A Comment on Tsebelis and Garrett

Alec Stone Sweet and Wayne Sandholtz

In their recent article, “The Institutional Foundations of Intergovernmentalism and Supranationalism in the European Union,” George Tsebelis and Geoffrey Garrett profess to offer a “unified model of the politics of the European Union.”(1) Unfortunately, the article provides something much more limited. Along the way, it misrepresents the current state of theoretical and empirical work on the EU. As a consequence, the article fails to recognize the areas of potential agreement between their research and that of others, as well as genuine differences.

Our point of reference is the European Integration and Supranational Governance project, collaborative research published in 1998.(2) It is this work that Tsebelis and Garrett crticize in their survey of the literature. Tsebelis and Garrett chose not to reference the full project, but rather a condensed overview published in article form in 1997.(3) In the following sections we (I) correct specific misrepresentations of our positions; (II) note potential areas of agreement; and (III) delineate the real points of disagreement, tying these to current theoretical debates.

I

Tsebelis and Garrett grossly misrepresent our arguments. Had they consulted European Integration and Surpanational Governance (or any of the many papers published pursuant to this project since 1996) instead of the abbreviated overview article they exclusively cite, they would have been compelled to develop different lines of criticism, and to scale back some of their own claims. Even the arguments in our 1997 article suffer distortion in their piece.

A. Tsebelis and Garrett write that we “explicitly seek to distance ourselves from classical neofunctionalism”(4) and, later, that “unlike Stone Sweet and Sandholtz, Burley and Mattli are happy to acknowledge their neofunctional heritage.”(5) What did we, in fact, say?

In the 1997 article cited by Tsebelis and Garrett (and again in the book), we state that “our theory has important affinities with neofunctionalism. . . . On crucial questions, we believe, they got it right.”(6) Further, we note that “the three constituent components of our theory are prefigured in neofunctionalism,” and we attribute to Haas the basis of an underlying “logic of institutionalization,” which is central to our project.(7)

We also discussed two main theoretical differences between classical neofunctionalism and our theory, which we were happy to call a “modified neofunctionalism.” First, we specified the dependent variable differently than did the original neofunctionalists: our project sought to explain variation (across time and policy domain) of modes of governance, which we define as the capacity of the EU polity to make authoritative rules. This variation was depicted in a continuum, stretching from intergovernmental to supranational, a heuristic device constituted by dimensions reflecting our concern for three variables.(8) Second, we reconceptualized, and gave explicit microfoundations, to the concept of spillover. We then discussed the causal mechanisms through which spillovers were likely to occur, and summarized the findings of subsequent chapters supporting these claims.

B. Tsebelis and Garrett write that we are “quite vague” about the nature of the self-reinforcing dynamic of European integration.(9) They allege that we mention only simple “functional demands” and propose a “loop of institutionalization,” but that “unfortunately [we] do not probe more deeply into this ‘loop of institutionalization’.” What did we write?

Like Tsebelis and Garrett, we emphasized how EU rules and procedures structure the integration process, and we focused on distinctions between intergovernmental and supranational politics, but we did so much more broadly than they do. We sought to explain why the authority to establish rules and policies has migrated from the national to the EU level, and why that authority has shifted more rapidly and extensively in some domains than in others; and we focused on the questions of whether, how, and the extent to which the extent to which the nature of EU politics had changed, over time, and across policy domains - again - in terms of the relationship between modes of governance (intergovernmental to supranational). Institutions are at the heart of that dynamic. We began, naturally, with a quite general statement of what typically happens when actors are successful in expanding the EU’s jurisdiction and capacities to govern:

As European rules emerge and are clarified, and as European organizations become arenas for politics, what is specifically supranational shapes the context for subsequent interactions: how actors define their interests, what avenues are available to pursue them, how disputes are to be resolved. This creates the ‘loop’ of institutionalization. Developments in EC rules delineate the contours of future policy debates as well as the normative and organizational terms in which they will be decided.(10)

Thus, we argued that institutional developments (changes in rules and organizations) at one time would structure politics and choices in later periods, and subsquent choices and activities could have an impact on institutional development.

Our notion of institutionalization amounts to, as we noted, a reconceptualization of Haas’ concept of spillover. Spillover was always fundamentally about feedback, which we see as one mode of institutionalization. Neofunctionalism, of course, focused on functional spillovers (integration in one policy area requires integration in substantively linked policy areas in order to retain the gains of the original integration). While we view the notion of functional spillovers as plausible, we do not see it as the primary mechanism driving integration.

Our notion is more general, and focuses on institutions. Positive feedback in European integration occurs to the extent that the activities of social actors (e.g., firms, interest groups, litigants, and so on) and the capacities of EU organizations to determine policy evolve symbiotically, reinforcing one another over time. How they have done so has varied across time and issue area, and this variance has been powerfully conditioned by the development of EU rules and procedures (institutions). Simplifying a great deal, we assume that actors will focus their political activity on sites where they see some chance of getting what they want. As the Commission and the ECJ showed that they could create rules that had real effects, actors invested more time and resources in Brussels and Luxembourg, and they lobbied national officials in new ways. They pressed claims that were not always functionally related to areas already covered by existing EC competences, and they pushed to deepen supranational governance in areas that were in fact covered. As more and more actors oriented their activities to the EC, more rules (legislative and judicial) were in fact produced, and the cycle was renewed, attracting more activity. Our reconceptualization preserved the key neofunctionalist insight that there was an expansive, endogenous logic to integration processes. Given certain assumptions (including self-seeking strategic actors), the scope of supranational governance is likely to widen and deepen, and is not likely to be rolled back substantially (barring exogenous shocks).

Several of the chapters of European Integration and Supranational Governance further refined the theoretical basis of feedback mechanisms and institutionalization in the development of supranational governance. In addition, they operationalized the concepts and tested them empirically. Sandholtz’s chapter shows how self-interested private actors, the Commission, and the Court were able to alter the institutional framework governing telecommuncations in ways that created a favorable context for subsequent efforts to liberalize at the EC level. He argued that specific Court rulings placed national governments in the shadow of litigation, making it difficult for them to resist later initiatives by the Commission. O’Reilly and Stone Sweet carried out a similar analysis for air transport.(11) With respect to the Court, Stone Sweet and Caporaso (1) specified the underlying interests of the main actors involved; (2) developed a basic principal-agent framework for analyzing the Court’s activities from the various institutional contexts in which it operates; and (3) derived a series of hypotheses from our theory of integration, including those on feedback. These hypotheses were then tested across the whole life of the EU, using econometric techniques and comprehensive data collected specifically for testing purposes. They further analyzed every ECJ ruling in two different policy domains, and assess the extent to which the Court's rulings preempted, required, or otherwise structured treaty revision, the production of secondary legislation, and changes in national law. They argued, notably, that the “institutional bases” for these dynamics flow from the fact that the ECJ operates in a relatively permissive strategic environment when it interprets the Treaty and secondary legislation adopted under unanimity rules, a point that Tsebelis and Garret pretend is their own.(12)

We find it difficult, then, to understand in what sense our project was “vague” regarding the dynamic of institutionalization. Furthermore, Tsebelis and Garrett go so far as to tell readers that their work can bring “precision” to our “plausible but vague intuitions.” This claim is particularly difficult to credit, since Tsebelis and Garrett have nothing to say about the causes of institutional change over time. In fact, theirs is not a dynamic theory in any sense. They, first, take as given certain changes in the rules governing EU legislation and, then spell out the consequences of these changes for relations among EU organizations (Court, Commission, Council, and Parliament) within those rules. How, exactly, does that bring precision to any dynamic theory of European institutional development?

C. Tsebelis and Garrett argue that “there is a clear analytic difference between Burley and Mattli's position [the second of the three ‘influential studies’ surveyed] and [ours]. Stone Sweet and Sandholtz's argument has no microfoundations.”(13) Again, Tsebelis and Garrett have simply overlooked what we actually said.

We explicitly declared that transantional actors (again, firms, interest groups, etc.) were one of the three primary components of our theory of institutionalization, along with EU rules and EU organizations.(14) We argued that two kinds of people would find it in their interest to push supranational organizations, like the Court and the Commission, to expand the scope of supranational governance: (1) transnational actors, like firms who exchange across borders and interest groups seeking EC regulation; and (2) those persons and groups who feel disadvantaged by national rules and legal regimes and strive to replace these with EU rules and procedures. Indeed, we state outright: “Our starting point is society, in particular, non-state actors who engage in transactions and communications across national borders, within Europe.”(15) In describing just one – and not the only - logic underlying the expansion of supranational governance, we established some premises regarding (general) preferences: “The absence of European-level rules will come to be seen as an obstacle to the generation of wealth and the achievement of other collective gains. Separate national regimes constitute the crucial source of transaction costs for those who wish to engage in exchanges across borders . . .”(16) In addition, we identified the broad preferences of EC organs like the Commission and the European Court of Justice: “Generally, EC organizations, such as the Commission and the Court, respond to this pressure [from transnational actors, and national actors seeking to destabilize national regimes through litigation] by working to extend the domain of supranational rules, in order to achieve collective gains and to accomplish the purposes of the Treaties, broadly interpreted.”(17)

In these and many other passages, we identified other factors (besides actors and their motives) that we believe must go into any analysis of institutional change. We argued that to explain any specific episode or series of episodes, in addition to actors and their interests, the analyst had to attend to (1) the relevant decisionmaking rules governing that policy domain, (2) the powers of the relevant EU organs (jurisdiction), and (3) the opportunities available to the various actors who would seek either to expand, or oppose expanding, supranational governance. That is exactly what Sandholtz, Pierson, Sbragia, Stone Sweet and Caporaso, and O'Reilly and Stone Sweet all did, to varying degrees, for varied purposes.(18) Given these points, how can Tsebelis and Garret assert that we “eschew the analysis of the strategies available to different actors and the constraints under which they operate”?(19)

In light of the foregoing discussion, what can Tsebelis and Garrett possibly mean when they declare that “supranationalists … view the EU’s institutions as actors, not dependent variables”?(20) We did not use and would in fact reject the label, “supranationalists,” to describe us, or “supranationalism” to describe the theoretical materials or ideas we elaborate and use. We treated the extent of intergovernmentalism and supranationalism in the EU, at any given point in time, in any given policy domain, as the central question organizing our research (Tsebelis and Garrett follow us, but proceed differently). Further, we explictly and consistently distinguished between “organizations” (like the European Court of Justice, the Commission, the Council of Ministers, and the Parliament), which can also be conceived as actors, and “institutions,” which are structures of rules and procedures. We always treated relevant institutional settings as independent variables (that is, we explored the extent to which different sets of institutions structure interactions differently), and at times (given feedback arguments) and as dependent variables (as integration proceeds, and as institutions are modified by interactions).

Finally, Tsebelis and Garrett’s misrepresentation of our arguments obscures the real relationship between our work and that of Burley and Mattli. The Burley and Mattli article focused on the Court’s interactions with national judges and private litigants, a topic on which Tsebelis and Garrett's various models have nothing to say. Stone Sweet and Caporaso's chapter concerned how these interactions impact concrete policy outcomes, including those produced by member-state governments and EU legislative processes. Contrary to Tsebelis and Garrett's assertion, Stone Sweet and Caporaso do not disagree with Burley and Mattli on any essential point; indeed, they wrote: “Our explanation [is] broadly consistent with the neofunctionalist account developed by Mattli and Slaughter.”(21) The key difference between the two analyses is that Stone Sweet and Caporaso updated the project and pushed it much further empirically. Stone Sweet and Caporaso (1) stipulated (what our critics refer to as) “microfoundations,” some of which overlap with those proposed by Burley and Mattli, (2) specified (what our critics call) the “institutional bases” of judicial politics in the EU, and (3) tested hypotheses about how the legal system operates and with what consequences.

We have, in this brief section, sought to correct the record, not to answer criticisms. Of course, how we proceeded in European Integration and Supranational Governance, and in related published papers, is subject to critique. Indeed, we do not always have satisfactory answers for various lines of criticism that have been well and fairly made. Unfortunately, Tsebelis and Garrett develop none of these.

II

Tsebelis and Garrett adopt various positions based on arguments that we ourselves made. It may be more constructive to discuss these issues as points of potential agreement.

A. Throughout their article, Tsebelis and Garrett argue for more carefully specified models of how the organs of EU governance, interacting under different sets of procedures and decision-making rules, produce various outcomes (e.g., the details of secondary legislation and treaty revision). We agree. In their article, Tsebelis and Garrett use theoretical materials drawn from game theory and principle-agent (p-a) frameworks to do so.

In the European Integration and Supranational Governance volume, four separate chapters discussed and used delegation theory to account for why various outcomes are patterned as they are. In the introduction, Stone Sweet and Sandholtz referred to “principal-agent relations” as one of the “two logics” that help to explain certain recurrent characteristics of EU politics, and then discussed how various contributors (Mark Pollack, Paul Pierson, and Stone Sweet and Caporaso) used p-a frameworks to organize their empirical research.(22) The arguments laid out in each of these chapters bear directly on Tsebelis and Garrett's piece (but are ignored).

Stone Sweet and Caporaso explicitly deployed an institutionally-derived p-a framework to correct a crucial flaw in Garrett's own prior work, a 1992 article in International Organization (cited by Tsebelis and Garrett). That article did not properly attend to the underlying institutional bases of EU politics, which partly accounts for why the propositions he offered in that piece found so little empirical support when subject to testing. Garrett argued that outcomes produced by the European legal system followed more or less directly from the preferences of the member-states, especially the more powerful ones. In their chapter, Stone Sweet and Caporaso evaluated these and other propositions about how the legal system was constructed and operates, and used p-a ideas to ground an argument as to why the Court should not be constrained in the way that Garrett had proposed. In their present article, Tsebelis and Garrett correct this flaw, in a more formal way than do Stone Sweet and Caporaso, but not completely (e.g, they inexplicably leave out private litigators and national judges, arguably the most important actors in the system, after the Court itself).

No non-specialist reader of the Tsebelis and Garrett treatment of our project could guess that their article, if anything, follows from ideas that we, and our collaborators (most particularly, Mark Pollack, whose seminal work get no notice in the piece), in fact had developed first.

B. Tsebelis and Garrett base their critique of Burley and Mattli on the latter's alleged failure to recognize that member-state governments have incentives to delegate, and on their alleged presumption that governments do not typically monitor their agents, including the Court. This critique of Burley and Mattli is overdrawn (Burley and Mattli in fact emphasize the well-known fact that member-state governments opposed many of the Court's most important “constitutional” rulings, before they were taken).

In our volume, we assumed that the EU's legal processes are constantly monitored by member-state governments. Some of the chapters presented data on the briefs that governments and the Commission file to tell the ECJ how it should decide pending cases. Indeed, Stone Sweet and Caporaso analyzed every brief that member-state governments submitted in two areas of litigation, as one means of assessing the impact of the expressed preferences of governments on the Court's rulings over time and across policy domain, and to test propositions Garrett had developed in an earlier article. They found no support for Garrett's claims which, in any case, had not been substantiated empircally.(23) Thus, again, our volume contained the innovation that Tsebelis and Garrett propose, namely, to take seriously that the Court knows its activities are being monitored by governments and other actors in the system. Yet, unlike Tsebelis and Garrett, Stone Sweet and Caporaso operationalized the point, and tested specific hypotheses against data.

By ignoring this clear overlap with our work, Tsebelis and Garrett deprive readers of a potentially useful discussion of these findings in light of their own theory-driven expectations.

C. Tsebelis and Garrett argue – indeed, they call it a “central contention” of their piece – that the Court and other supranational organs possess “more discretion . . . to move policy outcomes closer to their own preferences,” in situations in which it is more “difficult . . . for new legislation to be passed.”(24)

The point is also a “central contention” of European Integration and Supranational Governance. It informed the analysis in various parts of the introduction (e.g., pp. 17-20), as well the chapters of Pierson, Pollack, Sandholtz, Stone Sweet and Caporaso, and Stone Sweet and O'Reilly. Stone Sweet and Caporaso devoted more than three pages to the various dimensions of this issue (pp. 95-98). These discussions prefigured, in virtually every essential respect, the basic orientations - but not the conclusions - of the Tsebelis and Garrett article. We are puzzled, again, at Tsebelis and Garrett’s choice to ignore this obvious connection between their project and ours.

III

We now turn to specific areas of disagreement with the article published by Tsebelis and Garrett. Chief among these are the following: they have little constructive to offer those to whom they direct criticism; their models exclude from consideration actors, organizations, and institutions that those they criticize have argued are crucial to understanding the dynamics of EU politics; and they provide no empirical evidence in support of the contention that their “framework” performs better than the ones they criticize.

A. Tsebelis and Garrett situate their research in light of what they call “three influential studies”: Stone Sweet and Sandholtz; Burley and Mattli; and Marks, Hooghe, and Blank. But there is a chasmic disconnect between the literature Tsebelis and Garrett review, on the one hand, and what they actually give readers in the main body of their article, on the other. Burley and Mattli's legal neofunctionalism, the Sandholtz and Stone Sweet project, Andrew Moravcsik's intergovernmentalism, and Marks, Hooghe, and Blank's research on the development of multi-level governance, are broad-gauge research programs that seek to explain change in the most important features of European integration and EU politics over time. Tsebelis and Garrett's article deals with one, relatively small, piece of the puzzle; it makes a narrow, though potentially important, contribution on how the EU's organs interact in legislative processes. But while it contains several snapshots, it is not a moving picture. Some of the studies criticized by Tsebelis and Garrett seek to explain the evolution of the institutions, changes that Tsebelis and Garrett apparently take as unproblematic and given. Tsebelis and Garrett do not show why their framework (assuming properly specified models) could not be accommodated (as a kind of subprogram) within the approaches they criticize. It would seem that any existing theory of integration easily subsumes discreet models of legislative politics, including those proposed by Tsebelis and Garrett.

Curiously, the models Tsebelis and Garrett actually offer have little, and sometimes nothing, to do with the arguments surveyed and criticized in the first section of their article. This disconnect is especially glaring when one considers the research of Burley and Mattli, and Stone Sweet and Caporaso, on the legal system. Both sought to explain the construction of the EU's legal system, and both focused on interactions between litigants, national judges, the ECJ, and member-state governments, as these interactions are structured by the Treaty and the ECJ's rulings over time. This work has shown, conclusively that the activities of litigants, national judges, and the ECJ have combined to produce a self-sustaining system that routinely produces outcomes that member-states did not previously anticipate, or would not have produced on their own (given the pertinent decision-making rules in place). Tsebelis and Garrett do not model these interactions, and they neither mention nor defend this choice. Instead, they pretend that litigants, national judges, and relevant EU institutions (including the preliminary reference procedure laid down by the Treaty of Rome, the doctrines of direct effect and supremacy constructed by the Court, and others) do not exist. Thus, Tsebelis and Garrett's criticisms of the most important research on the Court are simply beside the point, if not gratuitous.

B. Many of the chapters in the European Integration and Supranational Governance volume explored the political impact of interstitial modification of the treaty through practice and interpretation (Pierson, Sandholtz, Sbragia, Stone Sweet and Caporaso, and O'Reilly and Stone Sweet). When the terms of jurisdiction and competence in any given policy domain shift as a result, then the character and dynamics of legislative politics in that domain will change. Tsebelis and Garrett have nothing to say about how or why such modifications occur, although we have shown them to be fundamental to the evolution of the “institutional bases” of EU politics. So have Burley and Mattli, and Marks, Hooghe, and Blank (see also the recent book by Adrienne Héritier,(25) and the follow-up volume to European Integration and Supranational Governance(26)).

In any case, the institutional “foundations” examined by Tsebelis and Garrett are exceptionally narrow, and are incapable of providing a basis for a theory of institutional change and politics in the EU. Take the linkages between litigation and legislation in the EU. Private litigants and national judges supply the Court with most of its caseload, and the Court has responded by reconstructing the treaty in crucial ways. But Tsebelis and Garrett ignore these actors, processes, and outcomes. When the Court preempts the EU legislature, by “constitutionalizing” the provisions of proposed or existing secondary legislation (through locating the substance of the legislative provisions in the Treaty), the ruling shifts the site of reversal from a legislative to a treaty-revision process; in Tsebelis and Garrett’s preferred language, but also in the language of our project, the politics shift from “intergovernmental” to “supranational.” Tsebelis and Garrett are oblivious to this fact, although it has happened systematically in some areas.(27) And the Court can all but require the EU legislator to act, when it places member-state governments in the shadow of litigation (documented in Sandholtz, and Stone Sweet and O'Reilly). Yet these outcomes, like so many others, are theoretical impossibilitites in the EU modelled by Tsebelis and Garrett.

Thus, Tsebelis and Garrett do not address much of what is at the very core of their own narrowly-focused agenda: how rules and procedures contained in the Rome Treaty condition legislative processes and outcomes over time, and how these institutions structure interactions within specific legislative processes.

C. There now exists a huge pile of evidence showing that integration has produced what Tsebelis and Garrett call “unintended consequences.” An important question is begged: unintended for whom? Surely, the outcomes of legislative and judicial processes in the EU are always intended, or desired, by some actors. The theoretically important issue for Tsebelis and Garrett concerns the extent to which EU processes produce outcomes that can be identified as unintended by the governments of the member states. The European Integration and Supranational Governance volume provided some of the evidence that the EU does so chronically, and explored in detail the underlying institutional logics for that claim. Tsebelis and Garrett do not address these arguments or findings, presumably because the dynamics that produce them fall outside the narrowly-drawn parameters of their models. Nonetheless, they bluntly assert that we and others (e.g., Pierson) exaggerated the extent or importance of such outcomes.(28) This is a reckless claim, since not only do the authors ignore our research, they offer no evidence in support of the proposition that they are right, and we are wrong; indeed, data is conspicuously absent from their piece. Instead, they derive their conclusion from their own models, none of which deals with the processes that we argue have routinely generated “unintended” outcomes.

More broadly, we see the question of “intentionality” as central to what remains of the debate between intergovernmentalism and modern versions of neofunctionalism, given the full-scale retreat of intergovernmentalists over the past decade. At issue is not whether important unintended consequences (i.e., unintended from the point of view of governments, at some ex ante moment) have been produced, but how one might assess the importance of these outcomes in light of other more consequentialist logics emphasized by intergovernmentalists like Moravcsik, by delegation theorists like Pollack, or by Mattli in his recent book on regional integration. Those who care about this issue should read Moravcsik,(29) Pollack,(30) and Mattli,(31) not Tsebelis and Garrett, since nothing in their piece addresses this question in a direct, or novel, way.

Conclusion

The intermittent debate between different theories of integration has rightly emphasized one of the fundamental questions regarding the evolution of the EU polity: to what extent are European organs and institutions best viewed as instruments of the member states, albeit flexible, creative instruments, and to what extent have they evolved through processes that routinely escape the collective control of governments? Tsebelis and Garrett raise exactly this question. They assert, as a central finding of their article, that the “purported ‘law of unintended consequences’ has empirically been riddled with many more exceptions than most commentators on European integration suggest.”(32) Unfortunately, it is impossible to assess their claim because (1) no one, to our knowledge, has announced such a “law”; (2) Tsebelis and Garrett do not define what they mean by “unintended consequences” so that their readers might know what events or outcomes would or would not fall into the category; and (3) Tsebelis and Garrett decline to invoke any evidence on their own behalf.

But empirical research on these matters does, in fact, exist. The intergovernmentalist strategy, best exemplified by Moravcsik’s work, has been to focus on EU treaty making and treaty revision, and – in more recent versions – to use delegation theory to link bargaining episodes (and the institutions thus designed) to what happens between bargains.(33) But this strategy raises two important issues. First, neofunctionalism was always, partly, a theory about delegation, and there is nothing a priori incommensurable between it and p-a approaches to the topic. Quite the contrary, p-a theory and ideas about pre-commitment can, and have been, profitably used by critics of intergovernmentalist theory, including by the heirs of neofunctionalism. Second, intergovernmentalists have not told us what kinds of outcomes are not predicted by their improved theory. Instead, they tend to construe innovations that do emerge in light of some prior decision on the part of governments to delegate to EU organs. In any case, intergovernmentalists now agree with us that the EU routinely produces “unintended consequences,” if what we mean by that phrase is the creation or development of institutions (rules, procedures, policies) and organizational capacities (the terms of jurisdiction of the EU organs) to which member-state governments would not have agreed, at a given moment in time, within existing institutions. That is, there appears tobe broad agreement that the more governments seek to enhance the credibility of commitments through extensive delegation to EU organs, the more EU organs can be expected to generate outcomes that governments would not have produced on their own, given existing decision rules.

For their part, Tsebelis and Garrett have noticed that the rules governing legislating in the EU have changed, and they seek to determine some of the consequences of these changes. But why have legislative institutions been restructured? How have changes in legislative institutions impacted, or been impacted by, the broad course of integration? Why has the EU steadily evolved from an interstate treaty toward a federal polity, that is, why have the institutional foundations of the EU become more supranational over time? We are still waiting for Tseblelis and Garrett to join a constructive conversation about existing attempts to answer these and other important questions that rightly claim the attention of the field.

Notes

References

Mattli, Walter. 1999. The Logic of Regional Integration: Europe and Beyond. New York: Cambridge University Press.

Héritier, Adrienne. 1999. Policy-making and Diversity in Europe: Escaping Deadlock. Cambridge: Cambridge University Press.

Moravcsik, Andrew. 1998. The Choice for Europe. Ithaca, NY: Cornell University Press.

Pollack, Mark A. 1997. Delegation, Agency, and Agenda Setting in the European Community. International Organization 51(1): 99-134.

Pollack, Mark A. 1998. The Engines of Integration? Supranational Autonomy and Influence in the European Union. In Wayne Sandholtz and Alec Stone Sweet, eds., European Integration and Supranational Governance, 217-49. Oxford: Oxford University Press.

Sandholtz, Wayne and Alec Stone Sweet. 1998. European Integration and Supranational Governance. Oxford: Oxford University Press.

Stone Sweet, Alec and Wayne Sandholtz. 1997. European Integration and Supranational Governance. Journal of European Public Policy 4(3): 297-317.

Stone Sweet, Alec, Wayne Sandholtz, and Neil Fligstein, eds. 2001. The Institutionalization of Europe. Oxford: Oxford University Press.

Stone Sweet, Alec and James A. Caporaso. 1998. From Free Trade to Supranational Polity: The European Court and Integration. In Wayne Sandholtz and Alec Stone Sweet, eds., European Integration and Supranational Governance, 92-133. Oxford: Oxford University Press.

Stone Sweet, Alec and Wayne Sandholtz. 1998. Integration, Supranational Governance, and the Institutionalization of the European Polity. In Wayne Sandholtz and Alec Stone Sweet, eds., European Integration and Supranational Governance, 1-26. Oxford: Oxford University Press.

Tsebelis, George and Geoffrey Garrett. 2001. The Institutional Foundations of Intergovernmentalism and Supranationalism in the European Union. International Organization 55 (2): 357-90.

(1)

Tsebelis and Garrett 2001, 357.

(2)

Sandholtz and Stone Sweet 1998.

(3)

Stone Sweet and Sandholtz 1997.

(4)

Tsebelis and Garrett 2001, 361.

(5)

Ibid., 362.

(6)

Stone Sweet and Sandholtz 1997, 300.

(7)

Ibid., 301, 310.

(8)

Stone Sweet and Sandholtz 1998, 7-9.

(9)

Tsebelis and Garrett 2001, 361-62.

(10)

Stone Sweet and Sandholtz 1997, 311.

(11)

Sandholtz and Stone Sweet 1998, chaps. 5 and 6.

(12)

Sandholtz and Stone Sweet 1998, chap. 4.

(13)

Tsebelis and Garrett 2001, 363.

(14)

Stone Sweet and Sandholtz 1997, 304.

(15)

Stone Sweet and Sandholtz 1997, 299.

(16)

Stone Sweet and Sandholtz 1997, 306.

(17)

Stone Sweet and Sandholtz 1997, 299.

(18)

See Sandholtz and Stone Sweet 1998, chaps. 2, 4, 5, 6 and 10.

(19)

Tsebelis and Garrett 2001, 385.

(20)

Tsebelis and Garrett 2001, 385.

(21)

Stone Sweet and Caporaso 1998, 105.

(22)

Stone Sweet and Sandholtz 1998, 19, 22-23.

(23)

Stone Sweet and Caporaso 1998, 115, 120-21, 122-23.

(24)

Tsebelis and Garrett 2001, 358.

(25)

Héritier 1999.

(26)

Stone Sweet, Sandholtz, and Fligstein, eds., 2001.

(27)

Stone Sweet and Caporaso 1998, 121-28.

(28)

Tsebelis and Garrett 2001, 360, 388.

(29)

Moravcsik 1998.

(30)

Pollack 1997, 1998.

(31)

Mattli 1999.

(32)

Tsebelis and Garrett 2001, 360, 388.

(33)

See Moravcsik 1998.