Power Sharing in Deeply Divided Places 9780812207989

Table of contents :
CONTENTS
CHAPTER 1. Power Sharing in Deeply Divided Places: An Advocate’s Introduction
PART I. Power Sharing and Electoral Systems
CHAPTER 2. Electoral Rules and Ethnic Representation and Accommodation: Combining Social Choice and Electoral System Perspectives
CHAPTER 3. The Track Record of Centripetalism in Deeply Divided Places
CHAPTER 4. Electoral Engineering for a Stalled Federation
PART II. Historical and Conceptual Forays into Power Sharing
CHAPTER 5. A Theory of Accommodation Versus Conflict: With Special Reference to the Israel- Palestine Conflict
CHAPTER 6. The Success of Religion as a Source for Compromise in Divided Empires: Ottoman and Safavid, Past and Present
CHAPTER 7. Geopolitics and the Long- Term Construction of Democracy
CHAPTER 8. Courts, Constitutions, and the Limits of Majoritarianism
PART III. Contemporary Power- Sharing Questions
CHAPTER 9. A Revised Theory of Federacy and a Case Study of Civil War Termination in Aceh, Indonesia
CHAPTER 10. We Forbid! The Mutual Veto and Power- Sharing Democracy
CHAPTER 11. Northern Ireland: Power Sharing, Contact, Identity, and Leadership
CHAPTER 12. Public Opinion and Power Sharing in Deeply Divided Places
CHAPTER 13. The Balkans: The Promotion of Power Sharing by Outsiders
CHAPTER 14. Governing Polarized Cities
CHAPTER 15. Power Sharing in Kirkuk: The Need for Compromise
CHAPTER 16. Power Sharing: An Advocate’s Conclusion
Contributors
Index
ACKNOWLEDGMENTS

Citation preview

Power Sharing in Deeply Divided Places

NATIONAL AND E THNIC CONFLIC T IN THE T WENT Y- FIR ST CENTURY Brendan O’Leary, Series Editor

Power Sharing in Deeply Divided Places

Edited by

Joanne McEvoy and Brendan O’Leary

U N I V E R S I T Y O F P E N N S Y LVA N I A P R E S S PHIL ADELPHIA

Copyright © 2013 University of Pennsylvania Press All rights reserved. Except for brief quotations used for purposes of review or scholarly citation, none of this book may be reproduced in any form by any means without written permission from the publisher. Published by University of Pennsylvania Press Philadelphia, Pennsylvania 19104-4112 Printed in the United States of America on acid-free paper 10 9 8 7 6 5 4 3 2 1

Library of Congress Cataloging-in-Publication Data Power sharing in deeply divided places / edited by Joanne McEvoy and Brendan O’Leary. — 1st ed. p. cm. — (National and ethnic conflict in the twenty-first century) Includes bibliographical references and index. ISBN 978-0-8122-4501-1 (hardcover : alk. paper) 1. Ethnic groups—Political activity. activity.

2. Minorities—Political

3. Representative government and representation. 4. Ethnic

conflict—Political aspects.

5. Cultural pluralism—Political aspects.

I. McEvoy, Joanne. II. O'Leary, Brendan. JF1061.P68 2013 320.90089—dc23 2012041497

. Hodgson

CON TEN T S

1. Power Sharing in Deeply Divided Places: An Advocate’s Introduction Brendan O’Leary

1

PART I. POWER SHARING AND ELECTORAL SYSTEMS 2. Electoral Rules and Ethnic Representation and Accommodation: Combining Social Choice and Electoral System Perspectives Bernard Grofman

67

3. The Track Record of Centripetalism in Deeply Divided Places Allison McCulloch

94

4. Electoral Engineering for a Stalled Federation Kris Deschouwer and Philippe Van Parijs

112

PART II. HISTORICAL AND CONCEPTUAL FORAYS INTO POWER SHARING 5. A Theory of Accommodation Versus Conflict: With Special Reference to the Israel-Palestine Conflict Ronald Wintrobe

135

6. The Success of Religion as a Source for Compromise in Divided Empires: Ottoman and Safavid, Past and Present Benjamin Braude

176

vi

Contents

7. Geopolitics and the Long-Term Construction of Democracy Randall Collins

198

8. Courts, Constitutions, and the Limits of Majoritarianism Samuel Issacharoff

214

PART III. CONTEMPORARY POWERSHARING QUESTIONS 9. A Revised Theory of Federacy and a Case Study of Civil War Termination in Aceh, Indonesia Alfred Stepan

231

10. We Forbid! The Mutual Veto and Power-Sharing Democracy Joanne McEvoy

253

11. Northern Ireland: Power Sharing, Contact, Identity, and Leadership Ed Cairns

278

12. Public Opinion and Power Sharing in Deeply Divided Places Colin Irwin

295

13. The Balkans: The Promotion of Power Sharing by Outsiders Florian Bieber

312

14. Governing Polarized Cities Scott A. Bollens

327

15. Power Sharing in Kirkuk: The Need for Compromise Liam Anderson

364

16. Power Sharing: An Advocate’s Conclusion Brendan O’Leary

List of Contributors Index Acknowledgments

386 423 425 437

CHAPTER 1

Power Sharing in Deeply Divided Places: An Advocate’s Introduction Brendan O’Leary

The Mafia makes offers that cannot be refused. In one peace process a politician was once accused of making offers that no one could understand (O’Leary 1990). Do these statements explain the difference between power and power sharing? Is power coercive capacity, whereas power sharing is incomprehensible? Power sharing is not incomprehensible, but it is frequently misunderstood. To aid comprehension a comparison is useful. In standard English, power is the ability to act, to be able to produce an intended effect (Russell 1992 [1938]). The powerless lack the capacity to do things they might want to do. The powerful are in the opposite situation. Power sharing, therefore, suggests spreading access to the capacity to get things done. Power is also a synonym for authority, jurisdiction, control, command, sway, or dominion, as well as the capacity to persuade, induce, constrain, oblige, or force. It follows that power sharing minimally means widening the access of persons or groups to the same domains or attributes. In standard usage power is also “a possession,” “held” by those with authority or influence over others, especially public officials, governments, officers, managements, or establishments who constitute what Paul’s Letter to the Romans described as “the powers that be.” Power sharing, therefore, broadens membership of “the powers that be.” It also requires that the included parties have access to key and observable “decision making.” There must be no important “non–decision making” taking place off stage, that is, no hidden possessors of power who

2

Brendan O’Leary

control the agenda or exclude some issues from being addressed. There must instead be an open and negotiable public agenda among the powersharers, or at least among their leaders. Any suppression of (controversial) issues must be mutually agreed upon among those who share power. Theorists contrast “power to” and “power over” (see Morris 2002; Parsons 1969). “Power to” is ability, “power over” is domination. The contrast resembles that between “positive-sum” and “zero-sum” relationships. “Positive-sum” power is joint, collaborative, or cooperative. All gain from its exercise, even if the benefits are not the same for all. “Zero-sum power,” by contrast, describes a distinct antagonism: if power could be measured, then A’s gain and B’s loss would sum to zero. Positive-sum and zero-sum conceptions do not exhaust the logical possibilities of power relations. The exercise of power may generate net losses (a “negative sum”) or the mutual ruin of the contending parties. It may create winners and losers; there may be disparities in benefits among the winners as well as in losses among the losers; and only one party may gain, while the others experience no net losses. Power sharing, for its proponents, is defended as “power to.” It enhances collective capacity; it is “positive sum.” Those who share will gain from a constructive way of making public decisions, from which all stand to gain, notably through the preservation of order and peace. Critics, by contrast, suggest that power sharing shapes public life at the expense of other and better kinds of politics— more competitive, individualist, or harmonious. The opposite of power sharing is power’s monopolization by a person, faction, group, organization, or party. On inspection, it is usually true that the chief power-holder has to delegate some power to organize and maintain the monopoly. But to delegate power is not to share it. The principal who delegates requires the delegated agent to perform specified tasks and may withdraw the mandate. Monopolies of power exist, at least formally, in tyrannies, despotisms, military autocracies, monarchies, lordships, papacies, theocracies, and oneparty dictatorships. They also exist, however, in democracies, a more unsettling idea. To say that democracy may coexist with monopolistic domination requires no commitment to theories suggesting that behind the façade of electoral competition lies the power of a ruling class or a power elite (see, e.g., Miliband 1980 [1969]; Domhoff 1990; Mills 1956). For example, no matter how competitive or free elections may be, critical political power can be monopolized between elections by the incumbent president, prime minister, cabinet, and nominated judges associated with the dominant party,

Introduction

3

ethos, or ideology. Even a temporary domination (between elections) is nevertheless domination, and the opportunities for elected leaders to dominate their societies against widespread or deeply held public preferences are significant (see, e.g., Nordlinger 1981, 92–94, 111–12, 130–32). That democracy might lead to domination was the theme of the “tyranny of the majority,” which deeply concerned eighteenth-century republicans, such as James Madison, and nineteenth-century liberals, such as Alexis de Tocqueville and John Stuart Mill. They were mostly preoccupied, however, with the impact of that possible tyranny on the individual’s property and liberty (including the individual’s religious beliefs) rather than on national, ethnic, or linguistic minorities as such (Madison, Hamilton, and Jay 1987 [1788]; Mill 1997, 5– 6, 81–82, 192–94; de Tocqueville 1988 [1835, 1840], vol. 1., chap. 7, esp. 250ff ). Democracy is, however, also straightforwardly compatible with the (temporary) tyranny of a minority, especially democracies with institutions that encourage the “winner” to take all. For example, an ideological faction, not supported by a majority of voters in a country, may nevertheless control a cabinet, which in turn controls a party, and which in turn controls a legislature. In consequence, law or public policy may be dictated in the interests of the faction as long as its control is maintained.

Defining Power Sharing, Deeply Divided Places, and Well-Ordered States These considerations suggest the following broad definition of power sharing: Any set of arrangements that prevent one agent, or organized collective agency, from being the “winner who holds all critical power,” whether temporarily or permanently. This suggestion explains why the synonyms of power sharing usually include the following generally positive connotations: “coalition” or “cooperative” government and “consensual” and “inclusive” decision making. Critics of power sharing just as powerfully insist upon negative connotations. They refer to power-sharing arrangements as “rudderless” or “leaderless,” and they complain of “stalemated,” “deadlocked,” or “blocked” decision making. The general definition of power sharing just suggested is broad if not vague. It does not, for example, specify how power is shared among the parties. It is capacious enough to include arrangements such as the Roman Republic’s executive, based on the annual election of two consuls, and its tribunes, who

4

Brendan O’Leary

were able to veto legislation; ancient Sparta’s two kings and ephorate; the mercantile republican aristocracy of Venice; and the Institutes promoted by Calvin in Geneva. These are examples of power-sharing arrangements, and the definition thereby displays a key advantage: it does not presume that all power-sharing arrangements are virtuous by our current standards. A definition is of considerable merit if it makes it possible to approve or disapprove of the use to which power-sharing systems are put. In this book our authors’ attention is mostly on contemporary powersharing systems. The major exception is Benjamin Braude’s discussion of limited power-sharing provisions under the Ottoman and Safavid empires (Chapter 6). In contemporary political science, to summarize a very large literature, power sharing is defined both by a regulatory goal and by specific instruments. The goal is the arrangement of political institutions to prevent the monopoly, permanent or temporary, of executive, legislative, judicial, bureaucratic, military, or cultural power. Four principal sets of instruments accomplish this goal. 1. The first are overtly political bodies (executive, legislative, judicial, and administrative) organized to ensure both “shared rule” and “selfrule” among the relevant agents. These political bodies are organized through partly self-governing communities or territories, or both. Differently put, and as we shall elaborate below, these political bodies may be consociational (based on communities) or federal (based on territories). These bodies usually respect some combination of the principles of parity, proportionality, and autonomy. 2. The second are security bodies: militaries, which normatively face outward for defense; police, which face inward to preserve order; and intelligence agencies, which engage in lawful surveillance and threat assessments. Security bodies must be organized so that power sharing within the political bodies is meaningful. 3. The third are economic policies, principally wealth-sharing formulae, that reinforce the power sharing within the political bodies through some combination of the principles of parity, proportionality, and autonomy that also apply within the political bodies. 4. The fourth are policies and practices that preserve cultural pluralism. Modern power sharing deliberately avoids the full-scale integration or coercive assimilation of “cultures” within the polity; that is part of its monopoly-rejecting ethos.

Introduction

5

The last set of instruments may be called “cultural protectionism,” both by their critics and their proponents. They distinguish modern power-sharing systems from the ancient, medieval, and early modern examples cited above: the Romans, the Spartans, the Venetians, and the Calvinists never intended to promote cross-cultural power sharing within their republics. In modern power-sharing systems, “cultures,” and their evolution, are not left to the free market or subject to the governing diktats of the largest group. Instead (at least some key components of) the cultures of the parties to the power-sharing system are protected. The overall power-sharing settlement may promote an inclusive overarching shared public identity, but that identity must complement rather than wholly replace the previously existing “cultures.” “Cultures” is put in quotation marks because agreeing what is entailed in “culture” is highly contested. As employed here culture encompasses the languages, national and ethnic traditions, religions and philosophies of life, customs, mores, and the ethos of peoples. As used here, no supposition is made that particular cultures are homogeneous, intrinsically holistic, static, wholly authentic and unaltered transmissions from antiquity, or mutually exclusive. It is, however, usually true that in deeply divided places at least some key agents believe that their “cultures,” in whole or in part, are threatened by others. Contemporary power-sharing systems promote the coexistence of at least some cultures. This idea is partly reflected in the language of “multiculturalism.” But to anticipate some false and facile criticisms, modern power sharing does not promote all cultures (e.g., headhunting, tribal scarring, genital mutilation of males or females, or foot-binding). Nor does it presume to freeze the cultures of the contending parties as they were when the power-sharing settlement was made. Rather, modern power sharing, through “encoded pluralism,” enables the partners to the political settlement to have the power to govern changes in their own cultures—through autonomy (self-rule) and through joint agreement with others (shared rule). “Cultural protectionism” works not through the wholesale freezing of certain practices but through empowering specified groups to control their own cultural evolution—both autonomously and jointly. “Deeply divided places” has more obvious connotations than power sharing, but warnings are in order. “Places” is a better expression than “societies” because it is a mistake to presume that a divided place contains just one society; that may be an issue in deep dispute, and a deeply divided place may be characterized by rival, parallel, or segregated societies. In a deeply divided place there may be more than one “civil society,” and their relations may be

6

Brendan O’Leary

far from civil. All moderately complex societies are divided (stratified) in ways that may matter politically, for example, by age cohorts, by sex or sexual preference, or by income, wealth, class, and status. But within deeply divided places these standard stratifications are superseded, or profoundly reinforced, by further divisions of nationality, ethnicity, race, tribe, language, or religion. Deeply divided places are, as the designation suggests, sites of actual or potential “civil” or intergovernmental wars. They are where genocide, ethnic expulsion, or coercive assimilation are threatened, or have taken place; they are the places for which power sharing is often recommended. There needs, however, to be some prospect of “stateness” or “governability” for power sharing to work as a recipe for deeply divided places. States matter more than societies in building inclusive power sharing because they define societies and their possibilities. Impersonalized institutions that have some degree of centralized and procedurally governed political decision making characterize functioning states. They have coercive capacities to ensure security: they can regulate all instruments of potential public violence and prevent or inhibit their own agents from being predators. They express authentic legal authority over persons, property, and their movements, and are recognized as such entities by their citizens, civil society organizations, and other states. Through self-help or alliances they can defend themselves. Lastly, functioning states are defined by their recognized sovereignty over their territory and its accompanying prerogatives: control over entry and exit of persons and entities. If states lack these capabilities they cannot protect human rights, promote human development, be inclusive, or share power effectively. Credible power sharing requires credible commitments to governability. Conversely, failing and failed states are personalized: previously dominated by rulers, a family, clan, or clique, which did not distinguish public from private realms. They have become “kleptocracies,” governments of thieves, before or during the collapses of their regimes. They lack coherent, institutionalized, rule-governed patterns that inhibit predation. The “rulers” are indeed predators. They have usually lost their monopoly on the regulation of coercion and are challenged by guerillas, paramilitaries, terrorists, Mafiosi; they may be invaded, looted, and occupied by other states. They neither make nor enforce law. Those over whom they have failed to rule despise them as much as they fear them. These properties of failed states remind us that inclusion and power sharing work best within wellordered states. Power sharing, inclusivity, and human development require more than the diff usion of the right values; they need the soil of functioning

Introduction

7

states because they are unlikely to grow in “anarchia.” Order, or its likely realization, is therefore a key condition of power sharing in a deeply divided place, a negotiated order that brings widespread human security. Thomas Hobbes was right to emphasize the necessity of order for a worthwhile human life but deeply wrong to mandate everywhere an authoritarian solution or a sole sovereign. Unlike Hobbes, we know that our states are far more lethal than the alleged “war of all against all,” and for that reason alone we need to prevent them from becoming Leviathans. Power sharing is one route to controlling the violence of states, not just the violence of civil wars. States are the most powerful agencies of exclusion, and governments the major murderers in human history; and many state-builders and nation-builders have been people-killers and nation-killers (Connor 1972). Rudolf Rummel has calculated that in the twentieth century governments killed nearly 170 million people within their borders, a figure that exceeds those killed in wars between states (Rummel 1997). Genocide—killing peoples because of their presumed ascriptive characteristics—has been more common than most countries’ official histories acknowledge, and governments have been the major perpetrators: the list is not confined to Ottoman Turkey, Nazi Germany, and Interahamwe Rwanda. “Politicide”—killing those deemed political opponents—has also been recurrent in modernity. The Soviet Union, especially under Stalin, was the major killing regime of the last century: nearly sixty-two million may have perished under its yoke. Maoist China was often as brutal, killing over thirty-five million people. “Democide”—the killing of peoples—is the ultimate form of exclusion, and government its major agency. Governments have also organized, encouraged, or not stopped the expulsion of whole categories of persons from their land borders or their shores—people whom they have helped defi ne as undesirable, nonindigenous, non-national, or disloyal. The twentieth century is justly described not just as one of death by government but as one of expulsions, of the “cleansing” of populations. Governments have also been the prime architects of policies of discriminatory control: orga nizing dominant national, ethnic, linguistic, or religious groups, and disorganizing and subordinating others through systematic discrimination, what we are now encouraged to call “exclusion.” These should be commonplace thoughts. Regrettably, they are not sufficiently appreciated. States define people’s life chances, and it is to their practices that we must look to see how greater inclusion, promised by power sharing, may best be facilitated. Exclusion, in the regulation of national, ethnic,

8

Brendan O’Leary

linguistic, and religious differences, is the product of coercive homogenization and special treatment of privileged communities: counting some categories of persons and communities “in,” as members, and others “out.” Genocide, expulsion, and the unilateral partition of territories are ways of homogenizing peoples that are now internationally outlawed, even if the laws are dishonored. If international norms against genocide and expulsion are rigorously enforced, through diplomatic engagements, economic sanctions, military embargoes and interventions, and the international criminalization of genocidal officials, then the worst forms of exclusion may eventually be halted or inhibited. But success in these endeavors requires moral universalism and a reorientation of the foreign policies of the great powers so that both their policymakers and domestic constituencies see the prevention and punishment of genocide and expulsion as in their interests. We are far from there. The record of recent history is, however, mixed: it need not occasion despair. Reforms and sanctions against exclusionary practices—against discriminatory racist, religiously intolerant, and xenophobic regimes—had major successes in the last century. The undermining of apartheid in South Africa had both international and domestic sources, both moral and political causes. Decolonization, desegregation, autonomy, the advancement of language rights, and civil rights movements were significant elements in the more inclusionary advances of the twentieth century. Power sharing is one of the most important instances of enhanced “inclusion,” and it was revitalized in the last century and in this one. It is a standard prescription for protracted national, ethnic, and communal conflicts in deeply divided places, especially ones focused on antagonistic self-determination claims, but it is not the sole one; and no sensible advocate of power sharing assumes it is a panacea. Power sharing in deeply divided places is the subject of this collection of essays. Our contributors address variations in power-sharing systems and the instruments used to calm deeply divided places to make them more peaceful, civil, and stable, and to provide credible commitments to well-ordered democratic states.

Constitutionalism: The Separation of Powers, Competing for Power, and Power Sharing Democracy can operate as a tyranny of the majority, and as the tyranny of a monopolistic faction, and therefore may worsen sociability and civility among

Introduction

9

the contesting peoples in a deeply divided place (McGarry 2010b). One response to such fears is to suggest that well-designed “constitutionalism” will block such tyrannies through protecting some core rights of citizens from potential violations by the governing majority (or faction) or from the stifling of individuality by crushingly conformist public opinion. Constitutionalism also works through making officeholders publicly accountable, requiring them to follow rule-governed and transparent procedures in their performance of their functions and through dividing public authority among multiple offices and institutions. Is constitutionalism therefore the relevant way to control monopoly? Is constitutionalism the key to power sharing? Power sharing is indeed normally constitutionalized, that is, encoded in formal constitutions, written agreements, treaties, or charters, but that does not mean that all constitutions are power-sharing systems, especially regarding the management of cultures. Constitutions in principle may assign termlimited authority to use power in a monopolistic fashion to a person, faction, party, or role, for example, “The President shall exercise the full plenitude of the executive power for a seven year term of office.” The constitution may grant emergency powers to a single-person executive, which may include provision for the suspension of basic rights, and fundamental freedoms, or dramatic “war powers.” A constitution may mandate that there is just one nation, one religion, one language, one ethnicity as the basis of eligibility for citizenship. A constitution may mandate that the territory of the state is indivisible, thereby blocking the formation of federal regions. A constitution may be nonamendable, or powerfully entrench some identities at the expense of others. The 1982 Constitution of Turkey has many of these features. But even when constitutions are more pluralistic they need not be power-sharing systems. If constitutional rules permit the domination of one person (faction, party, or national, religious, or ethnic group) over all the key institutions that shape major political decisions (and nondecisions), then the abuse of political power is possible. A powerful president or prime minister, backed by a regularly electorally endorsed party, may lawfully execute the powers of the office, promote legislation that reflects the preferences of the dominant nationality, race, religion, or linguistic group, and fi ll the judiciary and administration with their appointees. The “rule of law” is therefore often the rule of the dominant majority (or faction). Periodic elections may not prevent such domination. Accountability to some of the electorate, or to the courts, may not protect visible minorities, especially if the laws (and the constitution) have been drafted according to majority preferences.

10

Brendan O’Leary

Courts staffed by appointees of an executive and legislature controlled by the same party may not act as guardians of individuals’ rights, let alone collective minorities. Constitutionalism per se does not prevent cultural homogeneity from becoming the hallmarks of policy and the state. Constitutionalism, formal or informal, is probably a necessary condition for successful power sharing across cultures, but it is definitely not sufficient. Indeed modern power sharing is conceived of by some of its supporters as a necessary supplement to constitutionalism, required to prevent the monopoly of cultural power, and especially required by large minorities. Another way to make much the same argument distinguishes three remedies for despotic power found primarily in liberal thinking, namely the division of power, competition for power, and power sharing. All these three ways of preventing despotic power may be constitutionalized, although it is the separation or division of powers that is historically associated with “constitutionalism.”

The Separation of Powers

In the liberal political tradition, influenced by the arguments of Montesquieu and James Madison, and strongly present in American federalism, the separation of powers holds a famous place. Here dividing political power is seen as critical to preventing despotism. The tradition commends separating executive, legislative, and judicial institutions. Inhibiting a monopoly of power, especially in the executive, avoids kingly dictatorship. The separation of civilian from military power, of nomination from appointment, of police powers to arrest and interrogate from the judicial power to prosecute, of federal governments from state governments, or local governments from central governments are less recognized but just as important parts of the same logic. To divide power is to prevent its abuse; to check power with power controls public officials. Ambition tempers ambition. Some think that properly organizing the division of power to prevent domination is what really matters in deeply divided places (see Roeder and Rothchild 2005; Roeder 2007). A well-structured division of power, they say, inhibits national, ethnic, or communal majorities—or minorities—from dominating others. Instead of one majority, multiple majorities, or no majority, may be created. By placing power in different institutions, with different electoral procedures, and with staggered timings of elections, for fi lling key

Introduction

11

officials, different persons and different majorities (or pluralities) in different institutions may claim democratic mandates and thereby check one another. Proponents of integration and assimilation often make this claim. They celebrate the creation of dual executives (a president, and a prime minister and cabinet), a dual legislature (a house of representatives and a house of states or regions, or an elected house and a house of experienced experts), and autonomous legal institutions (perhaps with both a supreme court and a constitutional court). For such advocates, separating powers is power sharing as well as power-division because they think that the separation of powers in different institutions obliges power-holders to work cooperatively in anticipation of the checking and balancing capacities of the others. One tempting response to this apparently persuasive rhetoric might be to observe that the well-attested division of powers in the U.S. Constitution proved compatible with institutionalized and racialized slavery between 1787 and 1860; racism and a white tyranny over blacks in the deep South from the 1880s until the 1950s; and the termination, expulsion, and demographic erosion of Native Americans in an unfinished process of destruction. The U.S. Constitution has also proven wholly compatible with regular eruptions of persecutions of immigrant and religious minorities. Far from always supporting pluribus, the Philadelphia Constitution has enabled the unum of the dominant people palpably to take charge and avoided neither the tyranny of the majority nor the tyranny of faction. While that rebuttal is in order, it would be too quick a response because it does not address the relevant difficulty: it is a feature of our concepts that the expressions “power sharing” and the “division of powers” can be construed as synonyms by the informed, as well as the less well informed. In constitutional advisory work in Iraq, for the Kurdistan Region between 2003 and 2009, and in Sudan for the United Nations in 2009–10, I experienced the difficulty in the flesh and in texts. “Power sharing” in English would be translated into Arabic and then later translated back into English by another person to confirm that all negotiators and their advisors agreed the meaning of the relevant text. More often than not when translated back, “power sharing” would emerge as “the separation of powers.” There are philological and etymological similarities between sharing and division that explain this conflation: through dividing, after all, people or groups receive shares. But power sharing and the separation of powers are distinct, albeit potentially intersecting, notions. Power sharing mandates both coordinated jointness in shared decision making and autonomy

12

Brendan O’Leary

in group or territorial decision making. The separation of powers, by contrast, does not prescribe a coordinated policymaking system. Rather under the separation of powers, policy and order are expected to emerge (if at all) from the clash of ambitious power-holders scattered across multiple institutions, and the separation is not organized to facilitate group organization. Power-sharers seek to share power across national, ethnic, religious, and linguistic groups through their representatives making joint decisions in executives, legislatures, and judiciaries; power-dividers seek to break up the formation of such groups and to individualize what they disparage as communal politics. As we shall see, power-dividers are mostly integrationists— only centripetalists, whom we shall shortly describe, express the desire to compete as advocates of power sharing. But though they say that they support sharing power, centripetalists rarely wish to institutionalize autonomy among encompassing national, ethnic, linguistic, or religious communities, or to have their representatives jointly share power within an executive; what they want is power sharing among politicians incentivized to be moderate toward others who are different from them. That said, there can be a power-sharing coalition within a federal or central government that respects a formal separation of powers among executive, legislative, and judicial institutions, and between the federal/central government and the regional/local governments, and related variations.

The Competition for Power

In the liberal tradition, found also in its “constitutionalist” wing, there is a distinct focus on how public officials “win” powerful positions, be they executive, legislative, judicial, or bureaucratic. Competition, as in the marketplace, is seen as a way of preventing nefarious monopoly. In this tradition, competition for executive and legislative posts should occur through elections. Liberals, especially outside the United States, are more doubtful about elections to judicial and administrative positions, for which they generally favor competitive meritocratic appointment through professional associations and transparent and reviewable procedures. The minimal definition of representative government is a political system in which officials compete for authoritative positions in free and fair elections for citizens’ votes; in which elected officials hold office for limited terms, make laws, and give orders to unelected officials within constitutional norms that ensure accountability—

Introduction

13

both through the ballot box and recourse to the courts. Given these premises, the competition for power is a sine qua non of democratic government. The division of powers and competition for power are established and intelligent principles. But on their own, advocates of power sharing submit, they are unlikely to calm deeply divided places and may cause conflict. Where the competition for power resembles an ethnic, religious, or linguistic census, elections may not check governments; rather they may encourage the tyranny of the majority. The combination of the separation of powers and the competition for power may also be conducive toward the oppression of national, ethnic, and religious communities. After all, the competition for power expresses or creates majorities—and such majorities may be constructed from national, ethnic, or communal cleavages. A sustained majority from the same community may win control over all major offices and governments—even if the powers of those offices and governments are divided and checked— and then propose and implement discriminatory public policy or biased conceptions of merit. Integrationists and assimilationists often prescribe the division of powers and the merits of the competition for power but often presume that a nation of individuals is in existence, or that one should be built. They may forget (or ignore) that many states are multiethnic and multiconfessional— and that many are pluri-national—and thereby pass over the fact that the competition for power (with or without the division of powers) may be a recipe for conflict in deeply divided places. Sri Lanka has not wanted for a separation of powers or for elections. Neither has Kenya nor Northern Ireland. To commend procedures that advance the position of nationalizing majorities when there are rival national self-determination claims is partisan or, alternatively, utopian. It is partisan when one community seeks to nationalize the state or region in its image on no better claim than might (numbers) makes right; it is utopian when (potentially or actually) antagonistic communities are expected or instructed to fuse. Partisans and utopians have had opportunities many times in the last two centuries, too often producing bloodbaths. That is why many contemporary liberals commend power sharing as a supplementary approach to avoiding the worst outcomes in pluri-national, multiethnic, and multireligious states. Power sharing, however, commends not only the sharing of power but also the division of power and the competition for power. Power sharing should add to rather than subtract from the liberal, constitutional, and democratic experience. Power sharing commends “coalition” as a considered way

14

Brendan O’Leary

of doing things, but not as a wholesale substitute for the division of power or competition for power. If Rudolf Rummel’s calculations in Death by Government are even approximately correct, the last century was the most lethal in human history. Power sharing claims to offer some prospect of reducing domestic lethality and war in human affairs. The argument for power sharing is more sophisticated than acknowledging that what cannot be won on the battlefield is best allocated through a shared forum and a shared executive. Powersharers follow Rousseau’s declared method in The Social Contract that commends taking “men as they are, and laws as they might be,” but because they do not seek just one community they reject Rousseau’s particular proposals as disastrous, namely inalienable, indivisible, and absolute sovereignty, the rejection of partial associations, and one vigorous homogenizing civic religion. Power-sharers do not seek a social contract among a unified people; they seek social contracts for sociability among divided communities or between territorial governments. The first of these possibilities leads to what are called “consociational” directions; the second leads toward territorial or federative power sharing. These two possibilities can be combined in complex forms, as we shall see.

Power Sharing, Participation, Accommodation, and Their Alternatives Power sharing is intended to enhance effective and peaceful political participation, especially by minorities. Ineffective participation by majorities is rarely a major cause of violent conflict in an established democracy, though a majority’s potentially negative reaction to more effective participation by minorities may cause violence—the “backlash” phenomenon. What follows assumes that the relevant majorities are demographically and electorally dominant, an assumption that needs to be relaxed in deeper analysis. All other things being equal, we may assume that more effective political participation is desirable, for example, through electoral mobilization, interest-group or party membership, or through institutionalized functional or territorial autonomy. Power sharing is intended to hold the existing state together with the active participation and consent of its minorities, unlike strategies of genocide, expulsion, partition, and control. It may also bind states together, through

Introduction

15

confederations, to prevent war recurrence over deeply divided places. It should always be considered in the light of the possible alternatives. Genocide and expulsion are terminating and terminal strategies. Secessionists, when reasonable, wish to convert their status into that of majorities inside their own sovereign state. Their leaders and followers logically reject promises of better participation or power sharing in a unitary, a union- or a federal state, either because the historical record shows the promises lack credibility, as for the people of South Sudan, or because the merits of independence outweigh the expected benefits of greater participation in the existing state. Partitionists seek to divide territories to control conflict; in executing this strategy they invariably offend at least one national or ethnic community’s conception of its homeland, often more than one (O’Leary 2007, 2011). When sincere, partitionists promise better participation in all the post-partition entities, promises unfulfilled in the historical record. Controllers, by contrast, seek to organize the dominant and to disorganize the subordinate— and the potentially insubordinate. They wish to block effective participation by minorities in the core regime that matters and to confine whatever participation they enjoy to subordinated social organizations or to co-opted or indeed trivial institutions (Lustick 1979; O’Leary and McGarry 1996, chap. 3). Measured by these alternatives, power sharing merits a hearing, though it is never the sole strategy available or on offer. To understand power sharing fully we must compare it with its apparently reasonable alternatives, not just the grim prospects of genocide, expulsion, partition, and coercive assimilation. To do so, we may distinguish among strategies that respect the equality of human beings and minimal human rights by their degree of tolerance of heterogeneity (in race, language, ethnicity, religiosity, and nationality). At one extreme homogeneity is insisted upon in the aptly named “assimilationist” strategy; at the other end, recognition of heterogeneity may reach the point of accepting that it is better that the state be downsized. Policy and institutional instruments are logically linked to where policymaking elites’ preferences fit on this scale (see Figure 1.1). Assimilationist strategies encourage effective political participation solely on the majority’s terms: Join us and you can be part of us and cease to be a structural minority. They may work when targeted at migrants who have left their homelands and who wish to be assimilated, and where the dominant people welcomes egalitarian assimilation to take place. If the two groups wish to tango, assimilation occurs, either through acculturation, where the minorities adopt the dominant majority’s culture, or through fusion and

16

Brendan O’Leary

Insistence on Homogenization Assimilation

Integration

Recognition of Heterogeneity Accommodation

Disintegration

Figure 1.1. Scaling tolerance of difference among democratic elites.

amalgamation, where a novel collective identity is formed from a blend of the previous ingredients, including through extensive exogamy. Since Alexander the Great’s wedding of his Macedonian companions to the princesses of the Persian empire, formal and ambitious fusion strategies have been rare, and rarely successful. Fusions happen more than they are planned, as in Spanish-speaking America. Both acculturation and fusion have been evident in the Americas since 1492 and part of the histories of settler states elsewhere, such as Australia; but acculturation has also been present in immigrantwelcoming states, such as France was—when it wished to have as many citizens as Germany. Assimilationists emphasize majoritarian institutions for democracy, including single-person executive presidencies and winner-takes-all electoral systems in single-member districts (WTA-SMD), or expressly majoritarian electoral systems such as French-style double-ballot or two-round systems (TRS), or the majoritarian preferential vote in single-member districts, often known as the alternative vote (AV). (The varieties of electoral systems are lucidly appraised in Bernard Grofman’s chapter.) Formally, assimilationists emphasize inclusive civic citizenship: Nous n’avons pas de minorités, mais seulement des citoyens ou des étrangers. More circumspectly, they insist that their states only have political minorities, which are capable of becoming governing majorities if they become electorally persuasive. To establish homogeneity, assimilationists often seek to ban (minority) ethnic, religious, or linguistic parties and encourage the building of “national parties” based on programs, doctrines, and interests (Bogaards 2008). Assimilation is especially beneficial when voluntary, but if one or more party is a reluctant groom or bride, then different forms of coexistence may be demanded. Power sharing is one of these. Integrationist strategies, by contrast, do not seek the full cultural homogenization of minorities. They confine their homogenizing ambitions to the “public sphere,” as its proponents call it (following Habermas 1992; Calhoun 1992). In the public sphere, a common public language, public symbols, and common party politics are preferred. Integrationists are relaxed about the

Introduction

17

private preservation (and indeed the dynamic development of) minority cultures but do not hesitate expressly to bar them from the public domain. A duly elected Gaelic Irish nationalist and self-professed political republican may be admitted to the UK Parliament but must address it in English, if not the Queen’s English, and the republican must recognize the constitutional monarchy before being admitted to the floor of the house. Integrationists seek, at least in their own estimations, to design public institutions—executives, parliaments, courts, public mass media—to be inclusive, even if they recognize that they cannot be entirely culturally neutral. They encourage minority participation in public life; they do not oblige minorities to abandon their cultural identities, but they are expected to leave their cultures at the doors to public places in the way some institutions and peoples expect visitors to leave their shoes. Minorities may be monotheist, polytheist, or atheist in the synagogue, mosque, or club; they may be monolingual, bilingual, polylingual, or nonlingual in the bedroom; and they may be “ethnic,” fused, cosmopolitan, or plain dull in the kitchen; but respect is expected for the codes of the formal public culture. The sincere integrationist seeks to strip the state of as much of its inherited cultural presumptions as possible, but there are limits. Even Habermas’s ideal speech takes place within a language, and at least one person may need translation help (as all do with Habermas). Table 1.1 indicates the preferred goals, forms of party and party system, views of federalism, preferred political institutions, and favored electoral systems typically found among three types of integrationist: a (recognizably French or Turkish) republican, a (recognizably 1960s British Labour) socialist, and (a recognizably current U.S. or Canadian) liberal. Republican and socialist integrationists are strongly hostile to minority ethnic, linguistic, and religious parties, preferring national and class orientations, respectively. They favor party and electoral systems that discourage minority party formation and voting for ethnic, religious, and linguistic minority parties. Republicans and socialists are strong majoritarians. Liberals, however, have feared “tyrannies of the majority,” by which they have usually meant usurpationist labor movements, or religiously superstitious and ill-educated masses, rather than those of nationalizing majorities. Accordingly, they have generally preferred the constitutional division of powers and the creation of multiple “nonmajoritarian” regulatory institutions to unrestrained sovereign parliaments. Republicans and socialists dislike federalism if it weakens the center and its distributive and egalitarian powers; by contrast liberals favor “national federations”—because they may constrain socialists,

Alternative vote (AV) to encourage moderation; list proportional representation (PR) where socialists are strong Winner-takes-all in singlemember districts (WTASMD) or two-round system (TRS) where left is strong; otherwise list proportional representation (PR)

Difference-blind public sector appointments

Note: For discussion of voting systems, see Chapters 2–3.

Winner-takes-all in singlemember districts (WTASMD) or two-round system (TRS)

Affirmative action for the disadvantaged

Majoritarian democracy + Strong president

Preferred Institutions

Electoral System

Separation of powers Parliamentary executive Difference-blind appointments, temporary affirmative action

Majoritarian democracy + Parliamentary executive

Against if it divides the demos

View of Federalism

Against if it reduces the prospects for just redistribution

Majoritarian democracy modified by powerful courts and other delegated institutions

In favor if it enhances markets and protects rights

Programmatic, market regulation

Class-based

Cross-class, statewide, nonethnic; nonreligious; willing to ban ethnic or religious parties

Preferred Parties

Interests of individuals

Interests of labor

Interests of the nation

Liberals

Goal Orientation

Socialists

Republicans

Integrationist Strategies

Key Traits

Table 1.1. Integrationist Strategies Toward Minorities

Introduction

19

protect markets and liberties through power-dispersal, yet emphasize one nation. Accommodationist strategies are synonymous with power-sharing strategies. These favor the formal public recognition and organization of ethnic, religious, and linguistic minorities, as the Latin etymology of “accommodate” and its cognates suggests—namely, ad (toward) and commodare (to make fit), in turn derived from commodus (suitable). Genuine accommodation requires the political recognition of more than one ethnic, linguistic, national, or religious community in the state and aims to secure coexistence, though its proponents may occasionally support downsizing, secession, or even partition, if accommodation is impossible. Academic supporters of power sharing are not, contrary to their critics, “primordialists” who believe that national, ethnic, religious, and linguistic groups have existed since time immemorial. Accommodationists insist, however, that such groups’ identities, and differences from one another, are often resilient, durable, and hard, rather than malleable, fluid, soft, or quickly transformable, as suggested by self-styled “social constructionists,” who are often unconscious of their assimilationist or integrationist biases. Where divisions are enduring, advocates of power sharing believe that attempting either assimilation or integration is unjust, and unfeasible, unless much blood is shed, and that is usually counterproductive. John McGarry and I believe there are four distinct forms of democratic power sharing in real-world strategy and academic commentary, now known as “centripetalism,” “multiculturalism,” “consociation,” and “territorial pluralism.” Each of these sees itself as a mode of power sharing, and each is briefly treated below (see also Table 1.2). The order reflects, from left to right on the scale in Figure 1.1, the willingness of the respective advocates of these strategies to embrace national minorities on their own agendas, with centripetalists being the least enthusiastic and territorial multinational pluralists the most keen. The centripetalists are the closest to integrationists. They see themselves as responsible realists, though they regard some integrationists as wishful thinkers, except regarding policy toward voluntary immigrants. They are skeptical of romantic multicultural celebrations of difference. Centripetalism involves prescriptions first clearly articulated by the political scientist and legal scholar Donald Horowitz and a small number of other scholars who have followed in his tracks. Centripetalism emphasizes “convergence” toward the center, or “bringing together,” and is usually juxtaposed by its

Centripetalists

Conflict reduction Means: Promotion of convergence on moderate center; only as much group recognition as necessary for system maintenance

Nonethnic and cross-ethnic parties to be bolstered Ethnic parties to be incentivized to moderate their platforms Hard-line ethnic parties to be marginalized, and nationalist parties discouraged

Key Traits

Goal Orientation

Preferred Parties

Territorial Pluralists Recognition of multiple nationalities; support for power sharing Means: Territorial selfgovernment and shared rule in the central/ federal government; cross-border institutional linkages No expectation of a statewide party system Constitutional regional nationalist parties welcomed

Consociationalists Group recognition and power sharing Means: Cross-community executive power sharing; proportionality; autonomy; and veto rights

Parties based on cultural groups to be encouraged, but in liberal consociation room is left for nonethnic parties to flourish Favor proportionality and parity as organizing principles

Multiculturalists Goal: Group recognition Means: Proportionality, cultural autonomy

Ethnic and religious parties welcomed, nationalist parties are not

Power-Sharing Strategies

Table 1.2. Accommodationist Power-Sharing Strategies Toward Minorities

Unifying presidencies; dislike quotas and “ethnic preferences”

Alternative vote (AV) or “distributive requirements” to build statewide politics

Preferred Institutions

Preferred Electoral System

Collective executives, typically cabinets endorsed by parliaments but also collective presidencies List proportional representation (PR) or single transferable vote (STV-PR), or “set-asides” or quotas for corporate consociationalists)

Favor proportionality but no developed views

Open to federalism and highly compatible with pluralist federations

No developed views

Open to federalism, but cultural autonomy is their focus

Avoid winner-takes-all in single-member districts if it makes secessionist bid too easy

Collective executives

Very favorable toward federation provided the federation is pluralist (polynational, consensual, and decentralized)

less accommodation more accommodation

Valued if it disperses conflict and if it divides nationalities Prefer national federations to multinational or pluralist federations

View of Federalism

22

Brendan O’Leary

proponents to consociation, which is seen as generating centrifugal or fissiparous politics (in my view, contentiously and often inaccurately). Centripetalists claim majority-rule institutions may be tempered through “vote pooling” to facilitate the election of moderate ethnic politicians capable of reaching out to other ethnic communities (Harris and Reilly 1998; Reilly 2001; Reilly and Reynolds 1999). They counsel against electoral systems that they think favor ethnic partisans and often condemn proportional representation (PR) electoral systems, particularly party-list PR with low thresholds, or the single transferable vote (STV) in multimember constituencies with a significant number of seats because they permit “extremists” to flourish. “Vote pooling,” according to its proponents, should require at least some winning candidates to obtain votes from ethnically different voters and encourage campaigns focused on centrist moderate voters in heterogeneous constituencies. Such systems should “make moderation pay” (Horowitz 1989, 1991). Centripetalists advocate two approaches to electoral system design. The first relies on territorial “distributive requirements,” which is considered especially useful for presidential elections. An electoral college is an example. A majority of the electoral college votes need not be the same as a majority of the popular vote and may obligate successful candidates to focus their campaigns across a range of member states (and, by implication, groups). Horowitz’s favorite example is presidential elections in Nigeria circa 1979. The then new constitution stipulated that a winning presidential candidate needed at least a plurality of the popular vote and at least a quarter of the vote in at least two-thirds of the nineteen states, and tacitly required any competitive presidential and vice-presidential team to appeal across the north and south of Nigeria and, minimally, to at least two of its three largest ethnic communities. The second approach is the frequent advocacy of the “alternative vote,” a majority-preferential voting system that can be used for both legislative and presidential elections: the winning candidate must win an absolute majority of first-preference votes or a majority of votes after the transfer of lower-order preferences from the ballot papers of eliminated candidates (Bean 1997; Lijphart 1990; see also Allison McCulloch’s and Bernard Grofman’s chapters in this volume). Centripetalists support the “conciliatory potential of federalism” because it may undercut support for secessionists. But their reasons show the limits of their willingness to accommodate nationalism among minorities. Federalism, they argue, may disperse power away from the center, thus inhibiting authoritarianism; quarantine confl ict at the local level (such as

Introduction

23

Muslim-Christian disputes over sharia law in northern Nigeria); provide training for future central politicians to engage in ethnic bargaining at the local level; and promote cross-cutting politics through facilitating interand transethnic alliances based on common regional interests and facilitating intragroup divisions and group proliferation within regions (Horowitz 1991, 214, 2007, 1985). On federations, Horowitz’s ideas are close to those of “national federalists,” for example, for both South Africa and Iraq he commended strong political centers. He has also supported the forced partition of federative entities dominated by large ethnic communities, even against their will, to prevent regional majoritarianism. Horowitz commends Nigeria, where I grew up, as a paradigm case: the Nigerian First Republic was (primarily) made up of three regions, dominated by the Yoruba, Ibo, and Hausa-Fulani, respectively. Nigeria’s military dictators subsequently forcibly restructured the federation into twelve, and then nineteen states. (There are now thirty-six states, and perhaps forty-six soon.) Some are dominated by the three large communities, some by other smaller minorities, and some by no single community. None of the big three now commands an integrated contiguous homeland region. In contemporary Iraq, followers of Horowitz recommended a “national federation” based on the eighteen governorates established under Saddam Hussein’s Baathists and hoped that it would be imposed by the American-led Coalition Provisional Authority (Wimmer 2003–4; for similar ideas, see Dawisha and Dawisha 2003 and Makiya 2003). They believed that it would divide the Kurds, Sunni Arabs, and Shiite Arabs among several different governorates respectively, and thereby promote intercommunity and cross-provincial alliances, as well as an inclusive Iraqi national identity. Horowitz himself comes close to integrationists in several respects: he seeks to promote transethnic identities, to disadvantage politicians who focus exclusively on their own communities, and to favor those who “can find ways of transcending their own ethnic affi liations” (1991, 207). Yet it is difficult to see how a single-person presidential executive can fully represent distinct communities, as opposed to representing what they may share, unless the person is of mixed origins, such as President Obama. However, Horowitz is clearly not an integrationist (though some of his followers are): he does not favor engineering social mixing or the wholesale privatization of culture and has argued for bi- and multilingualism, and other forms of public support for minority cultures, in different contexts. He favors at least some self-government for minorities, particularly small ones that do not

24

Brendan O’Leary

threaten the state, though not “national self-government,” and his vote-pooling ideas are sincerely aimed at promoting some interethnic power sharing, albeit through a coalition of moderates, and albeit he strongly prefers transethnic to ethnic parties. Multiculturalists go closer toward meeting the aspirations of minorities; they affirm the desirability of protecting multiple communities in both public and private realms. Recognition is their theme. In most of North America and much of Western Europe, multiculturalism is not practiced literally or pervasively. That would mandate the full promotion of public schools for all cultural minorities in their languages, including programs designed deliberately to allow immigrants to retain their language of origin and other forms of cultural autonomy, such as those advocated by the Austro-Marxists Karl Renner and Otto Bauer and promoted today in a limited way in Hungary, Estonia, and parts of Russia (Hanf 1991; Bauer 2000). Western multiculturalism, by contrast, has aimed, generally, at the promotion of a mainstream public culture that is less one-sided, assimilationist, or conformist than previously. The idea has not been to support immigrants in publicly establishing their own separate political institutions or parties but to tolerate differences in private domains while actively promoting integration and mixing in common public institutions through, for example, programs of English as a second language and the provision of hospital and other public ser vices in the minority language. The object is to assist minorities in adaptation, and is often closer to acculturation than fusion. This is how one may understand decisions to allow Muslim schoolgirls to wear headscarves in public schools in the United Kingdom, to change public school curricula to reflect citizens’ multicultural origins in the United States (Takaki 1993), and to permit Sikh police to wear turbans in Canada. This “multiculturalism” does not advocate public support for cultural communities to remain enduringly viable and separate for the long term. A full-blooded, so to speak, multicultural strategy in Europe and North America, by contrast, would accommodate non-Western and illiberal practices. The American brand of multiculturalism, the absorption of all positive features in the cultures of the world, has been satirized by Stanley Fish as “boutique multiculturalism” (1999), which resists the force of the culture that is allegedly appreciated at precisely the point at which it matters the most to its committed members. Western multiculturalists, according to Fish, tolerate other cultures but only until they challenge liberal principles: child marriage, the inferior status of women in public or marriage laws, or

Introduction

25

clitoridectomy are out of the question. It is true that many Western European multiculturalists display a rather shallow tolerance of immigrants from the eastern part of the European Union, juxtaposing their own allegedly liberal nationalism with eastern or Balkan ethnic nationalism, deemed to be exclusivist, based on endogamous marital practices (“tribal”), hostile to group interaction, and segregationist. Such multiculturalism, we may suspect, is integrationism in disguise and, may, in certain limited circumstances, be no bad thing. “Credible multiculturalism,” by contrast, insists on two arrangements that require some minimum public support, through legislation or expenditures, and that reject the banning of political parties based on ethnicity, language, or religion. Both require some power sharing. First, public “respect” for a group’s selfgovernment in matters the group considers important amounts to parity in recognition; second, some broad application of the principle of proportional representation of all significant groups in key public institutions follows, not necessarily through quotas but certainly through public targets to enable the diverse representation of groups, such as in the military, in elite educational institutions, and the judiciary. The formula, “community recognition + proportionality,” is the minimum mark of credible multiculturalism. Whether autonomy requires public support for some form of autonomy, especially in educational institutions, is controversial: most, however, argue that multiculturalism requires “community autonomy + proportionality.” Consociation is rightly associated with the innovative work of Arend Lijphart, which he acknowledges revives a tradition found in the works of Althusius and indeed the “Austro-Marxists” of the Habsburg empire. To the multicultural formula, “autonomy + proportionality,” consociation adds two further elements, namely, “cross-community executive power sharing” and “veto powers.” A cross-community power-sharing executive joins representative elites from different communities to manage and prevent conflict. Consociation does not require every community to be represented in the executive, or that everybody within a particular community support all of its representatives in the government, or that all ethnic parties be present in the executive. “Complete consociational executives,” what Lijphart describes as a “grand coalition,” may be distinguished from “concurrent executives,” in which representatives of the majority within each of the main partner groups is in government, and “plurality executives,” in which at least a plurality of each significant group’s representatives are in government (O’Leary 2005a; McGarry and O’Leary 2006a).

26

Brendan O’Leary

More rigid consociations, constructed amid high historic mistrust or after deeply violent antagonisms, may endow each partner to the consociation with veto powers, enabling them to prevent executive, legislative, or constitutional decisions that threaten their fundamental interests. Sometimes executive power sharing and proportionality rules grant every major community de facto veto powers. Where they do not, however, minorities usually seek formal veto rights; for example, in Belgium the cabinet has parity of representation of Flemish and French speakers and operates by consensus. Likewise, autonomy sometimes accomplishes de facto veto provisions through the formation of a jurisdiction dominated by a given community that can nullify federal law (though it may be vulnerable to interventionist courts). Consociations do not need to be rigid; “opt-out” provisions that do not block others from pursuing a certain policy and that result in institutional and policy asymmetries may be distinguished from “veto rights.” Consociationalists flesh out the formula of “autonomy and proportionality.” Community self-government is perhaps a clearer description than “autonomy” because territorial autonomy is distinct from “pillarization,” the system Lijphart described as prevalent in the Netherlands until the mid1960s, that is, parallel pillars of Catholics, Protestants, Socialists, and Liberals, “equal but different” (Lijphart 1975). Consociational autonomy rejects the hierarchical ranking of partly self-governing groups, such as those evident in Islamic empires—described in Braude’s chapter—apartheid, and systems of control, but it allows for separate personal laws on marriage and inheritance, separate schooling and university systems, and separate publicly funded media. Provided adult membership and use of such institutions is voluntary, liberal consociation is possible. Territorial autonomy is generally supported by consociationalists (Lijphart 1979), but since it is distinctly characteristic of governance in federations or union states, it is examined below. The combination of territorial autonomy with consociational arrangements at the confederal (or federal or union center) is not uncommon; it notably may be used to describe the European Union (Chryssochoou 1994, 1997, 1998; Taylor 1993). Autonomy in consociations mandates public support for the maintenance of diverse communities, both now and into the future, but, contrary to its critics, it does not prevent voluntary de-pillarization; it does require that the pillars de-pillarize by their own consent. Consociationalists commend proportionality throughout the critical components of the public sector—not just the executive but also the legislature, the judiciary, administration, the police, and the army; that is, they

Introduction

27

insist on “representativeness” in all salient political institutions. Professionalism and merit-based impartiality occur, and are legitimate but in deeply divided places are only feasible once representativeness has been met. Consociationalists prefer PR electoral systems or, where that is not possible, systems that achieve similar outcomes, for example, “set-asides” or “quotas.” They differ over the merits of par ticular PR systems—or, rather, over which ones should be encouraged in particular cases. Lijphart generally favors listbased proportional representation (list PR) because it facilitates discipline and control by party leaders and eases the making and maintenance of consociational bargains. But it may have drawbacks: with a low threshold, list PR may enable hard-liners to wreck consociational deals because they can form their own party and win support without reducing the vote and seat share of their ethnic bloc. The STV version of PR has a higher effective threshold than most forms of list PR, given that constituencies rarely comprise more than six members, and therefore may inhibit fragmentation and facilitate transfers in favor of candidates and parties prepared to maintain power sharing (O’Leary 1999a, 1999b, 2002; see also Evans and O’Leary 1997a, 1997b, 2000; Mitchell, O’Leary, and Evans 2002, 200; Mitchell, Evans, and O’Leary 2009). Consociation can be democratic or authoritarian; formal or informal; liberal or corporate (McGarry and O’Leary 2007). In an authoritarian consociation, such as the Yugoslav federation, a communist elite from each nationality jointly controlled the federal government but was not democratically representative. A democratic consociation, by contrast, has open elections among competing elites, but each sizable community is represented in the executive along complete, concurrent, or plurality patterns. A formal consociation is entrenched by constitutional or statutory law, as in the 1998 Northern Ireland Act, but consociation may be organized through informal convention, for example, the Swiss “magic formula” for allocating seats on the federal collective presidency (Steiner 1970). A corporate consociation accommodates groups according to ascriptive criteria and tacitly assumes that group identities are and should be fi xed and that groups are both internally homogeneous and externally bounded (e.g., in Lebanon, BosniaHerzegovina, Belgium, and Cyprus at various junctures). Critics assail such constitutional primordialism. A liberal consociation, by contrast, allows groups (and individuals) to self-determine their organization and representation. It rewards whatever salient political identities emerge in democratic elections, whether these are ethnic, religious, linguistic, or based on other

28

Brendan O’Leary

criteria. South Africa’s transitional power-sharing arrangements (1994–96) allocated deputy presidencies and cabinet seats based on party strengths in seats won, not on ascriptive characteristics. Northern Ireland’s 1998 Agreement combines both corporate and liberal elements. Its first and deputy first ministers are guaranteed to one nationalist and one unionist. The other ten ministers, by contrast, win office through an allocation algorithm, the d’Hondt method, which is liberal or “difference blind” (O’Leary, Grofman, and Elklit 2005). Consociationalists generally commend formal over conventional arrangements, especially when deep divisions generate security preoccupations. Academic consociationalists are supporters of democratic and liberal consociations, though they are (unfairly) accused by integrationists of being “undemocratic,” “illiberal,” “essentialist,” and “primordialist.” Territorial pluralists, by contrast, seek to manage heterogeneity through a pluralist federation or union. A pluralist federation is polynational, is decentralized, and has consensual or consociational decision making in its federal government. Its internal boundaries respect nationality, ethnicity, language, or religion—it is “polynational.” Where the federation-wide majority is a majority in every federative entity, there is an integrated national federation, as in the United States, Germany, Australia, or the Latin American federations. Where all, or most, of a minority are a self-governing majority within its own single federative entity, as in Quebec, we have by contrast a pluralist federation. Between these polar types there exists an ambiguous federation, consisting of ethnic, linguistic, or national minorities divided across several federative entities, in each of which they may be majorities. Where such federations originate from the partition of nations without their consent, as occurred in Nigeria in the late 1960s, they should be categorized as national federations, and sometimes as not-so-disguised systems of “control,” intended to block secession. Where, by contrast, such a pattern developed organically, as with Switzerland’s language communities, two of which are divided across several cantons, we have a federation that is polylingual but not polynational. Pluralist federations vary in their pluralism. Fully pluralist federations entail three complementary arrangements. The first is significant and constitutionally entrenched autonomy for federative entities, both in the constitutional division of powers and in resource allocation. The federal government cannot unilaterally rescind the powers of the federative entities. Second, a pluralist federation may have consensual, indeed consociational, rules within the federal government, including its executive (O’Leary 2005b).

Introduction

29

Consensual federations create strong second chambers representing the constituent regions and have strong regional judiciaries and a regional role in the selection of federal judges. They do not create strong single-person executive presidencies or senates that are mirror images of the house of representatives. Third, they are fully polynational, with a pluralist rather than monist conception of sovereignty—flowing from multiple peoples not one, recognized in the constitution/treaty, flag(s), and symbols and in official bilingualism or multilingualism. The federation is multihomeland, a partnership between or among distinct peoples. A polynational federation often permits asymmetric institutional arrangements. Pluralist territorial accommodation extends beyond the federation format. Some “union states” recognize historic nationalities and their boundaries. This is the apt way to characterize the United Kingdom of Great Britain and Northern Ireland, the Kingdom of Spain and the Kingdom of Denmark, and perhaps India. In each case, jurists and constitutional tradition privilege a centralized sovereignty and treat autonomy as a rescindable gift of the central political institutions, but the state is or has become a composite that respects historically incorporated territories and grants them extensive autonomy and sometimes national recognition. In union states, territorial institutions of self-government are more likely to exist across only part of the state. Such asymmetry may result from majorities being happy with government from the state’s central authorities while minorities insist on distinct self-government (McGarry 2010a). Thus Zanzibar has home rule within Tanzania, but there is no analogue to Zanzibar’s institutions in Tanganyika: the Union parliament is its parliament. The UK has home rule institutions in Northern Ireland and Scotland, but Westminster is the sole parliament for England, and the sole body entitled to pass primary statutes for Wales, which has a National Assembly with limited powers to amend them. Denmark grants home rule to Greenland and the Faeroe Islands, but the common parliament is also Denmark’s. When autonomous and asymmetric institutions are entrenched, through the constitution or an international treaty, there may exist a “federacy,” namely a unit of government that enjoys a distinctive federal relationship with the state (the core of which may be a federation, a union state, or a unitary state), in which neither governmental unit can unilaterally alter the other’s powers (see Alfred Stepan’s chapter in this volume). Some academics and leaders of nationalities argue that what we may call standard territorial pluralism is not enough: it is too statist and cannot adequately meet the demands of mobilized national communities partitioned

30

Brendan O’Leary

by, or spread across, state borders, such as the Azeris, Basques, Irish, Kurds, Magyars, and Serbs. They are not satisfied by consociational or pluralist territorial accommodation and seek to establish or reestablish political linkages across state borders, that is, cross-border institutions, ranging from functional cooperation to confederal bodies. Irish nationalists in Northern Ireland, Magyar minorities in Romania and Slovakia, the Turkish minority in Cyprus, Serbian minorities in Bosnia-Herzegovina, and Austrian minorities in Italy (South Tyrol) seek or have cross-border links with their “kin-state.” Comprehensive territorial power sharing encompasses interstate as well as intrastate institutions.

Normative Disputes over Power Sharing Accommodation, whether called power sharing or not, is unpopular among advocates of a strong central government. Nevertheless, Belgium, Spain, the United Kingdom, Italy, Indonesia, and even France have moved toward systems that accommodate minorities through autonomy, whether through pluralist federations, devolution within pluralist union states, or federacies (e.g., Aceh, discussed by Stepan). States, and international organizations such as NATO, the OSCE, and the UN, have been prepared to encourage, or even impose, consociation, pluralist federations, and federacies in situations of conflict and ethnic polarization, including in Northern Ireland, BosniaHerzegovina, Iraq, Burundi, and, abortively, in Cyprus. On occasion, international organizations, including the Council of Europe and European Union, have moved beyond an integrationist defense of individual rights toward a more accommodationist perspective. Their reluctance to go further is not difficult to explain. The “international” system is composed of states, not nations, and power sharing is most likely to be sought by sizable nationalities that do not control their states, or perhaps only parts of their homelands. What unites advocates of power sharing is the public and private recognition of ethnic, linguistic, religious, or national categories and groups (see Table 1.2). They are prepared to address the demands and aspirations of such communities, if necessary at the expense of the politics of indivisible sovereignty, classes, or individuals. Centripetalists are positioned toward the integrative end because of their restricted focus on ethnic moderates rather than radicals, their preparedness to partition territorial units belonging to large nationalities, and their support for politicians who “transcend”

Introduction

31

their ethnic origins. Credible multiculturalists strongly value cultural pluralism per se and not just as a prudent strategy for achieving political stability. Consociation is much more accommodationist because it is inclusive toward radicals and because it may incentivize politicians whose appeal is restricted to their distinct community. It is more accommodative than multiculturalism because it combines the multicultural institutions of autonomy and proportionality with the additional devices of executive power sharing and vetoes. Territorial pluralism is at the most accommodative end when it converts minorities at the statewide level into majorities within regions. The most accommodative forms of territorial pluralism encompass not just self-government and power sharing in the federal or union government but linkages between minority-dominated regions and their kin in other states. Just as republicanism is the most likely of the integrationist strategies to become coercive assimilation, so territorial pluralism is the most likely of the accommodationist strategies to lead to state disintegration, but the placement of accommodation approaches to the left of the disintegration arrow in Figure 1.1 should remind us that all power-sharing strategies in a broad sense are aimed at holding states together. Integrationists charge that power sharing will increase instability, reinforce divisions, and cause violence. They assume that narrowly rational political elites sow or maintain divisions and that accommodating them will enable them to consolidate their position. Institutionalizing particular identities reinforces allegiance to such elites and may promote resentment among those who are not privileged. Integrationists think that ethnocentric elites will not reach—or at least not maintain—centripetal, consociational, or territorially pluralist bargains because the underlying divisions remain intact; they claim that consociations have broken down or are held together by external force; and they argue that pluralist federations have had a poor track record. Integrationists marshal such cases to argue that power sharing is unworkably brittle—and dangerous. They maintain that integration is more feasible, sometimes arguing that ethnic identities are seldom as longstanding or as deep as accommodationists suggest. Advocates of power sharing are accused of believing in “ancient hatreds,” exaggerating the homogeneity of ethnic groups, ignoring the existence of multiple and crosscutting identities and the capacity of people to change. Integrationists, by contrast, believe it is realistic and sensible to support progressive, or “bridging,” social forces against divisive elites and to support parties with crosscutting appeals.

32

Brendan O’Leary

Advocates of power sharing respond that promoting integration amid deep diversity creates at best an unstable equilibrium; either assimilation will occur after much blood is spilled, or communities will demand special consideration, such as power sharing or, if that is unavailable, secession. Without public support, weaker cultures or communities either acculturate, as hoped by some policymakers, or they react with a backlash, seeking public redress, subsidies, the institutionalization of their culture in the curriculum, political autonomy, or power sharing. Advocates of power sharing are usually skeptical that ethnic identities are universally malleable or superficial, or of recent vintage. They argue that the presence or otherwise of deep divisions is a matter of empirical verification and judgment, not to be dismissed through slogans such as “the myth of ancient hatreds.” Voters in deeply divided places are unlikely to accept standard integrationist protections and may be reluctant to place their faith in cross-cutting political parties, even if the electoral system deliberately makes it difficult for minority parties to win office. It is unrealistic, in such settings, to place much faith in the transformative ability of civil society organizations to counter elite ethnocentrism, precisely because civil society organizations are likely to be ethnocentric—“bonding” rather than “bridging” in character (McGarry 2001), with neither civil nor societal relations. Deeply divided places render the goal of social mixing problematic. Segregated schools, neighborhoods, and workplaces sometimes exist because they are, however regrettably, popular, and partly because they are considered safe by parents, residents, and workers. Deliberate “mixing” may be unrealistic within any short period or only likely to occur within stable power-sharing institutions that generate confidence in the survival of one’s own community. Advocates claim that the track record of power sharing is much better than integrationists allege, but there are notable and heated internal divisions as to which accommodationist institutions work best. For example, those who commend territorial pluralism maintain its success depends on a number of specified conditions being met. Failures are not evidence of an inherent flaw, they are often exaggerated, and long periods of success in staunching flows of blood are overlooked by the critics. Failures may flow from remediable design flaws, external interference, extraneous issues connected with managing the transition from war to peace, the coercive origins of a federation or union, or authoritarian rather than accommodative practices. It is an error, territorial pluralists believe, to hold up the former despotic communist federations as evidence that pluralist

Introduction

33

federations are bound to fail (McGarry and O’Leary 2009a). Accommodationists believe that Euro-federalists who seek to strip the European Union of its pluralist institutional characteristics are doomed to ignominious failure: the EU can work only if it combines territorial pluralism and the accommodation of its constituencies within its central institutions (O’Leary 2001a). There is an intense intra-power-sharing debate between centripetalists and consociationalists over what is most conducive to political order, and I am not a neutral observer in this debate. Horowitz argues that in a consociation it is unlikely that ethnic group leaders will agree to compromise sufficiently both to achieve and maintain consociational power-sharing institutions. He insists that because consociation is necessarily based on the idea of a “grand coalition” it cannot work: successful consociations are “as rare as the Arctic rose” (Horowitz 2002a, 197). Consociations are said to work, if at all, where they are not necessary, that is, where there are only mild divisions: they “are more likely the products of resolved struggles or of relatively moderate cleavages than they are measures to resolve struggles and to moderate cleavages.” Horowitz argues PR electoral systems may exacerbate divided places because they favor sectional appeals and radicals over moderates, facilitate political fragmentation, and promote “outbidding.” Horowitz also often argues for the alternative vote, but these arguments have not won widespread acceptance. He understands that his centripetal approach faces an “implementation problem,” that is, antagonistic elites are unlikely to accept institutions biased in favor of moderate politicians. He believes that this catch-22 can be overcome with the assistance of outside forces. In 2000, Horowitz suggested that the way out of the dilemma created by the failure of ethnic elites to accept centripetal institutions was for internal parties to put “constitutional decision-making in other hands,” including outside governments and international organizations (277). More recently he called for a “strong American push” in a centripetalist direction in America’s reshaping of Iraq, noted that “departing colonial powers left their imprints on new constitutions all over Asia and Africa, and many of these proved durable,” and argued, “[it] is time for the U.S. to do the same” (2005 A16). Centripetalists maintain that their implementation problem is no greater than that of consociationalists because both arrangements require intervention. The difference, they claim, is that centripetal institutions, once implemented, will be self-sustaining and require minimal intervention. Consociationalists regard centripetalism’s emphasis on vote pooling as based on unrealistic assessments of ethnic divisions, especially where these

34

Brendan O’Leary

have already deepened. They have demonstrated to their satisfaction that Horowitz’s claims for the properties of the alternative vote are theoretically and empirically unwarranted, especially in the case of Fiji (Fraenkel and Grofman 2004, 2006a, 2006b). Consociationalists argue that political leaders in polarized polities are far more likely to agree to consociational than to centripetal deals because they guarantee their group/party some direct share in power; they therefore reject the claim that consociation lacks incentives for cooperative conduct. Politicians are more likely to agree to PR than centripetal electoral systems because they can win votes and seats on their own preferred platforms. Proportionality norms enhance stability because they better match the rival parties’ respective bargaining strengths and their conceptions of distributive justice. Consociationalists think in consequence that there may be less need for external actors to play a maintenance role than in centripetal systems, which rig the political system toward unrepresentative moderates. Furthermore, PR electoral systems pose fewer barriers to entry for new political forces than do plurality and majoritybased alternatives, so they may contribute to long-term stability by allowing integrationist forces to emerge organically. Consociationalists flatly reject the merits of single-person executive presidencies. They point to the polarizing effects of the presidential institution in Sri Lanka, not modified by the supplementary vote (a truncated form of the alternative vote), and favor parliamentary and cabinet governments, or collective presidencies. They observe that many voluntary coalitions based on moderates collapse, faced with outbidding and destabilizing opposition from excluded radicals. Inclusion in power-sharing coalitions, on the other hand, may make radicals less extreme because it provides them with opportunities to have their concerns addressed and a stake in the system: tribunes can become consuls (Mitchell, Evans, and O’Leary 2009). It is not true, they observe, that consociations have uniformly failed: consociationalists point to the absence of violence in Bosnia and Herzegovina since 1995, in Lebanon for much of the period since 1943, and in Northern Ireland since 1998; they have also staked claims for Iraq and the European Union, while Lijphart has made them for India. As for Arctic roses, consociationalists ask why Sri Lanka, Fiji, and Papua New Guinea are coded as successes by centripetalists, whereas they treat most consociational cases as failures. Centripetalists and consociationalists obviously disagree on much, in par ticular on whether conflict regulation should be driven by building up

Introduction

35

moderates among the majority (centripetalism) or by building up moderates among hard-liners from all groups (consociation), but they are at one in insisting that democratic states in which ethnic, linguistic, and religious groups are politically mobilized into parties and civic associations are eventually obliged to experiment with elements of power sharing or federation. Integrationists claim that group recognition and public power for an ascriptive group may lead it to repress its own members, as happens in religious communities that discriminate against women (ignoring that statewide majorities are as capable of bad conduct). They argue such arrangements privilege or reinforce closed identities. Such arrangements are deemed unjust because they create a hierarchy of citizenship; the same is claimed of pluralist territorial arrangements, which are said to lead to the unfair treatment of regional minorities. Americans are particularly likely to make the latter argument, recalling how Southern whites used “states’ rights” to maintain slavery and then the Jim Crow segregationist regime. Yet, they also accept Native Americans’ rights on reservations, that is, they accept some “homeland rights.” Many critics of the communist federations argued that they privileged “titular” nationalities (as if “national federations” or unitary states do not privilege the titular nationality). Pluralist federation is also sometimes condemned as unfair if it interferes with the allocation of the state’s common wealth and resources. Advocates of power sharing respond that integrationist rhetoric frequently hides dominant interests under veneers of neutrality or impartiality. Dominant communities generally champion integration, while minorities generally prefer power sharing. Privatizing culture, accommodationists argue, is inherently biased against weaker cultures. Such reasoning is easiest to see in language disputes. Established languages, lingua francas, merely through resolving coordination difficulties, are likely to drive out less widely used languages, unless there is public protection for them. The same holds for integrationist state structures, executive design, and public sector composition. Unitary states and national federations will favor dominant communities, as will majoritarian single-person executives and public sectors with difference-blind composition rules. Integration, in other words, is often merely assimilation with good manners. Some multiculturalists go further and insist integrationism is a form of Western colonialism (Tully 1995). Liberal advocates of power sharing argue that territorial pluralism or consociations do not necessarily require corporate privileges. Many minority nationalities have “civic,” not just “ethnic,” identities and are more likely

36

Brendan O’Leary

to develop the former when in secure possession of their own territorial units. Quebec, Catalonia, Scotland, and Kurdistan are all no more likely, and perhaps far less likely, to repress or control minorities than are the broader states of which they are regions. What is important is not, therefore, to deny territorial autonomy to stateless nations but to ensure that such nations implement power-sharing arrangements for their own minorities, especially when they are large in proportion to the region. Liberal consociationalists grant that the privileging of particular communities occurs in corporate consociations but insist that such institutions may (or should) be made liberal, for example, by rewarding any party with electoral support rather than specified groups. Rules and mechanisms that are ex ante fair to all groups, including nonethnic groups, may be better for small groups than integrationist electoral or executive institutions. Consociationalists prefer “selfdetermination to pre-determination,” as Lijphart has put it (see Lijphart 1995 passim). They understand that parties to power-sharing pacts may entrench pacts that institutionally represent (and privilege) certain identities, but they do so because they have genuine existential anxieties about the security of the communities they represent. They observe, however, that it is sometimes possible to protect groups without corporatist consociation but recognize that liberalizing is more likely to occur after security anxieties are reduced. Consociationalists also respond with a “tu quoque”: centripetal vote-pooling institutions unfairly privilege the majority or largest group; politicians from such groups have to pool fewer votes to win office than do politicians from smaller groups, and that is especially evident with “distribution requirements.” Territorial pluralists reject the idea that justice necessarily requires centralized control of resources or a central government with a monopoly of, or even the preponderant share of, fiscal resources. Central control may be appropriate where citizens share a common national identity or are willing to regard themselves as part of a common insurance pool, but where no such unified nation or demos exists, the promotion of fiscal or resource centralization may be deeply divisive; it may encourage assimilationist fantasies in the capital city and secessionist movements in the periphery. Where the center has historically been a source of despotic abuse, regions are unwilling to grant the federal government significant fiscal powers until it has earned trust. That said, territorial pluralism may be impossible unless there is some distribution of resources across the state to enable some minorities to exercise autonomy; but, on the other hand, minorities that see themselves as

Introduction

37

penalized by unfair distributive mechanisms may protest or seek to secede. Power sharing has complex balancing acts to perform—adequate redistribution to make territorial pluralism feasible but without the centralization that renders it meaningless. Integrationists believe power-sharing arrangements undermine democracy. Republicans think that all types of federation, national or pluralist, are “demos-constraining,” interferences with the formation of the general will, and they regard consociations as the surrender of public power to ethnic or religious interest groups. Some integrationist critics, particularly on the left, focus on the consociational practice of negotiations among political elites and see such summit diplomacy as the negation of participatory democracy. Liberals condemn what they regard as the anticompetitive nature of consociational politics and regularly cite Lijphart’s one-off but unfortunate description of consociation as government by “elite cartel.” Furthermore, if everyone is in government, integrationists charge, how can governing parties be held to account? Consociational inclusion, they say, surely renders opposition politics impossible. Centripetalists share these concerns, arguing that vote-pooling coalitions, unlike grand coalitions, can be based on minimum-winning coalition government formation rules and therefore are consistent with significant opposition and competitive politics. Advocates of power sharing share the core position that majority rule in divided places is likely to be partisan, even when it is padded out with integrationist or power-dividing safety mechanisms. The electoral systems favored by integrationists, they observe, usually do not even produce majority rule but rather minority or plurality factional rule (from within the governing party) (Nagel 2000). Consociationalists observe that the alternative vote in single-member districts eliminates small parties (likely to be supported by minorities) and artificially boosts the legislative majority of the winning party—which may facilitate the abuse of power. An advantage of PR, by contrast, is that it makes it more likely that there will be a genuine legislative majority supported by a majority of voters. In any case, they observe, the Westminster model of democracy does not necessarily provide strong oppositions: plurality rule can convert electoral minorities into unassailable legislative majorities, and sometimes reduce the opposition to an ineffective rump that excludes important minorities, an outcome that happens frequently in Caribbean democracies. The accusation that consociational executives lack democratic opposition is only (partially) accurate when leveled at grand or complete consociational executives. In fact, in most alleged

38

Brendan O’Leary

cases of grand coalition small parties have been in opposition. Consociational executives may also be concurrent (restricted to parties commanding majority but not total support within their respective segments) or plurality based (if one of the segmental parties in office commands only plurality support within its community). They may also be based on some but not all segments of a divided place. In all such cases, significant parties exist outside the government to criticize its policies. In any case, mechanisms can be deployed to render the government more accountable, through empowering backbenchers to hold the ministers of other parties to account through interpellation and enhancing the powers of independent auditors, watchdogs, ombudsmen, and courts—and other nonmajoritarian institutions. Consociational institutions can, however, also foster competitive politics. If seats in the executive are based proportionately on seats in the legislature, then there are incentives for all parties to compete, within or across ethnic groups, to increase their share of legislative and executive office. Consociation’s supporters happily concede that their preferred system depends on the capacity of leaders to make compromises and to persuade their followers to accept such compromises, but they insist that effective settlements require leaders who are authentic representatives. Consociationalists, it must be said, are often skeptical about the merits of extensive participatory democracy. They observe that active participation from civil society may simply lay bare deep divisions and make it difficult to achieve agreement, but that said, the shaping of consociational settlements, their ratification, and their aftermaths need not be the exclusive property of political elites. In small Northern Ireland, the radical policing reform that followed the 1998 agreement resulted from a widespread consultation process that involved 2,500 written submissions and public meetings in every district council area (attended by 10,000 people and at which 1,000 spoke). The 1998 Agreement itself was ratified by simultaneous referenda in both parts of Ireland. If this settlement survives, it will owe no small part of its durability to its original popular endorsement. Lastly, it is likely that the benefits of deliberative democracy depend on the participants sharing a common language and frame of reference. That may be more attainable within linguistically homogeneous substate communities than in diverse polylingual states. To sum up, democratic states are limited normatively, and practically, in their responses to diversity. Within the past half century or so, integration and power sharing have internationally begun to displace, normatively and

Introduction

39

even in lip ser vice, rival nationalizing state strategies based on assimilation, control, partition, or ethnic expulsion. There is, however, a global predilection for integration, as well as some support for power sharing, across countries and public intelligentsia, and some evidence of growth in such support. Both integration and power sharing have empirical and normative merit in particular contexts. Integration’s appropriateness is facilitated when social cleavages are fully cross-cutting rather than reinforcing, that is, when each cleavage is of roughly equal potential salience. Integration is more likely when societies are not deeply polarized along national, ethnic, religious, or linguistic lines, that is, when there is already extensive hybridity and mixing. Integration may also be successful with minorities that are small in number and interspersed among others—and well-disposed to the strategy. A small and dispersed minority usually cannot realistically aspire to public recognition through territorial autonomy or consociation, and it is difficult to build winning coalitions out of such minorities. Territorially dispersed groups find it difficult to mobilize to defend their culture, even if technological and media revolutions make this easier than it once was. Integration is workable when the minorities involved are not “homeland people” because they are less likely to see themselves as national communities, entitled to some form of self-government. Not all migrant communities have the same potential to integrate. Those whose religion mandates bringing the sacred into the public domain find it more difficult, especially if the public domain has been defined by the public agenda of another religion or if the state is already secular. Likewise diasporas that maintain contact with their homelands and entertain a return of the exiles are less likely to integrate. There is also some evidence that the integration of many European immigrants into the United States, Argentina, Australia, Canada, and New Zealand in the nineteenth and early twentieth centuries was eased because they lacked what Ernest Gellner called a “high culture” (1983) and therefore treated emigration as social mobility. Many contemporary immigrants in wealthy democracies, by contrast, are professionals with cultural and educational capital from foreign universities who can maintain their “high culture.” Successful integration also depends on the willingness of the dominant community to accept the partial privatization of its culture and to accept new members in its political community. Even large, territorially concentrated minorities may be integrated if the offer of equal citizenship, with public laws that are equal and nondiscriminatory, is made early and credibly, but once a minority has become nationally

40

Brendan O’Leary

mobilized in reaction to blocked integration, it is likely too late for it to be won back except through power sharing. Lastly, integration may be facilitated if the state lacks a dominant community and if, as a consequence, its single public identity is genuinely composite. Countries that are very diverse and have no Staatsvolk, such as Tanzania, may find it easier to promote integration than those with two or three major communities, for example, Iraq or Nigeria. Power sharing is appropriate wherever integration cannot function well. Territorial power sharing works best for and is most sought by the middlesized battalions concentrated in their own national space. Territorially concentrated groups living on their own homelands reject integration as a policy of letting their culture die, but the demand for autonomy is likely to come not just from large territorially concentrated groups. Many Corsicans have called for autonomy from France; native communities in Canada have demanded autonomy since the late 1960s after the Canadian government, influenced by the American civil rights movement, tried to integrate them into a common Canadian citizenship. Large, nationally mobilized minorities are likely to insist not just on autonomy but also on power sharing within the federal government. The situation of the Quebecois, the Turkish Cypriots, and the Kurds of Iraq is different in this respect from that of the Corsicans, Moldova’s Gagauz, or the South Tyrolese. Their choice is not between integration and power sharing but between power sharing and secession. Territorially dispersed and interspersed minorities, including small ones, especially where politics has been deeply polarized, demand entrenched consociation—especially where the relevant minority has sufficient bargaining power. Burundi’s Tutsi and Cyprus’s Turks have such power, for now. In deeply divided places, it is doubtful that Horowitz’s preferred centripetal institutions can result from free negotiations between the parties representing mobilized ethnic, religious, or linguistic groups. That is not so of consociation, which, as Lijphart has correctly argued, is a recurring invention of politicians across time and place. Where nations spill over or have been partitioned by state borders, neither consociation nor pluralist federation may be sufficient to achieve a functioning accommodation. Northern Ireland’s peace agreement required institutions that linked the British and Irish governments, and Northern Ireland with the Irish Republic. In most contexts, such trans-state settlements are difficult to imagine, but they may become more feasible when the relevant states participate within an overarching confederation, such as the EU.

Introduction

41

That integration and accommodation are more feasible in some contexts than others helps explain why both strategies are adopted by democratic states and why governments may follow different strategies with different groups. Many states follow integrationist policies for immigrants and powersharing policies toward nationally mobilized communities. The international community usually preaches integration but has been prepared to back power sharing where that has been demanded, if only, and unfortunately, after rebellion has occurred.

Testing Power Sharing: Key Cases and Large-N Studies Sharing power to create fair and inclusive democracies after civil war or regime change is a political formula that is currently being tested in many places, including Iraq and Northern Ireland. If power sharing eventually fails in these places—and I am not suggesting it will—would that count as decisive evidence against the merits of promoting power sharing? Power sharing is a key component of current efforts to rebuild Cyprus as a unified federation. Any positive political renewal in Lebanon will adapt rather than abolish its power-sharing heritage. The prospective new constitution for Nepal, if a text is negotiated to a successful conclusion, will have power-sharing provisions. But how can we tell whether the primacy of power-sharing ideas to resolve these deeply divided places is well-founded? Both exponents and opponents of power sharing typically answer these questions through instructive examples. Proponents of power sharing argue that Northern Ireland, Bosnia and Herzegovina, South Tyrol, and Belgium, for example, would be in much worse shape without power sharing: civil war and secession (or partition) are the feasible alternatives. Opponents of power sharing typically argue that it just does not work. When confronted with examples of its success they insist that power sharing cannot work for long, and they cite breakdowns in Northern Ireland, Lebanon, and Cyprus. In any case, they say, power sharing is perverse because it entrenches the causes of conflict (group identifications and antagonisms) or jeopardizes key political principles (e.g., the equality of individual citizens). We should not disparage the importance of telling and detailed treatment of examples and cases as merely anecdotal—there is certainly no good reason to discount detailed knowledge of particular countries or regions. But we should all agree that the social scientific study of power sharing is

42

Brendan O’Leary

more advanced when it is both comparative (evaluating cases and non-cases of power sharing in a controlled way) and comprehensive (gathering as much evidence as possible to code like with like from across the world to build robust generalizations). Fortunately, within the fields of comparative politics and international relations, numerous large-N (many case) studies of power sharing are beginning to emerge. Moreover, the results of such studies are sufficiently convergent to enable us to have confidence in the merits of power-sharing prescriptions. Just three among numerous comparative and large-N studies will be briefly reviewed here: the swan-song book of Arend Lijphart, now an emeritus professor at the University of California at San Diego; the book of Dr. Pippa Norris, a lecturer at Harvard University; and an article of Drs. Michaela Mattes and Bruce Savun from the Universities of Vanderbilt and Pittsburgh, respectively. These suggest, seriatim, that power-sharing democracies work better than non-power-sharing democracies, that regimes with more power-sharing traits do better than their antonyms, and that power sharing reduces the likelihood of civil war recurrence. Lijphart. Arend Lijphart, a former president of the American Political Science Association and a U.S. citizen of Dutch origin, is the best-known analyst and advocate of power sharing. In Thinking About Democracy (2008) Lijphart collates some of his best articles and essays of the last thirty years. Within this collection, confirming his earlier comparative work on democracies, he demonstrates through a systematic evaluation of thirty-six democracies over thirty years that consensual (or power-sharing) democracies consistently outperform (or do no worse than) majority-rule (or winnertakes-all) democracies over a whole range of public policy performance criteria. Not merely do they regulate conflict and disorder better, but they also perform in superior ways across measurable economic outputs (growth, employment, and inflation rates) and in achieving a kinder and gentler form of democracy (in respect, for example, of women’s rights and rights to welfare). Power sharing pays off within democracies. Norris. Pippa Norris is of English origin but is a denizen of Harvard and has often worked with the United Nations. In her recently published (2008) book she asks, “Do power-sharing institutions work?” The answer is positive. She has coded the political development of most regimes in the world since the 1970s in a truly global large-N analysis. Her statistical studies strongly suggest that power-sharing institutions increase the likelihood of successful democratic governance and that they do so after statistically controlling for

Introduction

43

other factors that may also explain such performance (such as previous levels of economic development, relevant colonial history, or the relevant regional neighborhood). Power-sharing institutions are shown positively to affect conflict avoidance, maintain political stability, and increase governmental performance across a range of measurable and reasonable criteria (including human rights protection and control of corruption). Norris uses three dimensions to measure the degree of power-sharing incentives within regimes: the presence of proportional representation electoral systems, which enhance the likelihood of coalition or inclusive governments; the presence of parliamentary institutions, which make it more likely that the executive will be collegial rather than dominated by a prime minister or president; and the presence of federal or decentralized arrangements, which weaken the prospects of political monopoly throughout the state. These three dimensions are consistent with the arguments advanced in Lijphart’s work over the last thirty years and with how we have appraised power sharing above. Norris has also demonstrated the importance of pluralist media, though the relationships between this variable and other power-sharing mechanisms are not as clear; however, pluralist media are obviously essential if there is to be community self-organization. What is truly startling in Norris’s work is that nondemocratic regimes also perform better against performance outputs the better they score on power-sharing dimensions. Norris combines her large-N analysis with paired comparisons of carefully chosen and unusual case studies (particularly for Anglophone scholars). She examines the comparative performances of Benin and Togo, South Korea and Singapore, India and Bangladesh, and Ukraine and Uzbekistan, and successfully shows that that the presence or otherwise of powersharing institutions matters in explaining their respective governmental performances. Mattes and Savun. One of the quickest ways in which busy readers may see how large-N studies are confirming the merits of power-sharing prescriptions is to examine a recent article by Michaela Mattes and Burcu Savun (2009), both scholars of international relations, a subfield that increasingly overlaps with comparative politics precisely because most current wars are primarily civil wars rather than interstate wars. These authors are especially interested in the causes of war and of war recurrence. Mattes and Savun study the aftermaths of forty-six civil wars, which ended with peace agreements, between 1945 and 2005. They are especially interested in an argument advanced by Barbara Walter, which emphasized

44

Brendan O’Leary

the role of third parties and third-party guarantees in creating durable political settlements (Walter 1997), a result that is congenial to those who emphasize international over domestic factors in explaining political stability within countries. They note, however, that others have emphasized the importance of power-sharing and postwar military organization as crucial factors in explaining political stability. What Mattes and Savun therefore do is test the importance of all these factors and others in explaining the likelihood of civil war recurrence. Their results strongly suggest that elements of power sharing (separately and jointly), after controlling for other factors, lead to conflict reduction (less war and more peace). They confirm that third-party guarantees are helpful in reducing former belligerents’ fears that the other party will renege on its commitments, and they confirm that what they call “cost-increasing provisions,” that is, costs imposed on militaries, have strong conflict-reduction consequences, either through the separation of military forces (which makes renewal of fighting more costly) or through military integration (in which former rebels find a place in the new security order). Mattes and Savun code four dimensions of power sharing in peace agreements. The first is political (present if there are either proportionality or group parity provisions in the texts, or if there are quota allocations of public posts, e.g., in the civil ser vice or the police). The second is territorial (federal or autonomy arrangements). The third is military, which we have briefly discussed above. The fourth is economic: they examine peace agreements to see if they embed redistributive or wealth-sharing provisions in favor of the disadvantaged. What is especially interesting is that they find only the political and military dimensions of power sharing to be strongly statistically significant in reducing the likelihood of civil war recurrence: the territorial and economic dimensions do not achieve statistical significance. This is an interesting result for both international policymakers and advisors because it suggests that a focus on political power sharing and on separating or integrating militaries is crucial for sustaining peace agreements. Why no impact was found for territorial or economic dimensions of power sharing will no doubt lead to further research—though we know from the works of Lijphart and Norris that these dimensions, especially decentralization, positively shape the performance of established democracies. In the decade ahead we can expect further scholarly work on the properties of power-sharing systems and that both large-N work and work focused on controlled case studies will be done to test their effectiveness in affecting

Introduction

45

governmental performance and achieving peace and stability. The fact that, so far, large-N studies show that we can be confident that power-sharing matters, in making democracies perform better and in reducing the likelihood of civil war recurrence, is very important. However, the results come with no guarantees or simple recipes and are subject to future scholarly revision. Power sharing is not a panacea, but it is a political formula with some proven success in our times. The critical task of local negotiators of new democracies or new constitutions in deeply divided places, and those who advise them, is to build custom-designed power-sharing systems that are appropriate for the local antagonisms and neighborhoods. One size certainly does not fit all, but at least we know we should counsel policymakers, political parties, and peace activists to look at the power-sharing shelves when confronted with potential or actual violence in deeply divided places.

The Contributions in This Volume Readers disposed to explore the expanding library shelves and online work devoted to power sharing in deeply divided places should find within our contributors’ chapters realistic, imaginative, and critical ways of thinking more deeply about the subject. We have divided their contributions into three parts. In Part I, “Power Sharing and Electoral Systems,” three chapters focus on the role of electoral systems in promoting or inhibiting interethnic accommodation. In Chapter 2, Bernard Grofman provides a crisp appraisal of voting systems and their implications for power sharing. We do not need to display modesty on Grofman’s behalf: this is an exceptionally important essay from a master in the field of political science. It distills and merges what is important from two literatures, which are normally never integrated (I was tempted to say which never share power), namely, the formal studies of the properties of voting or decision rules and the comparative study of actual electoral systems. Electoral system advisors and concerned citizens will benefit tremendously from its admirable combination of comprehensiveness, concision, and fairness. In Chapter 3 Allison McCulloch provides an important overview of the claims of centripetalists that their prescriptions reduce conflict. In a judicious appraisal she finds these claims overstated. Drawing on evidence from Republika Srpska, Sri Lanka, and Fiji, McCulloch suggests that amid deep division, the alternative vote, centripetalists’ electoral system of choice, is unlikely to foster moderation. She argues that

46

Brendan O’Leary

preexisting heterogeneity and moderation are needed at high thresholds for centripetal incentives to be effective. The implication is that centripetalism can only positively shape places that already have moderate leaders and moderate followers, that is, places that are not deeply divided. Th is chapter may be a lasting contribution to knowledge cumulation in the social scientific study of power sharing, and I should emphasize that McCulloch’s contribution is not confined to centripetalist proposals on electoral system design. It should be read in conjunction with Chapter 4, that of Kris Deschouwer and Philippe Van Parijs’s “Electoral Engineering for a Stalled Federation.” They provide an imaginative proposal to make Belgium’s federal and consociational arrangements work better through a tweaking of the country’s electoral arrangements. Their proposal arguably matches the ethos of both the consociational and centripetalist schools. The core idea is to create a multimember district of seats with a pan-Belgian character but with internal quotas of speakers. The goal is to generate more politicians who will look to the interests of the federation as a whole while preserving parity and proportionality principles among Flemings and Walloons. In Part II, “Historical and Conceptual Forays into Power Sharing,” there are four contributions. In Chapter 5 Ronald Wintrobe considers whether rational choice can provide a plausible model to explain sustained moderate or conflict-regulating behavior in deeply divided places. In an original twist he adapts his prize-winning model of rational extremism to explain how much violence or accommodation occurs between contending groups (Wintrobe 2006). Wintrobe illustrates his model with two key events: the construction of the wall between Israel and the West Bank and the Israeli withdrawal from Gaza. He discusses how his model suggests an increased likelihood of accommodation can be realized—including through reframing, the use of rewards and sanctions, the marginalizing or the accommodation of extremists, and federalism. Students and general readers will find his an entirely fresh perspective, especially on Israel and Palestine—where original arguments are now rare. We stay in the Middle East in the next chapter, but Benjamin Braude takes us back in time in “The Success of Religion as a Source for Compromise in Divided Empires: Ottoman and Safavid, Past and Present.” Together with Bernard Lewis, Braude is a pioneer in the study of pluralism and religious autonomy in the history of the Ottoman empire (see Braude and Lewis 1982). Here he compares the two major empires of the pre-twentieth-century Middle East, contrasting Sunni and Shiite Islam in building legitimate and inclusive political institutions. He provides a perfect

Introduction

47

entry for those interested in subsequent power-sharing successes and failures among ethnic and religious communities in the former Ottoman and Safavid territories, for example, Israel, Cyprus, Lebanon, Syria, Iraq, and Iran. Next, in Chapter 7 Randall Collins provides a panoramic vision of the role of the international system in the likelihood or otherwise of the adoption of collegial or power-sharing institutions by states. Among other important conclusions, Collins suggests that the elites of great powers are much more likely to be reluctant to embrace collegial or consociationalstyle decision making for themselves than are the elites of small powers because of their respective security preoccupations. In Chapter 8 Samuel Issacharoff concludes this part by providing us with a lawyer’s perspective on how constitutional courts can act as a safeguard against majoritarianism in deeply divided places. Treating the South African example, he points to the role of courts as enforcers of constitutional commitments that constrain majoritarianism. Rather than the formal arrangements of consociation, which he regards as too rigid for the long term, Issacharoff focuses on “constrained democracy,” in which constitutional courts are key players in ensuring “strong constitutionalism.” He examines critical decisions of the South African Constitutional Court, which was entrusted with the power to ensure that the final constitution conformed to the thirty-four principles set out in the interim constitution agreed between the African National Congress and the National Party. The certification decision concerning the proposed constitution for post-apartheid South Africa is a particularly important example of the role that courts can play in limiting majoritarianism. Issacharoff emphasizes the possibility of stabilizing and constraining democratic governance without—or after—formal power sharing in deeply divided places but emphasizes that courts need to adopt an assertive stance for this to transpire. In Part III, “Contemporary Power-Sharing Questions,” there are four country-based studies with theoretical implications for power sharing and two treatments of deeply divided cities, one comparative, the other focused on the city of Kirkuk. In Chapter 9 Alfred Stepan achieves two goals. He clarifies the precise conceptual status of a “federacy,” a concept first issued by the late Daniel Elazar that has since enjoyed an underground existence in comparative politics (Elazar 1987; see also O’Leary 2005b, 2002; Rezvani 2007, 2003) and in which discussions of the Åland Islands between Finland and Sweden have always enjoyed a special currency. Then Stepan illustrates the inspirational role of federacy thinking (and the Åland Islands’ precedent)

48

Brendan O’Leary

in resolving conflict in Aceh. We have not heard the last on this concept, or its application, but Stepan’s essay will now be an authoritative source. He has added to the repertoire of constitutional design for deeply divided places. In Chapter 10 Joanne McEvoy moves us toward consociational rather than federal theory. She investigates the operation of the mutual veto, one of the four elements in Lijphart’s formulation of consociational democracy (Lijphart 1969). Veto powers are considered vital in providing group protection after conflict (or before a conflict breaks out), especially when interethnic trust may be negligible. Following an overview of the veto in the political science literature and its historical antecedents, McEvoy offers a framework encompassing the who, what, and where of the mutual veto (corresponding to veto players, veto issues, and veto points). She employs this framework to compare the operation of veto rules in Bosnia and Herzegovina and Northern Ireland. While veto rights are crucial for groups to feel secure under new political arrangements, she argues that specific veto procedures can make a difference in how much cooperation is fostered. A more refined mutual veto may offer more potential for cooperation by specifying the scope of the issues groups may be able to veto and the extent of veto procedures. In her assessment, veto players are preferably self-determined (rather than predetermined) groups; veto issues should be clearly defined in legislation and limited to identity questions, particularly the domains of symbols, language, culture, and education; and a complex set of veto rules at various points in the political process should be avoided. The next chapter is Ed Cairns’s “Northern Ireland: Power Sharing, Contact, Identity, and Leadership.” He suggests that analyses of power sharing would do well to consider insights from social psychology, particularly research on the “contact hypothesis.” Focusing on cross-community contact and leadership as important factors in the facilitation or obstruction of power sharing, Cairns explores insights from social identity theory and social categorization theory in Northern Ireland. The key question for him is whether elites who cooperate across the group divide can lead by example and encourage intergroup contact. He identifies a challenge for powersharing institutions: some community leaders may be more interested in group cohesion and the preservation of group identity than in pursuing intergroup contact to break down barriers. He maintains there is a policy need to focus beyond contact among ethnic elites to the promotion of contact among the general population, particularly in more disadvantaged places. For Cairns, power sharing is not a panacea but may at least have potential to

Introduction

49

increase intergroup contact, which may, in time, lead to the development of a shared identity. In Chapter 12 Colin Irwin, who established his distinction as a pollster in deeply divided places during key stages of the Northern Irish peace process, investigates the role of public opinion polling in the facilitation of the negotiation and implementation of power-sharing systems. With evidence from several cases—Northern Ireland, the Western Balkans, Kashmir, and Sri Lanka—he demonstrates the potential that “peace polls” may hold for political elites who are willing to consider negotiating a political settlement. In Northern Ireland, Irwin describes how testing opinion on variations on power-sharing, relative to other options, informed the interparty talks preceding the drafting of the Good Friday Agreement of April 10, 1998. Drawing on this experience, similar methods have been applied in the other places to test public attitudes toward modifications of various constitutional arrangements. Irwin argues that in Kashmir a peace poll showed the potential of a local/regional peace process, but that requires positive input from key figures in India, Pakistan, and internationally. A Sri Lankan peace poll also pointed to public support for power sharing, including devolution, though that was before the Tamil question in Sri Lanka was settled, at least for now, by the victory of the Sinhalese-dominated army over the Tamil Tigers. For Irwin, peace polls hold considerable potential in providing an indirect voice for the electorate at the negotiating tables but also as advisory mechanisms for negotiators. He does, however, stress the severe challenges in persuading governing parties to embrace their use. Florian Bieber assesses the various power-sharing systems in the Western Balkans as a “testing ground” for power sharing in Chapter 14. He has worked in the terrain for over fifteen years and has the credentials of a scholar and a policy advisor. Exploring the challenges of securing stable and effective power sharing in these cases, he points to four key factors: the role of international agents; decision-making rules including veto rights; forms of territorial autonomy; and the “thickness” of the state. He observes how difficult it is for power sharing to work when there is a lack of consensus on the state among the elites from different groups. For Bieber, however, much of the difficulty in making power sharing work in the Balkans lies with external powers—in both the design and operation of power sharing and the lack of consistency and follow-through in international organizations. He discusses how international agents affect legitimacy, particularly when core institutions and policies have been imposed on at least one community, observing that

50

Brendan O’Leary

sustained imposition undermines the potential for intergroup compromise, particularly when an international agent has the power to make decisions against the will of all parties. Arbitrators who become unaccountable sovereigns deny the meaningful self-government that must underpin power sharing. He reasons that the “thickness” of the state is an important factor for successful power sharing—if states are too “thin” or “minimalist,” groups may fail to invest in them. He reminds us of the limitations of any power-sharing design when one or more groups have, at best, a weak commitment to the state—as Bosnia has taught us for over fifteen years. Our final contributions are from Scott A. Bollens (“Governing Polarized Cities”) and Liam Anderson (“Power Sharing in Kirkuk: The Need for Compromise”). A planner and urbanist, in Chapter 14 Bollens provides an overview of conflict-regulation failures and successes in numerous ethnically, linguistically, and religiously divided cities from Belfast to Kirkuk and in between. He does what has not been done much within political science—he examines urban ethnic conflict regulation through the prism of the broader, usually state- or region-based explanatory theories of power-sharing success and failure. His chapter should provide an inspiration to others to expand his field. Anderson has already taken to that field. He concludes our volume by addressing a major disputed city, Kirkuk, which resembles many of those surveyed by Bollens, and he considers the prospects for compromise. Kirkuk is contested by four ethnic groups (Kurds, Arabs, Turkmen, and Assyrians) and at least three religious groupings that partly cut across these distinctions (Sunni Muslims, Shiite Muslims, and Christians). The city, astride major oil fields, is widely held to be a tinderbox, likely to blow up in the aftermath of the U.S. withdrawal from Iraq. In a subtle, robust, and modestly presented argument, Anderson employs federal and consociational thinking to project what a workable and fair power-sharing compromise over Kirkuk city and governorate (province) should look like. His arguments deserve to be heard within and outside Kirkuk.

Conclusion Nation-states, based on monistic conceptions of sovereignty, are, by disposition, integrationist or assimilationist, and they remain the role models in world politics. The impetus for power sharing comes not from generous nation-states but from those whose boundaries, placements, and numbers

Introduction

51

make them unhappy or repressed within the existing configurations of nation-states. Estimations of power usually decide whether a government will pursue integration or power sharing toward a particular community. The mobilization capacity of a community partly shapes its orientation toward either integration or power sharing—integration, usually, is the politics of the historically weak or the newly arrived, whereas power sharing may become the policy of those powerful enough to resist assimilation but not strong or united enough to achieve secession. Astute regimes promote power sharing from prudence, not generosity, when they recognize multinationality or multiethnicity or multilinguality as brute facts that cannot be altered except at unacceptable costs. They adopt power sharing to forestall violent politics or secession, and occasionally after the defeat of violent movements. Minorities may underestimate or overestimate their bargaining power—for example, pursue integration when they could achieve power sharing; or pursue power sharing when the regime is likely to offer them integration. Their success will depend on their political appraisals of the state that confronts them and exploiting opportunities that come their way. In making their estimates and appraisals the chapters that follow should help. They contain repertoires of ideas that governments and guerrillas would do well to ponder as alternatives to the costs of repression or revolution. Notes 1. The distinction is owed to Bachrach and Baratz 1963. 2. In some languages this contrast has different names, which makes them less confusing than their English equivalents: puissance and pouvoir in French; Kraft and Macht in German. 3. An exchange where at least one person is better-off and no one worse off is a “Pareto improvement.” 4. The potential rule of an unrepresentative faction within Westminster-style democracies has been ably set out by Nagel (2000). 5. After the ancient Romans terminated their monarchy they put the former power of their kings into the hands of a pair of magistrates who held office for one year. “The two consuls . . . [were] the head of the government, but policy was largely directed by ex-consuls,” the most prominent of whom would be called the principes civitatis (Syme 1974 (1939), 10). A helpful survey of interpretations of the Roman Republic and a defense of its democratic character may be found in Millar 2002. The Spartan dual kingship, checked by the council of ephors and by senators, as described in the writings of Herodotus, Plato, Thucydides, Plutarch, and Polybius, was another ancient dual executive intended to prevent the monopoly (and corruption) of power. The elaborate

52

Brendan O’Leary

procedures for the election of the doge of Venice ensured widespread consensus; see Finlay 1980, 141ff. For an introduction to the theory and practice of Calvin’s Geneva, start with MacCulloch 2004, 195–97, 237– 47. 6. For a critical overview, see Kuper 1999. 7. For example, any currently used language is usually in some state of change, e.g., in the growth, transformation, or decline of its vocabulary, syntactical rules, or norms; and it is influenced by other languages, and influences other languages; and it will share some lineages with other languages. All this is so, but it should not lead us to conclude that to speak of languages (and their users) is to reify, essentialize, or primordialize. Not all is flux. Likewise, orga nized religions are usually in some state of change. Consider debates on the status of women within Christianity, Judaism, and Islam, in which rules and practices are being negotiated, renegotiated, or imposed. Such changes, and the multivocal positions adopted within each religion, do not mean that these religions have no continuous attributes. 8. This requirement need not preclude power sharing between states over a deeply divided place, in a condominium; see, e.g., O’Leary et al. 1993. 9. This assertion takes issue with the emphasis on “social capital,” inspired recently by Robert D. Putnam (1995). The “social capital” school focuses on societies’ capacities to explain developmental differences between regions and states, including in political participation, citizens’ conceptions of their efficacy, and democratic development. The notion of social capital has merits—and difficulties—but it is states that determine the trajectories of their societies, including their levels of social capital, for good or ill. 10. Without order, we allegedly have Thomas Hobbes’s condition of “Warre”: “the time men live without a common Power to keep them all in awe . . . in which every man is Enemy to every man; . . . wherein men live without other security, than what their own strength, and their own invention shall furnish them. . . . In such condition, there is no place for Industry; because the fruit thereof is uncertain: and consequently no culture of the Earth, no Navigation, nor use of the commodities that may be imported by Sea; no commodious Building; no Instruments of moving, and removing such things as require much force; no Knowledge of the face of the Earth; no account of Time; no Arts; no Letters; no Society; and which is worse of all, continuall feare, and danger of violent death; and the life of man, solitary, poore, nasty, brutish and short.” Hobbes 1991 (1651), 88–89. 11. Here I draw upon O’Leary 2008, first written in 2002. 12. See Montesquieu 1989 (1748). Montesquieu famously mischaracterized key features of eighteenth-century Whig Great Britain as resembling the “mixed government” of Polybius’s Rome—with monarchical, aristocratic, and democratic features. See Polybius 1979. For strengths and weaknesses in Montesquieu’s thought, see Aron 1965 and Althusser 1982. For a powerful analysis of the strengths and weaknesses of Madison’s approach to democracy, see Dahl 1957, chap. 1.

Introduction

53

13. In negotiations in multiple languages it is good practice to divide translation activities among multiple translators to achieve concision and to identify and resolve ambiguities. 14. See Schumpeter 1987 (1944) for such an approach. 15. This was the bleak conclusion of a famous book that argued that plural societies could not function as democracies; see Rabushka and Shepsle 2009 (1972). 16. In Jalal-Alabad and Osh in the summer of 2010 on assignment for the United Nations I witnessed the build-up of what was excused as a preemptive backlash but what was in effect a pogrom, which ranged the titular nationality, the Kyrgyz, violently against the Uzbeks of Kyrgyzstan. 17. Americans regularly assimilate “majority” with “dominant” groups, and “minority” with “subordinate” groups; it is an error to do so. Dominant minorities and formerly dominant minorities in par ticu lar (such as southern white slaveholders or Sunni Arab Muslims in Iraq) are often major sources of conflict. 18. There are limits to desirable liberalism and pluralism; participation is not always benign, and overtly racist, chauvinist, theocratic, fascist, and Pol Pot–style communist parties, for most of us, would be unwelcome features of “deep diversity.” 19. What follows draws freely from our continuous taxonomic elaborations and evaluation of the strategies available to political elites intent on eliminating or managing national and ethnic differences: McGarry and O’Leary 1993; O’Leary and McGarry 1995, 2012; O’Leary 2001b. 20. Not infrequently, assimilation has started from a core minority, e.g., in Republican France or Kemalist Turkey. 21. The reverse also occurs, especially among conquering migrants: “Romans” assimilated into Greek culture in the Eastern Roman Empire, now known as Byzantium; see Millar 2006. The Mongol conquerors of China became Chinese; see Saunders 1972. The Normans became French, English, Welsh, Irish, and Sicilian; see Davies 2004, 1990; Gillingham 2003; Thomas 2003. 22. There are many useful discussions of assimilation, often focused on the United States and France: Alba 1990, 1999; Brubaker 2001; Glazer 1993; Horowitz and Noiriel 1992; Portes and Rumbaut 1990; Simpson 1968; van den Berghe 1981; Zolberg and Woon 1999. 23. A lucid account of liberals’ preference for nonmajoritarian or expert regulatory institutions may be found in Majone 2005. 24. For a neoclassical economics defense of federalism, see Weingast 1997; for the Austrian school of economics defense of the market-enhancing properties of federalism, see Gillingham 2003, channeling von Hayek 1973. 25. Facile rejections of primordialism are frequent and otiose because the existence of the thought crime is rare within the academy, though it is common among mass publics who are not available for academic cognitive therapy. Useful correctives on primordialism include Gil-White 2001, 1999 and Horowitz 2002b.

54

Brendan O’Leary

26. A critical summary of the literature on AV and other systems from the perspective of national minorities may be found in O’Leary 2010. 27. For criticism, see the contributions in O’Leary, McGarry, and Salih 2005. 28. Horowitz is also one of the finest political sociologists of ethnic conflict and not a facile social constructionist. He stresses the potential durability of ethnic divisions and denies that catchall parties or civic associations can easily supplant their ethnic counterparts once they have come into existence. 29. Hegel and a modernized Catholicism provide the underpinnings of the most famous Anglophone philosophical exponent of recognition; see Taylor, 1975, 1992. 30. See especially the edited collection on his work and his own books: Crepaz, Koelbe, and Wilsford 2000; Lijphart, 1977, 1985, 1996, 1999, 2008. 31. Although Lijphart originally identified a “grand coalition” in which all key communities are represented as the primary indicator for consociation, arguably what matters is some element of jointness across the most significant communities; see O’Leary 2005a. Lijphart has agreed on this matter in correspondence with the author (2004). 32. In many lunches with Paul Taylor at the London School of Economics & Political Science we compared our consociational analyses of the European Union and Northern Ireland. 33. STV, however, is only appropriate for literate and numerate electorates. 34. A critical overview of normative debate over consociation may be found in O’Leary 2005a, reviewing charges of perversity, jeopardy, and futility in the spirit of Albert Hirschman; see also Taylor 2009. 35. See the discussions in Mabry et al. 2013. 36. For a critical review of such claims in the case of Iraq, see McGarry and O’Leary 2007. 37. John McGarry and I are often condemned as primordial pessimists; see some of our critics in Taylor 2009 and our argument and response in McGarry and O’Leary 2009b, 2009c. 38. The Kurds of Turkey manage to coordinate to support a Kurdish minority party despite the existence of a high threshold and the formal outlawing of ethnic parties. 39. Henry Hale has argued that the failure of Nigeria’s first republic and the disintegration of the Soviet Union did not flow from pluralist federation per se but from the existence of a “dual power structure” in which a single “core ethnic region” dominated the federation. His recommendation is that institutional designers focus on dividing the largest group; he presents no evidence against granting minority groups control over their own regions (2004, 2005). 40. The expression, I think, was first used and elaborated by Rabushka and Shepsle (2009 [1972]). 41. With John McGarry I have written extensively on the current experiments in Iraq and Northern Ireland. 42. I review the rhetorical arguments over power sharing in O’Leary 2005.

Introduction

55

43. I would have liked to have reported on the dissertation of Dr. Brighid Brooks Kelly, “An Exploration of the Relationship Between Consociationalism and Stability in Plural Societies Throughout the World,” Trinity College Dublin, winner of the Basil Chubb Prize 2008 for the best politics Ph.D. thesis submitted to an Irish university in 2007. But it is not yet ready in book form. 44. The same result is found in Bingham Powell’s comparisons of proportional and majoritarian democracies; see Powell 1988 (1982), 2000. 45. The debate generated by these proposals can be followed in English at http:// www.rethinkingbelgium.eu/: Look for E-Book 4, which has the same title as the authors’ chapter in this volume. 46. See Wintrobe 1998; Breton and Wintrobe 1986; Breton et al. 1995. 47. During World War I Max Weber (1994) anticipated this thinking, arguing that a Machtstaat, as he wished Germany to be, required strong discretionary core executive institutions; he implied that those who wanted a peace-loving, powersharing democracy would condemn themselves to the international irrelevance of Switzerland. 48. For further work on the courts and variations in power-sharing systems, see McCrudden and O’Leary 2013. 49. For a survey of the concept across the history of the social sciences, see Forbes 1997. 50. See Irwin 2006. Shifts in public opinion and their relationships to consociational bargaining are also discussed in O’Leary 1992; Evans and O’Leary 2000; and Mitchell, Evans, and O’Leary 2009. 51. Others have called the international interventions imperial and neocolonial; see, e.g., Knaus and Martin 2004; Chandler 2006. 52. Together with Professor Gareth Stansfield, Anderson has treated the subject at monograph length, reinforced by extensive fieldwork; see Anderson and Stansfield 2009. References Alba, Richard. 1990. Ethnic Identity: The Transformation of White America. New Haven, CT: Yale University Press. ———. 1999. “Immigration and the American Realities of Assimilation and Multiculturalism.” Sociological Forum 14, 1: 3–25. Althusser, Louis. 1982. Montesquieu, Rousseau, Marx: Politics and History. London: Verso. Anderson, Liam, and Gareth Stansfield. 2009. Crisis in Kirkuk: The Ethnopolitics of Conflict and Compromise. Philadelphia: University of Pennsylvania Press. Aron, Raymond. 1965. Main Currents in Sociolog ical Thought. New York: Basic Books. Bachrach, P., and M. S. Baratz. 1963. “Decisions and Non-Decisions: An Analytical Framework.” American Political Science Review 57: 632– 42.

56

Brendan O’Leary

Bauer, Otto. 2000. The Question of Nationalities and Social Democracy. Ed. Ephraim Nimni. Trans. Joseph O’Donnell. Minneapolis: University of Minnesota Press. Bean, Clive. 1997. “Australia’s Experience with the Alternative Vote.” Representation 34, 2: 103–10. Bogaards, Matthijs. 2008. “Comparative Strategies of Political Party Regulation.” In Political Parties in Conflict-Prone Societies: Regulation, Engineering and Democratic Development, ed. B. Reilly and P. Nordlund. Tokyo: United Nations University Press. 48– 66 Braude, Benjamin, and Bernard Lewis, eds. 1982. Christians and Jews in the Ottoman Empire: The Functioning of a Plural Society. Vol. 2: The Arabic-Speaking Lands. New York: Holmes and Meier. Breton, Albert, Jean-Luigi Galeotti, Pierre Salmon, and Ron Wintrobe, eds. 1995. Nationalism and Rationality. Cambridge: Cambridge University Press. Breton, Albert, and Ronald Wintrobe. 1986. “The Bureaucracy of Murder Revisited.” Journal of Political Economy 94, 5: 905–26. Brubaker, Rogers. 2001. “The Return of Assimilation? Changing Perspectives on Immigration and Its Sequels in France, Germany, and the United States.” Ethnic and Racial Studies 24, 4: 531– 48. Calhoun, Craig. 1992. Habermas and the Public Sphere. Cambridge, MA: MIT Press. Chandler, David. 2006. Empire in Denial: The Politics of State-Building. 2nd ed. London: Pluto Press. Chryssochoou, Dimitris N. 1994. “Democracy and Symbiosis in the European Union: Towards a Confederal Consociation?” West European Politics 17, 4: 1–14. ———. 1997. “New Challenges to the Study of European Integration: Implications for Theory-Building.” Journal of Common Market Studies 35, 4: 521– 42. ———. 1998. Democracy in the European Union. New York: Tauris. Connor, Walker. 1972. “Nation-Building or Nation-Destroying?” World Politics 24: 319–55. Crepaz, Markus M. L., Thomas A. Koelbe, and David Wilsford, eds. 2000. Democracy and Institutions: The Life Work of Arend Lijphart. Ann Arbor: University of Michigan Press. Dahl, Robert A. 1957. A Preface to Democratic Theory. Chicago: University of Chicago Press. Davies, Rees R. 1990. Dominion and Conquest: The Experience of Ireland, Scotland and Wales, 1100–1300. Cambridge: Cambridge University Press. ———. 2004. “Nations and National Identities in the Medieval World: An Apologia.” Belgisch Tijdschrift voor Nieuwste Geschiedenis 34, 4: 568–77. Dawisha, Adeed, and Karen Dawisha. 2003. “How to Build a Democratic Iraq.” Foreign Affairs 82, 3: 36–50. de Tocqueville, Alexis. 1988 (1835, 1840). Democracy in America. Trans. George Lawrence. Ed. J. P. Mayer. New York: Harper Perennial.

Introduction

57

Domhoff, William. 1990. The Power Elite and the State. New York: Aldine de Gruyter. Elazar, Daniel J. 1987. Exploring Federalism. Tuscaloosa: University of Alabama Press. Evans, Geoff rey, and Brendan O’Leary. 1997a. “Frameworked Futures: Intransigence and Flexibility in the Northern Ireland Elections of May 30, 1996.” Irish Political Studies 12: 23– 47. ———. 1997b. “Intransigence and Flexibility on the Way to Two Forums: The Northern Ireland Elections of 30 May 1996 and Public Opinion.” Representation 34, 3– 4: 208–18. ———. 2000. “Northern Irish Voters and the British-Irish Agreement: Foundations of a Stable Consociational Settlement?” Political Quarterly 71, 1: 78–101. Finlay, Robert. 1980. Politics in Renaissance Venice. New Brunswick, NJ: Rutgers University Press. Fish, Stanley. 1999. “Boutique Multiculturalism.” In The Trouble with Principle, 56–74. Cambridge, MA: Harvard University Press. Forbes, H. D. 1997. Ethnic Conflict: Commerce, Culture and the Contact Hypothesis. New Haven, CT: Yale University Press. Fraenkel, Jon, and Bernard Grofman. 2004. “A Neo-Downsian Model of the Alternative Vote as a Mechanism for Mitigating Ethnic Conflict in Plural Societies.” Public Choice 121, 3/4: 487–506. ———. 2006a. “Does the Alternative Vote Foster Moderation in Ethnically Divided Societies? The Case of Fiji.” Comparative Political Studies 39, 5: 623–51. ———. 2006b. “The Failure of the Alternative Vote as a Tool for Ethnic Moderation in Fiji? A Rejoinder to Horowitz.” Comparative Political Studies 39, 5: 663– 65. Gellner, Ernest. 1983. Nations and Nationalism. Oxford: Basil Blackwell. Gil-White, Francisco J. 1999. “How Th ick Is Blood? The Plot Thickens . . . : If Ethnic Actors Are Primordialists, What Remains of the Circumstantialist/Primordialist Controversy?” Ethnic and Racial Studies 22, 5: 789–820. ———. 2001. “Are Ethnic Groups Biological ’Species’ to the Human Brain? Essentialism in Our Cognition of Some Social Categories.” Current Anthropology 42, 4: 515–55. Gillingham, John. 2003. The English in the Twelfth Century: Imperialism, National Identity and Political Values. Woodbridge: Boydell Press. Gillingham, John. 2003. European Integration, 1950–2003: Superstate or New Market Economy? Cambridge: Cambridge University Press. Glazer, Nathan. 1993. “Is Assimilation Dead?” Annals of the American Academy of Political and Social Science 530: 122–36. Habermas, Jürgen. 1992. “Citizenship and National Identity.” Praxis International 12, 1: 1–19. Hale, Henry E. 2004. “Divided We Stand: Institutional Sources of Ethnofederal State Survival and Collapse.” World Politics 56: 165–93. ———. 2005. “The Makeup and Breakup of Ethnofederal States: Why Russia Survives Where the USSR Fell.” Perspectives on Politics 3, 1: 55–70.

58

Brendan O’Leary

Hanf, Theodor. 1991.”Reducing Conflict Through Cultural Autonomy: Karl Renner’s Contribution.” In Uri Ra’anan, Maria Mesner, Keith Armes, and Kate Martin, eds., State and Nation in Multi-Ethnic Societies: The Breakup of Multi-National States, 33–52. Manchester: Manchester University Press. Harris, P., and B. Reilly. 1998. Democracy and Deep-Rooted Conflict: Options for Negotiators. Stockholm: International Institute for Democracy and Electoral Assistance. Hobbes, Thomas. 1991 (1651). Leviathan (or the Matter, Forme & Power of a CommonWealth Ecclesiastical and Civill). Ed. Richard Tuck. Cambridge Texts in the History of Political Thought. Cambridge: Cambridge University Press. Horowitz, Donald L. 1985. Ethnic Groups in Conflict. Berkeley: University of California Press. ———. 1989. “Making Moderation Pay: The Comparative Politics of Ethnic Conflict Management.” In J. P. Montville, ed., Conflict and Peacemaking in Multiethnic Societies, 451–75. Lexington, MA: Heath. ———. 1991. A Democratic South Africa? Constitutional Engineering in a Divided Society. Berkeley: University of California Press. ———. 2000. “Constitutional Design: An Oxymoron?” In Ian Shapiro and Stephen Macedo, eds., Designing Democratic Institutions, 253–84. New York: New York University Press. ———. 2002a. “Explaining the Northern Ireland Agreement: The Sources of an Unlikely Constitutional Consensus.” British Journal of Political Science 32: 193–220. ———. 2002b. “The Primordialists.” In Daniele Conversi, ed., Ethnonationalism in the Contemporary World: Walker Connor and the Study of Nationalism, 72–82. London: Routledge. ———. 2005. “The Sunni Moment.” Wall Street Journal, 14 December. ———. 2007. “The Many Uses of Federalism.” Drake Law Review 55: 953– 66. Horowitz, Donald L., and Gérard Noiriel, eds. 1992. Immigrants in Two Democracies: French and American Experiences. New York: New York University Press. Irwin, Colin. 2006. “The Northern Ireland ’Peace Polls.’ ” Irish Political Studies 21, 1: 1–14. Knaus, Gerald, and Felix Martin. 2004. “Lessons from Bosnia and Herzegovina: Travails of the European Raj.” Journal of Democracy 14, 3: 60–75. Kuper, Adam. 1999. Culture: The Anthropologists’ Account. Cambridge, MA: Harvard University Press. Lijphart, Arend. 1969. “Consociational Democracy.” World Politics 21, 2: 207–25. ———. 1975. The Politics of Accommodation: Pluralism and Democracy in the Netherlands. 2nd ed. Berkeley: University of California Press. ———. 1977. Democracy in Plural Societies: A Comparative Exploration. New Haven, CT: Yale University Press. ———. 1979. “Consociation and Federation: Conceptual and Empirical Links.” Canadian Journal of Political Science—Revue Canadienne de science politique 12, 3: 495–515.

Introduction

59

———. 1985. Power-Sharing in South Africa. Berkeley: University of California Press. ———. 1990. “The Alternative Vote: A Realistic Alternative for South Africa?” Politikon 18, 2: 91–101. ———. 1995. “Self-Determination Versus Pre-Determination of Ethnic Minorities in Power-Sharing Systems.” In Will Kymlicka, ed., The Rights of Minority Cultures. Oxford: Oxford University Press. ———. 1996. “The Puzzle of Indian Democracy: A Consociational Interpretation.” American Political Science Review 90, 2: 258– 68. ———. 1999. Patterns of Democracy: Government Forms and Per formance in ThirtySix Countries. New Haven, CT: Yale University Press. ———. 2008. Thinking About Democracy: Power Sharing and Majority Rule in Theory and Practice. New York: Routledge. Lustick, Ian S. 1979. “Stability in Deeply Divided Societies: Consociationalism Versus Control.” World Politics 31, 3: 325– 44. Mabry, Tristan, John McGarry, Margaret Moore, and Brendan O’Leary, eds. 2013. Divided Nations in the Age of European Integration. Philadelphia: University of Pennsylvania Press. MacCulloch, Diarmaid. 2004. The Reformation: A House Divided. New York: Viking. Madison, James, Alexander Hamilton, and John Jay. 1987 (1788). The Federalist Papers. Ed. and with an introduction by Isaac Kramnick. Penguin Classics. Harmondsworth: Penguin. Majone, Giandomenico. 2005. Dilemmas of European Integration: The Ambiguities and Pitfalls of Integration by Stealth. Oxford: Oxford University Press. Makiya, Kanan. 2003. “A Model for Post-Saddam Iraq.” Journal of Democracy 14, 3: 5–12. Mattes, Michaela, and Burcu Savan. 2009. “Fostering Peace After Civil War: Commitment Problems and Agreement Design.” International Studies Quarterly 53, 3: 737–59. McCrudden, Christopher, and Brendan O’Leary. 2013. Consociationalism and the Courts: Human Rights versus Power-Sharing (Oxford: Oxford University Press). McGarry, John. 2001. “Northern Ireland and the Shortcomings of Civic Nationalism.” In Northern Ireland and the Divided World: Post-Agreement Northern Ireland in Comparative Perspective, 109–36. Oxford: Oxford University Press. ———. 2010a. “Asymmetric Autonomy in the United Kingdom.” In Marc Weller and Katherine Nobbs, eds., Asymmetric Autonomy and the Settlement of Ethnic Conflicts, 148–82. Philadelphia: University of Pennsylvania Press. ———. 2010b. “Ethnic Domination in Democracies.” In Marc Weller and Katharine Nobbs, eds., Political Participation of Minorities: A Commentary on International Standards and Practice, 35–71. Oxford: Oxford University Press. McGarry, John, and Brendan O’Leary, eds. 1993. The Politics of Ethnic Conflict Regulation: Case Studies of Protracted Ethnic Conflicts. London: Routledge.

60

Brendan O’Leary

———. 2006a. “Consociational Theory, Northern Ireland’s Conflict, and Its Agreement, Part One: What Consociationalists Can Learn from Northern Ireland.” Government and Opposition 41, 1: 43– 63. ———. 2006b. “Consociational Theory, Northern Ireland’s Conflict, and Its Agreement, Part Two: What Anti-Consociationalists Can Learn from Northern Ireland.” Government and Opposition 41, 2: 249–77. ———. 2007. “Iraq’s Constitution of 2005: Liberal Consociation as Political Prescription.” International Journal of Constitutional Law 5, 4: 670–98. ———. 2009a. “Must Pluri-National Federations Fail?” Ethnopolitics (Special Issue: Federalism, Regional Autonomy and Confl ict) 8, 1: 5–26. ———. 2009b. “Part I: Argument: Power Shared After the Deaths of Thousands.” In Rupert Taylor, ed., Consociational Theory: McGarry and O’Leary and the Northern Ireland Conflict, 15–84. London: Routledge. ———. 2009c. “Part III: Response: Under Friendly and Less-Friendly Fire.” In Rupert Taylor, ed., Consociational Theory: McGarry and O’Leary and the Northern Ireland Conflict, 333–88. London: Routledge. Miliband, Ralph. 1980 (1969). The State in Capitalist Society. London: Quartet Books. Mill, John Stuart. 1997. Mill: Texts, Commentaries. Ed. Alan Ryan. Norton Critical Edition. New York: W. W. Norton, 1997. Millar, Fergus. 2002. The Roman Republic in Political Thought. Hanover, NH: University Press of New England. ———. 2006. Rome, the Greek World, and the East. Ed. Hannah M. Cotton and Guy M. Rogers. Vol. 3. Chapel Hill: University of North Carolina Press. Mills, C. Wright. 1956. The Power Elite. New York: Oxford University Press. Mitchell, Paul, Geoff rey Evans, and Brendan O’Leary. 2009. “Extremist Outbidding in Ethnic Party Systems Is Not Inevitable: Tribune Parties in Northern Ireland.” Political Studies 57, 2: 397– 421. Mitchell, Paul, Brendan O’Leary, and Geoff rey Evans. 2001. “Northern Ireland: Flanking Extremists Bite the Moderates and Emerge in Their Clothes.” Parliamentary Affairs 54, 4: 725– 42. ———. 2002. “The 2001 Elections in Northern Ireland: Moderating ’Extremists’ and the Squeezing of the Moderates.” Representation 39, 1: 23–36. Montesquieu, Charles de Secondat Baron de. 1989 (1748). The Spirit of the Laws. Trans. Anne Cohler, Basia Miller, and Harold Stone. Ed. Raymond Guess, Quentin Skinner, and Richard Tuck. Cambridge Texts in the History of Political Thought. Cambridge: Cambridge University Press. Morris, Peter. 2002. Power: A Philosophical Analysis. 2nd ed. Manchester: Manchester University Press. Nagel, Jack H. 2000. “Expanding the Spectrum of Democracies: Reflections on Proportional Representation in New Zealand.” In Markus M. L. Crepaz, Thomas A.

Introduction

61

Koelbe, and David Wilsford, eds., Democracy and Institutions: The Life Work of Arend Lijphart, 113–27. Ann Arbor: University of Michigan Press, 2000. Nordlinger, Eric A. 1981. On the Autonomy of the Democratic State. Cambridge, MA: Harvard University Press. Norris, Pippa. 2008. Driving Democracy: Do Power-Sharing Institutions Work? Cambridge: Cambridge University Press. O’Leary, Brendan. 1990. “Contribution to Symposium on Brooke Talks.” Fortnight. September. ———. 1992. “Public Opinion and Northern Irish Futures.” Political Quarterly 63, 2: 143–70. ———. 1999a. “The Implications for Political Accommodation in Northern Ireland of Reforming the Electoral System for the Westminster Parliament.” Representation 35, 2–3: 106–13. ———. 1999b. “The Nature of the Agreement.” Fordham Journal of International Law 22, 4: 1628– 67. ———. 2001a. “An Iron Law of Federations? A (Neo-Diceyian) Theory of the Necessity of a Federal Staatsvolk, and of Consociational Rescue: The 5th Ernest Gellner Memorial Lecture.” Nations and Nationalism 7, 3: 273–96. ———. 2001b. “The Elements of Right-Sizing and Right-Peopling the State.” In Brendan O’Leary, Ian S. Lustick, and Tom Callaghy, eds., Right-Sizing the State: The Politics of Moving Borders, 15–73. Oxford: Oxford University Press. ———. 2002. “The Belfast Agreement and the British-Irish Agreement: Consociation, Confederal Institutions, a Federacy, and a Peace Process.” In Andrew Reynolds, ed., The Architecture of Democracy: Constitutional Design, Conflict Management and Democracy, 293–356. Oxford: Oxford University Press. ———. 2005a. “Debating Consociation: Normative and Explanatory Arguments.” In S.J.R. Noel, ed., From Power-Sharing to Democracy: Post-Conflict Institutions in Ethnically Divided Societies, 3– 43. Toronto: McGill-Queen’s University Press. ———. 2005b. “Power-Sharing, Pluralist Federation and Federacy.” In Brendan O’Leary, John McGarry, and Khaled Salih, eds., The Future of Kurdistan in Iraq, 47–91. Philadelphia: University of Pennsylvania Press. ———. 2007. “Analyzing Partition: Definition, Classification and Explanation.” Political Geography 26, 8: 886–908. ———. 2008. “The Logics of Power-Sharing, Consociation and Pluralist Federations.” In Marc Weller, Barbara Metzger, and Niall Johnson, eds., Settling SelfDetermination Disputes: Complex Power-Sharing in Theory and Practice, 47–58. Leiden: Martinus Nijhoff. ———. 2010. “Electoral Systems and the Lund Recommendations.” In Marc Weller and Katharine Nobbs, eds., Political Participation of Minorities: A Commentary on International Standards and Practice, 363– 400. Oxford: Oxford University Press.

62

Brendan O’Leary

———. 2011. “Debating Partition: Evaluating the Standard Justifications.” In Kurt Cordell and Stefan Wolff, eds., The Routledge Handbook of Ethnic Conflict, 140– 57. London: Routledge. O’Leary, Brendan, Bernard Grofman, and Jorgen Elklit. 2005. “Divisor Methods for Sequential Portfolio Allocation in Multi-Party Executive Bodies: Evidence from Northern Ireland and Denmark.” American Journal of Political Science 49, 1: 198–211. O’Leary, Brendan, Tom Lyne, Jim Marshall, and Bob Rowthorn. 1993. Northern Ireland: Sharing Authority. London: Institute for Public Policy Research. O’Leary, Brendan, and John McGarry. 1995. “Regulating Nations and Ethnic Communities.” In Albert Breton, Jean-Luigi Galeotti, Pierre Salmon, and Ron Wintrobe, eds., Nationalism and Rationality, 245–90. Cambridge: Cambridge University Press. ———. 1996. The Politics of Antagonism: Understanding Northern Ireland. 2nd expanded ed. London: Athlone. ———. 2012. “The Politics of Accommodation: Surveying National and Ethnic Conflict Regulation in Democratic States.” In Adrian Guelke and Jean Tournon, eds., The Study of Politics and Ethnicity: Recent Analytical Developments. Leverkusen Opladen: Barbara Budrich. 79–116. O’Leary, Brendan, John McGarry, and Khaled Salih, eds. 2005. The Future of Kurdistan in Iraq. Philadelphia: University of Pennsylvania Press. Parsons, Talcott. 1969. “Political Power: A Reader in Theory and Research.” In Roderick Bell, David V. Edwards, and R. Harrison Wagner, eds., Political Power: A Reader in Theory and Research, 251–84. New York: Free Press. Polybius. 1979. The Rise of the Roman Empire. Trans. Ian Scott-Kilvert. Penguin Classics. Harmondsworth: Penguin. Portes, Alejandro, and Rubén G. Rumbaut. 1990. Immigrant America: A Portrait. Berkeley: University of California Press. Powell, G. Bingham. 1988 (1982). Contemporary Democracies: Participation, Stability, and Violence. Cambridge, MA: Harvard University Press. ———. 2000. Elections as Instruments of Democracy: Majoritarian and Proportional Visions. New Haven, CT: Yale University Press. Putnam, Robert D. 1995. “Tuning In, Tuning Out: The Strange Disappearance of Social Capital in America.” PS: Political Science and Politics 28, 4: 664–83. Rabushka, Alvin, and Kenneth A. Shepsle. 2009 (1972). Politics in Plural Societies: A Theory of Democratic Instability. Longman Classics in Political Science. New York: Pearson/Longman. Reilly, Benjamin. 2001. Democracy in Divided Societies: Electoral Engineering for Conflict Management. New York: Cambridge University Press. Reilly, Benjamin, and Andrew Reynolds. 1999. Electoral Systems and Conflict in Divided Societies, Papers on International Conflict 2. Washington, DC: National Academy Press.

Introduction

63

Rezvani, David A. 2007. “Shaping the Federacy Research Agenda.” Ethnopolitics 6, 1: 129–31. ———.2003. Federacy: The Dynamics of Semi-Sovereign Territories. Unpublished manuscript. Roeder, Philip G. 2007. Where Nation-States Come From: Institutional Change in the Age of Nationalism. Princeton, NJ: Princeton University Press, 2007. Roeder, Philip G., and Donald Rothchild, eds. 2005. Sustainable Peace: Power and Democracy After Civil Wars. Ithaca, NY: Cornell University Press. Rummel, Rudolph J. 1997. Death by Government, with a Foreword by Irving Louis Horowitz. London: Transaction Publishers, 1997. Russell, Bertrand. 1992 (1938). Power: A New Social Analysis. London: Routledge. Saunders, J. J. 1972. The History of the Mongol Conquests. Philadelphia: University of Pennsylvania Press. Schumpeter, Joseph. 1987 (1944). Capitalism, Socialism and Democracy. London: Allen and Unwin. Simpson, George E. 1968. “Assimilation.” In International Encyclopaedia of the Social Sciences, 438– 44. New York: Macmillan and the Free Press. Steiner, Jürg. 1970. “Switzerland: ’Magic Formula’ Coalitions.’ ” In Eric C. Browne and John Dreijmanis, eds., Government Coalitions in Western Democracies. New York: Longman. Syme, Ronald. 1974 (1939). The Roman Revolution. Oxford: Oxford University Press. Takaki, Ronald. 1993. A Different Mirror: A History of Multicultural America. Boston: Little Brown. Taylor, Charles. 1975. Hegel. Cambridge: Cambridge University Press. ———. 1992. “The Politics of Recognition.” In Charles Taylor and Amy Gutman, eds., Multiculturalism and “the Politics of Recognition.” Princeton, NJ: Princeton University Press. Taylor, Paul. 1993. International Organization in the Modern World: The Regional and the Global Process. London: Pinter. Taylor, Rupert, ed. 2009. Consociational Theory: McGarry and O’Leary and the Northern Ireland Conflict. London: Routledge. Thomas, Hugh M. 2003. The English and the Normans: Ethnic Hostility, Assimilation and Identity 1066–c. 1220. Oxford: Oxford University Press. Tully, James. 1995. Strange Multiplicity: Constitutionalism in an Age of Diversity. Cambridge: Cambridge University Press. van den Berghe, Pierre. 1981. “Assimilation.” In The Ethnic Phenomenon, 214–36. New York: Elsevier. von Hayek, Friedrich A. 1973. Law, Legislation and Liberty: A New Statement of the Liberal Principles of Justice and Political Economy. London: Routledge and Kegan Paul. Walter, Barbara F. 1997. “The Critical Barrier to Civil War Settlement.” International Organization 51, 3: 335– 64.

64

Brendan O’Leary

Weber, Max. 1994. “Between Two Laws.” In Peter Lassman and Ronald Speirs, eds., Political Writings, 75–80. Cambridge: Cambridge University Press. Weingast, Barry. 1997. “Federalism as a Commitment to Preserving Market Incentives.” Journal of Economic Perspectives 11, 4: 83–92. Wimmer, Andreas. 2003– 4. “Democracy and Ethno-Religious Conflict in Iraq.” Survival 45, 4: 111–34. Wintrobe, Ronald. 1998. The Political Economy of Dictatorship. Cambridge: Cambridge University Press. ———. 2006. “Extremism, Suicide Terror, and Authoritarianism.” Public Choice 128: 169–95. Zolberg, Aristide R., and Long Litt Woon. 1999. “Why Islam Is Like Spanish: Cultural Incorporation in Europe and the United States.” Politics and Society 27, 1: 5–38.

PA R T I Power Sharing and Electoral Systems

CHAPTER 2

Electoral Rules and Ethnic Representation and Accommodation: Combining Social Choice and Electoral System Perspectives Bernard Grofman

Decisions about voting rules can be regarded as one of the four most important choices structuring sociopolitical relationships, each of which has implications for ethnic representation, the central concern of this essay. Unlike the other three major choices—choosing between a unitary versus a federal system, choosing a parliamentary as opposed to a presidential system, and choosing between imposing a set of universally applicable laws on all citizens versus arrangements that allow for religious- or ethnic-specific laws and lawmaking bodies to govern many aspects of day-to-day life—specific electoral system rules are rarely constitutionalized. Thus while electoral systems tend to be “sticky,” they are usually easier to modify than these other important aspects of institutional design. Here no attempt will be made to address all the relationships between voting rules and ethnic relations. Furthermore, our discussion of the properties of various voting methods is limited to the expected consequences of the use of the rules in the situation of (largely or locally) bipolar ethnic conflict—where voter preferences can be treated as falling along a single ethnically defined (single-peaked) dimension—the nature of which will be explained. Insights are offered into the ways in which voting outcomes can incorporate minority preferences by examining competing normative criteria to judge voting rules that are directly or indirectly inspired by social choice

68

Bernard Grofman

theory. The goal is to integrate ideas from social choice with the mainstream political science electoral system tradition. Special attention will be paid to eight voting rules, which will be defined subsequently. In each case they enable electing a single alternative from among several. They are simple plurality, the alternative vote (AV), anti-antiplurality, approval voting, the Borda rule, the Coombs rule, the méthode majoritaire, and the two-round majority runoff. Then extensions of two of these rules are considered that permit the simultaneous election of multiple candidates. The electoral systems literature on the linkages between interethnic relations and choice of electoral rule has largely emphasized one of two dichotomies: proportional representation (PR) versus plurality rule, and PR versus so-called vote-pooling methods such as the alternative vote. Arend Lijphart, for example, advocates the use of pure list PR in divided societies because it facilitates proportional descriptive representation for minorities at the same time as it allows strong party leaders (perhaps from ethnically-ethnically based parties) to broker inter-ethnic deals. Lijphart (1968, 1969, 1977, 1991, 1996) also argues that stable democracy in plural societies requires other consociational arrangements, which are discussed elsewhere in this volume (1968, 1969, 1977, 1991, 1996, 2008). A related but still distinctive argument for PR is that in deeply divided societies, it allows powerful actors a stake in the system. Thus the focus is less on the moderates than on the ability to incorporate extremists who might have the ability to bring an end to the democratic process (Brendan O’Leary, personal communication, February 2008). Donald Horowitz, on the other hand, distinguishes what he calls votepooling methods from methods aimed at guaranteed or proportional descriptive representation of minorities. Horowitz wishes to create incentives for pan-ethnic coalitions, or broad multi-ethnic parties based on other interests (e.g., regional or economic) that transcend ethnic identities, or candidates who make cross-ethnic appeals. To achieve these ends, Horowitz has been rather insistent that the alternative vote is the best electoral rule to foster voting patterns that will cross ethnic lines and coalitions that will adopt accommodative or moderate policies with respect to ethnic divisions (Horowitz 1989a, 1989b, 1997). The social choice literature on voting rules is also often oriented around two competing, but rather different, normative perspectives. One is rooted in the view of the eighteenth-century philosophe the Marquis de Condorcet,

Electoral Rules

69

the other in ideas of his contemporary, Jean Charles de Borda. Each proposed a different way of aggregating individual preferences into a collective choice. A central argument of this essay is that all three dichotomies (PR versus plurality, PR versus vote pooling, and Condorcet versus Borda) are of limited utility in making choices about what electoral rules are sensible in situations of actual or potential ethnic divisions. This argument rests on seven claims sketched here. First, PR and plurality have much more in common than is usually supposed. Second, and relatedly, the differences between different types of PR (e.g., list PR and the single-transferable vote) can, for some purposes, be more important than differences between PR viewed generically and plurality rules viewed generically. Third, and similarly, the differences among different types of pluralitarian (or majoritarian rules) can be more important than the differences between such rules viewed generically and PR viewed generically. Fourth, when we consider which normative properties we would like voting rules to satisfy, the approaches of Borda and Condorcet are just two of the more important ways to think normatively about what we want a voting rule to achieve in the way of representing voter preferences. Fift h, how rules operate in practice and how they operate in theory can be quite different. For example, when we look at the formal properties of the alternative vote and when we consider evidence for its operation in Fiji, Donald Horowitz’s views of its expected moderating effects can be shown to confuse potential with actuality, and he is not sufficiently mindful of the contextual factors that must hold if the alternative vote is to work as he thinks it will to foster moderation (Fraenkel and Grofman 2004, 2006a, 2006b). Similarly, when plurality voting in the United States is considered its operation depends critically on contextual features, such as the degree of ethnic homogeneity in the constituency, the mechanism used to select party nominees, and the degree to which voting is polarized along racial lines in both primary and general elections (Grofman, Handley, and Lublin 2001), as well as contextual features such as the existence of a presidential system and various layers of simultaneous and nonsimultaneous elections. Sixth, in most societies, even if proportional methods are used for legislative elections, there is a substantial element of majoritarianism in legislative processes. Finally, and relatedly, to truly understand the implications of an electoral rule we must understand not just how it operates within par ticu lar constituencies but how it operates to generate the total set of legislators and

70

Bernard Grofman

what implications that may have for the creation of a ruling government or government coalition and the policies it chooses.

Some Important Voting Rules for Single-Seat Contests Below we consider eight important methods of converting votes into winners in single-seat contests, defined here in alphabetical order. a. Alternative vote. The alternative vote, known as the instant runoff in the United States, requires voters to submit a rank-ordered ballot. If no candidate has a majority of first-place votes, the candidate with the fewest number of first-place votes is dropped and her votes are reallocated to the next highest-ranked (eligible) candidate on the voter’s ranking. This process continues until one candidate has a majority of first-place preferences among the reduced set of candidates. b. Antiplurality. In antiplurality voting each voter indicates the candidate he least wishes to see elected. The candidate with the fewest antiplurality votes is then chosen. An equivalent way to generate an antiplurality outcome is to ask voters to vote for all but one candidate. Now the candidate with the most votes wins. c. Approval voting. In approval voting each voter may express support for as many candidates as s/he chooses, up to one fewer than the total number of candidates, n. The candidate with the most approval votes wins. d. Borda rule. As in the alternative vote, voters using the Borda rule must submit a rank ordering. Each candidate gets a Borda score, which is the sum over all voters of the number of candidates in each voter’s preference ranking that the given alternative ranks ahead of. The alternative with the highest sum is the Borda winner. e. Coombs rule. As in the alternative vote, the Coombs rule requires voters to submit a rank-ordered ballot. However, if no candidate has a majority of first-place votes, the candidate with the greatest number of last-place votes is dropped and her votes are reallocated to the next highest-ranked (eligible) candidate on that voter’s ranking. This process continues until one candidate has a majority of first-place preferences among the reduced set of candidates.

Electoral Rules

71

f. Méthode majoritaire. The recently invented méthode majoritaire is the least well-known. To determine the winner under this method, we begin by assigning scores (grades) to candidates from among some finite set of possible grades. Its inventors employ the six grades used in French education: excellent, very good, good, acceptable, poor, and reject (Balinski and Laraki 2007a, 2007b). Different candidates may receive the same scores (grades). The majority-grade of a candidate is his or her median grade. The méthode majoritaire begins by ordering the candidates according to their majority grades, that is, according to median scores. However, with many candidates and only a finite number of grades, some will probably have the same median rank. To break ties, three values attached to a candidate, called the candidate’s majority-values, are sufficient to determine the candidate’s place in the overall ranking. For each candidate, let pi = % of the ith candidate’s grades above the candidate’s majority-grade, αi = % of the ith candidate’s support at the candidate’s majority-grade, qi = % of the ith candidate’s grades below the candidate’s majority-grade. Candidate i ranks ahead of candidate j, when i’s majority-grade is better than j’s, or if i and j are tied in median rank, then i ranks ahead of j if pi > pj or, equivalently, if qi < qj. The way in which this rule operates may not be initially easy to grasp. It helps to recognize that the term majoritaire is somewhat of a misnomer. The method is really about establishing median preferences. We can show how the rule works with the data from a Balinski and Laraki experiment involving the set of candidates running in the 2007 French presidential election in Table 2.1. They first identify the median scale value for each candidate. Of the twelve, four are graded reject, four are graded poor, one is graded acceptable, and three are graded good. The four who receive a majority of ranking as rejects (and whose median scale value is therefore reject) are at the very bottom of the ratings, in ranks 9 through 12. Similarly the three whose median ranking is good are at the top, in ranks 1 through 3, and so on. To determine, for example, which of the three who are rated good

72

Bernard Grofman

Table 2.1. Méthode Majoritaire Experimental Results for the 2007 French Presidential Election, Round One

Besancenot (5) Buet (8) Schivardi (11) Bayrou (1) Bove (6) Voynet (4) Villiers (9) Royal (2) Nihous (10) Le Pen (12) Laguiller (7) Sarkozy (3)

Excellent (%)

Very Good (%)

Good (%)

Acceptable (%)

Poor (%)

to Reject (%)

4.1 2.5 0.5 13.6 1.5 2.9 2.4 16.7 0.3 3.0 2.1 19.1

9.9 7.6 1.0 30.7 6.0 9.3 6.4 22.7 1.8 4.6 5.3 19.8

16.3 12.5 3.9 25.1 11.4 17.5 8.7 19.1 5.3 6.2 10.2 14.3

16.0 20.6 9.5 14.8 16.0 23.7 11.3 16.8 11.0 6.5 16.6 11.5

22.6 26.4 24.9 8.4 25.7 26.1 15.8 12.2 26.7 5.4 25.9 7.1

31.1 30.4 60.4 7.4 39.5 20.5 55.5 12.6 56.0 74.4 40.1 28.2

Source: Balinski and Laraki 2007b, Table 10. Majority judgment results, three precincts of Orsay, April 22, 2007.

is at the number 1 rank, which at the number 2 rank, and which at the number 3 rank, we look at the sum of each candidate’s percentages above his or her median rank. For Bayrou this is 44.3 percent, for Royal it is 39.4 percent, while for Sarkozy it is 38.9 percent. Thus they come in 1, 2, and 3, respectively, under this method. g. Plurality. In plurality elections, also known as winner takes all, if there is a single candidate to be elected, then the candidate with the highest number of first-place votes wins. This rule is often mistakenly called first past the post—the mistake consists in supposing there is a fi xed post that candidates have to get past to win. h. Two-round majority runoff. In a two-candidate two-round majority runoff the two candidates who do best in the first round face each other in a second-round election unless the plurality winner of the first round has received a majority of the votes cast, in which case s/he is elected without the need for a runoff. There are three key methods of converting votes into electoral outcomes in multiseat contests, which are defined below in alphabetical order. In each case let m be the number of seats to be fi lled and n the voters of voters (n odd).

Electoral Rules

73

a. List PR. In a list PR system voters have a single vote that they may cast for the party of their choice. If there are m seats to be filled in a given constituency, then the parties share those seats proportionally to their share of votes, with the number of seats to which each party is entitled going to the top candidates on that party’s ranked list of candidates. We will limit ourselves to one particular form of list PR: the d’Hondt quota rule. In d’Hondt, we take the vote shares of each the parties and divide by the integers 1, 2, 3, and so forth. If there are m seats to be filled, the m highest quotients so obtained indicate a seat that will be assigned to the party achieving that quotient. b. Plurality bloc voting. Here each voter has m votes to give (singly) to each of up to m candidates, and the m candidates with the highest number of votes are elected. When the number of seats to be fi lled under plurality bloc voting is the same as the size of the legislature, then we have what is called an at-large election. c. The single transferable vote (STV). The single transferable vote, also known as the Hare system (and in Australia as the Hare-Clarke system), requires that voters rank order their preferences for candidates. If there are m seats to be fi lled and n voters, then winning candidates must receive a Droop quota of votes (the greatest integer bound of the quotient n/(m + 1)). If any candidate receives a Droop quota or more of first-place votes, the “excess” number of votes (i.e., in excess of the Droop quota) are reallocated to another (still eligible) candidate based on the next choices on the voter’s list of the voters casting “excess” ballots. If no remaining candidate receives a Droop quota of first-preference votes, the candidate with the fewest first-preference votes is eliminated, and for those ballots where she was fi rst choice, votes are reallocated to the next (still eligible) candidate on the voter’s list. Th is process continues until exactly m candidates have received a Droop quota or until the pool of eligibles is down to as many candidates as there are seats remaining to be fi lled.

Evaluating Voting Rules There are four important evaluative criteria for reconciling diverse preferences when choosing a single alternative.

74

Bernard Grofman

(1) The first-place criterion, that is, the rule should give as many voters as possible their first choice. (2) The Condorcet criterion, that is, the rule should give the majority what it wants. In particular, if there is an alternative that can defeat each and every alternative in head-to-head competition, choose that alternative (Condorcet 1785; Black et al.1998). Such an alternative is referred to as a majority winner or a Condorcet winner. (3) The acceptability criterion, that is, the rule should give as many voters as possible an “acceptable” choice. (4) The avoid the worst criterion, that is, the rule should give as few voters as possible their worst choice. We can now classify the eight major voting methods we focus on against these four evaluative criteria to see which one, on its face, they come closest to matching. Matching the first-place criterion with the voting rule to which it most closely corresponds is trivial; it is only plurality that limits itself to consideration of first-place preferences. In social choice theory the Condorcet criterion is often used as the litmus test for identifying a socially desirable alternative in situations where only a single alternative is being picked (Black et al. 1998). But there may not always be a Condorcet winner, as the famous paradox of cyclical majorities demonstrates. The simplest example of this paradox is three voters with respective preferences ABC, BCA, and CBA. A receives a majority over B, B receives a majority over B, yet C receives a majority over A, so majority rule preferences violate transitivity and create a cycle. There is no alternative among the three that receives a majority against each and every other alternative. A method that guarantees to select a Condorcet winner when one exists is called a Condorcet extension method (Young 1977). Most methods for selecting a single alternative, including all eight of those identified above, are not Condorcet extension methods. However, non-Condorcet extension methods may differ greatly in their likelihood of picking a Condorcet winner when one exists. Some methods have a very high probability of doing so, under certain distributional assumptions about voter preferences, whereas others have a much lower probability of doing so. The term “Condorcet efficiency” is used to refer to the likelihood of choosing a Condorcet winner when one exists. There is now a substantial literature evaluating voting methods, under specified distributional assumptions, as to their Condorcet

Electoral Rules

75

efficiency (Grofman and Feld 2004; Regenwetter and Grofman 1998; Merrill 1984, 1985, 1988; Fishburn 1976). Of our eight methods, the two that seem most obviously to be attempts to reflect “majority” preferences are the two-round majority runoff method and the alternative vote (instant runoff ) because each requires that the eventual winner receive a majority, not just a plurality, of the votes. But neither method guarantees to select a majority winner, even if one exists. Moreover, when there are a large number of alternatives being compared, both of these methods will probably look more like plurality in their Condorcet efficiency than they do like Condorcet extension methods with their 100 percent certainty of picking a Condorcet winner when one exists. Furthermore, under particular distributional assumptions, some of the other methods in our inventory may actually be more likely to give us a majority winner when there is one (i.e., have higher Condorcet efficiency) than methods like these that might seem, on their face, directly oriented toward doing so. The acceptability criterion is the vaguest of the four, and thus it is not surprising that there are several different methods that can be said to meet its ethos. The method that most obviously belongs in this category is “approval voting.” It is quite explicitly a search for an alternative with the greatest overall acceptability. But both the Borda rule and the méthode majoritaire also involve a search for outcomes that are acceptable on average. But just as there are different notions of a “representative” value for a distribution, for example, the mean and the median, among voting methods, the Borda rule may be said to behave like a search for the candidate with the highest average voter acceptability, while the misleadingly labeled méthode majoritaire might be said to reflect a desire to find a candidate who has the highest median level of acceptability. Strengthening the notion that the Borda rule is a search for an acceptable alternative, we can view the Borda rule as generating outcomes exactly halfway between plurality and antiplurality. All three methods are instances of “scoring rules.” For the case of three alternatives, Saari (1994, 1995) offers a simple way to represent all scoring rules by a single value, s, in the triple (1, s, 0), where weights have been normalized so that 0 is the weight given to an alternative that is placed last in a voter’s ranking, s is the weight given to an alternative that is placed second in a voter’s ranking, and 1 is the weight given to an alternative that is placed first in a voter’s ranking, with 0 < s < 1. The plurality rule is characterized by the triple (1, 0, 0), the antiplurality rule by the triple (1, 1, 0), and the Borda rule by the triple (1, 0.5, 0).

76

Bernard Grofman

Finally, when we look to see which voting method most closely matches the avoid the worst criterion it seems clear that antiplurality and the Coombs rule belong in that category, completing our assignment of evaluating the eight methods we introduced earlier against the four criteria. But it is helpful to say a bit more about how the two methods differ. Antiplurality is obviously the mirror image of plurality; the one looks to maximize first-place votes, the other to minimize last-place votes. In contrast, the Coombs rule is the mirror image of the alternative vote. It replaces “eliminating the alternative with the fewest first-place votes” with “eliminating the alternative with the most last-place votes.” However, both AV and the Coombs rule always satisfy the “mirror image” of the Condorcet criterion, the Condorcet loser criterion, namely, do not pick an alternative that loses to each and every other alternative in a head-to-head contest because the last pairing in each method guarantees that any Condorcet loser will be defeated. In contrast, it is easy to find examples to show that the antiplurality rule, like the plurality rule, violates the Condorcet loser criterion. Consider seven voters with preferences BAC, BAC, BAC, BCA, CAB, CAB, and CAB, respectively. Here antiplurality will choose A, yet A is a Condorcet loser. Since our focus in on the link between electoral rules and ethnic representation, we do not wish to consider the properties of the voting methods we identified only in the abstract (as we have just done) or in contexts where there are patterns of voting polarized along racial, ethnic, religious, or linguistic lines. To keep our discussion manageable we will look at the important case of bi-ethnic conflict. Our results will also apply in settings where, in any given constituency, there are only two major ethnic groups, even if there are many more than that in the state as a whole.

The Effects of Voting Rules for Selection of a Single Alternative in Bi-ethnic Situations Where Voting Is Single-Peaked Along an Ethnically Defined Single Dimension To simplify our initial analyses, we begin by looking at single-seat elections and confine analysis to settings where there are only two ethnies from which candidates are drawn, though there may be more than one candidate from each ethny. It might appear that there is not much to be said about ethnic representation in elections to pick a single representative in such a

Electoral Rules

77

context. In particular, if there are only two ethnic communities, one will elect a winner and the other will not, and presumably the group in the majority will be able to coordinate its votes to elect a candidate of its choice. But even leaving aside strategic and coalitional issues or the issue of candidate nomination procedures (for now), that point of view is too simplistic. In a bi-ethnic context, it is useful to think about candidates as lying along a continuum anchored by attitudes toward the two groups. Figure 2.1 illustrates the idea of bi-ethnic conflict between native Fijians and IndoFijians (descendants of those brought to Fiji from India in previous centuries to work the plantations) with single-peaked preferences along an ethnically defined dimension (Fraenkel and Grofman 2004). This illustration is based on a four-party situation. The four alternatives shown in the figure are: a radical pro-Indian party (RI), a moderate pro-Indian party (MI), a moderate pro–native Fijian party (MF), and a radical pro–native Fijian party (RF). Here moderation refers to the willingness to engage in compromise with the other ethnie. There are twenty-four possible linear rankings among the four parties, but only eight of those are single-peaked along the RI, MI, MF, RF continuum (see Table 2.2), that is, graphable as utility curves that have at most one inflection point, with that inflection point reflecting a change from an upward to a downward slope (Black et al. 1998; Fraenkel and Grofman 2004).

Order of Preference

1

2

3

4 Radical Indian

Moderate Indian

Moderate Fijian

Radical Fijian

Political Parties

Figure 2.1. Single-peaked preferences along an ethnically defi ned four-party continuum.

78

Bernard Grofman

Table 2.2. Single-Peaked Preferences Among Four Parties Along a Continuum Anchored by a Radical Native Fijian Party at One End and a Radical Indo-Fijian Party at the Other End Type A1 rI mI mF rF

Type B1

Type B2

Type B3

Type C1

Type C2

Type C3

Type D1

mI rI mF rF

mI mF rI rF

mI mF rF rI

mF mI rF rI

mF mI rI rF

mF rF mI rI

rF mF mI rI

Source: Fraenkel and Grofman 2004, Table 1.

Once we confine ourselves to single-peaked preferences (along a biethnic continuum), the properties of the various voting methods we have considered no longer look quite the same. In particular, we know from the work of Black et al. (1998) that when preferences are single-peaked, there always exists a Condorcet winner. Furthermore, we know that when preferences are single-peaked, the Condorcet winner will be whichever alternative is closest to the ideal point (the most preferred location) of the median voter on the spectrum, that is, it will reflect the preferences of the voter who is ideologically located so that (n − 1)/2 of the voters have ideal points that lie to her left on the spectrum and (n − 1)/2 have ideal points that lie to her right. Even when preferences are single-peaked, neither plurality, AV, nor Coombs guarantees the selection of a Condorcet winner. Still, when preferences are single-peaked, there are important differences in their expected Condorcet efficiencies among plurality, AV, and Coombs. Proposition 1: If we assume single-peaked preferences over a single dimension and no party holds a majority of first-place preferences, and posit that voter preferences are sincere (in the sense of Farquharson 1969), when we have four parties or fewer, the candidate of the median party is more likely (or at least no less likely) to win when voting is conducted under the alternative vote than when voting is conducted under plurality. Proof: Grofman and Feld 2004. On the other hand, once we have five or more alternatives, then the superiority of AV over plurality, even when preferences are single-peaked, is

Electoral Rules

79

no longer assured. We can demonstrate this with an example, for five alternatives, where the Condorcet winner is the plurality winner but is eliminated by the alternative vote sequential choice process. Consider fifteen voters who vote over five alternatives with single-peaked preferences, ABCDE, ABCDE, ABCDE, ABCDE, ABCDE, BCADE, BCADE, BCDAE, BCDAE, CBDAE, CBDAE, CBDAE, DCBAE, EDCBA, EDCBA, respectively. Here B is both the plurality winner and the Condorcet winner, yet the elimination sequence under AV will eliminate D, then E, then B, and then A, with C winning, even though B would defeat C by 9 to 6. Proposition 2: If voters have single-peaked preferences over a single dimension, and voter preferences are sincere (in the sense of Farquharson 1969), then the Coombs rule always selects the Condorcet winner (i.e., the alternative supported by the median voter). Proof: See Grofman and Feld 2004. The first proposition is important because it shows that in bi-ethnic conflict situations the superiority of AV over plurality is not guaranteed, although there are some circumstances where Donald Horowitz and others (Reilly 2001) are correct to suggest that AV can be reliably expected to outperform simple plurality in picking a Condorcet winner. The preceding example shows that even single-peakedness does not guarantee that AV will pick the Condorcet winner when one exists. It therefore casts doubt on Horowitz’s claims about the power of AV to result in moderation. The second proposition is important because it shows that a rule that might appear attractive in the abstract because it avoids the worst (i.e., meets the Coombs rule) becomes, under the distributional assumption of single-peaked preferences, a Condorcet extension method as well. This reinforces an extremely important point: how a given rule can be expected to operate strongly depends upon the particular context in which it is applied. The properties of a voting rule in general need not be the most important properties of that rule when we restrict ourselves to situations where particu lar types of ideological or ethnic preference patterns are found (or when other special features, e.g., particular candidate nomination rules or distributions of voter preferences across multiple constituencies, are in place).

80

Bernard Grofman

Extensions to Multicandidate Selection Processes Links Between Single-Seat and Multiseat Methods

Each of the three multiseat methods identified earlier has a single-seat analogue among the set of eight voting methods for selecting a single candidate. Let us define a voting method for choosing a single alternative as the root of a voting method for choosing multiple alternatives if the latter method reduces to the former method when we shift from a multiseat to a single-seat application of the method. It is easy to see that the alternative vote is simply the single transferable vote applied to a single seat, that is, for the case when m = 1 and the Droop quota is simply half the available votes. So AV is the root for STV. It is equally obvious that plurality is the root for plurality bloc voting. What is less well recognized is that plurality is also the root of (any form of) list PR because all forms of list PR, just like plurality, only pay attention to first preferences (Kurrild-Klitgaard forthcoming). But once we recognize this fact, we also recognize that a given method of electing a single candidate may serve as a root for more than one multiseat method.

Comparing Plurality and PR

The four normative criteria we looked at earlier were intended for single-seat constituencies. We sought to demonstrate that even though we need only pick a single alternative, that does not mean that it is obvious what choice we “ought” to make or that there is not a potential for seeking “compromise” or broader “consensus” even if there is only one office to be filled. In the multiseat context, in contrast, until quite recently, most of the electoral systems literature took as a given one particular normative principle, that of proportionality (notably Lakeman 1982, 1984). The contrast between PR and plurality with respect to proportionality of party representation is a staple of the historic debates in the electoral system literature about which electoral system is best. But the nature of the PR-plurality difference with respect to proportionality of representation of groups is more complex than is sometimes supposed. First, for PR systems, the greater the number of seats to be filled, m, the easier it is, in principle, for a party to nominate a diverse array of candidates.

Electoral Rules

81

Thus the closer m is to 1, ceteris paribus, the more a PR system will resemble plurality in single-seat constituencies. (Likewise, for STV, ceteris paribus, the closer m is to 1, the more that system will behave like AV.) Second, it is customary to distinguish PR from plurality in terms of the votes needed to assure minority representation. But comparisons across voting rules based on the threshold of exclusion, the largest vote share a party can receive and still be denied winning any seats, can be misleading (Grofman 2001). In comparing PR and plurality in terms of the number of votes needed to elect a candidate to office, we must control for district population. Consider plurality, with a threshold of exclusion of half, and its most direct m-seat extension, d’Hondt list PR, with a threshold of exclusion of 1/(m + 1). If we compare PR and plurality in terms of threshold of exclusion we might think that if m = 3, it takes only half as many voters in an m-seat district to guarantee the selection of a candidate of their choice as it does in a single-seat district with a plurality-based election (i.e., comparing a threshold of exclusion of half to one of one quarter). But that comparison neglects the fact that the three-seat district is probably roughly three times the size of the single-seat district. Thus it takes one and a half times as many voters to guarantee the election of a candidate in a legislature of m-seat districts using PR than in a legislature of the same size composed of single-seat constituencies using PR. In other words, contrary to our intuition, it is actually “easier,” in terms of votes needed, to elect candidates from single-seat constituencies than from multiseat ones—in the limit, twice as easy (Grofman 2001). But if that is so, why is the view that PR systems foster minority representation so universally accepted? The basic answer is that we need to pay attention to the distribution of minority voters across constituencies. Imagine that minority voters make up 26 percent of the total electorate and that they are evenly distributed across all constituencies. If all constituencies are single seat and voting is polarized along ethnic lines, then the minority will win no seats. In contrast, in threeseat constituencies using the d’Hondt rule, since the threshold of exclusion is 25 percent, the minority will be expected (given polarized voting patterns) to gain at least one seat out of three—actually a slight overrepresentation because of the lumpiness effects of small district size. It will take more raw votes to win the seat in the three-member constituency (roughly one and a half times as many) but a smaller vote share. In general, the more concentrated are the minority voters, the more resemblance to a PR outcome will be provided by plurality voting. Indeed, if minorities are “fortunately” distributed

82

Bernard Grofman

so that they comprise, say, 52 percent of the voters in 50 percent of the singleseat plurality-based constituencies, they might well do better under plurality than under PR. However, absent considerable geographic concentration, single-member district plurality methods will systematically and substantially underrepresent minorities—with the degree of underrepresentation expected to be greatest when the minority population is small (Grofman 1982). What this means is that the contrast between plurality and PR cannot be considered in the abstract; how different the two systems will be depends greatly on the geographic distribution of voter preferences. A third important complexity is the nature of the party system under plurality. There has been a largely uncritical acceptance of the universal applicability of Duverger’s Law that two-party competition (at least at the constituency level) will result from plurality-based elections. The empirical evidence is based primarily on the fact that for the vast bulk of pluralitybased elections, values of the Laakso-Taagepera (LT) index of effective number of parties (Laakso and Taagepera 1979) are relatively close to two, but that LT index conceals considerable variation in the actual number of parties regularly competing. In fact, only the United States (and a few small Caribbean island democracies) produces anything at all like two-party competition. Other democracies like Canada, India, and the United Kingdom are characterized by long-lasting third (or other minor) parties, and the struggle for power or influence at the national level can lead to repeated competition among more than two parties even at the district level (Grofman, Bowler, and Blais 2009). How disproportional plurality will be depends greatly on whether there are more than two parties competing. When there are, the distribution of third-party strength can yield perversities, such as nonmonotonic patterns of translating votes into seats or even a party winning a majority of the seats with fewer votes than a rival. But when there really are only two parties, as in the United States, the results are not that far from proportional at the federal level. A fourth reason to be careful about claims that PR systems are inevitably more proportional than elections under PR is the need to recognize that for PR, when there are thresholds imposed or when party systems are new and expectations unstable, there can be huge numbers of votes “wasted” on parties that do not gain representation in the legislature, leading to substantial divergence from proportionality.

Electoral Rules

83

Thus the PR-versus-plurality dichotomy is overblown when it comes to proportionality, at least in situations close to a two-party setting. Moreover, the other most common base of comparison of PR and plurality, the longstanding and still ongoing debate about whether it is best for political compromises to take place at the constituency level or in the legislature itself, is also more complicated than sometimes supposed. The standard view is that plurality makes it more likely that ideological or ethnic moderates will be elected, while the legislature elected under PR will be more like a microcosm of the society, mirroring the full range of political viewpoints, including extreme ones, and thus politics is both more stable and more moderate under plurality. However, this stylized portrait of differences between the two voting rules is problematic from both a theoretical and an empirical standpoint. On the one hand, the claim that single-seat elections involving plurality— at least ones with only two parties—lead to convergence toward the preferences of the median voter is theoretically suspect because it holds true only when a very large number of (mostly implicit) assumptions are met (Grofman 2004), and it can be shown that the claim is usually empirically falsified. In the United States we certainly do find that the more liberal the district the more liberal its representative will be. But rather than convergence between the parties, what we find instead is that for any given set of constituency characteristics, Democrats and Republicans elected from a district with similar characteristics (or from the same district at different times in close proximity to one another) do not look at all alike. In ideological terms the Democrat is always to the left of the Republican (Grofman, Griffin, and Glazer 1990), and, at least in recent decades, liberal districts will tend to elect Democrats and conservative districts will tend to elect Republicans (Grofman et al. 2000). Moreover, and equally important, divergence between parties at the district level can then translate into even more dramatic divergence at the legislative level because national party effects now prevent most conservative Democrats and most liberal Republicans from winning election, thus accentuating the differences between the parties in the legislature even beyond those between the parties at the constituency level. On the other hand, even though it is generally true that the ideological range represented in parliaments elected under PR is wider than that under plurality, this does not imply that governance in the countries that use PR is necessarily unstable (Lijphart 1994) or that policies in those

84

Bernard Grofman

countries are further away from the preferences of the median voter, since the coalitions that usually govern in PR countries are often based around a relatively centrist party and parties ideologically proximate to it (Powell 2000). Finally we would observe that because both plurality and list PR rely on first preferences, if there were no geographic constraints on how we assign voters into constituencies we believe that we can mimic PR outcomes in mseat districts with m appropriately constructed single-seat plurality districts. In a later section we will illustrate this idea of mimicking a multimember PR system with a single-seat system, along with how we might use AV to mimic STV.

Comparing List PR and STV

Although for ethnically divided societies both forms of PR allow for key groups to be represented in parliament, and that may be highly desirable, there are important differences between them that are not noted in the electoral systems literature. In that literature, discussion of differences between the two most important types of PR voting methods tends to be limited to enumerating the differences in ballot format and in rules for calculating winners or, at best, to pointing out that STV and the d’Hondt form of list PR have the same threshold of exclusion, 1/(m + 1), or that STV has incentives for politicians and parties to pay attention to second-place (and lower) preferences that list PR does not. Case studies tend to look at either STV settings or list PR settings but rarely compare the consequences of the two approaches to achieving proportionality (Reilly 2001 is a notable exception; see also Fraenkel and Grofman 2006a, 2006b). The differences between plurality and AV, the respective roots of list PR and STV, are central to a full understanding of the differences between list PR and STV. List PR, unlike STV, is about first preferences, not about consensus building. For example, consider eleven voters with single-peaked preferences: ABCDE, ABCDE, ABCDE, ABCDE, BCADE, BCADE, BCADE, CBADE, CDBAE, DCBAE, and EDCBA. If we need to elect two, in all versions of list PR, alternatives/parties A and B will be chosen. In contrast, in an STV election to choose two, the two winners are A and C, since after A is chosen because it has reached the Droop quota of 4 based on first-place votes, it will not

Electoral Rules

85

matter whether we eliminate DCBAE or EDCBA first. In either case, votes will eventually transfer to C, giving that candidate a Droop quota.

Mimicking List PR with Plurality and STV with AV

To better understand how STV can be thought of as a majoritarian system and how list PR has important resemblances to plurality, we provide two useful results. Both propositions are believed to be correct, but the proofs are not (yet) watertight, so they are identified here only as conjectures. (Conjectured) Proposition 3: In an electorate of size N, if there are m seats to be filled, (a) there exists a partitioning of the voters in that electorate into m + 1 groups, with m groups of size N/(m + 1) and the (m + 1)th group including the remaining voters, such that if the outcomes in each of the first m groups are decided by AV, the results will perfectly mimic the outcome of an STV election to fill those m seats conducted in the electorate as a whole; and (b) there will exist a partition of the voters in that electorate into m + 1 groups, with m groups of size N/(m + 1) and the (m +1)th group including the remaining voters, not necessarily the same partition as for STV, such that if the outcomes in each of the first m groups are decided by plurality, the results in the first m constituencies will perfectly mimic the outcome of a list PR election to fill those m seats conducted in the electorate as a whole using the d’Hondt method of PR. We believe an even stronger result is true. (Conjectured) Proposition 4: In an electorate of size N, where there are m seats to be fi lled, (a) there will exist a partition of the voters in that electorate into m groups each of size greater than or equal to N/(m + 1) such that if the outcomes in each are decided by AV, the results will perfectly mimic the outcome of an STV election to fill those m seats conducted in the electorate as a whole; and (b) there will exist a partition of the voters in that electorate into m groups each of size greater than or equal to N/(m + 1), not necessarily the same partition as for STV, such that if the outcomes in each are decided by

86

Bernard Grofman

plurality, the results will perfectly mimic the outcome of a list PR election to fi ll those m seats conducted in the electorate as a whole using the d’Hondt method of PR. In other words, if we could assign voters to constituencies in any way we chose, but in constituencies of roughly equal size, we could create AV constituencies that could be used to mirror the results of an STV election or plurality single-seat constituencies that could be used to mimic the results of a list PR election under the d’Hondt rule. This result shows just how important constituency assignments can be for determining outcomes. To see how such a partition process might work, let us consider different ways of partitioning the previous set of eleven voters with single-peaked preferences, ABCDE, ABCDE, ABCDE, ABCDE, BCADE, BCADE, BCADE, CBADE, CDBAE, DCBAE, and EDCBA, respectively, into two districts: one with five voters and one with six voters. To mimic the STV outcome of {A, C} we can put together the first five voters in one district and the second six in another and use AV. As with the results for STV in the electorate as a whole the outcome set does not depend upon the order of elimination of tied alternatives. On the other hand, if we created one district with ABCDE, ABCDE, BCADE, BCADE, and BCADE and put the remaining voters into the other district, AV would give us {B, C}, different from either the STV or list PR outcome. To mimic the outcome of list PR under d’Hondt, namely {A, B}, all we have to do is put the first four voters with the last two in one district and the remaining voters in the other district. But different ways of grouping voters into plurality could lead to different outcomes. If we put together the first six voters in one district and the second five in another we get the STV outcome of {A, C}. Of course, there are some outcomes we cannot get no matter how we distribute voters; for example, we cannot force the choice of E because it has only one first-place vote. Nonetheless, these examples show how districting is important in considering the consequences of alternative voting rules and how proportional or minoritarian multiseat voting rules can be mimicked by pluralitarian or majoritarian voting methods applied to single-seat constituencies.

Conclusion This chapter has identified several different ways of thinking about electoral systems according to four different normative criteria: realizing first-place

Electoral Rules

87

choices, implementing majority choices, finding widely acceptable choices, and avoiding picking alternatives at the bottom of the preference ranking of a large number of voters. It also has shown how the implications of rules might shift when we make specific (plausible) assumptions about the sociopolitical context in which they would be used, for example, bi-ethnic conflict along a single-peaked dimension. Then we demonstrated that such commonly used dichotomies in the electoral and social choice literature, such as that between PR and plurality, neglected some key similarities between these two methods, both in principle and in practice. We also showed that STV and list PR have important differences that can be linked to the singleseat method at the root of each (AV and plurality, respectively). The key aim of this essay has been to integrate some ideas (e.g., Condorcet winner, singlepeakedness) common in the social choice literature with the literature examining electoral systems but with a special concern for applications to the representation of minority (ethnic) preferences. Those interested in the role of electoral systems as tools for the promotion of power sharing or other forms of ethnic accommodation need to pay more attention to the social choice literature. They also need to understand that the commonly made stark distinctions between PR and pluralitarian-majoritarian methods are apt to be misleading since we need to pay much more attention to the consequences of political geography. Notes Thanks to Sue Ludeman and Clover Behrend-Gethard for bibliographic assistance. Research on this project was partially supported by funding from the Jack W. Peltason (Bren Foundation) Endowed Chair, University of California, Irvine, and by the UCI Center for the Study of Democracy. 1. The view that federalism is highly desirable for multiethnic societies is widely held but is not without dissenters. Pursuing this debate would take us outside the scope of this essay. For a general discussion of federalism, see Filippov,Ordeshook, and Shvetsova 2004. See also the introduction to this volume. 2. Prominent scholars of comparative politics, most notably Juan Linz, have asserted that presidential (as opposed to parliamentary) systems are not suited to plural societies and must be replaced. The existence of a single president and the inevitable winner-take-all confl ict for that office allegedly exacerbates ethnic tension and provides a power base to the winner to use for ethnic domination (Linz and Valenzuela 1994). Whether presidentialism is, per se, a destabilizing force has triggered an ongoing debate in the comparative politics literature whose pursuit would take us beyond the scope of this essay. See, e.g., Horowitz 1991; Lijphart 1992.

88

Bernard Grofman

3. The first two of these choices are almost always “constitutionally embedded,” and the third often is. In most countries, electoral systems choice is not constitutionally embedded but allows for change by legislative action. In none of the long-term democracies with recent dramatic changes of electoral system for national parliamentary elections—France in the 1980s, twice; New Zealand, 1993; Japan, 1996; Italy, 1994, 2004—was constitutional change required. However, in a number of new democracies (e.g., South Africa, Estonia), while electoral system details are not spelled out in the constitution, there is language along the lines of a requirement to “follow the principle of proportionality.” In some older democracies, e.g., Ireland, the precise electoral system is constitutionally mandated (Bunreacht na hÉireann 1937, Article 16.2). 4. The eight methods discussed are but a few of the many voting rules for a singleseat contest that have been proposed. The alternative vote, plurality, and two-round runoffs are among those used for real-world elections to legislatures or of executives. The others—antiplurality, approval, Borda, Coombs, and the méthode majoritaire— are important for theoretical, comparative, and evaluative reasons. There are several inventories of methods used in national elections (see, e.g., Reynolds and Reilly 1997). But human ingenuity in devising different rules, even for the case of selecting a single alternative, is quite remarkable. Even those broadly familiar with the methods in actual use may not be aware of the full spectrum of potentially usable voting methods. 5. There are, however, many variants of runoffs in addition to the two-round majority runoff. For example, the number of rounds required to select a winner can vary, as can the criterion used to determine which candidates advance from one round to the next (Grofman 2008). In legislative elections in France under the double ballot system, if there is no candidate with a majority of votes in the first round of balloting, only the candidates who receive at least 12.5 percent of the vote are eligible to run in the second ballot in which the winner is the plurality winner. In single-winner elections for some political party organizations in the United States and Canada, if no candidate receives a majority on the first ballot, a common form of multiballot runoff eliminates the candidate at each round who receives the fewest first-place votes and then continues in this fashion until some candidate receives a majority. This method is sometimes called lowest candidate out runoff (LCOR) or multiple round sequential elimination (MRSE). 6. These three multiseat methods include most of the key (families of) methods in common use. The two most important methods whose implications for ethnic representation space considerations do not allow us to discuss here are limited voting and cumulative voting. A useful discussion of these methods and ethnic representation may be found in Engstrom 1998. I will also not consider mixed methods, i.e., those that do not have the same rule used in all constituencies (Shugart and Wattenberg 2003). 7. More complex list PR rules permit voters to affect the ordering of candidates on a party’s list. Such methods are usually referred to as open list PR.

Electoral Rules

89

8. There are a variety of other methods for generating a “proportional” allocation (Balinski and Peyton Young 1982), but they raise issues beyond the scope of this essay. 9. In some computer implementations of STV these ballots are chosen at random from the set of ballots in which the candidate is ranked first (among eligible candidates); in the hand-counted version, still valid ballots are first sorted according to which candidate is ranked first, and a count continues until a Droop quota is met for a given first-place choice with the stack of “used” ballots eliminated from further consideration and the remaining ballots in the stack allocated to the highest ranked (still eligible) candidate on each voter’s ballot. Other variants create fractional ballots so as to sum in total to the excess of first-place ballots in excess of the Droop quota (Tideman and Richardson 2000). For present purposes none of these complexities matters. 10. Consider five voters with preferences ACB, ACB BCA, BCA, and CBA. The alternative vote will choose B because C, the Condorcet winner, will be eliminated in the first round. When there are only three voters, instant runoff and the two-round majority runoff are, in effect, identical; thus the latter method also fails to give us the Condorcet winner. (Note we are assuming that voters vote “sincerely” [Farquaharson 1969], i.e., they do not misrepresent their “true” preferences for strategic reasons.) 11. This way of understanding the normative power of these two methods on an intuitive level is my own and would not necessarily be shared by the leading contemporary advocates of each method (Balinski in the case of the méthode majoritaire and for Borda, Saari 1994, 1995). 12. The term ethny or ethnie means any religious, racial, ethnic, or linguistic grouping (Smith 1991; Williams 2003; Francis 1976). 13. For simplicity we neglect the possibility of voters with identical ideal points. 14. It is the alternative closest to the preference of the median voter. 15. AV will guarantee to pick the Condorcet winner when voter preferences are single-troughed (Grofman and Feld 2004), but single-troughedness (a mirror image of single-peakedness, where preferences can be graphed as curves that change their slope at most once, from a downward to an upward direction; Black et al. 1998) is a relatively obscure condition that is unlikely to be satisfied in real-world bi-ethnic settings. 16. Space constraints confine analysis to the three multiseat methods identified here. 17. The LT index is the inverse of the even better-known Hirschman-Herfindahl index of concentration, widely used in economics. The LT index of the effective number of parties at the seat (vote) level is given by the inverse of the sum of the squared values of each party’s seat (vote) shares. 18. In addition, differences between plurality and PR systems with respect to proportionality may not loom as large when we consider party (or group) representation over time. Taylor, Gudgin, and Johnston (1986) discuss the notion of proportional tenure, i.e., the average representation of a party over the course of many elections. The United States, for example, is far more proportional than it is commonly given credit because the alternation of power between Democrats and Republicans has led

90

Bernard Grofman

to a congressional seat share for each that closely tracks each party’s average share of the total votes cast for major-party candidates. 19. The STV, in addition to opening up the possibility of proportional representation of an ethnie’s voting strength, has been posited by some scholars to provide incentives for cross-ethnic choices by voters and even the formation of cross-ethnic preelectoral alliances by parties. Various authors familiar with the use of the STV in Northern Ireland have argued for its use instead of list forms of PR (Horowitz 1985; O’Leary et al. 1993; Reilly and Reynolds 1999; Reynolds 1999; Mitchell 2001; Mitchell, O’Leary, and Evans 2001, 2002). 20. The STV portions derive their inspiration from a previous contribution (Sugden 1984). 21. This proposition requires that voters cast complete ballots showing their ranking for all candidates; otherwise there may be seats won with less than a Droop quota, which complicates creating the single-district matchups to the m outcomes. 22. This proposition also requires that voters cast complete ballots showing ranking for all candidates. References Balinski, Michel, and Rida Laraki. 2007a. “A Theory of Measuring, Electing, and Ranking.” Proceedings of the National Academy of Science 104, 21. ———. 2007b. “Election by Majority Judgement: Experimental Evidence.” Ecole Polytechnique: Centre National de la Recherche Scientifique 28. 1– 41. Balinski, Michel L., and H. Peyton Young. 1982. Fair Representation: Meeting the Ideal of One Man, One Vote. New Haven, CT: Yale University Press. Black, Duncan, et al. 1998. The Theory of Committees and Elections. Ed. I. McLean, A. McMillan, and B. L. Monroe. 2nd rev. ed. Boston: Klewer Academic Press. Bunreacht na hÉireann. 1937. Constitution of Ireland, as Amended. Dublin: Stationery Office. Condorcet, Marquis de. 1785. Essai sur l’application d ’analyse à la probabilité des decisions rendues a la pluralité des voix. Paris: De l’imprimeul Royale. Engstrom, Richard L. 1998. “Minority Electoral Opportunities and Alternative Election Systems in the Unitred States.” In M. E. Rush, ed., Voting Rights and Redistricting in the United States. Westport, CT: Greenwood Press. Farquharson, Robin. 1969. The Theory of Voting. New Haven, CT: Yale University Press. Filippov, Mikhail, Peter Ordeshook, and Olga Shvetsova. 2004. Designing Federalism: A Theory of Self-Sustainable Federal Institutions. New York: Cambridge University Press. Fishburn, Peter. 1976. “An Analysis of Simple Two Stage Voting Systems.” Behavioral Science 21: 1–12. Fraenkel, Jon, and Bernard Grofman. 2004. “A Neo-Downsian Model of the Alternative Vote as a Mechanism for Mitigating Ethnic Conflict in Plural Societies.” Public Choice 121: 487–506.

Electoral Rules

91

———. 2006a. “Does the Alternative Vote Foster Moderation in Ethnically Divided Societies? The Case of Fiji.” Comparative Political Studies 39, 5: 623–51. ———. 2006b. “The Failure of the Alternative Vote as a Tool for Ethnic Moderation in Fiji? A Rejoinder to Horowitz.” Comparative Political Studies 39, 5: 663– 65. Francis, E. K. 1976. “Outline of a Theory: Ethnicity and Types of Society.” In Interethnic Relations: An Essay in Sociological Theory. Amsterdam: Elsevier. Grofman, Bernard. 1982. “For Single-Member Districts, Random Is Not Equal.” In B. Grofman, A. Lijhpart, R. McKay, and H. Scarrow, eds., Representation and Redistricting Issues. Lexington, MA: Lexington Books. ———. 2001. “A Note of Caution in Interpreting the Threshold of Exclusion.” Electoral Studies 20: 299–303. ———. 2004. “Downs and Two-Party Convergence.” Annual Review of Political Science 7: 25– 46. ———. 2008. “A Taxonomy of Run-Off Methods.” Electoral Studies 27: 395–99. Grofman, Bernard, Shaun Bowler, and Andre Blais. 2009. “Duverger’s Law in Canada, India, the U.S. and the U.K.” Berlin: Springer Verlag. Grofman, Bernard, and Scott L. Feld. 2004. “If You Like the Alternative Vote (a.k.a. the Instant Runoff ), Then You Ought to Know About the Coombs Rule.” Electoral Studies 23: 641–59. Grofman, Bernard, Robin Griffin, and Amihai Glazer. 1990. “Identical Geography, Different Party: A Natural Experiment on the Magnitude of Party Differences in the U.S. Senate, 1960–84.” In R. J. Johnston, F. M. Shelley, and P. J. Taylor, eds., Developments in Electoral Geography. London: Routledge. Grofman, Bernard, Lisa Handley, and David Lublin. 2001. “Drawing Effective Minority Districts: A Conceptual Framework and Some Empirical Evidence.” North Carolina Law Review 79: 1383–1430. Grofman, Bernard, William Koetzle, Michael McDonald, and Thomas Brunell. 2000. “A New Look at Split Ticket Voting for House and President: The Comparative Midpoints Model.” Journal of Politics 62, 1: 34–50. Grofman, Bernard, and Robert Stockwell. 2003. “Institutional Design in Plural Societies: Mitigating Ethnic Conflict and Fostering Stable Democracy.” In R. Mudambi, P. Navarra, and G. Sobbrio, eds., Economic Welfare, International Business and Global Institutional Change. New York: Edward Elgar. Horowitz, Donald L. 1985. Ethnic Groups in Conflict. Berkeley: University of California Press. ———. 1989a. “Ethnic Conflict Management for Policymakers.” In J. V. Montville, ed., Conflict and Peacemaking in Multiethnic Societies. Lexington, MA: Heath. ———. 1989b. “Making Moderation Pay: The Comparative Politics of Ethnic Confl ict Management.” In J. V. Montville, ed., Conflict and Peacemaking in Multiethnic Societies. Lexington, MA: Heath. ———. 1991. A Democratic South Africa? Constitutional Engineering in a Divided Society. Berkeley: University of California Press.

92

Bernard Grofman

———. 1997. “Encouraging Electoral Accommodation in Divided Societies.” In B. V. Lal and P. Larmour, eds., Electoral Systems in Divided Societies: The Fiji Constitutional Review. Canberra: Australian National University. Kurrild-Klitgaard, Peter. Forthcoming. “Voting Paradoxes Under Proportional Representation.” Scandinavian Political Studies. Laakso, M., and Rein Taagepera. 1979. “Effective Number of Parties: A Measure with Applications to West Europe.” Comparative Political Studies 12: 3–27. Lakeman, Enid. 1982. “The Case for Proportional Representation.” In Power to Elect: The Case for Proportional Representation. London: Heinemann. ———. 1984. “The Case for Proportional Representation.” In A. Lijphart and B. Grofman, eds., Choosing an Electoral System: Issues and Alternatives. New York: Praeger. Lijphart, Arend. 1968. The Politics of Accommodation: Pluralism and Democracy in the Netherlands. Berkeley: University of California Press. ———. 1969. “Consociational Democracy.” World Politics 21, 2: 207–25. ———. 1977. Democracy in Plural Societies: A Comparative Exploration. New Haven, CT: Yale University Press. ———. 1991. “The Power-Sharing Approach.” In J. V. Montville, ed., Conflict and Peacemaking in Multiethnic Societies. Lexington, MA: Heath Lexington Books. ———, ed. 1992. Parliamentary Versus Presidential Government. Oxford: Oxford University Press. ———, ed. 1994. Electoral Systems and Party Systems: A Study of Twenty-Seven Democracies. Oxford: Oxford University Press. ———. 1996. “The Puzzle of Indian Democracy: A Consociational Interpretation.” American Political Science Review 90, 2: 258– 68. ———. 2008. Thinking About Democracy: Power Sharing and Majority Rule in Theory and Practice. New York: Routledge. Linz, Juan J., and Arturo Valenzuela, eds. 1994. The Failure of Presidential Democracy. Baltimore: Johns Hopkins University Press. Merrill, Samuel. 1984. “A Comparison of the Efficiency of Multicandidate Electoral Systems.” American Journal of Political Science 28, 1: 23– 49. ———. 1985. “A Statistical Model for Condorcet Efficiency Based on Simulations Under Spatial Model Assumptions.” Public Choice 47, 2: 389– 403. ———. 1988. Making Multicandidate Elections More Democratic. Princeton, NJ: Princeton University Press. Mitchell, Paul. 2001. “Transcending an Ethnic Party System? The Impact of Consociational Governance on Electoral Dynamics and the Party System.” In R. Wilford, ed., Aspects of the Belfast Agreement Oxford: Oxford University Press. Mitchell, Paul, Brendan O’Leary, and Geoff rey Evans. 2001. “Northern Ireland: Flanking Extremists Bite the Moderates and Emerge in Their Clothes.” Parliamentary Affairs 54, 4: 725– 42. ———. 2002. “The 2001 Elections in Northern Ireland: Moderating ’Extremists’ and the Squeezing of the Moderates.” Representation 39, 1: 23–36.

Electoral Rules

93

Mozaffar, Shaheen, and Gail A. Maloney. 2007. “Are Multi-Ethnic Societies and DeeplyDivided Societies Different?” In 12th Annual World Convention of the Association for the Study of Nationalities. New York: Columbia University. O’Leary, Brendan, Tom Lyne, Jim Marshall, and Bob Rowthorn. 1993. Northern Ireland: Sharing Authority. London: Institute for Public Policy Research. Powell, G. Bingham. 2000. Elections as Instruments of Democracy: Majoritarian and Proportional Visions. New Haven, CT: Yale University Press. Regenwetter, Michel, and Bernard Grofman. 1998. “Approval Voting, Borda Winners and Condorcet Winners: Evidence from Seven Elections.” Management Science 44, 4: 520–33. Reilly, Benjamin. 2001. Democracy in Divided Societies: Electoral Engineering for Conflict Management. New York: Cambridge University Press. Reilly, Benjamin, and Andrew Reynolds. 1999. Electoral Systems and Conflict in Divided Societies. Washington, DC: National Academy Press. Reynolds, Andrew. 1999. Electoral Systems and Democratization in Southern Africa. New York: Oxford University Press. Reynolds, Andrew, and Ben Reilly. 1997. The International IDEA Handbook of Electoral Design. Stockholm: International Institute for Democracy and Electoral Assistance. Saari, Donald. 1994. Geometry of Voting. Berlin: Springer Verlag. ———. 1995. The Basic Geometry of Voting. Berlin: Springer Verlag. Shugart, Matthew Soberg, and Martin P. Wattenberg. 2003. Mixed-Member Electoral Systems: The Best of Both Worlds? Oxford: Oxford University Press. Smith, Anthony D. 1991. National Identity. Harmondsworth: Penguin. Sugden, Robert. 1984. “Free Association and the Theory of Proportional Representation.” American Political Science Review 78, 1: 311– 43. Taylor, Peter J., Graham Gudgin, and R. J. Johnston. 1986. “The Geography of Representation: A Review of Recent Findings.” In B. Grofman and A. Lijphart, eds., Electoral Laws and Their Political Consequences. New York: Agathon Press. Tideman, Nicolaus, and D. Richardson. 2000. “A Comparison of Improved STV Methods.” In S. Bowler and B. Grofman, eds., Elections in Australia, Ireland and Malta Under the Single Transferable Vote. Ann Arbor: University of Michigan Press. Williams, Robin. 2003. The Wars Within: People and States in Conflict. Ithaca, NY: Cornell University Press. Young, H. Peyton. 1977. “Extending Condorcet’s Rule.” Journal of Economic Theory 16: 335–53.

CHAPTER 3

The Track Record of Centripetalism in Deeply Divided Places Allison McCulloch

In ethnically divided places, exclusion from government tends to be associated with exclusion in the wider polity (Horowitz 1993). Centripetalism, as developed and defended by Donald L. Horowitz, is seen as a novel strategy for the design of political institutions intended to mitigate this sense of exclusion. Centripetalism, sometimes called the integrative approach, is thought to encourage minority influence on majority decision making through the use of electoral rules that necessitate cross-ethnic appeals on the part of political leaders. The underlying assumption is that leaders must be given incentives to “make moderation pay” (Horowitz 1990). Calling the electoral system “the most powerful lever of constitutional engineering for accommodation,” Horowitz advocates the adoption of preferential electoral systems, specifically the alternative vote (AV) (1991, 163, 189). AV is a majoritarian-preferential system typically used in single-member districts: voters rank the candidates, and the winning candidate must receive at least 50 percent plus 1 in order to be elected. Centripetalists claim that AV sets in motion a cycle of moderation and cross-community cooperation that reduces the potential for ethnic conflict in deeply divided places. Because it is both majoritarian and preferential, political leaders will have to campaign (in heterogeneous constituencies) outside their own ethnic group in order to receive enough votes to get elected. To capture the lower-ranked votes of members of other groups, they will have to make explicit cross-ethnic appeals and in so doing will have to moderate their own claims. This gives rise to “coalitions of commitment” composed of the ethnic parties willing to pool

Centripetalism

95

votes with other parties across the ethnic divide (Horowitz 2000, 379). It creates a scenario in which minorities help get majorities elected. Indeed, Horowitz’s argument hinges on the argument that minorities fare better when they can influence decision making even if they are not guaranteed their own representation: “in divided societies, officeholding is decidedly inferior to influence” (1997, 34). At first glance, the model appears to be compelling. It rests on an intuitive logic that suggests that political parties closer to the middle will find it easier to engage in cross-group cooperation, thereby enhancing political stability. The idea that moderate behavior and reciprocal relationships are the means by which to facilitate political stability and intergroup peace is both inherently plausible and normatively desirable. However, the model thus far has a poor track record. Its practice has shown that in cases of deep division, AV is unlikely to promote the sort of incentives imagined. Rather than promoting a politics of moderation, that logic is in many cases reversed, and a politics of extremism results. The disjuncture between the logic of moderation and the institutional outcome of extremism stems from two sources. First, deeply divided societies generally do not have the specific demographic configuration mandated by the model. Majoritarian-preferential voting arrangements are heavily dependent on both the spatial distribution of groups and on the number of groups engaged in confl ict. Where one group constitutes an overwhelming majority and can be elected without minority support, it may fi nd that there is little need to pool votes. Second, several factors related to how deeply divided the groups consider themselves to be make the implementation of the proposed recommendations difficult. The most important is the fact that cross-group voting is rare in situations of deep ethnic division where the stability of political institutions is precariously maintained in the face of high degrees of animosity between groups and where ethnic violence, on a small or large scale, is probable or already occurring. My intent in this chapter is first to demonstrate the problematic incentive structure at work in divided places that have adopted centripetal institutions, focusing on the cases of the Republika Srpska (RS) Entity of Bosnia and Herzegovina, Sri Lanka, and Fiji. The second goal is to sketch some contextual imperatives that help explain why the model is not appropriate for deeply divided places. I conclude with the possibility that Horowitz’s recommendations may still be of use to places that are mildly rather than deeply

96

Allison McCulloch

divided—ironically what Horowitz has always claimed is true of consociation applies in greater force to his own prescriptions.

The Practice of Centripetalism in Deeply Divided Places AV is intended to provide a “centripetal spin” to politics in divided places as it is thought to create “electoral incentives for broad-based moderation by political leaders and disincentives for extremist outbidding” (Sisk 1996, 41). Yet in very deeply divided places, these assumptions do not always hold. Below I briefly review three cases of deep division where AV was utilized. The results are discouraging. In RS, it reinforced the ethnic divide and consolidated support for the extremist war-oriented party. In Sri Lanka, it prolonged and exacerbated the self-determination dispute. In Fiji, it made extremism more politically rewarding and has precipitated two coups. Rather than facilitating political stability in these cases, AV represents an institutional impasse to constructive intergroup politics. In brief, it has served to make bad situations worse.

The Republika Srpska (RS) Entity of Bosnia and Herzegovina

Postwar Bosnia and Herzegovina, as stipulated by the terms of the Dayton Peace Accords, is organized as a rigid consociation. While this design has done much to help quell violence, following electoral successes for ultranationalist parties in 1996 and 1998, there was a growing concern that such an arrangement may be reinforcing rather than alleviating ethnic divisions (Manning 2008, 77–78; ICG 2000, 14). For this reason, the international community began to push for reforms that were “copybook Horowitzian in intent,” the most obvious example of which was the adoption of preferential voting for the 2000 presidential election in Republika Srpska (Bose 2002, 222). The president and vice president were elected on a single ticket on which voters were allowed to rank-order the candidates. According to the Association of Election Officials in BiH, “moderate candidates stand a better chance of winning the elections, as they will have greater support from a larger cross-section of the electorate” (quoted in Belloni 2007, 83). The assumption was that minority votes could come to mean as much as majority ones by way of cross-group voting.

Centripetalism

97

It was anticipated that the new electoral rules would favor Milorad Dodik of the Alliance of Independent Social Democrats (SNSD), whom the international community then viewed as “a moderate with sufficient nationalist credentials to be effective” (Manning 2008, 94). While Dodik was unlikely to receive a majority of first-preference votes, it was hoped that he could surpass the more extremist Serb Democratic Party (SDS) candidate, Mirko Šarović, and win on the basis of sufficient second-preference votes (Bieber 2006, 96–97; ICG 2000). Six political parties—three Serb and three primarily Bosniak—contested the election, with voting falling almost exclusively along group lines. The SDS obtained a near majority in the first round (49.8 percent) and easily passed the threshold on the second count; the SNSD, in turn, received only a quarter (25.7 percent) of the first vote, which increased to only 25.9 percent in the second round (OSCE 2000). This electoral outcome can be explained by the interrelated factors of ethnic homogeneity and deep division. First, the RS Entity was largely monoethnic. Its creation was a product of the 1992–95 war, and while the Serb community represents roughly 35 percent of the total Bosnian population, that percentage is largely concentrated in the RS (Manning 2008, 75– 76). The population of the territory now part of the RS fell from 46 percent non-Serb prior to the war to only 3 percent by 1997, and in the postwar period the return of refugees and internally displaced persons (IDPs) did little to alter the monoethnic environment; Serb returnees outnumbered Bosniaks two to one (ICG 2003, 3; UNHCR 2006). The monoethnicity on the ground was to be offset, it was hoped, by voting regulations that allowed refugees and IDPs to vote in their districts of origin, even if they had not yet returned to their original place of residence. Bieber estimates that 15–20 percent of participating voters were non-Serb, most of whom did not, at the time of the election, reside in the RS (Bieber 2006, 161n109). The hope was that these absentee voters would function as the swing vote between the moderate and radical Serb candidates (Bose 2002, 231). The minority vote, however, had little effect on the electoral outcome. Second, Horowitz’s argument that minority influence should trump minority officeholding did not resonate with voters. Each group united behind its respective ethnic parties. Bosniaks, outnumbered by Serbs in RS by a sixto-one ratio, opted to vote along ethnic lines despite the extreme improbability of a Bosniak party winning the election; likewise, only a handful of Serb voters extended first preferences to the primarily Bosniak Social Democratic Party, even though its presidential candidate was a Bosnian Serb

98

Allison McCulloch

(Bose 2002, 234). The groups were so deeply divided that they did not respond to the incentives to cross-vote, even when, in the case of the Bosniaks, centripetal logic would have suggested that it was to their benefit to do so because minorities are likely to fare better under a moderate government than under an extremist one. The Bosnian Serb voters in turn had no incentive to elect a moderate candidate who would appeal to the other group when they could elect a nationalist candidate who would guarantee their own ethnic interests. The nationalist hard-liners not only won the election but also obtained a greater share of the vote than they did in simultaneous elections to the National Assembly and to the statewide House of Representatives (Belloni 2004, 343). Disillusionment and dissatisfaction among voters at the international implementation of the new electoral rules, along with comments made by Richard Holbrooke suggesting that the SDS should be banned, may have added reactive strength to the party’s support. Following the election, the preferential rules were replaced with a system wherein a president and two vice presidents would represent the constituent nations. However, some potential long-term implications of the AV experience remain. One of the lessons that Dodik must have taken away from the election is that moderation does not always pay. In the period since the 2000 election, he has struck a much harder political line, most notably on the question of RS independence. Subsequent elections have also shown the SNSD outflanking the SDS as the primary representative of Bosnian Serb interests (Manning 2008, 83). In so doing, the moderate party has edged toward extremism. This, of course, is not solely a result of the 2000 election; other incentives exist for such ethnically exclusive positions. It is nevertheless important to stress that the use of AV in the RS, employed once and then abandoned, did not encourage moderate or cross-group voting and in the end only reinforced the inclination to vote for the extremist parties.

Sri Lanka

In Sri Lanka, the president is elected for a six-year term via the Supplementary Vote (SV), which requires an absolute majority for election and the ranking of up to three candidates. In this sense, SV is a version of AV: the sole difference is that voters are not able to exhaust their preferences if there are

Centripetalism

99

three or more candidates. Sri Lanka is divided into three main ethnic groups: the Sinhalese constitute a large majority, totaling approximately 75 percent of the population, while the Tamil population compose roughly 18 percent, and the Muslim community accounts for the remaining 7 percent (Mishler, Finkel, and Peiris 2007, 206), though hard-line Tamils tend to claim that Muslims are Tamils. The conflict occurs primarily along Sinhalese and Tamil lines. The use of SV, it was anticipated, would result in a situation where “prudent presidential candidates could hardly ignore Tamil interests” and where Tamil votes would become as important as Sinhalese ones (Horowitz 1990, 463; DeVotta 2000, 68). This has not happened. Of all the presidential elections held under SV—in 1982, 1988, 1994, 1999, 2005, and 2010—none has needed a second round. In each case, a majority was captured on the first count, with two Sinhalese-dominated parties, the United National Party (UNP) and the Sri Lanka Freedom Party (SLFP), serving as the only real contenders to fi ll the office. The 2005 election of SLFP hard-liner Mahinda Rajapaksa highlights the problem of electing a powerful president in deeply divided places. Rajapaksa’s platform included a renegotiation of the peace talks with the Liberation Tigers of Tamil Eelam (LTTE) (which collapsed shortly after the SLFP victory) and a reaffirmation of the unitary composition of the state. His election manifesto called for the suppression of separatism. He specifically promised to preserve the “sovereignty, territorial integrity and unitary structure of the state while pledging to abide by a majoritarian understanding of democracy” (SLFP 2005). He also courted the support of and signed preelectoral agreements with the Janatha Vimukthi Peramuna (JVP) and the Jathika Hela Urumaya (JHU), both parties with Sinhalese ultranationalist orientations. This prompted the LTTE to organize an unofficial election boycott and to suggest that it would be “a futile exercise to show any interest in the election” (EUEOM 2005). While turnout in the Sinhalese majority areas averaged between 75 and 80 percent, it was much lower in Tamil-populated areas. In the northern district of Vanni, turnout hovered around 34 percent while in Jaff na it was less than 2 percent. Rajapaksa and the UNP’s Ranil Wickremesinghe each captured eleven districts; Rajapaksa narrowly won with 50.3 percent of the vote (Mishler, Finkel, and Peiris 2007, 208). Had turnout been higher in the Tamil-majority north where the UNP performed reasonably well, it is estimated that Wickremesinghe would have won the election. The 2005 election serves as a reminder that

100

Allison McCulloch

where divisions are deep, minorities do not always adhere to centripetal logic and vote for the other side, nor do candidates from the majority always feel compelled to reach out to minorities. Wickremesinghe’s more moderate stance was not sufficient to attract significant minority support, but Rajapaksa’s extremist position won him the presidency and ultimately led to a resumption of war. While the causes of the conflict are many, deep ethnic division mixed with an unfavorable demographic context helps explain why the SV system has not improved intergroup relations. Centripetalists argue that AV provides incentives for moderation, not just constraints on extremism. Where ethnic majorities exist, political parties representing the majority may in fact require minority support in order to get elected, but it does not follow that minorities will provide that support at the expense of their own representation. The normative appeal of AV, then, is directed almost exclusively at the majority. The incentives are one-sided: ethnic majorities may conceivably be won over to moderate platforms, but there is little incentive for ethnic minorities to accept and be satisfied with such claims to moderation. Even in situations where there are two candidates from the majority, one moderate and one extreme, the more extreme candidate often gains election, showing that it often pays to be more extremist in order to secure support from one’s own group rather than to campaign on a moderate platform. The intragroup split, which centripetalists think compels moderation, often serves instead to make both sides less moderate. Worthwhile incentives, we may conclude, have to run in both directions— that is, they have to work for both majorities and minorities—if they are to have the intended centripetal effect. Even with preferential voting, singleperson presidencies effectively prevent minority groups from gaining representation in the top office and thus cannot achieve this bi-directional effect. The problem is that SV contains only (limited) unidirectional incentives for moderation. While Sinhalese candidates may garner transfer votes from the Tamil community, the same cannot be said for any potential Tamil candidates (Reilly 2001, 124). The unstated assumption underlying centripetal recommendations is that the Tamils should be satisfied with Sinhalese representation. Yet the tacit Tamil boycott of the presidential elections, both by not fielding candidates and by not casting votes, suggests that (alleged) influence is just not sufficient for aggrieved minorities. In each election, the

Centripetalism

101

contest has always been dominated by the UNP and SLFP. The Tamil parties have not fielded candidates because they recognize the improbability of minority officeholding in a single-person majoritarian office. Moreover, for the relevant incentives to be heeded by the majority, there must already be some level of moderation evident in the polity: the majority must actually want to reach out to the minority. Yet the political rewards of moderation have been largely absent in Sri Lanka, with both the UNP and SLFP engaging in outbidding tactics, as has arguably been the case in Sinhalese politics since the 1950s. Because it is possible to get elected without lower-order preferences, there is no motivation to produce moderate platforms. Rather, as in 2005 and 2010, when Rajapaksa reached out to Sinhalese extremists, it proved more rewarding to adopt an extremist position. Both the unwillingness of the minority to settle for proxy representation and the reluctance of the majority to concede autonomy demands are profoundly influenced by demographic considerations. Here the related issues of group size and the number of groups are significant. While the Tamils represent a sizable minority, their numbers nowhere near match the overwhelming majority of the Sinhalese. But because of the territorial concentration of both groups, that overwhelming majority is reversed in the north and eastern parts of the country, where the Tamils dominate. The Tamils, moreover, are a nationally mobilized group, resident in what they consider to be their homeland (Bose 2007, 20). Deep division compounds these demographic features. Given the history of “Sinhala Only” politics dating back to at least 1956, particularly in the area of language policy, the 1972 constitutional provision giving Buddhism a “foremost place” in the country, and the violent insurgency waged by the LTTE, a profound mistrust characterizes group relations. As a result, it remains highly unlikely that the Tamil population will settle for the sort of proxy representation that SV facilitates and just as unlikely that the Sinhalese population would accept a Tamil president. The Tamils want to exercise a right to self-determination either in a separate state or, at a minimum, within a federal Sri Lanka. The introduction of proxy representation represents the antithesis of what the Tamil population is seeking. Single-member presidencies are, by definition, winner-take-all institutions; presidential elections in situations of deep divisiveness are zero-sum games in which only one ethnic group can be victorious (cf. Reynolds 1995, 94). Where the winners and losers are defined in ethnic terms, the

102

Allison McCulloch

potential for instability and conflict is egregious. This has been the political route followed in Sri Lanka.

Fiji

Fiji was the first country to adopt AV at the national legislative level for the specific purpose of ethnic conflict management and held three elections (1999, 2001, and 2006) under the system (Fraenkel and Grofman 2006, 631). While inspired by centripetal logic, the experience in Fiji reveals the opposite incentive structure at work. Intergroup relations became progressively more polarized and divisive rather than more cooperative and moderate with each successive election. Fiji is ethnically bipolar with a population that is approximately 55 percent ethnic Fijian and 38 percent Indian, though the Indian population is on the decline, with around one hundred thousand emigrating since 1987 (Field 2005). Another 7 percent of the population is composed of “general voters,” consisting of Chinese, European, and other minorities. The 1997 constitution called for a mix of communal and open constituencies in elections to the House of Representatives. In terms of communal seats, 23 were reserved for ethnic Fijians, 19 for Indians, 3 for general voters and 1 for the Rotumans. There were also 25 seats open to all voters, where the incentives for cross-ethnic vote pooling were highest. The country is rightly seen as an important test case for centripetalism. In the first AV election in 1999, centripetal logic was on display, at least in part. Politicians responded to the electoral incentives and made preelection pacts, which resulted in the formation of two main multiethnic coalitions: the People’s Coalition (made up of the Indian-dominated Fiji Labour Party [FLP] and two smaller ethnic Fijian parties) and the incumbent ethnic Fijian Soqosoqo Ni Vakavulewa Ni Taukei (SVT) and its Indian ally, the National Federation Party (NFP). Both coalitions made arrangements for preference transfers and the sharing of seats, and in almost half of the open constituencies lower-order preferences were required for a majority (Fraenkel and Grofman 2006, 633). The country’s first Indo-Fijian prime minister, Mahendra Chaudhry, came to power, backed by an overwhelming victory for the People’s Coalition, which ran on a platform of multiethnic classbased issues. The FLP won all the Indian seats as well as 18 open seats; its coalition partners won a combined 15 seats (Elections Fiji 1999).

Centripetalism

103

Yet the centripetal effects were both short-lived and overstated. The victorious People’s Coalition, which made cross-ethnic deals for lower-ranked preferences, did so without any agreement on ethnically sensitive issues, such as land rights (Fraenkel 2003, 242). This was by no means a coalition of commitment. Lal goes so far as to argue that Chaudhry’s FLP “abused the spirit of the Alternative Vote” since it had very little in common with the parties with which it formed a coalition other than a desire to defeat the SVT (2000, 291). Moreover, the SVT and NFP, both of which specifically campaigned on policies of cross-ethnic moderation, performed poorly. Both lost support in their own communities and neither picked up enough cross-group votes to compensate for the within-group loss (Fraenkel 2003, 239). In the aftermath of the election, two forces—discontent among many within the ethnic Fijian population at the presence of an Indian prime minister, combined with his push for more extensive Indian rights—served to deepen mistrust and further polarize the groups over already politicized issues like the question of landownership (Tarte 2000, 512). The regime proved unstable and within the year was ousted by indigenous Fijian extremists led by George Speight, a part Fijian businessman who used the rhetoric of Fijian nationalism and the need to protect the indigenous populations from the dangers of “an Indian government” (Lawson 2004, 531). This in turn prompted a period of military rule and then the installation of a caretaker government. The constitution was reinstated and elections were held in August 2001. This time there was less willingness to engage in cross-ethnic moderation. Extremists captured the majority of seats; the hard-line ethnic Fijian party, the Soqosoqo Duavata ni Lewenivanua (SDL) led by the interim prime minister Laisenia Qarase, was elected with only negligible Indian support. It controversially cast aside the constitutional stipulation allowing any party above 10 percent to participate in the cabinet. Ethnicity became a major campaign issue, with the SDL arguing that the country was “not yet ready for a non-Fijian prime minister” and that it might take another twenty to thirty years before it would be ready (Lal 2007, 141). Moreover, not only were the moderate parties defeated, but the election also saw the number of first preferences for the radical ethnic parties increase. Interethnic vote transfers dropped from 70.5 percent in 1999 to 46.7 percent in 2001, and in both elections more vote transfers flowed from moderates to radicals than the other way around, 50.1 percent in 1999 and 39.8 percent in 2001, respectively (Fraenkel and Grofman 2006, 647).

104

Allison McCulloch

A similar trajectory followed in the May 2006 election where voting patterns confirmed further polarization between the two communities. Moderates fared poorly, and a two-party system emerged. All but four seats were split between the SDL and the FLP (Fraenkel and Grofman 2007, 7). The SDL and FLP swept their respective communal constituencies and split the open ridings. The SDL won all the Fijian communal seats and 13 open constituencies and attracted over 80 percent of the ethnic Fijian vote but only 3 percent of the Indo-Fijian vote. Similarly, the FLP captured all the Indian communal seats and 12 open seats. The party polled 81 percent among IndoFijians but only 6 percent among ethnic Fijians (Narsey 2006). The SDL formed a multiethnic cabinet, with the participation of FLP members, though it was made unstable by the proposal of controversial legislation on indigenous fishing rights and amnesty provisions for perpetrators of the 2000 coup (Pareti 2006). The end of the year brought another coup, with the military arguing that since bad governance and corruption were destroying the country, it had to embark on a “clean-up campaign” (Lal 2007, 151). The two coups since 2000 and the resulting instability cannot be solely attributed to the adoption of AV, yet it cannot plausibly be maintained that it has contributed to ethnic moderation or to long-term political stability. The lesson from the centripetalist experience in Fiji is that as groups become more polarized and divisions deepen, the moderating effects of AV are less likely to be realized. Thus while AV was adopted with the intent of fostering moderation, its implementation has had the opposite effect (cf. Stockwell 2005). Over the course of the three elections, the electoral process contributed to a consolidation of extremist support. Preference transfers tended to benefit the SDL and FLP rather than the moderates. In 1999, 36 seats were decided by preference transfers, but in only four of those ridings were moderates able to leapfrog over extremists (Elections Fiji 1999; Fraenkel 2003). In 2006, preferences were counted in 11 constituencies, but in only one riding did the winner overtake the candidate with the most first preferences (Elections Fiji 2006). Centripetalists argue that AV is appealing because it is based on incentives rather than on constraints. Politicians, as rational actors, will do what is required to get elected. That is, politicians “like being elected and reelected— not exactly farfetched assumptions” (Horowitz 1991, 197). And, in fact, the assumptions are not farfetched; politicians generally do enjoy being elected. But the logic works both ways; in many cases of severe division, politicians find it profitable to appeal to the extremes, not to the middle. Where first

Centripetalism

105

preferences are cast for extremist parties, there is no guarantee that moderate parties will benefit (Fraenkel and Grofman 2006, 630). This appears to have been the case in Fiji. Any marginal gains made in the 1999 election have since been lost. Voters consolidated behind the two main ethnic parties, which tended toward extremist platforms. AV did not benefit the moderate politicians. Instead, it made extremism and voting along ethnic lines politically rewarding and in so doing has contributed to political instability. Fiji may be the most centripetalist-inspired model yet, but its practice does little to prove the approach beneficial.

The Contextual Imperatives of Centripetalism As a whole, the cases highlighted here suggest that AV has an unpromising track record in terms of promoting moderation and political stability in deeply divided places. Specifically, a preexistent level of heterogeneity and a preexistent level of moderation are needed for centripetal incentives to be realized. In each of the cases discussed, centripetal institutions have been adopted in situations that lack these contextual factors. For centripetal logic to work, constituencies must be ethnically heterogeneous (cf. Reilly 2001). There should be an appropriate demographic breakdown wherein groups are geographically interspersed, otherwise AV is prone to overly majoritarian results and unlikely to mitigate conflict. AV privileges the parties with the largest numbers. Where voting occurs along group lines, party support may be read as proxy for ethnicity. Parties from the majority group have a tremendous advantage in this regard, as they are closer to (or may even surpass) the electoral threshold than are parties that represent minorities. It thus becomes easier to get elected, irrespective of whether they appeal to the minority or not. In the Fiji open constituencies, for instance, the respective groups achieved overwhelming majorities, lessening the need for cross-ethnic appeals. A corollary to heterogeneous constituencies is the idea of political party proliferation. Preferential voting will only be conducive to moderation if intragroup divisions are salient and there is a wide range of political parties vying for election (Horowitz 1991, 194). Where multiple parties compete for support within each community, it will be easier to form preelection coalitions across the ethnic divide (Horowitz 2000, 598). Some sections of a group will support extremists, while others will want to promote cross-group

106

Allison McCulloch

cooperation and will thus give their support to moderates. This rightly assumes that no ethnic group is a monolithic entity but is instead characterized by varying degrees of opinion on interethnic cooperation. A heterogeneous party system will allow for more fluidity in terms of voter preferences and will consequently open up the political space for moderation. Yet a major irony in the commendation of AV is that such a system contains disincentives for party proliferation: like all majoritarian systems it discourages party fragmentation because of the high threshold required to win representation. The effective number of parties is more likely to increase under systems of proportional representation than under a majoritarian one like AV. This is precisely what happened in Fiji. While the 1999 election was marked by a range of parties willing to make cross-community electoral pacts, by the time of the 2006 election the effective number of parties had been reduced to two, both with extremist tendencies. In addition, centripetal recommendations need to take into account deep division. For majoritarian-preferential systems to produce moderate outcomes, there must already be a well-established core of moderate politicians and voters (cf. Reilly 2001). The lesson of deep division thus suggests a paradox. While the prior establishment of moderation functions as a contextual imperative for the centripetal recommendation, deeply divided places are characterized by a dearth of moderation. Indeed, it is the extreme polarization between groups that sets them apart as deeply divided places. Most important, centripetalists concede not only that the sort of moderation needed is unlikely to be found in the most deeply divided places—“it may be difficult to create or even envisage conditions under which electors of one group would be prepared to vote for candidates from another”—but that “it cannot invent moderation where none exists” (Reilly 2001, 181, 167). Moreover, AV is intended, through its facilitation of coalitions of commitment, to exclude extremists from power. That is, it is meant to limit the electoral strength of those who may make cross-community cooperation difficult. While this may appear to be an attractive proposition, it is problematic at a practical level as the center is liable to be quite small in deeply divided places where it is generally the extremists who enjoy popular support. This is why the SDS was able to easily win the 2000 presidential election in the RS, why Rajapaksa and not Wickremesinghe won the 2005 presidential election in Sri Lanka, and why the SDL and the FLP have been able to monopolize the support of their respective communities in Fiji. The

Centripetalism

107

exclusion of radicals, then, is problematic both from a standpoint of democratic accountability and for the prospects of political stability. To exclude the parties democratically chosen by the voters not only denies large portions of the population authentic representation and is thus undemocratic, it also actively supplies the excluded groups with incentives for instability. The centrist coalition has to contend with the outbidding and opposition of omitted extremists and, in some cases, with the reactionary violence that the formation of such coalitions may cause. This is what happened to Northern Ireland’s Sunningdale coalition in 1974. The cross-community centrist coalition faced a number of destabilizing forces, including the fact that many Unionists were leery of their own party’s cooperation with nationalists. Extremist outflanking outside the government and the perpetuation of intergroup violence also delegitimized the coalition, which collapsed after only a few months (McGarry and O’Leary 2006). The coalition formed following Fiji’s 1999 election similarly caused a backlash from extremists who then overthrew the government in a coup. It is difficult to see how extremists would contribute to the political stability of a system from which they are excluded (cf. Reynolds 1995, 92). Excluding extremists does not make them go away. Finally, there is the difficulty in centripetalism’s depiction of voters as the vehicles of moderation. Centripetalism assumes not only moderation on the part of politicians but also that moderate tendencies will be displayed at the voter level (Reilly 2001, 176). Yet where levels of violence have been high, where atrocities have been committed, voters may not be, nor should they be expected to be, moderate in attitude, thought, or action, at least not initially. At a basic level, it is fundamentally unrealistic and intrinsically unfair to expect, for instance, a Bosniak in the RS or a Tamil in Sri Lanka to be satisfied with Serb and Sinhalese representation, respectively. Th is is the problem with prescribing majoritarian institutions like AV in deeply divided places. It overlooks the basic mistrust and insecurity that characterize intergroup relations. Moderation at the voter level is the outcome of extended periods of cross-community cooperation, not something that precedes it. This is not to suggest that moderation does not bring about political stability—in fact, quite the opposite. The willingness of groups and their leaders to engage in cross-community accommodation for the purpose of political stability presupposes some moderation. The point, however, is that centripetal institutions like AV cannot induce the sort of moderation they require in order to promote political stability.

108

Allison McCulloch

Wither Centripetalism? The analysis here suggests that centripetalism in general (and AV in particular) is contextually specific: the model is simply not appropriate in cases of deep division. This does not require an outright rejection of the model, but it does require the concession that the centripetal universe is deeply constrained by context. If the model cannot create the sort of moderation it requires to function, the implication is that it can only be pursued in cases that already have a core of moderate leaders and moderate followers. Further, the moderate middle should take up the most space on the political spectrum: that is, it should already have majority support in order to effectively counter the power held by extremists. Voters must also accept the pursuit of moderate politics on the part of their leaders. Th is precludes places where there has been severe and protracted ethnic conflict. Constitutional designers must also be mindful of demographic configurations. Specifically, AV is particularly inappropriate where one group has an overwhelming majority, as in the RS, or where there are two dominant groups, as in Fiji and Sri Lanka. Given these constraints, it is possible to speculate at least two demographic scenarios where centripetal measures might work: in places with extreme fractionalization, that is, a high number of ethnic groups, where no single group constitutes the majority, or in places with a low number of groups but where such groups are territorially dispersed (Reilly 2001, 187–92). Where groups are interspersed, as in many immigrant-dominated polities, constituencies may be designed to encourage majority candidates to take minority concerns into account and to encourage minorities to accept such majority representation. Territorially dispersed groups are also suitable for centripetalist measures because they face logistical obstacles to mobilization and are thus unable or disinclined to push for more accommodative measures (McGarry, O’Leary, and Simeon 2008, 85). The case of Papua New Guinea (PNG) may be illustrative. Benjamin Reilly has convincingly shown that in preindependence PNG, AV was conducive to moderation whereas a shift to a single-member plurality system after independence contributed to instability (Reilly 2000). Here the model was aided by demography: the country is characterized by ethnic fractionalization, with the number of groups estimated at about two thousand. This ensured that no group dominated government (Reilly 2000/2001). Yet it is useful to recall that PNG’s AV experience was at a time when the country was not deeply divided and when the political party system was not fully

Centripetalism

109

developed. The country reintroduced a limited form of AV for elections in 2007, which also displayed some moderate results. We may also look to Nigeria’s Second Republic, which did not employ preferential voting but did use creative distribution thresholds in its presidential elections that included incentives for cross-group appeals, as well as to the use of AV in mayoral contests in diverse cities, such as San Francisco and London. If the logic works as intended, centripetalism may create a scenario in which an ethic of multiethnicity would thrive, wherein both individual and group identities are respected. There is little to criticize in such a scenario; the point, however, is to find such a set of auspicious circumstances. The model rests on an inherently plausible logic—moderation as the avenue by which to facilitate political stability—but the institutions proffered by centripetalists do not promote the sort of incentives anticipated. Th is being so, the contextual imperatives severely curtail the universe of cases to which the model can apply. While this may appear pessimistic, it does not require us to be dismissive of Horowitz’s claims, but it does suggest that the model is in need of a reconfiguration as a model for the management of mild rather than deep division, perhaps as a preventive measure rather than as a strategy of postconflict accommodation. Where divisions are deep, minority officeholding, as under consociational or pluralist federal arrangements, may be required. In such places, however, minority influence does not suffice. References Belloni, Roberto. 2004. “Peacebuilding and Consociational Electoral Engineering in Bosnia and Herzegovina.” International Peacekeeping 11, 4: 334–53. Belloni, Roberto. 2007. Statebuilding and International Intervention in Bosnia. New York: Routledge. Bieber, Florian. 2006. Post-War Bosnia: Ethnicity, Inequality and Public Sector Governance. New York: Palgrave Macmillan. Bose, Sumantra. 2002. Bosnia After Dayton: Nationalist Partition and International Intervention. London: Hurst and Company. ———. 2007. Contested Lands: Israel-Palestine, Kashmir, Bosnia, Cyprus, and Sri Lanka. Cambridge, MA: Harvard University Press. DeVotta, Neil. 2000. “Control Democracy, Institutional Decay, and the Quest for Eelam: Explaining Ethnic Conflict in Sri Lanka.” Pacific Affairs 78: 55–76. Elections Fiji. 1999. “Ethnic Breakdown of Open Constituencies.” http://www.elections .gov.fj/results1999/open_ethnic.html (accessed 5 May 2007). ———. 2006. “Elections by the Count 2006.” http://www.elections.gov.fj/results2006 .html (accessed 30 May 2008).

110

Allison McCulloch

European Union Election Observation Mission (EUEOM). 2005. “Preliminary Statement: Sri Lanka Presidential Election 2005.” http://www.eueomsrilanka.com (accessed 7 March 2008). Field, Michael. 2005. “Fiji: The Indian Factor.” Islands Business. http://www.islands business.com (accessed 5 March 2008). Fraenkel, Jon. 2003. “Electoral Engineering and the Politicisation of Ethnic Friction in Fiji.” In Sunil Bastian and Robin Luckham, eds., Can Democracy Be Designed? The Politics of Institutional Choice in Conflict-Torn Societies, 220–45. London: Zed Books. Fraenkel, Jon, and Bernard Grofman. 2006. “Does the Alternative Vote Foster Moderation in Ethnically Divided Societies? The Case of Fiji.” Comparative Political Studies 39, 5: 623–51. ———. 2007. “The Merits of Neo-Downsian Modeling of the Alternative Vote: A Reply to Horowitz.” Public Choice 133, 1–2: 1–11. Horowitz, Donald L. 1990. “Making Moderation Pay: The Comparative Politics of Ethnic Conflict Management.” In Joseph Montville, ed., Conflict and Peacemaking in Multiethnic Societies, 451–75. Lexington: Lexington Books. ———. 1991. A Democratic South Africa? Constitutional Engineering in a Divided Society. Berkeley: University of California Press. ———. 1993. “Democracy in Divided Societies.” Journal of Democracy 4, 4: 18–38. ———. 1997. “Encouraging Electoral Accommodation in Divided Societies.” In Brij V. Lal and Peter Larmour, ed., Electoral Systems in Divided Societies: The Fiji Constitution Review. Canberra: National Centre for Development Studies. ———. 2000. Ethnic Groups in Conflict. 2nd ed. Berkeley: University of California Press. International Crisis Group (ICG). 2000. Bosnia’s November Elections: Dayton Stumbles. Sarajevo and Brussels: Balkan Report No. 104, 18 December. ———. 2003. Bosnia’s Nationalist Governments: Paddy Ashdown and the Paradoxes of State Building. Sarajevo and Brussels: Balkan Report No. 146, 22 July. Lal, Brij V. 2000. “Constitutional Engineering in Post-Coup Fiji.” In Andrew Reynolds, ed., The Architecture of Democracy, 267–92. Oxford: Oxford University Press. ———. 2007. “ ‘Anxiety, Uncertainty, and Fear in Our Land’: Fiji’s Road to Military Coup, 2006.” Round Table 96, 389: 135–53. Lawson, Stephanie. 2004. “Nationalism Versus Constitutionalism in Fiji.” Nations and Nationalism 10, 4: 519–38. Manning, Carrie. 2008. The Making of Democrats: Elections and Party Development in Postwar Bosnia, El Salvador, and Mozambique. New York: Palgrave Macmillan. McGarry, John, and Brendan O’Leary. 2006. “Consociational Theory, Northern Ireland’s Conflict and Its Agreement 2: What Critics of Consociation Can Learn from Northern Ireland.” Government and Opposition 41, 2: 249–77. McGarry, John, Brendan O’Leary, and Richard Simeon. 2008. “Integration or Accommodation? The Enduring Debate in Conflict Regulation.” In Sujit Choudhry, ed., Constitutional Design in Divided Societies: Integration or Accommodation? Oxford: Oxford University Press.

Centripetalism

111

Mishler, William, Steven Finkel, and Pradeep Peiris. 2007. “The 2005 Presidential and 2004 Parliamentary Elections in Sri Lanka.” Electoral Studies 26, 1: 205–9. Narsey, Wadan. 2006. “Fiji Elections: The People Have Spoken,” Islands Business. http://www.islandsbusiness.com (accessed 5 March 2008). Orga nization for Security and Cooperation in Europe (OSCE). 2000. “Uncertified Results Announced in Race for RS Presidency, Vice-Presidency.” 23 November. http://www.oscebih.org/public/default.asp?d=6&article=show&id=935 (accessed 18 June 2007). Pareti, Samisoni. 2006. “The Qarase Gamble.” Islands Business. http://www.islands business.com (accessed 5 March 2008). Reilly, Ben. 2000/2001. “Democracy, Ethnic Fragmentation and Internal Conflict: Confused Theories, Faulty Data and the ‘Crucial Case’ of Papua New Guinea.” International Security 25, 3: 162–85. ———. 2001. Democracy in Divided Societies: Electoral Engineering for Conflict Management. Cambridge: Cambridge University Press. ———. 2002. “Electoral Systems for Divided Societies.” Journal of Democracy 13, 2: 156–70. Reynolds, Andrew. 1995. “Constitutional Engineering in Southern Africa.” Journal of Democracy 6, 2: 85–99. Sisk, Timothy D. 1996. Power Sharing and International Mediation in Ethnic Conflicts. Washington, DC: United States Institute of Peace Press. Sri Lanka Freedom Party (SLFP). 2005. Mahinda Chintana: Manifesto of the SLFP. http://www.priu.gov.lk/mahindachinthana/MahindaChinthanaEnglish.pdf (accessed 27 August 2012). Stockwell, Robert F. 2005. “An Assessment of the Alternative Vote System in Fiji.” Commonwealth and Comparative Politics 43, 3: 382–93. Tarte, Sanda. 2000. “Fiji.” Contemporary Pacific 12, 2: 507–15. United Nations High Commissioner on Refugees (UNHCR). 2006. “Return Summary to Bosnia and Herzegovina from 01/01/1996 to 30/09/2006.” http://www.unhcr .ba/return/pdf%202006/SP_09_2006.pdf (accessed 28 December 2007).

CHAPTER 4

Electoral Engineering for a Stalled Federation Kris Deschouwer and Philippe Van Parijs

On June 13, 2010, a new Belgian federal parliament was elected. The elections had been held earlier than scheduled following the resignation of the federal government that had been unable to fi nd an agreement between the Dutch-speaking and the French-speaking parties on the boundaries of the Brussels electoral district. That issue remained on the table when negotiations started to form a new federal coalition, together with a possible change in the distribution of powers between the federal and the sub-state level, and a possible adjustment of the financial equalization mechanisms. Parties of both language groups in Belgium deeply disagreed on all these matters, and it took no less than 18 months—exactly 541 days—to find an agreement and put a new federal government into place. It was widely noticed that Belgian federal government formation managed to take even longer than in Iraq after its controversial elections of 2010. External pressure finally pushed Belgian parties toward a compromise, in particular the rising interest rates being paid for the financing of Belgium’s public debt and the downgrading of its status by the rating agencies. This story, following the long process of Belgian coalition-making of 196 days in 2007, shows that executive formation and the executive’s ability to function can be an exceptionally laborious enterprise. The most recent protracted government formation faced the usual challenge of bridging the different views and ideologies of the parties that have to govern together in a coalition, but the gridlock resulted from a clash between the conflicting de-

Electoral Engineering

113

mands of two sets of parties, each corresponding to one of Belgium’s two main language groups. The Dutch-speaking political parties had promised their electorate that a government could only be formed on the condition that further devolution would be secured. On the other hand, the French-speaking political parties had promised their own electorate that they would not accept these new demands. Both tried as long as possible to stick to these electoral pledges, resulting in a situation in which neither party would give in. This kind of governmental crisis is not a new phenomenon in Belgium. To the contrary: long and painful negotiations between the two language groups have become a normal feature of the system. The gradual transformation of the unitary Belgium into a federal state was a long and sometimes painful process. For example, between 1977 and 1981 there were no less than seven cabinets, all falling apart because they were not able to find an acceptable compromise about the institutional hardware of a new Belgium. When in 1993 the first article of the constitution was changed to define Belgium as a federal state, political stability seemed to have been restored. Between 1991 and 2007 all four federal governments went to the very end of their term without being torn apart by the tensions between Francophones and Flemings. Yet the spectacular return of political gridlock in the aftermath of the June 2010 election suggests that there is still something wrong with Belgium’s institutional capacity to deal with its linguistic and territorial divisions. In this chapter we argue that the design of the electoral system is one of the major problems because it offers insufficient incentives to display the spirit of accommodation that is needed for a divided society to be smoothly governed. The first section offers a short background sketch of the basic ingredients of the Belgian divide. The second section describes the institutional solution that was gradually put into place at the end of the twentieth century. The third section identifies the solution as typically consociational, with full emphasis on segmental autonomy and power-sharing devices. However, emphasizing autonomy and inclusion of both groups in the decision-making process does not guarantee smooth functioning and even less a high ability for decision making and change. The last section presents an electoral reform—a countrywide electoral district—that we have been advocating along with colleagues from all Belgian universities that would go some way, we shall argue, toward remedying the shortcomings of Belgium’s federal setup.

114

Kris Deschouwer and Philippe Van Parijs

Belgium’s Linguistic and Territorial Divide The conflict that led to the territorial transformation of the Belgian state is in the first place a linguistic matter. When Belgium was created in 1830 after seceding from the short-lived Kingdom of the Low Countries, the political elite of the new state spoke French throughout the country, and French was therefore also the obvious choice as the language of government and administration. However, the majority of the population spoke no French. Belgium is cut in two by an old language frontier than runs west to east (Geyl 1962, 211). It divides the country into a southern area where French is spoken (now called Wallonia) and a northern area where Dutch is spoken (now called Flanders). Nonetheless, the adoption of French as the official language was regarded as self-evident. Not only was French the language of the state-building elite— including those living in the north—but in 1830 French was also the language of modernity and liberalism, as well as the lingua franca of royal courts and diplomatic circles. Dutch, on the other hand, was the language of the northern Low Countries, that is, precisely the country from which the new Belgium had seceded. It was also perceived as the language of Protestantism, the dominant confession in the northern Low Countries, whereas both the Dutch-speaking and French-speaking parts of the new Belgium were, if religious at all, homogeneously Catholic. It also went without saying that Brussels would become the capital city of Belgium. The city, however, is located north of the language frontier (see Figure 4.1). Its role as the capital city of the newly independent state rapidly strengthened its predominantly Francophone character and fed its gradual expansion into its historically Dutch-speaking hinterland. These facts are the raw material for understanding Belgium’s modern language conflict, and conflict it becomes when in the course of the nineteenth century the inhabitants of the part of the country where varieties of Dutch that are the vernacular of the mass of the people started asking for the formal recognition of Dutch as a second official language of Belgium, in particular for the right to use Dutch for educational, administrative, and political purposes in the northern part of the country. This process almost naturally led to a territorial solution (Murphy 1988). From the 1920s on, the rules governing the use of language by public authorities and the language used as the medium of education relied on the creation—or rather the acceptance of the existence—of three linguistic territories: one for Dutch,

The Regions and Communities of Belgium

Flemish Community

Flemish Region

French Community

Walloon Region

German-speaking Community

Brussels Region

N

Figure 4.1. The regions and communities of Belgium. Source: http://www.Belgium.be.

116

Kris Deschouwer and Philippe Van Parijs

one for French, and one—the Brussels area—where both languages can be used. Obviously a territorial organization requires the drawing of boundaries. This is seldom easy when ethnic, linguistic, or religious tensions are present. Belgium has been no exception. Two conflicting principles have been and continue to be invoked. One stipulates that the language to be used for official business is determined permanently on the basis of the historical distinction between the northern and the southern parts of the country. The alternative principle stipulates that official linguistic boundaries can and should be adjusted in line with changes in the composition in the population. According to this principle, the boundaries can be altered in order to accommodate demographic movement and linguistic shifts. Whenever they did change, they led to the transfer of historically Dutch-speaking territory into the bilingual area and sometimes eventually into the Francophone area. It is not surprising that the historical principle has tended to be supported mostly by Dutch speakers, who feel that a protected territory is needed to safeguard their lower-status language. French speakers tend to invoke the principle that official boundaries should track real-life trends, including the spread of the stronger language. The use of language has therefore gradually been organized on territorial premises but without agreement on the operational principles for drawing the territorial boundaries. Different languages and different views on the way in which language shift needs to be given free rein or hemmed in are not the only differences between north and south Belgium. In terms of economic development, Flanders and Wallonia have differed from the very early days of the Belgian state. Industrialization came quite early and was very much concentrated in the south, while the north remained agricultural for much longer. But after the end of World War II the steel-and-coal-based Walloon economy started declining, whereas Flanders attracted investments in new economic activities— for which the harbor of Antwerp was, and remains, a major asset. The economic balance of the country therefore shifted. In the early 1970s, GDP per capita became higher in Flanders than in Wallonia, and since then the gap has continued to increase. In 2008 GDP per capita was € 24, 000 in Wallonia and € 32, 000 in Flanders, while the unemployment rate was 7 percent in Flanders and 12 percent in Wallonia. This emphasis on Belgium’s north-south economic divide is somewhat misleading, however, because well over one-quarter of the country’s GDP is produced on less than 1 percent of its territory, in the Brussels region and its immediate surroundings. But the dramatic shift in

Electoral Engineering

117

the balance of economic power between Flanders and Wallonia is nonetheless a crucial ingredient in Belgium’s present situation. By contrast, the political difference between north and south has remained relatively stable. As soon as all layers of the population were allowed to participate in parliamentary elections, the north and the south returned quite different results. At the first elections with some sort of universal male suffrage in 1894, the 72 Flemish seats and the 18 Brussels seats all went to representatives of the Catholic Party. Of the 62 Walloon seats, 14 went to the Catholic Party, 20 to the Liberal Party, and the remainder to the Socialist Party, which first entered Parliament with 28 representatives, all elected in Wallonia. Although the differences did not remain as sharp—not least because of the introduction of proportional representation in 1900—the two parts of the country still display significantly different electoral behavior. Table 4.1 shows the results of the federal elections of 2010 for Flanders and Wallonia separately per party family. These reveal that for each party family the results are very different. To these different electoral results one must add another crucial ingredient of the Belgian problem. The results in Table 4.1 are presented per party family, and not per party, because there are no countrywide parties any more. The traditional parties—Christian Democrats, Liberals, and Socialists—fell apart into two separate and unilingual parties between 1968 and 1978. The Greens and the populist radical right parties are younger but have never existed as Belgian parties. They have developed in the party system of each of the language groups separately. For the federal elections, it is only in the central electoral district of Brussels that the parties of the two language groups compete with each other.

Table 4.1. Results of the Elections to the Federal Parliament in 2010 for Flanders, Wallonia, and Brussels (percentage of the votes)

Christian Democrats Socialists Liberals Regionalists Populist radical right Greens Others

Flanders

Wallonia

Brussels

17.0 15.0 14.0 28.4 12.6 7.0 6.0

14.6 37.6 22.2

14.8 25.5 29.1 0.6 3.7 19.2 7.1

1.4 12.3 11.9

118

Kris Deschouwer and Philippe Van Parijs

The Institutions of the Belgian Federation The unitary Belgium of 1830 is now long gone. Several constitutional reforms have rebuilt the Belgian state into a federation. The linguistic regions that were created to regulate the use of language provide its building blocks, albeit in a fairly complex way. Belgium is both a federation of three territorial regions and three language communities. The regions are called Flanders, Wallonia, and the Brussels Capital Region, with clear (though not uncontested) territorial borders. They have been given a broad set of powers, for example, over environmental policy, public works, public transportation, housing, and important aspects of economic policy. The three language communities are called the Flemish Community, the French Community, and the Germanspeaking Community. In general, the communities offer services to individuals in the areas of education, culture, and welfare policy. The Flemish Community offers these services in Flanders and in Brussels. The French Community offers them in Wallonia and in Brussels. Hence in the territory of the Brussels region the operations of the two main language communities overlap. The Germanspeaking Community, composed of approximately seventy-three thousand people living in two areas next to the German border (transferred from Germany to Belgium after World War I), offers its services in those areas, which are part of the Walloon region. The twofold nature of the federation is rather awkward, but it is a subtle compromise. At a first level, it often presented a deal between pro-community Flemings—who prefer a one-against-one conflict— and pro-region Francophones—who might be advantaged by a two-againstone configuration. At a deeper level, it constitutes an attempt to articulate two types of concerns. The community component should assuage, at least for the time being, both many Flemings’ fear they will lose all control over Brussels, where the Flemish residential presence keeps shrinking, and many Francophones’ fear of a weakening of the solidarity between Brussels and Wallonia. On the other hand, the regional component reflects some awareness of the fact that efficient policymaking requires all decentralized powers to be exercised by one government, responsible to all those sharing the same territory. The first constitutional reform of 1970 laid down this double structure, but it was only in 1989 that Brussels was given the status of a region, and only in 1995 that the first direct election of the three regional parliaments took place. The constitutional reform of 1970 was also extremely important for the changes it introduced in the functioning of the central state. The

Electoral Engineering

119

rules laid down at that time define the way in which decision making operates at the federal level today, as well as the way in which the constitution can be further modified. Two principles were then deeply enshrined into the Belgian political system: a neat separation between the language groups and an obligation to govern together. The separation between the language groups was introduced both into Parliament and in the government. All members of the House of Representatives, whatever their origin, belong to either the Dutch- or the Frenchlanguage group. That membership is defined by the territory in which the members of Parliament (MPs) have been elected. Those elected in constituencies of the Flemish region belong automatically to the Dutch-language group, and those elected in Wallonia belong automatically to the Frenchlanguage group. For MPs elected in the central Brussels district the language in which they take their oath defines the group to which they belong. The full separation of the party system into Francophone and Flemish parties actually predefines the choice that these MPs will make. Subsequent reforms of the electoral system and of Parliament have only reinforced this split and the role ascribed to the descriptive representation of language groups. Since 1995, the Senate has been elected in two electoral districts: one for Flanders and the central Brussels district, and one for Wallonia and the central Brussels district. The Belgian members of the European Parliament are elected in the same way and, as in the Senate, with a fi xed number of seats available for each language group and an overlap in the Brussels area, where voters can choose either of the two districts. The members of the federal government also clearly belong to one of the two language groups. Again the split party system leaves no doubt about the membership. Since 1970, therefore, not one single politician formally represents voters outside his or her language group. Politicians might claim to do so, but their position in the institutions gives them a clear and unambiguous label. The neat separation of the language groups allows for the organization of the second principle: the obligation to include both groups within the federal decision-making process. This is done in a variety of ways. The federal government (i.e., Belgium’s cabinet) has to be composed of an equal number of Francophone and Dutch-speaking ministers. Only the prime minister is supposed to be linguistically a-sexué (as the semiofficial terminology puts it), but the party to which he belongs leaves no doubt about his

120

Kris Deschouwer and Philippe Van Parijs

linguistic status. Since 1970 nearly all prime ministers have been Dutch speakers, except for a one-year period in 1972–73, a two-month period in 1978, and the prime minister of the federal government formed after the long negotiations of 2010 and 2011. The rule of decision making in the federal government is unanimity. The cabinet never votes. This linguistic “parity” assures the governing of the country by the two major language groups. In Parliament the Flemish group is larger than the Francophone one, reflecting the 60:40 percent demographic ratios. However, it cannot use that majority to impose its will on the minority. The normal rule of decision making for the federal House of Representatives is simple majority, but the minority has a veto power. It is called the “alarm bell procedure.” Whenever three-quarters of a language group declares that a proposal might be accepted that harms the interests of that group, it can activate the alarm bell. The parliamentary procedure is then suspended for thirty days, during which the government needs to find a solution. And with parity in its composition and unanimity as the decision-making rule, the solution of the government can only be one that is acceptable for both language groups. If no solution is found, the government will have to resign. But to form a new government, possibly after electing a new parliament, both language groups will still need to find a compromise. Moreover, to change the constitution a two-thirds majority is needed. Yet for most articles that define the political institutions of regions and communities, and for the so-called Special Laws that implement these basic principles, a majority is needed in each language group, that is, a concurrent majority, as well as an overall two-thirds majority. The same logic of strict separation of the political personnel into language groups and the obligation to govern together and to avoid a veto by one of the language groups has been built into the institutions of the Brussels region. It has indeed become a full-fledged region, as the Francophone parties requested, but institutionally speaking it has not become a Francophone region, as the Flemish parties feared. Dutch-speaking parties are guaranteed 17 of the 89 seats in the regional parliament and two out of the four ministerial positions in the regional government, while the ministerpresident is supposed to be, like the federal prime minister, linguistically “asexual.” To some extent this picture is a mirror image of the federal institutions. The Brussels institutions display the Belgian logic of separation and inclusion. This logic is a consociational logic, albeit one in which the parity principle usually outweighs the proportionality principle. With two actors, the proportional distribution of power and resources is not the most impor-

Electoral Engineering

121

tant device. The common agreement needed for governing lays far greater stress on the right of both actors to be present in the decision-making process and hence on the veto power of each language group.

Consociational Belgium In a piece written three decades ago, Arend Lijphart left no doubt as to how he wanted to label Belgium: “What is remarkable about Belgium is not that it is a culturally divided society—most of the countries in the contemporary world are divided into separate and distinct cultural, religious, or ethnic communities—but that its cultural communities coexist peacefully and democratically. What is more, Belgium can legitimately claim to be the most thorough example of consociational democracy, the type of democracy that is most suitable for deeply divided societies” (1981, 1). If the Belgian federation— still very much in the making when Lijphart wrote—is consociational, it needs prudent leaders willing to accommodate and to govern with the leaders of the other language group. The devolution of powers to the regions and communities, however, has taken away from joint decision making quite a few powers for which the formulation and implementation of a common policy has been or would be difficult. For the remaining federal powers a common policy is required and therefore an agreement is needed. That is obviously also the case for all matters relating to the state structure itself. Only an agreement between elites willing to compromise can offer a way out. Functioning consociational democracy requires prudent leadership. Prudence may result from a learning process—from the awareness that a conflictual attitude leads to total gridlock and perhaps even to violent clashes (Lijphart 1977, 99). Prudent leaders are willing to bridge the gap between the potentially deep differences that divide the population. Functioning consociational democracy also requires that the elites want to keep the political system alive and value the latter’s survival above the interests of their own groups. It means that they are willing and able to play a double role, to be advocates on behalf of their own rank and file and compromise seekers at the elite level. Compromise therefore needs to come at an acceptable price. If compromise leads to a substantial loss of trust (and hence votes) from the followers, prudent leadership is not likely to develop (Horowitz 1985, 347). This is indeed one of the major problems facing Belgium. If we look back at the last fi ft y years, we can observe the capacity to find compromises

122

Kris Deschouwer and Philippe Van Parijs

when needed. It was never easy, but exactly at times when terms like “regime crisis” were being used by political commentators, a new, often unforeseeable compromise was found. It is important to note, though, that these agreements were reached in a political system that was not yet a fullfledged federation. Political agreements had to be found in the central government (and Parliament). The absence of an agreement acceptable to both language groups meant the end of the current central government or the nonformation of a government. That could go on for a while, but the longer it took, the more problematic it became for all parties. The very high systemic price to be paid for the absence of an agreement—for instance, in terms of pressure on the currency or public sector deficits—provided the incentive for the elite to be both creative and accommodating, and hence to concoct an acceptable compromise that could keep the system going again for a while. This institutional environment has changed in ways that have tended to reduce the pressure to find a compromise and increased the probability of long and enduring political crises. Since Belgium has become a federation, is has more than just a federal government. Many powers are now in the hands of the regions and communities. As a result, the formation or survival of the federal government is less important. In other words, the pressure to display an accommodating attitude in what used to be the only center of power is far weaker than it was before (Jans 2001; Swenden and Jans 2006; Deschouwer 2005, 2006, 2012). This is also the case because of the expansion of the powers of the European Union. The melting of the Belgian franc into the euro, for example, strongly limits the dangers of a financial crisis when the country is not able to produce or maintain a working government. With the shrinking of federal powers from above and below, it is both less important and more difficult to form federal governments and to keep them in place. No less relevant have been the changes in the pattern of party competition. For a long time—until the end of the 1990s—there were two dominant parties in Belgium. Christian Democrats were by far the largest party in Flanders and therefore almost always were a governing party. The Socialists were by far the largest party in the south. The most natural coalition was therefore one between the Christian democratic and the Socialist families. Since the turn of the century, however, this domination has gone. In each of the two party systems, competition is very high. All potential governing parties are very much afraid of losing votes. Even a slight electoral decline can

Electoral Engineering

123

have important consequences. Consequently, party elites are more fearful than ever of having to pay the electoral cost of the compromises they accept. This increased electoral competition materialized precisely when the electoral cycles for the different levels of the federation became desynchronized. In 1995 and 1999 the federal and regional parliaments were elected on the same day, but since then they have developed their own rhythm of five years for the regional parliaments and four years for the federal parliament. In the absence of countrywide political parties, distinct from the linguistically defined parties that compete for the regional elections, the federal elections and the regional elections are not really different. As explained earlier, by the late 1970s all three Belgium-wide parties had divided into two separate parties, one Flemish and one Francophone. Consequently, whatever the type of election, the same parties compete for the electorate of their own language group. The next election for all parties is not the next election at the same level but the next election tout court. There have been elections in 2003 (federal), 2004 (regional), 2007 (federal), 2009 (regional) and 2010 (federal). This is driving all political parties into a nearly permanent state of electoral campaigning. As a result, the likelihood of an accommodating attitude on the part of politicians governing, or wanting to govern, at the federal level has been dramatically reduced.

A Countrywide Electoral District There is definitely something wrong with the way the Belgian federation functions. Its federal governmental level lacks decision-making and problemsolving capacity, and most suggestions to improve the functioning of the federal state defend a further devolution of powers to the regions and communities: if the federal government does not work, it should be given less work to do. This thinking fits in neatly with the trend that has characterized Belgium’s institutions since the 1970s: the gradual hollowing out of the powers of the central government. Suggestions to improve the decision-making capacity at the federal governmental level are seldom heard. There is, however, one idea that has surfaced now and then over the last couple of decades, was worked out in some detail shortly before the 2007 federal election, and quickly became the subject of a lively debate: the idea of creating a federal or countrywide electoral district for the federal elections.

124

Kris Deschouwer and Philippe Van Parijs

When in 1979 Belgium had to decide on the procedure for the election of the Belgian members of the European Parliament (MEPs), the idea of a countrywide electoral district appeared for the first time. It was suggested by the Flemish Christian democrat leader and then Belgian prime minister Leo Tindemans, who was hoping to win votes among both linguistic groups. His Francophone coalition partners were diffident, and the government opted instead for an election of Belgian MEPs in two separate unilingual communitywide electoral districts, in line with the classical “splitting” logic outlined earlier. The idea reappeared in the 1990s, as Belgium was becoming a true federation, but this time applied to federal elections (see, e.g., Van Parijs 2000a, 2000b). It did not arouse much interest, however, until a group of academics, known as the Pavia Group and coordinated by the authors of this chapter, drafted a detailed scheme, tested it among politicians and lawyers, and then presented it to the press on February 14, 2007 (Deschouwer and Van Parijs 2007; http://www.paviagroup.be/). The proposal was picked up by some parties, was fiercely attacked by others, and eventually made it to the institutional agenda.

A Truly Federal Parliament for a Truly Federal Government

The basic idea is simple and straightforward. Of the 150 members of the federal House of Representatives, 15 should be elected in an electoral district that covers the whole territory of the Belgian state—henceforth called the federal district. Now the federal House is elected in 11 districts, coinciding with the provincial boundaries. Almost all MPs are therefore currently elected in unilingual districts where the parties of only one language group compete. Once a federal electoral district is created, voters will have two votes. Their first vote will be cast for one of the lists—or some of the candidates featured on one of the lists—presented in a provincial electoral district. The distribution of seats among these districts will be distributed, as now, in proportion to the population of each province. A second vote will be cast for one of the lists—or some of the candidates featured on one of the lists— presented in the federal district, common to all voters, irrespective of where they live. Any candidate will be allowed to stand on both a provincial list and a federal list. And most, if not all, of the candidates on a federal list can be

Electoral Engineering

125

expected to do so, for the following reasons. They may not be sure of being among the fifteen elected in the federal district, and therefore it provides extra safety for them to be in a good position on a provincial list. Or they are certain to be among the fifteen elected, but if they enjoy such popularity their party would be foolish not to also place them on a provincial list. Thus fifteen out of the 150 people elected to the House will have a claim to being truly federation-wide MPs. But a far greater proportion of the 150 eventually elected, in all likelihood a significant majority among them, will have been candidates in the federal district. To win as many votes as possible in this district, it will be in their interest to campaign also in the other language group, with a fair chance of success if they manage to highlight their commitment to causes that are not divisive along linguistic lines. This will hold, in particular, for the top politicians of all the parties with the ambition to form and lead the federal government, those whose promises and declarations will be most binding for the action of the next government. Not only will their total personal vote affect, as it does now, their pecking order in their party and in the country, but this vote and the way it is distributed across the country will affect the legitimacy with which they will claim and exercise the functions to which they aspire. For this reason, the number of seats to be allocated in the federal district is not that important. It could conceivably be increased beyond fifteen. But if this is done without a corresponding increase in the total size of the House, the district magnitude in the provincial districts would drop and that would create higher thresholds for the smaller parties. The degree of proportionality would be severely reduced, and the constitution requires the electoral system to be proportional. On the other hand, increasing the number of seats in the federal House would be an unpopular measure unless combined with an appropriate compensation. Bear in mind that the full implementation of the federal structure in 1995 increased the total number of parliamentary seats—federal and regional—from 369 to 503. However, increasing the number of seats in the House might possibly be compensated by a reduction of the number of seats in other assemblies. The most attractive and most probable version of such compensation would consist of scrapping the direct election of part of the Senate—25 Dutch speakers and 15 French speakers—thus leaving a Senate composed exclusively of people elected to the regional parliaments. Whether fi fteen or more members of the federal House are elected in the newly created federal district, the reform sketched would significantly

126

Kris Deschouwer and Philippe Van Parijs

alleviate the democratic deficit from which Belgium’s federal system suffers. The current organization of elections without federal parties does not offer the possibility of a true dialogue between the governing elite at the federal level and the population of the federation as a whole. All those competing seriously in the federal district will face incentives to propose mutually acceptable solutions for institutional matters instead of simply expressing the demands of their own language group. A federal district would reintroduce preelectoral incentives—absent since the Belgium-wide parties fell apart—to display a disposition to compromise that is needed to govern, in powersharing fashion, at the federal level. The proposal aims thereby to strengthen the potential for prudent leadership and political accommodation by compensating for institutional developments that have dramatically weakened it. Given the absence of federal political parties, the emergence of a federal system that reduced the importance of the central government has seriously reduced the capacity of the country’s political elites to promote or at least accept the principles of power sharing.

Quota per Language Group

As is often the case with institutional engineering, however, it is crucial to anticipate the various political actors’ response to the proposed setup and to fine-tune it so as to avoid perverse effects. For this reason, the Pavia Group’s proposal fi xes the number of seats allocated to each language group in the federal district before the election. The proportions simply match as closely as possible the proportions of members of the House belonging to the two language groups in the previous legislature. If 15 seats are to be allocated, this means that 9 will go to Dutch speakers and 6 to French speakers. The lists put forward by the various parties in the federal district will accordingly consist of a maximum of 6, 9, or 15 names. Only lists containing 9 Dutch speakers and 6 French speakers can present 15 candidates. Some simple and sufficiently uncontroversial criterion for recognition as a French speaker or a Dutch speaker will be required. In light of past experience and bearing the threat of political sanctions in mind, sponsorship by three members of the relevant language group of the previous House should work well. The allocation of seats between the lists and the candidates can proceed using the standard d’Hondt system, under the constraint of the linguistic quota. That means that a list can have its next candidate elected as

Electoral Engineering

127

long as he or she belongs to a language group for which the quota has not yet been reached. If this quota has been reached, the seat is allotted to the next candidate on the same list from the other language group. If the list is unilingual, the seat is allotted to the next list that can claim the seat and has candidates from that language group (see Table 4.2). The use of quotas might at first sight seem at odds with the spirit of the proposal; it is not. The aim is to offer electoral incentives for politicians to campaign in both language groups. In the absence of quotas, there is a risk—indeed a certainty in the foreseeable future—that many voters will be reluctant to support a politician from the other language group for fear of contributing to a reduction in the representation of their own group in Parliament. In the absence of quotas, the federal election would quickly degenerate into a race between the language communities. That is exactly what the federal district must not be. In the version of the federal district proposed by the Pavia Group, catching a vote from the other language group will not alter the numerical parliamentary representation of the language group to which a candidate belongs. It will not decrease but rather increase considerably the incentive for parties and candidates to court the voters across the linguistic border. If parties and voters’ strategies are no longer frozen by fear of disproportionality, there is far more to gain from making one’s promises and actions more palatable to others. The quotas make it possible to leave intact the existing power-sharing devices. All members of the federal parliament will still belong to one language group. This is needed to protect the Francophone minority and for the double majorities required for some institutional reforms. The intent of a countrywide electoral district is to strengthen the democratic legitimacy and the problem-solving capacity of the federal governmental level without destroying the existing power-sharing principles and devices. Its introduction would not ignore or attempt to erase the differences between the language groups. Nor is it intended to resurrect countrywide political parties. It is precisely because there are no such parties that other devices are needed to link the federal politicians to the population of the federation as a whole. Parties belonging to the same ideological family might decide to form common lists for the fifteen federal seats. This would make them look better, as they could present a full list, and would guarantee that they would never lose a seat in case one of the quotas is fi lled. Moreover, their leaders would be given a better chance of winning more votes across the linguistic frontier, as each voter can select several names on the same list. Parties belonging to the same ideological family could also present separate lists, while

5

Quota is full; seat cannot be fi lled

Fift h candidate

Fourth candidate

Third candidate

Second candidate

First candidate

6

Sixth candidate

Fift h candidate

Fourth candidate

Third candidate

Second candidate

First candidate

List B

Third candidate; must be Flemish Fourth candidate; must be Flemish 4

Second candidate; must be Flemish

First candidate (assume Francophone)

List C

9

7 8

6

5

4

3

2

1

Flemish

6

5

4

3

2

1

Francophone

Quota

Note: Imagine three lists are participating in the election. List A is a list with 6 candidates, all French speaking. List B is a list with 9 candidates, all Dutch speaking. List C has 15 candidates, 6 of whom are French speaking and 9 of whom Dutch speaking. The proportional distribution of seats between the lists—using the d’Hondt divisors—gives List A 6 seats (numbers 1, 3, 5, 8, 10, 13), List B 6 seats (numbers 2, 4, 6, 9, 12, 14), and List C 3 seats (7, 11, and 15). On each list the candidates are ranked according to their preference votes. That defi nes the order in which they can be elected. Seat number 16 also has to be allocated, since seat number 13 could not be fi lled by List A. List A thus loses one seat because it is unilingual. An extra seat goes to the bilingual List C.

TOTAL

Seat 16

Seat 14 Seat 15

Seat 12 Seat 13

Seat 8 Seat 9 Seat 10 Seat 11

Seat 1 Seat 2 Seat 3 Seat 4 Seat 5 Seat 6 Seat 7

List A

Table 4.2. Simulation of Seat Distribution for a Federal Electoral District

Electoral Engineering

129

deciding to pool their votes, as allowed in the Pavia Group’s formula. But in any event, they would still present unilingual party lists in both Flanders and Wallonia. Indeed, it cannot even be ruled out that as regionalization deepens, separate Brussels parties may arise within each political family. The proposal of a federal electoral district is fully consistent with such developments. Its purpose is to provide an electoral setup that facilitates the government of a divided society in the absence of countrywide political parties.

Conclusion By way of conclusion we offer two remarks, one strategic and one philosophical. It is seldom easy to get an electoral reform through, if only because those currently empowered to change the rules are in power thanks to the rules they are asked to change. The reform proposal described in this chapter is no exception. Its adoption requires small changes in two articles of Belgium’s federal constitution and hence a two-thirds majority in both the House and the Senate. Is it possible to convince two-thirds of Belgium’s top politicians that a change of this sort is in their personal interest? We doubt it. Is it nevertheless possible to convince enough opinion leaders that this is a remedy the Belgian system urgently requires, to convince enough political leaders that there is something in it for them, if not for the sake of gaining power, at least for the sake of exercising it, and to put enough moral pressure on the rest so that the required supermajority can be patched together despite the opposition of secessionist parties? The future will tell. If there is hope, it comes from linguistically well-balanced pressure from civil society. It was crucial for the proposal’s prospects to be associated with a bilingual set of academics rather than with a linguistically tainted political party. And it is crucial that it continues to be supported by journalists and other leaders from both sides of the linguistic frontier. The 2007 and especially the 2010 processes of government formation were extraordinarily laborious and if enough people are able to see in this sad sequence of events not the failings or bad luck of individual people but a major defect of the system in which they are caught, progress may not out of reach. Finally, let us briefly turn from political strategy to political philosophy. Among the many critiques expressed against the Pavia Group proposal in the course of the rich debate it triggered, the most profound is perhaps the

130

Kris Deschouwer and Philippe Van Parijs

one best articulated by Bart De Wever, president of the New Flemish Alliance (N-VA), a Flemish nationalist party that became the largest party of Flanders and of the country after the 2010 elections. Proposing a federal electoral district, on this view, is a form of “creationism.” The Belgian state failed to create a Francophone Belgian nation in the nineteenth century. It gave up the idea of creating a bilingual Belgian nation in the twentieth century. The federal district is too weak an instrument, and it comes too late, to create a Belgian nation. All it can do, if anything, is hinder the process through which the Flemish nation and, if such a thing exists, the Walloon nation can become full-fledged states. Only with the consolidation of two states matching these two nations will the never-ending process of transformation of Belgium’s institutions come to an end. The political philosophy that underlies the Pavia proposal is different. No one could deny that being able to function in one language makes life easier for a democratic polity. For this reason, devolution to linguistically more homogeneous entities was a wise decision. The survival of Belgium is no aim in itself, and if all matters could sensibly be devolved in this way, why not? But they cannot, essentially because any sensible management of Brussels and its hinterland requires them to be under a single authority and because neither an absorption of the Brussels region by either of the other two nor an absorption by the Brussels region of its hinterland (namely the richest provinces of both Flanders and Wallonia) belong to the realm of the possible. Instead of wasting one’s time dreaming about nation-states that will never and should never exist, one must design and implement institutions that improve the working of polities that are not and will never become nation-states, including for the sake of moving more smoothly, as Flemish nationalists wish, toward more thoroughgoing devolution. Belgium is one such polity, and the European Union is another. Such institutional engineering is not a losing battle against the democratic imperative of linguistic homogeneity. It is an essential part of the piecemeal shaping of the sort of institutions that the countries and super-countries of today’s world will increasingly need. References Deschouwer, K. 2005. “The Unintended Consequences of Consociational Federalism: The Case of Belgium.” In I. O’Flynn and D. Russell, eds., Power Sharing: New Challenges for Divided Societies, 92–106. London: Pluto Press. ———. 2006. “And the Peace Goes on? Consociational Democracy and Belgian Politics in the 21st Century.” West European Politics 29, 5: 895–911.

Electoral Engineering

131

———. 2012. The Politics of Belgium. Governing a Divided Society. London: MacMillan Palgrave Deschouwer, K., and P. Van Parijs. 2007. “Une circonscription fédérale pour tous les belges.” La revue nouvelle, no 4, April, pp. 12–23 ———. 2009. Electoral Engineering for a Stalled Federation (With Comments by Laurent de Briey, Donald Horowitz, Bart Maddens, and Brendan O’Leary and a Reply by the Authors.) Brussels: Re-Bel e-book 4, June. www.rethinkingbelgium.eu. Geyl, P. 1962. “The National State and the Writers of Netherlands History.” In Debates with Historians. Glasgow: Collins/Fontana. Horowitz, D. L. 1985. Ethnic Groups in Conflict. Berkeley: University of California Press. Jans, M. T. 2001. “Leveled Domestic Politics: Comparing Institutional Reform and Ethnonational Conflicts in Canada and Belgium (1960–1989).” Res Publica 43, 1: 37–58. Lijphart, A. 1977. Democracy in Plural Societies: A Comparative Exploration. New Haven, CT: Yale University Press. ———. 1981. Conflict and Coexistence in Belgium: The Dynamics of a Culturally Divided Society. Berkeley, CA: Institute of International Studies. Murphy, A. B. 1988. The Regional Dynamics of Language Differentiation in Belgium: A Study in Cultural-Political Geography. Chicago: University of Chicago Press. Pavia Group. 2007.Een federale kieskring voor het federaal Parlement? Une circonscription fédérale pour le Parlement federal? http://www.paviagroup.be/. Swenden, W., and M. T. Jans. 2006. “Will It Stay or Will It Go? Federalism and the Sustainability of Belgium.” West European Politics 29, 5: 877–94. Van Parijs, P. 2000a. “Power-Sharing Versus Border-Crossing in Ethnically Divided Societies: Comment on Horowitz” (followed by a reply by D. Horowitz).” In S. Macedo and I. Shapiro, ed., Designing Democratic Institutions, 296–320. New York: New York University Press; reprinted as chapter 6 in P. Van Parijs, Just Democracy. The Rawls-Machiavelli Programme, 79–98. Colchester: ECPR Press, 2011. ———. 2000b. “Must Europe Be Belgian? On Democratic Citizenship in Multilingual Polities.” In C. McKinnon and I. Hampsher-Monk, eds., The Demands of Citizenship, 235–53. London: Continuum, 2000; reprinted as chapter 7 in P. Van Parijs, Just Democracy. The Rawls-Machiavelli Programme, 99–116. Colchester: ECPR Press, 2011.

PA R T II Historical and Conceptual Forays into Power Sharing

CHAPTER 5

A Theory of Accommodation Versus Confl ict: With Special Reference to the Israel-Palestine Confl ict Ronald Wintrobe

Can rational choice provide a model that explains sustained moderate or conflict-regulating power-sharing behavior in deeply divided places? This is the question addressed in this chapter. I start with a model of extremist behavior developed previously (Wintrobe 2006a, 2006b). The first third of the chapter outlines the basics of that model. The second third applies it to the problem of conflict versus accommodation. I then provide two illustrations: a brief discussion of the circumstances under which federalism can work and then a more detailed one of the Israel-Palestine conflict. Specifically, I discuss the erection of the wall between Israel and Palestine and the devolution of Gaza to the Palestinians. With the help of the model I ask whether each of these policies tends to exacerbate or reduce the likelihood of conflict. The last third of the chapter discusses more general solutions to the problem of conflict between groups. In all of this, my basic point of view is that there is a single leadership of each group and that leadership acts rationally. From an economist’s point of view, rationality just means that, whatever the goal, a person chooses the best means to achieve it. The goal itself is neither rational nor irrational; we just take it as given. The leadership of the group may sometimes be extreme: The simplest way to think of an extremist is someone whose goals or views are outside the mainstream on some issue or dimension. In the twentieth century, extremists were typically persons on the extreme right or the extreme

136

Ronald Wintrobe

left, but the dimension could also be nationalism, religion, security, or any other politically important issue. However, there is another way to think of extremism in politics, in which it refers to the use of extreme methods of political competition, usually violent ones, such as assassinations or terrorism. What explains the attraction of violence to people with extreme goals? I argue that leaders whose views are outside the mainstream adopt extremist methods when there is indivisibility between what might be called the immediate goal of the group and its ultimate goal. The behavior of followers cannot be explained in the same way as that of leaders. The reason is that the goals of the organization are a pure public good to a follower. Consequently a rational follower would tend to “free ride,” no matter how much he believes in the goal of the group. To explain the participation of followers, one must turn to something else. In this chapter, I suggest that they are motivated by the desire for “solidarity” (or social cohesion or “belongingness”) with a group. Not all participants in conflict-ridden societies are extremist, and these same assumptions—that leaders are motivated by the goals of the organization while followers are motivated by a desire for solidarity—are employed to discuss the behavior of non-extremist- actors as well. Of course leaders may be motivated by other things, such as the desire to stay in office to enjoy the power and prestige that come with it. As long as these desires are correlated with the achievement of the goals of the organization, this simple shorthand description of them is adequate to describe their reasoning. Similarly, some followers may indeed believe in the goals of the regime, but as long as solidarity is obtained more easily for those followers who share the goals of the regime, then again, this suffices to describe their motivation. They will prefer to join an organization whose goals are similar to theirs, even if their only motive is solidarity because in that way they obtain solidarity at the lowest “price.” To illustrate the effects of various events on the tendencies of each of the two groups to use violence, I focus on the construction of the wall (the West Bank barrier) and on the Israeli withdrawal from Gaza. I then enumerate briefly the solutions to group conflict suggested by the model and look in more detail at some of them. The outline of the chapter is as follows. The next section discusses leaders and shows the circumstances under which they are most likely to use violence. The third section briefly describes the behavior of followers. The

Accommodation Versus Conflict

137

fourth section integrates the behavior of leaders and followers in a single model. The fifth section then applies this model to the problem of accommodation and conflict between two groups and provides two illustrations of this model—federalism and the Israeli-Palestine question—and then enumerates and discusses the various solutions to conflict suggested by the model. The last section concludes the chapter.

Leaders Extremist Methods Are Risky

Our starting point is that extremist methods—essentially forms of violence such as assassinations, terrorism, and so forth—are simply a form of political competition. From the rational point of view, the central point about extremist methods compared to normal democratic methods of political competition is that they are risky. Consequently the choice between extremist methods and moderation can be analyzed in the same way as the choice between a criminal career and a legitimate one, as in the model of criminal behavior pioneered by Becker (1968). This point is explored in the model that follows.

A Basic Model of the Calculus of Discontent

I assume a political organization with some ideological goal Z, which might be a state for the group that lacks a homeland, a communist society, a law banning abortions, throwing all people of a certain race out of the country, or an Islamic society governed by sharia law. I do not inquire into the rationality of the belief in this goal but take it as given, as is normal in economic theory. The group tries to further this goal by exerting political pressure. So the product of either moderate pressure or terrorism is an increase in Z. Of particular importance is that this goal is often indivisible or displays increasing returns. This property is illustrated in Figure 5.1. I use the PalestineIsrael conflict though the figure can apply to any conflict over some resource. The horizontal axis indicates the level of an intermediate goal—land to the Palestinians or Jews—and the vertical axis the relationship between this intermediate goal and the final goal of the group (respectively, a Palestinian or

138

Ronald Wintrobe

Jewish state). This is the relationship that displays indivisibility or increasing returns. There is a critical point where enough of the intermediate goal has been obtained such that the final goal is possible. Two other illustrations of extremism are communism and contemporary Islamic extremism. In the case of communism, the intermediate goal would be control over the means of production and the final goal a communist society. In the case of Islamic extremism the intermediate goal might be the extent to which foreign forces are thrown out of the homeland and the final goal an Islamic society. Thus Figure 5.1 could also be used to show how communism or contemporary Islamic extremism displays this “indivisibility” property by simply relabeling the axes appropriately. In each case there is an indivisibility or area of increasing returns between the intermediate goal and the ultimate goal of the group. How does the existence of indivisibility explain why a group would choose violence to pursue its objectives? At the beginning of this section I

N

N

N1 C

G

N0 E

L0

M

Figure 5.1. Increasing returns in Israel-Palestine.

L1

Land

Accommodation Versus Conflict

139

argued that the basic difference between violent and moderate pressure from the point of view of the group is that violence is risky. I try to capture this feature in the choice among methods of pressure, for example, that between moderate and extremist methods. I represent this as follows. Assume the organization has a production function that can produce either moderate (M) pressure or extremist incidents (I) in any combination from fi xed levels of labor (L), capital (K), and organizational capacity (O). In the second part of the chapter we will specify a precise meaning for the latter concept. Of course in reality there is a continuum of methods: voting, peaceful and lawful demonstrations, civil disobedience, violence toward property, assassination of political enemies, violence toward innocent civilians, and war. For the purpose of modeling I assume only two methods, one moderate and the other violent and risky. Thus the level of moderate and extremist pressures is: (1)

M = M (LM,K M,OM), I = I (LI,KI,OI)

in which I = the number of violent incidents and M = the level of moderate pressure. The organization’s total stock of L, K, and O is fi xed:

(2)

L = L M + LI , K = KM + KI , O = OI + OM.

In general, an organization can use any combination of moderate and violent methods. The more it chooses extreme or violent methods, the greater the level of risk undertaken. I first illustrate the general argument with a simple example in which I assume the organization uses all of its resources either in the extremist method or in the moderate one. I then sketch a more general and detailed argument that drops this assumption. Figure 5.2 shows the goal of the group Z on the vertical axis. Z therefore represents variables such as nationhood, communism, or an Islamic society. The horizontal axis shows the product of applying various methods of pressure. Suppose that from the risky method there are three possible “states of the world”. The first is success (and the achievement of a high level of pressure I, in which case the level of the goal achieved is Z + g. The second and third

140

Ronald Wintrobe

Z Z

Z0 + g

0

I0

M

I1

Pressure, Terror

Figure 5.2. The relationship believed to exist between pressure or terror and the level of the ultimate objective (Z) that is achieved. (Summary diagram 1.)

involve failure. In the first of these, the attempt to impose pressure fails and the outcome is simply the status quo Z. In the second, the attempt also fails, and the leadership is caught and punished, retarding the goals of the group. If the value of the sanction as measured by its cost to the goal of the group is –f, then the outcome in that case is Z − f. On the other hand the outcome of applying a moderate level of pressure is always the level of pressure M, with gains for the group equal to Z + m. Let g = the gains to the group as estimated by its leader from using its organization and other factors of production to produce successful terrorist incidents I, and m = the (certain) gain to the group from using only moderate methods of pressure. One dimension of the level of increasing returns may be summarized by the ratio g/m. This is the ratio of the gains from successful terrorist pressure to moderate pressure. The higher this is, the more the function displays increasing returns.

Accommodation Versus Conflict

141

q = the probability that violent methods succeed and the state accedes to the demands of the group 1 − q = the probability that the methods fail p = the probability that, in addition to failure, the leadership of the extremist group is caught, convicted, and sanctioned f = the cost of the sanction to the goals of the group Z = status quo level of the goal of the group U = the utility function of the leadership Then extremist methods will be chosen if: (3)

qU (Z + g) + (1 − q) pU (Z − f ) + (1 − q) (1 − p) U (Z) > U (Z + m)

This equation (along with Figure 5.2) shows how violence can be a rational choice. A moderate level of pressure may leave the group stuck in the region of increasing returns, with the goal hardly advanced. With violent or risky methods, on the other hand, it is possible that the group can achieve its goal. Of course it is also possible that the group will fail, but note that the costs of failure may not be that large if there are increasing returns, and Z − f is not that far from Z. Thus, given that the goal displays increasing returns, violence may be a rational choice. Whether violence is rational depends on the structure of opportunities. The greater the indivisibility, the larger the ratio g/m and the more likely extremist methods will be chosen, as shown in equation (3). An increase in the likelihood that the methods succeed (q) will also raise the likelihood that these methods are chosen. Similarly, an increase in the capacity to manufacture terrorist incidents (I) would, on the other hand, raise the level of terror by raising the ratio g/m. The other main determinants are the deterrence variables p and f. Increases in these variables are effective in deterring extremism if they can be raised high enough. But note that increasing returns may limit the effectiveness of these variables. If increasing returns are large, as depicted in Figure 5.2, the enormous potential gains from terror and the small potential losses to the goals of the group explain the indifference of many extremist groups to loss of life—either their victims or members of the group who sacrifice themselves for the cause. Similarly, the figure shows that sanctions and other punitive measures against the group may not be effective. In addition, raising p

142

Ronald Wintrobe

sufficiently high to act as an effective deterrent may involve a conflict with civil liberties.

The Argument in More Detail

To see the argument in a more detailed and comprehensive way, begin with Figure 5.3. This figure shows the relationship between the instruments chosen (and therefore the level of risk) and the level of pressure exerted. At the origin on the horizontal axis, all of the factors of production are employed in moderate pressure, so the level of risk is zero. As we move along the horizontal axis, more and more of the factors are employed in the risky method (terror). The vertical axis shows the expected level of pressure that results.

Expected Pressure

Q

Risk (I/M)

Figure 5.3. The choice among methods of pressure is essentially a decision about risk.

Accommodation Versus Conflict

143

Each point depicts the maximum level of pressure that is possible to produce corresponding to that level of risk. It seems reasonable to suppose that taking at least some risk increases the expected level of pressure, so the curve depicted is initially upward sloping. At some point, too much risk can be taken from the point of view of expending pressure and the slope of the curve turns negative. The maximum level of expected pressure is the point Q, but the actual level decided upon by the leader of the orga nization will also depend on his or her attitude toward risk, as discussed further below. The next issue we need to address is the relationship between pressure and power. This depends on the structure of political institutions or the rules of the political game in the society where terrorist activity is undertaken. To take the most obvious case first, suppose that the country is a democracy and that the assumptions underlying the median voter model are satisfied. Then pressure succeeds only when the median voter is “persuaded” and fails otherwise. Once it succeeds, further pressure does not produce any more power. In that case, pressure produces zero power until the median voter is persuaded, it produces “absolute” power at that point, and beyond that point further pressure produces no further increase in power. Of course this depiction is extreme. One way to relax the assumptions but stick to the median voter model is to allow for some uncertainty as to the location of the median voter. Then the curve displaying the relationship between pressure and expected power will again display increasing returns until the expected position of the median is reached and diminishing returns thereafter. Again there will be a critical point, depicted as A in Figure 5.4, and this will be at the location of the median voter if the estimate of this position is unbiased. Other possible models of democracy do not necessarily display such stark levels of increasing returns—for example, if political parties maximize expected votes, as in probabilistic voting models. Similarly, pressure group models do not display this property. Another possibility is that the regime is a dictatorship. Here once again we would expect that the curve would display increasing returns. The point of “extremist” protest against a dictatorship is presumably to launch a revolution, and the point at which the state is weakened sufficiently for a revolution to take place is obviously a critical point. The increasing returns do not imply that any individual’s action can get the bandwagon rolling, however. Short of having sufficient support to effect this revolution, most attempts at

144

Ronald Wintrobe

A

P

Figure 5.4. How pressure translates into power. The possibilities are median voter, interest group equilibrium, and pressure versus dictatorship (e.g., tinpot or totalitarian). Under most possibilities there is again a critical point (A). But this is not the case with a pressure group equilibrium or with models where parties maximize votes, as in probabilistic voting models.

protest will simply bring problems for those who attack the regime. Indeed, in recent years, the literature on revolution is replete with such things as the possibility of bandwagon effects or the achievement of “critical mass” as depicted in Figure 5.4 (see Rasler 1996; Opp and Ruehl 1990). If Figure 5.4 does have the shape depicted, this only reinforces the degree of increasing returns to extremism and the basic argument made here. However, it is not necessary to our argument. Only if Figure 5.4 displayed diminishing returns throughout would the picture we are developing be possibly undermined. Figure 5.5 then displays the relationship between power and the immediate objective, represented by the variable C, as in control of land for the Palestinians or Jews, control over the means of production, or ridding the country of non-Islamic authorities, domestic or foreign. There seems no compelling reason to believe that this relationship is nonlinear, hence it is depicted as a straight line.

Accommodation Versus Conflict

145

C

C1

CM

C0

∏0

∏M

∏1

Power ∏

Figure 5.5. Power and the immediate objective. ΠM = the level of power produced by “moderate” methods such as civil disobedience; Π = Π or Π = the level of power produced by extremist methods. The figure assumes a linear relation between power and the immediate objective C: control over land, control over the means of production, or ridding the country of non-Islamic authorities.

One way to illustrate our basic point can be seen in Figure 5.6. Since we assume that the group leader can choose any combination of extremist methods and moderate methods, the level of risk that can be undertaken is completely variable. The horizontal axis displays this level of risk and the vertical axis the expected total returns to it, that is, the value to the group of the achievement of its fi nal goals at different levels of risk. The curve EZ in Figure 5.6 displays the risk—total return relationship for a group that is contemplating various methods of pressure from fixed resources. U is the  utility function of the group leader. Equilibrium is at the point E , σ , if the group decision maker is risk averse. The indivisibility implies that from the point of view of the group’s decision makers, very little is to  be expected from moderate methods of pressure, and even switching some resources into extremist methods does not advance the goals of the

146

Ronald Wintrobe

E(Z)

U

E0

EZ

σ0

Risk (I/M)

Figure 5.6. Another way to see the relationship between pressure (or risk) and the level of the ultimate objective achieved. The figure shows an equilibrium (E) where the indivisibility is present. (Summary diagram 2.)

organization very much. As pressure is ratcheted up, the gains from it increase at an increasing rate over a substantial range. Ultimately the rate of increase of these gains tapers off, and they continue to increase but at a decreasing rate. So only at high levels of pressure do the gains become sufficiently large that the objective can be said to be reached. Finally a point is reached when so much risk is taken that it actually becomes counter productive, that is, after that point returns are negatively correlated with risk. But the essential point is that the larger the range of increasing returns or the greater the indivisibility, the more likely the group is to choose extremist or terrorist methods compared to moderate measures of pressure. Of course mistakes are possible. Extremist methods might have been chosen by mistake. For example, the curve may be mis-estimated so that moderate methods of pressure such as civil disobedience would actually have been sufficient. In this case, the production function actually has its critical

Accommodation Versus Conflict

147

point at a fairly low level of Z. But equilibrium (because of the mis-estimation) is at a high level of risk or extremism. The capacity for mistakes implies that terrorists sometimes end up on the downward-sloping portion of the curve. They go too far. Perhaps the most outstanding recent example was the killing by the Red Brigades of Aldo Moro in Italy in the 1970s, which seemed to everyone, ex post, a mistake and after which support for the terrorists dried up and the era of terror ended (Ginsborg 1990). However, recall that their basic objective was achieved in that the communist party never did join the government.

Followers Why do people join extremist organizations or participate in extremist acts? The most obvious reason is that they believe in the goals of the organization, and they participate in its activities in order to bring them about, just as we argued for leaders. However, in the case of potential followers, there is a “free-rider” problem: since one’s own contribution toward the achievement of the goals of the organization is likely to be small, why not “free ride” and hope that others will make the necessary effort? In the case of extremist organizations, which, as we saw in the last section, usually have goals that are grand and distant and therefore unlikely to be achieved, this problem is particularly acute. Whatever the goal of the organization—a national homeland for Palestinians or Kurds, the removal of “foreign” domination by the Indians in Kashmir or the Russians in Chechnya, or the removal of U.S. troops from Saudi Arabia, or the IDF from Lebanon or Palestine—the individual’s own contribution to this goal cannot be significant no matter how large his personal sacrifice. So we would not expect people to join such organizations, or even to participate in their activities, simply because they believe in their goals. Research into the internal workings of extremist groups has suggested two things. First, they are characterized by a high level of social cohesion or solidarity. Thus, as Post suggests, “For many, belonging to the terrorist group may be the fi rst time they truly belonged” (1990, 31). Similarly, suicide martyrs do not commonly act alone but are usually members of groups who “demand” their ser vices (Hoff man 1998; Pape 2002; Ricolfi 2005).

148

Ronald Wintrobe

Second, members of such groups usually do hold, in common, a set of extreme beliefs. Islam as used by al Qaeda is not a purely religious doctrine but one that has been intensely distorted to serve the ends of the group (Black 2001; Gunaratna 2002; Ruthven 2000). Some other extremist groups have bizarre beliefs: for example, the Christian Identity movement in the United States apparently believes that the lost tribes of Israel are composed not of Jews but of “blue-eyed Aryans” and that Jesus Christ himself was an Aryan (Hoffman 1998, 112). Mark Koernke’s 1993 video America in Peril states that “elements within the US government are working with foreign leaders to turn the United States into a dictatorship under the leadership of the United Nations” (Karl 1995, 69). It seems, then, that two remarkable features in many extremist groups are the depth of solidarity and the extremity of their beliefs. I contend that neither of these two phenomena is necessarily irrational and indeed that the key to understanding both of them is that they are related to each other. More precisely, they are the outcome of a process whereby beliefs are traded in exchange for solidarity or social cohesion. The person who gives up his beliefs loses something, which could be called his or her true “identity” or “independence of thought” or “autonomy.” On the other hand, he or she gains the experience of greater solidarity or social cohesion or “belongingness.” Note that the free-rider problem does not apply to the receipt of solidarity, which is a private good. At the same time, this idea can also explain why a person joins organizations with beliefs and goals that are similar to his or hers and why people who have the same beliefs as extremist organizations tend to join or participate more than those who do not. The more the beliefs of the individual are in agreement with those of the organization to begin with, the smaller the sacrifice required in terms of the individual’s own autonomy necessary to receive a given level of solidarity. Consequently an individual joins an organization whose beliefs are close to his not because he thinks that his own efforts will make any palpable difference to the achievement of the goals of the organization but because that way he obtains the desired solidarity at the lowest “price.” To sketch a model of how this process operates, assume that an individual is endowed with a certain set of beliefs and, corresponding to this, a certain identity. If a person agrees to join a group, the price of admission is, in part, that he or she must adopt certain beliefs that are sanctioned by the group. Additional requirements might be that he participate in group ac-

Accommodation Versus Conflict

149

tivities or in some other way demonstrate that he shares in the beliefs and goals of the group. The organization, in turn, supplies the individual with the sense of belonging to a community by organizing events or activities individuals can attend and participate in and meet and get to know others in the organization. The main implication of this way of thinking is that a person who holds a belief that appears on the surface to be irrational may not be irrational: the rationality may consist not in the content of the belief but in the reason for holding it. On this reading, the person who believes there is a UN plot to take over the U.S. government is no more irrational (in principle, if not in degree) than the professor who states to the officials in the administration of his university that this department, more than any other in the faculty, deserves more resources: in both cases, the reason for the belief may be solidarity or social cohesion, not the coherence of the belief itself. It is simple to formalize the basic proposition of the model, that is, that social cohesion (solidarity) and conformity (unity of belief) are positively related. To do so, assume that individuals have utility functions in which both autonomy and solidarity are positive arguments as depicted in Figure 5.7. Individuals are willing to trade autonomy for solidarity, and the way they do this is by adopting the beliefs demanded by one or more suppliers of solidarity. These suppliers may include religious organizations (organized religions and cults), gangs, political parties and movements, unions and business firms, or other organizations. The “industrial organization” of solidarity is complex because solidarity tends to be produced in the process of working toward some goal or participating in some activity and thus is usually supplied together with that activity. The individual maximizes utility subject to a constraint in the form of a production function depicted as the production possibility curve between solidarity and autonomy ES in Figure 5.7. A typical individual will have an endowment point like e and will trade autonomy for solidarity by giving up his own beliefs in the manner discussed, ending up at an equilibrium like E. The rate at which he can trade off autonomy for solidarity depends on the technology available for doing this, as summarized in the production function. Thus churches have a “technology” for conversion involving rituals, dogmas, and ceremonies by which individuals are assisted in becoming believers. Other organizations may have twelve-step programs, identification rituals such as “jumping in” to a

150

Ronald Wintrobe

A

E

e0

E1 Ui E2

U S

S

Figure 5.7. The choice between solidarity and autonomy and the solidarity multiplier.

gang, and so on. The production possibility curve is depicted as having the usual shape, implying diminishing returns to the conversion of autonomy into solidarity and vice versa. This is one way of specifying the group’s “organizational capacity” described in the previous section. However, this analysis leaves out something important: once an individual i has made the choice of giving up some of his autonomy (A) in exchange for solidarity (S), he has given up some of his autonomy and therefore his independent capacity to choose. For small changes this might not matter, but for large ones it obviously does: to some extent he has given up the control of the choices he might make to the leader of the group I show elsewhere (Wintrobe 2006a, 2006b) that this gives rise to a solidarity “multiplier”: As the individual chooses more solidarity, in order to get it he adopts beliefs and values that are more akin to those of the leader. But with these new values and beliefs, he finds that he prefers more solidarity than he did originally. In order to acquire still more solidarity, again his beliefs and values must change in order to conform to those of the other

Accommodation Versus Conflict

151

members of the organization. In turn, with this new utility function, he wants more solidarity than previously, leading him to change his values again, which again results in yet a further demand for solidarity, and so on. An interior equilibrium will result if these effects occur at a sufficiently diminishing rate, as shown in Figure 5.7. One need not join an extremist group to observe the solidarity multiplier in action. I first noticed it (but did not understand it) years ago in the behavior of academic friends (not all economists) when they assumed important administrative positions such as department chair or dean of the faculty. Within a short time their values seemed to undergo a transformation: previously highly individualistic in many cases, their conversation was now laced with phrases like “the good of the department,” the importance of promoting “institutional values,” and so on. Their behavior seemed to change as well, as they now began to promote collaborative research projects and “loyalty” to the department. Of greater importance is the possibility of temporary, rapid increases in solidarity such as those noted by Ricolfi (2005) in his empirical work on Palestinian suicide martyrs. Ricolfi found that suicide martyrdom was often motivated by revenge or the desire to avenge tragic events such as the death of a relative at the hands of Israeli forces. Similarly, “revenge” is the classic motive in many studies of solidarity (e.g., Gold 2000). At the mass level, in Palestine, as in previous resistance movements, the funeral of some important or tragic figure often becomes the occasion for stimulating revenge. The solidarity may be temporary, but that is enough to provide a mechanism for stimulating action. The analysis in Figure 5.7 is incomplete. It is easy to imagine that the self-reinforcing process just analyzed leads to a corner rather than an interior equilibrium. At a corner, individual i rationally chooses an equilibrium with all solidarity, zero autonomy. His utility function is simply the utility function of the leader. The individual has no independent thought but is completely under the leader’s control. His values are completely those of his leader, and he will do whatever maximizes his leader’s utility. If the leader wishes him to commit suicide for the goals of the group, he will do so. Note that he might do so even if he is not at a corner but close to it; the views of his leader or the values of the group contain great weight in his utility function. What is peculiar about the corner is not that only there is rational suicide possible but that at a corner, the individual will be especially resistant to

152

Ronald Wintrobe

change. In particular he will be resistant to pressure from outside sources such as threats or increases in the likelihood of prosecution or the size of the punishment for being a member of such a group. Further, even an individual who is at the corner may be extreme, but it is vital to note that he is not irrational. He possesses a well-behaved ordinal utility function and is perfectly capable of making choices that maximize his utility in the usual sense. Indeed, his behavior is merely an extreme version of a form of behavior that is extremely common, namely that, in part, he “internalizes” his values from the values of others, especially from those in a position of power over him. To obtain solidarity with the group of which he is a member, he adopts the group’s values and beliefs. This is precisely what members of religious groups do when they agree to or “internalize” the values and beliefs of their religion, or what members of ethnic groups do when they subscribe to the belief that they “belong” together in some sense because they have as ancestors people who held similar beliefs, or what economists do when they write papers based on a certain set of assumptions that they share about human nature (e.g., that people are always rational). The only difference in the behavior of the individual who is in equilibrium at a corner is the extent to which he behaves in this fashion. The behavior itself is perfectly “normal” and rational. And all of us are familiar with the internal struggle between doing what is right for the group and doing what is best for one’s “self” felt by individuals who are not at a corner but near it.

An Example of the Complete Model: What Happens to the Level of Extremism If the Technology for Producing Solidarity Increases? To see how the complete model of leaders and followers works, suppose for the purpose of illustration that there is an upward shift in the technology for producing solidarity. This leads the typical individual to take more solidarity than before. Consequently, individual followers are more willing to join extremist groups in order to get solidarity, and those individuals who are already members of such groups are more willing to be solidary with them. Figure 5.8 shows the effect of this choice on the group’s capacity to produce pressure. Because there is greater solidarity within the group, the organization can now produce more pressure for any given level of resources

Accommodation Versus Conflict

153

Pressure

Solidarity

Risk (I/M)

Figure 5.8. Increased solidarity within the group raises the likelihood of violence.

K and L. So at any given level of risk the amount of expected pressure increases. Because solidarity is more important for extremist methods of pressure than for moderate methods, the capacity of the group to produce pressure increases as the level of risk rises, also as shown in the figure. Figure 5.9 shows the effect of this change on the group leader’s choice between moderate and extreme methods of pressure. Th is can be represented as the choice between a riskless asset (moderate methods of pressure) and a risky one (extreme methods). The level of extremism here is represented by the proportion of the “portfolio” in risky assets (I/M, where I is the number of extremist incidents produced and M is the level of moderate pressure). Since expected pressure increases for any given level of risk, so does the capacity to achieve the goal of the orga nization (Z). In the figure, the relationship between risk and return (EZ) therefore shifts up (from EZ  to EZ). The increase in return for any given level of risk implies a substitution effect that favors risk taking and an income or “wealth effect” due to the increase in the orga nization’s capacity to produce pressure. The wealth effect favors risk taking (larger I/M) if the group leader’s

154

Ronald Wintrobe

E(Z)

UL1 E1 Z1

EZ1

UL0 E0

EZ0

Z0

Figure 5.9. Because of the improvement in organizational technology, the leader now faces a better trade-off between risk and return. Provided the substitution effect dominates the income effect or relative risk aversion is either constant or decreasing as wealth (organizational capacity) increases, the leader chooses more extreme methods.

coefficient of relative risk aversion is decreasing as wealth increases. It is neutral if this coefficient is constant and negative (lower I/M) if this coefficient increases with wealth. Thus, whether the degree of extremism increases or not depends crucially on the extremist leader’s attitude toward risk. If the substitution effect dominates the wealth effect, or if the wealth effect is positive or neutral, he or she chooses more extremism as a result of this change. We will see in the rest of this chapter that the direction of these wealth effects is crucial for understanding the problem of confl ict versus accommodation.

Accommodation Versus Conflict

155

Application to Power Sharing Among Ethnic Groups Now we apply this complete model to the problem of conflict. Figure 5.10 illustrates the general picture. We now shift the focus of the discussion from that of a society that is potentially threatened by extremists to an arena in which two groups (e.g., Palestinians and Israelis, Northern Irish Protestants and Catholics, Serbs and Croatians, French and English Canadians) both aspire to “nationhood.” The first issue is, how do they each frame the situation? For simplicity we focus on a fi xed territory over which both groups wish to be sovereign.

Framing

Whether there is conflict or accommodation between the groups depends on how the situation is framed. Figure 5.10 illustrates two possibilities. In Figure 5.10a the two curves intersect in the region of increasing returns to both groups. In Figure 5.10b they intersect in the region of diminishing returns. In the first situation, accommodation is difficult, in the second it is easy. Hamas and Jewish fundamentalists frame the situation in Israel-Palestine

Z Zj

Zi

(a)

Land to Palestinians

Accommodation Difficult

Figure 5.10. Land transfer and accommodation.

Zi

(b)

Zj

Land to Palestinians

Accommodation Easy

156

Ronald Wintrobe

this first way (conflict) but so does Amos Oz, a moderate, who argues simply that the basic problem in Israel-Palestine is simply the scarcity of land (2006). However, it is worth remembering that most of the time ethnic groups live in peace (Fearon and Laitin 1996), and in other countries and other situations Palestinians and Israelis live together harmoniously, as have Serbs and Croatians, Hutu and Tutsis, Protestants and Catholics, and French and English Canadians. In Jerusalem the three monotheisms are stuck together as it is the locus of three central religious “myths”: 1. the destruction of the Temple; 2. the ascent of Muhammad; and 3. the stations of the cross. Each of these myths is central, and monotheisms are mutually exclusive: if there is only one God, it has to be either mine or yours. Indeed the same problem can arise within denominations: in the Holy Land, the different branches of Christianity find themselves in conflict over the “ownership” of the Church of the Holy Sepulchre. The church (the fourteenth station of the cross) is dirty and in constant need of repair. The reason is that cleaning or repair implies ownership, and the different Christian denominations cannot agree on who “owns” the church. It was almost destroyed in a fire in 1840 because whoever calls in the fire department in a sense “owns” it so no one was allowed to do so. But “ownership” of the church or any of these major holy sites is obviously not a solution that will please any more than one of them. What determines how the situation is framed? This is obviously a very difficult question, and there is not much in the literature on it that is conclusive (although there is much that is interesting; e.g., see Ariely 2008 for some interesting experiments). But one thing that seems uncontroversial is that things that have happened in the past shape viewpoints. To illustrate with an example from Iraq, in April 1980 officials from Saddam’s Ba’ath regime arrested Moqtada al Sadr’s father-in-law and his sister. They raped and killed his sister and then set fi re to the ayatollah’s beard before driving nails into his head (Galbraith 2006). There were many such atrocities by the Ba‘ath regime, including the gassing of the Kurds. Events like these can attain the status of “myths.” Karen Armstrong defines a “myth” as “something that happened once, in some sense, and which also happens all the time” (2005, p. 10). Thus Saddam’s cruelty, which prob-

Accommodation Versus Conflict

157

ably became openly apparent in public policy only occasionally, became “mythical” in this sense, part of the way his image and that of his regime became “framed.” Nevertheless, Saddam was an ally of the United States until 1991, when his invasion of Kuwait caused the United States to back away from him. The first Bush administration publicly tried to build support for military action against Iraq by labeling Saddam a “modern-day Hitler.” Even more obviously, the Holocaust for the Jews and the naqba for the Palestinians have the status of “mythical” events: they happened once (the sense in which they happened is obvious in these cases but its interpretation is not), and, it is feared, they may happen all the time. Thus Benjamin Netanyahu, on his first visit to the Obama White House in May 2009, took that opportunity to “take the new president aback with an apocalyptic lecture on the 4,000 years of Jewish history which, he said, a nuclear-armed Iran would seek to end.” But “that evening while relaxing with an intimate group. . . . Mr Netanyahu gave vent to his discomfort. For all his efforts to set the scene in a Jewish historical perspective, he felt that the president focused more on the plight of the Palestinians. ‘What moves Mr Obama’ he wondered edgily aloud” (Economist, May 23, 2009, 49–50). In other words, he wondered how it would be possible to get Obama to reframe the Israeli-Palestinian confl ict in a manner more sympathetic to the Israeli cause as he sees it. Some argue (e.g., Kaufman 2006) that extreme ethnic violence (e.g., Rwandan genocide) can only be explained by group myths that generate hostility and fear, not by rational choice. But myths are not necessarily irrational. Like paradigms in science, they “frame” the problem. Jews and Palestinians fear each other, as do Serbs and Croats. Kurds and Shiites fear Sunnis in Iraq. These fears are based on events that happened in the past but, they fear, may happen all the time. These fears shape and position the curves and make accommodation difficult.

Equilibrium, Accommodation, and Conflict

What determines how much violence or accommodation actually occurs? We can use our model to answer this question. To do so, assume that each party decides on the optimal equilibrium level of moderate and extreme/violent forms of pressure. If there are only two parties, they might react to each other, and thus strategic considerations

158

Ronald Wintrobe

might arise that change this equilibrium. We will assume here that the equilibrium is Nash—that is, each side considers the other’s position as fi xed in calculating its own strategy. To illustrate, look at the Israel-Palestine conflict. For simplicity, let us initially ignore the division within Palestine between Hamas and Fatah and suppose a single Palestinian “leadership” that decides on an optimal allocation between moderate and extreme methods of putting pressure on Israel. Extreme methods refer to violent methods and include such things as firing rockets from Gaza or elsewhere, terrorist attacks within Israel, and so forth. We assume these can be aggregated into an index of the level of Palestinian violence against Israel. Figure 5.11 depicts Palestine’s optimal position, given the position of Israel. Thus, on the horizontal axis, the level of Palestinian violence (I/M)p increases as we move from left to right. (I/M)p is then the initial equilibrium Palestinian allocation between extreme and violent methods. In a similar

E1 i E(Zi)

E(Zp) E

0 Ei

0

p

E1p

0p

(I/M)1i (I/M)0i

(I/M)1p (I/M)0p

Figure 5.11. The construction of the wall. Palestinian violence decreases, Israeli violence increases. The net effect on the likelihood of accommodation is unpredictable.

0i

Accommodation Versus Conflict

159

manner, to begin with we assume away internal conflicts within Israel, such as that between secular Jews and the ultraorthodox (Haredim) and assume there is a “leadership” (the Israeli government) with a single position that decides on the optimal allocation between extreme and moderate methods of putting pressure on the Palestinians. Thus extreme methods include missile strikes, assassinations of Palestinian leaders, outright invasion of Palestinian territories, and so forth. Moderate methods would include diplomatic overtures, negotiations, aid to Palestine, and so on. Once again we assume a single index of the level of violence on the part of Israel (I/M)i. Israel’s allocation (I/M)i should be read in the opposite manner from Palestine, that is from right to left. Thus a movement from right to left implies a greater reliance on extreme methods for Israel. The initial equilibrium level of Israeli violence is (I/M)i. Figure 5.11 cannot be read in the same manner as Figure 5.10. That is, no significance can be attached to how close or far away the equilibria of the two parties are in I/M space. Nor does it matter whether the curves cross in the “increasing returns” range or not, since it is (I/M) (of both parties) on the horizontal axis, not land. One could say that the optimal positions of each party depict how “violent” that party is. For example, if we use an index of each party’s violence, which goes from 0 (that is, zero violence) to 1 (I/M = 1), and the equilibrium (I/M)p for Palestine is 0.4 and the equilibrium for Israel (I/M)i is 0.6, this could be interpreted to mean that 60 percent of Israel’s resources used in producing pressure on the Palestinians is devoted to violent methods and 40 percent for Palestine. However, since the parties in this case as in many other situations of conflict are so differently situated in terms of their control over resources and in terms of what constitutes a “violent” act, it is not clear that comparisons like this are all that meaningful. However, indices of violence have often been constructed, and it is possible they could be used in this context. But that is not our main purpose here. What is of interest and could be calculated is the effect of various events on the level of violence used by each party. Suppose for example that there is a change that causes the Palestinians to use less violence, that is, it reduces (I/M)p to some lower level (I/M)p. Then we could obviously say that the change makes Palestine less violent. Here I will go further and assume that this also implies Palestine is more interested in accommodation rather than conflict. Similarly, a change that, say, increases the Israeli equilibrium allocation between moderate and violent methods from (I/M)i to (I/M)i means that Israel becomes more violent. I will also assume that means Israel is less

160

Ronald Wintrobe

accommodative. In other words, I assume that the degree of accommodation is negatively correlated with I/M: the higher the level of violence used by one of the parties, the less likely that party is interested in peaceful accommodation. Given these assumptions, it seems also reasonable to conclude that if some change makes both parties less violent, accommodation between the two of them is more likely and the probability of conflict reduced. And vice versa. Let us now illustrate how this works by considering two of the most important changes in the recent history of the Middle East: the construction of the wall between Israel and the West Bank, and the Israeli withdrawal from Gaza.

The Wall and the Israeli Withdrawal from Gaza

First take the construction of the wall, which began in 2004. What effect does our model suggest this would have had on the propensities for violence on the part of Israel and Palestine? To answer questions like this, look back at the various elements that make up the model, as illustrated in Figures 5.1 and 5.3-5.6, and ask how one or more of them are affected by the change. It seems reasonable to suppose that the wall provides some protection for Israel against Palestinian attacks: the checkpoints there make any Palestinian terrorist attack on Israel more difficult to mount successfully, and the existence of the wall itself is obviously a barrier to many (not all) forms of violent attack. According to the article on the wall in Wikipedia (which calls the wall the Israeli West Bank barrier), “Members of al-Aqsa Martyrs’ Brigades, Hamas, and the Palestinian Islamic Jihad have been less able to conduct attacks in Israel, the numbers of which have decreased in areas where the barrier has been completed. In his November 2006 interview with Al-Manar TV, Palestinian Islamic Jihad leader Ramadan Salah claimed that the barrier is an important obstacle and that ‘if it weren’t there, the situation would be entirely different.’ In a March 23, 2008, interview, Palestinian Islamic Jihad leader Ramadan Abdallah Shalah complained to the Qatari newspaper Al-Sharq that his organization had been forced to switch from martyrdom missions to rocket attacks because the separation barrier ‘limits the ability of the resistance to arrive deep within [Israeli territory] to carry out suicide bombing attacks, but the resistance has not surrendered or become

Accommodation Versus Conflict

161

helpless, and is looking for other ways to cope with the requirements of every stage of the intifada.’ It follows that for Palestinians the return to pressure has fallen at any given level of risk, and this is especially so for violent methods of pressure. More precisely, using Figure 5.3 to depict the choice of Palestine first, it follows that at any given level of risk less pressure will be produced (not shown in Figure 5.3), and this effect will be larger at higher levels of risk. Tracing this effect through Figures 5.3-5.6 to its impact on the expected return to pressure we arrive at Figure 5.11. In that figure, the curve depicting the riskreturn relationship for Palestine shifts down from P to P. There are two effects: first, because only the return to violent methods has fallen (the yield on moderate methods is unchanged), there is a substitution effect against the use of violence. This implies a decrease in the use of violent methods of pressure. Second, Palestinians are in a sense poorer since some (violent) methods of pressure are less likely to be successful, while other (moderate) means no more likely to be so, after the creation of the wall. The wealth effect—the fact that the value of the entire “portfolio” of pressure options has fallen—also implies that less violence will be used if relative risk aversion is constant. Under these assumptions the new equilibrium allocation will be at a point like (I/M)p in Figure 5.11: Palestinians will want to use less violence after the creation of the wall. We can also derive the effects on the Israeli equilibrium allocation. Suppose the wall is also, as is commonly said, a “land grab” on the part of Israel. Looking at a diagram like Figure 5.5 for Israel depicting the relationship between power and the immediate objective, which in the case of Israel-Palestine is land, Israel would now have more land at any level of power than before. Thus the curve like the one depicted in Figure 5.5 for Israel would shift up (not shown). Again, tracing through the effect of this change on the final relationship between risk and expected return for Israel, we would anticipate that curve to shift up and to the left as depicted by the shift from I to I in Figure 5.11. The new equilibrium is at Ei , implying a greater reliance on violence compared to the original equilibrium at Ei. Both the substitution and the wealth effect favor increased risk taking on the part of Israel. Thus, Israel can be expected to use violence relatively more after the construction of the wall. Since one party (Israel) becomes less accommodative as a result of the change while the other party (Palestine) becomes more so, we cannot predict the effect on accommodation; that is, under these assumptions one cannot say that in general the creation of the wall promotes peace.

162

Ronald Wintrobe

What are the effects of Israel’s decision under Ariel Sharon to withdraw from the Gaza Strip? The results are depicted in Figure 5.12. The Israeli withdrawal would seem to make Israelis more vulnerable to rocket and other attacks, thus raising the return to violent methods of pressure by the Palestinians (see Figure 5.3; the curve in the figure would shift up [not shown]). Palestinians are now wealthier in the sense that their methods of pressure are more effective—at least the return to violent measures has increased. Again, assuming constant relative risk aversion, Palestinian violence tends to increase, as depicted in Figure 5.12. Israel has now given up land (in the sense that it has given up occupying Gaza), so this is the exact reverse of the construction of the wall as far as its effect on the level of Israeli violence is concerned. In a diagram like Figure 5.5, Israel has less land at any level of power than before. Consequently in Figure 5.12, the relationship between risk (I/M)i and return E(Z)i for Israel shifts down, and the new equilibrium is at a point like Ei instead of the original Ei. With constant relative risk aver-

E0i E(Zi)

E(ZP) E1p

1

E

i

E0p

0P

(I/M)0i (I/M)1i

(I/M)0p (I/M)1p

Figure 5.12. The Israeli withdrawal from Gaza. Palestinian violence increases, Israeli violence decreases.

0i

Accommodation Versus Conflict

163

sion, the level of Israeli violence will fall. So Palestinian violence increases, and Israeli violence decreases. Before we accept these conclusions about Gaza and the wall, note that there is one possibility that would change these conclusions in an important way. To see this point, consider first the case of the construction of the wall. Suppose the wall makes Israelis and the Israeli leadership genuinely secure. That is, as a result of the construction of the wall they feel themselves to be a “nation” with secure borders. (However just or unjust the path of the wall is a matter of heated, not to say incendiary, debate, and we are not passing judgment on that here; we are simply making an assumption about the way Israelis and/or their leaders might feel. Note that there is a great deal of evidence that Israelis do in fact feel more secure as a result of the building of the wall.) In that case, after the construction of the wall Israelis might find themselves on a relatively flat portion of the curve depicting the relationship between land and nationhood—that is, they have reached the point where, while more land would be nice from their point of view, they do not “need” it to be a nation. So the constraint is relatively flatter, and the higher indifference curve is tangent to the constraint at a flatter point. Thus in the new equilibrium, Israelis would take less rather than more risk, as depicted in Figure 5.13, that is, they would use less violence. In that case both Israelis and Palestinians would become more accommodative as a result of the construction of the wall. However justified or unjustified its construction or its actual route, we are simply suggesting that since it reduces the return to violent methods of pressure on the part of Palestine and if it also makes Israelis feel that the borders of their state are secure, the construction of the wall promotes peace. Could a similar argument change our conclusions about the Israeli withdrawal from Gaza? In this case it is Palestinians who became more violent as a result of their increase in “wealth,” while the Israelis became less so. Now one could argue that if Gaza met Palestinian objectives with respect to “nationhood,” that is, if Gaza alone were a sufficient amount of land to satisfy Palestinian aspirations to nationhood, then it could be argued that Palestinian violence would diminish, not increase. But Gaza is not a sufficient land base from which to form the Palestinian state. Thus there does not seem to be any reason to alter our conclusion about the effects of the Israeli withdrawal from Gaza. To sum up, we can conclude that if the construction of the wall along borders makes Israel “secure” then both Israel and Palestine would become

164

Ronald Wintrobe

E1i

E(Zp)

E(Zi) E0p

E

0 i

E1p

0p

(I/M)0i (I/M)1i

(I/M)1p (I/M)0p

0i

Figure 5.13. The construction of the wall (2). If the wall makes Israel sufficiently secure, Israeli violence decreases. Palestinian violence also decreases, and accommodation is more likely.

less violent. However unjust this might be from the point of view of Palestinians, it is easy to see the attraction of the policy to Israeli leaders. The Israeli withdrawal from Gaza induces Israel to be less violent and Palestinians to be more so. Here the attraction of the policy to the Israeli leadership is less obvious from the model, though that might lie in factors outside the model such as the difficulties encountered with the alternative policy of occupation of Gaza, something that is not modeled here. Let me emphasize that these conclusions are not intended to be nor can they be interpreted as endorsements or criticism of any of these policies. They are exercises in positive economics, statements about what the predictive implications of the model are with respect to these events under what appear to be the most reasonable set of assumptions. The next section turns to normative economics. From the normative point of view, it seems obvious that the best policy would be one that makes both parties satisfied with their “nationhood.” That might appear impossible with the simple policies that have been discussed so far and from the results

Accommodation Versus Conflict

165

shown in Figure 5.10. But it must be remembered that the curves as depicted there correspond to a set of beliefs about how much land Israel or Palestine “needs” to be a nation, not a set of facts. Some ways to ameliorate conflict are suggested in the next section. The argument there is conducted at a general level, as the details of the Israeli-Palestinian conflict are too specific for this purpose. Some reference is made to successful cases of accommodation (Canada, Northern Ireland, South Africa), though in the end I return briefly to the Israel-Palestine conflict.

Some Ways to Increase Accommodation

First it might be helpful to summarize all of the ways in which our model suggests that the possibility of accommodation between two parties might be increased: 1. “Reframe.” The effect of reframing is shown in Figure 5.14. Notice that accommodation is easier when the two curves spread apart rather than come together, a visual oddity. (See the discussion above on framing.) 2. Reduce solidarity within groups (e.g., by providing trusted external information). This is part of a more general strategy of raising the return to moderation. The effects of this are shown in Figure 5.15. 3. Use rewards and sanctions. 4. Make the indivisible divisible. 5. Try to create trust between the groups, or if that is too difficult, try to find an external guarantor of peace or, more generally, go to #6. 6. Economize on trust with “high-quality” institutions. 7. Get the opposing community on your side. For example, in Iraq the situation improved when the United States switched from a strategy of confrontation with the Sunni population to one of incorporating them in the fight against al Qaeda. 8. In the model depicted by Figures 5.3-5.6, attempt to break the link between any of the following: A) risk and expected pressure (overpay leaders, use sanctions to deter violence) B) pressure and power (e.g., consociation)

166

Ronald Wintrobe

C) power and the immediate objective D) the immediate and final objectives (e.g., federalism) 9. Marginalize or accommodate the extremists inside each community. We cannot discuss all of these here, but many of them are referred to in the specific examples below. On solution 3. Use rewards and sanctions, for example, by rewarding moderation or punishing extreme behavior. However, once we have a model, which explains extremist behavior, such as the one used here does, we can see why deterrence alone does not work: the goals of leaders are too large and indivisible. Moreover, the more millenarian the group, the less effective is either carrot or stick. The stick (sanctions) is easily counterproductive because, unlike the carrot, it can contribute to a group’s isolation and create solidarity versus an external enemy. Elsewhere (Wintrobe 2006a) I suggest that both the carrot and stick should be used, though there is no guarantee that this will have any effect. Nevertheless, if the relative returns to moderation can be raised, this will work as shown in Figure 5.15.

E(Zi)

Figure 5.14. Solution 1: “Reframe.”

E(Zp)

Accommodation Versus Conflict

167

1 Security

2

Control of the political system

Figure 5.15. Solution 2: Raise relative gains to moderation.

On solution 8. Break the link between pressure and power. For example, changing to a PR (proportional representation) system gives power at lower levels of pressure and reduces extremism. On solution 4. Make the indivisible divisible. If one can unbundle the goal or make the indivisible divisible, there may be ways to provide these goals in a way that dries up support for the grander ambitions of the leaders of extremist groups. In effect this is how Keynesian economics dried up support for communism. In the same way problems of ethnic conflict have been solved in Canada and in many other states through institutions that give different groups a share in power. Thus features like federalism, the division of powers, checks and balances, PR, and so forth all give groups some power without satisfying what they thought was an indivisible objective. A specific example of this in the current context is federalism. On solution 8. Federalism. This is the Canadian way: Make the indivisible divisible. It raises return to “moderation” by making some “nationhood” possible without separation. It was also the Biden (Galbraith) plan in Iraq. However, federalism also empowers ethnic groups, thus raising solidarity within them, which raises the return to extreme methods. So federalism works if the marginal effect of satisfying the demand for nationhood is greater than the marginal increase in the return to violence due to the increase in solidarity. Of course it helps if there is no religious dimension (especially the different

168

Ronald Wintrobe

N 1 2

2

1

I/M

Figure 5.16. Successful federalism.

monotheisms, which breed solidarity within and conflict without). An example of successful federalism is shown in Figure 5.16. Thus, to be successful, federalism should be accompanied by institutions or policies that reduce solidarity within the groups. Some Canadian examples of these are the following: • Reduce barriers to exit from ethnic enclaves (such as Trudeau’s provision of language rights for Francophones across Canada). • National programs with common outcomes (such as the fact that Medicare is portable across provinces in Canada, i.e., it is possible to obtain free medical treatment in any province). • Having a common external “enemy” (the fact that the United States doesn’t have a national health system is (was?) extremely good for Canadian solidarity). It has been suggested that the national health system (Medicare) currently plays a role analogous to that of the railways across Canada in an earlier time. Indeed if the United States succeeds in getting a national health system, this would have the unintended consequence of reducing Canadian patriotism, a possibility that no doubt greatly concerns American politicians. • The “Clarity act,” which mandated that in any new referendum on Quebec separatism, the question posed had to be “clear.” Th is

Accommodation Versus Conflict

169

separated hard-line from “soft” separatists. In previous referendum questions, Quebeckers were asked whether they favored “sovereigntyassociation,” that is, a “separation” from the Canadian federation combined with a proposed looser “association” with the rest of Canada. Some Quebeckers who favored separation thought that afterward they would still be able to use the Canadian dollar (this is not unlikely), their Canadian passports (definitely unlikely), have access to the Canadian health system (unlikely), and so on. On solution 6. Economize on the need for interaction and trust with High Quality InstitutionsTM (Easterly and Levine 1997). Most ethnic groups live in peace (Fearon and Laitin 1996), and, even more to the point, there is evidence that ethnic fractionalization does not cause conflict when institutions are “high quality” (Easterly and Levine 1997; Alesina 2005). The most important requisites for institutions to be high quality are: (a) freedom from expropriation, (b) rule of law, and (c) bureaucratic quality. High-quality institutions are not the same thing as democracies but are similarly correlated with GNP per capita. These institutions constrain the amount of damage one group can do to the other, and they provide legal enforcement; thus “ethnic” enforcement (i.e., relying on ethnic leaders to enforce contracts or arrangements) becomes less important. Of course, it is not easy to create such institutions, as the example of Kosovo demonstrates. On solution 9. Marginalize or accommodate the extremists. In an interesting analysis, Jung, Lust-Okar, and Shapiro (2008) look at three cases of conflicted societies (South Africa under apartheid, Northern Ireland, and Israel-Palestine). Using the successful negotiations in South Africa and Northern Ireland as a model, they suggest that one of the lessons in both of these cases is that the presence of a strong radical flank need not bode poorly for peace, provided efforts are made to incorporate its members if they cannot be marginalized. If the government conducting the negotiations has strong domestic support for a policy of accommodation, this increases the chance of success. De Klerk obtained this with his referendum, while Mandela’s policy of refusing to renounce violence made his position unassailable within the African National Congress (ANC) and kept the extremists within that group on the side of the ANC. Thus if the extremists feel represented by the party conducting the negotiation, there is a better chance of success; this explains such diverse actions as Nixon’s capacity to make a deal with China and Sharon’s ability to withdraw from Gaza.

170

Ronald Wintrobe

In Israel-Palestine, it has been clear to me for some time that the “solution” is obvious. A deal will eventually be struck in which Palestinians have a state, it will contain part of East Jerusalem and 92–97 percent of the West Bank, and so on. It is not the absence of a solution but the difficulty of getting there that poses the problem. Why is it so difficult to get there? Essentially, because as soon as progress is made toward some agreement, extremists on one side or the other do something to sabotage it. The most spectacular example of this was the assassination of Yitzhak Rabin, but there is no shortage of cases of sabotage from the other side, that is, by Hamas and other radical Palestinian factions. If this is accepted, then the question is how to accommodate or marginalize the extremist flank. Mitchell, Evans, and O’Leary (2009) argue that power-sharing institutions in Northern Ireland encouraged the development of electoral strategies based on “ethnic tribune appeals” in which extremist parties did not lose their ethnic identity but combined it with increased pragmatism. In Israel-Palestine, it is hard to believe the extremists on either side could be brought in. But the same thing might have been said about Northern Ireland, and it might have been said in Israel or Egypt before Sadat’s trip to Israel and the subsequent peace treaty. The key seems to be that whoever makes the deal has to have so much domestic support that the extremists have to either get onboard or become marginalized. No such figure with that capacity appears to be around as of this writing.

Conclusion In this essay I sketched a simple model to explain the demand and supply of extremist activity, developed previously (Wintrobe 2006a, 2006b), and then applied this model in a new way to the problem of accommodation versus conflict between groups in divided societies. For illustrative purposes, I focused specifically on the vexatious Israeli-Palestinian problem, though the analysis is general and could be applied to any situation where groups are in conflict. Each of the two groups is seeking a fundamental objective such as nationhood or autonomy. It can use moderate or extremist (violent) methods to pressure the other group with whom it is in conflict. From the economic point of view, the basic difference between moderation and violence is that violence is risky. It can backfire or result in defeat, humiliation, or condem-

Accommodation Versus Conflict

171

nation in a way that moderate methods usually cannot. On the other hand, violence can achieve goals that sometimes moderation cannot. I focus on the situation where the group is in conflict with another group, and their interests are diametrically opposed. The leader of each group decides on how much violence or moderation to use to pressure the other group. Pressure produces power, and power may be used to obtain some intermediate goal that is instrumental to achieving the final goal of the group. To illustrate with the example used throughout the chapter, Israelis and Palestinians may be thought of as in conflict over land. But land is just an intermediate goal for both groups. The fundamental goal is nationhood. The essence of the conflict as seen by many is that there is just not enough land; if Palestine, say, becomes a nation with secure borders and all the other trappings of a modern state, there is not enough land left over for Israel to be in a secure position. And vice versa. This point is familiar. The distinctive twist in this chapter is that there is a zone of “increasing returns” or what economists call an “indivisibility” on the way to reaching this objective. This “hill” or “mountain” on the way to nationhood is what makes violence so attractive in the same way, metaphorically, that a hill on the road may cause a driver to accelerate. I then asked what the effects of various events are on the tendencies for each of the two groups to use violence. This is an exercise in positive economics. I derive interesting implications of certain events or policies on the likelihood of either party’s use of violence. If an event results in one party wishing to use more violence, this is a prediction of the model and not a criticism or recommendation of that policy or event. For purposes of modeling I initially ignore differences of opinion within each group and assume each one has a “leadership” that makes decisions for the group, though informally it is obvious what the effects would be of relaxing this assumption. I further assume that if one group resorts less to violence, it is easier for the other group to have peaceful accommodation with that group. If both groups use less violence, I assume accommodation between the two of them is more likely. For illustrative purposes, and because of their importance, I focus on two events: the construction of the wall (the West Bank barrier) and the Israeli withdrawal from Gaza under Ariel Sharon. The wall makes Palestinians less violent because it reduces the effectiveness of violent methods in pressuring Israelis. Its effect on Israel is more complex. If it increases Israeli security but still does not satisfy the aspirations to nationhood of the Israeli leadership, it leads Israel to become more violent, not less. However, if

172

Ronald Wintrobe

it does satisfy these aspirations—it makes Israel a “secure nation”—it reduces Israeli violence. In this latter case, the wall promotes peace in the sense that both parties become less violent. Of course, it might be objected from the normative point of view that it does this in what might be called an “asymmetrical” or “dictatorial” manner in that it provides full Israeli security and nationhood while depriving Palestinians of their aspirations as well as the means to effectively contest the imposition of borders by Israel. But the analysis here is positive, not normative. The analysis of the Israeli withdrawal from Gaza is simpler. It makes Palestinians more violent and makes Israelis less so. If Gaza were sufficient to satisfy Palestinian aspirations to statehood, in the way that the wall may for Israel, then they would become less violent, but this does not strike me as a credible line of thought. Finally I looked at policy and normative economics, as well as some possible solutions to the problem of group conflict suggested by the analysis. I offered nine ways to promote peace suggested by the model and looked briefly at the experience of other groups that have successfully managed conflict—South Africa, Northern Ireland, and Canada—and asked what the lessons might be there, as interpreted through the lens of my model, for the resolution of conflict between other groups. I hope the analysis here provides some food for thought. Notes I am indebted to Brendan O’Leary and to seminar participants at the University of Pennsylvania and to Nathan Sussman for helpful comments. Responsibility for any errors remains mine. I wish to thank the Social Sciences and Humanities Research Council of Canada for financial assistance. 1. In turn, the indivisibility or zone of increasing returns arises because the intermediate goal can be likened to a missing “factor” in the production function of the ultimate goals. See Wintrobe 2006a for details. 2. Some other reasons why deterrence variables may be ineffective can be elaborated. Work done since Becker finds that there are other limits to the effectiveness of punishment besides the cost of the resources used in pursuing criminals and punishing them discussed by Becker (1968). The first is that overly severe punishments could lose the support of the community, thus reducing p in cases where such measures crucial to catching and convicting criminals (Akerlof and Yellen 1994). The second is that in the case of capital punishment, juries might be less willing to convict when the judge has the option of sentencing the defendant to death (Andreoni 1995). These are elaborated in Wintrobe 2006a and 2006b.

Accommodation Versus Conflict

173

3. See Mueller 2003 for a survey. 4. See Austin Smith in Mueller 1997 for a survey. 5. Note that the vertical axis depicts the total proceeds or return from a given “portfolio” of moderate and terrorist actions, not the average expected return on the portfolio. 6. Note that the constraint in Figure 5.7 describes the choices available to a subordinate or member and is not the constraint facing the leader. Hence the equilibrium autonomy and solidarity in Figures 5.7 is that of a member, not the leader. The leader’s equilibrium cannot be described with this apparatus. 7. Sandler and Lapan (1988) also consider the case of “fanatical” terrorists, defined as those who do not fear death, and suggest that deterrence is ineffective for such individuals. However, note that the point here is somewhat different: “fanatical” terrorists here are those who appear fanatically loyal or obedient to the organization’s wishes. It is worth noting that sufficiently large penalties can be effective. Finally, in this model solidarity, even when extreme, is always contingent. To illustrate this point, it is useful to recall that it is often suggested that no group ever demonstrated more loyalty to its leader than the SS did to Hitler. Yet toward the end of the war, when it was obvious that the Nazi regime was collapsing, these people deserted it in large numbers (see the analysis in Wintrobe 1998, chap. 13). 8. Conflicts between these groups are multidimensional, including the possibility of secular marriage, observance of various religious practices, and so forth, and many ultraorthodox are not Zionists at all, so they are not necessarily more extreme on the dimension of land for Israel. Some do not recognize the Jewish State of Israel. 9. http://en.wikipedia.org /wiki/Israeli _West _Bank _barrier. 10. For example, the Israeli Peace Now movement has stated that while they would support a barrier that follows the 1949 Armistice lines, the “current route of the fence is intended to destroy all chances of a future peace settlement with the Palestinians and to annex as much land as possible from the West Bank” and that the barrier would “only increase the blood to be spilt on both sides and continue the sacrificing of Israeli and Palestinian lives for the settlements.” http://en.wikipedia.org /wiki/Israeli _West _Bank _barrier.

References Akerlof, George. 1991. “Procrastination and Obedience.” American Economic Review 81: 1–19. ———. 2000. “Economics of Identity.” Quarterly Journal of Economics 115: 3 (August), 715–753. Akerlof, George, and Janet Yellen. 1994. “Gang Behaviour, Law Enforcement and Community Behaviour.” In H. J. Aaron, T. E. Mann, and T. Taylor, eds., Values and Public Policy. Washington, DC: Brookings Institution. Alesina, Alberto. 2005. “Ethnic Diversity and Economic Per for mance.” Journal of Economic Literatur 95: 913–35.

174

Ronald Wintrobe

Andreoni, James. 1995. “Criminal Deterrence in the Reduced Form: A New Perspective on Ehrlich’s Seminal Study.” Economic Inquiry 33 (July): 476–83. Ariely, Dan. 2008. Predictably Irrational. New York: Harper Collins. Armstrong, Karen. 2005. A Short History of Myth. New York: Canongate Books. Becker, Gary. 1968. “Crime and Punishment: An Economic Approach.” Journal of Political Economy 76: 169ff. Black, Anthony. 2001. The History of Islamic Political Thought. New York: Routledge. Easterly, William, and Ross Levine. 1997. “Africa’s Growth Tragedy: Policies and Ethnic Divisions.” Quarterly Journal of Economics 112, 4: 1203–50. Fearon, James, and David Laitin. 1996. “Explaining Interethnic Cooperation.” American Political Science Review 90 (December): 715–35. Galbraith, Peter. 2006. The End of Iraq: How American Incompetence Created a War Without End. New York: Simon and Schuster. Gambetta, Diego, ed. 2005. Making Sense of Suicide Missions. Oxford: Oxford University Press. Ginsborg, Paul. 2001. Italy and Its Discontents. London: Penguin Press. Gold, Riger V. 2000. “Revenge as Sanction and Solidarity Display: An Analysis of Vendettas in 19th Century Corsa Rica.” American Sociological Review 65: 682–704. Gunaratna, Robin. 2002. Inside Al Qaeda: Global Network of Terror. New York: Columbia University Press. Hoff man, Bruce. 1998. Inside Terrorism. New York: Columbia University Press. Jung, Courtney, Ellen Lust-Okar, and Ian Shapiro. 2008. “Problems and Prospects for Democratic Settlements: South Africa as a Model for the Middle East and Northern Ireland?” In Stathis N. Kalyvas, Ian Shapiro, and Tarek Masoud, eds., Order, Conflict and Violence. Cambridge: Cambridge University Press. Karl, Jonathan. 1995. The Right To Bear Arms: The Rise of America’s New Militias. New York: Harper Paperback. Kaufman, Stuart. 2006. “Symbolic Politics or Rational Choice? Testing Theories of Extreme Ethnic Violence.” International Security 30, 4: 45–86. Mitchell, Paul, Geoff rey Evans, and Brendan O’Leary. 2009. “Extremist Outbidding in Ethnic Party Systems Is Not Inevitable: Tribune Parties in Northern Ireland.” Political Studies 57: 397– 421. Mueller, Dennis. ed. 1997. Perspectives in Public Choice. New York: Cambridge University Press. Mueller, Dennis. 2003. Public Choice III. New York: Cambridge University Press. Opp, Karl-Dieter, and Wolfgang Ruehl. 1990. “Repression, Micro-Mobilization, and Political Protest.” Social Forces 69: 521– 47. Oz, Amos. 2006. How to Cure a Fanatic. Princeton, NJ: Princeton University Press. Pape, Robert. 2005. Dying To Win: The Strategic Logic of Suicide Terrorism. New York: Random House, 2005.

Accommodation Versus Conflict

175

Post, Jerrold M. 1990. “Terrorist Psycho-Logic: Terrorist Behaviour as a Product of Psychological Forces.” In Walter Reich, ed., Origins of Terrorism: Psychologies, Ideologies, Theologies, States of Mind. New York: Cambridge University Press. Rasler, Karen. 1996. “Concessions, Repression and Political Protest: A Model of Escalation in the Iranian Revolution. American Sociological Review 61: 132–52. Ricolfi, Luca. 2005. “Palestinians, 1980–2001.” In Diego Gambetta, ed., Making Sense of Suicide Missions. Oxford: Oxford University Press. Ruthven, M. 2000. Islam in the World. 2nd ed. London: Penguin. Sandler, Todd, and Harvey E. Lapan. 1988. “The Calculus of Dissent: An Analysis of Terrorists’ Choice of Targets.” Synthese 76: 245– 61. Sandler, Todd, and Walter Enders. 2002. “Patterns of Transnational Terrorism, 1970– 99: Alternative Time Series Estimates.” International Studies Quarterly 46 (June): 145– 65. Wintrobe, Ronald. 1997. “Modern Bureaucratic Theory.” In Dennis Mueller, ed., Public Choice: A Handbook. New York: Cambridge University Press. ———. 1998. The Political Economy of Dictatorship. New York: Cambridge University Press. ———. 2006a. Rational Extremism: The Political Economy of Radicalism. Cambridge: Cambridge University Press. ———. 2006b. “Extremism, Suicide Bombing and Authoritarianism.” Public Choice 128, 1–2: 169–95.

CHAPTER 6

The Success of Religion as a Source for Compromise in Divided Empires: Ottoman and Safavid, Past and Present Benjamin Braude

Contrary to the conventional assumptions of the moment, religion has been a means for political accommodation, integration, and stability in the Middle East for most of half a millennium, if not more. The forces that have converted it into a source for conflict over recent centuries have arisen not from the intrinsic qualities of religion itself but rather from the corrosive effects of French and Scottish notions, such as the Enlightenment, and a few of its consequences: notably the bizarre fancies that one can and should fathom the popular will, that it should exercise power, that there is such a thing as a nation, and that it deserves a polity of its own in which that popular will should be exercised. All of this should be familiar to those who read Elie Kedourie. The reality and benefit in contemporary politics of the nation-state’s supposed resulting system of rule, democracy, are much exaggerated. Many so-called democracies in fact depend on nonparticipation and antidemocratic mechanisms rather than democratic mobilization for their effectiveness and survival. A system in which fundamental and radical change is imposed upon a country over the repeated opposition of a majority of the electorate is not democratic. Yet such was the case in the Thatcherite United Kingdom. A system in which roughly half of the eligible population actually votes and of that number about 24 percent vote for one candidate and about 23 percent vote for his rival with the result that Mr. 23 percent wins can hardly be called democratic, as was the case in the 2000 U.S. election. Yet such polities

Religion as a Source for Compromise

177

offer themselves as models for emulation and propose to impose, willy-nilly, democracy around the world. In the early modern era, two major polyethnic, polyglot, and multireligious empires dominated the Near East: the Ottoman and the Safavid. Each faced a similar structural challenge: how to rule a majority population that in religious terms was different from themselves. Paradoxically each employed religion as one of the devices to create relatively stable long-lasting polities. A caveat is necessary before proceeding further. Repeatedly within this essay a variety of ethnic and other groups will be introduced and discussed in their contemporary and historical contexts. Such terms of identity are a necessary heuristic device. In fact as with most collective identities, the precise nature of the collectivity has been and remains constantly changing, subject to the bias of perspective and the political and material interest of the moment. Although the complexity of who is a Jew, for example, is wellknown as a regular crisis point in Israeli politics, it is no more and no less fraught and problematic than the question of who is a Turk, a Kurd, a Shiite, an Azeri, a Greek, a Macedonian, an Eastern Orthodox Christian, or any of the other many groups that will be discussed. It is precisely because such categories are insecure that conflicts arise. The Sunni Ottoman dynasty of mixed but predominantly Turkic origin ruled a territory that from about 1300 until 1517 was majority Christian, overwhelmingly followers of the Orthodox communion, a church dominated by Greek speakers but including various Slavic, Albanian, Vlach, and even Turkish speakers among others. Alongside these Orthodox faithful was the much smaller, theologically and hierarchically distinctive Armenian Gregorian Apostolic Church. In the closing decades of the fifteenth century because of their expansion into eastern Anatolia, the Ottomans started to acquire more Armenians in addition to Kurds, most of whom were like themselves Sunni, as well as smaller numbers of Shiite Turks. In the wake of the expulsions and persecutions of the Jews in the Iberian peninsula about that same time, the empire acquired a significant Sefardi community who settled in important urban centers and eventually overwhelmed preexisting Greekspeaking Jews. After 1517 with the conquest of the Arabic-speaking lands, the Ottoman Empire became for the first time a Sunni-majority state. As such, the dynasty outlasted its Romanov and Habsburg rivals past World War I, finally disappearing only in 1924. In 1501 the Shiite Safavids began a military and religious campaign that within a few decades encompassed and defined most of what today is Iran

178

Benjamin Braude

(Abisaab 2004). The Safavids themselves seem to have arisen out of a KurdishTurkish Sufi (at least nominally Sunni) order that gradually converted to a form of Shiism. Once in power they fashioned their political regime into a millenarian Twelver Shiite movement. The population of what became their domain was, in striking contrast to the Ottoman realm, largely Muslim but still highly diverse. Few would claim that Twelver Shiism had previously predominated. Varied, nominally Sunni-Sufi, Sunni, Shiite, and even Shamanist groups had flourished in the fifteenth century. Since Shiism had not heretofore predominated in this realm, they had to import Shiite scholars from Syria to establish this new state-sponsored religion. In ethnic and linguistic terms, the polity of Iran has boasted the most diverse compact population for the longest period of time of any entity within the Near East. Iran has encompassed speakers of Farsi, Turkish (of various dialects and types), Arabic, Baluchi, and Kurdish, all of whom persist today and doubtless existed half a millennium ago. In addition, the domain included Armenians and Jews, but these non-Muslim groups were insignificant in comparison to those under Ottoman rule. The 1722 Afghan invasion led to the fall of the last of the dynasty, Shah Abbas II, in 1732. Each dynasty faced similar challenges but developed different responses. The Ottomans established a policy of assimilation and co-option of Christian elites through the devshirme system (the draft of their young men) and occasional syncretism as evident in the Bektashi and other Sufi orders. The Safavids mounted an aggressive and highly successful campaign of conversion, transforming Iran into the only significant Shiite dominated state in the world, a status it has retained to this day. Though I have introduced the Safavids at this initial stage, my focus will be on the Ottomans, but ultimately I will return to both to analyze conflicts that have emerged in each region after the disappearance of their respective imperial regimes. My fi nal points of comparison will be: (1) Greece and the Macedonian question, (2) Turkey and the Kurdish question, (3) Iraq and the Kurdish question, and (4) Islamic Iran and the Azeri question. I will argue that the general decline of religion as a means of reducing communal strife has led to very different outcomes for Greece, Turkey, and Iraq, as opposed to Iran. I focus on the Ottoman Empire because it is my area of expertise but also because it contained three of the four examples that I wish to address. Although for most of its history the Ottoman Empire was a Muslim state both in terms of both its leadership and its population, for the formative first two centuries of its existence it was in fact a Christian state, at least as defined

Religion as a Source for Compromise

179

by its majority. Even after its Muslim majority was established with the addition of the Arab provinces, the older foundational reality persisted. True, subsequent Ottoman court chroniclers and early twentieth-century European scholarship have ignored this fact by trumpeting the Ottomans as a ghazi state, dedicated to spreading and promoting Islam at its frontier with Christendom (Wittek 1938). In this interpretation, ghaza has been regarded as synonymous with jihad. In fact whatever jihad can mean, it carries a valence different from ghaza. Although the latter can mean a war for the cause of Islam, it also has a more generalized meaning of raid for the cause of booty. That meaning persists in the etymologically related razzia, a word to be found in English and French, often used to describe local conflicts in Morocco and Algeria. Recent scholarship has challenged the original ghaza thesis, without necessarily constructing a clear alternative (Lindner 1983; Kafadar 1995). That is probably for the best. Given the complexity of the process of state formation and the sparsity of evidence, a simple alternative cannot exist and it would be best to entertain different hypotheses simultaneously. The Ottoman Empire was a Muslim state forced to deal with Christian demographic dominance. Except for the early caliphates, no other longlasting Muslim state has had to confront such a situation. This reality created a significant alternative hypothesis to the ghazi thesis. More than seventy years ago the prolific Rumanian historian and nationalist politician Nicolae Iorga published a slim essay, Byzance après Byzance, as a quasielegaic postscript to his monumental history of the Byzantine Empire. Its core argument asserted that Byzantium survived as a cultural and political reality through the quasi-autonomous Principality of Moldavia and Wallachia that ruled by Ottoman license and through a group known as the Phanariots, named after the district in the capital, Phanar (Fener in modern Turkish), where they settled. They were the Greek Orthodox communal leaders in Istanbul who wielded great influence through their own wealth, through their power over the Patriarchate of Constantinople, and through their own occasional control of that principality as well. More broadly he argued that the Ottoman Empire itself represented a continuation of many Byzantine traditions and could not be properly understood without that heritage. Iorga was not the first to make this claim. In fact it had been part of the arsenal of arguments mounted by earlier Moldavian and Wallachian cultural and political leaders who went even further by asserting, with some truth, that they were descendants of Byzantine aristocracy (Leal 2003, 410–13). It took a few centuries for it to emerge in the guise of the scholarly argument expressed

180

Benjamin Braude

by Iorga. For some Rumanian nationalists and historians, the tie to Byzantium was essential for the foundation myth of Rumania’s origin, part of the process by which they transformed themselves from Rumania to Romania, that is, from the land of Rum, the Arabized name for Byzantium and its varied regions and inhabitants of whatever identity, Muslim or Christian, Turkish or Persian, even Greek, to the land of Rome. Although Rumanians were among the most persistent in asserting GreekTurkish consociation, to use the jargon of the day, even the Greeks originally acknowledged a form of it. However, modern Greece has had much difficulty in accepting this historical fact. One of the first to suggest an Ottoman continuity of Byzantine and even Hellenic consciousness was the fifteenth-century Greek biographer of Mehmet the Conqueror of Constantinople, Kritovoulos of Imbros. To be sure his work should be seen more as a mendacious job application than as a candid and accurate assessment. The manuscript itself survived in only one copy, housed in the Topkapi Palace Archives, largely unremarked until the nineteenth century. Knowledge of its contents today owes much to twentieth-century Rumanian scholarship; so clearly its creation and diff usion depended on interested parties. But the attempt to imagine the Ottoman Empire in classical terms was not restricted to the occasional Greek. The Ottomans, particularly during the reign of Mehmed II the Conqueror of Constantinople, saw themselves as rightful heirs to the Roman imperial tradition that he self-consciously tried to emulate. They were content to maintain the fourth-century name of their capital (or some variation thereof). The formal adoption of Istanbul did not occur until 1924 when it ceased to be the capital of anything after the Turkish Republic moved its political center eastward to Ankara. Both geographically and ideologically, they were the most European of all the Islamic empires, internalizing European ideas of color prejudice and aspects of the European political cosmology (Braude 2011). But for the conquest of Arab lands, the Ottoman Empire might have embraced a form of Muslim-Christian religious syncretism, along the lines of the Hindu-Muslim-Sufi practice promulgated in Mughal India under Akbar. In fact what the Ottomans created through the devshirme, the Janissary corps, the Sufi orders, and the recognition of religious autonomy (i.e., the so-called millet system) was an elaborate and highly effective system for coopting its majority Christian population into a form of consociation (Braude and Lewis 1982). Beginning in the fourteenth century the Ottoman authorities sent agents into the non-Hellenic regions of the Balkans to identify and

Religion as a Source for Compromise

181

draft promising youths who would be converted to Islam and trained for ser vice in the Ottoman military or administration. The particular groups targeted were significant. In order to reduce friction with the overwhelmingly Greek-speaking hierarchy of the Orthodox church, the Ottomans did not regularly subject Greeks to devshirme recruitment. Although most of the recruits were Albanian and Slavic-speaking Orthodox Christians, Slavicspeaking Muslims, largely Bosnians, were eligible to volunteer. The latter constituted a community of recent converts to Islam without the connections and resources of the established Turkish Muslim elite. The Bosnians did not want to miss out on a good thing. The opportunities for advancement in and the privileges of this system were substantial. All those recruited became members of the askeri (military) class. As such they paid no taxes; instead they spent them. They could advance—at least in theory—to the very highest ranks of the Ottoman administration and military. Relatively few reached those heights, but the prospect remained alluring even for those who failed. Although they lost communion with their church, they did not lose communication with their community. Recruits could maintain contact with their families and villages of origin, defending their interests and helping others join the service in turn. This system reduced somewhat the alienation and subjugation that a subject population might normally feel toward an alien hegemon. Although the Orthodox hierarchy—overwhelmingly Hellenized— decried the loss of souls to Islam, they were assuaged somewhat by the fact that most were Slavs, thereby weakening these rivals in the political struggles within the church. The Balkan non-Greek Orthodox could not advance socially and economically above their own rural status without undergoing one or another form of deracination and self-abnegation. If they wanted to seek advancement in the church hierarchy, they had to abandon their native culture and become Hellenized. They also had to become monks, at least nominally celibate. However, they did retain and enhance their religious beliefs. On the other hand, if they wanted to advance in the Ottoman Empire, the terms were significantly easier. They could retain much of their native culture. There were no vows of celibacy. All they had to acquire was enough Turkish to function at their pay scale and abandon Christian Orthodoxy, the latter a conversion that was even easier than it might initially seem. If Paris was worth a mass, Constantinople was certainly worth a shehadah; at only seven words the Muslim testimony of faith was much shorter. Religiously the recruits rarely displayed the zeal of sincere converts. Many, particularly in the Janissary corps, joined the heterodox Bektashi Sufi Order, whose divine

182

Benjamin Braude

cosmology comprised a hash of Sunni, Shiite, and Christian beliefs that easily accommodated whatever Christianity the youths had remembered. In practice they played fast and loose with the Quranic prohibition against wine that most Muslims extend to include all alcohol. They were famous consumers of fermented drink. This Ottoman system had the simultaneous benefit of recruiting among the best, brightest, and most able of their Christian subjects for imperial service and denying them to the ser vice of any potential opposition. Much of the conflicts that undermined the empire in the eighteenth and nineteenth centuries arose after this recruitment had broken down during the seventeenth century. True, the Ottomans did employ Christians well into the eighteenth century, most notably as advisors on foreign affairs, but overall the numbers now recruited were much smaller than had prevailed in the past and their level of influence never reached the heights of their converted predecessors. The cause of the breakdown was that the opportunities afforded by this system were too attractive. In theory, those recruited by the devshirme were not to retain their privileged positions into subsequent generations, thereby opening the ranks to fresh talent. Instead, the offspring of these new Muslims were to seek new opportunities in other domains. However, once recruited, the fathers sought to pass on their privileged positions to their own descendants. Perhaps the Christian hierarchy was wise to insist upon celibacy? The examples discussed so far imagine a functioning and effective relationship between two supposedly hostile rivals, the Ottoman state and the Greek and/or Orthodox community, largely in political and economic terms. However, a high-ranking Greek of the late eighteenth century described this relationship in a somewhat a different set of categories, intertwined political-religious-institutional. A Greek Orthodox apologia for this quasi-concordat appeared in a tract published in Istanbul in 1798. The author was the Greek Patriarch of Jerusalem, Anthimos I. The year is significant, as the arguments that follow make clear: See how clearly our Lord . . . has undertaken to guard once more the unsullied Holy and Orthodox faith. . . . He raised out of nothing this powerful empire of the Ottomans. . . . The all-mighty Lord, then, has placed over us this high kingdom, “for there is no power but of God.” . . . For this reason he puts into the heart of the Sultan of these Ottomans an inclination to keep free the religious beliefs of

Religion as a Source for Compromise

183

our Orthodox faith and, as a work of supererogation to protect them, even to the point of occasionally chastising Christians who deviate from their faith, that they have always before their eyes the fear of God. . . . Brothers, do not be led astray from the path of salvation; but as you have always with bravery and steadfastness trampled underfoot the wiles of the devil, so now also close your eyes and give no hearing to those newly-appearing hopes of liberty “for now is salvation nearer to us.” . . . But let us analyse more scientifically the very name of this “liberty.” . . . True freedom is A, that disposition of the rational soul, which by the grace of God, leads man to the good without, however compelling him. Such liberty is called “freedom of the will.” B, it is freedom for man to be able, unhindered, to put into practice the appetites of his desires, which is insubordination. C, it is called freedom for someone to live according to divine and human laws, that is to live free of every reproach of conscience and free of civil discipline. . . . The only praiseworthy liberty is the third noted above. . . . Let us have steadfastness and prudence, let us not lose the unfading crowns of eternal blessedness for a false and non-existent liberty in this present life. Let us not deprive ourselves of the inexpressible rewards. This statement represents an Orthodox Christian philosophical response to the claims for freedom articulated by the French Revolution and its ideological forerunners. Articulation of this worldview was not limited to the Greeks at the end of the eighteenth century but emerges as well in the quixotic career of an Armenian adventurer, Joseph Emin, a slightly earlier contemporary of Patriarch Anthimos. Although the Armenians were a much smaller and later addition to the empire, they, too, acquired hierarchical autonomy. However, they were not subject to the devshirme system in any systematic fashion. Joseph Emin (1726–1809), a would-be Madras-based Armenian liberator, inspired by his years of military ser vice in England, embarked upon a secret mission in the mid-eighteenth century to raise a revolt of his brethren against Ottoman rule. In 1759 he made his way to Aleppo and hired Armenian servants, joining a caravan heading north into Armenian Anatolia. However, he soon tired of the large company and set out on his own, accompanied only by his servants. He was supremely confident, for he had two “instruments of guidance”: “a map of Asia made at Paris” and a compass, “the fruits of European wisdom.” His servants, if his own account can be

184

Benjamin Braude

believed, were now in awe of him, for he was “in every village respected by the Turks.” Unlike “the poor Armenian merchants” of the initial caravan, he “behaved in such a domineering way, that the Turks imagined he was a great Armenian, a favourite of the sultan, with a firman in his possession.” At their next halt he separated from his entourage to visit on his own a small Armenian village. Here his identity and reception came to be reversed: “When the [Armenian] countrymen saw him mounted on a fine grey horse, they took him to be a Turkish trooper; but when he spoke to them in their own language, it made them very angry; they ran to . . . beat him heartily, using menacing language, and asking, How he durst travel alone without a caravan, since he was a Christian?” He escaped the beating only by convincing his compatriots that he was in fact an Ottoman Turk. Immediately “the poor creatures were frightened out of their senses . . . down upon their knees, begging for mercy . . . expressing their fidelity to the Othmans, who are the only people able to travel alone.” Eventually he was able to conduct a private meeting with the village headman of whom he asked, “You, Christians, what is the reason of your objecting, if any of your countrymen should take a fancy to be a warrior? And why are you not free? Why have you not a sovereign of your own?” The response, “Sir, our liberty is in the next world; our king is Jesus Christ.” Emin said, “How came that about? Who told you so?” They answered, “The Holy Fathers of the Church, who say, the Armenian nation has been subject to the Mahometans from the creation of the world, and must remain so till the day of resurrection; otherwise we could soon drive the Othmans out of our country.” At this point Emin revealed himself as a Christian. He drew out from his pocket the Geographical History of the fift h-century Armenian scholar Movses Khorenatsi and “sent for a priest that could read a little.” He showed him the genealogy of the kings of the Armenians and posed Christ’s answer to the disciples’ question, “Who should inherit the kingdom of God?” “Whosoever shall leave behind him his father, mother, brother, and wife, lift up the cross and follow me.” He then noted that the “Christians of Frankestan” did not understand the Gospel in the manner in which the holy fathers in the Ottoman Empire have explained it and thus they are not “great slaves to the Mahometans as we are now.” He then explained Christ’s answer, according to his own lights: “The meaning of shouldering the cross is the ensign which the brave soldiers carry against the Infidels, to fight and die under it; those being

Religion as a Source for Compromise

185

the true Christians, who can inherit the kingdom of God; and not they that lead a cowardly life, like us, who are become cattle devoured by wolves. . . . For example, a rational being should not suffer himself to be a wilful slave to others; he ought not to be domineered over by his fellow-christians; since God has created them all free alike, to be ruled or governed by good laws, with the same justice to the rich or to the poor.” As the harangue continued the priest interrupted to exclaim, “He is in the right” and called all the villagers out to praise and embrace their erstwhile victim; then he added, “love and respect him; for he is the very man prophesied of by St. Nerses the Great, about six hundred and thirty years ago, who will be the instrument of delivering us from the hands of our oppressors, and of the enemies of our faith.” The headman was startled by this change of attitude and demanded, “What was that you pronounced? or why are we kept in ignorance?” The priest replied, “My dear people, what signifies pulling off shoes and stockings before we reach the bank of the rivulet; everything in good time; besides, the holy prophecy is for 666 years to be fulfilled; during that period, we must continue as in subjection; 638 years are expired, there remain 28 years more to complete our persecution; then we shall become free; then no power in the world can oppress us. Our guest must have seen a great deal of the world, as we may judge by his conduct as well as by his great Father; you may be judges yourselves: you were frightened at first, when you imagined he was a Turk. . . . I say, he is the very man; but he must wait, and go through various scenes of life twenty or thirty years more. I tell it to his face; it is not he that does these things, it is the great God above.” In the best Don Quixote tradition, Emin was delighted with this response, commenting by way of summary in his characteristic third person, “In this method he sowed the corn grain of true religion, and planted the admirable zeal of military spirit every where he travelled.” The priest’s skillful diff usion of what might have been a nasty confrontation is reminiscent of the story of the wonder-working hassidic rebbe and the czar. Stories of the miracles the rebbe wrought had reached all way to the Romanov court. So the czar demanded that he be brought to St. Petersburg to perform. The czar

186

Benjamin Braude

insisted that he teach his dog to talk. The rebbe agreed on one condition. Since the task was so difficult he needed extra time. He would accept the challenge as long as he had ten years in which to meet it. He returned home to his followers, who were aghast at the rebbe’s promise. As much as they admired him, even they did not believe he could make a dog talk. The rebbe reassured them: “In ten years either the dog will die, or the czar will die, or I will die.” So the priest doubtless realized that in twenty or thirty years his flock would die, or Emin would die, or the priest would die. And indeed while Emin still lived fifty more years, he never returned to the village to harvest “the zeal of military spirit” that he had planted. And no one else did either. This episode demands multiple levels of analysis. Contrary to expectations, Emin as an Armenian is treated better by the Turks, while Emin as a Turk is treated better by the Armenians. And then there is the esoteric quality of Armenian millenarianism and the subtle dueling between the priest and Emin. So subtle is the duel that he misinterprets a sly and skillful rebuff as endorsement. The contradictory receptions Emin received from Turk and Armenian alert us to the interplay of ethnicity and class in the structure of Ottoman society. Contrary to oversimplified interpretations that claim that one’s position was determined solely by religious identity and that all nonMuslims were constantly subjected to abuse and degradation from all Muslims, there was in fact a mix of two elements. Religious identity clearly played a role and, other things being equal, non-Muslims were inferior to Muslims, but other things were rarely equal. Proximity to power could elevate a non-Muslim to a status that could command more respect than a Muslim. For the few to achieve such high rank, access to the sultan was in fact the single most important attribute. Thus a proud and haughty Armenian Joseph Emin, as if a member of the sultan’s entourage, could command Turkish respect. However, that very same upper-class Armenian pride and haughtiness could arouse a very different reaction on the part of poor Armenians. They knew all too well that such uppity behavior ran the risk of a hostile response on the part of the lower Muslim orders. The sultan’s Armenian favorite himself could get away with it, but Muslim indignation and resentment would then be directed at his powerless coreligionists whose awareness of the consequences of such behavior caused them to lash out, defensively, against Emin. It was this very element of risk and precariousness of position that paradoxically led to the rise of Armenian financial advisors and bankers who served Ottoman officials in the eighteenth century in the first place (Barsoumian 1982). Contrary to some claims, Islamic law,

Religion as a Source for Compromise

187

particularly as it was interpreted in the Ottoman Empire, placed no significant restrictions on Muslim financial activity (see Braude 2007). It was not the supposed restrictions of Islamic law on Muslim financial transactions that gave Armenians an opportunity as court bankers; it was the shifting balance of power between the sultan and his high-ranking Muslim subjects who by the eighteenth century had so successfully entrenched themselves in different sectors of Ottoman life—witness the decline of the devshirme system—as to be almost immune from their ruler’s will. By contrast Armenians and other non-Muslims had no comparable independent base internally and proved far more loyal servants. However, once France, Great Britain, and the Romanov and Hapsburg empires started to offer non-Muslims their protection, they, too, could become almost comparably independent. The nexus of the priest’s millenarianism and Emin’s proto-nation-statism demonstrated in this passage is at the heart of religiously based collective consciousness. It is also to be seen in the Romaic—as opposed to Hellenic nationalist—impulse behind popular support for the Greek Revolution leading up to 1821. In that case deliverance was to come from a messiah-like blond-haired king (Skiotis 1978,). The priest’s dilatory prudence in the face of Emin’s filibustering (in its original sense) and his Enlightenment-linked radical appeal for equality and freedom should alert us to the resources that the religious tradition had at its disposal. St. Nerses was not only a twelft hcentury bishop, translator, scholar, theologian, and poet, he was also a prophet and the device by which to diff use a potentially explosive and dangerous situation. Emin could believe all he liked that he was sowing the seeds of true religion to yield a harvest of military spirit, but his religion was not true to its homeland and his harvest proved to be blighted. It was a supreme act of self-delusion for Emin to believe that he and the priest were talking the same language. These two different episodes from the eighteenth century should at the very least challenge the rampant conventional assumption that religious difference in and of itself creates deeply divided societies. On the contrary, in each instance traditional religious leadership and values acted as a source for compromise and acceptance that created a stable and long-lasting political system. The system may not have been democratic or just, but it survived. Of more significance for contemporary politics is the following: The Ottoman system lasted only to the first quarter of the twentieth century. The religious quietism that allowed a Muslim-dominated elite to rule for so long and, more generally, the sense of Islamic identity that gave coherence to a

188

Benjamin Braude

polity were overtaken by the Enlightenment, secularization, and the rise of nation-statism. So whatever lessons it may at one point have yielded have long since passed their sell-by date. Not so fast. The product has a longer shelf life than one might imagine. The product in question is the role of religion as a potential source of irenic political compromise. Again surprising as it may seem, that role is today most strikingly evident in the Safavid successor state, the Shiite Islamic Republic of Iran, while religion has lost that ability in the three Ottoman successor states, Greece, Turkey, and Iraq. The examples by which I propose to test this hypothesis, to recap what was presented at the outset, are (1) Greece and the Macedonian question, (2) Turkey and the Kurdish question, (3) Iraq and the Kurdish question, and (4) Islamic Iran and the Azeri question. All are structurally and historically remarkably similar, but the fourth yields a different outcome. First the structural. In the first three instances the respective state purports to be organized on linguistic national lines, claiming to deny the religious ties that had been so important a part of its heritage. In all instances the problematic community is a linguistic national minority whose bond with the dominant entity is in fact religious. In all instances that problematic community also has ties to an ethnically related state or potential state beyond the borders of the dominant nation-state that accordingly may or may not fear a secessionist movement. Herewith capsule summaries of the respective historical details. The Macedonian question first emerged in the late nineteenth century during the breakup of the Ottoman Empire as various regional and outside powers competed for hegemony in the disputed borderlands of what are today Albania, Bulgaria, Greece, Macedonia, Serbia, and Turkey. At the time the states involved were the Ottoman Empire and its former territories (Albania, Bulgaria, Montenegro, Serbia, Greece), as well as the Romanov and, less directly, the Habsburg empires. As Slavic speakers, the Macedonians were most closely related to the Bulgarians and politically, the Internal Macedonian Revolutionary Organization was uneasily allied to various Bulgarian governments. But the true identity of the Macedonians and their external alliances has been ever-changing. During the twentieth century peoples variously identified as Macedonians lived in practically all of the states of southeast Europe, with the largest concentrations in what had been southern Yugoslavia, Bulgaria, and Greece. After the Balkan war and World

Religion as a Source for Compromise

189

War I, the dispute diminished in intensity, but it continued to be an irritant in regional affairs, provoking assassinations and guerrilla warfare. During and after World War II the Macedonian question reemerged as a major international problem, poisoning relations between Yugoslavia, Bulgaria, and Greece, sparking the outbreak of the Cold War. Macedonians played an important part in the Greek Civil War (1944– 49), a key turning point in relations between the United States and the Soviet Union (Rossos 1997). Although that war has often been viewed through the perspective of the Cold War in ideological terms as a fight between communists against royalists and others allied with Great Britain and the United States, it also had a Macedonian dimension. At the height of the conflict in 1948, of the twenty thousand fighters orga nized into the opposition Democratic Army of Greece as many as fourteen thousand may have been Macedonians. In recognition of their role, the Greek communist party promised them selfdetermination. More recently the question has reemerged after the collapse of Yugoslavia and the transformation of its former province, the Socialist Republic of Macedonia, into independent Macedonia in 1991. Greece has attempted to block Macedonia’s use of that name principally on the grounds that it implies a claim on northern Greece, its own Macedonia. Although the European Union has rejected that reasoning, Greece has persisted in its objections, leading to various diplomatic absurdities, but in practical terms day-to-day relations between the two states have been peaceful. Lurking beneath the entire matter is the uneasy question of Greece’s own Slavic past and current population whose existence is denied by the Hellenic myth. For Turkey and Iraq, the historical details of the Kurdish question resemble those of the Macedonian problem for Greece. In both instances, a population dispersed over the states emerging from the Ottoman Empire laid claim to nationhood and territorial independence. In both instances, the very existence of that population as a separate and distinct entity has been denied by the host states; one notoriously dismissive formulation refers to them as “mountain Turks.” And the precise nature of the group has been difficult to define. One of the founders of Turkish nationalism, Ziya Gökalp, was the son of a Kurdish mother. Others of Kurdish origin have played significant roles in Turkish politics and culture, but they have done so largely by ignoring the complexity of their own identities. Whatever distinct identity each had was submerged in the overriding religious category to which

190

Benjamin Braude

most of them at least nominally belonged, respectively, Eastern Orthodox Christianity or Sunni Islam. After the Kurdistan proposed under the Treaty of Sevres in 1920 disappeared with the Treaty of Lausanne in 1923, Kurds in Turkey staged periodic uprisings over two decades that were finally suppressed on the eve of World War II (Harris 1977). The cause reemerged in the 1970s with the rise of the Kurdish Workers Party’s (Partiya Karkerên Kurdistan [PKK]) violent struggle in Turkey. Since the capture of the party’s leader, Abdullah Öcalan, in 1999 PKK activities have somewhat abated. However, the two Iraq wars have created an autonomous Kurdish enclave beyond Turkey’s control whose existence has periodically rekindled further military clashes. A major difference between the status of Kurds in Iraq as opposed to Turkey is that at no time has the Turkish Republic ever officially recognized their rights to any form of autonomy. Intermittently from 1919 to 1970, Iraqi Kurds fought against the British, the Hashemite, and the Republican regimes without gaining any significant concessions. However, in March 1970 the government of Iraq, largely in response to pressure from its major ally, the Soviet Union, agreed on paper to concessions for broader autonomy and Kurdish representation in government bodies. Whatever goodwill this might have fostered was soon undercut by a program of population resettlement of Arabs into the oil-rich areas that Kurds claimed. In 1975, the military support the Kurds had been receiving through Iran ended abruptly when, under the Algiers Pact, Iraq agreed to make major concessions to the Shah of Iran, particularly on the Shatt al-Arab border question. Thereafter Iraq began a major military campaign in the north that only grew more devastating with the outbreak of the Iran-Iraq War in 1980. The defeats of Iraq during the Gulf War of 1991 and the U.S. invasion of 2003 have permitted the emergence of the current autonomous Kurdish region in the north. Less well-known than the Macedonians and the Kurds and normally not compared to them are the Azeris. The simple explanation for this neglect is that they have, relatively speaking, presented the least contentious problem. The more cynical explanation is that their relatively tranquil status contradicts the dominant trope attributed by the United States and its allies to Islamic Iran as an expansionist state inherently threatening to its subjects and its neighbors. However, the structural similarity between the Azeris on the one hand and the Kurds and Macedonians on the other is almost as remarkable as their far greater social and political integration. All three are ethnic linguistic groups whose previous integration within the larger territorial

Religion as a Source for Compromise

191

community was originally largely achieved through religion. In the Ottoman Empire Kurdish leaders were as much a part of the Sunni Muslim elite as any other remote rural population could be, while Macedonians did have access to power through the devshirme system. In Iran, the Azeris were part of the ruling Shiite elite. It is notoriously difficult to estimate the subdominant populations of states in the Near East, but most observers put the Azeri population in Iran at around 25 percent, with some setting it as high as one-third, making it probably the largest of the three. Even more significant than numbers is the importance of individual Azeris in the Iranian power structure. After the Islamic revolution, one of Khumayni’s main rivals was Ayatollah Kazem Shariatmadari, who advocated a far more liberal and decentralized constitution. Having been implicated in a plot to overthrow the government in 1982, he was put under house arrest until his death in 1986. His disloyalty to the regime, however, did not automatically put other Azeri ayatollahs under suspicion. In fact the current Supreme Leader of the Islamic Republic is another Azeri, Ali Khamenei. And on a different note, the leading female pop singer of Iran over the past forty years has been Googoosh, an Azeri now in exile. Since the revolution no female soloist has been allowed to sing in public so for the past thirty years she has had little competition, but even without this Islamic backing, as it were, Googoosh’s talent has still been remarkable. And her appeal has been broad despite her willingness to sing to Azeri as well as Farsi concerns. However, this integration has its limits. In popular Iranian culture Azeris are the butt of politically incorrect humor since they are the Newfies, the country bumpkins. In 2006 Iran had its own cartoon scandal when a caricature appeared in a newspaper comparing them to cockroaches. The more common comparison is to donkeys. The structural similarities between the three groups become most evident in studying the events of the past sixty years, but the reasons for the significant differences between the Azeris and the others will emerge most clearly through recalling the history of the past six hundred years. In 1945, at about the same time as the Greek Civil War and the related Macedonian uprising, Jafar Pishevari proclaimed himself head of the Azerbaijan People’s Government, seceding from Iran. Unlike the Greek opposition, Pishevari had Stalin’s support, but his efforts were even less successful than theirs. Just a year later, backed by the United States and the British, the Iranian government reestablished its control in Tabriz and the rebel leaders either fled or were imprisoned. These wartime developments affecting the Azeris had

192

Benjamin Braude

significant implications for the Iranian Kurds, farther to the west. In 1941 a group of local Kurdish leaders had seized control of the town of Mahabad from the central government. The invasion of the country by British and Soviet forces and the consequent exile of the founding Pahlavi ruler, Reza Shah, had undermined central authority in Tehran, leaving a local power vacuum that the Mahabad Kurds rushed to fi ll. Soviet military occupation of the territory to the east blocked Iranian efforts to reassert central authority. Backed by modest Soviet aid, the Kurds maintained local autonomy without any pretensions to full independence. However, the proclamation of Pishevari’s People’s Republic upped the ante and prompted the Mahabad Kurds to announce their own People’s Republic, which they proclaimed was to be an autonomous region within the Iranian state. That survived for an even briefer period than the Azeri republic and it met with even more violent repression when Iran reasserted its control in December 1945. By this point the connections between the Azeris and the Kurds were not merely theoretical; they had become very practical. The reverberations of Mahabad’s fall were long-lasting. What little local military strength the Mahabad regime had was provided by Mustafa Barzani and his contingent of Iraqi Kurds. When the republic collapsed, some returned to Iraq where they were condemned to death, but others under Barzani fled to the Soviet Union, where they remained until 1958 when they, too, returned to Iraq. This return was more successful since it laid the foundation of the subsequent Kurdish revolt. Mustafa’s son, Massoud, now head of Iraqi Kurdistan, had been born in Mahabad. The flag of that republic remains a potent symbol in Kurdistan today. Just as there has been more than one Kurdistan and more than one Macedonia, so there has been more than one Azerbaijan. But while the multiple Kurdistans and Macedonias have aroused military and diplomatic anxieties among their respective Near Eastern and Balkan neighbors, the existence of two Azerbaijans today has had no such effect on Iran. And the difference between the Iranian reaction to the doubling of Azerbaijan and the Greek reaction to the doubling of Macedonia is significant. The immediate cause for this duplication of names was the end of the Cold War. The breakup of Yugoslavia turned a Macedonian province, which the Greeks could accept, into a Macedonian independent state, which they could not. The breakup of the USSR turned a province, the Soviet Republic of Azerbaijan, to the north of the longstanding Iranian frontier into an independent

Religion as a Source for Compromise

193

Azeri state. In striking contrast to Greece, Iran accepted this new state with equanimity. The serious military and political challenges that Greece and Iran had faced after World War II with their respective rebellious communities would be reason enough for both to be alarmed, but only one has been. The historical background to Azeri duality needs explanation. In the early nineteenth century czarist Russia repeatedly defeated Persia of the Qajars, the dynasty that had succeeded the Safavids. By 1828 the Romanovs had annexed all of the Caucasus including what is now independent Azerbaijan, leaving only the half of that region south of the Caspian Sea under Qajar control. Previously both halves had long been part of the Persian realm. Both Azerbaijans originally had the same cultural, linguistic, ethnic, and religious identity, but nearly two centuries of separation under radically different regimes had created some important divides. Nonetheless, during World War II and the Cold War, the Soviets had tried to exploit their control of the north to influence and control the south, without terribly much success, as the fall of the Pishevari People’s Government demonstrated. The very different demographic realities faced by the Ottomans and the Safavids during their rise to power explain the differences between the Azeri as opposed to the Macedonian and Kurdish questions today. Both empires integrated religion and religious institutions into their administrative structures, but they did so in different ways. The Ottomans could not make Islam the be-all and end-all of their identity and still negotiate successfully with their majority Christian subjects. So they had to delicately compromise between the two. For the Safavids creating Shiite Islamic unity was the only way to create and maintain their realm. Had the Ottomans been able to adopt a version of Safavid religious statecraft, it is conceivable that both Turkey and Iraq would be treating the Kurds today with the same relative success as Iran deals with its Azeris. On the other hand, had the Ottoman Empire attempted to be a militantly conversionist Sunni Muslim state (on the Safavid model), their Christian subjects would have long since overthrown them and they would have long since disappeared. Had the Greek Orthodox Church been less Greek and more Orthodox, more welcoming of the Slavic and other forms of cultural and linguistic diversity within its communion, it, too, might have been more unified and militant in dealing with its Muslim overlords. Under those circumstances, Macedonia for Greece would have been a matter no more difficult than Azerbaijan has been for Iran. But

194

Benjamin Braude

in contrast to the Supreme Leader of Iran, the Patriarch of Constantinople never spoke Macedonian or Serbian, his equivalent of Azeri. These Orthodox divisions made the Ottoman ability to divide and rule through the devshirme much easier and gave the later claims to political independence by Macedonians and other Slavs greater historical precedent, even earlier than the establishment of their autocephalic churches. In the post-imperial period, Turkey, Persia, and Iraq worked out the implications of their respective legacies. The most European and Christian of the Muslim empires adopted the most secular and ethnocentric version of nation-statism. Mustafa Kemal succeeded in effecting the Turkish revolution, the most significant and successful political and cultural transformation in world history since the French Revolution of 1789—both because of what the Ottomans bequeathed him and because of his own personal stature. He exploited his position as a genuine hero, one of the few successful military leaders on any side to emerge from World War I, to establish a radical agenda. Reza Khan tried to imitate him, but the Safavid- Qajar legacy blocked his efforts in many ways, and he lacked Kemal’s hero credentials to overcome it. The Pahlavis did push their pre-Islamic legacy in different ways and they toyed with the Indo-European Aryan origins of the Persians, aping Nazi-inspired racial ideas during the 1930s. Accordingly Reza attempted to restrict Azeri language and culture, one of the reasons that the cause of Azeri independence gained some support a few years later. However, the Pahlavi period was a brief detour from the main trends of Shia cultural dominance, first established by the Safavids more than five hundred years ago. Under the Islamic Republic, despite occasional tensions between the two groups, no systematic suppression of the Azeris has occurred. The Kurds, however, were the odd men out in all three states. Since they are overwhelmingly Sunni, they are outside the Safavid legacy of Shia consensus in Iran. Their Sunni Islam has done them little good in Kemalist Turkey or Baathist Iraq since secular nationalism has been the proclaimed ideology for both. The policies of the Islamic Republic toward the Kurds have been nearly as repressive as those of Turkey, though Iran has traditionally been willing to harness them as sticks to beat neighboring Iraq when it suits their interests. Given the dominance of secularism and nationalism in Turkey and Iraq for most of the past century, the possibility of wielding Islamic unity as a means of integrating their Kurdish populations seems unlikely. The last political leader to attempt that was Abdülhamid II and he, as

Religion as a Source for Compromise

195

a result, helped instigate the Armenian problem and was overthrown for his efforts. However, Islamic Iran’s relative success with the Azeris does suggest that such a strategy should not be dismissed out of hand. Notes 1. See the argument on the meaning of ghaza in my review of Lindner 1983 (Braude 1987). 2. Byzance aprés Byzance (1935). Parts of Iorga’s argument had been anticipated in Gibbons 1916. 3. Worth noting is the long reluctance of the Greeks themselves to publish a modern Greek version of this important text. The fi rst printed edition of the manuscript was Critoboulos 1870, in the series Monumenta Hungariae historica; next appeared a Turkish translation (Kritovoulos 1912); English was the third (Critoboulos 1954); and Rumanian fourth (Kritovoulos 1963). The Greek was reprinted two decades later, Critobuli Imbriotae historiae (1983), and a German translation by the same scholar followed, Mehmet II. erobert Konstantinopel: Die ersten Regierungsjahre des Sultans Mehmet Fatih, des Eroberers von Konstantinopel 1453: Das Geschichtswerk des Kritobulos von Imbros (ca. 1986). Finally, Vyzantiou halōsis: Xyngraphē historiōn . . . prologos eisagōgiko meletēma gia ton M. Kritovoulo (1999); and Vios tou Mōameth II, Historia, Kritovoulou tou Imvriou (2005). 4. Didaskalia Patriki (Constantinople, 1798), excerpted and quoted in Clogg 1976, 59 and extensively discussed and translated in full in Clogg 1969. Clogg acknowledges but effectively rebuts questions that have been raised about the true authorship of this pamphlet. 5. Emin 1918, 139– 46 (1st ed., London, 1792; fi rst Armenian translation by H. Khashmanean [Beirut: Tparan Mshak, 1958]). My attention was drawn to this passage by Suny 1993, 55–56. References Abisaab, Rula Jurdi. 2004. Converting Persia: Religion and Power in the Safavid Empire. London and New York: I. B. Tauris. Barsoumian, Hagop. 1982. “The Dual Role of the Armenian Amira Class with the Ottoman Government and the Armenian Millet (l750–l850).” In Braude and Lewis, Christians and Jews in the Ottoman Empire, 2:171–84. Braude, Benjamin. 1987. Review of Rudi Paul Lindner, Nomads and Ottomans in Medieval Anatolia. Speculum 62: 701–3. ———. 2007. “Christians, Jews, and the Myth of Turkish Commercial Incompetence.” Relazioni economiche tra Europa e mondo islamico: Secc. XIII–XVIII. Fondazione Istituto Internazionale Di Storia Economica “F. Datini” Prato, Serie II, Atti delle “Settiman di Studi” e altri Convegno 38, ed. Simonetta Cavociocchi, 219–39. Prato: Le Monnier

196

Benjamin Braude

———. 2011. “How Racism Arose in Europe and Why It Did Not in the Near East.” In Manfred Berg and Simon Wendt, eds., Racism in the Modern World, Historical Perspectives on Cultural Transfer and Adaptation, New York: Berghahn Books, 41–64. Braude, Benjamin, and Bernard Lewis, eds. 1982. Christians and Jews in the Ottoman Empire: The Functioning of a Plural Society. 2 vols. New York: Holmes and Meier. Clogg, Richard. 1969. “The ’Dhidaskalia Patriki’ (1798): An Orthodox Reaction to French Revolutionary Propaganda.” Middle Eastern Studies 5: 87–115. ———, ed. and trans. 1976. The Movement for Greek Independence, 1770–1821. London: Macmillan. Emin, Joseph. 1918. Life and Adventures of Emin Joseph Emin, Written by Himself. 2nd ed. Ed. Amy Apcar. Calcutta: Baptist Mission Press. Gibbons, Herbert A. 1916. The Foundation of the Ottoman Empire: A History of the Osmanlis Up to the Death of Bayezid I. Oxford: Clarendon Press. Harris, George S. 1977. “Ethnic Conflict and the Kurds.” Ethnic Conflict in the World Today. Annals of the American Academy of Political and Social Science 433 (September): 112–24. Iorga, Nicolae. 1935. Byzance aprés Byzance. Bucharest. l’Institut d’études byzantines. Kafadar, Cemal. 1995. Between Two Worlds: The Construction of the Ottoman State. Berkeley: University of California Press. [Kritovoulos] Critoboulos. 1870. Vios tou Mōameth II, Vie de Mahomet II. Trans. Philipp Anton Dethier. Budapest: Magyar Tudományos Akademia. ———. [Kritovoulos]. 1912. Tarih-i Sultan Mehmet Han-i Sani. Trans. Karolidi. Istanbul: Ahmet İhsan ve Şürekâsı. ———. Critoboulos 1954. History of Mehmed the Conqueror. Trans. Charles T. Riggs. Princeton, NJ: Princeton University Press. ———. [Kritovoulos]. 1963. Din domnia lui Mahomed al II-lea, anii 1451–1467 [de] Critobul din Imbros. Ed. and trans. Vasile Grecu. Bucharest: Editura Academiei Republicii Populare Romîne. ———. 1983. Critobuli Imbriotae historiae. Ed. Diether Roderich Reinsch. Berlin: W. de Gruyter. ———. 1986. Mehmet II. erobert Konstantinopel: Die ersten Regierungsjahre des Sultans Mehmet Fatih, des Eroberers von Konstantinopel 1453: Das Geschichtswerk des Kritobulos von Imbros. Trans. and ed. Diether Roderich Reinsch. Graz: Styria. ———. 1999. Vyzantiou halōsis: Xyngraphē historiōn . . . prologos eisagōgiko meletēma gia ton M. Kritovoulo. Ed. and trans. Phanēs Kalaitzakēs. Athens: Dēmiourgia. ———. 2005. Vios tou Mōameth II, Historia, Kritovoulou tou Imvriou. Ed. and trans. Roderich Reinsch Diether and Phōteinē Kolovou. Athens: Ekdoseis Kanakē. Leal, Karen Alexandra. 2003. “The Ottoman State and the Greek Orthodox of Istanbul: Sovereignty and Identity at the Turn of the Eighteenth Century.” Ph.D. diss., Harvard University. Lindner, Rudi. 1983. Nomads and Ottomans in Medieval Anatolia. Bloomington: Research Institute for Inner Asian Studies, Indiana University.

Religion as a Source for Compromise

197

Rossos, Andrew. 1997. “Incompatible Allies: Greek Communism and Macedonian Nationalism in the Civil War in Greece, 1943–1949.” Journal of Modern History 69 (March): 42–76. Skiotis, Dennis N. 1978. “The Romaic Ideal in the Greek Movement for Independence.” In Speros Vryonis Jr., ed., The “Past” in Medieval and Modern Greek Culture, 155–62. Malibu, CA: Undena Publications. Suny, Ronald G. 1993. Looking Toward Ararat: Armenia in Modern History. Bloomington: Indiana University Press. Wittek, Paul. 1938. The Rise of the Ottoman Empire. London: Royal Asiatic Society.

CHAPTER 7

Geopolitics and the Long-Term Construction of Democracy Randall Collins

To consider the future of democracy in Iraq, Pakistan, Russia, and other troubled places requires a theory of its social conditions. And in fact, the future of democracy anywhere cannot be taken for granted. Democracy is not an all-or-nothing condition: either you have it or you do not. Varying degrees of democracy have been created over the years and have declined as well. The United States, the United Kingdom, and the European Union will not necessarily be as democratic in the future as they are now. The degree of democracy needs explaining using comparative-historical theory. Some societies are said to have difficulty with democracy because they lack a democratic tradition. If this is our explanatory principle, it is a pessimistic one, presuming that societies are doomed to repeat the past. But the societies that we consider historically to be exemplars of democracy— England, France, the United States—have not always had what we would today consider democracy; their traditions, too, had to be created. And societies that we may think of as lacking a democratic tradition, such as Japan, Germany, and India, have been well-functioning democracies since the late 1940s or 1950s. I will argue that a democratic tradition, or democratic culture, is a subsidiary matter and that the conditions that promote or destroy democracy are in the structure and sequence of ongoing events. It often appears that conflict is destructive of democracy: whether the internal struggle of factions or the external geopolitics of power seeking and war. Nevertheless, conflict is the essence of politics, and war-making capacity is at the center of the state. Conflict can destroy democracy, but it also

Geopolitics

199

created it. We must look for particular configurations of conflict, those that create a balance among conflicting forces, the balance that we call democracy. To use a somewhat shaky metaphor, democracy is like riding a bicycle: in order to avoid falling over you have to be in motion, you have to feel how to shift your weight from side to side, and furthermore there is a moment when you suddenly find that you can ride. The bicycle metaphor is decidedly inadequate in the following respect: democracy is multidimensional, and we need separate theories to explain changes on each dimension. There are two major ones: first, the degree of collegially shared power, and second, the extent of the franchise. A third, subsidiary dimension is political rights. It is subsidiary because it tends to follow in time from the two main dimensions and is largely caused by them. The first dimension, collegial power, is an institutionalized structure of power sharing. It is power shared among colleagues, co-participants in decision making. The word “college,” which for Anglophones means an institution of higher education, goes back to a medieval organization in which the guild of teachers elected their administrative head. Medieval Europe was full of collegial structures: councils, assemblies, legislatures, electoral bodies, independent judiciaries; the origin of the word “president” was someone who presided at just such an assembly. Collegial power can pyramid upward at several levels: not only are there collegial bodies that decide collectively but there are concatenations of such collegial bodies that balance and contend with each other. And collegial power-sharing structures can be created higher up, adding further layers at the top, when federations and coalitions are formed. It is not easy to put this in quantitative form. We must visualize an abstract continuum ranging from zero collegial power—an ideal autocracy in which a single center makes all decisions—up through a potentially infinite number of such power-sharing structures. The second dimension, the extent of the franchise, is much simpler because we can put it in quantitative form. This is the percentage of people who have the right to participate in politics—to vote and to hold office. The ideal low point of this continuum would be a hereditary autocracy; slightly higher, a hereditary aristocracy. The franchise, although sometimes very restricted, is nevertheless important because it puts in place a structure that is capable of expansion. In the medieval German Empire, the emperor was not hereditary but was elected by eight heads of secular and ecclesiastical states, called Electors. The College of Cardinals, which elects the pope, has ranged in membership historically from 12 to 87; recent popes have expanded

200

Randall Collins

it to over 250. In the fifteenth century, the Conciliar movement attempted to make the Church more democratic by making the pope elective from a permanent general council, which furthermore would have ongoing legislative power. Catholic democracy failed, but the model was developed by Protestant churches such as the Presbyterians, which elect their own ministers. There is no absolute cutoff point in the size of the franchise between democracy and nondemocracy. We are inclined to think that democracy means universal franchise or, as used to be said, one man, one vote; but in fact the franchise has never been universal. Even today the franchise excludes young people below a certain age, as well as convicted felons and others. We may think this is just common sense—that children are not capable of voting responsibly—but in fact common sense is historically shifting. The same argument was made in England and elsewhere up through the nineteenth century, that only property-owning householders should vote, since everyone else—wives, servants, the younger generation of all ages—were not full members of society and were represented through the household head. Before 1920, even the most self-consciously universal democracy gave the franchise to 50 percent of the population at most. The exclusion of women from the vote was not taken very seriously by democratic theory, nor was the exclusion of American blacks, both in the period of slavery and in the period of segregation down to about 1970. France did not give the vote to women until 1946, Switzerland not until the 1970s. Conventional intuition here is partly correct; nineteenth-century states with a 40 or 50 percent franchise were democratic to some degree. That also means the structure of democracy goes much further back into the small but palpable franchise that we are prone to call oligarchy. We need two theories, then: one that explains the expansion—and sometimes contraction—in the extent of collegial power-sharing institutions and one that explains the rise and fall of the franchise. The second theory is nested in the first. The franchise can only be exercised if there is some collegial body whose members have it. The meaningfulness of the franchise depends on the extent of collegial power in which political participants can participate. Consider an ideal-typical combination of extremely high franchise with extremely low collegial power sharing: an absolute dictator, who is nevertheless elected by universal voting. This would be the situation if every member of the Catholic Church were allowed to vote for the pope. The symmetrically opposite combination is where collegial power sharing is high but the franchise is very restricted. This is the oligarchic republic,

Geopolitics

201

of which medieval Europe was full of examples. Because of our historical tendency to appropriate the term “democracy” for struggles to expand the franchise, we tend not to think of the Venetian Republic or the German selfruling free cities as democratic; our rhetorical tradition has been to consider them bastions of reactionary privilege. Nevertheless, these are the structures into which the expanding franchise was poured. Democracy, in its dimension as participatory franchise, is like pouring water into a container; if there are no containers to pour it into, it cannot be successfully poured at all. The third, subsidiary dimension of democracy is political rights. Freedom of speech, of the press, and of assembly, freedom from arbitrary arrest and punishment outside the law—these are freedoms that allow communication and political mobilization. The idea of these rights was created during the struggles over the structures of democracy; they are pragmatic slogans that have become elevated into abstract ideals. Democratic values are not themselves the primary causes of the degree of power sharing and extent of the franchise. The causes of these structural changes for the most part operate independently of whether the people involved intend to implement democratic rights. This is a good thing. People do not already have to be democrats in order to produce democracy. The point is illustrated by how federations institute a power-sharing structure. A federation can be made up of various kinds of units, which may themselves be internally quite undemocratic. In the era of the household property franchise, the paterfamilias could be an absolute despot over his family and servants (examples here come both from nineteenth-century England and ancient Rome), yet the domestic despot could take part in a republic of power-sharing institutions. Similarly, a group of hereditary monarchs who form a federation nevertheless limit their arbitrary power vis-à-vis their federal colleagues. Diplomatic alliances thus contain the germ of powersharing structures, laying the ground for expansion; this can occur even if their participants have no commitment to democratic values—even if they are antagonistic to democracy. If democracy is a structure rather than a belief, it has the potential to come to life in the strangest places. This is because power-sharing structures are the most fundamental of the three dimensions of democracy. Most theoretical effort in recent years has been devoted to giving conditions for the second and third dimensions of democracy. We have survey results that cause us to question the extent of commitment to democratic values, even in our own country. I do not think this is crucial. Deeper analytical

202

Randall Collins

attention has been paid to the causes of democracy understood as a wide franchise; a prominent theory is that democracy is caused by industrialism and economic growth. But this correlation is an historical artifact; the leading nation-states of modern democracy and modern industrial capitalism, England and the United States, were also those that had a long premodern background of collegial power-sharing institutions—the containers into which the expanding franchise could be poured. Some wealthy states in their eras—France between 1815 and 1875, Germany in the Nazi period— reversed a previously wide franchise. In fact the Weimar Republic in 1919 had the widest franchise in the world. Wealth does not protect against fluctuations. The growing wealth of China today does not automatically make it likely to become democratic. This is not to deny that for some historical periods there is a moderate empirical correlation between economic wealth and the second dimension, the size of the political franchise. This comes about because wealth is one of the resources for mobilizing groups into political movements. But another cause of mobilized political participation is the penetration of the modern state into society, initially in search of soldiers and taxes, which thus has an effect on the expansion of the franchise. Economic growth is thus not the only reason the franchise expands. It sometimes helps; alone, it is not sufficient. Let us shift attention now to the fundamental dimension that facilitates all the aspects of democracy: collegial power-sharing structures. The classic statements are those of Montesquieu and Tocqueville. The original powersharing institutions they extolled were medieval, not modern; on this dimension, modern industrialism is not necessary for collegial democracy, and Montesquieu and Tocqueville worried that modernity might destroy collegial democracy by replacing it with bureaucratic centralization. Some comparativists have argued that the societies that best retained their medieval institutions—notably England and its colonial offshoots—thereby became the archetypal modern democracies. Medieval Europe was historically the basis of collegial institutions, but this does not mean that societies lacking this medieval background are necessarily doomed. We must look at the medieval period analytically for the larger implications of its structural patterns. The extent of power sharing is an inherently sequential causal process; the crucial sequence is in the struggles that political actors go through in search of power. They do not have to be idealistic about this. Under the right sequence of events, the most hard-nosed, realist form of power seeking can lay down institutions that become the containers of democracy.

Geopolitics

203

Schematically, there are two steps: first, power sharing arises when a state—a militarily sovereign political unit—is weakened by geopolitical causes and is forced to enter into a power-sharing arrangement with allies or supporters; second, the power-sharing coalition becomes institutionalized because it is successful in meeting the geopolitical challenge, especially when it goes on to expand. Both steps depend on geopolitical conditions, that is to say, the rising or falling power-prestige of the state in the interstate arena based on its military success or failure. The fatefulness of military conditions goes deep into the structure of the state. The core of the state is its military, together with the organizational apparatus for extracting taxes, manpower, and other military resources. Historically, modern states grew out of military beginnings; the military revolution of increasing size and expense from the sixteenth century onward brought about modern tax extraction, bureaucracy, and state penetration and thus formed the apparatus that the modern state has turned to nonmilitary purposes as well. The great political revolutions began with a crisis of the state budget because of military strains and expenses. Revolutions today are still largely caused by geopolitical conditions, as in the collapse of the Soviet empire, which was set off by the unsustainable expenses of Soviet military overextension. Max Weber famously described the state as a structure monopolizing legitimate force upon a territory; less famously, he noted that rulers seek power-prestige in the interstate arena because it defines a state’s collective identity and the ruler’s own prestige. If we treat Weber’s statements as referring to variables rather than static ideal types, we can say that how much monopoly, legitimacy, force, and territory a state has are all historically variable and that geopolitics is a key determinant of their variation. For contemporary sensibilities, this may seem an overly militaristic view of the state, but notice how national anthems refer to “the rockets’ red glare, the bombs bursting in air,” “Allons enfants de la patrie,” and “Rule, Britannia! Britannia rule the waves.” The degree of legitimacy not only of rulers but of state institutions is thus powerfully affected by geopolitical success or failure. This applies to democracy as well. Thus democracy depends on, first, a nondemocratic ruler falling into trouble militarily, delegitimating the existing form of rule; second, the power-sharing coalition that takes over must go on to military victory, thereby legitimating democracy. Democracy to a large extent is determined from the outside in. I will illustrate the argument with brief sketches of the geopolitical sequences that

204

Randall Collins

produced a number of historic democracies, as well as a few that failed for lack of the right geopolitical conditions. Let’s first look at the medieval Republic of Venice, which lasted from the eleventh century until it was conquered by Napoleon in 1797. We don’t usually include Venice in our folk wisdom about democracy, in part because of our focus on England and France, in part because Venice restricted voting rights to a small number of elite families, at most 7 percent of the population. Though weak on the second dimension of democracy, Venice was very strong on the first dimension. Venice’s structure was designed to guarantee power sharing and prevent family oligarchy; it funneled voting through a series of indirect elections and split power among a series of councils and rotating magistrates. The electoral college by which the United States elects its presidents was copied from Venice; other Venetian innovations included the term “ballot” itself (i.e., voting by placing balls into urns [ballotti]), judicial rights, and the public defender to represent poor litigants. The doctrine of the separation and balance of powers came from Venetian institutions, although in taking over both ideology and structure Americans have forgotten its origins. There was no city-state of Venice in the tenth century, only a scatter of communities along the island waterways. The incentive for unifying was geopolitics—the opportunities for military expansion and state-protected trade in the geopolitical vacuum of the Adriatic and the eastern Mediterranean. Venice’s early weakness, like that of all the Italian city-states, was the tendency for anarchic conflict among rival families seeking to impose autocracy and, in the period of commercial expansion, class conflict from below. Such conflicts overthrew medieval democracy in Florence, Milan, and other Italian city-states. Venice escaped that fate because of its especially favorable geopolitical situation. Venice became a republic in the first place because no family was strong enough to impose itself against its rivals; this local geopolitical weakness was institutionalized in power-sharing structures, which in turn became legitimated by the geopolitical success of the Venetian Republic. It stayed clear of warfare in Italy and on the European continent, with its unstable conquests and tangled wars of German emperor versus pope, and stayed clear as well later on of the incursions into Italy by the French and Spanish monarchies. Venice found its geopolitical niche in expanding by sea into the fragmenting Byzantine Empire. The populace of Venice, most of them disenfranchised by the elite power-sharing oligarchy, nevertheless maintained their loyalty to the Republic because it worked.

Geopolitics

205

External power-prestige bolstered the legitimacy of the republican structure while providing wealth to keep the popular classes satisfied. The doge, as head of the government, directed military and foreign policy; the elaborate checks and balances to prevent tyranny did not hamstring government because these operated locally, leaving the state essentially as an overseas operation in whose success all Venetians had a stake. For six hundred years geopolitical success kept the republic alive. In contrast, on the other side of the Italian peninsula, republics fell prey to military conquest states (mediumsize fish eaten up by the big fish), which constituted the state-building process of land-based empires. It has sometimes been argued that maritime regimes are inherently prone to democracy, but this confounds the issue: the maritime states that become long-term democracies are those that go through the favorable geopolitical sequence. Similarly, the argument that middle-class, merchant states are inherently democratic overlooks the fact that a democracy must be geopolitically viable: small city-states surrounded by militarily expanding national states do not survive, and they make themselves all the more vulnerable by the turmoil of class struggle between rich and poor. Geopolitical success favors democratic institutionalization: both directly, by keeping up powerprestige and therefore legitimacy, and indirectly, by dampening internal class conflict. I have argued that institutionalizing power sharing depends upon a sequence: first, local weakness, which motivates power sharing; and second, geopolitical success abroad, which institutionalizes it. There are two major alternatives to this sequence. One is that states never go through the first stage of sharing power in collegial structures; this may happen because an autocratic state is quite successful in military expansion and has no need to concede power sharing. This was essentially the Russian route from the fifteenth through the nineteenth centuries. The other is that a state never gets the chance to try power sharing in its moments of crisis because it is swallowed up by a nearby conquest state, as happened to so many small European states, winnowing down from nearly a thousand to a few dozen between 1100 and 1900 as Charles Tilly showed. The first step in the democratic sequence is a rather delicate one: the state must be weakened so as to be forced to make concessions to power sharing but not so weakened that it does not survive. When the English barons in the thirteenth century forced the monarch to concede a parliamentary structure controlling the purse-strings of government, the English state

206

Randall Collins

nevertheless remained safe from external enemies due to its sea barrier to invasion and the chaotic condition of military regimes on the Continent. A more extreme case is Poland, which in 1572 institutionalized a weak electoral monarchy subject to a strong aristocratic parliament; here collegial power sharing was so jealously guarded that the king was given scant resources for his army, never going through the tax extraction and state penetration that characterized modern states, such as their neighbor Prussia. Poland’s weakness in military resources, together with its inherently weak geopolitical position in the midst of strong states on each side, caused it to disappear as a sovereign state in the conquests of the eighteenth century. Establishing power-sharing democracy is no guarantee that it will last; that depends on its geopolitical success. One might be tempted to infer from this argument that only big states can be successful democracies. In the long run, small states tend to get swallowed up by bigger ones; this is one of the basic principles of geopolitics; states with more military resources tend to expand their spheres of influence at the expense of states with fewer resources. Moreover, if the legitimacy of state institutions depends upon power-prestige in the interstate arena, wouldn’t small states, even if they survive, have little legitimacy among their citizens since they have no prominence in the world arena? The example of Switzerland helps us answer this question. The Swiss Federation developed within a favored geopolitical niche. It was one of several leagues that grew out of the fragmenting German Empire and was the only one to prosper and institutionalize as a republic. The Swiss mountain cantons and the surrounding ring of cities below the trade route passes of the Alps were at the geographical intersection of cross-cutting powers. As strong states grew up on several sides, notably France and Austria, their counterbalancing effects helped keep this buffer zone independent. In the thirteenth century, the threat of a Habsburg dynastic state expanding from southern Germany led to the formation of a small defensive confederation. Battle victories over the Habsburg and then against Burgundy started a bandwagon effect and gave prestige to a newfound Swiss patriotism and its slogan of liberty. More cantons joined the league, and by the fifteenth century the federal government was making its own conquests and incorporating surrounding petty states. The Swiss Federation promoted republican institutions from the top down; externally it kept the cantons and cities from being swallowed up by dynastic conquest states; and inter-

Geopolitics

207

nally it prevented its leading members from exercising authoritarian rule over the others. It is not surprising that during the Reformation the most radically populist forms of church government were in Swiss cities belonging to the federation. In its early centuries, the Swiss Federation enjoyed a string of military victories. Later, as the big states consolidated and vastly increased their military resources through state penetration, Switzerland became a very minor power, surviving chiefly by a purely defensive strategy and a policy of neutrality. With its military glory so long in the past, how did its democracy keep its prestige? The lesson seems to be that defeats are more dangerous to legitimacy than lack of victories; the mythmaking apparatus of government works fairly well on the materials of the distant past if they are not too strongly contradicted by the experience of the present. Switzerland in recent centuries has played sensible geopolitics, avoiding the dangers of military overextension that are so often a prelude to state crisis. I will discuss one more successful outgrowth of medieval collegial structures: the Dutch Republic. In this case, the region crystallized a national or cultural identity only during the period when it formed its political identity; prior to that time, it had no clear-cut ethnic or linguistic boundaries. The Low Countries were initially part of the German Empire, with its characteristic mix of self-governing cities, leagues of petty aristocrats, and dynastic principalities. The Netherlands was thick with cities of the Hanseatic League, which extended as far as the Baltic. The great geopolitical event of the later Middle Ages was the disintegration of the German Empire. Fragments of the Empire, such as Switzerland, had the opportunity to benefit from this geopolitical weakness, provided they could avoid being swallowed up by one of the expanding modernizing states. The Dutch misfortune was also their advantage: being fought over repeatedly by France, Burgundy, and the great power of the time, the Habsburg dynasty, with its properties ranging from Austria to Spain. The “cockpit of Europe” became just such a nexus of intersecting powers that promotes local autonomy. In effect the rival big states pushed each other out of the Netherlands. This autonomy was reinforced and ideologically inflamed by the adoption of republican forms of Protestant church governance. It helped, too, that merchant shipping and incipient capitalism generated wealth that could be used for Dutch military resources. A collection of local collegial institutions and military defense leagues confederated into the States-General, made up of deputies from

208

Randall Collins

seven republics. Like Switzerland and the United States, it was a republic of republics. The Dutch Republic was created in the revolt against Spanish Habsburg rule in 1568–84. This ought to be considered one of the great democratic revolutions, perhaps the first of the chain. It is not so considered, in part because of the tunnel vision of our Anglocentric and Francocentric political tradition and perhaps because, as the heir to big-state democracies, we only pay attention to the revolutions of big states. Yet the Dutch revolution was a model for the English revolutions of 1640 and 1688 and supplied the troops and the constitutional monarch for the latter. Like other successful republics, the Dutch Republic benefited first from the geopolitical weakness of its origins and then from its own geopolitical expansion. Its victory over the Spaniards came against a large but overextended enemy, stretched in wars across multiple fronts throughout Europe and overseas, operating at the end of very long supply lines, and reeling from financial strains of its wars. Once established, the Dutch Republic became institutionalized through the power-prestige flowing from its success. In Wallersteinian world-system theory, the Dutch are considered to be the hegemonic power of the seventeenth century. Its sea-based empire and commercial wealth came at just the right time to keep the young republic alive, in much the same way that the Republic of Venice was undergirded by its geopolitical success. Federations and other power-sharing structures were widespread during the European Middle Ages. Most of them failed, and we can draw a lesson from their failure. Federations or leagues, as well as free self-governing cities, were especially abundant in the German Empire, which, we should recall, at one time stretched down into Italy. Some of these power-sharing structures failed because of an internal drift to autocracy, usually because an autocrat arose to quell class conflict; but mainly they failed because they were not viable in the external geopolitical arena and were swallowed up. Many leagues or protective alliances grew up in the fragmenting German Empire to put down brigands, to control the disruptions of petty warfare, and to protect trade. But these advantages of providing internal peace were not enough to prevent virtually all of these leagues from disappearing into the maw of bigger states. The Swiss Federation and the Dutch Republic are virtually the only survivors of medieval federations, bolstered by their favorable geopolitical niches.

Geopolitics

209

The biggest failure of all took place in the largest structures. The entire German Empire, in the years between 1485 and 1520, made a move toward becoming one large federation, what I would call the Reichstag republic. You will recall that the German emperor was elected, albeit by an extremely small franchise of eight Electors. There was also a grand assembly of ambassadors from all the component parts of the empire: the dynastic states, the petty knights, free cities, and ecclesiastical principalities. This was the Reichstag, or diet. Its chief function was to vote subsidies for the emperor, who acted as the military arm of the empire. In the fifteenth century, as the emperor grew weaker and weaker, a movement grew within the Reichstag to increase its own power: it wanted to meet annually and form a permanent Imperial Council to control the army and approve all acts of the emperor. Toward the end of the century, Emperor Maximilian I was willing to consider these reforms, especially since the Reichstag proposed to levy taxes throughout its domains that would be collected centrally and be put at the disposal of the Imperial Council and the imperial army. In effect this was an alliance between a strong executive and a far-flung constituency of small powers against the upper-middle tier of expanding dynastic states. If the Reichstag Republic had succeeded, it would have been the great European republic, eclipsing Venice and the Swiss Federation. Indeed, it was because of the failure of the Reichstag Republic to consolidate that the Swiss Federation finally pulled out of the German Empire and went its own way. Why did it fail? It is a long, interesting, and forgotten story. In briefest summary, the Reichstag republic was never able to get off the ground as a geopolitically successful state. It was hampered by two of its potential allies, which were also rival structural possibilities. One was the emperor, usually elected from among the strongest dynastic monarchs. The Habsburg monarchs, who at this time usually served as emperor as well, had the choice between promoting their far-flung personal inheritances—many of which lay outside the German Empire and had been acquired by marriage or conquest—and acting to strengthen the empire. At a moment when his dynastic military fortunes had taken a turn for the worse, the Hapsburg ruler opted to encourage the Reichstag reforms—much in the way that the weaker English kings made concessions to the barons and to parliamentary institutions controlling the budget. The other rival was the pope. The papacy, too, was an elected office, and during the thirteenth through the fifteenth centuries there were often rival

210

Randall Collins

political factions in papal politics, drawing support from various kingdoms throughout Europe. Ecclesiastical powers were especially strong inside the German Empire—three of the eight Electors were archbishops, and it was these archbishops who took the lead in the movement for a Reichstag Republic. There had been a movement, too, inside the Church to turn the pope into the president of an ecclesiastical republic: this was the Conciliar movement, which in the fifteenth century had attempted to hold regular Church Councils, which would control church revenue and take over legislative power from the pope. If the Conciliar movement had succeeded, it would have preempted the Protestant Reformation, in effect establishing church democracy on a grand scale. But the Conciliar movement fell apart because it factionalized along national lines. In effect, both the Reichstag Republic and the Catholic republic failed at the same time and for similar reasons: both were made up of the same set of actors who tended to intrude in each other’s affairs—the empire tried to keep the church weak, and the church tried to keep the empire weak. Both succeeded. The result was a downsizing of political ambition on both sides: to the medium-size secular states and the localizing of the church in the Reformation. These forgotten examples of medieval power-sharing institutions are pertinent to what we think of as the archetypal modern democracy, the United States. The United States formed in much the same way as the Swiss and Dutch republics. The independence of the Atlantic coastal colonies of North America originated in geopolitical strains of the British Empire at a time when it was engaged in conflicts on multiple fronts. The British were at a further disadvantage in logistics costs in long-distance fighting against the settlers of their most populous and resource-rich colonies (i.e., who could put the most troops on the spot). The beginning of the American Revolution resembles the beginning of the revolutions in England and France: rejection of demands from the state to pay the costs of previous wars—in this case the conquest of Canada. But this time the result was not revolution inside the existing state—the American colonists did not succeed in increasing their democratic representation inside the English Parliament—but in the fragmentation of the state, splitting off to form a new state. Geopolitical opportunities in the form of an opposing weakness motivated the first coalition of the colonies. As Benjamin Franklin said, “We must all hang together or we will all hang separately.” But after the revolution succeeded—really a war of secession—why did centripetal forces not continue further? Those of us raised in American schools take it for granted

Geopolitics

211

that the thirteen colonies made up a natural unit and that they would expand to cover the continent from east to west. But there were in fact twentyfive British colonies in North America; only half of them joined the war of secession, and after the war, there were further rebellions of settlers demanding yet more independence from the U.S. coalition, especially over taxation. And there would be a number of other short-lived republics in continental America, including Texas, Utah, and California. Why did the U.S. coalition hold together instead of fragmenting? And why did it survive in the form of a democracy? Geopolitics surely helped. What institutionalized the federation from the 1790s onward was above all geopolitical opportunity: settlers’ demand to take Indian lands in the West that required help from the army of the federation. For its first century, the U.S. government performed few activities other than military ones. The spectacular geopolitical success of the U.S. coalition, which spread across the continent and built up its economic resources to become a world power, legitimated American democracy as well as American unity. This is not a very idealistic way of looking at it. We prefer to think that democratic institutions are intrinsically wonderful and that other countries ought to recognize this by looking at us and by listening to our idealistic rhetoric. As I have argued, the ideals and the rhetoric became popular because the geopolitics was favorable. If democracy resonates with slogans of liberty, the rights of the people, and success of the underdog, it is because democracies originated in times of geopolitical weakness. The slogans carried over and acquired new meaning when democratic power-sharing institutions were legitimated by military success. The ideology of democratic peacefulness may be an appealing one, but historically it is a myth. In conclusion, what does this conflict-oriented, geopolitical perspective say about the prospects for democratic power-sharing regimes today? Let me offer a few applications. To take a first example, the globe’s most notably failed regimes are in Africa. The legacy of European colonialism, combined with the use of external peacekeeping forces during the period of independence, prevented African states from going the route by which European states modernized. The successful European states underwent the military revolution because they were fighting and winning wars; this legitimated state penetration into society and the creation of a strong tax-extracting apparatus, which eventually became a benefits-providing apparatus as well. Where European states were democratic, their military victories legitimated democracy. The fact that

212

Randall Collins

autocratic states were defeated by the European democracies or collapsed under geopolitical strains has made democracy the prestigeful model of the past 150 years. But African states have not been allowed to go this route. Although peacekeeping forces have not been very successful in preventing internal war, they have prevented wars of conquest. This may well be a major reason why African states have built up so little legitimacy. A second example is Latin America. Miguel Centeno, in his book Blood and Debt: War and the Nation-State in Latin America (2002), has argued that the reason Latin American states are weak and prone to dictatorship is that they, too, failed to undergo the military revolution. Because geography provides good natural boundaries, it is difficult for Latin American states to fight each other. The result has been that armies are relatively small and inexpensive, used mainly in domestic politics. Without the pressure to support large armies, the apparatus of state penetration has been weak, and the mobilizing effects of strong patriotic rituals and ideologies are lacking. I would add that the lack of geopolitical success has eliminated a prime source of legitimating democracy. A third example is South Asia. The one successful democracy of the region is India. The unstable democracies, those rent with civil wars or prone to lapse into dictatorships—Pakistan, Sri Lanka, Bangladesh, Nepal—are those that have enjoyed no geopolitical success. Only India has been an expanding power, extending its sphere of influence and, despite its pacifist ideology, parading its military might. The operative cause is not merely the heritage of British institutions: most of the states in the region were under the British Raj. Indian democracy is legitimated, I suggest, because India has been able to expand its power-prestige in its external arena. A fourth case is Russia. As I have argued elsewhere, the USSR fell because of geopolitical strains, which delegitimated the existing regime: a nondemocratic one. Its successor turned to the opposite, high-prestige model: democracy. What will happen now? Would it be a good thing for democracy for Russia to fail militarily in Chechnya and to be humiliated by NATO expansion into its former possessions? Under the principle that geopolitical failure delegitimates state institutions, we should be wishing Russia’s electoral regime geopolitical successes, while it still exists. How do my arguments apply to the future of Iraq? Once the external force of the United States is removed from the equation, the initial phase of geopolitical weakness may well motivate power sharing of a kind. But would an Iraqi federation stand any chance of success in the local geopolitical en-

Geopolitics

213

vironment? From the perspective of a geopolitical theory, this is the crucial question. Note The argument of this chapter is an extension of that originally made in Randall Collins, Macro-History: Essays in Sociology of the Long Run (Stanford CA: Stanford University Press, 1999), chapter 4, “Democratization from the Outside In: A Geopolitical Theory of Collegial Power.”

CHAPTER 8

Courts, Constitutions, and the Limits of Majoritarianism Samuel Issacharoff

As my colleague Richard Pildes (2004) has proclaimed, we are in the Age of Democracy. Today more citizens participate in popu lar elections of government than at any other time in the history of world affairs. Democratization movements throughout the world have produced institutions of self-governance in regions where this was previously unthinkable. Even countries deep in the throes of tyranny or kleptocracy attempt to maintain a veneer of participatory engagement by their citizens, hoping perhaps that the act of casting a ballot serves as a sudden guarantee of legitimacy. The welcome expansive role of elections once again raises questions as to what properly constitutes democratic government, however. The term “democracy” is frequently used in its most minimal sense, as a system through which the majority, either directly or through representative bodies, exercises decision-making political power. And certainly, so long as the system is reasonably free and involves an uncorrupted popular selection of the head of state, a society is deemed democratic. Whether majoritarian selection is sufficient is another question altogether. Consider the Copenhagen criteria for accession to the European Union (EU) as a case in point. In order to be eligible for admission to the EU, a candidate state must have a demonstrated commitment to democracy. Little beyond this minimal condition is specified, leaving both the EU and potential entrants subject to ad hoc decision making on a country-by-country basis. The reason is largely historic and emerges from an era in which the democratic world in Europe could be quite aptly defined by the exclusion of the Soviet bloc, Francoist

Courts, Constitutions, and Majoritarianism

215

Spain, and the military regimes in Greece and Portugal. Little more analysis was needed, and such classification was straightforward. But when faced with claims for integration from the East, with divided societies such as the Baltics, Romania, and Bulgaria, or fratricidal societies such as the Balkans, this definition becomes predictably insufficient. In such countries, the question is not simply one of the majority being able to rule but of the limits on what the majority may do with its power. The emergence of new systems of self-rule forces us to ask whether such a spare definition is sufficient or whether further institutional constraints on the majority are necessary in order to produce stable governments capable of protecting other core liberal values, such as equality, rule of law, and freedom of speech. Once we turn to fractured societies marked by deep cleavages over race, ethnicity, language, or religion, the use of an elected head of state as the shorthand for our democratic aspiration becomes woefully inadequate. Popular selection only determines who shall rule; it does not speak of the constraints that must accompany the exercise of that power. In deeply divided societies, the emergence of stable democratic rule requires dampening animosities so that the population as a whole views the exercise of state authority as politically legitimate or, perhaps more modestly, does not rise in armed rebellion against the state. To this end, some guarantee is required that elections are not a one-shot, end-stage game in which access to state power will permanently define the relation of conqueror and vanquished. In a divided society, constraints on pure majoritarian democracy provide credible commitments that the faction in control of government will face limits in the use of state power against minorities in that country and that new majorities may emerge to dislodge the incumbent officeholders from power. For several decades now, the task of defining permissible bounds of constrained democracy has been the central challenge taken up by theories of consociationalism. In his classic account, Arend Lijphart (1977) identified the critical elements of the consociational experiment as turning on an ex ante allocation of political representation and other public benefits along the lines of the central cleavages of the society and the creation of a mutual veto or concurrent-majority voting rule to enforce that agreement. The key to the consociational model is that power will be allocated across competing interests in the society, independent of the political process. Thus, elections in consociational democracies typically can decide who among the candidates of a particular ethnic or racial group will hold an office that was predetermined to be assigned to that particular group; whether a particular group

216

Samuel Issacharoff

or interest should hold office is decided outside the electoral process through the formation of what Lijphart terms the “grand coalition.” Consociationalism is vulnerable to two primary lines of criticism. First, there are plenty of examples that demonstrate that formalizing power sharing along preset axes of social designations entrenches divisions within the society and ultimately does not work. Rather than diffusing the ethnic, racial, or religious antagonisms that led to the need to reconstitute a government, the consociational model may deepen or at least enshrine these divisions, with the danger of subsequent dissolution into further violence. The second and perhaps deeper criticism is that crystallizing confl ict at one point in time limits the range of electoral outcomes and is thus both inflexible and antidemocratic. Without some degree of accountability to the shifting social patterns within a society, a government structure runs the risk of losing its legitimacy as a democratic institution. As evident from the essays in this volume, consociationalism has developed a great deal since it was first formulated as a theory to explain government by the “grand coalition.” Over time, the concept of consociationalism has both been liberalized and expanded beyond the domain of the political. Many governmental arrangements have been brought within the rubric of consociationalism. These now extend beyond the formal power sharing of Lebanon or Cyprus or Belgium, characterized by rotation in office and formal assignment of duties along the fault lines of the particular country, to the more de facto arrangements by which dominant parties accommodate sectional demands for inclusion, as with the Liberal Party in Canada or the Congress Party in India. In some cases, it may even be applied to repeat-play coalitions that emerge ad hoc from the political process but that appear to approximate the same efforts to stabilize governance in fractured societies. In other areas, such as Northern Ireland, political consociationalism emerges in tandem with efforts at economic integration and job creation, a strategy that—thus far—seems to have stabilized a society seeking to overcome the misfortune of a time referred to as the “Troubles.” In prior writings, I have been quite skeptical of the successes claimed for the original visions of consociational power sharing, both as a matter of democratic theory and as a response to the empirical claims made for the stability of the efforts at formalized power sharing. As the literature has evolved, I find less to quarrel with, though I still maintain that with any theory that expands too far comes the risk of losing the parsimony that lends itself to testing. It is hardly a novel observation that a theory needs to be falsifiable

Courts, Constitutions, and Majoritarianism

217

in order to enrich our understanding of social phenomena. My concern, however, is not with the range of terminology or even with the balance sheet of the successes and failures of various experiments in formalized power sharing. I remain outside these debates and leave them to those with a dog in this fight. Instead, I want to present a constitutional lawyer’s take on what seems to be the critical insight of the consociational literature, from Lijphart to the present. That is, the critical frailty of majoritarian democracy in any fractured society is the risk that elections turn into a referendum on which group will wield the instrumentalities of state power against the other. Under such circumstances, the risk is always that an election in a fractured state will bear the characteristics of what the British ex-colonials sneeringly derided as elections in the postcolonial world they bequeathed: one man, one vote, one time. In more recent work on consociationalism there is a methodological shift from the clearer prescriptive forms of power sharing that characterized the early formulations into something new. Consociationalism has been recast as a system of political or structural overlays on the formal processes of democracy. On this view, consociationalism describes a series of institutional arrangements that are designed to constrain the majoritarian exercise of political power. On this account the early focus on prescribed arrangements such as executive power sharing, autonomy or self-government, proportionality, and veto rights are now seen as examples from among a non-closed menu of options all aimed at preventing democracy from becoming a one-off series of end-stage elections in fractured societies (McGarry and O’Leary 2004). To the critics of consociationalism, this should help avoid some of the sting of the charge that consociational mechanisms are antidemocratic in the simplistic sense that there is more room for politics. But as with the Copenhagen criteria for EU membership, sorting through the various institutional arrangements requires the use of a thicker conception of democracy. Democracy under such a conception both turns on an elected head of state and requires a repeat-stage election process that guarantees to the vanquished of today a chance to form part of a victorious coalition tomorrow. This is the concept of democracy as accountability that relies heavily on the idea of renewability of consent. Renewability of the consent of the governed provides temporary minorities sufficient expected payoffs in the long term to view the process as legitimate and prevents any attempt at withdrawal from society. Th is then becomes the critical issue for fractured societies, particularly the ones where the primary divisions of ethnicity, race, or religion

218

Samuel Issacharoff

are unlikely to yield to cross-cutting cleavages in subsequent election coalitions. In these deeply fractured societies, there are certain minority groups that will likely never constitute a sufficient segment of the population to claim political power—or at least not within any relevant time horizon. For example, the Turks in Bulgaria are a locked-in minority that cannot realistically aspire to participate in governance through mainstream political life. What, then, are the credible institutional arrangements that impose tolerable limits on majoritarian power in such divided societies? This chapter seeks to add to the consociational discussion by focusing on the role of courts as enforcers of constitutional commitments that may serve many of the same ends as more formalized conditions of power sharing. Although descriptions of consociational democracies have been recast to focus on the constraining force of institutional arrangements, the role of courts as enforcers of constitutional limitations on majoritarian prerogatives has yet to receive its proper due. For example, Lijphart (1999) amended his account of the pure majoritarian system to include the “absence of judicial review,” the implication being that, in contrast, constrained democracies include such review. However, the stabilizing impact of courts exercising constitutional review remains fairly underdeveloped and something of an afterthought in these theories. Rather than the formal power-sharing agreements of consociationalism, my focus is on the form of constrained democracy that has emerged in the past twenty-five years. All the newly emerging democracies of the late twentieth century have adopted some form of written constitution, along with some form of constitutionally based judicial review (Hirschl 2004). Indeed, in the recent rapid phase of democratization, epitomized by the former Soviet bloc and South Africa, the institutionalized power sharing that characterized the likes of Lebanon, Cyprus, Sierra Leone, and others has been largely absent. The dominant constraint has been the use of what I term “strong constitutionalism,” with constitutional courts emerging as the major institutional enforcers of the bargained-for constraint. There is already a recognition that constitutions serve what Ferejohn and Sager (2003, 1930) term restraints “against blatant majoritarian expropriation,” but courtenforced constitutions can further enable majorities in fractured societies to precommit to limitations on their power. These are complex institutional arrangements that may bear some features of consociational arrangements. A prominent example of consociational features embedded in constitutional courts is the composition of the

Courts, Constitutions, and Majoritarianism

219

Bosnian court outlined in the Dayton Accords. For six of the nine members, ethnic composition is formally divided among the major groups in the country: the Bosniaks (Bosnian Muslims), the Croats, and the Serbs. However, to prevent ethnic deadlock in adjudication, the remaining three members of the court must be noncitizens, must be selected by the president of the European Court of Human Rights after consultation with the presidency, and cannot be citizens of any neighboring country. A further example comes from the difficult negotiated arrangement on appointment to the South African Constitutional Court during the interim period between apartheid and adoption of the final constitution, a period of truly formalized power sharing. The claim that courts may enforce constitutional arrangements to constrain majoritarianism is really only the framing intuition behind this chapter and my other work in this area (e.g., Issacharoff 2004, 2008). My major concern is not so much the political science account of the institutional role of courts but the constitutional lawyer’s concern for what courts should do when called upon to play this role. To make this point, I contrast two critical decisions from the South African Constitutional Court: the certification decision concerning the proposed constitution for post-apartheid South Africa (In re Certification of the Constitution of the Republic of South Africa, 2006) and the subsequent lower-profile case involving the ability of legislators to shift party affi liation, in what is termed “floorwalking” (United Democratic Movement v. The President of the Republic of South Africa, 2003). The certification decision is a prime example of the role courts can play in successfully limiting the role of majoritarianism, while the later floorwalking decision offers an example of how the Constitutional Court declined to intervene in a clear situation in which the majority party employed state power to further entrench itself in the government.

South African Constitutional Certification There is perhaps no more inspiring story of the transition from authoritarianism to democracy than South Africa. Nor is there any clearer example of a society riven by formal boundaries of group oppression than South Africa struggling to emerge as a multiracial and multiethnic democracy in the aftermath of apartheid. In a true sense, the attempt to create a stable interim government was premised on formalized power sharing to ensure all groups a mutual veto over

220

Samuel Issacharoff

contested governmental action—precisely the formula for consociationalism identified by Lijphart and other proponents of this approach. The interim constitution provided detailed power-sharing arrangements, along with a critical list of thirty-four constitutional principles of democratic governance, to which I shall return. The primary mechanism was the election of a parliament by proportional representation and the assignment of the position of deputy president to the representatives of each party holding at least 80 of the 400 seats in the National Assembly. Among the powers conferred as a result of this representation in the executive was the ability to participate in the selection of some of the justices of the Constitutional Court, a power reserved to the executive branch. What was novel was precisely the way in which the thirty-four constitutional principles created an enforceable precommitment on the limitations of government, even once a permanent body was selected democratically. More innovative was the way in which this limitation on majoritarian power was to be implemented and policed. Rather than turn to the perhaps expected mechanisms of formalized divisions of power, as evidenced by the interim constitution, the task of policing ruling majorities was largely entrusted to the newly created South African Constitutional Court. Even there, the South African negotiated agreements anticipated an expanded conception of the role of an independent judiciary. Beyond the familiar powers of judicial review over the constitutionality of proposed legislation, the South African Constitutional Court had a power that, to the best of my knowledge, had never before been imparted on any court. Under the negotiated provisions of the interim constitution, the final constitution could not be adopted unless it faithfully adhered in its implementation to the negotiated general principles set out in the interim constitution (Gloppen 1997, 199). The Constitutional Court was entrusted with the power to ensure that the final constitution conformed to the thirty-four principles. The task of ensuring compliance was given in its entirety to the Constitutional Court. In effect, once South Africa emerged as a full constitutional democracy, the constitutional court would stand as the ultimate arbiter of the constitution, holding full powers of judicial review. Ironically, however, the court would predate the constitution and would serve as the final body approving the adoption of the constitution itself. Hardly customary yet innovative, this arrangement seemed to satisfy the security interests of all parties and was integral to the peaceful transition to constitutional democracy.

Courts, Constitutions, and Majoritarianism

221

In July 1996, the proposed permanent constitution was submitted for review to the constitutional court, which rendered its decision two months later. Of greatest significance for present purposes are the provisions that reaffirmed limitations on government and those that were struck down for what may be termed an excess of majoritarianism. These primarily concerned the attempt to preclude constitutional review from certain categories of statutes, the absence of federalist safeguards on centralized power, and the lack of supermajoritarian protection for certain components of the constitution itself, including the liberty protections of the Bill of Rights. With regard to the latter, the court found a violation of the principles of the interim constitution in the failure to “entrench” the rights in question. In summary form, the court’s majestic ruling turns on the following key understandings of permissible constitutional law: 1. Bicameralism and divided government. Invoking Montesquieu, the court reaffirmed the importance of checks and balances across the branches of government. The court pointed specifically to the creation of an upper house (the National Council of Provinces) that would not be based on equipopulational voting but on the election of ten representatives from each of the nine provinces. This has great practical significance because one of the provinces is majority Zulu (hence outside the political orbit of the African National Congress [ANC]) and two others have large concentrations of what are known in South Africa as “white” and “colored” voters. 2. Federalism. The court strictly enforced the requirement (found in Principle XXII) that the powers of the federal government would be offset by guaranteed limitations of federalism. Thus the court found unconstitutional those provisions that failed to provide the required “framework for LG [local government] structures” (p. 861), as well those that failed to ensure the fiscal integrity of political subdivisions. For the court, the South African constitution should provide only those powers to the national government “where national uniformity is required” (p. 845), and only economic matters and issues of foreign policy met this restrictive definition. 3. Supermajoritarianism. The court also strictly construed Principle XV, which required “special procedures involving special majorities” for constitutional amendments. According to the court, the purpose of this provision was to secure the constitution “against political agendas

222

Samuel Issacharoff

of ordinary majorities in the national Parliament” (p. 821). Various provisions of the proposed constitution requiring supermajoritarian action were nevertheless struck down for failing to create special procedures outside the framework of ordinary legislation. Thus, for example, a provision allowing a two-thirds majority of the lower house to amend parts of the constitution failed because “no special period of notice is required; constitutional amendments could be introduced as part of other draft legislation; and no extra time for reflection is required” (p. 822). Similarly, the court found that allowing the Bill of Rights to be amended by a two-thirds majority of the lower house failed the “entrenchment” requirement of Principle II, which, the court ruled, required “some ‘entrenching’ mechanism . . . [to give] the Bill of Rights greater protection than the ordinary provisions of the [constitution]” (pp. 822–23). 4. Judicial review. The rejection of judicial review for certain categories of statutes was found to violate the commitment to constitutional supremacy in Principle IV and the jurisdictional guarantees for judicial power contained in Principle VII. 5. International law. Although not a central issue in the ultimate approval of the constitution by the court, Article 39 of the final constitution provides that, in construing the Bill of Rights, a court “must consider international law.” Most observers will note only the valorization of international human rights norms. But the incorporation of a source of law beyond the control of the parliamentary majority again serves to constrain majoritarian prerogatives by providing an independent, nonparliamentary source of authority for courts to enforce. Thus in construing the obligations of the constitution, the court found an obligation to protect those rights recognized in open and democratic societies as being “inalienable entitlements of human beings” (p. 790). While the court recognized that no consensus exists as to what entitlements are inalienable, the court required the constitution, at a minimum, to guarantee rights that have achieved a wide measure of international acceptance. Thus, according to the court, in aggregating the fundamental rights of other societies, South Africa’s proposed constitution established a set of rights “as extensive as any to be found in any national constitution” (p. 790). 6. Minority party security. The interim principles and the ultimate constitution contained strong protections of proportional representation, and

Courts, Constitutions, and Majoritarianism

223

the constitution enshrined this provision as an “antidefection” principle in which a member of Parliament would have to resign if he or she attempted to switch parties. Although such provisions may restrict expression of beliefs by legislators, there is an overriding concern that minority legislators could be induced to sway from their constituents’ interests to support majoritarian policies. Since by definition there are fewer minority than majority representatives, any single minority defection would have a more severe impact on the representation of the minority population than the defection of a majority legislator would have on the representation of the majority. Such defection to the majority is not only more costly but also more likely. Minority caucuses are unlikely to be able to offer the same personal opportunities or chances for local blandishments as is the majority. In rejecting the civil liberties challenge to the antidefection clause, the court noted that antidefection clauses were found in the constitutions of Namibia and India and were therefore entirely consistent with democratic governance. The Constitutional Assembly then revised the constitutional draft to meet the court’s concerns in October 1996 and, following a second round of judicial scrutiny, the new constitution was signed and implemented by President Nelson Mandela in December 1996. The new constitution drew much fanfare, particularly among those taken with its ample guarantees of not just political rights but also the sort of economic rights that had frustrated American courts even in the most expansionist Warren Court era. The new constitution even assures South African citizens the right to “an environment that is not harmful to their health and well-being.” Cass Sunstein seized upon this feature of the South African Constitution to declare it “the most admirable constitution in the history of the world” (2001, 261). For many champions of activist courts, the broad recognition of judicial authority in implementing constitutional principles combined with the sweeping rights guarantees sparked expectations that postenactment cases would usher in an era of active judicial engagement with the process of dismantling the legacy of apartheid. Consistent with its origins as the handmaiden to constitutional democracy, however, the court has played a circumspect, pragmatic role, refusing to develop a broad mandate of rights jurisprudence. Thus much criticism was directed at decisions such as Republic of South Africa v. Grootboom (2001), in which the court distinguished between the rights guarantees of

224

Samuel Issacharoff

the constitution and the deference that should be accorded the government in crafting policies to achieve those objectives. To my mind, critics of the court largely fail to appreciate the true significance of the South African constitutional process and of the role played by the court in that process. The use of an interim constitution defused the pressure to resolve deeply contested governance questions on a once-and-for-all basis while providing a sense of security to minority constituencies. The interim constitution and the authority entrusted to the Constitutional Court enabled the transitional democratic organization of parties and constituencies and gave participatory legitimacy to the drafting and ratification of the final constitution. After helping create a fledgling democracy in a deeply divided and still violent society, the court was reluctant to substitute decrees for democratic deliberation and compromise, which seems well in keeping with the ambitious mandates of the transitional process.

Floorwalking All election systems run the risk of excess of majoritarianism if the capture of state institutions by an electoral majority is not tempered, most notably by constitutional constraints. Many countries confront minority-access issues in connection with matters such as thresholds for success in representation, particularly in nonproportional electoral systems relying on territorial units as the basis of representation. In most instances, the debates center on territorial units as the basis for attaining office for candidates preferred by minority constituencies. But merely ensuring the election of minority-preferred candidates does not end the tension between minority preferences and the global claims of the political process, as the South African Constitutional Court had to confront in the original certification decision. Once the new order was in place, the issue was joined again. In South Africa, the initial constitution required a member of Parliament to resign if he or she attempted to switch parties in what is known as “floor crossing” or “floor walking.” The provision was an explicit subject of negotiations in the transition from apartheid, reflecting fears that the likely parliamentary majority of the ANC could be used to woo minority legislators and overconcentrate political power. Once in office and once its political power was consolidated, however, the ANC used its legislative supermajority to repeal the antidefection provi-

Courts, Constitutions, and Majoritarianism

225

sion. Under the new law, defection was permitted as long as the defecting group constituted at least 10 percent of the party’s legislative delegation. Th is did little to placate critics, since this would pose a very large hurdle to defections from the ANC but would leave defection an individual choice for any party with less than ten members of Parliament. The constitutional amendment prompted a second constitutional challenge, this time a claim that the amendment would violate the principles of party integrity and separation of powers inherent in the entire constitutional structure. The question was whether the court would again view its role as a guardian of constitutional limitations of majoritarian power. In its second review of antidefection issues, however, the court’s reaction was less forceful. The court rejected the challenge both on the procedural ground that the mechanisms of constitutional amendment had been adhered to and, more significantly, on a deeper theory that representation after an election is not a right that attaches to a subset of voters: “The rights entrenched under section 19 [of the constitution] are directed to elections, to voting and to participation in political activities. Between elections, however, voters have no control over the conduct of their representatives. They cannot dictate to them how they must vote in Parliament, nor do they have any legal right to insist that they conduct themselves or refrain from conducting themselves in a particular manner” (United Democratic Movement v. The President of the Republic of South Africa 2003, p. 49). The court’s analysis was along conventional terms as a rights claim by electors rather than in structural terms for what the core constitutional tension was and remains in South Africa. To a great extent, the creation of strong constitutional courts in emerging democracies is an indication that they are expected to play a more direct role in superintending the institutions of democracy and, particularly, in defining the limits of democratic decision making. Nowhere is this clearer than in South Africa, where the interim compromise between the ANC and the apartheid rulers of the National Party required the immediate creation of a constitutional court as a structural safeguard against unbridled majority power. Viewed in this light, floor walking might be a minor example of the consolidation of majority coalition entrenchment, but it is nonetheless an unmistakable effort by the controlling party to enlarge majoritarian power. The somewhat cruel irony here is that this particular form of expansion of majoritarianism was anticipated and deliberate steps were taken to prevent it in the original political bargain. The court’s failure to assume a more forceful role is troubling.

226

Samuel Issacharoff

In some sense, the South African example raises the issue whether the triumph of a modern democracy in a divided society can avoid succumbing to the pressures toward majoritarianism. There are unfortunate historic examples of a social revolution yielding parties whose majority power portended the decline of true democracy. As the era of majority rule in South Africa continues to be dominated by the ANC, there should be concern that the party will turn into a variant of the Mexican Partido Revolucionario Institucional (PRI), which used its majority support and moral capital to saddle Mexico with single-party control for more than seventy years.

Conclusion The South African example demonstrates that courts can and do play a role in constraining majoritarian power as a mechanism for stabilizing democratic governance without the reinforcement of ethnic and racial divides created by formal power sharing, as is characteristic of consociationalism. Overall, this requires an adjudicative style that permits courts to use their authority to diff use ethnic or racial antagonisms and rule on the bounds of majoritarian politics in order to consolidate a constitutional order. Constitutional courts are by nature more assertive, and they must recognize their role in mitigating the dangers of majoritarianism present in democracy. Notes 1. For a more complete account of the certification decision, see Issacharoff 2004. 2. This is referred to as the requirement that there be “special procedures involving special majorities” for constitutional amendment and for any alteration of the constitutional guarantees of individual rights. 3. For example, Section 35 of South Africa’s interim constitution provides: “In interpreting the provisions of this chapter a court of law shall promote the values which underlie an open and democratic society based on freedom and equality and shall, where applicable, have regard to public international law applicable to the protection of the rights entrenched in this chapter, and may have regard to comparable foreign case law.” 4. For a summary of the criticisms of the South African Constitutional Court and a defense of the institutional paradigm, see Kende 2002, 753. 5. For a more complete account of this decision, see Issacharoff 2008. 6. India and Namibia adopted similar constitutional prohibitions on floor walking. New Zealand prohibited party switching by members of Parliament in the Electoral

Courts, Constitutions, and Majoritarianism

227

(Integrity) Amendment Act, 2001, but the prohibition was statutory and expired by design in 2005. 7. I am indebted to Pablo de Greiff for suggesting the parallel to the PRI. References Ferejohn, John, and Lawrence Sager. 2003. “Commitment and Constitutionalism.” Texas Law Review 81: 1929– 63. Gloppen, Siri. 1997. South Africa: The Battle over the Constitution. Sudbury, MA: Dartmouth Publishing Company. Hirschl, Ran. 2004. Towards Juristocracy: The Origins and Consequences of the New Constitutionalism. Cambridge, MA: Harvard University Press. Issacharoff, Samuel. 2004. “Constitutionalizing Democracy in Fractured Societies.” Texas Law Review 82: 1861–93. ———. 2008. “Democracy and Collective Decision Making.” International Journal of Constitutional Law 6: 231– 66. Kende, Mark S. 2002. “The Fift h Anniversary of the South African Constitutional Court: In Defense of Judicial Pragmatism.” Vermont Law Review 26: 753– 68. Lijphart, Arend. 1977. Democracy in Plural Societies: A Comparative Exploration. New Haven, CT: Yale University Press. ———. 1999. Patterns of Democracy: Government Forms and Performance in Thirty-Six Countries. New Haven, CT: Yale University Press. McGarry, John, and Brendan O’Leary. 2004. The Northern Ireland Conflict: Consociational Engagements. New York: Oxford University Press. Pildes, Richard H. 2004. “Foreword: The Constitutionalization of Democratic Politics.” Harvard Law Review 118: 29–154. Sunstein, Cass. 2001. Designing Democracy: What Constitutions Do. New York: Oxford University Press.

PA R T III Contemporary Power-Sharing Questions

CHAPTER 9

A Revised Theory of Federacy and a Case Study of Civil War Termination in Aceh, Indonesia Alfred Stepan

All independent democratic states have a degree of cultural diversity, but for comparative purposes we can say that, at any given time, states may be divided analytically into three different categories: 1. states that have strong cultural diversity, some of which is territorially based and politically articulated by significant groups that in the name of nationalism and self-determination advance claims to independence; 2. states that are significantly culturally diverse but whose diversity is nowhere organized by territorially based, politically significant groups mobilizing nationalist claims for independence; and 3. states in which a community, sufficiently culturally homogeneous to consider itself a nation, dominates the state, and no other significant group articulates similar claims. In this chapter, I will call states whose territory, in part, falls into the first category “robustly politically multinational.” These comprise par ticular deeply divided places. Canada (owing to Quebec), Spain (especially owing to the Basque Country and Catalonia), and Belgium (owing to Flanders) are robustly politically multinational. India, owing to the Kashmir Valley alone, merits classification in this category. Furthermore, at various times the Mezo

232

Alfred Stepan

movement in northeast India, the Khalistan movement in the Punjab, and the Dravidian movement in southern India, as well as others, have also given a multinational dimension to Indian politics. Switzerland and the United States are both sociologically diverse and multicultural. However, since neither country has significant territorially based groups mobilizing claims for independence, both countries clearly fall into the second and not the first category. Finally, countries such as Japan, Portugal, and Norway fall into the third category. In this essay, I restrict my focus to “politically robust multinational societies.” What type of political arrangements might be appropriate for such societies? Elsewhere I have argued that a hard nation-state model of the French Third Republic sort is particularly unsuited for robustly politically multinational societies (Stepan 2008) and have developed in a book, with Juan Linz and Yogendra Yadav, a new ideal type to contrast with “nation-state” that we call “state nation” (Stepan, Linz, and Yadav 2010). However, we need to expand our political imaginations and our sense of usable repertoires of political arrangements beyond the idea of nation-state and even of state nations. This essay presents a revised concept of federacy and a case study of the use of the idea of federacy for ending the long civil war in Aceh, Indonesia, on August 15, 2005. A central ingredient of the Stepan, Linz, and Yadav political formula of state nation involves federalism, indeed asymmetrical federalism. However, in this essay, I restrict myself to policies that might be useful and available in a unitary state whose political elites are either not willing or, for geopolitical or other reasons, unable to undertake a profound process of polity-wide reorganization that would be a necessary condition of becoming a federal state. With these constraints, could a federacy be a way of managing a deeply divided place in a unitary state?

Conceptualizing the Unitary State—Territorially Concentrated Minority Problem Let us first very briefly hypothesize the character of the political problem of territorially concentrated minorities in a unitary state. Let us imagine an existing small or medium-sized independent state with a unitary constitution, a relatively culturally homogeneous population, and a proud nation-state sense of identity. But let us further assume that this otherwise well-functioning

Revised Theory of Federacy

233

unitary state has a territorially concentrated population that does not identify with the history and culture of the unitary state. Its different identity will most likely stem from one or a combination of the following six factors: 1. 2. 3. 4.

physical separation and/or great distance from the unitary state; linguistic or religious difference; a previous self-governing tradition it wants to restore and/or expand; a geopolitical/cultural sense that it was, and should be, part of a neighboring irredentist state with which it identifies; 5. a radically different economy that the population believes needs special laws to help preserve their own livelihood and way of life; and 6. a history of warfare with, and/or coercive repression by, the unitary state within which it is located. Let us make a further (not unreasonable) assumption that this territorially concentrated minority population is unhappy with the status quo. At the very least, they want more autonomy and self-governing arrangements that give recognition to their culture and facilitate its development; some of the population may identify with leaders and organizations that demand independence or secession. This combination of factors gives a multinational dimension to the unitary state. Let us make some further constraining assumptions about two possible resolutions to this political problem: federalism or independence. Concerning federalism, if the population and dominant political actors of the unitary state agree to become an asymmetrical federal system, many of the problems of territorially concentrated minority population, short of independence, could be addressed reasonably well. However, if the vast majority of the state’s population and its major political actors do not want to change what they perceive as a well-functioning unitary state, we should assume that a federal state resolution, let alone an asymmetrical federal state, is highly unlikely. Let us also assume that full “exit” from the unitary state of the territorially concentrated minority, either by independence or joining an irredentist neighbor, is extremely unlikely for the foreseeable future because of the following factors singly or in combination: (1) strong sentiments in the unitary state opposing such an exit and the military capacity to back up such policy preferences; (2) geopolitical opposition by many neighboring states to the emergence of a new, small, and possibly weak independent state in the region

234

Alfred Stepan

or to its joining the irredentist neighbor; (3) divided sentiments in the minority population about the political and financial costs of independence. If we rule out federalism and complete exit, are there any other political arrangements that could possibly increase the “voice” and “loyalty” of the territorially concentrated minority community or nation (Hirschman 1972)? Can their desires for greatly enhanced autonomy and self-government be responded to from within the unitary state? Specifically, can we imagine a possible formula for managing this potential threat to a peaceful, inclusionary democracy in the existing state that could be acceptable to both the majority and the minority “nations”? I believe there is. At the theoretical, ideal-typical conceptual level, we can imagine what I call a “federacy” formula. (Later I will discuss alternatives in the literature to the ideal type and justify why ours is restricted to unitary, independent states.)

Federacy as an Ideal Type Our short ideal-type definition of a federacy is the following: A federacy is a political-administrative unit, in an independent unitary state, with exclusive power in certain areas, including some legislative power, constitutionally or quasi-constitutionally embedded, that cannot be changed unilaterally and whose inhabitants have full citizenship rights in the otherwise unitary state. The minimal agreed set of arrangements for such a federacy, as an ideal type, would have to satisfy five institutional requirements.

Federal-Like Agreement on Division of Powers and Responsibilities

In order for the federacy, unlike other parts of the unitary state, to pass special self-governing laws and administer its polity in such a way that it can address many of the areas of most tension with the unitary state, there must be a classic federal-like agreement concerning the division of powers in the polity. These federal-type arrangements are (1) explicit powers that fall in

Revised Theory of Federacy

235

the exclusive domain of the federacy, (2) powers that remain in the domain of the center, and (3) powers that might be shared or even remain with the center but can be progressively transferred permanently to the federacy. Powers that are in the exclusive domain of the federacy are the vital culture-making and culture-preserving powers, such as the right to establish the indigenous language as an official language of the federacy, control over the content and administration of education, the hiring and promotion of federacy civil servants, and possibly the granting of citizenship in the federacy to supplement their full citizenship in the unitary state. The federacy may have a range of prerogatives not found in any other part of the unitary state, such as the legal right to create extremely restrictive immigration policies in the federacy and to prohibit those citizens of the polity who are not also citizens of the federacy from buying land or establishing commercial enterprises. The federacy also might have the right to create its own rules governing the federacy’s legislature and executive and have extensive control over local development plans. Some powers, such as foreign affairs, defense, currency, and a final court of appeals, would generally remain the exclusive powers of the center. Finally, unlike most federations, there can be a presumption that some powers, especially the provision of costly ser vices such as pensions, welfare, and hospitals that may have been originally created by the center, may be increasingly transferred, by mutual agreement, to the federacy as it develops the fi nancial and administrative capability to be fully self-governing in these areas. The center would have cost-sharing formulas with the federacy and would normally contribute a significant amount to the federacy’s budget.

Quasi-Constitutionally Embedded Political Autonomy of the Federacy

The territorially concentrated minority population would have a constitutionally, or quasi- constitutionally, embedded degree of political autonomy well beyond that found in any other part of the unitary state. Ideally the legislatures of the unitary state and the newly created federacy would both agree to the new autonomy arrangements, but at an absolute minimum it would be stipulated that the act of autonomy, or federacy agreement, could not be unilaterally

236

Alfred Stepan

altered without exceptional majorities on both sides. The federacy situation might begin by negotiations and agreements between states, as a way of settling their disputes over the demands of two of the most important competing principles of modern politics: “the right of state sovereignty” and the “right of self-determination of populations.” Eventually, however, representative bodies of the territorially concentrated population and the unitary state would have to debate about, and vote for, the federacy arrangements. Legally, politically, and often internationally, federacy arrangements would be much more binding on the central government of the unitary state than devolution or decentralization. The latter two may be unilaterally reversed by parliamentary majorities, whereas a constitutionally embedded federacy can only be changed by mutual exceptional majorities.

Dispute Resolution Procedures

The federacy and the state, as part of the eventual autonomy agreement, would have dispute resolution procedures about their respective powers and prerogatives. Only in exceptional circumstances would a dispute go to the state’s highest court, and this court could make binding decisions for the entire polity only as long as it did not in any way violate the constitutionally embedded autonomy agreement. If the legislature of the center were considering a bill within its powers that might have a special impact on the federacy, the federacy would often have the right to present its views, and on some matters not only would consultation with the federacy be required but its express consent.

Citizens of the Federacy Would Vote for and Have Representatives in the Parliament of the Unitary State, and the Center Would Have Official Representation in the Federacy

The goal of the federacy arrangement is to create a high level of trust, voice, and loyalty between the federacy and the center; thus there is joint citizenship in the federacy and the center. The citizens of the autonomous unit would therefore be full citizens of the state, vote in statewide elections, and have representatives in the parliament of the state as well as in their own parliament. The center would have an official representative in the federacy

Revised Theory of Federacy

237

who would also help coordinate the activities in the federacy that fall under central state powers.

The Autonomy Agreement Would Be an Integral Part of a Single, Independent, Internationally Recognized State

The autonomy agreement would be an internal law of the unitary state and not part of international law. The autonomy agreement might be derived from international agreements but would not be part of international law (unless the state makes a specific exception) and would not create a “subject” of international law. If these five requirements are functioning, then federacy arrangements may additionally contain and legitimate two peace-facilitating features that are not normally found in unitary or federal states.

International Guarantors May Help Set Up the Federacy

A federacy arrangement, especially if it emerges in the context of geopolitical conflict, may well combine elements of two of the greatest principles in conflict since the late nineteenth century: “the right of state sovereignty” and “the right of self-determination of populations.” In such cases, federacy arrangements, more than in unitary or federal states, may involve some participation of concerned neighboring states and international organizations in their emergence and even in the guaranteeing of some arrangements, particularly the distinctive cultural rights of the federacy’s population.

The Federacy May Eventually Acquire Some Role in Approving and/or Structuring International Treaties Signed by the Center

More than in a federation, and totally unlike a unitary state, it is possible that the very existence of a federacy may facilitate its advocacy, and achievement, of some agreed “opt-out” arrangements of treaties into which the unitary state enters. This would be most likely if the leaders of the federacy, with the support of the unitary state, believe that certain provisions of the treaty would be harmful to the preservation of the distinctive economy and

238

Alfred Stepan

way of life of the population that the federacy had been created to protect. No other part of the unitary state would have the status to construct such opt-out arrangements with the participation and help of the center, which normally negotiates them.

Federacy’s Conceptual Distinctiveness from “Unitary States,” “Asymmetrical Federations,” “Confederations,” and “Associated States” Given the defining characteristics of our ideal type of federacy, it should be clear that it is quite different from the ideal type of a “unitary state.” This is so because of the high degree of constitutionally embedded autonomy and prerogatives that are not enjoyed by other jurisdictions of the unitary state. However, as we have seen, a unitary state can enter into a federacy; indeed it may be a particularly useful formula for managing a small or medium-sized unitary nation-state’s problems with a territorially concentrated minority population or nation. As an ideal type, a polity that contains a federacy is also analytically quite different from a federal polity, even an asymmetrical federal system. True, the unit that we call a federacy, especially in the first and second defining requirements of our ideal type, has a federal quality in its relations to the state. However, in the ideal type of federalism, the entire polity is federal. Some readers might think that what we call a federacy is close to the autonomy and special prerogatives that are found in what we have called asymmetrical federalism. This is true, but some differences are nonetheless fundamental. In asymmetrical federalism, every unit in the polity is part of a federation, and no part of the polity is part of a unitary state. Further, in our judgment, an asymmetrical federal system does not need a federacy because the specific prerogatives obtainable for minority populations can be provided for in an asymmetrical federal system. Witness the vast amount of special prerogatives India has negotiated and delivered for Mizoram, where a once war-torn secessionist society now has high degrees of voice and loyalty inside India. Contrary to what some readers may think, federacies are also fundamentally different from confederations. A confederation is an agreement between states. A federacy is an agreement within a state. In a confederation, a member state may make a unilateral decision not to participate in a collective foreign policy endeavor, unless such a decision violates the specific

Revised Theory of Federacy

239

treaty creating the confederation. However, in a federacy, opting out of a decision unilaterally is not constitutional. “Opt outs” by a federacy can only be attempted with the prior agreement and help of the center. Finally, some observers might feel that federacies and “associated states” that do not qualify for membership in the United Nations (or even some colonies) are analytically the same. However, our ideal-type federacy has the requirement that citizens resident in the autonomous unit participate fully in elections for the central government of the polity, have their own representatives in the parliament, and thus play a role in government formation. In a presidential system citizens resident in the autonomous unit have the right to participate in the election of the president. In any case, if the autonomous unit were actually still a colony, this would violate our idealtype stipulation that the federacy is an integral part of an independent state. We do not claim to have invented the word “federacy,” which has a long history going back to the Greeks and which has been briefly and variously used by social scientists and lawyers interested in discussing new forms of political autonomy for many decades. However, in our judgment, the term is used in so many ways that nothing like a cumulative literature in comparative politics has or can emerge, unless an attempt at definitional closure is made. I hope this chapter contributes to that task because the concept, and the practice, could be conceptually and politically useful, especially for managing multinational societies democratically in deeply divided polities. The scholar who did most to introduce the concept was probably Daniel J. Elazar. However, when Elazar advanced his definition in the introduction to his extremely useful compendium, he gave only two specific examples of federacies: Puerto Rico and the United States, and Bhutan and India (Elazar 1991, xvi–xvii). Neither meets my definition of a federacy. Puerto Rico’s commonwealth status with the United States is a very interesting arrangement, but Puerto Ricans, resident on the island, do not have their own representatives with a vote in the U.S. Congress, and they cannot vote in federal elections for the president of the United States, though they can vote in the party primary to select a candidate. In the case of Bhutan and India, both have been, since independence in 1949 and 1947, respectively, independent states. Bhutan has been a member of the United Nations since 1971. Bhutan’s elected bodies and judicial systems are fully separate from India’s. Bhutan, which has borders with China, has “outsourced” much, but not all, of its defense tasks to India. Bhutan is possibly thus an interesting case of an “associated state,” but it is not a federacy by my definition. Elsewhere

240

Alfred Stepan

Elazar classifies Jammu and Kashmir as a federacy of India, but in my view it is actually a disputed component of an asymmetrical federation. David Rezvani’s dissertation on federacy contains much useful documentation and analysis, especially of what we call the requirement in a federacy for constitutionally embedded political autonomy and the requirement for conflict adjudication procedures (Rezvani 2004). However, like Elazar, he does not stipulate that citizens of a federacy must have representatives in the legislature of the entire state or the right to vote for the president; this leads him to classify such entities as Puerto Rico, Northern Marianas Islands, and the Palau Islands as federacies. Like Elazar he often classifies what Linz and I deem cases of asymmetrical federalism as federacies. Thus, two of the four “federacies” he selects for special attention in an eleven-page appendix, along with the Åland Islands and Greenland, are Catalonia and the Basque Country, neither of which fits our definition of a federacy (Rezvani 2004, appendix C).

Did the Idea of Federacy Play a Crucial Role in the Aceh Peace and Democratization Process in Indonesia? In 1998, when the thirty-two-year-long military regime of General Suharto fell in Indonesia, the world’s most populous Muslim majority country, democratization was one possible outcome. But by 2003, a fourth Indonesian president since Suharto was already on the horizon. The new political managers of the state’s two-thousand-mile archipelago were trying to manage secessionism in its eastern flank of Papua and in its northwestern flank of Aceh. East Timor had already won its struggle for independence. MuslimChristian riots (with some military and police complicity) had occurred on islands such as Ambon and Sulawesi. Massive al-Qaeda-linked explosions had killed 202 in Bali and wounded approximately 170 in the Marriott Hotel in Jakarta. Thus while democratization was still one possibility, crises of “stateness,” intensified religious conflicts, and the return of military autocracy were other possibilities. To many in the country’s capital, Jakarta, the crisis in Aceh presented particularly dangerous problems for the unitary state of Indonesia. Aceh scores relatively high on all the variables (except for irredentism) that make it difficult to manage politics democratically and peacefully in a unitary state. Many of Aceh’s population pride themselves on espousing a purer form of Islam than in the Indonesian heartland of Java; many think of themselves as mem-

Revised Theory of Federacy

241

bers of the Acehnese, not Indonesian, nation; many want more control over their oil resources; and Aceh has had a long history of resistance to, and armed repression by, the unitary state. Many Acehnese gave active or at least passive support to the armed resistance to colonial rule in the late nineteenth century and to armed rebellions and separatist or independence movements in the 1950s and for much of the last thirty years. The most recent insurgent group, the Free Aceh Movement (GAM), started an insurrection in 1976 and intensified their struggle for independence in 1999 in the political and coercive spaces created by the fall of Suharto (Aspinall 2002, 2009). There are convincing arguments that for GAM, nationalism eventually became more important than Islamism (see Aspinall 2007a, 2007b). There were failed peace processes in 2000 and a Cessation of Hostilities Agreement (COHA) that began in December 2002. However, the government of Indonesia, backed and/or pushed by the military and many parliamentarians who were convinced that GAM was only using the ceasefire to strengthen its guerrilla forces for a further independence drive, declared martial law in May 2003. The chief of the armed forces, General Endriartono Sutarto, told troops their mission was “destroying GAM forces down to their roots” and “finishing off, killing those who still engage in armed resistance” (Aspinall and Crouch 2003, 1). Yet by March 2009 Aceh and the central government were in their fourth year of peace. The World Bank office in Aceh stressed that “political commitment to the peace process remains strong on both sides” (Clark and Palmer 2008, vi). Indonesia was ranked as one of the two most politically democratic of the world’s forty-seven Muslim majority polities (Freedom House 2008). Scholarly debates were now less about the challenges of “stateness,” or political Islam, than about what still needed to be done before Indonesia would be democratically consolidated (Liddle and Mujani 2009). Did federacy or, better, the idea of federacy contribute to this startling change, especially peace in Aceh? A strong case can be advanced that it did. One of the major reasons why the cessation of hostilities broke down in 2003 was the continuing clash of fundamental goals. The central government and the military were worried about the territorial fragmentation of Indonesia and were deeply suspicious of any formula other than a unitary state. GAM, for its part, was unable to see how it could advance its social and developmental goals outside of independence. In these circumstances four facilitating events happened that helped the peace outcome. Three of these have been well analyzed, but in our judgment none of the four events on their own are sufficient to explain the outcome.

242

Alfred Stepan

First, a tsunami killed nearly 150,000 people in Aceh on December 26, 2004. Within a week, GAM and the newly elected Indonesian president expressed interest not only in a ceasefire but in negotiations. However, as a single causal exogenous explanation, the tsunami disaster is not satisfactory. The same tsunami smashed neighboring Sri Lanka, which saw the opposite political effect: the struggle over relief resources worsened. Eventually the Norwegian peacekeeping mission was asked to leave so that the government could make one more attempt to eliminate the Tamil Tigers (Stepan, Linz, and Yadav 2010, chap. 4). Furthermore, it has now been documented that significant preparations for secret peace discussions about Aceh had begun at least a month before the tsunami (Kingsbury 2006, 15–22; Aspinall 2005). The second explanation often draws on Dankwart Rustow’s classic “dynamic model,” in which he advanced the argument that if two opponents are engaged in an exhausting major political struggle and one side can defeat the other, victory by one side, rather than democratic concession or democratic peace accords, may well be the outcome (1970). For Rustow it is precisely “prolonged and inconclusive political struggle” that can be conducive to the construction of new master frameworks of peace or democracy (1970, 352). Modern conflict settlement theory employs the analogous concept of “hurting stalemates” (Zartman 2001). Specialists on the military dimension of the conflict in Aceh have made a strong case that a stalemate had occurred by late 2003. The May 2002 military offensive against GAM had indeed taken a great toll: GAM had been forced to retreat to the mountains and had virtually ceased offensive actions. Nonetheless, even though the commander of the Indonesian armed forces had vowed they would root out GAM down to its roots, he later reluctantly acknowledged that they could not be defeated: “we cannot do what we hoped. . . . Two die but four take their place.” However, even if there were a complete stalemate, the theory is not necessarily a sufficient explanation in itself. Stalemate does not necessarily produce peace if the long-term goals of the two adversaries are still in a zero-sum relationship. GAM could and did withdraw to the mountains with virtually all its leaders and command structure intact. The military could and did continue to occupy all the urban space of Aceh. The new equilibrium was simply a less hurting stalemate, a possibly fragile ceasefire but certainly not yet peace. The third explanation of the origins of the surprising peace was the election in November 2003 of retired military general Susilo Bambang Yudhoyono (normally referred to as SBY) as president of Indonesia and Jusuf

Revised Theory of Federacy

243

Kalla as vice president, both of whom were creative, hands-on advocates of a serious new peace process in Aceh (Morfit 2007). This was certainly extremely helpful, and a new peace process began in Helsinki in January 2003, but the emergence of these new leaders was not sufficient for the achievement of peace. A strong case could be made that unless GAM radically and credibly renounced its goal of independence, even these new leaders could not have been able to build and sustain a winning peace coalition among the Indonesian nationalist majority within the military and the parliament, whose collaboration was necessary for the craft ing of a sustainable peace. Indeed, the first round of the Helsinki peace talks on January 27–30, 2005, yielded nothing except growing pessimism. In fact, in Aspinall’s judgment, “on the eve of the second round of talks on February 21–23, 2005, the gulf between the two sides seemed as wide as ever. Collapse seemed a real possibility.” However, soon after the second round began, a spokesperson of GAM announced at a press conference on February 23, 2005: “The demand for Independence is no longer on the table. They are demanding self-government now and the Indonesians [government representatives] understand this very clearly.” In Aspinall’s opinion, “This was a shift of historic proportions. It was the first time that GAM had ever indicated that it was prepared to accept anything less than independence or a referendum [on independence]. As such, it was widely viewed as major breakthrough, and it made all subsequent progress in the talks possible” (Aspinall 2007b, 21, 26, 27). What contributed to this major compromise on the part of GAM? It would seem that the idea of federacy-type arrangements broke the bargaining deadlock. Why, how, and with what consequence? More has to be written by key participants, but it seems that on the eve of the second round, the head of the Helsinki mediation team, the former president of Finland, Maarti Ahtisaari, in a television address on February 20, 2005, alluded to “self-government” for Aceh as a possible goal of the negotiations. On February 22 the GAM spokesperson said, rightly or wrongly, that self-government “is the main thing on the table.” The next day GAM renounced independence and accepted a goal of self-government. It may be that Ahtisaari’s informal conversations with GAM were even more important than his television address. One of the GAM negotiators in Helsinki, Nur Djuli, told Stepan the following account about why he personally came to believe that an Åland Island type of arrangement could lead to selfgovernment. One evening he and some other GAM negotiators were sitting with Ahtisaari looking out at the sea. They respected Ahtisaari, who, among

244

Alfred Stepan

many other accomplishments, had been nominated for a Nobel Prize for Peace for his peace-keeping achievements in the Balkans as a high official of the United Nations. Ahtisaari noticed a ship going by and asked Djuli if he recognized the flag the ship was flying. Djuli said he did not. Ahtisaari said the flag was from the Åland Islands. Ahtisaari went on to say that the Åland Islands were a part of the unitary state of Finland with special self-governing arrangements. He said that he, as president, could not send a Finnish ship to the Åland Islands without the permission of the legislature of the Åland Islands. He said further that no major domestic law or treaty affecting the Åland Islands could go into effect without the consultation and consent of the government and legislature of the Åland Islands. Djuli insisted that he and some other GAM negotiators literally did not sleep that night. They spent the night Googling the Åland Islands and then Greenland. They may or may not have heard the word “federacy” before, but they increasingly began to feel that the Finnish-Åland arrangement might produce a serious form of “selfgovernment” for Aceh (Nur Djuli, interviews with the author, Aceh, November 2 and 4, 2007). Whatever the exact reasons for GAM’s withdrawal of its demand for independence, the bargaining situation in Helsinki was no longer a zero-sum one. A positive-sum game was now on the table. GAM believed that if it could negotiate an Åland Islands type of federacy, many of its goals concerning selfgovernment could be met. If the central government’s negotiators were able to concede such a federacy, their central goal of keeping Aceh inside the unitary state of Indonesia would be met. A mutually acceptable treaty, called the Memorandum of Understanding between the Government of the Republic of Indonesia and the Free Aceh Movement (MoU), was signed on August 15, 2005, and witnessed by Maarti Ahtisaari (see Aspinall 2007a, 75–84). The major goal of the central government was achieved in the second paragraph of the MoU’s preamble: “The parties commit themselves to creating conditions within which the government of the Acehnese people can be manifested through a fair and democratic process within the unitary state and constitution of the Republic of Indonesia.” In addition to foregoing its independence claims, GAM agreed to demobilize “all of its 3,000 military troops” and to undertake “the decommissioning of all arms, ammunition and explosives held by the participants in GAM activities” under the supervision of the newly created EU-ASEAN-led Aceh Monitoring Mission no later than December 31, 2005. The quid pro quo for GAM was new federacy pre-

Revised Theory of Federacy

245

rogatives, many quite similar to those agreed to in the Åland Islands–Finland federacy, that were to be the principles of a new Law on the Governing of Aceh. Four of our five defining characteristics of a federacy, and both of our facilitating features, are explicitly involved in the MoU.

Defining Characteristic: Federal-Type Arrangements

For GAM negotiators some of the most important self-governing prerogatives were spelled out in Article 1 of the MoU, which set out federal-type arrangements: “Aceh will exercise authority within all sectors of public affairs . . . except in the fields of foreign affairs, external defense, national security, monetary and fiscal affairs, and justice and freedom of religion.” Indonesia’s general decentralization law of 1999 had given extensive new rights to provinces, but GAM’s negotiators evidently saw Article 1 of MoU as more significant and more binding on Jakarta.

Defining Characteristic: Dispute Resolution Procedures

In most policy areas where Aceh would not be self-governing because they involved exclusive central state powers, the MoU stipulated that policies would be crafted “in consultation with and with the consent of” the relevant authorities of Aceh. Specifically, Article 1.1.2.c stipulates that “Decisions with regard to Aceh by the legislature of the Republic of Indonesia will be undertaken in consultation with and with the consent of the legislature of Aceh,” and 1.1.2.d states that “Administrative measures undertaken by the Government of Indonesia with regard to Aceh will be implemented in consultation with and with the consent of the head of the Aceh administration.” In police and military affairs an important distinction was made in the MoU between “organic” and “non-organic” personnel. The government of Indonesia, subject to binding verification by the EU-ASEAN-run Aceh Monitoring Mission, agreed to “withdraw all elements of non-organic military and non-organic police forces from Aceh” (Article 4.5). “Non-organic” personnel meant any extra central government troops sent in to bolster the numbers and capacity of the regular “organic forces.” The distinction implied a significant reduction of the number of Indonesian government security

246

Alfred Stepan

forces in Aceh. In a significant passage, the MoU stated that “the appointment of the Chief of the organic police forces and the prosecutors shall be approved by the head of the Aceh administration” (Article 1.4.4).

Defining Characteristic: Federacy Citizens Vote in all Elections of the Unitary State and the Unitary State Has Official Representation in the Federacy

Under the section titled “Political Participation” (Article 1.2), it is explicitly stated that “Full participation of all Acehnese people in local and national elections will be guaranteed in accordance with the Constitution of the Republic of Indonesia.” The MoU is silent regarding whether the unitary state would have representation in the federacy; however, it does speak of central state functions being carried out by a variety of state representatives.

Defining Characteristic: Federacy Is Part of a Single, Internationally Recognized, Independent State

Indonesia, not Aceh, is a member of the United Nations.

Facilitating Characteristic: International Guarantors May Help Set Up the Federacy

The chair of the Helsinki Peace Process that created the federacy was not an Indonesian but Maarti Ahtisaari, the former president of Finland, who was also a major UN negotiator in Bosnia. Furthermore, the MoU explicitly states in Article 5.1 that “An Aceh Monitoring Mission (AMM) will be established by the European Union and ASEAN contributing countries with the mandate to monitor the implementation of the commitments taken by the parties in this Memorandum of Understanding.” In cases of a dispute between the government of Indonesia and GAM, Article 6.1 stipulates that “The Head of the Monitoring Mission will make a ruling which will be binding on the parties.” Some of the key issues that the AMM monitored were the disarmament of GAM, the exit of the “non-organic” Indonesian

Revised Theory of Federacy

247

military forces from Aceh, and the supervision of free elections in Aceh—all crucial issues in successfully setting up the federacy of Aceh.

Facilitating Characteristic: The Federacy May Acquire Some Role in Approving and/or Structuring International Treaties Signed by the Center

The MoU may be more explicit about this than any other federacy’s agreement with the center. Article 1.1.2.b asserts that “International agreements entered into by the government of Indonesia which relate to matters of special interest to Aceh will be entered into consultation with and with the consent of the legislature of Aceh.” While they are not defining conditions, some other aspects of the MoU helped end the civil war and facilitate the construction of the federacy and the deepening of democracy. In one of the most contested and important prerogatives, the MoU stated that “understanding the aspirations of Acehnese people for local political parties[,] the Government of Indonesia will create . . . the political and legal conditions for the establishment of local political parties” (Article 1.2.1). This was crucial for GAM because Indonesia’s parliament, fearing regional fragmentation, had passed an exceptionally restrictive law mandating that all political parties must have a recognized party committee in at least 60 percent of Indonesia’s provinces and in 50 percent of the districts in those provinces. If a party does not meet these requirements it is not recognized as a legal body (correspondence with Edward Aspinall). Such a law would, of course, have made the most important ethnoregional parties in Canada, Spain, Belgium, and India—which were crucial to the construction of a state nation—illegal. Without this special prerogative, GAM, being exclusively Aceh based, would not have been able to run for the upcoming elections as a political party. With this special prerogative, however, GAM members, running as independents but clearly representing GAM, were able to enter democratic political competition for the governorship and for all district heads and mayors in Aceh in the December 2006 elections. To ensure fair elections, there was an EU election monitoring team present. To many people’s surprise, the former Aceh-based head of intelligence for GAM, Irwandi Yusuf, won the governorship. Also to many people’s surprise, the president of Indonesia sent his immediate warm congratulations. The governor’s office, especially after the MoU (and the tsunami), was worth controlling. According to the

248

Alfred Stepan

World Bank, the budget of Aceh in 2008 was around $1.5 billion US, seven times that of 1999. (Data from the World Bank’s office in Aceh.) GAM also won about a third of the mayorships in Aceh (Aspinall 2007b). I visited a former GAM exile, Dr. Nurdin, who was elected district head of Bireuen, one of the most militarily contested war zones of Aceh. Dr. Nurdin said the process was working reasonably well, despite the local history of great conflict. Above his desk he had the flags of Aceh and Indonesia, as well as the standard photograph of the president of Indonesia that one sees in every public office. In addition to these federacy-like prerogatives, following the MoU (Section 3), a number of civil war–ending provisions were offered to GAM, such as amnesty, release of all political prisoners, and reintegration funds to be administered by GAM. Four months after the Helsinki agreement was signed and the deadline for disarmament arrived, one of Aceh’s most respected observers, Sidney Jones, released a report declaring “the Aceh peace process is working beyond all expectations. Guerrillas of the Free Aceh Moment, GAM, have turned in the required number of weapons. The Indonesian military, TNI, has withdrawn troops on schedule. The threat of militia violence has not materialized. Amnestied prisoners have returned home without incident. The International Aceh Monitoring Mission, AMM, led by the European Union’s Peter Firth, has quickly and professionally resolved the few violent incidents between GAM and the TNI. . . . The peace process has active support of the Indonesian government” (International Crisis Group 2005, 1). GAM was both proactive and active in the peace process. For example, one of the most senior former GAM commanders, Darwis Jeunieb, turned over to the police a former GAM soldier who was still trying to enforce GAM’s now renounced “intelligence tax,” and GAM leaders were in charge of the compensation program for the demobilized GAM troops (ibid., 6, 4; see also Aspinall 2005). In a survey conducted in 2006 among 1,075 former GAM militants by Macartan Humphreys, a professor of political science at Columbia University, former GAM militants were asked to indicate their degree of satisfaction or dissatisfaction “with MoU implementation.” “Satisfied” former GAM militants outnumbered “dissatisfied” former GAM militants nine to one: 46 percent were “very satisfied,” 44 percent were “satisfied,” 7 percent were “not satisfied,” 2 percent were “very dissatisfied,” and only 1 percent refused to answer. The prospect of a federacy, and of peace, was an acceptable alternative to the former fighters for independence. This chapter would be imbalanced, however, if I did not point out how Aceh falls short of the ideal-type federacy I discussed at the beginning of this essay.

Revised Theory of Federacy

249

Aceh is the only case of federacy-type arrangements being constructed by two forces actively engaged in civil war against each other. This contributed to one major limitation. The MoU was a peace agreement signed by the government of Indonesia and GAM. But in the new democracy of Indonesia, MoU could not become a law, much less be constitutionally embedded, until the parliament ratified it. The sitting parliament, without GAM representatives— because GAM as a revolutionary force turned provincial political party only participated in parliamentary elections for the first time in 2009—passed the Law on the Governing of Aceh (LOGA), which incorporated many but by no means all parts of the MoU (Republic of Indonesia, Law of the Republic of Indonesia, Number 11, 2006, Regarding Governing of Aceh). The most important watering down was that instead of LOGA stipulating that laws, treaties, or administrative measures affecting Aceh must be with the “consultation and consent” of the relevant Aceh authorities, only the word “consultation,” not “consent,” was included. Also, in our ideal type of federacy the arrangements should be constitutionally embedded. However, to GAM’s great disappointment, LOGA is a law that can be amended by an ordinary majority in the lower house of the parliament, the DPR. Nur Djuli felt particularly betrayed by the changes made by the Indonesian parliament and constructed and circulated a “matrix” illustrating how some key guarantees of autonomy were weaker in the LOGA than in the MoU. Indeed, the danger that the Indonesian parliament might attempt to erode MoU further was evident in January 2008 when (led by former president Megawati Sukarnoputri’s coalition-building efforts among opponents of MoU) the parliament passed a law recommending that in a future reorganization of Indonesia’s administrative units some units now a part of Aceh be transferred out of Aceh. This would be a violation of not only MoU but also LOGA. The bill, however, never went into effect because President Yudhoyono did not sign it (International Crisis Group 2008, 6–8). At the very least, I have to acknowledge that my second defining condition of a federacy, constitutional embeddedness, is not fully met. LOGA, by what constitutional theorists call embedded political “convention” (Marshall 1984), might have become de facto constitutionally embedded and therefore required exceptional majorities to amend. But the Megawati initiative indicates that she and her supporters did not view such a convention as binding. Another disturbing event, in circumstances still not explained, was the assassination of Nur Djuli in February 2009. In future studies of federacy, more consideration of the par ticu lar asymmetrical bargaining powers of armed secessionists versus the regime

250

Alfred Stepan

government in Time 1 (during negotiations) and in Time 2 (after the secessionists’ disarmament and participation in elections) should be analyzed. During Time 1, both the government and the secessionists may have strong incentives and capacities to make federacy arrangements. However, during Time 2, if the former secessionists want to return to armed violence, their disarmament, demobilization, and the knowledge of who their once secret members are will have greatly decreased their capacity to launch another rebellion. Also, because many of their most effective leaders may now be in important public positions of power, they will have less incentive to resist. Asymmetrically, the government’s relative capacity to resist would be increased. Despite the above reservations, I think the evidence presented in this essay strongly indicates that the surprising emergence of social peace and political incorporation in Aceh would not have been possible without the idea of federacy and without the utilization of many important federacytype arrangements in Indonesia. “Federacy” as a concept, and as a set of institutional arrangements, should be added to the repertoire of policies that should be considered in deeply divided places for questions of civil war termination, or avoidance, and for democratization. Notes 1. This essay builds upon a chapter on numerous existing and possible federacies in Stepan, Linz, and Yadav 2010. Juan J. Linz should be considered the intellectual coauthor of the section that follows on the theoretical concept and empirical characteristics of federacies. I carried out the fieldwork in Aceh. 2. India’s asymmetrical federal system has made provisions for Mizoram to protect its way of life through numerous measures, e.g., restrictions of some types of landownership for non-Mizos, and restricting eligibility for municipal voting and office holders to “Mizos only” (Stepan, Linz, and Yadav 2010, chap. 7). Such restrictions on the rights of citizens of the state, because they are not members of a par ticu lar ethnic or cultural group, violate the laws or legal traditions of most democratic unitary states. For Mizoram there are also extensive protective provisions for the use and development of the Mizo language in the educational system. 3. Some important constitutional theorists argue that political traditions, and a sense of what is politically appropriate, can give regular law de facto constitutionally constraining qualities, even without written constitutional status (Marshall 1984). Such constitutional accretion has occurred in some federacies, e.g., Denmark’s parliament accepts the de facto constitutional convention that its federacies have a right to negotiate their exit from Denmark. However, conventions are normally only produced over time so my focus here is on written constitutionally embedded agreements.

Revised Theory of Federacy

251

4. Bhutan retains the right to reject Indian security advice. For example, in 1961– 62, during a time of great Sino-Indian tension, Bhutan unilaterally denied India permission to guard the Bhutan-Tibetan border. 5. Also useful on GAM are Reid 2006, especially Nessen 2006. For analysis of the military and ideological dimensions of GAM as a separatist movement, see Schulze 2004. 6. Clark and Palmer (2008) use the word “dubious” to describe Indonesia’s democracy in order to focus on the poor quality of election procedures and practices. In the lead-up to the 2009 elections there were no signs of the resurgence of the civil war, but there was some use of intimidation and violence by many participants in the electoral struggle. 7. Statement of Armed Forces Chief Endriartono Sutarto cited in Aspinall 2005, 12. Although Sutarto did not make this public statement until 2005, the evidence suggests this was the situation by late 2003. 8. Djuli was a significant figure in many stages of the peace process. He was member of the GAM team that led the aborted 2002 Cessation of Hostilities Agreement and was later critical of its failure to address basic political differences. After being a member of the much more successful MoU negotiating team, Djuli was the architect of a GAM draft of the future Law on Governing Aceh. At the time of the interview Djuli was the head of BRA, the World Bank–supported Aceh Reintegration Board (International Crisis Group, 2007, 11–12). 9. Ahtisaari was awarded the Nobel Peace Prize in 2008, partly because his work was “central to the solution of the complicated Aceh question in Indonesia.” See http:// nobelprize.org /nobel _prizes/peace/laureates/2008/press.html. 10. Unpublished results provided to Alfred Stepan by Macartan Humphreys on 6 March 2009. References Aspinall, Edward. 2002. “Modernity, History and Ethnicity: Indonesian and Acehnese Nationalism in Conflict.” Review of Indonesian and Malaysian Affairs 36, 1: 3–33. ———. 2005. “The Helsinki Peace Process: A More Promising Basis for Peace in Aceh?” Policy Studies 20. East-West Center, Washington, DC. ———. 2007a. “From Islamism to Nationalism in Aceh, Indonesia.” Nations and Nationalism 13, 2: 245– 63. ———, ed. 2007b. “Special Issue: Aceh: Two Years of Peace.” Inside Indonesia 90 (October–December). ———. 2009. Islam and Nation: Separatist Rebellion in Aceh Indonesia. Stanford, CA: Stanford University Press. Aspinall, Edward, and Harold Crouch. 2003. “The Aceh Peace Process: Why It Failed.” Policy Studies 1. East-West Center, Washington, DC. Clark, Samuel, and Blair Palmer. 2008. “Peaceful Pilhada, Dubious Democracy: Aceh’s Post-Conflict Elections and Their Implications.” Indonesian Social Development Papers, No. 11. Jakarta: World Bank.

252

Alfred Stepan

Elazar, Daniel. 1991. Federal Systems of the World: A Handbook of Federal, Confederal and Autonomy Arrangements. Harlow: Longman. Freedom House. 2008. Freedom in the World: Political Rights and Civil Liberties: 2008. New York: Freedom House. Available at http://freedomhouse.org /. Hirschman, Albert. 1972. Exit, Voice and Loyalty: Responses to Declines in Firms, Organizations and States. Cambridge, MA: Harvard University Press. International Crisis Group. 2005. “Aceh: So Far, So Good.” Asia Briefi ng No. 44. 13 December. ———. 2007. “Aceh: Post Conflict Complications.” Asia Report No. 139. 4 October. ———. 2008. “Indonesia: Pre-Election Anxieties in Aceh.” Asia Briefing No. 81. Jakarta/Brussels, 9 September. Kingsbury, Damien. 2006. Peace in Aceh: A Personal Account of the Helsinki Peace Process. Jakarta: Equinox Publishing. Liddle, R. William, and Saiful Mujani. 2009. “Indonesian Democracy: From Transition to Consolidation.” Paper for the international conference “Islam and Democracy in Indonesia: Comparative Perspectives.” Center for Democracy, Toleration, and Religion, Columbia University, 2–3 April. Marshall, Geoff rey. 1984. Constitutional Conventions: The Rules and Forms of Political Accountability. Oxford: Oxford University Press. Morfit, Michael. 2007. “The Road to Helsinki: The Aceh Agreement and Indonesia’s Democratic Development.” International Negotiation 12, 1: 111– 43. Nessen, William. 2006. “Sentiments Made Visible: The Rise and Reason of Aceh’s National Liberation Movement.” In Anthony Reid, ed., Veranda of Violence: The Background to the Aceh Problem. Singapore: Singapore University Press. Reid, Anthony, ed. 2006. Veranda of Violence: The Background to the Aceh Problem. Singapore: Singapore University Press. Rezvani, David A. 2004. “Federacy: The Dynamics of Semi-Sovereign Territories.” D. Phil, University of Oxford. Rustow, Dankworth. 1970. “Transitions to Democracy: Toward a Dynamic Model.” Comparative Politics 2, 3: 337– 63. Schulze, Kirsten E. 2004. The Free Aceh Movement (GAM): Anatomy of a Separatist Organization. Washington, DC: East-West Center. Stepan, Alfred. 2008. “Comparative Theory and Political Practice: Do We Need a ‘State Nation’ Model as Well as a ‘Nation State’ Model’?” Government and Opposition (Winter): 1–30. Stepan, Alfred, Juan J. Linz, and Yogendra Yadav. 2010. Democracies in Multinational Societies: India and Other Polities. Baltimore: Johns Hopkins University Press. Zartman, I. William. 2001. “The Timing of Peace Initiatives: Hurting Stalemates and Ripe Moments.” Global Review of Ethnopolitics 1, 1: 8–18.

CHAPTER 10

We Forbid! The Mutual Veto and Power-Sharing Democracy Joanne McEvoy

On July 22, 2008, Serbian security forces surprised the world by finally arresting Bosnian Serb wartime leader Radovan Karadžić, captured on a bus as he traveled about Belgrade disguised as a New Age guru. The International Criminal Court for the former Yugoslavia (ICTY) had indicted Karadžić twelve years earlier on eleven counts of genocide, complicity in genocide, war crimes, and crimes against humanity during the Bosnian wars of 1992–95. The allegations included organizing the siege of Sarajevo where ten thousand people died and orchestrating the massacre at Srebrenica in 1995 where the Bosnian Serb army killed around eight thousand mainly Muslim men and boys. In 2007, with Karadžić and his war general Ratko Mladić still at large, the International Court of Justice (ICJ) ruled that Serbia had violated its obligations to prevent genocide and that it should transfer indicted war criminals to the ICTY. Yet progress remained elusive. Political elites in Bosnia and Herzegovina representing the Bosniak (Muslim) and Bosnian Croat communities grew increasingly impatient with Serbia’s failure to capture Karadžić and Mladić. The Bosniak member of the tripartite state presidency, Haris Silajdžić, backed by his Bosnian Croat counterpart, Željko Komšić, proposed to ask the UN Security Council to make Serbia comply with the ICJ’s ruling. But Bosnian Serb presidency member Nebojša Radmanović vetoed the proposal. It was then referred to the Republika Srpska (RS) National Assembly, and a majority of delegates supported Radmanović’s position that such a move would threaten Bosnian Serbs’ “vital national interests.” A war of words between

254

Joanne McEvoy

the Bosniak and Bosnian Serb political leaders ensued: Silajdžić claimed the RS was acting as Serbia’s “direct agent” while Radmanović accused the Bosniak leader of using the court’s ruling to strengthen his political influence. It is not surprising that the indictment of Karadžić and Mladić and their evasion of the authorities were highly contentious issues in Bosnia, Serbia, and the Western Balkans more generally. The episode also demonstrates how in deeply divided places unresolved issues from violent conflict can be framed in the vocabulary of a group’s “vital national interests” (VNI). When a group feels that a decision by the executive threatens its fundamental interests (e.g., issues relating to culture, language, education), it may be able to veto this decision in the name of minority group protection. Thus a major challenge for power-sharing systems is to provide protection for groups while preventing increased interethnic tension and unproductive stalemate among the governing parties. In addition to providing rules to protect a group’s vital national interests, power-sharing systems can be designed to protect groups by requiring a concurrent majority in the legislature: that is, a majority of delegates present and voting including a majority of members from the minority community. Such rules might also specify the thresholds for groups’ approval. Problems can occur, however, when representatives within a group choose to veto a decision to demonstrate their opposition to party policy. In November 2001, such use of the cross-community voting rules in the Northern Ireland assembly precipitated a crisis in the power-sharing system set up under the Good Friday Agreement of 1998. On a vote on the reelection of Ulster Unionist leader David Trimble as first minister alongside Mark Durkan, leader of the national Social Democratic and Labour Party (SDLP), as deputy first minister, two Ulster Unionist Party (UUP) members voted with the anti-agreement Democratic Unionists and against their own party, denying the cross-community support required for the election of the top two posts. Trimble had resigned from the executive months earlier over the lack of progress on the destruction of republican weapons but agreed to enter into power sharing after some IRA decommissioning in October. Despite some opposition within Trimble’s party, the executive was restored when members of the bi-communal Alliance Party redesignated from “other” to “unionist” and secured the vote (see Godson 2004, 696–700; Purdy 2005, 297–301). In this chapter I investigate the operation of the mutual veto, one of Arend Lijphart’s (1977) four elements in the original formulation of consociational democracy. The argument starts from the premise that the mutual veto may

We Forbid

255

be and often is an important and necessary institutional rule in a powersharing democracy. I address the institutional design of various veto rules and explore their impact on power-sharing practice. To date there has been no concentrated focus in the literature on the design and operation of the mutual veto in power-sharing systems. An argument might be made that this focus avoids accounting for the impact of the system as a whole and that it is ill-advised to focus on one institution in isolation. I suggest, however, that there is a need for academics and policymakers to explore the effects of various veto procedures. This microfocus on the mutual veto contributes to the book’s wider investigation of how constitutional design should respond to the political challenges in deeply divided places. Veto rights are arguably a crucial aspect of peace agreements and postconflict institutional design. If veto rights are misused they may lead to stalemate within the power-sharing executive and risk collapse of the political system. If they are used appropriately, they may help foster cooperation among governing parties. First, I provide an overview of the veto in political science, its historical precedents, and its treatment in the academic literature. Second, I explore how the mutual veto works in power-sharing systems by offering an analytical framework encompassing three constitutive elements of the procedure: the who, what, and where of the mutual veto. These elements correspond to veto players, veto issues, and veto points. Third, I use the analytical framework to compare the operation of various veto rules in Bosnia and Northern Ireland. I argue that although veto rights are crucial for groups to feel a sense of security in the new political arrangements following conflict (and arguably necessary for them to sign up for the deal), the specific veto procedures can make a difference in fostering cooperation among elites representing the contending groups. Finally, I argue that less is more, and I make some recommendations for a more refined mutual veto in terms of the scope of issues groups may be able to veto and the extent of veto procedures.

From the Roman Republic to Twenty-First-Century Conflict Regulation The practice of mutual veto powers is an underresearched and undertheorized aspect of power-sharing democracy. Of course, veto powers are not a new phenomenon, specific to power-sharing systems. The veto (Latin: I forbid) is an age-old political practice with historical antecedents stretching all

256

Joanne McEvoy

the way back to ancient Rome. In the fift h century B.C. the plebeian tribunes had the power to veto (intercessio) the acts of consuls and lower magistrates. The tribunes’ duty was to protect the people against acts of the magistrates that threatened their interests. According to Andrew Lintott, the possibility of one tribune vetoing his colleagues “would have frequently been a stimulus to consultation, but there was nothing abnormal in conflict per se.” He writes, however, that the tribunician intercessio veto “was an almost indefeasible means of obstruction against both magistrates and senate” (1999, 99–100). Though it was also employed by members of the ruling class (not its original purpose), the tribune veto allowed lower-class Romans “to make a significant impact in politics” (202). The veto in ancient Rome provided a precedent for the liberum veto (I freely forbid) in the Polish-Lithuanian Commonwealth. First officially used in 1652, delegates could trigger the veto in the Sejm to end the session and nullify its legislation. Jerzy Lukowski writes that the liberum veto “rapidly established itself as the keystone of Polish liberty, the triumph of the sovereignty of the individual over the sovereignty of the state” (1991, 7). The veto was used as a safeguard against potentially harmful laws enacted by the Polish monarchy. Yet the veto came to be criticized because “Technically, the veto put a moratorium on the whole session, so that even already agreed bills could not be enacted nor their final reading take place” (91). Following much debate on the virtue of the veto and its potential disruptiveness to sessions, the procedure was abolished with the adoption of the 1791 constitution. Veto powers are a key feature of democratic institutions in modern political systems. In separation of powers regimes, veto powers provide a central aspect of the checks and balances between the executive and the legislature. An extensive literature ranges from the debates between presidential and parliamentary systems (Linz 1990), to formal modeling of how separation of powers improves elite accountability via appropriate checks and balances (Persson, Roland, and Tabellini 1997), to a focus on how the separation of powers impacts party organization and behavior and creates “presidentialized” political parties (Samuels 2002), as well as analysis of veto powers in European Union decision-making processes (Tsebelis and Yataganas 2002). George Tsebelis’s work has advanced conceptual thinking on comparative government and veto players. In Veto Players: How Political Institutions Work, he focuses on how governments are structured, making distinctions in the extent to which political actors have veto power over policy choices. He argues that an increase in the number of veto players or

We Forbid

257

an increase in their ideological distance from each other increases policy stability, impeding significant departures from the legislative status quo. Tsebelis’s veto players may be institutional (the president or parliament) or partisan (parties in a coalition government). My focus here, however, is the operation of veto rights that are designed to protect the vital interests of groups previously in conflict. As noted above, the mutual veto is one of the four elements of consociational power sharing (the other three elements are grand coalition, proportionality, and segmental autonomy). For Lijphart, the mutual veto prevents the mutual from being outvoted by the majority. He argues that only the mutual veto can give each segment a complete guarantee of political protection. While Lijphart (1977) recognizes that veto powers may strain cooperation in the grand coalition, he suggests the situation is not as bleak as it might appear for three reasons. 1. The veto is a mutual veto, so its frequent use is unlikely as it can be turned against a minority. 2. The veto is a potential weapon for the minority and gives them a sense of security, so its use is improbable. 3. Groups will refrain from using the veto because they will recognize the danger of deadlock and immobilism. More recently, Lijphart revised the consociational model, positing the mutual veto and proportionality as secondary elements that reinforce the two main characteristics of executive power sharing and segmental autonomy (Lijphart 2004). Despite this revision, the mutual veto is considered a crucial aspect of power-sharing democracy with the potential to promote cooperation among rival ethnic elites. As Sujit Choudhry notes, the mutual veto is an incentive for “ethnic leaders to participate within politics, because they can rest secure that their fundamental interests will be protected” (2008, 19). While power-sharing theory confers much potential in the mutual veto for incentivizing cooperative behavior among elites, there is a surprising lack of empirical research on whether this is the case in the real world of power sharing. Ulrich Schneckener (2002) seeks to distinguish “better” powersharing regulations from less favorable ones, noting the importance of elite behavior shaped by institutions. He confirms the importance of veto rights as a typical institutional device in power-sharing systems. He writes that when “each group has the opportunity to block political decisions by using its veto rights,” the “aim is to foster consensus-building and the search for

258

Joanne McEvoy

compromises” (2002, 205). He distinguishes between various institutional options for veto rights: delaying veto; indirect veto; and direct veto. Schneckener’s distinction is helpful in illustrating the diversity of veto procedures. Some mechanisms require the decision to be reconsidered (delaying veto) via parliamentary mediation or referral to the constitutional court; others require certain conditions to be met, such as a concurrent majority in the assembly (indirect veto); and others allow groups to declare an issue pertinent to their “vital interest” and block the decision (direct veto). For Schneckener, the first two types of veto rights are preferable, and he suggests that “veto rights should be more restricted in order to prevent their misuse” (2002, 221). He notes that restrictions are necessary “since veto rights should foster and not prevent consensus-building” (222). Yet he does not say how these rights should be restricted, other than noting that they should not pertain to every issue. He warns against frequent use of veto rights, which “can actually be seen as a sign of severe crisis” (222). But he does not suggest how veto rules might be designed to prevent their frequent use. In a similar vein, Florian Bieber says veto procedures “constitute an important aspect of power-sharing arrangements, but at the same time they can have the most serious repercussions” (2005, 95). Bieber says there are qualifications to veto rules to limit the danger of immobilism. He categorizes these limitations in terms of the degree of support required to veto a decision; restrictions on the field of legislation subject to veto; and parliamentary mediation procedures to overcome conflict once the veto has been invoked. Bieber argues that on the basis of his analysis of the legal framework, these limitations to veto rights “do not accomplish much” and notes the need for an investigation of the effects of veto powers (100).

An Analytical Framework for the Mutual Veto There is much agreement among academics that restricted veto rights are a good idea. But we still need to know how veto rights should be restricted and the most appropriate way to design veto rights in deeply divided places. Arguably academics and policymakers still need to know whether there is a “better” institutional design for the mutual veto in power-sharing systems. Who should have veto power? What issues should be subject to veto? What are the appropriate procedures in the assembly and the executive? To understand better how the mutual veto works, I suggest an analytical framework

We Forbid

259

that encompasses three elements of the mutual veto in power-sharing systems: veto players (who has veto power); veto issues (what they can veto); and veto points (where they can veto). In terms of veto players, my framework differs slightly to that of Tsebelis (2002). For Tsebelis, veto players may be individual or collective, including formal institutions. Instead, I treat veto players as actors who have the capacity to block decisions in the name of group protection. There are two ways in which such groups are identified: by predetermination or selfdetermination. They may be predetermined groups specified in the peace agreement/constitution or self-determined groups on the basis of electoral success. The advantage of predetermination is that ethnic groups are guaranteed inclusion in executive power sharing. This will matter a great deal for groups that were excluded from government in the past. The advantage of self-determination is that it is up to the electorate to identify with and vote for communal or noncommunal parties, thereby determining the groups that enter power sharing. The literature suggests that self-determination of groups is preferable to predetermination. Lijphart (1995, 284–86) argues that there are several advantages to self-determination of groups: it avoids discriminatory choices among groups; it avoids the assignment of individuals to groups who may wish to self-identify otherwise; it gives equal chances to all ethnic or other segments; it avoids fi xing groups’ representation on a permanent or semipermanent basis; and self-determination is preferable to predetermination as mobility decreases groups’ geographical concentration. There is also a view that predetermined consociations are incompatible with the human rights of political participation under the International Covenant on Civil and Political Rights and that assigning a political office to a certain group “is an act of direct discrimination, prohibited by international human rights instruments” (Wheatley 2005, 163). John McGarry and Brendan O’Leary (2007) phrase this debate differently and note the distinction between corporate consociation and liberal consociation. A corporate consociation accommodates groups according to ascriptive criteria whereas a liberal consociation allows groups to self-determine their organization and representation (McGarry, O’Leary, and Simeon, 2008, 61– 62). Once identified, the pre- or self-determined groups then exercise their veto power via delegates in the assembly, ministers in the executive, and even presidents and vice presidents. In the assembly, institutional rules determine the proportion of a group’s delegates required to trigger the mutual veto. In Northern

260

Joanne McEvoy

Ireland, 30 out of 108 members of the Legislative Assembly (MLAs) can invoke a “petition of concern” for a decision to be taken on a cross-community basis. In Bosnia, a majority of Bosniak, Bosnian Serb, or Bosnian Croat delegates may trigger the VNI veto. Having considered who has the power to veto, an analytical framework needs to explore what issues these groups can veto. Veto issues are the decisions or legislation that groups may choose to block in power-sharing systems. The dilemma for policymakers is whether these issues should be defined in the constitution so groups know exactly what they are or whether the issues should be open to interpretation. In other words, do policymakers need to limit the scope of veto powers, or should groups be able to determine when a decision impacts their fundamental interests? Certainly important policy areas such as the budget may be subject to veto, requiring support from the main groups in parliament. Groups may also wish for areas subject to veto rights to be protected in the constitution. Such areas may include culture, language rights, and symbols. The important point is that the groups should agree on the policy issues that will be subject to veto rules to minimize dispute over the use of the procedure. Veto issues should not be imposed by external actors involved in the design of the peace agreement and/or constitution. Ideally, domestic elites should have responsibility for the negotiation of mutual veto provisions. Such rules will arguably have greater legitimacy as the outcome of interethnic compromise. Part of the problem with the Dayton Peace Agreement, after all, is that it was largely externally imposed and thus suffered from a legitimacy deficit (see Cousens 2002). It should be up to the parties in negotiations (and likely facilitated by an external actor) to agree on the areas of legislation that may impact their fundamental interests. While the particular issues may depend on the context and nature of divisions, there should be some restriction and, beyond agreement on the budget, they need not extend to more “normal” socioeconomic policy areas that do not have a direct bearing on groups’ interests defined in ethnic, religious, or cultural terms. Yet what happens when a group claims that an ad hoc issue impacts their national interests? On such controversial issues, groups may have recourse to parliamentary mediation to contest the legality of an invoked veto. In Bosnia, for example, when a majority of a constituent people’s delegates object to an invoked VNI, the chair of the House of Peoples convenes a joint commission comprising three delegates (one Bosniak, one Croat, and one

We Forbid

261

Serb) to resolve the issue. If the joint commission is unable to resolve the issue, the matter is referred to the Constitutional Court for review. Another solution might be for a group to opt out of another group’s proposed policy. McGarry, O’Leary, and Simeon distinguish between “opt-out provisions that do not block others from pursuing a certain policy and that result in institutional and policy asymmetries from veto rights that prevent majorities from pursuing their preferred course of action even within their own jurisdictions” (2008, 60). While an opt-out may allow a group to go ahead with their chosen policy and allow another group to proceed in an alternative direction, policy asymmetries among groups may well have contentious political and budgetary consequences and would arguably be less preferable to a moderated compromise on a shared policy among the different sides. The third element of the analytical framework concerns veto points, where groups have the capacity to veto the agreed issues covering vital interests. Veto procedures may exist in the assembly and/or the executive. In a bicameral system, institutional designers need to decide whether both chambers should have mutual veto rules. For a collective presidency, it makes sense that members of the presidency should agree on policy decisions, that there should be consensus among members, and that one group should have the capacity to block another group’s proposal if it threatens its national interests. Veto rules may also be exercised within the council of ministers. The objective of executive power sharing is for ministers to reach policy decisions on the basis of consensus. When a minister brings controversial proposals to the executive table, his/her ministerial colleagues should have the capacity to veto proposals if they threaten group interests. These rules should then incentivize members of the collective presidency and executive ministers to moderate their position and compromise with other groups in order to reach agreement. In the assembly, a specified number of delegates may have the capacity to veto a decision or refer it back to the executive for review. As noted above, there may also be a parliamentary mechanism for resolving disputes, deciding whether an issue should be subject to veto procedures in cases of ambiguity or hybrid legislation. The courts may also have the capacity to review procedures and render judgment on the legality of a group’s decision to veto legislation. While there is justification for veto procedures at both the assembly and executive levels, the literature is not clear on the extent of these procedures. Should there be some limitation on the number of veto points as well as veto

262

Joanne McEvoy

issues? An argument can be made that too many veto points can cripple the system. To promote efficient, streamlined decision making, it may be preferable for veto rights to be focused in one chamber in a bicameral system. The challenge for power-sharing institutions is to make executive decision making flexible without denying minority protection on important issues. Fewer veto points where mechanisms are focused and provide clarity may be better than a plethora of veto points that render the system overly complex, facilitating misuse by elites wishing to create obstacles for executive decision making. Employing this framework can help us investigate whether the mutual veto incentivizes groups’ political participation and interethnic cooperation as parties are confident the interests of their respective groups are protected. But can we argue that some veto arrangements are “better” than others? To demonstrate the variation in veto effects, I compare two power-sharing cases with veto rules: a set of complex veto procedures in Bosnia and Herzegovina, and a less rigid set of arrangements in Northern Ireland.

Reforming Complex Veto Powers in Bosnia and Herzegovina Veto powers have been an issue of considerable controversy in post-Dayton Bosnia. At the state level, members of the tripartite presidency can veto a decision that contravenes their respective group’s VNI. A majority of delegates representing each constituent people can veto legislation in the state parliament on the basis of their VNI. If a majority of Bosniak, Croat, or Serb delegates object to the veto, a joint commission is convened to resolve the issue. If unsuccessful within five days, the commission refers the matter to the Constitutional Court for review. In addition to the VNI procedure, decisions in the state parliament require that majority support includes at least one-third of delegates from each entity. Thus, in addition to the VNI veto, the power-sharing structures in Bosnia have an entity veto. In practice, this means that only ten members from Republika Srpska in the House of Representatives need to withhold their support to veto legislation. In the entities, two-thirds of a constituent people’s delegates can veto legislation in the respective assembly. The entity constitutions list areas of legislation subject to the VNI veto but also include a provision whereby two-thirds of a group’s delegates can trigger the veto on any issue.

We Forbid

263

Some scholars have noted the challenges posed by the use of veto powers in Bosnia (Bieber 2005; Zahar 2005). Marie-Joëlle Zahar writes that mutual vetoes “tend to carry with them the risk of ‘immobilism’ or state paralysis which can only be countered by the concerted action of responsible leaders willing to compromise in order to maintain stability” (2005, 127–28). As she argues, in Bosnia “elite intransigence or unwillingness to cooperate are more often than not the norm” (128). To acknowledge the pivotal role of political leadership, institutional design must identify arrangements that incentivize leaders’ willingness to cooperate or at least to constrain elite intransigence. Sofia Sebastián notes that “the many provisions aimed at protecting ethnic interests have . . . slowed down the state-level decision-making process to the point of outright paralysis at times” (2007, 3). The record of power sharing in Bosnia suggests that the misuse of veto powers by intransigent political elites has been an obstacle to efficient decision making. Some recent examples of veto rules triggered by groups include the attempt by Bosniak deputies in the RS assembly to block the Law on Holidays in May 2007, which was overruled by the Constitutional Court. In August 2006, Bosniak deputies in the RS argued that the privatization of Telekom Srpska threatened their VNI. The same year, Bosniak delegates in the state House of Peoples attempted to veto an agreement between Bosnia and Croatia whereby the latter would extend financial assistance to war victims in Bosnia who were members of the Croat Defence Council during the war. In March 2008, Bosnian Serb presidency member Nebojša Radmanović vetoed a proposal by his Bosniak and Bosnian Croat counterparts, Haris Silajdžić and Željko Komšić, to ask the ICTY to transfer its archives to Bosnia when it closes. And there have been a number of vetoes triggered by Bosnian Croats and Bosniaks on the reform of public broadcasting. In 2006 Bosnian Croats vetoed the draft Law on Public Radio Television System as they wanted to establish a channel in the Croat language. The veto was rejected by the Constitutional Court. In April 2008, Bosnian Croats and Bosnian Serbs agreed to ask the Council of Ministers to initiate a procedure for establishing a Croat channel. The Bosniak delegates in the House of Peoples declared this decision destructive of their VNI, a position rejected by the Constitutional Court on procedural grounds. In addition to the formal powers set out in the Dayton Agreement and the entities’ constitutions, elites in Bosnia have exercised a more informal veto by simply not showing up to participate in the institutions, thereby delaying the passage of legislation or frustrating decision making in the Council of

264

Joanne McEvoy

Ministers. The slow rate of progress on decision making led High Representative Miroslav Lajčák to announce reforms in October 2007 that would allow Council of Ministers sessions to proceed if a majority of members are present and voting in order to prevent blockage by absenteeism (OHR 2007). It is perhaps unsurprising, then, that the ongoing process of constitutional reform in Bosnia has included a focus on reforming decision-making procedures including the VNI procedure and the provision for entity voting. In March 2005, the European Commission for Democracy Th rough Law (Venice Commission 2005) published a document that has helped guide subsequent discussions on constitutional reform. The Venice Commission identified veto procedures as particularly problematic and in need of urgent reform. It stated that the vital interest veto “entails a serious risk of blocking decision-making” (2005, 9). The Venice Commission concluded that there is a need to clarify and restrict the scope of the VNI rules. It argued: “The main problem with veto powers is not their use but their preventive effect” as they may deter delegates from tabling proposals that are likely to be vetoed. Thus, “Due to the existence of the veto, a delegation taking a particularly intransigent position and refusing to compromise is in a strong position.” Indeed, being in such a strong position will hardly incentivize parties to cooperate and reach a compromise. The Venice Commission also believed that there was little justification for entity voting in the state parliament, which appears “redundant” given the existence of the VNI veto. Largely driven by external actors, reform of the Dayton structures has become a key priority for the EU. The European Commission has recently stated that institutional reform in Bosnia “is necessary before the Commission can recommend the granting of candidate status.” At an event marking the tenth anniversary of the Dayton Agreement, the political parties agreed to pursue discussions on reforming the constitution. Donald Hays and Jason Crosby recall that “the party leaders confi rmed their general agreement that the House of Peoples would handle VNI issues on a defined set of topics” (2006, 8). The Bosniak parties, the Party of Democratic Action (SDA) and the Party for Bosnia (SBiH), wanted to abolish entity voting while the Serb parties made clear that agreement was conditional on the retention of this procedure (Hays and Crosby 2006, 9). It is perhaps not surprising that Bosniak politicians want to see an end to entity voting; one delegate remarked that the entity veto in the House of Representatives is used mostly by Bosnian Serbs. By March 2006 the parties concluded their discussions on constitutional reform with a set of proposals dubbed the “April package” and presented

We Forbid

265

them to parliament for ratification. The package included agreement that the VNI procedure would be concentrated in the House of Peoples. The proposed amendments did not change the procedure very much, other than requiring the group invoking the veto to provide a written explanation and for any appeal from another group to be referred directly to the Constitutional Court rather than first being considered by a joint commission with a representative from each constituent people. In the reform package there was an attempt to clarify the scope of issues under the VNI, with “national symbols and flags” and “preservation of the integrity of Bosnia and Herzegovina” added to the list of areas. The provision did, however, retain the right for two-thirds of one of the caucuses in the House of Peoples to declare any issue to be a vital national interest. Moreover, the package did not eliminate entity voting as legislation in the House of Representatives would still require majority support of members present and voting, including one-third of members from each entity. The retention of the entity veto in the House of Representatives was part of the reason for the rejection of the April package, which failed to secure sufficient support in parliament by just two votes. The maintenance of entity voting in the reform package was a major stumbling block for SBiH leader Haris Silajdžić, who accused SDA leader and then Bosniak presidency member Sulejman Tihić of “selling out” Bosniak interests. Silajdžić rejected the package as he has favored the abolition of the entity structures, not their maintenance and continuation of entity voting rights in the state parliament. In the aftermath of the 2006 failed attempt to reform the Dayton constitution, several key external actors proposed reform of veto powers to improve the functionality of the state structures. According to Hays and Crosby, “Defining and restricting the circumstances in which the VNI veto could be used would remove one of the greatest obstacles to efficient and effective governance in Bosnia” (2006, 4). Acknowledging the importance of veto rights for the parties who want to say they can stop their rivals, international actors suggest the frequent use of the VNI procedure in the House of Peoples is employed as “an obstacle to getting anything done.” Deputy High Representative Raffi Gregorian maintains that the parties invoke the vital national interest veto “on absolutely anything,” which “creates a huge delay.” He suggests a need to “re-vamp the constitution or to limit or better define the issues under vital national interests.” There is a view among international actors that the April package was a step in the right direction, that there is an urgent need for progress, and that reform of entity voting

266

Joanne McEvoy

would be part of the process to facilitate decision making pertaining to EU integration (Sebastián 2007, 6). The stalemate following the failure of the April package led the high representative in Bosnia, Miroslav Lajčák, to announce streamlining reforms to stop “the praxis of delay and creating complications through pure absence” (RFE/RL Newsline, October 22, 2007). Following a protracted political dispute between the high representative and Republika Srpska, reforms were imposed that reduced ethnic quotas for voting in the Council of Ministers. On decisions that are finally decided by parliament, the reforms provided for a majority calculated on those present and voting. On final decisions of the Council of Ministers, the high representative’s reforms mean that if a consensus is not reached, majority support must include at least one member (rather than two hitherto) of each constituent people (OHR 2007). In September 2008 the Parliamentary Assembly of the Council of Europe (PACE) recommended the abolition of entity voting. Such a move would be supported by Bosniaks and Bosnian Croats but opposed by Bosnian Serbs. In response to PACE’s resolution, RS premier Milorad Dodik said Bosnian Serbs would never agree to the removal of entity-based voting, which would threaten the Dayton Agreement and balance within Bosnia and Herzegovina. On the basis of the analytical framework presented above, problems with veto rules in Bosnia arguably correspond to veto issues and veto points. In terms of veto players, groups have power to veto on the basis of the constituent people’s VNI or on an entity basis in the state parliament. As groups are predetermined, it is unlikely that the nationalist parties would agree to move to a system of self-determination as the Dayton Peace Agreement fi xed rights for the three constituent peoples. Veto issues, however, remain an unresolved problem. As acknowledged by international actors and domestic elites, the list of areas that covers legislation subject to the VNI procedure is too broad. Moreover, a two-thirds majority of delegates within a group can trigger a veto on any decision on the basis that it threatens their group’s vital interests. Thus there is a need to define and clarify the areas of legislation that may be subject to VNI. There is also a need to restrict the scope of these issues by removing the provision that two-thirds of a group’s delegates can trigger the VNI veto on any issue. In terms of veto points, there are arguably too many veto procedures at different levels in Bosnia—the VNI procedure in the presidency, the VNI procedure in the entity parliaments, and both VNI and entity vetoes in the state parliament. The procedure for entity voting in the House of Represen-

We Forbid

267

tatives is particularly contentious. The rules are further problematic as the House of Representatives lacks a mechanism to overcome gridlock as it does not have a mediation mechanism such as the joint commission in the House of Peoples. Arguably, without such a procedure, entity delegates in the House of Representatives are incentivized to block legislation rather than compromise to meet the recommendations from a parliamentary mediation body. The dangers of the entity voting procedures are already acknowledged and pose a challenge for EU integration. Moreover, entity voting is arguably unfair as Bosniaks and Bosnian Serbs each have the numerical strength to reach the two-thirds threshold in their respective entities to veto while Croat delegates do not. Though Bosnian Serbs may be reluctant to agree to the abolition of the entity veto for fear of relinquishing some control, this procedure is likely to be a topic for future negotiation. Discussions on the effect of veto rules are part of a wider process of constitutional reform that needs to account for the parties’ divergent positions on state structures. As well as technical problems relating to the specifics of the mutual veto powers, the challenges for power sharing in Bosnia are compounded by these wider issues. Groups (unsurprisingly) seek to defend what they view as their national interests. Yet in Bosnia the three main groups have divergent national interests reflecting their respective positions on what kind of state Bosnia should be. Bosniaks, by virtue of being the plurality, would like to see the abolition of the entities and a move toward a citizen- rather than group-based democracy. They see the creation of Republika Srpska as illegitimate and a reward for ethnic cleansing. Bosnian Serbs resist any attempt to tinker with the entity structure, preferring to maintain control in Republika Srpska and even, at times, threatening a referendum on independence. Bosnian Croats, in an increasingly weak demographic situation, seek a functional state and genuine decentralization, occasionally calling for their own entity. In the absence of defining and limiting the scope of veto issues and veto points, policy issues can very easily be interpreted by elites as issues affecting a group’s VNI, especially when parties are so far apart on the appropriate state structures.

Formal and Informal Veto Powers in Northern Ireland The Bosnian case demonstrates how a complex set of mutual veto arrangements can hinder efficient decision making. Perceptions persist that elites

268

Joanne McEvoy

employ veto rules to obstruct the political system. In contrast, powersharing structures have arguably been more successful in Northern Ireland. Both cases have consociational structures agreed in a peace agreement designed to resolve a self-determination dispute. While Bosnia is a “rigid” consociation based on the three constituent peoples, Northern Ireland is a more flexible consociation based on power sharing between the nationalist and unionist communities. Despite some considerable challenges in Northern Ireland, the use of veto powers has not been particularly contentious for power sharing. I argue that veto rules have been more successful in Northern Ireland than in Bosnia because procedures are less complex and because elites are more committed to the political system. While Sinn Féin may see the arrangements as transitional to a united Ireland, and unionists consider the structures necessary for the maintenance of the Union, both groups are content to work with power sharing, bolstered by the institutional links with the Republic of Ireland and Britain that are vital for nationalists and unionists, respectively. Even though the St. Andrews Agreement of 2006 introduced additional veto mechanisms, the parties have not (to date) triggered their usage to hinder decision making. As discussed below, however, delaying tactics have been used as an informal veto to prevent decisions on certain contentious issues including the devolution of policing and justice powers from Westminster to the Stormont assembly. In 1998, unionists and nationalists, loyalists and republicans agreed to the establishment of new political institutions including a Northern Ireland assembly and power-sharing executive set out in the Good Friday/Belfast Agreement. In terms of formal rules for veto players, the agreement provides for a mix of self-determination and predetermination power sharing among nationalists and unionists. McGarry, O’Leary, and Simeon suggest the peace agreement “combines both corporate and liberal elements” (2008, 62– 63) because the first minister and deputy first minister are elected via cross-community support of national and unionist members while the other ten ministers are selected according to the d’Hondt method, an allocation algorithm “that is liberal or ‘difference-blind.’ It operates according to the strength of representation won by parties in the Assembly, not their national identity” (63). The St. Andrews Agreement changed the election of the first minister and deputy first minister. Instead of a cross-community vote, the largest party of the largest designation in the assembly nominates the first minister and the largest party of the second largest designation nominates the deputy first minister. While this provision might also be seen

We Forbid

269

to be “difference-blind,” it is likely to result in a unionist first minister and nationalist deputy first minister unless the largest party is other than the largest party of the largest political designation. In the assembly, groups are represented by members who designate as “nationalist,” “unionist,” or “other.” A number of institutional safeguards protect the interests of the nationalist and unionist communities. For example, key decisions in the assembly are taken on a cross-community basis: via either parallel consent (a majority of members present and voting, including a majority of nationalists and unionists present and voting) or a weighted majority (60 percent of members present and voting, including at least 40 percent of each of the nationalist and unionist designations). In contrast to Bosnia, the groups in Northern Ireland do not veto decisions on the basis of vital national interests set out in legislation. Under the Good Friday Agreement (GFA), cross-community voting procedures relate to key decisions including the election of the chair of the assembly, the first minister, and deputy first minister (since revised), standing orders, and the budget. Otherwise, thirty assembly delegates can trigger a “petition of concern” for a vote on any other issue to be taken on a cross-community basis. The evidence in Northern Ireland suggests that the parties have not sought to veto each other’s proposals as has been the case in Bosnia. That is not to say that the record of power sharing has been without controversy. The 1999–2002 executive was tested on several occasions by difficulties relating to the inability of the executive and the assembly to check ministerial executive authority. In a controversial decision (Sinn Féin) education minister Martin McGuinness announced the abolition of the 11 plus transfer test, an examination for school pupils transferring from primary to secondary education. While Sinn Féin and the SDLP preferred a comprehensive education system, the UUP and the Democratic Unionist Party (DUP) protested against McGuinness’s decision and wanted to retain academic selection. UUP leader David Trimble later suggested that McGuinness’s decision was a “side effect” of the impending suspension of power sharing in November 2002 and regretted that his party “didn’t get somebody to do a judicial review of the department’s decision. . . . [It] probably would have been struck down because that could not have been taken as a decision under devolution in accordance with the Ministerial Code.” The UUP and DUP also objected to the decision by (Sinn Féin) health minister Bairbre de Brún to site Belfast maternity ser vices at the Royal Victoria Hospital rather than the City Hospital and in opposition to a vote in the assembly. The DUP’s Gregory Campbell

270

Joanne McEvoy

said de Brún’s decision “crystallized the problem of lack of ministerial accountability”; she decided the maternity unit was going to the Royal and “there was nothing anybody could do to stop her.” In the post-suspension period, between November 2002 and the St. Andrews Agreement of October 2006, the DUP consistently cited these two examples of ministerial “solo runs” and argued that any new power-sharing executive would need to be subject to new institutional rules that would limit a minister’s power in the face of opposition from his/her coalition colleagues and/or the assembly. Following a series of interparty discussions and proposals by the British and Irish governments, new veto mechanisms were introduced in the St. Andrews Agreement that ultimately paved the way for restored power sharing in May 2007. The agreement included a provision for any three executive ministers to trigger a cross-community vote in the executive when a vote is required on an issue. The reforms also included a statutory ministerial code and a provision allowing thirty delegates to refer important ministerial decisions back to the executive for review (British and Irish Governments 2006). For the DUP, amendments relating to ministerial accountability were crucial for the party’s agreement to enter an executive with Sinn Féin. In addition to demanding an end to IRA activity and Sinn Féin’s support for the police, the party insisted on new checks and balances to constrain other parties’ ministers via accountability to the assembly and ensure the protection of unionist interests on controversial policy issues (see DUP 2005, 2007). But what difference have these reforms made to the operation of power sharing in Northern Ireland? While the new provisions have not yet been tested, a reading of the text suggests that they do not do much to assert assembly or executive control over ministerial authority. On one hand, the St. Andrews Agreement provides for the assembly to refer important decisions back to the executive and extends the topics for discussion and agreement within the executive to include significant or controversial issues as agreed by the first minister and deputy first minister. On the other hand, the provision for three ministers to trigger a cross-community vote within the executive when a vote is required limits the potential for one group to veto a minister from the other group. For instance, if the DUP seeks to veto a decision of a nationalist minister by referring a decision back to the executive for review, Sinn Féin (with three ministers plus deputy first minister) can request a vote to be taken on a cross-community basis suggesting that the minister’s decision would stand. Arguably, then, ministers will continue to

We Forbid

271

enjoy executive authority over their departments as set out in the GFA without needing assembly support. While formal veto rules have not been as contentious in Northern Ireland as they have been in Bosnia, elites in both cases use informal vetoes. Rather than triggering the formal vetoes in the St. Andrews Agreement, the Northern Ireland parties have recently practiced an informal veto over difficult policy decisions. For example, the DUP refused to agree to the timing of the transfer of policing and justice powers from Westminster. While the DUP claimed there was not sufficient community confidence for the Stormont assembly to have responsibility for policing, Sinn Féin argued that the devolution of policing should have happened in May 2008, as planned in the St. Andrews Agreement. While the two parties agreed that there would be a single justice department with a minister elected via cross-community support and that neither party would take up the post, they failed to agree on further specifics. Following lack of movement from the DUP, Sinn Féin threatened to withdraw from the executive, precipitating reports of “deadlock” at Stormont and the failure to convene executive meetings over the summer of 2008 (BBC News Online 2008). While the DUP exercised an informal veto by refusing to agree on the timing of devolution of policing, Sinn Féin responded in kind by boycotting executive meetings. Thus, as in Bosnia, absenteeism as a form of informal veto can delay and frustrate the decision-making process. In the Northern Ireland case, the deadlock over the devolution of policing and justice was eventually resolved via interparty bargaining. Following talks at Hillsborough Castle in January 2010, chaired by British prime minister Gordon Brown and Irish Taoiseach Brian Cowen, the two parties finally agreed on timing with April 12 set as the target date for devolution. David Ford, leader of the bicommunal Alliance Party, was approved in the assembly as the minister for justice. The record of power sharing in Northern Ireland appears to be one that justifies measured optimism. While the system does provide for group protection via cross-community voting procedures, these provisions have not been used by elites on a regular basis to create gridlock. Compared to Bosnia, mutual veto rules in Northern Ireland are not overly contentious. There is no demand for a procedure designed to protect the communities’ explicit “vital national interests.” Even though thirty MLAs can trigger a vote to be taken on any issue, the motion requires cross-community support to succeed. For example, although the DUP in the 1999–2002 executive attempted to exclude Sinn Féin over the lack of progress on IRA decommissioning by

272

Joanne McEvoy

triggering a “petition of concern” on the issue, the motion failed to secure cross-community support. Arguably this need for cross-community support limits the potential for blockage and promotes accommodation. While the parties agreed to new veto procedures under the St. Andrews Agreement, they have not been compelled to use them to gridlock the system. Yet controversial decisions on academic selection, the Irish language, and the site of the sports stadium are yet to be agreed on and may test elites’ incentives to compromise.

Conclusions and Policy Implications: A More Refined Mutual Veto While the mutual veto may be a secondary element in Lijphart’s consociational framework, the operation of such rules clearly matters for powersharing practice. Veto powers may be crucial for ethnonational parties when agreeing to new postconflict institutions as they provide them with the sense that their vital interests will be secure. Groups need to feel that they have the capacity to control issues important to them and prevent decisions favored by other groups that may threaten their fundamental interests. But what does this analysis of various veto rules mean for power-sharing practice? It argues for more refined veto powers on the basis of the who (veto players), the what (veto issues), and the where (veto points) of the mutual veto. First, veto players should preferably be self-determined groups, though that might not always be possible, as per the rights of the three constituent peoples in Bosnia. Second, when groups want to ensure the protection of their fundamental interests, veto issues should be clearly defined in legislation and be agreed upon among the political parties. Such issues should be limited to identity issues, symbols, language, culture, education, the budget, and security. Th ird, veto points should also be limited—a complex set of veto rules at various points in the political process without sufficient clarification and limitation will likely not incentivize ethnic parties to cooperate. While veto procedures may be necessary in a collective presidency as well as the executive, veto procedures should be focused in one chamber in a bicameral system. The chapter’s findings support the view that veto rules have the potential to foster interethnic cooperation and constrain elite intransigence. If politi-

We Forbid

273

cal parties taking part in peace negotiations feel their group’s interests will be protected in the new institutions and they will have the power to block “harmful” decisions, they may have a greater incentive to support the political system. There is an important qualification here, however, as institutional rules can only go so far. Much depends on elites’ political willingness to work the system in a positive manner. As the example of Bosnia shows, when parties disagree over state structures and constitutional reform, they may be incentivized to veto decisions to demonstrate their political preferences. The challenge for the reform of the Dayton constitution is to design a political system in which the three groups will have a sense of ownership and believe that working the system is their best option for achieving their political objectives. Moreover, as the Office of the High Representative in Bosnia moves toward closure (to be replaced by a “reinforced” EU presence), the task will be to ensure that elites choose to take responsibility for decisions within the existing arrangements. Yet elite willingness is not the whole story. Elite motivations do not render institutions obsolete. Even when elites choose to delay decision making or boycott meetings, as has been the case in Northern Ireland, progress is worked out “behind the scenes” in ongoing negotiations. When parties believe power sharing is the most acceptable (or the least bad) political system, when there are less extensive veto rules and an agreed set of veto rules, they will have less incentive to create gridlock. These insights may help guide policymakers responsible for the design of veto powers in potential powersharing cases. For example, lessons might be drawn for the UN-led negotiations on finding a solution to the stalemate in Cyprus. Under the 1960 constitution, the Greek president and Turkish vice president had extensive veto rights and could block any decision. The Annan Plan of 2004 (rejected by Greek Cypriots) proposed a reformed set of veto procedures. Decisions in the federal parliament would require majority support in both the Chamber of Deputies and the Senate, including at least one-quarter of senators from each constituent state (the Greek Cypriot state and the Turkish Cypriot state). The proposals also set out a list of “specified matters” requiring at least twofift hs support of senators from each constituent state. Veto rights were also proposed for the presidential council and the council of ministers (see UNFICYP 2004). To secure a new settlement, negotiators need to consider the appropriate power-sharing institutions including mutual veto powers. Appropriate institutional rules should enable the communities to block “harmful”

274

Joanne McEvoy

decisions at the federal level and provide for group protection, efficient decision making, and interethnic accommodation. Notes 1. Bosnia and Herzegovina (hereafter Bosnia) encompasses two entities: the (Bosniak-Croat) Federation of Bosnia and Herzegovina and the Bosnian-Serb dominated Republika Srpska (RS). 2. HINA Croatian News Agency, 22 February 2008; BBC Monitoring European, 25 February 2008, available at Bosnian Institute, http://www.bosnia.org.uk. 3. At a news conference at Parliament Buildings announcing the election of first minister and deputy first minister on 6 November 2001, fighting broke out between pro- and anti-agreement members, later termed the “brawl in the hall.” See Belfast Telegraph, “Scuffles After Leaders Elected,” 6 November 2001. 4. Thanks to Brendan O’Leary for pointing me to these historical examples. 5. The Dayton Peace Agreement accorded veto rights to the constituent peoples and the two entities in the state parliament; the VNI veto in the entities was introduced following a Constitutional Court decision in 2000 and the subsequent imposition of constitutional amendments by the high representative in April 2002; see OHR (http://www.ohr.int/). The definition of VNI in the constitutions of the two entities includes representation of constituent peoples; identity of a constituent people; constitutional amendments; organization of public authorities; equal rights of the constituent peoples in decision making; education, religion, language, culture; territorial organization; and the public information system. 6. The caucus of Bosniak delegates in the state House of Peoples maintained that the creation of a Croat channel within the state public broadcasting system would be destructive to the interests of the Bosniak people as it would threaten dissolution of the state institution, increase ethnic tension, and possibly give rise to the dissolution of Bosnia and Herzegovina. See Constitutional Court of Bosnia and Herzegovina, Decision U 9/08, available at http://www.ccbh.ba.eng /odluke/povuci _zip?pid+138211. The court concluded that the 23 April 2008 Draft Conclusion of the House of Peoples supported by Bosnian Croats and Bosnian Serbs to request the Council of Ministers to consider the creation of a Croat channel is the start of a procedure and does not, therefore, pose a constitutional question. 7. European Commission, Progress Report Conclusions on Bosnia and Herzegovina, available at http://ec.europa.eu/enlargement/pdf/key_documents/2009/conclu sions _on _bih _en.pdf. 8. Author interview, Sarajevo, May 2008. 9. Correspondence with Professor Bruce Hitchner, Dayton Project, Tufts University. Professor Hitchner is chair of the Dayton Peace Accords project based at Tufts and participated in the constitutional reform process supported by the international community from 2005. See Hitchner 2006.

We Forbid

275

10. Author interview, British Embassy, Sarajevo, May 2008. 11. Author interview, OHR, Sarajevo, May 2008. 12. SETimes, “PACE Issues Resolution in Bosnia and Herzegovina, Angering RS Leaders,” 7 October 2008, available at http://www.setimes.com/ (accessed 13 October 2008). 13. On 13 October 2008 the RS National Assembly convened a special session to discuss political options including the right to a referendum on independence. Against severe criticism from the OHR and the U.S. embassy, Dodik reportedly said he was prepared to proclaim independence for the RS should the OHR seek to remove him from office. See BalkanInsight, “Bosnian Serbs Warned over Secession,” 13 October 2008, available at http://www.balkaninsight.com/. 14. Author interview, October 2005. 15. Author interview, July 2005. 16. Author interview, December 2007. References BBC News Online. 2008. “SF Threatens to Collapse Assembly.” 24 August. Bieber, Florian. 2005. “Power Sharing after Yugoslavia: Functionality and Dysfunctionality of Power-Sharing Institutions in Post-war Bosnia, Macedonia, and Kosovo.” In Sid Noel, ed., From Power Sharing to Democracy: Post-Conflict Institutions in Ethnically Divided Societies. Montreal: McGill-Queen’s University Press. British and Irish Governments. 2006. St. Andrews Agreement. Available at http://www .nio.gov.uk /st _andrews _agreement.pdf. Choudhry, Sujit. 2008. “Bridging Comparative Politics and Comparative Constitutional Law: Constitutional Design in Divided Societies.” In Sujit Choudhry, ed., Constitutional Design for Divided Societies: Integration or Accommodation? Oxford: Oxford University Press. Cousens, Elizabeth M. 2002. “From Missed Opportunities to Overcompensation: Implementing the Dayton Agreement on Bosnia.” In Stephen John Stedman, Donald Rothchild, and Elizabeth M. Cousens, eds., Ending Civil Wars: The Implementation of Peace Agreements. Boulder, CO: Lynne Rienner Publishers. Democratic Unionist Party (DUP). 2005. Leadership That’s Working. Westminster and Local Government Elections Manifesto. Available at http://www.dup.org.uk. ———. 2007. Getting It Right. Northern Ireland Assembly Election Manifesto. Available at http://www.dup.org.uk. Godson, Dean. 2004. Himself Alone: David Trimble and the Ordeal of Unionism. London: Harper Collins. Hays, Donald, and Jason Crosby. 2006. From Dayton to Brussels: Constitutional Preparations for Bosnia’s EU Accession. Special Report, United States Institute of Peace. October. Available at. http://www.usip.org /publications/dayton-brussels -constitutional-preparations-bosnias-eu-accession.

276

Joanne McEvoy

Hitchner, R. Bruce. 2006. “From Dayton to Brussels: The Story Behind the Constitutional and Governmental Reform Process in Bosnia and Herzegovina.” Fletcher Forum of World Affairs 30, 1: 125–35. Lijphart, Arend. 1977. Democracy in Plural Societies: A Comparative Exploration. New Haven, CT: Yale University Press. ———. 1995. “Self-Determination Versus Pre-Determination of Ethnic Minorities in Power-Sharing Systems.” In Will Kymlicka, ed., The Rights of Minority Cultures. Oxford: Oxford University Press. ———. 2004. “Constitutional Design for Divided Societies.” Journal of Democracy 15, 2: 96–109. Lintott, Andrew. 1999. The Constitution of the Roman Republic. Oxford: Clarendon Press. Linz, Juan. 1990. “The Perils of Presidentialism.” Journal of Democracy 1, 1: 51– 69. Lukowski, Jerzy. 1991. Liberty’s Folly: The Polish-Lithuanian Commonwealth in the Eighteenth Century, 1697–1795. London: Routledge. McGarry, John, and Brendan O’Leary. 2007. “Iraq’s Constitution of 2005: Liberal Consociation as Political Prescription.” I.CON 5, 4: 670–98. McGarry, John, Brendan O’Leary, and Richard Simeon. 2008. “Integration or Accommodation? The Enduring Debate in Conflict Regulation.” In Sujit Choudhry, ed., Constitutional Design for Divided Societies: Integration or Accommodation? Oxford: Oxford University Press. Office of the High Representative (OHR). 2007. “Decision Enacting the Authentic Interpretation of the Law on Changes and Amendments to the Law on the Council of Ministers of Bosnia and Herzegovina Enacted by the Decision of the High Representative of 19 October 2007.” 3 December. Available at http://www.ohr.int/. Persson, Torsten, Gerard Roland, and Guido Tabellini. 1997. “Separation of Powers and Political Accountability.” Quarterly Journal of Economics 112, 4: 1163–1202. Purdy, Martina. 2005. Room 21: Stormont—Behind Closed Doors. Belfast: Brehon Press. Samuels, David J. 2002. “Presidentialized Parties: The Separation of Powers and Party Orga nization and Behavior.” Comparative Political Studies 35, 4: 461–83. Schneckener, Ulrich. 2002. “Making Power-Sharing Work: Lessons from Successes and Failures in Ethnic Conflict Regulation.” Journal of Peace Research 39, 2: 203–28. Sebastián, Sofia. 2007. Leaving Dayton Behind: Constitutional Reform in Bosnia and Herzegovina. FRIDE working paper, November. Tsebelis, George. 2002. Veto Players: How Political Institutions Work. Princeton, NJ: Princeton University Press. Tsebelis, George, and Xenophon Yataganas. 2002. “Veto Players and Decision-Making in the EU After Nice.” Journal of Common Market Studies 40, 2: 283–307. United Nations Peace Keeping Force in Cyprus (UNFICYP). 2004. The Comprehensive Settlement of the Cyprus Problem (“Annan Plan”). 31 March. Available at http://www.unficyp.org /nqcontent.cfm?a _id=1637.

We Forbid

277

Venice Commission, European Commission for Democracy Th rough Law. 2005. Opinion on the Constitutional Situation in Bosnia and Herzegovina and the Powers of the High Representative. CDL-AD 004. 11 March. Wheatley, Steven. 2005. Democracy, Minorities and International Law. Cambridge: Cambridge University Press. Zahar, Marie-Joëlle. 2005. “The Dichotomy of International Mediation and Leader Intransigence: The Case of Bosnia and Herzegovina.” In Ian O’Flynn and David Russell, eds., Power Sharing: New Challenges for Divided Societies. London: Pluto Press.

CHAPTER 11

Northern Ireland: Power Sharing, Contact, Identity, and Leadership Ed Cairns

Some commentators argue (pace McGarry and O’Leary 2006) that power sharing cannot provide a durable solution to intractable identity-based conflicts (Sisk 2003) but instead “provides only a temporary lull” in the conflict and may even “freeze group boundaries” and “heighten latent ethnic identities” (Norris 2005, 3). Perhaps what is needed to ensure the ultimate success of power-sharing initiatives is more input from social psychology. Th is is what Abrams and Hogg might argue given their claim that social psychology occupies a “pivotal position” in the social sciences and is ideally placed to link the micro- and macrolevels of analysis and to demonstrate how social and individual variables “become expressed in social situations within a societal context” (2004, 98). They therefore argue that in terms of framing policy, social psychology should be enormously attractive. However, the problem is determining which parts of social psychology are relevant to analyzing power sharing. In this chapter I argue, with illustrations from research in Northern Ireland, that we need to look at the part of social psychology that has specialized in bringing together people from different backgrounds—known as the contact hypothesis or contact theory (Allport 1954; Hewstone and Brown 1986; Pettigrew 1986). In its simplest form, contact theory proposes that bringing together individuals from opposing groups can, “under optimal conditions,” reduce intergroup conflict. In the context of power sharing, therefore, the hope would be that the vast amount of evidence now available as a result of empirical research based on contact theory (see Pettigrew and Tropp 2000)

Northern Ireland

279

can be used to build on the (temporary) lull in conflict afforded by the current power-sharing arrangements in Northern Ireland. Searching for a point of contact between social psychology and the short- to mid-term impact of power sharing, I was struck by the following quotation: “there were lots of words written on the flow chart, but I just wrote one word—leadership.” These are the words Martin McGuinness uttered during a television interview (BBC NI television interview, September 5, 2007) upon returning from meeting with a delegation representing the warring factions in Iraq in which he and others provided insights into the peace process and power sharing in Northern Ireland. In fact any attention given to leadership in the peace process has tended to come from nonacademic sources. This is all the more surprising given the fact that it is now more than forty years since Rose pointed out that the history of Ireland, including Ulster, offers copious examples of the importance of political leadership (1971). This sentiment was echoed by Heskin, Cairns, and McCourt, who claimed that “one area of political psychology which would appear to be of relevance to understanding the conflict in Northern Ireland is that concerned with people’s perceptions of political leadership” (1990, 355). Despite this they noted that work on leadership was “a glaring omission from the existing research base.” Today it would appear that the position has not changed. In her book on the role of leaders in the current peace process, Gormley-Heenan concluded that the role of political leaders has apparently been viewed, at least by academics, “as both decidedly obvious and warranting little further explanation” (2005). Further, she notes that journalists and others who did write about Northern Irish political leaders tended to focus mostly on such things as the leaders’ motivations and their personalities. These emphases on leaders’ personal attributes closely reflect early leadership research in psychology, which tended to concentrate on the leader’s personality. For example, do leaders with authoritarian attitudes appeal to certain followers? Early research suggested that situational determinants are critical while subsequent theorizing adopted an interactional approach and claimed that it is the interaction of personality and situation, which means that some leaders are best suited to certain tasks (Bales and Slater 1955). What has tended to be ignored until recently is the psychological impact the leader has on his/her followers. More recent theorizing in social psychology (see Haslam 2004 for a review) claims that this is critical because the way we think about ourselves, our self-concept, our social identity, has

280

Ed Cairns

important effects upon the way we think, feel, and behave. In turn leadership has the ability to change the way followers think about themselves and come to define their social identities. In this chapter I have therefore chosen to focus on the topics of leadership and cross-community and to illustrate how both can be relevant to power sharing in Northern Ireland. I will begin by outlining what has come to be known as the “the social identity approach,” which informs the theoretical bases for both recent work on contact theory in the social psychology literature as well as more recent social psychological insights involving leadership. This will be followed by a brief review of existing work on both of these topics—contact and leadership. A final section will attempt to sketch the intimate relationship between leadership and contact. I examine the possibility that in Northern Ireland, this relationship may erect barriers to power sharing and make tentative suggestions about the policy implications this might have.

Theoretical Background: The Social Identity Approach One part of social psychology that has come to be seen as critical in attempts to “unravel aspects of the mind-society interaction” (Turner and Reynolds 2001, 149) and that impacts both the contact hypothesis and social psychology’s approach to leadership is what has come to be known as “the social identity approach” (Abrams and Hogg 2004, 100). This approach began with social identity theory (Tajfel and Turner 1986), which explained, using normal psychological processes, our need, as group members, to positively differentiate our own group from relevant out-groups in order to achieve a sense of positive identity. Today the social identity approach has subsumed social identity theory and –self-categorization theory as subtheories, the latter focusing on our need to categorize ourselves and others in order to reduce subjective uncertainty (Abrams and Hogg 2004, 102).

Social Identity Theory

Social identity theory (SIT) proposes that social identification with a specific social group leads to the need to differentiate the in-group from comparable out-groups, which in turn may lead to in-group favoritism and out-group

Northern Ireland

281

hostility (Tajfel 1978; Turner and Onorato 1999). The degree of group identification determines the individual’s inclination to apply a specific social category to the self in a specific social context limited by fit and accessibility (Haslam 2001). According to SIT, individuals identify with social groups in order to increase their self-esteem by comparing their in-group with out-groups along chosen comparison dimensions. There is now ample evidence that discrimination elevates self-esteem, as opposed to the earlier view that lowered selfesteem motivates discrimination (Abrams and Hogg 2004, 102). Because SIT has attempted to explain intergroup cooperation and especially intergroup conflict and consequent discrimination and prejudice, SIT has played a key role in facilitating an understanding of the psychological dimensions of ethnic conflict in Northern Ireland (Cairns 1982; Cairns et al. 1995; Tausch et. al 2006; Trew and Benson 1996; Whyte 1990).

Self-Categorization Theory Self-categorization theory, or SCT (e.g., Turner 1982; Turner et al. 1987), was developed as an extension to social identity theory. SCT moved from the motivational question asked by SIT—“Why do people in groups discriminate against each other?”—to the more fundamental question—“What is a psychological group?” (Turner and Reynolds 2001, 135). As a result SCT focuses on the basic social cognitive processes, especially social categorization. This means that when we categorize people (including ourselves) we come to see them not as individuals but as members of a common social category (to a greater or lesser extent). In SCT this has come to be known as “depersonalization.” In other words people are not viewed as unique individuals but as representatives of their group. Categorization also leads us to accentuate intergroup differences and to minimize intragroup differences. This has fundamental consequences, leading not only to the categorization of other people as members of social groups (depending of course on the particular social context) but also, most important, to self-categorize. This happens, self-categorization theory claims, because the self can be defined at different levels of abstraction including an individual and a collective self, which reflect different categories of identity. In social situations, self-categorization depends on which social identity is salient.

282

Ed Cairns

Self-categorization operates in exactly the same way as categorization of others in that “it depersonalizes self-perception in terms of the in-group prototype” (Hogg and Reid 2006, 11). As a result, not only do particular social identities become salient, as people see themselves in terms of shared features that define membership, but also they are more likely to adopt group-defining attitudes and norms. Critically, the more highly identified people are with a particular category or group the more likely they are to feel and behave according to group attitudes and norms. In the next sections the two theoretical approaches reviewed above, SCT and SIT will be used to examine, respectively, the short-term impact of leadership during power sharing and the long-term role of contact theory.

The Short Term: Leadership According to Schneckener (2002, 224), if power sharing is to succeed, leaders (he refers to them as “elites”) have not only to convince their followers to enter into power sharing but also forge links with moderate elements in the out-group, while at the same time dealing with radical opponents. This makes it clear, as Hogg et al. (2005, 991) have pointed out, that leadership is such an essential feature of social groups that it is virtually impossible to think about groups without thinking about who the leader is and how good the leadership is. What is different about the SCT approach is that leadership “is not simply a matter of leaders, or even of leaders and followers. Rather it has to do with the relationship between leaders and followers within a social group” (Reicher, Haslam, and Hopkins 2005, 551). This, Reicher, Haslam, and Hopkins (2005) note, is in fact so obvious it is often overlooked. If leadership is viewed as both an intragroup and indeed an intergroup phenomenon, then following SCT, which sees group structure in terms of prototypicality, more prototypical members are more influential (Abrams and Hogg 2004, 103). Further, individuals who are group prototypical, that is, represent a collective reference point in terms of attitudes and behavior, are the members of the group most likely to be seen as leaders. None of this comes about unless people identify with the collective— that is, unless they possess a strong social identity. This social identity can moderate the effectiveness of leadership group prototypicality and leader group orientedness. Leaders impact the parts of the self that are salient in a

Northern Ireland

283

par ticular context. In turn, leaders can either create these elements of the self-concept or strengthen or weaken existing parts of the self. From an SCT perspective, therefore, one of the most important accomplishments of effective leadership is tying the sense of self to the collective. That is, leadership can convert followers’ motivation from personal selfinterest to personal collective interest—in other words, make individuals feel that they are members of the group. This, it has been suggested, may be the very essence of transformational or charismatic leadership in that “charismatic leaders are capable of redefining group norms and objectives. They not only direct, but actively transform, their followers’ attitudes and behavior” (Reicher, Haslam, and Hopkins 2005, 550). Incidentally, it is important to understand that although there is evidence that leaders are group prototypical in SCT, prototypicality is not a fi xed category. Rather the “relative prototypicality of an individual varies with the dimensions of comparison” (Turner 1987, 80). According to SCT, prototypicality of a member of any group depends on (among other things) that individual being minimally different from other group members that are salient in any given context (Turner 1987). This is critical for understanding leadership because it implies that “the prototypicality of exactly the same exemplar will vary as a function of the social context within which categorization takes place” (Haslam 2004, 46). This idea can be used, for example, to explain why in Northern Ireland, as the social context changed, the Unionist electorate began to see Ian Paisley as their leader, in other words, as “more prototypical” than David Trimble. As the Nationalists/Republicans appeared to make more ground this forced Protestants to categorize themselves as ultra-unionist and Ian Paisley became the prototype of their group and therefore a potential leader. Similarly, it is interesting to speculate in these terms as to why Martin McGuinness was chosen by his party to be Deputy First Minister rather than Gerry Adams. Was this because as an exemplar of militant Republicanism he was a more appropriate leader for those Republicans least likely to agree with the Good Friday Agreement?

The Long Term: Contact Inspired by social identity theory, the contact hypothesis was further developed over the last twenty years. Three contrasting models can be identified

284

Ed Cairns

as encompassing the contact hypothesis (Pettigrew 1998). The “decategorization model” (Brewer and Miller 1984) calls for contact between individuals of two opposing groups, hence encouraging individuals to disregard their social identity and thus distance themselves from the in-group in order to promote personalized contact between in-group and out-group members. Gaertner et al. (1993) focused on recategorization into broader social categories and the establishment of a common in-group identity for social groups in intergroup conflict, again a process that requires a relevant social identity to be salient. Hewstone and Brown (1986) propose that the most effective form of contact is intergroup contact, by which they mean contact that takes place with each group’s social identity salient. The main limitation of the contact hypothesis is that often attitude change does not generalize from contact with a single member of the out-group to the out-group as a whole (Hewstone and Brown 1986; Hewstone and Cairns 2001; Pettigrew 1998). On a more practical note, as Hewstone and Cairns (2001) point out, it is often difficult to bring about intergroup contact on a large scale in real-life contexts. Aron et al. (1991; Wright et al. 1997) tried to overcome this limitation by proposing that “extended contact” (the individual experiencing his or her friends having close out-group friends) may also lead to more positive out-group attitudes. Our current research on cross-community contact in Northern Ireland has focused on psychological processes (e.g., Hewstone et al. 2006; Hewstone et al. 2008). We have clear evidence that segregation restricts crosscommunity contacts and that absence of contact is associated with stronger ethnoreligious identities; these are, in turn, associated with greater anxiety about intergroup contact and more sectarian attitudes that include greater support for violence as a means of solving the ongoing conflict. We also have evidence (Paolini et al. 2006) from surveys in Northern Ireland that the size of the friendship out-group prejudice relationship is influenced not only by direct friendship effects but also by indirect friendship effects, a phenomenon that will be discussed later in the context of power sharing.

A Common In-Group Identity in Northern Ireland? Before considering the possibility that Catholics and Protestants in Northern Ireland could develop a common in-group identity, it is important to emphasize that doing so does not mean that former identities are abandoned.

Northern Ireland

285

Instead this means that people come to categorize their world in such a way that they have available to them an additional (new) social identity that they hold in common with the out-group. This is possible because “psychological group memberships are not fi xed, but fluid and context dependent” (Platow, Reicher, and Haslam 2008, 3). And there is encouraging evidence emerging from Northern Ireland that a common or shared in-group identity is beginning to develop and that this may be related to intergroup contact. Survey evidence over the period 1968–94 shows that when asked to choose the one term that “best describes the way you usually think of yourself” Catholics consistently chose “Irish” while more Protestants were increasingly likely, with the passage of time, to see themselves as “British” (Trew 1996). However, analyses of more recent survey data from 1993 to 1995 suggest that a “Northern Irish” identity is beginning to develop, especially among more middle-class respondents, both Catholic and Protestant. More recently, two studies have provided evidence that the development or strengthening of a national-political identity that can be embraced by both Catholics and Protestants—“Northern Irish”—may in fact be the result of successful intergroup contact. In the most extensive of these studies, Hayes, McAllister, and Dowds (2005), using data from random sample surveys of the Northern Irish population, were able to show that after a range of background variables were included in the analysis, Protestants who had attended an integrated school were significantly less likely to hold a British-Unionist identity than those who had experienced a segregated education. A similar pattern emerges among the Catholic community, albeit among those who attended an informally mixed school. Catholics who had attended an informally integrated school were significantly less likely to hold both an Irish and nationalist identity than were their colleagues who had attended a religiously segregated school. In other words the adults (Catholics and Protestants) who had more intergroup contact at elementary or high school were more likely to use the inclusive label “Northern Irish” when describing themselves in nationalpolitical terms. Cairns, Hewstone, and Tausch (2006) used data from a random sample of Northern Irish sixteen-year-olds to investigate this phenomenon. Of these, 78 percent of Catholics described themselves as Irish, while 51 percent of Protestants described themselves as British. However, 17 percent of Catholics and 33 percent of Protestants indicated they preferred the superordinate

286

Ed Cairns

identity “Northern Irish.” Crucially, in line with the Hayes, McAllister, and Dowds (2005) finding, the young people who chose the Northern Irish identity reported that they had had significantly more intergroup contact opportunities both at school and where they lived, compared to their peers who described themselves as Irish or British. Unfortunately, while integrated education may be succeeding as an instrument to promote a common in-group identity and therefore strengthen power sharing, it is, as I have suggested, a policy that will have impact only in the long term. Therefore, while worth pursuing, it is not going to help Northern Ireland’s power-sharing government through the rocky days that undoubtedly lie ahead. Something therefore needs to be done to encourage intergroup contact among the general population and, in particular, among the sections of the population that are clinging stubbornly to segregationist anti-power-sharing views.

Leaders and the Contact Hypothesis While the main impact of contact will undoubtedly be long term it is possible that contact may also play a more modest role in the short term in reducing sectarian attitudes and behaviors. Remembering the work on “extended” contact, as the two political leaders, Ian Paisley and Martin McGuinness appeared so at ease in each other’s company, to the extent that the local media referred to them as the “chuckle brothers,” their conduct may have impacted the attitudes of their party rank-and-fi le members. That is, party members who have no out-group friends may have been influenced by the fact that their friend (Ian or Martin) had a friend from the other community. And there is evidence to support the possibility that out-group leaders play a role in implementing the extended contact effect. Hajnal (2001) used data from the United States to show that the election of African Americans to mayoral office, while it does not have a major impact on the black community, does improve racial attitudes among the white electorate. More particularly, black mayors have a profound impact on Democrats (but not Republicans). Among this sector of the U.S. population the election of a black mayor leads to a decrease in racial tension, increased racial sympathy, and increased support for black leadership (Hajnal 2001, 603). Hajnal suggests that this effect is due to the fact that “every new Black leader provides

Northern Ireland

287

additional information to the White community and reduces fear of Black representation in general” (2001, 614). Hajnal’s (2001) data indicate a polarizing effect of out-group leadership because the gulf between white Democrats and Republicans on racial attitudes increases under black mayors. In the context of Northern Ireland this draws attention to an important confluence of contact and leadership that springs from the role of community leaders who hold negative attitudes toward power sharing. Theoretically this is linked to the fact that from a social identity perspective, in-group prototypes play two crucial roles in determining how people behave. First they describe behavior and second, and perhaps more important in the context of divided societies, they prescribe (and proscribe) it—telling people how they ought to behave as group members. In other words, norms that define an in-group that we identify with also have the power to influence our actual behavior (Hogg and Reid 2006, 13). And in this context it is important to remember that SCT indicates that leaders play a central role in defining identity and influencing behavior in their role as “entrepreneurs of identity” (Reicher et al. 2005, 556). This is key because a leader’s entrepreneurial role may involve not just mobilizing certain identities but also “manipulating the social comparative frame” (Hogg and Reid, 2006, 19). This change in the comparative context “can make both particular identities and particular leaders viable” (Reicher, Haslam, and Hopkins 2005) by altering the group prototype and the leader’s own prototypicality. In turn this may influence group members’ norms and behavior. In other words, leaders do not simply define identities; they also “construct” them (Klein, Spears, and Reicher 2007, 42). All of this is of particular importance because in divided societies, according to Livingstone and Haslam, out-group conflict may not simply be a consequence of identification with the in-group but “may itself actually be an important part of in-group identity” (2008, 3). An important part of conflict is, of course, avoiding socializing with the out-group. This can be seen in Northern Ireland, where the norms demand subtle and not-so-subtle levels of segregation. Examples of subtle segregation can be seen in sport and other pastimes (Niens, Cairns, and Hewstone 2003) and indeed in the way both communities continue to patronize different ser vice providers such as schools, family doctors, dentists, and lawyers even though, in Northern Ireland as a whole, residential segregation may be less common than it once was (Shuttleworth 2007). Less subtle evidence for

288

Ed Cairns

this desire for avoidance can be seen in working-class interface communities, for example, in the erection of Belfast’s so-called peace walls and a general increase in “chill factors” including increased marking of sectarian boundaries with lamppost painting, flags, graffiti, and other expressions of ingroup identity such as paramilitary association (OFMDFM 2002). With the advent of the peace process in Northern Ireland it is not surprising, therefore, that evidence suggests that “leaders” at the “community” level, particularly in working-class interface communities, attempt to ensure the continuation of current sectarian ideas about what is considered normative, in particular ensuring that the key norm of avoiding contact with the out-group is maintained. Based on interviews in working-class communities in Belfast, Shirlow concluded that the peace process is stalling precisely because in these communities “the capacity to reconstruct identity and political meaning is obviated by political actors who mobilize fear in order to strengthen unidimensional classifications of political belonging. Community based selfrepresentation assumes the form of a mythic reiteration of purity and self-preservation” (2003, 89). Further, these “political actors” are essentially “motivated by a sectarian consciousness” that regards “involvement in cross-community schemes as perfidious and as a betrayal of community based loyalty” (Shirlow 2003, 87). They certainly represent a strong deterrent to intergroup contact. And these “political actors” are not benign leaders. In these communities many respondents fear the paramilitary leaders in their own community at least as much as members of the ‘outgroup’ (Shirlow 2003). This means that in these communities there is a reluctance to be involved in intergroup contact activities. As a result people in interface areas report that they often choose to “lie to their neighbours” rather than reveal that they had been engaging in contact with members of the “other” group (Shirlow 2003, 88). This throws new light on the contact hypothesis given that one of the items in Allport’s original set of conditions necessary for optimal contact is that contact should receive positive “institutional support” (Allport 1954). What I am suggesting here is that it is at the community level, as well as at the political level, that we should look for the critical “social climate” (Amir 1969) that will promote intergroup contact. In other words, if contact is to succeed, the role of community leaders in providing “institutional support” must not be overlooked.

Northern Ireland

289

Recently Dixon et al., in what they claim is a “reality check for the contact hypothesis,” criticized contact researchers because of their inability and/or reluctance to shoulder the “burden of explaining the specific nature, form and consequences of failed contact within a given set of historical, material and political circumstances” (2005, 709). Based on the evidence presented in this chapter, in Northern Ireland, one of the “circumstances” that prevent optimum contact from taking place, at least in certain communities, may relate to the way in which (some) people in Northern Ireland represent their social identity. These social identities are in turn allied to the question of leadership—not political leadership in its formal sense but leadership at the community or local level.

Conclusion Livingstone and Haslam (2008) suggest that one way to improve intergroup relations in settings of chronic social conflict, such as Northern Ireland, is to create conditions that allow group members to “reconstruct and reconstrue the (identity) relationship between the in-group and outgroup, such that it comes to be seen as benign and even positive and productive, rather than conflictual” (2008, 17). This presents leaders in Northern Ireland with a particular problem: that of leadership “across boundaries” (Platow, Reicher, and Haslam 2008). However, as we have noted, the possibility of developing a shared identity is not impossible at least among the middle classes. Unfortunately in Northern Ireland a major obstacle to the universal acceptance of power sharing is the fact that “weak community infrastructure and lack of community development in the most disadvantaged communities, often means that there are few ‘collectives,’ other than those with a separatist agenda, to drive action at local level” (Hughes et al. 2007, 48). In light of this, if policy initiatives are to attempt to transform identities in disadvantaged communities in Northern Ireland, it is now necessary to adopt a “dynamic view of the leadership process [that] is underpinned by a dynamic model of the relationship between social reality, social identity and collective action” (Reicher, Haslam, and Hopkins 2005, 556). It is crucial that such policy initiatives must not be one-sided. Instead, it is critical to remember that “out-group norms can have a significant impact on how we construct in-group norms” (Hogg and Reid 2006, 13) and that this may be particularly

290

Ed Cairns

true in a conflictual society like Northern Ireland where, it has been claimed, “the in-group’s relationship with the out-group constitutes an important part of in-group identity” (Livingstone and Haslam 2008, 3). Such a change of emphasis may help bridge the gulf between contact as it is represented in the social psychological literature and contact as it is practiced, experienced, and regulated in everyday life (Dixon, Durrheim, and Tredoux 2005, 709). At the same time it still remains true, as Pettigrew has pointed out, that social psychology can “contribute only part of the answer to the riddle of social life” because most social-psychological processes do not typically have causal primacy (1986, 191). Rather, as Pettigrew notes, if the optimum conditions for implementing effective contact can be met “it usually signifies that the important structural issues, from power to resource allocations, have already been equitably worked out” (191). I would argue that in Northern Ireland all or most of these issues have been worked out—thanks to power sharing. As a result the time is now ripe to promote contact not just among schoolchildren but also among the general population and particularly in those underprivileged working-class enclaves that have been, for the most part, immune to community relations efforts to date. I am therefore arguing that power sharing is a short-term solution to Northern Ireland’s sectarianism and that among the obvious benefits it is bringing to Northern Irish society, including dramatically lower levels of violence and increased economic prosperity, is the basis for promoting increased contact between the rival communities which, if handled correctly, should in turn lead to the development of a shared identity between Catholics and Protestants. When this happens, then, and only then, in Northern Ireland will “conventional socio-economic politics [become] more prevalent than identity politics” (McGarry and O’Leary 2006, 276). Notes I am grateful to Alex Haslam and Miles Hewstone for their very helpful comments on an earlier draft of this chapter. 1. Paralleling Gormley-Heenan’s observations about the lack of attention to leadership in the peace process in Northern Ireland. 2. Pettigrew (1998) has attempted to combine these theoretical approaches by suggesting that all three are important though the order in which they are introduced would determine the effectiveness of contact interventions. According to Pettigrew (1998), contact should first be between individuals from different groups, then intergroup contact should be emphasized, and finally a superordinate category of identification established.

Northern Ireland

291

Editors’ Note: We deeply regret to note that Professor Cairns died in a car accident in 2012, before the copyediting of his draft chapter was completed. Aside from required changes to conform to American English and conventions, no changes have been made to his final text. We sought to trace all his citations with complete accuracy and have assumed that Tausch et al. 2006 is the citation he intended because his note indicated that Cairns, Hewstone, and Tausch were coauthors. References Abrams, D., and M. A. Hogg. 2004. “Metatheory: Lessons from Social Identity Research.” Personality and Social Psychology Review 8, 2: 98–106. Allport, G. W. 1954. The Nature of Prejudice. Reading, MA: Addison-Wesley. Amir, Y. 1969. “Contact Hypothesis in Ethnic Relations.” Psychological Bulletin 71: 319– 42. Aron, A., E. N. Aron, M. Tudor, and G. Nelson. 1991. “Close Relationships as Including Other in the Self.” Journal of Personality and Social Psychology 60: 241–53. Bales, R. F., and P. E. Slater. 1955. “Role Differentiation in Small Decision-Making Groups.” In T. Parsons and R. F. Bales, eds., Family, Socialization and Interaction Process. New York: Free Press. Brewer, M. B., and N. Miller. 1984. “Beyond the Contact Hypothesis: Theoretical Perspectives on Desegregation.” In N. Miller and M. Brewer, Groups in Contact: The Psychology of Desegregation. Orlando, FL: Academic. Cairns, E. 1982. “Intergroup Conflict and Northern Ireland.” In H. Tajfel, ed., Social Identity and Intergroup Relations. Cambridge: Cambridge University Press. Cairns, E., J. Kenworthy, A. Campbell, and M. Hewstone. 2006. “The Role of In-group Identification, Religious Group Membership and Intergroup Conflict in Moderating In-group and Out-group Affect.” British Journal of Social Psychology 45: 701–16. Cairns, E., R. Wilson, T. Gallagher, and K. Trew. 1995. “Psychology’s Contribution to Understanding Conflict in Northern Ireland.” Peace and Conflict: Journal of Peace Psychology 1: 131– 48. Dixon, J., K. Durrheim, and C. Tredoux. 2005. “Beyond the Optimal Contact Strategy: A Reality Check for the Contact Hypothesis.” American Psychologist 60: 697–711. Gaertner, S., J. F. Dovidio, P. A. Anastasio, B. A. Bachevan, and M. C. Rust. 1993. “The Common Ingroup Identity Model: Recategorization and the Reduction of Intergroup bias.” In W. Stroewe and M. Hewstone, eds., European Review of Social Psychology, 4:1–26. Chichester: Wiley. Gormley-Heenan, C. 2005. Political Leadership and the Northern Ireland Peace Process: Role, Capacity and Effect. Basingstoke: Palgrave Macmillan. Hajnal, Z. L. 2001. “White Residents, Black Incumbents, and a Declining Racial Divide.” American Political Science Review 95, 3: 603–17. Haslam, S. A. 2001. Psychology in Organizations: The Social Identity Approach. London: Sage Publications.

292

Ed Cairns

———. 2004. Psychology in Organizations: The Social Identity Approach. 2nd ed. London: Sage Publications. Hayes, B. C., I. McAllister, and L. Dowds. 2005. “The Erosion of Consent: Protestant Disillusionment with the 1998 Northern Ireland Agreement.” Journal of Elections, Public Opinion and Parties 15, 2: 147– 67. Heskin, K., E. Cairns, and D. McCourt. 1990. “Perceptions of Political Leaders in Northern Ireland: A Factor-Analytic Study.” Irish Journal of Psychology 11, 4: 354– 66. Hewstone, M., and R. Brown. 1986. “Contact Is Not Enough: An Intergroup Perspective on the Contact Hypothesis.” In M. Hewstone and R. Brown, eds., Contact and Conflict in Intergroup Encounters, 3– 44. Oxford: Basil Blackwell. Hewstone, M., and E. Cairns. 2001. “Social Psychology and Inter-group Conflict.” In D. Chirot and M.E.P. Seligman, eds., Ethno-Political Warfare: Causes and Consequences, and Possible Solutions, 319– 42. Washington, DC: American Psychological Association. Hewstone, M., E. Cairns, A. Voci, J. Hamberger, and U. Niens. 2006. “Intergroup Contact, Forgiveness, and Experience of ’The Troubles’ in Northern Ireland.” Journal of Social Issues 62: 9–120. Hewstone, M., J. B. Kenworthy, E. Cairns, N. Tausch, J. Hughes, T. Tam, A. Voci, U. von Hecker, and C. Pinder. 2008. “Stepping Stones to Reconciliation in Northern Ireland: Intergroup Contact, Forgiveness, and Trust.” In A. Nadler, T. E. Malloy, and J. D. Fisher, eds., Social Psychology of Intergroup Reconciliation, 199–226. New York: Oxford University Press. Hogg, M. A., R. Martin, O. Epitropaki, A. Mankad, A. Svensson, and Karen Weeden. 2005. “Effective Leadership in Salient Groups: Revisiting Leader-Member Exchange Theory from the Perspective of the Social Identity Theory of Leadership.” Personality & Social Psychology Review 31, 7: 991–1004. Hogg, M. A., and S. A. Reid. 2006. “Social Identity, Self-Categorization, and the Communication of Group Norms.” Communication Theory 16: 7–30. Klein, O., R. Spears, and Stephen Reicher. 2007. “Social Identity Per for mance: Extending the Strategic Side of SIDE.” Personality and Social Psychology Review 11: 28– 45. Livingstone, A., and S. A. Haslam. 2008. “The Importance of Social Identity Content in a Setting of Chronic Social Conflict: Understanding Intergroup Relations in Northern Ireland.” British Journal of Social Psychology 47: 1–21. McGarry, J., and B. O’Leary. 2006. “Consociational Theory, Northern Ireland’s Conflict, and Its Agreement 2: What Critics of Consociation Can Learn from Northern Ireland.” Government and Opposition 41, 2: 249–77. Niens, U., E. Cairns, and M. Hewstone. 2003. “Contact and Conflict in Northern Ireland.” In O. Hargie and D. Dickson, eds., Researching the Troubles: Social Science Perspectives on the Northern Ireland Conflict, 123– 40. Edinburgh: Mainstream Publishing.

Northern Ireland

293

Norris, P. 2005. Stable Democracy and Good Governance in Divided Societies: Do Powersharing Institutions Work? February. RWP05- 014 Harvard John F. Kennedy School of Government. Working Paper. Office of the First and Deputy First Minister (OFMDFM). 2002. Review of Community Relations Policy, Report of the Review Team. Available at http://www.ofmd fmni.gov.uk /communityrelationsunit/harbisonreportannex.pdf. Paolini, S., M. Hewstone, A. Voci, J. Harwood, and E. Cairns. 2006. “Intergroup Contact and the Promotion of Intergroup Harmony: The Influence of Intergroup Emotions.” In R. J. Brown and D. Capozza, eds., Social Identities: Motivational, Emotional, Cultural Influences. Hove, England: Psychology Press. Pettigrew, T. F. 1986. “The Intergroup Contact Hypothesis Reconsidered.” In M. Hewstone and R. Brown, eds., Contact and Conflict in Intergroup Encounters. Oxford: Basil Blackwell. ———. 1998. “Intergroup Contact Theory.” Annual Review of Psychology 49: 65–85. Pettigrew, T. F., and L. R. Tropp. 2000. “Does Intergroup Contact Reduce Prejudice?: Recent Meta-analytic Findings.” In S. Oskamp, ed., Reducing Prejudice and Discrimination. Mahwah, NJ: Erlbaum. Platow, M .J., S. D. Reicher, and S. A. Haslam. 2008. “On the Social Psychology of Intergroup Leadership: The Importance of Social Identity and Self-Categorization Processes.” Unpublished paper. Reicher, S., S. A. Haslam, and N. Hopkins. 2005. “Social Identity and the Dynamics of Leadership: Leaders and Followers as Collaborative Agents in the Transformation of Social Reality.” Leadership Quarterly 16: 547– 68. Rose, R. 1971. Governing Without Consensus. London: Faber. Schneckener, U. 2002. “Making Power Sharing Work: Lessons from Successes and Failures in Ethnic Conflict Regulation.” Journal of Peace Research 39, 2: 203–28. Shirlow, P. 2003. “ ‘Who Fears to Speak’: Fear, Mobility, and Ethno-Sectarianism in the Two ‘Ardoynes.’ ” Global Review of Ethnopolitics 3, 1: 76–91. Shuttleworth, I. 2007. “Are Northern Ireland’s Communities Dividing? Evidence from the Census of Population 1971–2001.” Unpublished paper, Queen’s University Belfast. Sisk, Timothy D. 2003. “Power Sharing.” In Guy Burgess and Heidi Burgess, eds., Beyond Intractability. Conflict Research Consortium, University of Colorado, Boulder. www.BeyondIntractability.org Spotlight. 2007. Television program, BBC NI, Northern Ireland, 5 September. Tajfel, H. 1978. Differentiation Between Social Groups: Studies in the Social Psychology of Intergroup Relations. London: Academic Press. Tajfel, H., and J. V. Turner. 1986. “The Social Identity Theory of Intergroup Behaviour.” In S. Worchel and W. G. Austin, eds., Psychology of Intergroup Relations. 2nd ed. Chicago: Nelson-Hall. Tausch, N., M. Hewstone, J. Kenworthy, E. Cairns, and O. Christ. 2006. “CrossCommunity Contact, Perceived Status Differences, and Intergroup Attitudes in

294

Ed Cairns

Northern Ireland: The Mediating Roles of Individual-level versus Group-level Threats and the Moderating Role of Social Identification.” Political Psychology 28 (1):53– 68. Trew, K. 1996. “Complementary or Conflicting Identities?” Psychologist 9: 460– 63. Trew, K., and D. Benson. 1996. “Dimensions of Social Identity in Northern Ireland.” In G. M. Breakwell and E. Lyons, eds., Changing European Identities, 123–43. Oxford: Butterworth-Heinemann. Turner, J. C. 1982. “Towards a Cognitive Redefinition of the Social Group.” In H. Tajfel, ed., Social Identity and Intergroup Relations. Cambridge: Cambridge University Press. ———. 1985. “Social Categorization and the Self-Concept: A Social-Cognitive Theory of Group Behavior.” In E. J. Lawler, ed., Advances in Group Processes: Theory and Research, 2:77–122. Greenwich, CT: JAI Press. ———. 1987. “A Self-Categorization Theory.” In J. C. Turner, M. A. Hogg, P. J. Oakes, S. D. Reicher, and M. S. Wetherell, eds., Rediscovering the Social Group: A SelfCategorization Theory. Oxford: Blackwell. Turner, J. C., M. A. Hogg, P. J. Oakes, S. D. Reicher, and M. S. Wetherell, eds. 1987. Rediscovering the Social Group: A Self-Categorization Theory. Oxford: Blackwell. Turner, J. C., and R. S. Onorato. 1999. “Social Identity, Personality and the Self-Concept: A Self-Categorization Perspective.” In T. R. Tyler, R. Kramer, and O. Johns, eds., The Psychology of the Social Self. Mahwah, NJ: Erlbaum. Turner, J. C., and K. J. Reynolds. 2001. “The Social Identity Perspective in Intergroup Relations: Theories, Themes and Controversies.” In R. Brown and S. Gaertner, eds., Blackwell Handbook of Social Psychology. Vol. 4: Intergroup Processes. Oxford: Blackwell. Whyte, J. 1990. Interpreting Northern Ireland. Oxford: Clarendon Press. Wright, S. C., A. Aron, T. McLaughlin-Volpe, and S. A. Ropp. 1997. “The Extended Contact Effect: Knowledge of Cross-Group Friendships and Prejudice.” Journal of Personality and Social Psychology 73: 73–90.

CHAPTER 12

Public Opinion and Power Sharing in Deeply Divided Places Colin Irwin

Power sharing is generally thought of in terms of various constitutional arrangements made for different political parties representing contending groups to share power. But does it need to end there? Can power sharing be extended to include “the people” in some way, and if this is done is it a good idea? Will it help the political process or will it make decision making in deeply divided places more difficult? Public opinion and public opinion research were used to help the Northern Ireland parties negotiating the Belfast Agreement reach their historic accord in April 1998. Since then these same methods have been used in Macedonia, Bosnia and Herzegovina, Kosovo and Serbia, Kashmir, and Sri Lanka with varying degrees of success. Th is chapter will review these experiments in public opinion and power sharing with a view to better understanding how, why, and when bringing “the people” into the decision-making process can have political benefits and when and why the results are sometimes disappointing.

Northern Ireland All aspects of cultural, social, and political life tend to get polarized in deeply divided places, and in most such places this process extends to a partisan media, the public opinion polls they run, and the research done in support

296

Colin Irwin

of those polls. Even the academic research can be polarized in this way. Sometimes I would say to a close colleague and friend in Belfast, to underline this point, “Jim [it wasn’t Jim but someone else], that is just a Protestant piece of research carried out by a Protestant researcher for a Protestant audience” (“Jim” was Protestant). For “Protestant” substitute “Catholic,” “Israeli,” “Palestinian,” or an ethnonational identity as required. Such research does not help the conflicting communities understand each other and the deepest concerns of their respective constituencies. Unfortunately, sometimes when academics do get together across a divide to address each other’s problems in a joint piece of research they can also miss their target by carefully avoiding the most sensitive questions that may be the most critical issues that need to be addressed. Between the failings of partisanship on the one hand and good manners and the politics of the street on the other, efforts to bring the views of the people to the negotiating table in a useful and constructive way is both difficult and rare. But it was done in Northern Ireland by bringing best practice in power sharing to the research process. I have explained these methods at length elsewhere (Irwin, 2002a, 2001, 1999) so I will not go into any great detail here except to review some of the essential features of the process as follows: 1. The parties elected to negotiate the Belfast Agreement were all asked if they would like to participate in a joint piece of public opinion research and associated program of public diplomacy. 2. Each party nominated a member of their negotiating team to work with me on the project as the facilitator. 3. All aspects of the research and public diplomacy were agreed upon among the parties including timing of the research, means of publication, polling contractors, sources of funding, sampling, and the questions to be asked. 4. All matters so agreed were done through private, confidential, faceto-face meetings so that sensitive issues could be explored without the risk of interference from “spoilers.” 5. A new style of polling, better adjusted to the needs of negotiators, was used that abandoned the social psychology/market research model developed in the United States in favor of a “problems” and “solutions” model that tested options against a simple five-point scale of “essential,” “desirable,” “acceptable,” “tolerable,” and “unacceptable.”

Public Opinion

297

Nine polls were run with the parties in this way to solve specific problems in the negotiations at the time of each poll. Poll 1 was a demonstration poll dealing with a wide range of issues to attract the interest of both the public and politicians. Polls 2 and 3 addressed procedural or “shape of the table” issues. Poll 4 explored all the critical elements of a comprehensive settlement, and poll 5 tested the results of that poll as a “package.” Polls 6, 7, 8, and 9 dealt with problems of implementation following the signing of the agreement and referendum. For the purposes of comparison and because this book focuses on power sharing, I will illustrate the methods used with one of the key questions from the Northern Ireland polls that deals with this issue (there were hundreds on all manner of topics) by comparing the central constitutional proposal against its alternates (Table 12.1). Critically, “power-sharing with North-South institutions but no joint authority” comes in with the lowest percentage of “unacceptable” at 40 percent. When this is broken down for the Catholic and Protestant communities (Table 12.2), it is easy to see that this option was the only workable solution for an agreement. But the result of the referendum was 71 percent “Yes” and only 29 percent “No” so how did we get from 40 percent “unacceptable” to 29 percent “No”? The answer is by adding in other features to the comprehensive peace agreement so that the overall “package” was greater than the sum of its individual parts; this “package” (Table 12.3) was tested against Northern Ireland public opinion shortly before the agreement was made. Table 12.4 brings together the results of this poll and the referendum of May 22. Because an additional 26 percent of the Democratic Unionist Party (DUP) electorate said they would not vote for the Belfast Agreement if their party did not support it (50 percent down to 24 percent), the difference between the public opinion poll and referendum results can be calculated with some precision. With the anti-power-sharing DUP taking about 18 percent of the vote, a reduction of about 6 percent from 77 percent in the poll to 71 percent in the referendum was to be expected. These results are all within the margins of error. The results were published in the Belfast Telegraph on March 31, 1998, eleven days before the deal was struck on Good Friday, April 10, 1998, and fi ft y-three days before the referendum of May 22, 1998. To lose the referendum would have been political suicide for any party supporting the agreement, but the risks had been calculated with considerable precision and the parties knew they could win providing they

Separate Northern Irish state: The complete separation of Northern Ireland from both the United Kingdom and the Republic of Ireland and the establishment of a separate state within the European Union. Full incorporation into the British state: Direct rule from Westminster and local government similar to the rest of the United Kingdom with no Northern Ireland Assembly or separate laws for Northern Ireland and no Anglo-Irish Agreement. Continued direct rule (no change): The continuation of direct rule from London in consultation with the Irish government under the terms of the Anglo-Irish Agreement. Power sharing and the Anglo-Irish Agreement: Government by a Northern Ireland Assembly and power-sharing executive under the authority of the British government but in consultation with the Irish government under the terms of the Anglo-Irish Agreement. Power sharing with North-South institutions but no joint authority: Government by a Northern Ireland Assembly, power-sharing executive, and a number of joint institutions established with the Republic of Ireland to deal with matters of mutual interest. (But these arrangements will not include joint authority between the British and Irish governments.) Joint authority and power sharing: Government by joint authority between the British and Irish governments in association with an elected powersharing executive and assembly. Separate institutions for the two main communities: Creation of separate structures for the government of each of the two main communities in Northern Ireland, subject to joint authority by the British and Irish governments. Full incorporation into the Irish state: Full incorporation of Northern Ireland into the Republic of Ireland to create a single state within the European Union.

8

14

8

8

11

13

5

12

13

2

3

3

4

3

14

Desirable

3

Essential

9

17

20

23

24

21

18

17

Acceptable

9

20

14

23

23

25

16

15

Tolerable

56

55

49

40

42

44

39

57

Unacceptable

Table 12.1. Constitutional Proposals for the Future of Northern Ireland Tested Against Each Other as Percent “Essential,” “Desirable,” “Acceptable,” “Tolerable,” and “Unacceptable”

Public Opinion

299

Table 12.2. Catholic and Protestant Percent “Unacceptable” for Full Incorporation into the British State, the Irish State and the Alternative of Power Sharing

Full incorporation into the British state Power sharing with North-South institutions but no joint authority Full incorporation into the Irish state

Catholic Unacceptable

Protestant Unacceptable

75 27

15 52

10

90

Table 12.3. A Comprehensive Settlement Tested Against Public Opinion as a “Package” A COMPREHENSIVE SETTLEMENT • A REGIONAL ASSEMBLY made up from elected members who share responsibilities in proportion to their representation and employing a voting system, with other checks and balances, to ensure the fair participation of the whole community in government and the prevention of abuse of power. • NORTH/SOUTH BODIES strictly controlled by the elected politicians who establish them to deal with a wide range of issues using various functions and powers appropriate to the areas of government policy being managed. • Replace the Anglo-Irish Agreement and establish a “COUNCIL OF THE ISLES” to create a new relationship between London, Dublin, Cardiff, Edinburgh, and Belfast appropriate to the needs of the region as a part of Europe. • CONSTITUTIONAL REFORM that embraces the principle of consent of a majority of the people of Northern Ireland to keep or change its status, guaranteed rights of British and/or Irish citizenship, and any other balanced changes required to implement the various agreements made at the Stormont Talks. • A BILL OF RIGHTS that deals specifically with the political, social, and cultural problems that have aggravated the conflict and a Human Rights Commission with responsibilities and powers to educate, monitor standards, and bring cases to court. • REFORM THE RUC [Royal Ulster Constabulary] to create community policing units as part of a two-tier ser vice restructured with a view to recruiting more Catholics and improving community relations under the authority of a new Department of Justice in a Regional Assembly.

stuck together in sufficient numbers. When it comes to undertaking research in support of power sharing, this is, I would like to suggest, “rocket science.” So why have these methods not been systematically adopted in all such negotiations since the signing of the Belfast Agreement with similar success?

77 23

All of NI

74 26

Protestant 81 19

Catholic 50 50

DUP 68 32

PUP + UDP 83 17

UUP 96 4

Alliance 95 5

SDLP

61 39

Sinn Féin

50 50

All of NI

71 29

All of NI

46 54

Protestant 53 47

Catholic 24 76

DUP 24 76

PUP + UDP 50 50

UUP

87 13

Alliance

70 30

SDLP

22 78

Sinn Féin

Parties are as follows: Democratic Unionist Party (DUP); Progressive Unionist Party (PUP); Ulster Democratic Party (UDP); Ulster Unionist Party (UUP); Social Democratic and Labour Party (SDLP); and Alliance Party of Northern Ireland (Alliance).



Yes No

Percent

Referendum: May 22, 1998

Yes No

Percent

Question: “If you said ‘Yes’ would you still accept these terms for a settlement even if the political party you supported was opposed to them?” (March 1998)

Yes No

Percent

Question: “If a majority of the political parties elected to take part in the Stormont Talks agreed to this settlement would you vote to accept it in a referendum?” (March 1998)

Table 12.4. Support for the Belfast Agreement and Referendum Result

Public Opinion

301

The Balkans Following the signing of the Belfast Agreement, the Queen’s University of Belfast, its Institute of Irish Studies, and its new chancellor, Senator George Mitchell, helped arrange a fellowship for me with Atlantic Philanthropies to extend my work in Northern Ireland to other deeply divided places. I followed Senator Mitchell to Israel and Palestine, but following the end of the Clinton administration and because of serious limits on freedom of association in the West Bank and Gaza, the prospects of progress in the Middle East were very poor indeed. My first opportunity to apply the Northern Ireland methods elsewhere came in the Balkans with a Greek-based NGO, the Centre for Democracy and Reconciliation in South East Europe (CDRSEE). The first in a series of what were now being termed “peace polls” was run in Macedonia in 2002. It was used to analyze the political situation in the country and identify possible solutions to difficulties. It was successful inasmuch as it drew attention to the desire of both the Albanian and Macedonian communities for free and fair elections, which the European Union was more than willing to support (Irwin, 2002b, 2002c). No attempt was made to explore new constitutional arrangements for Macedonia in this poll given the existence of power sharing since the early 1990s and increased autonomy/decentralization for the ethnic Albanian community under the Ohrid Agreement of 2001. In a poll run in Bosnia and Herzegovina in 2005 (Irwin 2005a) all the major possibilities for constitutional reform were tested against each other in the context of increased debate over the reform of the Dayton Peace Agreement brokered by the United States and the European Union in 1995 (Table 12.5). Here the point of least resistance was for “Bosnia and Herzegovina with decentralized regions in accordance with European standards” with only 24 percent considering this option “unacceptable” overall and only rising to a maximum of 39 percent for the Croats. Everyone wanted to be a part of Europe, but everyone also wanted the terms of settlement under the Dayton Agreement to be simplified (Table 12.6). Although the members of the international community are pursuing these objectives, they are made more complicated by the situation in Kosovo and Serbia where, unlike the Serbs in Republika Srpska, the Kosovo Albanians have been given independence (Irwin 2005b). A peace poll run in Kosovo and Serbia just before the negotiations that led to the independence of Kosovo began clearly demonstrated that the situation there was very difficult indeed.

302

Colin Irwin

Table 12.5. Options for Constitutional Reform in Bosnia and Herzegovina (percent “unacceptable”)

Bosnia and Herzegovina as it was before the war during the existence of the Socialist Federal Republic of Yugoslavia SFRJ Bosnia and Herzegovina with decentralized regions in accordance with European standards Bosnia and Herzegovina as a decentralized state with powers going to the municipalities instead of the entities and cantons, which will go. Bosnia and Herzegovina as it is now with two entities and the District of Brcko. Bosnia and Herzegovina with three entities, one each for three constituent peoples The abolition of cantons and a federation between Bosnia Herzegovina and Republika Srpska as two entities Bosnia and Herzegovina made up of just cantons without entities or a district Bosnia and Herzegovina made up of a large number of federal units with equal powers Bosnia and Herzegovina as it is now with two entities and the District of Brcko but with the higher levels of responsibilities given to the state Separation and union of some parts of Bosnia and Herzegovina with neighboring states

All

Bosniak

Serb

Croat

26

11

32

60

24

17

28

39

37

17

52

66

32

40

10

60

39

55

22

22

41

40

31

69

38

22

53

51

39

25

47

58

35

34

25

62

59

71

36

72

Similar to Table 12.1 for Northern Ireland, Table 12.7 lists all the constitutional arrangements that could possibly be considered. The best option appeared to be “A protectorate of the EU,” and in practice this seems to be what they now have. Here levels of “unacceptable” rise as high as 60 percent for the Serbian Serbs. Strengthening the package with effective reforms to protect the rights of the Serb minority would reduce this level of resistance. The peace poll explored such options, many of which were incorporated into the Final Status Agreement. But in the end the international community chose to impose their will in the case of Kosovo rather than finesse these arrangements to achieve broader Serbian support. The EU and NATO will be in Kosovo for a very long time.

Public Opinion

303

Table 12.6. Priorities for Reform of the Bosnia and Herzegovina Constitution

Fewer levels of government Establish a Constitutional Commission to advise on reform Simplify government ser vices to municipal level Do not duplicate ser vices in entities, cantons, and municipalities Reform the constitution through Parliament Zagreb, Belgrade, and Sarajevo should cooperate to join the EU together

Essential

Desirable

Acceptable

Tolerable

Unacceptable

47 45

31 32

17 16

3 4

2 2

44

39

13

3

1

44

34

15

4

2

41

32

17

6

3

51

27

15

4

4

The final question in this poll asked persons living in Kosovo and those who might return there a very idealistic question that did not focus on constitutional arrangements as such but on security and equality as best understood in terms of human rights. In this context 83 percent of Kosovo Albanians said they would stay while 71 percent of Kosovo Serbs and 61 percent of Serb Internally Displaced Persons (IDPs) would return (Table 12.8). This result might indicate that power sharing—or for that matter any other set of constitutional arrangements in deeply divided places—is not an end in itself but only a means to an end and that the story of Kosovo, Serbia, and power sharing may not yet have reached its conclusion.

Kashmir and Sri Lanka In 2008 the methods used in Northern Ireland and the Balkans were extended to Kashmir on the Indian side of the Line of Control as well as to Sri Lanka. Doubts had been raised about the ability of informants to answer what appeared to be complex constitutional questions in the rural areas of Kashmir and Sri Lanka. But in practice this did not prove to be an insurmountable problem. Providing the questions addressed concrete issues (e.g., who shared

Union of all Serbian lands Full integration of Kosovo into Serbia A republic in Serbia Montenegro with control of all aspects of government locally (1974 constitution) A republic in Serbia Montenegro with control of all aspects of government locally and regional status in the EU A protectorate of the EU An economic union of independent states of Kosovo, Montenegro, and Serbia A state as part of the EU but North Kosovo joins Serbia A state as part of the EU with choice of citizenship for Serbs who will share their canton in North Kosovo with Serbia under joint authority A state as part of the EU with choice of citizenship for Serbs who will have their own canton in Kosovo A state as part of the EU with choice of citizenship for Serbs Full independence and no choice of citizenship for Serbs in Kosovo Union of Kosovo with Albania Union of all Albanian lands

13 3 66 63 58 88 57 59

48 64 94 98 98

92 20 34 96 87

63 52 13 10 8

Kosovo Serb

96 98 96

Kosovo Albanian

42 33

81 44

70

89 100

57 54

91

79 77 94

Kosovo Other

96 95

67 93

65

55 62

60 68

43

13 4 40

Serbia Serb

Table 12.7. Constitutional Arrangements for the Future of Kosovo and Serbia (percent “unacceptable”)

97 97

65 88

62

60 61

58 66

46

13 1 35

Serbia Serb IDPs

88 91

45 81

44

41 45

39 42

31

34 16 29

Serbia Others

Public Opinion

305

Table 12.8. Ideal Conditions to Stay in/Return to Kosovo (percent “yes”)

It does not matter so much about the constitution. I would (stay in/return to) Kosovo providing I felt completely safe there, could choose my citizenship, and was free to work and practice my culture, language, and religion without any fear of discrimination.

Kosovo Albanian

Kosovo Serb

Kosovo Other

Serbia Serb IDPs

83

71

100

61

what powers) and not abstract concepts (e.g., “federalism”), informants were generally able to answer. Other research seems to confirm this observation as a universal global phenomenon (Miller 2008). The research in Kashmir (Irwin 2008a) threw up one slightly surprising result, which was the strength of opposition to a union with Pakistan by the Muslim community in Indian-administered Kashmir (IaK) at 71 percent “unacceptable” (see Table 12.9). What they wanted was independence. But the Hindu and Buddhist minorities were strongly opposed to this solution, particularly the Buddhists who had left Tibet because of Chinese occupation and saw what happened to their world-famous Buddhist shrines in Afghanistan under the Taliban. So the best solution for this part of Kashmir may be increased autonomy with local/regional power sharing under the protection of the Indian secular constitution. All who were interviewed were asked if they would like the poll to be run in Pakistan-administered Kashmir (PaK). Seventy-four percent said “Yes” (77 percent Muslim, 69 percent Hindu, and 74 percent Buddhist), but this has not been done yet owing in part to the increase in tensions in the region and the change of government in Pakistan in 2008. The Kashmir peace poll clearly demonstrates the viability of such work as a positive contribution to a local/regional peace process, but until the international community takes a more proactive role in this regard little progress can be made. The Sri Lanka peace poll (Irwin 2008b, 2008c) clearly demonstrates wide public support for power sharing through a process of devolution that would provide the Tamil minority in the north and east with control of a regional provincial government and seats in an all-island government in Colombo where there is also a significant Tamil population (Table 12.10).

Join Pakistan: All of Jammu and Kashmir should become a part of Pakistan like any other Pakistan province. Full independence: All five districts should join to become the independent state of Kashmir with responsibility for both their domestic and foreign policy and protecting their borders with Pakistan, India, and China. Disintegration: Each of the five districts should be allowed to choose their own future with Pakistan or India. Regional integration and devolution: Pakistan and Indian Kashmir should function like a cofederation with an open border and decentralization/local control in all regions, districts, and blocs. No change: The status quo should stay the same with present central, state, and regional arrangements for governance. Autonomy: Full implementation of Article 370 and return to the status existing in J and K before 1953 with a parliament and prime minister, leaving only defense, foreign policy, and communications to India. Join India: All of J and K should become a part of India like any other Indian state. 71

16

49

27

47

27

49

32

49

32

34

23

34

Muslim

74

All

10

15

12

40

50

58

78

Hindu

13

61

3

55

63

74

84

Buddhist

Table 12.9. Constitutional Arrangements for the Future of Kashmir (percent “unacceptable”)

63

34

58

22

53

7

69

Valley

13

14

17

39

47

50

77

Jammu

13

33

3

49

48

62

87

Ladakh

Two states: Two completely separate independent states of Tamil Eelam and Sri Lanka. Confederal state: Two autonomous units comprising the northeast and the rest of Sri Lanka with a minimum of functions for the joint central government. Federal state: A number of autonomous units comprising the northeast and existing provinces in the rest of Sri Lanka with a joint central government sharing power with the autonomous units. Enhanced devolution: Full implementation of the 13th and 17th amendments plus the devolution of significant powers to autonomous provinces negotiated at a peace conference. 13th Amendment devolution: Present constitution with full implementation of the 13th and 17th amendments. Unitary state: Pre-1987 constitution.

40 12

12

33

38 54

91

68

31

24 16

Tamil

95

Sinhala

83

12

7

25

38

27

Northern Tamil

Table 12.10. Constitutional Arrangements for the Future of Sri Lanka (percent “unacceptable”)

37

47

48

21

19

53

Up-Country Tamil

63

38

42

41

58

90

Muslim

308

Colin Irwin

The ongoing war at the time this poll was run in 2008 necessitated the separate organization of two polls (using the same questionnaire): one undertaken by Social Indicator of Colombo for all of Sri Lanka (including Tamils) but without the Northern Province and one completed by the academic community with connections to the University of Jaff na (Northern Tamils). The results of these polls suggest the government of Sri Lanka could easily win a referendum on the proposals being developed by their All Party Representative Committee (APRC) for constitutional reform and that such a referendum (as in Northern Ireland) could help politically marginalize the extremists in both communities.

Conclusion Table 12.11 brings together the levels of “unacceptable” for the various forms of power sharing offered in these different polls in their broadest possible constitutional terms. All of these results can be improved by adding to the respective “packages” measures such as strengthened good governance, minority protections, effective policies to deal with discrimination, and police reform. For instance, power sharing in Northern Ireland (including institutionalized arrangements with the Republic of Ireland), which polled at 40 percent “unacceptable,” was reduced to only 29 percent “No” to the agreement in the May 1998 referendum. The only case that appeared to be more difficult than Northern Ireland from a public opinion perspective was Kosovo and Serbia. Here levels of “unacceptable” did not fall below 60 percent for Serbians for the “EU Protectorate” option and rose to 93 percent “unacceptable” for “Full independence” (Table 12.7). But this is what the states that support NATO in Kosovo wanted so this is what they got. From a public opinion perspective, reforming and simplifying the Dayton Agreement would be welcomed in Bosnia and Herzegovina. Here the obstacles to such reform are the political elites who have a vested interest in the various pieces of the political pie they were given under the terms of that agreement and the complications introduced by the independence of Kosovo and the international implications this has for Republika Srpska. In this case the smooth course of power sharing is clearly not being held up by the people but by their local political representatives, some of whom are beholden to their NATO, U.S., EU, and Russian allies, respectively.

Public Opinion

309

Table 12.11. Resistance to Peace/Power-Sharing Agreements and State of Peace Processes (percent “unacceptable”) Northern Ireland (relatively stable) Belfast Agreement

All 40

Catholic 27

Protestant 52

Bosnia and Herzegovina (unstable) Decentralized EU regions

All 24

Bosniak 17

Serb 28

Croat 39   

All 23

Kosovo Albanians 20 Muslim 27

Kosovo Serbs 58 Hindu 15

Serbia Serbs 60 Buddhist 61

Sri Lanka (unresolved) Enhanced devolution

Sinhala 31

N. Tamil 7

Cyprus (unresolved) “Bi-communal, bi-zonal entity with strong central government” (Office of Research 1999)

Greek Cypriot 31

Turkish Cypriot 38

Kosovo (imposed solution) EU Protectorate Kashmir (IaK) (unresolved) Autonomy

Serbia Serb IDPs 58

Kashmir is caught up in a regional conflict. A power-sharing agreement of some form would seem to be within reach. But as with Kosovo the local political elites have played the ethnic/sectarian card to stay in power and curry favor with Pakistan and India, respectively, or rather the more radical Muslim and Hindu elements in Pakistan and India. The Buddhists are caught in between both this conflict and relations between themselves, India, and China. Cyprus, arguably, is also another victim of local and regional politics, but at least the people there are not killing each other (see Lordos 2004 and 2005 for peace polls in Cyprus). The most tragic and unnecessary conflict of them all that I have studied is Sri Lanka. It is arguably an island paradise whose people would be quite content with a well governed system of provincial power-sharing arrangements agreed to at a peace conference, and to this end they have been failed by a succession of governments that have manipulated their constitution for reasons of electoral advantage. In most of these cases the local politicians blame the international community and/or their people when they fail to share power while the

310

Colin Irwin

international community, if they have a vested interest, blames other sections of the international community, the local politicians, and/or the people if power sharing is not achieved. But the peace polls run in Northern Ireland, the Balkans, Kashmir, Sri Lanka, and Cyprus suggest that the people (with the possible exception of Kosovo) are not to blame; they are being used as scapegoats for the selfish interests of third parties, and perhaps that is why independent polling of this kind is so difficult to fund because those in power who can pay for these polls do not want this truth to come out while those whose only interest is peace do: the Joseph Rowntree Charitable Trust in Northern Ireland, the Centre for Democracy and Reconciliation in South East Eu rope in the Balkans, Yashwant Deshmukh and the Team CVoter Foundation in Kashmir, Alexandros Lordos in Cyprus, and members of the international community who presently wish to stay anonymous in Sri Lanka (for a discussion of the ethics of peace research, see Irwin 2008d).

Postscript Peace polls have since been run on the Pakistan side of the Line of Control (LoC) in Kashmir; in Israel and Palestine; in Sri Lanka to test the All Party Representative Committee’s reform proposals; and tested in Darfur. Muslim respondents prefer to stay with Pakistan, unlike their coreligionists on the Indian side of the LoC who seek independence, and since Hindus and Buddhists prefer India, a simple self-determination referendum would not resolve Kashmir. In Israel and Palestine, polls suggest that domestic publics are amenable to an agreement but regional and international questions make a settlement difficult. The Sri Lanka poll shows concurrent Tamil and Sinhalese support for reform. The Darfur-Darfur Dialogue and Consultation established hearings and a research program including peace polling. These developments confirm that objective polls can help parties reach a settlement, provided they wish to do so. (Editor’s Note: Colin Irwin has since produced two short books: see http://www.peacepolls.org/peacepolls/documents/002539. pdf and http://www.peacepolls.org/peacepolls/documents/002903.pdf.) References Irwin, C. J. 1999. “The People’s Peace Process: Northern Ireland and the Role of Public Opinion Polls in Political Negotiations.” Security Dialogue 30, 3 (September): 105–17.

Public Opinion

311

———. 2001. “How Public Opinion Polls Were Used in Support of the Northern Ireland Peace Process.” Global Review of Ethnopolitics 1, 1: 62–73. http://www .ethnopolitics.org/ethnopolitics/archive.html (accessed 6 August 2006). ———. 2002a. The People’s Peace Process in Northern Ireland. New York: Palgrave Macmillan. ———. 2002b. “Forum Macedonia: An Opinion Poll and Its Implications.” Global Review of Ethnopolitics 2, 1 (September). http://www.ethnopolitics.org/ethnopolitics /archive.html (accessed 6 August 2006). ———. 2002c. “Forum Macedonia: Reply—Making Dreams Come True.” Global Review of Ethnopolitics 2, 1 (September). http://www.ethnopolitics.org/ethnopolitics /archive.html (accessed 6 August 2006). ———. 2005a. “A People’s Peace Process for Bosnia and Herzegovina?” Ethnopolitics 4, 3: 1–18. http://www.ethnopolitics.org/ethnopolitics/archive.html (accessed 6 August 2006). ———. 2005b. Coming to Terms with the Problem of Kosovo: The Peoples’ Views from Kosovo and Serbia. http://www.peacepolls.org (accessed 6 August 2006). ———. 2008a. Peace in Kashmir: Myth and Reality. http://www.peacepolls.org/docu ments/peacepolls/000295.pdf (accessed 30 September 2008). ———. 2008b. Peace in Sri Lanka: From Symbols to Substance. http://www.peacepolls .org/documents/peacepolls/000277.pdf (accessed 30 September 2008). ———. 2008c. Peace in Sri Lanka: Negotiating with Northern “Separatists”? http://www .peacepolls.org/documents/peacepolls/000321.pdf (accessed 30 September 2008). ———. 2008d. “Research Ethics and Peacemaking.” In D. Mertens and P. Ginsberg, eds., The Handbook of Social Research Ethics. Los Angeles, London, New Delhi, Singapore, Washington, DC: Sage. Lordos, A. 2004. Can the Cyprus Problem Be Solved? Understanding the Greek Cypriot Response to the UN Peace Plan for Cyprus. http://www.help-net.gr/download.htm (accessed 6 August 2006). ———. 2005. Civil Society Diplomacy: A New Approach for Cyprus? http://www.help-net .gr/download.htm (accessed 6 August 2006). Miller, J. D. 2008. Cross-national Attitude Mea surement: The Impact of Salience and Information. WAPOR Quality Criteria in Survey Research Seminar, Cadenabbia, Italy, 10–12 July. Office of Research. 1999. Cypriot Opinion Differences Persist, Limited Public Support for Resolution Apparent. Washington, DC: Department of State. 29 November.

CHAPTER 13

The Balkans: The Promotion of Power Sharing by Outsiders Florian Bieber

Former Yugoslavia has been a fertile ground for experimentation with power sharing since the mid-1990s. Proposals for varying forms of power sharing have been made by international actors in numerous peace plans and by domestic actors for nearly every country or region that emerged from Yugoslavia. The power-sharing arrangements that emerged in former Yugoslavia were established to accommodate competing self-determination claims and/or to provide nondominant groups better and guaranteed access to governance. As such, these arrangements were all tools of confl ict management, even if not all of them came about as a consequence of violent conflict. Today Bosnia and Herzegovina (BiH), Kosovo, and Macedonia are governed by varying forms of power sharing. In addition, Serbia and Montenegro (SCG) briefly experimented with aspects of power sharing before disintegrating in 2006 following Montenegro’s referendum on independence. All experiments with power sharing owe their origins to foreign imposition and have relied on continued third-party intervention for their survival. The power-sharing arrangements vary greatly in terms of the degree to which they provide for the inclusion of different ethnic groups to parliament and government and the measure of decentralization. It would thus be misleading to characterize power sharing in former Yugoslavia as following a single template. As I have argued elsewhere, the systems apply different forms of inclusion and cooperation (Bieber 2005, 87–88). This variation is a function of (a) the nature of the conflict; (b) the group power and demands at the

The Balkans

313

conclusion of the power-sharing arrangement; and (c) the proposals and approaches of international actors. This chapter argues that there are no viable and legitimate alternatives to power sharing in most cases. As such, it rejects the arguments of some recent works that have dismissed power sharing outright (Rothchild and Roeder 2005, 1–25). Nevertheless, the success of power-sharing systems in former Yugoslavia has been modest: power sharing mostly continues to rely on strong external intervention, and the political systems and countries remain profoundly contested. Some of the blame needs to be placed on badly designed systems, but not all the problems of postconflict stability derive from power sharing alone. The chapter discusses four core aspects that are often identified with the failure (or at least lack of success) of power-sharing regimes in former Yugoslavia: the role of international actors in the establishment and implementation of the different power-sharing arrangements; veto rights and decision-making mechanisms; forms of autonomy and its relationship to the strength of the state. In conclusion, the chapter explores the underlying problem of a number of power-sharing arrangements, namely the lack of consensus among the political elites from the different communities over the state and the institutional setup. The discussion seeks to identify the difficulties that the countries of former Yugoslavia are still confronted with today and to what degree the experience with power sharing has a broader significance for our understanding of accommodating diversity in deeply divided places.

The Role of International Actors Despite relying on foreign imposition, power sharing and interethnic accommodation have a domestic tradition in former Yugoslavia. Yugoslavia at large had experimented with a decentralized federal system, albeit under an authoritarian framework. Furthermore, ethnic representation—known as the ethnic key—and group rights were prominent in the Yugoslav system until its dissolution in 1991. This legacy has been a source of both accommodation and instability. In particular, territorial decentralization is often rejected by majorities as the dissolution of Yugoslavia appears to suggest the latent secessionist threat deriving from any form of territorial autonomy. The power-sharing arrangements established in BiH, Kosovo, Macedonia, and SCG have all been the result of different forms of international

314

Florian Bieber

intervention. The pattern of intervention has varied. In the case of Kosovo, the first moderate power-sharing formula was established in 2001 by the Constitutional Framework imposed by the Special Representative of the UN Secretary General (SRSG), Hans Haekkerup. While some consultation with political actors in Kosovo took place, the institutions contained in the Constitutional Framework can be considered an outright imposition. The Dayton Peace Agreement (DPA) for BiH was also based on extensive external—in par ticular U.S.—pressure. Consequently, the constitution contained in the DPA was written in English largely by U.S. State Department lawyers. Still, the parties negotiating the peace settlement in Dayton, Ohio, were part of the process and more profoundly involved than in just a consultative role. In fact the weak central institutions and extensive group protection mechanisms reflect the involvement of the wartime parties. The power-sharing arrangement in Macedonia and SCG saw a considerably less forceful role of international actors. In Macedonia, the Ohrid Framework Agreement was mediated by the EU and NATO in 2001, and a number of core principles of the power-sharing system were proposed by external actors. Nevertheless, the third-party role was diminished by (a) the preexisting elements of power sharing in the system of government, such as the grand coalition of Macedonian and Albanian parties (since 1991), and (b) the fact that the Ohrid Agreement merely outlined the principles for power sharing whereas the details and legally binding provisions were passed by the Macedonian parliament subsequently. The State Union of Serbia and Montenegro was similarly established through the EU-mediated Belgrade Agreement in 2002, which set out the basic principles for the establishment of the new state. The details were subsequently negotiated by the two member states over a one-year period, resulting in the formal establishment of SCG in 2003. Finally, the new constitution of Kosovo, which replaced the 2001 power-sharing arrangement, was formally passed by the Kosovo parliament in April 2008. Although the document was formally adopted by domestic institutions, international actors were heavily involved in its drafting. Unlike the provisions in the other power-sharing systems discussed here, the inclusion of the Serb community in the Kosovo institutions was not negotiated with the community itself. The degree of imposition or meditation of international actors is important in assessing the legitimacy of the agreement to political elites and citizens. Imposed arrangements are more easily dismissed by parties and often do not reflect a commitment to either the system itself or the state at large that they govern.

The Balkans

315

Of equal significance for the functioning of the power-sharing system is the role of international actors after the establishment of the system itself. In both Kosovo and BiH international actors are formally and informally deeply embedded in the power-sharing system. The Office of the High Representative (OHR) in BiH and the SRSG in Kosovo (and between 2008 and 2012 the International Civilian Representative [ICR]) have formal powers to impose and annul the decisions of local institutions. The problems arising from international officials overriding democratically elected institutions have been extensively discussed, in particular for the case of BiH in terms of undermining democracy (see Venice Commission 2005), democratization (Bieber 2006), and power sharing (Knaus and Martin 2003). The patterns in which the power of imposition has been used vary greatly between BiH and Kosovo. In BiH most external intervention took place in the shape of OHR impositions, which broke a deadlock between the different communities, whereas in Kosovo, the SRSG frequently had to annul decisions, as the large Albanian majority in parliament took decisions beyond their mandate or in disregard of the Serb or other minority communities. The post-independence ICR has been largely passive despite extensive formal powers. In both cases, the external imposition has undermined the compromise-seeking processes of local actors and created incentives for confrontational political posturing as imposition could break deadlock. In addition to exercising formal powers, the international actors in the two countries have been informal arbiters and mediators— depending on the circumstances—between the different communities. In BiH (and to a lesser degree Kosovo) international actors have been integrated into domestic institutions. Of particular significance has been the often decisive vote of the three international members of the Constitutional Court (in addition to two Bosniak, two Croat, and two Serb members) in BiH. International actors have had a more subdued function in Macedonia. Whereas the military aspects of the peace agreement were monitored by NATO, the Ohrid Agreement established no comparable civilian authority. The roles of the Organization for Security and Cooperation in Europe (OSCE) and the EU Special Representative (EUSR) have been limited in the postconflict period. The EUSR and key embassies have only informally been key mediators in conflicts over the implementation of the Ohrid Agreement. The lowest degree of external intervention took place in SCG. The implementation of the State Union was entirely left to the parties, and external mediation occurred only occasionally and on an ad hoc basis. Tellingly, no EUSR for the State Union was appointed, and the EU as the key mediator

316

Florian Bieber

lacked a clear engagement with the parties. As the disintegration of SCG was a clear and legitimate (after three years) outcome, the efforts to render the joint state functional were considerably more restrained than in BiH and elsewhere. In fact the EU only reengaged as a mediator to negotiate the terms of the state dissolution. The strong role of international actors in Kosovo and BiH is a reflection of the provisional nature of the institutional arrangements in both countries. In Kosovo, the 2001 Constitutional Framework and the 2008 constitution are clearly interim arrangements, both in terms of the sovereignty of Kosovo and the interethnic bargain—or lack thereof—between Kosovo Albanians and Serbs. Formally, the Dayton constitution appears to be a permanent settlement, but the strong nature of the international actors, in particular since the OHR’s powers were enhanced at the Peace Implementation Council meeting in Bonn in 1997, suggests a temporary nature to the current settlement. In fact, the Dayton framework has changed significantly de facto and de jure as a result of a series of international interventions that have strengthened the state institutions to the detriment of the entities and that have also resulted in the establishment or broadening of power sharing in the entities themselves. Since 2006 efforts to enshrine and broaden these changes in the institutional structure failed after a constitutional reform package was narrowly defeated in parliament and a subsequent crisis in the power-sharing arrangement resulted in delays to the planned closure of the OHR (Sebastián 2007). The core problem of international intervention in the power-sharing structure in BiH and to a lesser degree in Kosovo has been the ability of international actors to suspend the consensus-based principles of the institutional arrangement. Frequently the intervention has been characterized not by mediation or arbitration but by imposition against the will of at least one party. As a result, political elites of varying communities have argued that their rights are not adequately protected by the institutional framework as international intervention and the power-sharing system become conflated. The practice of consensus building and trust in the system to protect group interests have thus been frequently undermined.

Decision Making and Power Sharing The formal institutions of power sharing in BiH, Kosovo, Macedonia, and SCG share varying degrees of consociationalism. As noted earlier and dis-

The Balkans

317

cussed below, the technicalities of the power-sharing systems vary significantly. Nevertheless, the systems can all be characterized as consociational arrangements rather than other forms of power sharing (e.g., centripetal power sharing). With different mechanisms in place, all four cases have included the key groups in government, provided for tools to prevent outvoting by the majority, established forms of proportionality in the public administration, and provided for some degree of autonomy. It is in the latter that one can identify the greatest degree of difference across the cases and will be discussed in the next section after examining the other most contentious aspect of power sharing, namely the decision-making mechanisms that prevent minority exclusion. BiH has been struggling since the Dayton constitution came into force with extensive and destructive veto mechanisms. Bosnian parliamentary procedure recognizes two forms of veto rights: the entity veto and the veto by constituent people. All laws thus require the support of at least one-third of the MPs from each entity in both chambers of parliament (House of Peoples and House of Representatives) and the support of a majority of all three constituent peoples in the upper chamber. This vital interest, which can be invoked, is not defined in the constitution, but if a majority of another delegation challenges the invocation of the vital interest clause, an ad hoc commission is formed, and if it fails to resolve the matter, the Constitutional Court is responsible for deciding on the matter. As a result, the Constitutional Court has been developing jurisprudence on what constitutes a vital interest and has taken a restrictive line. In the Velimir Jukić case, for example, the court found the law on establishing a public broadcasting system not to be destructive of the vital interests of the Croat community. In effect most laws that failed in parliament were blocked through entity veto which allows a two third majority from each entity to block any decision without further recourse. Furthermore, in BiH, it has often been the threat of the veto and its option, rather than its usage, that have brought decision making to a standstill. Serbia and Montenegro had voting mechanisms similar to the entity veto in Bosnia and Herzegovina. Here, all decisions required majority support from both republics in addition to the overall majority. As the Constitutional Charter required majority support from all 126 deputies (91 from Serbia and 35 from Montenegro), parliamentary sessions were often repeatedly canceled because many MPs from Montenegro did not attend sessions. This blockage was not a direct veto to the decisions (most of which involved

318

Florian Bieber

the ratification of international treaties) but was employed as a tool to render the already weak institutions even less effective. In Macedonia, on the other hand, veto rights were established by the introduction of the so-called double majority, according to which parliamentary decisions in fields of particular relevance to the Albanian and other communities and in areas of broad significance (parts of the constitution, law on municipalities) require the consent of an overall majority of MPs and a majority among the MPs who are elected from minority communities. The voting principle, widely known as the Badinter majority, named after the French constitutional lawyer responsible for the proposed mechanisms in 2001, has been effective in securing Albanian and other minority consent to key legislation. The elections of 2006 and subsequent government formation highlight the fact that this voting system also has had an impact on government formation and the ability of governments to have a majority for all laws in parliament. The conservative VMRO-DPMNE (Internal Macedonian Revolutionary Organization– Democratic Party for Macedonian National Unity) won the 2006 elections with a coalition of smaller parties (including parties representing Roma, Turks, Vlachs, and Bosniaks) and decided to form the governing coalition with the second largest Albanian Party, the Democratic Party of Albanians (DPA), rather than the Democratic Union for Integration (DUI), which had received more votes in the elections. This triggered a parliamentary boycott by DUI, which argued that it had the right the strongest Albanian party to join government (Gaber-Damjanovska and Jovevska 2007). Including Albanian parties has been a practice, not a legal requirement, in Macedonia and neither there nor elsewhere could the inclusion of the largest minority party been seen as a requirement for power sharing more broadly. The impact of the decision, however, was that the government lacked a clear majority in parliament for decisions to be voted by double majority. Of the 36 (of 120) seats in parliament held by minority community representatives, the DUI controlled 17. Laws and other decisions to be voted by double majority thus required consent not only from the Albanian community for the DPA but also from smaller communities. In addition to enhancing the power of smaller communities and having rendered governing more difficult, this constellation also sheds light on the tension in Macedonia between the binational structure of most interethnic tensions and the inclusion of additional communities in the structure.

The Balkans

319

The constitution of Kosovo, passed in April 2008, draws heavily on the arrangements in Macedonia. Not only does it not foresee any meaningful territorial autonomy besides enhanced municipal powers, it also establishes similar voting mechanisms to those of Macedonia. Thus laws affecting minority communities require majority consent from communities that hold reserved seats. This provision is unlikely to be very effective in Kosovo, as a transitional clause in the constitution sets forth that the relevant laws can “initially” be adopted by a simple majority. This, of course, severely undermines the ability of minorities to intervene in the establishment of the legal framework. The different mechanisms to prevent majority outvoting and the other aspects of power sharing impact the performance of the system in terms of elites’ ability to compromise and effectively govern the country. I would argue, however, that the problems of veto mechanisms are not the root of blocked and ineffective governance.

Autonomy and Power Sharing Power sharing by definition needs to work in tandem with other forms of interethnic accommodation. In particular autonomy, which is understood to form an integral part of power sharing, can take many different forms: from cultural and nonterritorial autonomy to a federal or confederal arrangement. The performance of power sharing to a large degree rests on the interrelationship with autonomy. The record of former Yugoslavia suggests that we need to examine the interaction between the degree of territorial decentralization and the thickness of state institutions governed by power sharing. In terms of territorial decentralization, we can identify two regional patterns: both BiH and SCG are/were states with a great degree of territorial decentralization. The entities or member states have clearly defined borders, have broad competences, have powers to legislate and levy taxes and to determine their economic policy, and are governed by their own constitution. While in BiH the entities do not enjoy the right to self-determination, in SCG the Constitutional Framework and both member states acknowledged the right of the constituent units to secede within three years of the country’s formation. The other model of substate units is the considerably lower degree of decentralization employed in Macedonia and Kosovo. Neither has an in-

320

Florian Bieber

termediate level of government between the central state and municipalities. This has meant that the state remains the sole and unchallenged (at least de jure) legislator. On the other hand it has also resulted in a greater empowerment of municipalities. In fact, debates and controversies in Kosovo and Macedonia over decentralization have been highly contentious; in Macedonia they even resulted in a 2004 (failed) referendum initiated by nationalist Macedonian groups against the decentralization plans. Municipal decentralization has been a meaningful development in the evolution of power-sharing systems. Following a regional pattern of centralization, this form of autonomy contains three core aspects: (a) the effective devolution of competences (i.e. in the sphere of education or policing) and financial resources to municipalities; (b) the redrawing of boundaries to create municipalities that accommodate the needs of minority communities; and (c) establishing mechanisms for minority inclusion at the municipal level. The advantage of municipal decentralization has been twofold. First, it can alleviate majority fears of secession and “federalization” of the country, which autonomy otherwise often evokes. Second, it can more effectively follow the population distribution than are larger territorial autonomies. Municipal decentralization is not without its problems, however. Lacking the power to legislate and being inherently less powerful than regional autonomies, decentralized municipalities clearly remain subordinate to the central state. Some protection is secured in both Macedonia and Kosovo by preventing changes to the borders, competences, and funding of municipalities without consent of the minority community. As they do not effectively allow for aggregation of minority interests but rather fragment the political representation of the minority community into multiple constituencies, it is doubtful whether municipal decentralization can be considered a full-fledged form of autonomy. While alleviating majority fears of secession, it might also increase conflict and contestation at the center, as groups do not necessarily enjoy autonomous areas of decision making. In Macedonia this has indeed been the case, as legislative powers remain with the central state institutions. The high degree of de facto partition of Kosovo and territorial fragmentation has meant that since 2001, especially in 2008, the constitutional setup and reality have drifted apart. The second aspect under discussion here is the thickness of the state or other levels of governance subject to a power-sharing arrangement. This is crucial for two reasons. If the scope and the strength of the state are limited, the blockages in the power-sharing system, such as vetoes and other means

The Balkans

321

of delaying decisions, will have a lesser impact on the overall structure of governance, as lower levels of government can continue functioning undisturbed. On the other hand, if the state has broad decision-making competences, power-sharing mechanisms can paralyze much of the political system. Thus in Macedonia and Kosovo, where only municipalities exist below the state-level government, veto mechanisms are restricted to particular areas of decision making. The problem that “thin states” or states I call “minimalist” are confronted with is that because of their limited scope, groups fail to invest in them and might be more likely to further undermine them. BiH and SCG are examples of minimalist states, which are limited in both scope and strength of state functions. According to Francis Fukuyama, the scope of the state describes the “ambition” of the state in terms of the fields in which it engages, whereas the state strength refers to the ability of states to enforce its policies. Here both formal and informal constraints need to be considered (Fukuyama 2004). Minimalist states thus firmly fall into the category of weak states in terms of both strength and scope. Unlike failed states, however, minimalist states hold minimal functions and do not have the constitutional ambition to exercise broader functions or possess greater enforcement mechanisms. The constitutional frameworks of BiH and SCG thus institutionalized the stateness problem they faced in light of the challenges from secessionist units. As the existence of the state as such was challenged prior to the agreements, a key feature is not only the governance of the state (e.g., power sharing) or the territorial organization (e.g., confederation or federation) but the scope and strength of the state. In terms of the scope of the state, both BiH and SCG have held competences that have been generally limited to foreign policy and basic human rights policies. SCG has held more competences in the field of defense (at least prior to the creation of the BiH army and ministry of defense in 2004), while BiH has been able to establish a single currency. Neither state has had a unified economic space. In terms of the strength of the state, both have also been limited. The ability to enforce their competences has been severely constrained. With no tax-raising abilities (in the case of BiH until the introduction of the VAT in 2006), limited security structures as a result of control by the entities, and a judicial system with weak recourse mechanisms to the nonenforcement of decisions, the central state institutions have often been unable to enforce decisions. Furthermore, central state institutions in both states had weak decision-making capacity, as the substate units possess the ability to paralyze

322

Florian Bieber

state-level decision making. The parliament of the State Union, for example, met only irregularly and passed only twenty acts during a two-year period (2003–5) (European Commission 2005, 6). A third dimension that might be added to the features of minimalist states under consideration here is their contested nature. Consequently, they lacked symbolic cohesion and identification. The very fact that BiH has no lyrics to its national anthem, as well as the fact that during the short life of the state union Montenegro and Serbia could not agree on a national anthem (a medley of the two states’ anthems was under discussion), is evidence of the weak commitment to the state. Just as with federal arrangements, minimalist states can be centrifugal or centripetal. SCG is an example of a centrifugal minimalist state, where the limited competences gave rise to a level of compliance with the state institutions’ decisions that was lower than formally foreseen. In the first postwar years BiH displayed similar features, where the institutional reality of the state was weaker than the formal constitutional shape of the state. Since the late 1990s, the dynamics of institutional development in BiH have become increasingly centripetal. It is important to note, however, that this process has been externally driven, and the crisis in the Bosnian state-building project since the failure of the constitutional reform in 2006 suggests that the centripetal process is not based on the nature of the institutional setup but on the consequence of externally imposed state building. Kosovo and Macedonia have functioned very differently. With no intermediate layers of governance, as discussed, the competences of the state are significant. This does not mean, however, that this formal power of the state translates into reality. In Kosovo in particular, it might be argued, the formal strength and scope of the state remain profoundly out of touch with the reality of quasi-inexistent state institutions in most Serb-populated areas.

Conclusion: Consensus Democracy Without Consensus Power sharing in BiH, Kosovo, and SCG has been a response to competing self-determination claims. The conflicts were not over dominance within the state but whether the state should exist at all. In BiH, the Dayton Peace Accords secured a weak commitment of the Bosnian Serb and Croat elite to the joint state. In Kosovo, the Serb community participated only briefly in the power-sharing institutions and only when they did not challenge the Serbian

The Balkans

323

sovereignty over Kosovo. SCG only came about after the EU exerted intense pressure on Montenegro not to hold a referendum on independence in 2002. This situation differs from that of Macedonia, where the Albanian National Liberation Army (NLA), after a brief campaign for self-determination, struggled to achieve greater rights for Albanians in Macedonia rather seeking secession altogether. The weak or nonexistent commitment in the other cases of one or several actors to the political unit has severely undermined the functioning of power sharing. Kosovo has not seen any agreement between Serbia and Kosovo Albanians or between Kosovo and the Kosovo Serbs. All peace plans have been rejected by one side. As a result, all institutions and power-sharing arrangements have been a direct imposition rather than internally driven compromises or settlements. The first institutions of power sharing were set up in the Constitutional Framework for Kosovo, which was imposed in 2001 by the UN SRSG Hans Haekkerup. Kosovo Serbs bought into the new institutions en masse in 2001 when the Serb coalition Povratak (“Return”) gained a surprising 11.34 percent of the vote and the coalition secured 22 seats in parliament (12 gained through PR and 10 as additional reserved seats). As the institutions moved toward consolidating Kosovo statehood and as Serbia’s stance on Kosovo hardened, Serb participation in the institution dwindled. The disengagement of Kosovo Serbs from the institution was partially caused by the lack of protection from being outvoted, despite some weak veto mechanisms, but, more important, it was also due to the lack of consensus between Albanians and Serbs in Kosovo over the status of Kosovo. Institutions of consensus building thus broke down in the absence of a minimal agreement on the larger status of the polity. BiH might rest on a somewhat stronger degree of commitment by the three communities because the DPA was signed by all parties (even if not by Bosnian Serbs and Croats but rather by their handlers in Croatia and Serbia) and committed the parties to a joint state. This commitment has been limited, however, and all communities perceive a degree of injustice in the postconflict development of the state: Croats feel disadvantaged for being a minority at the state and entity levels; Bosniaks consider the recognition of the Republika Srpska (RS) as an entity of BiH unjust because it was created through ethnic cleansing; and many Bosnian Serbs feel that the steady erosion of their entity runs against the principle and letter of the DPA. The structure of postwar BiH has thus been a ceasefire but not a broadly accepted compromise among all three communities.

324

Florian Bieber

None of the power-sharing arrangements of political units has been legitimized by referendum. However, the legitimacy of the mediators and the confirmation of the agreement through elections have helped in some cases, such as Macedonia, while election results have undermined the agreement in BiH, where one of the dominant parties has explicitly or implicitly rejected the agreement almost constantly. Consensus politics in a political system that lacks a basic consensus has brought about a number of challenges to power sharing, which have effectively undermined the ability of the system to function autonomously. First, “spoilers” are encouraged as they can challenge incumbents for excessively buying into the system. Compromising with other communities can be interpreted not only as giving up on the substance of the matter but as supporting the respective political system. In BiH, there has been a constant outbidding within each community as the dominant party is pressured by international actors to compromise with other communities whereas the challenger can accuse the party of jeopardizing group interests by empowering the state or another entity. In Kosovo, when the moderate Serb List for Kosovo in 2004 decided to participate in elections, the more nationalist Serb parties boycotted the elections, and as turnout among Serbs was limited to a few thousand, the moderate group did not take up its seats in the Kosovo parliament for fear of being termed traitors. Second, even if intragroup spoilers are co-opted or if no outbidding takes place, decision making is generally perceived not on the basis of interethnic competition but on whether or how it legitimizes the contested political unit. In 2007, for example, Bosniak ministers repeatedly boycotted government sessions to prevent a decision that would allow the sale of a refinery in the RS. The resulting deadlock in the institution was not based on vital interests but because the sale and its (potential) commercial success would help legitimize the RS. Third, the existence of a reserve power base through a high degree of territorial autonomy allows parties to disengage from power-sharing institutions. The autonomy of the entities in BiH and the states in SCG meant that the deadlock and stalled decision making do not necessarily hurt the parties in question but might in fact be the purpose of blocking decisions. In conclusion, the acceptance of the political settlement and the mechanism through which the settlement has been legitimized are crucial for its success. If the cost of disengaging from power-sharing institutions is low owing to alternative power bases and limited international engagement, and

The Balkans

325

the joint institutional project lacks legitimacy, the risk to power sharing is high. Among all power-sharing systems discussed here, Macedonia has been the most successful in building a legitimate institutional arrangement that has avoided breakdown or the opting out by an entire community. In BiH and Kosovo the power-sharing arrangements are largely viewed as an interim solution that will yield to some more permanent settlement. The visions of a post-power-sharing system among the parties are diametrically opposed and often reduce the incentives to render the existing institutional setup effective. Notes I would like to thank the International Policy Fellowship, which has supported research for this chapter. 1. See my discussion on this matter in Bieber 2003. 2. In fact, the legal status of the Ohrid Framework Agreement remains controversial in Macedonia. Some actors, in par ticu lar from the majority, argue that it has no domestic legal standing and all its significance derives from the legal implementation in domestic law, whereas mostly Albanian parties argue for granting the agreement a legal binding status. 3. See, for example, Letter of Republika Srpska prime minister Milorad Dodik to the European Parliament, 20 November 2007, in “Inside the Bosnian Crisis” 2007. 4. Article IV, Constitution of BiH. 5. Constitutional Court, U-10/05, July 22, 2005. 6. Article 23, Constitutional Charter of Serbia and Montenegro. 7. The parliamentary rules of procedure consider laws in the following fields as requiring a double majority: “culture, use of language, education, personal documents and use of symbols.” Article 164, Rule of Procedure, Parliament of Macedonia. 8. In a controversial move, one MP from the VMRO-DPMNE changed the declaration of her community affi liation from Macedonian to Vlach after the first declaration in an effort to reduce the advantage of DUI. Izveštaj 2006, Sobranie RM, p. 22. 9. Article 81, Constitution of Macedonia. The laws covered by the provision are those pertaining to municipalities, communities, the use of language, local elections, cultural heritage, religious freedom, education, and symbols. 10. Article 149, Constitution of Kosovo. References Bieber, Florian. 2003. “Institutionalizing Ethnicity in Former Yugoslavia: Domestic vs. Internationally Driven Processes of Institutional (Re-)Design.” Global Review of Ethnopolitics 2, 2: 3–16. ———. 2005. “Power Sharing After Yugoslavia: Functionality and Dysfunctionality of Power-Sharing Institutions in Post-war Bosnia, Macedonia and Kosovo.” In Sid

326

Florian Bieber

Noel, ed., From Power Sharing to Democracy: Post-conflict Institutions in Ethnically Divided Societies. Montreal and Kingston: McGill-Queen’s University Press. ———. 2006. “After Dayton, Dayton? The Evolution of an Unpopu lar Peace.” Ethnopolitics 5, 1: 15–31. European Commission. 2005. “Staff Working Document: Report on the Preparedness of Serbia and Montenegro to Negotiate a Stabilisation and Association Agreement with the European Union.” SEC 478 final. http://edz.bib.uni-mannheim.de/www -edz/pdf/sek /2005/sek-2005-0478-en.pdf. Fukuyama, Francis. 2004. “The Imperative of State-Building.” Journal of Democracy 15, 2: 21–22. Gaber-Damjanovska, Natasha, and Aneta Jovevska. 2007. “Current Events and Political Parties: Development in the Republic of Macedonia.” Institute for Sociological, Political and Juridical Research, Skopje, No. 16 (June). http://library.fes.de/pdf -fi les/bueros/skopje/05321/barometer16 -2007.pdf. “Inside the Bosnian Crisis.” 2007. Journal of Intervention and State-Building 1 (November): 68–70. Knaus, Gerhard, and Felix Martin. 2003. “Travails of the European Raj.” Journal of Democracy 14, 3: 60–74. Rothchild, Donald, and Philip G. Roeder. 2005. “Dilemmas of State-Building in Divided Societies.” In Philip G. Roeder and Donald Rothchild, eds., Sustainable Peace: Power and Democracy After Civil Wars. Ithaca, NY: Cornell University Press. Sebastián, Sofia. 2007. “Leaving Dayton Behind: Constitutional Reform in Bosnia and Herzegovina.” FRIDE Working Paper No. 46. Venice Commission (Eu ropean Commission for Democracy Th rough Law). 2005. “Opinion on the Constitutional Situation in Bosnia and Herzegovina and the Powers of the High Representative.” CDL-AD 004, 11.3.2005. http://venice.coe.int/.

CHAPTER 14

Governing Polarized Cities Scott A. Bollens

This chapter provides a comparative analysis of different institutional approaches to dealing with antagonistic group identity claims on the city. I discuss Brussels, Johannesburg, Belfast, Sarajevo, Jerusalem, Baghdad, and Kirkuk. These cities are broken down into three categories: (1) cities that have utilized power sharing and forms of transitional democratization effectively enough that stability of the local and national state has occurred, (2) cities that have made some progress but are vulnerable to regression because local political arrangements are not sufficiently stabilizing, and (3) cities where power sharing is itself contested and a potential contributor to further instability.

Divided Nations, Divided Cities In a deeply divided society, the configuration of local political power in its major cities can create either solid or faulty foundations for the future stabilization and sustainability of the national state. Cities can be critical spatial, economic, and psychological contributors to national ethnic stability and reconciliation. I have argued elsewhere that urban power sharing and democratic compromises should be considered a necessary supplement to national political agreements (Bollens 2007). This is so because many immediate and existential foundations of intergroup conflict frequently lie in daily urban life and across local ethnic divides and, importantly, that it is at this microlevel that antagonisms can be most directly influenced by government interventions aimed at their amelioration. More than at larger geographic scales

328

Scott A. Bollens

where segregation of ethnic communities is more likely due to historic settlement patterns, the economic pull of the city means that urban areas will frequently contain contestable fault lines between ethnic, religious, and/or linguistic groups. Cities are significant depositories of material resources and culture vulnerable to penetration or implosion by nationalistic ethnic conflict and violence. They are focal points of urban and regional economies dependent on multiethnic contacts, social and cultural centers and platforms for political expression, and potential centers of grievance and mobilization. They are suppliers of important religious and cultural symbols, zones of intergroup proximity and intimacy, and arenas where the size and concentration of a subordinate population can present a direct threat to the state. Cities in unsettled societies are susceptible to intense intercommunal conflict and violence reflecting ethnic or nationalist fractures. In these cities and societies, ethnic identity and nationalism combine to create pressures for group rights, autonomy, or even territorial separation. Political control of multinational cities can become contested as nationalists push to create a political system that expresses and protects their distinctive group characteristics. Whereas in most cities there is a belief maintained by all groups that the existing system of governance is properly configured and capable of producing fair outcomes, assuming adequate political participation and representation of minority interests, governance amid severe and unresolved multicultural differences can be viewed by at least one identifiable group in the city as artificial, imposed, or illegitimate. Polarized cities are where two or more ethnically conscious groups—divided by religion, language, and/or culture and perceived history— coexist in a situation where neither group is willing to concede supremacy to the other (Hepburn 2004). Characterized by ethnic/nationalist saturation of what are typically mundane urban management issues, the unsettled nature of such cities “reveals the contested and limited nature of the national settlement in its schoolrooms and town halls” (Keith 2005, 3). The urban areas discussed in this chapter are as follows: Sustainable Brussels (Belgium) Johannesburg (South Africa)* Fragile Belfast (Northern Ireland) Sarajevo (Bosnia-Herzegovina)*

Polarized Cities

329

Combustible Jerusalem (Israel/Palestine) Baghdad and Kirkuk (Iraq)* Cities in the first set have utilized power sharing and forms of transitional democratization effectively enough that some amelioration of political conflict and stability of the local and national state have occurred. The second set of cities, in contrast, is more vulnerable to relapse because local political arrangements are not sufficiently stabilizing. Cities in the third set are in highly inflammatory settings and power sharing is itself contested and a potential contributor to further instability. The cities identified by an asterisk illuminate the particular challenges of local democratic management during major transitions associated with regime change or postwar reconciliation and reconstruction.

The City and Conflict Management The fundamental problem in polarized cities, as in conflictual national states, is that it is difficult to reconcile majoritarian concepts of democracy with the reality of large, multiethnic cities where group identity is a primary driver. Majoritarian democracy may breed frustration and alienation and intensify conflict and the potential for violence. Beyond majoritarian democracy exist alternatives that either promote sharing urban political power across identity groups or moderate the potency of group allegiance in local politics. These nonmajoritarian forms of local democracy have two broad aims that can clash: (1) accommodation—treat the city as a mosaic of groups living essentially apart and thus provide local ethnic groups autonomy in their own affairs and in representation at municipal level; and (2) assimilation— promote the city as a melting pot and create integrated political coalitions so that democracy and political activity are not based along identity lines (International IDEA 2001). In the accommodative strategy, urban governance would protect group autonomy and minority rights. Mechanisms used would be decentralization of city authority to neighborhoods, minority vetoes on issues of particular importance to group, proportionality requirements in areas such as budgeting and civil ser vice appointments, and the use of powersharing grand coalitions to govern the city. In the assimilative strategy, urban governance would seek integration, provide incentives for multiethnic

330

Scott A. Bollens

cooperation, use electoral systems that encourage preelection pacts across group lines, create a nonethnic federalism that diff uses points of power, and enact policies that promote political allegiances across group lines. National-level models of conflict management aimed at diff using or moderating intergroup conflict have important implications for urban governance. These national models have tended to treat smaller-scale forms of conflict management prevalent in urban areas—such as those dealing with discrimination and segregation, demographic policies, and community relations—as subordinate and reflecting the playing out of broader imbalances of power (O’Leary and McGarry 1995). This de-emphasis on city-based dynamics notwithstanding, these models of conflict management inform about urban possibilities. O’Leary and McGarry (1995) outline four main methods for managing ethnic differences and three methods that seek to eliminate ethnic differences. With hegemonic control, one group dominates the state apparatus and channels decision-making outcomes toward the favored ethnic group (Lustick 1979; Smooha 1980). In urban settings of partisanship, the proximity and interdependence of urban ethnic populations may necessitate greater cooperation or co-optation between political leaders than would be found at national levels. Third-party intervention relies on there being an arbiter whose claim of neutrality must be broadly accepted by contending ethnic groups. At the urban scale, this perceived joint neutrality can be difficult because historic imbalances and inequalities are highlighted by a relative deprivation effect induced by physical proximity. Cantonization and federalization involve, in the first case, devolution of some government authority to homogeneous ethnonational territories and, in the second, separate domains of formal authority between levels of government. Urban applications of these concepts include, in the first case, the creation of community or neighborhood-based groups that would advise or decide on local issues. In the second case, there would be the creation of a metropolitan government and subordinate municipal governments. The last model of ethnic management is consociation or power sharing. At the national level, this has been the most closely scrutinized option for deeply divided societies (see Lijphart 1968, 1977; Nordlinger 1972; Horowitz 1985). Primary characteristics of power-sharing approaches applicable to municipal settings are inclusive government (municipal legislative councils, especially leadership positions, that mirror the ethnic configuration of society), group self-government (neighborhood self-government concerning particular issues), and proportionality

Polarized Cities

331

in resource allocation (especially regarding civil ser vice position appointments and budgetary decisions) (Roeder and Rothchild 2005). Urban settings can also be key terrains upon which strategies of ethnic difference elimination can be applied. Urban ethnic cleansing of entire cities or of specific ethnic neighborhoods illustrates the application at the urban and regional scale of forced population transfers. A second means of eliminating ethnic differences, political partitioning of urban space (such as Nicosia, cold war Berlin, and Jerusalem from 1948 to 1967), can be an important feature of national-level agreements regarding territorial separation; yet at the urban level, it tends to introduce a host of practical problems not found at a broader, national scale. Integration or assimilation strategies in many ways have greater salience at the smaller-scale urban level than they do at more dispersed national scales. The three main political alternatives that acknowledge group identity in the urban arena—political (or physical) separation, two-tier federated governance, and consociational city government—run the gamut from least to most interethnic cooperation. In a politically partitioned city, sovereignty is divided and ethnic groups are isolated from one another (sometimes dramatically so if physical walls accompany political separation). Political segmentation of an urban area presents numerous logistical problems, especially if competing ethnic groups, while being segregated from each other, are not concentrated in particular sectors or directions. Either drastic relocation must occur to ethnically sort the urban region or local boundaries must be drawn in disfigured, noncontiguous ways that dampen ethnic community cohesiveness. In the second alternative—creation of a two-tier system of local government—two ethnic-specific local authorities share the sovereignty of the urban area. There is unity or cooperation at the higher level of government (metropolitan or city) but functional and political division at the lower level (city or borough). Again, the creation of ethnic local governments (or boroughs) becomes logistically problematic where urban ethnic geographies are intertwined. The pressure associated with ethnically dividing urban governance is alleviated somewhat in this alternative because there is an umbrella government for the whole area (either city or metropolitan level), and minority rights and guarantees can be built into these integrative institutions. In the third alternative, a consociational or powersharing city government, a local conflict-accommodative government is established that utilizes power sharing, ethnic proportionality within the public sector, community autonomy, and minority vetoes. Such a consociational

332

Scott A. Bollens

arrangement would likely be part of the higher level of the second alternative (two-tier governance).

Politically Sustainable Governance Th is fi rst set of case-study cities is most politically sustainable because power-sharing structures and forms of transitional democratization have effectively stabilized the local state to such a degree that most group-based conflicts are channeled into political processes where there is opportunity for some compromise.

Brussels (Belgium) COMPLE X INSTITUTIONAL ACCOMMODATION

Brussels is at the fault line between northern Dutch-speaking and southern Francophone areas, a frontier contested by the strong Francophone majority in the city and historic claims of its Flemish (Dutch-speaking) past (see Figure 14.1). In dealing with the strongly bi-national nature of the urban region (and the country as a whole), there has been for more than forty years a series of reforms aimed at increasing representation and autonomy along linguistic and ethnic lines. This is an example of the group building-block approach, seeking to accommodate in the Brussels urban region both the 85 percent Francophone majority and the Dutch-speaking minority. Belgium as a country has a Dutch-speaking majority, which has encouraged a type of power-sharing “trade-off ” where some parity is provided for minority Dutch speakers in the Brussels region while parity is provided for minority French speakers country-wide. An officially bilingual Brussels Capital Region (BCR) was created to provide institutional space between the monolingual Dutch-speaking Flanders region to the north and Francophone Wallonia region to the south. The BCR has powers related to town planning, environment, housing, employment, and economic policy, and other territorial issues. The directly elected regional parliament for the BCR is chosen from candidates put forth by each of two main linguistic communities and Parliament decisions require a ma-

Polarized Cities

333

Antwerp East Flanders

West Flanders

Limburg

Brussels

Lan

guage B order

Hainaut

Brabant

BEL G IU M

Liège

Namur

Luxemburg

N 0

25

50

75

100 Kilometers

Figure 14.1. Bilingual Brussels. Adapted from Donaldson 1983, 32.

jority in each language group. The political rights of the Dutch linguistic minority group in Brussels are constitutionally protected by affording them equal power sharing in the executive branch of the city-region government. The regional government preserves two of its minister positions and a secretary of state position for a Dutch speaker. Additional institutions in the BCR include a bi-communitarian public authority, the Common Community Commission, responsible for implementing cultural policies of common interest, and two linguistic community-specific public authorities—the Flemish Community Commission and the French Community Commission—that implement policies of the respective communities in the BCR, including

334

Scott A. Bollens

cultural issues, health and social assistance, education, and the use of language in administrative and workplace relations. In contrast to the three “regions” in Belgium that are territorial, “communities” are nonterritorial and exercise their legislative authority over cultural, educational, and health matters within linguistically determined geographical boundaries, including within the Brussels urban region. Compromises through the years between Flemish and French politicians seeking to increase and maintain their power within the Brussels urban region have created this complex layering of local governance (Terhorst and van de Ven 1997). Some have criticized this local governance system as excessively disjointed and disarticulated, describing the urban area as existing within a “provincial and parochial institutional straightjacket” (Swyngedouw and Moyersoen 2006, 172). Nevertheless, the messy institutional structure means that tensions that inevitably occur often become dispersed between these various forms and scales of governance, thus making improbable the establishment of stable urban hegemonic coalitions that might further inflame ethnic passions (Hooghe 1995). Some consolidation of local public services has occurred; in 2001, police services formerly based in each of nineteen communes/municipalities were combined into six police zones in the Brussels Capital Region. Another curious aspect of Brussels’s governance is that while the urban region experiences constrained capacities due to complex political/linguistic bordering and administrative fragmentation, the Brussels metropolis has been exceptionally successful in international city competition over economic and financial assets (Kesteloot and Saey 2002). Since Brussels is officially bilingual and the two other regions are monolingual, the boundaries of the Brussels district become “language borders.” Dutch speakers worry that if Brussels region expands territorially that the “oil stain” of bilingualism will expand into Flanders. Francophones, meanwhile, criticize “iron collar” constraints on the regional expansion of Brussels as unfairly stopping the bilingual region from spreading (Hepburn 2004). There are six peripheral boroughs outside Brussels, and spatially in the Flanders region, where there is a significant and growing Francophone minority. In these cases, a compromise has been worked out whereby there are permanent guarantees for “language facilities” for French, but these areas are to remain part of Flanders.

Polarized Cities

335

Johannesburg (South Africa) TR ANSITIONAL POWER SHARING, BOUNDARY DR AWING, AND ME TROPOLITAN RESTRUC TURING

Johannesburg and South Africa provide a positive lesson of power sharing used as an effective transitional device on the way to eventual majoritarian democracy. The use of multiple bargaining arenas (state, regional, and local) during the transition from apartheid facilitated the development of a new constitutional order (Sisk and Stefes 2005). At the local level, explicit use of boundary drawing and metropolitan government as social justice mechanisms are also of importance. On-again, off-again multiparty national negotiations from 1991 to 1993 reached agreement on a transitional constitution and executive council and on the procedures for the country’s first democratic elections—for national and provincial legislatures—to be held in 1994. National negotiations were successful in establishing the makeup of a multiparty power-sharing executive cabinet to be configured based on election results and the process by which the final constitution would be created by the national legislature. With the election April 26–29, 1994, a five-year “government of national unity” headed by Nelson Mandela’s African National Congress (ANC) formally replaced the old apartheid regime. With the end of apartheid, it became evident that local governance had to be transformed to overcome the virtual coincidence of race, residential area, and local government in South Africa. The geographic distance between races created by apartheid urban policy worked in parallel with administrative separation and subordination (Beavon 1992). Apartheid categorized cities and towns into group areas for exclusive occupation by single racial groups. In Johannesburg, mass displacement of black populations took place from western ghettos to black townships (most notably to Soweto, constituting anywhere from 30 to over 50 percent of the total population of the Johannesburg urban region). These townships were remote and disconnected from the now white city (see Figure 14.2). Their political and financial detachment from white areas undermined their fiscal base and made self-government fiscally unviable (Hart 1995). While white local authorities of Johannesburg city, Sandton, Randburg, and Roodepoort contained substantial commercial and industrial tax bases that enabled good municipal services at moderate tax rates, black local authorities and other nonwhite areas were fiscally depleted owing to

336

Scott A. Bollens

Sandton

Marlboro Gardens Alexandra

Randburg Davidsonville Roodepoort

J OHA NNE SBU R G Newclare

Edenvale

Coronationville Riverlea

City Centre

Bedfordview

MayfairFordsburg

Germiston

S O WE T O Protea

Alberton

Eldorado Park

Lenasla Katlehong

Vosloorus

Asian

N

Black Coloured White

0

5

10 Kilometers

Figure 14.2. Apartheid Johannesburg. Adapted from Parnell and Perry 1991, 135.

restrictions on nonresidential uses and the illegality of homeownership. Meanwhile, open-field informal settlements outside of black townships faced marginalization and even exclusion from the local governance system. Pass laws regulated movement of black Africans in urban areas. Outside designated “homelands,” black South Africans had to carry passbooks at all times that showed that they were authorized to live or move in “white” South Africa. Transition-period Johannesburg was characterized by a local consociational form of power sharing between “statutory” officials of the old regime, black political leaders, and nongovernmental organizations. Johannesburg

Polarized Cities

337

(and other South African cities) also emphasized the metropolitan scale as a focal point for local government transition negotiations and used metropolitanism as a means to integrate and transcend old local authority boundaries that had separated races. The impetus for the early consideration of local government reform in Johannesburg was crises brought on by the boycotting of the payment of rent and ser vice charges in Greater Soweto in the late 1980s. The boycott called attention to the illegitimate form of local governance in Soweto, the racial compartmentalization of local government financing, and the resulting inadequate levels of urban services. Negotiations to end the boycott resulted in the Greater Soweto Accord and an agreement to establish in 1991 a Central Witwatersrand Metropolitan Chamber (CWMC) that would be a negotiating body to formulate nonracial and democratic structures for local and metropolitan government. This Metropolitan Negotiating Forum contained 50 percent nonstatutory and 50 percent statutory representation. This forum then appointed members to an interim council—the Transitional Metropolitan Council (TMC)—which would manage urban affairs until local elections took place based on newly demarcated, more equitable local and ward boundaries. Metropolitan negotiators debated how to politically redraw municipal borders to integrate what was torn apart under apartheid. There were basic disagreements about the boundaries and roles of new local governments (called Metropolitan Sub-Structures, or MSSs) in post-apartheid Johannesburg. An initial ANC proposal for local government restructuring sought the “stitching of townships to cities” so that they would no longer be marginalized and the creation of a strong overarching metropolitan authority. Statutory representatives, meanwhile, proposed eight MSSs that would de-link some black areas from white ones. After significant contention, a Special Electoral Court in 1995 approved a 4-MSS model, stating that it most effectively eliminated the racial political geography of old group areas. In each of the four MSSs— Johannesburg, Sandton, Randburg, and Roodepoort—there was an existing and functioning administration. At the same time, each of these administrations would have responsibility for managing black townships within its MSS borders. The enacted 4-MSS configuration sought to balance the richer north and the poorer south, share the responsibility for managing Greater Soweto across three different MSSs, and evenly distribute votes, commercial and industrial activities, and tax bases across the four MSSs. A strong metropolitan government was created with the ability to redirect budgetary resources across MSS boundaries. This was added assurance to equity advocates who

338

Scott A. Bollens

claimed that a strong metropolitan government role in regional budgets was needed to reverse the extreme inequalities of apartheid. This 4-MSS configuration provided the framework for local and metropolitan elections in 1995. Electoral rules specified that 60 percent of the seats in each MSS were to be ward based; the other 40 percent would be based on proportionate representation (PR) rules. For the metropolitan council, 60 percent of the councillors were to be appointed by the MSSs while 40 percent were directly elected by PR. In a compromise with white authorities, the national local government law specified that at least half of electoral wards had to come from preexisting white (including Indian and Coloured) authority areas. This basically locked in a 30 percent (half of 60 percent) representation in each jurisdiction from nonblack areas. This was agreed upon by the multiparty negotiators to ensure white minority representation during the transition period (Ewing 1995). Further governmental reform in 2000 created a unified City of Johannesburg Metropolitan Municipality. Apartheid-era cities such as Sandton and Roodepoort no longer have separate municipal governments, having been subsumed within the metropolitan municipality of Johannesburg. At the same time, administration of some city ser vices such as health care, housing, and social development was decentralized to seven precincts. Criminal violence has been a horrible problem in post-apartheid South Africa. Faced with the problems posed by crime and the resulting loss of confidence among residents, Johannesburg city has reformed its police services. Whereas Johannesburg’s policing previously was the responsibility of sixteen different law enforcement agencies, five local government security agencies, and various traffic law enforcement agencies, the Johannesburg Metro Police Ser vice was created to join together these fragmented agencies into one entity. The Metro Police augments the work of the South African Police Ser vice (SAPS) in the municipal area of Johannesburg. Transformation of the racial composition of the Police Ser vice is to occur pursuant to the Employment Equity Act of 1998, which requires affirmative action in public and private sector employment so previously disadvantaged black Africans “are equitably represented in all occupational categories and levels in the workforce of a designated employer” (section 15.1). By 2004, about 68 percent of the SAPS police force in Johannesburg’s province (Gauteng) was black Africans, nearing proportionality with the black percentage of the overall provincial population (about 74 percent).

Polarized Cities

339

Politically Fragile Governance This set of cities is more vulnerable to relapse because local political arrangements have not sufficiently developed to a point where they can contribute to stabilization of the local and national state.

Belfast (Northern Ireland) LEGACIES OF THIRD- PART Y INTERVENTION, IMPOTENT LOCAL GOVERNMENT

A historic alteration of Northern Ireland’s governing institutions and constitutional status was specified in the April 1998 Agreement Reached in the Multi-Party Negotiations (i.e., the Belfast Agreement). This agreement, approved by over 70 percent of Northern Ireland voters in May 1998, transfers day-to-day rule of the province from Britain to a new directly elected Northern Ireland Assembly, in which Protestants (unionists/loyalists) and Catholics (nationalists/republicans) have shared power in that decisions require concurrent majorities within both Protestant and Catholic camps. After several attempts, this reconstituted Northern Ireland Assembly began to function in early 2007. The Belfast Agreement states that Northern Ireland is to remain within the United Kingdom as long as a majority in the province wants to remain there. In response to nationalist desires, the new assembly and the Irish Parliament are to form a North-South Council to coordinate and encourage cross-border cooperation. To reassure unionists, a Council of the Isles is to link the governments in Northern Ireland and Ireland with the British government and with new legislative assemblies being set up in Scotland and Wales. Prior to the Belfast Agreement, from 1972 to 1998, legislative power for Northern Ireland was held directly by the British House of Commons. The Northern Ireland Parliament, the governing body for the province up until 1972 and linked to the formulation of discriminatory and unjust laws, was held to be incapable of fair and capable governance and the British enacted “direct rule” in the midst of sectarian conflict in 1972. Executive power for Northern Ireland under direct rule was possessed by the secretary of state for Northern Ireland, who was chosen from the party ruling at Westminster. The thinking behind direct rule was that the removal of policy formulation

340

Scott A. Bollens

% 0-9 10 - 29

Belfast Lough

30 - 49 50 - 69 70 - 89 90 - 100

N 0

1

2

3 Kilometers

Figure 14.3. Belfast’s ethnic segregation (percentage Catholic by neighborhood ward). Adapted from Boal 1994.

and implementation from the bitter sectarian confl ict would make it more efficient and effective (Loughlin 1992). Ministers in charge of Northern Ireland governance took their political cues from Westminster and tended toward an inherently conservative, non-risk-taking approach to the province’s controversial issues (Sweeney 1995).

Polarized Cities

341

Belfast city is hypersegregated and of strict sectarian territoriality, with antagonistic groups both proximate and separate (see Figure 14.3). During the years of sectarian conflict, the city became barricaded and divided. The city to this day is one of physical barriers that symbolically separate proximate Catholic and Protestant residential neighborhoods. The extent and severity of intercommunity hostilities since the late 1960s necessitated the building of physical partitions, so-called peace walls, between neighborhoods at sixteen locations. Most of these are located in west and north Belfast, where population shifts following the outbreak of violence in 1969 were greatest. The physical dividers are built of varied materials—ranging from corrugated iron fences and steel palisade structures, to permanent steel or brick walls, to more aesthetically pleasing “environmental” barriers of landscaped railings and multicolored walls, to “buffer” zones of vacant space or alternative nonresidential development. Police stations are often located on or near these peace wall interfaces. “Direct rule” substantially eroded the authority of local governance in Belfast and elsewhere in Northern Ireland. It was at the local level of government that sectarian bias was most evident, especially in the fields of public employment, service delivery, and housing (Loughlin 1992). The locally elected fift y-one-member Belfast city council has thus had severely constrained policymaking power in planning, urban ser vice delivery, and housing; it has been predominantly an advisory body (Hadfield 1992). Instead, most power in these policy areas was located in appointed boards—such as the Northern Ireland Housing Executive—or in central executive agencies—such as the Department of the Environment for Northern Ireland—which are responsible to British ministers rather than to local politicians. Th is centralized policymaking structure was viewed as capable of depoliticizing local planning issues and holding in abeyance the larger community power struggles (Douglas 1982; Blackman 1991; Cunningham and Byrne 2006). Urban administrators in Belfast viewed themselves as above local conflicts and often took pride in a technocratic approach (Loughlin 1992). Despite the impotence of local elected officials, one of the greatest obstacles to effective urban policymaking in Belfast under “direct rule” was local city councillors, whose relative lack of power freed them to be extreme in their interactions with government. They often had little to lose from being scaremongers who emphasize division, confl ict, and single ethnic identity. Direct-rule government units became easy targets for local councillors who would rant and rave at consultation meetings as a way to show

342

Scott A. Bollens

their communities that they care (Hendry 1995). Indeed, the confrontational and divisive rhetoric was often key to their being elected. Local politicians increased their “leadership” role most easily by tapping into separate constituencies, not seeking to span them (Fitzduff 1995). In their ability to thwart efforts by government to move communities forward, local politicians in Belfast would “lead from the back” (MacBride 1995). In the local politics of contested Belfast, urban issues often became subordinated to arguments over nationalism, constitutionalism, and symbolism. One interviewee described the “tragedy of the masquerade” represented by monthly city council meetings that resembled more juveniles on a playground than locally elected officials in a forum. There is hope that the local government of Belfast will eventually be one of shared and genuine authority across sectarian groups, that it would mimic the power-sharing Northern Ireland Assembly. Such local-level power sharing would likely build upon voluntary “responsibility sharing” arrangements used in about a dozen of Northern Ireland’s twenty-six district councils, primarily nationalist-majority ones (Knox and Carmichael 2007; Hazleton 1999). The Belfast Agreement called for a comprehensive review of local governments in Northern Ireland. A “Review of Public Administration,” initiated in 2002, called for the existing twenty-six local councils in Northern Ireland to be reduced to seven in such ways that would reduce fragmentation in local government and equalize populations and tax bases across local bodies (Knox and Carmichael 2006). Critical public functions such as planning, zoning, and transportation would be devolved to these reconstituted local units. One idea, put forth by the secretary of state for Northern Ireland, was that there should be three “nationalist” councils and three “unionist” councils across the province, with Belfast constituting a “swing” council owing to its mixed population and key anchoring position in the province. Along the tenuous path toward peace in Northern Ireland, policing and security have been extremely contentious issues that have disrupted advances in the larger peace process. The Belfast Agreement renamed the Royal Ulster Constabulary the Police Service of Northern Ireland and required that 50 percent of new hires be Catholic. Although Sinn Féin agreed to sit on the advisory Policing Board, deep distrust of police remains in Catholic neighborhoods. Another issue that has long been a source of local tension and violence has been loyalist triumphalist parades that pass by Catholic neighborhoods. A cross-sectarian Parades Commission has been established that works in cooperation with police in approving parade routes and developing codes of conduct for both parade participants and spectators.

Polarized Cities

343

Direct rule utilized an urban policy approach for twenty-five years primarily aimed at stability, neutrality, and maintenance (Byrne and Irvin 2002). This protection of the status quo defended a rigid and sterile territoriality of significant segregation and reinforced the physical and psychological correlates of urban civil war. Such an approach in the future will not help Belfast and Northern Ireland advance to a greater level of sectarian coexistence. Indeed, since the signing of the Belfast Agreement, none of the “peace walls” built by security forces during the “Troubles” has been torn down, and some new ones have even been built (Pringle, Boal, and Royle 2007). Devolution of genuine political power to city government will thus likely need to introduce a more proactive and progressive ethnic agenda able to move this urban society forward, one that is responsive to the differential and changing needs of both Catholic and Protestant communities.

Sarajevo (Bosnia-Herzegovina) MULTICULTUR AL CIT Y UNR AVELED

Sarajevo demonstrates the difficulty of sustaining the multiculturalism of a city after it has experienced the trauma of massive war aimed at its death. With a prewar mixed ethnic population in 1991 of 540,000 Bosnian Muslims (40 percent), Bosnian Serbs (30 percent), and Bosnian Croats (20 percent), Sarajevo in the early postwar years became an approximately 80 percent Muslim city of about 340,000. Despite international community (IC) acknowledgment of the city’s importance as an anchor in rebuilding Bosnia and local power-sharing efforts that sought electoral representation of displaced residents and assured representation to minority groups, the essence of the city as a multicultural entity expired soon after the end of the 1992–95 Bosnian war. The war created numerous new municipalities in Bosnia, most along the Inter-Entity Boundary Line that now separates the two autonomous regions of the Bosnian state—the Muslim-Croat Federation and Republika Srpska. In numerous cities, including Sarajevo, prewar municipalities were split into separate ethnic municipalities. Sarajevo is located at the boundary interface between the Muslim-Croat Federation and Republika Srpska. There are few restrictions on movement between the two political regions. Since policing is divided between federation and republic jurisdictions, difficulties can result when pursuing suspects that cross this boundary. A European Union Police

344

Scott A. Bollens

Mission has been present since 2003 to support police reform, strengthen police accountability, and support the fight against organized crime. Pursuant to international community requirements, power-sharing rules for local governments in the federation and the Serb republic endeavored to allocate municipal council seats proportionate to 1991 ethnic distribution, yet reconstituting local politics at the urban scale has been even more difficult than at regional and state levels. This was illustrated by the need to postpone municipal elections in postwar Bosnia and Herzegovina in August of 1996, while national elections—although flawed—were carried out the following month. Massive war-caused displacement on the scale of Bosnia-Herzegovina’s created substantial impediments to conducting municipal elections, tied as they are to specific geographies. Forty to 50 percent of Bosnia’s population no longer lived where they were registered for the 1991 census but were nonetheless provided the right to vote in 1997 municipal elections in their place of 1991 residence. The ability of displacees to vote in their former municipality meant that many city councils represented “virtual populations” of ethnic members who voted but had not physically relocated to their former place of residence (Jokay 2001). Appointments to administrative positions and hiring in public employment were also to mirror the registered 1991 ethnic distribution in the city. Such proportionality in public employment constituted “wages of peace,” aimed at intergroup stability more than government effectiveness (Jokay 2001). Given the fundamental lack of trust among ethnic groups and the ethnic fragmentation of the professional class, administrative decisions became commonly distorted by ethnic considerations. In terms of Sarajevo specifically, an initial strategy during early diplomatic efforts to counter possible ethnic claims on the city was to create a special status for it as a district under United Nations or European Community administration. This corpus separatum strategy resembled the unsuccessful proposal in 1947 to protect the city of Jerusalem through United Nations oversight and was premised on Sarajevo being different and special in BosniaHerzegovina. The urban region was to be preserved as a multiethnic capital for the entire state and a place where collective rights would be protected (Ivanovic 2003). The Owen-Stoltenberg proposal of August 1993 proposed that Sarajevo city—including all ten prewar municipalities except the Serbian stronghold Pale—would be under UN governance. One year later, the UN neutral zone idea still appeared alive in diplomatic discussions; the Contact Group Plan of July 1994 recommended UN administration of a spatially expansive Sarajevo district. However, by the time of the 1995 Dayton Accords

Polarized Cities

345

that ended the war, the idea for international governance or oversight of the city had been overtaken by the give-and-take negotiations of ethnic leaders. When peace finally came to Bosnia, peace-making paradoxically started processes that unraveled Sarajevo as a multicultural space amid a fracturing state. The Dayton Accords structured and empowered the two entities and the federation’s cantonal governments in order to facilitate ethnic self-rule. On the one hand, distributing power in a postwar state in such a way responds to a compelling logic of how to reconstruct a collapsed state. On the other hand, Dayton’s ethnic circumscription of space after the war had detrimental effects on Sarajevo, catalyzing a mass exodus in early 1996 of some sixty-two thousand Sarajevo Serbs from inside what would be the Dayton borders of Sarajevo city and its suburbs and creating today’s more monoethnic postwar city (Internal Displacement Monitoring Centre 1996). To the credit of Dayton negotiators, there was a strategy to “reunify” postwar Sarajevo and not let it become ethnically fragmented. The transfer of certain Serbian-populated districts and suburbs into postwar Sarajevo city boundaries was an effort to maintain the Bosnian Serb population within the postwar city (see Figure 14.4). Yet this “reunification” of Sarajevo was a proposition with significant ethnic salience because to be reunified within the city, the Serbs would also under Dayton be simultaneously incorporated into the Muslim-Croat Federation. This psychological factor spawned the substantial outmovement of the Bosnian Serb population from the transferred districts and suburbs to nearby Serb Republic land and to other places in that republic. If, instead, the “reunified” city was a spatially expansive zone that spanned entity boundaries and thus was not fully contained within the federation, it is probable that more Serbs would have stayed in these neighborhoods. The holding of the Serb (and Croat) population within a functional and de-ethnicized urban space was critical to the multicultural vitality of the city; the fact that such did not come to pass represents a significant lost opportunity. Even after outmigration of Serbs, another method of sustaining the city’s multiculturalism would be to construct a cross-ethnic power-sharing agreement for city governance. For the first several years after Dayton, the city had a joint governance and administrative structure that sought fair representation by minority groups. A mayor and two deputy mayors had been selected in a way that each of the main nationalistic groups—Bosniak (Muslim), Croat, and Serb—had one of these positions. From 1997 to 2000, the allocation of city council seats was also engineered to ensure multiethnic representation. In the run-up to the first municipal elections in September

346

Scott A. Bollens

SARAJEVO Illijas

Bo

Vogosca

sna

Centar Sarajevo

er Riv

I

H

Zuc Hill

a ck lja Mi

R

Snip

r

B A

K

DC E FG

M J NL Stari Grad

Grbavica

Novi Grad Lukavica

Air

Hadzici

ey

er All

ive

iver rinja R Dob

Novo Sarajevo

Rajlovac

po

Ilidza

rt Dôbrinja

Ze lje Butmir

Tunnel

Mount Jahorina

er Riv ica zn

Mount Igman

Mount Bjelasnica

Major Sites

N 0

1

Key

2 Kilometers

Roads Line of Bosnian Serb encirclement

Railroad Rivers

Suburb transferred

Mountains Built-up area

A B C D E F G H I J K L M N

Oslobodenje building PTT building Tito barracks Holiday Inn The “Unis” building Museum Parliament Zetra, Kosevo stadiums Radio/TV building Presidency building Residency Bascarsija Library Marketplace

Figure 14.4. Sarajevo’s suburbs transferred in post-Dayton transition. Adapted from King 2003.

1997, the international community carefully managed the ethnic balance for the city. The Protocol on the Organization of Sarajevo (Office of the High Representative [OHR] 1996) specified that at least 20 percent of city council seats go to Croats and another 20 percent go to other minorities. In March 1997, the constitution of the city was amended to reflect these ethnic quotas.

Polarized Cities

347

The intent was that the city would be governed on a shared basis and this would demonstrate a way forward for the Bosnian state. With the 2001 Election Law for Bosnia, however, this ethnically engineered electoral system for Sarajevo was replaced by an electoral system where council seats are allocated proportionately to popular votes garnered. As of 2004, without minimum representation quotas, an ethnic party representing Croats or Serbs held only one of the twenty-four city council seats. The days of ethnically shared governance of Sarajevo city appeared to be over. The Sarajevo and Bosnian cases also illustrate the problems associated with early elections after war. The Dayton accords, although critically beneficial from a military perspective, may have created a faulty foundation from which to politically rebuild Bosnia (Saura 2004). The push by the international community for early elections meant that they often took place amid feelings of threat and mistrust; in such circumstances, extremists and war profiteers were most likely put into office in the early years (Mier 2003). These “forces for separation,” once incorporated into the state’s structures of governance, then sought to obstruct and separate the country. One OHR interviewee suggested that the IC should have come in with greater power and acted as a protectorate instead of accommodating local ownership and cooperating with internal leaders (Mier 2003). Under this scenario, after a number of years of social and economic stability amid stronger international intervention, the electoral process might have had a greater chance of producing a more balanced assortment of local politicians.

Politically Combustible Governance These cities are the most problematic, existing in highly inflammatory settings where power sharing itself is contested and a potential contributor to further instability.

Jerusalem (Israel/Palestine) HEGEMONIC CONTROL

Israel exercises de facto hegemonic control over Jerusalem, dominating the state apparatus and channeling decision-making outcomes in partisan ways toward the privileged Israeli Jewish population. This city of almost eight

348

Scott A. Bollens

hundred thousand residents is a site of demographic, physical, and political competition. The recent social and political geography of Jerusalem has consisted of a multicultural mosaic under British control (1920–48) and then two-sided physical partitioning into Israeli and Jordanian-controlled components (1948–67). Since 1967, it has been an Israeli-controlled municipality three times the area of the pre-1967 city (due to unilateral annexation) and encompassing formerly Arab east Jerusalem. The international status of east Jerusalem today remains strongly contested. Jewish demographic advantage (of approximately 65 to 35 percent) within the Israeli-defined city borders translates into Jewish control of the city council and mayor’s office. Arabs within the annexed eastern part of Jerusalem are considered by Israel as residents of the city but not citizens of Israel. Although they can vote in municipal elections, they resist participating in municipal elections they deem illegitimate. Over the past forty-five years Israel has created an urban landscape of visible and stark inequalities. “From the very first, all major development represented politically and strategically motivated planning,” admits Israel Kimhi (1994). Equating demographic dominance with political control, large Jewish communities have been built in strategic locations throughout the annexed and disputed municipal area (see Figure 14.5). Of the approximately twentyseven square miles unilaterally annexed after the 1967 war, the Israeli government has expropriated approximately 33 percent, and neighborhoods built in these areas are home today to approximately 200,000 Jewish residents. At the same time, Israeli planners have curtailed the growth of Palestinian neighborhoods through land expropriation, restrictive environmentally based “green area zoning,” road construction that fragments Arab neighborhoods, “hidden guidelines” within Israeli plans that cap building volume in some areas, and the intentional absence of plans in others. The cumulative impact of Israeli restrictions on Palestinian growth is that only 11 percent of annexed east Jerusalem in 1995 was vacant land where the Israeli government allowed Palestinian development. Although lacking a physical partition since 1967, Jerusalem is a city functionally and psychologically divided. The building of the Israeli separation barrier since June 2002, including what will be a forty-mile wall extending alongside and beyond the city’s politically contested municipal boundaries, is the most recent and imposing manifestation of Israel’s use of land development and planning for security and political goals. Policing and security in the Jerusalem area are tightly controlled by Israel. Local police functions are the domain of the Israeli Police Force and the border police, which is the combat arm of the police that serves in trouble-

Armistice Line 1949 Jerusalem Municipal Boundary Palestinian Neighborhood Enclave Israeli Settlement Separated from Palestinian Neighborhood West Jerusalem & Settlements Continuity Open Area Israeli Settlement Palestinian Settlement Al Qubeiba Biddu

Ramallah Al Bireh Governorate

Ramallah Kokhav Ya’aqov

Rafat

Kafr ’Aqab

Al Juderia

Ar Ram

Al Jib

Dahiyat al Bareed

Bir Nabala

Jerusalem Palestinian Governorate

Har Shmuel

Jaba

Neve Ya’aqov

Beit Hanina al Balad Beit Hanina Pisgat Ze’ev

Har Adar

Beit Surik Beit Iksa

Jerusalem Israeli District

EAST Pisgat ’Omer JERUSALEM

Ramot

Rekhes Shu’afat

Shu’fat

Meyatseret Tsiyon

’Anata

French Hill

Givet Shaoul

Castel Industrial Zone

Shu’fat Camp

Kiryat moshe

Romemma

OLD CITY At Tur

Beit Zayit

WEST JERUSALEM Ras al Amud

Ein Karem Even Sapir

Kiryat Hayovel

Givat Mordechai

Silwan Baq’a

Amindav

Eir Ganim Malha

Qatamon East Talpiot

Taipiyot Industrial

Jabal Al Mukkabber

Beit Safafa Al Waiaja

Sur Bahir Gilo

Bethelem Governorate

0

Abu Ghneim

Har Gilo

N

Battir

1

2 Kilometers

Bethlehem Beit Jala

Figure 14.5. Jerusalem’s political geography. Adapted from International Peace and Cooperation Center 2005.

350

Scott A. Bollens

some areas such as the border between Israel and the West Bank. A Palestinian National Security Force now operates, with limitations, in parts of the West Bank but not in east Jerusalem. There are fourteen permanently staffed Israeli checkpoints in Jerusalem district along the city borders that monitor and restrict Arab/Palestinian mobility (B’Tselem 2008). Arab “permanent residents” of Jerusalem carry blue (Israeli) identification cards that allow travel throughout Israel. However, residency permits can be revoked if “links” to the city are not maintained. In addition, it is becoming increasingly difficult for partners of cardholders to obtain blue IDs. Arab residents of the West Bank and Gaza Strip must carry orange ID cards and are subject to substantial restrictions when attempting to enter Israel sovereign territory and Arab east Jerusalem. Further, Arabs who are barred outright from entering Israel are issued green ID cards. An alternative, special arrangement for the governance of Jerusalem was considered at an important historic juncture in the city’s trajectory. When the British Mandate period (1917–48) came to a close, there was a UN resolution that the city of Jerusalem be a demilitarized and neutral corpus separatum (separate entity) governed by a special international regime and administered by the UN. This resolution (181 II, the Future Government of Palestine. 29 November 1947) called for the internationalization of Jerusalem within a context of a recommended partitioning of Mandatory Palestine into Jewish and Arab states. The resolution was approved by the national leadership of the Jewish community in Palestine and rejected by the Arab Higher Committee. Intense warfare between Israeli, Transjordan (now Jordan), and Egyptian armies, centered on the Holy City, turned Jerusalem instead into a physically divided city for almost twenty years, with Israeli west and Arab east parts separated by concrete barriers and “no-man’s lands.” It was this physically divided city that was transformed by the beginning of Israeli control and annexation in 1967. In the early years of contested Jerusalem under Israeli control, a borough plan was debated from 1968 to 1977, envisioning a single municipal government under dual sovereignty, the representation of Palestinians in the running of the city, and the creation of separate semiautonomous borough governments to manage local affairs in different ethnic neighborhoods. The possibility of creating spatially cohesive local ethnic boroughs has become increasingly problematic with the construction of large Israeli neighborhoods in the contested eastern part of the city. Either drastic relocation

Polarized Cities

351

would need to occur to ethnically sort the urban region or local boundaries must be drawn in disfigured, noncontiguous ways that dampen ethnic community cohesiveness. Through the years, numerous other ideas have been put forth to address the conflicting sovereignty claims of Israelis and Palestinians, oftentimes by nongovernmental organizations and sometimes in formal negotiations. At the Camp David Summit in 2000, for instance, Ross (2004) reports that key elements of a Jerusalem proposal provided Palestinian sovereignty over specified outer neighborhoods and over the Muslim and Christian quarters of the Old City, meaningful Palestinian self-government (including planning and zoning, security, and dispute resolution responsibility) in inner neighborhoods (although under Israeli sovereignty), and Palestinian custodianship over the Haram al-Sharif. A neighborhood outside the Israeli borders of Jerusalem was seen as a possible site for a Palestinian capital, with the municipal border enlarged to encompass this area in order to meet Palestinian demands that “Jerusalem” be their capital. Other ideas floated less officially have included a model of “scattered sovereignty” (Baskin and Twite 1993). Jewish communities (in both west and east Jerusalem) would be under the sovereignty of Israel, while Arab communities in east Jerusalem would be within Palestinian sovereignty. Sovereignty would be “scattered” because of the dispersed and complex spatial mosaic patterns of Arab and Jewish communities. A simpler geographic demarcation of sovereignty would be to use the pre-1967 green line to delineate west from east. This “simple solution,” however, is made difficult by the now Jewish majority in east Jerusalem. One response might be to reestablish an Arab majority in east Jerusalem by expanding eastern Jerusalem borders to include Arab communities now just outside the Israeli-drawn border. In this plan, Jewish neighborhoods in eastern Jerusalem would be under Palestinian control, although residents would remain Israeli citizens. Another idea would enlarge the jurisdictional scale of Jerusalem. A metropolitan expansion of Jerusalem’s borders would encompass within the new larger city approximately equal Arab and Jewish populations, thus likely increasing the parity of sovereignty solutions. There could then be two ethnically based municipalities under a joint umbrella metropolitan council. K. Tufakji (1994) states that the chances that Jerusalem can be solved geographically increase with a metropolitan federalism of two sovereignties. Sovereignty issues within today’s Jerusalem would remain, however.

352

Scott A. Bollens

Baghdad and Kirkuk (Iraq) BAGHDAD: CAPITAL CIT Y OF A FEDER ALIST IR AQ?

The governance and management of Baghdad present two primary challenges: (1) how to locally govern a city that has been segregated, cleansed, and sorted during war and (2) how to govern the city in a way that might hold together a country that likely faces some federalist devolution of national power to ethnic autonomous zones. Can Baghdad constitute a multiethnic capital district or zone that holds together a fragmented or federalized state? From early 2006 to November 2007, over 65 percent of Iraqis displaced due to sectarian fighting were in ethnically mixed Baghdad, the site of bitter fighting between Sunni Arab insurgents and Shiite Muslim militias (United Nations OCHA 2007). The city extensively segregated with substantial sectarian homogeneity of neighborhoods, with a discernible stronghold of Shiite neighborhoods east of the Tigris River and in two swaths west of the river. Blast walls and physical partitions were constructed by occupation forces at particular sectarian friction points. A green zone with its perimeter secured by Iraqis, Americans, and security contractors provided some greater security but also was a target of mortar blasts. The increased “surge” of American troops beginning in early 2007 had some success in decreasing violence, yet there was continued takhalasu (purging) of formerly mixed neighborhoods. Violence was especially chronic in mixed Shiite-Sunni neighborhoods in western Baghdad. The stability of Baghdad was consistently a key plank of American military and political planners. In this view, the protection of the Iraqi population in Baghdad is primary because it would allow breathing space to Iraqi leaders to achieve needed political reconciliation. “Localized security” through agreements at the local level, including with militias and former insurgents, would be a necessary complement and encouragement to national compromises (Gordon 2006). These local arrangements would then be stitched together to establish a broader sense of security on a nationwide basis. The American troop surge of 2007 was a larger manifestation of this desire to stabilize the city as a key foundation of larger national accords. In an effort to introduce local democratic institutions in Baghdad, the United States attempted to establish a three-tier system utilizing neighborhood, district, and city council representation. First, there was election of councils in the neighborhoods officially designated before the war, elected

Polarized Cities

353

by neighborhood caucuses. When all eighty-eight neighborhood councils were in place in the city, each neighborhood council then elected representatives from among their members to serve on one of the city’s nine district councils. Then, each of these nine district councils elected representatives from their membership to serve on the thirty-seven-member Baghdad City Council. There is also a thirty-five-member Baghdad Regional Council representing the city and the other communities in Baghdad Province outside of the city itself. Finally, there is a Baghdad Provincial Council, elected from the lower councils in numbers proportional to the population of the districts they represent. Significant displacement and sectarian purging occurred also at the national level. By late 2007, there were an estimated 2.3 million internally displaced people (IDPs) in Iraq. Since the intensification of sectarian conflict in February 2006 (with the bombing of the Al-Askari shrine in Samara), the majority of IDPs have been displaced from central provinces such as Baghdad (see Figure 14.6). Before then, displacement occurred most frequently in northern and southern regions (United Nations OCHA 2007). With such sectarian sorting of the country comes the opportunity to devolve substantial political powers to ethnic autonomous regions as a way to encourage buy-in to national agreements by the country’s Sunnis, Kurds, and Shiites. The 2005 constitution defines the country for the first time in its history as a “federal” country, and its federalism law allows semiautonomous regions to be created through referenda out of one or more existing provincial governorates. One possibility is that there would be reconstitution of the country’s eighteen provinces as three self-governing entities and reconstruction of Iraq as a loose confederation of these governments. A key centerpiece of any sustainable federalist arrangement for Iraq would be the creation and protection of Baghdad city as a multiethnic capital district with special status. Within this special district, power sharing of local governance and protections afforded minority residents would be needed to avoid continued fragmentation and dismemberment of Baghdad into sectarian districts of autonomy linked more to regional ethnic geographies than to a citywide political corpus. KIRKUK: NORTHERN FL ASHPOINT

Kirkuk province is a northern oil-rich region of ethnic diversity and competition whose governance is crucial to the sustainability of the Iraqi state and to the allegiance of the Kurdish population to it (see Figure 14.6). Much of the

354

Scott A. Bollens

Figure 14.6. Iraq’s provinces. Adapted from Central Intelligence Agency (CIA) map of Iraq available at www.cia.gov.

Kirkuk population comprises Kurds, who consider it their holy place, but there is a significant minority population of Turkmens and Arabs. The provincial capital is the city of Kirkuk, whose population makes up about 90 percent of the provincial population. Saddam Hussein displaced thousands of Kurds from Kirkuk and parts of Nineveh and Diyala provinces as part of his “Arabization” plan to control the region’s oil. Arab tribes from southern Iraq were enticed to move to the north with government benefits and offers of housing. In addition, there was gerrymandering of the province’s borders during this time to lessen its size and detach it from four other traditionally Kurdish districts. Since the end of Hussein’s rule, Kurds have been empowered through

Polarized Cities

355

the creation of a semiautonomous Kurdish region. Whether the contested lands of Kirkuk will join this region of Kurdistan or remain with the rest of Iraq positions Kirkuk as a flashpoint of ethnic and sectarian conflict and a key element of national negotiations over the future status of the country. There are three provinces currently fully under the authority of the Kurdistan Regional Government; however, the regional government claims in whole or in part four other provinces, including Kirkuk. Kurds have agreed on a set of national issues contingent on favorable resolution of the Kirkuk issue. Provincial elections in 2005 produced a Kurdish majority in the province (26 of 41 seats). The intent was that there was then to be a “normalization” process to include the reintegration of the four districts detached from Kirkuk province by Hussein. This would then be followed by a referendum (by November 2007) to decide whether the province would become part of the Kurdistan regional government. Efforts have been under way by all sides to create demographic “facts on the ground” in advance of the referendum. Kurdish negotiators have proposed a binding political pact between the leadership of the Shia, Sunni, and Kurdish blocs to return the administrative boundaries of Kirkuk to the 1970 map, thereby adding the four Kurdishmajority surrounding districts to Kirkuk and ensuring Kurdish success in any future referendum (Kritz, al-Sarraf, and Their 2007). In July 2008, an Arabsponsored plan was put forth to delay elections in Kirkuk province and city and impose a quota-based power-sharing arrangement that would apportion power in the provincial government equally among Kurds, Arabs, and Turkmens. This proposal further inflamed the situation and jeopardized provincial elections in all of Iraq.

Conclusion In robustly nationalistic cities where group identity matters substantially, leaders need to accommodate the unique needs of each salient group while building and protecting a citywide public interest. There is little possibility of reliance on the traditional model of ethnic assimilation; there must rather be governance that accommodates ethnic difference and focuses on coexistent community viability. Further, majoritarian democracy will not work in these polarized cases because cross-group coalition building is limited, and members of an aggrieved minority group may boycott the political process because they see it as unfair or illegitimate. Power-sharing elements often employed at

356

Scott A. Bollens

the national level—including multitier governance, collective executive bodies, communal legislative bodies, reserved seats or quotas in legislatures, and formal rules mandating proportional resource allocation and public sector hires—must come into play and take local and metropolitan forms. The structuring of local government and elections can be useful in allowing a minority group with little power at the national level to potentially find greater inclusion and different partners in the local arena. Beyond this, local and metropolitan governance arrangements that utilize multiple levels of representation can be key elements in efforts to divide political power among and within levels of government so as to create “multiple majorities” with possibilities for alignments that cut across group identity. Such a powerdividing approach “begins at the bottom” and creates diverse, numerous, and nonoverlapping political bodies able to foster multiple and cross-cutting constituencies (Roeder and Rothchild 2005). The seven cases presented here involving the local governance of polarized cites point to the frequent fragility of such efforts, the evolutionary nature of even the “best case” examples, and the role of local governance reform amid national political progress. Even in the sustainable cases—Brussels and Johannesburg—where there has been some amelioration of political conflict and stability of the local and national state, sustainability does not necessarily connote stability of institutions and arrangements. Brussels and Johannesburg have changed and modified governing arrangements to accommodate new needs and circumstances. In these cases, there is an institutional learning process sufficiently embedded and open to democratic innovation such that long-term sustainability is likely. Brussels’s governance structure has been a work in progress now for over four decades. And in Johannesburg, the establishment of a sustainable system of governance did not happen overnight but was created over a five-year period out of numerous and multilayered transitional forms of governing institutions. As late as 2000, local and metropolitan government was still being substantially reorganized. In the fragile cases of Belfast and Sarajevo, local governance arrangements are not sufficiently stable and are subject to relapse. Local governance is emergent and untested in both cases. These cases will likely need to undergo significant restructuring and experimentation regarding how local governance structures can effectively address local problems and obstacles. Fragile local democratic institutions must be able to create visible on-theground outcomes (in terms of job opportunities, better ser vices, and safer neighborhoods) that can reinforce and gain public acceptance of shared lo-

Polarized Cities

357

cal governance. Only then can we be assured that these precarious city institutions survive and become key agents of intergroup coexistence and local anchors to national stability over the longer term. Yet, cities such as Belfast present a difficult predicament—that in the short term the constraints on local democracy that are necessary for shared power may make local government less effective in producing such tangible change in the city. Jerusalem and the Iraqi cities are cases of unformed shared governance where the question of how the city is to be governed is itself part of larger political negotiations. These cities can be major roadblocks and obstacles to larger national peace agreements and constitutional arrangements. The future of Baghdad and Kirkuk is one of debilitating sectarian fragmentation unless strong and effective integrative institutions are put in place. And in Jerusalem, any sustainable peace agreement between Israelis and Palestinians will need to directly confront the mechanisms and outcomes of forty-five years of partisan governance in the Holy City. Shared urban policies and institutions can set important local precedents that can positively shape broader longterm social and political development. Absent effective cross-group urban governance, however, ethnic antagonists who recognize the power of the city will likely be successful in their efforts to submerge and fragment the peaceconstitutive potential of the city in pursuit of their own group aspirations. Four of the case-study cities also allowed for the analysis of local democratic management during major transitions associated with regime change or postwar reconstitution: Johannesburg, Sarajevo, and the two Iraqi cities. Two additional lessons emerge. First, from Sarajevo, international agreements that stop war must be cognizant of the new ethnic geographies of local and substate governments that they create so that local governments are not unintentionally stunted in their inherent ability to bring people together, over time. Second, from Johannesburg, two-tier governance utilizing metropolitan and local levels can be particularly useful during political transitions. By creating a metropolitan level of governance during a period of political transition, new distributions of power can be created that transcend older, more ossified ethnic power structures. It can open up opportunity spaces for conflict management by dividing power between levels of government and by enabling new alignments of constituencies that are not necessarily group based. Metropolitanism can be an effective mechanism not only during times of major political transitions but also in more stable arrangements, such as Brussels, that have used multitier governance (often in complex and evolutionary ways) to ensure the territorial and political expression of identity groups.

358

Scott A. Bollens

Local political arrangements that allocate power to identity groups in some shared and engineered way are likely required for resolution of urban tensions amid contested sovereignty. By building ethnic group power into local institutions, however, these systems can ethnicize many policy issues, dampen crosscutting cleavages, and concentrate institutional power in the hands of ethnic politicians who then have the means to escalate demands (Roeder and Rothchild 2005). Thus these structural solutions may provide autonomy and sovereignty at the expense of ethnic separation and isolation and, ultimately, urban and regional dysfunctionality. Something more is needed. Political and physical separation tears at the heart and soul of the urban region, hermetically sealing antagonistic sides within their respective political chambers or behind physical walls of hatred. Two-tier structuring of local government that establishes both integrated and separate institutions is more moderate, yet it also can lead to a functionally disconnected and economically stagnant urban area for one or both groups. Even joint, power-sharing political control of the city can disintegrate into a condition of urban paralysis amid policy vetoes. Essential in these circumstances are on-the-ground urban policies and strategies that facilitate mutual tolerance and coexistence. Any local political arrangement, no matter how constituted, must have the capacity and means to implement programs and produce outcomes that make a meaningful difference to all groups in the divided society. Under either shared local rule or international management, some effective city policymaking must occur that moves urban life toward mutual accommodation and away from rigidity and the status quo. Seen in this light, political negotiations that restructure local political power represent a first, but by no means sufficient, step toward normalizing polarized cities. Notes 1. Attempts to incorporate a partnering local component into national restructuring would need to contend, however, with the possibility of conflict between a national and local state. Cities are institutionally established by a national state whose interests may not be primarily city based. Further, national constitutions can be stingy in granting authority to urban governments. India and South Africa are among the few countries whose national constitutions explicitly recognize local government in their clauses. 2. Although I focus on its contemporary manifestations, Hepburn (2004) reminds us that contested cities are a historical fact; for example, urban regions such as Trieste and Prague (under the Habsburgs), Helsinki (under tsarist rule), and Danzig/Gdansk (under Prussian and German rule).

Polarized Cities

359

3. Montreal is an additional exemplar of the first category, Mostar (Bosnia) and Nicosia are further examples of the second type, and Mitrovica (Kosovo) is representative of the inflammatory category. 4. Other comparative overviews of ethnic management techniques include Smith 1969 (modes of collective accommodation); Esman 1973 (regime objectives); and Palley 1979 (constitutional devices). 5. Swyngedouw and Moyersoen (2006) describe four different stages of state reform in Belgium: 1970 (creation of regions), 1980 (creation of communities), 1989 (creation of Brussels Capital Region), and 1993 (creation of Belgium as federal state). 6. Direct transitions from minority (or authoritarian) governments to majoritarian democracy would likely cause major strains and disruptions and could be susceptible to setback. In cases such as Barcelona (Spain), however, such an unfi ltered transition occurred without major setbacks. 7. Because of this guarantee of white minority representation, local and metropolitan councils elected in Johannesburg and elsewhere were still technically transitional councils. The interim stage of local government restructuring did not end until local elections occurred pursuant to new constitutional principles in 1999. 8. Identity of interviewee withheld upon request. 9. The most common mechanisms in these agreements are office rotation of mayor and vice mayor positions and proportional arrangements for assigning committee chairs and/or membership. 10. The end of guaranteed representation of Croats and Serbs on the city council did not lead to dominance of Sarajevo city politics by the Bosniak nationalist Party of Democratic Action (SDA). In 2000 and 2004 city elections, Sarajevo voters opted more for the multiethnic Social Democratic Party and the moderate Bosniak “softnationalist” Party for Bosnia and Herzegovina than they did for the SDA. 11. An example of such a protectorate role for the international community is the Brcko district in northeast Bosnia, which the IC placed under strong international supervision and declared a unified and neutral District in 1999. 12. An Israeli metropolitan planning study found that within the functional commuting region of Jerusalem 54 percent of the 1.14 million population were Arab and 46 percent were Jewish (Mazor and Cohen 1994). 13. The Kurds in the north already have extensive regional government and autonomy in their Kurdistan. Several southern provinces have significant Shiite majorities. 14. Another example of evolutionary change in institutional arrangements is bicultural Montreal, which, since 1996, has reorganized metropolitan-level government, amalgamated local governments in the urban core, and decentralized some political power to boroughs (Collin and Robertson 2005). 15. In other examples of fragility, local public authority has faced severe obstructions by nationalist leaders in Mostar, has been hamstrung by rigid power-sharing formulas in Beirut, and has been divided in Nicosia (Bollens 2007; Salamey and Tabar 2008).

360

Scott A. Bollens

16. Similarly situated is the divided Serb-Albanian city of Mitrovica, which presents a critical challenge to the sustainability of the disputed newly independent country of Kosovo. UN recognition of Serb group aspirations in this city intended to ameliorate confl ict may at the same time lay a foundation for a damaging institutional separation of Serbian parts of Kosovo (Spahiu 2002). References Baskin, Gershon, and Robin Twite, eds. 1993. The Future of Jerusalem: Proceedings of the First Israeli-Palestinian International Academic Seminar on the Future of Jerusalem. Jerusalem: Israel/Palestine Center for Research and Information. Beavon, K. S. 1992. “The Post-Apartheid City: Hopes, Possibilities, and Harsh Realities.” In David M. Smith, ed., The Apartheid City and Beyond: Urbanization and Social Change in South Africa. London: Routledge. Blackman, Tim. 1991. Planning Belfast: A Case Study of Public Policy and Community Action. Aldershot, UK: Avebury. Boal, Frederick. 1994. “Belfast: A City on Edge.” In Hugh Clout, ed., Europe’s Cities in the Late Twentieth Century. Amsterdam: Royal Dutch Geographical Society. Bollens, Scott A. 2007. Cities, Nationalism, and Democratization. London: Routledge. B’Tselem. 2008. “Restrictions on Movement: Information on Checkpoints and Roadblocks.” http://www.btselem.org/English/Freedom_of_Movement/Statistics.asp (accessed 18 December 2008). Byrne, S., and C. Irvin. 2002. “A Shared Common Sense: Perceptions of the Material Effects and Impacts of Economic Growth in Northern Ireland.” Civil Wars 5, 1: 55–86. Collin, Jean-Pierre, and Melanie Robertson. 2005. “The Borough System of Consolidated Montreal: Revisiting Urban Governance in a Composite Metropolis.” Journal of Urban Affairs 27, 3: 307–30. Cunningham, Chris, and Sean Byrne. 2006. “Peacebuilding in Belfast: Urban Governance in Polarized Societies.” International Journal on World Peace 23, 1: 41–73. “Districts of Iraq.” 2007. http://en.wikipedia.org /wiki/File:Iraq _Dist.png. GNU Free Documentation License (accessed 4 January 2009). Donaldson, B. C. 1983. Dutch: A Linguistic History of Holland and Belgium. Leiden, Netherlands: Martinus Nijhoff. Douglas, J. Neville. 1982. “Northern Ireland: Spatial Frameworks and Community Relations.” In Frederick W. Boal and J. Neville Douglas, eds., Integration and Division: Geographical Perspectives on the Northern Ireland Problem, 105–35. London: Academic Press. Esman, M. J. 1973. “The Management of Communal Conflict.” Public Policy 21, 1: 49–78. Ewing, Deborah. 1995. Guide to Local Government Elections. Durban, South Africa: Y Press. Fitzduff, Mari. 1995. Interview with author. Director, Northern Ireland Community Relations Council. 13 January.

Polarized Cities

361

Gordon, Michael. 2006. “U.S. Seen in Iraq Until at Least 2009.” New York Times, 24 July, pp. A1, A10. Hadfield, Brigid. 1992. “The Northern Ireland Constitution.” In Brigid Hadfield, ed., Northern Ireland: Politics and Constitution, 1–12. Buckingham: Open University Press. Hart, Tim. 1995. Interview with author. Urban geographer, SRK Engineers, Johannesburg. 10 August. Hazleton, William A. 1999. “Local Government and the Peace Process.” In John Harrington, Elizabeth J. Mitchell, and American Conference for Irish Studies, eds., Politics and Per formance in Contemporary Northern Ireland, 174–96. Amherst: University of Massachusetts Press. Hendry, John. 1995. Interview with author. Professor of Town and Regional Planning, Department of Environmental Planning, Queen’s University of Belfast. 18 January. Hepburn, A. C. 2004. Contested Cities in the Modern West. New York: Palgrave Macmillan. Hooghe, L. 1995. “Belgian Federalism and the European Community.” In J. B. Jones and M. Keating, eds., The European Union and the Region, 134–65. Oxford: Oxford University Press. Horowitz, Donald L. 1985. Ethnic Groups in Conflict. Berkeley: University of California Press. Internal Displacement Monitoring Centre. 1996. “More Population Displacement in 1996.” Oslo: Norwegian Refugee Council. International IDEA (Institute for Democracy and Electoral Assistance). 2001. Democracy at the Local Level: The International IDEA Handbook on Participation, Representation, Conflict Management, and Governance. Stockholm: International IDEA. International Peace and Cooperation Center (IPCC). 2005. Jerusalem on the Map. Jerusalem: IPCC. Ivanovic, Dragan. 2003. Interview with author. Deputy Speaker, Sarajevo Canton Assembly; Member, Federation Parliament (Chamber of Peoples); Director, Center for Policy Research and Development. 24 November. Jokay, Charles. 2001. “Local Government in Bosnia and Herzegovina.” In Emili Kandeva, ed., Stabilization of Local Governments, 91–140. Budapest: Local Government and Public Ser vice Reform Initiative. Keith, Michael. 2005. After the Cosmopolitan? Multicultural Cities and the Future of Racism. London: Routledge. Kesteloot, Christian, and Pieter Saey. 2002. “Brussels: A Truncated Metropolis.” GeoJournal 58: 53– 63. Khuri, I. Fuad. 1975. From Village to Suburb: Order and Change in Greater Beirut. Chicago: University of Chicago Press. Kimhi, Israel. 1994. Interview with author. Jerusalem Institute of Israel Studies. City planner, Municipality of Jerusalem (1963–86). 20 October.

362

Scott A. Bollens

King, Curtis S. 2003. “The Siege of Sarajevo, 1992–1995.” In William G. Robertson and Lawrence A. Yates, eds., Block by Block: The Challenges of Urban Operations, 235– 90. Fort Leavenworth, KS: Combat Studies Institute, U.S. Army Command and General Staff College Press. Knox, C., and P. Carmichael. 2006. “Bureau Shuffling? The Review of Public Administration in Northern Ireland.” Public Administration 84, 4: 941– 65. ———. 2007. “Making Progress in Northern Ireland? Evidence from Recent Elections.” Government and Opposition 33, 3: 372–93. Kritz, Neil J., Sermid al-Sarraf, and J. Alexander Their. 2007. “Constitutional Reform in Iraq: Improving Prospects, Political Decisions Needed.” USIPeace Briefing. September. Lijphart, Arend. 1968. The Politics of Accommodation: Pluralism and Democracy in the Netherlands. Berkeley: University of California Press. ———. 1977. Democracy in Plural Societies: A Comparative Exploration. New Haven, CT: Yale University Press. Loughlin, John. 1992. “Administering Policy in Northern Ireland.” In Brigid Hadfield, ed., Northern Ireland: Politics and Constitution, 60–75. Buckingham: Open University Press. Lustick, I. 1979. “Stability in Deeply Divided Societies: Consociationalisation vs. Control.” World Politics 31: 325– 44. MacBride, Deirdre. 1995. Interview with author. Housing and Projects Officer, Community Development Centre, North Belfast. 30 January. Mazor, Adam, and Shermiyahu Cohen. 1994. Metropolitan Jerusalem: Master Plan and Development Plan. Summary Document. June. Merrifield, Andy, and Erik Swyngedouw, eds. 1997. The Urbanization of Injustice. New York: New York University Press. Mier, Javier. 2003. Interview with author. Criminal Institutions and Prosecutorial Reform Unit, Office of the High Representative, Sarajevo. 21 November. Nordlinger, Eric A. 1972. Conflict Regulation in Divided Societies. Boston: Center for International Affairs, Harvard University. Office of the High Representative (OHR). 1996. Protocol on the Organization of Sarajevo. Federation Forum Meeting. 25 October. Sarajevo: OHR. O’Leary, Brendan, and John McGarry. 1995. “Regulating Nations and Ethnic Communities.” In A. Breton, G. Galeotti, P. Salmon, and R. Wintrobe, eds., Nationalism and Rationality, 245–89. Cambridge: Cambridge University Press. Olver, Crispian. 1995. Interview with author. Director, RDP Development Planning, Ministry in the Office of the President, Pretoria. 25 August. Palley, Claire. 1979. Constitutional Law and Minorities. London: Minority Rights Group. Parnell, S. M., and G. H. Pirie. 1991. “Johannesburg.” In Anthony Lemon, ed., Homes Apart: South Africa’s Segregated Cities. London: Paul Chapman Publishing. Pringle, M. E., F. W. Boal, and S. A. Royle. 2007. The Enduring City: Belfast in the 20th Century. London: Blackstaff Press.

Polarized Cities

363

Reid, Graeme. 1995. Interview with author. General manager/lawyer, Planact. 28 July. Roeder, Philip G., and Donald Rothchild, eds. 2005. Sustainable Peace: Power and Democracy After Civil Wars. Ithaca, NY: Cornell University Press. Ross, Dennis. 2004. The Missing Peace: The Inside Story of the Fight for Middle East Peace. New York: Farrar, Straus and Giroux. Salamey, Imad, and Paul Tabar. 2008. “Consociational Democracy and Urban Sustainability: Transforming the Confessional Divides in Beirut.” Ethnopolitics 7, 2–3: 239– 63. Saura, Jaume. 2004. Interview with author. Member, Elections Monitoring Team, Bosnia, and Professor of International Law, University of Barcelona. 20 May. Sisk, Timothy D., and Christoph Stefes. 2005. “Power Sharing as an Interim Step in Peace Building: Lessons from South Africa.” In Philip G. Roeder and Donald Rothchild, eds., Sustainable Peace: Power and Democracy After Civil Wars, 293– 317. Ithaca, NY: Cornell University Press. Smith, M. G. 1969. “Some Developments in the Analytic Framework of Pluralism.” In Leo Kuper and M. G. Smith, eds., Pluralism in Africa. Berkeley: University of California Press. Smooha, Sammy. 1980. “Control of Minorities in Israel and Northern Ireland.” Comparative Studies in Society and History 22, 2: 256–80. Spahiu, Nexhmedin. 2002. “ ‘Legalized’ Division of Mitrovica.” Balkan Crisis Report. Report No. 393. Institute for War and Peace Reporting. 23 December. Suara, Jaume. 2004. Interview with author. Member, Elections Monitoring Team, Mostar, June 1996; Professor of International Law, University of Barcelona. 20 May. Sweeney, Paul. 1995. Interview with author. Advisor, Department of the Environment for Northern Ireland. 2 February. Swyngedouw, Erik, and Johan Moyersoen. 2006. “Reluctant Globalizers: The Paradoxes of ‘Glocal’ Development in Brussels.” In M. Mark Amen, Kevin Archer, and M. Martin Bosman, eds., Relocating Global Cities: From the Center to the Margins, 155–77. Lanham, MD: Rowman and Littlefield. Terhorst, P. J. F., and J. C. L. van de Ven. 1997. “Fragmented Brussels and Consolidated Amsterdam: A Comparative Study of the Spatial Orga nization of Property Rights.” Netherlands Geographical Studies 223. Tufakji, Khalil. 1994. Interview with author. Geographer, Arab Studies Society; Member, Palestinian-Israeli Security Committee. 10 December. United Nations. Office for the Coordination of Humanitarian Affairs (OCHA). 2007. “Humanitarian Crisis in Iraq: Facts and Figures.” http://www.uniraq.org /docu ments/ Amman. 13 November.

CHAPTER 15

Power Sharing in Kirkuk: The Need for Compromise Liam Anderson

On July 22, 2008, the Iraqi Parliament passed the provincial elections law with 127 of the 142 members present voting in favor. Kurdistan’s members of the Baghdad Parliament did not attend the vote, having staged a walkout in protest over the inclusion of Article 24 of the law. This delayed elections in Kirkuk but mandated, in the meantime, power sharing among the governorate’s three main ethnic groups (Kurds, Arabs, and Turkmen) on the basis of parity. A day later, the law was vetoed by the Presidency Council, throwing into jeopardy the entire provincial elections process, a process viewed by many as a vital prerequisite for Iraq-wide political reconciliation. This incident offers a graphic illustration of potential for the Kirkuk issue to unravel the tenuous political consensus in Iraq. It is no exaggeration to state that the future stability of Iraq revolves around finding an acceptable solution to the problem of Kirkuk. At its core, the dispute centers on whether or not Kirkuk city and governorate (province) should be incorporated into the Kurdistan Region and how it should be governed in the future. Any mutually acceptable compromise seems destined to require some form of power-sharing arrangement for Kirkuk, but the precise nature of this remains subject to debate. By most reliable measures, the Kurds comprise a majority in Kirkuk, and the extant literature on power sharing in ethnically divided societies, though voluminous, offers little in the way of guidance for dealing with majority-minority situations (the so-called 60–40 problem). Both major theoretical approaches—consociation and integration—struggle to provide a convincing explanation for why a majority group would voluntarily accept

Kirkuk

365

institutions that enhance the power of minority groups at its own expense. Bluntly put, as the majority group in Kirkuk, the Kurds have no compelling incentive to share power with other groups, and any proposed solution to the Kirkuk problem that fails to acknowledge this cannot plausibly succeed. Ironically, it is the very inability of integrationist and consociational approaches to deal with the 60–40 problem that yields the crucial insight necessary to resolving the Kirkuk issue. The analysis proceeds as follows. In the first section I provide a brief synopsis of the various dimensions of the struggle for Kirkuk and explain why the dispute has reached its current impasse. The section that follows examines the two major theoretical approaches to governing ethnically divided societies and illustrates how both struggle to deal convincingly with majority-minority situations such as that in Kirkuk. That both approaches are “motivationally inadequate” in this respect yields a seemingly obvious but largely unacknowledged conclusion. Majority groups cannot be expected to share power unless furnished with appropriate incentives. The subsequent section shifts the focus back to Kirkuk and argues that the logical solution, therefore, is a deal whereby Kirkuk is incorporated into the Kurdistan Region (the incentive) in return for which the Kurds agree to share power with Arabs, Turkmen and the small Christian community (the trade-off ). The concluding section briefly assesses the broader significance of such a deal.

The Struggle for Kirkuk It is seemingly impossible for a journalist to write an account of Kirkuk that does not include the terms “oil rich” and “ethnically divided.” Certainly Kirkuk’s oil and its demographic composition are central to an understanding of its past history and current predicament, but the logical implication that the struggle for Kirkuk can, therefore, be distilled down to a struggle among ethnic groups for control over oil is seriously misleading. Kirkuk’s oil has made it a strategic prize since production first began in the late 1920s, and the key concern for a succession of Arab regimes during the twentieth century was to ensure that one of Iraq’s major assets remained in Arab hands. By 1957, however, census data indicated that Arabs were in a clear minority in Kirkuk governorate and city. The same data showed the Kurds on the cusp of majority status in the governorate, while Turkmens

366

Liam Anderson

held a plurality (though not a majority) in the city. The numerical dominance of Kurds in Kirkuk evolved into a tangible strategic threat to the government after 1961 following the outbreak of hostilities between Kurdish forces and the Iraqi army. In 1970, the newly installed Baath regime negotiated an autonomy agreement, directly between Saddam Hussein and Mustafa Barzani, the president of the Kurdistan Democratic Party, that would have allowed the Kurds extensive powers of self-government and cultural autonomy in majority Kurdish areas. A census would determine the final boundaries of the Kurdish autonomous area. The census was never conducted, however, because Saddam Hussein foresaw it would lead to Kirkuk becoming part of the Kurdistan Region, and war resumed between the Kurdistan Democratic Party and the Baathists. After the military defeat of Barzani (occasioned by the withdrawal of support by the shah of Iran and the United States) the Iraqi government embarked on an ambitious program of demographic engineering designed to increase Kirkuk’s Arab population at the expense of Kurds and, to a lesser extent, Turkmens. Known as “Arabization,” the program was multidimensional and included payments of money (10,000 dinars per family) and land to hundreds of thousands of Arabs from southern Iraq to relocate to the region; the forced expulsion of hundreds of thousands of Kurds and Turkmens from the city and its environs; the physical destruction of several Kurdish-inhabited neighborhoods in Kirkuk; redrawing Kirkuk’s boundaries to exclude predominantly Kurdish districts (Chamchamal, Kalar, and Kifri) and include heavily Arab districts; and nationality correction, which required non-Arabs to register as Arabs during the census (Anderson and Stansfield 2009, 30–42; Human Rights Watch 2003). Collectively, these measures ensured that by the 1990s, Arabs appeared to comprise roughly three-quarters of Kirkuk’s total population. Kirkuk had no prior extensive or long-rooted history of ethnic conflict. The tensions of contemporary Kirkuk are mostly attributable to the effects of this systematic, often brutal program of demographic manipulation. The overthrow of Saddam Hussein’s regime in April 2003 bought the issue of Kirkuk to the front burner. From a position of political strength, Kurdish leaders were able to negotiate the terms of Article 58 of the Transitional Administrative Law (TAL). Article 58 required the “normalization” of Kirkuk—in effect a blueprint for the reversal of Arabization. The Arabs introduced into Kirkuk under Ba’athist rule were to return to their place of origin with compensation, displaced Kirkukis (Kurds, Christians, and Turkmens) were to be allowed to return, again, with compensation, a mech-

Kirkuk

367

anism was set up to handle Kirkuk’s myriad property disputes, and a formal procedure was established for restoring gerrymandered boundaries to their pre-Ba’athist status. As part of a compromise deal and under pressure from U.S. authorities, the Kurds agreed to defer a final resolution on the status of all disputed territories, including Kirkuk, until after a permanent constitution had been ratified. In the meantime, the interim Iraqi government was instructed to “act expeditiously” to implement the normalization process. By the time the permanent constitution was drafted in 2006, the government had failed to make any discernable progress on normalization. Hence, Article 140 of the permanent constitution incorporated all parts of Article 58 but made an additional three key stipulations. First, it assigned explicit responsibility to the “executive authority” to implement all parts of Article 58; second, it specified a three-stage process for resolving the status of Kirkuk and other disputed territories (normalization, census, and referendum); and third, it established a deadline of December 31, 2007, for completion of all thee stages. The overwhelming approval of the constitution by popular vote in October 2006, including in Kirkuk province, appeared to signify an important victory for the Kurds on the issue of Kirkuk. Article 140 imposed a constitutional obligation on the Iraqi government to implement all three stages of the process and to do so according to a specified timeline. The failure of the government to fulfill this obligation can, in part, be attributed to the sheer complexity of the process. For example, just to resolve property disputes involved negotiating a labyrinthine legal process that would have taxed the resources of the world’s most efficient judicial systems. According to the June 2004 statute that established property claims commissions throughout Iraq, each individual disputed claim required the preparation of a case report by the relevant Regional Commission summarizing the factual basis of the case, the legal issues involved, the parties’ arguments, and a recommendation made by a legal advisor to the commission. The commission’s subsequent decision could be appealed within sixty days to the Appellate Division located in Baghdad, where the Appellate Division Clerk’s Office was required to prepare another case report to present to the appellate court for final adjudication. The process was thorough and fair but was clearly not equipped to deliver timely decisions on the scale that was needed to implement the normalization stage of Article 140 within the time allotted. Riddled with corruption and mired in civil war from February 2006 onward, the Iraqi federal legal system was not adequate to the task. At the same time, it seems clear that the strength of opposition to implementing

368

Liam Anderson

Article 140 among Turkmens (and Turkey) and most Arab political factions provided few incentives for the beleaguered governments of Ibrahim alJaffari and Nuri al-Maliki to move expeditiously on the Kirkuk issue. Moreover, those seeking to thwart Kurdish ambitions had ample opportunity to disrupt the process. To return Arabs to their governorates of origin, for example, required the active cooperation of bureaucrats at multiple levels of government to change ration cards and identification documents. More often than not, those in charge of implementing these procedures were deeply hostile to the entire 140 process and did everything possible to obstruct progress. For example, in 2007, the minister of the interior and staunch opponent of Article 140 Jawad Bollani issued an order prohibiting the transfer of personal registration documents from one governorate to another. This was in response to the large-scale displacement of citizens due to escalating sectarian violence, but it applied equally to the individuals covered by Article 140. From a Kurdish perspective, therefore, the failure of the government to implement Article 140 was a betrayal that violated both the spirit and letter of the new constitution. In turn, this betrayal heightened Kurdish concerns regarding the overall integrity of the constitution, on which the Kurds would ultimately depend for their future autonomy in Iraq. After all, if today’s relatively weak Arab-dominated government could ride roughshod over a critical constitutional provision, what guarantee was there that a future Arab government would respect those parts of the constitution that safeguarded Kurdish rights and the Kurds’ autonomous status? Hence the issue of Kirkuk acquired a deeper significance for the Kurds as a litmus test of the Arab majority’s willingness to respect constitutional norms and the rule of law.

Core Issues of Concern in Kirkuk

Of the three key elements that constitute the Kirkuk dispute—oil, future administrative status, and future governance—oil is by far the least important and easiest to resolve. Though oil is pivotal to understanding Kirkuk’s historical evolution, its contemporary relevance is marginal. The Kurds have already bargained away any exclusive claim to the revenues accruing from Kirkuk oil fields. Specifically, Article 112 of the Iraqi constitution states that “the federal government, with the producing governorates and regional governments, shall undertake the management of oil and gas ex-

Kirkuk

369

tracted from present fields, provided that it distributes its revenues in a fair manner in proportion to the population distribution in all parts of the country.” The term “present” here is intended to refer to fields already producing oil, which would obviously include Kirkuk. If, as many argue, the Kurds want Kirkuk in order to acquire the resource base necessary to sustain an independent state, then this posture is simply inexplicable. The argument that the Kurds only want Kirkuk because its oil would provide a resource base for an independent Kurdish state does not withstand close scrutiny. The most obvious reason to dismiss a link between Kirkuk’s oil and Kurdish independence is that oil is virtually useless to an independent Kurdistan. The oil could not be exported to generate revenue because Kirkuk is wholly dependent on an oil infrastructure that is controlled by others (namely Turkey and Arab Iraq). Conversely, an independent Kurdish state would obviously sacrifice funding from Baghdad, which currently accounts for over 90 percent of the Kurdistan Regional Government’s budget. Kirkuk’s oil is, therefore, not the issue because it has already been resolved. Of the other two issues, the most straightforward, in principle, is resolving Kirkuk’s future administrative status. Kirkuk is either incorporated into the Kurdistan Region, or it stays part of Arab Iraq. Beyond this basic dichotomy, there are various permutations regarding Kirkuk’s autonomy relative to Irbil (if inside the Kurdistan Region) or Baghdad (if outside), but at heart, the issue is clear-cut, as are the positions of the various parties to the dispute. The Kurds believe that Kirkuk’s population should be given the opportunity (in line with the terms of Article 140) to express a preference over the future territorial status of the governorate. Obviously, underpinning this belief is the assumption that a majority of Kirkukis (especially Kurds and, perhaps, Christians) would vote to join the Kurdistan Region. Despite differences at the margins, most Turkmens and Arabs are unified in their opposition to the Kurds’ efforts to incorporate Kirkuk into the Kurdistan Region. The issue of future governance is more complex. A broad array of powersharing arrangements could, theoretically, be implemented in Kirkuk. The most appropriate institutional design for any given society is a function of multiple variables, including the ethnic composition of the population, the depths of divisions among groups, and the willingness of group leaders to work together in some form of accommodative power-sharing arrangement. A precise figure for the breakdown of Kirkuk’s population along ethnic lines is not available. The 1957 census is considered by all groups to be the most recent reliable record of Kirkuk’s population and, as noted above, this

370

Liam Anderson

put the Kurds on the verge of majority status in the governorate as a whole. Turkmens were the smallest of the three groups in the governorate but a plurality in the city. Subsequent to the 1957 census, the best recent indicator of population was the December 2005 election results. Assuming, as seems reasonable, that the population voted primarily along ethnic lines, the results suggest that by the end of 2005, Kurds comprised some 53 percent of the population, Arabs, approximately 27 percent, and Turkmens, around 13 percent. Hence the available evidence suggests that the Kurds are in the majority in Kirkuk, even without the restoration of Kirkuk’s pre-Ba’athist boundaries or before the changes in the electoral register that have since taken place through the partial implementation of Article 140. Arabs, meanwhile, constitute the second largest group, while Turkmens comprise a relatively small minority. Kirkuk is, therefore, an example of the type of majority-minority situation that most challenges designers of power-sharing arrangements.

Two Approaches to Power Sharing The Integrationist Approach

The two best-known and theoretically coherent bodies of literature on institutional remedies for ameliorating ethnic tensions in divided societies are integrationism and consociationalism. Integrationists advocate the use of institutions to either create the fluid dynamics of pluralism, or to replicate its desirable consequences (Sisk 1996; Horowitz 1985, 1991). The term “pluralism,” much like “federalism,” has been defined in a variety of ways, but in essence all definitions equate the term with the fragmentation and dispersal of political power. All societies are divided to some extent, but the significance of these divisions for political stability varies according to their number, their depth, and the degree to which divisions are cross-cutting or mutually reinforcing. Mutually reinforcing cleavages will tend to produce a small number of segregated, monolithic groups whereas multiple, crosscutting cleavages result in a “fine partition of society into a large number of relatively small preference clusters” (Miller 1983, 735). According to most theorists of pluralism, the former pattern of cleavage distribution is dangerously destabilizing to the political system, while the latter is conducive to stability. A number of explanations for why multiple cross-cutting cleavages

Kirkuk

371

should generate political stability have been advanced by advocates of pluralism. First, pluralism fosters moderation. In a pluralist society, individuals must inevitably interact across groups and interact with different groups at different times depending on the issue at stake. The prospect that today’s enemy could become tomorrow’s ally provides incentives for moderating political interactions. As Bailey puts it, “those who are enemies in one situation are sometimes required to act as allies in another situation. With an eye on future cooperation, they restrain their behavior in present competition” (1970, 129). A second, related explanation is that where there are multiple, cross-cutting lines of conflict, none is likely to develop the depth and intensity sufficient to endanger the stability of the entire system. Conversely, where cleavages are mutually reinforcing, every single issue risks intensifying the divisions that separate groups, and every political battle becomes existential. Pluralism, therefore, helps “de-intensify” existing divisions. The relevance of pluralism to societies divided along ethnic lines is not immediately apparent. Indeed, it is precisely the absence of cross-cutting cleavages that makes ethnically divided societies among the most difficult to govern democratically. Where ethnic divisions are deep, the tendency of other cleavages to reinforce rather than cut across existing ethnic cleavages and the paucity of systematic interaction among members of different groups makes political extremism and the intensification of divisions more plausible than the converse. For integrationists, institutions provide the solution. Specifically, vote-pooling electoral systems such as the alternative vote (AV) that allow voters to express preferences over candidates should encourage those running for office to campaign on moderate platforms that aim to attract rather than alienate second (and third) preference votes from members of other ethnic groups. Assuming that candidates cannot win without votes across ethnic lines, then those who campaign on extremist platforms condemn themselves to probable electoral defeat. AV, in the view of proponents, encourages not just moderate campaign platforms but also preelection “preference-swapping” deals across ethnic groups that may help break down ethnic divisions and foster norms of compromise and consensus. Since AV is a single-member district system, for its moderating effects to function, electoral districts must be drawn to maximize ethnic heterogeneity. Ideally, in no district should a single ethnic group comprise a majority. Alternatively, as some suggest, intraethnic divisions may lead rival politicians of the same ethnicity to compete for the second-preference votes of voters from other ethnicities. This presumably would have a similar moderating effect

372

Liam Anderson

on the platforms of rivals from the same ethnic group. Much the same logic underlies the integrationists’ preferred design of federal system. Integrationists advocate the engineering of federal systems to maximize subunit heterogeneity with the aim of depriving any single group of majority status within a subunit. In turn, this provides political leaders with strong incentives to form coalitions with other groups in order to form the majority necessary to govern. Alternatively, where this is not feasible, they prescribe drawing subunit boundaries to deliberately “crack” large, ethnically homogenous regions into smaller units. The intention of this approach is to use boundary lines to create cleavages that cut across ethnic groups. The focus is on the deliberate fragmentation of monolithic ethnic blocs in order to manufacture potentially competitive subethnic identities. An AV electoral system and some form of territorial federalism are the hallmark institutions of the integrationist approach. Both are driven by the logic of pluralism and both seek to use institutions to create incentive structures for ethnic leaders (and populations) to favor moderation over extremism. Implicit in this approach is a belief in the transformative power of institutions. The goal of institutions is to use short-term (electoral) incentives to transcend ethnic divisions and build nations in the long term.

The Consociational Approach

The consociational approach, closely associated with the work of Arend Lijphart, is rooted in significantly different assumptions about the malleability of ethnic identity and, therefore, the capacity of institutions to transcend ethnic divisions. Originally developed as a critique of pluralism, or at least the claims of its advocates that multiple, cross-cutting cleavages were a societal prerequisite for political stability, the concept of consociational democracy now finds itself dueling for supremacy with pluralism’s institutional progeny. For consociationalists, ethnic divisions reflect real, durable identity differences that cannot simply be engineered out of existence. Hence, where ethnic divisions are deep, it is preferable to segregate ethnic groups rather than attempt to engineer their integration artificially. As Lijphart states, “subcultures with widely divergent outlooks and interests may coexist without necessarily being in conflict; conflict arises only when they are in contact with each” (1969, 219). Hence Lijphart’s contention is that “good social fences may make good political neighbors” and that “a kind of volun-

Kirkuk

373

tary apartheid policy” is the “best solution for a divided society” (219). Underlying this argument is a theory of political conflict that views interaction among heterogeneous groups as likely to increase mutual tension rather than generate norms of compromise and consensus. From the consociationalist perspective, therefore, the preferable approach is to use ethnic groups as the building blocks of a stable political order while ensuring that all groups govern themselves in matters of core concern and that all participate in government. Voluntary segregation and power sharing are the key concepts that define consociational democracy. Institutionally, these concepts are expressed through executive power sharing (coalition governments in which all major groups are represented), group autonomy (invariably, though not inevitably, involving a form of federal system comprising ethnically homogenous subunits), minority veto power over important issues, and proportionality in terms of electoral system and share of administrative appointments. Interethnic interaction occurs only at the elite level in the original consociational vision of things; indeed in early formulations of the concept, Lijphart maintains, “The essential characteristic of consociational democracy is not so much any particular institutional arrangement as the deliberate joint effort by the elites to stabilize the system” (1969, 213). This requires that elites understand the dangers of political fragmentation, that they are committed to preserving the system, that they have the ability to transcend cleavages to cooperate with elites of other subcultures, and that they can adequately represent the interests of their respective subcultures. This necessarily brief outline of two dominant approaches to the problem of democratic governance in ethnically divided societies suggests sharply divergent solutions to the problem of governing Kirkuk.

The 60-40 Problem

The debate over the respective merits and defects of the two approaches has already spawned an impressive body of literature, and it serves little purpose to retread old ground, not all of which is relevant to a substate entity like Kirkuk. A more efficient way to proceed is to focus on the key problem for which neither approach offers a convincing solution—namely, in a majority-minority situation, what incentives does the majority have to share power with other group? Specifically, what incentives do political leaders of

374

Liam Anderson

the majority group have to agree to power-sharing arrangements in the fi rst place and, second, what incentives do they have to make power-sharing arrangements function effectively rather than, say, engage in ethnic outbidding? Though both are important questions to address, the first of these is more relevant in the current context. Except in unusual cases where a society’s ethnic groups are both/all of approximately equal size, power sharing will necessarily involve one ethnic group ceding power to another group or groups. Where no one group constitutes a majority, the incentives issue is a much less acute concern because there may be plausible incentives for even the largest ethnic group to agree to share power with other groups; but under majority-minority conditions, why should a majority group willingly acquiesce in arrangements that intentionally dilute its power? All powersharing theories struggle to answer this question. They are, in Horowitz’s words, “motivationally inadequate” (2002, 20). Though this is often viewed by integrationists as a serious flaw with consociationalism, much the same argument can be leveled at the institutions advocated by integrationists. AV is a majoritarian electoral system that rewards moderation. Theoretically, then, it should produce a legislature dominated by moderates from the majority ethnic group. For minorities, this may be preferable to a parliament dominated by ethnic extremists, but why would the leaders of the majority ethnic group, who are assumed by integrationists to be incapable of moderation absent institutional incentives, voluntarily accept the implementation of an electoral system that rewards moderation? Moreover, for AV to exhibit its moderating influence requires that candidates cannot win election with just the votes of their ethnic cohort. It is the need to appeal to members of other ethnic groups that moderates the positions and behavior of political leaders, and, absent this need, it is unclear why AV would have any moderating influence. In fact, if rival candidates from the majority ethnic group are campaigning for the same ethnic votes, it seems more plausible that AV would lead to the ethnic outbidding it is designed to avoid. The only way to address this problem is through the creative drawing of electoral constituency boundaries such that the majority ethnic group forms a majority in as few constituencies as possible. This will be extremely difficult to achieve in contexts where ethnic groups are geographically concentrated, but even if logistically feasible, it is highly implausible that a majority group would willingly agree to the drawing of boundaries designed to minimize its electoral influence. Equally implausible is that the leaders of a majority ethnic group would agree to territorial federalism, a system that is designed spe-

Kirkuk

375

cifically to break up the power of a dominant ethnic group and exacerbate intraethnic divisions. Unfortunately, integrationists seem unwilling to even recognize this as a problem, let alone offer a coherent response. The case of Nigeria is instructive here because the evolution of its federal system provides powerful empirical evidence against ethnic federalism and in favor of territorial federalism, and because its system of presidential elections requires the winning candidate to obtain a plurality of the votes overall but also at least one-quarter of the votes in two-thirds of Nigeria’s states. This combination of territorial federation and a vote-pooling system for electing presidents makes Nigeria an ideal proving ground for the claims of integrationists. Though not without critics, these institutions appear to have served Nigeria reasonably well, but in the current context, the more relevant concern is how these institutions came to be implemented in the first place. Why would the majority group (in Nigeria’s case, the Hausa Fulani) have consented to the introduction of a federal system that deliberately broke up the large Hausa Fulani–dominated northern region into a larger number of smaller states, thereby fostering intraethnic divisions and diluting the Hausa Fulani influence at the center? Moreover, why accept a presidential election system that effectively gives minority groups a veto over the result? The answer to these questions is instructive. Of the various iterations of new state creation, only that of 1963 (which created the midwest region) was achieved under democratic conditions and according to specified constitutional processes. All other bouts of state creation took place under military governments and were then imposed on the Nigerian people. Predictably, the expansion of 1967 was strongly opposed by the Hausa Fulani–dominated Nigerian People’s Congress, then the dominant political force in the country, because it feared (legitimately, as it transpired) that the division of the north into multiple units would threaten its regional and national hegemony (Olugbemi 1983, 276). The presidential election system was introduced in the 1979 constitution after a sustained period of military dictatorship and, once again, was imposed on the Nigerian people. In short, in the case of Nigeria, the majority group did not consent to the implementation of integrationist institutions; they had to be imposed by a military dictatorship. Dealing with what Horowitz terms the “60-40 problem” is, therefore, a challenge for both consociational and integrationist approaches. The institutions associated with the integrationist approach might be easier for a majority group to swallow because they require the majority to yield less power than under consociational arrangements, but this still does not remove the

376

Liam Anderson

issue of incentives from the equation entirely. Horowitz himself asks the question, “Why should majority-group leaders, with 60 percent support, and the ability to gain all of political power in a majoritarian democracy, be so self-abnegating as to give some of it away to minority-group leaders?” (Horowitz 2002, 20) but seems unwilling to recognize that his institutional remedies are vulnerable to exactly the same argument. This matters in the case of Kirkuk because efforts to devise appropriate political institutions encounter precisely the majority-minority problem to which neither approach offers a convincing solution.

Power Sharing in Kirkuk Kirkuk presents what might most accurately be termed a “53-27-13 problem.” As the majority group, the Kurds can win elections and govern Kirkuk alone, regardless of the electoral system used. The Kurds have a strong geographical, historical, and moral claim to Kirkuk, and Kurdish leaders have variously described Kirkuk as the “heart of Kurdistan” or the “Kurdish Jerusalem.” Why, then, would the Kurds voluntarily share power with Arab and Turkmen political leaders who have dedicated themselves since 2003 to thwarting the Kurds at every turn? In the absence of an external player with the capacity to impose power sharing on a majority group, the prospects for power sharing in majorityminority situations are not good. The architect of consociational democracy, Arend Lijphart, concedes that the majority-minority issue creates problems and that a “multiple balance of power among the subcultures instead of either a dual balance of power or a clear hegemony by one subculture” offers the optimal context for consociational arrangements to take root (1969, 217). It is left to another consociational advocate, Brendan O’Leary, to offer the most convincing explanations for why a majority group might volunteer to share power with a minority group. Though couched as a defense of consociationalism, O’Leary’s arguments apply equally to any power-sharing arrangement that confronts the 60-40 problem. At best, O’Leary claims to offer only a “partial qualification” to Lijphart’s pessimistic assessment of the potential for power sharing in the presence of a majority group (2005, 21). The conditions under which a “hegemonic segment” may indeed have an incentive to share power, according to O’Leary, include when minorities have resources and blackmail potential, when there is a differential population

Kirkuk

377

growth rate among segments such that the majority risks losing this status at some point in the future, and when the majority group feels morally obliged to compensate the minority group for past injustices (2005, 22). The second and third of these explanations are irrelevant in the case of Kirkuk. The majority group has higher birthrates than those of the other groups, and the Kurds are also the historical “victims” in Kirkuk. The first explanation is relevant to Kirkuk but unlikely to provide sufficient incentive for the Kurds to agree to the type of power sharing demanded by the Turkmens and Arabs. According to O’Leary, a minority may have “credible bargaining power” when it enjoys the support of coethnics/linguals in a “neighboring and significant power” or when it comprises an “economically vital minority, with high levels of human capital and the means to emigrate” (2005, 21). Though a mass exodus of qualified Arab and Turkmen administrators and professionals from Kirkuk may create short-term difficulties, particularly in institutions dominated by these two groups, such as the Education Directorate and the Northern Oil Company (NOC), the Kurdistan Region has more than fifteen years of (virtually) independent governing experience and a cadre of experienced officials capable of stepping into the breach. More important, the large-scale outmigration of Arabs and Turkmens would only make Kirkuk more Kurdish, which is precisely why it is unlikely to happen and, if it did, why it would probably be welcomed by most Kurds. In terms of powerful external support, neighboring Turkey and the Arab Iraqi state are obvious potential sources of pressure on the Kurds to share power equitably. It is not clear, however, that these “external” players have a vital interest in how Kirkuk is governed, still less that they could do anything meaningful to affect it. Moreover, it is difficult to see why the Turks and Arab Iraqis have any more right to push for power sharing for their coethnics in Kirkuk than the Iraqi Kurds have to call for equal power-sharing arrangements in Ankara for Turkey’s Kurds or to demand an equal share of power in Baghdad. The absence of compelling incentives for the Kurds to share power in Kirkuk is reinforced by two compelling incentives not to share power. First, if Article 140 is not implemented in full, the Kurds will justifiably feel that they have been cheated of their chance to reclaim Kirkuk. After securing the inclusion of Article 140 in a document that was democratically endorsed by an overwhelming majority of the Iraqi population in a free and fair vote, the Kurds can be forgiven for believing that they had at least earned the right to take the issue to a vote. Denied this, there is absolutely no reason to suppose

378

Liam Anderson

that the Kurds would, or should, look favorably on the idea of sharing power with precisely those groups that have battled the hardest to prevent the implementation of Article 140. Second, the only type of power-sharing model that seems to be acceptable to Kirkuk’s other groups is at the extreme end of the spectrum. There is an approximate form of PR power sharing in place in Kirkuk already, and this has been rejected by other groups as inadequate. The most coherent replacement power-sharing scheme advanced by the Iraqi Turkmen Front envisages a 32-32-32 split in power positions among the three major ethnic groups, with the Christians receiving the remaining 4 percent share. Hence, the Kurds are expected to give up incorporating Kirkuk into the Kurdistan Region and not just to share power but to share power on the basis of parity. Equal power sharing may be preferable to the alternatives under certain conditions, but it is inefficient in terms of legislative output, and it is certainly not a “fair” democratic system in that the allocation of power bears no relation to relative group size. Moreover, if, as Arab and Turkmen leaders demand, seats on the governorate council are also allocated according to the equal power formula, the Kurds can actually be outvoted by a coalition of the other two groups. With no legal, constitutional, or moral obligation to share power with other groups, the Kurds cannot realistically be expected to adopt voluntarily a form of power sharing that significantly diminishes their power and that is neither fair nor efficient.

Cutting a Deal To summarize briefly, Kirkuk is an example of the majority-minority situation that neither integrationists nor consociationalists can comfortably accommodate within their frames of reference. Both fail to offer convincing explanations for why a majority group would voluntarily acquiesce in the implementation of institutional arrangements designed to reduce its own power for the benefit of other groups. This problem is even more acute in the case of Kirkuk because Arabs and Turkmen political leaders demand an extreme form of power sharing that would require the Kurds to share power equally with other groups. At the same time, a durable and stable resolution to the Kirkuk issue would seem to require institutional arrangements that allow all groups to participate to some degree in its future governance. Ironically, the motivational inadequacy of the two approaches to deal with

Kirkuk

379

the majority-minority problem yields the critical insight necessary to cut a compromise deal on Kirkuk. If the Kurds currently lack an incentive to share power with other groups, then the obvious solution is to provide them with one. Though seemingly obvious, this solution has yet to be grasped by the Iraqi government, Arab and Turkmen political leaders in Kirkuk, and even many Western observers (O’Hanlon and Taspinar 2008, A15). A series of reports by the influential International Crisis Group (ICG) provides a typical example here. Initially, the ICG advocated keeping Kirkuk outside the Kurdistan Region as a special status region for an interim period of ten years, with governing power in the meantime to be shared among all ethnic groups in the governorate (ICG 2006). This superficially plausible solution falls apart on deeper analysis precisely because it fails to furnish the Kurds with an incentive to agree. In return for sharing power, the trade-off for the Kurds appears to be that the administrative status is not resolved immediately to their detriment. But ten years from now, the prospect of the Kurds being in a position to force the issue of Kirkuk to a favorable resolution is negligible. Hence this proposed deal (effectively) requires the Kurds to sacrifice permanently the incorporation of Kirkuk into the Kurdistan Region and to share power with other groups. The ICG’s most recent report appears to recognize the incentive problem because it advocates an “oil-for-soil” deal that would allow the Kurds to manage oil fields within the Kurdistan Region in return for allowing the UN to find a solution to the problem of disputed territories (ICG 2008). Pending a resolution, power is to be shared on the basis of parity in Kirkuk. To the extent that this deal at least recognizes the need to offer the Kurds something in return for sharing power and giving up Kirkuk, it is a step forward from previous reports but only marginally. By any reasonable reading of the constitution, the Kurdistan Region already has the power to manage its own oil and gas fields, so this deal gives them nothing they do not already have. In return, the Kurds are required to share power equally and accept an open-ended UN-led process of arbitration that may or may not lead to a favorable outcome. This is not a deal the Kurds can or should be expected to accept. Of the various Turkmen proposals for Kirkuk, the most coherent would keep Kirkuk outside the Kurdistan Region as a freestanding autonomous region and would require power to be shared equally among the three major groups. Kirkuki Arabs, meanwhile, have proposed an arrangement whereby Kirkuk remains under the administrative control of Baghdad, administrative positions and seats on the governing council are allocated equally among

380

Liam Anderson

the three main groups, and all groups share key executive positions on a rotating basis. Common to all these proposals is the assumption that the Kurds must accept parity power-sharing arrangements without being offered anything in return. Though couched as “compromise solutions,” they are, in reality, nothing of the sort. Using a standard definition of “compromise” as the “settlement of a dispute by concessions on both or all sides,” it is self-evident that a solution that requires one side to make all the concessions and the other side to make none does not meet the definition of compromise. A true compromise must include incentives for the Kurds to share power, and the one incentive likely to prove acceptable to the Kurds is a deal whereby they share governing power with other groups in return for Kirkuk’s incorporation into the Kurdistan Region. The precise details of a compromise deal are less important than the broad outlines. First, Kirkuk in its pre-1968 borders would be recognized as falling within the official boundary of the Kurdistan Region, subject to approval in a referendum. Second, Kirkuk would be granted roughly the same level of autonomy with respect to the Kurdistan Regional Government as the Kurdistan Region enjoys relative to Baghdad. This means, for example, that Kirkuk would have its own constitution (or statute of autonomy) that could not be unilaterally amended or revoked by Baghdad or Irbil and control over its own security forces. Third, Kirkuk would receive its budget directly and automatically from the government in Irbil, and a portion of this would be distributed, again automatically, among Kirkuk’s various ethnic groups according to a formula based on either proportionality or equality. Fourth, administrative posts and positions in the security forces at all levels would be allocated equally among groups. Fift h, groups would be granted autonomy over important symbolic and cultural issues, such as religion and education. Finally, the passage of important legislation would require a supermajority mechanism to prevent the Kurds from passing laws against the wishes of all other groups. There are various ways to achieve this; establishing an executive council with representation for each group wherein each group possesses veto power is one obvious way. Requiring a two-thirds majority to overturn an executive veto would then require at least one other group to side with the Kurds in order to pass legislation. The broad principles, therefore, are that Kirkuk is officially part of the Kurdistan Region; that it has a high level of guaranteed autonomy from other levels of government; that its autonomous status is guaranteed in the Kurdistan constitution; and

Kirkuk

381

that power and resources are shared through a mix of proportionality and equality. The result is a compromise that gives the Kurds “ownership” of Kirkuk in the sense that it is officially recognized as territory of the Kurdistan Region but denies them full political control; in this sense, Kirkuk belongs to no one group. This is not an elegant outcome. Problems such as how to identify groups, which groups are significant enough to warrant treatment on the basis of equality, and the propensity of such an arrangement to produce political gridlock mean that this option is inevitably “messy.” However, as part of a package that definitively settles the questions of Kirkuk’s official administrative status, and who gets what, when, and how, the most contentious and intransigent issues associated with Kirkuk are resolved. With the resolution of these issues, most of the sources of ethnic tensions in Kirkuk are eliminated.

Conclusion: An Absence of Alternatives Power-sharing arrangements are invariably “messy” solutions to the problem of governing deeply divided societies. Their major selling point, however, is that they are often less bad than any plausible alternative. In the case of Kirkuk, the most plausible alternative is, unfortunately, a solution imposed by one side on the other. A non-negotiated solution imposed by the Iraqi government on the Kurds would raise important questions about the wisdom of future Kurdish participation in the state of Iraq. To date, Kurdish leaders have pursued a political strategy—that of securing maximum autonomy but within the borders of Iraq—that has put them at odds with the majority of their own population. Constitutional guarantees were viewed as a safeguard against the resurgence of a powerful Arab-dominated government at the center, and Article 140 was, in many ways, the litmus test of this strategy. The fate of Article 140 demonstrates that the integrity of the constitution is questionable and that, realistically, the “guarantees” embodied in its provisions are somewhat less than guaranteed. If an adverse resolution to the Kirkuk issue is simply imposed on the Kurds, it is difficult to see how Kurdish leaders could continue to pursue an obviously failed strategy. One plausible outcome, therefore, is that the Kurds would unilaterally declare the border of the Kurdistan Region, to include certain of the socalled disputed territories and all, or parts, of Kirkuk governorate. Clearly this would not be positive for the future stability of Iraq. The broader point

382

Liam Anderson

is that it is in the interests of those who oppose the Kurds on Kirkuk to fi nd a compromise position that Kurdish leaders can accept and plausibly sell to their population. Kirkuk’s Turkmen and Arab political leaders, as well as many in the West, appear to believe that a solution that requires the Kurds to yield on everything offers a viable alternative. They are mistaken. The most plausible alternative to compromise on Kirkuk is armed conflict and, ultimately, the breakup of the Iraqi state. Notes 1. Kirkuk’s ever-dwindling population of Christians (Assyrian and Chaldean) is often included as a fourth ethnic group when experts discuss Kirkuk. For current purposes, however, the Christian population is simply too small to be considered politically relevant in the determination of Kirkuk’s future status. 2. As part of a compromise deal, elections were held in all governorates except Kirkuk (and the Kurdistan Region) in January 2009. At the time of writing, the date of the Kirkuk governorate election has yet to be finalized. 3. The census data from 1957 indicate that of Kirkuk city’s total population of 120,000, just over 45,000 were Turkish speakers, 40,000 were Kurdish, and 27,000 were Arabic speakers. In the province as a whole, the Kurds composed over 48 percent of the population, Arabs 28 percent, and Turkmens 21 percent. 4. Republican decree 608 of December 1975 formally detached Chamchamal and Kalar from Kirkuk and reattached them to Sulaimaniya, while Kifri was reapportioned to the governorate of Diyala. Republican decree 41 of January 1976 detached the mainly Turkmen Tuz Khurmato and Qadir Karam from Kirkuk, reapportioning them to Salahadin governorate. 5. For example, Republican decree 33 of December 1976 saw the subdistricts of Kandenawa and Qaraj detached from the Makhmour district of Erbil governorate and reapportioned to the Arab-dominated Dibis district of Kirkuk. Later the Mosul governorate district of Zab was formally annexed into the Hawija district of Kirkuk. Further additions occurred in the 1980s when Arab-dominated subdistricts of Mosul were attached to Kirkuk, creating a clear Arab-majority subregion in the west of the governorate linking directly into Arab-majority areas in Mosul and Salahadin and Diyala governorates. 6. Though censuses conducted after the advent of the Ba’ath regime in 1968 were notoriously unreliable, the available data indicate that by 1997 Arabs composed over 72 percent of the population of Kirkuk governorate. Kurds, meanwhile, composed approximately 21 percent and Turkmens 7 percent. 7. Article 58 instructed the Iraqi Transitional Government to “act in accordance with Article 10 of the Iraqi Property Claims Commission statute to ensure that such individuals may be resettled, may receive compensation from the state, may receive new land from the state near their residence in the governorate from which they came,

Kirkuk

383

or may receive compensation for the cost of moving to such areas.” With regard to boundary changes, Article 58 stated, “The Presidency Council of the Iraqi Transitional Government shall make recommendations to the National Assembly on remedying these unjust changes in the permanent constitution. In the event the Presidency Council is unable to agree unanimously on a set of recommendations, it shall unanimously appoint a neutral arbitrator to examine the issue and make recommendations. In the event the Presidency Council is unable to agree on an arbitrator, it shall request the Secretary General of the United Nations to appoint a distinguished international person to be the arbitrator” (Governing Council of Iraq 2004). 8. Though most media sources refer to a singular Kirkuk oil field, the reality is that Kirkuk oil is located in a chain of fields, some of which are located in Erbil governorate and, hence, fall within the existing boundary of the Kurdistan Region. 9. That the voters of Kirkuk voted primarily along ethnic lines cannot, of course, be proven. Nonetheless, there is no reason a priori to assume that Kirkukis behaved any differently from the vast majority of Iraqis who voted their ethnicity or sect in the December 2005 election. The percentages provided here are approximations based on the votes polled by obviously ethnic parties that participated in the December 2005 elections. 10. Many advocates of this perspective refer to themselves (and their preferred institutions) as centripetalists (centripetal). In line with Sisk, I use the term “integrationism” to describe the broad approach that includes centripetal institutions. 11. Logistically, the mechanisms associated with the integrationist approach, which is designed to operate at the nation-state level, would present serious though not necessarily insurmountable challenges to implement at the substate level. To implement AV in Kirkuk, for example, would require the governorate to be divided into equally sized electoral districts, which, at a minimum, would require an accurate census to be conducted. 12. In Nigeria, the serious opposition to the initial redesign of the federation was notably absent during subsequent redesigns; indeed, with the creation of an upper House (Senate) with equal state representation under the Second Republic and the introduction of a formula for resource distribution that stressed equality among states, the problem became how to satisfy demands for new states from all parties. 13. The percentage of Kurds is, by now, almost certainly higher than 53 percent, given that displaced Kirkuki Kurds have continued to return to Kirkuk under the provisions of Article 140 since the December 2005 election. 14. This proposal was presented in the form of a handout, “Kirkuk. . . . The Alternative Soluation [sic],” distributed by the head of the Iraqi Arab Kirkuk Front (IAKF), Akram al Obady, at the conference “Kerkuk Problem and Article 140: Defining Alternatives,” Brussels, 22 June 2008. 15. When questioned by the author at the Brussels conference as to why he thought the Kurds would be willing to accept the IAKF’s proposal, the IAKF representative suggested that it offered the Kurds “30 percent power, whereas now they have nothing.”

384

Liam Anderson

16. At the time of the January 2005 elections, for example, voters in the Kurdistan Region were also asked to participate in a nonbinding referendum on independence from Iraq. Over 95 percent of the population voted in favor. 17. Many of the disputed territories, Sinjar, Makhmur, and Aqra, for example, are already under the de facto control of the KRG and Kurdish security forces. It is also plausible that the Kurds would take the opportunity to include Mosul’s large Kurdish population within the expanded boundary of the Kurdistan Region. Th is would inevitably provoke a serious confrontation with the city’s Sunni Arab majority, but the alternative—leaving a sizable population of Kurds beyond the protective cover of Kurdish security forces—may justify the risk in the eyes of Kurdish leaders. 18. Obviously, this is only one possible response to an imposed solution, but the broader point is that it is unthinkable that Kurdish leaders would passively consent to an imposed solution on Kirkuk and the disputed territories. References Anderson, L., and G. Stansfield. 2009. Crisis in Kirkuk: The Ethnopolitics of Conflict and Compromise. Philadelphia: University of Pennsylvania Press. Bailey, F. G. 1970. Strategems and Spoils: A Social Anthropology of Politics. Oxford: Basil Blackwell. Governing Council of Iraq. 2004. Law of the Administration for the State of Iraq for the Transitional Period (Transitional Administrative Law [TAL]). Baghdad Governing Council of Iraq and Coalition Provisional Authority. http://www.cpa-Iraq .org /government/TAL .html. Horowitz, D. L. 1985. Ethnic Groups in Conflict. Berkeley: University of California Press. ———. 1991. A Democratic South Africa? Constitutional Engineering in a Divided Society. Berkeley: University of California Press. ———. 2002. “Constitutional Design: Proposals versus Processes.” In A. Reynolds, ed., The Architecture of Democracy: Constitutional Design, Conflict Management, and Democracy. Oxford: Oxford University Press. Human Rights Watch. 2003. “Iraq: Forcible Expulsion of Ethnic Minorities.” Human Rights Watch Report 15.3 (E). Washington, DC. International Crisis Group (ICG). 2006. “Iraq and the Kurds: The Brewing Battle over Kirkuk.” Middle East Report 56, 18 July. ———. 2008. “Oil for Soil: Toward a Grand Bargain on Iraq and the Kurds.” Middle East Report 80, 28 October. Lijphart, A. 1968. The Politics of Accommodation: Pluralism and Democracy in the Netherlands. Berkeley: University of California Press. ———. 1969. “Consociational Democracy.” World Politics 21: 207–25. Miller, N. R. 1983. “Pluralism and Social Choice.” American Political Science Review 77, 32: 734– 47. O’Hanlon, M., and O. Taspinar. 2008. “Time for Kurdish Realism.” Washington Post, 9 February.

Kirkuk

385

O’Leary, B. 2005. “Debating Consociational Politics: Normative and Explanatory Arguments.” In S. Noel, ed., From Power Sharing to Democracy: Post-Conflict Institutions in Ethnically Divided Societies. Montreal: McGill-Queen’s University Press. Olugbemi, S. O. 1983. “The Ethnic Numbers Game in Interelite Competition for Political Hegemony in Nigeria.” In W. C. McCready, ed., Culture, Ethnicity, and Identity: Current Issues in Research. New York: Academic Press. Sisk, T. D. 1996. Power Sharing and International Mediation in Ethnic Conflicts. Washington, DC: USIP.

CHAPTER 16

Power Sharing: An Advocate’s Conclusion Brendan O’Leary

Political science is the science not only of what is, but of what ought to be. . . . Utopia and reality are . . . the two facets of political science. Sound political thought and sound political life will be found only where both have their place. —Edward Carr, The Twenty Years’ Crisis

Leading analysts of power sharing have presented some of their recent and current research in this volume. They were not requested to provide comparative and empirical case studies directly to test commonly held hypotheses about power sharing; and they were not asked to supply case studies to illuminate mechanisms and processes that explain the correlations found in large-N studies. No shared research design was attempted here in which power-sharing successes and failures across the world were to be systematically compared to draw explanatory and policy conclusions from a representative sample. Such projects have begun, with varying degrees of success and rigor. Efforts to continue and refine them are entirely to be welcomed, and further comment on such projects occurs below. The chapters in this volume are intended, however, to illuminate the current state and range of scholarship on power sharing across multiple disciplines. Rather than summarize each of the chapters, a task already performed in the introduction and accomplished more extensively by the authors themselves, this conclusion focuses on general questions raised by the contributors’ chapters and

Conclusion

387

indicates future research avenues, especially among political scientists. Lastly, some normative questions are addressed. The jilted woman in “The Winner Takes It All,” sung by the Swedish pop group Abba, suggests she is wrong to lament her fate too long: only one lover can have her former partner; fate dictates that she has to fall while her rival stands tall. Her folk wisdom may be true of love between persons, at least in cultures that reject polygamy or polyandry, but does it have to be true in politics? Must the winner take it all?

Power Sharing and Proportional Representation Part I of this book focuses on the role of electoral systems in promoting or inhibiting interethnic accommodation. It is widely agreed among political scientists, whether advocates of power sharing or otherwise, that power sharing is more likely to be encouraged if proportional representation (PR) systems are used to elect politicians. There is, however, a contrary claim, advanced by “centripetalists,” notably Donald Horowitz, who argue that power sharing may be best advanced through the so-called alternative vote (AV), more accurately known as the “majority-preferential vote.” This claim is skeptically scrutinized by Allison McCulloch (Chapter 3), who expresses the more common view within political science that the use of AV does not promote power sharing because the majority-preferential vote is a majoritarian system and as such is not conducive toward power sharing as standardly conceived. One response maintains that Arend Lijphart and his followers should not monopolize the use of the expression “power sharing” (Horowitz 2002). That may be so, but centripetalists’ conception of power sharing is idiosyncratic and not just distinct from that of consociationalists and multinational federalists. As standardly conceived, power sharing is directly inclusive, that is, the key partners to a power-sharing system, including minorities, represent themselves and jointly make key decisions in the polity. By contrast, centripetalists prefer a pattern of accommodation in which candidates and parties from ethnic majorities, at all tiers of government, are incentivized to accommodate minorities through obtaining lower-order preferences on the ballot papers of voters from minority backgrounds. In short, they prioritize better behavior by representatives from majorities rather than the direct election of minorities. Centripetalists place no strong value on minorities representing themselves and often emphasize a common and undifferentiated conception of citizenship. Regularly the metaphor of “pooling votes” is

388

Brendan O’Leary

used as part of the advocacy of AV, which suggests “sharing.” Yet whether the lower-order preferences on ballot papers cast with first preferences for minorities constitute sharing may reasonably be questioned. They permit voters to rank unpalatable alternatives, not quite the everyday meaning of “pooling” or “sharing.” To avoid profitless dispute over the meaning of power sharing, it may be said that PR systems facilitate the forms of power sharing in which ethnic groups are free to represent themselves, if they wish to do so, through their own parties and candidates. They do so for straightforward reasons that do not need confirmation in statistical tests. PR systems have lower barriers to entry for parties or groups and make it less likely that any one party or group will command a parliamentary or assembly majority. They thereby render more probable the formation of coalition governments (though they do not guarantee this outcome, as the domination of post-apartheid South Africa by the African National Congress demonstrates). By contrast, nonPR voting systems, including the majority-preferential vote, usually underrepresent minorities in assemblies and parliaments, especially when the minorities are territorially dispersed. Non-PR systems disproportionally benefit plurality “winners,” often converting a less than 50 percent vote share obtained by one party in an election into a governing parliamentary or assembly majority. They are systematically disproportional at the district or constituency level. Indeed, in political systems saturated with majoritarian thought, such as France, the winner of the most votes is described as the holder of a “relative” majority, even though the candidate is merely the leading candidate. Advocates of power sharing, where it is necessary, are therefore usually advocates of PR systems, though they are aware, as Bernard Grofman (Chapter 2) observes, that under certain territorial (“districting”) configurations of the electorate, non-PR systems may facilitate minority representation and indeed their overrepresentation. In the introductory chapter the broadest definition of power sharing proposed was given as “Any set of arrangements that prevent one agent, or organized collective agency, from being the ‘winner who holds all critical power,’ whether temporarily or permanently.” Under AV (or its variants, such as the supplementary vote or the limited preferential vote) a majority party (or group) may well hold all legislative power for an electoral term and may permanently exclude ethnic minorities from self-representation. AV does not just fail to prevent this possibility; it may encourage it. PR systems,

Conclusion

389

provided thresholds and districts are designed appropriately, prevent the permanent or temporary exclusion of significant minorities from access to the legislature under their own banners. The only electoral system that absolutely guarantees no monopoly of legislative power, either temporary or permanent, may be called “every party gets the same number of seats” (Taagepera and Shugart 1989, 34). This hypothetical system would be entirely uncompetitive and would favor parity at the complete expense of proportionality. For the same reason that advocates of power sharing prefer PR to nonPR systems, they prefer multiperson to single-person executives. The former facilitate power sharing and minority self-representation, whereas the latter can only accommodate those different from themselves (unless they happen to represent all relevant ethnicities in their family history). Advocates of power sharing need not, however, be mission committed to parliamentary (or cabinet) systems of government rather than presidential systems. That is because collective presidencies are a viable institutional form, as demonstrated by the history of the Swiss Federal Executive Council and by the European Union and its precursors (O’Leary 2003; Schneckener 2002). Readers should, however, beware the temptation to classify advocates of power sharing as blindly antimajoritarian or as naive enthusiasts of maximum feasible consensus in decision making. Power sharing’s advocates are usually democrats, but they are as sensitive to the nature of the demos or demoi as they are to the form of decision-making rule employed. They properly observe that precisely because PR election systems make coalition governments more likely, they make it more probable that such coalitions will consist of parties that between them have a mandate from a majority of those who recently cast their votes. Such a mandate is made more legitimate when a program of government is negotiated, before or after the election, based on the public manifesto commitments of the coalition parties (Nagel 2000). Usually such coalitions allocate cabinet portfolios in proportion to each party’s share of membership of the assembly or parliament. In short, advocates of power sharing insist that PR election systems strongly increase the probability that any government, single-party or multiparty, elected under its auspices will be, in some sense, mandated by a majority of votes and include in the government those for whom individuals voted as their first choice. In short, there need be no absolute clash between majority rule and advocacy of PR systems.

390

Brendan O’Leary

Proportionality and Parity Principles May Coexist But May Also Clash Power sharing based on coalition governments that enjoy majority public support among individual voters within a demos should not be controversial among democrats, provided that their conduct toward those outside the coalition is not racially, nationally, religiously, or ethnically discriminatory. Heat arises, however, when advocates of power sharing insist that the relevant political system contains more than one people and that in these circumstances an undifferentiated model of equal or difference-blind individual citizenship is inappropriate. Advocates of power sharing may agree that majority rule is a good democratic decision-making rule where there is a unified demos, provided that the decision rule really means majority rule rather than plurality or factional rule. Yet where there are multiple peoples in deeply divided places they insist that majority rule is not appropriate. Such claims challenge the merits of institutions based on treating the relevant unit of voting power as one voter with a vote of equal value in the political system as whole. If the premise is granted that there are multiple peoples in the political system, then the inference may be drawn that there is a case for parity in voting (or decision-making) power among the relevant peoples, at least on some matters, just as each independent government may count for one in the decisions of some international organizations. The principle of parity and the principle of proportionality imply roughly the same outcome for representation, or for decision making, only when the relevant peoples have roughly the same number of valid voters. Imagine two key peoples in a polity are roughly balanced in size. For example, people A comprises 47 percent of the electorate, and people B comprises 43 percent, whereas others (neither As nor Bs) comprise 10 percent. Imagine further that all voters in each category vote for one party of their category under pure PR. In this example, achieving both proportionality, according to equality among individual voters, and parity between the peoples A and B seems to be not too difficult. In this case, the use of PR to elect legislators, and of a PR system to determine the executive, can be combined with a concurrent majority decision-making rule among the two peoples, A and B, over key matters. Such concurrent majority decision making can be achieved explicitly through corporate naming (or “designation”) of the peoples, for example, “A majority of those deputies who represent people A and a majority of those deputies who represent people B as well as a majority in the

Conclusion

391

parliament shall agree before legislation is passed regulating any aspect of policing.” This rule, however, has consequences for the voting power of the others (the neither As nor Bs) who may not be pivotal. A concurrent majority requirement may also be accomplished implicitly through a differenceblind qualified majority decision-making rule, for example, “No legislation regarding policing may pass without the consent of two-thirds of all elected representatives.” This rule would seem to make it difficult to pass legislation against the preferences of a majority in group A or B, though it would treat the non-As and non-Bs equally with the As and Bs. Note, however, that this difference-blind rule does not provide the same measure of reassurance to the larger peoples as does the corporate naming rule. Take B. In a 100-seat assembly with 47 members being in group A, 43 in group B, and 10 among the others, a two-thirds majority (67) can be formed (47 As + 10 Bs + 10 others) against the wishes of a large majority of the B group (33 out of 43). For this reason, even large minorities may prefer designated or corporate forms of decision making to difference-blind rules. This example illustrates why advocates of power sharing in deeply divided places usually favor laws being made by concurrent majorities among the key peoples who make up the relevant polity, provided that they are working within a viable and just settlement. When that standard cannot be met, advocates of power sharing may settle for the use of concurrent majorities (“co-decision making”) for key matters, such as constitutional change, provided that the initial constitution is judged fair. These arguments, for either ubiquitous or partial concurrent majoritarianism, become more controversial when one people is significantly larger than another. Imagine, for example, that people C comprises 75 percent of the electorate, people D comprises 20 percent, and 5 percent comprises the others (neither Cs nor Ds), and imagine again that all voters in each category vote for one party of their category under pure PR. In this scenario it is far more difficult to combine parity and proportionality principles without sharp tension. In a 100-seat assembly with 75 members in group C, 20 in group D, and 5 among the others, then under a concurrent majority rule giving parity to C and D would mean that a blocking majority among D would comprise just 11 assembly members. In the real world, when a people’s position resembles that of D but is sufficiently powerful in its resources, its representatives will insist on parity in the executive and a full veto right over all executive and legislative decision making. Otherwise the representatives of people D face the vista of being constantly outvoted by representatives of people C, which could easily

392

Brendan O’Leary

happen, even under PR, and even under a difference-blind qualified majority rule for the passage of legislation set at 75 percent of the legislature or executive. Ceding parity for decisions is the most difficult power-sharing principle to accept for the representatives and negotiators of a people with a large or significant majority. This example is not abstract. Where the ratios of people C to people D are roughly 4 to 1, people C resembles the position of Greek Cypriots confronting Turkish Cypriots, the Hutus in Rwanda and Burundi confronting Tutsis, the Sinhalese confronting Tamils in Sri Lanka, and Anglophone Canada confronting the Quebecois, to name some obvious examples of largely dualistic antagonisms. When ceding full parity to another people is very difficult for the majority group, autonomy may become even more significant as a means of building a stable power-sharing system. Maximizing the domains in which the majority and minorities exercise cultural or territorial self-government and reducing the importance and the number of functions the central or federal government monopolizes may ease the potential clashes that would otherwise occur between the principles of parity and proportionality. Significant symmetrical powers of autonomy are of course one version of parity: each autonomous region or community has the same powers of self-government. Granting increased (and asymmetrical) powers of autonomy to a minority group that dominates a given region or to a minority community may compensate the minority for lack of full parity in the central or federal government. In international organizations and confederations the normal decisionmaking rule has historically been parity: one member-state, one vote, and one veto. Moves away from this principle toward giving greater weight to the size of member-states’ populations usually have had to be justified, for example, as reflecting the different resource contributions of member-states in revenues or in soldiers supplied to the joint institutions. Granting bigger vote shares to bigger contributors recognizes proportionality. Decision making in such an alliance or confederation, however, is usually based on qualified majority rules. That is because there are insufficiently shared interests to permit simple majority voting. The twenty-eight-member European Union (EU), as it will be beginning January 1, 2013, is perhaps the most institutionally complex confederation in history. Its treatment of votes and peoples is topically instructive for advocates and critics of power sharing. The idea of directly electing a powerful executive president of the EU by its citizens has so far had neither major

Conclusion

393

proposers of standing nor any serious prospect of success, precisely because there has been no profound and shared sense of a common demos to underpin the legitimacy of such a powerful office. The same lack of support has applied to proposals that the European Parliament, elected under PR, should elect a powerful executive. Currently only a collegial executive, significantly controlled by the executives of the member-states, appears consonant with the Union’s multipeople identity. The current executive has three faces: the European Council (the Council), comprising each of Europe’s heads of government (in parliamentary systems) and heads of state (where there is an executive president); the European College of Commissioners (the Commission), each of whom is nominated by a member-state and approved by the European Parliament; and the Council of Ministers, which consists of the ministers of Europe’s member-state governments making decisions together in those functions where the EU has competence. These three collegial executive institutions, which have additional legislative roles in the case of the Council of Ministers and the Commission, reflect the difficulties in applying majority-rule decision making within the European Union. The Council and the Council of Ministers illustrate the principle of parity. The Commission currently reflects the principle of parity (one commissioner per member-state), but in the past it reflected the principle of proportionality, albeit approximately (large states nominated two commissioners, whereas small states nominated one). The decision-making rules of the Council of Ministers reflect parity (unanimity) and proportionality (the precise rules actually underweight the comparative population of the largest member-state, Germany) and require qualified concurrent majorities of the total EU population and of its member-states. So far all efforts to create a majoritarian executive in the European Union (directly or indirectly elected) or simple majority decision-making rules within it have proved fruitless. That the Parliament now approves the choice of President of the Commission and allocation of functional portfolios to commissioners by the President of the Commission are not significant exceptions to this summary because the Council nominates the President and the commissioners. Proposals to reduce the number of commissioners, in the interests of efficiency, so that the small states would rotate nominations to incumbency of such posts, recently went down to defeat, triggered by Irish citizens’ decision to reject the first draft of the Lisbon Treaty in a referendum. In short, power sharing and consensual, concurrent, or qualified majority voting, far from being abnormal, are exactly what one should expect when

394

Brendan O’Leary

multiple peoples negotiate their coexistence and interdependence. If this is so for the European Union, which is not deeply divided in the sense of being on the verge of war, then one must expect similar institutional arrangements to emerge where multiple claims to peoplehood exist in deeply divided places. Europeans, for good or ill, have delegated significant powers to the institutions that govern them under the treaties that bind them, particularly to the Commission, to the European Court of Justice, and, in the case of members of the Euro-zone, to the European Central Bank. This delegation has been to “nonmajoritarian institutions,” that is, not to democratically elected power-sharing institutions where federal or consociational representation operates with qualified majority voting. Rather, these nonmajoritarian bodies are expected to operate technocratically, not politically, and according to rules construed to be in the general interests of all, rules that are suggested to have no directly redistributive or political consequences. The implementation of decisions by such bodies, with the major exception of some functions of the European Central Bank, rests with the member-states. These features of European decision making show how the principle of autonomy is often drawn upon to relieve the pressure created by unwelcome powersharing decision making at the center. They also suggest an underexplored feature of research on power sharing in deeply divided places: parties’ willingness to delegate technical decision making to professional bodies in the hope that they may reduce political antagonisms. The major point is that the world’s largest multipeople political system, the European Union, has been quite unable to function without power-sharing arrangements that recognize its multipeople character. The EU’s lack of democratic character in its key nonmajoritarian institutions may legitimately be questioned, but equally we may doubt whether it would be possible to run EU institutions with majoritarian or winner-takes-all arrangements. The formal existence but rare use of qualified majority decision-making rules in European lawmaking reinforces this judgment. A qualified majority in the Council of Ministers in favor of a measure may not fi nd a qualified majority in the European Parliament to support it, not least because the two bodies are composed of people elected in different election cycles. When there is support in both bodies for a measure then the likelihood is that the implicit qualified majority support across the EU is extremely high. Yet even so, such measures are still subject to a possible member-state veto, whatever the Union’s lawyers may say, including the type of veto involved in reluctant compliance or outright nonimplementation.

Conclusion

395

The European Union has been used to illustrate how voting rules and institutions may be intimately related in a power-sharing system, one that may combine diverse versions of proportionality, autonomy, veto rights, and parity. The EU, so far, is not a deeply divided place, though because one of its official justifications is to keep the peace in Europe its evolution is pertinent to the agenda of this book. The EU and its predecessors, the European Community and the European Economic Community, have had many policy successes but also distinct failures. European integration has contributed to peace in Europe since 1958, though its salience in this regard may be less than suggested by Euro-enthusiasts. It has generated a free-trade zone, and then a common market in goods, ser vices, and labor. Its foreign policy, not least in the Balkans, has, however, been a site of frequent coordination failures. The crisis of the euro, one of the key initiatives at the founding of the Union under the Maastricht Treaty, is one of the obvious failures. If the euro crisis is to be resolved, that is, if “the currency without a country” is to be given a viable political roof, it will require the Euro-zone’s peoples to agree a power-sharing fiscal union, in which each of the member-states will have parity in rules over revenue raising as well as some type of proportionate share of mutualized debts and assets. Whether that can be accomplished remains an open question. These remarks about the European Union and the Euro-zone remind us of what all contributors to this book have emphasized: power sharing is neither a panacea nor easy. In discussing electoral system design or qualified majority lawmaking rules it is necessary to address a foundational question, namely, whether the goal is to make many peoples into one or simply to enable many peoples to cooperate. The former goal is integrationist; the latter is accommodationist. These goals may or may not be contradictory. In the short run, fast-paced integration may damage accommodation and provoke conflict, which implies contradiction. In the long run, sustained accommodation may lead to voluntary integration. Honest and evidence-based electoral systems and lawmaking advice by political scientists to policymakers has to be clear. Will system X enhance accommodation or integration? Will it provoke conflict and disharmony? Will it do all of these things in some measure, for example, provoke some while appeasing others? Bernard Grofman’s chapter has provided the formal framework within which such questions should be posed for voting systems, but, as he has insisted, context-sensitive knowledge is required for effective advice. With less than a century’s experience of widespread use of electoral systems for competitive elections it is

396

Brendan O’Leary

not clear that political scientists can predict the very long-run consequences of electoral system choice on the integration or accommodation of peoples; the majority, however, confidently predict that PR arrangements, ceteris paribus, are more likely to encourage power sharing. In deeply divided places there may be a majority, which may call itself the majority. It may insist that the polity consists of one people. There may also be at least one minority in population size that may refuse to describe itself as a minority and insist on its distinct recognition as a people. The former want integration, the latter accommodation. Political science may expand the imaginative range of possible institutional compromises for deeply divided places, as both Bernard Grofman and Alfred Stepan have suggested here, but it cannot promise to transcend genuine incompatibilities. Whether the relevant polity consists of one or more peoples may be an incompatibility; it may be negotiable, it may not. Creative ambiguity has its limits. It matters whether power sharing is seen as transitional toward integration of all into one people or whether it is seen as a goal in itself that permanently respects pluralism.

Power Sharing and the Courts Samuel Issacharoff (Chapter 8) raises the issue of the role of courts in powersharing systems. Courts are par excellence a delegated institution, expected to operate at some remove from democratic decision making. Two general issues arise that will likely produce further arguments and research. The first is one of constitutional design. Should courts formally be part of the relevant power-sharing arrangements in a deeply divided place? To wit, in a consociation, should the courts be consociational? Should parity and proportional representation apply to the judges appointed to senior positions in the major courts? Should unanimous, qualified majority or concurrent consent apply to the rulings of high, supreme, or constitutional courts? Should judicial veto powers be granted? If so, to each member of the supreme or constitutional court, a qualified majority, or the court as a whole? Should the courts honor consociational principles in hard cases, that is, use and balance the principles of inclusivity, parity, proportionality, autonomy, and veto rights in their jurisprudence? The answer for consociationalists at least is “yes.” How otherwise will a consociation operate? They say it is no different from asking

Conclusion

397

whether the courts in a federation or a confederation should be federal or confederal in character. Should each member-state have representation on the con/federal supreme or constitutional court? Should con/federal supreme courts operate according to qualified majority rules, which achieve significant territorial consensus across the con/federation? Should there be con/ federal or regional legal supremacy or parity between the federal courts and the courts of the federative entities? Should the courts honor federal or confederal principles and firmly police the agreed division of powers? What, it may be asked, is the alternative to making the courts part of the relevant power-sharing arrangements, designed just like them? It is to attempt to design them as depoliticized, technocratic, and entirely nonmajoritarian institutions. Yet how likely are such courts to develop and retain credibility in deeply divided places? If such courts overrepresent some groups on the judicial bench, how legitimate will their difficult decisions be? To ask all these questions is to identify at least one lacuna in political and legal science. There has been insufficient focus on the place of courts in consociational power-sharing systems. By contrast, there has been an abundance of literature on the place of courts in federations (but not always of a comparative or cumulative kind). The second general issue is whether courts, domestic or international, should consider themselves entitled to modify or strike down the par ticular power-sharing bargains reached between national, ethnic, or religious groups, for example, if they believe that such bargains have violated human rights, such as the human rights of those categorized as others. This question is not abstract. At the mundane level, U.S. courts, for example, rule on electoral laws and electoral administration that may encourage affirmative districting or that may affect the voting turnout of minorities (Grofman, Handley, and Niemi 1994). At the metaconstitutional level in 2009, the European Court of Human Rights ruled against the provisions of the power-sharing Dayton Agreement of 1995 contained in the Constitution of Bosnia. The latter mandates that a three-person collective presidency be composed of a Serb, a Croat, and a Bosniak. The court ruled against the mode of electing the collective presidency because, it said, the provisions violated the human rights of Roma and Jewish petitioners. The court ruled as it did, despite evidence heard that this provision and others like it were part of the compromises that ended the bloodiest war in Europe since 1945 and that nothing in law prevented a Jew or a Roma (or a Hungarian or Vlach, or any other

398

Brendan O’Leary

citizen of Bosnia) from identifying as a Bosniak, Serb, or Croat by nationality and thereby rendering themselves eligible for presidential office. Christopher McCrudden and I have argued that courts should be more prudent than the European Court of Human Rights was in this case about overturning or modifying power-sharing bargains. Such bargains may be necessary to end or prevent a war, and abstract difference-blind rules may not offer the assurances needed. Power-sharing bargains may acquire legitimacy through inclusive negotiations or from endorsements through electoral or referendum processes (though this was not so of the making and ratification of the Dayton Agreement and not discussed in the court’s opinion). Courts should avoid allowing themselves to become forums for “lawfare” by national, ethnic, or religious protagonists, who may seek to reverse what they previously conceded at the negotiating table (in this case the petitioners’ goals coincided with those of the Bosniaks who want a civic and integrated state). Court decisions of this kind arguably make it more difficult for future power-sharing settlements to be made precisely because they render key provisions regarding executive and legislative design vulnerable to subsequent modification or abrogation. They therefore make it more difficult for negotiators to give credible commitments—and so in the future they may respond by trying to exclude courts from having adjudicative powers over the agreement being made (McCrudden and O’Leary 2013). In this case, the European Court of Human Rights did not treat parity between the three constituent peoples of Bosnia, as they were defined in the Dayton agreement, as an overriding principle of a hard-won (and still precarious) settlement. An alternative decision for the court to have followed would have been simple. It could have observed the difficulty of the status of others, endorsed the ruling of Bosnia’s own constitutional court, and left it to the democratically elected successors to those who made the Dayton Agreement to change it according to the rules by which their predecessors had agreed to bind themselves. Our view, which argues for considerable judicial restraint in the regulation of power-sharing bargains, is, of course, not the only possible perspective about these matters. Courts and power-sharing arrangements are therefore one subject on which future research and arguments should be expected from scholars and practitioners. Simply put, should courts have any role in the unwinding of power-sharing agreements, and, if so, what might that be? Alternatively, are such agreements best unwound by the parties to those agreements, according to the rules of amendment to which they agreed?

Conclusion

399

The Promotion of Power Sharing: Arbitration and Self-Appointed Arbiters Florian Bieber (Chapter 13) assesses the novel power-sharing systems in the Western Balkans that have replaced parts of the former Yugoslavia, itself fairly described as a failed power-sharing system, albeit not a democratic one. Exploring the challenges of securing stable and effective power sharing in these cases, he points to four key factors: the role of international agents; decision-making rules including veto rights; forms of territorial autonomy; and the “thickness” of the state. Of these I want to highlight here the role of international agents as a likely subject for future research. For Bieber much of the difficulty in making power-sharing work in the Balkans lies with external powers—in both the design and operation of power sharing and the lack of consistency and follow-through in the conduct of international organizations. He discusses how international agents affect legitimacy, particularly when core institutions and policies have been imposed on at least one community, observing that sustained imposition undermines the potential for intergroup compromise, particularly when an international agent has the power to make decisions against the will of all parties. Arbitrators who become unaccountable and usurp sovereignty or authority deny the meaningful self-government that must underpin successful power sharing. The activist intervention of the European Court of Human Rights discussed above, purporting to strike down the constitutional provisions governing the formation of Bosnia’s collective presidency and its senate (its House of Peoples), is just one of a number of spectacular examples of paternalistic arbitration by outsiders in the Balkans. Others have included designing Bosnia’s currency, flag, anthem, and symbols of the state and removing elected politicians from office who have not been found guilty of crimes. Power-sharing systems must be distinguished by whether they are sovereign (like Lebanon) or subsystems of wider polities (like Northern Ireland), since this status has implications for the legitimate roles of outsiders. Table 16.1 suggests that the role of outsiders in power-sharing systems can be assessed as balanced, partisan, imposed, or destabilizing (or indeed as moving between these possible orientations). The ideal role for outsiders is as voluntarily agreed balancers who play the welcome role of positive support for a power-sharing system, mediate differences that arise upon request, and arbitrate only when asked to do so. Sometimes the outsiders are empathetic to a different partner to the system (the governments of the United

400

Brendan O’Leary

Kingdom and Ireland play these matching roles in Northern Ireland), but they coordinate to ensure balance. When, by contrast, outsiders are partisans of particular local partners and do not coordinate their actions in the interests of stability, the prospects for the power-sharing system are not good (think of Greece and Turkey over Cyprus, so far). The imposition of power sharing may in principle be welcomed by some parties to a civil war, but others may reject the unwelcome reduction in their status (think of formerly dominant minorities, such as the Serbs of Bosnia or the Sunni Arabs of Iraq). In any version of imposed power sharing one of the most delicate tasks is the withdrawal of the great or regional power that imposed the settlement. The withdrawal, if inept, may lead to a renewal of civil war. If, by contrast, the imposition is prolonged, sovereignty may be usurped. Table 16.1 expresses the view that “the Bonn powers” granted to— should that be “taken by”?—the High Representative of the Peace Implementation Council in 1997 made Bosnia a protectorate of the EU and NATO rather than a fully sovereign country, a status in which it remains. This explains why Bosnia is placed under both the sovereign and the regional categories of power-sharing system in the table. Moreover, many of the outsiders involved in the government (not governance) of Bosnia have been determined to dismantle its power-sharing arrangements to promote an integrated state, quite different from what was agreed at Dayton, albeit under duress, in 1995. One act of duress does not obviously excuse another. A more subtle critique of imposed power-sharing’s prolongation under an external arbiter also deserves emphasis. Too regular unilateral arbitration by outsiders means that the parties to a power-sharing agreement will not bear the costs of their disagreements and not have the incentive to resolve them. Instead all difficulties that arise can be blamed on the arbiter, thereby perpetuating both a standoff and dependency. Political irresponsibility, in short, may be encouraged by too regular interventions against political irresponsibility. The difficulties posed for power-sharing systems by those overtly intent on destabilization (whether driven by irredentist or other motives) need no elaboration. It bears recalling, however, that outsiders may unintentionally destabilize power-sharing settlements. Finland’s accession to the EU on January 1, 1995, and its consequent assent to and incorporation of the entire body of accumulated EU law (the acquis communautaire) would have had unfortunate consequence for the Åland Islanders. It was only late in Finland’s negotiations that it was realized that EU law would override the Åland treaties and statutes of autonomy. This difficulty eventually led Finland to

Cyprus (1960– 63) (Greece pro-Greek Cypriots, Turkey pro-Turkish Cypriots)† Lebanon (1989–) (Iran pro-Shiite, Saudi Arabia pro-Sunni; Syria claimed to be impartial; Israel had proxies)

Bosnia (1994–95)‡ (United States, NATO, and EU, formation of federation and Dayton Agreement) Iraq (2004–5–) (United States imposed Transitional Administrative Law in 2004; Iraqis made constitution of 2005 under U.S. mediation)

Lebanon (1967–) (Israel, PLO, Syria, Iran, Saudi Arabia)

Partisans

Imposers

Destabilizers

Åland Islands (European Union after accession of Finland to the EU)

Northern Ireland (UK after March 1972 and again in 1985 [but with Ireland] insisted that there would be no restoration of a devolved parliament except under local power sharing)

Bosnia (1997–) (Croatia pro-Croats, Serbia pro-Serbs; Turkey, EU, and United States mildly pro-Bosniak; after the high representative has “Bonn Powers,” Bosnia is no longer sovereign but rather a protectorate of the EU)

Northern Ireland (1998–) (UK and Ireland regarding 1985, 1998, and 1999 agreements) South Tyrol/Alto Adige (Italy and Austria regarding 1946, 1972, and 1992 agreements)

Regional Power-Sharing Polity (devolved, autonomous, or federacy)

* IGAD: Intergovernmental Authority on Development (Djibouti, Eritrea, Ethiopia, Kenya, Somalia, Uganda, and Sudan). † The UK, the decolonizing power, was a joint guarantor of the treaty establishing the independence and constitution of Cyprus and portrayed itself as impartial between Greek and Turkish Cypriots. ‡ The United States “put together” the Federation of Bosnia and Herzegovina in 1994, and then put it together with Republika Srpska in 1995 at Dayton.

Macedonia (2000–) (Ohrid Agreement, NATO, and EU) Sudan (2005–11) (Kenya, Uganda, Egypt, Italy, Netherlands, Norway, UK, United States, African Union, European Union, Arab League, IGAD,* UN)

Sovereign Power-Sharing Polity

Balanced

Outsiders’ Role in Power-Sharing System

Table 16.1. Possible Roles of Outsiders in Power-Sharing Systems

402

Brendan O’Leary

attach a protocol to European treaties safeguarding the islanders’ special rights over restricting non-islanders’ access to property rights and residence. The role of outsiders in the promotion of power-sharing arrangements will continue to be controversial. Functioning power-sharing systems plainly benefit from dispute-resolution procedures. Ulrich Schneckener has even proposed that arbitration mechanisms are typical of power-sharing systems, but it is clear when he elaborates that what he has in mind are domestic professional and nonmajoritarian mechanisms: “Measures include informal meetings among the group leaders, ombudspersons, formalized mediation committees, independent commissions or special arbitration courts in which all sides are represented” (Schneckener 2002, 205). It will be far easier to get external arbitrators, international organizations such as the UN, the EU, the Organization for Security and Cooperation in Europe (OSCE), and the African Union (AU), to perform dispute-resolution tasks well when they are fully attuned to the special characteristics of power-sharing systems and when they learn to avoid unintended destabilization, usurpation of sovereignty, or illegitimate imposition. These organizations can mediate effectively, but they need to take greater care either in assuming arbitration roles or in accepting them. One current danger in international diplomacy is opportunistic efforts to promote power sharing when rival parties dispute the outcome of a general election. If the incumbent party steals a presidential or parliamentary election or if the opposition insists that the incumbents have done so (or if the incumbents insist that the opposition has stolen the elections), then there is an obvious danger of disorder, which could lead to conflict and civil war. In these circumstances to promote power sharing between the incumbent(s) and the opposition, however, is to risk sacrificing respect for constitutional or democratic regularity. Such impulses may be fairly called the promotion of false or pseudo power sharing. The “power sharing” may not be a genuine response to divisions based on nationality, ethnicity, religion, or race. Whoever has stolen (or ignored) the elections then retains or receives a share of power at the expense of the rule of law and thereby increases political cynicism in the electorate. In any case power may not be meaningfully shared in such scenarios. Debate continues to this day over the merits of such temporary power-sharing arrangements in Kenya, Zimbabwe, and Madagascar, significantly brokered by international mediators. Power sharing’s democratic advocates rightly believe its merits are besmirched if international organizations do not actively uphold validly conducted elections or democratically

Conclusion

403

elected incumbents and instead promote deals with little normative merit because they involve cheaper resource commitments. In Cote d’Ivoire the international community did better.

Power Sharing and Federacy Arrangements Alfred Stepan (Chapter 9) seeks to clarify the precise conceptual status of a “federacy.” The idea was first issued by the late Daniel Elazar and has since enjoyed an underground existence in comparative politics, as well as a special currency in discussions of the Swedish-speaking Åland Islands, located between Finland and Sweden but part of Finland’s sovereign jurisdiction (Elazar 1987; O’Leary 2005b, 2008; Rezvani 2007). Stepan accurately illustrates the inspirational role of federacy thinking (and the Åland Islands’ precedent) in resolving conflict in Aceh in Indonesia. The federacy concept certainly adds to the repertoire of power-sharing constitutional designs for deeply divided places. According to Stepan, the ideal-typical definition of a federacy is the following: “a political-administrative unit in an independent unitary state with exclusive power in certain areas, including some legislative power, constitutionally or quasi-constitutionally embedded, that cannot be changed unilaterally and whose inhabitants have full citizenship rights in the otherwise unitary state” (Stepan, Linz, and Yadav 2011, 204). Let me propose a definition with four amendments to Stepan’s and then justify it, namely, “A federacy is an autonomous political region in an otherwise unitary or union state that has specifically federal institutional relationships with the political center of its host state, embedded in an internationally binding treaty, a constitution, or an organic law, which cannot be unilaterally altered by either the political center or the political officials of the federacy.” “Autonomous political region” is proposed because without selfgovernment, through possession of its own legislative powers, an entity would not have the element of self-rule necessary for federal qualities: federal is here understood to imply both self-rule and shared-rule relationships and bilateral processes for changing the specific conjunctions of self-rule and shared rule. The phrase “in an otherwise unitary or union state” is added because it is important not to dichotomize the world of independent states into unitary or federal states (elsewhere Stepan argues that Spain, India, and the United Kingdom are federal states: Stepan 2001; Stepan, Linz, and Yadav 2011). A

404

Brendan O’Leary

unitary state may be centralized or decentralized; it may possess a constitution that restrains its central government, or it may not. But what makes the state unitary is that the political center possesses the unilateral capacity in law wholly to restructure regional and local government and that it will normally structure it in a uniform (symmetrical) way. A “union state” also has a political center with the formal legal capacity unilaterally to restructure regional and local government, but, by contrast, it is necessarily decentralized, and its forms of decentralization recognize historical regions that were previously incorporated within it and often protect their legal and cultural particularities (and thereby it accepts asymmetries). The United Kingdom is a union state in its official self-descriptions. The Westminster parliament in London, according to its own official views, may restructure regional (“devolved”) and local self-government in England, Scotland, and Northern Ireland, but it may not abolish Scotland or Northern Ireland and incorporate them into England without violating the Acts of Union with Scotland and Ireland (later Northern Ireland), overriding their historic distinctiveness, including their distinct systems of law and their different patterns of church-state relations (McGarry 2010). Spain too is a union state, initially built through dynastic agglomeration and conquest like the UK. In Spain national sovereignty “belongs to the Spanish people, from whom all state powers emanate,” and “the indissoluble unity of the Spanish nation” is emphasized as “the common and indivisible [i.e., nonfederal] homeland of all Spaniards.” The constitution “recognizes and guarantees the right to self-government of the nationalities and regions of which [Spain] is composed and the solidarity among them all” and elsewhere provides for mandatory and possible forms of decentralization: “The State is organized territorially into municipalities, provinces and the self-governing [autonomous] communities that may be constituted. All these bodies shall enjoy self-government for the management of their respective interests.” The Constitution of India similarly officially describes India as a union of states (and not as a federation) and grants to the union government the power to create new states, increase or diminish their size, alter their boundaries, or alter their names, but it does not permit India to have no states. Since 1956, however, India has legislated that linguistic (though not religious) states may be created, each of which recognizes one titular majority language community (besides Hindi speakers). Since then the center has generally only modified the state structure of non-Hindi-speaking majority

Conclusion

405

states to reflect the interest of other language users; India thereby recognizes the historicity of its constituent language spaces. This elaboration obviously dissents from Stepan’s description of the United Kingdom, Spain, and India as federal states. They are not legally so. By my proposed definition, union states can establish federacies (for Stepan these would be federal states establishing federacies, which is why he would code such entities under the category of asymmetrical federalism). Th is matters because if it is correct to describe the UK, Spain, and India (and perhaps Denmark) as union states, then certain power-sharing entities stand in need of classification. Arguably the United Kingdom has established a federacy regarding Northern Ireland (O’Leary 2008), Spain may have established federacies through bilateral agreements with the historic nationalities of the Basque country, Catalonia and Galicia, and India has been intermittently willing to establish federacy-style arrangements in Kashmir. Readers may reasonably think not much may matter if Stepan and others code these countries as asymmetrical federal states, whereas I and others code them as union states. Certainly it is unlikely that anyone will ever die, I hope, for these distinctions. They matter, however, because whereas federations are obliged to share sovereign and constitutional amendment powers between the federal and regional (or state) governments, unitary and union states are not, except if they so bind themselves in federacy arrangements. This thought leads to the third proposed amendment of Stepan’s definition, namely, “a specifically federal institutional relationship with the political center of its host state, embedded in an internationally binding treaty, a constitution, or an organic law, which cannot be unilaterally altered by either the political center or the political officials of the federacy.” It is the agreement of a unitary or union state to bind itself in a federal relationship with the federacy that gives the latter entity its distinctive quality, its legal essence. The amendment recognizes that an international treaty may credibly commit a unitary state or a union state to establishing self-rule arrangements in a distinctive region that it will not unilaterally alter, as occurred, for example, with the treaty resolving the status of the Åland Islands between Sweden and Finland. It also arguably occurred through the United Kingdom’s signing of a treaty with the government of Ireland in 1999, which recognized the new institutions of Northern Ireland as an act of selfdetermination of the people of Ireland, North and South, and committed the UK government not to alter North-South institutions on the island of Ireland

406

Brendan O’Leary

or the new power-sharing institutions within Northern Ireland, except through the agreed processes of amendment contained in the 1998 agreement (O’Leary 2008). The third amendment also recognizes that an organic law, that is, a law with a higher threshold for passage at the political center, may serve the function of constitutionalizing the status of the federacy and disagrees with Stepan’s desire to count mere conventions as “quasi-constitutional,” an unnecessary concession to the myths of British public law. The last proposed amendment to Stepan’s definition is a suggested deletion, removing the requirement that the “inhabitants [of the federacy] have full citizenship rights in the otherwise unitary state.” One can, after all, imagine circumstances in which a federacy with significant self-rule reduces its entitlement to proportional representation in the central government and identical citizenship rights (and their taxation and conscription duties), that is, it trades having more autonomy for less formal say at the center. Just as important, one can imagine a differentiated model of citizenship in which the citizens of the federacy enjoy the right to exclude citizens from the rest of the polity from having, for example, rights of residence or voting in the federacy (as is true of the Åland Islands) and in which citizens of the federacy only enjoy full citizenship rights in the rest of the polity when they ordinarily reside somewhere in the rest of the polity. These are intended as entirely friendly amendments to the concept of a federacy. Do they matter? They may have the empirical consequence of slightly increasing the number of entities to be coded as federacies in the past and present. Further discussion of these matters and of the relationships between federacies and the full array of past and current autonomy agreements will doubtless continue among power-sharing specialists (Coakley 1994; Hannum 1996; Lapidoth 1997; Weller and Wolff 2005; Weller and Nobbs 2010).

Rational Choice and Power Sharing Bernard Grofman (Chapter 2) and Ronald Wintrobe (Chapter 5) show the virtues of formal rational choice modeling, especially when disciplined by real-world applications. It is, however, quite striking how little comparative rational choice work has been done on the making or breaking of powersharing systems. This is especially surprising given that one of the pioneering books in this field, Politics in Plural Societies: A Theory of Democratic

Conclusion

407

Instability, which argued that democratic stability was highly unlikely to be feasible in deeply divided places, was coauthored by Ken Shepsle, a wellknown exponent of rational choice in political science. Writing in 1972, Alvin Rabushka and Shepsle identified four configurations of ethnic politics: (1) competitive (where no group enjoys dominance); (2) majority domination; (3) minority domination; and (4) fragmented. In each configuration they suggested that the prospects for stable democracy “appear dim as the historical record has indicated” (Rabushka and Shepsle 2009 [1972], 206). Competitive ethnic configurations were predicted to polarize, to become centrifugal through “ethnic outbidding,” and to generate civil wars or coups; dominant majorities were predicted to stay dominant (with moderate reformers among them kept in line by the threat of hard-liners’ outbidding them); dominant minorities were thought unlikely to concede democracy in which they would lose; and fragmented ethnic configurations were considered unlikely to develop statewide brokerage parties and likely to succumb to military dictatorships or one-party rule. The authors did consider Switzerland a counterexample to the thesis that deeply divided places cannot be democratically stable. They took care of the anomaly, at least in their eyes, by attributing its stability to its being a confederation (in which largely homogeneous cantons hold most power, and potentially divisive ethnic or linguistic competition is therefore minimized). The take-home message from Politics in Plural Societies was that neither majoritarian nor power-sharing democracy offered much hope of stability for deeply divided places (what they called “plural societies”). When one looks at this book forty years after its publication, its authors were correct in 1972 to doubt the prospects for democratic stability in Ceylon (Sri Lanka) and in some Caribbean cases (e.g., Guyana and Trinidad). The same may certainly be said for Cyprus, Rwanda, and some of Zanzibar’s history, as well as for the Congo, Sudan, Nigeria, Lebanon, and Yugoslavia. Yet the subsequent picture across the cases they picked has not all been doom and gloom. In their terms, Northern Ireland has moved from majority domination closer to the competitive configuration and now appears to have a stable power-sharing system (Mitchell, Evans, and O’Leary 2009). Independent Mauritius has not been wracked by violent ethnic or religious politics and has been a developmental success, certainly by regional standards (Carroll and Carroll 2000). Belgium has morphed from a consociation in a unitary state to a consociational federation, as Kris Deschouwer and Philippe Van Parijs discuss (Chapter 4), but it has had no severe ethnic

408

Brendan O’Leary

violence, and it has not yet broken up. Among what were dominant minority configurations in 1972, namely, South Africa, Rhodesia, and Burundi, severe ethnic conflict has certainly occurred. They are, however, no longer dominated by their minorities, and South Africa has transitioned to an apparently stable consensus democracy, partly through consociational devices, while Burundi, after a horrific genocide, has a power-sharing arrangement, albeit an insecure one, in which Tutsis no longer have overt dominance (Lemarchand 2009, chaps. 9–11). Zimbabwe, of course, is not a postcolonial democratic power-sharing success story, but it is not clear whether that development relates to its ethnic configuration, which is one of majority dominance, whether thought of as blacks versus whites or Shona versus Ndebele (the population ratio favors the former by over five to one). Power sharing has, however, sometimes worked, even among the grim cases Rabushka and Shepsle looked at in 1972, but it is not clear that their approach can explain why, though it can help explain instability in power-sharing bargains. Future rational choice–driven comparative work on power sharing in deeply divided places should have at least three goals. The first is to generate novel insights from modeling the dynamics of power-sharing successes and failures in deeply divided places (varying the institutional and ethnic configurations more systematically). Rabushka and Shepsle’s original work selected failures, not a balance of successes and failures. Such work would need to contrast places with and without deep divisions and be underpinned with a credible sociology of the causation of conflict. Such modeling needs to be sophisticated in its treatment of institutions, for example, integrating the work of Lijphart on patterns of government, Cox and Grofman on electoral systems and their effects, or the work of Tsebelis and others on veto points, or using power indices to assess how power is shared within cabinet or multiperson executives. The innovative formal work of Samuel Merrill and James Adams suggests that though power-sharing institutions may not lead rational parties to moderate their policy platforms under PR and may encourage them to be extreme, the policy outputs encouraged by power-sharing systems are moderate by comparison with what one would expect under single-party dominance (Lijphart 1999; Cox 1997; Tsebelis 1990, 2002; Merrill and Adams 2007). The latter results diverge from what Rabushka and Shepsle suggest. Second, the development of plausible formal models of the situations faced by those who consider sharing power in deeply divided places after conflict is surely worth developing further (Walter 2002; Hartzell and Hoddie 2007).

Conclusion

409

Here the fields of comparative politics, international relations, and rational choice intersect but are not very advanced. Steve Brams and Marc Kilgour have written on the instability of power sharing in which they model two possible power-sharing partners as a pair of duelists (Brams and Kilgour 2010b). They show in two simple models why it is rational for each partner to seek to kill the other preemptively: shoot or be shot is the moral. In a third model, however, they maintain that if the costs of losses (including the catastrophic loss of life) can be properly assessed by the duelists then they are less likely to risk initiating ruinous conflict. Fear of damage, perhaps even of mutual damage, what our age has come to call “mutually assured destruction,” may incentivize cooperation, which they treat as power sharing. They conclude, however, pessimistically, “We see no way to speed up the process [of learning the costs of conflict]. Sometimes the damage must be inflicted to sink into collective memories and bring people to their senses, whether the conflict is interpersonal, international, or something in between” (2010b, 242). In a follow-up work Brams and Kilgour suggest that to deter conflict initiation it is best if the possible partners can respond nearly simultaneously to a preemptive defection and attack (Brams and Kilgour 2010a). It may be appropriate to develop this vein of thought, that is, modeling the incentives for power sharing as opposed to conflict initiation, albeit against a richer institutional backdrop than that provided by a tale of prospective duelists. What Brams and Kilgour suggest is sobering: stable power sharing requires credible massive deterrence capabilities against defection among the partners, combined with credible willingness to use such capabilities. If they are right, conflict is frequently initiated by risk takers who do not appreciate its likely costs and power sharing is best conducted by those who know they live under a balance of terror. To adapt Machiavelli, it may be better for power-sharers to fear one another than to love one another. A third possible domain for rational choice scholars to explore is whether the type of work initiated by Rabushka and Shepsle can provide microfoundations on the causes of ethnic conflict and its regulation that are compatible with recent large-N statistical work (and its prevention through power sharing). Here I have in mind the impressive work produced by Lars-Erik Cederman and Andreas Wimmer and their collaborators (Cederman and Girardin 2007; Wimmer, Cederman, and Min 2009; Wimmer 2008; Wimmer and Min 2006). The ethnic configurations used in their work are historically informed and are quite different from those in Politics in Plural

410

Brendan O’Leary

Societies, and their policy implications are more welcome to those who favor inclusive forms of government.

Empirical Research on Power Sharing Rabushka and Shepsle did not pretend to do large-N statistical analysis in Politics in Plural Societies. In their 2009 retrospect on their 1972 book they honestly recognized the selection bias in their sample of eighteen countries (238). Since 1972 large-N statistical work has been done, which is more alert to selection biases. In the introduction I suggested that the work of Arend Lijphart across all functioning democracies in a thirty-year period, Pippa Norris across all recent kinds of regimes, and Michaela Mattes and Burcu Savan across a representative sample of postconflict regimes provide plausible reasons for believing that power sharing enhances the performance of democracies (and autocracies, according to Norris) and reduces the likelihood of civil war recurrence. Minimally, this work, and work resembling it (Hartzell and Hoddie 2007), suggests that there is no reason why scholars or policymakers should conclude that power sharing is a pointless or unsuccessful way of regulating divisions in deeply divided places, before or after violent conflict. To the contrary. Yet there is little doubt that better large-N work can be done in future, both as the quality of data collection and treatment improves in political science and as scholars work over longer time periods dealing with more cases, including going back before World War II (from which too much social science begins). In particular, I have two suggestions. Power sharing itself must be consistently operationalized in ways that test explicitly articulated theories: Lijphart, Norris, and Mattes and Savun all operationalize power sharing rather differently, perhaps only proportionality is commonly coded across these three sets of authors, and perhaps only Lijphart has a full-fledged theory (in this case of consensus as opposed to consociational democracy). There are, of course, variations in the forms and scope of power-sharing arrangements, and these subtleties need to be reflected in further large-N work so that we may assess what appears to work best in what kinds of particular configurations. Second, large-N work on power sharing and its alternatives in the subregions of independent states remains terra incognita, partly because most databases on confl ict, and on institutions, are collected and organized at the level of the state, not the substate. Spatial variation exists within states, both in manifest and latent conflict and in the distribution and concentration

Conclusion

411

of national, ethnic, linguistic, and religious groups, and therefore we should expect the demand and the supply of power-sharing institutions to vary within states. Above all social scientists need to work with (or indeed as) historians to provide detailed knowledge of why some cases of power sharing have proven successful while others have failed, and why some places alternate between success and failure. Negotiation histories and implementation histories need to be systematically compared. In the English-language literature some cases are arguably overstudied (Northern Ireland) whereas many others are radically understudied (particularly in Africa and the Pacific). Sites of current violence attract inordinate scholarly attention, whereas peace (or nonviolence) is boring. The updated potted illustrations from eighteen countries in Rabushka and Shepsle’s book regrettably confirm some of the stereotypes of the critics of rational choice theory, namely that its exponents search for confirming evidence with “just so” stories (Green and Shapiro 1994); that they select from newspapers rather than specialist works on the relevant regions; that they get basic facts wrong; and that sometimes they do not mention when the primary sources or secondary specialists provide divergent accounts of the causes of conflict in deeply divided places or of how institutions have caused or calmed conflict, or, if they do, what principles should govern the interpretation of such sources (see Lustick 1996). Sound social science must be consistent with and be an ally of sound history. Not just professional history or appropriate respect for professional history is required but comparative history, which will focus on the negotiation and implementation of power-sharing arrangements in search of robust generalizations and examine to what extent and whether the same formal institutions operate in the same way in different places. The late Eric Nordlinger’s simple extraction of three necessary conditions for successful conflict regulation—appropriately motivated leaders, leaders autonomous enough to lead their followers, and stable bases of support for parties and leaders—showed what can be achieved from a careful historical comparison of six cases (Nordlinger 1972). More of such work needs to be done.

Conclusion Normative debate on power sharing, whether inside or outside deeply divided places, is not likely to stop. Its scholarly opponents condemn power

412

Brendan O’Leary

sharing as futile, as perversely likely to reinforce the causes of confl ict, or as certain to jeopardize key values, for example, individualism, equality of opportunity, or democratic competition (O’Leary 2005a). Many maintain that the alternatives are better—whether that be war to the finish (Luttwak 1999), partition (Kaufmann 1996, 1998), power division that rejects ethnic, religious, or linguistic bases for politics and exports a refined version of the U.S. experience (Roeder and Rothchild 2005; Roeder 2007), or other alternatives considered in the introduction. Proponents of power sharing consider these arguments unconvincing and argue that the alternatives have usually been much worse. To argue that war to the finish is better than (democratic and authentic) power sharing is to argue that death is better than using medicine with limited negative side effects; to argue for partitions is perverse when their track record is disastrous (O’Leary 2012, 2007, 2011); and to insist solely on the merits of power division overgeneralizes optimistically from the U.S. experience as an immigrant state (in which assimilation has become easier) and forgets the severe costs of that experience for many minorities and the original natives and the frequent dysfunctionality of institutions that do not strongly encourage coordination. Advocates of power sharing maintain that it works (Lijphart 2008; Norris 2008; Hartzell and Hoddie 2007; Mattes and Savun 2009); that it is a realistic response to likely or sustained conflict; and that it can lead to a more stable, better-performing, and more inclusive form of democracy (McGarry and O’Leary 2009a, 2009b). These arguments do not seem likely to be settled by appeals to evidence, whether based on case studies or large-N studies, or mixed methods. In particular conflicts, both outsiders and insiders will cherry-pick among the repertoires of existing rhetoric. Outsiders who wish not to intervene will argue that it is best to let conflict run its course. Those who wish to intervene are more likely to look well on the merits of power sharing. Insiders who calculate that they can win will argue that power sharing is impossible or too costly (a loser-takesall system) and press for victory through domination (or through partition). Patient readers, however, should have learned from the chapters in this volume of the wide variety of past, existing, and functioning power-sharing systems and their attendant difficulties. They will have been introduced to the wide repertoire of both tested and untested ideas about how to improve particular power-sharing arrangements. If this book encourages the policymaker, or the policymaker’s advisor, or, more likely, the teacher of the policymaker’s advisor to consider the merits of power sharing, as well as its potential

Conclusion

413

drawbacks, it will have served its intellectual function. Abba’s lines from the jilted lover who complains that “the winner takes it all” need to be deleted in deeply divided places. Let me offer the following emendations, in the hope of attracting better poets to the cause: The wise winner shares Those who want it all must postpone their cares It’s complex, and yet elegant and plain, If you are not included you should certainly complain. Notes My thanks to Bill Finan, Jon Fraenkel, John Hall, Joanne McEvoy, and John McGarry for their comments on this chapter; they are excused responsibility for its contents. 1. See also the introduction to this volume. 2. The majority-preferential vote is the description used in the country that has used the system the longest; see Wright 1986. 3. The supplementary vote (SV) is an abbreviated version of AV, used to elect mayors in England. There are two columns on the ballot paper: one to mark a first choice with an X, and another in which to mark a second choice with an X (making a second choice is optional). If no candidate receives a majority in the count of first-choice preferences, the top two candidates continue to a second and fi nal round. All other candidates are eliminated. The second-choice votes of everyone whose first choice has been eliminated are counted, and votes for the remaining candidates are then added to their first-round totals. Whoever then has the highest vote total is the winner. In limited preferential voting, used in Papua New Guinea (PNG) and to elect the president of Sri Lanka, voters are entitled to up to three rankings. In PNG ranking is obligatory for voters, and in Sri Lanka it is optional (and is little used). For information on PNG, thanks to Jon Fraenkel. 4. Under majoritarian systems the government formed after the second round (in two-round systems) or after the allocation of lower-order preferences (in majoritypreferential systems) may also be construed to have the mandate of a majority, but a majority of voters may not have directly mandated those who are in the government with a first-round or first-preference vote. Jon Fraenkel, John Coakley, and I have wondered in dialogue why centripetalists don’t favor two-round systems. After all, the interval between the two rounds in principle offers extensive opportunities for those who did well in the first round to make transparent (or covert) bids for lower-order preferences. Perhaps the spectacle of center-right candidates appealing for the second-round votes of eliminated National Front candidates in France too visibly suggests the power of centrifugal over centripetal dynamics in a majoritarian electoral system.

414

Brendan O’Leary

5. This, however, is the case in Israel, where most multiparty coalitions (except under the Yitzhak Rabin premiership) have discriminated against the Arab citizens of Israel (who live within Israel’s internationally recognized borders); see Lustick 1980; Beinart 2012, chap. 1; Yiftachel 2006. 6. For a fluent and logical defense of majority rule, see McGann 2006, especially chaps. 2, 3, 8. He argues that political equality requires proportionality (one vote of equal value), that equality implies majority rule (in the making of laws), and that the good performance of the consensus democracies (e.g., in Scandinavia) on a range of policy outputs is explained by their conjunction of PR electoral systems and simple majority rule in the legislature. Many of his arguments are convincing, but his understanding of (cycling) majorities is that their interests and identities are tradable rather than quite fi xed, i.e., he assumes that a people rather than two or more peoples exist. His minorities are within a people rather than different peoples. McGann (and others, including Joanne McEvoy in this volume) is right to caution that some types of supermajoritarianism and the creation of multiple institutional veto points may privilege minorities. The logical conclusion for advocates of power sharing is that building supermajoritarian procedures and requirements into a political settlement should be done only if the new constitution reflects a just status quo across the constituent peoples or when it credibly and programmatically requires that such a constitution be implemented. Advocates of power sharing should not forget that the first lucid thinking on concurrent majoritarianism came from a committed slaveholder, John C. Calhoun. See Kateb 1969; Read 2009. 7. Northern Ireland currently approximates this picture (though two major parties exist among the nationalists and unionists, respectively). 8. In Northern Ireland STV (PR) is used to elect the Assembly, and the proportional d’Hondt rule is used as a sequential allocation rule for forming the cabinet; see O’Leary, Grofman, and Elklit 2005. 9. In Northern Ireland, “In this Act . . . ‘cross-community support’, in relation to a vote on any matter, means—(a) the support of a majority of the members voting, a majority of the designated Nationalists voting and a majority of the designated Unionists voting.” Northern Ireland Act 1998 c. 47, Part 1, § 4 (5). 10. For discussion, see McGarry and O’Leary 2004; for a rigorous treatment, see Schwartz 2011. 11. Freeman 1863, 1893. 12. Checkel and Katzenstein 2009 passim. 13. It is more conventional to regard the EU as having a dual executive; see, e.g., Hix and Høyland 2011, chap. 2. 14. For the pioneering critical analysis of nonmajoritarian institutions within the EU, see Majone 1996, 2005, 2009. 15. See the discussion of arbitration below. 16. For other pioneering work on designing better voting procedures and better ways of sharing out public offices fairly, e.g., ministries in cabinets, see Brams 2008,

Conclusion

415

especially chaps. 9 and 10. Particularly on majoritarian judgments (though not applied in the context of ethnic divisions), see Balinski and Laraki 2010. 17. Sejdić and Finci v. Bosnia and Herzegovina (Applications Nos. 27996/06 and 34836/06), Judgment of the Grand Chamber, 22 December 2009. 18. Identification with one of three constituent peoples is required by the Dayton provisions, but they do not require any person to repudiate any ethnic, linguistic, or religious identity. 19. For a judicious metareview of the enormous literature in multiple languages on the breakup of Yugoslavia, see Ramet 2005. 20. Elsewhere attention has been drawn to the neglected role of external agents in the promotion and operation of consociational settlements (McGarry and O’Leary 2006). The emphasis in classical or traditional consociational theory was on how power sharing might be encouraged by the joint perception of a common external threat (e.g., Lijphart 1977; Nordlinger 1972). 21. Others have called the international interventions imperial and neocolonial; see, e.g., Knaus and Martin 2004; Chandler 2006. 22. One writer accuses outsiders of “faking democracy” after Dayton (Chandler 2000). It is certainly accurate to accuse outsiders of faking Bosnia’s sovereignty. 23. On Kenya and Zimbabwe, see Cheeseman and Blessing-Miles 2010; Allen 2009. These are cases of incumbent usurpations. Madagascar by comparison is a case of the lawful elected incumbent being ousted in a coup by a challenger and then obliged to accept a short-lived power-sharing settlement (see Bearak 2009). On anxieties about the rule of law as a result of peace agreements between governments and rebels, see Levitt 2005– 6, and compare with Bell 2006. There is an emergent literature among Africanists suggesting that power sharing perversely incentivizes greedmotivated (rather than grievance-driven) rebel groups (e.g., Tull and Mehler 2005). For other work on power sharing in Africa, see Lemarchand 2009; Spears 2002. 24. Constitution of Spain, 1978, §1 (2), §2, §137. 25. Constitution of India, 1951, as amended, Part 1. 26. States Reorga nization Act 1956, Constitution (Seventh Amendment) Act, 1956; see Adeney 2007 passim. 27. For some of the typically unnoticed difficulties in distinguishing between symmetrical and asymmetrical federations and the argument that all federations are asymmetrical in some respects, see O’Leary 2010. 28. A separate metareview needs to be written by an intrepid scholar on the interpretation and misinterpretation of Switzerland among comparativists, not least among advocates and critics of power sharing. One basic fact is often forgotten: Switzerland’s original deep division was religious, pitting urban Protestants against rural Catholics, and this cleavage underlay its nineteenth-century civil war. Switzerland’s concurrent majority rule in referendums (a popu lar majority overall combined with a majority of support among cantons) arguably arose from regulating this confl ict. Language conflict in Switzerland, by contrast, has rarely been severe. Moreover,

416

Brendan O’Leary

though Switzerland is multilingual, multiethnic, and multireligious, it is overwhelmingly mono-national: the Swiss regard themselves as a nation, and its linguistic minorities do not generally regard themselves as national minorities. Moreover, while Switzerland may call itself a confederation, it is a federation; indeed its current constitution was significantly modeled on that of the United States. I have been told that Kofi Annan wisely told Greek Cypriots that Switzerland was a federation that is called a confederation, whereas Turkish Cypriots were informed it was called a confederation. 29. One did not need to be an exponent of rational choice theory, however, to predict instability in these places. 30. Sisk 1995; Koelbe 2000; for different takes, see Guelke 1999; and Jung and Shapiro 1995. See also comments in Koelbe and Reynolds 1996 and O’Leary 2005a. 31. On Northern Ireland, on which I count myself a specialist, Rabushka and Shepsle open their update as follows: “The roots of conflict in Northern Ireland go back to the Battle of the Boyne in 1690, which established Protestant supremacy over Catholics in the north of Ireland. When Ireland became a separate nation as stipulated in the Government of Ireland Act in 1920, the six counties of Northern Ireland became part of the United Kingdom of Great Britain and Northern Ireland” (2009 [1972], 226). Most historians, however, would agree that the roots of the conflict go back to the colonization of Ulster at the beginning of the seventeenth century, if not earlier. Many historians would code the confl ict as between natives and settlers rather than between rival religious groups. Under the Government of Ireland Act of 1920, the UK government unilaterally partitioned Ireland and sought but failed to create two new dependent home rule parliaments in Northern and Southern Ireland. It manifestly did not therein recognize Ireland as a nation (the authors mean as a sovereign state). Only after further war between the UK government and the IRA was a treaty negotiated in 1921, under which the Irish Free State seceded from the UK, and under which Northern Ireland was granted the right to secede from the Irish Free State (which it did) and rejoin the United Kingdom, subject to a boundary commission to redraw the new border between Northern Ireland and the Irish Free State. If this specialist can quibble with so many arguably context-sensitive errors in his area, the illustrations from the other seventeen cases are necessarily put in doubt. 32. A similar tight focus on territorial settlements of ethnic conflicts in Europe shows what can be done; see Wolff 2003. References Adeney, Katharine. 2007. Federalism and Ethnic Conflict Regulation in India and Pakistan. New York: Palgrave Macmillan. Allen, Karen. 2009. “Has Kenya’s Power-Sharing Worked?” BBC News. http://news.bbc .co.uk /2/hi/africa/7921007.stm Balinski, Michel, and Rida Laraki. 2010. Majority Judgment: Measuring, Ranking, and Electing. Cambridge, MA: MIT Press.

Conclusion

417

Bearak, Barry. 2009. “Madagascar Political Rivals Agree to Power-Sharing Deal.” New York Times, 10 August, A7. Beinart, Peter. 2012. The Crisis of Zionism. New York: Times Books/Henry Holt and Co. Bell, Christine. 2006. “Peace Agreements: Their Nature and Legal Status.” American Journal of International Law, no. 100: 373– 412. Brams, Steven J. 2008. Mathematics and Democracy: Designing Better Voting and FairDivision Procedures. Princeton, NJ: Princeton University Press. Brams, S. J., and D. M. Kilgour. 2010a. “Stabilizing Power Sharing.” In A. Van Deemen and A. Rusinowska, eds., Collective Decision Making, 169–84. Berlin: Springer. ———. 2010b. “12: The Instability of Power Sharing.” In Matthew Braham and Frank Steffen, eds., Power, Freedom, and Voting, 227– 43. Berlin: Springer. Carr, Edward H. 1981 (1939). The Twenty Years’ Crisis. 2nd ed. London: Macmillan. Carroll, B. W., and T. Carroll. 2000. “Accommodating Ethnic Diversity in a Modernizing Democratic State: Theory and Practice in the Case of Mauritius.” Ethnic and Racial Studies 23, 1: 120– 42. Cederman, L. E., and L. Girardin. 2007. “Beyond Fractionalization: Mapping Ethnicity onto Nationalist Insurgencies.” American Political Science Review 101, 1: 173– 85. doi: 10.1017/s0003055407070086. Chandler, David. 2000. Bosnia: Faking Democracy After Dayton. 2nd ed. London: Pluto Press. ———. 2006. Empire in Denial: The Politics of State-Building. 2nd ed. London: Pluto Press. Checkel, Jeff rey T., and Peter J. Katzenstein. 2009. European Identity. Cambridge: Cambridge University Press. Cheeseman, Nic, and Tendi Blessing-Miles. 2010. “Power-Sharing in Comparative Perspective: The Dynamics of ’Unity Government’ in Kenya and Zimbabwe.” Journal of Modern African Studies 48, 2: 203–29. Coakley, John. 1994. “Approaches to the Resolution of Ethnic Confl ict.” International Political Science Review 15, 3: 297–314. Cox, Gary W. 1997. Making Votes Count: Strategic Coordination in the World’s Electoral Systems. Cambridge: Cambridge University Press. Elazar, Daniel J. 1987. Exploring Federalism. Tuscaloosa: University of Alabama Press. Freeman, Edward Augustus. 1863. History of Federal Government: From the Foundation of the Achaian League to the Disruption of the United States. Vol. 1. London: Macmillan & Co. Freeman, Edward Augustus. 1893. History of Federal Government in Greece and Italy. Ed. J. B. Bury. 2nd ed. London: Macmillan & Co. Green, Donald P., and Ian Shapiro. 1994. Pathologies of Rational Choice Theory: A Critique of Applications in Political Science. New Haven, CT: Yale University Press. Grofman, Bernard, Lisa Handley, and Richard G. Niemi. 1994. Minority Representation and the Quest for Voting Equality. Cambridge: Cambridge University Press. Guelke, Adrian. 1999. South Africa in Transition: The Misunderstood Miracle. London: IB Tauris.

418

Brendan O’Leary

Hannum, Hurst. 1996. Autonomy, Sovereignty, and Self-Determination: The Accommodation of Conflicting Rights. Rev. ed. Philadelphia: University of Pennsylvania Press. Hartzell, Caroline, and Matthew Hoddie. 2007. Crafting Peace: Power-Sharing Institutions and the Negotiated Settlement of Civil Wars. University Park: Pennsylvania State University Press. Hix, Simon, and Bjørn Høyland. 2011. The Political System of the European Union. Basingstoke: Palgrave Macmillan. Horowitz, Donald L. 2002. “Constitutional Design: Proposals Versus Processes.” In Andrew Reynolds, ed., The Architecture of Democracy: Constitutional Design, Conflict Management and Democracy, 15–36. Oxford: Oxford University Press. Jung, Courtney, and Ian Shapiro. 1995. “South Africa’s Negotiated Transition: Democracy, Opposition, and the New Constitutional Order.” Politics and Society 23, 3: 269–308. Kateb, George. 1969. “The Majority Principle: Calhoun and His Antecedents.” Political Science Quarterly 84, 4: 583– 606. Kaufmann, Chaim. 1996. “Possible and Impossible Solutions to Ethnic Civil Wars.” International Security 20, 4: 136–75. ———. 1998. “When All Else Fails: Ethnic Population Transfers and Partitions in the Twentieth Century.” International Security 23, 2: 120–56. Knaus, Gerald, and Felix Martin. 2004. “Lessons from Bosnia and Herzegovina: Travails of the European Raj.” Journal of Democracy 14, 3: 60–75. Koelbe, Thomas A. 2000. “The New South African Constitution: A Case of Consensus Democracy?” In Marjus M. L. Crepaz, Thomas A. Koelbe, and David Wilsford, Democracy and Institutions: The Life Work of Arend Lijphart, 129–53. Ann Arbor: University of Michigan Press. Koelbe, Thomas A., and Andrew Reynolds. 1996. “Power-Sharing Democracy in the New South Africa.” Politics and Society 24, 3: 221–36. Lapidoth, Ruth. 1997. Autonomy: Flexible Solutions to Ethnic Conflicts. Washington, DC: United States Institute of Peace Press. Lemarchand, Réné. 2009. The Dynamics of Violence in Central Africa. Philadelphia: University of Pennsylvania Press. Levitt, Jeremy L. 2005– 6. “Illegal Peace: An Inquiry into the Legality of Power-Sharing with Warlords and Rebels in Africa.” Michigan Journal of International Law 27: 495–577. Lijphart, Arend. 1977. Democracy in Plural Societies: A Comparative Exploration. New Haven, CT: Yale University Press. ———. 1999. Patterns of Democracy: Government Forms and Per formance in ThirtySix Countries. New Haven, CT: Yale University Press. ———. 2008. Thinking About Democracy: Power Sharing and Majority Rule in Theory and Practice. New York: Routledge. Lustick, Ian S. 1980. Arabs in the Jewish State: Israel’s Control of a National Minority. Modern Middle East Series. No. 6. Austin: University of Texas Press.

Conclusion

419

———. 1996. “History, Historiography, and Political Science: Multiple Historical Records and the Problem of Selection Bias.” American Political Science Review 90 (September): 605–18. Luttwak, Edward. 1999. “Give War a Chance.” Foreign Affairs 78, 4 (July–August): 36–44. Majone, Giandomenico. 1996. Regulating Europe. London: Routledge. ———. 2005. Dilemmas of European Integration. Oxford: Oxford University Press. ———. 2009. Europe as a Would-Be World Power: The EU at Fifty. Cambridge: Cambridge University Press. Mattes, Michaela, and Burcu Savun. 2009. “Fostering Peace After Civil War: Commitment Problems and Agreement Design.” International Studies Quarterly 53, 3: 737–59. McCrudden, Christopher, and Brendan O’Leary. 2013. Courts and Consociations: Human Rights Versus Power-sharing. Oxford: Oxford University Press. McGann, Anthony J. 2006. The Logic of Democracy: Reconciling Equality, Deliberation, and Minority Protection. Ann Arbor: University of Michigan Press. McGarry, John. 2010. “Asymmetric Autonomy in the United Kingdom.” In Marc Weller and Katharine Nobbs, eds., Asymmetric Autonomy and the Settlement of Ethnic Conflicts, 148–82. Philadelphia: University of Pennsylvania Press. McGarry, John, and Brendan O’Leary. 2004. “Stabilising Northern Ireland’s Agreement.” Political Quarterly 75, 3: 213–25. ———. 2006. “Consociational Theory, Northern Ireland’s Conflict, and Its Agreement, Part One: What Consociationalists Can Learn from Northern Ireland.” Government and Opposition 41, 1: 43– 63. ———. 2009a. “Part 1, Argument: Power Shared After the Deaths of Thousands.” In Rupert Taylor, ed., Consociational Theory: McGarry and O’Leary and the Northern Ireland Conflict, 15–84. London: Routledge. ———. 2009b. “Part III, Response: Under Friendly and Less-Friendly Fire.” In Rupert Taylor, ed., Consociational Theory: McGarry and O’Leary and the Northern Ireland Conflict, 333–88. London: Routledge. Merrill, Samuel, III, and James Adams. 2007. “The Effects of Alternative PowerSharing Arrangements: Do ‘Moderating’ Institutions Moderate Party Strategies and Government Policy Outputs?” Public Choice 131: 413–34. Mitchell, Paul, Geoff rey Evans, and Brendan O’Leary. 2009. “Extremist Outbidding in Ethnic Party Systems Is Not Inevitable: Tribune Parties in Northern Ireland.” Political Studies 57, 2: 397– 421. Nagel, Jack H. 2000. “Expanding the Spectrum of Democracies: Reflections on Proportional Representation in New Zealand.” In Markus M. L. Crepaz, Thomas A. Koelbe, and David Wilsford, eds., Democracy and Institutions: The Life Work of Arend Lijphart, 113–27. Ann Arbor: University of Michigan Press. Nordlinger, Eric A. 1972. Conflict Regulation in Divided Societies. Vol. 29. Occasional Papers in International Affairs. Cambridge, MA: Center for International Affairs, Harvard University.

420

Brendan O’Leary

Norris, Pippa. 2008. Driving Democracy: Do Power-Sharing Institutions Work? Cambridge: Cambridge University Press. O’Leary, Brendan. 2003. “Consociation: Refining the Theory and a Defence.” International Journal of Diversity in Organisations, Communities and Nations 3: 693–755. ———. 2005a. “Debating Consociation: Normative and Explanatory Arguments.” In S. J. R. Noel, ed., From Power Sharing to Democracy: Post-Conflict Institutions in Ethnically Divided Societies, 3– 43. Toronto: McGill-Queen’s University Press. ———. 2005b. “Power-Sharing, Pluralist Federation and Federacy.” In Brendan O’Leary, John McGarry, and Khaled Salih, eds., The Future of Kurdistan in Iraq, 47–91. Philadelphia: University of Pennsylvania Press. ———. 2007. “Analyzing Partition: Definition, Classification and Explanation.” Political Geography 26, 8: 886–908. ———. 2008. “Complex Power-Sharing in and over Northern Ireland: A SelfDetermination Agreement, a Treaty, a Consociation, a Federacy, Matching Confederal Institutions, Inter-Governmentalism and a Peace Process.” In Marc Weller, Barbara Metzger, and Niall Johnson, eds., Settling Self-Determination Disputes: Complex Power-Sharing in Theory and Practice, 61–124. Leiden: Martinus Nijhoff. ———. 2010. “Thinking About Asymmetry and Symmetry in the Remaking of Iraq.” In Marc Weller and Katharine Nobbs, eds., Asymmetric Autonomy and the Settlement of Ethnic Conflicts, 183–212. Philadelphia: University of Pennsylvania Press. ———. 2011. “Debating Partition: Evaluating the Standard Justifications.” In Kurt Cordell and Stefan Wolff, eds., The Routledge Handbook of Ethnic Conflict, 140– 57. London: Routledge. ———. 2012. “Partition.” In Thomas Wilson and Hastings Donnan, eds., The Blackwell-Wiley Companion to Border Studies, 29– 47. Oxford: Blackwell-Wiley. O’Leary, Brendan, Bernard Grofman, and Jorgen Elklit. 2005. “Divisor Methods for Sequential Portfolio Allocation in Multi-Party Executive Bodies: Evidence from Northern Ireland and Denmark.” American Journal of Political Science 49, 1: 198– 211. Rabushka, Alvin, and Kenneth A. Shepsle. 2009 (1972). Politics in Plural Societies: A Theory of Democratic Instability. Longman Classics in Political Science. New York: Pearson/Longman. Ramet, Sabrina P. 2005. Thinking About Yugoslavia: Scholarly Debates About the Yugoslav Breakup and the Wars in Bosnia and Kosovo. Cambridge: Cambridge University Press. Read, James H. 2009. Majority Rule Versus Consensus: The Political Thought of John C. Calhoun. Lawrence: University of Kansas Press. Rezvani, David A. 2007. “Shaping the Federacy Research Agenda.” Ethnopolitics 6, 1: 129–31. Roeder, Philip G. 2007. Where Nation-States Come From: Institutional Change in the Age of Nationalism. Princeton, NJ: Princeton University Press.

Conclusion

421

Roeder, Philip G., and Donald Rothchild. 2005. Sustainable Peace: Power and Democracy After Civil Wars. Ithaca, NY: Cornell University Press. Schneckener, Ulrich. 2002. “Making Power-Sharing Work: Lessons from Successes and Failures in Ethnic Conflict Regulation.” Journal of Peace Research 39, 2: 203–28. Schwartz, Alex. 2011. “How Unfair Is Cross-Community Consent? Voting Power in the Northern Ireland Assembly (June 2010)?” Northern Ireland Legal Quarterly 61, 4: 349– 62. Sisk, Timothy D. 1995. Democratization in South Africa: The Elusive Social Contract. Princeton, NJ: Princeton University Press. Spears, Ian S. 2002. “Africa: The Limits of Power Sharing.” Journal of Democracy 13, 3: 123–36. Stepan, Alfred. 2001. “Toward a New Comparative Politics of Federalism, (Multi)Nationalism, and Democracy: Beyond Rikerian Federalism.” In Arguing Comparative Politics, 315– 62. Oxford: Oxford University Press. Stepan, Alfred, Juan J. Linz, and Yogendra Yadav. 2011. Crafting State-Nations: India and Other Multinational Democracies. Baltimore: Johns Hopkins University Press. Taagepera, Rein, and Matthew Soberg Shugart. 1989. Seats and Votes: The Effects and Determinants of Electoral Systems. New Haven, CT: Yale University Press. Tsebelis, George. 1990. “Elite Interaction and Constitution Building in Consociational Democracies.” Journal of Theoretical Politics 2, 1: 5–29. ———. 2002. Veto Players: How Political Institutions Work. Princeton, NJ: Princeton University Press. Tull, Denis, and Andreas Mehler. 2005. “The Hidden Costs of Power-Sharing: Reproducing Insurgent Violence in Africa.” African Affairs 104: 375–98. Walter, Barbara F. 2002. Committing to Peace: The Successful Settlement of Civil Wars. Princeton, NJ: Princeton University Press. Weller, Marc, and Katharine Nobbs. 2010. Asymmetric Autonomy and the Settlement of Ethnic Conflicts. Philadelphia: University of Pennsylvania Press. Weller, Marc, and Stefan Wolff. 2005. Autonomy, Self-Governance and Conflict Resolution: Innovative Approaches to Institutional Design in Divided Societies. London: Routledge. Wimmer, Andreas. 2008. “The Making and Unmaking of Ethnic Boundaries: A Multilevel Process Theory.” American Journal of Sociology 113, 4: 970–1022. Wimmer, Andreas, Lars-Erik Cederman, and Brian Min. 2009. “Ethnic Politics and Armed Conflict: A Configurational Analysis of a New Global Data Set.” American Sociological Review 74, 2: 316–37. Wimmer, Andreas, and Brian Min. 2006. “From Empire to Nation-State: Explaining Wars in the Modern World, 1816–2001.” American Sociolog ical Review 71, 6: 867–97. Wolff, Stefan. 2003. Disputed Territories: The Transnational Dynamics of Ethnic Conflict Settlement. Studies in Ethnopolitics. New York: Berghahn Books.

422

Brendan O’Leary

Wright, Jack. 1986. “Australian Experience and Majority-Preferential and QuotaPreferential Systems.” In Arend Lijphart and Bernard Grofman, eds., Choosing an Electoral System: Issues and Alternatives, 124–38. New York: Praeger. Yiftachel, Oren. 2006. Ethnocracy: Land and Identity Politics in Israel/Palestine. Philadelphia: University of Pennsylvania Press.

. Hodgson

CONTRIBUTORS

Liam Anderson (Ph.D., Georgia) is Professor of Political Science at Wright State University. His latest book (with Gareth Stansfield) is Crisis in Kirkuk: The Ethnopolitics of Conflict and Compromise. Florian Bieber (Ph.D., Vienna) is Professor of Political Science at the University of Graz, Austria. His latest book is Post-War Bosnia: Ethnic Structure, Inequality and Governance of the Public Sector. Scott A. Bollens (Ph.D., North Carolina) is Professor of Planning, Policy and Design and Warmington Chair in Peace and International Cooperation at the University of North California, Irvine. His latest book is City and Soul in Divided Societies. Benjamin Braude (Ph.D., Harvard) is Professor of History at Boston College. With Bernard Lewis he coedited Christians and Jews in the Ottoman Empire, the Functioning of a Plural Society: vol. 1: The Central Lands; vol. 2: The Arabic-Speaking Lands. Ed Cairns (Ph.D., Queen’s University Belfast) was Professor of Social Psychology at the University of Ulster. Tragically, he died in a road accident in February 2012. He had recently coedited the Handbook on Peace Education. Randall Collins (Ph.D., University of California, Berkeley) is the Dorothy Swaine Thomas Professor in Sociology at the University of Pennsylvania. His recent books include Violence: A Micro-Sociological Theory. Kris Deschouwer (Ph.D., Vrije Universiteit Brussel) is Professor of Political Science at Vrije Universiteit Brussel. He edits Re-Bel with Phillipe Van Parijs.

424

Contributors

Bernard Grofman (Ph.D., Chicago) is Jack W. Peltason Endowed Chair in Political Science and Director of the Center for the Study of Democracy at the University of California, Irvine. Among the many books he has coauthored is A Unified Theory of Party Competition. Colin Irwin (Ph.D., Syracuse) is Research Fellow at the University of Liverpool and author of The People’s Peace Process in Northern Ireland. Samuel Issacharoff (J.D., Yale) is Bonnie and Richard Reiss Professor of Constitutional Law at New York University. He is the author of The Law of Democracy. Allison McCulloch (Ph.D., Queen’s University, Kingston) is Assistant Professor of Political Science at Brandon University. She is researching the design of political institutions after ethnic conflict. Joanne McEvoy (Ph.D., Queen’s University Belfast) was Sawyer Mellon PostDoctoral Research Fellow at the Penn Program in Ethnic Conflict, University of Pennsylvania, and is now Lecturer in Politics at the University of Aberdeen. Brendan O’Leary (Ph.D., London School of Economics and Political Science) is Lauder Professor of Political Science at the University of Pennsylvania. In 2009–10 he was the Senior Advisor on Power-Sharing in the Standby Team of the Mediation Support Unit of the United Nations. His most recent book (coauthored with Christopher McCrudden) is Consociation and the Courts. Alfred Stepan (Ph.D., Columbia) is Wallace Sayre Professor of Government and the founding Director of the Center for the Study of Democracy, Toleration, and Religion at Columbia University. His most recent coauthored book is Crafting State-Nations. Philippe Van Parijs (D.Phil., Oxford) is Professor at the Faculty of Economic, Social and Political Sciences of the University of Louvain, where he has directed the Hoover Chair of economic and social ethics since its creation in 1999. He is the author of Linguistic Justice for Europe and for the World. Ronald Wintrobe (Ph.D., Toronto) is Professor of Economics at the University of Western Ontario and the author of The Political Economy of Dictatorship.

. Hodgson

INDE X

The relevant chapter is cited in bold before pagination; citations to the contributors within their own chapters are not recorded. Name Index Abba 16: 387, 413 Abbas II, Shah 6: 178 Abdülhamid II 6: 194 Abisaab, R. 6: 178, 195 Abrams, D. 11: 278, 280, 291 Adams, G. 11: 283 Adams, J. 16: 407 Adeney, K. 16: 416 Ahtisaari, M. 9: 243– 46, 251 n.9 Akerlof, G. 5: 172, 173 Alba, R. 1: 53, 55 Alesina, A. 5: 169, 173 Allen, K. 16: 416 Allport, G. W. 11: 278, 288, 291 Althusius 1: 25 Althusser, L. 1: 52, 55 Amir, Y. 11: 288 Anastasio, A. 11: 291 Anderson, L. 1: 50, 55; 15: 364–85 Andreoni, J. 5: 174 Annan, K. 16: 416 Ariely, D. 5: 156, 174 Aspinall, E. 9: 241, 243, 247, 251 Armstrong, K. 5: 156, 174 Aron, A. 11: 284, 291, 294 Aron, E. N. 11: 284, 291 Aron, R. 1: 52, 55 Bacharach, P. 1: 51, 55 Bachevan, B. 11: 291 Bailey, F. G. 15: 371, 384 Bales, R. F. 11: 279, 291 Balinski, M. 2: 71–72, 89, 90; 16: 416

Baratz, M. 1: 51, 55 Barsoumian, H. 6: 195 Barzani, Mustafa 5: 192; 15: 366 Baskin, G. 14: 351, 360 Bauer, O. 1: 24, 56 Bean, C. 1: 22, 56 Bearak, B. 16: 415, 417 Beavon, K. 14: 335, 360 Becker, G. 5: 137, 174 Beinart, P. 16: 414, 417 Bell, C. 16: 415 Belloni, R. 3: 96, 98, 109 Benson, D. 11: 281, 294 Bieber, F. 1: 49; 3: 97, 109; 10: 258, 263, 275; 13: 312–26; 16: 399 Black, A. 5: 148, 174 Black, D. 2: 74, 77, 78, 89 Blackman, T. 14: 341, 360 Blais, A. 2: 82 Blessing-Miles, T. 16: 415 Boal, F. 14: 343, 360, 362 Bogaards, M. 1: 16, 56 Bollens, S. 1: 50; 14: 327– 63 Borda, J. 2: 69 Bose, S. 3: 96, 97, 98, 101, 109 Bowler, S. 2: 82 Brams, S. 16: 409; 16: 414–15, 417 Braude, B. 1: 4, 56; 6: 176–97 Breton, A. 1: 56, 62 Brewer, M. B. 11: 284 Brooks-Kelly, B. 1: 55 Brown, G. 10: 271 Brown, R. 11: 278, 292 Brubaker, R. 1: 53, 56

426 Brunell, T. 2: 91 Byrne, S. 14: 341, 360 Cairns, E. 1: 48– 49; 11: 278–94 Calhoun, C. 1: 16, 56 Calhoun, J. 16: 414 Campbell, A. 11: 291 Campbell, G. 10: 269 Carr, E. H. 16: 386, 417 Carroll, B. 16: 417 Carroll, T. 16: 417 Cederman, L-E. 16: 409, 417, 421 Chandler, D. 1: 55; 16: 415, 417 Chaudhry, M. 3: 102, 103 Checkel, J. 16: 414, 417 Cheeseman, N. 16: 415 Choudhry, S. 10: 257, 275 Clark, S. 9: 241, 251 Clogg, R. 6: 196 Coakley, J. 16: 406, 411, 417 Cohen, S. 14: 362 Collin, J-P. 14: 360 Collins, R. 1: 47; 7: 198–213 Condorcet, N. 2: 68– 69, 90; 2: passim Connor, W. 1: 7, 56 Cordell, K. 1: 62 Cousens, E. 10: 260, 275 Cowen, B. 10: 271 Cox, G. 16: 408, 417 Crepaz, M. 1: 54, 56 Crosby, J. 10: 264– 65, 275 Crouch, H. 9: 241, 251 Cryssouchoou, D. 1: 26 Cunningham, C. 14: 341, 360 Dahl, R. 1: 52, 56 Davies, R. 1: 53, 56 Dawisha, A. 1: 23 Dawisha, K. 1: 23, 56 De Brún, B. 10: 269–70 de Grief, P. 8: 227 De Klerk, F. 5: 169 Deschouwer, K. 1: 46; 4: 112–31; 16: 407 Deshmukh, Y. 12: 310 DeVotta, N. 3: 99, 109 Dixon, J. 11: 289–90, 291 Djuli, N. 9: 243, 249, 251 n.8 Dodik, M. 3: 97, 98 Domhoff, W. 1: 2, 57 Donaldson, B. 14: 360

Index Douglas, J. 14: 341, 360 Dovidio, P. A. 11: 291 Dowds, L. 11: 285–86, 292 Durrheim, K. 11: 290, 291 Durkan, M. 10: 254 Easterly, W. 5: 169, 174 Elazar, D. 1: 57; 9: 239– 40, 252; 16: 403, 417 Elklit, J. 1: 28, 62; 16: 414, 420 Emin, J. 6: 183–86, 196 Engstrom, R. 2: 90 Epitropaki, O. 11: 292 Esman, M. 14: 360 Evans, G. 1: 27, 34, 57, 60; 2: 90, 92; 5: 174; 16: 407 Ewing, D. 14: 360 Farquharson, R. 2: 78, 79, 90 Fearon, J. 5: 156, 169, 174 Feld, S. 2: 75, 78, 89 Ferejohn, J. 8: 218, 227 Field, M. 3: 102, 110 Filippov, M. 2: 87, 90 Finkel, S. 3: 99, 111 Finlay, R. 1: 52, 57 Firth, P. 9: 248 Fish, S. 1: 24, 57 Fishburn, P. 2: 75, 90 Fitzduff, M. 14: 360 Forbes, H. 1: 57 Ford, D. 10: 271 Fraenkel, J. 1: 34, 57; 2: 77, 84, 90; 3: 102, 103, 104, 110; 16: 413 Francis, E. 2: 89, 91 Freeman, E. 16: 417 Fukuyama, F. 13: 321, 326 Gaber-Damjanovska, N. 13: 318, 326 Gaertner, S. 11: 291 Galbraith, P. 5: 156, 174 Galeotti, J-L. 1: 62 Gallagher, T. 11: 291 Gambetta, D. 5: 174 Gellner, E. 1: 39, 57 Geyl, P. 4: 114, 131 Gibbons, H. 6: 196 Gillingham, J. (UK) 1: 53, 57 Gillingham, J. (US) 1: 53, 57 Gil-White, F. 1: 53, 57 Ginsborg, P. 5: 174

Index Girardin, L. 16: 417 Glazer, A. 2: 83, 91 Glazer, N. 1: 53, 57 Gloppen, S. 8: 220, 227 Godson, D. 10: 254, 275 Gold, R. 5: 151, 174 Gordon, M. 14: 352, 361 Gormley-Heenan, C. 11: 279, 291 Gregorian, R. 10: 265 Green, D. 16: 410, 417 Griffi n, R. 2: 83, 91 Grofman, B. 1: 16, 22, 28, 34, 45, 57, 62; 2: 67–93; 3: 102–5, 111; 16: 388, 395–96, 397, 406, 408, 414, 417, 420 Gudgin, G. 2: 89, 93 Guelke, A. 1: 62; 16: 417 Gunaratna, R. 5: 148, 174 Habermas, J. 1: 16, 57 Hadfield, B. 14: 341, 360 Haekkerup, H. 13: 314, 322 Hajnal, Z. L. 11: 286–87, 291 Hale, H. 1: 54, 57 Hall, J. 16: 413 Hamberger, J. 11: 292 Hamilton, A. 1: 3, 59 Handley, L. 2: 91 Hanf, T. 1: 24, 58 Hannum, H. 16: 406, 418 Harris, G. 6: 190, 196 Harris, P. 1: 22, 58 Hart, T. 14: 335, 361 Hartzell, C. 16: 407, 410, 412, 418 Harwood, J. 11: 293 Haslam, S. A. 11: 279, 281, 282, 285, 287, 289–93 Hayek, F. von 1: 53, 63 Hayes, B. 11: 285–86, 291 Hays, D. 10: 264– 65, 275 Hazleton, W. 14: 361 Hendry, J. 14: 341, 361 Hepburn, A. 14: 328, 361 Herodotus 1: 53 Heskin, K. 11: 279, 291 Hewstone, M. 11: 278, 284, 287, 291, 292, 293 Hirschl, R. 8: 218, 227 Hirschman, A. 1: 54; 9: 234, 252 Hitchner, R. B. 10: 274, 276 Hix, S. 16: 414, 418 Hobbes, T. 1: 52, 58

427

Hoddie, M. 16: 407, 410, 412, 418 Hoff man, B. 5: 147, 148, 174 Hogg, M. A. 11: 278, 280, 282, 287, 289, 292, 294 Holbrooke, R. 3: 98 Hooghe, L. 14: 334, 361 Hopkins, N. 11: 282, 283, 287, 289, 292 Horowitz, D. 1: 19, 20, 22, 23, 40, 53, 58; 2: 68, 69, 79, 87, 90, 91; 3: 94, 95, 104, 105, 109, 110; 4: 121, 131; 14: 330, 361; 15: 370, 374, 375–76, 384; 16: 387, 418 Høyland, B. 16: 414, 418 Hughes, J. 11: 292 Humphreys, M. 9: 248 Iorga, N. 6: 179, 196 Irvin, C. 14: 342, 360 Irwin, C. 1: 49, 55, 58; 12: 295–311 Isaacharoff, S. 1: 47; 8: 214–27 Ivanovic, D. 14: 344, 361 Jans, M. 4: 122, 131 Jawad Bollani 15: 368 Jay, J. 1: 3, 59 Jeunieb, D. 9: 248 Johnson, N. 1: 61; 16: 420 Johnston, R. 2: 89, 93 Jokay, C. 14: 341, 361 Jones, S. 9: 248 Jovevska, A. 13: 318, 326 Jung, C. 5: 169, 174; 16: 416, 418 Kafadar, 6: 179, 196 Kalla, J. 9: 242 Karadžić, R. 10: 253–54 Karl, J. 5: 148, 174 Kateb, G. 16: 414, 418 Katzenstein, P. 16: 414 Kaufman, S. 5: 157, 174 Kaufmann, C. 16: 410, 418 Kedourie, E. 6: 176 Keith, M. 14: 361 Kemal, M. 6: 194 Kende, M. 8: 226, 227 Kestleloot, C. 14: 334, 361 Khamenei, A. 6: 191 Khan, R. 6: 194 Khorenatsi, M. 6: 184 Khumayni, R. 6: 191 Khuri, I. 14: 361

428

Index

Kilgour, M. 16: 409, 417 Kimi, I. 14: 361 King, C. 14: 362 Kingsbury, D. 9: 242, 252 Klein, O. 11: 287, 292 Knaus, G. 1: 55, 58; 13: 315, 326; 16: 415, 418 Knox, C. 14: 362 Koelbe, T. 1: 54; 16: 416, 418 Koetzle, W. 2: 91 Komšić, Z. 10: 253, 263 Kritovoulos of Imbros 6: 196 Kritz, N. 14: 355, 362 Kuper, A. 1: 52, 57 Kurrilld-Klitgard, P. 2: 80, 92 Laakso, M. 2: 92 Laitin, D. 5: 156, 169, 174 Lajčák M. 10: 264, 266 Lakeman, E. 2: 80, 92 Lal, B. 3: 103, 104, 110 Lapan, H. 5: 175 Lapidoth, R. 16: 406, 418 Laraki, R. 2: 71–72, 90; 16: 416 Lawson, S. 3: 103, 110 Leal, K. 6: 179, 196 Lemarchand, R. 16: 415 Levine, R. 5: 169, 174 Levitt, J. 16: 415, 418 Lewis, B. 6: 180, 196 Liddle, R. W. 9: 241, 252 Lijphart: A. 1: 22, 25, 26, 36, 40, 42, 43, 54, 58; M. 2: 68, 83, 92; 4: 121, 131; 8: 215, 217, 220, 227; 10: 254, 257, 259, 272, 276; 14: 330, 362; 15: 372–73; 16: 387, 408, 412, 415, 418 Lindner, R. 6: 179, 195, 196 Lintott, A. 10: 256, 276 Linz, J. 2: 87, 92; 9: 232, 240, 250 nn.1–2, 252; 10: 256, 276 Livingstone, A. 11: 287, 289–90, 292 Lordos, A. 12: 309, 311 Loughlin, J. 14: 340, 341, 362 Lublin, D. 2: 91 Lukowski, J. 10: 256, 276 Lust-Okar, E. 5: 169, 174 Lustick, I. 1: 15, 59, 61; 14: 362; 16: 411, 414, 418, 419 Luttwak, E. 16: 412, 419 Lyne, T. 1: 62, 2: 93

Mabry, T. 1: 59 MacBride, D. 14: 341, 362 MacCulloch, D. 1: 52, 59 Madison, J. 1: 3, 59 Majone, G. 1: 59; 16: 414, 419 Makiya, K. 1: 23, 59 Maloney, G. 2: 93 Mandela, N. 5: 169; 14: 335 Mankad, A. 11: 292 Manning, C. 3: 96, 98, 110 Marshall, G. 9: 249, 252 Marshall, J. 1: 62; 2: 93 Martin, F. 1: 55, 59; 13: 315; 16: 415, 417 Martin, R. 11: 292 Mattes, M. 1: 43– 44; 16: 410, 419 Maximilian I 7: 209 Mazor, A. 14: 362 McAllister, I. 11: 285–86, 292 McCourt, D. 11: 279, 292 McCrudden, C. 1: 59; 16: 398, 419 McCulloch, A. 1: 22, 45– 46; 3: 94–111; 16: 387 McDonald, M. 2: 91 McEvoy, J. 1: 48; 10: 253–77; 16: 413 McGann, A. 16: 414, 419 McGarry, J. 1: 9, 15, 19, 25, 29, 32, 33, 53, 54, 59, 61, 62; 3: 107, 108, 110; 8: 217, 227; 10: 259, 261, 268, 276, 292; 11: 278, 290; 14: 330, 362; 16: 404, 413, 414, 419, 420 McGuinness, M. 10: 269; 11: 279, 282, 286 McLoughlin-Volpe, T. 11: 294 Mehler, A. 16: 415, 421 Mehmed II 6: 180 Merrifield, A. 14: 362 Merrill, S. 2: 75, 92; 16: 407, 419 Metzger, B. 1: 61; 16: 420 Mier, J. 14: 347, 362 Miliband, R. 1: 2, 60 Mill, J. S. 1: 3, 60 Millar, F. 1: 51, 52, 60 Miller, J. D. 12: 305, 301 Miller, N. 11: 284, 291 Miller, N. R. 15: 370, 384 Mills, C. W. 1: 2, 60 Min, B. 16: 409, 421 Mishler, S. 3: 99, 111 Mitchell, G. 12: 301 Mitchell, P. 1: 27, 34, 60; 2: 90, 92; 5: 174; 16: 407, 419

Index Mladić, R. 10: 253–54 Montesquieu, C. 1: 52, 60; 7: 202; 8: 221 Moore, M. 1: 59 Morfit, M. 9: 243, 252 Moro, A. 5: 147 Morris, P. 1: 2, 60 Moyersoen, J. 14: 334, 363 Mozaffar, S. 2: 93 Mueller, D. 5: 174 Mujani, S. 9: 241, 252 Murphy, A. 4: 114, 131 Nagel, J. 1: 37, 51, 60; 16: 389, 419 Narsey, W. 3: 104, 111 Nelson, G. 11: 291 Nessen, W. 9: 251 n.5, 252 Netanyahu, B. 5: 157 Niens, U. 11: 287, 292 Nobbs, K. 1: 59; 16: 406; 16: 420 Noel, S. 1: 61; 16: 420 Noiriel, G. 1: 53, 57 Nordlinger, E. 1: 3, 60; 14: 330, 362; 16: 411, 415, 419 Norris, P. 1: 42– 43, 60; 11: 278, 293; 16: 410, 412, 420 Nurdin, Dr. 9: 248 Oakes, P. J. 11: 294 Obama, B. 1: 23; 5: 157 O’Leary, B. 1: 1– 66; 2: 90, 92, 93; 3: 107, 108, 111; 5: 172, 174; 8: 217, 227; 10: 259, 261, 268, 274, 276, 292; 11: 278, 290; 14: 330, 362; 15: 376–77, 385; 16: 386– 422 Olugbemi, S. 15: 375, 385 Olver, C. 14: 362 Onorato, R. S. 11: 281, 294 Opp, K-D. 5: 144, 174 Ordeshook, P. 2: 87 Oz, A. 5: 156, 174 Paisley, I. 11: 283, 286 Palley, C. 14: 362 Palmer, B. 9: 241, 251 Paolini, S. 11: 284, 293 Pape, R. 5: 147, 174 Pareti, S. 3: 104, 111 Parnell, S. 14: 362 Parsons, T. 1: 2, 62 Peiris, P. 3: 99, 111 Persson, T. 10: 256, 276

429

Pettigrew, T. F. 11: 278, 284, 290, 293 Pildes, R. 8: 214, 227 Pinder, C. 11: 292 Pirie, G. 14: 362 Pishevari, J. 6: 191 Plato 1: 51 Platow, M. J. 11: 285, 289 Plutarch 1: 51 Polybius 1: 51, 62 Portes, A. 1: 62 Post, J. 5: 148, 175 Powell, B. 1: 55, 62; 2: 84, 93 Pringle, M. 14: 343, 362 Purdy, M. 10: 254, 276 Putnam, R. 1: 52, 62 Qarase, L. 3: 103 Rabin, Y. 16: 170, 412 Rabushka, A. 1: 54, 62; 16: 407, 409, 416, 420 Radmanović, N. 10: 253–54, 263 Rajapaksa, M. 3: 99, 100 Ramet, S. 16: 415 Rasler, K. 5: 144, 175 Read, J. 16: 414, 420 Regenwetter, M. 2: 75, 93 Reicher, S. 11: 282, 283, 285, 287, 289, 292, 293, 294 Reid, A. 9: 251, 252 Reid, G. 14: 363 Reid, S. A. 11: 282, 287, 289, 292 Reilly, B. 1: 22, 57, 62; 2: 79, 84, 88, 90, 93; 3: 100, 105, 106, 107, 108, 111 Reynolds, A. 1: 22, 62; 2: 88, 90, 93; 3: 101, 107, 111; 16: 416, 418 Reynolds, K. J. 11: 280, 294 Rezvani, D. 1: 47, 63, 403, 420; 9: 240, 252 Richardson, D. 2: 89, 93 Ricolfi, L. 5: 147, 151, 175 Robertson, M. 14: 360 Roeder, P. 1: 10, 63; 13: 326; 14: 356, 363; 16: 412, 420, 421 Roland, G. 10: 256, 276 Ropp, S. A. 11: 294 Rose, R. 11: 279, 293 Ross, D. 14: 351, 363 Rossos, A. 6: 189, 197 Rothchild, D. 1: 10; 13: 326; 14: 356, 363; 16: 412

430 Rousseau, J-J. 1: 14 Rowthorn, B. 1: 62 Royle, A. 14: 343, 362 Ruehl, W. 5: 144, 174 Rumbaut, R. 1: 62 Rummel, R. 1: 7, 15, 63 Russell, B. 1: 63 Rust, M. 11: 291 Rustow, D. 9: 242, 252 Ruthven, M. 5: 148, 175 Saari, D. 2: 75, 89, 93 Saddam Hussein 5: 156–57; 14: 354–55; 15: 366 al-Sadr, M. 5: 156 Saey, P. 14: 334, 361 Sager, L. 8: 218, 227 Salah, R. 5: 160 Salamey, I. 14: 363 Salih, K. 1: 54, 61, 62 Salmon, P. 1: 62 Samuels, D. 10: 256, 276 Sandler, T. 5: 175 Šarović, M. 3: 97 al-Sarraf, S. 14: 355, 362 Saunders, J. 1: 63 Saura, J. 14: 347, 363 Savun, B. 1: 43– 44; 16: 410, 418 Schneckener, U. 10: 257–58, 276; 11: 282, 293; 16: 389, 402, 421 Schulze, K. 9: 251, 252 Schumpeter, J. 1: 53, 63 Schwartz, A. 16: 414, 421 Sebastián, S. 10: 263, 276; 13: 326 Shalah, A. 5: 160 Shapiro, I. 5: 169, 174; 16: 410, 416, 417, 418 Shariatmadari, K. 6: 191 Sharon, A. 5: 162 Shepsle, K. 1: 54, 62; 16: 407–9, 416, 420 Shevetsova, O. 2: 87, 90 Shirlow, P. 11: 288, 293 Shugart, M. 2: 93; 16: 388, 421 Shuttleworth, I. 11: 287, 292 Silajdžić, H. 10: 253–54, 263, 265 Simeon, R. 3: 108, 111; 10: 259, 261, 268 Sisk, T. 3: 96, 111; 11: 278, 293; 14: 335, 363; 15: 370, 385; 16: 416, 421 Skiotis, D. 6: 187 Slater, P. E. 11: 279

Index Smith, A. 2: 89, 93 Smith, M. 14: 363 Smooha, S. 14: 363 Soberg, M. 2: 93 Spahiu, N. 14: 363 Spears, I. 16: 415, 421 Spears, R. 11: 287, 292 Speight, G. 3: 103 Stansfield, G. 15: 366 Stefes, C. 14: 335, 363 Steiner, J. 1: 27, 63 Stepan, A. 1: 29–30, 47; 9: 231–52; 16: 396, 403– 6, 421 Stockwell, R. 2: 91; 3: 104, 111 Suara, J. 14: 363 Sugden, R. 2: 90, 93 Suharto (general) 9: 240– 41 Sukarnoputri, M. 9: 249 Sunstein, C. 8: 223, 227 Sussman, N. 5: 173 Sutarto, E. (general) 9: 241, 251 n.7 Svensson, A. 11: 269 Sweeney, P. 14: 340, 363 Swenden, W. 4: 121, 131 Swyngeedouw, E. 14: 334, 362, 363 Syme, R. 1: 51, 63 Taagepeera, R. 2: 82, 92; 16: 388, 421 Tabar, P. 14: 363 Tabellini, G. 10: 256, 276 Tajfel, H. 11: 280–81, 293 Takaki, R. 1: 24, 63 Tam, T. 11: 292 Tarte, S. 3: 103, 111 Tausch, N. 11: 281, 285, 292, 293 Taylor, C. 1: 63 Taylor, Paul 1: 26, 54, 63 Taylor, Peter 2: 89, 93 Taylor, R. 1: 54, 63 Terhorst, P. 14: 334, 363 Their, A. 14: 355, 362 Thomas, H. 1: 53, 63 Thucydides 1: 53 Tideman, N. 2: 89, 93 Tihić, S. 10: 265 Tilly, C. 7: 205 Tocqueville, A. de 1: 3, 56; 7: 202 Tredoux, C. 11: 290, 291 Trew, K. 11: 281, 285, 291, 294 Trimble, D. 10: 254

Index Tropp, L. R. 11: 278, 293 Tudor, M. 11: 291 Tufakji, K. 14: 351, 363 Tull, D. 16: 415, 421 Tully, J. 1: 35, 63 Turner, J. C. 11: 280, 283, 293, 294 Twite, R. 14: 351, 363 Valuenzela, A. 2: 92 Van den Berghe, P. 1: 53, 63 van de Ven, J. 14: 334, 363 Van Parijs, P. 1: 46, 4: 112–31; 16: 407 Voci, A. 11: 292, 293 Von Hecker, U. 11: 292 Wallerstein, I. 7: 208 Walter, B 1: 43, 63; 16: 408, 421 Wattenberg, M. 2: 93 Weber, M. 1: 55, 64; 7: 203 Weeden, K. 11: 292 Weller, M. 1: 59, 61; 16: 406, 420, 421 Weingast, B. 1: 53, 64 Wetherell, M. S. 11: 294 Wever, B. de 4: 130 Wheatley, S. 10: 259, 277 Whyte, J. 11: 281, 294 Wickremesinghe, R. 3: 99, 100 Williams, R. 2: 89, 93 Wilsford, D. 1: 54 Wilson, R. 11: 291 Wimmer, A. 1: 23, 64; 16: 409, 421 Wintrobe, R. 1: 46, 56, 62, 64; 5: 135–75; 16: 406 Wittek, P. 6: 179 Wolff, S. 1: 62; 16: 406, 416, 420, 421 Woon, L. 1: 53 Wright, J. 16: 422 Wright, S. C. 11: 294 Yadav, Y. 9: 232, 250 n.2, 252 Yataganas, X. 10: 256 Yellen, J. 5: 173 Yiftachel, O. 16: 422 Young, H. P. 2: 74, 90, 93 Yudhoyono, S. B. 9: 242, 249 Yusuf, I. 9: 247 Zahar, M. J. 10: 263, 277 Zartman, I. W. 9: 242, 252 Zolberg, A. 1: 53, 64

431 Subject Index

Aaland (Åland) Islands 1: 47; 9: 240, 243; 16: 400– 401, 403, 405 acceptability criterion 2: 74 accommodation 5: 157–70; 14: 329; accommodationist strategies 1: 19–30; 5: 165–70. See also power sharing Aceh 9: passim; 16: 403; Free Aceh Movement (GAM) 9: 241– 49 African Union (AU) 16: 402 Albanian National Liberation Army 13: 323 Albanians 12: 301, 303 Alliance Party of Northern Ireland; 10: 271; 12: 300 All Party Representative Committee 12: 308 al Qaeda 9: 240 alternative vote (AV) (majority-preferential vote or instant run-off ) 1: 16, 33, 37, 45– 46; 2: 68, 70, 75, 78, 79, 80; defi ned 70; 3: 94–111 passim; 15: 371–72, 374; 16: 387–88 Anglo-Irish Agreement 12: 298 Anthimos I 6: 182 antiplurality voting 2: 70, 76 approval voting 2: 70 (defi ned), 75 Arabization 15: 366 arbitration 16: 399– 403 Argentina 1: 30 assimilation 1: 11, 13, 50; 6: 178; acculturation 1: 15; assimilationist strategies 1: 15–16; and devshirme 14: 329; fusion 1: 16 Australia 1: 30 Austria/Austrians 1: 30; 7: 206 autonomy 3: 101; 6: 180; 9: passim; 13: 319–22; 12: 306; 16: 392. See also consociation; federacy; federation avoid the worst criterion 2: 74 Azeris, 30 Badinter majority 13: 318 Baghdad 14: 352–53; displacement, 14: 352; map of Iraq 14: 354; politically combustible, 352–53 Bali 9: 240 Balkans 1: 49; 6: 180–81; 8: 215; 12: 301–3; 16: 395, 399 Bangladesh 1: 43; 7: 212 Basques 1: 30; 9: 231, 240; 16: 405

432

Index

Belfast 11: 288; 14: 339– 43; direct rule 14: 339; map 14: 340; responsibility sharing 14: 339– 43 Belfast Agreement 10: 254, 268– 69; 11: 283; 12: passim; 14: 339; aka Good Friday Agreement 1: 49; May 1998 referendum 12: 300, 308 Belfast Telegraph 12: 297 Belgium, 1: 26–27; 4: passim; 9: 231, 247; 16: 407; consociation in 4: 121–23; 8: 216; federal institutions 4: 118–21; linguistic & territorial divisions 4: 114–18; from unitary to federal state 4: 113 Belgrade Agreement 13: 314 Benin 1: 43 Bhutan 9: 239, 251 n.4 bicameralism 8: 221 Borda, J-C./Borda rule, 2: 68; defi ned 2: 70, 75 Bosnia and Herzegovina 1: 27, 34, 48; 3: 95, 97; 8: 219; 10: passim; 12: passim; 13: passim; 14: 343; 16: 397–98, 400, 415; Constitutional Court 13: 315, 317; constitutional reform 12: 302–3; 13: 316; House of Peoples 10: 260, 264– 65, 267; House of Representatives 10: 262, 264– 65, 267; Republika Srpska 10: 262, 267; 12: 301, 308 Bosniaks. See Bosnia and Herzegovina British state. See United Kingdom Brussels 4: 114, 117, 118, 120–21, 130; 14: 327, 328, 332–34; communities 4: 118; evolution of institutions 4: 115 (fig. 4.1), 118 ff; 14: 333 Burundi 1: 40; 16: 392, 408 Byzantine empire 6: 179; 7: 204 Canada/Canadians 1: 39, 40; 2: 82, 88; 5: 155, 165, 167– 68; 7: 210; 8: 216; 9: 231, 247; 16: 392 Catalonia 1: 36; 9: 231, 240; 16: 405 Catholic 12: 296–97, 299–300 Centre for Democracy and Reconciliation in South East Europe (CDRSEE) 12: 301, 310 centripetalism 1: 12, 19–24, 30–31, 33, 37, 45– 46; 3: passim; 16: 387–88 Chechnya 5: 147 China 7: 202; 9: 239; 12: 309 Christ, O. 11: 293 citizenship 9: 235–36

civil society 1: 5– 6 competition for power 1: 12–14 conciliar/conciliar movement 7: 199–200, 210 concurrent majority (rule[s]): qualified majority voting 4: 120; 16: 390 ff; supermajoritarianism 8: 221 condominium 1: 52 Condorcet criterion 2: 74 and passim; Condorcet efficiency 2: 74; Condorcet extension method 2: 74; Condorcet loser criterion 2: 76; Condorcet winner 2: 74 confederal state/confederation 9: 238– 40; 12: 307 consociation/consociationalism 1: 4, 14, 19, 25–28, 31, 33, 37, 38, 47; 2: 68; 8: 214–19; and courts: 8: 218–26; 10: 257; 13: 316–17; 14: 330; 16: 396–99; authoritarian consociation 1: 27; in Belgium 4: passim; 8: 216; corporate consociation 1: 27, 10: 259, 268; democratic consociation 1: 27; liberal consociation 1: 27; 10: 259, 268; under the Ottomans 6: 180 constitutionalism 1: 8–14; 8: 218 ff; certification 8: 219–24; constitutional reform 12: 302; proposals/agreements 12: passim contact hypothesis 1: 48; 11: passim control, hegemonic 14: 330 Coombes rule 2: 68; defi ned 2: 70, 76, 78 Corsica/Corsicans 1: 40 Cote d’Ivoire 16: 403 Council of Europe 10: 266 courts 8: passim; 16: 396–98 Croats 5: 155; 8: 219; 12: 301; 16: 397 cross-border institutions 1: 30. See also North-South institutions culture(s) 1: 5 Cyprus 1: 27, 30, 47; 8: 216, 218; 10: 273; 16: 392, 400, 416; Annan Plan 10: 273 Darfur-Darfur Dialogue and Consultation (DDDC) 12: 310 Dayton Peace Agreement (DPA) 8: 219; 10: 260, 263, 274 n.5; 12: 301, 308; 13: 314, 322; 14: 345; 16: 397, 400, 415 decentralization 9: 236; 13: 319–20. See also devolution deeply divided places 1: 5– 6; 16: 406–7 democracy 7: passim

Index Democratic Party of Albanians (DPA) 13: 318 democratic transitions 14: 357 Democratic Union for Integration (DUI) 13: 318 Democratic Unionist Party (DUP) 10: 254, 269–71, 275; 12: 297 Denmark 1: 29; 9: 250 n.3 devolution 9: 236; 12: 305–7 D’Hondt method 1: 28; 2: 73, 81; 10: 268 direct rule 12: 298 dispute resolution procedures 9: 236 Droop quota 2: 73, 80 Dutch Republic. See Netherlands Duverger’s Law 2: 82 East Timor 9: 240 electoral engineering 4: passim; and pre-electoral incentives 4: 124 ff Estonia 2: 88 Euro 16: 395 European Commission for Democracy Th rough Law (Venice Commission) 10: 264; 13: 315 European Court of Human Rights 16: 397–98, 399 European Union (EU) 9: 247; 12: 301, 302, 308; 13: 314; and consociation 1: 26; and Copenhagen criteria 8: 214, 217; European Central Bank 16: 394; European College of Commissioners 10: 264; 16: 393; European Council 16: 393; European Council of Ministers 16: 393; European Court of Justice 16: 394; European Parliament 16: 393; and territorial pluralism 1: 33; 7: 198; 8: 214; 16: 392–95 evolution of local governance 14: 356 executive(s)/executive formation 4: 112 ff; executive power(s) 16: 389, 393 Faroe Islands 1: 29 federacy 1: 54; 9: passim; 16: 403– 6 federalism 5: 167– 68; 8: 221; 9: passim; 14: 330 federation/federal state 2: 87; 7: 201, 208; 14: 343; 16: 403; in Belgium 4: passim; and cantonization 14: 330. See also territorial pluralism Fiji 1: 34; 2: 69; 3: 95, 102–5

433

Finland 1: 47; 16: 400, 403, 405 first place criterion 2: 74 Flanders 9: 231 floorwalking 8: 224–26 followers 5: 147–52 framing 5: 155–57 France 1: 53; 2: 71–72, 88; 7: 198, 200, 202, 204, 206; 16: 388; Th ird Republic 9: 232 Gagauz 1: 40 Galicia 16: 405 Geneva (Calvin’s) 1: 4, 5 geopolitics and collegiality and power sharing; 7: passim Germany 1: 28; 7: 198, 202; 16: 393 ghazi state 6: 179 Gökalp, Z. 6: 189 Good Friday Agreement 10: 254, 268– 69; 11: 283. See also Belfast Agreement Greece and the Macedonian question 6: 178, 188 Greenland 1: 29; 9: 240 Guyana 16: 407 Hausa-Fulani 15: 375 Helsinki 9: 243, 246, 248 Hutus 5: 156; 16: 392 immigration 9: 235 independence 12: passim index of the effective number of parties (Laakso-Taagepeera) 2: 82 India 1: 29, 34, 43; 2: 82; 7: 198, 212; 8: 216, 226; 9: 231, 239, 247; 12: 309; 16: 403, 404; Mizoram 9: 250 n.2 Indian-administered Kashmir (IaK) 12: 303–8; 16: 405 integration/integrationists 1: 11, 12, 13, 31–32, 35, 37, 50–51; appropriateness 1: 39– 40; strategies 1: 16–18 Internally Displaced Persons (IDPs) 12: 303 Internal Macedonian Revolutionary Organization-Democratic Party for Macedonian National Unity (VMRODPMNE) 13: 318 International Criminal Court for the Former Yugoslavia (ICTY) 10: 253, 263 international law 8: 222; 9: 237 Iran 1: 47; 5: passim; and Azeri question 6: 178, 190–91; and Kurds 6: 192

434

Index

Iraq 1: 47, 53; 7: 198; 212–13; 11: 279; 16: 400 Ireland, Irish state, Irish Republic 1: 30; 2: 88; 12: 298–99 Irish Republican Army (IRA) 11: 270–71 Islam 9: 240– 41 Israel 1: 46; 5: passim; building the wall inside and around the West Bank, 5: 160– 65; withdrawal from Gaza 5: 160– 65; 12: 301; 16: 414 Italy 2: 88; 7: 204–5, 208 Jakarta 9: 240, 245 Jammu 12: 306 Japan 2: 88; 7: 198; 9: 232 Jerusalem 5: 156; 14: 347–52, 349 (fig. 14.5); hegemonic control 14: 347–52; politically combustible 14: 347–52 Johannesburg apartheid geography 14: 335–38, 335 (fig. 14.2); transitional power sharing 14: 335–38 joint authority 12: 298–99 Joseph Rowntree Charitable Trust (JRCT) 12: 310 judicial review 8: 222 Kashmir 1: 48; 5: 147; 9: 231; 12: 295, 303, 305– 6, 309–10 Kenworthy, J. 11: 291, 292, 293 Kenya 1: 13; 16: 402 Kirkuk 1: 49; 14: 353–55; 15: passim; Article 58 of the TAL and 15: 382 n.7; Article 140 and 15: 367– 68, 377–78, 381; and consociation 15: 372–73; future governance 15: 369–70; and integrationism 15: 370–72; in map of Iraq 14: 354; oil 15: 368– 69; politically combustible 14: 354; power sharing in 15: 376–78; sixty-forty problem 15: 364– 65, 373–76; struggle for Kirkuk 15: 365– 68 Korea, South 1: 43 Kosovo 5: 169; 12: passim; 13: passim; Serb List for Kosovo 13: 324 Kurdistan/Kurds 1: 30, 36, 40, 54; 5: 147; 156; 6: 178; 189–90, 194; and Iraq, 6: 190; 15: passim. See also Kirkuk Kurdistan Democratic Party (KDP) 15: 366 Kuwait 5: 157 Kyrgyzstan/Kyrgyz 1: 53

leaders/leadership 5: 137 ff; 11: passim Lebanon 1: 27, 34, 47; 5: 147; 8: 216, 218; 16: 399, 407 Macedonia 6: 188–89; 12: 301 ff; 13: passim; EU Special Representative 13: 315 Madagascar 16: 402 Magyars 1: 30 Mauritius 16: 407 median voter argument questioned 2: 83 méthode majoritaire 2: 68; defi ned 71, 75 metropolitanism 14: 357 Mitrovica 14: 360 n.16 Moldova 1: 40 Montreal 14: 359 n.14 Mostar 14: 359 n.15 multiculturalism 1: 19, 24–28, 31 Namibia 8: 226 Netherlands 1: 26; Dutch Republic 7: 207–8 New Zealand 1: 39; 2: 88; 8: 226 Nicosia 14: 359 n.15 Nigeria 1: 22, 23, 40; 15: 375; Nigerian People’s Congress 15: 375; 16: 407 nonmajoritarian institutions 16: 394, 397, 402 North Atlantic Treaty Orga nization (NATO) 12: 302, 308; 13: 314–15; 16: 400 Northern Ireland 1: 13, 27–30, 38, 40, 48, 54; 3: 107; 5: 155, 165, 169; 8: 216; 10: passim; 11: passim; 12: 295–300; 16: 399, 405, 407, 414, 416; Assembly 10: 260, 268; Bill of Rights 12: 299; Council of the Isles (officially British-Irish Council) 12: 299; North-South institutions 12: 298; public opinion 12: 295–300 Northern Marianas Islands 9: 240 Norway 9: 232 Office of the High Representative 10: 273; 13: 315–16 Ohrid Framework Agreement 12: 301; 13: 314–15 Orga nization for Security and Cooperation in Europe (OSCE) 3: 97; 13: 315; 16: 402 Ottoman empire 1: 4; 5: passim package 12: 297 Pakistan 7: 198, 212; 12: 305– 6, 309 Palestine/Palestinians 1: 46; 5: passim; 12: 302

Index Papua New Guinea 1: 34; 3: 108; 16: 413 parity 4: 120–21; 16: 390–96 Pavia group 4: 124, 129–30 Peace Implementation Council 13: 316 peace polls 12: passim “people,” the/multi-people 12: 295; 16: 390 plurality rule (aka winner-takes-all, “relative” majority voting but wrongly known as “first past the posts”) 2: 68; comparing PR and plurality 2: 80–84; defi ned 2: 72; plurality bloc voting 2: 73, 78 Poland 7: 206 polarized cities 14: 328 Polish-Lithuanian Commonwealth 10: 256 Portugal 9: 232 Povratak (“Return”) 13: 323 power 1: 1–2 power sharing: accommodationist strategies 1: 19–30; appropriateness 1: 40– 41; and collegial power 7: 199 ff; competition for power 1: 12–14; contributions in this book 1: 45–50; defi ned 1: 1–5; defi nitional issues 16: 387–88; and electoral systems 2: 67–93; 3: 94–111; 4: 112–31; among ethnic groups 5: 155–70; and geopolitics, 7: passim; goal and instruments 1: 4; and normative debate 16: 411–12; and opposition 1: 37–38; 12: 309–10; pseudo power sharing 16: 402–3; and rational choice 5: passim; 16: 406–9; testing and empirical research on 1: 40– 45; 16: 410–11; urban approaches 14: 329–32 predetermination 10: 259, 268 preferences, single-peaked 2: 67, 76–79 Progressive Unionist Party (PUP) 12: 300 proportionality and proportional representation (PR) 1: 27, 33, 37; 2: 68, 73; in Belgium 4: 120–21, 123–29; 5: 167; 16: 387–89, 390–96; comparing PR and plurality 2: 80–84; comparing list PR and STV 2: 84–85; list PR 1: 27; 2: 73; mimicking list PR with plurality and STV with AV 2: 85–86; STV PR 1: 27; 2: 73 Prussia 7: 206 public opinion 12: passim Puerto Rico 9: 239 Quebec/Quebecois 1: 28, 36, 40; 9: 231; 16: 392

435

rational choice/social choice 2: 68 ff; 5: passim; 16: 406–9 recognition 1: 24 Republika Srpska (RS)/Serb Republic 3: 95, 96–98; 12: 301 ff; 14: 343 Roman Empire of the German Nation (Holy) 7: 199, 206–10 Romania (Rumania) 1: 30; 6: 179–80 Rome 10: 256 Royal Ulster Constabulary (RUC) 12: 299 Russia 7: 198, 205, 212; 12: 308; and Azeris 6: 193 Rwanda 16: 407 Safavid empire 1: 4; 5: passim Sarajevo 14: 343– 47; Dayton accords 14: 344, 346 (fig. 14.4); politically fragile “reunification” 14: 343– 47 Saudi Arabia 5: 147 scoring rules 2: 75 Scotland/Scots 1: 29, 36 self-categorization theory 11: 281–84, 287 self-determination 9: 231, 236–37; 10: 259, 268 segregation 11: 284, 287 separation of powers 1: 10–12; 8: 221 Serbia/Serbs/Bosnian Serbs 1: 30; 3: 97–98; 5: 155; 8: 219; 10: 254; 13: 323; 12: passim; 16: 397; Serbia and Montenegro 13: passim Sierra Leone 8: 218 Singapore 1: 43 single-seat and multiseat methods of voting 2: 80 ff single transferrable vote (STV)(Hare or Hare-Clarke system). See proportional representation Sinhalese/Sinhala 3: 99–102; 12: 303–8; 16: 392 Sinn Féin 10: 268– 69, 271 sixty-forty problem 15: 364– 65, 373–76 Slovakia/Slovaks 1: 30 Social Democratic and Labour Party (SDLP) of Northern Ireland 10: 254, 269; 12: 300 social identity approach 11: 280–84 social psychology 11: passim solidarity multiplier 5: 150 South Africa 1: 47; 2: 88; 5: 165, 169; 8: 218, 219–26; 16: 388, 408 sovereignty 9: 236–37

436

Index

Soviet Union (USSR) 1: 54; 7: 203 Spain 1: 29; 9: 231, 247; 16: 403, 405 spoilers 13: 324 Sri Lanka 1: 13, 34, 49; 3: 95, 98–102; 7: 212; Sri Lanka 9: 242; 12: 295, 307–9; 16: 392, 407, 413 St. Andrews Agreement 10: 268, 270–72 state/stateness 1: 6–7, 49–50; minimalist states 13: 321–22; state nation 9: 232; weakness of African states 7: 211–12 Sudan 16: 407 supplementary vote 3: 98 ff; 16: 413. See also alternative vote Sweden 1: 47; 16: 405 Switzerland 1: 26, 27, 28; 7: 200, 206–7, 208; 9: 232; 16: 389, 407, 415 Syria 1: 47 Tamil(s) 3: 99–102; 12: 305; 16: 392; Tamil Eelam, 12: 307 Tamil Tigers (LTTE) 3: 99; 9: 242 Tanganyika 1: 29 Tanzania 1: 29, 40 Team CVoter Foundation 12: 310 territorial pluralists/pluralism 1: 28–32, 36 third-party intervention 14: 330 Tibet 12: 305 Togo 1: 43 Transitional Administrative Law (of Iraq) 15: 366–77 tribunes 10: 256 Trinidad 16: 407 tsunami 9: 242 Turkey 1: 53; 6: 180; and Kemalism 6: 194; and Kurdish question 6: 189–90, 195 Turkish Cypriots 1: 30, 40 Turks in Bulgaria 8: 218 Tutsi 1: 40; 5: 156; 16: 392

two-round system (TRS) 1: 16; 2: 68, 72, 75; defi ned 2: 72 tyranny of the majority 1: 3, 8–12, 13, 19; 8: 215 Tyrol, South 1: 30, 40 Ukraine 1: 43 Ulster Democratic Party (UDP) 12: 300 Ulster Unionist Party (UUP) 10: 254, 269 union states 1: 29; 16: 404–5 United Kingdom/Great Britain/British state 1: 29, 52; 2: 82; 6: 176; 7: 198; 16: 403– 4 United Nations 5: 148, 149; 9: 239, 246; 10: 273; 16: 402 United States 1: 11, 28, 35, 39; 2: 82, 88, 5: 147, 148, 157; 7: 198, 200, 208, 210–11; 9: 232, 239; 12: 296, 301, 308 urban policies and strategies, importance of 14: 358 Uzbekistan/Uzbeks 1: 43, 53 Venice, Republic of 1: 4; 7: 201, 204; and influence on United States 7: 204 veto rights, rules 1: 48; 10: passim; 13: 317–19; 16: 396 veto rules 10: passim vital national interests 10: 253–54, 258, 265, 274 n.5; 13: 317 vote pooling 1: 22; 2: 68 Wales 1: 29 winner-takes-all in single-member districts (WTA-SMD) 1: 16 Yugoslavia 16: 399, 407 Zanzibar 1: 29; 16: 407 Zimbabwe (previously Rhodesia) 16: 402, 408

. Hodgson

AC KNOW LEDG MENTS

The editors are extraordinarily grateful to all our contributors; we hope our readers will benefit as much as we have from their experience and intellect. With equal gratitude we acknowledge the Sawyer-Mellon Seminar grant, which made this volume possible. At the University of Pennsylvania, Dr. Roy Eidelson helped in the initial grant application and in its initial administration, as did Jenna Laske. Most of the seminar’s logistics were smoothly run by the Sawyer-Mellon doctoral fellows David Bateman and Stephan Stohler, who have bright futures ahead of them in the academy. Symeon Braxton proved most helpful in liaison with the Mellon Foundation. The faculty, staff, graduates, and undergraduates of Penn’s Political Science Department, and those of many other social science and humanities departments at Penn and elsewhere in the Philadelphia region (notably at Haverford, Bryn Mawr, and Temple), participated extensively in the enterprise that led to this book. Special thanks are also owed to the Penn Law School. We must also express our profound gratitude to Ambassador Peter W. Galbraith, Professor Richard Johnston of the University of British Columbia, Professor John McGarry of Queen’s University, Canada, Professor Jack Nagel of the University of Pennsylvania, Professor Pippa Norris of Harvard, Professor Philip Roeder of the University of California at Davis, and Dr. Gwen Sasse of Oxford University for their contributions. The University of Pennsylvania Press, especially Bill Finan, Peter Agree, and Erica Ginsburg, have been true professionals as always. Joanne McEvoy wishes to say that she is indebted to her coeditor, Brendan O’Leary, for giving her the opportunity to spend a year at the University of Pennsylvania as the Sawyer-Mellon Postdoctoral Research Fellow at the Penn Program in Ethnic Conflict. His sage advice, sense of academic rigor, and commitment to imaginative and critical thinking on postconflict institutions are most helpful. The Andrew W. Mellon Foundation funded her fellowship, research, and travel. She warmly thanks doctoral fellows David Bateman and Stephan Stohler for their support and friendship.