1.   Constitutional Courts as Deliberative Institutions:
Towards an Institutional Theory of Constitutional Justice
John Ferejohn and Pasquale Pasquino
The collapse of the communist empire coincided with a new wave of constitution making. AO the Eastern European countries included in their new constitutions an organ in charge of constitutional adjudication that is modeled on what we define below as the Kelsenian type. This was neither surprising nor original. The same phenomenon occurred in Southern Europe about 20 years ago, after the collapse of fascist authoritarian regimes in Greece,1 Portugal and Spain. Likewise, after the Second World War, a similar process took place in Austria, Italy and Germany. For this reason the following discussion of constitutional justice in post-authoritarian regimes is pertinent to the recent developments in Eastern Europe.
This institutional revolution can be understood in terms of a simple and more or less unified model.2 The idea is straightforward: each of the European states is committed to maintaining parliamentary authority over the executive and judicial departments. And, as a matter of sociological fact, each of the post-authoritarian states exhibited distrust either of the judiciary (in the post fascist states) or the parliament (in France) or both. Each therefore chose to introduce a model in which constitutional review of parliamentary actions would take place in a specialized court, outside the judicial system. That model, because it puts some legislative policy making in the hands of constitutional courts (and not merely negative legislative authority as Kelsen himself was forced to admit3), places the additional burden of legitimation on those courts. To some extent the burden of legitimation can be addressed by insulating the justices from political
1 Greece has been familiar with judicial review since the 19th century; it is interesting though to note that only the constitution enacted in 1975 at the end of the military regime introduced (art. fOO.le) a Special [and specialized] Supreme Tribunal overarching the diffuse system of judicial review.
2 It may be useful, in making sense of the success of the Kelsenian model, to separate parliamentary sovereignty theory with its hierarchical doctrine of the separation of powers naturally hostile to any judicial review, from sociological expectations, like the distrust of the constitution makers towards judges educated in authoritarian regimes, or the distrust towards parliaments that have a record of complicity with fascism or communism.
3 See 'Wer soil der Hiiter der Verfassung sein?' (1931), Italian translation in: H. Kelsen, La giustizia costituzionale (Milano: Giuflfre 1981), p. 260.
CONSTITUTIONAL JUSTICE, EAST AND WEST
pressure and by permitting them to craft procedures (such as holding sessions in private, issuing opinions on behalf of the court, etc) that produce impartiality or the appearance of impartiality. But, beyond these more or less structural assurances, constitutional courts need to provide reasoned justifications for their decisions. We argue, in short, that constitutional courts face special and demanding deliberative expectations. In fact, to some extent, these expectations are embodied in the constitutions that created these courts. For example, the organic law of the Fifth French Republic regulating its procedures requires that the Constitutional Council provides a reasoned justification for its holdings.4
In Political Liberalism, John Rawls described courts as exemplary deliberative institutions - forums in which reasons, explanations, and justifications are both expected and offered for coercive state policies.5 The authority of courts is supposed, on this view, to rest in large part on the qualities of judicial reasoning - reasons linking court decisions to legal or moral authority - especially since courts as institutions lack democratic credentials and often lack the means to implement their decisions. So, deliberation and reason-giving seem especially valuable (and familiar) aspects of adjudication. If, therefore, we are trying to locate the institutions where reasoning and deliberation play an important role in public life, it is apt to begin with courts and especially with courts dealing with constitutional issues.
In this chapter, therefore, we compare European and American constitutional courts as deliberative forums. We argue that constitutional courts are very differently situated in various political systems. They are asked different kinds of questions by different political actors, and are faced with different expectations, histories and cultural and political constraints. In view of this diversity of circumstances it is to be expected that constitutional courts adopt different kinds of deliberative practices even when treating quite similar issues. Still, despite the diversity, we think there is an important sense in which each of the constitutional courts we examine - the French, German, Italian, Spanish and U.S. courts -have retained the exemplary deliberative character that Rawls describes.
1. DELIBERATIVE EXPECTATIONS
What kind of deliberation are courts asked to undertake? Aristotle's conception of deliberation puts the main emphasis on the requirement that deliberation
4 The Constitutional Council shall give a reasoned decision' (Ordinance no. 58-1067 of 7 November 1958 incorporating an Institutional Act on the Constitutional Council; art. 20). See also, concerning Italy: Law no. 87 of March 11th 1953, The composition and procedures of the Constitutional Court, Art. 18: 'Judgements are issued "In the name of the Italian people", and set out the reasons for the decision...'.
5 John Rawls, Political Liberalism (New York: Columbia University Press), pp. 231-36: 'The Supreme Court as Exemplar of Public Reason'.
CONSTITUTIONAL COURTS AS DELIBERATIVE INSTITUTIONS
aims at critically evaluating, and perhaps changing, goals or preferences. In deliberating, people come to recognize and embrace reasons for action that they may not have understood previously. In public deliberation, moreover, the kinds of goals or preferences that agents might be expected to bring to the forum are especially likely to be subject to revision insofar as, at least within a liberal legal system, private and public purposes must fit together in complex ways.
But whether goals or purposes actually change as a result of deliberation, or whether they merely remain open to revision, the way that deliberation changes or reinforces goals or purposes is by giving reasons or arguments. Deliberation in this sense is participating in the process of reasoning about public action. This entails being open to reasons - that is, being willing to alter your preferences, beliefs or actions if convincing reasons are offered to do so - and being willing to base attempts to persuade others on giving reasons rather than threatening coercion or duplicity. Rather than asking, like Aristotle, whether minds are changed, we would ask only whether public action is conducted through reasoning. Minds may rarely actually change as we suspect they seldom do on the American Supreme Court, at least not at the level of deep political commitments. Nevertheless, even if its members remain quite fixed in their general purposes, the Court is nonetheless a forum_of reasonand deliberation in another sense. The fact that court decisions have effects on real or hypothetical cases, and are aimed at deciding what can or should be done in concrete settings, permits judges who disagree about fundamental principles to find agreement on more concrete levels. While abstract ideas may not be open to revision, ideas about particular cases may be much more flexible and pragmatic.
Moreover, high courts, unlike other political institutions, do not simply publish orders or decisions. They are expected also to publish plausible rationales for their holdings: arguments that others can be expected to respect and embrace, vvhether or not their own interests have been vindicated. The expectation that courts explain their holdings can be seen as a deliberative expectation in two senses. The first, just given, is that reasons are given that can be understood and embraced as, in some normative sense, our own reasons for action. Courts offer, in: Rawls' idiom, public reasons for action - reasons of a kind each of us can be expected to embrace from our own moral vantage point. Secondly, since courts are collegia! institutions, these reasons are arrived at through an internal process of deliberation, guided by the particular court's decision-making norms. This process may or may not be regulated by a shared expectation that the court will publish a single opinion or that multiple opinions will be published as well. A court's published reasoning is, in this sense, negotiated within a normative framework ranging from consensus seeking to majoritarian.6
The changing practices of the U.S. Supreme Court regarding the (increasingly frequent) publication of dissents and concurring opinions is an example of the historical flexibility of: these decision-making norms. Perhaps, in an increasingly pluralist culture, there are good; reasons to offer multiple rationalizations for holdings and that is the best explanation f<,r why the Court's normative practices permit them.
constitutional justice, east and west
Reason-giving seems especially important for the authority of courts in two respects. The first we might term democratic: because (most) judges are not electorally accountable, the reasons offered in their decisions - especially if some | of those reasons are rooted in the (positive) acts of electorally accountable bodies - can provide indirect democratic justifications for public actions. And even if judicial deliberations are not traceable to legislative acts, if they are rooted in (constitutional or moral) principles that in some way underlie democratic government or that are presupposed by a democratic people, judicial reasons can still Jf be seen as indirectly or transitively democratic. Judicial opinions and decisions f are, in this respect, the working out of democratic principles in new circumstances 'I and in particularities of specific cases. : |
In this sense, deliberative judges may actually enhance the powers of democratically elected officials by offering them a more flexible and intelligent system to J implement and refine their own legislation and public ordinances. If this is what ~j j legislators expect of judges, it should follow that democratic legislators would f be eager to expand the judiciary and enlarge judicial powers, purely in the | pursuit of empowering the democratic branches, and in enlarging the people's 9 capacity for collective action. Alternatively, insofar as they are acting on constitu- . I tional or moral principles, deliberative judges may enhance the authority of the I people (either as a collectivity or as individuals) as distinct from their democratic I representatives. Rawls' own account of public reason might be seen as emphasizing the democratic character of judges in the sense that public reasons are , | reasons that we can all, whatever our comprehensive views of the good, be 4 expected to endorse. Public reasons, on this view, have a kind of (non positivistic) j democratic character at least insofar as most citizens adhere to one or another comprehensive view of the good. But, of course, judicial deliberations have risks for the legislature as well. Judges, acting on reasons, can modify or nullify t democratic commands, and demand of the pattern of legislation a kind of ) coherence or rationality that forbids elected officials from making certain kinds of policies, however politically attractive they may be. But whether such risks are to be seen as pro or anti-democratic depends on connections between the legislature and the polity, amongst other things.
The second way in which reason-giving is important for judicial authority is the way that judicially provided reasons allow others - state officials, other _: judges, lawyers, ordinary citizens, etc. - to anticipate the implications of the current decision for future cases. Reason giving, in this sense, is efficiency enhancing: it plays a predictive or coordinating role rather than a justificatory one; it helps to permit others to choose their actions intelligently in light of their likely consequences. It helps to perfect the rule of law in light of experience. Reason giving in this second sense aims at a certain kind of efficiency - permitting courts j to channel or coordinate lots of private and public actions without too much '% wasteful litigation. Of course, efficiency can be a justification as well. If the deliberative activity of courts permits society to more effectively coordinate -Jf
constitutional courts as deliberative institutions
actions and to achieve important common purposes, there is more to be said for according respect and allegiance to courts. But this kind of justification has nothing to do with democracy and would be available to courts in any legal system.
While deliberation and giving reasons are important sources of legitimacy for judges in any legal system, they seem to bear particular weight in democracies. The expectation or demand that political authority be seen as flowing, directly or indirectly, from the people together with the notion that democratic institutions must respond to popular demands places judges in a peculiarly weak position. This is, of course, at the root of the worries over the countermajoritarian aspects of judicial review.
2. DEMOCRATIC DELIBERATION
In contrast to what we expect from courts, we do not demand as much by way of reason-giving from our other, more evidently democratic, institutions. True, we might expect that unelected bureaucrats provide some rational connection between their decisions and those of the legislature. Such a demand arises because of the weak democratic pedigree of public agencies: their connection with ordinary voters is remote, operating through a chain of election and delegation that is only as strong as its weakest link. In any case, at least as things have evolved in the United States and most western democracies, we do not demand that rationales for bureaucratic action be either clear or convincing. Administrative actions should be reasonable in the sense of using means more or less proportional to legislated ends, as interpreted by the executive or the agency itself. Usually we trust that the fact that agencies are at least indirectly accountable to voters permits us to rely on their exercise of authority, without always demanding extensive deliberation or reason giving.7
Slill less do we ask that the legislature itself provide reasons for its actions, :at least not in any legal forum. Citizens in most democracies expect political parties to elucidate some kind of thematic program that might serve to explain or justify the legislative program they would push if given the chance. But such manifestos or platforms fall very far short of providing closely reasoned justifications for actual legislative proposals. In any case, the only recourse for a poor justification is subsequent electoral rejection at a more or less distant time and
7 Obviously, the extent to which such trust is placed in administrative decisions depends on V the nature of the administrative actions in question. When agencies are judging individual claims concerns with due process lead to deliberative expectations of the kind imposed on courts. Where they are setting general or abstract policies, such expectations tend to be displaced to the electoral process. If voters think that the current administration is not being vigorous in protecting environmental values, they can replace agency leaders by installing another administration.
CONSTITUTIONAL JUSTICE, EAST AND WEST
in circumstances in which the reasons for electoral failure are likely to be obscure. We may expect to see, but hardly demand, a preamble before a legal text, but the content of that preamble is seldom taken as seriously as the actual provisions of the law that follows. Almost never is the understanding of the law itself seriously modified by such preambular expressions. While some legal scholars urge that legislative actions be held to some kind of deliberative test - as a way of forcing legislative deal-making to be responsive to public rather than private values - as things stand neither the courts nor the general public have been willing to establish or enforce such expectations. Thus, in spite of some hand-wringing by legal scholars, we permit the legislature to engage in logrolls, build checkerboard statutes, enact Christmas tree bills and make other kinds of bargains pretty freely and we impose legal restrictions only where there is a conflict between the final legislative action and a constitutional value or norm.
Finally, when it comes to electoral decisions, far from requiring deliberation or reason giving, our public norms seem to forbid it. In this context we are thinking specifically of the secret ballot, and restrictions on campaigning near polling booths, but the more general endorsement of the one-person-one-vote view of electoral democracy and the growing efforts to regulate and restrict campaign contributions and spending, work to make democracy at the electoral level a matter of numbers, independent of reasons or reasoning. Moreover the reluctance of courts to look behind electoral results (including popular initiatives and referenda) to probe for inadmissible intent of the kind that would routinely be sought in legislative records is another suggestion that we take the ballot box to be a reason free zone - at least in the sense that the exercise of the franchise ought not to depend on the capacity of the voter to give or have good reasons. She just has the right to vote and is free to use it however she wants. In effect, if reasons are to be reflected in the electoral process, they must work through the individual voter, by persuading her to cast her ballot based on one set of considerations rather than another. But in the end, however they choose to cast (or not cast) their votes, it is numbers and not reasons that count.
The reasons for such limitations are pretty obvious. Theorists since Plato have worried about the corruptibility of public deliberation. They feared that ordinary citizens would be too easily swayed by seductive rhetoric, or by bribes, or that they would be intimidated by public officials or private interests. Such fears historically led to a deep suspicion of political parties, interest groups, or other formations that could overwhelm the ordinary citizen. Rousseau and Madison, to take two prominent examples, regarded the public arena as sufficiently corruptible that they went to great lengths to seek institutional reforms capable of channeling these powerful tendencies. Rousseau, in the Social Contract, would prohibit the formation of parties and other factions if possible, and prohibit legislative proposals from addressing private individuals and groups. And while he deplored the need for the Romans to have adopted secret balloting, he recognized that the corruption of the Republic was too advanced to permit open
CONSTITUTIONAL COURTS AS DELIBERATIVE INSTITUTIONS
expressions of the franchise. Madison, who was more sceptical of limiting or regulating private interests directly, proposed multiplying them in the interest of limiting their influence in the public forum. It took another hundred years for Americans to adopt secret balloting as a way of limiting the corruption of the franchise.
We have sketched a rough range of deliberative expectations of institutions in .a democracy. The most democratic institutions, in the sense of closeness to the people, are not expected to be deliberative at all and, indeed are surrounded with impediments to deliberation, whereas the least democratic decision -making institutions are expected to be conducted as more or less pure forums of public reason. There are several possible explanations for these divergent expectations.
In one sense, deliberation or reasoning might be regarded as a substitute for more directly democratic legitimacy. If you have the votes, you don't need to persuade others to follow you because they have essentially already promised to do that by electing you to office. If not, persuasion and reasoning is the only way to go. Deliberation and reason-giving are, in this sense, ways in which non -democratic institutions can go about getting people to go along with their decisions. In this sense, deliberating and giving reasons are special burdens that courts and agencies must bear if they are to make authoritative decisions.
Another possible interpretation is that, built deeply into western democratic traditions is the suspicion that deliberative requirements are somehow antidemocratic or elitist. To tell the people or the legislature that having a majority is not enough and that you must also have good reasons to act is to disable or weaken something essentially important in democratic life. This position is not necessarily anti-deliberative, but it is sceptical about the imposition of deliberative requirements on majority action. If reasons and deliberations are to play a role in shaping the law, they must operate through the votes of citizens and their representatives. If reasons are good, then they will sway majorities one way or the other and be reflected in their votes, at least if the tendencies to corruption are limited. If not, well, the reasons must not have been very good in the first place. Reasons, in a democracy, should not get counted twice.
A third possibility is this: in a constitutional democracy, public policy choices take place within a hierarchy of norms - moral and constitutional norms that appropriately limit democratic choices. When engaged in regulating the boundaries: between such norms and democratic choices, reason-giving and deliberation is especially important. After all, what is being contemplated in such regulation is telling the people or their representatives that a certain policy may not be permitted. Courts and administrative agencies (when acting like courts) therefore are burdened with deliberative expectations in precisely those areas where constitutional values may be at stake. The legislature, insofar as it acts as a constitutional: interpreter, would face similar deliberative expectations, but not when it is engaged in making policies of the kind that are safely .within the bounds of constitutional acceptability.
CONSTITUTIONAL JUSTICE, EAST AND WEST
CONSTITUTIONAL COURTS AS DELIBERATIVE INSTITUTIONS
3. CONSTITUTIONAL COURTS IN MODERN DEMOCRACY
Roughly speaking, we may describe the institutional position of constitutional courts as situated aloug a single dimension, with pure parliamentary sovereignty regimes, such as Britain or the French Third and Fourth Republics at the left and a Montesquieuian separation of powers regime on the right. Between these two poles are a variety of constitutional systems that mix aspects of legislative supremacy with other more or less independent institutions. Parliamentary sovereignty regimes by definition regard both the executive and the courts as subordinated to the legislature, with their role being to implement and enforce the legislator's commands. The notion of judicial review of legislation is, for this reason, completely alien to such legal regimes and, as a result, the way in which legal institutions can lead to statutory revision takes on a particular institutional form in such settings.
In general, parliamentary sovereignty regimes must place constitutional review inside the legislature. Otherwise parliament would not be sovereign. This implies that in parliamentary sovereignty regimes constitutional review can only be a priori - prior to the promulgation of a legislative proposal as law - and abstract, in the sense of involving only the comparison of legislative and constitutional texts. Moreover, in such systems, while advice as to constitutional principles may be widely sought, the authority to invoke constitutional principles will have to be concentrated in a small number of hands internal to the legislature. Thus, in Britain, France, Finland and Sweden (until very recently) whatever constitutional review takes place occurs wholly inside the legislature. Traditionally, this review may be concentrated in an upper chamber, which may have some kind of negative legislative authority (such as a suspensive veto), or a judicial committee of some kind, or perhaps in a separate institution such as the French Conseil d'Etat, which can advise the legislature on constitutional issues. More recently, the French Fifth Republic has devised what may be termed a third legislative body - the Conseil Constitutions - that can modify or veto proposed legislation.8 This institution was originally invented by the Gaullists as a way of rationalizing or controlling parliament, but it has evolved in such a manner as to permit general constitutional review of governmental legislative proposals.9 But wherever the review or advice originates, the authority to apply constitutional principles to legislation rests with the legislature itself.
8 Alec Stone, The Birth of Judicial Politics in France (Oxford: Oxford University Press), 1992, and John Bell, French Constitutional Law (Oxford: Clarendon Press), 1992.
9 This transition occurred at two critical moments in recent French history. First, in 1971 the Conseil asserted the authority to strike down governmental legislation based on broad (and uncodified) constitutional principles (the so called bloc de constilutionnalite). In this particular case, the Conseil asserted that there was a right to free association that could: be discerned from the 'fundamental principles underlying the republic' In the same opinion the 1789 Declaration of the Rights of Man and the preamble to the 1946 Constitution were elevated to the status of constitutional norms that could be used to overturn legislative
Division of power regimes, again by definition, hold the legislative, executive and judicial powers to be separated horizontally rather than vertically, and such regimes tend to attribute constitutional authority to institutions exercising each of these powers. Montesquieu, of course, famously defined despotism as a circumstance without separated powers, arguing that undivided powers inevitably produced arbitrary and unpredictable rule. He thought it particularly important that courts, when applying law to particular cases, were not legislating in any important sense. The judicial power was in this sense a pouvoir nul. But this did not mean that judges (or, perhaps more precisely, juries) exercised no discretion in applying the laws. A court could refuse to apply a law to a particular case if its application would produce injustice. This would not nullify or abolish the law but would set its effects aside in that particular case. This kind of review authority, the power to interpret law and facts in application, produces a characteristic kind of judicial review that is quite distinct from that found in parliamentary sovereignty regimes. Review authority is dispersed throughout the judiciary - any court must interpret the law, constitutional or statutory, in order to apply it. It is exercised in the context of concrete cases and is neither abstract nor a priori. Finally, it is not legislative in the sense that laws are not abolished but i only refused application to the case (and, depending on the legal system to similar cases).10
As an example, Article III of the U.S. Constitution places the Supreme Court at the head of the federal judicial department, permits constitutional review of statutes to arise only out of genuine cases or controversies, and permits any court to engage in such an inquiry.11 The U.S. system of constitutional adjudication is, in this respect, completely different from the French and most other European models. Courts, in applying statutes must always read them in view of the Constitution, and never apply them in ways that would violate constitutional protections. In view of this requirement statutes are often given interpretations: fitting them into the constitutional scheme and are occasionally given no authority at all. While such actions may have something like legislative effects,
proposals. The second critical moment occurred in 1974 when the government, fearing electoral defeat, successfully urged a constitutional amendment permitting any 60 members .of the legislature to refer legislation to the Conseil Constitutions!. Thus, after 1974, the Conseil focused increasingly on reviewing governmental initiatives with the powerful constitutional tools it created in 1971.
The nim-lt'^i.ilcitivc character of judicial review is a theoretical idea. Judicial systems with y .dispersed review will typically give order and predictability to judicial actions by devising [ methods of hierarchical control of lower courts that create coherent rules or doctrines out ; of diverse decisions. The effect of such developments is to make judicial action more 1 legislative as statutory rules are supplemented with judicially crafted ones. In rnjiny respects, the Norwegian system of constitutional adjudication resembles the .American in that such review takes place in concrete cases and is dispersed throughout ordinary courts,
CONSTITUTIONAL JUSTICE, EAST AND WEST
they do not formally change or eliminate statutory texts in the manner of the French system. Unconstitutional statutes are not repealed or eliminated. They are simply given no application to particular disputes. Putting matters this way emphasizes the judicial rather than legislative aspects of judicial review.
While most European legal systems have parliamentary sovereignty traditions - in the sense that both the executive and the judiciary are subordinated to parliament - developments in the twentieth century, particularly following the two world wars, have tended to erode these commitments. Thus, Austria after WW and Germany and Italy after WW II (and Spain and Portugal after the collapse of fascist regimes), adopted constitutions that departed from the hierarchical parliamentary model in important ways. For convenience, we may call all of these regimes Kelsenian, after the eminent Austrian jurist who invented its distinct institutional form. Each of the European constitutions reflect, in various ways, Kelsen's central idea that constitutional adjudication is more of a legislative than a judicial function. When a constitutional court strikes down a statute it is not only legislating in the negative sense of abolishing a law, but insofar as it must reconstruct the legal situation before the statute, it is legislating positively as well.13
Kelsen's notion of constitutional adjudication emphasized this legislative aspect by conceiving it as involving a comparison of a statutory and a constitutional text. Such abstract review arises not out of a fact specific case with real (harmed) litigants claiming rights, but as an a priori and abstract comparison of texts. Constitutional adjudication seen this way seems inherently political, in the sense that a constitutional court must deliberate and choose from among alternative normative rules for regulating social conduct. As a result, Kelsen thought that constitutional courts should be placed outside the judiciary as well as the other governmental departments. Their powers were to be exercised by politically appointed judges, usually drawn from people particularly competent at making abstract comparisons among texts, and with the capacity to deliberate
12 Austria was the first country on the European continent to introduce in 1920 a Constitutional Court. Hans Kelsen played a crucial role in establishing this institution. It is important to take into account that historical antecedents of such an organ were courts adjudicating conflicts between the central government and the Länder in the Austro-Himgarian Empire as in the Holy German Empire (Reichskammergerich).
13 This is even more obviously the case when a constitutional court construes a statute in light of constitutional values. Kelsen himself hesitated in defining the role of the Constitutional Court. At the beginning he spoke of 'negative legislation', but later on, answering C. Schmitt, he accepted, in his book Wer sol! der Hütet- der Verfassung sein? (1931), that the Court plays a positive legislative function. See P. Pasquino, 'Gardien de la constitution ou justice constitutionnelle? C. Schmitt et H. Kelsen'; in 1789 et ['invention de la constitution, sous )a direction de M. Troper et L. Jaume (Paris: Bruylant - L.G.D.J.); 1994, pp. 141-152.
CONSTITUTIONAL COURTS AS DELIBERATIVE INSTITUTIONS
about norms and explain decisions, and not necessarily from those with judicial experience.
3.1. Constitutional Adjudication in Post-Authoritarian Regimes
Kelsen's ideas have proved especially attractive to post-authoritarian regimes. Not only were they adopted in Austria, Germany, Italy and Spain, but they have also taken root throughout Eastern Europe after 1989. Each of the post-authoritarian constitutions put in place institutions of constitutional adjudication that permitted constitutional review of legislative, and sometimes of executive and judicial acts. But in each of these cases, the very fact that there was a transition under way from an old and distrusted regime to a new one, meant that judges were viewed with particular suspicion, as potential holders of constitutional review authority.14 As a result, there were powerful political reasons to place constitutional adjudication outside the judiciary - in effect the reviewing body was placed above each of the other institutions in position to review any governmental action from a constitutional perspective - and given to a specialized and politically appointed body.
While Kelsen emphasized abstract constitutional review, all of the modern post-authoritarian constitutional courts have concrete a posteriori review powers as well. Access to these courts is controlled not only by governments and political minorities, but also by ordinary litigants in the context of specific cases, or, as in Italy, by ordinary law courts. Thus courts may be asked, in a Kelsenian fashion, to compare constitutional and statutory texts abstractly (by direct referral of a constitutional issue), or they may be presented with a constitutional issue that arises in an ongoing case before a lower court, or they may be presented with a whole decided case (as happens in both Spain and Germany).15 In any of these situations, the actual authority to nullify or modify legislation is generally concentrated in the constitutional court and not dispersed throughout the judiciary. If an ordinary court doubts the constitutionality of a law, it must stay the proceedings before it and refer the question to the constitutional court fór. determination.16
The makeup of these constitutional courts is distinctive as well. Because the
M- The judiciaries of each of these systems are essentially closed career hierarchies that are particularly insulated from outside influence. This extreme insulation of judges made it even less likely that important additional powers would be vested in them following the collapse of authoritarianism.
15 This is the traditional practice also for appeals to the European Court of Human Rights in Strasbourg, which reviews cases only after litigants exhaust domestically available remed-
j^i;: ies. With the incorporation of the European Convention on Human Rights, domestic courts increasingly give direct application to human rights law.
Ifv Courts applying European Community Law follow the same practices.
CONSTITUTIONAL JUSTICE, EAST AND WEST
separation of powers systems tend to have dispersed and concrete review, ordinary judges can be expected to develop competence in constitutional adjudication. In post-authoritarian systems, however, where constitutional review is concentrated and often abstract, ordinary judges have no special claims to authority. Moreover, because of their authoritarian past, judges in these systems were at least initially distrusted as arbiters of constitutional and democratic values. Thus, in all of the post-authoritarian systems, law professors tend to occupy many of the seats on the court together with some judges.
3.2. A Special Case: the French Conseil Constitutional
The institution of constitutional review in the French Fifth Republic is worth considering separately. The French republican tradition is solidly in the parliamentary sovereignty tradition and, it has been hostile to constitutional adjudication since the Revolution.17 But De Gaulle and his supporters insisted on placing institutional restraints on parliament and one of these was the Conseil Constitutions. Within the parliamentary sovereignty system, the only way that this could be done was to place the court effectively within the legislature. Thus, to a greater extent than Kelsen recommended, this placement emphasized the Conseil's legislative function. It is permitted to review statutes only prior to their promulgation, and then only on referral from the government or (since 1974) significant political minorities of deputies or senators. Thus, legislative proposals cannot become law if the Conseil strikes them down. But, the Conseil has no capacity to review a statute after it has become law, and regular courts are not empowered to undertake constitutional review in the course of ordinary litigation. In the French view, the statute law is sovereign and subject to no external checking, certainly not by judges, and neither by a constitutional court. Rather, the constitutional court, by becoming part of the legislature itself, plays an essential role in preserving the idea of sovereignty of the law. Legislative action cannot occur in the presence of an objection by the Conseil, but once it takes place, is unchecked by constitutional mechanisms.
Because French legislative proposals cannot become law in the presence of Conseil refusal, if it is requested (and this is virtually always at the request of a political minority appealing against government sponsored legislation) constitutional review must take place immediately after the legislative action and in the face of a sitting government whose proposed law has been challenged. This means such review has to happen quickly and, in view of the majoritarian nature of French political institutions and political culture, in a potentially politically charged situation. By contrast, constitutional challenges to U.S. statutes must await a case raising the issue in a genuine manner and this often occurs long
17   In 1795 the post-Jacobin Convention unanimously rejected a proposal presented by Sieyes: to introduce a jury constitutionnaire.
CONSTITUTIONAL COURTS AS DELIBERATIVE INSTITUTIONS
after the legislature that enacted the statute has disappeared. In post-authoritarian systems, such as Germany, Spain and Italy, constitutional courts cannot prevent a law from taking effect, even while review has an abstract character. Thus, unlike the French situation, since the government is enforcing the disputed law, there is not much political pressure to resolve constitutional issues quickly (indeed, political pressures might work in the opposite direction), and so abstract review may take place after the heat of political battle is somewhat dissipated.
Perhaps because of its placement within the legislature, the makeup of the Conseil is comparatively quite distinctive. Those on the court only occasionally have substantial legal or judicial backgrounds and there are also few law professors. Rather, the Conseil is dominated by political allies of the appointing authorities. This is not to say that the members are unable to understand or make legal arguments, but only that a high value is placed on their political judgement. This seems especially crucial in view of the urgency with which they must reach decisions.
4. DELIBERATION IN CONSTITUTIONAL COURTS
The argument in this section is that the three distinctive systems of constitutional review produce three separately identifiable deliberative patterns. In separation of powers systems, in which constitutional review takes place in ordinary courts, there is a great need for coordination and communication among judges. Such systems tend to develop open forms of deliberation with public hearings, published opinions and votes, and doctrines of precedent. In the Kelsenian model, with concentrated constitutional review powers in a specialized tribunal, clarity : of constitutional decisions is crucial if ordinary courts are to be able to apply them, so such courts tend to speak with a single voice and to articulate reasoned rationales. Such systems rarely have public hearings or published dissents. In parliamentary sovereignty regimes, constitutional review is a part of the legislative process aimed at ensuring that legislation conforms to constitutional requirements. How exactly this is done depends on the particular way in which constitutional review is integrated into the legislative process.
In France, the parliamentary sovereignty system with the most institutionalized system of constitutional review, the court deliberates in secret session, without ;public hearings or presentations, over the legislative proposals that are placed on its docket. The important feature of French constitutional review is that the statute voted by parliamentary majority cannot be promulgated until the Conseil assents to it. So there is an enormous time pressure on the review process (the delay is normally one month, but 'this period shall be reduced to eight days where the Government declares the matter urgent').18 The Conseil issues its
Art. 25 of I he Ordinance no. 58-1067 of November 1958.
CONSTITUTIONAL JUSTICE, EAST AND WEST
opinions in the name of the court, with no reported votes or published dissents. Its opinions are typically very brief and do not elaborate constitutional objections to legislation in any great detail. Instead, the Conseil may report that the proposed legislation is, in whole or part, constitutionally objectionable in view of several sources of constitutional law: the 1789 Declaration, the Preamble to the 1946 Constitution and to the 'Fundamental Principles underlying the laws of the [Third] Republic.'19 Not surprisingly, many of the most far-reaching decisions concern themselves with overly broad or unconstrained delegations of authority to administrative agencies.20 Typically, in the face of such concerns,
19 Representative in this perspective is the very important opinion of July 16, 1971 that rules on the right of association and modifies substantially the role of the Council. The decision is not only short, but it did not really spell out the reasons for including the bloc de constitutionnalite among the sources for constitutional adjudication:
In the light of the Constitution and notably of its Preamble;
Tn the Sight of the ordonnance of 7 November 1958 creating the organic law on the Conseil Constitutionnel, especially chapter 2 of title II of the said ordonnance; In the light of the loi of 1 July 1901 (as amended) relating to associations; In the light of the loi of 10 January 1936 relating to combat groups and private militias; Considering that the loi referred for scrutiny by the ConseiS constitutionnel was put to the vote in both chambers, following one of the procedures provided for in the Constitution, during the parliamentary session beginning on 2 April 1971; Considering that, among the fundamental principles recognized by the laws of the Republic and solemnly reaffirmed by the Constitution, is to be found the freedom of association; that this principle underlies the general provisions of the loi of 1 July 1901; that, by virtue of this principle, associations may be formed freely and can be registered simply on condition of the deposition of a prior declaration, that, thus, with the exception of measures that may be taken against certain types of associations, the validity of the creation of an association cannot be subordinated to the prior intervention of an administrative or judicial authority, even when the association appears to be invalid or to have an illegal purpose;
Considering that, even if they change nothing in respect of the creation of undeclared associations, the provisions of the art. 3 of the loi, the text of which is referred to the Conseil before its promulgation for scrutiny as to its compatibility with the Constitution, is intended to create a procedure whereby the acquisition of legal capacity by declared associations couSd be subordinated to a prior review by a court as to its compliance with the Saw;
Considering that, therefore, the provisions of the article 3 of the loi are declared not to be compatible with the Constitution ...
(English trans, by J. Bell, French Constitutional Law (Oxford: Clarendon Press, 1992), pp. 272-73).
20 See the vehicle search cases (J. Bell, op. cit., p. 141 and 308-09) and the Nationalization decisions. The 1971 decision that discovered that freedom of association was a constitutional; principle was also of this form. Implicitly, a group could be made illegal if there were principled reasons relating to public security that could be produced.
CONSTITUTIONAL COURTS AS DELIBERATIVE INSTITUTIONS
the government can re-draft the legislation by including some standards governing subsequent administrative actions.
While the French Constitutional Council has invalidated legislative proposals with some frequency, it has more often held parts of proposed laws offensive to constitutional principles,21 and has increasingly issued opinions suggesting that a law would be constitutionally valid only if it is interpreted in a particular way.22 This last strategy is particularly problematic in view of the separation of the Conseil Constitutionel from the judiciary. How can the Conseil offer any assurance that a promulgated law will be interpreted in a constitutional fashion? As there are no formal mechanisms or doctrines (such as stare decisis) to ensure this, the only guarantee is the persuasiveness of the reasons offered by the Conseil either to the ordinary courts, the Cour de Cassation or to administrative agencies. There is some evidence that agencies, perhaps seeking clear guidance from a tribunal that can be troublesome, have tended to use Conseil decisions in formulating internal guidance to their personnel. And, ordinary courts, seeking legal stability have frequently taken up Conseil interpretations as authoritative.
In the Kelsenian or post-authoritarian systems, the courts typically deliberate in secret, only rarely hold public hearings, and opinions are issued in the name of the whole court without recorded votes. Few of these systems permit the publication of dissenting opinions. Those courts that permit direct access to litigants (such as Spain and Germany) usually institute an internal division of labor (meeting in panels for considering such cases). Some of these courts permit or encourage subject matter specialization (commonly in Italy but opposed in Germany). All of these courts are internally deliberative in the sense that they do much of their work in a collegia! and face to face fashion where real attempts are made to convince others and produce a collectively reasoned decision. This is especially so in Italy where the court aims at, and usually achieves, consensual decisions (as far as can be told from discussions with justices and other observers).
The deliberative practices on the US Supreme Court are quite different and we may term them externally rather than internally deliberative. In part this is made necessary by the fact that the Court has no monopoly on constitutional interpretation and mostly acts to regulate the process by which the Constitution is applied by other courts. This coordinating or regulatory role forces the Court to do its work in a public and transparent manner that permits other agents -Judges, agencies and lawyers - to anticipate clearly and (mostly) successfully how the Court itself would rule on cases not yet before it. Moreover, the notion that .statutes may raise constitutional issues as they are applied rather than on
From 1981 to 1993, about 50% of the referred laws have been censured by the Council; see A. Stone, 'Constitutional Politics and Malaise in France', in J.T. Keeler and M.A. Sehain (eds.), Chirac's Challenge (New York: Saint Martin 1996), p. 65. Alexandre Viala, Les reserves d'interpretation dans la jurisprudence du Conseil constitutionnel (Paris: L.G.D.J. 1999).
CONSTITUTIONAL JUSTICE, EAST AND WEST
their face, sometimes leads the Court to permit or encourage experimentation in lower courts, rather than seeking an immediate resolution of the constitutionality of a statute restricted to the text itself. Obviously, this strategy is available only in systems permitting concrete review and can arise in Kelsenian systems only in considering constitutional complaints (amparo and Verfassungsbesch-werde).
The American Supreme Court acts publicly in several senses. First, access to its docket is open to litigants generally and its decisions on whether to take a case are public. On rare occasions members of the Court even announce their individual views on this issue. Secondly, public hearings are the norm, at least for the most important part of the Court's docket, votes are recorded and dissenting (and concurring) opinions published. The Court rarely tries to speak with one voice, apparently preferring to let conflict and disagreement ferment. On most accounts the actual level of internal deliberation on the court may be low, in the sense that the Justices do not spend very much time in conference deliberating specific issues in cases (observers usually report that justices interact mostly with their own clerks and less often with other justices and rarely with the aim of actually changing minds or votes). Internal deliberation seems mostly to take place after the initial voting has taken place in conference, when the justices decide whether they can agree in whole or part with the opinion written for the Court. Such deliberation takes place in writing for the most part and not through face to face interaction. But this very fact permits the deliberative ^process to be much more transparent to outsiders than face to face deliberation \ would be. The public manner of conducting its business puts the Court at the Icenter of wider deliberative processes within the judiciary, the legal, and political ^communities.
2.   Some Conditions for the Success of Constitutional Courts: Lessons from the U.S. Experience
Martin Shapiro
In recent years we have been experiencing a global flourishing of constitutional courts wielding the legal authority to declare legislative and executive acts unconstitutional Such judicial review has flourished even in nations whose legal culture was long thought to be antithetical to it such as France. It occurs even in a few non-Western nations such as Korea, India and Japan. It occurs in Israel which does not have a written constitution. It occurs in trans-national settings such as the European Union and the European Convention on Human Rights system. It has now appeared in states that have emerged from former Soviet domination.
Yet very clearly to encounter the legal forms of judicial review is not necessarily lo encounter successful judicial review, granting that in this instance success is difficult to define or measure. My definition of success is a purely institutional one involving whether a constitutional court has achieved acquiescence in its judgments by other public and private institutions, organizations and individuals. I do not concern myself at all with issues of the goodness or justice of the policies pursued by such courts. At a minimum successful judicial review would require that constitutional judgments are routinely, if not always, obeyed by bo tli governmental and private actors, and that relatively significant acts of government are judicially invalidated on constitutional grounds, at least occasionally. Measurement difficulties occur along a number of dimensions. Even :constitutional courts that are usually considered highly successful have experienced extended periods of massive disobedience to some of their decisions even ■ while being routinely obeyed as to others. The U.S. Supreme Court, for instance, encountered long resistance by autonomous local government authorities and < even state governments to its school prayer and desegregation decisions.1 Its judicially pronounced national code of police conduct is frequently evaded by police perjury and other misconduct.2 Yet that code generally has been effective in changing police conduct; school prayers are not said in many places where they: would be in the absence of Supreme Court decisions; legally sanctioned
1 Stephen Wasby, The Impact of the U.S. Supreme Court (Horaewood, 111. Dorsey Press, J970); Gerald Rosenberg, The Hollow Hope (Chicago: University of Chicago Press 1991).
2 Wayne La Fave, Search ami Seizure: A Treatise on the Fourth Amendment (1978). (St. Paul: West 1996).