Cambridge University Press
0521832535 - Procedural Politics - Issues, Influence, and Institutional Choice in the European Union - by Joseph Jupille
CHOICE, CONSTRAINT, AND EUROPEAN UNION INSTITUTIONS
Procedure hasn't simply become more important than substance - it has, through a strange alchemy, become the substance of our deliberations. Who rules House procedures rules the House.
- Robert H. Michel, R-Ill.1
This is a book about "procedural politics," the everyday conduct of politics not within, but with respect to, political institutions. The questions it asks are fundamental to political science, and indeed to "institutional" approaches across the social sciences: why, when, how, and with what effects do actors attempt to influence their institutional environment? Why, when, how, and with what effects, by contrast, do rules constrain them? The book develops and tests answers to these questions in the context of EU politics. The results of this inquiry paint a novel picture of EU politics and policymaking, suggesting most importantly (but somewhat paradoxically) that the EU exhibits a more profound degree of rule governance than is usually recognized. But these results generalize far beyond the EU, not only to other international organizations, but also to domestic political systems and, indeed, to all institutionalized political and social systems.
My general argument can be succinctly summarized. I assume that actors seek to ensure the usage of institutions (rules) that maximize their political influence. They are constrained, however, by the strategic nature of institutional choice - the need to interact with others - and by the availability of institutional alternatives. As a result, I argue that incentives (in the form of potential influence gains) and strategic opportunities (in the form of the availability of institutional alternatives) combine to produce procedural politics, everyday politics with respect to rules. Actors play procedural politics through predictable means (most notably, procedural coalition formation and "gaming" the criteria that govern institutional selection) with predictable effects. These effects include, but are not limited to, the rules themselves, the content of public policy and the efficiency with which it is made, and long-run changes in higher-order rules, including the most fundamental rules of all: constitutions. This procedural political cycle reinforces the political primacy of higher-order rules, paradoxically entrenching rather than undermining rule governance.
Subsequent chapters will develop this argument at length. The bulk of this introduction will locate it within the broader institutional literature and relate it to existing work on European Union institutions. The introduction ends with a plan of action for the rest of the book.
Douglass North has defined institutions as "the rules of the game in a society, or, the humanly devised constraints that shape human interaction" (North 1990, 3). Among its other advantages, this definition embraces what Robert Grafstein (1992) has dubbed the "dual nature" of institutions. On the one hand, institutions represent objects of human creation or choice. They are, as North puts it, "humanly devised." On the other hand, institutions also represent sources of independent effects - they are "constraints that shape." This book gets to the heart of this tension: when are the rules available to strategic actors (as [endogenous] objects of choice), and when, by contrast, are they unavailable (as [exogenous] sources of constraint or independent effects)?
Existing institutional work in political science and on the European Union provides the elements of an answer but not the answer itself. Institutional work differs along many dimensions. One key dimension involves the theoretical place of institutions as either endogenous (explicable in theoretical terms) or exogenous (external to the theory in question) (Jupille and Caporaso 1999; Shvetsova, 2003). A second key dimension involves the level of analysis, with some analysts privileging higher-order institutions (such as constitutions) and others emphasizing lower-order institutions (such as legislative procedures) (Kiser and Ostrom 1982, 208; Ostrom 1990, 1995). These treatments tend to correlate with the explanatory use of time, with those allowing endogeneity and focusing on higher-order rules more likely to take a diachronic (over-time) approach, and those emphasizing exogeneity and focusing on lower-level rules more likely to develop synchronic (at-a-given-time) explanations (Knill and Lenschow 2001). Figure 1 identifies four modal types of institutional analysis (design, change, selection, and effects), which offer different mixes of institutional endogeneity and exogeneity, higher-order and lower-order institutions, and diachronic and synchronic explanatory approaches. Decomposing the literature in this way lays bare the raw materials that must be pieced together into a coherent explanation of institutions, including procedural politics, everyday politics with respect to rules.
Figure 1 Approaches to Institutions
Theories of institutional design (alternatively, institutional creation) explicitly treat institutions only as dependent variables, that is, as phenomena to be explained with reference to noninstitutional factors such as preferences, power, functional needs, and strategic interactions (Goodin 1996a, 1996b; Koremenos, Lipson, and Snidal 2001). Hence, as shown in Figure 1, they explain over-time movement from an institution-free environment (Ø) to one governed by a new set of higher-order rules (constitution A), at which level design theories tend to operate almost exclusively.2 One variant views institutions as the by-products of other interactions such as economic exchange (Schotter 1981) or distributional bargaining (Knight 1992). A second, perhaps more common, variant views institutions intentionally, created to serve functional or more overtly political goals (Coase 1937/1988; Williamson 1975, 1985; Weingast and Marshall 1988; Horn 1995; Keohane 1984).
The functional design approach has been widely applied in the literature on the EU, the by-product variant less so. Those working in the idiom of "constitutional choice" portray EU member states in an institution-free environment choosing the rules that will best serve their future contracting and political needs (König and Bräuninger 2000; Bräuninger et al. 2001). More influentially, Moravcsik's liberal intergovernmental approach (1991, 1993, 1998) embraces and extends Keohane's (1984) functional theory of international regimes, which explains international institutions as functional responses to international market failures permitting states to capture otherwise unavailable gains from cooperation. Moravcsik's (1998) tripartite explanation of European integration begins with a liberal theory of state preference formation (see Moravcsik 1997), proceeds to model intergovernmental bargaining, and concludes by theorizing institutional choice (design) in the EU.3 On this last element, he explains institutions as functional responses to the transaction problems that member states face. Institutional design, and in particular the pooling of sovereignty and the delegation of authority to supranational agents, enhances the credibility of commitments by reducing the possibilities for ex post opportunism.
Design approaches are necessary for full institutional analysis. They respond directly to the definitional requirement that institutions be the objects or products of human creation. However, they suffer from problems with respect to each element of the dual nature of institutions. On the "humanly devised" side, they downplay the broader social and historical determinants of institutions (Oberschall and Leifer 1986; Granovetter 1985, 1992; Bromley 1989; Wendt 2001) and the prevalence of collective action problems or multiple efficient equilibria (Krasner 1985, 1991; Garrett 1992; Garrett and Weingast 1993). On the "constraints that shape" side of North's definition, design approaches, which seek to model institutions only as dependent variables, suffer from two problems. First, though design theorists seek fully to endogenize institutions to noninstitutional parameters, they almost invariably end up smuggling rules into the right-hand (independent variable) side of their explanatory equations (Field 1979, 1981, 1984; Bromley 1989; Scharpf 1999, 165). Second, though they rely quite heavily on anticipated institutional effects to explain institutional design, they often fail explicitly to theorize those effects. To this extent, the connections they draw between design and constraint (effects) resemble assumptions or, at most, inferences, rather than theoretical claims. As Bates puts it, design approaches tend to "confound the analysis of the role of institutions [i.e., anticipated institutional effects] with a theory of their causes" (Bates 1988, 387).
Theories of institutional change have tended to take seriously the criticisms of the design approach. In particular, while they also work at the level of higher-order rules and portray institutions as endogenous only diachronically, they embrace preexisting higher-order rules as constraints on institutional change. That is, they explain the modification over time of existing institutions or the movement from one set of institutions to another (as from constitution A to constitution B in Figure 1). Existing institutions - sometimes the very institutions that form the objects of choice - themselves constrain institutional innovation. They thus reside on both sides of the explanatory equations offered by theorists of institutional change. These institutional constraints on institutional change derive from numerous sources, including but not limited to institutionally determined risk; uncertainty; discounting, transaction, and other costs; countervailing interest and power; and increasing returns to scale (North 1981; Krasner 1984; Levi 1988; Shepsle 1989; David 1985, 1994; Thelen 1999; Pierson 2000a, 2000b).
Pierson (1996) offers a compelling account of EU institutional change. Pierson's theory explains why formal institutional change following a design logic - that is, with weak prior institutional constraints - is difficult, and why (and in what ways) informal change can and does occur. He argues that although institutions initially reflect their creators' (i.e., EU member states') interests, control "gaps" arise that are difficult (costly) to close. Such gaps result from the partial autonomy of EU institutions, national leaders' restricted time horizons, unanticipated institutional consequences, and exogenous (e.g., electorally induced) shifts in national leaders' preferences. Several factors militate against closure of these gaps, including resistance by (institutionally created) supranational actors, institutional barriers to reform (e.g., supermajority rules for institutional change), and sunk costs and positive feedbacks (see also Marks, Hooghe, and Blank 1996; Pollack 1996, 1997). The rise of difficult-to-close gaps, Pierson concludes, explains both the trajectory of and limitations to EU institutional change (but see Farrell and Héritier 2003).
The institutional change approach, taken generally, exhibits many strengths. It recognizes that institutions are potential objects of choice. Like design approaches, it works at the level of higher-order rules. Unlike them, it theorizes the limits to (constraints on) choice at this level. Most important, among the limits that it identifies are institutions themselves, which can constrain change in any number of ways, as described above. At the same time, though, institutional change approaches suffer from some characteristic weaknesses. They fail to consider the ways in which the lower-order rules established by higher-order institutions might themselves constitute objects of political choice. (In the language of Figure 1, they fail to consider institutional selection.) What is more, they generally fail explicitly to connect lower-order and higher-order rules by theorizing the feedback between the former and the latter. It seems fair to say that they recognize the possibility of such a connection, and that it is consistent with their approach. But to this point this insight remains underdeveloped.
Analyses of institutional effects focus on the "constraints that shape" clause of North's definition. They tend to operate at the level of lower-order institutions such as legislative procedures (e.g., procedures b1, b2, and b3 in Figure 1) and tend to work at shorter time scales. Institutions reside only on the right-hand side of effects equations, usually intervening between preferences and power on the one hand and strategies, bargaining, and substantive outcomes on the other, and they are used to explain noninstitutional outcomes (y1, y2, y3) such as power, policies, budgets, and the like. Americanist scholars pioneered and developed the approach (Shepsle 1979, 1986; Denzau and Mackay 1983; Shepsle and Weingast 1984a, 1987; Krehbiel, Shepsle, and Weingast 1987), and while relatively few comparativists have followed suit (Huber 1992, 1996), early efforts by Garrett (1992), Tsebelis (1994), and Steunenberg (1994) spawned dozens of follow-on efforts in the EU literature. EU institutional effects analysts have closely followed the parent literature, taking higher-order rules as given and generating comparative static results of the effects of procedural variation on bargaining dynamics, political power, and policy outcomes.
Institutional effects approaches fulfill a critically important role in institutional analysis by theorizing the many and varied ways in which institutions can independently shape political outcomes. They can claim many successes. However, a key omission (failing to consider institutional causes) runs the risk of producing errors of commission. For example, it may be fallacious to infer political power from an effects analysis without considering institutional causes (Krehbiel 1988, 1991; Krehbiel and Rivers 1988; Cox and McCubbins 1993). Similarly, the neglect of institutional choice poorly serves policy analysis. "Far from being exceptional occurrences," Majone writes, "attempts to modify procedural rules and other institutional constraints are so pervasive that no descriptive or prescriptive policy analysis can be complete that does not explicitly take institution-changing behavior into consideration" (Majone 1989, 114). Finally, failure by effects analysts to consider institutional causes unjustifiably - and, given their rationalist foundations, puzzlingly - truncates strategic actors' behavioral repertoires and the range of factors over which they might exercise choice. Effects analyses tend to portray agents as fully sophisticated within institutions but incompletely sophisticated (i.e., naïve or myopic) with respect to them (Krehbiel 1988, 310-311). If institutions produce outcomes in the ways predicted, and if actors prefer the outcomes produced under available institutional alternatives to those produced by rules currently in (or proposed for) use, they face incentives to engage in politics with respect to, and not just within, those rules (Goldberg 1974; Riker 1980; McKelvey and Ordeshook 1984; Moe 1990a, 1990b).
Institutional Selection and Procedural Politics
This book, focusing as it does on everyday politics with respect to rules, relates to existing work in two ways. First, it fills a gap in the literature by explicitly theorizing institutional selection, that is, the synchronic choice of lower-order rules. However, second and perhaps more important, it provides new explanatory leverage on institutional effects, and by explicitly contemplating cross-level, over-time feedback between lower- and higher-order rules, it improves our understanding of institutional change. With respect to the EU, it paints a more coherent and complete picture of the operation of EU institutions than is currently available, moving beyond the stale dichotomy between "everyday politics" (within rules) and "historic grand bargaining" (with respect to rules). In particular, it finds that opportunities for institutional choice present themselves more frequently than previously suspected. At the same time, and somewhat paradoxically, it also bespeaks the EU's transformation from a diplomatic system governed by balance of power principles to a constitutional system governed by the rule of law. Even the most egregious manipulations surrounding lower-order choice tend to reinforce higher-order constraint. More generally, it opens promising new avenues for thinking about the EU in comparative perspective, for thinking about rule governance in the international system, and for thinking about institutions in their "dual nature" and at multiple levels.
Institutional selection involves the choice of a lower-order rule (procedure b1, b2, or b3 from Figure 1, for example) from among a menu of alternatives delineated by the higher order rules (e.g., constitution B). The constitution also usually outlines the process by which procedures will be selected, connecting some antecedent condition to some procedural consequence. Some procedure must be selected if outcomes are to be generated. That selection process can occur more or less automatically (e.g., some condition is met that unproblematically triggers the use of a given procedure), in which case actors are contenting themselves to action within rules. Alternatively, the selection process can become politicized, in which case actors find themselves in the realm of everyday politics with respect to rules - the realm, that is, of procedural politics.
How does this process unfold? Procedural politics, I have suggested, responds to both incentives and opportunities. On the incentive side I model actors as influence maximizers. That is, I suggest that they seek to ensure the usage of rules that give them the most power in the political/legislative process. This position is controversial, and others are possible. Most generally, there exists little consensus as to whether policy, office, or votes primarily motivate political actors (Müller and Str⊘m 1999). More narrowly, we can imagine that efficiency, functionality, habit, indifference, or any number of other "motivations" might undergird institutional selection. Happily, the empirical pudding can provide some proof as to which of these potential motivations best explains observed behavior. However controversial, this argument implies that if all rules are politically equivalent - that is, if all provide the same level of political influence - then no incentive to act with respect to them exists. By contrast, as institutional alternatives differ in their influence properties, incentives to engage in procedural politics increase.
Opportunities for procedural politicians to pursue influence are not unlimited, however. In the first instance stands the strategic context. Absent dictatorship, institutional selection is a strategic choice process in which the choice of each actor depends in part upon expectations about the preferences, power, and strategic choices of others. A second, more properly institutional, set of constraints also presents itself. Higher-order rules (constitutions) define the menu of lower-order rules (procedures) and set forth the conditions or criteria for selecting from among them. However, these criteria often correspond with less-than-perfect clarity to the situation in the "real world." Situations can often be "jurisdictionally ambiguous," not determining the use of a single procedure, but potentially inviting the consideration of many. (Ultimately, only one rule can be selected and used.) As this jurisdictional ambiguity increases, procedural alternatives present themselves and procedural political opportunities expand.
In contemplating again the questions posed at the outset - why, when, how, and with what effects do actors attempt to influence their institutional environment? - I have already suggested answers to the first two. Why do actors engage in procedural politics? They do so in pursuit of influence. When do they engage in procedural politics? They do so when incentives and opportunities come together, with influence differences among alternatives defining incentives (gross incentives, in any case), and strategic context and jurisdictional ambiguity defining opportunities.4 As influence differences among alternatives and jurisdictional ambiguity increase, the likelihood of procedural politics likewise increases. Two questions remain: how do actors play procedural politics, and how does all of this affect important and interesting political outcomes?
Behavior and process will involve the pursuit of influence within strategic and higher-order institutional constraints, but it will also involve attempts to loosen those constraints (equivalently, to create opportunities). Starting with strategic constraints, actors will face incentives to forge coalitions with those sharing their procedural goals. While a straightforward proposition, the theory in Chapter 2 and the empirics in Chapters 5 and 6 suggest and confirm that this often involves the creation of otherwise unexpected partnerships. Procedural politics can make for strange bedfellows.
Strategic, influence-maximizing procedural politicians will also face incentives to loosen the higher-order institutional constraints that they face by "gaming" the constitutional criteria governing institutional selection. This will involve manipulating the degree (or perceptions of the degree) of jurisdictional ambiguity, with those seeking to promote new alternatives increasing it (or reducing it in a favorable direction) and those with the opposite preference opposing them. In the case of the EU, the treaty (constitution) forges a link between policy issues and the procedures used to pass laws dealing with them. Different procedures apply in the making of agricultural, environmental, transport, taxation, foreign, and other policies. In this context, "gaming the criteria" means manipulating issue definitions so as to establish the applicability of more favorable rules. Different criteria would produce different games, and while the specific criteria differ from one institutionalized system to another, the procedural political logic should apply equally well to institutional selection in firms, international organizations, European Unions, or nation-states.
Finally, what will be the effects of all of this? First, institutional selection determines which rules will be used in the making of policy, and if the new institutionalism as a whole can agree about one thing, it is that rules influence outcomes. To the extent that institutional selection involves procedural politics, both rules and outcomes will have been affected. Second, however, procedural politics reduces the efficiency of the decision-making process. While Goldberg (1974) and Majone (1989) liken procedural politics to an investment decision, this is arguably investment not in production but in rents. Procedural politics involves expenditures of time and money on influence rather than on the efficient making of good or functional policy.5 Third, and perhaps most important, procedural politics feeds back into constitutional change. It responds to jurisdictional ambiguities, which represent circumstances with respect to which constitutional contracts are incomplete. It may thus inform contracting parties (in the EU case, member states) about infirmities in their agreements and incite them subsequently to modify higher-order rules.
Plan of the Book
Summarizing, the approach pursued in this book coheres with prevailing approaches to institutions in the European Union and more generally. With design and change approaches, and consistent with North's definition, it treats institutions as at least potentially endogenous. With the effects approach, and also consistent with North's definition, it treats institutions as at least partly exogenous at any given point in time. Thus, like the institutional change approach, it models institutions as both independent and dependent variables, as predicates and outcomes of politics. But unlike most prevailing work, it explicitly embraces multiple levels of institutional analysis and suggests clear causal chains (including feedbacks) linking higher-order to lower-order institutions. Indeed, if anything it turns on its head the implicit claim that higher-order rules are best seen as endogenous and lower-order rules best treated as exogenous. This fuller picture of the operation of institutions, I suggest, will advance understanding not only of the EU but of rule governed systems more generally.6
© Cambridge University Press