems were cause for anxiety.56 By 1978 the target of a 60 per cent-Inuit work-force had not been met (only 20 per cent of the workers were Inuit); this negated the supposed benefit of increased local employment. The consistent failure of Nanisivik in this respect resulted in the omission of aboriginal employment targets for Canada's most northerly mine, Polaris, on Little Cornwallis Island, which commenced operations in 1982. The first-order unintentional consequences of the environmental, economic, and social costs and benefits of the Nanisivik project cannot be assessed until long after the mine has ceased to operate. However, the first-order intentional impact was that the need for large-scale economic development overcame both the need to assess thoroughly the possible environmental consequences and the views of local communities. A second-order intentional impact was the extent to which the project failed to maximize economic benefits to Canada. Keewatin Uranium Exploration and the Baker Lake Case

In the Baker Lake case the federal government defended the interests of foreign uranium companies against the interests of Canadian citizens. The area involved

258/S. Mclnnes

is about 70,000 square kilometres near the town of Baker Lake, some 1,600 kilometres north of Winnipeg. Its people are the only Inuit not living within 100 kilometres of the coast; their main source of food is caribou. Two caribou herds frequent the Baker Lake area: the Beverly herd, some 120,000 strong, roams from Baker Lake to northwest Saskatchewan, and the Kaminuriak herd, numbering 60,000 animals, ranges south to points due west of Churchill, Manitoba. Caribou account for about 30 per cent of real income of heads of households in Baker Lake.57 The availbility of caribou thus affects both the nourishment and the pocketbooks of the people. Uranium, which had suffered a slump in Canada beginning in the mid-1950s, suddenly became an important alternative fuel source following the 1973 OPEC oil embargo. The Baker Lake Hamlet Council feared that exploration activities would disturb the caribou calving-grounds and water-crossings. In April 1974 ITC sent a petition to Ottawa and the territorial government in Yellowknife requesting that ITC and Baker Lake be kept informed of future exploration activities. In June 1975 the Hunters and Trappers Association and the hamlet council proposed a land freeze to prevent the damaging exploration activity. In August, Judd Buchanan proposed a land-management zone with local input from Baker Lake to monitor all exploration activities. However, the Inuit wanted control of land use. In their view there was insufficient knowledge of the migratory and reproductive patterns of caribou. At the same time DIAND was funding an extensive Inuit land-use and occupancy study to determine the extent of Inuit aboriginal use of the land, to serve as the factual basis for land-claims negotiations. This study had yet to be completed.58 Warren Allmand became minister of Indian affairs and northern development in 1976. He responded to the concerns of the Baker Lake residents by imposing a one-year land freeze in March 1977 to permit a special $ 100,000 study of the impact of exploration on the caribou herds.59 This freeze was greeted with howls from the mining industry. The Chambers of Mines of the Northwest Territories, the Yukon, Alberta, and British Columbia asked Prime Minister Trudeau to reverse Allmand's decision.60 DIAND officials defended the land freeze on the ground that the caribou study was needed to verify the allegations of the residents of Baker Lake.61 The mining interests circumented the freeze by moving their exploration camps before they reached the limit of 100 person-days at a single location (beyond which land-use permits were required). The new minister, Hugh Faulkner, said that 'while this is contrary to the spirit and intent of the Land Use Regulations, it is not illegal.'62 As 1977 drew to a close, the Hamlet Council of Baker Lake requested an extension of the land freeze. Faulkner rejected this on the ground that 'successive stages of exploration, development, and production commonly take from 10 to 30 years, and consequently exploration must be encouraged now.'63 The

Northern Development / 259

hamlet also requested Urangesellschaft Canada Ltd, a German-owned company, 'to refrain from working on our land until land claims are negotiated by the federal government'.64 Because the land freeze was to expire at the end of March 1978, the uranium companies planned extensive summer exploration. Essex Minerals, Urangesellschaft, and Western Mines, among others, were to map and take rock samplings, and to carry out aerial surveys. These activities, in the eyes of Baker Lake residents, would have frightened the calving caribou away from the region, affecting Inuit hunting. On 8 March, 87 per cent of the voting population of Baker Lake petitioned Prime Minister Trudeau, asking that exploration await the settlement of land claims and the approval of the Baker Lake community. ITC lawyers advised the community that a court action to extend the freeze ought to commence at once, before the current freeze expired. Hugh Faulkner was asked to attend a community meeting in Baker Lake on 7 April 1978 to hear the residents' views. Pending the outcome of this meeting, the 1977 land freeze was extended to 14 April.65 At the meeting, Faulkner repeated the August 1975 offer of the land-management system. This proposal was rejected as inadequate.66 On 18 April ITC announced legal action to prevent any further exploration on a portion of the Baker Lake lands. DIAND deferred prospecting and land-use permits near Baker Lake until 24 April to consider the ITC application for an injunction. On that day the federal court granted an interlocutory injunction pending consideration of the action for a permanent injunction. Companies had to cease activities within 4.8 miles of caribou watercrossings. The government announced its intention to comply. Western Mines asked why the government did not do something about the overhunting of caribou instead of 'coming down on the miners.'67 Western Mines also suggested to ITC that Western was in a position to persuade foreign companies of the need to maintain good communications with the residents of Baker Lake, and that exploration would create local opportunities.68 The 1978 summer exploration season was a success. Urangesellschaft reported finds in the Baker Lake area of twenty pounds per ton of uranium oxide, as good as the Wollaston Lake find in northern Saskatchewan.69 The prospects for further exploration in 1979 and beyond looked extremely favourable. The Baker Lake community faced a difficult situation. The 1977 freeze had expired in 1978, even though the government-commissioned caribou study had recommended controls to restrict the location of exploration facilities, the flying of aircraft, the timing of activities, and the possible outright prohibition of industrial facilities in calving areas.70 The court action of 18 April 1978 to seek a permanent injunction would not necessarily produce an outcome favourable to Baker Lake. The possibility of an out-of-court settlement was explored with DIAND in April 1979. Baker Lake proposed, first, that an Inuit advisory committee to

260/S. Mclnnes govern land use in the caribou calving season be created. Second, the same committee should govern campsites in areas of spring migration. Third, no mine, mill, or other semi-permanent activity was to be located within twenty-five miles of water-crossing areas. Fourth, the companies should match the dollars spent in exploration in calving areas by spending equal amounts in exploration outside those areas. Fifth, and most important, a full-scale socio-economic and environmental enquiry should be conducted into the impact of uranium mining in the Baker Lake area. Sixth, the Baker Lake residents requested that the mining regulations be amended to prohibit the issuance of mining leases in the calving areas.71 These proposals were unacceptable to the government, and the Inuit proceeded with court action to obtain a permanent injunction. The Baker Lake court case, as it came to be called, was to join those few Canadian cases that have spoken to the issue of aboriginal rights. The hamlet council, the local Hunters and Trappers Association, 113 of the 1,000 residents of Baker Lake, and ITC sought a permanent injunction prohibiting exploration on 58,000 square kilometres. The defendant was the government of Canada; Urangesellschaft, Pan Ocean Oil, Essex Mining, Noranda Exploration, Cominco, and Western Mines were interveners. Counsel for the Inuit argued that the activities of mining companies were illegally violating aboriginal rights. Inuit lawyers sought to demonstrate exclusive Inuit use of the area since earliest times; they argued that use implied possession of the lands, that possession was a sufficient demonstration of aboriginal rights, and that no law had abrogated those rights.72 The government lawyers contended that the Inuit had no basis to claim aboriginal rights; those had been extinguished by federal and provincial land laws. Finally, they argued that interference with the caribou was not relevant to the case.73 The strategy of the Department of Justice contradicted the efforts of DIAND to achieve landclaims settlements in those areas of Canada where none existed. If there were no aboriginal rights, it might be asked, how could one explain the government's anxiety to 'extinguish' aboriginal rights through land-claims negotiations? The arguments of industry were more straightforward. A permanent injunction would block all mining in the Northwest Territories. The Hudson's Bay Company charter gave the company absolute control of Rupert's Land in 1670, and Rupert's Land was ceded to Canada in 1870. According to legal counsel for Pan Ocean Oil, Cominco, and Western Mines, that charter gave the Hudson's Bay Company both mineral and fishing rights. They also argued that hunting was now only a part-time occupation, since Inuit were increasingly entering the wage economy. The case was heard in Baker Lake from 14 to 18 May and in Toronto from 28 May to 1 June 1979.75 The Inuit argued that caribou hunting was a major part of the socio-economic and cultural fabric of Baker Lake society. The mining

Northern Development / 261

companies' witnesses presented evidence that exploration activities had no permanent impact on the caribou herds. Mr Justice Mahoney's decision, delivered on 15 November, held that the Inuit had had exclusive use of the area since earliest times. He found that aboriginal rights did exist, and that laws had never extinguished those rights in the Baker lake area. But he also found that there was little proof that the ground operations of the companies had harassed caribou. While Mr Justice Mahoney did agree that low-flying aircraft were a harassment, he did not find sufficient evidence to demonstrate that aboriginal rights had been interfered with. Aboriginal rights were not property rights, because those had been given by Charles n to his German cousin Rupert. Aboriginal rights were the rights to hunt, fish, trap, and move about the land. These may have been interfered with, but the Inuit had not sought damages. In short, while the court found that the Inuit had aboriginal rights, there was insufficient evidence that these had been interfered with by mining companies' activities, and that the decline in the population of the herds was caused by the companies' exploration. All parties tried to interpret the judgment as a victory. The Inuit were pleased that aboriginal rights were recognized.76 In the government's view, 'a company's right to explore under a mining permit takes precedence over the aboriginal rights to hunt and fish',77 and 'aboriginal rights... do not override parliamentary legislation.'78 Industry was delighted with the decision. The interim injunction was lifted on 17 December 1979, and over the winter months the mining companies flew in supplies and set up camps for the 1980 exploration season, which brought further findings so encouraging that Oilweek predicted that an operating mine would be in existence by 1985.79 The Inuit were awarded costs, which meant a payment of about $29,000 from the federal government.80 The territorial councillors unanimously agreed to donate $30,000 to the Inuit.81 But the case had cost the Inuit over $120,000,82 and the costs to the defendants and interveners cannot have been less. Given the inconclusiveness of the decision, neither the defendants nor ITC appealed. ITC placed its hopes on an out-of-court settlement, but did not rule out future court action.83 ITC proposed to DIAND in January 1980 that an agreement be developed to control access to the Baker Lake lands, but this was not warmly received.84 In March, ITC again put forward a multi-point plan similar to the April 1979 proposal asking for protection of calving areas and water-crossings, and for cutbacks in the activities of biologists.85 In a novel move, ITC placed an advertisement in leading newspapers-a 'Notice to Developers'- inviting the industry to discuss projects with ITC in order to avoid the difficulties encountered by the uranium companies in Baker Lake, and stressing Mr Justice Mahoney's recognition of aboriginal rights. Legal counsel for Cominco and Pan Ocean Oil replied that the decision had clearly given precedence to exploration over aboriginal rights.86

262/S. Mclnnes

The Baker Lake case demonstrates the pre-eminence of economic development over the interests of local residents and environmental concerns. Perhaps the most important unintended impact of northern development demonstrated here was to mobilize the Inuit effectively, and ultimately to enlist indirectly the support of the territorial government against Ottawa. Another unintended consequnce was the realization that the resolution in the courts of problems arising between developers and aboriginal peoples is an expensive process, and one that in the long run may do more to diminish than to affirm aboriginal rights. Because it is in the best interests of government, developers, and aboriginal peoples to avoid expensive disputes, the ultimate consequence of the Baker Lake case may be to encourage negotiation rather than litigation. Policy Impacts: The Value of the Analysis

This study has shown that northern resource development has reinforced rather than questioned the dominant position of primary resource extraction. This predominance has been a first-order impact of northern development. Government intended that a second-order consequence of northern development policy would be the finding and extracting of resources, ideally to the benefit (yet to be measured) of all Canadians. Government also intended to allay enviromental concerns and the lives of Northerners. These objectives, perhaps incompatible with the goal of resource development, have not been met adequately. The need to satisfy Canada's demand for energy increasingly dominates decisions affecting the North, and through the 1980s satisfying that demand will be the primary intended consequence of northern development. Several unintended consequences have resulted from northern development. In the first order, resource developments politically awakened aboriginal Northerners. The awakening was aided by the advent of southern-type education for the Inuit, creating a new class of Inuit able to deal with and respond to government decision-making. The Inuit, like all Northerners, were influenced by the Berger inquiry,87 and additionally had the example of Indian political mobilization. The Inuit knew how to respond when resource developments threatened their way of life. However, because the Inuit do not control any power structure comparable to Ottawa's, their response has been (and will probably remain) largely unsuccessful. As long as the NWT remains a territory this is likely to be the case. A second-order unintentional impact of northern development is that the desire of Northerners for responsible government will mount as long as they believe that their needs and those of the environment are being ignored. This could encourage Ottawa to respond positively to Northerners' political aspirations, particularly in light of the conclusive results of the April 1982 plebiscite on

Northern Development / 263

division of the Northwest Territories which provide some support for those who argue that at present the territories are ungovernable. The price of gaining a separate territory above the tree-line, Nunavut, may be the federal government's insistence that it retain control and management of non-renewable resources. This may not be incompatible with responsible self-government, however. If a divided NWT has exclusive regulatory powers over matters such as land use and renewable resources and achieves a greater degree of financial self-sufficiency, local political aspirations may be largely addressed, while concomitantly leaving Ottawa room to satisfy national energy and economic goals. The contribution of this study to understanding policy-making in Canada, it is hoped, is twofold. First, it alerts analysts to the need to spend more time assessing policy consequences. The comprehensive understanding of policy consequences can be invaluable in understanding the formation of new policies. Second, in the North, as elsewhere, policies must be readjusted in light of unintended consequences. No fundamental reassessment of Canada's North: 1970-1980 has as yet taken place.88 The demand for energy and resources is not easily compatible with the needs of Northerners, particularly aboriginal peoples. The devolution of greater governmental powers to the Northwest Territories, however, could bring about the reconciliation of these presently dichotomous goals. Having already moved slowly in that direction since the GNWT was created in 1967, Ottawa has set in motion developments that now cannot easily be stopped. The Northwest Territories could very well be the setting of some interesting political and economic developments throughout the 1980s. NOTES I am grateful to Francois Bregha and reviewers of this volume for their comments on an earlier draft. Any remaining errors of interpretation are mine. The opinions herein are not necessarily those of my past or present employers. This essay was written in April 1982. 1 'Northern development' as used here refers to the political, resource, and industrial policies and activities of governments and business that transform part or all of the Northwest Territories above the tree-line. 2 Ottawa: Department of Supply and Services 1980. 3 Kim Richard Nossal, 'Allison through the Looking Glass,' Canadian Public Administration (Spring 1980). 4 Ottawa: Department of Supply and Services 1972. 5 See The Arctic in Question, ed. E.J. Dosman (Toronto: Oxford University Press 1976), 2. 6 EMR, Statement of Policy (mimeograph 1976). 7 Report of the Advisory Committee on the Development of Government in the

264/S. Mclnnes

8 9 10 11

12

13 14

15 16 17 18 19 20 21 22 23 24

25 26 27 28 29 30 31

Northwest Territories (Ottawa: Department of Supply and Services 1966). For a comprehensive review of northern politics, see Gurston Dacks, A Choice of Futures (Toronto: Methuen 1981). Report of the Special Representative (Ottawa: Department of Supply and Services 1979). Political Development in Nunavut (Igloolik: Inuit Tapirisat of Canada 1979). 'NWT Division Possible - Munro' Tapwe, 21 May 1981. Hon. Richard Nerysoo, NWT minister of energy and renewable resources, Notes for an address on bill c-48 (Standing Committee on Natural Resources and Public Works, Ottawa, 2 April 1981). Canadian Arctic Resources Committee, 'Canadian Arctic Marine Energy Projects,' presentation to the Special Committee of the Senate on the Northern Pipeline, Ottawa, 16 March 1982, 9-10. 'Offshore Drilling for Oil in the Beaufort Sea: Preliminary Environmental Assessment' (Ottawa: Department of the Environment 1975), 43. Everett B. Petersen, 'Biological Productivity of Arctic Lands and Waters' in Inuit Land Use and Occupancy Project, vol. 2 (Ottawa: Department of Supply and Services 1976), 85-100. Press release, 8 Feb. 1974. 2 April 1974. Memorandum to settlement councils, 7 June 1974. 'Offshore Drilling,'40. Letter to Judd Buchanan, 25 Feb. 1976. Dome Petroleum 'Key Points Supporting Beaufort Sea Drilling' (mimeograph 22 March 1976), 3. Globe and Mail, 16 April 1916. Press releases, 26 April and 7 May 1976 respectively. Letter, 3 June 1976. Social and Economic Impacts of Canadian Marine Drilling Limited's 1976 Operations on the Beaufort Sea Communities (Toronto: Mary Collins Consultants Ltd 1977). The report noted the rise in the incidence of liquor consumption and gambling among Tuk residents. Canadian Arctic Resources Committee, press release, 30 March 1977. Interview, 'Canada AM', 5 May 1977. Globe and Mail, 29 April 1977. Letter to ITC from Murray Hardie, special assistant to Warren Allmand, minister of Indian affairs and northern development. Vancouver Province, 13 July 1977. Financial Post, 1 Oct. 1977. Dome Petroleum, press release, 3 Oct. 1977.

Northern Development / 265 32 DIAND, press release, 4 May 1978. 33 Northern Perspectives 8, no. 2 (1980). 34 'COPE / Government Working Group Joint Position Paper on the Inuvialuit Land Rights Claim' (mimeograph May 1978). 35 'COPE'S Land Claims Proposal,' ITC (mimeograph 1978). 36 Ken Hatt, 'The COPE / Government Joint Position Paper: Blueprint for Internment' (mimeograph 9 Aug. 1978), 20. 37 Canadian Press, 15 Nov. 1979. 38 'Provinces' land rights in jeopardy if Inuit pact allowed, Yukon warns,' Edmonton Journal, 10 July 1980. 39 Ray Guay, 'Problem of land claims which cross provincial borders,' Regina Leader-Post, 23 July 1980. 40 Walter Brown, 'COPE agreement-in-principle not acceptable to Cabinet, says Munro,' News of the North, 16 Jan. 1981. 41 Senate,Debates( 15 Dec. 1981), 3456. 42 Hon. Marc Lalonde, Notes for an address to the Senate Standing Committee on Banking, Trade and Commerce (Ottawa, 9 Dec. 1981). 43 Canadian Arctic Resources Committee, 'Beaufort Sea production Overview Discussion Draft' (undated mimeograph) 15. 44 Watts, Griffis and McQuatt Ltd, 'Feasibility Study of the Strathcona Sound Project for Mineral Resources International Ltd,' 1973, 35. 45 Watts et al., 'Feasibility Study,' 47-8. 46 Correspondence, Kenn Harper, Arctic Bay, 1974 and 1975. 47 Letter to Judd Buchanan from James Arvaluk, president of ITC, 19 Feb. 1975. 48 Letter to Arvaluk from Buchanan, 19 March 1975. 49 Letter to ITC from J.B. Sprague, 9 June 1975. 50 Letter from Strathcona Minerals, 5 Dec. 1975. 51 Other investors were MRI (59.5 per cent), Metallgesellschaft AG of Germany (11.25 per cent), and Billiton BV of Holland (11.25 per cent). 52 Watts et al., 'Feasibility Study,' 326. 53 'Preliminary EMR Evaluation of Feasibility Study Prepared for MRI Ltd on Proposed Strathcona Sound Lead-Zinc project in the Arctic Islands' (Ottawa: Department of Energy, Mines and Resources 1973). 54 House of Commons, Debates, vol. 5 (5 April 1975), 4554. 55 Globe and Mail, 6 Nov. 1976. 56 J.B. Sprague Associates Ltd, 'Evaluation of Water Quality at Nanisivik Mine,' report presented to ITC, 27 Feb. 1978. 57 Interdisciplinary Systems Ltd, 'Effects of Exploration and Development in the Baker Lake Area), vol. 1 (1978), executive summary. 58 Inuit Land Use and Occupancy Project.

266/S. Mclnnes 59 DIAND, press release, 30 March 1977. 60 Jon Ferry, 'Mining freeze draws protest,' News of the North, 5 April 1977. 61 Standing Committee on Indian Affairs and Northern Development, Minutes of Proceedings and Evidence, 20 April 1977. 62 Letter to ITC from Hugh Faulkner, 28 Nov. 1977. 63 Letter to ITC from Hugh Faulkner, 16 March 1978. 64 Letter from Baker Lake Settlement Council, 16 March 1978. 65 Order in Council 1978-944. 66 Record, Baker Lake Hamlet Council Special Meeting, 7 April 1978. 67 Anne M. Sansom, 'NWT Mining industry airs gripes,' News of the North, 29 May 1978. 68 Letter to ITC from Western Mines, 10 July 1978. 69 Northern Miner, 21 Sept. 1978. 70 Interdisciplinary Systems Ltd, 'Effects of Exploration and Development,' vi. 71 Agreement proposed in Ottawa between the Baker Lake delegates and DIAND, 19 April 1979. 72 Marc Denhez, 'Baker Lake: What are Aboriginal Rights' Inuit North (1979) 57. 73 Canadian Press, 9 Aug. 1979. 74 CBC News, 10 Aug. 1979. 75 Hamlet of Baker Lake et al. v The Minister of Indian Affairs and Northern Development et al. [1979] 1 FC 487 (Trial Division). 76 Doug Earl, 'Native rights upheld in land trial,' News of the North, 21 Nov. 1979. 77 'Native rights secondary to mining,' Oilweek, 26 Nov. 1979. 78 Earl, 'Native rights upheld.' 79 Oilweek, 20 Oct. 1980. 80 Federal Court of Canada, Trial Division, Order of 18 Dec. 1979 re Baker lake et al. v. DIAND et al. 81 GNWT, press release, 16 June 1980. 82 Earl, 'Native rights upheld.' 83 News Inuit, 29 Feb. 1980. 84 Letter to DIAND from ITC, 17 Jan. 1980. 85 Letter to DIAND from ITC, 19 March 1980. 86 Letter to ITC from McCarthy and McCarthy, 17 April 1980. 87 Report of the Mackenzie Valley Pipeline Inquiry (Ottawa: Department of Supply and Services 1977). 88 In All Fairness-A Native Claims Policy (Ottawa: Department of Supply and Services 1981) reaffirms existing policy.

12 / Will Program Evaluation Be Used in Formulating Policy? JOHN MAYNE ROBERT S. MAYNE

In recent years governments have increasingly come under attack for waste, inefficiency, and poor management. This criticism has often been directed at specific programs and in particular the inability of governments to demonstrate that their programs are effective. As a result, both governments and their critics have become more aware of and interested in the empirical study of the outcomes of government programs. This increased interest in the evaluation of programs also stems, in part from a realization that, given the growth and complexity of government intervention in society, it is no longer adequate to base the assessment of past government performance solely on common sense, experience, and available administrative data.1 Complex programs require extensive investigations to find out exactly what is going on within a program and what has happened as a result of a program. These are the products of program evaluation. This essay discusses some of the ways in which program evaluation might play a useful role in government decision-making in the future. In the United States, program evaluation has, in the last ten years, become somewhat of a growth industry.2 In certain states, evaluation has played a legitimizing role for programs through the use of 'sunset laws,' whereby programs are terminated unless an evaluation shows them to be effective and new enabling legislation is passed.3 In Canada, there has been a growing recognition of the need for more and better evaluation of government programs. Recent reports on the management of government at the federal level, including the 1976, 1977, and 1978 reports of the auditor-general4 and the 1979 report of the Royal Commission on Management and Accountability5 have called for more evaluation of government programs. The need for more extensive evaluation of programs has been reinforced by the choices forced on politicians as the result of

268 /J. Mayne and R.S. Mayne continuing restraint in government spending.6 Restraint implies that trade-offs among existing programs must be made; it becomes even more important, therefore, to know what has been achieved by existing programs. In addition, the short-lived Clark government had intended, in late 1979, to introduce a Canadian version of a sunset law, which would have enabled Parliament to scrutinize better the efficiency and effectiveness of government programs, in large part through program evaluation. 7 The Liberals have responded by attempting to make program evaluation an integral part of the management of the federal government. 8 Evaluation studies had been carried out on specific federal programs for some time, but on an ad hoc basis and with no systematic attempt to ensure that findings were considered in formulating policy. In 1978, the first comptroller-general of Canada was appointed, and was given broad authority and responsibility in the area of program evaluation. It is expected that by the end of 1982, most federal departments and agencies will have a program evaluation capability. In addition, the newly established cabinet committee system for policy and expenditure management includes as an important element a requirement for the systematic reporting of evaluations of government programs. 9 But even given this strong interest on the part of governments and at least some of their critics, can we expect that program evaluation will have a significant influence on future government decisions? Or will it, like other techniques for improving management in government, such as management by objectives or planning, programming, and budgeting systems, experience a brief popularity among the powerful interests in the bureaucracy and then be quickly relegated to a minor supporting role? The answer depends in part on how one defines 'significant influence.' Here the term 'influence1 is used to refer to the direct impact that a particular study makes on a specific policy decision. While focusing on this type of direct influence, we also recognize that other types of indirect influence may be of interest and importance. The question of the impact of social science research on policy decisions has been addressed frequently in the literature, and a number of indirect ways in which an empirical study can influence policy decisions have been identified.10 For example, it has been argued that, over time, the accumulated evidence from empirical studies results in an indirect but pervasive influence on government decisions as the knowledge gained becomes more generally accepted by the public and various decision-makers." Others point out that empirical studies can also be used as a tactical tool in an advocacy role, or as a legitimizing mechanism after tactical tool in an advocacy role, or as a legitimizing mechanism after the decision has been made. Program evaluation, in this last view, contributes to the implementation of a policy decision by providing

Program Evaluation / 269 'scientific' evidence confirming the decision to others who must be brought on board in order for the decision to take effect.12 This essay will focus on the direct impact that an evaluation study can have on specific policy decisions. It is this direct type of influence that is usually assumed when there is a call for more analysis in the public sector.13 The underlying assumptions about the direct influence of program evaluation are simple enough. If an empirical study is conducted that is able to determine the extent to which a certain program works -that is, whether a specific government intervention in society has achieved what was intended - then this information should directly influence future decisions in the same policy area. Some have argued that so far in the United States, the real impact of evaluation findings has been very limited. 14 Others take the more optimistic view that evaluation already has been influential in the policy formulation process, and will be more so in the future.15 What are the possibilities for direct influence? If they appear limited, what can or should occur to increase the likelihood of evaluation having an influence on policy formulation? We will argue that for evaluation to be successful in directly influencing policy, the practice of program evaluation will have to be modified, perhaps fundamentally, to adapt to the realities of the decision-making environment within government. Only in this way will evaluation succeed in playing a major role in future government decisions. The next section of this essay will discuss the main features of the policyformulating environment in government that influence the extent to which program evaluation is likely to be used. These features will be compared to the characteristics of program evaluation as it is usually carried out, and the inconsistencies between the two discussed. Modifications to traditional program evaluation that would lead to a better integration with the policy formulation environment will be examined, and conclusions presented. The Policy Formulation Environment in Government In the Canadian system of government, policy is usually formulated not in abstract ideological terms, but in the context of specific programs or legislation.16 From time to time, governments issue policy statements of general intent, but the significant decisions made by governments tend to be in relation to individual program or legislative changes. If program evaluation is to have a significant influence, it must affect these program decisions. In order to understand the potential influence of evaluation in these decisions, it is first necessary to understand how the decision mechanisms in government operate and to identify those characteristics of the decision process that are likely to inhibit or encourage the use of evaluation results in the policy development process.

270 / J. Mayne and R.S. Mayne In the cabinet system of government, decisions are based on the consideration of one or more program options, which typically are written in the form of a formal memorandum to cabinet. The cabinet memorandum is usually presented by the minister concerned with the program area to his cabinet colleagues. The document is reviewed and discussed, and a decision is taken to proceed either with one of the options presented in the document or with a modification of one or more of these options. Alternatively, no decision is taken and the document is set aside for future consideration pending further analysis or discussion at the political or the bureaucratic level or both. In the parliamentary system, the cabinet provides a convenient place to concentrate attempts to influence policy decisions. This is in contrast to the United States, where several adversarial decision-making bodies have veto power. This makes it more difficult for a particular study to have an influence, especially since the study will probably have been sponsored by one of those bodies. It seems more likely that empirical studies will influence decisions in a system with a single decision-making focal point, such as the cabinet in a majority government. Nevertheless, the presentation of a memorandum by a minister to cabinet and the ensuing policy debate are typically only the final steps in a long and complex process involving extensive consultations both inside and outside government. Many revisions will have been made to an initial draft memorandum. If an evaluation is to have a direct influence on policy formulation, it must have an impact on the often complex process leading up to the presentation of the cabinet document to ministers. Policy decisions often involve elaborate consultations, reviews, and revisions; this is because government programs deal with complex economic and social problems, and the types and ranges of interventions used by governments are themselves often complex. A government program may consist of a range of goods and services provided through a variety of mechanisms to a number of different segments of the population. This means that programs often are not well-defined entities, but rather consist of a loose collection of diverse activities directed at a number of different objectives. In addition, program objectives are often phrased only in general terms. Politically, a program is more likely to be approved not because of the precision of its objectives, but rather because of their broadness or even their ambiguity. This allows various special-interest groups to be satisfied that they will benefit from the program, or at least that the program will not be detrimental to them. 17 Because of this complexity, evaluation studies can only focus on certain aspects of a government program, such as their administrative or financial characteristics, social impacts, or economic benefits. Who is to decide which characteristics (which activities? which objectives? which effects?) are important

Program Evaluation / 271 and therefore should be the subject of study? Consider a program that provides industrial grants to firms. Outcomes of this program could include new jobs, export sales, reduced regional disparities, enhanced defence production capability, technology development, or a distorted market. Which ones are to be examined? If program evaluations are to be successful in influencing the decision process, these evaluation design decisions must be made in such a way that the information provided by the study will be directly related to the programming concerns of decision-makers involved in the process. Another reason for the complexity of the decision process is that it often involves a large number of interest groups. Both inside and outside government, interest groups will have differing ideological, cultural, and political perspectives. The policy process will therefore involve a debate among these various groups, each of which will try to influence the formulation of policy or program options before a decision is taken by cabinet. Different interests involved in the policy formulation process will have different viewpoints, and will interpret the objectives and range of programming alternatives in different ways. This means that the information about the program needed by different interest groups varies. If a study is to have an effect on this type of complex decision process then it must take into account as wide a range of views as possible. Otherwise, the influence of the study's results on the final decision may be reduced or eliminated by those who feel that the study is not relevant or not credible. Another key characteristic of the government decision process relates to the timing of decisions. Predicting when decisions are to be made is difficult given the fluid nature of the decision process. Even when the timing of upcoming decisions can be determined with some certainty, the time available for conducting a relevant study is often short. To be successful, studies must be planned in such a way as to anticipate as much as possible when decisions will be made. This requires good communication between those scheduling evaluation studies and those involved in the decision process. Often shorter studies that can adapt to a changing timetable will be more successful. An additional characteristic of the decision process is the difficulty of communication with decision-makers because of the hierarchical structure of government. Access to senior decision-makers (ministers and deputy ministers) is limited. Although many interest groups are represented in the decision process, a few key decision-makers have, in effect, a veto power. Therefore, the interests of persons in these key positions need to be taken into account in designing studies intended to influence the process as a whole. Unfortunately, those involved in the design of research and the planning of evaluation studies have tended in the past not to have direct access to senior decision-makers. To ensure that a study will remain relevant through what may be a long and

272/J. Mayne and R.S. Mayne complex decision process, it is necessary to establish some form of communication between the evaluation planners and the key decision-makers. Finally, government decisions may not be solely or even primarily based on rational-empirical considerations. Many other factors enter into most policy decisions, including bureaucratic and political considerations. In most large bureaucratic organizations, a wide variety of non-empirical factors directly affect the outcome of a decision process. Personal biases, personal ambition, hidden agendas, bureaucratic territoriality, power relationships, tradition, and existing structures can all play an important part in influencing policy decisions. In addition, it is important to recognize the political context in which policy formulation occurs. Many policy or program changes are politically controversial, since the proposed changes usually affect some groups more than others. Special interests, ideological and cultural perspectives, political timing, and media reaction all influence the decision process. In some situations these factors will play a dominant role and will virtually eliminate the possibility of an empirical study having a significant direct influence on the policy decision. In summary, we have seen that policy formulation within government is a complex process. Any attempt to influence it must take into account the nature of the process and the interests of those involved in it. Specifically, we have seen that (1) government activities are often complex and that the structure and objectives of government programs may be ambiguous; (2) government policy processes involve different and competing interest groups; (3) the timing of decisions is often difficult to predict; (4) access to senior decision-makers is usually limited; (5) many decisions are not based primarily on rational-empirical considerations. To be successful in influencing the decision process, any evaluation must take into account the characteristics that are inherent to the process. In the next section we will look at the nature of program evaluation and identify those elements that are inconsistent with these decision-process characteristics. Program Evaluation as a Research Activity Current program evaluation practice derives from the tradition of applied social research. Other disciplines, such as management science and public-policy analysis, have an overlapping interest in the assessment of government programs. However, the aims and methods of social research are the predominant influence in most current program evaluation practice. This means, for example, that evaluation is often defined simply as another area of applied social research: ... evaluation research is the application of social science methodologies to the assessment of human resource programs, so that it is possible to determine, empirically and

Program Evaluation / 273 with the confidence that results from employing scientific procedures, whether or not they are useful. 18 Evaluation research is, first and foremost, a process of applying scientific procedures to accumulate reliable and valid evidence on the manner and extent to which specified activities produce particular effects or outcomes.19

The social research approach to evaluation carries with it an elaborate set of implicit assumptions about the evaluation process, including the overall purpose of the activity, acceptable procedures and methodologies, and acceptable rules of evidence. As we shall see, these implicit assumptions or operating biases are not consistent with the requirements for strategic information in the decision-making process. The social research approach usually involves the following steps: (1) the determination of the objectives of the program to be evaluated; (2) the specification of the indicators by which the achievement of the objectives is to be measured; (3) the collection of data on the indicators using some evaluation design;20 and (4) the formulation of findings, which are compared to the objectives so that an! appropriate conclusion can be reached as to the success of the program. The research methodology used in this approach is typically some form of experimental or quasi-experimental design, whereby two groups are compared-a group whose members are directly affected by a program and a control group whose members are not affected at all. By carefully comparing the two groups and correcting for differences between them, the specific effects of the program can, in theory, be isolated.21 This traditional approach to program evaluation relies on the classical scientific method as its operating paradigm: program evaluation is viewed as scientific research. 22 This has several important implications, the most important of which is that program evaluation defined in this way entails a search for scientific authoritativeness. The aim is to provide evidence that is as objective, as reliable, and as valid as possible. As research, program evaluation is the search for objective 'truth' about the outcomes of programs and therefore can be carried out quite independently from any current decision process. In addition, the specific issues to be examined in any particular study will be defined in terms of research criteria-that which is most interesting from a theoretical point of view or that which fits most readily into a given research methodology. As a result, the focus in this approach is often on the more esoteric aspects of a program rather than on issues that are more obvious or of more immediate concern. For example, an evaluation of an industrial grants program that concentrated on the general equilibrium effects of corporate investment behaviour would be irrelevant if the policy concern was the number of jobs to be created.

274/J. Mayne and R.S. Mayne Furthermore, evaluation as research tends to concentrate on a relatively few issues which are examined in depth rather than, for example, gathering limited evidence on a range of issues. This approach tends to assume away many of the complicating real-world features of any problem in order to try to capture the essence of the situation. The result is that research evaluation tends to require clearly articulated objectives as a basis for developing research hypotheses. Non-rational factors are often treated as unfortunate obstacles to be overcome rather than as necessary constraints to be dealt with and adapted to. The resulting findings are often inappropriate for policy decision-making because they do not lead to recommendations on specific courses of action.23 A final consequence of this search for scientific authoritativeness is that the studies tend to be long-term and relatively inflexible with respect to both timing and methodology once the research design is established. As a result the requirements for legitimate evidence are demanding. The current approach to program evaluation does not always exhibit all these characteristics, but the implicit research paradigm that is used does bring with it significant biases about what should be examined, how the study should be carried out, and what constitutes valid evidence. Furthermore, at least some of these characteristics are part of most current evaluation research efforts. The problem is that these characteristics of the research paradigm are not consistent with the information requirements stemming from the nature of the policy formulation process. As we have seen it is the nature of the decision process itself that determines to a large extent the type of information that will be used as an input to any final decision. For example, because government programs and objectives are typically complex and often ambiguous, it is difficult or impossible to translate program characteristics and objectives into precise research hypotheses. Attempts to do so often end up either focusing on program impacts that are not of major concern in the decision process, or producing complicated findings that contain long lists of assumptions and are difficult to interpret. Also, since the decision process involves many competing interest groups, a study that is broad in scope will often be better received and achieve a wider credibility. In the previous section, the problems associated with the timing of the decision process were identified. Again, a different approach from that usually followed in the research paradigm is more likely to be successful in directly influencing the process. Long-term studies based on an inflexible research design are likely to be lost in the shuffle as political and bureaucratic priorities develop and change. One example of this is the MINCOME Manitoba negativeincome experiment conducted during the mid-1970s, which is still being analysed even though the climate favourable to policy change in this area has long since passed.

Program Evaluation / 275 Another characteristic of the decision process is the difficulty in gaining access to and developing a dialogue with senior decision-makers. As was pointed out, communication is necessary to ensure that a study's results will remain relevant. The orientation of the research paradigm is exactly the opposite. There may seem to be little reason to communicate with decision-makers since the search for 'scientific truth' can be carried out in isolation from any current policy formulation process. There is no requirement in the research approach to ensure that the results will be relevant in any direct way to a particular decision. As a result of his training, the researcher typically defines the research problem primarily to suit himself. Important factors affecting the policy decision that are not based on empirical-rational grounds are often ignored, even though they may in fact play a key role in the final decision. Given this discrepancy between the information needs of the decision process and the characteristics of the research approach to program evaluation, our general conclusion must be that this type of program evaluation is unlikely, in general, to be able to influence directly specific policy decisions. However, even if this conclusion is correct, it does not necessarily mean that we should therefore abandon the attempt to make program evaluation play a significant role in the policy development process. What is needed, in our view, is a re-examination of the methodology and assumptions underlying current program evaluation practice to see how changes can be made to improve the chances for success. In the next section we will discuss some possible changes in current evaluation practice. Adapting Program Evaluation to the Policy Formulation Environment

We have seen how many of the characteristics of policy decision-making militate against the direct use of program evaluation - or indeed any empirical study - in the formulation of policy. This is particularly true of evaluation carried out from a research perspective. But it is this research perspective that has characterized traditional evaluation, and that continues to dominate the way in which much program evaluation is currently conducted. A better recognition of the realities of decision-making and of the limiting characteristics of traditional approaches provides a basis for modifying program evaluation so that it can play a part in the government decision process.24 On the surface, what needs to be done seems quite straightforward. First, program evaluation needs to be planned and managed so that it becomes integrated into, and hence part of, the decision-making process. Second, there is a need to design evaluation studies that will be more policy-relevant and decision-oriented than has usually been the case in the past. Both of these changes are, however, extremelv difficult to implement, the first because it

276/J. Mayne and R.S. Mayne involves a different role for the evaluator and implies changing, at least in part, the decision-making process itself, and the second because it involves changing the way evaluators think about their work. Integrating program evaluation into the decison-making process implies the need for a way to schedule and report the findings of the study in a timely manner, and for a way to focus the study on policy-relevant issues. Effective communication must be established between those responsible for policy decisions and those responsible for conducting the evaluations, so that relevant policy issues (and the programs and specific questions to be evaluated) can be identified. Studies can then be designed to meet the information and timing needs. We have seen earlier that, owing to the complexity of government programs, assuring the relevance of evaluations is not easy. The difficulty is compounded because, in the quest for relevance, the role of the evaluator changes from that of researcher to that of management advisor. In this role, the evaluator must become involved with 'management' and its concerns. This implies trade-offs and compromises in order to achieve relevance. Such a role is often not attractive to those trained in the classical social sciences: 'The traditional academic values of many social scientists lead them to want to be nonpolitical in their research. Yet they always want to affect government decisions. The evidence is that they cannot have it both ways.'25 Without a major effort on the part of the evaluator to find out what issues are relevant, a particular program evaluation is likely not to be useful in a specific policy decision.26 While the need for planning and management of an evaluation to ensure its integration into the decision process is perhaps clear, what may not be appreciated is that this requires in all but the simplest cases an elaborate set of procedures and incentives. This is true for even the apparently simple requirement of appropriate scheduling and reporting of studies. Time is at a premium for senior decision-makers and, as we have seen, the decision process is often complex. In addition, those responsible for the design of a study are often seen more as technical advisors than as staff advisors and therefore are not involved in a direct relationship with decision-makers. One example of an attempt to forge links between evaluations and the decision process is the current Canadian federal initiative to build an evaluation planning and reporting system into its decision process and to establish a permanent interaction between those planning the evaluation studies and those involved in policy decisions. Departments are required to submit their plans for evaluations and the findings from important evaluations to the central agencies. Based on their understanding of the policy issues coming before cabinet committees, central agencies can influence the schedule and focus of an evaluation and

Program Evaluation / 277 ensure that relevant findings are considered by cabinet committees.27 Cabinet committees can also call for specific evaluations. This system envisages all programs being evaluated over a three- to five-year cycle - a strategy primarily intended to improve the accountability of programs. It should also have the effect over time of making available current evaluation findings for almost all programs. This would reduce the problem of decisions being made before an evaluation study can be completed. A different approach to linking evaluation and policy-making has been tried in the Florida Department of Health and Rehabilitative Services.28 There, a systematic formal mechanism was created, which requires that a review of evaluation and policy analysis results take place involving the policy developers, evaluators, and program managers. This leads to recommendations for policy changes. Although perhaps not feasible in a larger federal context because of the many organizations that would be involved, this process does result in evaluation having a direct impact on policy decisions.29 It should be noted, however, that if linkages are established, they necessarily will interact with and affect the policy process itself. The process will be modified to accept the evaluation findings as a legitimate part of decision-making. The current trend toward fiscal restraint by governments and the continuing demand for more effective management should encourage this acceptance, but it will by no means assure it. The decision system itself may not be ready to adjust to the new demands and implicit judgments that would be entailed by the extensive use of evaluation studies. We have already touched on the second reform, the need to conduct policyrelevant studies. This means more than simply communicating with policymakers. It requires a rethinking of what evaluation should entail and of the purpose of the evaluations. The traditional approach of a rigorous study that aims at testing the extent of achievement of the program's objectives will likely discourage the consideration of evaluation findings in the policy decisionmaking process. The presence of multiple decision-makers, non-empirical factors, and various political actors suggests that the scope of the study must be broad enough to include the interests and concerns of all significant parties. Typically, there is interest in much more than whether a stated objective has been met: the program may be causing a number of outcomes relevant to those affected; there may be better ways of achieving the intended results; or the situation the program was addressing may have changed. A study that does not aim solely at producing single absolute measures (for example, benefit-cost ratios) but rather tries to provide credible data on a wide variety of outcomes and questions concerning a program is more likely to be used in a decision process. Proponents of different

278/J. Mayne and R.S. Mayne points of view can use such results to back up their arguments. In this way, the study should at least reduce the uncertainty surrounding the outcomes of programs and produce findings of interest to all parties. Furthermore, by considering all significant aspects of the program and by not emphasizing one finding over another, the study will appear to be more balanced. This may be the only way to gain a measure of acceptance. The possible scope of a program evaluation, therefore, must be broadened to allow the study to focus on any aspect of a program or its results and must not be limited to examining the achievement of objectives. A truly useful evaluation allows for any policyrelevant effect to be examined. In addition to broadening the scope of program evaluation, the level of rigour required in establishing evidence must be reconsidered. Typically, the timing of decisions will mean that a shorter (and hence less 'scientific') study may be all that can usefully be completed in time. Precision may have to be sacrificed for relevance. Indeed, a good approach may be to try to gather several lines of evidence that point to the same conclusion. While each may be subject to debate, the cumulative evidence gathered may be sufficient to be used in the decisionmaking process. This evidence is more likely to be credible to policy-makers than attempts to establish definitively a single line of proof concerning the outcomes of a program. Finally, the purpose of the study must be changed from trying to establish 'facts' to providing information for use in a policy debate. In cases where non-empirical and political factors are of major importance, the impact of any study may simply be to raise the level of the discussion by injecting some relevant empirical evidence into a policy debate. An evaluation is unlikely to be used as the sole basis for a policy decision. A study carried out with the aim of being used as an information base for the debate on a policy is less likely to be dismissed and will be seen as less threatening, since it is presenting relevant information and evidence rather than purporting to have produced authoritative conclusions. As a result, the study will have more credibility and various groups involved in the policy process will tend to use at least part of the study in the debate. These changes in the way a program evaluation study is carried out imply a significant if not a fundamental change in how evaluation is defined. Evaluation that tries to have a direct influence on policy formulation cannot be conducted as if it were searching for 'scientific truth' on which decisions will be based. Rather, it must 'get involved' and be much more an exercise aimed at filling information gaps identified by policy-makers, and reducing to some degree the uncertainty about the results of government programs. In summary, we have suggested that program evaluation can be a useful input to policy decisions if the planning and management of evaluations is integrated

Program Evaluation / 279 into the policy decision process, and if evaluations are designed to provide appropriate information for policy deliberations. If evaluations are intended to play an ongoing role in policy formulation, there is a need for a formal system to permit and encourage communication between the evaluator and those involved in the policy decision process. This will aid not only in designing relevant studies but also in enhancing their timeliness. Furthermore, to be compatible, evaluations conducted in this environment will have to be guided by a paradigm other than that typically employed when evaluation is viewed as applied social research. Conclusions In the fiscal year 1981-2, the federal government spent some $20 million carrying out program evaluations. Can we expect to see a corresponding increase in the utilization of evaluation findings in the policy formulation process? We have tried to address this question by examining the nature of the decision process itself, and by contrasting the demands arising from the decision process with the characteristics of program evaluation as it has been traditionally perceived. In order to understand how program evaluation could influence policy formulation, it is first necessary to understand the process and to recognize those factors that affect its operation. The decision-making environment determines in large part the type of information that will have a direct influence on the decision process. The complexity and ambiguity of government programs, the diffuseness of decision-making, the difficulty in predicting the timing of decisions, and the influence of non-rational factors all help to define the degree of influence that can be exerted by any empirical study. When we compared the characteristics of the decision-making process with those of the traditional approach to evaluation, we found that, for the most part, they were not consistent with one another. The social research tradition implies a need for precise hypotheses, an orientation toward long-term in-depth studies using sophisticated methodologies, and a striving for scientific authoritativeness. The decision process must deal with imprecise program objectives, short lead times for decisions, and an orientation toward provision of advice to decision-makers and mediation among competing interests. We concluded from this comparison that program evaluation as research is unlikely, in general, to be able to influence directly specific policy decisions. In our view, a shift in the current approach to program evaluation practice is necessary. Specific changes are needed: program evaluation should be planned and managed so that it becomes part of the decision-making process, and program evaluation studies must be designed to ensure that they will be relevant

280/J. Mayne and R.S. Mayne to current policy formulation problems. It is difficult but not impossible to achieve these aims. To do so will involve a significant review and perhaps a redefinition of what program evaluation entails and how it is carried out. We have argued that evaluations can be of direct use in policy formulation if they aim at providing relevant information as one input into the decision process rather than attempting to provide definitive answers to the policy issues being discussed. An evaluation carried out for this purpose would look at much more than whether the objectives of the program have been met. It would be directed toward ensuring relevance and timeliness of the information for decisionmaking, at the expense, if necessary, of some scientific rigour. The evaluator's role should be 'to illuminate the situation, not to dictate the decision.'30 This role for evaluation is in keeping with the basic nature of policy formulation-a process of conflict and accommodation. NOTES 1 For a discussion see Edward A. Such man, Evaluative Research (New York: Russel Sage 1967), ch. 1, and Social Sciences in Policy Making (Paris: Organization for Economic Co-operation and Development, 1979). 2 Howard E. Freeman, 'The Present Status of Evaluation Research,' in Evaluation Studies Review Annual, vol. 2, ed. M. Guttentagand S. Saar (Beverly Hills: Sage Publications 1977), 18-22. 3 For a review, see Dan R. Price, 'Sunset Legislation in the United States,' Bayor Law Review 30, no. 3 (1978), 401-62. 4 The auditor-general, for example, has stated that 'The scope and quality of effectiveness evaluation will have to be increased significantly before management, the Government and Parliament, each with its respective interests, can be reasonably informed on the achievements of public programs.' Auditor-General, Conspectus of the Annual Report 1978 (Ottawa: Supply and Services Canada 1978), 11. 5 Royal Commission on Financial Management and Accountability, Final Report (Ottawa: Supply and Services Canada 1979) 102-3, 409-12, and recommendations 6.6, 6.7, and 22.9 6 See, for example, Timothy E. Reid, 'The failure of PPBS: Real incentives for the 1980s' Optimum 10, no. 4(1979), 23-36, and Auditor-General, Conspectus, 10. 7 Perrin Beatty, former minister of state for the Treasury Board under the Clark government, tabled a private member's bill in the House of Commons in May 1980, bill c-403, entitled 'An Act Respecting the Termination and Evaluation of Existing Programs, Regulations and Agencies of the Government of Canada.' 8 A federal government policy on program evaluation was issued in 1977 (Treasury Board Circular 1977-44), but guidelines for formal evaluation were not published

Program Evaluation/281 until recently: Office of the Comptroller-General, Guide on the Program Evaluation Function (Ottawa: Supply and Services Canada 1981). The linking of evaluation to the management system is described in Treasury Board of Canada, Accountable Management (Ottawa: Treasury Board 1981) and Government of Canada, Guide on the Policy and Expenditure Management System (Ottawa: Supply and Services Canada 1980). For a discussion of the evolution of evaluation in the federal government see J.M. Jordan and S.L. Sutherland, 'Assessing the Results of Public Expenditure: Program Evaluation in the Canadian Federal Government,' Canadian Public Administration 22, no. 4 (Winter 1979), 581-609. 9 This new system and its development is described in The Policy and Expenditure Management System (Ottawa: Privy Council Office 1981). See also Treasury Board, Accountable Management, and Government of Canada, Guide on the Policy and Expenditure Management System. 10 See, for example, Using Social Research in Public Policy Making, ed. Carl H. Weiss (Lexington: Lexington Books 1977), Charles E. Lindblom and David E. Cohen, Usable Knowledge (New Haven: Yale University Press 1979), Craig N. Locatis, Jeffery K. Smith, and Virgil L. Blake, 'Effects of Evaluation Information on Decisions' Evaluation Review 4, no. 4 (1980) 809-23, Martin Rein and Sheldon H. White, 'Can Policy Research Help Policy?,' The Public Interest 49 (Fall 1977), 119-36, and The Utilization of the Social Sciences in Policy Making in the United States (Paris: Organization for Economic Co-operation and Development, 1980). 11 For a discussion and examples, see Martin Rein, Social Science and Public Policy (New York: Penguin Books 1976) 119-23. 12 Lindblom and Cohen, Usable Knowledge, 6-10. 13 Ibid. 14 See, for example, Carol H. Weiss, Evaluation Research (Englewood Cliffs, NJ: Prentice-Hall 1972), Suchman, Evaluation Research, and H.R. Davis and S.E. Salasin, 'The Utilization of Evaluation,' in Handbook of Evaluation Research, ed. E.L. Struening and M. Guttentag (Beverly Hills: Sage Publications 1975). 15 This view is evident in the emphasis placed on evaluating the impact of programs for Parliament in the Royal Commission Final Report and is clearly the federal government's belief as evidenced in the linkages being formed between program evaluation and the new policy and expenditure management system. See, for example, Treasury Board, Accountable Management, 3-6. See also Charles L. Schultze, The Politics and Economics of Public Spending (Washington: Brookings Institute 1968), 60-4. 16 Indeed, policy is often defined in terms of the actions taken by governments rather than the expressions of interest which constitute policy statements. See, for example, H. Hugh Helco, 'Review Article: Policy Analysis,' British Journal of Political Science 2, part 1 (Jan. 1972), 83-108.

282/J. Mayne and R.S. Mayne 17 Fora discussion of the inherent fuzziness of objectives in government, see Herman R. van Gunsteren, The Quest for Control: A Critique of the Rational-CentralRole Approach in Public Affairs (New York: John Wiley and Sons 1976), ch. 1, and Harry S. Havens, 'MBO and Program Evaluation, or Whatever Happened to PPBS,' Public Administration Review 36, no. 1 (1976), 40-5. 18 Freeman,'Present Status,'25. 19 Leonard Rutman and Joe Hudson, 'Evaluating Human Services: A Process Approach,' in Social Work Processes, ed. Beulah R. Compton and Burt Galaway (Homewood: Dorsey Press 1974)410. 20 Program evaluation almost always does involve the collection of new data. There is rarely adequate information already existing on the effects of a particular program. In particular, social indicators - on which there is a large literature - are usually of limited value since they try to measure national, or at best regional, indicators and cannot be related to a particular program. 21 For a review of classical evaluation designs, see Donald T. Campbell, 'Reforms as Experiments' in Readings in Evaluation Research ed. Francis G. Caro (New York: Russel Sage Foundation 1977), 172-204, and Thomas D. Cook, Fay L. Cook, and Melvin M. Mark, 'Randomized and Quasi-experimental Designs in Evaluation Research,' in Evaluation Research Methods, ed. Leonard Rutman (Beverly Hills: Sage Publications 1977) 101-40. Methodological limitations on program evaluation are discussed in, for example, Program Evaluation: An Introduction (Ottawa: Office of the Comptroller-General 1980), Sharon Sutherland, 'On the Audit Trail of the Auditor General: Parliament's Servant 1973-1980,' Canadian Public Administration 23 no. 4(1980), 616-44, and Jordan and Sutherland, 'Assessing the Results of Public Expenditure,' 598-607. 22 An interesting discussion of evaluation as science versus evaluation as technology can be found in Jonathan A. Morell, Program Evaluation in Social Research (New York: Pergamon Press 1979). 23 See Glen G. Cain and Harold W. Watts, 'Problems in Making Policy Inferences From the Coleman Report,' in Evaluating Social Programs, ed. Peter H. Rossi and Walter Williams (New York: Seminar Press 1972). 24 Alternative strategies to the traditional approach to evaluation are the subject of several recent books. See, for example, Michael Q. Patton, Utilization -Focussed Evaluation (Beverly Hills: Sage Publications 1978) and Lee J. Cronbach and Associates, Toward Reform of Program Evaluation (San Franciso: Jossey-Bass Publishers 1980).

25 Patton, Utilization, 46. 26 The inability of academic economists to focus on relevant issues and hence to contribute to the formulation of the u.s. Clear Water Act of 1972 is described in Marc J. Roberts, The Political Economy of the Clear Water Act of 1972: Why No

Program Evaluation / 283 One Listened to the Economists,' in Utilization of the Social Sciences, 79-119. 27 See Office of the Comptroller General, Guide on Program Evaluation, Government of Canada, Guide on the Policy and Expenditure Management System, and Treasury Board, Accountable Management. 28 See Larry Polivka and Eric Steg, 'Program Evaluation and Policy Development: Bridging the Gap' Evaluation Quarterly 2, no. 4 (1978), 696-707. 29 Ibid.,700-3. 30 Cronbach, Toward Reform, 11.

This page intentionally left blank

Note on Contributors

Arpad Abonyi is in the Department of Politics at Queen's University. He has published widely on issues in contemporary international political economy. His research focuses on East-West trade in technology-intensive commodities and on the behaviour of multinational corporations. Michael Atkinson is associate professor of political science at McMaster University. He is the co-author of The Canadian Legislative System, and his articles have appeared in several journals, including the Canadian Journal of Political Science, Canadian Public Administration, and the Journal of Canadian Studies. Marsha Chandler is associate professor in the Department of Political Science and the Faculty of Law at the University of Toronto. In 1981-82 she was visiting scholar at the University Consortium for Research on North America, Harvard University. She is the co-author of Public Policy and Provincial Policies, and has contributed articles to the Canadian Journal of Political Science and Canadian Public Administration. William Coleman is an assistant professor of political science at McMaster University. He has written extensively on language in Quebec, and has contributed articles to the Canadian Journal of Political Science and Studies in Political Economy. He is currently taking part in a cross-national research project on the role of business interest associations in Canada and other countries. Pranlal Manga is a National Health Research Scholar in health administration at the University of Ottawa. He has published widely on various aspects of health policy. John Mayne is assistant director, Policy Division, Program Evaluation Branch, Treasury Board of Canada. Since he received his PHD in operations

286 / Note on Contributors research in 1973 from Northwestern University, he has held a number of positions in policy analysis in the federal government. He is also a sessional lecturer at the School of Public Administration, Carleton University. Robert S. Mayne holds master's degrees in applied statistics and sociology. He has held various positions in evaluation and policy analysis in government departments in Canada and the United Kingdom. He is currently senior consultant with DPA Consulting Ltd, Ottawa. Susan McCorquodale is associate professor of political science at Memorial University of Newfoundland. She has written on provincial politics and public administration. Her recent work includes studies of Newfoundland fisheries management for the Economic Council of Canada's regulation reference. Simon Mclnnes is currently an energy advisor with the Department of Energy, Mines and Resources. He holds a PHD in political science from Carleton University. He taught political science for several years (at Saskatchewan, York, and McMaster Universities) before becoming a policy analyst with the Inuit Tapirisat of Canada. Dale Poel is an associate professor at Dalhousie University with appointments in the Department of Political Science and the School of Public Administration. He has written on various aspects of public policy and is currently completing an evaluation of legal aid in Nova Scotia for the Nova Scotia Legal Aid Commission and the federal Department of Justice. J. Robert S. Prichard is associate professor of law at the University of Toronto. He is currently visiting professor of law at Yale University. He has served on a number of enquiry commissions and task forces, and has written several articles and books on various instruments of policy. Mark Sproule-Jones is professor of political science at McMaster University, where he holds the V.K. Copps Chair in Urban Studies. He has published widely in the fields of public-choice theory, philosophy of science, federalism, citizen participation, and resource management. Michael J. Trebilcock is professor of law at the University of Toronto. He is director of the Law and Economics Program, and the author of a number of books and articles on public policy, including The Calculus of Instrument Choice. Geoffrey Weller is associate professor and chairman of the Department of Political Studies at Lakehead University. He has published extensively in the areas of Canadian and comparative health policies and the politics of northern development. Kenneth Woodside is an assistant professor at the University of Guelph. He has published in the area of taxation policy in Canada and Britain and in the area of communications policy, with particular emphasis on the introduction of pay-TV in Canada.