Introduction The Decline and Fall of Parliamentary Sovereignty THE DECLINE OF PARLIAMENTARY SOVEREIGNTY The idea of the sovereignty of Parliament was long seen as the core of democratic practice. The superior position of the popularly elected legislature and its corollary of majority rule have been central principles for democratic revolutionaries since the notion was appended to the unwritten English constitution.1 At that time, the threat to liberty was monarchical power, and the subjugation of monarchical power to popular control was the primary goal. The resulting doctrine was that Parliament had "the right to make or unmake any law whatever; and further, that no person or body is recognized by the law of England as having a right to override or set aside the legislation of Parliament,"1 In the continental tradition, the intellectual underpinning of parliamentary sovereignty was provided by the Rousseauian concept of the general will. The people were supreme, and their general will as expressed through their republican representatives could not be challenged. This theory, combined with the regressive position of the judicial parlements in the French Revolution, led to a long tradition of distrust of judges in 1 i'lse original focus in England during the Glorious Revolution was on control of the crown rather than the rule of the people per se, because the democratic franchise was quite restricted. Jeffrey Goldsworthy, The Sovereignty of Parliament: History and Philosophy (1999). Rakove distinguishes the supremacy of Parliament from the idea that representative bodies were primarily designed to be law-making bodies. Jack Rakove, "The Origins of Judicial Review: A Plea for New Contexts, 49 Statu L. Rev. 1031, 1052. (1997). 1 Albert V. Dicey, The Lam of the Constitution 3-4 (8th ed., 1915). z Judicial Review in New Democracies France.3 The government du juges replaced the crown as the primary threat to popular will in French political thought.4 It was natural that the early proponents of democracy supported parliamentary sovereignty. They saw threats to liberty from the traditional sources: the ancien regime, the monarchy, and the church. Once these formidable obstacles to popular power had been overcome, theorists could hardly justify limitations on the people's will, the sole legitimate source of power. As democratic practice spread, however, new threats emerged. In particular, Europe's experience under democratically elected fascist regimes in World War II led many new democracies to recognize a new, internal threat to the demos. No political institution, even a democratically legitimate one, ought to be able to suppress basic liberties. Postwar constitutional drafting efforts focused on two concerns: first, the enunciation of basic rights to delimit a zone of autonomy for individuals, which the state should not be allowed to abridge; and second, the establishment of special constitutional courts to safeguard and protect these rights. These courts were seen as protecting democracy from its own excesses and were adopted precisely because they could be coun-termajoritarian, able to protect the substantive values of democracy from procedurally legitimate elected bodies. The ideal of limited government, or constitutionalism, is in conflict with the idea of parliamentary sovereignty.5 This tension is particularly apparent where constitutionalism is safeguarded through judicial review. One governmental body, unelected by the people, tells an elected body that its will is incompatible with fundamental aspirations of the people. This is at the root of the "countermajoritarian difficulty," which has been 3 Jeremy Jennings, "From 'Imperial State to I'Etat de Droit': Benjamin Constant, Blandine Kriegel and the Reform of the French Constitution," in Constitutionalism in Transformation: European and Theoretical Perspectives 76, 78 (Richard Bellamy and Dario Castigtione, eds,, 1996). The parlements had engaged in a kind of judicial review themselves. Mauro Cappz[\etti, Judicial Review in the Contemporary World 33-34 (1971). The activation of the Cotiseil Constitutioitnel in the Fifth Republic, especially because it unilaterally read the preamble of the constitution as being legally binding in 1971, has radically changed French practice in this regard. See Alec Stone, The Birth of Judicial Politics in France (1992), 4 This distrust of a judicial role in governance, beyond applying legislation, led the French to create a special system of administrative courts in 1871. This system of special courts applying a separate law for the government led Dicey to argue that the French droit administratifwas less protective of individual liberties than the English institutional manifestation of the rule of law. Dicey, supra note 2, zio-xi, 266. 5 Paul W. Kahn, The Reign of Law: Marbury v. Madison and the Construction of America aiS (i997). Introduction the central concern of normative scholarship on judicial review for the past three decades.6 ■ Although the postwar constitutional drafting choices in Europe dealt parliamentary sovereignty a blow, the idea retained force in terms of political practice. More often than not, the idea was used by undemocratic regimes. Marxist theory was naturally compatible with parliamentary sovereignty and incompatible with notions of constitutional, limited gov-ernmenf. Similarly, new nations in Africa and Asia reacting to colonialism often dressed their regimes in the clothes of popular sovereignty, though oligarchy or autocracy were more often the result. Today, in the wake of a global "wave" of democratization, parliamentary sovereignty is a waning idea, battered by the legacy of its affiliation with illiberalism. Judicial review has expanded beyond its homeland in the United States and has made strong inroads in those systems where it was previously alleged to be anathema. From France to South Africa to Israel, parliamentary sovereignty has faded away. We are in the midst of a "global expansion of judicial power," and the most visible and important power of judges is that of judicial review.7 Even in Britain, the homeland of parliamentary sovereignty and the birthplace of constitutional government, there have been significant incursions into parliamentary rule. There have been two chief mechanisms, one international and the other domestic. The first mechanism is the integration of Britain into the Council of Europe and the European Union (EU), which has meant that supranational law courts are now regularly reviewing British legislation for compatibility with international obligations. The domestic subordination of legislation of the British Parliament to European law was established when the House of Lords disapplied a parliamentary statute in response to the European Court of Justice's (ECJ) Factortame decision of 1991.8 More recently, the incorporation of * The term, and the terrain of the debate, were laid out by Alexander Bickel, The Least Dangerous Branch: The Supreme Court at the Bar of American Politics (2d ed., 1986}. 7 NealTate and Thorsten Vallinder, eds., The Global Expansion of Judicial Power (1995}. s R. v. Secretary of State for Transport, ex parte Factortame Ltd (No. z) [1991] r A.C. 603. The case concerned parliamentary legislation aimed at preventing primarily Spanish-owned but British-registered ships from operating in particular quota areas. This violated various EU law principles of nondiscrimination. The House of Lords asked the ECJ whether it could issue a preliminary injunction against an act of Parliament anil was told that it had an obligation to do so where legislation violated EU treaty rights. For a detailed discussion of the case, see Josef DrexL "Was Sir Francis Drake a Dutchman? - British Supremacy of Parliament after Factortame," 41 Am. J. Comp. L- 55i (i993>- 4 Judicial Review in New Democracies Introduction 5 the European Convention of Human Rights into United Kingdom domestic iaw by the Human Rights Act 1998 has led to greater involvement of courts in considering the "constitutionality" of parliamentary statutes (and administrative actions) under the guise of examining compatibility with Convention requirements.9 Although as a matter of domestic law the Human Rights Act attempts to preserve parliamentary sovereignty in that it allows an explicit parliamentary derogation from the convention, it has not been wholly successful. The Parliament now tends to scrutinize legislation for conformity with the convention, and this is a source of constraint; furthermore, even explicit parliamentary derogrations may still lead to a finding by the European Court of Human Rights that Britain has violated its obligations. Thus, it cannot really be said that the Parliament is truly sovereign in Dicey's sense of being unchecked by other bodies. The second mechanism is the growth of domestic judicial review as shown by an expanding body of administrative law. According to many observers, United Kingdom (UK) courts are exhibiting growing activism in checking the government, especially since the 1980s.10 This administrative law jurisprudence has grown in recent years. The practice of international courts reviewing British legislation no doubt played a role in undermining the primary objection to domestic judicial review. The British objection to domestic courts exercising judicial review was not that judges were incapable of it or that the rule of law was a secondary goal. Indeed, it was the assertion that government was subject to ordinary law applied by ordinary judges that was at the heart of Dicey's celebration of the English constitution. Rather, the traditional objection to judicial review was that the people acting through Parliament possess complete sovereignty. This argument has now lost force. If the will of the Queen in Parliament is already being constrained by a group of European law professors sitting in Strasbourg, then the objection to constraint by British judges is much less potent. 5 See, for example, Ian Leigh, "Taking Rights Proportionately: Judicial Review, the Human Rights Act and Strasbourg," PublicLaw 2.65-87 (2.001), and David Feldman, "Parliamentary Scrutiny of Legislation and Human Rights," Public Law 32.3-48 (2.002). 10 See, for example, Jerold L. Waltman, "Judicial Activism in England," in Judicial Activism in Comparative Perspective 33-52 (Kenneth Holland, ed., 1991); Susan Sterett, Creating Constitutionalism? The Politics of Legal Expertise and Administrative Law in EnglandandWales (1997). For an older doctrinal exegesis of judicial review in UK courts, see C.T. Emery and B. Smythe, judicial Review (1986). Even if one believes that Parliament is still sovereign in the United Kingdom, the adaptability of the always-anomalous British unwritten constitution as a model is clearly declining. In Britain itself, academics widely agree that there is a crisis of constitutional legitimacy.11 Furthermore, several countries that were historically recipients of the British model have recently departed from it. In the Caribbean, several former British colonies have joined together to establish a new supranational court of final appeal, the Caribbean Court of Justice, discontinuing the practice of appeal to the Privy Council in London. Other former colonies have adopted constitutional acts or amendments entrenching new rights in the constitution.11 In some countries, such as New Zealand and Israel, these acts are amendable by ordinary majorities and not entrenched as in other polities. Nevertheless, they maintain great normative power as constitutional legislation and politically speaking are more difficult to amend than legislation concerning routine matters of governance, even if not institutionally protected. There has even been a step in this direction in Saudi Arabia, although the Saudi government continues to take the formal position that it has neither a constitution nor legislation other than the law of Islam.13 The major bastions resistant to judicial involvement in constitutional adjudication have lowered their resistance in recent years. The concept of expanded judicial power has even crept surreptitiously into the international system, where there has been recent consideration as to whether there is a sort of inherent power of judicial review in international law.14 The issue under consideration concerns whether the United Nations Security Council's findings that it is acting to defend peace and security under Chapter VII of the United Nations Charter (UN Charter} are reviewable by the International Court of Justice. There is no explicit ," For cites, see Tony Prosser, "Understanding the British Constitution," in Constitutionalism in Transformation; European and Theoretical Perspective 61, 68 n.33 (Richard Bellamy and Dario Castiglione, eds., 1996). 11 For example, the Israeli Basic Laws of 1992, the Canadian Bill of Rights Act (i960), the Canadian Charter of Rights and Freedoms (1981), and the New Zealand Bill of Rights Act (1992}. " In 1992, the government adopted a Basic System of Rules that defines the structure of government and establishes a new mechanism for succession. See Rashed Aba-Namay, "The Recent Constitutional Reforms in Saudi Arabia," 42 Int'l & Comp. 2-95 (1993)- M Dapo Akande, "The International Court of Justice and the Security Council: Is There Room for Judicial Control of Decisions of the Political Organs of the United Nations?," 46 Int'l & Comp. L.Q. 309 (1997); see also Jose Alvarez, "Judging the Security Council," 90 Am. J. Int'l L. 1 (1996). 6 Judicial Review in New Democracies provision for judicial review in the UN Charter, and a Belgian proposal to establish it during the drafting of the UN Charter was rejected. The International Court of Justice has, however, considered the issue in dicta. The court has thus far carefully avoided making an express finding that the security council has acted outside of the scope of its powers, but it refused to explicitly deny that the court has the power to review the security council's actions.15 The United Nations, of course, is not a democratic system, nor one wherein majority rule has ever been unconstrained, by virtue of the institutional entrenchment of particular founding nations through the veto power on the Security Council. It is nevertheless interesting that some of the same questions that confront new democracies are being asked at the international level as well. Is there any action by supreme organs in a legal system that are ultra vires} If so, who has the power to decide whether an action crosses the line? And if the answer is a judicial body, who guards the guardians of legality? As the "third wave" of democracy has proceeded around the globe, it has been accompanied by a general expansion in the power of judges in both established and new democracies. Virtually every post-Soviet constitution has at least a paper provision for a constitutional court with the power of judicial review.16 New constitutional courts have been established in many new democracies. The following table (Table i.i) demonstrates the spread in new democracies of constitutional courts, that is, bodies with the explicit power to overrule legislative acts as being in violation of the constitution. Countries listed in the table are those characterized by the Freedom House survey as democracies in zooo that had not been so as of 1986, plus other well-known "third wave" democracies. Table 1.1 shows that although there are institutional variations, providing for a system of constitutional review is now a norm among democratic constitution drafters. Indeed, that such a norm exists is also evidenced by the fact that new constitutions in countries that still fall fairly short 15 See "Questions of Interpretation and Application of die 1971 Montreal Convention Arising from the Aerial Incident at Lockerbie (Libya v. US; Libya v. UK)," 3, 114 l.C.J. {1992) (Provisional Measures). The issue was also raised in "Application of the Convention on the Prevention and Punishment of the Crime of Genocide (Bosnia/Herzegovina v. Yugoslavia (Serbia and Montenegro))," 3 LC.J. {1996} (Request for Provisional Measures). 16 See, for example, Rett R. Ludwikowski, "Constitution Making in the Countries of Former Soviet Dominance: Current Developments," 13 Ga. J. Int'l & Comp. L. T55 (1993), and Rett R. Ludwikowski, Constitution Making in the Countries of Former Soviet Dominance {1996). Ill Introduction 7 table i.i Constitutional Review in Third Wave Democracies Form of Constitutional Review (Key: CR = review by special body; JR = Freedom House review by courts; L = Rating 2000-01 scope of review or access Country (*= amendment only) (average) limited) Year of Constitution/ Last Major Amendment amendment only) Albania 1991* 4-5 CR Argentina 1853 i-5 JR Armenia 1995 4 CR Bangladesh 1972/1991 3-5 JR Benin 1991 2 LCR Bolivia 1994 2 JR Bosnia-I lerzegovina 1995 4-5 CR Brazil 1988 2 JR/CR Bulgaria 1991 JR/CR Burkina-Faso 1991 4-5 LCR Cape Verde 1992 i-5 JR Central African 1994 3-5 CR Republic Chile 1981 LCR/LIR Colombia 1991 3-5 CR Croatia 1990 2-5 CR Czech Republic 1993 i-5 CR Dominican Republic 1996 2 JR Ecuador 1979 M JR/CR LI Salvador 1983 2-5 JR Estonia 1992 i-5 JR Ethiopia 1995 4 LCR Fiji 1990/1997 3-5 JR Ciabon 1991 4-5 LCR Georgia 1995 3-5 CR Ghana 1993 3 JR Greece 1975 2 CR Guatemala 1985 3-5 JR/CR Guinea-Bissau 19 84/1990 4-5 JR Guyana 1992. 2 JR Honduras 1982 2.5 LJR Hungary 1949/1990 i-5 CR Indonesia 1949 3-5 CRt Jordan 1952 4 LJR Korea 1988 2 CR Kyrgyz Republic 1993 5 CR Latvia 1922/1991 r-5 LCR Lesotho 1993 4 JR {continued) Judicial Review in New Democracies table i.i {continued) Form of Constitutional Review Year of (Key: CR = review by Constitution/ special body, JR = Last Major Freedom House review by courts; L — Amendment Rating 2000-01 scope of review or access Country {*= amendment only) (average) limited) Lithuania 1992. i-5 CR Macedonia 1991 3 CR Madagascar 1992. 3 CR Malawi 1994 2.5 JR Mali 1992 2-5 CR Moldova 1994 3 CR Mongolia 1992 i-5 CR Morocco 1971/1996 4-5 LCR Mozambique 1990 3-5 JR/CR Namibia 1990 2.5 JR Nepal 1990 3-5 JR Nicaragua 2000* 3 LJR Panama 1972/1994 i-5 JR Paraguay 1992 3-5 LJR Peru ^993 3 JR/CR Philippines 1987 i-5 JR Poland 1997 i-5 CR Portugal 1976 i JR/CR Rumania 1991 2 LCR Russia 4 LCR Sao Tome &C 1990 i-5 JR Principe Senegal 1991* 3-5 LCR Seychelles 1993 3 JR Sierra Leone 1991 4-5 JR Slovakia 1993 1 LCR Slovenia 1991 i-5 CR South Africa 1994 i-S JR/CR Spain 1978 i-5 LCR Suriname 1987 i-5 JR Taiwan 1947/1997 2 CR Tanzania 1992* 4 JR Thailand 1997 2.5 CR Ukraine 1996 4 CR Uruguay 1997 i JR Zambia 1991 4-5 LJR/LCR t A Constitutional Court was proposed for Indonesia in 2001. Source: Robert Maddex, Constitutions of the World (1995); United Stares Department of State, Human Rights Reports {1997); Freedom House, Freedom in the World. Dates of Constitutions were supplemented through the CIA Factbook at http:ffwww.tbeodora.com/wfb/. Note that a lower Freedom House rating indicates a higher level of democracy. Introduction 9 of the conventional definition of democracy (such as Cambodia (1993), Mozambique (1590), Ethiopia (1995), and Eritrea {1996)) contain provisions for constitutional review that remained unimplemented for several years after their passage. Like democracy itself, constitutionalism commands such normative power as an aspiration that it is invoked by regimes that make no pretense of submitting to constitutional control. The table shows that the centralized system of constitutional review, designed by Hans Kelsen for Austria and subsequently adopted in Italy and Germany, has been predominant in the recent wave of democratization.17 In contrast, a 1978 study of constitutions found that only 2.6% of constitutions included provision for a designated constitutional court with the power of judicial review.18 The centralized system reflected Kelsen's positivist jurisprudence, which incorporated a strict hierarchy of laws. Because constitutional rules are provided only to parliament and ordinary judges are subordinate to the parliament whose statutes they apply, only an extrajudicial organ could restrain the legislature.19 This extra-judicial organ was solely responsible for constitutional review. In new democracies, there may be particularly strong reasons to distrust a decentralized system.10 After all, the judiciary was typically trained, selected, and promoted under the previous regime. While some judges may have been closet liberals, there is little ability to ensure that these judges will wield power in a decentralized system. Furthermore, there may be significant popular distrust of the judiciary. Giving the ordinary judiciary the power of constitutional review risks dragging the prestige of the ,? Because designated constitutional courts in this tradition use adjudicative methods, we consider the term judicial review to apply to them as well as to systems of decentralized constitutional control. For a discussion of whether systems of abstract review are better characterized as engaging in a legislative or judicial process, see Stone, supra note 3, at 209-21. '* Henc van Maarseveen and Ger van der Tang, Written Constitutions (1978). ', Kelsen made his argument in Hans Kelsen, "La garantie jurisdictionnel de la constitution," 44 Revue de Droit Public 197 (1928). There, Kelsen characterized the Constitutional Court as a kind of negative legislature. For a discussion, see Elena Marino-Bianco, The Spanish Legal System 96-97 {1996) and Stone, supra note 3, at 118-30. 10 One hybrid variation is to adopt a single hierarchy of courts, with a supreme court that is exclusively charged with constitutional control. See, for example, Constitution of Yemen (1991), Article 124; Constitution of Estonia (1992), Article 152 (ordinary courts can refuse to apply an unconstitutional act, but only the National Court can declare it null and void); Constitution of Eritrea (1997), Article 49{i)(a). io Judicial Review in New Democracies Introduction ii constitution down to the level of the adjudicators in the public eye. Setting up a specialized body, by contrast, designates constitutional adjudication as a distinct, important function. So one explanation for the shift toward centralized review may be that widespread democratization has occurred and that decentralized review is particularly unattractive in new democracies. Accompanying the institutional spread of judicial review has been a normative turn in its favor in western scholarship on democratization. Conventional analysts of democracy are increasingly frustrated with the illiberal tendencies of democratically elected regimes and suggest that elections are not enough. Zakaria notes that "[t]he trouble with .. .winner-take-all systems is that, in most democratizing countries, the winner really does take all."21 Huntington notes that thirty-nine "electoral democracies" are deficient in protecting civil and political liberties.11 There is increasing concern for the constitutional elements of democracy, leading some analysts to distinguish between electoral democracy and liberal democracy, with the latter guaranteeing civil rights to a greater degree.13 Despite this fundamental shift in democratic practice and scholarship, there has been little inquiry into questions about the expansion of judicial review. We know very little about the conditions leading to the establishment of judicial review and about the successful exercise of judicial power, This is particularly acute with regard to non-European contexts, outside the core.14 With development banks, scholars, and politicians insisting on the importance of the rule of law as a universal component of "good governance,"15 the issue of judicial power merits more attention. We ought to know where judicial power comes from, how it develops in the crucial early stages of liberalization, and what political conditions support the expansion and development of judicial power. This study is an effort to examine these questions by focusing on the most visible and important 11 Fareed Zakaria, "The Rise of Illiberal Democracy," Foreign Aff. zz, 42 (November/December 1997). 11 Samuel Huntington, "After Twenty Years: The Future of the Third Wave," 8 J. Democracy 3, 10 (1997). 1J See Larry Diamond, "Is the Third Wave Over?" 7 /. Democracy 20 {1996); Huntington, supra note 2,2, at 3-11 {1997); Guiflermo O'Donnell, Horizontal Accountability in New Democracies, 9 J. Democracy 112,117 (1998); Andreas Schedler, Larry Diamond, and Marc F. Plattner, eds., The Self-Restraining State (1999). 14 C. Neal Tate, "Book Review of Paula Newberg's Judging the State: Courts and Constitutional Politics in Pakistan," 6 L. &Pol. Book Ret/. 109-12 {1996). 15 Thomas Carothers, "The Rule of Law Revival," 35 Foreign Aff. 23 (1997). institutional manifestation of judicial power, constitutional constraint by I courts. One theory argues that the spread of judicial power is a reflection of a broader extension of rights consciousness around the globe.16 This theory focuses on the demand for judicial protection of fundamental rights. The achievements of the human rights movement, the shift toward markets I that rely on notions of private property, and the spread of democracy all reflect the importance of ideas of fundamental rights. As rights consciousness has spread, the argument goes, so, too, does the importance of courts as the primary political actors with the mission to protect rights. I do not wish to contest the basic contours of this story. It would be difficult to deny that globalization and democratization have been accompanied by a dramatic spread in awareness of the importance of fundamental rights. What I wish to do is to supplement this story by ex- ■ amining specific contexts of judicial review, rather than simply accepting that a single uniform process is affecting the entire globe. In doing so, I will introduce considerations of power into the analysis, showing how politics shapes and is shaped by judicial review. If we were to accept the conven- 1 tional argument that a shift in consciousness is the key factor behind the spread of judicial review, it would follow that differences in the way \ judicial review is structured and operates could be explained by variations i in consciousness. My analysis shows that interests, as mobilized through * institutions and politics, are at least as important in dictating outcomes in new democracies as rights ideology. In doing so, I shift attention from the demand for institutions of judicial review to the supply side, asking " why it is that politicians would be interested in providing it. CONSTITUTIONALISM IN EAST ASIA I approach the problem of courts in new democracies by focusing on understudied constitutional contexts, particularly in East Asia. Asia has -I been called the home of illiberal democracy and represents perhaps the most difficult regional context for establishing the rule of law.17 Although Asia has deeply rooted indigenous legal and political traditions, the assumptions and orientation of these traditions are often contrasted with 16 See, for example, Heinz Klug, Constituting Democracy: Law, Globalism and South 1 Africa's Political Reconstruction (2000); Charles Epp, The Rights Revolution {1998). I " 17T>aniel Bell, David Brown, Kanishika Jayasuriya, and David Martin Jones, Towards Illiberal Democracy in Pacific Asia (1995); Huntington, supra note 22, at 10. ■ iz Judicial Review in New Democracies Introduction *3 the western ideals associated with constitutionalism. Confucianism, in particular, would seem to present a difficult cultural environment for the development of judicial review. In contrast with western legal traditions organized around the notion of the autonomous rights-bearing individual, the Imperial Chinese legal tradition is usually depicted as emphasizing social order over individual autonomy and responsibilities over rights/8 Law exists not to empower and protect individuals from the state, but as an instrument of governmental control. Any rights that do exist are granted by the state and may be retracted. Furthermore, power is conceived as indivisible in the Confucian world-view, flowing solely from the emperor, who is the center of the cosmolog-ical and political order. No human force can check the emperor's power if he enjoys the mandate of heaven.19 The notion of an intergovernmental check on the highest power is foreign to traditional Confucian thought. The emperor has "all-encompassing jurisdictional claims over the social-political life of the people."30 The only human constraint on the emperor's power is the duty of scholar-officials to remonstrate the leader where he errs (a practice that varied in its practical impact in different periods of Chinese history}.31 This unified conception of power is a very different one from that of modern constitutionalism with its distrust of concentrated authority.31 18 See the classic presentation of this position in Derk Bodde and Clarence Morris, Law in Imperial China (1967). 19 See, generally, Til Wei-ming, ed., Confucian Traditions in East Asian Modernity: Moral Education and Economic Culture in Japan and the Four Mini-Dragons {1936). 30 Benjamin Schwartz, "The Primacy of Political Order in East Asian Societies: Some Preliminary Generalizations," in Foundations and Limits of State Power in China 1 (Stuart Schram ed., 1987), quoted in A. King, "State Confucianism and Its Transformation in Taiwan," in Confucian Traditions in East Asian Modernity: Moral Education and Economic Culture in Japan and the Four Mini-Dragons 228, 230 (Tu Wei-ming, ed.,1996}. ■'' See Thomas Gold, "Factors in Taiwan's Democratic Transition," paper presented at Consolidating the Third Wave Democracies: Trends and Challenges, Institute for National Policy Research 11 (Taipei, Taiwan, August 17-30, 1995); Andrew Nathan, "China's Constitutionalist Option," 7 J. Democracy 43 (1996). 31 See, for example, R. Fox, "Confucian and Communitarian Responses to Liberal Democracy," 59 J.Pol. 561, 572. (1997); Daniel Bell, East Meets West: Human Rights and Democracy in East Asia (2000). Of course, Confucianism offers a more general critique of law as a means of social ordering. For example, the Analects express disdain toward "guiding the people by edicts and keeping them in line with punishments." The classical opposition between Fa and Li is discussed in virtually every account of Chinese Saw. See, for example, Bodde and Morris, supra note 2.8; Janet E. Ainsworth, "Categories and Culture: On the 'Rectification of Names' in To the extent that these traditional ideas about law and power continue to operate in East Asia (a highly contested question), they would seem to pose a challenge to the establishment of judicial power. Some authors have pointed to modern law as a reflection of a particularly western configuration of values and ideals.33 A set of strong, secular, autonomous legal institutions capable of checking legislative and executive authority took centuries to develop in Western Europe.34 With much less experience with the legal machinery of the modern nation state and with a legacy of strong and concentrated political authority, similar institutional development would seem to be a difficult proposition in Asia. Despite increasing public scrutiny and pressure from foreign donors and international financial organizations, reciprocity and personalism remain central to many descriptions of East and Southeast Asian politics and economies.35 Many scholars and professionals remain skeptical about the possibility of the rule of law taking root, even after the economic crisis of 1997-98 led to political reforms in some countries in the region.36 This discussion echoes the now decade-old debates over the question of whether Asian values are incompatible with western notions of human rights and democracy.37 Several leaders in the region have argued Comparative Law," 82 Cornell L. Rev. 19 (1996); S. Lubman ed., China's Legal Reforms iiyytV); Ralph Folsom, John Minan, and Lee Ann Otto, Law and Politics in the People's Republic of China 13-18 (1992). Li refers to morality, custom, and propri- ; ety, while Fa is usually translated as criminal law, but refers more broadly to formal rules backed by sanctions. " Roberto linger, Law in Modern Society: Toward a Criticism of Social Theory (1976); see also Samuel Huntington, "After Twenty Years: The Future of the Third Wave," S J. Democracy 3 (1997). 34 Harold Berman, Law and Revolution (1985). " On donor efforts, see the Bulletin on Law and Policy Reform maintained by the Asian Development Bank at http://www.adb.orgJdocuments/periodicals/lawJbulletin/. On personalism, see, for example, David I. Steinberg, "The Republic of Korea: Piuralizing Politics," in Politics in Developing Countries: Comparing Experiences with Democracy 396 (Larry Diamond et al., eds., 1995). Sec Lester Thurow, "Asia: The Collapse and the Cure," N.Y. Review of Books, February 5, 199S, at %%. See also Enrique Carrasco, "Rhetoric, Race and the Asian Financial Crisis," L.A. Times, January 1,1998; Enrique Carrasco, Tough Sanctions: The Asian Crisis and New Colonialism," Chi. Trib.t January 3, 1998; H. Patrick Glenn, Legal Traditions of the World 297 (2000), For contributions to the debate on "Asian Values," see William Theodore de Bary, Asian Values and Human Rights: A Confucian Communitarian Perspective (2000); v: Kishore Mahbubani, Can Asians Think (1998); Joanne R. Bauer and Daniel Bell, <■'<'*•> The East Asian Challenge for Human Riglns (1999); and Michael C. Davis, : ; "Constitutionalism and Political Culture: The Debate over Human Rights and Asian S Values," 11 Ham Hum. Rts. L.J. 109 (1998). 14 Judicial Review in New Democracies Introduction 15 that Asian political traditions, especially the Confucian legacy, are fundamentally incompatible with, and offer an alternative to, western-style liberal democracy. The western emphasis on civil and political rights, it is asserted, does not take into account an alleged Asian preference for economic well-being and communal goods. Asians prefer order over freedom, hierarchy over equality, and harmony over conflict. Hence, authoritarian governments in Asia actually reflect different cultural values that constrain democratic and constitutional development in the Chinese and more broadly Asian tradition.38 Others have challenged these views as simplistic and have called into question the cultural determinism that underlies the Asian values position.39 The notion that Asian values are distinct presupposes an orientalist dualism between a monolithic Asian tradition of hierarchy and a western tradition of individualism. This dualism does justice to neither tradition, ignoring individualistic and liberal elements in the Confucian tradition as well as collective, hierarchical, and conflict-avoiding elements in the western tradition.40 In terms of thinking about the development of particular institutions, one problem with using culture as an explanatory category is that a tradition such as Confucianism is so broad it contains elements that might either support or hinder any institution under consideration. For example, Confucianism, once thought to be a hindrance to modernization, has in recent years been used to explain economic success in Asia.41 Similarly, one might argue that certain aspects of the Imperial Chinese tradition, such as government by elite generalists, are compatible with judicial review.41 }R Samuel Huntington, The Clash of Civilizations and the Remaking of World Order (1996), Lee Teng-hui's reflection on the contribution of Chinese culture to Taiwan's democratization is found in Lee Teng-hui, "Chinese Culture and Political Renewal," 6 ]. Detnoc. 3 (1995). y) See Davis, supra note 37, and Randall Peerenboom, "Answering the Bell: Round Two of the Asian Values Debate," 42 Korea Journal 194 (1002). 40 William Theodore de Bary, The Liberal Tradition in China (1983); Tatsuo Inoue, " Critical Perspectives on the 'Asian Va lues' De ba te," in The East Asian Challenge for Human Rights 27, 37-45 (Joanne Bauer and Daniel Bell eds., 1999). 41 See, for example, Gary Hamilton and Kao Cheng-shu, "Max Weber and the Analysis of the Asian Industrialization," Working Paper No. 2, University of California, Davis Research Program in East Asian Culture and Development (1986); Benjamin A. Elman, "Confucianism and Modernization: A Reevaluation," in Confucianism and Modernization: A Symposium 1 (Joseph P. L. Jiang, ed., 1987); Cal Clark and K. C. Roy, Comparing Development Patterns in Asia 61-93 (1997). 41 See Tom Ginsburg, "Confucian Constitutionalism? The Emergence of Judicial Review in Korea and Taiwan," 27 Law and Social Inquiry 763 (2001). The point is that, because of their very breadth, cultural and legal tradi- I (ions do not dictate outcomes in predictable ways. The Confucian legacy as: conventionally interpreted poses barriers to the emergence of constitu- I tionalism and judicial review of legislation in Chinese society. But cultural 1 and legal traditions are flexible and dynamic and can provide rationales i for a wide range of political institutions.43 This suggests the difficulty of I building a workable theory of the adoption and function of judicial review I on cultural factors. I This study will explain the emergence of judicial review as a result of i institutions and politics, rather than culture. By focusing on the spread i and transfer of a central practice of constitutional democracy, judicial !| review, outside of its core areas in the United States and later Western I Europe, this study is an effort to broaden the empirical and theoretical I base of comparative constitutional law. The core areas have been at the * center of comparative projects documenting the vast expansion of judicial \ review in recent decades.44 Studies of nonwestern countries have been far i less frequent. By demonstrating that judicial review can function outside : the core, this study will challenge culturally deterministic accounts of the rule of law and judicial power. J AMERICAN EXCEPTIONALISM? I How ought one approach the study of judicial review in countries '■ beyond the core? There may be several dangers in treating the American i experience as the benchmark against which other countries' practices are I measured. One way that American constitutionalism is distinctive is the J fact that there is no explicit constitutional provision for judicial review in i the American constitution. This has consequences that may not apply to ' other systems, including the embedding of the constitution into ordinary ; law.4' (Technically, there is a distinction between judicial review, in which I ordinary judges play the role of constitutional check, and constitutional : review, in which the function is given to specialized judges or political i actors. This study uses the terms interchangeably.) The primary role of the i : " Cf. Huntington, supra note 38. See de Bary, supra note 40; William Theodore de Bary, í "The ^Constitutional Tradition' in China," 9/. Asian L. (1995); Davis, supra note 37; I Michael C. Davis, "The Price of Rights: Constitutionalism and East Asian Economic I Development," 20 Hum. Rts. Q. 303-37 (1998). See also Michael C. Davis, ed., . Human Rights and Chinese Values: Legal, Philosophical and Political Perspectives {1995). ; 44 The Global Expansion of Judicial Power, supra note 7. 'j 45 Stephen Griffin, American Constitutionalism: From Theory to Politics (1996). i6 Judicial Review in New Democracies Introduction 17 United States federal judiciary is resolving disputes among private parties, and it need not exercise judicial review to do so. Because judicial review is incidental to the basic functions of the courts, the legitimacy of judicial review is always in doubt. Scholars of American constitutionalism have responded by focusing almost exclusively on normative issues of judicial legitimacy rather than positive issues of judicial power. But these issues may be less important in contexts where there is a clear constitutional moment and a designated court whose only role is to safeguard the constitution. Another risk of focusing exclusively on the American origins of judicial review is that one might overcharacterize the insular, purely national character of the practice. American courts are notoriously reluctant to acknowledge the normative or legal importance of other countries' case-law or international instruments.46 Yet, in the international context, domestic practices of judicial review draw extensively on international treaties, other countries' case-law, and normative rhetoric from other national experiences. The danger of beginning with the American experience is missing the significant international dimension of contemporary judicial review. The rule of law ideal has strongly universalist overtones, and courts may invoke their fraternal duty to defend it in specific cases. This often involves an examination of how other judiciaries have dealt with a particular problem. This practice of borrowing has long been a feature of the common law tradition, but also occurs in civil law jurisdictions.47 Citing cases from other contexts is a strategy of legitimation for courts.48 46 See, for example, Sei Fujii v. California, 38 Cal. nd 718, 242 P. id 617 (1952.). But see United States v. Then, 56 F. 3d 464, 469 (2d Cir, 1995) (Calabresi, J., concurring). 47 See, for example, T. Koopmans, "Comparative Law and the Courts," 45 Am. ]. Comp. L. 545, 550-55 (1996); Anne-Marie Slaughter, "The Real New World Order," 76 Foreign Aff. 183 (1997) (arguing that such "transgovernmentalism" by both judges and bureaucrats is the primary response to globalization, and represents the future of governance in an era when the traditional territorial state seems less able to cope with growing regulatory demands). Another form of judicial use of comparative law involves looking to practices consistent with notions of a "free and democratic society," an approach reflected in Israeli Supreme Court practice as well as in the case-law of the European Court of Human Rights. The European Court of Justice itself engages in comparative law exercises under Article 2.87 (formerly Article 215) related to noncontractual liability of the community, where it must compensate based on principles common to the laws of the member states. See T, Koopmans, supra. 48 See, for example, Herman Schwartz, "The New Courts: An Overview," 2 E. Eur. Const. Reu. 28 (1993). Finally, the origin of the practice in the United States may lead us to look for Marbury-type "grand cases" wherein the court asserts its power to overrule political authorities.49 The danger is that a grand case is not the only way judicial review can be established. Beginning with an American orientation may lead us in the wrong direction by focusing our attention on the search for nonexistent "grand cases" in new democracies. This approach may misread Marbury, which after all did not include any command to a political branch.50 More accurately, observers looking for "grand cases" that establish institutions of judicial review have in mind Brown v. Board of Education, where the Supreme Court overturned the American caste system with a single blow.51 But Brown is another highly atypical case. First, it explicitly overrules a precedent in contrast with the usual characterization of common law courts. Second, Brown's rhetoric is primarily moral rather than legal. Only in the sense that the Warren Court was highly conscious of the political ramifications of its decision was Brown a "normal" constitutional case. And it is precisely here that the U.S. experience is helpful. For studies of courts in new democracies will have to consider the delicate political contexts in which they operate. Just as the American courts are concerned about securing compliance with their decisions, so courts in new democracies face the same fundamental political problem: how to convince the losing party to abide with their decisions.51 APPROACH AND PLAN OF THE BOOK This book addresses three questions concerning judicial review. First, why is it that countries adopt judicial review during periods of democratization and constitutional design? After all, if judicial review is undemocratic as scores of scholars have argued, it should be unattractive to newly empowered democrats. Second, what explains variation in the design and powers of new constitutional courts? One might think that there would be little variation in the design of new courts across different countries, but in fact there is variation, as Table 1.1 Marbury v. Madison, 5 U.S. (1 Cranch) 137 (1803). 50 See Michael J. Klarman, "How Great Were the 'Great' Marshall Court Decisions?" 87 Va. L. Rev. mi (2,001). 5' Brown v. Board of Education, 347 U.S. 483 (1954) (overruling Plessy u. Ferguson, 165 U.S. 537(189«)). 51 Martin Shapiro makes a similar argument for courts in all times and places. See Martin Shapiro, Courts: A Comparative and Political Analysis (1981). , ,, f'fiO } 18 Judicial Review in New Democracies suggests. Third, why is it that some constitutional courts exercise the power of judicial review more aggressively than others? Variation in institutional design plays a role, but there may be other more important factors. The answer I offer to all three questions is that politics matters. I begin by treating the first two questions together: Why is it that judicial review is adopted in the democratic constitution, and why does it take the form it does? I consider why judicial review makes sense from the point of view of those who write the constitution. The answer has to do with the time horizons of those politicians drafting the constitution. If they foresee themselves in power after the constitution is passed, they are likely to design institutions that will allow them to govern without encumbrance. On the other hand, if they foresee themselves losing in postconstitutional elections, they may seek to entrench judicial review as a form of political insurance. Even if they lose the election, they will be able to have some access to a forum in which to challenge the legislature. I argue that the particular institutional design of the constitutional court will tend to reflect the interests of powerful politicians at the time of drafting, with optimistic politicians preferring less vigorous and powerful courts so they can govern without constraint. The third question concerns the operation of the system of judicial review after it has been established. Here I focus on the decisions by judges, but also on the political constraints in which they operate. I show that the more diffused politics are, the more space courts have in which to operate. In contrast, where a dominant disciplined political party holds power, judicial review is more constrained. Drawing a distinction between systems with active judicial review and those where it appears relatively dormant, we can see a clear correlation between active review and diffused politics. The second half of the book consists of historical analysis of the emergence of judicial review in three transitional political systems. The objective here is both descriptive and theoretical. Descriptively, I present data on the development of judicial review in unlikely and understudied contexts. Theoretically, my goal is to use the studies to test some of the propositions developed in the first part of the book and to demonstrate the utility of the theoretical framework for understanding the exercise of judicial power in new democracies. The three cases selected for full study are Korea, Taiwan, and Mongolia. These cases are particularly useful given the influence of Introduction 19 i Imperial Chinese legal institutions 011 all of them." Judicial review has •1 grown in all three environments in recent years as democratization has I proceeded, a significant result given the supposed aversion of Asian so- cieties to legal ordering. Although this selection of cases may be termed intraregional because all three countries are in Northeast Asia, the three represent very different environments with regard to a number of other important independent variables that might plausibly affect the develop-; mcnt of judicial review. Of special importance are political and institu- tional variations. The book concludes with a comparative analysis of the three cases and argues that political and institutional structure, rather than cultural factors, are the keys to understanding the development of judicial review 1 in new democracies. The evidence in the case studies is consistent with the political theory of constitutional court design and performance offered in the first part of the book. Political uncertainty leads to the adoption I of judicial review as a form of insurance to protect the constitutional bargain. Political diffusion after the bargain is concluded allows courts to exercise greater power. By increasing uncertainty, democratization leads to greater demand for judicial review; the extent of political diffusion determines how successful courts can be in asserting the power. " Two of the case studies, Taiwan and Korea, are conventionally viewed as Confucian societies (with Confucian influence even stronger in Korea than in the Chinese society on Taiwan). Although it was a part of the imperial Chinese system that promoted Confucianism as official ideology, Confucian influence on Taiwan was probably less pervasive than on the mainland. After 1895, Chinese Confucian influence was subordinated under Japanese rule to State Shinto ideology and growing militarism. Some scholars therefore argue that Confucian influence was minimal on Taiwan. See, for example, Lucien Pye, Asian Power and Politics: The Cultural Dimensions of Authority (1985). Others, including a prominent former grand justice, assert r that Taiwan is a Confucian society. See Herbert Han-pao Ma, The Rule of Law in a Contemporary Confucian Society: A Reinterpretation, presentation to Harvard Law School's East Asian Legal Studies Program (spring 1998). It is difficult to reconcile :; these two views. As the issue of Chinese and Confucian influences touches on the question of national identity, it is subject to intense contestation within Taiwan. Tn any case, the precise level of Confucian influence on Taiwan is not empirically I verifiable. Nevertheless, as Taiwan is universally acknowledged to be a part of the f ::■ "greater Chinese cultural system," it seems reasonable to consider the possible ef- : ::fects of the dominant Chinese legal and political philosophy on developments there. Mongolia, by contrast, has a strong historical aversion to Chinese culture and is not conventionally included in the Confucian world. Mongolia was, however, a former ;;; part of the Manchu Empire, which ruled China and has a long history of interaction I with Chinese culture. All three cases, then, were historically influenced by Imperial I Chinese legal institutions to varying degrees. zo Judicial Review in New Democracies Ultimately, an examination of the development of judicial power in Asia can help us understand one of the most important questions of sociolegai studies, namely how a political system can transform itself from one governed by personalistic forms of authority toward one in which the rule of law prevails. In a region where prevailing traditions have emphasized an instrumental approach to law, the emergence of law as a constraint on political authority is a remarkable development with potentially broader implications. Cultural and legal traditions are not insurmountable barriers to institutions of liberal democracy. While this is good news for advocates of liberal democracy, the account offered here also suggests limits on the ability of outside intervention to facilitate institutional change. How does judicial power emerge? The answer suggested by this book is that domestic political diffusion is a necessary condition for the development of judicial power. Why Judicial Review? Modern scholarship on judicial review begins with the countermajoritar-ian difficulty.1 This famous problem focuses on the propriety of unelected judges, who lack democratic legitimacy, overturning duly enacted decisions of democratic assemblies. This normative challenge has been bolstered by theorists of democracy who argue that judicial power comes at the expense of representative institutions.1 Judicial review, from these perspectives, is not only unnecessary for democracy, but in fact suspect. In the face of these critiques, most legal scholars discussing judicial review have self-consciously adopted a defensive tone at the outset, trying to justify the role of courts in terms of democratic theory. The conventional move to solve the problem of courts in democratic theory is to celebrate the role of judicial review in democracy as a check on majority power. Judicial review in this view can facilitate the democratic process by clearing out obstacles to its advancements Such obstacles can emerge, for example, through majority impositions on the electoral process: It may be in the narrow self-interest of permanent majorities to disenfranchise political minorities, who then have no recourse through ordinary legislative processes. In such instances of systemic failure, the Originally identified in Alexander Bickel, The Least Dangerous Branch: The Supreme : Court at the Bar of American Politics (id ed., 1986). See Barry Friedman, "A History of the Countermajoritarian Difficulty, Part One: The Road to Judicial Supremacy," 73 N.Y.U. L. Rev. 333 (1998) for a history of the problem. Most prominently, Robert Dahl, Democracy and Its Critics 188 (1989). F.ly is the most well-known proponent of this view, elaborating on footnote 4 of United States v. Carolene Products Co., 304 U.S. 144, 152-53 n.4 (1938). See John Hart Ely, Democracy and Distrust (1980). 2.2 Judicial Review in New Democracies | llll courts can clear the channels of the political process by striking statutes. By serving as a countermajoritarian institution, judicial review can ensure that minorities remain part of the system, bolster legitimacy, and save democracy from itself. Several scholars have recently articulated a more majoritarian view of constitutionalism that emphasizes the need to empower rather than restrict majoritarian processes.4 Democracy is at bottom about deliber- I ation and debate, they argue, and the function of a constitution is both )\ to set boundaries for and facilitate this debate. The function of judicial review in these accounts is to provide another perspective on questionable policies. Courts are not the ultimate determiner of constitutionality but merely another governmental institution that helps deliberation take place through institutional dialogues with other branches of government. Judges, because of their special training and selection, can ruminate on fundamental principles of the democratic system. Although normatively attractive, both of these accounts raise a fundamental difficulty, namely how it is that judicial review is adopted in the constitution in the first place. After all, why would a political majority adopt an institution that constrains itself in policy making? And why would it rely on judges to undertake the task of constraint? The recent wave of constitution drafting around the globe invites inquiry into the political logic of judicial review, beginning with the fundamental question of why it is adopted. JUDICIAL REVIEW AS INSURANCE Why would constitutional drafters choose to include provisions for judicial review in the constitutional text? To answer this question, we must begin with foundational questions about the constitution and whose interests it reflects. Since Locke, constitutional theorists have thought of the constitution as a contract between citizens and government. We imagine that citizens empower a state and develop a system of constitutional democracy as a mechanism to satisfy individual preferences through 4 See Cass Sunstein, Designing Democracy; What Constitutions Do (2001); Jed Rubenfeld, Freedom and Time: A Theory of Constitutional Self-Government (2001); Bruce Ackerman, Wethe People: Transformations (1998); see also Amy Gutmann and Dennis Thompson, Democracy and Disagreement {1966); Jürgen Habermas, Between Facts and Norms: Contributions to a Discourse Theory of Law and Democracy (1996); Carlos Santiago Nino, The Constitution of Deliberative Democracy (1995). Why Judicial Review? 23 collective action. This device enables us to ask normative questions about what institutions most approximate the good society, what citizens might have chosen behind a veil of ignorance, or what institutions best help the citizenry resolve collective-action dilemmas.5 The contractarian perspective analogizes the democratic constitutional scheme to a series of principal-agent relationships wherein the people rely on politicians as agents to satisfy their collective demands. If the people are the principal on whose behalf the constitution is created, constitutional adjudication should reflect the need to monitor their political agents. Judicial review of legislation exists to prevent politicians from reneging on the founding bargain with citizens. This contractarian perspective is normative rather than positive, and it is open to criticism on empirical grounds. There are numerous reasons to be suspicious that actual constitutional design reflects the interests of citizens. Most obviously, constitutional design would only reflect citizen interest if the designer-politicians who actually draft and agree on the constitutional text were themselves pure agents of those citizens. But that can hardly be the case because citizens are subject to collective-action problems that prevent them from organizing to monitor constitutional debates. Under such circumstances, politicians who draft the constitution can seek to design institutions that benefit themselves, their institutions, or their interests narrowly rather than those of citizens more broadly. Much empirical evidence supports the assertion that constitution making is dominated by short-term interests of the designers rather than the long-term interests of the citizenry/ In light of the agency problem of constitutional design, we must ask why self-interested politicians would design a system of judicial review. It is not sufficient to describe constitutional review as a device to protect citizens from future politicians without explaining why it serves the interests of present politicians who serve as a veto gate for the constitution. Although constitutional designers are subject to the same constraints of bounded rationality as everyone else, there are reasons for assuming that they consider their institutional choices carefully. Constitutional choices typically have a great impact on subsequent political outcomes, so there '■.James Buchanan, The Limits of Liberty (1975); Robert Cooter, The Strategic Constitution 243 (2000); Dennis Mueller, Constitutional Democracy 61-67 (J99^)i J0'111 Rawls, A Theory of Justice (1973). 6 Stefan Voigt, "Positive Constitutional Economics: A Survey," 90 Public Choice 11, 26 (1997); Mueller, Constitutional Democracy, supra note 5,at 316-18; Jon Elster, "Forces and Mechanisms in the Constitution-Making Process," 45 Duke L.J. 364 (1995). 1 24 Judicial Review in New Democracies j are strong pressures on designers to choose institutions that wili benefit their constituencies in the future. I argue that the answer to the question of why self-interested politicians would design a system of judicial review depends on the prospective | power positions of constitutional designers in postconstitutional govern- L ment. Assume that constitutional drafters are themselves politicians, who are interested in governing after the adoption of a new constitution. It follows that they will seek to design institutions that maximize their ability to govern under the new constitutional order. The key factor from the -rj drafters' perspective is the uncertainty of the future political configuration j at the time of constitutional drafting.7 Consider two extreme constitutional scenarios. Where a single party believes it is likely to hold on to political power, it has little incentive to set up a neutral arbiter to resolve disputes about constitutional meaning. It would rather retain the flexibility to dictate outcomes without constitutional constraint. Flexibility allows policy change and maximum exercise of power. The absence of independent judicial review institutions under j authoritarian constitutions reflects this desire to maintain the exclusive role of constitutional interpretation. By contrast, where many political forces are vying for power, no party r can have confidence that it is likely to continue to win future elections. A constitutional design allowing unlimited flexibility for electoral 7 This theory is related to J. Mark Ramseyer's work on judicial independence. | See J. Mark Ramseyer, "The Puzzling (In)Dependence of Courts: A Comparative ' Approach," 23 /. Leg. Stud. -jit. (1994). Drawing on evidence from Japan and j the United States, Ramseyer suggests that independent courts will be supported by - \ politicians where they believe two conditions exist: {1) Continuing elections are likely, and {2) the ruling politicians are likely to lose a future election. In such an instance, it is in the interest of the ruling party to create independent courts to protect its policy preferences that are enacted as laws. The courts serve as the agents of politicians who are now out of office. Ramseyer's first condition is constitutional; the second is related to the character of the democracy. Where either one of the conditions does not hold, a ruling party will not choose independent courts that can only hinder | that party's ability to act decisively. See also William Landes and Richard Posner, "The Independent Judiciary in an Interest-Group Perspective," 18 J. L. & Econ. 875 (1975). In Ramseyer's presentation, this decision's impact on the survival of the constitutional regime is exogenous to the model. The party making the choice to in- ! stitutionalize independent courts makes a judgment about continuing elections and based on that judgment chooses to make courts independent or not. It is possible, even likely in the context of fragile new democracies, that such a decision will itself affect the probabilities of continued elections and maintenance of the constitutional order. Active systems of judicial review are not often associated with democratic failure. Why Judicial Review? 15 winners, as in the model of parliamentary sovereignty, is much less attractive in a politically diffused setting than in a setting wherein a single party holds sway. While prospective governing parties would [ike flexibility, prospective opposition parties value limited government. Opposition parties want to minimize their maximum losses. They also value an alternative forum in which to challenge the policies of the majority, because they do not expect to win in the legislature. When the designers' party cannot count on reelection and may end up an opposition party, it will prefer a judicial forum in which to challenge its newly empowered political adversaries. These considerations lead to a general prediction about judicial power and constitutional rights: Explicit constitutional power of and access to judicial review will be greater where political forces are diffused than where a single dominant party exists at the time of constitutional design. This is because dominant parties are likely to anticipate continued success in postconstitutional elections and therefore to prefer majoritarian institutions. Where political forces are deadlocked, or scattered, no party can confidently predict that it will be able to win postconstitutional elections. Because there are no parties that will be confident in their ability to win, all parties will prefer to limit the majority and therefore will value minoritarian institutions such as judicial review. The key factor in explaining variation in iheextcnt of judicial power in constitutional design is the structure of the party system and the configuration of political forces at the time of constitutional drafting. I call this the insurance model of judicial review. By serving as an alternative forum in which to challenge government action, judicial review provides a form of insurance to prospective electoral losers during the constitutional bargain. Just as the presence of insurance markets lowers the: risks of contracting, and therefore allows contracts to be concluded that; otherwise would be too risky, so the possibility of judicial review lowers the risks of constitution making to those drafters who believe they may not win power. Judicial review thus helps to conclude constitutional bargains that might otherwise fail. Let us consider three objections to the insurance theory. One might argue that other minoritarian devices exist that can substitute for judicial review, such as difficult procedures for constitutional amendment, bicameralism, and proportional representation.8 Because provisions about For some evidence of these propositions, see Elster, supra note 6, at 377-82; Jon Eister, "Limiting Majority Rule: Alternatives to Judicial Review in the Revolutionary Epoch," in Constitutional Justice under Old Constitutions (Eivind Smith, ed., 1995). 2.6 Judicial Review in New Democracies Why Judicial Review? 27 judicial review are embedded in the larger constitutional bargaining process, we cannot predict a perfect correlation between strong parties and weak judicial review. Judicial review can be traded off against other mi-noritarian institutions in the larger constitutional bargain. The precise configuration of the constitutional bargain will reflect tradeoffs across a number of different substantive and institutional issues. Nevertheless, in the proverbial state of other things being equal, more-optimistic parties will prefer less judicial constraint. Why might judicial review be an attractive minoritarian institution for designers? As a form of insurance, judicial review is relatively inexpensive because it can be exercised by a court staffed with a few members. While a court, like other branches of government, may seek to expand its budget, it is certainly cheaper to run than, say, a second house of a legislature that could serve to protect the constitutional bargain because of a different representational system. Thus, judicial review, to the extent it serves the interests of the founders in constraining majorities, is cheap minoritarianism. This might explain why it is that judicial review may have been adopted more universally than other minoritarian institutions that could serve the interests of prospective losers. Another reason judicial review may be a particularly desirable form of insurance is the international context of constitutional drafting. Constitutional designers do not operate in a vacuum, and there is a growing international norm that constitutions include some sort of institution to exercise constitutional oversight. The success of the institution elsewhere enhances its reputation. The formal submission of political power to the dictates of the rule of law is one of a package of institutions designed to express the break with the past. To a certain extent, like legislatures and presidencies, judicial review forms part of the "script" of modernity and is adopted for reasons of both external legitimacy and internal political logic.3 Although other institutions might provide equally good protection for minorities, judicial review has a reputation for effective minoritarianism that makes designers particularly likely to adopt it. When designers must choose among alternative institutions that address a particular problem, one solution can stand out and become focal, even if other solutions would be effective substitutes.10 The international context helps make 9 John W. Meyer, John Boli, George M. Thomas, and Francisco O. Ramirez, "World Society and the Nation State," 103 Am J. Soc. 144 (1997}. 10 On focal points, see Thomas C. Schelling, The Strategy of Conflict (i960). See also Richard McAdams, "A Focal Point Theory of Expressive Law," 86 Vir. L. Rev. 1649 (2000). judicial review a particularly focal solution to the problem of constitutional protection of political minorities. Besides providing positive examples, foreign countries also play a more direct role in the spread of judicial review. By providing technical assistance to constitution-drafting exercises, foreign assistance programs subsidize the cost of evaluating different institutions. In some cases, international actors can also play an indirect or even direct role in the constitutional bargaining process. The extreme case is that of the Japanese constitution, which included provisions for American-style judicial review because it was imposed by the occupation authorities. All these factors help to make judicial review an attractive form of insurance, but they do not dictate institutional choices. If we hold these international factors constant, weak judicial review is unlikely where political forces are evenly balanced; conversely, dominant parties are less hkely to desire strong judicial review. Furthermore, foreign observers of constitutional drafting processes may pay scarce attention to the details of institutional design. The inclusion of a constitutional court may satisfy foreign interests concerned with rights protection and controlling legislative power, but the institutional details of standing law may be off the screen of foreign observers and hence susceptible to manipulation by local political actors.11 Thus, the insurance theory still has a large explanatory role to play. A second objection to the insurance theory concerns the use of the insurance analogy, which strictly speaking implies risk aversion. A risk-averse party is one that would prefer, for example, a sure chance to govern for one year over a one-in-four chance to governing for four years. White such risk-averse parties are sure to value judicial review, because they know they will be out of power, the assumption that parties are risk averse is not necessary for the theory as I have articulated it. All that is necessary is that there is intertemporal uncertainty between the time institutions are chosen and the time they will actually begin to operate. I use the term insurance in this looser sense. Certain other technical elements of insurance may in fact fit the analogy to judicial review, but they are not necessary for the theory,11 " Jodi S. Finkel, "The Implementation of Judicial Reform in Peru in the 1990s," paper presented at the American Political Science Association Meeting, San Francisco, : August 2001. 11 For example, one might argue that the effective insurance of judicial review can create a kind of moral hazard for political party members, who may work less hard to win the next election because they do not fear the consequences of loss as severely z8 Judicial Review in New Democracies Why Judicial Review? 2-9 A third potential objection concerns signaling. Even a dominant party controlling constitutional design may wish to provide for a system of judicial review as a way of signaling its serious intention to abide by the constitution. This illustrates a competing rationale for judicial review that I call the "commitment theory." Whatever the problems they seek to solve, constitutional drafters face the challenge of making their commitments credible.13 Judicial review is an answer to problems of constitutional commitment. By setting up an independent institution to adjudicate disputes arising under the constitution, the drafters signal that they are serious about upholding their promises. Judicial review is thus a form of self-binding on the part of constitutional designers.14 Of course, this signal of self-binding is only effective to the extent that the threat of independent judicial review is itself credible: The court must have both power and insulation from political control. Although they are similar in many respects, this "commitment" rationale can be contrasted with the insurance theory in terms of its empirical implications. Whereas the insurance theory predicts less powerful institutions of judicial review with a dominant party, the commitment theory might predict more powerful institutions of judicial review with a dominant party. This is because demand by smaller parties for commitment during constitutional design will increase with party strength of the dominant party. On the other hand, if a dominant party is strong enough, it will be able to dictate the constitution without concern for the smaller parties' desires. To illustrate, imagine a constitutional bargain among two parties where a two-thirds majority is needed to pass a constitution. The first column in Figure i.i represents the relative strengths of the two parties. Under the commitment theory, the level of predicted judicial review rises as one party becomes stronger, so long as it needs the cooperation of the weaker party to pass a constitution. This is reflected in the move from Low to Medium along the first two rows. Once the dominant party has a secure enough majority to dictate a constitution, there is no more need to accede to the minority by including judicial review in the constitution. Under the as a party without the protections of judicial review. Thanks to Eric Rasmusen for this point. 13 Stephen Holmes, "Precommitment and the Paradox of Democracy," in Constitutionalism and Democracy (Jon Elster and Rune Slagstad, eds„ 1988); Landes and Posner, supra note 7. 14 See, generally, Jon Elster, Ulysses Unbound 88-174 {2000). "Parry Configuration of Scats in Constitutional Assembly Predicted Level of Judicial Review Power under Commitment Theory Predicted Level of Judicial Review Power under Insurance Theory "50^50 Low High 60-40 Medium Medium 70-30 Low Low figure i.i. Competing Theories of Institutional Design insurance theory, the predicted level of judicial review declines consistently as one party becomes stronger. The commitment theory focuses on the stronger party in constitutional negotiations, while the insurance theory focuses on the weaker, prospectively losing parties. Other than the different empirical implications for situations of evenly divided parties, the commitment theory is not really very different from the insurance theory. Both theories have elements of commitment, in that a truly dominant party that can dictate a constitution has little need for any form of judicial review. Only when cooperation is required does judicial review enter the picture. In short, some of the objections to the insurance rationale really provide supplementary theories that complement rather than replace the insurance framework. Other minoritarian institutions may indeed render judicial review less attractive from the perspective of constitutional drafters; but judicial review is a relatively cheap form of minoritarianism so we should see it included in many constitutional bargains. The international context is also an important consideration and actually helps make judicial review a kind of focal point for drafters concerned about minoritarian interests. Ultimately, domestic politicians form a veto gate, so we should also expect their interests to be reflected in the details of institutional design. The commitment rationale is in part an alternative theory in that it predicts greater constitutional constraint where designers are stronger, up to a point. It remains to be seen whether this is a superior account of the design of judicial review. Although the insurance theory is clearly minoritarian in character, it need not rely on a view that courts will always serve minority interests or that courts will always be effective when they do so. Recent work on 30 Judicial Review in New Democracies deliberative democracy ties in with literature emphasizing that judicial decisions are not in fact final, but rather involve a kind of dialogue with political branches of government. The court is one actor among several that participate in the governmental conversation. The crucial point that all these theories share is that the court provides an alternative forum to the legislature and can thus allow the articulation of views that would otherwise not be heard. Whether or not this increases the quality of democracy, as the deliberative theorists argue, is not our concern here. What is important for present purposes is that, as a positive matter, judicial review potentially expands the range of voices to include political losers. Two fora are always better than one for the party that loses in the legislature. Thus, the insurance rationale is compatible with a variety of normative theories. INSURANCE IN NEW DEMOCRACIES Now let us consider more carefully judicial review in the context of new democracies and political transitions. There are two features of such contexts that contrast with more established democratic regimes. First, future political outcomes are more uncertain relative to autocracy. The presence of electoral competition means that even the most dominant and popular party faces a relatively higher chance of losing power than it would under a one-parry system. Information on future outcomes is more difficult to assess. There is ample empirical evidence that constitutional designers sometimes misjudge the probabilities of their electoral success.15 Second, by definition the institutional structure of the political system is in a period of transition, of movement from one equilibrium toward another. I do not mean to suggest in teleological fashion that all democratic transitions lead to the same place. There is a wide range of institutional configurations that are possible even within the category of democracies. Indeed, this range itself is a source of uncertainty. As institutional structure changes, parties are even less certain that their power will remain intact. In sum, changes in the party system and institutional structure characterize transitional environments, so that outcomes are more uncertain. Other things being equal, uncertainty increases demand for the political insurance that judicial review provides. Under conditions of high uncertainty, it may be especially useful for politicians to adopt a system 15 See, for example, Jon Elster, "Introduction," in The Roundtable Talks and the Breakdown of Communism i, 17 {Jon Elster, ed., 1996). J -1 Why Judicial Review? 31 of judicial review to entrench the constitutional bargain and protect it from the possibility of reversal after future electoral change. Because rational politicians believe they may not remain in power under the terms of the constitution, they choose to set up independent courts to protect i their bargain from repeal.16 Judicial review in such circumstances pro- vides insurance for the past against the future. In short, the presence of elections - the sine qua non of democracy - increases uncertainty and thus the demand for judicial review. The expansion of judicial power around the globe reflects democratization and is not antidemocratic as suggested by some analysts.17 Judicial review may be countermajoritarian but is not counterdemocratic. By relating judicial review to political uncertainty, this account provides a new perspective on the spread of judicial review around the globe in the latest "wave" of democratic constitution writing. The spread of judicial review does not merely reflect a norm among constitution drafters, but a response to the particular problems of electoral uncertainty that they face. ! This is not to argue that the international context is irrelevant. Given that no insurance contract is perfect or infallible, constitution writers must consider various institutions that might achieve their goal of reducing risk. Demand for any particular institution will rise with the perception that thcinsurance it provides is likely to be effective. As judicial review spreads to new environments and appears to function successfully, it becomes \ easier for new democracies to adopt it as they engage in constitutional reform and drafting. The spread of judicial review is self-reinforcing as its j institutional reputation grows. But the account offered here supplements ; the international story with an account of the domestic political logic of t review. j Note that the "insurance" rationale for judicial review is not an orig- in 1 list theory. Politicians need not anticipate that judges always inter- | prct the constitution in accordance with the founders' wishes. There are agency costs associated with judicial review as in any situation where one body (constitution drafters) appoints another (court) to monitor a ■j third (government). Judges may impose their own constitutional prefer- j trices on the polity, even where they are appointed by politicians. Certain * asP?cts of the institutional design of judicial review, such as political con- * trol oyer, appointments or the budget, are designed to minimize agency 1 costs from the point of view of current politicians. If judges act too Raniscyer, supra note 7. ^Robert Dahl, Democracy and Its Critics (1989). 3* Judicial Review in New Democracies Why Judicial Review? 33 outrageously, politicians can punish them through constitutional and extraconstitutional means. These mechanisms to reduce agency costs create a new puzzle. If the tools of political control over judges are available to political majorities after the constitution enters into force, why should a prospective political loser set up judicial review to constrain those majorities? Notions of legality and fidelity to text are crucial to reducing perceptions of judicial agency costs. Judicial agency costs are relatively low compared to those of other kinds of political functionaries because of the typical (though not universal) requirement of legal training to serve on a constitutional court. If judges were unconstrained in asserting their own policy preferences, as suggested by proponents of the attitudinal model in political science studies of courts, constitutional law becomes politics by other means, and there would be no inherent reason that constitutional interpretation should be limited to lawyers and judges. Legal training is a form of ideology that can help to reduce agency costs. From the perspective of politicians, it is an inexpensive mechanism because legal academia subsidizes the costs of training judges and developing jurisprudential solutions to particular problems and also subsidizes the cost of monitoring the court by rewarding commentators who analyze the work of the court. Judicial review may also be attractive to minorities even in the face of majority dominance because political pressures on judges are costly. Authoritarians from Zimbabwe to Malaysia have been criticized for improper interference with the judiciary, and though the threat of such criticism does not always protect courts, it is effective much of the time. The presence of a third-party adjudicative body whose explicit mission is to safeguard the constitution raises the costs of violating the constitution, even if it does not provide perfectly complete protection against all contingencies. Judicial review, like insurance, is a risk-reduction device. No risk-reduction device is foolproof: Insurers can go bankrupt just as courts can be ineffectual. But if the expected gains from a relatively inexpensive insurance contract outweigh the potentially catastrophic risk of a failed constitutional scheme, judicial review should be adopted. The discussion so far can be understood in terms of a simple inequality. Constitutional designers will choose judicial review if and only if the expected costs of electoral loss (the probability of electoral loss times the average expected cost} exceed the net agency costs of judicial review. As the risk of electoral loss increases, the incentive to adopt judicial review increases as well. Similarly, any increase in perceived loyalty of the judiciary to the constitutional designer, either for ideological or political reasons, Si í will increase the incentive to adopt judicial review, holding electoral risks constant. Judicial review will be adopted where comparative institutional analysis suggests that courts' agency problems are likely to be less than the costs of having no third-party monitor at all. Simply because I focus here on the self-interested motives of politicians does not mean that I believe that all constitutional design can be explained as the product of self-interest. Other forces clearly play a role. Constitutional designers may sometimes be motivated by passions about certain ideas and institutions; they may even on occasion try to choose the best possible institutions for their polity.18 Accidents and miscalculations are no doubt more frequent than social scientists like to recognize. My argument is one of probability: Those who need political insurance will tend to prefer to adopt judicial review, but this theory does not purport to explain every case. We have now sketched the outlines of a theory regarding why judicial review is adopted in a democratic constitution. Although judicial review is associated with the global ideal of the rule of law, the adoption of a constitutional court may reflect in large part the insurance needs of the founders. This hypothesis suggests a corollary, that the particular design of judicial review institutions reflects local political realities. In particular, we predict that where dominant parties control the constitutional drafting process, we should expect a weak, low-access form of judicial review. Where constitutions are designed in conditions of political deadlock or diffused parties, we should expect strong, accessible judicial review. The next chapter will consider this hypothesis in greater depth. CONCLUSION Judicial review reflects the incentives of constitutional designers to adopt a form of political insurance. By ensuring that losers in the legislative arena will be able to bring claims to court, judicial review lowers the cost of constitution making and allows drafters to conclude constitutional bargains that would otherwise be unobtainable. As democratization increases electoral uncertainty, demand for insurance rises. Although other institutions can also serve to protect minorities, judicial review has become particularly focal. This theory goes a long way toward explaining the rapid spread of judicial review in recently adopted constitutions. 18 Jem Elster, "Forcesand Mechanisms in the Constitution-Making Process," 45 Duke ■■: LJ- 364 (1995)1 Constituting Judicial Power 35 Constituting Judicial Power What determines the character of judicial review as it operates in new democracies? One important factor is the institutional design of the court and access to it, which is the subject of this chapter. The institutional design of the judicial review mechanism is generally, though not always, a product of the written constitution itself. As such, it reflects, in large part, the choices of the constitutional designers. The political bargain struck at the outset of the democratic regime and embodied in the constitutional text will frequently include some provisions for judicial review. Some important features of the judicial review body, such as its jurisdiction, composition, and selection method of its members, may be detailed in the constitutional text. Text is not the only source of judicial power, however. This qualification is necessary both because some systems of judicial review are not derived from constitutional text (the systems in Israel and the United States are two well-known examples), but also because nonconstitutional norms may be important in shaping the environment of judicial review. Frequently, matters such as terms and procedures are listed in ordinary organic statutes of the judicial review body. Furthermore, judicial decisions themselves will fill in many of the gaps in these frameworks. Particularly important are decisions related to jurisdiction and standing that play a major role in a court's self-articulation of its political role. This chapter is primarily concerned with the institutional choices embodied in written constitutions and their importance in setting the stage for judicial review. It focuses on explicit design of judicial institutions from the perspective of politicians seeking insurance. By tying judicial review to the politics of constitutional drafting, this chapter offers a theory 34 witfv"more specific implications than the general argument that the spread of judicial review reflects a global rights consciousness. TAILORING THE INSURANCE CONTRACT: DIMENSIONS OF INSTITUTIONAL DESIGN Chapter i argued that the decision to include judicial review in the constitution reflects the political needs of constitutional drafters. We still need ro consider why politicians choose a particular institutional design for the judicial review that they do. This choice can be analogized to tailoring the insurance contract to fit specific local conditions. Why, for example, do sortie constitutional designers choose to adopt a system with open access so that any individual citizen may invoke the machinery of constitutional control, while other designers limit access? This question, and others of institutional design, can only be addressed after an exploration of the dimensions along which systems of judicial review differ. This chapter addresses five major dimensions on which systems of judicial review vary: access to the court; effect and timing of judicial decision; the institutional mechanisms for accountability to the political environ-ment; the term length of constitutional justices; and the size of the court. Access refers to how cases are brought to the court, effect refers to the consequences of a finding of unconstitutionality, and accountability concerns how the court as an institution is composed and the mechanisms for political control and influence over the court. Term length concerns the length of time judges can serve and whether they may be reappointed. The size of the court refers to the number of judges. We do not devote much explicit consideration here to what is perhaps the highest-order choice constitutional designers face with regard to court design - namely, whether to grant the power of judicial review to the ordinary judiciary (as in the United States) or to limit it to a designated body (as in the Kelsenian model). Rather, we consider this decision as being essentially related to access to judicial review. A word of explanation is in order here. Countries that allow ordinary courts to conduct judicial review are almost exclusively those that for historical reasons were subject to Anglo-American legal influence. We see this decision as being driven in part by considerations of history or legal tradition, though it is important to note that many of these countries have opted for hybrid solutions to the question of centralization. For example, the South African Consti-tutional Court is really a court of constitutional appeals from ordinary 36 Judicial Review in New Democracies Constituting Judicial Power 37 courts, and the Indian Supreme Court has a designated constitutional bench. The common law courts, with their myriad roles and long tradition of autonomy, are a resource in the Anglo-American world that is simply unavailable to countries with other legal traditions. Ordinary courts in most new democracies seldom have such institutional credibility. For this reason, the default choice for many countries is to adopt a German-style designated constitutional court, but to tailor the design along the various dimensions discussed as follows. Therefore, we do not spend much time considering the political incentives to adopt centralized or decentralized review: In many constitutional design situations, there is no real choice to be made here. The particular relationship between the ordinary courts and constitutional review does have significant effects on the operation of the system of judicial review, and this is a theme that will be apparent in the case studies. Access Constitutional review systems differ widely on the question of who has standing to bring a claim. One can array access to the court on a spectrum from very limited access, as in the original design of the Austrian model in 1920, in which only state and federal governments could bring cases, to the present design of the German Constitutional Court, where not only political bodies but individuals may enjoy direct access through constitutional petitions and ordinary judges may refer questions as well. The Indian Constitution guarantees direct access to the supreme court on questions of fundamental rights and also allows the court to hear advisory questions so that its jurisdiction is much broader than its American counterpart, whose jurisdiction is limited to concrete cases. The present Hungarian Constitutional Court has perhaps the widest access of any such body in the world today, as the right of abstract constitutional petition is not even limited to citizens.1 Like other elements of institutional design, access can change over time. For example, 1974 constitutional amendments in France extended the right of petition to any group of sixty parliamentary deputies, allowing 1 However, the court does not engage in concrete review in the classical sense of intervening in the context of ordinary court decisions. As for other courts with wide access, the Slovak court allows petition by "anyone" whose rights are the subject of inquiry, but this right is probably limited to citizens. Slovak Constitution (1991), Articles 130(f), 117. Centralized Decentralized Open Access Germany, Italy United States, Canada, Japan, Scandinavian Countries Limited Access France 5th Republic, Austria 1920-29 figure 2.1. Type of Judicial Review Body and Access minority parties to challenge governmental action on constitutional grounds. Judicial decisions can also expand or contract standing.2 Standing doctrine in the United States Supreme Court has changed over time, reflecting different judicial agendas.' Figure 2.1 describes these features for some of the major systems of judicial review, again keeping in mind that hybrids are possible between these ideal types. Access to the court is perhaps the most important ingredient in judicial power, because a party seeking to utilize judicial review as political insurance will only be able to do so if it can bring a case to court.4 Setting up a designated constitutional court, accessible only to a narrow set of organs, has the effect of limiting the insurance function of the constitutional court. But some designated courts, such as that of Hungary, have wide access. Figure 2.2 arrays access to constitutional courts on a spectrum, from very limited access to very open. Open access decentralizes the monitoring function widely and makes it more likely that politicians will be challenged in court should they fail to abide by constitutional limitations. The design choice on access has much to do with the prospective position of political forces in the constitutional system. Other things being 1 See, for example, Plasty. Cohen 39z U.S. 83, 88 S. Ct. 1942, zo L. Ed. 947 (1968) (taxpayer standing in the United States). 5 Maxwell Stearns, Constitutional Process: A Social CJjoice Analysis of Supreme Court Decision Making (2000), especially Chapter 6. 4 Note, however, that some of the new courts also have limited power to initiate proceedings on their own prerogative, without a formal petition from outside. 38 Judicial Review in New Democracies Constituting Judicial Power 39 Access Mechanism Examples Special bodies only Austria 1920-29, France before 1974 Special bodies + legislative minorities France after 1974, Bulgaria, Rumania Special bodies + any court Taiwan, Poland before 1997 Any litigant United States Special bodies + any court + citizen petition Germany, Mongolia Special bodies + any court + open petition Hungary figure 2.2. Accessibility of Constitutional Adjudication (Lower on Figure More Accessible) equal, a dominant party will seek to limit access to judicial review, perhaps by restricting it to major political institutions. Political forces in rough balance will seek to maximize access to legislative minorities and ordinary citizens to provide insurance in the event of an electoral loss. Because they expect to lose in the legislatures, the availability of constitutional review provides the prospective minority with another forum in which to contest policies of the majority. This may be achieved by extending access to the court to minority groups in the legislature or to ordinary citizens. Open access also allows watchdog groups that might share the policy preferences of the politicians to make claims and assist in monitoring the government. We should thus expect a correlation between political uncertainty and open access. Another distinction is whether the court can hear constitutional questions only in the context of concrete legal cases (as in the U.S. Supreme Court), or whether it can consider constitutional issues in the abstract. Concrete review requires litigation of constitutionality in the context of a particular case. Abstract review determines the constitutionality of a statute without a specific case. The French Conseil Constitutionnel may only hear issues in the abstract. The German and Spanish Constitutional Courts practice both abstract and concrete review.5 In practice, the distinction between abstract and concrete review is not as important as it may appear, but it is a widely used theoretical construct. A related issue concerns the timing of review: In the French system, review can only take place ex ante promulgation of legislation. This means that the law can be modified by the legislature to conform with the decision of the Conseil Constitutionnel^ this form of review makes the 5 For a discussion, see Alec Stone, The Birth of Judicial Politics in France 231-35 (1992). c0,i Article r49(5). 9 See, for example, Constitution of France (1958), Articles 58-60; Basic Law of Germany (1949), Article 41(2); Constitution of Lithuania (i992)> Article 105(3)) 10 See, for example, Constitution of Bulgaria (1991), Article 149(8); Constitution of Hungary (1949}, Article 31(a); Constitution of Mongolia (1992), Article 35(1); Basic : Law of Germany (1949), Article 61. " Rett R. Ludwikowski, "Constitution Making in the Countries of Former Soviet Dominance: Current Developments," 23 Ga.J. Int'l & Comp, L, 155, 190 (1993). The constitution was passed in 1995 without these provisions. 4° Judicial Review in New Democracies Constituting Judicial Power 41 legislation.11 The Thai Constitutional Court has the power to approve recommendations of the Counter-Corruption Commission to ban politicians from office for failing to accurately report income and assets, a power that it has already used several times in the short period of time since the adoption of the 1997 constitution. Indeed, in one notable case the constitutional court was asked to ban the incoming candidate for prime minister. All of these powers can be very important for understanding the political role of constitutional courts in a particular system. Effect Systems of judicial review also vary in the effect of their pronouncement on legislation in concrete cases. American courts, bound by the rule of stare decisis, do not actually void laws that they find to be unconstitutional. Rather, because subsequent similar cases must follow the rule in previous cases, the voided law remains on the books, if dormant for all practical purposes. In centralized systems, by contrast, the court has the power to declare the laws unconstitutional and immediately void. This feature of direct annulment of laws in centralized systems is often said to follow from the lack of a stare decisis doctrine.13 Without a clear principle that precedents must be followed, ordinary courts could vary in their application of the constitution, hampering predictability and consistency in the legal system. To avoid such a result, the declarations of some constitutional courts are given erga omnes effect, meaning they are binding for all future cases. A variation found in the German tradition is that the constitutional court has two choices in rendering a finding of unconstitutionality.14 It can either find legislation null and void (nichtig) or incompatible {unvereinbar) with the basic law. In the latter case, the court declares the law unconstitutional but not void and usually sets a deadline for the legislature to modify the legislation. Sometimes these decisions admonish the legislature to modify the legislation within particular guidelines.15 The court becomes deeply involved in "suggesting" to the legislature language that ultimately finds its way into the statute. For example, in its 1975 decision voiding 11 Herman Schwartz, "The New Courts: An Overview," 2 E. Eur. Const. Rev. 28, 30 (r993>> quoting Constitutional Court Act of the Russian Federation (1993), Article 9. 13 Mauro Cappelletti, Judicial Review in the Contemporary 'World (1971). 14 Donald Kommers, The Constitutional Jurisprudence ofthe Federal Republic of Germany (1989). Ibid, at 53. ■1 permissive statute allowing abortion, the German Constitutional Court engaged in extensive suggestions for rewriting of the statute.1^ In other cases, the court will sustain a challenged statute, but warn the legislature that it is likely to void it in the future, or suggest conditions for the constitutional application of the statute. These types of decisions are typically understood as pragmatic, designed to give the legislature time to adjust the content of major legislation for which a judicial declaration of unconstitutionality would cause too much social disruption.'7 Partial findings of unconstitutionality are indeed less politically dramatic, and courts in new democracies that have adopted this technique have been more willing to send legislation back to the legislature than to void it completely.18 There is no doubt that the availability of "lesser" options to voiding a law allows courts to take a more nuanced view of the political process in which they are engaged and therefore facilitates a more subtle range of interactions with the political bodies. Their pragmatism notwithstanding, such decisions are problematic from a rule of law perspective. After all, the court finds a law unconstitutional, but allows its continued application. Although the delay in voiding the legislation may provide some advantages in terms of predictability, it appears odd to allow an unconstitutional act to stand. In my view, these techniques can best be understood by viewing the court as a quasi-legislative actor engaged in democratic dialogue with political branches of government. This typically is a negative form of legislating, guarding the limits of the process rather than promoting policy initiatives. Nevertheless, through suggestions for revision the court may have significant impact on the shape of legislation. In some systems with a legacy of parliamentary control of constitutionality, the decision of the constitutional court as to unconstitutionality is not binding, but rather is advisory to the legislature. The legislature retains some power to reject or accept the court's finding, either by majority or supermajority vote. A version of this model was extant in Poland during the life of its first Constitutional Tribunal, 1988-97, and remains intact in Mongolia, as will be discussed in Chapter 6. Similarly, the Brazilian Senate ■ Ibid. See also Mary Ann Glendon, Abortion and Divorce in Western Low (1987). 17 Kommers, supra note 14, at 54. ,s For example, the Korean Constitutional Court has begun to use this technique as a matter of course, preferring it to outright striking of legislation. See Constitutional Court of Korea, the Constitutional Court 17 (1997); and discussion in Chapter 7, infra. 4* Judicial Review in New Democracies Constituting Judicial Power 43 can choose to accept as binding erga omnes a decision rendered inter partes in a specific case, allowing it to convert a finding of unconstitutionality.19 Courts in Latin America make use of a device called amparo, wherein a successful constitutional complainant will be free from the application of the offending law or government act, but the act will continue to apply to others. This device is desirable from the perspective of politicians who do not want much judicial constraint, a fair characterization of many governments in Latin America during the twentieth century. The amparo channels constitutional protest into the courts, perhaps relieving a source of broader political pressure on the regime, but at the same time does not really limit the government's freedom of action. An unconstitutional act that affects 1,000 people might require up to 1,000 suits, with all the expense they entail, before it no longer has effect. The amparo may work well to provide redress against government actions that provide substantial burdens on small numbers of citizens, such as measures affecting property rights. But actions that provide only minor burdens, or those that affect populations less able to mobilize for legal action, are likely to remain effective tools. Mechanisms of Appointment and Accountability Appointments are among the most crucial of design issues. Constitutional designers are unlikely to adopt constitutional review unless they believe it will be carried out by impartial appointees. If designers believe they are : likely to lose postconstitutional elections, they will not be in a position ; to appoint judges. So overly partisan mechanisms are especially unattractive. The normative task is to select an appointment mechanism that will maximize the chances that the judge will interpret the text in accordance with the intentions of the constitution writers. This, in turn, requires considering judges' utility functions, an issue concerning which there is no consensus in the literature.10 Appointment mechanisms are designed to insulate judges from short-term political pressures, yet ensure some accountability. The United States 19 Constitution of Brazil, Article 52(X). » jeffrey A. Segal and Harold J. Spaeth, The Supreme Court and the Attitudmal Model (i993) {judges vote their political preferences); Lee Epstein and Jack Knight, The Choices Justices Make (1998) (judges are strategic maximizes); Lawrence Baum, Tte Puzzle of Judicial Behavior (1997) (reviewing evidence and discussing poor state ot knowledge on this question); Richard Posner, Overcoming Law (1995)' federal judicial system has lifetime appointments for insulation, but puts tremendous effort into screening potential candidates in the appointment process. Other systems set up mechanisms for ensuring accountability for judicial performance ex post by providing for renewable terms. Many American states use a system of elections that allows a judge to be appointed by a governor upon recommendation by a committee of mixed composition.11 Judges are then subjected to recall elections where they "run on the record," that is, without opposition. Judges in these systems are very rarely recalled, so the threat may not be much of a constraint in reality. Mueller persuasively argues that a supermajority requirement for judicial selection will tend to protect the minority from losing in both the courts and the legislature and by extension will tend to produce more-moderate, acceptable judicial candidates.11 Mueller also considers the merits of having the judiciary and the chief executive serve as appointing authorities for the judiciary. He favors such professional appointments by existing judges, noting that the judiciary has internal incentives for competent selection.13 A judiciary that appears incompetent invites modification of the appointment system. Indeed, one design suggested by Mueller would allow judiciary-nominated judges to take office barring legislative intervention by supermajority.14 This proposal combines accountability and independence, because most appointments would be tontine, but there is a mechanism for political intervention should judges nominate candidates who are far out of step with political opinion. : I divide appointment mechanisms into three broad types: professional appointments, as in Mueller's proposal; cooperative appointing mechanisms; and representative appointing mechanisms. Theoretically, one can also have single-body appointment mechanisms where, for example, an executive can appoint all members of the constitutional court without legislative oversight. An example that is close to this is the Council of Grand Justices in Taiwan, whose members are appointed by the president from a list of nominees prepared by a committee he picks. Approval is required by the legislature, but because the president was historically the lr This is the so-called "Missouri plan." Mary Volcansek and Jacqueline Lucienne Lafon, Judicial Selection (1987). 1 Dennis C. Mueller, Constitutional Democracy 281 (1996). 1 Dctinis C. Mueller, "Fundamental Issues in Constitutional Reform: With Special References to Latin America and the United States," 10:2 Const. Pol. Econ. 119,125 U.999). H [bid. 44 Judicial Review in New Democracies Constituting Judicial Power 45 head of the largest political party, this was not an effective check, and i he mechanism was a de facto single-body appointment mechanism. Single-body mechanisms of this type are unusual in democracies because they can lead to all-or-nothing composition of the court. If the president can appoint all the judges, the presumption of effective constitutional con- j straint disappears. Therefore, the insurance rationale for judicial review i loses its appeal. Cooperative appointment mechanisms require the cooperation of two bodies to appoint constitutional justices; the American, Russian, and! Hungarian procedure of presidential nomination followed by legislative j confirmation is one example. These systems seem consistent with the objective of supermajoritarian requirements to ensuring broad support I (institutional or political) for those who are to interpret the constitution. They risk deadlock, however, because they require the agreemeni of | different institutions to go forward. Although there are no institutional! barriers to such bargains being concluded, it is possible that in circuin-. stances of political conflict, appointments would not be made. Finally, representative mechanisms utilize multiple appointing authorities: For example, in Italy a third of the nine-member court is nominated by the president, a third by the parliament, and a third by the supreme court.15 This system has been copied in such diverse places .is Bulgaria, Korea, and Mongolia. Alternative versions provide for one-third of appointments by each house of a bicameral legislature and one-third by the chief executive. Representative systems can be distinguished from cooperative systems in that, theoretically, appointees cm be much closer to pure agents of the appointors. Because no other insii- J tution must agree to the appointment, there is no need for compromise. There may also be, however, a dynamic that prevents politicized appointments where there are three appointing bodies. Each appointing bodv may seek to appoint persons sympathetic with its institutional interest1-. However, if it is too blatant in doing so, the other appointing bodice will respond by appointing their loyal partisans. Because only one-third:-^ of the membership is appointed by any one body, each can be assured that it will be unable to dictate outcomes if each judge acts as a pure agent. I characterize this institutional design as "mutually assured politi-cization." Each body that appoints a person who appears to be a pure agent signals that it may plan to engage in extraconstitutional action 15 Volcansck and Lafon, supra note 21. J needs to influence the court to uphold its action. By appointing "oineone who appears "neutral" and nonpartisan, the appointing authority signals that it does not anticipate needing or using the court to uphold its own controversial actions. Thus, representative mechanisms may P rovide, like cooperative mechanisms, an incentive for moderate appointments. Despite their popularity, representative systems have a disadvantage compared with cooperative systems. Although a dynamic of moderation as described above may come into play, there is some possibility that politicians will simply nominate pure agents. Opinions issued by a court of pure agents are likely to be internally fragmented and of lower quality than those issued by a more centrist, consensual deliberative body as appointed through cooperative mechanisms. Cooperative mechanisms more closely approximate the supermajority principle of constitutional economics but risk deadlock in the appointing process. Representative systems ensure a smooth appointment process but risk deadlock on the court. In the German system, wherein each house of the legislature can appoint an equal number of members to the Constitutional Court, supermajority requirements are used in selecting judges.16 This has led to a norm of reciprocity that has established de facto party seats held by the three -major parties. The norm produces a stable court that reflects broad political preferences without overrepresenting either of the two main factions. This version of the legislative-centered system turns parties, not institutions, into the important players. The system is stable because the party system is stable. The dynamics of party systems are a crucial variable in evaluating selection systems. A system of self-appointments by the professional judiciary may be the most likely to produce accurate review if we assume judicial neutrality, but it can lead to a court that dominates the legislature if the party system is too fragmented and unstable to provide a constraint on judicial decision making. In stable party systems, supermajority requirements will produce moderate judges, but appointments may not be made if there is deadlock. Representative systems ensure appointments will be made but create other risks on the court. For example, if the chief executive is the head of the majority party in one or both houses of Parliament, The Bundestag appoints its members through a two-thirds vote of the Judicial Selection Committee with party representation proportional to that of the body as ■a whole and the Bundesrat through a two-thirds vote of the body as a whole. Kommers, supra note 14. 46 Judicial Review in New Democracies Constituting Judicial Power 47 this system will lead to a court that is allied with the chief executive. Where there is little party discipline or where the chief executive is independently elected, however, institutional rivalries can lead to a more divided court.*? Term Length Term length is typically seen as being a key component of judicial independence. zS Other things being equal, it is argued, the longer the term of appointment, the freer a judge will be in exercising discretion. U.S. federal court judges serve for life, and this is considered an important guarantee of their independence. The longer the appointment, the more independent a judge can be of prevailing political sentiments. Like central bank governors, judges are at risk from undue pressure to advance short-term political interests rather than the long-term collective good. We should thus expect longer terms to correlate with politicians who value judicial accuracy and independence, namely pessimistic politicians with insurance needs. Although one might think that lifetime appointments are always longer than designated terms, this is not the case because virtually all other systems with "lifetime" appointments provide for a mandatory retirement age of sixty-five to seventy years of age. Even if this were not the case, appointments could come late in life as a reward for political loyalty rather than an incentive for independent adjudication. Thus, actual time served on such courts may in fact be lower than judges on courts with specific and limited terms. For example, Japanese judges on the supreme court serve until mandatory retirement at age seventy, but this in fact produces very short terms, averaging around six years.19 Politicians in these systems exercise preappointment scrutiny over prospective judges.30 Other constitutional judges have limited terms. French members of the Conseil Constitutionnel serve a single nine-year term, and judges on the 27 Bailey proposes that constitutional issues be decided by a legislature, possibly the previous sitting legislature that appointed judges if the issue is legislation passed by the current legislature. Martin Bailey, "Toward a New Constitution for a Future Country," 90 Pub. Choice 73, 99 (1997). 18 See, for example, William Landes and Richard Posner, "The Independent Judiciary in an Interest-Group Perspective," 18 /. L. &Econ. 875 {1975). 19 J. Mark Ramseyer and Eric B. Rasmusen, "Judicial Independence in a Civil Law Regime: The Evidence from Japan," 13 /. L., Econ. & Org. 259 (1997). 30 Masaki Abe, "Internal Control of a Bureaucratic Judiciary: The Case of Japan," 23 Infi. J. Soc. L, 303 (1995); Ramseyer and Rasmusen, ibid. German Constitutional Court serve a single twelve-year term. Judges of other constitutional courts, including that of Spain, are allowed to be reappointed. Other things being equal, the possibility of reappointment has the potential to reduce judicial independence, as judges late in their term who seek to remain in office must be sensitive to the political interests uf those bodies that will reappoint them. Of course, judges serving a single limited term also have an incentive to act with an eye toward future employment possibilities, so to the extent political authorities have control over entry into the professorate or other postjudicial positions, judges may be subject to political discipline in such systems as well. Court Size The constitutional designer may specify in the constitution the number of judges on the court. The major tradeoff here is between speed and accuracy. The greater the number of judges, the higher the costs of deliberation. At the other extreme, a single judge deciding all cases would be a relatively inexpensive method of judicial decision making. The problem with a single judge is that the potential error costs of such a system are high.3' Hence, it is common for judicial panels to grow larger as an issue rises through a system of appeal. For example, United States federal courts of appeals frequently decide cases in panels of three judges with appeal to the court en banc. One might think that larger courts would always be more accurate and hence better able to fulfill the insurance function for constitutional designers. After all, it seems plausible to assume that error costs are reduced by deliberations, and there is ample empirical evidence that group decision making is of higher quality than individual decision making.31 However, others have argued that once a group expands beyond a certain size it tends to make poorer decisions. For example, Richard Posner has recently argued that an expansion in court size may be associated with a decline in quality of decisions, in part because norms of work are less sustainable with larger groups. However, his evidence is not dispositive 3' At an extreme, in the United States, we let the trial judge decide the initial matter himself or herself even though his or her preferences may not reflect those of the court as a whole or of the median judge. Warren F. Schwartz and C. Frederick Bcckner III, "Toward a Theory of the 'Meritorious Case': Legal Uncertainty as a Social Choice Problem," 6 Geo. Mason L. Rev. 801 (1998}. J1 At least in certain contexts. See Stephen Bainbridge, "Why a Board? Group Decisionmaking in Corporate Governance," 55 Vand. L. Rev. 1 (2002). 48 Judicial Review in New Democracies on the question.'^ Furthermore, Posner considers overall court size on an appeals court that initially hears cases in panels, so his argument is not directly relevant to constitutional designers that are creating courts that hear matters en banc. In the context of new democracies, we believe smaller courts should be associated with more dominant political parties. This is because there are less factions concerned with representation on the court, and hence less of a need for ensuring balance among the membership. Furthermore, each additional judge increases the budget of the court, and there is little reason a dominant party would want to incur these extra costs, other things being equal. One might argue that the salient variable to examine is panel size rather than court size. But the size of panels is typically a matter left to ordinary law or the organic statutes of a constitutional court, rather than being specified in the constitutional text. Furthermore, because important cases will often be heard en banc, the overall size of the court is a relevant variable subject to influence by constitutional designers. There is some empirical support for the proposition that designated constitutional courts are larger than their counterparts that are the courts of final appeal for all issues. For new constitutional courts set up after 1989 (n = 25), the mean number of justices was 11.25. For supreme courts given the power of constitutional review in the same period (n = 8), the mean size is 8.25. The fact that supreme courts are smaller even though they have nonconstitutional cases to consider might indicate that first-instance consideration of the issues by lower-level courts saves time later on. Summary To summarize the argument so far, each dimension of design choice has certain effects on the capacity of the court to render accurate review. Table 2.1 summarizes three prototype constitutional courts along these dimensions. Because there are numerous dimensions upon which the institutional design of a system of judicial review may vary, there is an almost infinite array of configurations, and no two courts share exactly the same design and institutional environment. The diversity of systems 33 Richard Posner, "Is the Ninth Circuit Too Large? A Statistical Study of Judicial Quality," 29 /. Leg. Stud, jii (2000). See also Kaushik Mukhopadhaya, "Jury Size and the Free Rider Problem," forthcoming, /. L., Econ., & Org. (1003). Constituting Judicial Power 49 table 2.1 Dimensions of Design Choice Germany - Constitutional United States - France - Consetl Court Supreme Court Constitutionnel Access/Standing Petition, courts, Access through Restricted requests from courts only standing government Justiciable Concrete and Concrete review Abstract review Questions abstract review only only Appointments Representative - Cooperative - Representative - 2 houses president, president, parliament parliament 2 houses (supermajority) parliament 1 erm Length 12 Life - no age limit 9 Size 16 9 9 of judicial review can be seen in Table 2.2, which presents the structural features of selected new constitutional courts established after 1980. EXPLAINING VARIATION IN JUDICIAL REVIEW Judicial Review as Insurance: Anecdotal Evidence Does actual design of judicial review reflect the insurance model? There is strong anecdotal evidence to support the hypothesis that judicial review will be more accessible and powerful where political forces are diffused at the time of. the constitutional bargain and more limited when a single party controls the process.34 Take as an initial example the French system, sometimes referred to as limited constitutional review. Constitutional review is restricted to abstract, ex ante review by a centralized body. At the time of the establishment oi the Fifth Republic, standing was restricted to certain designated governmental bodies, a fact perfectly consistent with the insurance theory. The conseil was adopted at the instigation of General De Gaulle, who wanted a strong executive to prevent the deadlock that had characterized the Fourth Republic. The constitutional scheme features a dual system of J1 There is similar evidence that centra! bank independence is strongly correlated with politicians' time horizons. As politicians' time horizons shorten, independence increases. John B. Goodman, Monetary Sovereignty: The Politics of Central Banking in Western Europe (1992). table z.z Constitutional and Supreme Courts in Selected New Democracies Term in Term Total # Country Years Renewable? Judges Appointing Authorities Access Albania 9 no 9 president with assembly approval certain bodies, citizens Argentina life+ 5 president with senate approval courts Armenia life+ 9 5 assembly, 4 president certain bodies, 33% of deputies Bangladesh life+ N/A president courts, limited original jurisdiction Benin 5 once 7 4 assembly, 3 president certain bodies, citizens Bolivia 10 no 12 parliament courts, limited original jurisdiction Bo snia-Herzegovina life® yes, initially 9 4 house, 2 by Serb assembly, 3 by certain bodies, 25% of deputies of European Court of Human Rights either chamber, ordinary courts in consultation with president Brazil life ri president with senate approval special bodies, political parties, labor unions, and any citizen Bulgaria 9 no 12. 1/3 president, 1/3 parliament, certain bodies, 20% of deputies 1/3 top courts Burkina-Faso 9 no 9 1/3 president, 1/3 assembly, certain bodies, zo% of deputies 1/3 judges nominated by justice minister Cambodia 9 yes 9 1/3 king, 1/3 assembly 1/3 sup certain bodies, courts judicial council Cape Verde life 5 1 president, 1 assembly, 3 judges courts, limited original jurisdiction Chile 8 N/A 7 3 supreme court, 1 president, certain bodies, 25% of members of 1 senate, 2. national security either chamber council Colombia 8 no 9 senate from lists given by president, certain bodies, citizens supreme court, council of state ,, , , Croatia 8 yes 13 parliament certain bodies,.33?b of deputies of either chamber, any person Czech Republic IO once 15 president selects from parliament list certain bodies, open to citizens with senate consent Djibouti 8 no 6 1/3 president, 1/3 assembly, certain bodies, 10 deputies 1/3 judges Dominican Republic iife-f- 16 senate courts Ecuador life 9 national judicial council courts El Salvador 9 yes 5 national judicial council certain bodies, citizens Estonia life+ 17 court president nominated by any court president; others by court president Ethiopia life+ 11 6 president, 3 house members and certain bodies supreme court president and VP Fiji life+ N/A president certain bodies Gabon 7 once 9 1/3 president, 1/3 senate, certain bodies, 10% of deputies 1/3 assembly Georgia IO yes 9 1/3 president, 1/3 parliament, certain bodies, open to citizens 1/3 supreme court Ghana fife+ at least 10 judicial council courts Greece 2, yes 11 courts, state councils, and professors citizens Guacemala 5 yes 13 congress selects from list of 2.6 courts Guinea-Bissau life 17 assembly on presidential nomination courts Honduras 4 yes 9 congress certain bodies, open to citizens Hungary 9 once 15 parliament certain bodies, open to anyone Korea 6 yes 9 1/3 president, 1/3 parliament, certain bodies, open to citizens 1/3 supreme court Kyrgyz Republic 15 yes 9 assembly certain bodies, courts (continued) table 2.2 (continued) Term in Term Total # Country- Years Renewable? Judges Latvia 10 no, but later 7 reelection OK Lithuania 9 no 9 Macedonia 9 no 9 Madagascar 6 no 9 Malawi life+ at least 3 Mali 7 once 9 Mauritania 9 no 6 Moldova 6 yes 6 Mongolia 6 yes 9 Morocco 6 yes 9 Nepal life+ Nicaragua 6 N/A at least 7 Niger 6 no 7 Panama IO no Paraguay Iife+ 9 Peru 5 no j Philippines life+ 15 Poland 8 no 12 Rumania 9 no 9 Russia life+ 19 Senegal 6 no 5 Sierra Leone life+ at least 5 Slovakia 7 once 10 Slovenia 9 no 9 South Africa 12 no II Taiwan 9 no after 2003 15 Tanzania life+ 5 Thailand 9+ no 15 Ukraine 9 no r8 Appointing Authorities Access Uruguay 10+ no, but later reelection OK 3 assembly, 2 cabinet, 2 supreme court 1/3 president, 1/3 parliament, 1/3 supreme court assembly 3 president, 2 assembly, 1 senate, 3 by supreme judicial council president 1/3 president, 1/3 president of assembly, 1/3 judges 3 president, 2 assembly, 1 senate 1/3 president, 1/3 assembly, 1/3 supreme council 1/3 president, 1/3 parliament, 1/3 supreme court 4 4- president by king, 4 by assembly king on judicial council nomination assembly on presidential nomination 1 president, 1 assembly, 2 judges, 1 professor, 1 bar association, 1 human rights organization president, with cabinetand legislative appmva\ president with senate approval ■■ assembly president parliament r/3 president, 1/3 each house of parliament appointed by federal council on presidential nomination president president president appoints from candidates national assembly on presidential nomination president after consultation with judicial commission; 4 must be judges president from list prepared by committee judicial council 5 supreme court judges, 2 supreme administrative court judges, 8 by committee 1/3 president, 1/3 parliament, 1/3 assembly assembly certain bodies certain bodies, 20% of deputies certain bodies, open to citizens president any court courts certain bodies, 10% of deputies certain bodies, 33 % of deputies of either chamber certain bodies certain bodies, open to citizens certain bodies, 25% of deputies courts courts certain bodies solicitor general courts certain bodies, groups of citizens courts certain bodies, open to citizens limited to certain bodies or legislative minority open to citizens certain bodies courts open to citizens open to citizens any court certain bodies, ordinary courts special bodies, 10% of deputies, or any court certain bodies courts Key: 4- = includes some age limitation; @ = after a five-year transitional set of appointments. Source: Adapted from constitutional texts at http://www.oefre.unibe.ch/law/icl/index.html with country-specific supplementary s 54 Judicial Review in New Democracies Constituting Judicial Power 55 law making, with certain subjects to be the province of executive decrees rather than parliamentary legislation. De Gaulle's confidence was such that he drafted the entire constitution around his personal popularity and did not trust parties or parliamentarians. By allowing the conseil to consider only statutes before promulgation, he placed a check on the Parliament's ability to dictate policy. Restricted standing allowed De Gaulle and government agencies to bring cases, but not ordinary citizens, who might challenge legislation that the government wanted. Furthermore, eliminating concrete review meant that the government would he able to act without constitutional scrutiny once policies were adopted. This scheme changed radically when standing was broadened in 1974 to include any minority group from the Parliament. This change was initiated by President Giscard d'Estaing, who headed the small Republican Party that governed briefly. As a minority party heading a coalition government, the Republicans valued expanded standing that would provide a guarantee of access once they were out of power. These changes have had a profound effect on French constitutional law.35 Predictably, expanded standing led minority groups in Parliament to complain to the conseil with greater frequency and to judicialize the very issues they had lost in the legislature. The Gaullists themselves were able to take advantage of this in the early 1980s, after the election of Francois Mitterand and the Socialist Party: The Socialists' extensive program of nationalization was challenged in and ultimately modified by the conseil*6 The German system features a centralized body that can engage in both abstract and concrete review. Standing is broad and includes constitutional petitions, as well as the so-called concrete norm control that allows ordinary courts to refer questions to the constitutional court in the context of ongoing legal cases. The design of the German system reflected a strong ideological desire to maintain an open and effective system in the wake of the Nazi experience and in this sense reflects the importance of the rights theory.37 The strong emphasis on basic rights and the distrusi of the ordinary judiciary meant that the centralized constitutional couri was an attractive option. However, the insurance theory also has a role to play in explaining institutional design. The German Basic Law was in 35 Stone, supra note 5. 36 Ibid, at 140-72. 37 Mauro Cappelletti, The Judicial Process in Comparative Perspective (1989). m.inv respects a compromise between those who emphasized "positive" economic and social rights and those who emphasized "negative" rights, such as the right to property. An easily accessible constitutional court served the interests of both groups in circumstances where neither felt thjf it could be assured of a victory in the political arena. Compared with France, a more divided political configuration led to a more powerful constitutional court, The adoption of judicial review in South Africa in the early 1990s provides a textbook illustration of the insurance theory.38 One might think that the African National Congress (ANC), as the dominant political force among the black majority, was the paradigm case of a dominant party lhat would prefer an unconstrained legislature after democratization. The ANC, however, needed to provide assurance to the white and Zulu minorities that it would respect their views or else risk the very stability of the transition process. These minorities, in turn, sought to ensure that the ANC would not ride roughshod over their interests after the inevitable transition to majority rule. The National Party, in particular, sought to ensure a system of rights protected by constitutional review, as well as other minoritarian devices such as group rights and decentralization.^ These competing interests led to numerous deadlocks in realizing the transfer of power to the black majority. The configuration of the South African transition, with one dominant party that was unable to dictate a constitution, is such that either the insurance or commitment theory provides an intelligible explanation for the emergence of constitutional review. Where the insurance analogy is perhaps more helpful is in explaining how the presence of judicial review enabled the transition to go forward, when it otherwise might not have been able to. The;key point in South Africa's negotiated transition occurred with the decision to use a two-stage constitution-making process.40 The parties would establish an interim constitution based on certain agreed principles, during which period a final constitution would be drafted. Not only wouldthe interim stage include a bill of rights and a constitutional court, 1 his section draws on data presented in Richard Spitz with Matthew Chaskalon, The Politics of Transition: A Hidden History of South Africa's Negotiated Settlement 191-209 (2000). w Spitz, ibid, at 14. Heinz Klug, Constituting Democracy: Law, Globalism and South Africa's Political Reconstruction 140 (2000). I j 56 judicial Review in New Democracies 1 but this constitutional court had the power to certify the proposed final ! constitutional text before it would take effect.41 The presence of judicial review in the menu of constitutional design resolved a deadlock in the negotiation of South Africa's transition, just as the possibility of insurant- allows the conclusion of private contracts that might otherwise not occur. ; Although the ANC might have preferred an unhindered majoritarian con- ' stitution, it was unable to dictate that result to distrustful minorities that I were sure to lose. Because the National Party had an effective veto on the ; timing of the transition, the design of the constitutional order reflected its demand for insurance. The constitutional court became the alierna- ] five forum in which minorities could - and did - challenge the draft final f constitution. The particular design of judicial review, in the form of a special consii- i tutional court, also reflected insurance dynamics. There were significant [ debates over whether constitutional review should be performed by the \ ordinary courts, dominated by appointees of the previous government or : by a designated body. The chief advocate of decentralized review was the | smaller Democratic Party, which had no hope of winning a major share ■ :' seats after the election and was unable to muster support for its position. The debates were resolved in favor of a designated body that would hear cases on appeal from the ordinary courts. Both sides sought to ensure > some control over the composition of the proposed constitutional court \ in further debates on the qualification of potential appointees. The go\ ernment argued for a ten-year period of service as an advocate or judj;c, j which would have effectively barred many nonwhite candidates. Advo- i cates of wider participation argued for the inclusion of academics and <* others in the pool of potential justices. ; The final mechanism, agreed to by both the Nationalists and the ANC, \ was that justices would be appointed by the president of the country, r sure to be Nelson Mandela. Some justices would come from the ranks, [ of the supreme court, and others would be chosen by the president af- j ter consultation with the cabinet. The decision to give the president the j dominant role in forming the court made sense to the ANC; it apparently | also reflected the National Party's mistaken belief that it would have 1 significant role in the first posttransition cabinet and thus influence ov 1 : :4|i|i8|| ■:-:-^fSill :SllJB 41 Republic of South Africa Constitution Act (1993) $71(2) ("The new constitutions text passed by the Constitutional Assembly, or any provision thereof, shall riot ho , of any force and effect unless the Constitutional Court has certified that all of ihi _| provisions of such text comply with the Constitutional Principles....") 1 Constituting Judicial Power 57 court composition.41 If the National Party had some influence over the cabinet appointments and had a strong presence in the ordinary judiciary, it might have significant representation on the court. These demands were based on overoptimism by the National Party. They understood that they were sure to be a minority in the midterm and hence desired constitutional review; but they hoped to be able to influence appointments. The Democratic Party, with no hope of influencing appointments, argued against this proposal and in favor of a role for the nonpartisan Judicial Service Commission. Apparently, this argument convinced the National Party that it had made a mistake in allowing the president such a prominent role in making appointments. Although a role for judges of the supreme court would provide some insurance against an executive-dominated constitutional court, the National Party and the Democrats made a iast-minute, ultimately successful push to expand the role of the Judicial Service Commission in the court appointments. The need for the court in the first place and features of its institutional design reflect political insurance demanded by minorities certain to lose postconstitutional elections. The Israeli system illustrates how judicial review can also be adopted in established democracies as political configurations change.43 Demand for insurance should increase when established political forces believe that they will no longer be able to remain in power. In a deeply divided society at independence in 1947 (as today), Israel's founders chose not to adopt a constitution but rather to use a series of incrementally enacted nonen-trenched Basic Laws to embody the nation's central political principles. For many years, a secular Ashkenazi elite dominated Israeli politics, and the Labor Party ruled uninterrupted for the first decades of the country's history. The election of Menachem Begin in the late 1970s initiated an alternation of power between Likud and Labor Parties. As political outcomes became less predictable, the Israeli Supreme Court became more assertive as the expositor of the constitution. This move was tolerated, and in fact institutionalized, by secular politicians who passed two Basic Laws protecting civil rights and explicitly empowering the court to void any 41 This belief was mistaken. See Spitz, 104-5. * See Ran Hirschl, "The Political Origins of Judicial Empowerment through Constitutionalization: Lessons from Four Constitutional Revolutions," L, Soc. Inquiry ?r (1000). Israel's system of judicial review is structurally similar to the American system, with the exception that judges must retire at age 70. 58 Judicial Review in New Democracies Constituting Judicial Power 59 legislation not in accordance with their provisions and the basic values of the State of Israel.44 These politicians faced increased political uncertainty caused by the rise of religious parties in conjunction with a massive wavu of immigration from Russia. Judicial review was an attractive way of ensuring that the values of the secular Ashkenazi elite remained protected from future attack. What of the American founding? Any general theory of judicial review ought to be able to account for the premier case, namely that of i the United States, though one must also recognize that the theory as 1? have articulated it assumes that judicial review is already on the menní of constitutional design. The conventional account suggests that judicial review in the United States flows not from constitutional text but rather* from the early case of Marbury v. Madison.^ This emphasis on the self-articulation of judicial review by judges is somewhat unfortunate because^ it draws attention away from the important question of how the founders thought about judicial review. This is a complex question; indeed, some consider it to be the central question of American constitutional scholarship.46 Despite these complications, let us consider briefly whether there might be an insurance rationale behind the institution of judicial review. It is important to remember that the United States Constitution was drawn up in an era before the existence of political parties. Therefore,; framing the insurance issue as being one considered by formal political parties makes little sense. Nevertheless, there is plenty of evidences that judicial review was seen to be a minoritarian device, and those demanding judicial review were concerned with minimizing the maxi-f mum harm that could be imposed on them by a majority Furthermore,;; 44 Basic Law: Human Dignity and Liberty and Basic Law: Freedom of Occupation (1991)* 45 5 U.S. (1 Cranch.) 137 (1803). For a discussion of Marbury as central, see, for example, Paul Kahn, The Reign of Law: Marbury v. Madison and the Construction of: America (1997). It is safe to say that this is the orthodox position by examining the central position of Marbury at the outset of the standard American textbooks in constitutional Jaw. See also Robert McCloskey and Sanford Levinson, The American Supreme Court (1994). But see Robert L. Clinton, "Game Theory, Legal History and the Origins of Judicial Review: A Revisionist Analysis of Marbury v. Madison," 38 Am. J. Pol. Sci. 285 (1994) (arguing against the conventional understanding of Marbury); and Robert L. Clinton, Marbury u Madison and Judicial Review {1989): (stating that Marbury only stands for the proposition that judicial review is justified when Congress interferes with judicial power). 46 Jack Rakove, "The Origins of Judicial Review: A Plea for New Contexts," 49 Stan. L. Rev. 1031 (1997}. the configuration during the constitutional bargaining process was one in which thirteen states of various sizes sought to negotiate a union. None of the thirteen was sufficiently large to be able to dominate the others. Rather, each state was concerned that its own welfare would he in jeopardy. In this sense, the key factor was political uncertainty among constituent political units, rather than a dominant party precom-mittifig itself to constitutional constraint, as the "commitment" theory described. The need for insurance is particularly acute in federal systems where a free trade regime is contemplated. Most federal systems provide for some kind of judicial review mechanism to police the law-making boundary between national and local levels of government. This not only reassures the component parts that the center will not trample their rights, but also solves a collective action problem among the components themselves. Free trade in federal systems is endangered by problems of securing credible commitments to the free flow of goods among the component parts.47 Each state in federalist polities, the reasoning runs, would like to sell goods to all other states but, other things being equal, would like to protect its own market. Without a guarantor to ensure that states cannot enact protectionist legislation, this configuration will soon lead to a high-protection, low-trade outcome. It may be in the interests of each state to accept constraints imposed by independent courts as the price for keeping the other states in line as well. |1 Federalism provides an important rationale for active judicial review in comparative terms, evidenced by American history and also emphasized by "realist" protagonists in debates over the role of the European Court of Justice in European integration.48 Federal polities illustrate how political diffusion promotes judicial power. The free trade rationale can be stated in terms of insurance needs or in terms of precommitment on the part of the constituent units of the federation, illustrating that the commitment and insurance theories need not always be inconsistent. 4?::See Martin Shapiro, "Federalism, the Race to the Bottom, and the Regulation-Averse Entrepreneur," in North American Federalism in Comparative Perspective (Harry Scheibcr, ed., 1992.). 4it Geoffrey Garrett, "From the Luxembourg Compromise to Codecision: Decision Making in the European Union," 14 Electoral Stud. 289 (1995); Geoffrey Garrett, "The Politics of Legal Integration in the European Union," 49 Int'l Org. 175 (1995); Geoffrey Garrett and George Tsebelis, "An Institutional Critique of Intergovern-mentalism," 50 Int'l Org. 269 (1996); cf. Anne-Marie Burley and Walter Mattli, ."Europe Before the Court," 47 Int'l Org. 41 (1993). 60 Judicial Review in New Democracies The Design of Judicial Review: Empirical Evidence The above examples illustrate that the insurance theory has explanatory power in several prominent cases of the establishment of judicial review. In this section, we develop a more systematic empirical test of the insurance model by examining the constitutional courts adopted in Latin America and the former Soviet bloc in recent years. Nearly every postcommunist country has adopted a constitutional court, usually following the German model of a centralized body. Latin American countries also began to move to this model, though some countries retain the decentralized model of review. The details of institutional design vary across countries. Table 2.3 presents some data on these countries and their constitutional courts. We consider some of the dimensions of institutional design mentioned above. To examine whether demand for political insurance is a determinant of constitutional court design, we must evaluate the relationship between demand and those features of court design predicted to produce more-accurate constitutional review. To capture demand for insurance, we use a proxy variable "Party Strength," the difference in the first postconsti-tutional election between the seat shares of the strongest and second-strongest parties or blocs of parties in the legislature. This captures the extent to which there is a dominant party and should correlate wilh the degree of political uncertainty during constitutional drafting.49 The lower the differential between seat shares, the less certain will be the leading party or bloc that it will end up in power. Note that in most cases we cannot use the political configuration before democratization, as the former configuration may have been a one-party system that did not reflect 45 For our purposes, this indicator is superior to another one frequently used in comparative political studies, namely the effective number of parties. The effective number of parties is A£ — i/Spf where pj equals the percent share of seats in the legislature of the < th party. Markku Laakso and Rein Taagepera, "Effective Number of Parties: A Measure with Application to West Europe," 12 Comp. Pol. Stud. 3 (1979); Rein Taagepera and Matthew Soberg Shugart, Seats and Votes: The Effects and Determinants of Electoral Systems (1989); John Ishiyama and Matthew Velten, "PresidenU.il Power and Democratic Development in Post-Communist Politics," 31 Comm. Post-Comm. Stud. 2.17, 222 {1998). Effective number of parties might correlate inversely with political uncertainty as the smaller number of parties indicates a greater chance of each to capture seats in government. However, it would not capture the situation of political deadlock between two equally large parties, which would create high uncertainty but a low number of parties. Thanks to Omri Yadlin for point- : , ing out this problem in an earlier version of this chapter. See also Tom Ginsburg, "Economic Analysis and the Design of Constitutional Courts, 3 Theoretical Inquirii L. 49 (2.002.). Constituting Judicial Power 61 Table 2.3 Constitutional and Supreme Courts in Postsocialist Countries and Latin America Constitution Court Term in Access Party Country Year Size Years (Dummy) Strengt Postsocialist Albania 199T 9 9 i 0.37 Armenia J995 9 life 0 0.58 Belarus 1994 11 11 0 0.03 Bulgaria 1991 12 9 0 0.17 Czech Republic 1993 10 0 0.04 Estonia 1992 *7 life i 0.21 Georgia 199S 9 10 i 0.31 . Hungary 1949/1990 15 9 1 0.18 Lithuania 1992 9 9 0 0.39 Macedonia 1991 9 9 r 0.24 Moldova 1994 6 6 i 0.37 Mongolia 1992 9 6 i 0.2 Poland 1997 12 8 i 0.05 Rumania 1991 9 6 0 0.59 Russia r993 J5 life 1 0.06 Slovakia *993 10 7 i 0.28 1 Slovenia 1991 9 9 i 0.09 Ukraine 1996 19 9 i 0.19 Latin America Argentina 1994 9 life i 0.07 Bolivia 1967/1994 12 10 1 0.05 Brazil 1998 11 life i 0.04 Chile 1997 7 8 0 0.05 Colombia 1991 9 8 1 0.47 Dominican Republic 1994 16 life 0 0.23 Ecuador 1998 9 life i 0.07 El Salvador 19S3 5 9 i 0.33 Guatemala 1993 T3 5 0 0.28 Haiti 1987 10 10 i 0.12 Honduras 1982 9 4 i 0.12 . Mexico 1917/2001 it life i O.II Panama 19514 9 10 0 0.03 Paraguay 1992 9 life i O.I ■ Peru 1993 7 5 0 0.44 Uruguay 1997 5 10 i a. i Venezuela 1999 15 9 r 0.28 Mean 10.57 8.97 0,68 0.22 Note: Constitution dates reflect major amendments for Albania, Bolivia, Hungary, and Mexico. Certain institutional features, such as the life terms for Russian justices and the size of the Hungarian court, may have been modified subsequent to the date given here. Because our argument concerns initial design, we do not reflect these changes in the table. For purposes pointment to the constitutional court. The prediction is that as the level of party dominance rises, the term of judges will fall. This is because reappointments and short-term length give politicians the ability to influence judges, especially if a party anticipates staying in power through multiple reappointment cycles. In practice, judges may not actually serve as long as provided in nominal appointments, but the constitutional courts of Eastern Europe are too young to have reliable data on actual time served. There is the additional problem of assigning term length for purposes of statistical tests to judges with lifetime appointments. In the data analyses that follow, we therefore assume, somewhat arbitrarily, that "lifetime" np-pointments are eleven-years long, precisely the same length as the longest designated term in the dataset. Figure 2.3 shows the relationship between term length and parly strength in scatterplot form. In Figure 2.3, the countries tend to cluster in either the lower-right or upper-left quadrants. The lower right represents c 11) Brazil Russia Paraguay Ecuador Estonia Dominican Republic □ □ no 00 □ □ Armenia □ Republic Bolivia Uruguay Haiti Panama Georgia aa a a aa Slovania Bulgaria Hungary Ukraine Macedonia Venezuela El Salvador Albania Lithuania 0 □ □ a a □ Chile Poland a Colombia □ Slovakia □ Mongolia Moldova □ a Rumania a Guatemala □ Peru a Honduras □ 0.0 Party Strength figure 2.3. Term Length and Party Strength 20 18 16 SCzech Republic Russia + + in a ®. S3;-- z Poland Bolivia Bulgaria Brazil + + + Belarus Mexico + "'"Haiti + Ukraine + Estonia + Dominican Republic Hungary + Venezuela + + Guatemala + Slovakia + Ecuador Paraguay Mongolia Macedonia Georgia Albania Lithuania Colombia Armenia Argentina +++*■ + ++++++ + ++ Slovenia Honduras Panama Rumania Chile + Uruguay + Moldova El Salvador + Peru + Party Strength figure 2.4. Court Size and Party Strength a strong party with short terms, while the upper left represents a weaker party with longer terms for constitutional or supreme court justices. We note that Armenia is somewhat of an outlier, featuring a dominant party with lifetime appointments. Figure 2.4 presents a similar scatterplot diagramcomparing the size of the court with the strength of the dominant p.irty and shows similar results. ''Access" is a dummy variable that captures standing. The value is 1 if citizens have the right to petition the court or if ordinary courts can refer constitutional questions to the court. Thus, both a decentralized system such as that of Israel and a centralized system like that of Germany would carry an "Access" value of 1. Systems of limited access where only designated political institutions can bring questions to the court have an "Access" value of o. This is a feature of the French model but also is found in some courts that otherwise look like the German model. The predicted relationship between extent of party dominance and access is negative. The stronger the dominant party in constitutional drafting, the less incentive there is to design an open system of access to the court. To test whether the relationship is as predicted, Figure 2.5 presents the resqlts of four separate least-squares regression operations with "Party Strength" as the sole independent variable. The dependent variables are ■ 64 Judicial Review in New Democracies Constant Regression Coefficient (t-stat) Regression Confidence Level ill ■ B Regression one: y = court size 11.70 -5.26 (-1.46) 85% Regression two: y = term length 9.94 -4.77 (-2.30) 97% Ilfi i Regression three: y = access dummy .87 -.89 (-1.78) 91% Regression three: y = normalized index of court size, term length, and access 1.26 -5.85 (-3.08) 99% 11 IIB ili Si 1 ill N = 35 figure 2.5. Regression Results: Insurance Model of Design court size, term length, access, and an index variable, summing each of the other three variables after normalizing them. The regressions demonstrate strong results for all three dependent vari- j ables. All coefficients have the predicted sign, and the results for term ; length and the index variable are statistically significant. Three features J thought to enhance independence and accuracy of the court are those that ! are chosen in diffused party systems, where politicians should have an ii centive to do so. This suggests that the insurance model has substami explanatory power. If the precommitment model were superior, we would ' have expected to see that in many cases stronger parties led to more open i access, longer terms, and larger courts because there would be greatn ' need for precommitment. - ,| To summarize the argument so far, judicial review provides an insur- j ance policy for prospective losers in the electoral arena. The design oi die system will reflect in part the configuration at the time of constitutional drafting, with the availability and power of judicial review increasing with political diffusion. In this sense, judicial review reflects democratization and is not antidemocratic, as asserted by theorists who focus on the coun^ termajoritarian difficulty. While the precise institutional design has been less uniform than the spread of the practice itself, there are strong trends toward adopting a German-style designated constitutional court who: members have limited terms, open not only to particular political actors but also to courts or ordinary citizens as well. This open-access design noi only ensures access to judicial review by prospective minorities but al: provides courts with opportunities to become involved in a wide range of -cases and to build up their power over time. It is to this process that w-turn in the next chapter. 3 Building Judicial Power Chapter 2 focused on the creation of constitutional courts and argued that the design of judicial review is to a large degree a function of politicians' insurance needs. However, courts that are created to do one thing can gradually adopt new roles for themselves. Courts are not passive players 111 the judicial review "game." Although politicians design judicial review with their own interests in mind as a way of reducing future political uncertainty, there is substantial evidence that courts behave strategically once they are established, both with respect to individual cases and with respect to their own position in the constitutional system. judicial activism leads to a potential problem with regard to the insurance theory. If courts are able to assume roles that differ from those anticipated by constitutional designers, would not constitutional designers discount the value of the insurance provided by constitutional courts? In other words, would not prospectively weak constitutional designers want to specify in some detail the norms to be used by courts in constraining political authorities? From the point of view of prospective minorities, however, this is not a problem as long as there is some positive probability that the court will use its powers to constrain political majorities. While the designers will try to channel judicial decision making into certain areas, for example, by specifying jurisdiction, enumerating rights to he protected, and listing sources of law to be considered - the intertemporal nature of the insurance contract means that they cannot do so with perfect confidence.1 The question for designers is always whether they are ' See Stephen Holmes, "Precommitment and the Paradox of Democracy," in Constitutionalism and Democracy 195-240 (Jon Eister and Rune Slagstad, eds,, 1988). 65