Neopluralism - McFarland - 2007 PDF

1 Apr 2007 12:22 AR ANRV312-PL10-03.tex XMLPublishSM (2004/02/24) P1: KUV 10.1146...

1 Apr 2007 12:22 AR ANRV312-PL10-03.tex XMLPublishSM (2004/02/24) P1: KUV 10.1146/annurev.polisci.10.072005.152119 Annu. Rev. Polit. Sci. 2007. 10:45–66 doi: 10.1146/annurev.polisci.10.072005.152119 Copyright  c 2007 by Annual Reviews. All rights reserved First published online as a Review in Advance on November 16, 2006 NEOPLURALISM Andrew S. McFarland Political Science Department, University of Illinois at Chicago, Chicago, Illinois 60607; email: [email protected] Key Words pluralism, power, public policy, political process Annu. Rev. Polit. Sci. 2007.10:45-66. Downloaded from www.annualreviews.org Access provided by 114.108.219.115 on 07/26/22. For personal use only. Abstract Neopluralism is one of a class of research findings or social science models—such as elitism, pluralism, and corporatism—that refer to the structure of power and policy making in some domain of public policy. Originating from Robert Dahl’s pluralism model in Who Governs? (1961), neopluralism evolved in the study of American politics through discarding or modifying some of Dahl’s ideas, while adding new concerns about agenda building, the logic of collective action, special-interest subgovernments, social movements, advocacy coalitions, and the theory of political processes. Neopluralism is normally a finding of complex action in policy systems, but neopluralism does not assume that complexity implies fairness of representation, nor does it assume interest group elimination of autonomous action by governmental agencies. INTRODUCTION “Neopluralism” has several overlapping meanings. The term can refer to one of several models of power in a local, national, or policy system: elitism, pluralism, neopluralism, corporatism, statism, clientelism, consociationism. The term can focus on the pluralist theory of political power as set forth by Robert A. Dahl in Who Governs? and hence refer to revived formulations of this outlook, setting forth more sophisticated methods and conclusions. Neopluralism more specifically can refer to the observation that Dahl’s (1961) formulation of theoretical and methodological elements, known as pluralist theory, has been used in modified forms by researchers in the areas of public policy making, urban politics, and interest groups ever since the publication of Who Governs? in 1961. Accordingly, this essay begins with an analysis of Dahl’s pluralist theory. It then reviews the challenge to pluralist theory as set forth by Lowi (1964, 1969, 1979), Olson (1965), and others in the viewpoint variously described as theories of subgovernments, clientelism, and interest group liberalism, although I use the term multiple elitism. Next, the essay proceeds to the challenge posed to multiple elitism termed neopluralism, which only partially resembles Dahl’s pluralism. Then I examine characteristics of neopluralism that make it something more than just another model in the line-up of models of power. A major conclusion is that 1094-2939/07/0615-0045$20.00 45  1 Apr 2007 12:22 AR ANRV312-PL10-03.tex XMLPublishSM (2004/02/24) P1: KUV 46 MCFARLAND this sequence of scholarship proceeding from Dahl provides the basis for a general theory of the political process. To prevent confusion, one must first point out that there are no fewer than five usages of the term pluralism in the literatures of the social sciences and philosophy. As used herein, pluralism refers to the most common usage by political scientists since 1961, that is, referring to the theory of political power set forth by Dahl in Who Governs?. Dahl’s usage was emulated by numerous political scientists until this pluralist theory came under serious criticism at the end of the 1960s. The usage referring to Dahl’s theory differs from reference to the group theories of Truman (1951) and other political scientists in the 1950s, who stressed the pri- macy of interest groups in determining policy outcomes in American politics. A Annu. Rev. Polit. Sci. 2007.10:45-66. Downloaded from www.annualreviews.org common error among scholars not very familiar with this literature is to assume Access provided by 114.108.219.115 on 07/26/22. For personal use only. that Dahl was also a pluralist in the group theory sense, but in truth, he was not (see below). Truman’s group theory was dubbed pluralism because it was an empirical theory that stressed the role of groups, as had the political theory of the British philosopher Harold Laski a generation earlier (Laski 1921). Laski downgraded the normative role of state sovereignty in comparison to the importance of social groupings, and his theory was called pluralism (although Laski himself later be- came a leading Marxist). Sociologists also use the term pluralism when referring to social diversity, such as in reference to pluralist societies, perhaps containing many ethnic and religious groups. Sociologists sometimes use the term in referring to social theory that stresses the significance of social groups, as did Tocqueville. Finally, during the past 20 years, “pluralism” has come to mean political theo- ries that emphasize the value of promoting distinctive gender and ethnic identity, implying that public policies should promote such pluralism as social diversity (Connolly 2005). Having pointed out these sources of confusion, I reiterate that pluralism and neopluralism in this essay refer to the pluralist theory of political power set forth by Dahl and to revisions of Dahl’s theory by a generation of research successors. DAHL’S PLURALISM: FOUR BASIC CONCEPTS Dahl’s pluralism is posited on four basic concepts, which have formed a template for much political science research since the publication of Who Governs?. They have served as assumptions for many of the subsequent critics of Dahl’s theory, such as Lowi; and they are still used by neopluralists and others as a basis for a political process paradigm. Power as Causation A definition of power was a key element in Dahl’s construction of pluralist theory. With the definition of power as causation, Dahl brought together two strands of academic discussion during the 1950s. The Power Elite by Mills (1956)  1 Apr 2007 12:22 AR ANRV312-PL10-03.tex XMLPublishSM (2004/02/24) P1: KUV NEOPLURALISM 47 argued that America is ruled by three connected elites: the executives of the largest corporations, those at the top of the executive branch of the U.S. government, and leading officers of the U.S. military. This book immediately was assigned in many undergraduate classes and was widely read by intellectuals outside of academia. A few years earlier, social science leaders Harold Lasswell (Lasswell & Kaplan 1950) and Herbert Simon (1953) had put forth a definition of power intended to serve as a basis for a more analytical political science discipline. Dahl picked up the Lasswell-Simon definition of power as causation and made it a basis of his Who Governs? study, the pluralist alternative to Mills’ power-elite study. The definition of power as causation states that A has power over B to the Annu. Rev. Polit. Sci. 2007.10:45-66. Downloaded from www.annualreviews.org extent that A causes changes in B’s behavior in the direction of A’s intentions. For Access provided by 114.108.219.115 on 07/26/22. For personal use only. instance, when Dahl studied the urban renewal decision-making process he found that the New Haven mayor tended to realize his intentions by getting other actors to do what he wanted. Dahl (1961) defines this as political power. Power here is seen as causing changes in the behavior of others. Clearly the role of intentions is a sticky issue. Actor A might cause changes in the behavior of others simply by acting, but those changes might not be the ones A favors. In such cases, most would speak of A exercising “influence” by causing changes in behavior, even if A opposed those changes. But it seems, following Max Weber, that it is best to use “power” to refer to changes intended by the actor, and to reserve “influence” to refer to any behavioral change. Lasswell, Simon, March, and Dahl introduced the causal definition of power to American politics researchers without at first making a distinction between power and influence. Dahl and others later came to see the distinction as necessary (Weber 1946; Lasswell & Kaplan 1950; Simon 1953; March 1955; Dahl 1957, 1961, 1968; McFarland 2004). The definition of power as causation is certainly not universal. Structural theo- rists tend to use “power” to refer to the control of resources, such as money and property (Polsby 1980), or, in the realm of international relations, military and economic capability (Morgenthau 1964). However, pluralists refer to these items as “resources” to be used in the pursuit of power; for example, military capability might be a resource used by one nation-state to threaten an opposing nation-state and make it comply with the policy of the first. The Political Process Model A second basic concept in Dahl’s pluralism is the political process model. A re- searcher may describe a political or policy-making process or activity and then refer to the actors within that process as having power, in terms of whether the ac- tors caused others to change their behavior. Dahl thus described political processes in Who Governs?. In 1961, concepts of political process were seen as empirical and innovative, as part of “the behavioral revolution” of empiricism within political science. Studies of political process relied on empirical observation of activities over time, in contrast to the preceding modes of inquiry based on textual analysis  1 Apr 2007 12:22 AR ANRV312-PL10-03.tex XMLPublishSM (2004/02/24) P1: KUV 48 MCFARLAND of law and constitutional traditions, description of the legal forms of state institu- tions, and philosophical inquiry into the nature of justice and other social values. Accordingly, two leading American exponents of empiricism in political science, Arthur F. Bentley and David B. Truman, entitled their major works The Process of Government and The Governmental Process, respectively (Bentley 1967 , Truman 1951). During the 1950s, Truman’s book was often regarded by schol- ars as the leading theoretical work on American politics, and political scientists of that era often used the term “process” in book titles and course descriptions. Dahl adopted the basic Bentley-Truman model of the political process but did not adopt the more extreme assertions of this model, in which process was claimed to be everything, and in which institutions, laws, and political culture were viewed Annu. Rev. Polit. Sci. 2007.10:45-66. Downloaded from www.annualreviews.org as artifacts of the political process. Dahl described political campaigning, public Access provided by 114.108.219.115 on 07/26/22. For personal use only. policy making, bargaining among elites, and interest group behavior as political processes, but he viewed political institutions and political culture as more fixed and as responding to social changes such as economic and technical changes, cul- tural factors, and demography. Moreover, even within such areas as public policy making and elite bargaining, Dahl rejected the extreme Bentley-Truman viewpoint that political groups could be assumed to be the most important variable in caus- ing outcomes, as, for instance, individual voting decisions might play an important role in elections. Nevertheless, the political process model is a basic concept of Dahl’s pluralism. Following Bentley and Truman, this model is characterized by a combination of the following elements: 1. Empirical observation indicates agents acting within a policy system. 2. These agents are seen as groups and individuals representing group interests. 3. The agents interact and affect one another’s behavior. 4. The agents pursue their interests, defined according to the agents’ own def- initions of interest, although sometimes these must be inferred from their behavior. 5. Interests frequently change in the process of interaction among the agents over time. 6. Implicit in the foregoing, empirical observation should continue over a period sufficient to understand fluctuations in power, interest groups, and policy- making activities. As Polsby (2006) notes, pluralist research usually tries to trace events back in time to decide who caused what to happen in the search for power. The political process model is normally complex; it normally has many units, many interrela- tionships that shift over time, and interests (goals) that are frequently redefined by the interacting agents. In some cases, however, there might be only a few agents, as for instance, it might be observed that one group consistently defeats a second group or just a few groups. The existence of a power elite can be empirically observed, if one group consistently defeats other groups.  1 Apr 2007 12:22 AR ANRV312-PL10-03.tex XMLPublishSM (2004/02/24) P1: KUV NEOPLURALISM 49 Separate Domains of Political Processes The third element of Dahl’s pluralism is that power as causation and the political process model are applied to specific domains of politics and policy making. This is linked to the definition of power as causation and the empirical assumption that power might differ in different domains or fields of activity. Alternatively, of course, power might be the same in major domains, corresponding to the empirical finding that a power elite exists in a given system. In Dahl’s pluralism, the political system is seen as divided into numerous separate policy areas (there may not be so many at the local level). Dahl assumed that the structure of power and the nature of the process might vary in different policy areas. One has to observe the various areas to find out. Of course, previous research may lead one to expect a particular Annu. Rev. Polit. Sci. 2007.10:45-66. Downloaded from www.annualreviews.org pattern of power and process, but someone sometime had to do an empirical study Access provided by 114.108.219.115 on 07/26/22. For personal use only. to establish such a conclusion. This third assumption represents a commitment to one view of empiricism, emphasizing the need to observe (or at least study) actual political processes, and a reluctance to accept quick generalizations about power and policy making. It is a rejection of research postulating that the accumulation of wealth (Marx) or the holding of top hierarchical positions (Michels 1959) is all one needs to know about to describe the holding of political power. Reliance on Actors’ Definition of Their Own Interests Dahl’s fourth assumption is that researchers should accept the definitions of interest given by the subjects themselves, or perhaps as inferred from observations of subjects’ patterns of behavior. Clearly, this assumption rules out notions of class in the Marxist sense (Balbus 1971). But this definition of interest as understood by the subjects themselves also rules out concepts of quiescence—the idea that some subjects should make an issue out of something but are not observed doing so. A writer such as John Gaventa might argue that poor coal miners should be protesting working and housing conditions, even if they are not actually doing so, and that such quiescence is the key issue in studying power. The pluralists of the 1960s called this the problem of “nonissues” and maintained that specifying a list of nonissues embodied value assumptions that could lead to biased research (Bachrach & Baratz 1962, Lukes 1974, Gaventa 1980, Polsby 1980). WHO GOVERNS? THE PLURALIST THEORY OF POWER From a perspective of 45 years, the statement of these interlinked four concepts is a contribution to political science as important as Dahl’s refutation of Mills’ power- elite theory in Who Governs?. The criticism of Mills and the statement of pluralist political power are of course based on the use of these four related concepts. These concepts were retained as a theoretical template by many of the critics of Dahl’s pluralism and have continued to be assumed by the later writers termed neoplu- ralists. Later in this essay I mention misinterpretations of Dahl’s statement of  1 Apr 2007 12:22 AR ANRV312-PL10-03.tex XMLPublishSM (2004/02/24) P1: KUV 50 MCFARLAND political pluralism and discuss at greater length the better-substantiated criticisms of this pluralist theory. In Who Governs? Dahl assumed a political process model for his three case studies of the urban renewal, public education, and primary elections domains for the exercise of political power. Dahl traced the recent histories of these three issue-area domains and identified individuals wielding power by gaining their way against opposition, thereby changing the course of events within the case-study issue domain. He took self-identified interests and goals for granted as the subject of research, and accepted the views of political activists that these were likely the three most important issues. As is widely remembered, Dahl found that different persons held power in the different issue domains, with the exception of the mayor, who was Annu. Rev. Polit. Sci. 2007.10:45-66. Downloaded from www.annualreviews.org powerful in all three of them. There was not a single Millsian power elite. On the Access provided by 114.108.219.115 on 07/26/22. For personal use only. other hand, the urban political system was not authoritarian, because the mayor was checked by voters, who could eliminate his power in competitive primary and general elections. Accordingly, the mayor was accountable to the voters and was checked by the need to anticipate the opinions of the voters. One could then conclude that power in New Haven is not held by a power elite but is “pluralist.” In other words, power is fragmented, although sometimes partially concentrated in the hands of the mayor, who is, however, held accountable through competitive elections. Thus Dahl established this pluralist research procedure, implying the impor- tance of doing case studies of the political process. Political issues are identified as those important to the people being studied. Power as causation implies the necessity of observing a flow of events or reconstructing a history of past events. Power is seen as divisible by issue area and so cannot be generalized from one issue area without empirical confirmation. The techniques for case studies of politics were straightforward. Essentially, the scholar assembled a history of a sequence of political events. To do a history, one interviewed participants, collected documents issued by political participants, read newspapers and official records, and, if possible, directly observed political meetings and other events that were part of the history. The concern about causation did not rule out the communication of the ambience of political events; good writing and readability were valued. Pluralists sometimes conducted relatively simple surveys about political par- ticipation and elections in local communities. Following Dahl, pluralists tended to pay special attention to individual politicians, their motivations, and their political strategies. Presidents, mayors, members of the U.S. Congress, state legislators, and bureaucrats were studied as political entrepreneurs assembling political resources to gain power, often to please constituents and thus to keep their jobs. City and town political organizations were another favorite topic of study; these were re- garded as organizational entrepreneurs assembling resources to gain support and to win elections, sometimes thereby representing the working classes and assim- ilating immigrants into the American system (Wildavsky 1964a; Polsby 1971, 1980).  1 Apr 2007 12:22 AR ANRV312-PL10-03.tex XMLPublishSM (2004/02/24) P1: KUV NEOPLURALISM 51 Such methods and concepts could be applied to most sectors of political life in America. There were pluralist studies of cities, local communities, and different issue areas both nationally and locally (Polsby 1980). Pluralist studies of the pres- ident and of mayors sometimes found them to be powerful, yet contained within a bureaucratic system of politics (Allison 1971). Political party organizations were studied in pluralistic terms (Greenstein 1963). Congress and other legislatures were studied in terms of centralization and decentralization of power (Polsby 1971). In studying federalism, pluralists assumed a political process divided into issue ar- eas and conducted an empirical study of the relative power (causation) of central government versus state and local governments (Grodzins 1960). Pluralists stud- ied attempts to centralize power in administrative management systems: planning, Annu. Rev. Polit. Sci. 2007.10:45-66. Downloaded from www.annualreviews.org budgeting, management by objective, government reorganization, and so forth Access provided by 114.108.219.115 on 07/26/22. For personal use only. (Braybrooke & Lindblom 1963; Wildavsky 1964b, 1979; Lindblom 1965). This pluralist theory of power did achieve some apparently permanent success. For the next 30 years, political scientists eschewed elitist or power-elite explana- tions of political processes. Power-elite theories of community power have been supplanted. Those interested in social elites started in the 1970s to analyze elites in terms of network theory, based on aggregating the frequency of communication among individuals or units, and then depicting such communication networks as a mathematical theory of social structure (Laumann & Knoke 1987). MULTIPLE-ELITE THEORY Although Dahl’s pluralist theory had several major problems, most have been dealt with in subsequent neopluralist and political process research (see below). One problem in Dahl’s work provided an opening for what I term “multiple-elite the- ory.” Dahl found fragmented power in the three issue areas of Who Governs? except that the mayor exercised some power in all three issue areas. In turn, the mayor was controlled by voters in competitive elections. But could there not be policy systems in which a few persons or groups exercised power without any higher democratic authority to check them? Could not a particular group or coalition rule a specific issue area to the exclusion of higher democratic authority, such as chief executives, legislatures, or even higher courts? Moreover, a political system might have com- petitive elections for its chief executive and for its legislature, but such officials, selected by the entire citizenry, might lack control over some particular policy areas. This was the argument made by multiple-elite theorists. Such writers sided with Dahl against Mills; they seldom observed single power elites in the United States. However, they differed from Dahl in arguing the general significance of multiple independent elites, each particular to specific policy areas, within the fragmentation of power found by Dahl. Multiple-elite theorists normally argued that this fragmentation of power led to unrepresentative government of special interests, controlling policy in a particular area to the detriment of diffused public interests.  1 Apr 2007 12:22 AR ANRV312-PL10-03.tex XMLPublishSM (2004/02/24) P1: KUV 52 MCFARLAND Without using the terminology of “special interests,” Dahl’s colleague at Yale, Herbert Kaufman, joined Wallace Sayre of Columbia, along with research assistant Theodore Lowi, to study policy making in New York City. These authors observed islands of power within the city’s policy processes, meaning that the Board of Education and the Port Authority had great power within their respective issue areas and that the mayor lacked control over these semiautonomous bodies. Sayre & Kaufman (1960) and Lowi (1964b), however, used the same research procedures as Dahl. In a classic article, Lowi (1964a) argued that pluralism-elitism varied with the type of policy process in a discrete area. Pluralism might indeed be descriptive, Lowi argued, in policy areas focused on regulation, such as when the federal Annu. Rev. Polit. Sci. 2007.10:45-66. Downloaded from www.annualreviews.org government regulates working conditions, producing an interest group politics of Access provided by 114.108.219.115 on 07/26/22. For personal use only. countervailing business and labor groups. But, Lowi argued, other policy areas might ordinarily be controlled by a narrow coalition seeking to direct the distribu- tion of discrete benefits by government, such as the purchase of expensive military aircraft. Multiple-elite theorists emphasized the finding that oligarchical coalitions tend to control a particular area of policy making. For instance, one might say that a coalition of trucking companies, the teamsters’ union, and sympathetic Interstate Commerce Commission personnel controlled the regulation of interstate trucking. This is an example of what might be termed an “island of oligarchy.” Rule by the few exists, but it is limited to a single policy area. Such multiple-elite writers found islands of oligarchy in many areas: federal public works construction (pork barrel); the Port Authority in New York City; the antimonopoly policy of the Federal Trade Commission; and the regulation of airlines, trucking, local and long-distance telephone carriers, and railroads before their defeat by trucking interests (Maass 1951, Bernstein 1955, Sayre & Kaufman 1960, Redford 1969, Stigler 1975). Defense contracting was another area in which collaboration among defense industries, the Pentagon, and members of Congress with relevant constituencies seemed to be common. Agriculture was seen as rife with islands of oligarchy: the Farm Bureau (the center of many little govern- ments), tobacco production, sugar production, orange production in the Sunkist cooperative, peanut production, the distribution of cheap grazing rights on fed- eral lands, the distribution of subsidies for conservation projects such as farm ponds, the regulation of water sales from federal irrigation projects, and the dis- tribution of benefits of the multipurpose Tennessee Valley Authority’s economic development program (Selznick 1953; Foss 1960; Cater 1964; McConnell 1966; Lowi 1969, 1979). The regulation of the insurance industry by state governments was seen as another example (Orren 1974). Such research posits an alternative to Dahl’s theory of political power. However, multiple-elite theory research ex- plicitly or implicitly was based on the template of four assumptions set forth by Dahl: the need for empirical study of political process, power as causing change in the process, separability of policy domains, and interests as defined by the actors themselves.  1 Apr 2007 12:22 AR ANRV312-PL10-03.tex XMLPublishSM (2004/02/24) P1: KUV NEOPLURALISM 53 Besides islands of oligarchy, another major new idea put forth by multiple- elite theory was Olson’s (1965) logic of collective action. Olson’s was the best single theory to explain the existence of islands of oligarchy. Olson based collective action theory on the sort of economic theory just considered. Oligopolies produce a collective benefit for constituent units (although not for the public), as a few units cooperate to restrict production and raise prices. But as the number of producer units increases, it becomes more difficult to collude to restrict production, because some firms will leave the production restriction to others and produce more to obtain higher profits for themselves in the short run. Similarly, Olson reasoned, a few units will contribute to support a political lobby, but as the number of units increases, they will all try to be free riders, and the lobby will get no contributions. Annu. Rev. Polit. Sci. 2007.10:45-66. Downloaded from www.annualreviews.org Olson concluded that a lobby with a large number of units could exist only if an Access provided by 114.108.219.115 on 07/26/22. For personal use only. interest group offered benefits that were restricted to its members— for example, the discounts available exclusively to members of the American Association of Retired Persons—and then put some of the group’s resources to work in lobbying. The logic of collective action therefore predicted that lobbies for consumers, victims of pollution, and other constituencies of perhaps a hundred or more units would not exist unless they provided members with selective benefits worth more than the cost of joining. Collective action theory predicted, in contrast, that small groups of producers would collude to form lobbies to work in their interests, and therefore “the few would defeat the many,” in Olson’s terminology. The few would defeat the many by controlling particular policy areas of regu- lation, for instance, through lobbying and forming a dominant coalition, whereas the more numerous constituents on the public side would be unable to organize. Such an argument would apply to much government regulation of business, such as the regulation of prices, pollution, service, and safety. It could also be applied to agricultural policy, especially insofar as local farm elites captured policy ad- ministration. It could be applied to public works construction, as beneficiaries of construction form public coalitions to maintain spending. The logic of collective action was a persuasive argument on behalf of multiple-elite theory. Olson later published an application of his logic of collective action to explain declines in economic growth in societies other than the United States, as well as declines of growth in some areas within the United States, by positing that institu- tionalization of special interest political power creates oligopolistic inefficiencies (Olson 1982). Because it was tied to explaining economic growth, Olson’s general work had little influence on discussions about neopluralist political power. Along with Olson’s The Logic of Collective Action, Lowi’s The End of Lib- eralism [1979 (1969)] has been the most widely read and influential work of multiple-elite theory. Lowi did not refer to Olson’s theory but alluded to about 30 case studies of political power in specific issue domains, apparently agreeing to the four assumptions used by Dahl and the pluralists in the study of political process. But Lowi’s findings were quite different from those of the pluralists. Lowi drew a general interpretation of American politics based on multiple- elite theory. He argued that the American political process had become suffused  1 Apr 2007 12:22 AR ANRV312-PL10-03.tex XMLPublishSM (2004/02/24) P1: KUV 54 MCFARLAND with the public philosophy of interest group liberalism, not in the sense of the Democratic Party, but in the sense of political philosophy. In its academic usage, “liberal” refers to scholars who define democracy according to some process of decision making, rather than referring to standards of justice in the substance of government action. Lowi argued that interest group liberalism in the process sense had come to permeate the values of legislative, executive, and judicial decision makers in American government, so that laws were written and interpreted without useful reference to clear standards of justice and administration. He argued that, as a consequence, lower-level executive decision makers interpreted the practical meaning of legislation after a process of bargaining with organized interest groups, thereby forming a special interest policy-making coalition specific to a particular Annu. Rev. Polit. Sci. 2007.10:45-66. Downloaded from www.annualreviews.org area of public policy. Consequently, American public policy making is rife with Access provided by 114.108.219.115 on 07/26/22. For personal use only. particularistic policy coalitions, forming multiple elites within the overall system. To reform interest group liberalism, Lowi advocated institutional changes to pro- mote the statement of clear standards in legislation and administration of public policy, such as judicial prohibition of legislation lacking such clear standards. NEOPLURALISM VERSUS MULTIPLE-ELITE THEORY The multiple-elite researchers observed elite coalitions controlling specific issue domains without checks from executive or legislative authority, thereby contra- dicting Dahl’s pluralist theory. Analogously, neopluralist researchers observed that such subgovernments of multiple-elite research were not the most frequent pattern of power within issue domains, more frequently characterized by multiple opposed interest groups. Moreover, the neopluralists derived theories to deal with the implication of Olson’s logic of collective action that few groups representing diffused interests would organize and maintain themselves. Neopluralists were careful to note that their observation of a proliferation of groups did not imply a fair and representative policy-making process in some issue domain. Neopluralist observations were largely based on individual studies of policy making in par- ticular issue areas, supplemented by theories of why collective action is possible despite free riders in respect to public goods provided by lobbies. Such studies tended to appear between 1975 and 1985, in response to the research sequence set forth by Dahl’s political pluralism and its critique by Lowi and Olson. For example, in the issue domain of air pollution, studies by Jones (1975), Ornstein & Elder (1978), Ackerman & Hassler (1981), and Wenner (1982) all found that environmentalist groups have significant effects on public policy. Other au- thors (Berry 1977, McFarland 1984, Mitchell 1984, Nadel 1971, Vogel 1980/1981) chronicled the significant strength of public interest lobbies. Still others (Free- man 1975, Costain & Costain 1983, Klein 1984, Gelb & Palley 1982) found women’s lobbies to be important. McCarthy & Zald (1977) generalized from such observations to develop the resource mobilization theory of social move- ments, which implies that movements frequently create lobbying organizations.  1 Apr 2007 12:22 AR ANRV312-PL10-03.tex XMLPublishSM (2004/02/24) P1: KUV NEOPLURALISM 55 Beer (1976), Haider (1974), and others described the phenomenon of governments lobbying other governments—associations of cities, counties, and so forth lobby- ing in Washington and elsewhere. Walker (1983) and his collaborators (Gais et al. 1984) found a great increase in lobbying by nonprofit organizations (e.g., churches, hospitals, and colleges); by professions, especially groups of government employ- ees; and by ideological cause groups. Schlozman & Tierney (1986, pp. 55–57) found a great increase in government lobbies since 1960, although they argued that this did not substantially decrease the business domination of other lobbies. Neopluralists thus observed a complex welter of group participation in pub- lic policy making (Jordan 1981, Jordan & Maloney 1996, Gray & Lowery 2004, Lowery & Brasher 2004). Within a single issue domain such as air pollution regu- Annu. Rev. Polit. Sci. 2007.10:45-66. Downloaded from www.annualreviews.org lation, one might find producers of electricity, coal mining companies, auto man- Access provided by 114.108.219.115 on 07/26/22. For personal use only. ufacturers, the United Auto Workers, railroads (carriers of coal), environmentalist groups, the American Lung Association, the American Public Health Association, various state governments, the association of state public utility regulators, and probably other groups. Neopluralists would expect to observe this situation in nu- merous issue areas at the national level, rather than to observe a narrow coalition of an interest group, government agency, and legislative committee. Concomitant with this finding of a complexity of groups active in national policy domains was the need of explaining why such groups mobilized and maintained themselves in light of Olson’s persuasive theory of collective action, which stated that lobbies make public goods (public policies) available even to groups and individuals who do not contribute to the lobby and thereby act as free riders. As mentioned above, Olson posited that groups would organize if they provided selective benefits to contributors, benefits available only to members, such as the Farm Bureau providing insurance and fuel discounts to its farmer members. But the concept of selective benefits did not provide a persuasive explanation for the existence of the full panoply of interest groups. Why did so many interest groups exist despite the logic of collective action? Political scientists made this question a priority, particularly during 1970–1990. A number of answers received widespread acceptance and were incorporated into neopluralist views of the policy-making process. First, a few years after the publication of The Logic of Collective Action, Chamberlin (1974) and Frohlich et al. (1971) pointed out that interest groups can be started and maintained by political entrepreneurs, to use economic lan- guage. One or more individuals take it upon themselves to start a group and to bear the time costs of organizing and doing the necessary fund-raising. Walker’s (1991) surveys found that about half of the interest groups responding had been organized by such political entrepreneurs. Walker then extended this mode of thought by demonstrating the key role of patrons in organizing interest groups. Patrons were particularly helpful in under- standing the organizing of citizen groups, in which members contributed to lobby- ing for public policy issues even when such policies did not have a great immediate impact on the members’ own pocketbooks. Patrons are seen as providing money  1 Apr 2007 12:22 AR ANRV312-PL10-03.tex XMLPublishSM (2004/02/24) P1: KUV 56 MCFARLAND and other resources to the entrepreneurs who do the organizing. Patrons include wealthy individuals, foundations, other organizations (creating spin-off groups, for instance), and governments, particularly the federal government. Walker em- phasized the government as a patron, noting that the federal government actually played a key role in organizing several major groups, such as the Farm Bureau, the U.S. Chambers of Commerce, and the National Organization of Women (Walker 1983, 1991; McFarland 2004, pp. 44–45). A third source of interest groups is social movements. Although they use non- institutionalized tactics, social movements eventually encompass supporters who prefer to work within established political institutions to attain movement goals, thus forming lobbies for large but seemingly diffused groups, such as women, Annu. Rev. Polit. Sci. 2007.10:45-66. Downloaded from www.annualreviews.org environmentalists, and religious fundamentalists. Social movements are a major Access provided by 114.108.219.115 on 07/26/22. For personal use only. reason for the sudden expansion of organized lobbies occurring in Washington, DC around 1970. Although efforts have been made to explain social movements using a rational choice framework similar to that of Olson’s collective action theory (Chong 1991, Lichbach 1995), most writers emphasize a variety of causes for so- cial movements (McAdam 1982, Costain & McFarland 1998), thereby employing variables outside of Olson’s theory. A fourth theory addressing the proliferation of interest groups was provided by policy theorist Hugh Heclo (1978), whose concept of “issue networks” found immediate, widespread acceptance among political scientists. Heclo observed that although the particularistic coalitions described by multiple-elite writers do indeed exist, they are only part of a complex network of communication and interaction within some issue domain. Heclo termed such an overall network an “issue net- work.” He argued that the subgovernments or little governments of multiple-elite theory existed in only a minority of cases, and that ordinarily such limited coali- tions, only part of an issue network, are checked by other actors within the issue network (sometimes called a “policy network”). In addition, from the standpoint of collective action theory, the issue network provides a basis for mobilizing lob- bies to represent interests more diffuse than the oligopolistic coalitions implied by Olson’s theory. For instance, in the case of air pollution policy, there is a potential subgov- ernment consisting of eastern coal mining businesses; eastern miners; legislators representing Kentucky, Ohio, Illinois, and West Virginia; and the Department of Energy. Such a subgovernment would favor the use of high-sulfur coal, which would require the purchase of expensive emissions-scrubbing equipment or regu- latory variation to meet the requirements of air pollution legislation. But in addition to the electric utilities directly affected, there exists a network of communication of political, economic, and technical information within this issue area. The federal Environmental Protection Agency and state EPAs regularly communicate about air pollution issues, as do university researchers, environmental group personnel, recreation industry personnel, American Lung Association staff, journalists, fed- eral and state legislators specializing in environmental legislation, and the staffs of such legislators. Because of the overall issue network, the potential high-sulfur coal  1 Apr 2007 12:22 AR ANRV312-PL10-03.tex XMLPublishSM (2004/02/24) P1: KUV NEOPLURALISM 57 subgovernment cannot control information and public attention regarding policy in this area. A countervailing lobby representing widespread but diffuse interests can be organized from other elements in the issue network. A potential issue-area elite is checked by forces within the more general issue network. Heclo’s viewpoint has been adopted by neopluralists and probably by most political scientists. A fifth reason for the proliferation of interest groups follows from an expansion of Olson’s concept of selective benefits, meaning that one can organize a large group on the basis of “members-only” benefits. Professional and occupational groups normally organize conventions or other members-only meetings, in which the selective benefit is valuable information about professional opportunities or the presentation of significant recent research (Moe 1980). Annu. Rev. Polit. Sci. 2007.10:45-66. Downloaded from www.annualreviews.org Access provided by 114.108.219.115 on 07/26/22. For personal use only. COUNTERVAILING POWER AND SELECTIVE ATTENTION Neopluralists thus observe a world of political domains that normally contain com- peting interest groups. One is immediately reminded of the “balance of power” concept in the history of international relations. The analogous concept in domestic politics is Galbraith’s (1952) “countervailing power.” Balance of power or coun- tervailing power implies initially observing a dominant power that subsequently is checked by one or several balancers of power. This is a primitive but useful concept for observing public policy making. In the study of American policy making, the next step is to observe how the countervailing power groups communicate policy information to various political entities in the process of calling public attention to the observed policy domain. In other words, how do the countervailing power groups exercise power? They have the potential to influence the executive, legislative, judicial, and federal decision- making institutions in American government. For example, in the 1970s, when environmentalists opposed the coalition supporting nuclear power generation, they initiated lawsuits to point out the economic and safety problems of nuclear power to the public, and directly and indirectly to regulatory commissions and to legislators not on the Joint Committee on Atomic Energy. Thus, the politics of countervailing power is not usually one of group gridlock, but rather of groups maneuvering to call attention to their policy preferences as they attempt to influence decision makers in different political venues (Baumgartner & Jones 1993, Jones 1994, Jones & Baumgartner 2005). This is a politics of influencing the selective attention of decision makers, who cannot attend to the scores of potential issues that might come before them. Two models of policy making follow from the observation of countervailing group power. The first is something of a paradox: The existence of countervailing groups prevents the capture of a government agency by a dominant group, a basic part of the multiple-elitist model of the controlling coalition of interest group, leg- islative committee, and government agency. In other words, countervailing power groups enhance the autonomy of government agencies, which, for instance, are  1 Apr 2007 12:22 AR ANRV312-PL10-03.tex XMLPublishSM (2004/02/24) P1: KUV 58 MCFARLAND more likely to design and execute policy reflective of professional norms, such as those of lawyer, economist, educator, or biologist (Wilson 1980; McFarland 2004, p. 48). More groups lead to more autonomy for government agencies (controlled in varying degrees by executive or legislative officials). Countervailing group power is related to the theory of advocacy coalitions, as set forth by Sabatier & Jenkins-Smith (1993). In the history of international relations, one refers to bipolar systems; in American domestic politics, one fre- quently observes competing advocacy coalitions, such as competition between environmentalists and development interests in Western communities. An advo- cacy coalition is a stable alliance, probably lasting ten years or more, among actors within a policy network, including individuals representing government agencies Annu. Rev. Polit. Sci. 2007.10:45-66. Downloaded from www.annualreviews.org at all levels of federalism. The concept of advocacy coalition within a policy net- Access provided by 114.108.219.115 on 07/26/22. For personal use only. work, often in competition with each other, has proved useful in scores of studies of policy making (Sabatier & Jenkins-Smith 1999). CHARACTERISTICS OF NEOPLURALIST POLITICAL PROCESS The basic meaning of neopluralism is a finding about power structure within the framework of four assumptions underlying Dahl’s theory of political power (pro- cess, power as causation, domains, subjective definition of interests). Using these and perhaps other assumptions, political science has set forth seven findings about power structure: power elite theory, Dahl’s pluralist theory, multiple-elite theory, neopluralism, corporatism (Schmitter 1974), statism (Krasner 1978), and conso- ciationism (Lijphart 1969). Neopluralism was derived from a research sequence that started with Mills versus Dahl and proceeded through multiple-elite theory, finally resulting in neopluralism. Explicitly or implicitly, these seven research findings can be based on the use of Dahl’s concepts and methods. Dahl (1958) criticized Mills for not using these methods, but a scholar can readily find a single elite ruling a single issue domain, a community, or even a nation by using Dahl’s methods and arguing that the same group controls the core of public policy. The multiple-elite theorists did this in respect to single policy domains. Neopluralism is thus a type of finding in research about political power. But it is more than this; neopluralism can be called a theory, because researchers assuming neopluralist power domains have begun to make other theoretical observations about political and policy-making processes. As of 2006, according to my reading of the literature, most political scientists and other public policy researchers assume a condition of neopluralism in the policy domains they study. Dahl’s pluralist theory and multiple-elite theory are normally seen to be dated as general theories, although they might provide accu- rate explanations for outcomes in a minority of policy domains. Disproving the existence of a power elite or a state of Dahl’s pluralism within a policy domain  1 Apr 2007 12:22 AR ANRV312-PL10-03.tex XMLPublishSM (2004/02/24) P1: KUV NEOPLURALISM 59 is no longer a priority in research studies, although in urban studies, the question of political power co