A political economy perspective of judicial review in the European Union. Judicial appointments rule, accessibility and jurisdiction of the European Court of Justice. by George Tridimas Abstract The paper examines the benefits the sovereign member states of the EU expect to derive by granting the European Court of Justice the power to review the collective policy making decisions of the EU legislative bodies. Using the methodology of constitutional political economy it investigates the one-country one-judge rule of judicial appointments in the ECJ, the restrictions imposed on litigants to access the ECJ and the limits on the jurisdiction of the ECJ to review EU legislation. It also analyses how the presence of judicial review affects the size of the policy measures taken by the policy makers. JEL Classification: D72 – Economic Models of Political Processes D74 – Conflict; Conflict Resolution D78 – Positive Analysis of Policy–Making and Implementation K33 – International Law K41 – Litigation Process Key Words:
European Court of Justice, constitutional political economy, judicial review, judicial appointments, accessibility to court, jurisdiction of court, independent judiciary
George Tridimas, School of Economics and Politics, University of Ulster, Shore Road, Newtownabbey Co. Antrim, BT37 0QB e-mail:
[email protected] February 2004 FORTHCOMING IN THE EUROPEAN JOURNAL OF LAW AND ECONOMICS
A political economy perspective of judicial review in the European Union. Judicial appointments rule, accessibility and jurisdiction of the European Court of Justice. * Under the constitutional arrangements of the European Union, in order to adopt legislation, it is not sufficient to have the consent of the policy making institutions. The institution must also have the power to act in the respective field. Since the Community institutions may only adopt legislation in accordance with the provisions of the founding Treaty a measure must make clear the legal basis under which it was adopted. Failure to do so, may lead to its annulment by the European Court of Justice (ECJ). The European Union is governed by the so-called principle of enumerated competencies (as is the case with numerous national legislatures). The principle means that the presumption of competence lies with the Member States. The Union only has those powers which are expressly or by implication granted to it by the Treaty1. The validity of policy measures and more generally the limits of EU competence are determined by the European Court of Justice. The ECJ can annul a policy measure if it judges that the supranational organs have acted beyond or in excess of their jurisdiction. Why have the sovereign national political actors of the member states set up an independent judicial authority with the power to review and therefore enforce or annul the policy measures of the supra-national decision making bodies? After all, agreeing on a collective choice outcome in the EU is a complex and cumbersome process. Briefly and at the risk of oversimplification, the European Commission prepares legislation and has executive powers. Legislation has then to be approved by two bodies. First, the European Council, which represents the sovereign governments of the member states and votes by unanimity or qualified majority voting as the case may be. Second, the European Parliament, which in an embryonic form represents the peoples of Europe and whose powers have increased dramatically since its inception, and votes by simple majority2. The European Court of Justice is the judicial arm of the European Union and in many respects it resembles the supreme court of national polities. Its fundamental function is to ensure that in the application and interpretation of the Treaty the law is observed (Art. 220 EC). The ECJ has jurisdiction to interpret the treaties on which the European Communities and the European Union are founded and also to annul Community legislation if it conflicts with these treaties. Furthermore, in accordance with the principle of primacy and direct effect, the ECJ may rule on the compatibility of national legislation with EU law. Setting up an independent court with the power to review legislation has had profound implications for policy making in the EU. In many occasions the rulings of the ECJ clashed with long standing domestic policy measures of some governments of member states, while in others they reversed policy measures taken by the supranational legislative authorities of the EU. The ECJ has no equivalent counterpart in international relations3. However, there are fundamental differences Article 5 EC states that “The Community shall act within the limits of the powers conferred upon it by this Treaty and of the objectives assigned to it therein”. However, there is a potential conflict between this pronouncement and the broad formulation of Articles 2 and 3 of the Treaty, which lay down the aims and the activities of the Community. 2 For a recent critical account of the evolution of decision making rules and shifts in the balance of powers between those three sets of decision makers see Tsebelis and Garrett (2001) and Tsebelis and Yataganas (2002). 3 For example, NATO has not set up a judicial structure. Similarly, the International Court of Justice in The Hague has had a notably more limited role in shaping international relations. The European Court of Human Rights is another example of a court with significant influence in international relations. The latter does not have jurisdiction on the allocation of competence between the nation-state 1
1
between the ECJ and its national counterparts. Three of those institutional arrangements stand out. First, each country-member of the Union appoints one judge, so that over time the number of judges has not been fixed but rising with the number of member states. Second, accessibility to the Court is restricted. Only sovereign states – members of the EU, as well as its supra-national political organs have an automatic right to bring an action to the Court; other disputants first have to prove locus standi, that is, the policy measure they are challenging has a direct impact on their welfare. Third, the jurisdiction of the Court to hear cases is strictly prescribed by the founding treaties; only issues relating to the scope of the EU can be brought to the Court. What benefits do the sovereign states – members of the Union expect to gain by subjecting the outcome of the legislative process to the judiciary review by the ECJ? What reasoning justifies the one country – one judge rule of judicial appointments? Why the scope of judicial review is constrained in the way it was just described? Given the far reaching implications of judicial review for policy making in the EU, it is rather surprising that these questions have not received systematic attention in the literature. The present paper aims to answer those questions using the methodology of constitutional political economy. Specifically, it analyses the choices that the framers of the founding treaties make in setting up a court and granting review powers. Although the paper focuses on the EU, it is hoped that the analysis will be of more general interest. The structure of the paper is as follows. Section I provides a brief account of the reasons of the existence of the Court encountered in the interdisciplinary literature; such reasoning is often implicit rather than explicit. Section II analyses formally the question of why the sovereign actors - members of the union agree to set up a court with the power to review and annul legislation in a setting of uncertain majority voting outcomes. Section III explores the components of judicial review focusing on the onecountry one judge rule of judicial appointment, the accessibility to the judicial process and the jurisdiction of the court. Section IV explores how the presence of judicial review affects the size of the policy measures taken by the policy makers. A short conclusion follows. I.
A selective review of political economy accounts of the ECJ
As the effective constitutional court of the “quasi-federal polity” of the EU, the ECJ performs several functions reserved for the supreme courts of national polities. It interprets the founding treaties of the Union, which define the set of rules that determine the union and its policy making institutions; it checks the compatibility of legislative acts with the founding treaties; it adjudicates in disputes between the state and the citizen and in disputes between different branches of government; it rules on the division of powers between the supranational (central) authorities and the national governments and in disputes between different states4. A large and still growing interdisciplinary literature, combining law scholarship, political science and economics has developed to explore the role of the ECJ as a constitutional player in the collective decision making process of the EU5. Several complementary strands can be distinguished. and the supranational organs. However, it can hear cases brought by private litigants regarding the violations of their human rights (provided that their countries have adopted the European convention on human rights), see Helfer and Slaughter (1997) for a comparison between the two and other international courts. I owe this point to an anonymous referee. 4 See T. Tridimas (1999) for a detailed analysis of the role and functions of the ECJ. 5 For a detailed and systematic survey of the benefits and costs that office motivated politicians may expect when they delegate decision making power to appointed bodies such as a national constitutional court, see Voigt and Salzberger (2002).
2
The first strand emphasises the savings on transaction costs and other efficiency benefits made from delegating decision making power to the ECJ. Since the international treaties founding the union are incomplete contracts beset by problems of asymmetric information and high transaction costs, appointing an independent judiciary with powers of interpretation and enforcement resolves the problem of credibility of commitment of the sovereign member-states to the supranational polity, securing thereby the benefits from international co-operation. Further efficiency gains can be had as the ECJ develops specialised expertise in collecting information, interpreting vague aspects of EC law and resolving disputes regarding violations of the treaties by member states (for details see Pollack, 1997; Stone Sweet, 2000; Majone, 2001; and Tallberg, 2002). A second strand of the literature considers the ECJ as the institution which preserves the economic and political gains of the winning coalition at the time of establishing the union. This is the so-called inter-governmentalist school of thought, which emphasises the pre-eminent role of member states in the integration process and the political subordination of the judiciary. It views the ECJ as lacking the autonomy to act against the interests of its creators, and more specifically those of the most powerful members, typically, France and Germany (see Garrett, 1992 and 1995, Garrett and Weingast, 1993, Moravcsik, 1991 and 1993)6. At the opposite extreme of the spectrum, the neofunctionalist view of European integration argues that the Court possesses significant autonomy which can be used to rule against powerful member states and pursue a pro-integration agenda serving the interests of a developing transnational polity (see Burley and Mattli, 1993 and 1995; Stone Sweet and Sandholtz, 1997 and Stone Sweet and Brunell, 1998). Recently, more elaborate views have gained ground, which model the ECJ as an additional strategic player in the policy game of the EU; see amongst others, Mattli and Slaughter (1998), Alter (1996), (1998) and (2000), Garrett et al. (1998) and Voigt(2003). In the same vein, a number of studies have used the spatial model of rational choice to analyse the interaction between the Court as maximising an objective function subject to the relevant institutional constraints and the other arms of governance; see Cooter and Drexl (1994), Bednar et al. (1996), Tsebelis and Garrett (2001) and Tridimas and Tridimas (2004a). A third thread in the literature points to benefits that politicians of the member states may derive when they shift to the Court the responsibility for implementing policy measures unpopular with domestic voters (see Fiorina, 1977, and Vaubel, 1994). Criticism against ECJ rulings by some national governments may lend credence to this view. Yet, drawing on constitutional political economy one adds a further most important benefit that the sovereign actors expect to obtain from setting the Court, notably, the protection of the rights of each sovereign actor against abuses by the supranational bodies of governance7. In order to pursue the efficiency objectives of international coThere are obvious similarities between this view and the view of Landes and Posner (1975) of an “independent” judiciary as the institution, which secures the durability of the distributional gains achieved by the winning coalition of interest groups. Interest groups will support legislators who can safeguard the permanence of their gains. The latter can be accomplished when the enacting legislature delegates review powers to the constitutional court, which then applies the “original intent” of the legislator and ratifies the privileges awarded to interest groups independently of democratic controls. Mueller (1996) points out that such an explanation of judicial independence is, however, invalid when the framers of the constitution are neither members of a sitting legislature nor are seeking political office in the future.
6
7
For clarity we confine our analysis to disputes about the rights of the nationstate as a member of the Union vis-à-vis the supra-national authorities. In reality,
3
operation the member states of the EU have established supranational legislative and executive institutions with significant regulatory power (their tax and spending powers although not insignificant are less pronounced). To protect from the risks that the supranational agencies will abuse their authority and violate the political and economic rights of the members of the polity, in practice, their powers are not unlimited but subject to constraints as specified in the founding treaties8. The ECJ then emerges as an institution to keep the Commission from exceeding its authority and decide on charges of non compliance raised by the Commission or member states (see for example Alter 1998). Adopting majority rule (simple or qualified) for collective decision making by the member-states of the union implies that some of them will be on the losing side of the collective choice outcome suffering financial and / or other utility losses. Application of the unanimity rule, which requires the consent of all members of the polity, would have prevented the passing of the measures which affect negatively some members of the polity. However, the unanimity rule may be impractical, because it imposes large decision-making costs. Mueller (1991) demonstrates that granting the members of the polity inviolable constitutional rights to undertake (or abstain from) certain actions, “provides the same protection, with lower decision-making costs, as does the implicit veto each [member] possesses under the unanimity rule” (p.321). He shows that citizens wish to protect certain individual rights, that is, freedoms of action, under two conditions. (a) The cost to someone denied that right is very large relative to the gain secured by others when the right is denied, and (b) the framers of the founding charter, who grant the right, are uncertain whether they will be protected or harmed by the right. Lacking the ability to foresee changes in individual preferences and expectations, technology and possible future conflicts of rights, the founders of the constitution establish various procedures for resolving disputes about rights. The judiciary is such an institution, which may protect those on the losing side of the collective choice outcome from the “tyranny of the majority”. Under these circumstances, conflicts about losses and gains may not be resolved by majority voting. On the above reasoning, an independent judiciary with the power to review and annul legislation emerges as an essential element in interpreting the nature of rights and the effective exercise of checks and balances on the executive and legislative branches9. An independent judiciary carries out such responsibilities without regard to the power and political preferences of the parties appearing before it, or the power and preferences of since the formation of the EU has granted rights to the citizens of the member states, additional disputes arise regarding the rights of the individual voter against its own national government, which are also dealt by the ECJ. We intend to address this issue in future research. 8 This problem is well known in the political economy literature since Hobbes. Mueller refers to it as the Sorcerer’s Apprentice Problem, “the spectre that government once created turns upon its creators” (1996 p.3). Weingast calls it “the fundamental political dilemma of an economic system: a government strong enough to protect property rights is also strong enough to confiscate the wealth of its citizens” (1993, p.287). 9 The importance of the independence of the judiciary has been recognised since the time of Plato and Aristotle. The contemporary notion of the judiciary as a separate branch of power is traced to Montesquieu. Rather than mere functional separation, checks and balances require that each branch of government jointly exercises the powers, which, inevitably, leads to some confusion of powers. See Stone Sweet (2000) for a critical discussion of the usefulness of the traditional notion of the separation of powers and an analysis of the role of the constitutional judge with special attention to European countries. For a recent discussion of various arrangements to safeguard judicial independence, see Mueller (1996) and Salzberger and Voigt (2002). Prominent amongst the factors discussed are the method of appointment of judges, the length of their term of service and their financial security.
4
any other branch of government with an interest in the case10. How does the addition of judicial review change the calculus of expected benefits from participating into the collective choice process11? II.
A formal model of collective decision making under judicial review
This section addresses the question of why at the constitutional stage sovereign states – decision makers, who join a supra-national union, establish an independent court with the power to review the policies collectively decided. The formal model builds on the constitutional political economy framework used by Mueller (1991), (1996) and (1997), who in turn builds on the original study of efficiency and decision costs in a democracy by Buchanan and Tullock (1962). The model explores the decision of the sovereign actors to subject the outcome of the costly legislative process to the risk that it may be overturned by the judicial process. For any collective choice, the larger the majority required for passing legislation, the larger the costs of reaching a decision, in terms of time and other resources required to reach and implement agreement by the members of the polity. In practice, a majority below unanimity may be optimal. However, using a voting rule less than unanimity implies that there are some losers from the collective action. In the absence of unanimity rule, the losers may wish to have recourse to another means of preventing these losses. An independent judiciary, whose function is to rule on the legality of the policy and either ratify it or annul it, provides a most effective means for this purpose. Judicial review, however, gives rise to transaction costs. These relate to the resource of costs of setting up a court, selecting judges and ensuring their financial independence and the costs of the deliberations of the court. For each sovereign actor let P be the probability that he benefits from the collectively decided policy at the legislative stage, with 0≤P≤1, and let G>0 be the corresponding benefit. Conversely, the probability that he loses from the collectively decided policy is 1-P and let X>0 be the loss inflicted. Following Mueller (op.cit.), the probability of passing favourable legislation depends positively on the majority M required to reach collective action, where 00 and PMM≡∂2P/∂M2 0 and VMM≡∂2V/∂M2>0. After the legislature passes a policy measure, its compatibility with the founding treaties may be challenged at the ECJ. Let N be the probability that it will be challenged, where 0≤N≤1. The probability that a policy measure will be enacted without being subject to court scrutiny is then 1-N. Not all legislative acts are reviewed by the judiciary. Specifically, two kinds of restrictions may affect whether or not it is subject to judicial review. (a) Accessibility to the constitutional court, also known as the locus standi, that is, the conditions under which disputants can bring an action for annulment of legislation. (b) Jurisdiction of the constitutional court, that is, which legislative measures is the judiciary allowed to review. Denoting accessibility by A and See Pollack (1997) for an account of the autonomy of ECJ decision making using the economic theory of the principal-agent relationship. For a detailed analysis of the arrangements which in practice aim to secure the political independence of the ECJ from the national governments of the member states and the supra-national authorities of the EU see Tridimas and Tridimas (2004b). 11 Empirically, until the completion of the Single Market the ECJ has promoted integration, but after that it has taken a more critical stance with greater deference to national sovereignty, see amongst others Alter (1998) and (2000). 10
5
jurisdiction by J, we may then write N=N(A,J). N is assumed to be increasing in its attributes, but at a decreasing rate; formally, NA≡∂N/∂A>0, NAA≡∂2N/∂A20, NJJ≡∂2N/∂J20, CAA≡∂2C/∂A2>0 and CJ≡∂C/∂J>0, CJJ≡∂2C/∂J2>0, and is again assumed that CAJ≡∂C/∂A∂J=0. In addition, it is reasonable to assume that the cots of the judiciary system rises, as the number of serving judges increases. Let Y be the number of judges appointed to the Court; thence, we have CY≡∂C/∂Y>0, CAA≡∂2C/∂Y2>0 and assume again that CAY(≡∂C/∂A∂Y)=CJY(≡∂C/∂J∂Y)=0. The expected net benefit for the individual sovereign actor is E(B) = P×N×Q×G + P×N×(1-Q)×0 + P×(1-N)×G + (1-P)×N×(1-Q)×(-X) + (1-P)×N×Q×0+(1-P)×(1-N)×(-X)–V–C, or more compactly E(B) = P(1-N+QN)G – (1-P)(1-QN)X–V–C (1) Obviously, the sovereign actor will agree to establish an independent judiciary with policy review powers when the expected benefit in equation E(B) is (a) positive and (b) greater than the expected benefit in the absence of judicial review. Note that, unlike for example the choice of the majority rule M, setting an independent court is not a continuous variable; rather it is a binary yes-or-no variable. Either a court is set up with the power to review, in which case there is a Q probability of a favourable ruling, or no court is established. The probability Q may of course vary from policy measure to policy measure reviewed by the court. For the sovereign actor the expected benefit without judicial review is W(B) = PG–(1-P)X–V (2) The additional benefit from judicial review is then ∆B≡E(B)-W(B)=-PN(1-Q)G + (1-P)NQX-C Thus, an independent judiciary with the power to review legislation will be established for parameter values such that the following two inequalities are satisfied simultaneously 12
In reality, the judiciary may be more influential than this specification allows. Specifically, court rulings may offer guidelines as to what constitutes the correct interpretation of the constitution and therefore indirectly shape new policy measures, if the government decides to re-introduce legislation. On this, see Tridimas and Tridimas (2002) for an analysis of an ECJ ruling which annulled a Directive banning the advertisement and sponsorship of tobacco at the Community level.
6
E(B) = P(1-N+QN)G – (1-P)(1-QN)X–V–C > 0 ∆B = -PN(1-Q)G + (1-P)NQX-C > 0 Solving the two inequalities we obtain that the court will be established for values of G such that G0 < G < G* where G0≡[(1-P)(1-NQ)X+(V+C)]/P(1-N+NQ), from the E(B)>0 inequality and G*≡[(1P)NQX-C]/PN(1-Q) from the ∆B>0 inequality. For the (G*,G0) interval to be non-empty it must be G*>G0, which upon manipulating yields (1-P)N(2Q-1)X > C+N(1-Q)V (3) A necessary condition for the inequality in (3) is that Q > ½. That is, a necessary condition for establishing an independent court with the power to review legislation passed by the collective decision making body of the sovereign players is that the probability of a judicial ruling that favours the sovereign actor under consideration is higher than 50%. In sum, the analysis points to the following three components of judicial review of the policies decided by the political organs of the union: (a) the probability of a favourable court ruling (where the court either ratifies a policy measure resulting in a gain or annuls one resulting in a loss), Q; (b) the accessibility to the court, A, by potential litigants; and (c) the jurisdiction of the court, J, to rule on disputes. III.
How much judicial review?
Having explained the conditions under which the member-states of the Union agree to subject their collective decision making to an independent judiciary, the present section explores the constituents of judicial review, Q, A, and J, which maximise the expected benefit from judicial review. The one-country one-judge appointment rule Strictly speaking, judicial review by an independent judiciary implies that the ruling of the Court does not depend on who the disputants are, but only on whether or not the policy measure under consideration is consistent with the constitutional arrangements which authorise the adoption of the measure. Hence, at the constitutional stage the interests of the members of the polity are best served by setting up a Court, which will not discriminate against any of the members of the polity. Given that the EU is a polity of sovereign states, a possible way to achieve this is to allow each member-state to appoints one (or more) of its own nationals as a judge. Specifically, it is presumed that the appointing national government expects that a judge form the same country will have preferences close to the preferences of the appointing government, and therefore such a judge will protect the rights of the country. This is not to say that each judge sides with his/her own country. But it is to acknowledge that the sovereign governments are more likely to have confidence in the system of judicial review, and be prepared to be bound in it, if they can appoint one of their nationals to serve at the court, rather than if they cannot. If a country does not appoint one of its nationals to the bench, there may be occasions when some of its interests, views and legal traditions may be overlooked. Moreover, it may be very difficult for the citizens of the country which is not represented at the bench to accept that they are not discriminated against the citizens of other member-states or that they are ‘lower class citizens’ in the union13. Let Yi be the number of judges that country member i appoints to the Court. The non discrimination argument implies that the probability that the Court will reach a decision favouring the sovereign actor i is14 The failed French proposal that each one of the “big four” countries will appoint two judges (see below footnote 15 for details) lends support to this supposition. 14 For a recent use of the non-discriminating probabilities in judicial rulings see 13
7
(4) Qi = Yi/∑iYi Maximising (1) with respect to Yi yields ∑j≠iYj/(∑iYi)2 = Ci/Φ I equations (5) where Ci≡∂C/∂Yi>0 denotes the marginal resource cost of an extra judge and Φ≡N[PG+(1-P)X]. Assuming that Ci is constant and the same across all i, so that Ci=θ for all i, the set of equations (5) yields that (6) Y1 = Y2 = … = Yi That is, for political optimality each country member should appoint the same number of judges. Further, adding up the I equations of (5) and rearranging we obtain = [(I-1)/I][Φ/θ] ∑iYi which upon using Yi=Y and normalising so that [Φ/θ]=1 for all i yields [7] Y = [(I-1)/I2] Rounding the latter to the nearest non-zero integer yields Yi = 1 for each i. That is, each country member should appoint one judge. Thus, the total number of judges is equal to the number of member states (currently standing at fifteen). This political optimisation argument may well explain why, unlike national polities, the total number of ECJ judges has not been constant, but it has increased with every enlargement of the EU. The rule of one member-state – one-judge implies that, at the judicial level, small and big countries carry equal weights15. Interestingly, although virtually ignored by commentators, the rule of one judge per member-state has been retained in the Nice Treaty of 2001. In anticipation of the entry of ten new countries the rules for qualified majority voting in the Council and the composition of the Commission and the European Parliament have been revised, while the total numbers of the latter two have been capped. No such revisions will affect the Court giving a total of twenty-five judges. Restrictions on the accessibility to the Court and the jurisdiction of the Court At the constitutional stage the sovereign actors are assumed to choose the accessibility to and jurisdiction of the court in order to maximise the expected benefit E(B). To appreciate how the sizes of expected gains and losses from collective choice vary with the process of judicial review we again borrow a leaf from Mueller (op.cit.) and express losses as a proportion of gains, X = k×G with k>0. (In the special case where collective action decides pure redistribution issues the policy gains exactly match the policy losses, so that G = X and k=1). The expected benefit from collective decision making under majority voting and judicial review then is written as E(B) = G[P(1-N+QN) – k(1-P)(1-QN)]–V–C (1′) Maximising E(B) with respect to A and J16 we derive the first order conditions: EA ≡ GNA[k(1-P)Q-P(1-Q)]-CA = 0 (8.1) EJ ≡ GNJ[k(1-P)Q-P(1-Q)]-CJ = 0 (8.2) Notice that since both NA and CA are positive (or equivalently, NJ and CJ) the expression F ≡ k(1-P)Q-P(1-Q) is also positive. Recalling the assumptions that EAC = JAC = 0, the second order conditions for maximisation are (9.1) EAA ≡ GNAA[k(1-P)Q-P(1-Q)]-CAA < 0 EJJ ≡ GNCC[k(1-P)Q-P(1-Q)]-CJJ < 0 (9.2) The term GNA[k(1-P)Q-P(1-Q)] in equation (8.1) shows the expected marginal benefit Rubin, Curran and Curran (2001). 15 In 1980 responding to a judgment that went against French restrictions on sheep meat imports the then French president Valery Giscard d’Estaign suggested that each one of the then big four countries would appoint an additional judge on the Court. However, in the end no such institutional change took place, for details see Garrett et al (1998). 16 Obviously, maximisation of the benefit, say, with respect to jurisdiction is not the same as maximisation of jurisdiction. Salzberger and Voigt (2002) argue that unrestricted jurisdiction of the judiciary to hear cases may embroil it into too many political disputes and undermine its perceived non-partisan character.
8
from increasing accessibility to the court (in terms of judicial rulings that will benefit the sovereign actor), while the term CA denotes the marginal cost of using resources to operate the judiciary system. An equivalent interpretation applies to equation (8.2) too. In equilibrium, the founders of the union choose accessibility and jurisdiction so that the marginal benefits exactly match the respective marginal costs. Equations (8.1) and (8.2) reveal that the determinants of the optimal degree of accessibility to the judiciary and its jurisdiction to review legislation are (a) the probability of passing a policy measure, and therefore, the majority required to pass it, P(M). (b) The probability that the judiciary will enforce a policy measure which benefits the sovereign actor, or annuls a policy measure which inflicts losses on the actor, Q. (c) The size of the gain from the policy measure, G. (d) The size of the loss from the policy measure in relation to the potential gain, k. However, no explicit algebraic solution can be obtained at this level of generality. We are nevertheless able to explore the comparative statics properties of the politically optimal solution. Differentiating totally the system of equations (8.1) and (8.2) with respect to its determinants and rearranging we have (10.1) dA/dM = NAGPM(kQ+1-Q)/EAA < 0 (10.2) dJ/dM = NJGPM(kQ+1-Q)/EJJ < 0 The lower the majority required to enact a collective decision, like passing legislation, the higher the optimal degree of accessibility to and jurisdiction of the court. Consequently, the fewer restrictions will be imposed on the process of judicial review, and the more cases will be reviewed, and vice versa. This implies that ceteris paribus, legislation passed with weak majorities will be challenged at the court, while at the limit, legislation passed with unanimity will not be challenged at the court. This result formalises the intuition of Mueller’s conclusion on the relationship between majority voting and granting of rights. (11.1) dA/dQ = - NAG[k(1-P)+P]/EAA > 0 (11.2) dJ/dQ = - NJG[k(1-P)+P]/EJJ > 0 The higher the probability that the judiciary will make a favourable ruling (one which either ratifies a policy that benefits the individual actor, or rejects a policy that inflicts losses), the larger the optimal size of court accessibility and jurisdiction. (12.1) dA/dG = - NAF/EAA > 0 dJ/dG = - NJF/EJJ > 0 (12.2) That is, the higher the expected gain from collective decision, the higher the optimal size of judicial review, in the sense of a higher optimal size of accessibility to and jurisdiction of the court. (13.1) dA/dk = - NAG(1-P)Q/EAA > 0 (13.2) dJ/dk = - NJG(1-P)Q/EJJ > 0 That is, the higher the loss from collective decision making as a proportion of the benefit, the higher the optimal level of accessibility to and jurisdiction of the court. Accessibility and jurisdiction are laid down in the founding treaties. Briefly, privileged applicants, like the Commission, the Council of the EU and the (governments of the) member-states have an automatic right to seek the annulment of an EU policy measure by the ECJ17. By contrast, non-privileged applicants, i.e., physical persons and companies have only limited access based on whether they can show that their interests are affected by the EU policy measure. As far as jurisdiction is concerned, in accordance with the economic rationale of the EU, and unlike the supreme courts of national polities, the jurisdiction of the ECJ covers primarily economic issues and encompasses all matters relating to the single market. Following the 1997 Amsterdam Treaty, a limited expansion of its jurisdiction to home affairs and security issues has Note that the Treaty of Nice added the European Parliament to the group of privileged applicants.
17
9
taken place, but the national governments retain substantial control over those areas18. Mueller’s analysis of majority voting and rights implies that legislation passed by unanimity in the Council will not be overturned by the ECJ. At first sight, this prediction is not always borne in practice. There have been occasions where, at the instigation of the European Parliament or the Commission, the ECJ has annulled measures voted unanimously by the Council on the ground that they were adopted under the wrong legal basis19. The ECJ held that the Council should have adopted them under a different procedure, which provided for greater participation by the European Parliament. However, rather than undermining the validity of the hypothesis, this experience only serves to highlight the insight of the analysis. As already explained, a policy measure by the supranational authorities can be adopted only after the correct legal basis is ascertained. Otherwise, the road is open for abuse of power. By annulling such measures the Court confirmed the rule of law, acknowledging in the process the bi-cameral nature of the legislative process of the EU. IV.
The effect of judicial review on the size of policy
The previous section studied the conditions for maximising the expected net political benefits from subjecting the legislative process to judicial review. This section explores how the presence of judicial review may affect the size of the policy measure decided by the political process. Quantitatively, the size of the policy measure may be best understood as the level of a standard public policy instrument, like the level of a tax rate, or of a public expenditure programme or of a regulatory control. Specifically, we are interested in comparing the level of G decided under judicial review, GR, with the level of G without judicial review, GW. For this reason we adopt the standard assumption that the utility function of the typical sovereign actor is an increasing, continuous twice differentiable function of the size of policy, and write it as U(G) with UG>0 and UGG0 and E(B)>W(B) yield rather complex formulas which do not allow easy generalisations. In what follows, we focus on the range of parameter values for which the two inequalities hold. Maximising equation (14) with respect to GR and equation (15) with respect to GW, and assuming that ∂V/∂G = ∂C/∂G = 0, we obtain ∂U(-kGR)/∂GR = (1/k)[(P/(1-P)] [(1-N+NQ)/(1-NQ)] ∂U(GR)/∂GR (16) W W ∂U(-kG )/ ∂G = (1/k)[(P/(1-P)] ∂U(GW)/∂GW (17) The second order conditions for maximisation are respectively ERGG ≡ P(1-N+NQ]U(GR)GG + k(1-P)(1-NQ)U(-kGR)GG < 0 EWGG ≡ PU(GW)GG + k(1-P)U(-kGW)GG < 0 Comparing equations (16) and (17) we have that whether or not ∂U(-kGR)/∂GR ≥ (
