Gender, Politics and Institutions
9780230_245884_01_pre.indd i
10/5/2010 5:01:28 PM
Gender and Politics Series Series editors: Johanna Kantola, University of Helsinki, Finland and Judith Squires, University of Bristol, UK This timely new series publishes leading monographs and edited collections from scholars working in the disciplinary areas of politics, international relations and public policy with specific reference to questions of gender. The series showcases cutting-edge research in Gender and Politics, publishing topical and innovative approaches to gender politics. It will include exciting work from new authors and well-known academics and will also publish high-impact writings by practitioners working in issues relating to gender and politics. The series covers politics, international relations, and public policy, including gendered engagement with mainstream political science issues, such as political systems and policymaking, representation and participation, citizenship and identity, equality, and women’s movements; gender and international relations, including feminist approaches to international institutions, political economy, and global politics; and interdisciplinary and emergent areas of study, such as masculinities studies, gender and multiculturalism, and intersectionality. Potential contributors are encouraged to contact the series editors: Johanna Kantola, (
[email protected]) and Judith Squires (judith.squires@ bristol.ac.uk) Series Advisory Board: Louise Chappell, University of Sydney, Australia Joni Lovenduksi, Birkbeck College, University of London, UK Amy Mazur, Washington State University, USA Jacqui True, University of Auckland, New Zealand Mieke Verloo, Radboud University Nijmegen, the Netherlands Laurel Weldon, Purdue University, USA Titles include: Jonathan Dean RETHINKING CONTEMPORARY FEMINIST POLITICS Mona Lena Krook and Fiona Mackay (editors) GENDER, POLITICS AND INSTITUTIONS Towards a Feminist Institutionalism
Gender and Politics Series Series Standing Order ISBNs 978–0–230–23917–3 (hardback) and 978–0–230–23918–0 (paperback) You can receive future titles in this series as they are published by placing a standing order. Please contact your bookseller or, in case of difficulty, write to us at the address below with your name and address, the title of the series and one of the ISBNs quoted above. Customer Services Department, Macmillan Distribution Ltd, Houndmills, Basingstoke, Hampshire RG21 6XS, England
9780230_245884_01_pre.indd ii
10/5/2010 5:01:28 PM
Gender, Politics and Institutions Towards a Feminist Institutionalism Edited by
Mona Lena Krook Assistant Professor of Political Science and Women and Gender Studies, Department of Political Science, Washington University in St. Louis, USA and
Fiona Mackay Senior Lecturer in Politics and International Relations, School of Social and Political Science, Edinburgh University, UK
9780230_245884_01_pre.indd iii
10/5/2010 5:01:28 PM
Editorial matter, selection, introduction and conclusion © Mona Lena Krook and Fiona Mackay 2011 Foreword © Joni Lovenduski 2011 All remaining chapters © respective authors 2011 All rights reserved. No reproduction, copy or transmission of this publication may be made without written permission. No portion of this publication may be reproduced, copied or transmitted save with written permission or in accordance with the provisions of the Copyright, Designs and Patents Act 1988, or under the terms of any licence permitting limited copying issued by the Copyright Licensing Agency, Saffron House, 6-10 Kirby Street, London EC1N 8TS. Any person who does any unauthorized act in relation to this publication may be liable to criminal prosecution and civil claims for damages. The authors have asserted their rights to be identified as the authors of this work in accordance with the Copyright, Designs and Patents Act 1988. First published 2011 by PALGRAVE MACMILLAN Palgrave Macmillan in the UK is an imprint of Macmillan Publishers Limited, registered in England, company number 785998, of Houndmills, Basingstoke, Hampshire RG21 6XS. Palgrave Macmillan in the US is a division of St Martin’s Press LLC, 175 Fifth Avenue, New York, NY 10010. Palgrave Macmillan is the global academic imprint of the above companies and has companies and representatives throughout the world. Palgrave® and Macmillan® are registered trademarks in the United States, the United Kingdom, Europe and other countries. ISBN: 978–0–230–24588–4 hardback This book is printed on paper suitable for recycling and made from fully managed and sustained forest sources. Logging, pulping and manufacturing processes are expected to conform to the environmental regulations of the country of origin. A catalogue record for this book is available from the British Library. A catalog record for this book is available from the Library of Congress. 10 9 8 7 6 5 4 3 2 1 20 19 18 17 16 15 14 13 12 11 Printed and bound in Great Britain by CPI Antony Rowe, Chippenham and Eastbourne
9780230_245884_01_pre.indd iv
10/5/2010 5:01:28 PM
Contents Foreword Joni Lovenduski
vii
Acknowledgements
xii
Notes on Contributors
xiv
1
Introduction: Gender, Politics, and Institutions Mona Lena Krook and Fiona Mackay
1
2 Gender and Institutions of Political Recruitment: Candidate Selection in Post-Devolution Scotland Meryl Kenny
21
3 Discursive Strategies for Institutional Reform: Gender Quotas in Sweden and France Lenita Freidenvall and Mona Lena Krook
42
4 Gendered Institutions and Women’s Substantive Representation: Female Legislators in Argentina and Chile Susan Franceschet
58
5 Gendering the Institutional Reform of the Welfare State: Germany, the United Kingdom, and Switzerland Michelle Beyeler and Claire Annesley
79
6 Gender and Institutions of Multi-Level Governance: Child Care and Social Policy Debates in Canada Joan Grace
95
7 The Institutional Roots of Post-Communist Family Policy: Comparing the Czech and Slovak Republics Hana Hašková and Steven Saxonberg
112
8 Gendering Federalism: Institutions of Decentralization and Power-Sharing Jill Vickers
129
9 Gendered Institutionalist Analysis: Understanding Democratic Transitions Georgina Waylen
147
v
9780230_245884_01_pre.indd v
10/5/2010 5:01:28 PM
vi
Contents
10 Nested Newness and Institutional Innovation: Expanding Gender Justice in the International Criminal Court Louise Chappell
163
11 Conclusion: Towards a Feminist Institutionalism? Fiona Mackay
181
References
197
Index
225
9780230_245884_01_pre.indd vi
10/5/2010 5:01:28 PM
Foreword Does institutionalism need a concept of gender? And does feminism need institutionalism? Probably the answers to these questions will turn on what we think is good social science. Good feminist social science is simply good social science; it is no more or less than good practice. It should concomitantly be impossible to imagine a good social science that ignores gender. Yet this is precisely what most political science does, and the new institutionalism, despite its concern with power relations in institutions, is no exception. Arguably, good institutionalists should realize the importance of gender relations to the configuration of institutions. But they do not. They need to be reminded, and feminist institutionalism, exemplified by the essays in this groundbreaking volume, is the reminder. How does the incorporation of a feminist perspective change institutional approaches to the study of politics? To set the scene it is useful to ask what feminist scholars want from political science. Broadly there are two possibilities: the explanation of gendered outcomes, including the current order or state of play, and predictions of future developments in the gender regimes that are so identified. Both entail the specification of contexts, the collection and analysis of evidence, and the theorization of political relationships, but for feminists these are inevitably gendered in some way. They may be gender neutral or balanced in their configurations and effects. Gender may be defined simply as a scale of masculinity and femininity along which behaviour and attitudes may be ordered. This is evidenced by the presence of codes, norms, and behaviour that reflect accepted (but possibly changing) dimensions of masculinity and femininity. Gender is always present in social life. In short, institutions are gender regimes (Connell 1987, 1990) and feminists use institutionalist approaches to answer questions about power inequalities in public life. When feminists adopt institutionalist research strategies that include gender, they seek to illuminate and change the status of women. In this endeavour the crucial feminist contribution to institutionalism is the addition of concepts of gender. Feminism therefore genders institutionalism. Why do feminists need new institutionalism? Probably the answer to this question starts with the politics of social science. Steinmo reminds us that the important battles in the social sciences are struggles over vii
9780230_245884_01_pre.indd vii
10/5/2010 5:01:29 PM
viii
Foreword
the meaning of science defined in terms of its predictive capacity. But such a focus often distorts research and obscures the importance of explanation. In general institutionalists are more interested in explanation than in prediction (Steinmo 2008). I think the same is true of feminist institutionalists, and that preference is one of the bases of a scholarly affinity between the two approaches that I highlighted more than a decade ago (Lovenduski 1998). Both explanation and prediction require an ability to characterize the institutional context and environment in order to identify their most salient features. Feminist political scientists want to discover and explain gender effects in political life, a project that inevitably leads them to focus on how political institutions are formed and sustained and how gender is embedded in them. Many feminists are also interested in prediction but are unwilling to pay the high cost of overlooking the gendered dimensions of institutions – which is a risk when concentrating on prediction at the expense of explanation. Institutions are the rules that structure political and social life. They are configurations of ideas and interests which are expressed as the ‘formal rules, compliance procedures and standard operating practices that structure relationships between individuals in various units of the polity and the economy’ (Hall 1986: 19–20). Political institutions express and necessarily contain a normative element – the norms, principles, and ideas that hold a given institutional structure together and provide the ‘compass’ for the assessments of attempts at change. This order consists of collectively constructed values and principles that are protected and maintained by accepted rules of the game. The rules take precise meaning through the actions of the individual organizations (parliaments, executives, and political parties) that they constitute, and they also structure, largely but imperfectly, the interactions that take place between and within these organisations (March and Olsen 1989: 107). March and Olsen, the founders of modern or ‘new’ institutional theory, provide the central insight that history is encoded into institutions. The ‘new institutionalism’ therefore permits a focus on process and offers a conceptual toolkit that includes formal and informal institutions, critical junctures, path dependence, feedback mechanisms, logics of appropriateness, and, more recently, institutional convergence and layering. It invites consideration of the roles of ideas in determining the interests of actors operating in a specific institutional context. All social scientists make methodological choices that carry advantages and disadvantages. The advantages of new institutionalism are frequently set out as the expansion of definitions of institutions (usually
9780230_245884_01_pre.indd viii
10/5/2010 5:01:29 PM
Foreword ix
in relation to old institutionalism, in which institutions were narrowly defined in organizational terms) to include norms, values and ideas, incentive systems, and its ability to explain the persistence of social structures. The disadvantages turn around the problem of explaining change, a particular challenge for a feminist project to explain how changing gender relations might alter institutions. An equally important criticism is that institutionalists tend to underestimate agency because the repertoire of action is so constrained by the rules of the game that actors may be thought of as trapped by institutions. Critics regard institutionalism, and particularly historical institutionalism, as an approach that overprivileges stability, pointing to the frequency with which change is explained as an effect of an exogenous shock and the resulting crisis. This was once fair comment. Institutionalism has been a weak approach to explaining internally generated change because institutions are defined partly in terms of their capacity to inhibit change. But as Thelen points out, institutional stability may conceal considerable adaptation, rendering the differences between change and stability rather less stark. Moreover ‘sometimes power begets power and institutions reinforce and magnify the position of their creators; but sometimes institutions provide interesting and unintended opportunities for marginal groups to exercise leverage well beyond their apparently meagre power resources’ (Thelen 2003: 216). Feminist institutionalists recognize that political explanation is about ideas, interests, and institutions, which are intertwined. In common with their colleagues they want better scholarship and better explanation. The proponents of feminist and institutionalist political science share a desire to answer real-world questions. But the shared interest in how institutions work in general has not to date extended to a common interest in how that working is gendered. While feminist and institutionalists agree that such answers require empirical investigations of institutions and their effects on political decisions and outcomes, for feminists those institutions are gendered in various ways. For the most part new institutionalists do not use a gendered approach, and feminist studies of institutions so far available, if acknowledged at all, are rarely engaged by them. The exclusion of consideration of gender risks crucial elements of ideas, interests, rules and processes, portents and causes of change, and instances of agency being ignored, and hence impairs the institutionalist project. If a central question of institutionalism is what it means to maximize power in a given situation then the institutionalist researcher is required to provide a detailed and complete account of the aspects of the political environment that frame the choices and
9780230_245884_01_pre.indd ix
10/5/2010 5:01:29 PM
x Foreword
strategies of the political actors who operate within it. They must also have an understanding of the perceptions of political actors, of their interests and their (ideas about) environment, and have an understanding of both internally generated change and the agency of institutional actors. In making gender a central concern and conceptualizing institutions as, inter alia, gender regimes, feminism contributes to each part of the institutionalist project. Feminist institutionalists think that changing gender relations are an important source of both internally and externally generated change. They investigate these components of institutions for the ideas they express about women and men, femininity and masculinity. Mapping the formal architecture and informal networks, connections, conventions, rules, and norms of institutions is the first step in constructing an institutional analysis. The way the analyses then proceeds depends on which of at least four different strands of new institutionalism is adopted: historical, sociological, rational choice, or discursive (Mackay and Waylen 2009).1 Each suggests a slightly different map of gendered processes. There is considerable common ground among the different kinds of institutionalists. They all see institutions as rules that structure behaviour. Where they differ is about the nature of those beings whose behaviour is being structured (Steinmo 2008: 126). While all of the new institutionalisms have something to offer feminist scholarship, historical institutionalism is especially adaptable to the concerns of feminist political scientists who seek to explain gendered outcomes in different contexts. Historical institutionalists think that human beings are both norm-abiding rule followers and self-interested rational actors whose behaviour depends on which rules and which contexts obtain. These are all matters for empirical investigation. To gender historical institutionalism it is necessary first to identify and record differences in the effects of the rules and the nature of interests among and between women and men. Next the researcher must construct an account of processes through which these differences came to be present. Finally their impact on gender relations in the institution should be assessed. The feminist sociological institutionalist would identify the social norms and explicate their gendered effects, producing an account of the mutually constitutive character of the gender regime of the institution they are studying. For example they might explore how logics of appropriateness may support or undermine gender stereotypes of performance in a legislature. In a feminist rational choice institutionalist analysis, gender relations would form part of the account of the context in which individuals calculated costs and benefits of their actions. Differences in the
9780230_245884_01_pre.indd x
10/5/2010 5:01:29 PM
Foreword xi
decisions between and among women and men would be recorded and explained in terms of the differences in their calculations that might, in turn, depend on gendered labour markets. Finally feminist discursive institutionalists consider the gender ideologies found in the institutional discourses focusing on how ideas about women and men and masculinity and femininity are present in its rules. They echo the initial insight of Schattschneider (1960) that control of the definition of a problem determines its solution, recasting this idea in the language of discourse analysis. Hence conflict over meaning is termed an interpretive struggle. Notably, discursive institutionalism explains change in institutions as the result of changes in ideas; hence changes in ideas about gender relations are predicted to change institutions. Inevitably the foundations of feminist institutionalist analysis are fine-grained descriptions of gendered environments accompanied by explanations of how gender constrains or enhances agency and affects stability and change. This might take the form of consideration of the options not discussed, actors marginalized, or micro-political economies not identified because they are obscured by unexamined assumptions about gender relations. In short, feminist institutionalism both enhances analysis and makes for more effective explanation. Feminists bring to the study of institutions a specific lens that makes visible constitutive, gendered power relations and the processes that support and undermine them. In identifying changing gender relations as a potential cause of institutional change feminism increases the capacity of ‘new’ institutionalists to model causality. There is now a foundational body of feminist research that offers gendered institutional analysis (see Lovenduski 1998). Since 2006 the Feminist Institutionalist International Network (FIIN) has explored the interplay between feminist approaches to gendered institutions and new institutional theory, publishing work that synthesizes insights from each approach in order to address issues of gender, politics, power, and change (www.femfiin.com). This volume builds on those foundations, offering new research on gender in institutions that is theoretically grounded in terms of both feminism and of the new institutionalism. JONI LOVENDUSKI Notes I am grateful to Dermot Hodson and Alan Ware for comments on this essay. 1. See also the recent ‘Critical Perspectives on Feminist Institutionalism’ in Politics & Gender, 5: 2, 2009.
9780230_245884_01_pre.indd xi
10/5/2010 5:01:29 PM
Acknowledgements This volume is the product of many conversations, workshops and conference panels, and an ongoing seven-year dialogue about the promise of a feminist institutionalism. The idea was sparked at the European Consortium of Political Research (ECPR) Joint Sessions of Workshops in Edinburgh, Scotland, in 2003, by a workshop convened by Fiona Mackay and Petra Meier, and bore fruit with the establishment in 2006 of the Feminism and Institutionalism International Network (FIIN; for more details, see www.femfiin.com). Since the founding of the network, we have debated and shared findings at the Gender Research Network Launch Conference in Manchester, England, in June 2007; the European Consortium for Political Research General Conference in Pisa, Italy, in September 2007; the European Conference on Politics and Gender, organized by the Gender and Politics Standing Group of the European Consortium for Political Research, in Belfast, Northern Ireland, in January 2009; and the Midwest Political Science Association National Conference in Chicago, Illinois, in April 2009. In April 2008, we also organized a six-day workshop on ‘Gender, Politics, and Institutions: Towards a Feminist Institutionalism?’ for the ECPR Joint Sessions of Workshops in Rennes, France. The task of putting together this volume was made possible thanks to the encouragement and assistance of a number of important people. For financial assistance, we are grateful to the School of Social and Political Science at the University of Edinburgh and the Centre for Political Economy at Washington University in St. Louis for funding our attendance at several key workshops and panels. We also appreciate support from the University of Edinburgh Campaign Small Projects Scheme, and the School of Social and Political Science Strategic Research Fund for enabling us to set up the FIIN website, run by FIIN coordinator Meryl Kenny, which has facilitated communications among members of the network and beyond. All panel participants over the years, and especially those who participated in the ECPR workshop in Rennes, have contributed to the broader conversation that has led to the publication of this book. We thank our co-directors, Louise Chappell, Meryl Kenny, and Georgina Waylen for their support. We acknowledge our debt to Joni Lovenduski, whose work inspired us to embark on the project, and we thank her also for writing the foreword. We record our appreciation xii
9780230_245884_01_pre.indd xii
10/5/2010 5:01:29 PM
Acknowledgements xiii
of Alison Howson, Amber Stone-Galilee, and Liz Blackmore at Palgrave, as well as the Gender and Politics Series editors, Judith Squires and Johanna Kantola, for their enthusiastic support and patience. Finally, we very much appreciate the editorial assistance of Mindy Sher and Meryl Kenny in the very last stages of pulling this manuscript together. We hope that the final product reflects a shared collective enterprise to forge a synthesis between feminism and institutionalism.
9780230_245884_01_pre.indd xiii
10/5/2010 5:01:29 PM
Contributors Claire Annesley is Senior Lecturer in European Politics at the University of Manchester (United Kingdom). She has been a fellow of Georg-August-Universität-Göttingen and a visiting researcher at the Wissenschaftszentrum Berlin für Sozialforschung. Her research interests include gender and politics, comparative European politics, German politics, and welfare capitalism. She is the author of articles on the adult work model appearing in the Journal of European Social Policy and Parliamentary Affairs, as well as co-editor of Women and New Labour: Engendering Politics and Policy? (Policy Press 2007). Michelle Beyeler is Lecturer in the Institute of Political Science at the University of Bern (Switzerland), where she teaches research methods and comparative politics. Her research covers various topics such as care policies, monetary policies, comparative welfare states, and social movements. Her overall focus is on the mechanisms and conditions of political reforms and change. She has published articles in Global Social Policy and Mobilization and is the author of The Paths to Price Stability (Haupt 2007) and the co-editor of The OECD and European Welfare States (Elgar 2004, with Klaus Armingeon). Louise Chappell is Professor and Australian Research Council Future Fellow in the School of Social Sciences and International Studies at the University of New South Wales (Australia). In 2008 she was a visiting fellow in Political Science at Leiden University (the Netherlands). Her book Gendering Government (UBC Press 2002) was awarded the American Political Science Association’s Victoria Schuck Award in 2003 for the best book published in the field of women and politics. Her recent publications include The Politics of Women’s Interests (Routledge 2006, co-edited with Lisa Hill) and The Politics of Human Rights in Australia (Cambridge University Press 2006, with John Chesterman and Lisa Hill). Susan Franceschet is Associate Professor in the Department of Political Science at the University of Calgary (Canada). She specializes in gender and politics in Latin America and is the author of Women and Politics in Chile (Lynne Rienner 2005). She is currently engaged in a project funded by the Social Sciences and Humanities Research Council xiv
9780230_245884_01_pre.indd xiv
10/5/2010 5:01:29 PM
Contributors xv
entitled ‘Gendered Mandates: The Impact of Women in Politics in Latin America’, exploring the impact of female legislators, ‘femocrats’, and women’s movement organizations on policy outcomes that advance women’s rights in Argentina and Chile. Lenita Freidenvall is a researcher in the Department of Political Science at Stockholm University (Sweden). She has published widely on women and politics, gender quotas, and candidate selection. Recent publications include Electoral Gender Quota Systems and Their Implementation in Europe: Report to the European Parliament (2008, with Drude Dahlerup and in cooperation with International IDEA) and Kvotering (2008, with Drude Dahlerup). She has also written reports on gender and constitutional reform on behalf of the Swedish National Working Commission on Constitutional Reform. Joan Grace is Associate Professor in the Department of Politics at the University of Winnipeg (Canada). Her research interests include women’s relationship to the state, women’s collective action, and feminist critiques of institutionalism. She was awarded a grant from the Social Sciences and Humanities Research Council for her project ‘Coming to Grips with the State: Women’s Political Activism in Western Canada’, which analyses women’s policy advocacy in Manitoba, Saskatchewan, and Alberta. Recent publications have appeared in Canadian Woman Studies and Canadian Public Administration. Hana Hašková is Researcher in the Gender and Sociology Department in the Institute of Sociology of the Academy of Sciences (Czech Republic). Her research focuses on gender, reproduction, intimate lives, and family policies, as well as women’s civic organizing in post-socialist Europe. She coordinated the Czech research team for the ‘Enlargement, Gender and Governance’ project on women’s civic and political participation in post-socialist Europe and is currently coordinating the Czech research team for the international project ‘Gendered Citizenship in Multicultural Europe: The Impact of Contemporary Women’s Movements,’ funded by the European Union’s Sixth Framework Program. Meryl Kenny is an Economic and Social Research Council Postdoctoral Fellow at the University of Edinburgh (Scotland, UK). Her doctoral thesis, entitled Gendering Institutions: The Political Recruitment of Women in Post-Devolution Scotland, used feminist and new institutional theory to explore the gendered dynamics of political recruitment in Scottish political parties. Her research interests include women and comparative politics, political recruitment, political representation, and territorial
9780230_245884_01_pre.indd xv
10/5/2010 5:01:29 PM
xvi
Contributors
politics. Her recent publications have appeared in Politics & Gender, Politics, and Scottish Affairs. Mona Lena Krook is Assistant Professor of Political Science and Women, Gender, and Sexuality Studies at Washington University in St. Louis (USA). In 2008–2009, she was a Fellow at the Radcliffe Institute of Advanced Study at Harvard University. She is the author of Quotas for Women in Politics: Gender and Candidate Selection Reform Worldwide (Oxford University Press 2009), which employs a combination of feminist and institutionalist theory to analyse the adoption and impact of quotas for the selection of female candidates to national parliaments around the globe. She is also co-editor with Sarah Childs of Women, Gender, and Politics: A Reader (Oxford University Press 2010). Joni Lovenduski is Anniversary Professor of Politics at Birkbeck College, University of London (United Kingdom), and a Fellow of the British Academy. Her recent books include Feminizing Politics (Polity 2005), State Feminism and Political Representation (Cambridge University Press 2005), and Political Recruitment: Gender, Race and Class in the British Parliament (Cambridge University Press 1995, with Pippa Norris). She is co-editor of The Politics of the Second Electorate (Law Books 1981, with Jill Hills) and Gender and Party Politics (Sage 1993, with Pippa Norris), as well as the author of many articles and essays in edited collections on issues of gender and politics. Fiona Mackay is Senior Lecturer in Politics and International Relations in the School of Social and Political Studies at the University of Edinburgh (Scotland, UK). Her research focuses on gendered dimensions of constitutional change and devolution in the United Kingdom. She is the author of Love and Politics: Women Politicians and the Ethics of Care (Continuum 2001), co-author of Women, Politics and Constitutional Change: The First Years of the National Assembly for Wales (University of Wales Press 2007, with Paul Chaney and Laura McAllister) and co-editor of Women and Contemporary Scottish Politics (Polygon 2001, with Esther Breitenbach). Steven Saxonberg is Professor of Sociology at the Department of Social Policy and Social Work at the Masaryk University in Brno (Czech Republic). He has written extensively about the collapse of communism in Central Europe and the transformation process, as well as postcommunist gender relations. He is the author of The Fall: A Comparative Study of the End of Communism in Czechoslovakia, East Germany, Hungary
9780230_245884_01_pre.indd xvi
10/5/2010 5:01:29 PM
Contributors xvii
and Poland (Routledge 2001) and The Czech Republic Before the New Millennium (Columbia University Press 2003). Jill Vickers is Distinguished Research Professor in Political Science at Carleton University in Ottawa (Canada) and was named a Fellow of the Royal Society of Canada in 2003. Her books include Politics As If Women Mattered (University of Toronto Press 1994, with L. Pauline Rankin and Christine Appelle), Reinventing Political Science: A Feminist Approach (Fernwood 1997), Gender, Race and Nation (University of Toronto Press 2002, with Vanaja Druhvarajan), and The Politics of Race: Canada, Australia and the United States (Golden Dog 2000). She has led several Canadian organizations and was Parliamentarian of Canada’s National Action Committee on the Status of Women. Georgina Waylen is Professor in Politics at the University of Sheffield (United Kingdom). She is the author of Gender in Third World Politics (Lynne Rienner 1996) and Engendering Transitions (Oxford University Press 2007), which was co-recipient of the American Political Science Association’s Victoria Schuck Award in 2008 for the best book published on women and politics. She is also the co-editor of Gender, Politics and the State (Routledge 1998, with Vicky Randall), Towards a Gendered Political Economy (Macmillan 2000, with Joanne Cook and Jennifer Roberts), and Global Governance: Feminist Perspectives (Palgrave 2008, with Shirin Rai).
9780230_245884_01_pre.indd xvii
10/5/2010 5:01:30 PM
9780230_245884_01_pre.indd xviii
10/5/2010 5:01:30 PM
1 Introduction: Gender, Politics, and Institutions Mona Lena Krook and Fiona Mackay
Gender, Politics, and Institutions – and the international network of scholars involved in this collaborative enterprise1 – is animated by the desire to find new tools and analytical frameworks to help us to answer some of the big questions and real-world puzzles about gendered power inequalities in public and political life. For example, how are formal structures and informal ‘rules of the game’ gendered? How do political institutions affect the daily lives of women and men, respectively? By what processes and mechanisms are such institutions produced, both reflecting and reproducing social systems, including gendered power relations? How do institutions constrain actors, ideas, and interests? Finally, what is the gendered potential for institutional innovation, reform, and change in pursuit of gender justice, and what are its limits? The ‘institutional turn’ in feminist and mainstream political science has been well documented. However, despite some common preoccupations, there has been little dialogue between mainstream new institutional scholars and institutionally oriented, feminist political scientists. As such, this edited collection breaks new ground by seeking to integrate gender and neo-institutionalist perspectives.2 In initiating this conversation, Gender, Politics, and Institutions sets the agenda for a ‘feminist institutionalism’. We begin in this chapter by providing a brief overview of institutionally focused research in feminist political science (FPS) and new institutionalism (NI), mapping key concepts, theories, and methods within each broad approach. We then identify concerns that are common to both fields and argue that each offers innovative tools for understanding the dynamics of political stability and change, and that there is considerable potential for the two approaches to mutually inform one another to devise innovative tools for understanding 1
9780230_245884_02_cha01.indd 1
10/5/2010 5:01:37 PM
2
Mona Lena Krook and Fiona Mackay
dynamics of political stability and change. We conclude with a brief discussion of each of the subsequent chapters in the volume, which taken together cover a number of major topics in feminist and institutionalist research, affording an opportunity to explore the contours, and assess the added value of a synthesized approach.
An institutional turn in feminist political science From its earliest days, an important strand of FPS has been concerned with institutions of state and society, particularly in explaining the chronic minority of women in ‘conventional’ public institutions and political life. This research has evolved over time in response to changing political developments. Academic interest has mirrored and paralleled growing political activism and advocacy around the political under-representation of women, the promotion and institutionalization of gender equality as a social and political goal, and the increased presence of women – including feminists – in political and state organizations, and as legislative, political, and bureaucratic actors. Fuelling this interest, the last three decades have seen substantial institutional change with significant consequences for women, for gender relations, and for gender equality. These changes include broad restructuring trends in many advanced, democratic welfare states involving processes such as marketization, regionalization, decentralization, and constitutional reform (see for example, Banaszak et al. 2003). Meanwhile, efforts at institutional redesign have been experienced globally in both peaceful and violent transitions to democracy, sparking new processes of state- and institution-building in successive waves of democratization across Eastern Europe, Africa, the Middle East, Latin America, and South Asia (Waylen 2007). In the course of these developments, demands for gender equality have become part of broader political discourses of democracy and modernization, as new international norms have diffused through global institutions and networks (Krook 2009; Towns 2010). Early feminist work on gender and institutions, however, generally overlooked the role of institutional processes and practices in reinforcing and reproducing gender inequality (Witz and Savage 1992). More specifically, the causes of gender inequality were understood to exist at the macro-level, rooted in a stratifying system or structure known as ‘patriarchy’. Institutions and organizations, therefore, were not the direct cause of inequality in, and of themselves. Rather, they were interesting only insofar as they illustrated or reflected a ‘more general set of
9780230_245884_02_cha01.indd 2
10/5/2010 5:01:38 PM
Introduction: Gender, Politics, and Institutions
3
patriarchal practices’ (Witz and Savage 1992: 7). Indeed, a line of feminist research influenced by Ferguson’s (1984) critique of bureaucracy viewed the institutions of the state and state bureaucracy as inherently and uniformly patriarchal. Subsequent literature shifted from a focus on universal structures – such as ‘power’ and ‘oppression’ – to the dynamic structuring of gender relations, in which relations of power operating through institutions are seen as a crucial part of the ‘structure’ of gender (Connell 2002: 59; see also Connell 1987; Walby 1990; Watson 1990). As a result, gender scholars increasingly scrutinized the ‘black box’ of the state and its institutions, against a backdrop of growing theoretical and empirical interest in variations within and across countries. In particular, real-world feminist engagements in and against the state provoked a theoretical reconsideration of political institutions, challenging preconceptions of ‘the state’ as essentially and monolithically patriarchal. The experience and study of the ‘femocrat’ phenomenon in Australia, in particular, led the way to the framing of new and more nuanced conceptions of states and gendered power relations (Franzway et al. 1989; Sawer 1990, 2003; Watson 1990). Several interrelated insights can be drawn from this and other work. First, state institutions are not monolithic but can only be understood when broken down into a number of institutional arenas or spaces (Chappell 2002a; Connell 1987, 2002; Franzway et al. 1989; Watson 1990). Second, while state institutions, variously defined, are historically the products of the ‘mobilization of masculine bias’ (Burton 1991: 14), and both produce and reproduce patriarchal power relations, they are historically variable in their composition and effects, and are theoretically open to change (Eisenstein 1991; Franzway et al. 1989; Sawer 1990; Watson 1990; see also Abrar et al. 1998; Ferree and Martin 1995; Halford 1992; Kantola 2006). In parallel to these trends, more complex understandings of ‘gender’ have emerged. This has shifted focus from near-exclusive attention to women at the individual level, ‘documenting sex imbalance and sexism’, to increasing interest in institutional-level analysis, analysing the underlying structures which underpin ‘institutionalised advantages and disadvantages’ according to gender (Duerst-Lahti and Kelly 1995: 44). Because ‘gender’ is a social construction, simply attending to the status of women in politics can only describe gendered patterns of access. It ‘cannot explain the persistence of inequality and permutations of exclusionary practices that operate within political institutions’, nor capture the gender norms and discourses that underpin institutional dynamics (Lovenduski 1998; Mackay 2004: 111).
9780230_245884_02_cha01.indd 3
10/5/2010 5:01:38 PM
4
Mona Lena Krook and Fiona Mackay
As the centre of gravity of FPS scholarship has gradually moved from a focus on ‘women and politics’ to more relational and institutionallevel conceptions of ‘gender and politics’ (Beckwith 2005; Childs and Krook 2006a; Kenny 2007; Lovenduski 1998; Mackay 2004), studies have begun to map the complexities of institutional gender dynamics. They highlight, in particular, the multiple ways in which gendered power relations and inequality are constructed, shaped, and maintained through institutional processes, practices, and rules (Annesley and Gains forthcoming; Chappell 2002a; Hawkesworth 2003; Kathlene 1995; Katzenstein 1998; Mackay 2006). Research in this area is theoretically and empirically eclectic, frequently drawing upon insights from other disciplines, most notably from work on gender and organizations in the field of sociology (see Acker 1990, 1992; Cockburn 1991; Savage and Witz 1992). Yet, while institutionally focused FPS is internally diverse, work in the field shares a number of common preoccupations and insights. As a body of literature, FPS is characterized – at a minimum – by three key features (Driscoll and Krook 2009): expanding the definition of ‘politics’ to include not just formal processes related to government and elections, but also informal actions embodied in social movements and interpersonal relations (Fraser 1989; Okin 1979; Phillips 1998); incorporating ‘gender’ as a relational concept and an analytic category (Beckwith 2005; Hawkesworth 2005; Kulawik 2009; Lovenduski 1998; Scott 1988); and generating insights that may be used to pursue political change (Hesse-Biber and Leavy 2007; Lovenduski 1998; Waylen 2009). Institutionally focused FPS extends this work by raising fundamental questions regarding the scope and nature of the ‘political’. It not only highlights interconnections between formal and informal institutions (Lovenduski 1998; Phillips 1998; Randall 2002), but also calls attention to the boundary work involved in demarcating the ‘political’ from arenas, issues, and actors deemed ‘social’ or ‘private’, and therefore ‘nonpolitical’ (Fraser 1987; Kulawik 2009; Weldon 2002). Nonetheless, research on the gendered effects of formal institutions is by far the most extensive. FPS scholars, for example, have analysed the impact of electoral systems on women’s political representation (Caul 1999; Kenworthy and Malami 1999; Matland 1995; Norris 2004; Reynolds 1999; Rule and Zimmerman 1994). They have explored the role of political parties in promoting and constraining women’s behaviour as legislators and social movement activists (Kittilson 2006; Lovenduski 2005b; Lovenduski and Norris 1993; Norris and Lovenduski 1995; Opello 2006), as well as the complex relationship between state feminism,
9780230_245884_02_cha01.indd 4
10/5/2010 5:01:38 PM
Introduction: Gender, Politics, and Institutions
5
women’s movements, and female politicians in pursuit of substantive policy change (Annesley 2010; Chappell 2002a; Chappell 2002b; Lovenduski 2005a; Mazur 2001, 2002; Outshoorn 2004; Outshoorn and Kantola 2007; Stetson and Mazur 1995; Weldon 2002). In light of their interest in change, institutionally oriented FPS researchers have also paid close attention to the gendered implications of the changing institutional and legal arrangements as a result of devolution and European integration (Brown et al. 2002; Chaney et al. 2007; Mackay 2006; Ostner and Lewis 1995; Vogel-Polsky 2000); the gendered impact of regime change, such as transitions to democracy (Waylen 2007); and the political opportunity structures that inform women’s movement interactions with the state, including territorial divisions of power, and processes of state reconfiguration (Banaszak et al. 2003; Brown 2001a; Chappell 2002a; Dobrowolsky and Hart 2003; Grace 2006, 2008; Haussman et al. 2010; Kantola 2006; Vickers 1994, 2010). Although they rarely use the language of ‘institutions’ and ‘institutionalism’, FPS scholars have similarly examined the importance of informal institutions, which NI researchers define as the practices, discourses, and norms that shape formal institutions, but which may also contradict or undermine formal rules (Helmke and Levitsky 2004). Like mainstream institutionalists, therefore, many studies in FPS confer the same status on the formal and the informal aspects of political life, although they often do so implicitly rather than explicitly. Numerous scholars, for example, analyse the gendered impact of political practices, such as the parliamentary procedures and cultures that act as constraints on the ability of female legislators to represent women (Carroll 1984, 2001; Childs 2004; Dodson 2001, 2006; Franceschet 2005; Rosenthal 1998); informal practices and condoned rule-breaking through which male politicians may reassert masculine political dominance (Kathlene 1995; Liddle and Michielsens 2000; Puwar 1997, 2004a; Shaw 2000); and the de facto rules of political parties that influence candidate selection procedures (Freidenvall 2006; Lovenduski and Norris 1993; Kenny 2009; Krook 2009; Norris and Lovenduski 1995; Perrigo 1996). Along similar lines, work by institutionally focused FPS researchers explores the constraining power of – but also the political opportunities afforded by – existing political discourses (Bacchi 1996, 1999, 2005). They note, for example, how actors of various kinds articulate their positions for and against gendered political change in relation to widely accepted values like ‘fairness’, ‘democracy’, and ‘equality’ (Freidenvall 2005; Holli et al. 2006; Krook et al. 2009). Finally, a wide range of studies in FPS have assessed the powerful role of norms ‘that prescribe and proscribe
9780230_245884_02_cha01.indd 5
10/5/2010 5:01:38 PM
6
Mona Lena Krook and Fiona Mackay
“acceptable” masculine and feminine forms of behaviour, rules, and values for men and women within institutions’ (Chappell 2006: 226; see also Duerst-Lahti and Kelly 1995; Kathlene 1995). This includes work on the ‘gender regimes’ and ‘gender contracts’ that underpin various kinds of welfare state arrangements (Connell 2002, 2006; Hirdman 1990; Jenson 1997; Lewis 1992; Sainsbury 1996), as well as Chappell’s (2006; see also Chappell 2002a; Stivers 1993) analysis of the norm of bureaucratic neutrality demonstrating that it is profoundly gendered, such that the more embedded and enforced the norm of ‘neutrality’ is, the harder it will be for feminists to advance ‘biased’ claims of gender equality. This feminist scholarship is diverse, presenting a wide variety of ways of understanding and conceptualizing gender, politics, and institutions. The key point of agreement across these approaches, however, is that conceiving of political (and social) institutions as ‘gendered’ is crucial to understanding the practices, ideas, goals, and outcomes of politics. Grasping the ways in which institutions reflect, reinforce, and structure unequal gendered power relations in wider society, in turn, offers insights into the dynamics of continuity and change – and the means for interrupting them to promote or undermine feminist goals. For FPS scholars, gender provides a central structuring dynamic of institutions, such that gender relations and gender norms are key legacies with which to contend in feminist analysis and strategies for political change (Mackay 2004: 112–113; see also Annesley and Gains forthcoming; Chappell 2002a; Duerst-Lahti 2002; Duerst-Lahti and Kelly 1995; Hawkesworth 2003; Kenney 1996; Kenny 2007; Krook 2009; Lovenduski 1998, 2005b; Mackay and Meier 2003; Mackay 2008). To say that an institution is gendered means that constructions of masculinity and femininity are intertwined in the daily culture or ‘logic’ of political institutions, rather than ‘existing out in society or fixed within individuals which they then bring whole to the institution’ (Kenney 1996: 456). While constructions of masculinity and femininity are both present in political institutions, the masculine ideal underpins institutional structures, practices, discourses, and norms, shaping ‘ways of valuing things, ways of behaving, and ways of being’ (Duerst-Lahti and Kelly 1995: 20). Yet, gender relations are also plural and cross- cutting, playing out in different ways across various types of institutions, and on distinct institutional levels. These range from the construction of images, symbols, and ideologies that justify, explain, and legitimize institutions, and their gendered patterns of hierarchy and exclusion (Acker 1992: 568) to interpersonal day-to-day interactions,
9780230_245884_02_cha01.indd 6
10/5/2010 5:01:38 PM
Introduction: Gender, Politics, and Institutions
7
where the continuous performance of gender takes place (Kenney 1996: 458; see also Acker 1990, 1992; Connell 1987, 2002; Lovenduski 1998). On the one hand, formal and informal institutions shape and constrain gendered political behaviour, defining the parameters of what action is possible and intelligible. On the other hand, political institutions are themselves constituted by these embodied social practices of ‘doing gender’ on a daily basis (Connell 1987, 2002; see also Kenny 2007). Recognizing these reciprocal relationships, in turn, opens up possibilities for agency and change. For, as Beckwith (2005: 133) notes, if institutions are gendered, the potential exists for them to be ‘regendered’, including in ways that might promote gender equality (see also Hawkesworth 2005; Kenny and Mackay 2009). Uncovering possibilities to reshape the status quo in this manner is critically important to FPS, which is explicitly concerned with recognizing how institutions reproduce gendered power relations, but perhaps more importantly, with how these institutions can be challenged and reformed (Mackay and Meier 2003). Theorizing institutions in their multiple forms, as well as recognizing their gendered dimensions, presents considerable challenges for FPS scholars, provoking a reconsideration of appropriate methods, frameworks, and research directions. As Kenny (2007: 7) observes, institutionally oriented FPS ‘has begun to map the complexities of institutional gender dynamics’ but ‘continue[s] to grapple with the implications of incorporating “slippery” and complex understandings of gender into their frameworks and analyses, and work in this area is eclectic and, often, uneven’ (see also Lovenduski 1998; Mackay 2004). One of the most ambitious projects on gender and political institutions to date is the research, and series of publications associated with the Research Network on Gender Politics and the State (RNGS). Adopting a ‘feminist empiricist’ approach (Harding 1986; cf. Mazur 2002), this work has provided extensive case studies of gendered structures and feminist agency around the state; the origins, impact, and changing role of women’s policy machineries within state structures; and the complex relationship between state feminism, women’s movements, and women political representatives in pursuit of substantive policy change (Mazur 2001; Lovenduski 2005a; Outshoorn 2004; Outshoorn and Kantola 2007; Stetson and Mazur 1995). However, there is growing recognition among feminist scholars that a more systematic approach is needed in order ‘to bridge the gap’ between complex theories of gender and workable theories for empirical research (Mackay 2004: 114; see also Childs 2004; Childs and Krook
9780230_245884_02_cha01.indd 7
10/5/2010 5:01:38 PM
8
Mona Lena Krook and Fiona Mackay
2006a; Kenny 2007; Lovenduski 1998), as well as to get beyond the description stage in order to begin to analyse, and understand gendered patterns and effects (Lovenduski in this volume). Innovative conceptual tools and approaches, in particular, are needed in order to address the considerable challenges posed by the turn to institutions; and, in particular, the turn in focus to gendered institutions. Initiating a new research agenda requires asking – and answering – questions related to interactions between gender and institutional effects, the origins and shape of gendered institutional change, and the reasons for positive outcomes in some contexts, but little or no change in others. A wave of recent studies suggest that there is strong potential for engagement with other institutionally oriented scholars, particularly those working in the broad field of ‘new institutionalism’ (Lovenduski 1998; see also Adams and Orloff 2005; Annesley and Gains forthcoming; Chappell 2006; Driscoll and Krook 2009; Kantola 2006; Kenny 2007; Kenny and Mackay 2009; Kulawik 2009; Lovenduski 2005b; Mackay 2004; Mackay et al. 2009; Mackay and Meier 2003; Waylen 2009; Weldon 2002).
New institutionalisms in mainstream political science The basic premise of new institutionalism (NI) is that institutions ‘matter’, an ‘argument that the organization of political life makes a difference’ (March and Olsen 1984: 747). Yet, while the term ‘new institutionalism’ is widely used in political science, there is considerable debate over ‘just what [it] is’ (Hall and Taylor 1996: 936). Work that identifies with NI has largely converged around four main approaches: rational choice institutionalism, historical institutionalism, organizational or sociological institutionalism, and more recently, discursive institutionalism.3 This proliferation of perspectives has served to expand the reach of institutionalist theory into the study of a wide range of political phenomena. At the same time, however, it has led to a compartmentalization and fragmentation of institutionalist research (cf. Crouch 2003), organized according to schools of thought, types of institutions, and patterns of institutional transformation. The four main schools of NI share a common focus on ‘institutions’, but are characterized by distinct theoretical and methodological commitments, leading them to analyse political phenomena using slightly different sets of analytical assumptions. Rational choice institutionalists work mainly at the micro-level, aiming to understand the origins of institutions, the mechanisms of their survival, and the nature of their effects on macro-level political outcomes (Weingast 2002). Drawing on
9780230_245884_02_cha01.indd 8
10/5/2010 5:01:38 PM
Introduction: Gender, Politics, and Institutions
9
the insights of game theory, they view institutions as conventions created by actors in order to overcome collective action problems (Ostrom 1990), either by reducing uncertainty (North 1990), or by restructuring incentives so that individuals are more motivated to cooperate (Weingast 2002). Some argue, as a consequence, that institutions endure when they provide more benefits to the relevant actors than alternative institutional forms. However, others are careful to stress that these dynamics do not necessarily ensure the most efficient outcomes (North 1990), and that in many cases, institutions are not only structures of coordination, but also structures of power and domination (Knight 1992; Moe 2006). Historical institutionalists, in contrast, conduct their research primarily at the meso- or macro-level, focusing on the long-term ramifications of largely contingent events (Pierson 2003; Pierson and Skocpol 2002). Interested in apparent historical inefficiencies, they approach institutions as the various formal and informal procedures, routines, norms, and conventions embedded in the organization of politics, society, and the economy. While concerned with how these institutions permit and exacerbate various asymmetries of power (Thelen and Steinmo 1992), most of the theoretical attention in this literature relates to the concept of path dependence and the role of unintended consequences (Mahoney 2000). In the course of grappling with questions of sequencing and possible period effects (Lieberman 2001), however, other scholars have sought to nuance the division between institutional creation and institutional consequences. Seeking to call attention to dynamics of institutional development over time (Pierson 2004), they introduce concepts such as institutional conversion and layering to capture what they see as more complicated relationships between continuity and change (Mahoney and Thelen 2010; Schickler 2001; Streeck and Thelen 2005; Thelen 2003, 2004). Organizational or sociological institutionalists, in turn, present a middle position centred on the links between micro- and macro-level political interactions (DiMaggio and Powell 1991). Adopting a more interpretive lens, they define institutions to include formal rules, procedures, and norms, as well as symbol systems, cognitive scripts, and moral templates that provide the ‘frames of meaning’ guiding human behaviour (March and Olsen 1989). More than the other two schools, they emphasize the interactive and mutually constitutive character of the relationship between institutions and individual actions. As such, they understand change not in relation to cost–benefit analysis or the legacies of past events, but instead in terms of attempts to enhance the
9780230_245884_02_cha01.indd 9
10/5/2010 5:01:38 PM
10
Mona Lena Krook and Fiona Mackay
social legitimacy of a particular institution (DiMaggio and Powell 1991; Douglas 1986; Meyer and Rowan 1991). These concerns may extend as far as the global arena: as proponents of the ‘world polity’ school note, countries themselves are subject to pressures to conform to the standards of a ‘modern’ state, leading many to adopt similar institutions in order to establish or enhance their legitimacy in the international sphere (Meyer et al. 1997). Discursive institutionalists, finally, engage with multiple levels of analysis, ranging from – and connecting – the micro- to the macrolevel. Although these scholars embrace a relatively broad definition of institutions, compatible with rational choice, historical, and sociological perspectives, their approach differs from the others in a number of significant ways. First, discursive institutionalists place greater emphasis on the role of ideas and ideas-as-discourses in influencing actor interests, preferences, and behaviour, although their definitions and uses of ideas and discourse vary widely (compare for example Campbell and Pedersen 2001; Hay 2006; Lynggaard 2007; Schmidt 2008). Institutions are built on ideational foundations, and these ideas shape institutional design and development, constraining possibilities for agency (Hay 2006; Schmidt 2008). Second, discursive institutionalists move beyond powerdistributional perspectives, arguing that discourse itself is a medium of power (Freidenvall 2009; Lynggard 2008). In this view, power is not simply defined in terms of positional power within a particular institutional context; rather, ideas and discourses construct and shape the very ‘exercise of power’, including ‘(subjective) perceptions of position’ (Schmidt 2010: 18). Third, discursive institutionalists outline a theory of institutional change ‘from the inside’, drawing attention to the interactive processes through which ideas and discourses are constructed, legitimated, and communicated within particular institutional contexts (Schmidt 2008, 2010). Yet, most discursive institutionalists also distinguish ‘discourses’ from ‘institutions’, referring to the discursive effects of institutions (Fischer 2003) or the institutional contexts ‘in which and through which ideas are communicated via discourse’ (Schmidt 2010: 4). Thus, while they stress the importance of discourse in institutional innovation, dynamism, and transformation, they stop short of treating discourses on a par with institutions themselves (Freidenvall and Krook this volume). Despite these important differences, there is a growing trend towards rapprochement and synthesis among NI scholars (see for example Bates et al. 1998; Katznelson and Weingast 2005), such that it is possible
9780230_245884_02_cha01.indd 10
10/5/2010 5:01:38 PM
Introduction: Gender, Politics, and Institutions
11
to draw out continuities across these approaches, even if the authors in this volume engage with one or more specific NI variants in their respective chapters. The common feature uniting all these approaches, for example, is their attention to formal and informal institutions. Indeed, perhaps their most notable feature is their effort to confer the same theoretical, empirical, and methodological status to both kinds of institutions. North, for example, expands on his widely cited definition of institutions as ‘the rules of the game in a society or ... the humanly devised constraints that shape human interaction’ (1990: 3) by describing these more specifically as ‘formal constraints – such as the rules that human beings devise – and informal constraints – such as conventions and codes of behavior’ (1990: 4). Similarly, in an overview of various types of institutionalist approaches, Peters identifies an institution as a ‘structural feature of the society and/or the polity ... that may be formal (a legislature, an agency in the public bureaucracy, or a legal framework), or may be informal (a network of interacting organizations, or a shared set of norms)’ (1999: 18). These types of statements have led other scholars to characterize the object of inquiry in institutional analysis as a continuum between formal and informal institutions, ‘mov[ing] from the study of such intangible phenomena as ideas, meanings, signifiers, beliefs, identities, attitudes, worldviews, discourses, and values to such tangible entities as states, constitutions, bureaucracies, churches, schools, armies, parties, and groups’ (Ethington and McDonagh 1995: 470). While recognizing that institutions may take a variety of forms, some researchers do note, however, important differences between formal and informal institutions. A great deal of attention, for example, has focused on why formal institutions do not have the effects that their creators intended (Goodin 1996; Pierson 2004), and conversely, how and why informal institutions have the effects that they do. In an attempt to understand why informal institutions exist, some hypothesize that informal rules emerge when formal institutions are incomplete; when actors prefer, but cannot achieve, a formal institutional solution; or when actors are pursuing goals that are not publicly acceptable, either because they are unlikely to stand the test of public scrutiny, or will attract international condemnation (Helmke and Levitsky 2004: 730–731; cf. Leach and Lowndes 2007). This distinction is captured well in North’s observation that ‘formal rules may change overnight as the result of political or judicial decisions, [but] informal constraints embodied in customs, traditions, and codes of conduct are much more impervious to deliberate policies’ (1990: 6).
9780230_245884_02_cha01.indd 11
10/5/2010 5:01:39 PM
12
Mona Lena Krook and Fiona Mackay
Although both kinds of institutions delineate rules of appropriate behaviour (March and Olsen 1989), they may be subject to distinct processes of creation and alteration, at the same time that they work together to shape political outcomes. Given the focus on institutions as structures and constraints, most research focuses on institutional stability, seeking to explain how and why institutions lock the expectations and behaviour of individuals into relatively predictable, self-reinforcing patterns, even in the face of major changes in background conditions. As Pierson (2000a) notes, these tendencies are especially characteristic of political institutions: while market mechanisms may potentially disrupt dynamics of increasing returns in the economy, various features of politics make increasing returns more likely, namely the central role of collective action, the possibilities for employing political authority to magnify power asymmetries, and the absence or weakness of efficiency-enhancing mechanisms of competition and learning. Nonetheless, institutional change does sometimes occur. To account for this, scholars have developed two broad models that present different views regarding the form and content of institutional innovation. The first, and by far the most common, draws a sharp analytical distinction between moments of change and mechanisms of reproduction. According to this perspective, new institutions emerge rarely, and often only in the context of crisis or great uncertainty. These ‘critical junctures’, which cannot be predicted from prior events or initial conditions, serve as major turning points, in which the – often contingent – decisions of actors establish certain directions of change (Capoccia and Kelemen 2007; Mahoney 2000). ‘Path dependence’ then precludes the emergence of other options over time through increasing returns, sunk costs, durable commitments, personal and social conservatism, and learning curves (North 1990; Pierson 2000a). Crucially, these dynamics may reinforce movement along a given trajectory, even when other choices appear better or more efficient (Aminzade 1992; Tilly 2001). This approach resembles a ‘punctuated equilibrium’ model in which long periods characterized by path dependence alternate with brief and dramatic turns of events (cf. Krasner 1984). While still widely applied, this perspective has come under increasing challenge in recent years from scholars working in all four traditions, who criticize this strict separation between institutional creation and institutional reproduction. Some rational choice theorists, for example, prefer a model of endogenous ‘institutional refinement’, whereby ‘institutions organically evolve (or are intentionally designed) through changing, introducing, or manipulating institutional elements while
9780230_245884_02_cha01.indd 12
10/5/2010 5:01:39 PM
Introduction: Gender, Politics, and Institutions
13
supplementing existing elements (or responding to their failure to generate desire behavior)’ (Greif and Laitin 2004: 640). Along similar lines, several historical institutionalists note that ‘when institutions have been in place for a long time most changes will be incremental’ (Pierson 2004: 153). Consequently, they theorize that institutions more typically survive through dynamics of institutional conversion, where existing institutions are directed to new purposes, and institutional layering, where some elements of existing institutions are renegotiated, but other elements remain (Streeck and Thelen 2005; Thelen 2003). These intuitions are echoed by some sociological institutionalists, who observe that ‘ideas about appropriate behavior ordinarily change gradually through the development of experience and the elaboration of worldviews’ (March and Olsen 1989: 171). They are also the focus of various discursive institutionalists, who call attention to how elements of one discourse are translated into another through ‘discursive alliances or bricolage’ (Campbell and Pedersen 2001: 12). Taken together, these views are closer to a ‘bounded innovation’ model, in which periods of institutional reproduction overlap with moments of institutional creation in partial, and often unanticipated ways (cf. Weir 1992; see also Thelen, 2003, 2004).
A common ground in feminist institutionalism? On the face of it, there is considerable potential for productive dialogue between NI and institutionally focused FPS. Although each body of literature is internally diverse, they share important theoretical orientations and research goals. For example, both employ broad conceptions of the political and its interconnection with the social. Each is concerned with analysing interactions between actors and institutions; uncovering the interplay between formal rules and informal practices, discourses, and norms; and probing the consequent effects of these dynamics. Another mutual interest, at least among FPS and variants of NI that integrate the cultural,4 is their ‘value-critical’ stance, sharing an understanding that seemingly neutral institutional processes are influenced by practices, discourses, and norms, and thus always operate within wider cultural contexts. Finally, both FPS and NI are centrally concerned with explaining patterns of institutional creation, continuity, and change, as well as the scope for agency within institutional constraints. These common concerns and commitments suggest that it may be possible to forge a shared approach, drawing on elements of both feminism and institutionalism to devise a means for better understanding
9780230_245884_02_cha01.indd 13
10/5/2010 5:01:39 PM
14
Mona Lena Krook and Fiona Mackay
the gendered dynamics of political life. Yet, what might a feminist institutionalism look like – and furthermore, what would it mean in practice? To explore the potential for a feminist institutionalism, contributors to this volume take up three sets of questions. First, how might NI approaches contribute to better understanding the puzzles addressed by FPS? What specific concepts or perspectives from institutionalism might be useful to feminist scholars? Second, how might work in FPS inform NI understandings of structure and agency in political life? What elements of feminism might be helpful for these researchers? Third, is there – or can there be – a feminist institutionalism? If so, what does it comprise and how might it be operationalized? What is the ‘added value’ of this approach for both mainstream and feminist political analysis? The next nine chapters tackle these questions in relation to empirical studies of candidate selection, political representation, welfare state policies, democratic transitions, multi-level governance, and international law. They are sequenced in the volume according to three lines of research: political recruitment/representation, state–family relations, and state structures and political innovation. Within these categories, individual studies are ordered thematically, as a means to build on and also facilitate comparisons with the preceding and subsequent analyses. As a group, the authors converge on the opinion that NI and FPS would mutually benefit from closer engagement, but draw upon and combine a range of different insights. Some adopt and operationalize frameworks from specific schools of NI, while others take a more eclectic approach by borrowing and adapting tools from different types of NI. Similarly, they employ distinct feminist definitions of ‘gender’, ‘politics’, and ‘power’. Read together, therefore, they provide insight into the FPS and NI concepts, theories, and methods that might be applied and combined to answer a wide range of theoretically, and substantively important questions in political science. The first three chapters address questions of political recruitment and political representation. In Chapter 2, Meryl Kenny explores gendered dynamics of institutional reform and innovation in the institutions of political recruitment in post-devolution Scotland. Highlighting their gendered and institutional dimensions, she extends the ‘supply and demand’ model – proposing that the number of women elected to political office depends on the supply of women willing to stand for office and the demands of party ‘gatekeepers’ who select candidates – via an illustrative case study of a Scottish Labour Party constituency seat selection contest. Given the high numbers of women elected since 1999, the
9780230_245884_02_cha01.indd 14
10/5/2010 5:01:39 PM
Introduction: Gender, Politics, and Institutions
15
Scottish Parliament is often considered to be a ‘success story’ in terms of women’s representation. However, evidence from the case study indicates that the reality is far more complex. Overall, the findings suggest that while some innovations have been made in the candidate selection process, the underlying trend is one of erosion, decline, and even reversal following the initial reforms. In particular, they illustrate the difficulties of institutionalizing a ‘new’ gender-balanced politics within a pre-existing institutional and cultural context. The case therefore highlights the complex and gendered dynamics of institutional design, continuity, and change, pointing to the difficulties of reforming recruitment in the face of powerful institutional and gendered legacies. Chapter 3 similarly takes up issues of candidate selection, but analyses these in a comparison of the origins and effects of electoral gender quotas in Sweden and France. Introduced in countries around the world, these policies aim to increase the proportion of female candidates, but have diverse effects on the numbers of women elected. Combining elements of feminism and discursive institutionalism, Lenita Freidenvall and Mona Lena Krook argue that quota campaigns should be seen as discursive struggles to define – and then maintain or reformulate – existing gendered institutions, which interact with a range of other institutions in processes of candidate selection. Individual campaigns succeed and fail in these tasks to varying degrees: quota policies may or may not be adopted, and when they are, may or may not have major effects on the numbers of women elected. To unravel these causal processes, Freidenvall and Krook examine and compare two cases where ostensibly similar policies produce quite different results: Sweden, where 50 per cent party quotas adopted by most political parties generated the election of 47 per cent women in 2006, and France, where a 50 per cent legislative quota imposed on all political parties led to the election of only 18 per cent women in 2007. The analysis reveals that discourses may enable gendered policy change, but also constrain its form, content, and impact, together with reigning rules, practices, and norms. Shifting attention to the next stage in the political process, policymaking, Chapter 4 examines the extent to which the tools of historical institutionalism might contribute to a better understanding of an enduring question in FPS: how the presence of women in politics may or may not alter policy outcomes in ways that benefit women as a group. Adequate answers, Susan Franceschet suggests, require, first, analysing the actions of legislators and the outcomes of those actions and, second, theorizing the ways in which formal rules and informal norms in legislative institutions may facilitate or obstruct the successful
9780230_245884_02_cha01.indd 15
10/5/2010 5:01:39 PM
16
Mona Lena Krook and Fiona Mackay
promotion of women’s rights. Focusing on the cases of Argentina and Chile, she shows how formal and informal legislative institutions in the two countries reinforce unequal power relations between men and women, leading to differences in how female legislators represent women’s interests with varying degrees of success. These gendered institutions explain why legislators in Argentina are more active on a range of women’s rights issues than in Chile, at the same time that Chilean legislators are more successful in getting women’s rights laws adopted, even though they introduce fewer initiatives. Elaborating further on public policy, the next three chapters tackle issues of state–family relations. Chapter 5 observes that, due to feminist interventions, the literature on welfare states has long recognized the gendered impact of ‘welfare regimes’, understood as the intersecting institutions in the structure of the welfare state. Feminist scholars have been particularly successful in widening the scope of ‘institutions’ that are considered as significant in these regimes, including the gendered dynamics of the family and caring relationships, and the way in which gender relations play out in other socio-economic spheres such as the labour market. Yet, as Michelle Beyeler and Claire Annesley note, there has been less attention paid to how the gendered institutional architecture of welfare states may be reformed in ways that promote greater gender equality. A more dynamic account, they argue, must be able to account for variations in the extent of reforms undertaken in previously strong ‘male-breadwinner’ states, as well as the effects of ostensible women-friendly initiatives that, in the end, reproduce existing gender inequalities. Extending what they view as the feminist institutionalist perspective already present in welfare state analysis, Beyeler and Annesley propose that a pluralist approach – combining rational choice, sociological, and historical institutionalism – can be merged with feminist concepts to develop useful tools for the analysis of gendered welfare reforms, which they illustrate through a comparison of Germany, Switzerland, and the United Kingdom. Sharing an interest in how institutions shape prospects for changes in gendered public policies, Chapter 6 explores the importance of federalism – also known as ‘multi-level governance’ – in either facilitating or blocking women’s policy demands. In the case of Canada, Joan Grace suggests that attention to the division of powers between the provinces and territories, on the one hand, and the federal government, on the other, is vital for understanding women’s socio-economic realities. More specifically, policy areas important to women – like child care, social assistance, job training, and education – are provincial responsibilities,
9780230_245884_02_cha01.indd 16
10/5/2010 5:01:39 PM
Introduction: Gender, Politics, and Institutions
17
at the same time that the federal government retains the major share of financial resources needed to effectively fund these programme. This division of powers has required a politics of accommodation across the two levels of government that, depending on the period, has been either collaborative or quite fractious. Melding tenets of historical institutionalism with feminist transformative politics, Grace develops a feminist institutionalist approach to theorize how this politics of accommodation – and the resulting permeability of federal governance structures – provides both opportunities and challenges to women’s policy agendas. Her two case studies, debates over child care and social policy, which took place during the Social Security Review in 1994 and the Social Union Framework Agreement in 1999, highlight the significance of attending to institutions in terms of feminist policy analysis, as well as in terms of a critical, gendered understanding of the policy context in institutionalist research. Chapter 7 also addresses state policies related to the family, but goes further than the previous two chapters in tracing the historical legacies shaping the current form taken by family policies in the Czech and Slovak Republics. Observing that the two countries have the most similar policies among post-communist states, Hana Hašková and Steven Saxonberg draw on the concept of ‘critical junctures’ or ‘formative moments’ from historical institutionalism to identify four key points – reaching back to when these countries were part of the Austro–Hungarian Empire and then Czechoslovakia – when government officials made decisions which continue to shape and influence present-day family policy. The authors also incorporate a feminist perspective, however, which emphasizes how ideas about gender inform state policies, at the same time that state policies may alter or reinforce gendered power relations. In the Czech and Slovak Republics, Hašková and Saxonberg note, early family policies aimed to confine women to the home and establish men’s primary roles as breadwinners. A shift in this approach appeared to take place during the communist period, when nearly all women had to work outside the home, but the lack of changes in the private sphere – combined with a series of decisions by the government regarding maternity leave, child care, and schooling – had the effect of encouraging traditional gender roles. Further developing these themes, the last three chapters explore the potential for institutional arrangements, including new institutions, to empower women as a group, focusing on opportunities that have been seized, as well as missed, at moments of political innova-
9780230_245884_02_cha01.indd 17
10/5/2010 5:01:39 PM
18
Mona Lena Krook and Fiona Mackay
tion. In Chapter 8, Georgina Waylen draws on an eight-country study of transitions to democracy to explain why outcomes for women have been more positive in some countries than in others with the advent of new political arrangements. Tracing developments in each country over a period of 20 to 30 years, she focuses on changes, and lack of change, from the previous non-democratic regime, through the democratic transition, and through at least one decade of post-transition. Countries varied in characteristics at all three points in time, permitting comparisons of the legacies of decisions made during each particular period. On the basis of her analysis, Waylen argues that feminist research has cast important light on how transitions to democracy are gendered. However, she also suggests that recent theoretical developments in historical institutionalism – in particular, such concepts as institutional layering and conversion – can help feminist scholars better understand the complexities of gendered institutional change, including bigger questions about how to promote positive outcomes for women. She thus calls for greater attention to interactions between structure and agency, highlighting the potentially distinctive contribution and advantages of a feminist historical institutionalism. Chapter 9 adopts a similar comparative approach, developing and testing six hypotheses concerning the presence of a ‘federalism advantage’ for advancing feminist projects in seven federal states. The intuition is that, in federations, institutions dividing powers across levels of government provide multiple access points to citizens, and thus might be favourable towards feminists seeking avenues of change. Although the unique designs of many democratic federations make comparative analysis difficult, Jill Vickers suggests that gaining a better gendered understanding of federalism requires assessing the validity of two competing views: one which sees federalism as an obstacle to women’s participation, and the other which claims that the multiple points of access to decision- makers in federations constitute an advantage that women can exploit to develop multi-level strategies. Engaging in a twoway comparison of federal versus unitary states and then across different types of federations, Vickers draws on an eclectic mix of feminist interpretations of institutionalist approaches – historical, post-structuralist, sociological, and rational choice – to outline the six hypotheses, related to the role of multiple points of access, high levels of federalization, competition among levels of government, division of powers, and asymmetrical versus symmetrical federal arrangements. She concludes that a federalism advantage is most easily realized in centralized, symmetrical federations, but that there are also opportunities in asymmet-
9780230_245884_02_cha01.indd 18
10/5/2010 5:01:39 PM
Introduction: Gender, Politics, and Institutions
19
rical federations under conditions of competition or conflict between levels of government. Expanding the reach of multi-level governance, Chapter 10 investigates the impact of the International Criminal Court (ICC) for the pursuit of gender justice at the global level. The Rome Statute, providing the legal foundation of the ICC, is groundbreaking in its codification of international criminal law and the recognition of the crimes committed against women in times of war and conflict. It not only criminalizes acts of rape, sexual slavery, and enforced pregnancy in the most advanced articulation ever of gender-based violence under international law, but it also embeds a gender equality mandate into the structures and processes of the ICC to ensure that women, and their rights are considered by the judicial, prosecutorial, and administrative arms of the Court. In many respects, therefore, it appears that the ICC is a ‘new’ institution offering a clean slate for rewriting women’s experience of gender justice. Yet, as Louise Chappell observes, the Court has produced a number of contradictory outcomes in its first five years. Combining feminist attention to gender with insights from historical institutionalism on institutional origins and development, she elaborates Fiona Mackay’s concept of ‘nested newness’ to capture the way in which a new institution relates to others, whose legacies and continuities with the past profoundly effect their operation, and can be shaped, modified, and constrained – and conceivably enhanced – through the new institution’s interaction with these other institutions. Despite its newness, the ICC carries gender legacies from the past which may influence its capacity to operate as a reform venue to promote the aims of gender justice. Together with the other contributions in the volume, the chapter reveals the analytical leverage afforded by a combination of feminism and institutionalism, with potentially crucial implications for strategies to improve women’s status around the globe. Seeking to pull these various contributions together, in Chapter 11 Fiona Mackay draws on the preceding empirical chapters to provide a provisional set of answers to the questions around which the book is structured. Mackay argues that the collection provides important conceptual and empirical blueprints for a feminist institutionalism – or institutionalisms – in the making. Such an approach, she concludes, has considerable potential for enhancing understanding and analyses of institutional dynamics, gender power, and the patterning of gendered inequalities in political and public life. To this end, she sets out an agenda for future theory building and empirical research.
9780230_245884_02_cha01.indd 19
10/5/2010 5:01:40 PM
20
Mona Lena Krook and Fiona Mackay
Notes 1. This network was launched in 2006 as the Feminism and Institutionalism International Network (FIIN). Details on members and activities can be found at the network website at www.femfiin.com. 2. This chapter is the product of conversations, shared ideas, and common insights, as well as joint writing projects, with our colleagues in FIIN. In particular, we draw upon work with Lenita Freidenvall (Freidenvall and Krook 2007), Meryl Kenny (Kenny and Mackay 2009), and Petra Meier (Mackay and Meier 2003). 3. Until recently, most scholars recognized three broad schools of new institutionalist thinking: rational choice, historical, and sociological (Hall and Taylor 1996; Immergut 1998). However, others have distinguished as many as seven different types of NI, adding normative, empirical, interest representation, and international institutionalism (Peters 1999). 4. There is considerable debate about whether rational choice institutionalism is compatible with value-critical approaches within NI and FPS (Driscoll and Krook 2009).
9780230_245884_02_cha01.indd 20
10/5/2010 5:01:40 PM
2 Gender and Institutions of Political Recruitment: Candidate Selection in Post-Devolution Scotland Meryl Kenny
This chapter adopts a feminist institutionalist approach in order to explore the gendered dynamics of the institutions of political recruitment in post-devolution Scotland. The term ‘feminist institutionalism’ (FI) is understood here as an approach that synthesizes elements of the new institutionalism (NI) – specifically historical institutionalism (HI) – and feminist political science (FPS). The chapter sees this is a relationship of mutual benefit. On the one hand, NI can offer a number of useful tools to FPS, particularly with regards to key issues of institutional innovation, continuity, and change. On the other hand, a gendered approach can enrich NI theory, drawing attention to the complex ways in which gender power relations play out in political institutions. The chapter assesses the promise of an FI approach in the context of the comparative literature on gender and political recruitment. Current research in the women and politics field has focused increasingly on political recruitment, pointing to the candidate selection process as a critical area in need of further investigation and research (see for example Krook 2009; Shepherd-Robinson and Lovenduski 2002). The chapter begins with a brief theoretical critique of the dominant framework used in the literature – the supply and demand model (Norris and Lovenduski 1995). It contends that the combined insights of feminist and new institutional theory offer a promising way to take the supply and demand model forward, building on, and developing gendered and institutional trends already present implicitly in the literature. 21
9780230_245884_03_cha02.indd 21
10/5/2010 5:01:47 PM
22
Meryl Kenny
The chapter evaluates these theoretical issues through an illustrative case study, candidate selection in post-devolution Scotland, and presents findings from a micro-level case study of a Scottish Labour Party constituency seat selection contest. While on the face of it the Scottish Parliament is often considered to be a ‘success story’ in terms of levels of women’s representation and the promotion of a more gender-balanced politics, evidence from the case study suggests that the reality is far more complex. Findings from the case study highlight ongoing tensions and contradictions in the political recruitment process in Scottish Labour post-devolution, while also drawing attention to an increasing gap between the formal rules of candidate selection and the implementation, enactment, and enforcement of these rules on the ground. The chapter concludes with a brief discussion of the implications of the Scottish case for the FI theoretical project more generally.
Gender and political recruitment: Evaluating the supply and demand model Norris and Lovenduski’s (1995) supply and demand model is the dominant framework used in studies of political recruitment (see also Norris and Lovenduski 1993; Randall 1987). Supply-side factors such as resources – time, money, experience – and motivational factors – ambition, interest, confidence – influence who decides to run for office. On the demand side, selectors evaluate applicants in accordance with a wide range of factors, including both formal and informal criteria. The effect of gender on political recruitment can be seen as both a supply-side and demand-side effect. Due to wider systemic factors such as the public– private divide, the sexual segregation of the work force, and patterns of gender socialization, women are likely to have less time, money, ambition, and confidence than their male counterparts. Alternatively, the effect of gender on the selection process can be a product of demand, either through direct or indirect discrimination (Norris and Lovenduski 1995: 106–108). While the supply and demand model offers a compelling framework for the study of political recruitment, it has two key shortcomings. First, it underestimates the extent to which gender norms shape the dynamics of supply and demand. The supply and demand model represents a significant improvement of previous work on political recruitment, in that it highlights the role that gender norms play in the selection process. However, Norris and Lovenduski view gender as only one of many factors influencing the dynamics of supply and demand. On the one
9780230_245884_03_cha02.indd 22
10/5/2010 5:01:47 PM
Gender and Institutions of Political Recruitment
23
hand, they find some evidence of both direct and imputed discrimination in the selection process. Yet they ultimately conclude that there is little evidence of pervasive discrimination against women candidate applicants. These findings sit at odds with both previous and later work on political recruitment, including their own, which highlights the structural barriers constraining potential female candidates and the discriminatory effect of masculinist norms which pervade the candidate selection process (Chapman 1993; Lovenduski and Norris 1989; Shepherd-Robinson and Lovenduski 2002; Lovenduski 2005b). For example, in their earlier work, Lovenduski and Norris are generally sceptical of supply-side explanations for women’s political under-representation, stressing the extent to which gender norms shape the selection process, favouring the stereotype of a middle-class, white, male, professional candidate: Arguably, both the way in which the role of candidate is defined and the candidate qualities sought tend to penalize many women. The system has been designed to select a standard model candidate who is articulate, well-educated, and typically employed in a professional career. (Lovenduski and Norris 1989: 559) Ongoing empirical work on political recruitment also highlights the limiting power of demand, despite increases in the supply of female candidates. Shepherd-Robinson and Lovenduski’s (2002) Fawcett Society report following the 2001 General Election found that British political parties were institutionally sexist, meaning that gender bias was entrenched and well established in the parties in terms of ‘personnel, outcome, and practices’ (Lovenduski 2005b: 53). The report found that the key factor explaining low levels of women’s political representation was a lack of demand on the part of candidate selectors and that incidences of both direct and indirect discrimination were widespread in all of the parties. Other studies highlight the extent to which seemingly neutral recruitment and selection practices are, in fact, gendered (see for example Chapman 1993). Second, while the supply and demand model attempts to theorize the interconnections between the institutions of political recruitment, representing a significant improvement from previous approaches to studying candidate selection, disaggregating factors in this way oversimplifies the dynamics of the selection process. In its later uses, the supply and demand model is presented as a straightforward interaction model, assuming a linear and logical relationship between the institutions of
9780230_245884_03_cha02.indd 23
10/5/2010 5:01:47 PM
24
Meryl Kenny
political recruitment (see in particular Norris 1997). However, studies of political recruitment reveal multiple directions of causality (Krook 2009). For example, the demands of gatekeepers predominantly shape the supply of available candidates, as candidates often self-select themselves out of the recruitment process on the basis of their perception of selection criteria: a demand-side factor (see for example Fox and Lawless 2004). Moreover, while the supply and demand model acknowledges that there are institutional interconnections, it fails to explicitly theorize the effect of institutional configurations, in other words, how the interaction of particular institutions may help or block the selection of women political candidates (Krook 2009). The supply and demand model assumes that the institutions of political recruitment work together relatively straightforwardly in cases of institutional reform (see Norris 1997). Changes in the political system – for example, changes to the electoral system – impact the selection process within political parties. The resulting changes in the recruitment process, then, impact the dynamics of candidate supply and selector demand.
Beyond supply and demand: Towards a feminist institutionalist model? The chapter contends that an FI approach offers a way forward for research on gender and political recruitment. First, an FI approach would establish gender as a central dimension of the institutions of political recruitment. Feminist theoretical and empirical work on gender and institutions suggests that gender relations are cross-cutting, that they play out in different ways in different types of institutions and on different institutional levels, ranging from the symbolic level to the ‘seemingly trivial’ level of interpersonal, day-to-day interaction, where the continuous performance of gender takes place (Kenney 1996: 458; see also Acker 1990, 1992; Connell 1987, 2002; Lovenduski 1998). Gender, then, cannot be conceptualized as simply a ‘factor’ that affects the dynamics of supply and demand, manifested at the individual level through, for example, direct or imputed discrimination by party gatekeepers. Rather, gender needs to be integrated throughout; these gendered interactions take place within a framework of formal and informal party rules and practices which are shaped and structured by masculinist gender norms (see for example Chapman 1993; Lovenduski and Norris 1989), as well as a wider context of systematic and structural barriers which have a differential effect on men and women as institutional actors.
9780230_245884_03_cha02.indd 24
10/5/2010 5:01:47 PM
Gender and Institutions of Political Recruitment
25
Second, an FI approach offers the tools to systematically theorize the complex, dynamic, and often contradictory ways in which the institutions of political recruitment interact with each other at different stages of the process. Insights from HI offer a more nuanced approach to theorizing institutional interconnections that moves beyond conceptions of supply and demand as a straightforward interaction model, highlighting the ways in which institutions are profoundly shaped by legacies of the past, as well as ongoing interactions with other institutions. In particular, HI – unlike the rational choice and sociological schools of NI – points to the need to take issues of timing, sequence, and context seriously, highlighting the ways in which particular configurations of institutions shift over time, promising and/or limiting possibilities for innovation and change (Hall and Soskice 2001; Pierson and Skocpol 2002; Waylen 2007, 2009; Weir 1992). In addition, recent work in the field offers new and more dynamic conceptions of institutional stability and change, such as institutional layering, which are especially useful for feminist work on gender and institutions (see in particular Schickler 2001; Thelen 2003, 2004). The intent, however, is not just to selectively borrow certain NI conceptual tools or to apply NI frameworks uncritically. The focus, then, is not only on what NI can contribute to FPS, but also how a feminist approach can enrich and transform NI theory (see Kenny 2007; Kenny and Mackay 2009; Mackay and Meier 2003). FPS and wider feminist social science provide rich and complex insights into gender and power, two concepts that are underdeveloped – if not virtually absent – in mainstream NI theory (Kenny 2007; Kenny and Mackay 2009). The relationship between NI and FPS, then, is one of mutual benefit. While NI can offer a number of mid-level conceptual tools to FPS, a feminist approach establishes gender as a crucial dimension of political institutions, while also yielding key insights into the dynamics of power and change.
Setting the context: Gender and political recruitment in post-devolution Scotland The following section evaluates the potential of an FI approach in the context of an illustrative case study – candidate selection in postdevolution Scotland. The new Scottish Parliament brought with it new expectations, opportunities, and problems with regards to the candidate selection process. For all of the Scottish political parties, a clean slate of candidacies with no incumbents, as well as a new, more proportional electoral system, opened up possibilities for innovation and change in
9780230_245884_03_cha02.indd 25
10/5/2010 5:01:47 PM
26
Meryl Kenny
established selection procedures. Women activists working inside and outside the parties were able to frame their demands within the context of these wider reform trajectories, successfully introducing a gendered perspective to the prospect of a ‘new’ Scottish politics, a more inclusive politics that was distinct from that of Westminster.1 Responding to these pressures, three of the four main political parties – Scottish Labour, the Scottish National Party (SNP), and the Scottish Liberal Democrats – accepted the principle of gender balance and all of the parties improved their recruitment and selection procedures, for example, incorporating equal-opportunities considerations into the process (Mackay 2003). In the run-up to the 1999 elections, Scottish Labour implemented strong equality guarantees, instituting a practice called ‘twinning’ in constituency seats, while its main opponent, the SNP, used proactive unofficial measures encouraging women to stand for office, for example, by holding a women’s training day for prospective parliamentary candidates, and also placing women at the upper end of their regional Party lists (Brown 2001b). The first elections for the Scottish Parliament were held on 6 May 1999, and a record number of female Members of the Scottish Parliament (MSPs) were elected, 48 out of 129 MSPs, or 37.2 per cent. In the subsequent 2003 elections, contrary to many expectations, a record and higher number of female MSPs were returned to the second parliament, 52 out of 129 MSPs, or 39.5 per cent (Mackay 2003). The results of the 1999 and 2003 elections established the Scottish Parliament as a world leader in terms of women’s representation and suggest that women’s political presence has become institutionalized to some degree. However, overall, indications are that progress remains fragile and contingent, and that previous gains have been the result of ‘accident rather than design’ (Mackay and Kenny 2007: 89; see also Mackay 2003, 2006). While the presence of women in the Scottish Parliament has, to some extent, become a ‘routinized’ or common-sense feature of post-devolution Scottish politics, the results of the recent May 2007 elections demonstrate that equality guarantees and positive-action mechanisms are far from embedded, and raise questions as to the future sustainability of women’s representation in Scottish politics. The 2007 elections resulted in the first drop in the number of female MSPs since the creation of the parliament. Women took 43 out of the 129 seats (33.3 per cent), an overall drop of eight women MSPs from the previous election. Analysis of the 2007 elections indicates clear gendered patterns of selection and recruitment. The general trend among the parties was one of either stasis or decline in the number
9780230_245884_03_cha02.indd 26
10/5/2010 5:01:47 PM
Gender and Institutions of Political Recruitment
27
of women elected (Mackay and Kenny 2007). None of the parties used strong equality guarantees, although some did implement softer measures, for example, Labour’s use of gender-balanced shortlists in constituency selections. While both Scottish Labour and the Green Party implemented quota-type mechanisms on the regional lists, in all of the parties male candidates dominated the top winnable places. Women were also generally less likely to be selected to fight safe and winnable constituency seats. Above all, the 2007 results demonstrate the continuing effects of Scottish Labour’s use of equality guarantees in 1999. Labour has consistently maintained a strong gender balance, ranging from 50 per cent women in 1999, to 56 per cent women in 2003, and finally back to 50 per cent women in 2007. Over half of the female MSPs returned to the Scottish Parliament in 2007 belong to Scottish Labour, the majority of whom were first selected in constituency seats in 1999 under the since-abandoned twinning scheme. Yet while most of Labour’s incumbent female MSPs have continued to successfully defend their seats election upon election, patterns of turnover within the Party suggest that there has been a ‘re-masculinization’ of Labour candidacies (Mackay and Kenny 2007). For example, of the seven Scottish Labour MSPs elected for the first time in 2007, only one was a woman. If those Scottish Labour MSPs returning to the Parliament after a term away are included, only two out of nine new Labour MSPs were female. The following section explores these gendered puzzles and patterns in more detail, providing an in-depth analysis of the candidate selection process within the Scottish Labour Party at the micro-level. It tells the ‘story’ of a selection, reconstructing the sequence of events in a Scottish Labour constituency seat selection contest in the run-up to the 2007 Scottish Parliament elections.
The story of a selection: Tracing the process The micro-level case study – in this case, the constituency seat of City North and Greenside2 – was selected for a number of reasons. Historically, City North and Greenside is a very safe Labour seat and, until the 2007 elections, had been held by Scottish Labour since 1999 by a consistent margin of over 6,000 votes. The seat was one of only four open Labour seats in the 2007 elections and, as a result, a large and relatively strong field of potential candidates declared interest in standing for selection. Finally, it was a seat in which a sitting Labour woman MSP – first selected under the twinning scheme in 1999 – was standing
9780230_245884_03_cha02.indd 27
10/5/2010 5:01:47 PM
28
Meryl Kenny
down. The focus of the case study is on reconstructing a very specific set of events within a particular temporal context and spatial location, and is not intended to be representative of every candidate selection process within the Scottish Labour Party. Rather, the aim is to draw out the implications of the case study for the wider FI theoretical project. The City North and Greenside constituency selection contest was triggered when the sitting female incumbent MSP decided to stand down before the 2007 elections, first personally notifying the Constituency Labour Party (CLP) Chair and then making a formal announcement at a CLP meeting. The announcement came very late in the election season – several months after the retiring MSP had already been re-selected – and as a result, the CLP had a limited period of time in which to select a candidate. In accordance with Labour Party procedure, the central Scottish Party – based at John Smith House in Glasgow – set a formal timetable for the process lasting 11 weeks, a standard time frame for constituency selections. Due to the fact that a female MSP was retiring, there was some discussion within the central Scottish Party as to whether steps should be taken to ensure that a woman candidate was selected through, for example, an all-women shortlist (AWS). The possibility of instituting an AWS in City North and Greenside was ultimately discounted, a decision that was framed, in part, as an issue of timing. As one of the CLP officers put it: ‘we couldn’t afford to end up in a long, drawn-out, contentious process when we had a small number of weeks to put a candidate in place.’3 Stage one: Declaration of interest and application process Once the timetable was established, aspiring candidates were given a two- to three-week period in which to declare their interest in standing for selection. Formally, there is little evidence of the central Scottish Party or the CLP taking active measures to recruit particular candidates within the constituency. In contrast to the highly centralized procedures used in 1999, the Scottish Party did not make an active effort to recruit women candidates for the seat in question, or to encourage them to run, although there is some evidence of active steps being taken to encourage certain ethnic-minority candidates to stand. For example, one of the ethnic-minority candidates reported being directly approached and encouraged to stand by the then-First Minister, Jack McConnell. All in all, 12 Scottish Labour members formally declared an interest in standing for selection, four of whom were women, and two of whom were ethnic minorities (both men). The aspiring candidates were from
9780230_245884_03_cha02.indd 28
10/5/2010 5:01:48 PM
Gender and Institutions of Political Recruitment
29
a range of different backgrounds: the majority with local government experience, while others were activists, community leaders, or active in the local Party organization. Most were local Party members, although several prominent applicants were from outside the constituency. Stage two: Branch and union nominations Once the application stage closed, the CLP held an informal ‘meet and greet’ event for the candidates open to all Party members, combining both a brief five-minute presentation for each candidate and an informal meeting in which candidates could circulate with members. Two weeks later, the Party branches, trade unions, and relevant affiliates met to formally decide on candidate nominations.4 In the case of City North and Greenside there were five Party branches involved in nominations as well as five trade union branches. Per Labour Party rules, local branches are required to put forward one man and one woman in their nominations. Of the 12 original candidate applicants, five received branch nominations, two of which were women. All but one of the five branch-nominated candidates also received trade union nominations. The trade unions also nominated two additional applicants, one of whom was an ethnic-minority candidate applicant. The ethnic-minority candidate applicant, a former member of the Scottish Executive Committee (SEC), was also endorsed by the local Party executive. As per Labour Party rules, the local Party executive had to endorse one ethnic-minority candidate to go forward for shortlisting. Stage three: Shortlisting Candidate applicants with branch and/or union nominations then went forward to a meeting of the CLP’s executive. At this stage of the selection process, the Executive Committee (EC) made several key decisions. First, it decided how many candidates it was going to shortlist, and second, it made recommendations as to whether the Party should implement any equality measures to ensure that either female or ethnic-minority candidates were included on the final shortlist. Labour Party rules require that the candidate shortlist be ‘genderbalanced’, meaning that there must be an equal number of male and female candidates on the list. The EC can also make the decision as to whether to explicitly reserve a shortlist place for an ethnic-minority candidate. In the case of City North and Greenside, the EC’s recommendation was to implement a gender-balanced shortlist of four. This recommendation
9780230_245884_03_cha02.indd 29
10/5/2010 5:01:48 PM
30
Meryl Kenny
was put to the full General Committee (GC) of the CLP, which includes approximately 40 members, about 30 of whom were branch delegates, and the remainder of whom are delegates from trade unions or other organizational affiliates. The GC endorsed the EC’s recommendation with regards to how the shortlist should be made up, and finally, voted on which of the nominated candidates it wanted to include in that shortlist. As the final shortlist of four candidates had to be genderbalanced, this left five male nominated candidate applicants competing for two spots on the final shortlist. As a result, there were mixed feelings within the CLP as to whether the male ethnic-minority candidate should be automatically shortlisted as well. As one of the other male candidates put it: Given that it had to be gender-balanced, the final shortlist kind of scuppered the rest of us, for Pete’s sake. If one of the ethnic candidates had got through, there would only be one male left in the last four.5 Ultimately, the GC decided not to automatically shortlist the ethnicminority candidate and the meeting then went to a vote, using Single Transferable Vote (STV). From the shortlist, two of the candidates stood out as front-runners, having received a majority of the branch nominations. The ‘front-running’ female candidate, a high-profile woman from outside the CLP with substantial political experience, had received the most branch nominations, while the ‘front-running’ male candidate, an active Party official, was seen to be popular within the local Party. The remaining candidates had both received one branch nomination each. The ‘second-ranked’ male candidate was a former councillor from the town of Greenside, while the ‘secondranked’ female candidate was a trade union office-holder and women’s activist from a neighbouring CLP. Stage four: Canvassing and campaigning Following the shortlisting, candidates were allowed to begin formal campaigning. Candidates were allowed to distribute a canvassing leaflet, which could be mailed to all local Party members eligible to participate in the ballot, and could also be accompanied by a single-page letter asking for support. Candidates could also approach members personally or by telephone. While Party rules specifically prohibit formal campaigning until after shortlisting, in reality, campaigning was already well underway from the very beginning of the selection contest,
9780230_245884_03_cha02.indd 30
10/5/2010 5:01:48 PM
Gender and Institutions of Political Recruitment
31
with many of the candidates already canvassing for votes informally throughout the CLP. As one candidate stated: The rules were acknowledged and ignored. As long as you don’t do anything really blatant. You just had to be careful how you went about it. [One of the candidates] just went walking about ... and just happened to pass the doors of Party members and chat to say hello. I did most of mine by telephone.6 While there was a general understanding amongst the candidates that there would be canvassing outside of the formal rules, several additional rule violations occurred throughout the selection process. A few days before the Greenside branch nomination meeting, a local MP distributed a constituency report in which they explicitly endorsed the ‘second-ranked’ male candidate, the former councillor from Greenside. Additionally, the head of the Greenside branch distributed a letter to all members of the Greenside BLP, on official headed paper, endorsing the same candidate immediately prior to the final selection meeting. Both interventions were in violation of the Party’s formal rules and were reported to both central and local Party officials. While no action was taken by the central Scottish Party on the violation by the local MP, the central Party did issue a formal statement regarding the Greenside branch letter at the final selection meeting, instructing Party members that they should not be ‘under the impression that they are mandated how to vote tonight and ... should discard the contents of the letter’.7 Stage five: Final hustings The final selection meeting was open to the full local Party membership, which stands at approximately 230 members eligible to vote. Party members at the meeting voted according to STV, ranking candidates in order of preference. In the first round of voting, the ‘front-running’ male and female candidates came in second and third, respectively, while the other male candidate received the majority of the first-round votes, a third of which were postal votes. The other female candidate received the lowest number of votes, and her votes were distributed according to second preferences to the other candidates. In the second round, the ‘front-running’ female candidate was also eliminated. Ultimately, the final vote was won by the male local councillor from Greenside, who won more than half of the votes, while the ‘front-running’ male candidate came in second. The two female candidates came in at third and fourth place.
9780230_245884_03_cha02.indd 31
10/5/2010 5:01:48 PM
32
Meryl Kenny
Many of the candidates and Party members were surprised by the results on the night. The City media had largely presented the contest as a ‘two-horse race’ between the front-running male and female candidates, depicting the contest as a choice between a high-profile versus a local candidate: ‘it will boil down to [the female candidate’s] high profile and her ability to appeal across a wide spectrum of voters against [the male candidate’s] knowledge and understanding of the politics of the area’.8 Others present throughout the selection process saw a slightly different picture, highlighting the divisions between the two sides of the constituency, and the Greenside ‘dynamic’ running through the selection contest. While the town of Greenside is a small part of the Scottish Parliament electorate, it has the biggest Labour Party membership in the City North and Greenside constituency seat, and historically, has tended to dominate the selection process. Greenside was seen within the constituency as a very traditional Labour Party, in which loyalty and local politics played a large role: ‘[Greenside’s] quite a traditional Labour Party in that key players pretty much say this is how we’re going to vote, we’re going to support our local man and everybody just gets on board and does it’.9 Stage six: Election While the male Greenside candidate was successfully selected, he ultimately lost the seat for Scottish Labour in the 2007 Scottish Parliament elections. To some, this was seen as an extension of a surprising selection in a divided constituency, in which Party activists from the City North wing of the CLP were reluctant to support a candidate that the majority of them had not voted for. Activist turnout was extremely low in the elections, and Party members expressed difficulty in ‘getting the candidate known’ on the City North side of the constituency.10 The election results were also influenced by wider issues within the Scottish Labour Party, reflected in an increasing decline in Labour Party membership, an overall loss of Labour seats, and a rise in support for the SNP. In particular, the election results highlight the tension between selectability and electability in the Scottish Labour candidate selection process. Several of the candidate applicants highlighted the tension between the repeated emphasis on selecting a ‘local’ candidate, versus the need to appeal to, and also represent a wider electorate: Local does matter. But you’re also selecting somebody who’s going to be part of a national parliament. You’re not only there as a local
9780230_245884_03_cha02.indd 32
10/5/2010 5:01:48 PM
Gender and Institutions of Political Recruitment
33
representative. You will be making legislation and policy for the whole of Scotland. And every group within that. So I think that ... people are very bad at looking at the bigger picture.11 These issues were exacerbated by a sense of complacency on the part of the CLP and the central Scottish Party as to City North and Greenside’s status as a ‘safe’ seat. As the CLP was going through the selection process, the SNP was already mobilized and campaigning within the constituency, having selected a high-profile, well-funded candidate at a much earlier stage. Due to the late timing of the City North and Greenside selection contest, Scottish Labour was already well behind the SNP’s campaign. In addition, Labour Party activists within the CLP were poorly organized. Some attributed the lack of organization to the ‘old-fashioned’ and traditional nature of the City North and Greenside CLP: What struck me about the constituency was that it was a very oldfashioned constituency in terms of its politics ... . New Labour was really not around. It was a creaking constituency in the sense that the level of campaigning was frankly abysmal.12 The lack of organization and activist turnout within the CLP can be linked, in part, to a wider trend of declining Labour Party membership in Scotland. While declining membership is a problem for British and Scottish political parties more generally, Scottish Labour, in particular, has experienced a significant drop in Party membership in recent years. In the case of City North and Greenside, this overall drop in membership is reflected in a return to the core ‘old guard’ of the CLP Party membership: You have a situation now where ... you’re back to your core old folk that are eighty plus ... . I mean guys that were old when I joined – when I joined the Party in [Greenside], I was just twenty-three, turning twenty-four – and these folk are still around and make up the bulk of the membership.13 Finally, the election results in City North and Greenside reflect a wider electoral swing in favour of the SNP, who won 20 seats more than it had in 2003, ending over 50 years of Labour dominance of Scottish politics. Following the election, media coverage of the City North and Greenside electoral result was generally subsumed within the wider media coverage
9780230_245884_03_cha02.indd 33
10/5/2010 5:01:48 PM
34
Meryl Kenny
of Labour losses and SNP victories. There appears to be no specific coverage of Labour’s electoral loss in City North and Greenside, and any general coverage of electoral results mentioned the seat only in passing, providing few additional details.
Assessing the case study: A feminist institutionalist approach At first glance, findings from the case study reinforce trends within the mainstream literature on post-devolution Scotland, which argues that the Scottish Labour Party can be characterized by an increasing degree of territorial autonomy in the area of candidate selection (Hopkin and Bradbury 2006; Laffin and Shaw 2007; Laffin et al. 2007). However, the case study highlights an overall trend of decentralization in the candidate selection process that goes substantially beyond the ‘revised structure of power’ highlighted by devolution scholars (Hopkin and Bradbury 2006: 149). Running parallel to this is an apparent trend of informalization, in which there seems to be an overall drift from, as well as lack of enforcement of, the formal rules, regulations, and procedures of the selection process. It is clear, then, that there is more ‘going on’ in the institutions of political recruitment in post-devolution Scotland than the mainstream literature seems to suggest. The following section explores these issues in more depth and draws upon the insights of both NI and feminist theory to provide a more thorough interpretation and analysis of the microlevel case study. From centralization to decentralization Findings from the case study suggest that tensions between central intervention and local control are ongoing, despite declarations to the contrary within the mainstream literature (see for example Mitchell and Bradbury 2004). As has been well documented elsewhere, the original candidate selection procedures put into place by Scottish Labour in 1999 were highly centralized (see for example Bradbury et al. 2000a, 2000b; Shaw 2001). The official objectives of the centralization of the candidate selection process were to ‘raise candidate calibre’, while also formally enforcing strong equality guarantees to increase the number of women MSPs (Laffin and Shaw 2007). Pressures for increased central control created tensions in the Labour Party, where decentralized constituency selection had long been the standard practice (Denver 1988; see also Mitchell and Bradbury 2004).
9780230_245884_03_cha02.indd 34
10/5/2010 5:01:48 PM
Gender and Institutions of Political Recruitment
35
While different factions of the Party were generally supportive of the move to reform candidate selection procedures, there were concerns that these changes in fact masked ‘more pragmatic power-related concerns’, selecting candidates who would ‘support the leadership and toe the party line’ (Bradbury et al. 2000b: 152). Much of the academic literature on political recruitment in post-devolution Scotland continues to focus on the controversies of the 1999 selection process, alternately describing the process as ‘control-freakery’, or as an ‘ideological cull’ seeking to consolidate New Labour support within the Scottish Party14 (Hopkin and Bradbury 2006; Laffin and Shaw 2007; Laffin et al. 2007). However, the mainstream literature’s continued focus on the alleged ‘control-freakery’ of the 1999 candidate selection process misses out on important ‘processes unfolding on the periphery’ (Thelen 2004: 295; see also Clemens 1993; Orren and Skowronek 1994; Weir 1992). Institutions do not just generate positive feedback, but also ‘generate grievances’ (Schneiberg and Clemens 2006: 218). In moments of institutional change, the ‘losers’ do not necessarily disappear: When institutions are founded they are not universally embraced or straightforwardly ‘adapted to’, but rather continue to be the object of ongoing conflict, as actors struggle over the form that these institutions should take and the functions they should perform. (Thelen 2004: 32) In the case of post-devolution Scotland, those who ‘lost out’ in the highly centralized process in 1999 frequently appealed to principles of decentralization, attacking the ‘centralization of control’ in the process, which resulted in a ‘restriction rather than a broadening of democracy’ (Bradbury 2000b: 158; see also Jones 2001). However, evidence from the City North and Greenside case highlights a distinct shift in internal Party debates surrounding centralization and decentralization. CLP members and candidate applicants frequently reframed the issue of central Party intervention in terms of locals versus outsiders: ‘the last thing we would want is somebody from the outside’.15 The ongoing decentralization of the candidate selection process, then, can be seen not only as a gradual trend of inter-Party devolution, but rather as an active process of institutional resistance to central Party control. Evidence from the case study also suggests that this process of decentralization has a distinctly gendered dimension. Feminist research, in particular, has exposed the extent of gendered resistance in established institutions, providing ‘clear evidence of powerful, pervasive, and
9780230_245884_03_cha02.indd 35
10/5/2010 5:01:49 PM
36
Meryl Kenny
specific resistance to positive policies for women’ (Halford 1992: 160; emphasis in original). Throughout the City North and Greenside selection process, Party members consistently reframed the issue of gender equality in terms of local versus central control. For example, as several CLP members explained, the issue was not about ‘women’ candidates specifically, rather, the problem was the central imposition of ‘outsider’ candidates. As one CLP officer stated: If we take the subject of an all-women shortlist, there’s an almost hostility to it, on most occasions when the topic comes up ... . It’s less about opposition to the idea in principle. It’s much more about creating a situation in which we end up with candidates and members from outside.16 This repeated linkage of gender equality with ‘imposed intervention’ by the central Party then positions women candidates as outsiders to the process, and correspondingly, positions men as ‘local’, or insiders to the process. One of the shortlisted female candidates, who was a member of a neighbouring CLP, described her experience as an ‘outsider’ at a candidate event in City North and Greenside: One of my kind of pitches was that what I brought to the constituency was people, really, from different movements all over Scotland who would come and help ... And you could feel a wee ripple going through a section of the room. And it was – I hadn’t meant it that way at all – but afterwards a couple of people said to me they could just feel the people next to them getting all prickly about this outsider apparently suggesting that what we needed was a lot of women and gay people coming in, when they could run things themselves, thank you.17 The underlying ‘local versus outsider’ discourse disadvantaged male candidates as well as female candidates, positioning certain types of political masculinities in opposition to the ideal type of the loyal, local male candidate. Almost all of the candidates were relatively ‘local’, living within the wider City area, if not within the specific City North boundaries. ‘Localness’, then, became a matter of degrees: ‘local only means some things when it suits people’.18 One of the unsuccessful male candidates reflected on the process: ‘I was an outsider because I lived more than 500 yards from the edge of the constituency. It’s the local champion thing. It counts for a lot.’19 Both male and female
9780230_245884_03_cha02.indd 36
10/5/2010 5:01:49 PM
Gender and Institutions of Political Recruitment
37
‘outsider’ candidates perceived this tension in clearly gendered terms, as a ‘favourite-son selection’.20 As one of the female shortlisted candidates reflected: I think there was all this loyalty thing, this male loyalty ... You see, this is the trouble with men. Men are always in favour of a woman being selected, so long as it’s not them that’s got to stand aside for a woman, or to say well, we want to support a woman.21 From formalization to informalization Evidence from the case study also draws attention to an overall trend of informalization in the candidate selection process, highlighting the extent to which the selection process increasingly operates in accordance with informal rules, norms, and understandings of ‘how things are done’, for example, with regards to candidate canvassing and campaigning. This trend of informalization in candidate selection and recruitment appears to be compounded by the overall lack of intervention into the process by the central Scottish Party and the inconsistent and uneven enforcement of formal rules by both central and local Party officials. Institutions require ‘active maintenance’, they need to be continually ‘reset and refocused, or sometimes more fundamentally recalibrated and renegotiated’ (Streeck and Thelen 2005: 24). This is particularly true of ‘new’ or ‘young’ institutions which: ... require elaboration of their meaning in practice, by a sequence of decisions on the part of rule makers as well as rule takers. The ‘path’ along which an institution is ‘worked out’ in this sense is shaped by exogenous circumstances as well as a myriad of strategic choices, deciding together which of the many possible meanings of a young institution are practically explored and which are foreclosed or left by the wayside. (Streeck and Thelen 2005: 30) The failure, then, to actively ‘maintain’ an institution, to enforce particular rules, regulations, conventions, or practices, can essentially ‘amount to actively allowing it to decay’ (Streeck and Thelen 2005: 25). Findings from the case study draw attention to the active cultivation of drift and erosion in the institutions of political recruitment, highlighting repeated incidences of rule-breaking as well as a general lack of rule enforcement on the part of key Party actors. Formal and informal complaints regarding rule violations were made both during and after the
9780230_245884_03_cha02.indd 37
10/5/2010 5:01:49 PM
38
Meryl Kenny
selection process by candidate applicants as well as other Party members, but central Party officials do not appear to have taken any direct action on these complaints in most cases. One candidate applicant, for example, contacted the central Scottish Party about a rule violation involving another candidate, who as a CLP officer had direct access to the membership lists before they were officially available by virtue of his position within the CLP: One of the issues we had early on was that [he] started canvassing before the lists were officially available. Because he was [a CLP officer], he had the lists anyway. And I said to him that I thought that was an inappropriate use of the lists. And it is in breach of the rules, absolutely in breach of the rules, and he just laughed ... and I said something to somebody at the Scottish Labour Party office and they’re just like, well, it’s a contest.22 Evidence from the case study again suggests that this trend of informalization has a distinctly gendered dimension. A key criterion of the success of institutional innovation is the extent to which reforms have become ‘common-sense’, creating new ‘routinized, taken-for-granted patterns of behaviour’ (Inhetveen 1999: 406). Yet, feminist research on gender and institutions demonstrates the particular difficulties of institutionalizing and maintaining gender-equality reforms, highlighting the constant possibility for norm erosion, drift, and reversal. As Mackay writes: ‘resistance is entrenched, and without vigilance, gender equality issues can slip off the agenda’ (2006: 19). For example, evidence from the case study highlights an increasing informalization of candidate selection criteria, and points to a general reassertion of traditional gendered norms of an ‘ideal candidate’. Despite efforts to rewrite and open up candidate job descriptions in 1999, findings from the case study indicate that these were not actively used in the selection process. Job descriptions were not distributed to the members of the CLP, and participants in the selection process commented on the lack of agreement or discussion among Party selectors as to what a ‘good’ candidate was, both in a general sense, and in a particular sense for the constituency in question: I mean, I don’t know what they were looking for. Nobody tells you what they’re looking for ... . I mean, the reverse of that is nobody gives you an advert. Nobody tells you. So the CLP ... they didn’t ever have a discussion even on the Executive, even in a smaller group.
9780230_245884_03_cha02.indd 38
10/5/2010 5:01:49 PM
Gender and Institutions of Political Recruitment
39
They never had a discussion about what they were looking for in an MSP. 23 Following the lack of maintenance and enforcement of existing, formalized selection criteria, findings from the case study suggest that there has been a drift back to the gendered model of the ideal candidate – the ‘local man’. While there was no explicit use of enforcement of formal selection criteria, interviews with both candidate applicants and selectors suggest an implicit sense of what selectors were looking ‘for’. All of the candidates highlighting the perceived importance throughout the process of being ‘local’: ‘Political experience was totally irrelevant. It was all due to local factors.’24 In addition, both candidate applicants and CLP members continuously referred to the selection process in gendered terms, repeatedly highlighting the importance of ‘male loyalty’, ‘favourite sons’, and ‘support for the local man’. As one of the female candidates put it: I’ve said ... to some people, you wonder if you were a man with the experience I’ve had. You know, if you look at my CV, it’s pretty impressive. And yet they don’t select you ... But at the end of the day it was all loyalty kind of stuff to the man who was there.25
Conclusion Academic commentators increasingly question the success of the ‘new politics’ in Scotland, highlighting the extent to which post-devolution political practices resemble those of Westminster and critiquing the ‘unrealistic’ hopes and expectations of devolution campaigners (see for example Arter 2004; McGarvey 2001; Mitchell 2004). Findings from the case study, however, suggest that the reality is far more complex, highlighting evidence of innovation and change in the institutions of political recruitment, while also drawing attention to underlying continuities. As such, the Scottish case points to a need to rethink conventional models of political recruitment, highlighting the complex and gendered dynamics of institutional design, continuity, and change within political institutions, and illustrating the difficulties of reforming and redesigning the institutions of political recruitment in the face of powerful institutional and gendered legacies. The combined insights of feminist and HI theory, then, may offer a way forward for research on gender and political recruitment. For example, work in the HI field offers critical insights into the bounded nature
9780230_245884_03_cha02.indd 39
10/5/2010 5:01:49 PM
40
Meryl Kenny
and contradictory outcomes of institutional innovation in existing systems, highlighting the ways in which ‘new’ (or partially reformed) institutions continued to be shaped by past legacies, as well as initial and ongoing interactions with already existing institutions within which they are ‘nested’ (Goodin 1996; Mackay 2009: 3; Pierson 2004). The Scottish Parliament is not a blank slate, it is a ‘new’ institution framed within a pre-existing institutional and cultural context, nested within an existing system in which many of the key elements of the Westminster model have survived (Mackay 2006; see also Chappell in this volume). Findings from the case study suggest that the ‘success’ of institutional innovation in candidate selection and recruitment is a complex and contingent question, and that elements of the ‘old’ continue to uneasily co-exist with elements of the ‘new’, constraining and shaping each other. At the same time, FPS offers several critical insights which are often ignored in mainstream analyses. Evidence from the Scottish case demonstrates that the gendering and regendering of the institutions of political recruitment are ‘active processes with palpable effects’ (Hawkesworth 2003: 531), occurring through the actions of embodied institutional actors, as well as through institutional rules, regulations, norms, and standing operating procedures. These findings demonstrate not only that gender norms and gender relations are particularly ‘sticky’ institutional legacies with which to contend, but also that gender – at both the symbolic level as well as the level of day-to-day interaction – is a primary means through which institutional reform and innovation can be resisted (Mackay 2009: 26). This suggests that the dynamics of institutional power relations, resistance, reproduction, continuity, and change need to be filtered through a gendered lens (Mackay 2004: 113; see also Chappell 2006; Kenny 2007; Mackay and Meier 2003). FI, then, is a promising theoretical approach for future work on gender and political recruitment, one that has the potential to overcome the key limitations of existing work in the field.
Notes 1. The ‘story’ of women’s mobilization in the 50:50 campaign in Scotland has been extensively detailed elsewhere (see in particular Breitenbach and Mackay 2001; Brown 2001a, 2001b). 2. Names have been changed for reasons of anonymity. 3. Interview 3 with CLP officer, 17 March 2008. 4. This is not always standard practice. In some cases, the branch and union nomination stage takes place at the end of the informal ‘meet the candidates’ event. 5. Interview 4 with candidate applicant, 24 March 2008.
9780230_245884_03_cha02.indd 40
10/5/2010 5:01:49 PM
Gender and Institutions of Political Recruitment 6. 7. 8. 9. 10. 11. 12. 13. 14.
15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25.
41
Interview 8 with candidate applicant, 17 March 2008. Internal Party correspondence, 2006. CLP member quoted in City newspaper, 2006. Interview 8 with candidate applicant. Interview 8 with candidate applicant. Interview 7 with candidate applicant, 8 April 2008. Interview 6 with candidate applicant, 1 April 2008. Interview 6 with candidate applicant. Cutts et al. (2008) address similar dynamics at the UK level with regards to all-women shortlists (AWSs). In the aftermath of the 2005 General Election, the dominant frame was one of ‘AWS backlash’, with the controversial case of the Labour seat of Blaneau Gwent offering ‘proof’ that AWS candidates lose votes. Cutts, Childs, and Fieldhouse critique this frame, arguing that there is no evidence of any significant ‘AWS effect’ in 2005. Candidate applicant, quoted in Greenside newspaper, 2006. Interview 3 with CLP Officer. Interview 7 with candidate applicant. Interview 7 with candidate applicant. Interview 6 with candidate applicant. Interview 6 with candidate applicant. Interview 5 with candidate applicant. Interview 7 with candidate applicant. Interview 7 with candidate applicant. Interview 6 with candidate applicant. Interview 5 with candidate applicant.
9780230_245884_03_cha02.indd 41
10/5/2010 5:01:49 PM
3 Discursive Strategies for Institutional Reform: Gender Quotas in Sweden and France Lenita Freidenvall and Mona Lena Krook
Gender quotas have emerged in recent years as a key strategy for increasing women’s representation in electoral politics around the globe. Appearing today in more than 100 countries, these measures take several forms, including changes by individual political parties to their party statutes, as well as reforms initiated by national legislatures to revise constitutions and electoral laws (Dahlerup 2006b; Krook 2009). The shared goal is the nomination of more female candidates, most often to national parliaments. The spread of gender quotas to diverse contexts, as well as their mixed effects on the numbers of women elected, has inspired a large literature on quota policies. To understand quota adoption, scholars often begin with an analysis of quota debates, exploring how proposals are framed and justified in specific contexts. While advocates and opponents may vary across countries, a common approach is to map and unpack the content of these discourses, noting how actors articulate their positions in relation to existing political values (Freidenvall 2005; Holli et al. 2006; Krook et al. 2009). In many cases, these controversies lead to quota reform, but the apparent resolution of these debates does not ensure that quotas will be implemented fully. Rather, some reforms produce jumps, while others experience stagnation, or even decreases in the numbers of women elected. These patterns suggest that while discursive struggles may play a crucial role in enabling quota reform, they are constrained in their ability to enact change. This raises two sets of questions. First, what shapes the discourses employed in quota debates? Why are some more effective than others? Second, what explains the limits of discursive strategies? How do elements of the broader institutional context shape quota 42
9780230_245884_04_cha03.indd 42
10/5/2010 5:01:55 PM
Discursive Strategies for Institutional Reform 43
implementation? In an attempt to answer these questions, this chapter combines feminism and discursive institutionalism (DI) to analyse the origins and effects of quota reform. DI differs from other types of institutionalism: it places greater emphasis on the role of ideas and discourses in influencing actor interests, preferences, and behaviours (Campbell and Pedersen 2001), and as a consequence, it offers a more dynamic theory of institutional change by calling attention to the role of discourse in generating and legitimizing ideas about political action (Schmidt 2008). From a feminist perspective, however, there are elements missing from the DI framework, namely an explicit focus on gender, power, and institutional configurations. A combined approach thus presents an opportunity to better theorize complex patterns of political stability and change, at the same time that it helps explain puzzles surrounding quota reform. The chapter begins with an overview of research on gender quotas, focusing on how scholars have analysed quota adoption through the lens of discourses, and quota implementation through attention to the role of political institutions. It then outlines the key features of DI and addresses its potential and limits for feminist analysis. On the basis of this evaluation, it proposes a synthesized approach – a ‘feminist discursive institutionalism’ (FDI) – that adds gender, power, and institutional configurations to the existing repertoire of concepts used in DI, namely institutions, discourses, and change. The final part of the chapter applies this approach to analyse the origins and outcomes of quotas in two countries, Sweden and France. Largely similar quota policies were adopted in both countries but produced quite different results at the national level, despite striking parallels in the two countries’ debates. To unravel these effects, the analysis maps the institutional environment prior to quota campaigns to explore how these shape the content and resonance of discourses used in quota debates, and in turn, how these discourses succeed or not in changing existing institutional configurational dynamics.
Gender quotas and candidate selection The rapid diffusion of gender quotas raises several empirical puzzles. First, methods of positive action are controversial in many contexts, but these measures have gained the support of parties and legislatures in countries with a diverse range of characteristics. Second, quotas for women in elected positions require a reduction in the proportion of men, but most policies are passed by largely male-dominated parties
9780230_245884_04_cha03.indd 43
10/5/2010 5:01:56 PM
44
Lenita Freidenvall and Mona Lena Krook
and legislatures. Third, quotas are criticized on the grounds that they reify ‘women’ as a group, but have been embraced by many groups as a new basis for women’s organizing. Focusing on a diverse range of actors and motivations for quota reform, a growing number of studies explain quota adoption with reference to discursive framing, arguing that how quotas are defined and discussed has crucial implications for their approval. Once these measures are passed, however, research on quota implementation tends to turn its attention to the interactions between these measures and existing political institutions. Campaigns for quota adoption: Discourses and change Case studies of quota campaigns point to a number of actors and motivations behind quota reform. One explanation is that women’s groups support quotas because they believe that women in politics are necessary to achieve justice, promote women’s interests, and make use of women’s resources for the good of society (Phillips 1995). Noting that there does not appear to be a ‘natural’ trend towards change, they recognize that any break with the status quo is likely to be achieved only through specific, targeted actions to promote female candidates (Krook 2006a). A second argument is that political elites pursue quotas for strategic reasons, either in an effort to attract female voters (Meier 2004), or as a means to maintain control over political rivals (Baldez 2004). A third possibility is that quotas gain support among a variety of actors when they align with reigning political norms. In some countries, party quotas in left-wing parties are attributed to the match between quotas and party ideologies embracing the goal of social equality (Opello 2006). In others, quotas are framed as an extension of guarantees given to other groups (Meier 2000). A final explanation is that quotas are supported by global norms, as international organizations have embraced the goal of 30 per cent women in decision-making, and transnational networks have facilitated the exchange of information about effective quota strategies across national borders (Krook 2006b). Seeking to understand how actor support is cultivated and transformed into political action, an expanding literature analyses the content of quota debates. A key concern is to identify the discourses employed in arguments for and against quota policies. Discourses refer to language, concepts, and categories used to frame a particular issue (Bacchi 1999). Quotas are controversial for many reasons. Three common conflicts arise over competing ideals of equality, representation, and gender (Freidenvall et al. 2006; Krook et al. 2009). Positions on equality oppose equal opportunities, attributing responsibility for
9780230_245884_04_cha03.indd 44
10/5/2010 5:01:56 PM
Discursive Strategies for Institutional Reform 45
inequalities to individuals, and equal results, recognizing that unequal outcomes derive from broader structures (Dahlerup 2007). Meanings of political representation vary in terms of emphasis on ideas versus identities: principal-agent notions view it as when one person acts on behalf of another, while descriptive conceptions see it as when office-holders reflect the broader population’s physical, social, or other ascribed characteristics (Phillips 1995). Notions of gender, finally, concern the unity of the category ‘women’ and its status as a political identity on a par with other recognized groups. Quotas for women are most consistent with arguments about equal results, descriptive representation, and gender as a salient political category. For this reason, they are vulnerable to objections placing priority on equal opportunities, principal-agent notions of representation, and diversity among women, and the precedence of other identities. One of the most effective critiques is that quotas discriminate against men, and thus violate the constitution and sex-equality law (Guadagnini 2005; Mossuz-Lavau 1998). These concerns intersect with discourses about qualifications, with opponents suggesting that quotas undermine ‘merit’ in candidate selection, promoting ‘unqualified’ individuals to positions who otherwise would not have made it there ‘on their own’ (Murray 2008). Advocates often respond by questioning these assumptions and asserting that quotas are likely to facilitate renewal by electing more women, bringing new values and backgrounds to political deliberation (Franceschet and Piscopo 2008). They thus link these questions to political representation. Other controversies revolve around whether women constitute a group deserving representation like identities based on class, race, language, and other categories (Jenkins 1999). The presence of these and other competing discourses are not only important at moments of quota adoption. They can also shape processes of implementation if doubts linger as to their legitimacy and legality, or if their outward acceptance masks ongoing resistance to quota reform (Bird 2003; Holli et al. 2006; Meier 2008). Dynamics of quota implementation: Institutions and reforms Despite some attention to discourses in influencing quota implementation, the majority of research focuses on interactions between these policies and various types of political institutions. These insights are rooted in the three explanations often given for variations in quota effects. The first is that quota impact is linked to their details, like type, wording, requirements, sanctions, and legitimacy (Meier 2004; Murray 2004). The second is that impact depends on features of the
9780230_245884_04_cha03.indd 45
10/5/2010 5:01:56 PM
46
Lenita Freidenvall and Mona Lena Krook
electoral and party systems in which they are introduced, with greater possibilities in countries using proportional representation (Tremblay 2008), and parties in which leaders are able to enforce party or national regulations (Caul 1999). The third is that impact reflects the balance among actors who support and oppose quota policies, including elites, women’s groups, courts, and ordinary citizens (Baldez 2004). Viewed alongside the literature on women’s legislative representation, this work suggests that effects of quotas are shaped by rules, practices, and norms, which can be conceptualized as three broad categories of institutions: systemic, practical, and normative (Krook 2009). The advantage in using the term ‘institution’ to refer to these dynamics is that it confers roughly equivalent causal status, rather than privileging one of these effects over the other a priori. Systemic institutions constitute the formal features of political systems. Their gendered effects are among the most studied in the field of women in politics. One of the most important factors explaining cross-national variations in women’s representation, as well as the impact of quota policies, is the electoral system (Matland 2006; Tremblay 2008). While the role of systemic institutions is most obvious in the second account outlined above related to electoral and party systems, it is also evident in the other two explanations. The first points to the importance of oversight bodies responsible for reviewing quota requirements and procedures for rectifying non-compliance, while the third identifies national and international courts as crucial arenas for challenging non-compliance and requiring parties to redo candidate lists that do not comply with a given quota law. Practical institutions refer to the formal and informal practices shaping the criteria and procedures that parties employ to select their candidates. These influence perceptions regarding who is deemed to be a ‘qualified’ candidate (Norris and Lovenduski 1995) and are related to methods of ballot composition (Caul 1999). The first story reflects the importance of selection practices, drawing attention to the degree to which quotas bind the hands of elites by reducing ambiguities in quota requirements, establishing placement mandates, and imposing sanctions for non-compliance. The second and third highlight the role of political will, suggesting that elite support or opposition is crucial in determining how quota policies translate into reality. Normative institutions, finally, embody formal and informal principles such as equality and representation. These can be gendered to the extent that they justify differences among women and men, despite rhetoric of neutrality and inclusion (Okin 1979). The first explanation
9780230_245884_04_cha03.indd 46
10/5/2010 5:01:56 PM
Discursive Strategies for Institutional Reform 47
above observes that effective implementation often hinges on the perceived legitimacy of a particular quota policy. The second and third are more indirect, but nonetheless allude to the values and commitments that inform the actions taken by various types of political actors with regard to quota reform. The interwoven nature of these explanations, in turn, suggests that ascertaining reasons behind the success and failure of quota policies is a task that requires attending to the ways in which these measures alter and interact with existing rules, practices, and norms (Krook 2007a).
Feminism and discursive institutionalism The existing literature on quota policies thus addresses questions of quota adoption and implementation, respectively, by focusing largely on the discourses that frame quota demands and the institutions that mediate quota effectiveness. On their own, these trends suggest that the tools of DI may hold promise for developing more integrated accounts of quota reform. Taken a step further, drawing on central tenets of DI also presents an opportunity to explore in greater depth two further puzzles: the varied forms taken by discourses and the differential impact of quotas across national contexts. However, the gendered nature of quota reforms requires a key modification to the DI framework, namely the incorporation of feminist insights related to gender, power, and institutional configurations. The resulting synthesis offers ‘added value’ over the application of feminist and institutionalist concepts on their own, calling attention to – and compensating for – gaps in each framework taken on its own. Discursive institutionalism: A feminist evaluation For the last decade, new institutionalism (NI) has been broadly understood to encompass three major approaches: rational choice, historical, and sociological (Hall and Taylor 1996; Immergut 1998). In recent years, however, DI has emerged as a fourth alternative, seeking to bring in a focus on ideas and discourse to better understand mechanisms of institutional emergence, survival, and reform (Campbell and Pedersen 2001; Schmidt 2008). DI is similar to the three other forms of NI in its focus on a broad definition of institutions. However, DI differs in its emphasis on discourse and its attempts to offer a theory of institutional change. These features make DI a particularly attractive option for feminists, many of whom share an interest in how language, concepts, and categories shape the content and conduct of politics, and who seek to
9780230_245884_04_cha03.indd 47
10/5/2010 5:01:56 PM
48 Lenita Freidenvall and Mona Lena Krook
study and develop strategies for altering the status quo. Yet, in order to fit feminist goals, key insights need to be integrated from the feminist toolkit. The defining feature of NI, as noted elsewhere in this volume, is its attempt to broaden the scope of ‘institutions’ to include both formal and informal institutions (March and Olsen 1989; North 1990). Researchers explore why formal institutions often do not have the effects that their creators intended (Pierson 2004) and conversely, how and why informal ones have the effects that they do (North 1990). Recent work among NI scholars also analyses how formal and informal institutions work together to shape political outcomes (Helmke and Levitsky 2004). Some feminists develop this point further by drawing on configurative methods (Ragin 2000) to more explicitly theorize how institutions, both formal and informal, work together as part of institutional configurations. The intuition is that the effects of one institution are shaped by the presence or absence of others (Krook 2009, forthcoming). The concept of ‘discourse,’ in turn, has many meanings. DI scholars themselves offer slightly different definitions, but converge on discourses as a medium for conveying ideas. For Vivien Schmidt (2008), ideas are the substantive content of discourse and occur at three main levels: policies, programmes, and philosophies. Discourse refers to the interactive process of conveying ideas and serves to communicate dominant perspectives and taken-for-granted assumptions. However, it can also provide a means for expressing and communicating new worldviews, thereby facilitating the process of political change. The first intuition, related to the structuring nature of discourse, fits closely in line with how discourse is understood by many discourse analysts, reflecting upon the ways in which the terms of a discourse can constrain what can be talked about (cf. Foucault 1971; Potter and Wetherell 1987). The second, proposing that discourse can be strategically employed to articulate new ideas, constitutes a departure from many versions, which reject intentionalism and rationality in how discourses are expressed. Consequently, the concept of ‘discourse’ as used by DI scholars does not simply import a consensus definition from discourse analysis, but rather seeks to add and expand what discourses can ‘do’ in political life. Although many feminists adopt more traditional approaches (Stapleton and Wilson 2004), some develop frameworks that also depart from conventional understandings. Bacchi (1999), for example, studies the construction of policy problems. She argues that the ways of framing a problem affect how policy-makers think about how that problem should be resolved. Uncovering these assumptions requires analysing
9780230_245884_04_cha03.indd 48
10/5/2010 5:01:56 PM
Discursive Strategies for Institutional Reform 49
competing interpretations to ascertain what effects follow from them, what is left unproblematic, and the ways in which identities are constructed through discourses. The goals of the problem-representation approach, however, go beyond simply documenting the structure of policy discourses. They also entail focusing on the ways in which power operates through discourse to fix certain constructions of gender relations as dominant and to marginalize or exclude counter-discourses. In other words, issues of gender and power are central to the articulation and effects of discourse. The third element of a DI approach, a focus on ‘change’, is a major departure from other schools of NI, which focus primarily on questions of stability (Pierson 2000a; Thelen and Steinmo 1992). To account for instances where change does occur, NI scholars present two models of innovation: one centred on critical junctures and path dependence (Mahoney 2000; Pierson 2000a), and the other criticizing strict separation between institutional creation and reproduction (Greif and Laitin 2004; Pierson 2004; Thelen 2004). For DI scholars, these options do not exhaust possible sources of change. Discourse may play a role in promoting political action as ‘policy makers, political leaders, and the public are persuaded, or not, of the cognitive necessity and normative appropriateness of ideas’ (Schmidt 2006: 11). Shifts in worldviews may also occur via conflicts over meaning and processes of translation (Lynggaard 2007). Feminist scholars and activists share a similar interest in possibilities for change (Driscoll and Krook 2009). Feminist discursive institutionalism: Concepts and innovations The discussion above suggests that there are many points of contact among feminism and DI in terms of attention to institutions, discourses, and change. At the same time, critiques voiced by feminists indicate that it is not sufficient to simply import the insights of DI into feminist research (Kulawik 2009). Rather, these tools must be modified to match feminist commitments by incorporating attention to three additional components: gender, power, and institutional configurations. The advantage of combining feminism and DI is that the revised approach can help better theorize how institutions and prospects for institutional reform shape, and are shaped by, dynamics of inequality and the broader institutional environment. As such, its benefits are not restricted to feminists, but could also be employed by mainstream NI scholars interested in understanding complex patterns of institutional stability and change.
9780230_245884_04_cha03.indd 49
10/5/2010 5:01:56 PM
50
Lenita Freidenvall and Mona Lena Krook
The concept of ‘gender’ is perhaps the key contribution of feminist research. Although often elided with ‘women’, feminists distinguish between ‘sex’, biological differences between women and men, and ‘gender’, the social meanings given to these distinctions. For this reason, Joan Wallach Scott describes gender as ‘a constitutive element of social relationships based on perceived differences between the sexes’ (Scott 1988: 42). While gender is a central category, feminists disagree as to what it is: some argue that it is something that people ‘have’ and others frame it as something that people ‘do’ (Connell 1987). Still others describe it as an institution, seeking to capture the nature and social origins of gendered practices and ideas (Martin 2004). Attention to gender in feminist analysis thus intersects in many ways with institutions and power, but is not entirely reducible to either phenomenon. Related to these discussions, a second major axis of an FDI is a focus on ‘power’. As discussed in relation to Bacchi (1999), feminists are aware of how pervasive relations of power are in shaping dynamics of political life. Hegemonic discourses can restrict the issues and interpretations that are possible and desirable to talk about. Attending to questions of power can expose the potential for non-hegemonic discourses, recognizing nonetheless that efforts may face an uphill battle given entrenched interests in maintaining the status quo. This focus provides an important corrective to optimistic beliefs about the potential for alternative discourses to enact change. An emphasis on power also draws attention to the fact that all subject positions are not equal: some confer greater capacity than others to articulate, or mute, new discourses. As such, marginalized groups may be less able to effect change, despite innovations in discursive frames. Finally, an FDI would also highlight the importance of ‘institutional configurations’. As noted above, NI researchers often allude to a range of different institutions. Yet, many restrict their focus to a single institution to understand its origins and effects (Pierson 2004). Adopting such an approach, however, misses an opportunity to explore how various kinds of institutions fit together as part of a ‘complicated ecology of interconnected rules’ (March and Olsen 1989: 170). This is especially important when the effects of one institution may depend on the shape of other institutions (Krook 2009). Although this insight is not restricted to feminists, attention to questions of gender is more closely attuned to interactions. Thus, viewed together, a combination of feminist and DI tools offers prolific prospects for understanding the complexities inherent in the study of institutional effects, reform, and change.
9780230_245884_04_cha03.indd 50
10/5/2010 5:01:56 PM
Discursive Strategies for Institutional Reform 51
Institutions, discourses, and gender quota reform Existing research suggests that discursive strategies are crucial to the passage of quota reforms, at the same time that institutional contexts play a decisive role in shaping the impact of quota policies. A synthesized FDI approach thus presents an opportunity to link moments of adoption and implementation, and better grasp the shape and impact of discourses put forward to promote reform. To illustrate the benefits of a combined approach, the analysis examines and compares debates and outcomes in two countries, Sweden and France, where framing processes and reforms are relatively similar, but where there are differences in impact. These parallels and divergences can be explained through attention to the inspiration behind the particular discursive strategies employed in each case, as well as shifts and stability reflected in reigning institutional configurations. This account thus considers (1) how institutional contexts shape discourses; and (2) how discourses alter some, but possibly not all, parts of the broader institutional environment. Introducing ladies’ choice: Varannan damernas in Sweden In Sweden, the word ‘quota’ has a difficult history due to norms of equality, consensus, and cooperation, eventually giving way to the concept of varannan damernas (‘every other one for the ladies’). When implemented, this new policy approach interacted in favourable ways with existing institutions, leading to the election of 47 per cent women to parliament in 2006. The first quotas were proposed in the 1920s, soon after women gained the right to vote, when the Social Democratic Party’s Women’s Federation suggested quotas so that women might be placed in safe seats on the Party’s lists. The proposal was rejected on the grounds that all positions should be based on equal opportunity and subject to open competition (Karlsson 1996). Over the next 50 years women continued to mobilize inside the parties, but generally succeeded in only getting one woman – often described as the ‘obligatory woman’ – on electoral slates, most often at the ‘decorative places’ at the bottom of the lists. By the 1970s, however, Sweden witnessed the ‘sex role debate’, focused on changing both women’s and men’s roles in society. It inspired new efforts to get more women included on party lists, but this time focused on gaining equal – not just token – representation for women in politics. A discursive strategy over the course of the 1980s and 1990s was to argue that women were not a ‘minority interest’ but the ‘majority of the citizenry’ and merited equal inclusion
9780230_245884_04_cha03.indd 51
10/5/2010 5:01:56 PM
52
Lenita Freidenvall and Mona Lena Krook
on those terms (Sainsbury 2004). As a result, policies proposed for ensuring women’s representation radicalized over time, moving from recommendations to targets, introduced by a number of parties from across the political spectrum (Freidenvall 2006). However, these measures stopped short of formal quotas, given the desire to avoid ‘special treatment’ at the expense of men. While the numbers of women in elected office continued to increase, reaching 32 per cent by 1985, the proportion of women in state committees continued to be low, leading the government to convene a commission on women’s under-representation in this sector. In an important shift, the authors of the final report rejected the view that advances were the result of ‘natural development’ that would continue until equality was achieved. They stopped short of calling for ‘quotas’, however, and proposed voluntary measures with mandatory provisions to be introduced only if women’s numbers did not rise (Varannan damernas 1987). To offer guidance, the report proposed a system of alternating between men’s and women’s names, a system it dubbed varannan damernas. Referring to a countryside dance custom where every other song was women’s turn to invite the men, the term literally meant ‘every other one for the ladies’. While functionally equivalent to a quota, this concept appeared to soften demands for a radical redistribution of power by alluding to more sympathetic rules of ‘taking turns’. As such, it avoided any reference to conflicts of interest between women and men, enabling calls for equal representation to resonate with dominant discourses of ‘consensus’ and ‘equality’ in Swedish society (Bergqvist 1994; Freidenvall et al. 2006). Although the report was not intended to apply to electoral politics when it was published in 1987, it gained renewed visibility in 1991 when women’s representation dropped for the first time since 1928. The election created shockwaves, because many had believed the upward trend to be irreversible. Women’s groups responded by exerting pressure within their parties, as well as by becoming involved in a crossparty network known as Stödstrumporna (the ‘Support Stockings’) that threatened to form a women’s party if established parties did not elect more women. In response, the Social Democratic Party announced in 1993 that it would apply the principle of varannan damernas on all lists for all elections. Similar policies were soon adopted in nearly all other parties, some opting for a policy of strict alternation, and others preferring softer formulations that achieved the same effect through policies of ‘at least 40 per cent of either sex’. The result was the election of 41 per cent women in 1994, 43 per cent in 1998, 45 per cent in 2002, and 47
9780230_245884_04_cha03.indd 52
10/5/2010 5:01:56 PM
Discursive Strategies for Institutional Reform 53
per cent in 2006, approximating the demand for equal representation (Inter-Parliamentary Union 2009). Before these debates, Sweden was characterized by a mix of institutions more and less favourable to women’s representation. Electoral rules involved proportional representation (PR) with high district magnitudes and closed party lists. Scholars consistently associate all three rules with higher numbers of women in politics (Caul 1999). Party practices were also encouraging: many parties felt obligated to include women, although varied greatly in their commitment and the tools applied to ensure the selection of female candidates. In general, the parties with the most binding measures were those with the highest proportions of women in parliament (Wängnerud 2001). Norms were perhaps the most contested. Women were initially considered ‘represented’ as long as one was selected. Over time, this definition expanded as women moved from being considered a ‘special interest’ to being recognized as ‘one half of the citizenry’. Yet, women’s groups continued to struggle with the term ‘quota’, which many framed as ‘undemocratic’ and implying the selection of ‘unqualified’ women. These issues were played out through discourses of ‘consensus’ and ‘equality’ that undermined attempts to enact strict quotas (Freidenvall 2005). When women’s numbers declined in 1991, however, the notion of varannan damernas enabled a shift in thinking, drawing on the very discourses that had previously undermined quota adoption. As electoral rules entailed closed-list PR, alternation in selection practices improved opportunities for women to be placed in higher spots on party lists. These changes influenced a reformulation of norms, transforming calls for ‘proportionate’ representation into demands for the election of equal numbers of women and men. Lingering ambivalence over the term ‘quota’, nonetheless, reflected continued normative resistance to positive action that might create unfair advantages. The reference to a dance-floor tradition, however, enabled supporters to frame a 50 per cent quota as a measure aimed at the equal division and sharing of political power, a more radical claim that was, all the same, more consistent with reigning political norms and discourses. Thus, while quota adoption was aimed mainly at reforming party selection practices, debates over these policies intersected in various ways with existing rules, norms, and discourses. In this case, the content of institutions shifted in mutually reinforcing ways, leading to nearly equal numbers of women and men in the Swedish Parliament.
9780230_245884_04_cha03.indd 53
10/5/2010 5:01:57 PM
54 Lenita Freidenvall and Mona Lena Krook
Sexing the universal citizen: Parité in France In France, the word ‘quota’ had similarly negative connotations, related to the strong emphasis on universalism in French conceptions of citizenship. After several failed attempts to institute quotas, supporters developed a universalist critique based on the concept of parité (‘parity’). When enacted, parity interacted with existing institutions, leading to the election of 12 per cent women to parliament in 2002 and 18 per cent in 2007 (Inter-Parliamentary Union 2009). Proposals for quotas in France first emerged in the 1970s, when women inside the Socialist Party (PS) began to demand special measures to promote women’s representation. Advocates argued that quotas were consistent with party ideology and would demonstrate commitment to achieving gender equality. In 1974, the PS approved a 10 per cent quota in Party leadership positions and for elections by PR, a proportion that increased in 1977 and 1979. Further proposals to raise the quota to 30 per cent were not passed until 1990, however, and it was not until 1996 that this quota was extended to elections run by majoritarian vote (Appleton and Mazur 1993; Opello 2006). As these debates took place, several female ministers and members of parliament pursued quotas for women in local politics. In 1982, they succeeded in passing a law as part of a bill on local government limiting to 75 per cent the proportion of candidates of the same sex. Deputies voted nearly unanimously in favour of the measure, but several months later the Constitutional Council declared the quota article unconstitutional on the grounds that it violated the principle of ‘equality before the law’, prohibiting the ‘division’ of voters and candidates into ‘categories’ for the purpose of voting. This verdict reaffirmed a principle of ‘equal opportunities’ that precluded attempts to institute ‘equal outcomes’. Nonetheless, by the end of the 1980s this conception came under scrutiny, as scholars began to attribute women’s absence in politics to the principle of the ‘universal citizen’ (Fraisse 1989). This led to a manifesto, Au pouvoir citoyennes! Liberté, égalité, parité, which proposed changing the law to state that elected assemblies be composed of as many women as men (Gaspard et al. 1992). While the concept of parité emerged in discussions in the Council of Europe, it gained momentum in France as a means to get beyond the impasse of ‘quotas’ (Krook 2007b). In the 1990s, supporters began to argue that reigning understandings of ‘equality’ and ‘representation’ – as well as their subject, the ‘universal citizen’ – were originally deemed to apply only to men. They suggested that inscribing parité in
9780230_245884_04_cha03.indd 54
10/5/2010 5:01:57 PM
Discursive Strategies for Institutional Reform 55
the constitution was the only way to recognize the ‘two sexes of the universal citizen’, a reform that was crucial to general welfare, because ‘sex’ was the universal difference cutting across all other groups, categories, and communities (Agacinski 2001). They emphasized that parité differed from ‘quotas’, because while quotas implied special representation rights, parity called for equitable sharing of power between women and men. Inscribing parité in the constitution thus would not reverse the accomplishments of the French Revolution, but rather would fully realize them (Scott 2005). Following public debates, the constitution was amended in 1999 and the electoral law reformed in 2000. Yet the final reforms disappointed many advocates, as ‘parity’ was replaced by ‘equal access’, while the verb ‘guarantees’ was substituted with ‘favours’. The more radical claim for equal representation was thus reduced to the milder goal of increasing the number of female candidates (Bird 2003). The limits of these reforms were apparent in divergent outcomes across levels of government: while women’s representation increased dramatically from 26 per cent to 48 per cent in local councils in 2001, in the National Assembly it inched from 11 per cent to 12 per cent in 2002, and then to 18 per cent in 2007, due largely to decisions by the major parties not to implement the law (Murray 2004). These variations stem from the fact that lists not complying at the local level were not eligible to contest the election, while those that did not comply at the national level were subject to a financial penalty that could be overcome by winning more seats (Krook 2009; Sineau 2001). In the years leading up to the parity debates, most institutions of candidate selection were largely disadvantageous to women’s representation. The electoral rules blocked major gains, as the two-round majoritarian system enabled parties to exclude women. However, this pattern was also true under PR: when applied in 1986, women comprised nearly 25 per cent of candidates, but were placed in less than 6 per cent of electable positions (Mossuz-Lavau 1998: 24–25). Similarly, only the PS adopted a quota, and in that case, its commitment was weak, excluding legislative elections. In spite of these considerable barriers, norms again were the most problematic. Legal decisions in the 1980s reinforced conceptions of ‘equality’ and ‘representation’ that undermined attempts to pursue ‘quotas’ at the statutory level, which were discursively construed as being a threat to the values of French republicanism. Indeed, ‘quotas’ were described as distinctively ‘un-French’ – even disparagingly as ‘American’ – measures that would not improve women’s status, but that would instead escalate claim-making by other groups, and subsequently
9780230_245884_04_cha03.indd 55
10/5/2010 5:01:57 PM
56
Lenita Freidenvall and Mona Lena Krook
erode the cultural assimilation at the heart of the French Republic (Millard and Ortiz 1998; Scott 1998). In this context, advocates employed the discourses of ‘universalism’ and ‘equality’ that had blocked quota implementation to gain the adoption of a 50 per cent quota law. Interestingly, the need to work within prevailing discourses enabled – and, indeed, required – supporters to press for more radical reforms, because these could be framed as more consistent with existing values. Whereas quotas had been rejected in the 1980s on the grounds that they ‘divided’ citizens into ‘categories’, a focus on parité presented an opportunity to frame women not as a ‘special interest’, but rather as one of the ‘two halves of the human race’. However, the campaign’s concern with norms and discourses was not matched by similar success in shifting rules and practices of candidate selection. In fact, during debates over reform, all major parties opposed change to the electoral system. At the same time, legislators devised very loose regulations for implementation in legislative elections, imposing no placement requirements and a weak financial penalty for parties who did not conform to the law. As a result, the adoption of quotas in France achieved shifts in discourses and formal norms, but made little impact on the rules and practices shaping women’s access to political office.
Gender, discourses, and institutional change Gender quotas constitute one of the widest-reaching electoral reforms of recent years. The origins and outcomes of these policies, however, are not yet well understood. Existing research on these questions indicates that quota policies are largely the result of discursive strategies for political reform. At the same time, however, these measures do not have uniform effects: some lead to increases, while others produce stagnation and even decreases in the numbers of women elected. Seeking to contribute to these discussions, this chapter sought to combine the tools of feminism and DI to better theorize the shape of the discourses employed in quota debates, as well as the role of existing institutional contexts in facilitating – or hindering – the effects of quota policies. The synthesis links together the focus on discourses, institutions, and change from DI with ideas about gender, power, and institutional configurations from feminist research. This revised framework was then applied to a comparison of the cases of Sweden and France, where similar policies produced quite distinct results, despite striking parallels in the quota debates. The analysis reveals the value of engaging
9780230_245884_04_cha03.indd 56
10/5/2010 5:01:57 PM
Discursive Strategies for Institutional Reform 57
in a dialogue across feminism and DI, as well as more generally across work on gender, discourses, and political institutions. It suggests that there is space to develop a ‘feminist institutionalism’ that builds on shared concerns, but also imports insights from each approach taken on its own in order to tackle emerging puzzles in political dynamics witnessed around the globe.
9780230_245884_04_cha03.indd 57
10/5/2010 5:01:57 PM
4 Gendered Institutions and Women’s Substantive Representation: Female Legislators in Argentina and Chile Susan Franceschet
Women have made impressive gains worldwide in terms of their access to elected office, a development that has led to two distinct research agendas. The first investigates the factors that shape women’s participation in electoral politics and the second explores whether women’s presence makes a difference, particularly in terms of advancing women’s rights. This chapter combines the main contributions of historical institutionalism (HI) and feminist political science (FPS) to show how a ‘gendered institutionalist’ approach fills in the lacunae in the existing literature. The chapter is divided into three sections. The first provides a critical overview of the existing literature on women’s substantive representation, highlighting some of the gaps in our knowledge about whether, why, and how female legislators ‘act for’ women. The second section outlines the features shared by HI and FPS that address these gaps. The third section applies a ‘gendered institutionalist’ (GI) framework to the contrasting cases of Argentina and Chile, showing how formal and informal institutions structure the legislative process in ways that influence the actions of female legislators and the likelihood that women’s actions will succeed or fail.
Studying women’s substantive representation Over the last few decades, a vast literature has emerged that explores both the attitudes and actions of female office-holders. Much of this research asks whether women in politics contribute to women’s 58
9780230_245884_05_cha04.indd 58
10/5/2010 5:02:05 PM
Gendered Institutions and Substantive Representation 59
substantive representation. Other researchers ask whether female legislators differ from their male counterparts in terms of their attitudes or legislative activity. Yet, even researchers who seek merely to uncover sex differences frequently infer that if men and women display different attitudes or behaviours, then this is an indication of women’s substantive representation (see Schwindt-Bayer 2006). Alternatively, researchers who find few sex-related differences in attitudes often conclude that electing more women will likely not have an impact on women’s substantive representation (Htun and Power 2006; Tremblay 1998). The conflation of sex-based differences among legislators and women’s substantive representation is problematic for at least two reasons. First, the concept of women’s substantive representation is inherently complex and impervious to straightforward definition.1 Second, taking sex differences as an indicator of substantive representation is problematic because ‘sex as a simple, dichotomous variable will distort unless it is located in a gendered frame of reference’ (Lovenduski 1998: 339). Third, even when narrowing the focus to one site of representation – the legislature – women’s substantive representation occurs at multiple stages of the legislative process (Mackay 2008; Tamerius 1995). Women’s interests are promoted when legislators introduce women’s rights bills, when they try to mobilize support for these bills (either behind-the-scenes, or publicly in committee and chamber debates), and when women’s rights laws are adopted. Women’s substantive representation can thus be conceived as both ‘process’ and ‘outcome’ (Franceschet and Piscopo 2008). A number of existing studies, however, focus on either the legislative process, or policy outcomes, while drawing conclusions about women’s substantive representation as a whole. Studies that focus on process, in particular those that compare the role of men and women in initiating women’s rights bills, tend to find that female legislators are more likely than their male colleagues to take action (Jones 1997; Schwindt-Bayer 2006; Taylor-Robinson and Heath 2003). But while these studies show that women’s presence influences the legislative agenda, they tell us little about legislative outcomes. Without looking at the outcomes of women’s actions we miss out on a fuller account of how the gendered dynamics within legislative institutions affect different aspects of women’s substantive representation. This is important because research that focuses on legislative outcomes, whether defined as women’s empowerment within legislative bodies or the adoption of women’s rights laws, tends to be more pessimistic about the kind of difference that women make (Goetz 2003; Marx et al. 2007; Vincent 2004). Researchers have found that even when
9780230_245884_05_cha04.indd 59
10/5/2010 5:02:05 PM
60 Susan Franceschet
the proportion of women in the legislature grows, political practices and policy outcomes may not change. However, studies that explore both women’s actions (bill introduction) and their consequences (the adoption of women’s rights laws) demonstrate that the two aspects of women’s substantive representation can occur separately (Franceschet and Piscopo 2008). Similarly, Childs distinguishes between the ‘feminization of the political agenda (where women’s concerns and perspectives are articulated) and a feminization of legislation (where output has been transformed)’ (Childs 2006: 9; see also Lovenduski 2005b). A second problem is that scholars use different methods to measure women’s substantive representation (see Mackay 2004; 2008). Scholars have variously measured substantive representation through surveys of legislator attitudes (Htun and Power 2006; Schwindt-Bayer 2006), through bill introduction (Swers 2002; Taylor-Robinson and Heath 2003), or through voting records (Vega and Firestone 1995). Although these research strategies likely reflect pragmatic decisions by scholars seeking to simplify a highly complex process, they can nonetheless lead to mistaken assumptions that a multi-faceted concept such as women’s substantive representation can be measured by a single indicator. Selecting a single indicator implies a relatively straightforward relationship among attitudes, legislator behaviour, and policy outcomes, whereas in reality, any such relationship is highly contingent. Legislators may hold favourable attitudes toward women’s rights and prioritize gender issues, but might not act on these attitudes by introducing or supporting women’s rights bills (Childs 2004). Even when attitudes are translated into action, these actions may fail (Franceschet and Piscopo 2008). As such, women’s substantive representation is best conceived as a multi-faceted and dynamic process that includes both legislative activities2 and outcomes.3 A conceptual framework focusing on both process and outcome has at least two advantages. First, it allows a more precise analysis of women’s roles in promoting women’s substantive representation and of the gendered environment in which they act. Although existing studies that focus on process (such as bill introduction) are explicitly addressing women’s roles, studies of policy outcomes may be missing out on the role of women and the gendered environment in bringing about these outcomes. A focus on both process and outcomes permits a firmer grasp of the gendered institutional dynamics in legislatures that either facilitate or obstruct favourable policy outcomes. Distinguishing between process and outcome also draws attention to a broader array of factors that affect women’s substantive representation.
9780230_245884_05_cha04.indd 60
10/5/2010 5:02:05 PM
Gendered Institutions and Substantive Representation 61
Existing studies have examined individual-level variables such as feminist orientation and party affiliation, finding that feminist attitudes and membership in a left party are correlated with ‘acting for’ women (Htun and Power 2006; Tremblay and Pelletier 2000; Vega and Firestone 1995). Scholars examining ‘critical mass’ (the proportion of women in a legislative body) have tested whether higher proportions of women increase legislative activity on behalf of women (Grey 2002), and also whether the increased descriptive representation of women produces more gender-friendly outcomes (Bratton and Ray 2002). Partly because this latter group of studies has produced such divergent findings, critical-mass theories have come under increased scrutiny in recent years. There seems to be an emerging consensus that a focus on numbers alone is problematic (Beckwith 2007; Childs and Krook 2006b; Dahlerup 2006a; Mackay 2008; Trimble 2006). Other scholars have expanded the research agenda to investigate contextual factors such as party dynamics, women’s seniority in the legislature (for example, committee leadership positions), and the influence of women’s movements and public opinion (Beckwith and CowellMeyers 2007; Dodson 2006; Swers 2002). While these studies expand the research focus and permit a more multi-faceted understanding of how and why women’s substantive representation occurs, none draws specifically on institutionalist theories of politics to analyse women’s substantive representation. Moreover, with the exception of Beckwith and Cowell-Meyers’ (2007) work, none uses cross-national comparisons, although there are a growing number of edited volumes that shed light on the possible reasons for cross-national variation in patterns of women’s substantive representation (Chappell and Hill 2006; Sawer et al. 2006). The lack of an explicitly comparative and institutionalist focus inhibits the development of a broader conceptual framework for analysing how gendered institutions shape women’s substantive representation (see Waylen 2009).
Gender and historical institutionalism Institutionalists have defined institutions as ‘relatively enduring collection[s] of rules and organized practices’ that are ‘embedded in structures of meaning and resources’ (March and Olsen 2005: 1). Historical institutionalists (HI) explore both formal organizations as well as the ‘informal rules and procedures that structure conduct’ (Thelen and Steinmo 1992: 4). Informal institutions are ‘socially shared rules, usually unwritten, that are created, communicated, and enforced outside of
9780230_245884_05_cha04.indd 61
10/5/2010 5:02:06 PM
62 Susan Franceschet
officially sanctioned channels’ (Helmke and Levitsky 2006: 5). A focus on informal institutions is something shared by gender scholars and is crucial to uncovering the gendered dynamics in legislative politics. Many of the norms that govern parliaments, such as the preference for an adversarial style or bargaining within ‘old boys networks’, are informal. Importantly, it is the informal nature of these rules that makes them more resistant to change. Informal rules are unwritten and not sanctioned through official channels. Instead, non-compliance might involve career costs such as less prestigious appointments or being shut out of informal power networks. Feminist scholars have highlighted the ways in which gender shapes and is shaped by institutions (Acker 1992; Chappell 2006; Lovenduski 1998, 2005; Waylen 2007). According to Beckwith, ‘gender as a process is manifested as the differential effects of apparently gender-neutral structures and policies upon women and men, and upon masculine and/or feminine actors’ (2005: 132). Gender scholars have noted that institutions, created by men in response to the interests of elite men, often function to constrain the behaviour of women (and, of course, other ‘outsiders’). But feminists have also explored the way that women’s movements and strategically placed ‘insiders’ or ‘femocrats’ can exploit institutional openings and even re-orient institutions and practices to feminist goals (Chappell 2002b). As Beckwith argues, ‘gender as process suggests not only that institutions and politics are gendered but also that they can be gendered’, that is, women inside and outside the state may succeed in ‘recast[ing] the gendered nature of the political’ (2005: 133; emphasis in original). Using the definitions of institutions and the insights of FPS outlined above, the institutions that structure and gender the legislative environment, and influence both legislator behaviour and policy outcomes, include things like electoral rules, party discipline, candidate nomination procedures, executive–legislative relations (especially the formal rules that define the authority and scope of action of each branch). Informal institutions include representative–constituent relations; internal and cross-party dynamics, for example, how parties distribute committee assignments among their legislative caucuses and how participation in floor debates is allocated; and norms of appropriate parliamentary behaviour, for example, whether and how a combative or consensual style is encouraged. In defining institutions broadly and placing them in their historical and gendered context, a GI approach has much to offer. Specifically, there are four aspects of this approach that enhance studies of women’s
9780230_245884_05_cha04.indd 62
10/5/2010 5:02:06 PM
Gendered Institutions and Substantive Representation 63
substantive representation: (1) problematization of preference formation, (2) a focus on the gendered power asymmetries embedded in institutions, (3) a focus on ideas and gendered norms, and (4) research methods that are more contextually sensitive. Problematizing preferences Both feminist and HI scholars problematize preference formation, rejecting the assumption that political interests are automatically given by one’s social location or economic position. For HI, institutions shape interests because they embody a society’s norms and values, thereby legitimating the expression of preferences that reflect those values. Political institutions thus reflect the political values held by the society in which they exist, or at least the dominant sectors of society who played a role in constructing those institutions. For example, political systems with extensive checks and balances respond to norms about the concentration of power.4 Likewise, the very names given to institutions express the values they embody. For example, in recent years most Latin American countries have created parliamentary committees to address gender issues. These committees often have names that signify whether women’s autonomy, or their family roles are to be prioritized. In Chile, the parliamentary committee created to address women’s issues was named the Family Committee, whereas in Mexico it is called the Equity and Gender Committee. In the former case, the committee’s name implies that women’s rights ought to be consistent with that society’s valuing of the family. In the second case, committee members are serving on a committee that is more explicitly geared toward equity promotion, thus not constraining the preferences that can be legitimately expressed on the committee to those consistent with a traditional view of the family. In this way, institutions ‘affect the very identities, self-images, and preferences of the actors’ (Hall and Taylor 1996: 939), thereby defining what it means to be a member of a particular institution, for example, a parliamentarian, a judge, or a bureaucrat. These norms, in turn, define appropriate conduct for individuals in different parts of the state and in society. For HI, what is most important is that the meanings inscribed in particular institutions are always contextual, and therefore what it means to be a legislator differs not only cross-nationally, but also across time. Institutions are not static, but are themselves subject to change brought about by new processes and actors’ collective responses to them. Indeed, having parliamentary committees dealing with gender issues is a recent development. The very creation of these committees
9780230_245884_05_cha04.indd 63
10/5/2010 5:02:06 PM
64 Susan Franceschet
signals that these issues have been accorded a legitimacy that had been denied in earlier historical periods, a good example of how political spaces can be re-gendered. In addition to a concern with preference formation, feminists and HI explore how preferences are translated into action (or inaction) and how political outcomes are derived from these actions. A GI approach thereby provides the causal links that are missing in much of the existing scholarship on women’s substantive representation, while also drawing attention to the multi-faceted and dynamic processes that constitute substantive representation. As noted earlier, some of the existing literature assumes a relatively straightforward relationship between legislator attitudes and substantive representation. A GI approach, on the other hand, shows how institutions mediate the relationship between individual attitudes and actions. According to Immergut, ‘institutionalist theory aims to expose and analyse the discrepancy between “potential” interests and those that are expressed in political behaviour’ (1998: 7). There are a number of ways in which legislative institutions encourage or discourage the articulation of perspectives that advance women’s rights. Women’s caucuses can create space for female legislators to work together to promote issues of concern to women (Archenti and Johnson 2006). The level of legislator activity (the quantity of bills introduced) is also shaped by institutional factors, namely, the human and technical resources provided to legislators. Where legislators have resources to hire highly skilled assistants, they can introduce a greater quantity of bills than in places where legislators do not have such resources. Institutions also determine what kind of bills legislators can introduce. For example, executives in some countries, whether due to formal or informal rules, enjoy exclusive rights to legislate in certain areas, thereby reducing the scope of activity for legislators. A GI approach to studying women’s substantive representation thereby avoids the voluntarism that is implicit in some accounts of substantive representation. Some studies of women in legislatures imply that if female legislators do not ‘act for’ women it is because they do not hold feminist attitudes or do not prioritize women’s rights policies. In contrast, a GI approach assumes that individual action does not spring directly from attitudes and preferences but instead, ‘individuals turn to established routines or familiar patterns of behaviour to attain their purposes’ (Hall and Taylor 1996: 939). Prevailing norms and patterns of behaviour can provide either opportunities or obstacles to the promotion of gender issues: party leaders may exercise strict control over the activities of legislators or put few constraints on legislators to develop
9780230_245884_05_cha04.indd 64
10/5/2010 5:02:06 PM
Gendered Institutions and Substantive Representation 65
their legislative priorities autonomously. Thus, in order to explain the kind of actions female legislators take on behalf of women, we need to explore the rules and norms that structure their environment. Gendered power asymmetries A second component of GI involves exploration of the way institutions reflect and reinforce gendered power relations. Rather than viewing institutions as neutral, both feminists and HI assume that institutions mobilize bias in ways that make certain outcomes more or less likely. As historical creations, institutions embody a particular set of power relations present at the time of their creation. Institutions thereby ‘structure politics’, privileging the expression of certain interests over others (Thelen and Steinmo 1992). Feminist scholars add further complexity to HI, pointing to gender as an integral feature of social relations (Acker 1992; Lovenduski 1998). The gendered origins of legislative institutions have important implications for women’s substantive representation. The hierarchies and power relations in political institutions are those that privilege male-defined interests and priorities. The most prestigious and influential legislative committees, for example, correspond to those policy areas that have been defined as most relevant to the ‘national interest’, such as finance, foreign affairs, and industry. According to Kathlene, the fact that men dominate these committees ‘reinforces the given that these are the power committees’ (1995: 168). This may lead women seeking to advance their political careers to eschew the less prestigious ‘women’s issues’ in favour of policy issues considered more prestigious.5 This is another example of how institutions shape individual preferences. The hierarchy of policy issues, where women’s issues are considered of less national or strategic importance, makes it less likely that female legislators seeking to advance their political career prefer to work in these areas. On the other hand, the common assumption that the ‘soft’ policy areas, such as education and welfare, are of greater interest to women has meant that female legislators are often placed on these committees by male party bosses (Heath, Schwindt-Bayer, and Taylor-Robinson 2005). The policy divisions and legislative specializations present in modern legislatures constrain the substantive representation of women in other ways, too. Most of the policy specializations relate to the economy, internal and external security, or social welfare, while many of the issues relating to women’s rights and equality, for example, violence against women or equality in the workforce, are difficult to address as singular policy issues (see Weldon 2002). This complicates the trajectory
9780230_245884_05_cha04.indd 65
10/5/2010 5:02:06 PM
66 Susan Franceschet
of legislative initiatives on these issues. Often, gender issue bills are sent to multiple committees, which has the effect of increasing the veto points for these bills. Thus, a gendered institutional bias is expressed in the organizational rules that structure legislative institutions. Or, women’s rights initiatives, such as candidate gender quotas or equality provisions, may require a change to a country’s constitution, in which case the threshold for success is much higher, because constitutional change almost always requires more than simple majority support in parliament.6 Gendered power relations are also expressed in legislative institutions through male dominance of leadership positions on legislative committees, making it more difficult for women who do prioritize women’s rights initiatives to move their bills successfully through the committee process (Norton 1995). Women are under-represented in other leadership positions as well, and seldom exert influence over the legislative agenda. The scarcity of women in leadership posts can slow down the progress of women’s rights initiatives, constituting another way that gendered power relations entrenched in political institutions can weaken the link between women’s presence and women’s substantive representation. Ideas and gendered norms A third feature of a GI approach is a focus on the gendered aspect of ideas and norms. For HI, ideas have important ‘framing effects’ that shape policy debates in particular contexts (Steinmo 2008). Feminist scholars take this insight even further, showing how ideas and norms about gender are reflected in myriad ways in institutions (Acker 1992; Chappell 2006). Precisely because ‘institutions have distinctly gendered cultures’ (Lovenduski 1998: 348), women seeking to be both accepted and effective in politics encounter ‘assimilative pressure to conform to the behavioural norm’ (Puwar 2004b: 77). Where behavioural norms and values are masculine, as they are in most legislatures, women may respond by disavowing distinctly feminine (and feminist) concerns, instead favouring both the style and substantive issues of the dominant group. One of the best examples of feminist research into the gendered nature of norms is Chappell’s (2002a; 2006) work on gender and institutions, which shows that even the norm of bureaucratic impartiality ‘is profoundly gendered’. Neutrality norms ultimately support the (unequal) status quo by delegitimizing the efforts of ‘feminist activists who seek to use state institutions – including the bureaucracy – to advance their equality claims’ (Chappell 2006: 227).
9780230_245884_05_cha04.indd 66
10/5/2010 5:02:06 PM
Gendered Institutions and Substantive Representation 67
On the other hand, ideas can also serve as powerful mobilizing forces (Sanders 2006). The global diffusion of norms around violence against women legitimized the efforts of women’s movements and their parliamentary allies to achieve comprehensive domestic violence laws (Hawkins and Hume 2002). In Latin America, the adoption of domestic violence laws had more to do with emerging norms about state obligations to address violence against women than the increased presence of women in congress. Likewise, emerging global norms about the importance of equity in political representation have made it somewhat easier for female legislators to mobilize domestic support for gender quota legislation, although this norm remains more contested than norms concerning violence against women (Krook 2009). Contextualized research methods and complex causality The final aspect of a GI approach is an emphasis on contextualized and historical methods and a more complex view of causality. According to Sven Steinmo, ‘rather than treating all political action as if fundamentally the same irrespective of time, place, or context, historical institutionalists explicitly and intentionally attempt to situate their variables in the appropriate context’ (2008). Likewise, for feminist scholars, employing gender as a concept necessarily gives rise to a research strategy that is contextual, comparative, and relational (in that femininity cannot be analysed without reference to masculinity) (see Waylen 2009). This has at least two implications for studying women’s substantive representation. First, it means that the search for correlates of substantive representation that would hold across time and place ought to be abandoned. Taking predefined indicators, especially specific policies such as the existence of reproductive rights, public child care, or anti- discrimination laws, as indicators of women’s substantive representation overlooks the extent to which the historical and political context in each country shapes what substantive representation means for women in that country at that point in time. This taps into what Mackay (2008) has termed ‘thick’ conceptions of substantive representation. This does not make cross-national comparisons unworkable. Indeed, it is only through cross-national comparison that it is possible to appreciate the importance of different institutional factors that affect women’s actions and their outcomes. It does mean that we must avoid predefining the content of those actions and outcomes. Instead, comparisons must be fully grounded in the empirical realities of each case.
9780230_245884_05_cha04.indd 67
10/5/2010 5:02:07 PM
68 Susan Franceschet
Second, a GI approach means abandoning the quest for a general theory, searching instead for the ‘complex configurations of factors’ that have some causal significance for women’s substantive representation (Immergut 1998: 19). This way, women’s substantive representation becomes a dynamic concept, involving both procedural and substantive aspects (that is, agenda-setting and policy outcomes). Analysis along these lines would lead to ‘dense empirical descriptions’ of how institutional context produces specific preferences among female legislators, encourages their articulation, and generates particular outcomes. Researchers would aim to ‘tell a story’ about the factors that produce (or inhibit) women’s substantive representation in a particular context. To summarize, a GI framework for studying women’s substantive representation includes an investigation of: (1) the formal and informal organizational rules in legislative bodies and the ways in which gendered power relations are reflected and reinforced by these rules; (2) the role of ideas and norms, both in their institutional expressions and as a mobilizing force for facilitating certain actions and outcomes; and (3) a dynamic and comparative approach, where women’s actions and their outcomes are rooted in the historical legacies of an existing institutional context.
Legislative institutions and women’s substantive representation in Argentina and Chile7 This section applies a GI approach to women’s legislative behaviour and policy outcomes in Argentina and Chile, and shows how the formal organizing rules and the informal institutions that structure politics are gendered in ways that influence the promotion of women’s interests. Argentina and Chile offer interesting contrasts. In 1991, Argentina was the first country in the world to adopt a gender quota law (Ley de Cupos), requiring parties to include a minimum of 30 per cent female candidates (and in electable spots) on all candidate lists (Chama 2001). The law has been effective in increasing women’s descriptive representation. Between 1993 and 2007, women’s seat share in the lower house rose from 5 to 40 per cent. When a reform of Senate electoral rules led to the application of quotas to the 2001 Senate elections, women’s presence rose to over 35 per cent. Chile lacks a quota law, and women’s representation remains low.8 Women won 5.2 per cent of the seats in the Chamber of Deputies in the country’s first post-dictatorship elections in 1989, and that figure has risen to 15 per cent in 2006. In the Senate, women’s share of seats has ranged from three to five per cent
9780230_245884_05_cha04.indd 68
10/5/2010 5:02:07 PM
Gendered Institutions and Substantive Representation 69
(Franceschet 2008). Numerical differences are not sufficient to explain differences in legislator behaviour or legislative outcomes, however. A GI framework draws attention to the different organizational rules, informal institutions, and ideas in the two countries. Formal rules and organization A comparison of Argentina and Chile reveals two important aspects of formal rules that affect women’s substantive representation: (1) legislative–executive relations, particularly the extent of legislative power exercised by the executive; and (2) the presence and location of ‘policy gate-keepers who control the legislative agenda’ (Alemán 2006). The differences between the two countries on these aspects of legislative organization explains why Argentine legislators are more active than their Chilean counterparts in terms of introducing gender rights bills across a wide array of areas, but also why Chilean legislators have greater opportunities for their bills to succeed, if they are willing to ally with the executive women’s agency, the National Women’s Service. Both countries have presidential systems with presidents enjoying a broad range of executive prerogatives. Chile’s Constitution gives extensive agenda-setting and legislative powers to the executive branch. Recent studies of Chilean politics, however, show that a focus on these formal rules, without an understanding of how informal institutions actually curtail the president’s use of these powers, leads to erroneous conclusions about the weakness of congress (Ferraro 2008; Siavelis 2006). Nonetheless, Chilean legislators perceive that they lack power. Many interviewees highlighted the numerous restrictions on their ability to legislate. A deputy from the governing coalition explained that ‘we can’t introduce bills that affect the national budget, which are in fact the most important because everything has to do with money and resources’.9 Another said, ‘I feel that here [in congress] our capacity is very limited’.10 These perceptions, along with real limitations on the types of bills that legislators can introduce,11 is one of the factors that accounts for lower levels of legislative activity on women’s rights issues. In Argentina, none of the legislators interviewed reported feeling constrained in terms of their legislative initiatives, and indeed, women in Congress introduce a far greater quantity of women’s rights bills than their counterparts in Chile.12 A second aspect of formal organization, namely, the location of policy gate-keepers, influences the outcome of women’s legislative initiatives in different ways in the two countries. Policy gate-keepers control the legislative agenda and can advance initiatives they favour
9780230_245884_05_cha04.indd 69
10/5/2010 5:02:07 PM
70 Susan Franceschet
and block or stall those they oppose (or those that simply are not priorities). But different legislatures entrench different rules that either increase or decrease the power of gate-keepers. In Chile, the executive enjoys substantial agenda-setting powers. Presidents can declare ‘urgencies’ requiring that a bill be discussed in committee or chamber within a specified time period. Argentine presidents do not enjoy these powers (Alemán 2006: 130). But it is not the executive prerogatives of the Chilean president per se that affect policy outcomes on women’s rights. Rather, it is the existence of a relatively well-resourced women’s policy agency, the National Women’s Service (SERNAM), whose director enjoys ministerial status and can thus directly lobby the president to use urgencies to advance women’s rights bills. The existence of SERNAM, and more specifically, the legislative influence possessed by the agency, has compelled feminist legislators to develop a cooperative working relationship with the agency. SERNAM can help to move women’s rights bills through the committee process and on to chamber debates. Indeed, successful women’s rights legislation has almost always been supported by SERNAM, and, in many cases, has been introduced by the executive rather than by legislators (Haas 2006). While this reality can be frustrating for female legislators who would prefer to have their own names associated with particular women’s rights bills, SERNAM’s influence and greater access to technical and human resources13 make cooperation with SERNAM critical for the success of women’s rights initiatives.14,15 Women in Argentina’s congress cannot count on an influential ally whose role it is to promote gender equality in the executive branch. Although there is a women’s policy agency in the Argentine bureaucracy (the National Women’s Council), it has a very low rank and does not enjoy any legislative powers. In the Argentine congress, it is the president of each branch and the leaders of the legislative blocs (party caucuses) that act as policy gate-keepers. Because there are no formal deadlines for committees to address bills within a specified time, committees only deal with bills when caucus leaders or committee chairs take action (Alemán 2006: 133). Although there is a formal body (Commission on Parliamentary Work), comprised of the chamber’s president as well as all the caucus leaders, which sets the legislative agenda, a former staffer explained that it is the president of each chamber that enjoys the ‘full capacity to choose which bills are addressed and which are not’.16 Similarly, a senator explained in an interview that ‘it’s totally up to the committee chair [whether a bill advances], and even if there’s political will there, someone higher up – the president for example – will send signals to
9780230_245884_05_cha04.indd 70
10/5/2010 5:02:07 PM
Gendered Institutions and Substantive Representation 71
committee chairs about whether a bill should move or not’.17 This is why women’s rights bills almost always die in committees rather than being rejected in floor votes. Indeed, data show that caucus leaders’ control over the legislative agenda filters bills in such a way that those that make it to floor votes already have the support of the majority party (Alemán 2006: 144). In sum, the formal organizational rules in Argentina and Chile create different incentives for female legislators to introduce women’s rights initiatives and different opportunities for these initiatives to succeed. In Argentina, where caucus leaders play a key gate-keeping role, it is more difficult to have gender issues considered a high priority. Party leaders rarely prioritize gender issues, and party leaders tend to be men. In Chile, the executive has important agenda-setting powers. A legislator explained in an interview that the most common strategy for seeking progress on a legislative initiative is to generate support from the relevant ministry and lobby the minister to have the bill declared urgent.18 Hence, the existence of a well-resourced women’s policy machinery in the executive means that female legislators have an ally in that body.19 Although a women’s policy agency exists in Argentina as well, it does not play a legislative role, and has virtually no resources that could be deployed in the research, design, or lobbying in favour of women’s rights legislation.20 Gendered norms and informal institutions Women in Argentina’s congress have been far more active on certain gender issues, for example, reproductive rights, than their Chilean counterparts. The explanation for this is not that Argentina has a substantially larger proportion of female legislators, or even that there are more stridently feminist legislators in congress. Instead, there are certain informal rules and norms that govern the political process in the two countries that account for this difference. One of the most deeply entrenched informal institutions governing the Chilean political process since the return of democracy in 1990 is the norm of consensusseeking and conflict avoidance (Borzutzky and Weeks, forthcoming; Siavelis 2006). In practice, this means that the formal powers of the executive, for example, the ability to force debate on a bill by declaring ‘urgencies’, are only used after a president has already built support for the bill within the coalition parties and the opposition as well. According to Peter Siavelis, ‘reaching a pre-legislative acuerdo [agreement] is an integral part of the Chilean informal political game, with consequences for violation’ (2006: 50).
9780230_245884_05_cha04.indd 71
10/5/2010 5:02:07 PM
72 Susan Franceschet
Chile’s ‘democracy by agreements’ produces a substantial degree of self-censorship, whereby even female legislators with well-known feminist views avoid introducing bills on conflictual issues. In a context where abortion is highly controversial, few congresswomen have taken up the issue of decriminalizing abortion.21 The avoidance of policy initiatives on abortion was entrenched after a female legislator, Adriana Muñoz from the governing coalition, introduced a bill to decriminalize therapeutic abortion in 1991. She was vilified in the press and by conservative legislators (Blofield 2006). She also failed to win immediate re-election (although she did go on to win subsequent elections). This intensely negative reaction has reinforced the norm of self-censorship among feminist legislators, none of whom have associated themselves with explicitly pro-choice positions on abortion. In recent years, male legislators have been more active in pursuit of liberalizing abortion law, and a male legislator expressed some concern about his female colleagues’ hesitancy in pushing harder on this issue.22 This implies that it is easier for men to challenge informal institutions, especially on issues related to women’s gendered roles. But it is also important to note that both male and female legislators sympathetic to liberalizing Chile’s abortion law are more likely to be members of the centre-left governing coalition, a factor that further reduces women’s propensity to rebel or violate the norms of consensus-seeking. The costs of defection or rebellion are higher for members of governing parties, as Swers’s (2002) research on the USA finds. There, female Republicans were more likely to cross-party lines to ally with Democrats when their party was in opposition. In contrast, when the Republicans controlled Congress, women were less likely to defect from their party’s position on women’s issues. In Argentina, in contrast, the norms of politics do not include an emphasis on consensus-seeking or conflict avoidance. Indeed, political leaders often act in a unilateral fashion, and it is not uncommon for legislators to put forward initiatives that are controversial, or potentially divisive. In this context, women are more active on abortion than in Chile, even though Argentina already has a slightly more liberal law. In Argentina, there have been 37 bills to liberalize the abortion law between 2000 and 2008, with women introducing 76 per cent of the bills (Franceschet and Piscopo 2008). In Chile, there have been only six initiatives to expand reproductive rights in that same time period (and only two of these bills can be considered explicit attempts to liberalize the law).23 These bills carry 26 signatures in total24 with women’s signatures representing 38 per cent of the total. This difference cannot
9780230_245884_05_cha04.indd 72
10/5/2010 5:02:07 PM
Gendered Institutions and Substantive Representation 73
be explained by numerical differences in women’s presence in the two bodies. Instead, the strong norms of conflict avoidance help to explain women’s behaviour in Chile. In Argentina, legislators perceive greater freedom to introduce bills on a range of issues, even if these issues are unlikely to receive their party’s support. If legislator actions (and inactions) were unproblematic indications of individual preferences, one would conclude that female legislators in Argentina are contributing to women’s substantive representation, while Chilean legislators are not (at least on this issue). But if the gendered institutional context is explored, it becomes evident that female legislators in Chile face greater constraints in acting on their preferences. The norms of conflict avoidance in Chilean politics, and the fact that most feminists are members of the governing coalition, encourage legislators to focus on issues that are less divisive and that have a greater chance of attracting broader support. Hence, female legislators have been more active on violence against women than reproductive rights due to the cross-party support for expanding the state’s role in protecting women from violence and for increasing sanctions for violent acts. Norms like consensus-seeking and conflict avoidance produce distinctly gendered outcomes, particularly because Chile’s parties are programmatic, well-disciplined, and organized around ideology. As such, gender issues are subject to partisan conflict, with conservatives taking positions that prioritize the traditional family and the more constrained gender roles that go along with it. The norm of consensus-seeking has thus led consecutive governments to avoid taking up potentially polarizing issues, especially women’s reproductive rights. In this way, the informal norms governing the political process in Chile reinforce unequal gender relations and mobilize bias to keep certain issues from assuming a spot on the policy agenda. Ideas Both HI and FPS emphasize the mobilizing potential of ideas. A good example is Argentina’s quota law (Ley de Cupos), whose impact goes beyond the greater numbers of women elected since its adoption in 1991. Quota laws can also create the idea that women are needed in politics precisely to act as substantive representatives for women. Hence, in countries where quota laws are adopted because of broad-based mobilization campaigns led by women, female legislators may perceive an obligation to represent women, an outcome that Franceschet and Piscopo (2008) term a ‘mandate effect’. Interviews
9780230_245884_05_cha04.indd 73
10/5/2010 5:02:07 PM
74 Susan Franceschet
with female legislators in Argentina reveal a sense of debt to women as a group because it was women’s collective action that produced the institutional changes that facilitated women’s access to elected office. As one legislator noted, ‘I would not have received the second spot on the list had it not been for the quota’. Not surprisingly, as women’s presence in both chambers grew as a result of the quota, the number of women’s rights bills introduced expanded considerably (Franceschet and Piscopo 2008). In this way, legislated gender quotas are a potential source of institutional change. To use Kathleen Thelen’s term, quotas may be a form of ‘layering’, whereby ‘institutional arrangements are renegotiated periodically in ways that alters their form and functions’ (2003: 213). Following Beckwith (2005), quotas are also a way of re-gendering politics in ways that legitimate both women’s presence and their actions as substantive representatives. At the very least, the incorporation of legislated quotas constitutes an explicit recognition that gender discrimination exists and needs to be remedied by institutional mechanisms. But, by explicitly acknowledging that women’s absence in politics is a problem, quotas also create the idea that women’s presence is needed precisely to promote issues of concern to women. This, in turn, legitimizes the idea that women are a politically relevant collective, which entails an explicit re-gendering of political institutions. In countries where the adoption of a quota law is preceded by extensive mobilization and political debate, the ultimate adoption of the law can therefore strengthen the link between women’s presence and women’s substantive representation. This occurred in Argentina. In Chile, on the other hand, while proposals for a quota law have been introduced by female legislators (on three occasions) and in 2008 by President Michelle Bachelet, advocates have been unsuccessful at generating much support, either across the political parties, or in civil society. Unlike in Argentina, women’s movements have not embraced the idea of gender quotas, perhaps reflecting the growing distance between the political class and social movements. Within the political class, quota opponents reinforce this division by framing quotas as a mechanism that would primarily benefit ‘elite’ women while doing little for ordinary Chilean women. Quota critics also emphasize that affirmative action policies are inconsistent with Chile’s deeply rooted meritocratic norms (Franceschet 2008). The nature of these debates, and the scant interest in quotas among organized women, do little to entrench the kinds of ideas and norms that would foster a sense
9780230_245884_05_cha04.indd 74
10/5/2010 5:02:07 PM
Gendered Institutions and Substantive Representation 75
of mandate or obligation for women in congress to focus on gender issues. By looking at informal institutions and norms, it is possible to understand why patterns of bill introduction differ across the two countries, particularly why women in Chile’s congress have been less active in certain areas. On the other hand, the contrast in the organizational rules, and particularly, the role of policy gate-keepers in the two countries, provides insight into why there have been more policy successes in Chile than in Argentina, despite the greater legislative activity by women in the Argentine Congress. The existence of SERNAM in Chile, combined with the executive’s greater agenda-setting powers means that, at least on issues where a cross-party consensus is possible, women’s rights issues can become legislative priorities. In sum, while the norms of consensus-seeking and conflict avoidance constrain the articulation of legislator preferences on potentially divisive issues, there are other organizational rules and institutional features that create opportunities for legislative success. In Argentina, in contrast, there are fewer constraints on female legislators to act on their preferences by introducing bills (even on controversial issues), but women’s lack of power in legislative institutions (i.e. their lack of leadership positions in congress), and their lack of an influential ally in the executive, means that their legislative initiatives face greater obstacles to success.
Conclusion This chapter argues that combining HI and FPS provides valuable tools for studying women’s substantive representation. A GI approach encourages a broader conception of substantive representation, one that includes both the actions of female legislators and the outcomes of those actions. This focus draws attention to gendered power asymmetries as well as the informal institutions that shape the legislative environment. A further strength of a GI framework is that it encourages scholars to abandon the search for a general theory in which specific variables are linked in a relationship believed to hold across space and time. Instead, this approach would seek a more contextualized and richly detailed account of how gender relations are inscribed in institutions in ways that shape women’s preferences and actions in legislative bodies and how those actions produce (or fail to produce) their desired outcomes. Cross-national comparisons are useful in this task because they draw
9780230_245884_05_cha04.indd 75
10/5/2010 5:02:08 PM
76
Susan Franceschet
attention to the importance of factors that may not be obvious to those who study single cases. The analysis above reveals not only that scholars of gender and politics ought to make more use of the concepts and methods used by HI, but also that HI would benefit from borrowing tools from FPS. Explicit attention to the gendered origins and underpinnings of institutions provides greater specificity to the power asymmetries and the mobilization of bias that HI seeks to uncover. Likewise, while HI emphasizes the importance of ideas and norms, FPS draws attention to specific ideas, namely, ideas about gender and expectations about the appropriate behaviour for men and women in specific gendered institutional environments. Finally, attention to gender offers a potential explanation for institutional change, something that HI is sometimes accused of lacking (Waylen 2009). A GI approach assumes that institutions can be regendered, as feminist activists and their allies seek change both from within and outside of institutions. The outcomes of these actions can produce new institutions (such as women’s policy machineries) or produce changes to existing institutions (such as candidate gender quotas).
Notes 1. Theoretical debates about women’s substantive representation have raised concerns about essentialism and the coherence of ‘women’s interests’ given the heterogeneity of women as a group. Due to space constraints, the chapter does not delve into these debates. The author’s view, however, is that the concept of ‘women’s interests’ remains useful so long as it is understood that the content of women’s interests is shifting, contextual, and open to democratic dialogue, rather than static and essential (see Vickers 2006). 2. This can include a number of things such as bill introduction, behind-thescenes lobbying to generate support for a bill, communication and sharing information with women’s organizations in civil society, and speaking out in support of gender issues in committees and parliamentary debates. 3. The distinction between substantive representation as process and outcome is drawn from Franceschet and Piscopo (2008). Some of the best work on women’s substantive representation in the USA, such as Dodson (2006), Swers (2002), and Tamerius (1995), also examines multiple stages of the legislative process. 4. Of course, because HI views institutions as historical creations, the organizational rules also reflect the outcome of conflict and power struggles. This is discussed in the next section. 5. This is not meant to imply that gender interests cannot be expressed in foreign affairs or finance. Indeed, these areas have important implications for gender relations. But even within these policy domains, the gendered aspects tend to be accorded less attention, particularly because they are defined as
9780230_245884_05_cha04.indd 76
10/5/2010 5:02:08 PM
Gendered Institutions and Substantive Representation 77
6.
7.
8.
9. 10. 11.
12.
13.
14. 15. 16. 17. 18.
19. 20.
21.
areas of ‘national’ interest and are therefore less amenable to the advancement of particularistic or group-based claims. In a number of Latin American countries, legislative initiatives have been introduced (and adopted in some cases) that add ‘right to life’ provisions to the Constitution, thereby substantially increasing the level of support needed to decriminalize abortion. This is an example of how institutional rules create obstacles to particular women’s rights policies. The analysis in this section is based on 32 interviews with legislators, staffers, and legislative assistants in the Argentine and Chilean Congresses. These interviews were conducted between July and December 2006 and May and June 2008. Interviewees were asked general questions about the legislative process, their strategies for pursuing their legislative goals, their motivations for taking up certain issues, and their perceptions about the role of women in legislative politics. The research was supported by the Social Sciences and Humanities Research Council of Canada. The main reason for women’s under-representation in politics is not the lack of a quota law, but rather an electoral system that sets up very high barriers to women’s entry (Franceschet 2005). Interview, 20 May 2008, Valparaíso, Chile. Interview, 20 May 2008, Valparaíso, Chile. The executive enjoys the exclusive right to introduce legislation affecting the budget. Indeed, comparative politics scholars have noted that ‘the budgetary powers of Chile’s president are among the strongest in Latin America’ (Ferraro 2008: 103). A recent comparative study of legislators in Argentina and Chile found that Chilean legislators, on average, introduced 10.6 bills in the period under study, compared with 28.9 in Argentina (CERC 2005). SERNAM has a legal reform department whose role is to research and design legislative initiatives that promote women’s rights and equality. Individual legislators in Chile do not have access to the same resources. See Haas (2006) for a discussion of the tensions that have defined the relationship between feminist legislators and SERNAM. Interview, 6 June 2008, Buenos Aires, Argentina. Interview, 6 September 2006, Buenos Aires, Argentina. Interview, 20 May 2008, Valparaíso, Chile. SERNAM does not always promote the issues favoured by female legislators, and even when it does, it does not always succeed in convincing the president to prioritize the issue and use executive prerogatives to advance a bill. As will be discussed in the next section, there are a number of informal institutions governing the political process that determine what kinds of issues are more likely to advance. This was confirmed in interviews with a number of legislators and legislative assistants in Argentina. Under Chilean law abortion is illegal in all instances. While therapeutic abortion had been legal prior to 1989, the Pinochet dictatorship, as one of its final acts, re-criminalized abortion in all instances, including danger to the mother’s life. Interview, 23 May 2008, Santiago, Chile.
9780230_245884_05_cha04.indd 77
10/5/2010 5:02:08 PM
78 Susan Franceschet 22. There was another attempt to decriminalize abortion in October 2006, but the bill, sponsored by a group of male legislators, was deemed inadmissible. (Interview with legislator from the Socialist Party, 21 November 2006, Valparaíso, Chile, and interview with one of the bill’s sponsors, 23 May 2008, Santiago, Chile.) 23. The records kept by the Chilean Congress do not distinguish between a bill’s primary sponsor and co-sponsors. 24. Interview, 2 August 2006, Buenos Aires, Argentina.
9780230_245884_05_cha04.indd 78
10/5/2010 5:02:08 PM
5 Gendering the Institutional Reform of the Welfare State: Germany, the United Kingdom, and Switzerland Michelle Beyeler and Claire Annesley
Due to the intervention of feminist scholars, the gendered dimensions of welfare states have increasingly been recognized. In welfare state analysis there is now a clear understanding of the role of intersecting institutions in the structure of the welfare state, referred to as ‘welfare regimes’. Very importantly, feminists showed how different social policy systems enforce or weaken the ‘male-breadwinner principle’, which defines a gendered division of labour within the family (see for example Lewis 1992; Lewis and Ostner 1994; O’Connor et al. 1999; Sainsbury 1996). The contribution of feminist scholars has widened the scope of institutions considered significant in these regimes to include, for example, the gendered dynamics of the family and caring relationships, as well as the way in which these gender relations play out in other socio-economic spheres such as the labour market. In an unprecedented fashion, the significance of gender has become recognized and adopted by mainstream scholarship (Amenta 2003; Bussemaker and Van Kersbergen 1994; Esping-Andersen 1996, 1999; Pierson 2000b). Yet, while much progress has been achieved in classifying the gendered nature of welfare institutions, the literature is less adept at understanding the process of reforming the gendered institutional architecture of the welfare state in a way that promotes more gender equality. The male-breadwinner ideology clearly has lost its influence in the Western world. In many welfare states, policies were changed and services extended to improve the possibilities of reconciling family and work for both sexes. Examples are the recent far-reaching reforms in the 79
9780230_245884_06_cha05.indd 79
10/5/2010 5:02:14 PM
80
Michelle Beyeler and Claire Annesley
United Kingdom and Germany. The male-breadwinner ideology, however, has not completely disappeared and, indeed, it still receives open political support from conservative forces in countries like Switzerland. Moreover, not all policy changes really lead to more gender equality. Some reforms of policies and institutions that at first sight look womenfriendly may turn out to set wrong incentives, or have unintended consequences that in the end reproduce gender inequalities. In this chapter we argue that a feminist institutionalist (FI) perspective should emphasize the dynamic aspects of welfare state analysis so that it is not only possible to understand the gendered institutional relations of the welfare state, but that it also becomes possible to theorize the ways in which these can be altered for more gender equality. An FI approach would thus focus on theorizing and analyzing (1) reasons, (2) agents, (3) opportunities and obstacles, and (4) directions for change. This approach draws on the conceptualization of institutions in new institutionalism (NI) as complexes of formal and informal rules. Therefore, we claim that an FI perspective should take a pluralist attitude towards the different schools of NI. Arguably, an open and pluralist attitude was one of the reasons why feminists have had such a significant impact on the welfare state literature thus far. In the first section, we outline our perspective on FI. In the second section, we briefly introduce the debate on welfare regimes and gender regimes, where we argue that such an FI perspective already exists, even if it has not previously been labelled that way. We claim, however, that the focus on reform could still be enhanced. In the third section, we therefore sketch different possible research questions related to the transition of the ‘male-breadwinner model’ based on our outline of FI. We illustrate our arguments with examples from the United Kingdom (UK), Germany, and Switzerland. We then conclude by summing up and presenting some lessons that can be drawn from the FI as it has been developed, and is still developing in the welfare state literature for other research fields.
Outline of a feminist institutionalism Feminist political theory is concerned with the reasons and effects of gender inequality and the unjust power structures that perpetuate them. Distinct from other theoretical positions, it also seeks to develop a political agenda aiming at more gender equality. Partly stemming from this political agenda, women’s agency and feminist agency, in particular, constitute core preoccupations of the feminist perspective. In our
9780230_245884_06_cha05.indd 80
10/5/2010 5:02:15 PM
Gendering the Institutional Reform of Welfare State
81
view, feminism has the potential to improve mainstream institutionalist analysis by bringing in the issues of gender, gender inequality, and women’s agency, as well as the awareness of the implications of explicit or implicit gender norms in society, policies, and political research. Can and should feminist scholars learn from NI, and to what benefit? In a recent issue of Politics & Gender (volume 5, number 2, 2009), feminist scholars involved in the FI development project have discussed concepts from rational choice institutionalism (RCI), sociological institutionalism (SI), historical institutionalism (HI), and discursive institutionalism (DI) and how these could be used, reformulated, and changed to fit the demands of feminist political science (FPS). We argue for a different path. FI should be developed as a different NI variant, developing and presenting its own concepts. The important questions, in our view, are not whether feminists should learn form different NI concepts or whether NI should rather incorporate feminist concepts of gender and power (Kenny and Mackay 2009; Krook and Mackay, this volume). The important issue in our view is to develop FI as an NI variant that is broad enough to capture as many issues and questions related to feminist concerns as possible, and at the same time is close enough to other NI variants to make a dialogue with the mainstream possible. To get there, we claim for an FI that takes a pluralist attitude towards the different NI variants. At the same time, FI should be ready to incorporate the core claim of NI, namely its conceptualization of institutions as systems of rules that can be both formal, like written norms, procedures, laws, and contracts, and informal, like norms of behaviour, codes of conduct, and conventions (Hall 1986; Hall and Taylor 1996; North 1990; Steinmo et al. 1992). Conceptualizing institutions as complexes of rules and norms helps to disentangle agency and structure in our analytical and empirical work. Institutions do not, in themselves, act or behave. Institutions are structures of incentives (Scharpf 2000) that influence perceptions and options of actors, but do not determine them. One should not lose sight of two core implications of this conceptualization. First, institutions consist of formal and informal rules, and second, institutions are complex systems of rules. These two aspects are particularly relevant for feminists and their strong interest in the issue of change and development. One always has to be aware that even if some formal rules are radically changed, the outcome will not necessarily be as radical, since many other – formal and often informal – rules still persist (North 1990). Similarly, while formal rules may be highly
9780230_245884_06_cha05.indd 81
10/5/2010 5:02:15 PM
82 Michelle Beyeler and Claire Annesley
resistant towards reforms, institutions may nevertheless be changed in more subtle ways through developments in informal structures. Although sharing commonalities in the conceptualization of institutions, NI variants greatly diverge in their conceptualization of the relationship between actors and institutions. Feminist scholars often consider HI, and sometimes DI, to be the NI variants that are most compatible with feminist research. We diverge from this view and call for a pluralist attitude towards different NI variants. We contend that the different NI variants tend to complement rather than to contradict each other. HI, for instance, can be most fruitfully used to analyse the origins of gendered institutions and policy regimes, as well as the reform paths and the sources of stability. It provides, however, only incomplete analytical guidance regarding the chances and directions of change. In addition, HI has little to say about how institutions affect actor behaviour. In fact, when it comes to the discussion of institutional effects, HI often implicitly or explicitly relates its arguments to other institutionalist perspectives. In RCI analyses, institutions are seen primarily as external constraints that affect actors’ choices independently from their values, beliefs, and tastes. The advantage of treating institutions as independent from values is that they become amenable to relatively simple and straightforward choice models. The disadvantage is that RCI obscures indirect, unconscious, and long-lasting effects of institutions in perpetuating power imbalances and value structures. At first sight this seems to be incompatible with feminist concerns. Still, the simple and straightforward models derived from the RCI perspective should not be neglected in the analysis of gender-related issues. There are gains from analysing institutions in their capacity to set incentives and disincentives for gender-relevant decisions. From an analytical, scientific point of view, it is interesting to analyse how political regulations constrain and influence the choices of politicians, parliaments, individuals, and families in ways that affect gender relations and gendered outcomes. Due to their simplicity, such analyses are likely to gain political influence and point to specific reform needs. Political institutions constrain the decision-making process in a polity. In this function, institutional structures determine under what conditions and to what degree new policy solutions can be implemented. Furthermore, they specify the type and size of the policy coalition needed for a reform. Institutionalists pay attention to these various effects of decision-making rules. From a feminist perspective it makes
9780230_245884_06_cha05.indd 82
10/5/2010 5:02:15 PM
Gendering the Institutional Reform of Welfare State
83
sense to analyse the effects in terms of how gender issues and ideas are constrained or promoted within the general institutional setting. However, RCI arguments explain behaviour only up to a certain point. More complex, cultural models have to be applied when the simple, rationalist assumptions reach their limits. SI emphasizes the fact that rules and norms not only affect choices, but are often also internalized by actors, thereby shaping their perceptions, observations, and interpretations (Edeling 1998). SI and its conceptualization of actor behaviour as following a logic of appropriateness within institutional contexts (March and Olsen 1989) offers helpful tools for analysing the construction of gender-specific identities within specific organizational settings. Moreover, by pointing to internalized constraints, an SI lens helps uncover hidden gender effects of institutions stemming from processes of socialization and learning. Since the different institutionalist schools provide different lenses and perspectives that are useful for the analysis of institutions and gender issues, FI should not be contained to one variant. Openness towards the different disciplines and an emphasis on commonalities and complementarities of the different approaches is likely to offer greater breadth for motivating a diverse and interesting research agenda. Moreover, a pluralist attitude also makes it easier for feminist researchers to communicate and collaborate with mainstream researchers. The core distinguishing element of the FI approach per se, in our view, is the preoccupation for issues of gender equality and the question of how gender equality is achieved. A focus on political reforms requires including political actors and political power structures in the analysis. From a feminist perspective, particular attention should thus be given to political activity of, and for women. Extending simple group and coalition theories, institutionalist analysis points to the role of formal and informal constraints in mediating between political actors and their interests on the one hand, and political outcomes (for example, regulations, laws, administrative structures, and polices) on the other. Power and conflict in FI analysis should therefore be conceptualized as constrained and shaped by institutions.
FI already exists: The literature on gender and welfare regimes We discuss and elaborate these claims based on examples of welfare state research, noting that FI already has been developed and applied very successfully in feminist contributions to the welfare regime literature.
9780230_245884_06_cha05.indd 83
10/5/2010 5:02:15 PM
84
Michelle Beyeler and Claire Annesley
One fruitful way of conducting institutionalist research that is attentive to complexes of rules and norms is to identify ideal-typical institutional regimes. A regime perspective helps to compare, classify, understand, and analyse countries. In welfare state analysis this approach has been very influential, especially since Esping-Andersen’s seminal Three Worlds of Welfare Capitalism (1990), whose scholarship introduced an approach towards the welfare state which argued that complexes of different policy measures reflect similar political ideologies and principles. Crucially, his approach highlighted the ways in which institutional configurations are interrelated to form distinctive yet coherent regimes. These were, for him, the social democratic, conservative, and liberal welfare state regimes. Esping-Andersen initially paid little attention to gender issues, thereby provoking forceful and lively feminist critique from scholars such as Lewis (1992), Orloff (1993), and Fraser (1994). More generally, the exclusion of the family and caring relationships from Esping-Andersen’s original analysis was particularly obvious in his core concept of ‘decommodification’. Feminist scholars quickly realized that this concept was flawed with regard to the situation of women. ‘Decommodification’ refers to formal policies which enable citizens to enjoy a certain standard of living independent of their labour market participation. However, for many women, welfare policies concerned ones which enabled them to take up a job in addition to performing care responsibilities. The paradox was that non-working women had to become ‘commodified’ before they could become ‘decommodified’, and no welfare state value was attached to the roles they performed in the family. Finally, Three Worlds failed to recognize the crucial role that non-working women performed to enable men to be breadwinners, freeing them up from domestic work, or the fact that formally employed women would be less likely to benefit from such free domestic assistance. Feminist scholars did not contain themselves to criticisms, but also offered alternatives and amendments to the Three Worlds typology. Esping-Andersen’s basic idea of analysing welfare states by identifying complexes of rules, norms, and policies was, in general, not challenged by his feminist critics. On the contrary, their concepts and arguments indicate that an FI approach already exists. Instead of analysing the degree of ‘decommodification’ of social policies, these scholars focused on the degree to which the male-breadwinner model was implemented in different welfare systems (Lewis 1992; Sainsbury 1994, 1996, 1999). Lewis (1992), for example, characterized Western welfare states according to the degree to which the welfare state is prepared to support a
9780230_245884_06_cha05.indd 84
10/5/2010 5:02:15 PM
Gendering the Institutional Reform of Welfare State
85
woman independent of a male breadwinner, labelling these as either strong, modified, or weak male-breadwinner states. Sweden counted as an example of a weak male-breadwinner country, France as a modified male-breadwinner state, and Ireland and the UK as strong malebreadwinner countries. Along similar lines, Sainsbury (1994, 1996) showed that contributionbased social protection systems in particular tend to assume a gendered division of labour in the family with an earner, who gets social protection through compulsory contributions to social insurances, and a carer, whose social rights are derived from a spouse’s contributions. Although this is, in general, not formally regulated, the basic assumption is that women take the part of the carer and men the part of the earner. Welfare states that strongly rely on contribution-based protection systems are frequently also characterized by strong measures to protect employees from dismissal. These labour market regulations reflect a preoccupation of the state for securing the (male) breadwinner income. In addition, other policy principles, such as the joint taxation of couples, as well as the definition of care responsibilities as private, and their assignment to the tasks of the family, contribute to perpetuating the male-breadwinner model through the welfare state. In later contributions, Esping-Andersen (1999, 2003) accepted the core of the feminist critique and included familialism as a second dimension to his welfare regime analysis. He managed to add this dimension without changing the most fundamental of his 1990 claims by showing that familialism is particularly strong in countries belonging to the conservative cluster and particularly low in countries belonging to the socialdemocratic countries. He argued that the welfare regime also defined the relationship between state, market, and the family. For the organization of care work this implies that liberal welfare regimes (like the United States) emphasize private solutions, while the social-democratic welfare regimes (like Sweden) prioritize public services. Conservativecorporatist regimes (like Germany) rely on a familial division of labour with one earner, who is protected on the labour market and entitled to social security benefits, and a carer, whose social rights are derived from those attained by the earner. The Three World typology and its relevance for gender outcomes, however, remained controversial (Adams and Padamsee 2001). Lewis (1992) had already showed that gender regimes do not necessarily coincide with Esping-Andersen’s welfare regime: the male-breadwinner principle was quite modified in France (classified as a conservative welfare state) and rather strongly implemented in the UK (classified as a liberal welfare
9780230_245884_06_cha05.indd 85
10/5/2010 5:02:15 PM
86
Michelle Beyeler and Claire Annesley
state). Similarly, O’Connor et al. (1999) argued that within the liberal countries (Canada, Australia, the UK, and USA), considerable variation exists with regard to gender policies. Yet, studies including wider ranges of countries showed that differences in family policies, at least partially, cluster with the conventional regime typologies (Gauthier 2002; Gornick et al. 1997). Further work emphasized the societal organization of care work, identifying care arrangements or care regimes (Bettio and Plantenga 2004; Daly and Lewis 2000; Lewis 1998; Pfau-Effinger 2005). In all these contributions, a preoccupation for gender equality issues was clearly visible. There appears to be, however, a shift in the core preoccupations raised by the scholars. While in its first wave, feminist contributions to the welfare regime discussion were primarily concerned with bringing women into the labour force (that is, ‘commodification’), with newer accounts increasingly shifting towards policy measures that incite men to take up an equal share with regard to caring responsibilities. This turn is connected to the growing awareness among feminists that, for many mothers working in the formal labour market the price has been a much higher workload than that of men. Feminist researchers of policy debates realized that the issue of care and of who does what amount of care work in a society is frequently neglected. Mahon (2006) shows that the discourses on reconciliation policies in the OECD emphasize the role of mothers in formal labour and in the economy, and neglect the roles of parents (of both sexes) as carers for dependent children and frail elderly. Lewis (2006) identifies a very similar policy discourse at the EU level. There are a number of remarkable features with regard to these debates. A feminist lens clearly shapes the way in which institutions and institutional mechanisms are analysed. Feminists shared the view that Esping-Andersen’s analysis of welfare regimes was deficient and they contributed with interesting and influential approaches to the debate. These contributions, even if they had not been presented under that label up to that point, can be conceived as FI analyses of the welfare state. They built upon, or provided alternatives to the existing welfare state concepts (see also Sainsbury 2008). The scholars were open towards different cultural, historical, and rational (political economy) approaches. And finally, most importantly for our classification of FI, they were preoccupied with the institutional effects on women as a category and, more specifically, with gender equality. And as, for instance, Morgan (2008) or Lewis et al. (2008) nicely demonstrate, the analysis may lead to concrete policy recommendations for the wider feminist cause.
9780230_245884_06_cha05.indd 86
10/5/2010 5:02:15 PM
Gendering the Institutional Reform of Welfare State
87
What is especially noteworthy about feminist interventions into welfare regime analysis is the fact that it has been taken up by the mainstream, not least Esping-Andersen, who has firmly adopted the institution of the family into his analysis. Gender issues have, as a consequence, become a central pillar in welfare state research. According to Pierson, feminist scholarship has made a major contribution to welfare state research as ‘it has explained why the outcomes that welfare state scholars have traditionally investigated need to be reconceptualized and expanded’ (2000b: 800).
Pushing the FI agenda in welfare state research: Making sense of recent reforms The male-breadwinner model as a guiding principle is increasingly losing its relevance in social policy programmes. Many advanced welfare states currently face massive changes with regard to family policies and a considerable expansion of publicly subsidized child-care programmes. Among the welfare states that faced the largest reforms are Germany and the UK. Both countries have been traditionally strong male-breadwinner states and have now strongly changed the basic policy settings towards the dual-earner principle. How can we understand and explain these massive reforms? How can FI contribute to analysing these reforms? We discuss a number of issues and possible research questions that arise from an FI lens as we specified it above. We illustrate our arguments with examples from Germany, the UK, and Switzerland. According to Esping-Andersen’s Three World typology, Germany belongs to the conservative cluster and the UK to the liberal one. Switzerland does not clearly belong to either one: its contributionbased schemes point at the conservative side, but the labour market has many liberal characteristics. In contrast to Germany and the UK, there have been only smaller changes towards more women-friendly family policies in Switzerland. High reform capabilities and favourable reform discourses in Germany One core issue of institutional analysis has always been to show how institutional structures constrain and enable policy change (Hall 1993). A simple institutionalist model of reform would emphasize two issues needed to bring about change: first, the reform capabilities defined by the political power distribution as well as the decision-making institutions and, second, the awareness of reform needs defined by the
9780230_245884_06_cha05.indd 87
10/5/2010 5:02:15 PM
88
Michelle Beyeler and Claire Annesley
dominant political discourse. Large reforms are possible, once the ideas that underpin the policy programme and the policy instruments come under strain (Baumgartner and Jones 2002). If a frame of a failed policy dominates the debate, new policy proposals are likely to mark a very clear departure from the previous solution (Palier 2005: 141). Germany, in our view, is such a case, where large reform capabilities and a frame of policy failure have led to significant reforms. German family policy after reunification rested on high cash transfers to families but only few services (for example, child care). By the end of the 1990s the system was increasingly criticized for its ineffectiveness, given persisting child poverty and low birth rates. In addition, marriage-related benefits were considered problematic as they incited partners in childless couples to stay at home, and generally reduced women’s labour market participation (Ostner, 2010). The new policy solutions that were proposed marked a clear departure from the maternalist, family-based values. The new agenda to be met was primarily based on employment (mothers should take up paid work) and to a lesser degree educational (public high-quality care of children) goals. The policies were not primarily framed as measures to increase gender equality. The policy ideas and the framing strategies were in line with proposals debated in the Organization for Economic Co-operation and Development (OECD) and the European Union (EU), that emphasize the activation of mothers, while neglecting the wider issue of parenting and caring (Lewis 2006; Mahon 2006). This discourse, together with a large reform capability due to a coalition government between the Christian and the Social Democrats, resulted in extensive reform programmes. In 2007, a new tax-funded parental leave was introduced, granting a wage replacement of 67 per cent of former earnings for 14 months. The leave is only fully paid for a lone parent or if the second partner takes at least two months of the leave entitlement. This new scheme was designed to encourage higher qualified, well-paid women to continue employment when having children and for skilful employed women to have children. Moreover, by introducing two socalled daddy months, the reform took up the newest feminist concerns of engaging fathers more in the care work. Contrary to the older leave scheme, the new one discourages longer labour market exits of mothers. While rewarding the well-paid working parents, poorer parents are worse off with the new scheme, because benefit rates and duration of the leave were cut down. Still, combined with larger programmes to extend the provision of day-care facilities, the new policy is clearly consistent with the goal of having mothers in the labour market.
9780230_245884_06_cha05.indd 88
10/5/2010 5:02:16 PM
Gendering the Institutional Reform of Welfare State
89
While classical NI models may suffice to explain why there have been large reforms, there are still many open questions that should be analysed from a feminist perspective. How can we explain this strong focus on the activation of women? Why have arguments about the support for carers (Betreuungsgeld) been marginalized (see Ostner, 2010)? Maybe the reason is a lack of mobilization of women and the marginalization of gender issues in the political debates. But these issues still need further investigation and comparative assessments. In particular, we should aim at developing new concepts that help to classify, and evaluate, and compare such reform efforts with regard to their impacts on gender equality. Reform inconsistencies in Switzerland Given the large reforms in family policy, influential scholars have cast doubt on the further applicability and usefulness of welfare regime typologies for studying gender issues and the welfare state (Lewis 2006). Regime analysis, as Esping-Andersen (1999: 73) points out, has a strong static component. The approach is very helpful for identifying large, general patterns, but does not perform particularly well as a tool for grasping development and change. This is certainly true. But the regime approach also teaches us that there are complex interrelationships between formal rules and informal constraints, as well as between different sectors. These interlocking institutions have to be kept in mind when it comes to analysing change. Women-friendly changes in one policy field – even if they are significant – in the end may only have little impact on gender equality, if inequality persists in other domains. Thus, even a massive expansion of public day-care facilities may have only limited effects on gender equality if the institutional set-up of the labour market still contributes to the resilience of gender inequalities. If career and working models are still based on the idea of continuous, full-time engagement for the firm, parenthood is likely to put dualearner couples under a high strain to reconcile the demands of their children and their bosses. If this is the case, it is plausible that the couple may decide to concentrate on one career and that the second earner may only work part-time and give up his or her career aspirations. This is known as the ‘one-and-a-half-adult-worker-model’.1 In almost all cases, the part-time component, which is more likely to be linked to the renouncement of a career, is taken by the woman. This phenomenon is widespread in Switzerland. Although female employment is rather high in Switzerland compared to most other OECD countries, many women, especially when they have small
9780230_245884_06_cha05.indd 89
10/5/2010 5:02:16 PM
90
Michelle Beyeler and Claire Annesley
children, only work short, part-time hours. Twenty-five per cent of all women work less than 19 hours a week and another 22 per cent work between 20 and 29 hours. These are the highest short, part-time rates in the OECD world with the exception of the Netherlands, where working part time is, however, also quite common among men.2 Gender-specific role models, as well as societal values linked to motherhood, may explain this pattern to a certain degree. Of large concern from a feminist perspective should, however, be that policy regulations are set up which set incentives towards such a gender-specific, workcare pattern: Buetler (2007) calculated for the city of Zurich3 the direct household income effects of a change in working hours of a dual-earner couple with small children in formal day care. Because subsidies are only given up to a certain threshold household income, an increase in working hours may reduce total household income. This negative effect on the household income is particularly large if both parents have a middle-high income, for example, as teachers or in positions in lower management. Moreover, the effect is particularly pronounced if the second earner switches from a 40 per cent to a 60 per cent workload. In the Swiss labour market, shorter-hour, part-time jobs, for example below 60 per cent of an equivalent full-time position, are mostly incompatible with executive functions and responsibilities. Because there is no deduction of tax for private contributions towards child-care facilities, high-income couples also face negative incentives (to a certain degree) in choosing to work longer hours. Such perverse incentives partially explain why shorter-hour, part-time employment for women is so high in Switzerland. On the other hand, the system also reflects the institutionalized principle of a model that gives the mother the possibility to work part time in order to stay in the labour market, but no incentives to keep her career aspirations intact. The likely result is a strong persistence in horizontal labour market segregation. This example points out that it is not sufficient to focus on seemingly gender-friendly policy developments, such as public measures to expand day-care facilities, to decide whether the welfare state has become more ‘gender friendly’ or not. New policies may also create new disincentives to gender equality. Such disincentives may arise from cost–benefit calculations at the level of households, which may be influenced by policies. This example also shows that the simple and straightforward models derived from the rationalistic perspective should not be neglected in the analysis of gender-related issues. There are gains from analysing institutions in their capacity to set incentives and disincentives for
9780230_245884_06_cha05.indd 90
10/5/2010 5:02:16 PM
Gendering the Institutional Reform of Welfare State
91
gender-relevant decisions. From an analytical, scientific point of view, it is interesting to analyse how political regulations constrain and influence the choices of individuals and families in ways that affect gender relations and gendered outcomes. If incentives create problematic gendered outcomes, one might also be able to develop a highly compelling political argument for why, and in what direction the regulation should be changed. Due to their simplicity, such analyses are likely to gain political influence and point to specific needs for reform. Women’s agency and welfare reforms in the UK One strand where gender analysis and feminist research has also contributed greatly to the welfare literature relates to analyses focusing on the role of women in the formation and development of welfare states. To date, feminist historical studies on the origins of welfare states have found evidence of the role played by women’s political agency in shaping social policies in the period from 1880 to 1920. Studies by Koven and Michel (1990), Skocpol (1992), Skocpol and Ritter (1991), and Lewis and Ostner (1994) have identified the important role of women’s agency in the period of welfare state construction in the UK, the USA, France, and Germany. Koven and Michel (1990), for example, argue that in those four countries large-scale welfare policies and programmes emerged at the same time as women’s social action movements. Although women lacked the formal political power to shape the welfare state according to their vision, they nonetheless played a significant role via women’s grassroots organizations and national and international lobbying groups, which pressed for maternal and child welfare benefits. More recent studies (Daly and Lewis 2000; Daly and Rake 2003) make some important claims about women’s social and economic agency as clients of the welfare state. However, there is a gap in the literature on the capacity of women as political agents to shape and reform the programmes and policies of the contemporary welfare state (Annesley 2007). There are important indications that feminist agency can play an important role in welfare reconfiguration and this can be illustrated by the significant reforms in the domain of work–life reconciliation policy in the UK since 1997. These have extended to women’s and men’s rights to leave when a child is born, as well as the right to request flexible employment to parents of young children and those with other caring responsibilities. From 1998 to 1999, the government extended the entitlement to paid maternity leave from 14 to 18 weeks with and additional 29 weeks’ unpaid leave, and implemented the European Union parental-leave directive, offering parents 13 weeks’ unpaid parental
9780230_245884_06_cha05.indd 91
10/5/2010 5:02:16 PM
92
Michelle Beyeler and Claire Annesley
leave. Further reforms were introduced incrementally, with the rights being extended and improved at each stage. The first major wave of reform, set out in the Work and Parents: Competitiveness and Choice Green Paper (2000), extended the right to maternity leave to 52 weeks in total, 26 of which were paid. It also introduced to fathers the right to two weeks’ paternity leave. A second Green Paper, Work and Families: Choice and Flexibility (2005) led to a further set of reforms which extended the right to paid maternity leave to 39 weeks (with a commitment to 52 weeks by the end of the legislative period) and introduced the possibility of developing legislation to allow a mother to transfer her second six months of maternity leave to the father. These reforms are significant in at least two ways. In the context of a liberal welfare regime, they introduce generous periods of paid leave, currently 39 weeks with a promise of 52. The flip side of this development is that the level of payment during maternity leave is not high: 90 per cent of previous earnings for the first six weeks, but then a flatrate statutory maternity pay – at £123.06 per week at the time of writing – kicks in. This rate is low. To put it in perspective, if one worked 35 hours on the government’s National Minimum Wage (NMW) one would earn £203.00. The second significant feature of these reforms is that the emphasis is strongly on extending extensive rights to mothers, rather than to parents, or to mothers and fathers on an equal basis. Fathers have just two weeks’ paid leave at the statutory rate of maternity pay, and the proposed six months’ transferable maternity leave, if it ever comes to fruition, would be in the mother’s gift and paid at below the NMW rate of statutory maternity pay, if at all. As such, the reforms reinforce the social view that women care for children, and the reforms do little to redistribute the division of informal labour in the institution of the family. The developments pose two questions about (1) how the extension of welfare rights is possible in a liberal welfare state in an era of liberal political economy and welfare retrenchment, and (2) why these policy developments take the gendered direction that they do. The answer is feminist political agency. The welfare reforms were articulated, developed, and pushed through by an informal, but tightly-knit, group of Labour Party feminists. This group forged close links during Labour’s period in opposition from 1979–1997, campaigning for improvements to women’s representation in parliament. Politicized by this, the group pushed from 1997 onwards for feminist reform of the economy, welfare state, and international relations. In welfare policy they advocated
9780230_245884_06_cha05.indd 92
10/5/2010 5:02:16 PM
Gendering the Institutional Reform of Welfare State
93
reform to support women, but paid less attention to men, as they were aware that women’s votes won Labour the election in 1997 and there were no similar demands from men. Despite calls from gender equality groups such as the Equal Opportunities Commission and the Fawcett Society, work–life balance policies which emphasized men were not pursued. The key to policy change was the presence of key feminist actors in the institutional sites most important for the process of UK policy-making. This gets at the important finding that Labour Party feminists held office in the key sites of decision-making in the UK: the core executive, with ministers, the Prime Minister’s Office (No. 10), and the Treasury, as particularly significant. For example, Patricia Hewitt was Secretary of State for Trade and Industry and Minister for Women from 2001–2005 when the second wave of maternity reforms was being pushed through, while Carey Oppenheim was Social Policy Advisor to Prime Minister Tony Blair and played an important role in developing the agenda and fostering support internally and externally (Annesley 2010).4 This is an example of where feminist actors in a political system promoted a certain gendered discourse on welfare reform – a gender difference discourse, extending rights for mothers – and secured concrete policy outcomes. However, discourse on welfare reform comes from different sources, including the EU and other international organizations. Moreover, the UK government faces comparatively low levels of institutional constraints to reform. In contrast to Switzerland, where strong federalism and especially direct democracy often make reforms difficult and timely, the reform capabilities were very high.
Conclusions FI, according to our conceptualization, brings together feminist theory and institutionalist concepts not only to analyse differences and change, but also to contribute to feminist reforms. It takes a pluralist attitude towards the different NI variants and keeps an eye on its conceptualization of institutions. We argued that FI analysis is already very present and influential in welfare research. We also pointed at some puzzles and questions related to the newer welfare state dynamics, where we would like to encourage further FI work to develop. In particular, we call for the development of concepts that allow us to evaluate and compare the recent reform programmes with regard to their effects on gender equality. Such concepts should not only take into account activation of mothers, but also focus on the concept of
9780230_245884_06_cha05.indd 93
10/5/2010 5:02:16 PM
94
Michelle Beyeler and Claire Annesley
care and a more equal sharing of caring tasks by women and men. A clearer evaluation scheme will also be helpful for studying the conditions and political contexts under which reforms that lead to greater gender equality are possible. Such studies should be attentive towards women’s agency and the structural conditions under which women are likely to have an effect on reforms. Feminist research was, and very likely will continue to be influential in welfare state research. A last question we therefore want to address is whether this feminist success story might be transformed to other areas of FPS. Welfare policies do have obvious, observable effects on women and gender role models. In most other areas of political research, the importance of women’s issues is potentially less obvious and less visible to the mainstream, making it certainly more difficult to awaken gender awareness. Moreover, welfare state researchers tend to be more preoccupied with their issue than with developing a particular methodology or theoretical approach. This contributed to a research field that is more open than others to different methodologies and interdisciplinary arguments (Amenta 2003). Still, there is a lesson for feminist researchers in other fields: one thing that made feminist arguments about the welfare state compelling to mainstream researchers was that they not only criticized, but also built upon the concepts used by the mainstream. This made the dialogue possible and fruitful.
Notes 1. An equal sharing of care responsibilities in such a labour market setting is of course possible, but in such cases both parents risk losing their career opportunities. 2. OECD family database (www.oecd.org/els/social/family/database) accessed October 2009. 3. Public day-care subsidies are regulated at the communal level, and the tax system varies between cantons. The mentioned study calculates income effects in the context of the city of Zurich. While there have been recent attempts to change the financing rules, similar systems still persist in other Swiss communes, especially in the German-speaking part of the country. 4. This case study derives from Claire Annesley’s Economic and Social Research Council (ESRC) funded project RES–000–22–1615, ‘Gendering Welfare Reform in Emerging Adult Worker Model Welfare States: The UK’.
9780230_245884_06_cha05.indd 94
10/5/2010 5:02:16 PM
6 Gender and Institutions of Multi-Level Governance: Child Care and Social Policy Debates in Canada Joan Grace
It could be said that to study Canadian politics is to study institutions. While much has been written about the formal and informal practices that have animated the Canadian institutional landscape, analysing macro-governing structures through a feminist lens remains remarkably understudied. To be sure, there have been important contributions. Early work by Teghtsoonian (1992) investigated the influence of federal institutions on child-care policy development, a key policy sector of the feminist movement activism, while the contribution of Vickers (1994) squarely placed the tension between feminism, women’s policy objectives, and federalism at the centre of analysis. More recently, Chappell’s (2002a) analysis of feminist policy advocacy in Canada and Australia added comprehensive and significant comparative insights about the impact of federal structures. I propose to build on this seminal work by marrying the theoretical and empirical tenets of historical institutionalism (HI) with the transformational objectives of feminism to provide a gender critique of the idea and practice of Canadian federalism. From one vantage, HI draws our attention to the institutions that comprise the state, the consequences of history regarding the institutionalization of practices, and the relative permeability of multi-level governance structures. From the other, feminist political science (FPS) brings to the analysis a gendered conceptualization of federalism and a focus on how the institutions of the state contribute to perpetuating women’s inequality. For my purposes, federalism is conceptualized as macro-political structures which 95
9780230_245884_07_cha06.indd 95
10/5/2010 5:02:24 PM
96 Joan Grace
have institutionalized power sharing and multi-level governance as ‘a system of continuous negotiation among nested governments at several territorial tiers’ (Marks 1993: 392). I apply feminist institutionalism (FI) to critically examine the principles of Canadian federalism and the practice of multi-level governance highlighted through two policy debates: the Social Security Review (SSR), which took place between 1994 and 1995, and the 1999 Social Union Framework Agreement (SUFA). These two policy events were selected because they were defining moments in restructuring the social safety net in English Canada and because they represent the concretization of neo-liberal politics. As such, the SSR and SUFA are interconnected policy developments as part of a historical trajectory towards the development of flexible intergovernmental relations. The SSR and SUFA underscore the necessity to analyse the public policy landscape from an FI perspective. Although largely gender blind, Canadian academic enquiry has often had an institutional focus amply demonstrating how federalism has structured federal and provincial intergovernmental dynamics, policy outcomes, and state–society relations (Bakvis and Chandler 1987; Bakvis and Skogstad 2002; Rocher and Smith 2003). A feminist analysis brings to the fore overlooked and under-conceptualized dimensions of the politics of multi-level governance in Canada. The SSR experience reveals, for example, the importance of looking to the dominant political discourse, which ironically obstructed a gendered understanding of child care. A gendered analysis redirects institutional attention to women’s policy agencies to assess whether they were able to promote a progressive agenda and whether state feminism was able to extend beyond the institutional confines of the two levels of government to advance the goals of the women’s movement. With respect to SUFA, FI highlights the gendered dimensions of fiscal arrangements between the levels of government and the way in which women’s voices were marginalized from policy discussions within the closed practices of accommodation associated with executive federalism. Overall, I have two objectives. The first is to make the case for FI as a critical approach to political analysis, which brings to new institutionalism (NI) a radical understanding of institutions, the political context, and policy outcomes. The second goal is to explore Canadian federalism through this rigorous lens. I ultimately conclude that the SSR and SUFA, framed by a politics of accommodation, were critical junctures which provided an opportunity to develop a child-care policy that could have addressed some of the English Canadian women’s movement policy
9780230_245884_07_cha06.indd 96
10/5/2010 5:02:24 PM
Gender and Institutions of Multi-Level Governance 97
demands, yet the result was no, or little policy outcome due to degendered intergovernmental machinations.
From historical to feminist institutionalism HI attends to the consequences of policy trajectories and the ways in which institutions mediate politics. It is an approach that has been highly productive in reminding analysts that institutions are embedded in a particular socio-political context and that history matters (Thelen and Steinmo 1992: 3). In building the case for a feminist critique of federalism and multi-level governance, I look to HI to recognize institutions as ‘the legacy of concrete historical processes’ (Thelen 1999: 382). It is a perspective, Waylen reminds, which demands careful analysis through ‘structured comparison and historical process tracing’ (2009: 246). HI is also a fruitful approach because it highlights how past choices often hinder future policy developments and institutional change (Wilsford 1994: 251–283). Path dependencies are of particular interest to social policy advocates, who often seek dramatic change to existing public policy regimes. From a feminist perspective, policy and institutional continuity pose a formidable challenge for women’s groups, who not only advocate policy reform, but who also wish to contribute to a rewriting of history that counters and subverts institutionalized practices and normalized policy relationships. This is often difficult given the complex architecture of Canadian federalism which presents to feminist policy advocates a wide stretch of government to navigate (federal, provincial, territorial, and sometimes municipal), and hence a broad array of political elites to persuade. This situation is often vexing to women’s groups, who neither possess significant policy influence, nor the financial resources to engage in sustained political campaigns. However, as Thelen (2004) notes, part of the analytical endeavour of HI is to uncover interactions among institutionally located actors or to expose negotiations and events which take place within various sites of institutional activity which open up possibilities of change. In other words, path dependencies can be disrupted or derailed due to critical events or junctures, and because institutions continue to ‘evolve in response to changing environmental conditions’ (Thelen 1999: 387). HI, however, does not theoretically consider institutions to be gendered, nor does the approach empirically analyse institutions of government and the machinery of the bureaucracy as perpetuating gendered power relations. From a methodological, theoretical, and empirical
9780230_245884_07_cha06.indd 97
10/5/2010 5:02:24 PM
98 Joan Grace
perspective, feminist theorizing intensifies HI by critically unravelling understandings of institutional effects. From this point of view, we come to understand institutions as ‘instruments of social organization that exercise collective power over a number of generations’ (Vickers et al. 1993: 133–134), with power understood to be stratified by gender, sex, sexuality, race, ability, and economic status, which is unevenly distributed within society and the economy. Indeed, one of the great strengths of FI is that gender and power are analytically anchored (Kenny 2007; Kenny and Mackay 2009) by attending to women’s different relationship to the welfare state from men’s (Vickers 1994). That is, while at times protecting and advancing women’s equality, the Canadian welfare state institutionalized women’s systematic discrimination by engendering particular attitudes about women’s roles in the private and public spheres. Women were recognized in emergent Keynesian welfare states largely as members of a gender regime ‘through the lens of the nuclear family as wives, mothers, and widows ... to reinforce a particular family form and model of social reproduction’ (Brodie 2008: 150). Although many of the most egregious, sexist laws have been repealed and the patriarchal regime has shifted and even muted in some sectors, gendered power dynamics remain relative yet steadfast. Women activists continue to confront a political landscape that is at times hostile to feminist policy goals, or is simply blasé about women’s sustained inequitable treatment and socio-economic situation. As a consequence, women’s groups persist to subvert gendered relations of power, and along the way, reveal why, and why not, governing practices either accept or obstruct their policy objectives. Although path dependencies are a policy and institutional reality, FPS cautions that the state should not be taken to be static nor overly determinative; it is ‘mutable and multi-faceted’ and hence can be a ‘terrain of struggle’ (Dobrowolsky and Saint-Martin 2005: 4). We see this in the Canadian context, with the state form shifting during the 1970s from a Keynesian welfare state influenced by social liberalism, to the more contemporary neo-liberal, social investment state. With this in mind, this study applies a rather loose conception of path dependency to be mindful of ‘unpredictable outcomes and ongoing contestation’ (Waylen 2009: 251). FI is adept in highlighting the often challenging interface between women’s groups and the state by further revealing how gendered power relations are embedded in bureaucratic rules and practices located within the state, and how women’s policy agencies, as an inside ally to the women’s movement, attempt to act as a counter-balance. From an FI perspective, informational resources and norms that structure
9780230_245884_07_cha06.indd 98
10/5/2010 5:02:24 PM
Gender and Institutions of Multi-Level Governance 99
individual participation and behaviour within institutions are understood to be gendered. Here, gender is understood in terms of the work of Scott’s two propositions that gender is a constitutive element of social relationships based on perceived differences between the sexes, and gender is a primary way of signifying relationships of power (1986: 1067). To argue that a policy outcome is gendered refers to the way in which ‘advantage and disadvantage,