When Protest Makes Policy

Front matter

: When Protest

Makes Policy

HOW SOCIAL MOVEMENTS REPRESENT DISADVANTAGED GROUPS

Index

Acknowledgments

A book-length project necessarily incurs many debts, so I have many people to thank. First, I thank my colleagues at Purdue University who helped me repeatedly and in many different contexts with this book. Leigh Raymond read most of the chapters, and his enthusiasm for this project shored me up when my confidence lagged. I have also been grateful for his appreciation for a certain local Indian restaurant. Jay McCann is astoundingly generous with his time, excellent advice, and fresh coffee. Suzie Parker made important suggestions about how to think about the Democratic Party in the Southern states in the United States and provided invaluable advice about survey design and analysis. In addition to teaching me to really appreciate a Margarita, Rosalee Clawson read several of these chapters and taught me much about surveys and intersectionality. Bill Shaffer is a great friend and colleague, reading and critiquing just about every chapter in this book at some point and offering suggestions, advice, and beer. Bert Rockman has been a mentor of mine since graduate school, taking time to provide me advice on questions big and small, including many of the chapters and ideas presented in this book. His influence in setting me on the institutional road is no doubt substantial. Bob Bartlett also read every chapter here and provided sage advice about the proposal. Pat Boling gave me expert feedback on many of these chapters as only a fellow scholar of feminist theory and public policy could. Harry Targ provided invaluable expertise on labor organizing and politics. Aaron Hoffman argued with me about nearly everything in this book and nearly drove me crazy by finding things I could improve when I thought I was finished. This made the book much better. Aaron also helped shape the idea for the book and supported my nascent vision of the book when I was not sure it would be of interest. Other Purdue colleagues helped me solve a problem with a particular part of the analysis (Eric Waltenburg and Roger Larocca) or provided important feedback on drafts of chapters and the proposal (Ann Clark, Niambi Carter, and Daniel Aldrich).

I am also lucky to have an extended network of brilliant and creative colleagues in the broader discipline who are unfailingly generous with their time.

X ACKNOWLEDGMENTS

Jane Mansbridge provided suggestions on many parts of the project, helping with important refinements in the argument. Karen Beckwith generously provided feedback at key moments. Amy Mazur also talked with me at length about many of these ideas, providing encouragement and feedback. The wideranging discussions I have had with Mala Htun have helped me better to articulate many of the ideas in these pages. David Meyer is a reliable source of feedback and enthusiastic support for a more junior colleague; his collegiality is admirable. Melissa Williams has been an inspiration and support for more than a decade.

For helpful discussion, assistance with data or technical matters, or incisive questions about particular issues or chapters, I thank David Alexander, Olga Avdeyeva, Debbie Bahu, Maryann Barakso, Sylvia Bashevkin, Anya Bernstein, Lisa Brush, Lisa Baldez, Amy Caiazza, Suzi Dovi, William Dunn, Rachel Finwohner, Amy Elman, Brian Fuerst, Carol Gould, Bernie Grofman, Ange-Marie Hancock, Susan Hansen, Don Kerchis, Cheryl Kerchis, William Jacoby, Judson Jeffries, Jyl Josephson, Katsuo Nishikawa, Pam Paxton, Andrew Rehfeld, Sue Tolleson Rinehart, Kira Sanbonmatsu, Oxana Shevel, Evelyn Simien, Dorothy Stetson, Dara Strolovitch, Terri Towner, and Larry Weldon. I thank Pamela Lubin and Manjusha Gupte for research assistance. I am also very glad that Maura Bahu and Patrick Jessee, who were undergraduates at Purdue, each worked with me on a chapter of the book. They made the work much more fun than it would otherwise have been, and I learned a great deal from both of them. I thank Lori Norris, Brianne Belliveau, Jackie Strange, and Vicki McKinley for helping with household matters and child care so that I could devote my attention more fully to these pages.

This work has been presented at various stages to the Midwest Political Science Association and the American Political Science Association and at Purdue University (especially at the Public Policy and Political Theory Workshop); the University of Pittsburgh; the University of Toronto (Canada); the University of Victoria (Canada); the University of California, Irvine; and Georgetown University. I thank the people who participated in those sessions for their helpful comments and questions. Two of these chapters were published in journals (the Journal of Politics and Political Research Quarterly), and the process of review improved those chapters and helped me to think about this project. I thank the Institute for Women’s Policy Research (especially Amy Caiazza) and the Center for American Women and Politics for making so much valuable data available.

ACKNOWLEDGMENTS xi

I thank Melody Herr at University of Michigan Press for her support for this work and for her attention to detail. Anonymous reviewers at the University of Michigan Press, as well as Susan Carroll and Kira Sanbonmatsu, series editors, provided important suggestions and comments on the manuscript.

Having received all this help and support from such a wide range of people, it would be shocking if errors remained in the text. Indeed, at this point, I can hardly be blamed for such errors, and I suggest that any problems be laid squarely at the feet of one of the aforementioned colleagues.

We all know books require emotional as well as intellectual resources, and in my case, my friends and family so often shored me up, made me laugh, or helped out with my kids that I feel lucky to have such a big group to thank. Laura Barker and Cathy Flynn picked me up from a distance when I needed a hand. I thank Leigh, Teresa, Ximena, and Chris for being our Lafayette “family” and making our life here fuller of love, laughter, and care. Thanks also go to my exceptionally wonderful family: to my mother for being an inspiration and unflagging support, not to mention a source of great jokes; to my father for encouraging me and standing behind me every step of the way (and for helping out with free statistical advice sometimes); to my brother (Kaljo) and sisters (Christine, Allison, and Katie) for making my family so much richer and for being a good uncle or auntie when their sister needed a little help with child care. My mother’s partner, Don Walker, has always been there for me; my father’s wife, Jill Weldon, has also been a great support. Great big thanks go to my family-in-law: no one could ask for more love and support than I have been lucky enough to enjoy from the Hoffmans.

I am sustained by my loving husband and children, and I dedicate this book to them. I thank my amazing children, Zed and Audrey, for being themselves. Words cannot express my debt, intellectual and personal, to my wonderful partner, Aaron Hoffman, but I am foolish enough to try in the brief space available here. He is the best partner anyone could possibly imagine, and he has given generously of his time to offer feedback on the book. He is a wonderful father to our children. He cooks delicious meals and picks out fabulous wines. He makes me laugh and makes my life fun.

Finally, as I was writing this book, my friend and mentor (and former dissertation advisor) Iris Young passed away. I am sorry she will not get to see this book. I hope this book contributes, in its own small way, to furthering her legacy of pressing for both social-structural analysis and attention to the ways power and difference are entwined. I also dedicate this book to her memory.

Introduction: Movements, Marginalization, and Representation

In the aftermath of the 1999 “Battle in Seattle,” when more than 50,000 people gathered in Seattle to protest globalization and the World Trade Organization (WTO), protesters claimed success in representing the marginalized and disenfranchised of the world. Indeed, activists of all stripes see themselves as “representatives,” giving voice to perspectives that would otherwise be excluded.’ By providing better representation of historically disadvantaged groups, social movements further inclusion, produce greater trust in government, and reduce social conflict (Mansbridge 1999; Williams 1998; Wolbrecht and Hero 2005; Costain 2005). But social movements are frequently criticized for overrepresenting the privileged and lacking democratic accountability.” Moreover, critics allege that movements deepen exclusion and marginalization when protest tactics disrupt deliberation, prevent compromise, and emphasize differences over commonalities, fragmenting the polity into separate “identities” or “special interests.” Given these concerns, what are we to make of social movement claims to represent or “speak for” the excluded? Do social movements offer a pathway to a more inclusive democracy or toward a more polarized, fragmented, elite-dominated polity?

Examining a variety of public policies, I show in this book that social movements are important avenues of political representation, especially for excluded and disadvantaged groups in established democracies. Social movements, I find, constitute a critical avenue of policy influence for women, workers, and women of color, an avenue often more important than political parties, interest groups, or electing group members to government office (descriptive representation). Indeed, movements are so important that we ought to further such mobilization wherever we can. I propose the ideal of the advocacy state as a guide to deepening inclusive democracy. Such a state is aimed at furthering the mobilization of disadvantaged groups both within the formal institutions of government and in civil society.

Social movements enable representation and inclusion by facilitating the articulation of group consciousness, by organizing and mobilizing the very groups to be represented. Disadvantaged groups define themselves and identify their priorities through such movements.* Without social movements, there are no constituencies or group perspectives to be represented. Furthermore, through a variety of activities and agents (movement organizations, intellectuals, publications, theater, protests, etc.), movements articulate and diffuse to the broader citizenry alternative, otherwise marginalized perspectives on political issues, enriching deliberation about policy issues (Mansbridge and Flaster 2007; Fraser 1995; Young 2000).

Mobilization of socially and economically disadvantaged groups is more easily undertaken in the informal, fluid world of social movements, because formal institutions tend to disempower and exclude these groups. The “entry costs” are lower for social movement mobilization than for other types of political mobilization. Moreover, social movements can be very influential. In fact, the research presented in this book shows that social movements can be more effective avenues of policy influence than electing larger numbers of women or minorities to political office, voting, traditional lobby groups, or political parties. They also enhance the effectiveness of these other mechanisms of representation.

Empirical and theoretical studies of democratic representation, even those focused on representation for disadvantaged groups, have mostly overlooked this important contribution to democracy. The literature on representation for marginalized groups has traditionally focused on the value of descriptive representation in the legislature.° This is an important area of research, to be sure, but an overemphasis on descriptive representation in the legislature leaves other important avenues for representation unexamined, ignores the way different avenues of representation interact, and results in a tendency to overlook the institutional context (Weldon 2002; Childs 2006; Mansbridge 2003; Poggione 2004; Wolbrecht and Hero 2005). Indeed, an emerging literature is revealing the complexities involved in processes of representation. For example, social movements and government agencies can be mechanisms of representation for marginalized groups, and there are politically important links between feminist organizations and legislators. This is not to suggest that the literature on descriptive representation in the legislature claims that electing women or people of color is the only way to obtain policy change (no scholar says such a thing); rather, the point is that descriptive representation in other places and other mechanisms of substantive representation have been understudied in relation to their importance in securing policy change of significance for marginalized groups.

The main empirical literature focusing on the political significance of social movements (political sociology) tends to focus on questions relating to the causes of social mobilization, their success or failure in maintaining mobilization, or perhaps their ability to obtain particular policy goals; democracy tends not to be a “dependent variable” for sociologists.’ As a result, few scholars of social movements have assessed these movements in terms of their potential contribution to democracy, especially for marginalized groups. Democratic theorists, meanwhile, have mostly overlooked or devalued the representative role of social movements. Although many political theorists assign an important role to social movements, they mostly see movements as providing an avenue for or model of participatory democracy.’ Some theorists see the very existence of social movements as providing a critique of representative democracy (Kitschelt 1993).

There is a growing recognition of the important connections between participation and representation (Young 2000). In this instance, thinking of social movements as offering avenues of representation as well as participation suggests new ways to value these organizations and new ways to promote democracy. Instead of criticizing social movement organizations for focusing on representation rather than participation (e.g., Acker 1995; Putnam 2000), scholars should focus on which types of movement organizations or institutions provide which democratic functions (Edwards et al. 2001). Indeed, moving beyond the traditional pluralist focus on representation through a system of interest groups (Dahl 1962; Truman 1951; Walker 1991), contemporary theorists of representation have come to emphasize civil society actors such as voluntary associations as a critical avenue of democratic representation (Cohen and Rogers 1992; Warren 2001). Drawing on this work, I propose that the democratic ideal requires an “advocacy state” that both includes agencies designed to promote marginalized groups (e.g., women’s policy machineries) and also encourages autonomous organizing by these groups in civil society. As a way of showing the practicability of the ideal, I offer concrete examples of real policies that encourage autonomous organizing by marginalized groups. These arguments point to new ways to improve and deepen democracy and further inclusion of the most marginalized segments of society.

The Plan of This Book

In the rest of this introduction, I outline the conceptual bases for arguing that social movements provide democratic representation, despite the absence (much of the time) of formal mechanisms of authorization and accountability that many see as the sine qua non of democratic representation. I use this argument to outline the advocacy state as an ideal: this is the idea that democratic states ought to foster the self-organization and mobilization of disadvantaged groups within and outside of state structures. In the chapters of this book, I provide further elaboration of and empirical support for these points, using a variety of primary sources (interviews, surveys) and secondary ones (existing data sets, newspaper articles, other published works).? Drawing on a series of analyses of women’s movements and labor movements both in the United States and in established democracies more generally, I aim to show that social movements provide a vital form of democratic political representation, through both conventional and newer avenues of political expression. I supplement the cross-sectional, statistical analyses of multiple countries or states at a single point in time with case evidence from Canada, the United States, Norway, or other established democracies, to illustrate the dynamic relation between protest and policy over time.'®

Social movements organize political mobilization through traditional lobby or interest groups, voter registration and education initiatives, and leadership institutes. They also work through the everyday politics of language change, institutional reforms, consumer boycotts, street theater, cultural criticism, and grassroots efforts to promote broader social change (Mansbridge 1995; Mansbridge and Flaster 2007; Rochon 1998; Beckwith 2007; Katzenstein 1998). Through these efforts, social movements create the constituencies and corresponding group perspectives that can be represented, a critical (if often overlooked) aspect of the process of representation (Williams 1998; Young 2000). More generally, social movements create and revitalize democratic civil society and fundamentally shape the architecture and operations of political institutions. These ongoing interactions between the state and civil society can have lasting impact far beyond the particular movement in question (Meyer 2003; Meyer et al. 2005; Costain and McFarland 1998).

At the same time, however, analysts have raised a number of concerns about social movements that bear directly on this argument about social movements as democratic representatives. The first, most practical concern is whether social movements actually have any lasting influence. Some posit that resilient social structures such as patriarchy, capitalism, white privilege, or other structures that work through the coercive power of the state obstruct efforts to transform policy outcomes (Bell 1987; MacKinnon 1989; Abramovitz 1992). Others argue that direct influence through formal electoral avenues of representation is the most direct, effective way to influence policy, contending that it is elected officials who set the agenda (Haider-Markel et al. 2000; Kingdon 1984; Spill et al. 2001). In chapters 1, 2, and 3, I seek to establish that for women and for workers, social movements can be more effective avenues of influence than traditional legislative representation and can prompt policy changes that matter a great deal to the everyday lives of these groups. These chapters examine movements and policies cross-nationally and in the United States.

Another concern focuses on the internal politics of social movements, pointing to evidence that movements best represent privileged groups and do a poorer job of “speaking for” the marginalized. I address this question in chapters 4 and 5, pointing out that while they are not perfect, social movements are still the best avenues of representation for disadvantaged groups, especially for groups, such as women of color, who tend to “fall through the cracks” of more generally focused organizations and institutions. Indeed, I show that women’s movements are critical for policy developments that promote the distinctive interests of women of color. Women’s movements do a better job of representing all women, though, when disadvantaged subgroups of women can organize themselves separately.

In chapter 5, I aim to provide evidence about the tripartite relationship between (1) the constituency “women” and various subgroups of women (women of color, working-class women, professional women, lesbians); (2) women’s organizations; and (3) the public sphere (media and policy-making). The chapter presents a more detailed look at local women’s organizations and their access to the media in the U.S. city of Chicago. I find that women’s organizations, taken together, do a better job of representing the diversity of women—both descriptively and substantively—than do formally representative organizations such as legislatures and political parties. This suggests that social movements taken as a whole (but not necessarily any particular activists or organizations) are more representative of the diversity of social groups and attend more to diverse interests than previous research suggests. In addition, different sorts of organizations perform different democratic functions.

In chapter 6, I draw on the preceding analyses to suggest some ways to improve representation by social movements and to create a more inclusive democracy. I argue that states should actively encourage and foster the development of social movements and provide opportunities for activist participation in policy-making processes. Given the structured inequalities present in all contemporary and foreseeable democracies, the ideal democratic state would have to be an advocacy state, one that fostered independent movements especially of the most marginalized.

There are at least two reasons that scholars might be skeptical that such an advocacy state would work. First, some argue that inclusion of social movements in the state too often results in a depletion of civil society and a weakening of democracy (Dryzek et al. 2003). However, I contend that the risk of co-optation and depoliticization occurs not when social movements participate in policy-making processes (through public consultation, lobbying, bargaining, etc.) but, rather, when social movements are entirely absorbed by the state. When all the major social movement organizations and institutions are inside the state and there are few autonomous organizations, activists risk losing the leverage they gain from a broader base of public support. But it is possible for state policy to encourage strong, autonomous social movements without compromising movement autonomy, as I show in this introduction and in chapter 6. Moreover, if a dual strategy of organizing inside and outside the state is pursued, the dangers of depletion are far smaller, and the potential payoff in terms of the representation of otherwise excluded groups is tremendous."

A second possible objection to an advocacy state focuses on whether it is possible, in practice, to distinguish systematically disadvantaged social groups from those that are not so disadvantaged. Although many scholars are skeptical that government policies can effectively discern disadvantaged groups, scholarly discussions tend to ignore the many examples of such policies that work reasonably well in practice. I offer the Court Challenges Program of Canada and the Disadvantaged Business Enterprise Program of the U.S. Department of Transportation as examples of programs that effectively targeted disadvantaged groups and facilitated their political mobilization.

Let us turn, now, to considering the conceptual bases for arguing that social movements can provide a kind of democratic representation, despite the absence (much of the time) of formal mechanisms of authorization and accountability.

Social Movements, Marginalization, and Representation

Substantive representation is partly a process of agenda setting, or prioritizing problems (Jones and Baumgartner 2005). Asking whether governments reflect the priorities of citizens is one way of getting at this sort of representation. Of course, the sets of priorities of citizens—indeed, the view of the appropriate role of the state—vary systematically by group. For example, a large body of scholarship shows systematic differences between women and men in public opinion and voting behavior both in the United States and cross-nationally.

Numerous studies establish the relationship between race and ethnicity and public opinion, both in the United States and cross-nationally. It is well established that class affects political attitudes, and in some contexts, religion is also an important factor. Political attitudes vary systematically across subgroups as well (i.e., Latinos v. Latinas, African American men v. white men, Latinas v. white women, etc.).!* The distinctive attitudes and priorities associated with each social group can be thought of as a sort of group perspective (I describe this idea more fully in chapter 1). To assess how well particular groups are represented, we can ask whether their attitudes and views are articulated as part of the political process and whether these priorities are reflected in policy agendas and outcomes.

Comparing policy issue agendas to group agendas or perspectives to examine the degree of correspondence or disjuncture can be done at either an individual or collective level, by asking whether the priorities of this individual or legislative body or organization reflect the priorities of this particular group. So we can think of representation as a systemic or macropolitical process, asking how well this particular political system represents citizens. Such an approach, however, has rarely been undertaken at the macro or systemic level (Baumgartner and Jones 2005; Costain 1994).

This analytic approach is especially valuable for studying the representation of systematically disadvantaged groups, sometimes called “marginalized.” Although there are many differences across marginalized groups defined by race/ethnicity, gender, class, sexuality, and the like, they share the characteristic of being disadvantaged in multiple arenas (social, economic, political), of having an experience or group history of formal exclusion from public life (and sometimes economic and social exclusions as well), and of being defined by characteristics that are relatively immutable (Williams 1998). Note that pointing to the disadvantaged position of these groups need not imply that the form and degree of disadvantage are the same (Cohen 1999, chap. 2). The argument I advance here does suggest, however, that similar measures will improve representation for groups suffering a wide variety of forms of disadvantage (for a contrasting view, see Htun 2004).

If groups have distinct perspectives and concerns, substantive representation for those groups can be evaluated by examining the extent to which these views influence public discussion of policy issues and, ultimately, policy outcomes themselves (Jacobs and Shapiro 1994). To do so, however, we must be clear about what counts as a group perspective and how such a perspective might influence public debate and policy. Much social science research demonstrates considerable variation across and within the groups I discuss here, in terms of their priorities, concerns, and so on. Despite this heterogeneity, however, groups can be thought of as being characterized by distinctive points of view. One way to square this seeming contradiction is to note that group perspectives are characteristics of groups, not individuals. A group perspective may not find itself perfectly instantiated in any particular individual, although the vast majority in a group will share some set of overlapping concerns. Moreover, a group perspective emerges in interactive contexts, where members of the group discuss their status or fate with each other. This process brings to the foreground aspects of individual experience that are shared with other group members; it also highlights those aspects of experience that are different from other group members. In such contexts, group members discuss those elements of social or political life that loom large in their experience (police brutality, discrimination, etc.), although it is unlikely that they agree on the specific actions that individuals or collectivities ought to take to address these concerns. This set of issues can be thought of as being a group perspective even if the group is not unified in terms of ideology or interests in every case (I say more about this idea in chapter 1).

Avenues of Social Group Representation

How are group perspectives best articulated in established democracies? If we examine the question of representation at the macro level, we can examine a variety of mechanisms for the articulation of social group perspective, explore the relationships between them, and ask which ones (or which combinations) seem to have the most influence on deliberative processes and policy outcomes. Although descriptive representation is often valued for its symbolic effects (i.e., contributing to the social construction of groups as, e.g., powerful or powerless) and for legitimating government decisions, it is also thought to have substantive effects. Indeed, it is perhaps the single most examined avenue of substantive representation in the literature. Other possible avenues of influence for marginalized groups include political parties, interest groups, and social movements. In this section, I discuss the advantages and disadvantages of each avenue of representation and theorize the relationship between them. I contend both that social movement representation is the most effective avenue for marginalized groups and that it also improves representation by traditional, intralegislative means (descriptive representation in legislatures and political parties).

DESCRIPTIVE REPRESENTATION IN PUBLIC OFFICE

Much of the literature on the political representation of marginalized groups focuses on women and minorities in public office, asking whether or under what circumstances legislators or bureaucrats from these groups differ from those of other ascriptive groups in terms of their policy priorities, ideology, or style (Weldon 2002b). This literature importantly illustrates why it matters that women and minorities are present among the representatives of a diverse democratic public. The presence of members of such groups (1) increases trust in government, political mobilization, and participation by marginalized groups; (2) increases the likelihood that a more diverse set of interests and experiences will be articulated during policy deliberations; and (3) undermines the social construction of marginalized groups as incapable of leadership (Mansbridge 1999; Williams 1998; Swers 2002; Thomas 1994). While valuable, however, this literature is overly focused on only one dimension of the process of political representation: intralegislative representation as accomplished by individual legislators. But representation is a multidimensional process, and important aspects of the representative process are obscured by this focus on intralegislative representation. Institutions that mediate between social groups and formal political processes (parties, interest groups, social movements, etc.) are a critical aspect of this relationship. In terms of influence on policy outcomes (one of the most concrete manifestations of effective representation), extralegislative avenues of representation can be the most important ones.

POLITICAL PARTIES

Political parties are “mediating institutions” that potentially link marginalized groups to democratic political processes (Wolbrecht and Hero 2005). Some scholars have argued that party competition advances inclusion, as parties who fail to respond to excluded social groups will be at an electoral disadvantage (Key 1949; Dahl 1962). Multiparty systems better enable links with particular social groups, such as class and religious groups (Powell 1982; but see Chhibber and Nooruddin 2004). Further, research on women and politics suggests that left political parties are more likely to increase women’s descriptive representation and, at least sometimes, appear to be more favorable to advancing women’s substantive representation as well (Mazur 2002). However, it is clear that for some issues important to women (violence against women, reproductive rights), left or right party ideology is a poor guide to whether the party will support women’s interests (Weldon 2002; Elman 1996). Indeed, Frymer (2005) points out that party competition has not advanced representation for African Americans in the United States, and Sanbonmatsu (2002) shows that the importance and alignment of parties varies across issues. Leighley (2005) argues that since parties are strategic actors, they aim at mobilizing supporters and are not necessarily the equalizing, inclusion-promoting institutions they are sometimes thought to be. Leighley emphasizes the importance of studying the relationship between parties and other mobilizing groups (e.g., voluntary associations) as critical for sorting out the complex relationship between parties and inclusion.

Parties are most likely to advance substantive representation when they are pressured by organizations independent of the parties themselves. Strong links between political parties and organizations of marginalized groups are likely to advance substantive representation, but it is important that marginalized constituencies maintain an independent organizational base outside the political party (Weldon 2002; Wolbrecht 2000). This can be seen most clearly in the case of strong social movements of marginalized groups that are transformed into dedicated political parties, as in the Icelandic women’s party or in the case of parties that focus on particular ethnic or racial groups, like the Bloc Québécois in Canada. When such movements are completely subsumed in the party system, their mobilizing potential decreases, their links to marginalized constituencies weaken, and substantive representation is decreased. Extralegislative avenues offer the most mobilization potential and political leverage on existing political parties, regardless of the number of parties.