The dynamics of institutionalization: transformation processes in Norwegian fisheries - includes appendix

Administrative Science Quarterly, Sept, 1995 by Petter Holm

INTRODUCTION

What are institutions? How do they work? Are they created intentionally, or do they come about as unintended by-products? When and why do they change? Despite the renewed interest in institutions across the social sciences during the last decade or so, no consensus has emerged on these questions. There are basically two ways of thinking about institutions. From a rational perspective, institutions are perceived as efficient solutions to predefined problems (Olson, 1965; Williamson, 1975). Institutions are instruments and they can be understood in the context of the tasks for which they were created. This view may contribute insights about the function of institutions and why they change but is still problematic. It disregards an important aspect of what institutions are, namely, frameworks for action and, as such, outside the scope of strategic manipulation. The second way of thinking about institutions, represented by the new institutionalists in organizational theory (Meyer and Rowan, 1977; DiMaggio and Powell, 1983), focuses particularly on the aspects of institutions that are ignored by the rational school. The new institutionalists define institutions as "socially constructed, routine-reproduced, program or rule systems" (Jepperson, 1991: 149). The study of institutions is the study of norm-governed behavior. In this perspective, the processes by which institutions are formed and reformed, which tend to be interest-driven and highly political, have been ignored. The result is an institutional theory that cannot explain how institutions are created and how they change (DiMaggio and Powell, 1991).

I remain sympathetic to the new institutionalism in organizational theory. I believe, however, that institutionalists should not focus exclusively on situations that are ignored by rationalists, i.e., those in which actors do not see or are prevented from pursuing their interests (DiMaggio, 1988). Institutionalist analysis must include all types of behavior, including those driven by interests and power (DiMaggio and Powell, 1991). This can be achieved if we take seriously the insight that institutions, while they are products of action, also constitute action. To handle both sides of this equation, I propose that institutions be seen as nested systems, drawing a distinction between actions guided by the established institutional order, on the one hand, and actions geared toward creating new or changing old institutions, on the other. This perspective makes it possible to retain the insight that institutions are products of action, and therefore constructed for some purpose, without giving up the notion that institutions are frameworks for action, and therefore taken for granted.

Theoretical Issues

From a rational perspective, institutions help align individual and collective interests. When a group of interdependent rational individuals act independently and with regard to their own interests, and the pay-off structure is a prisoner's dilemma, the best option for each individual produces a suboptimal solution for the group as a whole. In this situation, the group may establish an institution in the form of an agreement that aligns individual and collective interests. The institution is the solution to the collective-action problem. For instance, each member of a village knows that a wall around it would make everyone more secure. Each also knows that if the others fail to contribute to the wall, it will not be built, and his or her contribution will be wasted. Since all of them prefer to free-ride, the wall will not be built. They therefore might agree to form an institution that compels everyone to contribute (Bates, 1988).

From a rational perspective, then, an institution is a means to reconcile the inherent contradiction between individual and collective interests. This way of thinking about institutions has been influential in rational choice theory (Elster, 1979; Taylor, 1987), agency theory (Alchian and Demsetz, 1972; Libecap, 1989), economic history (Chandler, 1962; North and Thomas, 1973), and institutional economics (Olson, 1965; Williamson, 1975) as well as in organizational theory (Thompson, 1967).

While this model can explain how institutions work, it cannot serve as an explanation for their creation. It ignores the second-order collective-action problem that arises in the attempt to solve the original problem. Even if every member of the group in question realizes that the original problem, the building of a wall, can be solved by some institution, the same incentive problem occurs in the institutional project itself. Each member prefers the others to take on the cost to initiate, construct, and monitor the institution. Since all members think this way, the institution will not be established (Olson, 1965; Bates, 1988; Ostrom, 1990). This conclusion, which is somewhat problematic, since institutions in fact are established, stems from the implicit assumption that the structures of the first-order and second-order collective-action problems are identical. The underlying frame of reference is the image of a society constituted by presocial, autonomous, and rational individuals. Institutionalists do not confront this problem. From their perspective, the individual, rational actor is not regarded as a universal constant but as a social construct defined within a particular set of institutions (Thomas et al., 1987). New institutions are not created from scratch but are built upon older institutions and must replace or push back preexisting institutional forms. There is no problem of infinite regress here. The question of the first institution has no meaning within institutional theory, since neither society nor individual actors can exist without institutions. Instead of the theoretical problem of the first institution, institutional theory must handle the empirical problem of describing and analyzing the institutional structure within which action takes place.