Rediscovering Institutions

James G. March, Johan P. Olsen

• Publisher: Free Press
• Year: 2010
• ISBN:
  9781451602401
• Topic: Politics & International Relations
• Subtopic: Politics
• Index: Politics & International Relations

The authors propose a new theory of political behavior that re-invigorates the role of institutions—from laws and bureaucracy to rituals and symbols—as essential to understanding the modern political and economic systems that guide contemporary life.

1

Institutional Perspectives on Politics

In most contemporary theories of politics, traditional political institutions, such as the legislature, the legal system, and the state, as well as traditional economic institutions, such as the firm, have receded in importance from the position they held in earlier theories of political scientists such as J. W. Burgess or W. W. Willoughby, economists such as Thorstein Veblen or John R. Commons, and sociologists such as Max Weber. From a behavioral point of view, formally organized political institutions have come to be portrayed simply as arenas within which political behavior, driven by more fundamental factors, occurs. From a normative point of view, ideas that embedded morality in institutions, such as law or bureaucracy, and that emphasized citizenship and community as foundations for political identity, have given way to ideas of moral individualism and an emphasis on bargaining among conflicting interests.

In recent years, however, institutional perspectives have reappeared in political science. They reflect an empirically based prejudice, an assertion that what we observe in the world is inconsistent with the ways in which contemporary theories ask us to think, that the organization of political life makes a difference. This resurgence of concern with institutions is a cumulative consequence of the modern transformation of social institutions and persistent commentary from their observers. Social, political, and economic institutions have become larger, considerably more complex and resourceful, and prima facie more important to collective life. Many of the major actors in modern economic and political systems are formal organizations, and the institutions of law and bureaucracy occupy a dominant role in contemporary life.

Attention to political institutions has increased in the literature on politics, in particular in studies of legislatures (Shepsle and Weingast, 1983; 1987a; 1987b), local government (Kjellberg, 1975), public law (Smith, 1988), political economy (Kiser and Ostrom, 1982), political culture (Wildavsky, 1987), budgets (Padgett, 1981), public policy making (Ashford, 1978; Scharpf, 1977b), rational choice (Ferejohn, 1987), and political elites (Robins, 1976). It is manifest in studies of the origin of the state (Wright, 1977), the development of national administrative capacity (Skowronek, 1982), the nation-state’s relation to a world polity (Thomas et al., 1987), and the development of the welfare state (Ashford, 1986), as well as in analyses of the breakdown of democratic regimes (Potter, 1979), and in discussions of corporatism (Schmitter and Lehmbruch, 1979; Berger, 1981; Olsen, 1981). It is reflected in the Marxist rediscovery of the state as a problem in political economy (Jessop, 1977) and of the importance of organizational factors for understanding that role (Therborn, 1980). It is present in studies of formal organizations (Scott, 1987a; Zucker, 1987), including those of the place of such organizations in the implementation of public policy (Hanf and Scharpf, 1978). It is visible in attempts to link the study of the state to natural science (Masters, 1983) and to the humanities (Geertz, 1980), as well as in historical-comparative studies of the state (Hayward and Berki, 1979; Evans, Rueschemeyer, and Skocpol, 1985; Krasner 1984; 1988).

Renewed interest in institutions is not peculiar to political science. It is characteristic of recent trends in public law (Smith, 1988) and in economics, which has discovered law, contracts, hierarchies, professional codes, and social norms (Williamson, 1975; 1985; Furubotn and Richter, 1984; Akerlof, 1980). It is also seen in sociology (Meyer and Rowan, 1977; DiMaggio and Powell, 1983; Meyer and Scott, 1983; Thomas et al., 1987), although noninstitutionalist visions never succeeded in that field to the extent that they did in political science and economics. Cycles in ideas have brought us back to considerations that typified earlier forms of theory. The new and the old are not identical, however. It would probably be more accurate to describe recent thinking as blending elements of an old institutionalism into the noninstitutionalist styles of recent theories of politics.

STYLES OF CONTEMPORARY THEORIES OF POLITICS

Although the concept of institution never disappeared from theoretical political science, it has been largely supplanted in recent years by a conception of political life that is noninstitutional. The vision that has characterized theories of politics since about 1950 is (1) contextual, inclined to see politics as an integral part of society, less inclined to differentiate the polity from the rest of society; (2) reductionist, inclined to see political phenomena as the aggregate consequences of individual behavior, less inclined to ascribe the outcomes of politics to organizational structures and rules of appropriate behavior; (3) utilitarian, inclined to see action as stemming from calculated self-interest, less inclined to see action as a response to obligations and duties; (4) instrumentalist, inclined to define decision making and the allocation of resources as the central concerns of political life, less attentive to the ways in which political life is organized around the development of meaning through symbols, rituals, and ceremonies; and (5) functionalist, inclined to see history as an efficient mechanism for reaching uniquely appropriate equilibria, and less concerned with the possibilities for maladaptation and nonuniqueness in historical development.

Politics as Subordinate to Exogenous Forces: Contextualism

Historically, political scientists and political philosophers have tended to treat political institutions, particularly the state, as independent factors, important to the ordering and understanding of collective life (Heller, 1933). Modern political scientists, with few exceptions, do not. The state has lost its position of centrality in the discipline; interest in comprehensive forms of political organization has declined; political events are defined more as epiphenomena than as actions necessary to an understanding of society; politics mirrors its context (Easton, 1968).

The most conspicuously cited contextual factor cited is the social class structure (Dahrendorf, 1959). The social stratification of a modern society with its associated distribution of wealth and income has obvious major effects on political events. Class differences translate into political differences with great reliability across time and across cultures; differences in the organization and ideology of social class seem to lead to predictable differences in political organization and institutions (Tilly, 1978). Other analyses at the same level of aggregation make the structure and process of politics a function of physical environment, geography, and climate; of ethnicity, language, and culture; of economic conditions and development; or of demography, technology, ideology, or religion. Plausible arguments which make political life a derivative of one or more of these broad contextual forces have been developed, and it is not hard to find empirical data to support the arguments. Although there are a number of relatively precise contextual theories, the major theoretical significance of these ideas from the present point of view is less the specific forms of the theories than their general inclination to see the causal links between society and polity as running from the former to the latter, rather than the other way around. It is assumed that class, geography, climate, ethnicity, language, culture, economic conditions, demography, technology, ideology, and religion all affect politics but are not significantly affected by politics.

The Macro Consequences of Micro Behavior: Reductionism

Historically, political theory has treated political institutions as determining, ordering, or modifying individual motives, and as acting autonomously in terms of institutional interests. In contrast, substantial elements of modern theoretical work in political science assume that political phenomena are best understood as the aggregate consequences of behavior comprehensible at the individual or group level.

Such theories depend on two presumptions. The first presumption is that a polity consists of a number (often a large number) of elementary actors or events. Human behavior at the level of these elementary actors or events may be seen as conscious, calculated, and flexible, or as unconscious, habitual, and rigid. In either case, the preferences and powers of the actors are exogenous to the polity, depending on their positions in the social and economic system. The timing of events is exogenous, depending on external flows of problems and solutions. The second presumption is that collective behavior is best understood as stemming from the (possibly intricate) interweaving of behavior understandable at a lower level of aggregation. Discovering, or deducing, the collective consequences may be difficult, even impossible; but the central faith is that outcomes at the collective level depend only on the intricacies of the interactions among the individual actors or events, that concepts suggesting autonomous behavior at the aggregate level are certainly superfluous and probably deleterious.

Within such a perspective, for example, the behavior of an organization is the consequence of the interlocking choices by individuals and subunits, each acting in terms of expectations and preferences manifested at those levels (Niskanen, 1971). The behavior of a market is the consequence of the interlocking choices by individuals and firms, each acting in terms of a set of expectations and preferences manifested at those levels (Stigler, 1952). It is not necessary that the micro processes involve choice, of course. Aggregate behavior in a group can be defined as the consequence of the interlocking of trial-and-error learning occurring at the individual level (Lave and March, 1975). Or the aggregate behavior of an industry can be defined as the consequence of the interlocking of standard operating procedures and accounting rules followed at the level of the individual firm, and of death and growth rates of those firms (Nelson and Winter, 1982). Or aggregate behavior in a decision process can be seen as the merging of independent flows of problems, solutions, decision makers, and choice opportunities (Cohen, March, and Olsen, 1972).

As is clear from the last example, there is nothing intrinsic to a perspective that emphasizes the macro consequences of micro actions which requires that the elementary units be individuals. All that is required is that the behavior of a more comprehensive system be decomposable to elementary behaviors explicable at a less comprehensive level. In practice, however, in most of the social sciences, the actions of individual human beings are considered to determine the flow of events in a larger social system. Outcomes at the system level are thought to be determined by the interactions of individuals acting consistently in terms of the axioms of individual behavior, whatever they may be. Thus, we make assumptions about individual consumers to understand markets, about voters to understand politics, and about bureaucrats to understand bureaucracies.

Action as the Making of Calculated Decisions: Utilitarianism

Historically, political science has emphasized the ways in which political behavior was embedded in an institutional structure of rules, norms, expectations, and traditions that severely limited the free play of individual will and calculation (Wolin, 1960). In contrast, modern political science has, for the most part, described political events as the consequence of calculated decisions. Not just in political science, but throughout modern theoretical work in the social sciences, the preeminent vision of human behavior is a vision of choice. Life is characterized as deliberate decision making.

The details of the choice metaphor vary from one treatment to another, but the characteristic form is one that assumes choices stem from two guesses about the future. The first is a guess about the uncertain future consequences of possible current action. Decision theorists recognize that human limitations may restrict the precision of the estimates, that the estimates may be biased, and that the information on which the estimates are based may be costly; but information about probable consequences is assumed to be important to a choice (Savage, 1954; Lindley, 1973). From this assumption comes an emphasis on the power of information and expertise (Crozier, 1964) and the importance of reliable and unbiased information sources (Nisbet and Ross, 1980). Although numerous psychological experiments have indicated that the guesses of human subjects are biased (Kahneman, Slovic, and Tversky, 1982; Schoemaker, 1982), it has not been easy to formulate alternatives to the simple notion that the guesses of experienced humans are, on average, accurate. As a result, most theories of choice present decisions as being, on average, sensible. In their political versions, choice theories assume that, on average, voters vote intelligently with respect to their interests; legislators organize sensible coalitions, given their interests; and nation-states voluntarily enter alliances that, on average, improve their positions.

The second guess on which intentional, anticipatory choice is based is a guess about a decision maker’s uncertain future preferences for possible future outcomes. In any theory of deliberate choice, action depends on the decision maker’s values. Since the consequences of interest are to be realized in the future, it is necessary to anticipate not only what will happen but how the decision maker will feel about those outcomes when they are experienced (March, 1978). The complexities of the second guess are largely ignored by theories of choice. In their standard forms, the theories assume that preferences are stable, thus that current preferences are good predictors of future preferences; that preferences are unambiguous and consistent, thus that a choice will be clearly indicated, given the first guess; and that preferences are exogenous, thus that whatever process generates preferences it precedes choice and is independent of the choice process. In one of the best-developed forms of choice theories, these assumptions about preferences are taken as axioms, and preferences are discovered not by asking decision makers to report them but by defining a “revealed preference” function that satisfies the axioms and is consistent with choices made by a decision maker (Luce and Raiffa, 1957). Although the empirical existence of consistent revealed preferences has been the subject of considerable debate (Becker and Stigler, 1977; Kahneman, Slovic, and Tversky, 1982), the theoretical idea forms the basis of extensive analytical development and empirical exploration.

The Primacy of Outcomes: Instrumentalism

Historically, theories of political institutions portrayed political decision making primarily as a process for developing a sense of purpose, direction, identity, and belonging. Politics was a vehicle for educating citizens and improving cultural values. Although there are exceptions, the modern perspective in political science has generally given primacy to outcomes and either ignored symbolic actions or seen symbols as part of manipulative efforts to control outcomes, rather than the other way around.

Modern polities are as replete with symbols, ritual, ceremony, and myth as the societies more familiar to anthropological tradition. Politicians announce public support for positions they fail to defend in private (Edelman, 1964). Legislators vote for legislation while remaining indifferent to its implementation (Pressman and Wildavsky, 1973). Administrators solicit public participation in decision making in order to secure public support for policies to which they are already committed. Chief executives advocate reorganization of the public bureaucracy, announce plans for making reorganizations, and regularly abandon the plans (see Chapter 5). Information is gathered, policy alternatives are defined, and cost-benefit analyses are pursued, but they seem more intended to reassure observers of the appropriateness of actions being taken than to influence the actions (Feldman and March, 1981).

In modern discussions of politics, these symbolic actions are characteristically portrayed as strategic moves by self-conscious political actors. Rituals and ceremonies are defined as window dressing for the real political processes, or as instruments by which the clever and the powerful exploit the naive and the weak. The hiring of experts lends legitimacy to policies (Meyer and Rowan, 1977); associating unpopular moves with popular symbols is reassuring (Edelman, 1964). Control over symbols is a basis of power, like control over other resources (Pfeffer, 1981a); and the use of symbols is part of a struggle over political outcomes (Cohen, 1974).

The Efficiency of History: Functionalism

Historically, political theory has been ambivalent about the efficiency of history. Like other social scientists, students of political development have been inclined to accept an idea of progress, the more or less inexorable historical movement toward some more “advanced” level. At the same time, political histories have often emphasized the unique significance of a particular sequence of events or choices, the impact of a particular campaign strategy or speech, or the particular tactics of international negotiation. In modern usage, the terminology of progress has been largely replaced by a terminology of survival, but for the most part, in contemporary theoretical political science, institutions and behavior are thought to evolve through some form of efficient historical process.

An efficient historical process, in these terms, is one that moves rapidly to a unique solution, conditional on current environmental conditions, and is thus independent of the historical path. The equilibrium may involve a stochastically stable distribution or a fixed point, but we require a solution that is achieved relatively rapidly and is independent of the details of historical events leading to it. The assumption of historical efficiency is a standard, although usually not explicit, assumption of much of modern social science. Economic theories of markets and ecological theories of competition, for example, are concerned with the characteristics of an equilibrium, if one exists. They are used to predict differences (e.g., in markets, organizational structures, populations, technologies) that will be observed, at equilibrium, in different environments. It is no accident that the most common principle of theories in the social sciences is the optimization principle, and that one of the greatest concerns in such theories is showing that a process has a unique optimum that is guaranteed to be achieved.

Similarly, some postwar theories of political parties see party orientation and organization as equilibrium solutions to problems of survival in a competitive political environment (Downs, 1957; Ordeshook, 1986). The assumption of historical efficiency makes such theories largely indifferent to the behavioral reality of the micro processes that are assumed. For example, competition can be assumed to eliminate action that is inconsistent with the logic of survival. Examples include theories of market equilibria, such as those found in recent ideas of efficient capital markets (Sharpe, 1970); theories of organizational structure, such as those found in recent ideas of industrial organization (Williamson, 1975); and theories of political parties, such as those found in ideas of political economy (Olson, 1965).

DISCOVERING ORDER IN AN INSTITUTION-FREE WORLD

The remainder of this book considers how styles of theories of politics that focus on institutions differ from the styles described above. Before proceeding to that topic, however, we examine briefly two examples of theories that portray politics as contextual and reductionist: theories that focus on the competitive interaction of rational actors and theories that focus on the temporal sorting of problems, solutions, decision makers, and choices. Although the two examples differ in other respects, they are both initially reductionist and contextual in style. That is, they imagine that the key attributes of a macro system are understandable as the consequence of the confluence of a large number of simple micro events within an exogenously determined social context. These theories are powerful contributors to our understanding of politics and should be seen as necessary correlates of the perspective that is offered in this book.

Politics as Rational Competition

Classical theories of competitive conflict in economic affairs assume that markets (particularly labor, capital, and product markets) mediate among conflicting demands through prices. Within any particular economic institution, entrepreneurs are imagined to impose their goals on other participants in exchange for mutually satisfactory wages paid to workers, rent paid to capital, and product quality paid to consumers. The behavior of a market is assumed to be understandable as a consequence of these individual actors making choices that, in aggregate, fit together into market phenomena.

The proc...

Table of contents

• Cover
• Acknowledgments
• Chapter 1: Institutional Perspectives on Politics
• Chapter 2: Rules and the Institutionalization of Action
• Chapter 3: Interpretation and the Institutionalization of Meaning
• Chapter 4: Transforming Political Institutions
• Chapter 5: Institutional Reform as an Ad Hoc Activity
• Chapter 6: Institutional Reform as Public Policy
• Chapter 7: The Search for Appropriate Institutions
• Chapter 8: Preferences, Power, and Democratic Institutions
• Chapter 9: The Role of Political Institutions
• Notes
• References
• Index
• Copyright