The dynamics of European integration reactivate unresolved questions that have been worked on by practitioners and theoreticians for centuries. What are the `driving forces' that form and change political orders? What is the role of human intention, reflection and choice in the development of political institutions and good government? In institutional matters, do we know how to reform? How, and under what conditions, can political actors rise above, and get beyond, existing institutional structures?6
Students of institutional dynamics have given very different answers to these questions. In particular, they have disagreed about political agency, the relative importance of environmental imperatives and intrinsic dynamics which go beyond the comprehension and control of political actors, and historical processes of gradually evolving systems of meaning and incremental change. Thus, different perspectives will suggest different answers to what kind of processes are likely to produce radical change, of the kind suggested by Fischer, in the European polity.
Pawns, organic development and imperative processes. Political actors are sometimes portrayed as largely impotent pawns. They are captives of imperative (technological, economic, demographic etc.) processes in their environments, or of intrinsic institutional dynamics beyond their comprehension and control. They may codify, through formal reorganisation, change that has already happened, but they are unable to structure future institutional developments deliberately. The key processes of change are external competitive selection or internal organic processes of institutional birth, growth, stagnation and death. In the first case, only comparatively efficient institutions and political orders survive. The others lose support and disappear. In the second case, all institutions have their heydays. Then, they wither and die, irrespective of whatever reform plans political leaders present (Kimberly & Miles 1980, Olsen 1992).
Engineers, design and institutional choice. In contrast, the concept of governance is about how differently, over any given period of time, our social and political life, can be purposefully shaped (Dunn 1990:161). An institutional reform policy is about explicitly changing social and political life through new institutional arrangements. A constitutional reform policy is about changing the basic institutions and principles of governance in order to change the identity and character of the polity.
Portraying political leaders as institutional engineers, and institutions as malleable, is consistent with a democratic ethos of governance. Democracies are supposed to be able to design and choose institutions in order to improve the welfare of citizens. The key questions involved in explaining institutional dynamics, then, are: who are the significant actors? what do they want an institutional arrangement to accomplish? what do they believe different arrangements will accomplish? what resources can they mobilise?
Under special circumstances, `We the people' can form a constitutional convention and deliberately rearrange the whole political order (Ackerman 1991). Under normal conditions, political intention, will and power secure rational adaptation of institutions that are not working well. Institutional dynamics become a question of bargaining and building that win coalitions among competing interests. In a short-term perspective, however, constitutive institutions and rules limit the legitimate space of institutional design-what can be changed, how fast, and in what ways. Heterogeneous societies, in particular, demand strongly qualified majorities to change the power of different branches and levels of government or the relative power of public authorities and citizens (Weaver & Rockman 1993:464).
To understand the dynamics of European integration, however, we have to go beyond institutional change as a simple reflection of differences in the comparative functional efficiency of alternative forms. In other words, we have to question the idea that political institutions normally adapt fairly quickly to changes in external conditions and human purposes through processes of competitive selection and rational adaptation.7
The pawn and engineering perspectives lead to different assessments of reform plans. Consider, for instance, the constitution-writing aspect. In a period of flux, uncertainty and ambiguity, an engineering approach suggests that the time is ripe for deliberate intervention, to give more structure to current developments. The pawn perspective suggests the opposite. A period of flux, uncertainty and ambiguity is definitely not the right moment for codification and constitution writing. Both perspectives, however, suppose that the comparative efficiency of forms of governance and organisation is the key factor that determines their chances of survival.
Gardeners, incremental reforms and meanders. Studies of comprehensive institutional reform in large-scale, complex and dynamic systems with unresolved conflict suggest that reorganisation of the polity with a single scheme is unlikely to be politically digestible. Change is not well understood and controlled, and actual reforms are usually incremental rather than comprehensive. Governance is less a matter of engineering than of gardening.8 Existing institutional configurations are usually the result of long historical processes, involving conflicts, victories, defeats and compromises, as well as processes of interpretation, learning and habituation.9 It is difficult to subject institutional evolution to tight control, and history becomes a meander (March 1994).
In this perspective, reforms are influenced by environments and political actors. Yet, institutions do not adapt instantaneously or efficiently to minor changes in will, power, or circumstances. Institutions cannot be changed into any arbitrary form and comprehensive reform requires strong organisational capabilities to stabilise attention, mobilise resources and cope with resistance (March & Olsen 1983,1989). Change does not start with clear problem definitions and objectives, which lead to tailor-made institutional designs, as suggested by instrumental-functional approaches. Often, change takes the form of deliberation and `sounding out' processes, involving the use of ambiguity, `soft laws' and tacit agreements (Blichner & Sangolt 1994, Sverdrup 1999).10
6 Hamilton, Jay & Madison (1787, 1964:1); Mill (1861, 1962:1); March & Olsen (1989,1995,1998); Olsen (1997); Brunsson & Olsen (1993); Sartori (1997:xi).
7 March & Olsen (1989), North (1990). Brennan and Buchanan also criticise the `hidden hand' assumption in economic theory: ... `great damage has been and is being done by modern economists who argue, indirectly, that basic institutional change will somehow spontaneously evolve in the direction of structural efficacy,' (Brennan & Buchanan 1985:149).
8 Szanton (1981:24). See, also, March & Olsen (1983:287,1989,1995); Olsen (1996); Benz & Goetz (1996), Knill (1999).
9 This is certainly true for state and nation building processes in Europe (Eisenstadt and Rokkan 1973), (Rokkan 1975,1999), and also for the development within the EU (Pierson 1996:126-7).
10 Sounding out, in contrast to Habermasian force-free deliberation, involves the systematic use of ambiguity. It is important for each participant to avoid taking an early stand. While the participants will try to reveal the trend in their beliefs and preferences, and attempt to move the final outcome toward a desired end result, they will avoid very accurate indications of beliefs and preferences. They will always retain some degree of counter-argument and contradiction in their statements. The process is time consuming. The outcome is the result of more and more participants accepting a certain alternative as the best solution, while other alternatives `fade away' (Olsen 1972:273). The behaviour is purposeful, but it is reflecting what in a culture is defined as appropriate behaviour and process, not strategic calculation.