50290 Agriculture and Rural Development Discussion Paper 43 The World Bank Getting Good Government for Women: A Literature Review Agriculture & Rural Development Department Leah Horowitz World Bank 1818 H Street, NW Washington, DC 20433 http://www.worldbank.org/rural Horowitz_cover.indd 1 6/10/09 5:22:30 PM Horowitz_FM.qxd 6/11/09 1:22 PM Page i Agriculture and Rural Development Discussion Paper 43 Getting Good Government for Women: A Literature Review Leah Horowitz Horowitz_FM.qxd 6/11/09 1:22 PM Page ii © 2009 The International Bank for Reconstruction and Development/The World Bank 1818 H Street, NW Washington, DC 20433 Telephone 202-473-1000 Internet www.worldbank.org/rural E-mail ard@worldbank.org All rights reserved. This volume is a product of the staff of the International Bank for Reconstruction and Development/The World Bank. The findings, interpretations, and conclusions expressed in this paper do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Rights and Permissions The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The International Bank for Reconstruction and Development/ The World Bank encourages dissemination of its work and will normally grant permission to reproduce portions of the work promptly. For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA, telephone 978-750-8400, fax 978-750-4470, http://www.copyright.com/. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA, fax 202-522-2422, e-mail pubrights@worldbank.org. Cover photo: Simone D. McCourtie Horowitz_FM.qxd 6/11/09 1:22 PM Page iii Getting Good Government for Women Contents Acknowledgments ..................................................................................................iv Abstract........................................................................................................................v 1. Introduction ........................................................................................................1 1.1 Conceptual Framework: Strengthening Accountability to Improve Governance for Women ........................................................2 1.2 Scope ............................................................................................................5 2. Women and Politics: Obstacles and Opportunities Along the Long Route of Accountability ....................................................................................7 2.1 Explaining Women's Presence in the State ............................................8 2.1.1 Women as Legislators ........................................................................8 2.1.2 Women as Administrators ..............................................................15 2.2 More Than Numbers: Women's Influence on the State ......................15 3. Translating Mandates to Outcomes: Public Administration Responsiveness to Women ............................................................................23 3.1 Women's Machinery ................................................................................23 3.2 Gender in Administrative Policy and Planning ..................................25 3.3 Anticorruption Efforts ..............................................................................26 3.4 Public Sector Reform and Privatization ................................................28 4. Decentralization and Local Politics: A New Space for Accountability? ..........................................................................................30 4.1 Women and Local Politics ........................................................................30 4.2 User Committees and Community-Driven Development ..................36 5. Discussion..........................................................................................................38 5.1 Questions for Further Study....................................................................40 5.1.1 Women and Politics ........................................................................41 5.1.2 Decentralization and Local Politics ................................................42 5.1.3 Public Sector Capacity and Responsiveness ....................................43 5.1.4 Operations and Outcomes ..............................................................44 6. Conclusion ........................................................................................................45 7. References ..........................................................................................................46 Endnotes ............................................................................................................57 iii Horowitz_FM.qxd 6/11/09 1:22 PM Page iv Agriculture and Rural Development Acknowledgments This report is a component of a project led by Agriculture and Rural Development (ARD)/World Bank, implemented by the International Food Policy Research Institute (IFPRI), and funded by the Bank-Netherlands Partnership Program (BNPP) titled "Gender and Governance in Rural Services." The task manager of the overall project was Eija Pehu (ARD). The project was coordinated by Catherine Ragasa (ARD). This paper was authored by Leah Horowitz with scholarly advice and invaluable contributions from Regina Birner. Reviews were provided by Arun Agarwal, Caroline Andrew, Marc Cohen, J. Devika, Nora Dudwick, Nathasha Hayward, Penda Mbow, Mansah Prah, Agnes Quisumbing, and Doris Voorbraak. Participants of the brown bag lecture series on gender and governance, held on January 8, 2009, provided comments and suggestions. The paper was edited by Mike Donaldson and Sonia Madhvani provided support for publication. Leah Horowitz, author of this Discussion Paper, met with a fatal accident in Ghana on May 23, 2009. It was her wish that this review would inform the research and development community during their efforts to get good government for women. iv Horowitz_FM.qxd 6/11/09 1:22 PM Page v Getting Good Government for Women Abstract This review assesses what is known and not known about the interface of gender and governance. More specifically, the purpose of this review is to identify mechanisms that have been used to make the outcomes of public sector governance more responsive to women. The review focuses on the features of institutional design that underlie service delivery, and it assesses those that strengthen women's representation in political and governance processes or build the capacity and incentives of the public sector to provide quality services that are accessible to and fit the needs of women. After presenting an organizing conceptual framework, more than 150 papers on women and gender issues in national and local politics and the public administration are reviewed to assess commonly used strategies, instruments, and institutions for gender-sensitive governance as well as the frame conditions that predict success or failure. An array of mechanisms have been used to strengthen women's voice and to incentivize attention to women, but the preponderance of the literature is focused on so-called demand-side strategies at the expense of supply-side mechanisms. Several general themes emerge from the review. First, initiatives generally fail when gender equity is not recognized to be a political project. Second, women's political empowerment cannot be divorced from their economic empowerment, and social and cultural transformation also lies at the root of governance transformation. Third, political, social, and institutional contexts will affect the shape and success of any single mechanism across multiple places. However, much remains to be learned about getting good government for women, and directions for future research are suggested. v Horowitz_FM.qxd 6/11/09 1:22 PM Page vi Horowitz.qxd 6/11/09 1:22 PM Page 1 Getting Good Government for Women 1. Introduction The global community formally recognized gender inequality as one of the major hurdles to development when it set Millennium Development Goal 3: Promote gender equality and empower women. Yet beyond the midway point between the millennium and the 2015 target date, women in the developing world still remain disadvantaged in relation to men. They are more likely to die before their fifth birthday (in some regions), less likely to be enrolled in secondary school, more likely to be employed in vulnerable sectors, and face high risk of maternal mortality (UNIFEM 2009). Mechanisms to fully address women's poverty are thus still not fully understood. Development research conducted over the last two decades suggests the need for a convergence between economic and political explanations for and solutions to women's poverty. That women have differential access to the resources, assets, and opportunities necessary for social and economic well- being is now well established. Researchers have demonstrated that the distribution of consumption resources within households is frequently unequal and is dependent on an individual's internal bargaining power (Haddad, Hoddinott, and Alderman 1997; Sen 1990a). This bargaining power has been shown almost always to favor men, since it is affected by economic entitlements that have historically been biased against women, such as education or the resources partners bring to the marriage, as well as by qualitative gender norms, such as perceptions of relative status or who is "contributing" to household output (Quisumbing 2003). Women also often have less--or less secure--access to productive inputs, partially due to intrahousehold dynamics. Women's differential access to or control over agricultural inputs such as land, hired or household labor, credit, and productivity-enhancing inputs has been particularly well documented (Goldstein and Udry 2008; Doss 2001; Quisumbing et al. 2001; Meinzen-Dick et al. 1997; Goetz and Gupta 1995; Gladwin and McMillan 1989; Boserup 1970). Because one way that unequal allocation practices can exacerbate women's poverty is by limiting the range of resources women can draw on to meet their needs and their market engagement, development agencies have primarily responded by building women's private entitlements to goods such as credit, agricultural inputs, and education (Quisumbing 2003; Quisumbing et al. 1995). It is assumed that such interventions will strengthen women's position within the household and the market. However, allocative inequalities in the private sphere imply potential political failures too. Women's differential access to resources and opportunities means that gender-blind public interventions, even if targeted to the poor, cannot be assumed to reach women equally or well. Policy mechanisms targeted at household-level consumption, such as food aid or education subsidies will benefit women and men 1 Horowitz.qxd 6/11/09 1:22 PM Page 2 Agriculture and Rural Development differently; economic livelihood programs that ignore differences in access to or the need for productive resources are likely to be biased against women. The importance of public services to development and poverty alleviation has been well established (Besley and Ghatak 2004; World Bank 2003), and the particular importance of good service provision for women has been highlighted (UNIFEM 2009). The good governance agenda ascendant in the international finance and donor community since the mid-1990s emphasizes the proper functioning of institutions as a precondition for sustainable development. If eliminating women's poverty is important to development and if the public sector is to play an important role in combating poverty, then how to make the provision of public goods gender sensitive is an essential development question. The purpose of this review is to identify mechanisms that can make the outcomes of public sector governance more responsive to women. While taking into account technical issues of gender-sensitive service design and delivery, the review assumes that services are conceived of and delivered in broader institutional environments in which the mandates, priorities, implementing practices, budgets, and cultures of policy-making and implementation institutions (public and private) impact governance outcomes. The role of women's voices in public arenas and women's involvement in public decision making also impact decision makers' awareness of women's exclusion from entitlements and rights and help to define the solutions. The review thus focuses on the features of institutional design that underlie service delivery, to assess those that strengthen women's representation in political and governance processes or build the capacity and incentives of the public sector to provide quality services that are accessible to and fit the needs of women. In every case, attention is paid to the frame conditions and contextual factors that mediate the answers to these questions in particular places and circumstances. Ultimately, gender equality in governance demands that governments ensure that women and men have equal entitlement to resources and equal opportunities to flourish in their productive and private lives. 1.1. Conceptual Framework: Strengthening Accountability to Improve Governance for Women How can individuals in a society act to improve public governance and, through that, the services, rights, and protections provided by the government? To develop answers to this question, this review uses a conceptual guide that overlays a gender lens on the framework for service provision introduced in the World Development Report 2004: Making Services Work for Poor People (World Bank 2003). The World Development Report's framework starts from the proposition that successful services for poor people emerge from "institutional relationships in which the actors are accountable to each other" (46). Accountability is understood as a cycle involving (1) delegation of responsibility from a principal to an agent, (2) provision of finance for that goal, (3) performance of the delegated task, (4) information about performance ("giving an account"), 2 Horowitz.qxd 6/11/09 1:22 PM Page 3 Getting Good Government for Women Figure 1.1. Key Relationships of Accountability The state Politicians Policy makers C om e of accountabi g rout ice lity Lon pa Vo ct Citizens/clients Short route Providers Coalitions/inclusion Client power Management Nonpoor Poor Front line Organizations Services Source: World Bank 2003. and (5) enforceability/sanctions. The three types of actors involved in the chain of service delivery--citizens, policy makers, and service providers--are linked in relationships of accountability, as depicted in Figure 1.1. Yet despite the generic nature of the actors in Figure 1.1, Anne Marie Goetz, the lead author of the United Nations Development Fund for Women's (UNIFEM) Progress of the World's Women 2008/2009 report, Who Answers to Women? Gender and Accountability, demonstrates that the depicted relationships are not gender blind.1 Rather, women are frequently disadvantaged in using any of these three accountability systems due to their subordinate status to men at home and to men as decision makers and power holders in the community. They may not be free to use either their political voice or their money according to their own preferences, but must mediate their demands of governments through men. Getting good governance for women, then, depends on making accountability systems work for women within the specific social contexts in which they can operate. This framework suggests that the major entry points for improving gender- sensitive service delivery follow the three sets of relationships shown in Figure 1.1. 1. Women to policy makers: Using political voice. Citizens hold their elected representatives accountable through a process that has been labeled "voice." They delegate the task of public governance to their representatives and specify the nature of this delegation by articulating their values and preferences through collective action and the political process. Are women equally engaged in this process of voicing their heterogeneous preferences? Are policy makers (both male and female) equally receptive to their 3 Horowitz.qxd 6/11/09 1:22 PM Page 4 Agriculture and Rural Development demands? Do women have access to information about the performance of their political representatives on issues that are important to them, and can they use the range of vertical enforceability mechanisms, including both periodic elections and protest, to reward or sanction policy makers for their political performance? Strategies that strengthen the ability of citizens to have their voice heard in the political process and to hold policy makers to account are frequently referred to as "demand-side" mechanisms because they affect citizens' ability to demand good government. 2. Policy makers to service providers: Compacts or instructions. Once policy makers set an agenda, they authorize institutions in charge of implementation--for example, health services, education authorities, or agriculture ministries--to implement it. If service providers are outside the state, the delegation of responsibilities may take the form of a compact, or contract. If they are inside the state, policy implementers receive legal or administrative instructions. Do these instructions include a mandate for gender-sensitive service delivery? Do service providers have the capacity to implement the given agenda, including the availability of time, money, information, skills, and other human capital to cater to women's and men's specific needs? Do they report back to elected representatives on their results in achieving women's human rights? Are gender-disaggregated monitoring and evaluation data available to enable gender-responsive reporting systems within the public administration, and do these activate management sanctions for poor performance? Do anticorruption strategies take into account the types of corruption that disproportionately affect women? Strategies that build the capacity of the public administration or a nongovernmental provider to efficiently, equitably, and honestly perform its tasks are sometimes considered "supply-side" mechanisms, since they build the supply of good governance. 3. Women to service providers: Client power and choice. Citizens also interact directly with service providers as their clients and can use this forum to reveal preferences and to monitor outputs. Privatization and decentralization of frontline service delivery relies on the assumption that citizens will use market tools such as user fees or moving between competing service providers to reward service providers who meet their preferences. In this case, the strategy employed to improve service delivery is improving citizens' choice in service providers. Do women have the ability to use these market mechanisms, given their often subordinate status in the market and their inability to make decisions about the use of household income? Do private or decentralized institutions have the same incentives to serve women as the public sector does? This paper surveys what is known about how to improve the gender sensitivity of public service.2 Specifically, it looks at how all the male and female actors in this framework--as citizens, public officials, and service providers--can strengthen accountability systems to avert gender-specific service failures and to provide good government for women. It also gives attention to the social contexts in which women operate as they attempt to access various accountability mechanisms. 4 Horowitz.qxd 6/11/09 1:22 PM Page 5 Getting Good Government for Women Box. 1.1. One Review Can't Do It All Although such a broad literature search requires the setting of boundaries, it is nonetheless important to note four main limitations posed by the scope as defined: 1. Gray literature. By reviewing only peer-reviewed literature, this paper excluded much of the gray literature, especially project documentation, which may be the earliest source of information on governance innovation. 2. Sectoral literature. Sectoral literature may also reveal particularly innovative mechanisms for improving governance for women. It would be advantageous to review sector-specific literature at a later date. 3. Nonanglophone literature. Focusing on English language literature may bias the review away from observing trends in particular regions, such as Latin America and francophone Africa. 4. Faltering or crisis states. The governance relationships discussed in the paper assume a basic level of government functionality and accountability. The particular challenges inherent in governing a crisis or postcrisis state are not treated here. 1.2. Scope The political science, public administration, development, and feminist literatures were surveyed for empirical studies of gender-relevant governance reforms in developing countries.3 The search focused on reforms targeted at changing the nature of political inclusion and/or political and administrative accountability rather than those targeted at gender disadvantage in a specific sector (e.g., education, water, land) because it is assumed that changing underlying governance relationships will also improve sectoral governance. Where applicable, research conducted in industrialized countries was consulted for purposes of identifying the origins of specific discursive threads. Both quantitative and qualitative methods were included, and recent work was favored. Only peer-reviewed literature has been included (reports, discussion papers, presented papers, journal articles, and books), although the review rigor is sometimes minimal (e.g., discussion papers). Table 1.1 indicates the methodological and regional focus of the 175 studies reviewed.4 Table 1.1. Count of Peer-Reviewed English-Language Articles on Gender and Governance in the Global South Quantitative Qualitative Theory/Overview Totals Africa 3 54 n/a 57 Asia 7 22 n/a 29 Latin America 6 12 n/a 18 Cross-regional 14 9 48 71 Totals 30 97 48 175 5 Horowitz.qxd 6/11/09 1:22 PM Page 6 Agriculture and Rural Development The remainder of the paper is organized as follows. Section 2 reviews the substantial literature on women in politics, looking at the numerical and substantive aspects of women's formal representation in government. Section 3 provides an overview of the fairly limited literature on mechanisms to increase public sector capacity to serve women. Section 4 discusses the gendered aspects of decentralization and local government reform. The discussion in section 5 points to agreements as well as gaps in the literature and suggests research questions for future work on this topic. Section 6 concludes. 6 Horowitz.qxd 6/11/09 1:22 PM Page 7 Getting Good Government for Women 2. Women and Politics: Obstacles and Opportunities Along the Long Route of Accountability When the composition of decision-making assemblies is so markedly at odds with the gender make-up of society they represent, this is clear evidence that certain voices are being silenced and suppressed. Anne Phillips (1991) The majority of literature on the topic of gender and good governance centers on women as governors or public officials. Observing the feminist canon, this is a logical finding. The second wave of Northern feminism ushered in a highly politicized movement, and feminist political observers became keenly aware of the gender disparity in political representation around the world.5 Increasing the number of women who walk the halls of power has been described as a democratic imperative, and activists question the legitimacy of polities that tolerate political inequality (Htun 2004). Two seminal contemporary milestones for gender equality--the 1979 Convention on the Elimination of All Forms of Discrimination Against Women and the Platform for Action adopted by governments at the 1995 Fourth World Conference on Women in Beijing--endorse affirmative action for political positions. In addition to the democratic rationale for formally empowering women, one can also recognize within the women-in-politics debate a second, often implicit, assumption: increasing some women's political involvement can improve the way government serves other women in the general population. Following the conceptual framework in Figure 1.1, this effect would occur through strengthening women's political voice. Policy makers sit at one point of the framework of service delivery relationships. They are key intermediaries in what the World Bank (2003) calls the "long route of accountability," a stylized sequence in which citizens voice their preferences to policy makers, who then act to build the accountability and capacity of service providers and other public actors accordingly. The argument then follows that when such policy actors are female, they are more cognizant of and/or receptive to women's needs and preferences, and they may be more willing to allocate resources to meeting them. Hassim writes (2006b, 172), "without representation in legislatures, women citizens have a diminished ability to hold governments accountable." Accordingly, over the past two decades, political scientists have developed a significant body of knowledge on women and politics.6 Although most of the early work focused on the western democracies, in the last decade a handful of studies have made use of expanded data sets to test the determinants of women's representation generally accepted in industrialized democracies for 7 Horowitz.qxd 6/11/09 1:22 PM Page 8 Agriculture and Rural Development a larger set of countries, including less developed countries. A surprisingly small number of case studies on women's entry into politics in developing countries and several major comparative works deepen and contextualize the elements highlighted in the quantitative studies (e.g., Bauer and Britton 2006; Nelson and Chowdhury 1999). Yet in terms of content, much of the research on women's political representation in developing countries to date has concentrated on the (arguably easier) quantitative question: what factors increase the numbers of women in the structures of government? There is limited research on the qualitative impact of women's political representation, that is, the link between women's presence and their influence over political outcomes. Available literatures on these topics are now considered. 2.1. Explaining Women's Presence in the State 2.1.1. Women as Legislators At the beginning of 2009, women made up only 18.4 percent of world parliaments. Yet within this average lies significant variation: whereas Qatar and Saudi Arabia have no women in parliament, Rwanda and Sweden have nearly or more than half female parliaments (Inter-Parliamentary Union 2009). Cross-country studies have exploited this variance to delineate the determinants of women's numerical presence in legislative bodies across the world. Factors examined include those affecting both the supply of female candidates, which is determined by women's structural access to opportunities, and political and cultural demand for female representation (Randall 1987). Labor force participation and greater education does not necessarily help women enter politics No study that included measures of women's educational attainment as a predictor of female presence in national legislatures found it to have a significant effect (Paxton and Kunovich 2003; Kenworthy and Malami 1999; Matland 1998; Paxton 1997; Moore and Shackman 1996), although this may be partially attributable to the lack of international data measuring women's achievements in particularly relevant fields, such as law (Paxton and Kunovich 2003). Women's labor force participation has also not been shown to be a strong predictor of women's political presence in developing countries (Kenworthy and Malami 1999; Matland 1998).7 As for the direct impact of national wealth on women's standing, none of the studies that included an aggregate measure of economic development found a significant or positive effect on women's representation levels (Paxton and Kunovich 2003; Kenworthy and Malami 1999; Paxton 1997; Moore and Shackman 1996). On the other hand, in the only study explicitly examining representation trends in developed versus developing countries, Matland (1998) found striking differences across the two samples, which led him to posit the existence of a development threshold. These results suggest that women's access to politics is not simply a matter of development or of increasing the numbers of women participating in the workforce or higher education. Even in places where women have made 8 Horowitz.qxd 6/11/09 1:22 PM Page 9 Getting Good Government for Women significant gains in socioeconomic status, they do not appear able to translate socioeconomic power into political power. In recent years, many have examined the institutional barriers and resistant political cultures that seem to "block women in the tracking process to the top" (Paxton 1997, 460). Electoral and party structures, affirmative action mechanisms, and cultural attitudes about the idea of women in public life are all part of the story on women's political participation in the developing world. Instead, the structure of electoral systems is what makes the difference Evidence from studies of the wealthiest democratic nations strongly suggests that the structure of the electoral system is one of the most important factors, if not the most important factor, in accounting for variations in women's representation (Matland 1998; Rule 1987; Lovenduski and Hills 1981). Specifically, female politicians are elected in much greater numbers under proportional representation systems where voters choose among closed party lists in multimember districts rather than individual candidates in single- member districts.8 Women's success in proportional representation systems has traditionally been attributed to the incentives political parties have to front more female candidates. Multimember districts and proportional representation enable parties to include both women and men on the ballot without significant costs to intraparty peace.9 The zero-sum nature of majoritarian electoral politics creates incentives for parties to front those they feel to be the safest candidates, which are often men.10 Does this intuition hold in developing countries? The available evidence strongly suggests that it does. Party list/multimember district systems were found to be strong and significant predictors of women's national representation in all but one cross-country sample that included a number of developing countries (Yoon 2004; Lindberg 2004; Paxton and Kunovich 2003; Kenworthy and Malami 1999; Reynolds 1999; Paxton 1997). Bauer (2004), in her case study of women's relative success in gaining entry into Namibian politics, attributes a large measure of women's electoral gains to the use of a closed-list proportional representation electoral system (combined with party quotas).11 Proportional electoral systems have also been found to positively influence the numbers of women serving in mesolevel legislative bodies (Vengroff, Nyiri, and Fugiero 2003).12 In Senegal, when the electoral system for local governments changed to a proportional representation ballot, the number of women rural councilors jumped from 12 to 25 percent (Patterson 2002). Women won more than a third of party list seats in KwaZulu-Natal in South Africa's 2000 election, but only 12 percent of ward seats (Mbatha 2003). Political parties--as the gatekeepers to candidate selection--are critical Because candidate selection operates through parties (at least in democracies with active multiparty competition), gender dynamics within political parties are extremely important.13 Political parties are often described as the "gatekeepers" of women's political inclusion (Norris and Lovenduski 1993). Variations in such aspects as candidate selection mechanisms, recruitment, branch structure, systems of mentoring and leadership development, and the procedures by which party activists make it to the national executive 9 Horowitz.qxd 6/11/09 1:22 PM Page 10 Agriculture and Rural Development committee are all likely to strongly affect the political prospects of female party members (Goetz and Hassim 2003).14 Women's wings or sections can also help promote women's interests within parties, or they can simply segregate women's influence and activities while remaining excluded from central decision-making bodies (Gray 2003; Geisler 1995). Women who reach elite positions in political parties may also be able to aid other female candidates, though the specific form of this aid is shaped by opportunities provided by the electoral system (Kunovich and Paxton 2005).15 Unfortunately, while women in many countries activate their political participation as party members, the record of translating this participation into formal power within the party is poor. More broadly, the level of party institutionalization and internal democracy matters. Several authors have noted the difficulty that women have had in institutionalizing their political legitimacy in informal but centralized political party organizations; these rely mainly on patronage systems and a dominant individual leader rather than on transparent rules for candidate selection and choosing policy priorities (Compaore 2005; Goetz 2002; Waylen 2000). So- called benevolent autocrats can use centralized leadership to overcome conservative opposition to women's presence, but women's positions can be just as easily eroded when patronage is withdrawn. Informal rules within minimally institutionalized parties also exclude women in practice. In the Thai case, the prerequisite for being considered an attractive candidate depends on how strong a local network of canvassers one commands at the constituency level. Such networks are constituted and controlled almost wholly by men (Bjanegard 2008). Despite their theoretical importance, the gendered internal cultures of political parties have not been well studied outside of western democracies. The masculinity of party cultures in both the North and the South and the even physical danger that political competition within this environment can pose for women has been noted (Goetz 2007; Beck 2003; Geisler 1995). Concerning political party orientation, there is substantial evidence of left- leaning or progressive political parties being more receptive to feminist public policy goals as well as to higher levels of female representation, at least in late-twentieth-century western nations (Beckwith 2000; Kenworthy and Malami 1999; Lovenduski and Hills 1981). Yet Goetz and Hassim (2003) note that women's movements may still cautiously engage with leftist political parties. Some leftist parties have a tendency to "alter feminist demands and use them for their own purpose." Thus they cite evidence in which parties have made "some short-term concessions in areas that are less costly and less contentious . . . while not necessarily shifting the basic gender equalities in access to the labor market" (9). Democracy does not necessarily create real spaces for women The relationship between women's representation and prevailing political ideology is not uncomplicated. Quantitative studies do not find a significant effect of democratization on the level of women's representation in cross- regional data sets (Paxton and Kunovich 2003; Reynolds 1999; Kenworthy and 10 Horowitz.qxd 6/11/09 1:22 PM Page 11 Getting Good Government for Women Malami 1999; Paxton 1997). In fact, Tripp and Kang (2008) argue that the use of gender quotas in many nondemocratic regimes, which can be implemented for symbolic purposes, to create new avenues for patronage or to gain legitimacy within the global community increasingly make regime type unimportant for explaining the numbers of women in parliament. Conversely, women can be excluded from newly democratic politics when the rise of multiparty systems raises the cost of campaigning, requiring candidates to draw on the informal patronage networks for campaign financing that women are often excluded from (Brown 2001). Women in authoritarian systems may have a relatively high degree of representation without having any real power (Norris and Inglehart 2001). A substantial literature also exists on gender and transitions that examines both the role of organized women in the breakdown of nondemocratic regimes as well as the gendered outcomes of democratic transitions. Comparative work has identified a range of contextual factors that influence whether women's mobilization during periods of transition translates into sustained activism and positive gender outcomes afterward (for an overview, see Waylen 2007; Waylen 2003).16 Although there are many cases in which women's involvement in national liberation struggles failed to translate into their increased political representation in new states, regime change can also break existing networks of state-related clientelism and provide women with new opportunities to represent themselves (Tripp 2001a; Tripp 2001b; Tripp 2000). Rwanda, which has boasted the world's highest percentage of women in parliament in the years since its genocide, appears to exemplify this point (Powley 2006). It is important, however, to avoid applying generalizations about women's political opportunities in new or postconflict/transition states to those in long-standing democracies, where women can continue to face frustrating hurdles and entrenched interests. Public perceptions of women's capabilities matter to their political inclusion Despite the strong theoretical evidence for the importance of cultural preferences and social norms in explaining disparities in women's representation, little early work on women in politics privileged this explanation. Many of the first cross-national statistical studies found little explanatory power for the importance of national gender ideologies (e.g., Kenworthy and Malami 1999; Matland 1998; Moore and Shackman 1996). Weak data may be at fault. Most of these studies had only crude proxies of ideology. The two studies that did use an explicit measure of national gender ideology derived from the 1995 World Values Survey found that ideology strongly affects the number of women in national legislatures (Paxton and Kunovich 2003; Norris and Inglehart 2001). Clearly, the way in which the general public imagines women's capabilities and opportunities matters to their political inclusion. Overriding sentiments about gender equality in a society also influence voters' and political elites' impressions about how viable women are as candidates and leaders. Indeed, one extensive review of the social science 11 Horowitz.qxd 6/11/09 1:22 PM Page 12 Agriculture and Rural Development literature suggested that where attitudes toward female leaders or potential leaders are less positive than toward males, it is due to a specific perception of incongruity between the characteristics of women and the requirements of leadership roles, rather than to a generalized negative attitude toward women (Eagly and Karau 2002).17 The literature disagrees over who is actually carrying this prejudice: parties or voters. Polls throughout Latin America have shown voters to be supportive of women in government (Goetz 2005), and there is evidence elsewhere that voters consider women good governors (e.g., Honour, Barry, and Palnitkar 1998). However, Duflo and Topalova (2004) show women's leadership to be systematically "underappreciated," at least in Indian local government. In a nationally representative survey of public service quality and satisfaction, they found women leaders in local government able to deliver equivalent or better public goods than men and less likely to take bribes. Yet both male and female respondents were less likely to be satisfied with the public goods they received in villages with female presidents. Although no researcher has conducted the same type of empirical study of women in national government in the developing world, such results suggest that women face an uphill battle in politics where populations generally expect women to be less effective leaders and are slow to adjust even in the face of facts. . . . But these perceptions can change with exposure At the same time, evidence suggests that gender norms are dynamic over time. Qualitative studies of the political environment in Rwanda have emphasized the importance of women in power as role models for younger women and for transforming general societal attitudes about the capability of women (Powley 2006). Beaman et al. (2008) found that in India, exposure to a female panchayat leader weakens men's stereotypes about gender roles in public and the home, and alters their perceptions about the effectiveness of female leaders. Eagly and Karau (2002) call this a "dynamic stereotype": as women continue to adopt the characteristics that have been traditionally associated with leaders, the perceived disjuncture between women's characteristics and those of a leader diminishes. Calls for affirmative action seek to bring to power a "critical mass" of women, partially in the hopes of fast-tracking this process of cultural change. Supporting this claim is Beaman et al.'s (2008) finding that, after 10 years of the reservation policy in India's local government, women are more likely to stand for and win free seats in villages where the chief councilor position had been reserved for women in the previous two elections. Affirmative action mechanisms can help women enter politics but may undermine their legitimacy to govern Gender-based affirmative action mechanisms have become an increasingly prominent solution to the underrepresentation of women in electoral politics.18 In Africa, Ghana introduced a quota system for women in 1960, and after recently implementing a quota system, Rwanda superseded Sweden for 12 Horowitz.qxd 6/11/09 1:22 PM Page 13 Getting Good Government for Women the highest level of women's parliamentary representation in the world. Argentina enacted the first gender quotas in Latin America in 1991, and 11 other Latin American countries followed in the 1990s. Today, 97 countries currently have some combination of constitutional, electoral, or political party quotas in practice, according to the International Institute for Democracy and Electoral Assistance (IDEA), which maintains a Global Database of Quotas for Women.19 Unlike some of the other factors that have been identified as influencing the levels of women in government (e.g., gender norms, political culture, economic development), institutional rules are viewed as relatively easy to change and therefore as a promising gender-based governance intervention (Hassim 2006b). Gender quotas have thus been institutionalized in the good governance program that receives Northern donor support. Affirmative action regimes vary along two major axes. First, who mandated? Legal gender quotas are mandated either by the constitution or by electoral law, while in many places quotas are adopted voluntarily by political parties themselves, effectively mandated by party leadership. Second, who is promoted? Nearly 170 parties in 62 countries employ gender quotas for the nomination of candidates to be placed on party ballots.20 The other major form of positive discrimination is seat or constituency/district reservation. Reservation systems vary according to whether the seats are filled by a direct election processor indirect election process, such as appointments (Dahelerup 1998). Most scholars discuss the success of affirmative action systems in terms of the ability of a quota regime to actually translate into increased numbers of women in power (Tripp and Kang 2008; Htun 2004). Whether the women brought to power through positive discrimination are politically effective (i.e., able to implement their chosen political agenda) or consciously seek to improve the well-being of women is also an important question. Party list quotas. At the national level, the evidence suggests that quotas on party lists can indeed increase women's numerical representation, but only with proper frame conditions and full implementation. Cross-national analysis shows quota legislation to be most successful in closed-list proportional representation systems because, in such systems, the party controls who is put on and selected from the electoral list and can implement this power to aid female candidates (Htun and Jones 2002). The greater the number of legislators elected per district, the greater the impact of quotas seems to be (Gray 2003; Jones 1998). Meaningful quota legislation in proportional representation systems must also require that gender placements be in electable positions (Jones 2004). Examples of countries that have passed mandates for list position include Bolivia where women must be every third candidate on an electoral list, Paraguay where they are every fifth, and Argentina where they must be placed in an "electable position" (or at least third on the party list). Many women's movements, especially those associated with the 50/50 campaign led by the Women's Environment and Development Organization, argue that genuine gender equality within party returns is assured only with the use of a 13 Horowitz.qxd 6/11/09 1:22 PM Page 14 Agriculture and Rural Development zebra list, in which men and women alternate from the top to the bottom of the ballot and therefore win equal percentages of seats proportional to the party's win. When women are placed at the bottom of the list or when voters can select from among a menu of candidates as in open-list systems, women generally fare poorly unless they are well-known (Tinker 2004). The legal mandate for the quota and associated sanctions for noncompliance are also important determinants of its success (Jones 1998). In Costa Rica, legislation that "merely exhorted parties to voluntarily increase the participation of women in public office [e.g., the 1990 Law of Real Equality] fell on deaf ears, exercising no significant impact on the election of women" (Jones 2004). Later, when a minimum percentage quota was coupled with a placement mandate, the number of women in the parliament increased greatly. Globally, 15 countries have constitutional quotas for national parliaments, and 42 countries regulate election law quotas (IDEA 2007). Finally, quotas must be implemented to have an effect. The literature and policy debate has focused largely on the politics surrounding the introduction of quotas while neglecting the implementation process (Krook 2006). As Dahelerup writes, "In the worst case scenario, quotas may be introduced after a heated debate, but then have no effect on increasing women's representation because there are no mechanisms to ensure their implementation" (1998, 3). Gray (2003) describes such a scenario in Argentina. Many writers who focus on implementation issues also highlight the importance of sanctions for noncompliance, such as the withholding of government campaign subsidies, a method used to good effect in France, Argentina, Costa Rica, and Bolivia (Goetz 2005; Gray 2003). Reservations. The small literature that exists on reservations for women in developing countries has focused on whether the institutional structure of the reservation system actually politically empowers women. Of the 32 countries using some form of affirmative action that Htun (2004) examined, only 3 out of 22 electoral democracies used reservations.21 All 10 of the nondemocracies used reservations.22 As she notes, "part of the motive for choosing the right [representational] remedy is to reinforce connections between leaders and the base that nourishes them politically . . . Regimes that lack a commitment to substantive representation have little incentive to promote the right remedy (450)." Goetz (2002) and Hassim (2006a) note that in Uganda, President Museveni's nominal commitment to gender equality brought hundreds of women into office through reserved seats in parliament and local government. Yet Goetz suggests that "the political value of specially created new seats has been eroded by their exploitation as currency for the National Resistance Movement's patronage system, undermining women's effectiveness as representatives of women's interests once in office (549).23 Isolating women from a popular base (i.e., strengthening their upward accountability to regime elites rather than downward accountability to constituents) can accentuate accusations of tokenism--that women are less effective and less accountable representatives than those elected on open tickets. 14 Horowitz.qxd 6/11/09 1:22 PM Page 15 Getting Good Government for Women Where women are appointed rather than elected to reserved seats, the links between representative and constituent are further weakened (Kunovich and Paxton 2005). The ruling parties in Pakistan and Bangladesh appoint women to seats in parliament in proportion to party wins, which serves simply to consolidate government majorities and undermines the perceived legitimacy of the women who fill them (Goetz 2005). In Tanzania, each political party ranks female candidates for the reserved seats and then sends a list to the National Electoral Commission for appointment. Special seats were also found to diminish women's chances of entering parliament via normal seats (Meena 2003). Finally, it should be noted that, although affirmative action has been an important rallying point for women's movements around the world, feminist scholars have offered strong critiques of such methods. Some authors suggest that women's political mobilization around quotas represents a deceptively easy but ultimately ineffective focus on women's numerical presence in politics and that it diverts attention from more transformative goals to challenge deeply entrenched institutional biases and ways of working that prevent women's political effectiveness (Goetz 2002; Hassim 2006a).24 Where affirmative action mechanisms are used "as a means of incorporating potential political protest into a manageable form" (Tamale 1999, 25) or where women's participation ends up bolstering an objectionable status quo, they may make it harder for women to gain real legitimacy.25 Such arguments allude to bigger questions about women's political effectiveness, a topic that is considered next. 2.1.2. Women as Administrators The literature on women in leadership positions in the civil service is sparse, but what does exist indicates that women face extraordinary challenges in reaching management and leadership positions in most countries (Omar and Ogenyi, 2004; Zafarullah 2000; Turner and O'Connor, 1994). Some countries, such as Bangladesh, have implemented gender quota or target systems for public employment, but evidence suggests that these are often capriciously administered, especially where there is no genuine commitment to gender equity from the administration (Zafarullah 2000). In such cases, bureaucratic quotas can be a substitute for questioning basic assumptions, objectives, or practices. In others, such as in India and postapartheid South Africa, affirmative action in the context of a broader agenda for political and social inclusion appears to have had a genuinely transformative effect (Rao and Kelleher 2005). As well, further study of the gender-differentiated impacts of civil service restructuring would be useful because practitioner evidence suggests that retrenchment often affects women civil servants disproportionately and can cause the elimination of informal gender equality practices in the bureaucracy and the marginalization of gender issues. However, on the topic of civil service reform, the sole empirical work is Rama's (2002) ex ante analysis of the gender impacts of public sector downsizing in Vietnam. 2.2. More Than Numbers: Women's Influence on the State The literature on women and politics in developing countries has developed a substantial knowledge base on the institutional barriers and solutions to women's presence in the political sphere. Yet the easy model described in 15 Horowitz.qxd 6/11/09 1:22 PM Page 16 Agriculture and Rural Development this literature--that increasing women's numerical presence in politics assures the representation of (poor) women's interests in policy making and implementation--makes a number of unexamined assumptions. In terms of the relationship between citizens and politicians, does a political community of women exist that has unitary interests defined by sex/gender rather than by other identity characteristics such as class, ethnicity, geography, and the like? If so, can groups of disempowered women effectively voice their claims and/or hold female representatives accountable? Are women (and/or men) in power linked and receptive to such a feminist constituency? Are female (or feminist) politicians politically effective? If they do have feminist agenda items in mind, are they able to implement these within the power structures of the political arena? In practice, these are empirical questions, and the available evidence must be assessed. Do women vote for women? Assuming a shared perspective among women as a unified group is a key basis for the critical mass argument widely employed by international organizations, transnational women's networks, and party activists.26 However, the empirical evidence for this claim should be tested. Western political theorists have debated the question of whether women should and do represent women for several decades now (Swers 2002; Dovi 2002; Mansbridge 1999; Swers 1998; Thomas 1994; McAllister and Studlar 1992; Thomas 1991).27 Only a handful of studies address how this issue plays out in developing countries. Female legislators in multiple Latin American parliaments have been found more likely to sponsor women's rights legislation and often to emphasize children and family bills (Schwindt-Bayer 2004; Taylor-Robinson and Heath 2003; Jones 1997). Female parliamentarians in Rwanda, after the massive gains for women in the 2003 election, saw a clear advocacy role for themselves in holding their government accountable on issues of importance to women and children. Women parliamentarians, especially the female chair of the Budget Committee, have been credited with, in particular, advocating for increased spending on health and education (Devlin and Elgie 2008; Powley 2006). At the local level, Chattopadhyay and Duflo (2004) found that local council presidents in India invest more in infrastructure than is directly relevant to the needs of their own genders.28 But assuming that gender is a fixed identity that causes women to legislate on some shared and well-defined set of interests appears to be an oversimplification. Qualitative reports from India suggest that women representatives often align their policy emphasis along caste rather than gender lines, which means increasing women's representation has only minimally shifted the caste and class hold on power (Vyasulu and Vyasulu 2000). In southern Africa, "women politicians find the idea that they should support a women from an opposition party, because she is a women and against their own party candidate, even if he might be a man, ludicrous" (Geisler 1995, 572). And Chattopadhyay and Duflo (2004) argue against assigning an inherent feminist altruism to the female presidents they surveyed. They suggest that it is not that individual women are particularly more 16 Horowitz.qxd 6/11/09 1:22 PM Page 17 Getting Good Government for Women responsive to the needs of some "gender constituency" in their communities, but rather that it is because women leaders' own preferences are more aligned with the preferences of other women that they end up serving them better. This literature stresses the importance of parsing out which sectors powerful women really represent (e.g., class, region, ethnic group, family). Whose interests are being served by women in politics is not always obvious. And do women act for women? Another interesting set of papers has focused on the behavior of women in the front line of service provision, or the so-called street-level bureaucrats who exercise significant discretion in who gets services (Lipsky 1980).29 At issue is whether the numerical presence of women in service delivery organizations (passive representation) translates into better service to women clients (active representation) (Wilkins 2006). The limited evidence of active representation for gender in developing countries presents a mixed picture. Goetz (2001) shows that women field- workers and managers in two rural credit and development programs in Bangladesh did identify with some of the problems of their poor women clients and acted as advocates for them within their organizations. They were also more willing than male development workers to engage with clients on matters that overstepped the bounds of the credit program, such as domestic violence, reproductive health, children's well-being, and property rights. Similarly, in Tendler and Freedheim's (1994) study of a highly successful community health program in northeastern Brazil, one key factor was that the female preventative health workers were able to build trust with their women clients by helping with seemingly unrelated household tasks, such as cooking and child care, that men did not have access to. Female farmers preferred female extension agents in Tanzania (Due, Magayane, and Temu 1997). But studies of nurses and midwives in maternity wards in both South Africa and Benin document critical, judgmental, or abusive behavior of women nurses toward female patients, including significant levels of physical abuse (Walker and Gilson 2004; Jewkes, Abrahams, and Mvo 1998; Sargent 1989). Whether sex becomes a salient variable for active representation depends on the political and institutional context. In Brazil, the time-intensive care practices of female preventative health workers received positive reinforcement within a broader program of state government reform that focused on showing public servants respect and trust. In contrast, implicit incentives in the microcredit program in Bangladesh rewarded staff who rapidly disbursed loan funds and recovered interest payments, which discouraged many from targeting poor women who were perceived as risky investments. Gender biases in the structures and cultures of the development organizations there also prevented women staff, who were in the minority in the organizations, from promoting women's rights more aggressively. Walker and Gilson (2004) locate nurses' abuse of patients in South Africa within a broader context of class and racial struggle born of both apartheid and demoralizing working conditions. They explain that "nurses are engaged in an unremitting struggle to claim a status and respect as middle class professionals 17 Horowitz.qxd 6/11/09 1:22 PM Page 18 Agriculture and Rural Development within environments in which political, professional, historical, and personal factors continuously undermine this claim" (1259). They deploy violence as a means of creating social distance from their poor clients and of maintaining fantasies of identity and power. Violence then becomes commonplace because managers do not take action against nurses who abuse patients. Several scholars also note that institutional contexts that embrace field-worker discretion may increase the gender-responsiveness of service provision. Tendler and Freedheim (1994) and Tendler elsewhere (1998) attribute a whole range of programmatic successes to governance reforms that increase worker discretion and, in so doing, create an unusual sense of "calling" among state workers and that elicit strong worker commitment to the job. They suggest that this type of management conforms to contemporary theory on industrial performance in advanced industrial countries, which "has been discovering the importance of increased worker autonomy and discretion in explaining high performance." Unfortunately, the development literature "has been concerned with reducing the discretion of the government worker or manager--because of his reputedly `inherent' propensity to serve his own interest rather than the public good" (Tendler and Freedheim, 1994, 1772). In addition to lowering worker satisfaction, Goetz notes, "these constraints go against the spirit of rural development service, where there is a huge variety in local conditions, requiring flexibility of response, and where field workers Box 2.1. Can Women Combat Corruption? The argument that women in power can be an instrument for gender-sensitive governance has been supplemented by a second case for the instrumentality of women in politics. A pair of cross-country studies finding that a greater representation of women in parliament is associated with lower levels of corruption, launched women into the global debate on anticorruption and good governance (Swamy et al. 2001; Dollar, Fisman, and Gatti 2001). Swamy et al. (2001) also suggested women in general have less tolerance for and involvement in bribery. Do these studies show that women are intrinsically more honest and thus can be a tool to combat corruption? The literature argues no. Rather, the political scientists who have engaged in the debate suggest that it is opportunities for corruption that are gendered, not people's reactions to it. In particular, in socially conservative societies, it is difficult for women to become either clients or patrons in the male-dominated patronage networks through which corrupt exchanges occur. Where corrupt acts are condoned by social networks or even required by social convention, women have been shown to be no less willing than men to engage in such behavior, especially if required to do so to create a sustainable livelihood (Goetz 2007; Alhassan-Alolo 2007; Goetz and Jenkins 2005; Budlender 2000). Tripp suggests that where such findings are true, they may be because women as a group have "generally been excluded as benefactors in state-linked clientelistic relations" (2001b, 35), so they have less to lose in challenging them. These scholars argue that the central question is not whether women or men are less corrupt as a group, but how to combat gender-specific accountability failures. 18 Horowitz.qxd 6/11/09 1:22 PM Page 19 Getting Good Government for Women are in the best possible position to interpret changes in their clients' needs and devise appropriate responses to them" (2001, 21). Overall, the comparative literature on representative bureaucracy echoes that on descriptive representation. All over the world, some individual women politicians and some individual women bureaucrats may have innate empathy for the women they serve. However, whether or not they act systematically to improve service provision for women in their constituency or clientele depends on broader relationships of accountability and institutional norms. This work thus challenges the overarching assumption that the simple numerical presence of women will make representative democracies more responsive to women's needs. Rather, it would support Childs and Krook's (2006) proposal to "reformulate the research question from when women make a difference to how the substantive representation of women occurs." This would explore the various actors (both male and female), strategies, and outcomes consistent with the substantive political representation of women across space and time (524).30 It would also suggest an analysis of the broader institutional factors that undermine or bolster professional and cultural biases against women. Although the next section takes up the first challenge by looking at issues of political voice and effectiveness, Chapter 3 reviews the available literature on institutional responses to gender in policy implementation. How does the representation of women's issues in politics occur? As Childs and Krook's (2006) call to arms suggests, a small cluster of political scientists have been pushing the women and political discourse beyond elections toward issues of political effectiveness and accountability. The evident leader of this dialogue is Anne Marie Goetz. In her recent review written for UNIFEM, she depicted a cycle of political accountability (Figure 2.1) Figure 2.1. The cycle of political accountability Representation Legislation and policy Mobilization Implementation Trans- forming politics 19 Horowitz.qxd 6/11/09 1:22 PM Page 20 Agriculture and Rural Development such that "stronger political participation leads to better representation and accountability, and gradually to a transformation and deepening of democratic politics" (UNIFEM 2009, 18). This section looks at the extant literature on these components of political accountability to women in the developing world. Civil society mobilization. Goetz, among others, argues that "building political accountability to women, `like any accountability project' begins with women and their allies identifying and mobilizing around an issue of common concern" (UNIFEM 2009, 19). Feminist political scientists have contributed to a growing literature on women's movements in comparative perspective, and a good deal is now known about how women in different places come to politicize their identity and to choose organizational and advocacy strategies, especially in Latin American and Eastern Europe (for a thorough overview, see Beckwith 2005, 2000; also Nelson and Chowdhury 1999; Ray and Korteweg 1999; Randall and Waylen 1998; Jaquette and Wolchik 1998).31 Central to this literature has been attention to the political opportunity structures that enable or hinder women's mobilization, especially at the national level. In the developing world, the political opportunities studied have generally been periods of regime change or state transformation--particularly processes of decolonization and democratization as already discussed. Yet there is debate about whether women's revolutionary activism necessarily becomes state participation (Nelson and Chowdhury 1999), and less is known about how and to what effect women enact their political demands on the increasing number of relatively stable states in the global South, where consistency would allow such social movements to develop a strategic repertoire over time.32 Representation. Goetz reminds us that "`voice' does not easily and simply lead to better outcomes for women, because public institutions can have strong gender biases which undermine the impact of women's voice and presence in public" (2003, 35). The second step in the cycle of political accountability thus relies on the institutions that link voice to power, especially elections that give parties and politicians the mandates of the general citizenry and/or particular constituencies. How well women as a constituency leverage influence over the two main mediators of political representation--political parties and elected representatives--is at issue in this step. For women to gain leverage over the policy machinery within political parties (not just the candidate selection mechanism where individual favoritism could override), they must be seen to be instrumental as a group to the party's political success. Wielding a women's vote effectively, such that women can prove they will vote en bloc in response to a party's policy platform, is the clearest tool for leverage here.33 Building women's basic political participation, especially voting, is thus a first step in improving political responsiveness.34 Unfortunately, lack of data may be preventing both academic and political analyses of such a voting bloc. The International IDEA project, which collects the most extensive database of voter turnout statistics worldwide, is able to present gender-disaggregated statistics for only nine 20 Horowitz.qxd 6/11/09 1:22 PM Page 21 Getting Good Government for Women countries.35 At the same time, women need to be educated about the power of their electoral leverage. Compaore (2005) reports that in Burkina Faso too many women remain unconscious of their electoral potential and thus are often still labeled "electoral cattle." Another mechanism that has been discussed in the literature is the so-called women's manifesto--a basic list of policy priorities that parties are pressured to sign on to and include in their campaign platforms. A Botswana women's organization drew up the first women's manifesto in Africa, which prompted political parties to change their primary election procedures over a decade (Selolwane 2008). Such processes can have other effects too. The partnerships forged during the writing of a women's manifesto for Ghana provided the foundation for an ongoing independent women's movement, which had not existed in Ghana in recent years. However, parties must have the incentives to mind women's concerns, a point that goes back to the primary issue of electoral leverage. Representation also requires that individual elected representatives are receptive to women's political demands and willing to act on them. In Africa, where big men politics are common, the perception that big women in public office neglect their grassroots sisters suggests that this link is not automatic (Orock 2007). Goetz and Hassim (2003) use the distinctive cases of South Africa and Uganda to demonstrate that the institutions that govern both political life and the competition over public ideas affect the perceived legitimacy of feminist politicians (female or male) in politics as well as the connections between women representatives and those in civil society. For example, although the closed-list proportional representation electoral system has been recognized for its ability to put women in office, the intraparty politics required for women to get themselves on the ballot and/or to get their campaigns financed often entrenches upward accountability, with women more accountable to the party elites than to voters. But in ward-based first- past-the-post systems, publicly advancing gender equity sentiments can be political suicide for women who were tenuous candidates in the first place. Affirmative action mechanisms may help put women in office, but they are rarely enacted so that women assuming reserved seats are to explicitly represent women constituents. The authors note several measures that women's movements have used to build accountability relationships with women entering politics, such as the creation of party-independent bodies that provide financial and moral support to feminist candidates, as with Emily's List in the United States, or more informal encouragement of women's movement activists to enter politics, as in South Africa where it is hoped activists' "historical commitment to the women's movement will hold its own against the call of party loyalty once in office" (Goetz and Hassim 2003, 53). Legislation and policy. Civil society voice and accountable politicians do not alone effect policy change: legislative efficacy also matters. Yet once women are in office, they must face the difficult internal gender dynamics of parliamentary institutions, which undermines their effectiveness vis-ŕ-vis 21 Horowitz.qxd 6/11/09 1:22 PM Page 22 Agriculture and Rural Development male representatives (Bauer and Britton 2006; Tamale 1999). Rai and Ayaz (2007) suggest that even prevalent modes of dialogue disadvantage women inside parliaments when they are adversarial/aggressive rather than deliberative. Karam and Lovenduski (2005) outline a "rules strategy" to address what they describe as the "institutional masculinity" of parliament: learn the rules, use the rules, and change the rules. Such a strategy implies a number of mechanisms that could help women and/or feminist legislators become more effective inside parliament: orientation programs and trainings on parliamentary processes for new legislators, mentoring programs between new and experienced female MPs, media training, and encouragement for women to take up parliamentary leadership positions. Other mechanisms commonly discussed in the literature are cross-party women's caucuses, parliamentary committees on women's issues or gender equality, and institutional linkages with women's machinery in the bureaucracy. As Tinker writes (2009, 13), "building coalitions for specific legislation is essential if the political elite are to consider new policy," as is casting gender equality issues as benefiting entire communities rather than only women. Such mechanisms can build cross-cutting support. More broadly, paying attention to the types of legislative issues on which women work is important for the political efficacy of a gender equality agenda. Although female legislators around the world appear to be concerned with social policy, such as the well-being of women and children, they also place a high priority on issues traditionally thought to be in men's domain, such as economic, fiscal, and foreign policy. This may arise from personal interests; from a commitment to representing all of their constituents and an awareness that they were elected on a party ticket, rather than a gender ticket; and/or from strategic considerations, to avoid being marginalized as single- issue politicians (Geisler 1995). Also, if gender bias is entrenched in the basic economic institutions of society, it is in realms such as budgeting for the implementation of policies or the regulation of institutions that govern the economy where a real gender equality agenda must play out. Having women or feminist legislators in these debates is essential to gender equality outcomes. Implementation. As Goetz writes, "The test of political accountability to women is whether laws and policies are put into practice and make a difference to women's lives. Strong political participation, powerful representation and even ground-breaking laws and policies will change little for women unless policies are actually enforced" (UNIFEM 2009, 28). The challenge of "supplying" accountability to women through the public administration is considered in the next chapter. 22 Horowitz.qxd 6/11/09 1:22 PM Page 23 Getting Good Government for Women 3. Translating Mandates to Outcomes: Public Administration Responsiveness to Women The literature in the last section discusses determinants of gender-sensitive governance related to political voice. Whether women enter the public realm to act as policy makers themselves or are linked to elected representatives through civil society movements, the key focus of this scholarship is how women articulate their preferences in the public realm in an effort to see these enacted as public policy. Yet as scholars of administration point out, the implementation of public policy through such mechanisms as administrative procedures, budget allocations, incentive systems, and reporting, monitoring, and oversight systems determines how citizens at the grassroots actually experience policy directives or intentions developed at a central level (e.g., Wilson 1989; Lipsky 1980). It is thus important to consider issues of capacity within the public administration to implement any given policy agenda--the so-called supply-side factors. Services may also be reorganized so that citizens can articulate preferences directly to service providers, through a so-called short route of accountability. As Goetz writes, the public administration in development not only "reflects gender relations in society, but . . . as a part of social relations, is itself a gendering process. It has played an active, discretionary role in the construction and maintenance of men and women's unequal access to social and economic resources" (1992, 15). At the same time, public service agencies "are an important site for making a public stance against this process, and for transforming it in women's interests" (6). As both a cause and cure of gender inequalities, such arguments suggest that public sector management must be an essential consideration in the alleviation of poverty for women. Emphasizing supply-side factors in getting good government for women also removes from women the burden of agency for rectifying institutional biases against them. Rather, it reinforces the expectation that public institutions serve all of their publics equally. Yet, although getting institutions right has been a key preoccupation of major development institutions since the 1990s, gender inequality and mechanisms to rectify it are rarely mentioned in these policy debates. The scholarly literature on gender and administration is also fairly thin and focuses on the major mechanism for evoking gender-sensitive administration: special institutionalized gender units. How to integrate gender sensitivity into the norms, incentives, and practices of mainstream agencies in the public administration of developing countries is a topic little studied. 3.1. Women's Machinery Since the First World Conference on Women in the mid-1970s, the international women's movement and donor community have pressed 23 Horowitz.qxd 6/11/09 1:22 PM Page 24 Agriculture and Rural Development countries to establish state institutions specifically tasked with the promotion of the status of women. As of 2004, 165 countries had established some type of national machinery for promoting gender equality (Jahan 2005), including self-standing ministries, gender units within existing ministries, or central advisory bodies within the executive offices.36 Often this so-called national women's machinery become states' primary administrative response to demands for gender-sensitive governance. Available evidence suggests that the task of promoting a cross-cutting issue like gender in the average bureaucracy is a supremely difficult task for even the best situated agency.37 National women's machineries in developing countries have struggled to do this while facing a range of political, institutional, and financial constraints. Many of these stem from a lack of indigenous political commitment. Because many national women's machineries grew out of UN mandates, they have often been criticized as externally imposed. Where machineries were established by undemocratic governments, especially the first ones in the 1970s, they were viewed as nothing more than political support units for ruling party interests and enjoyed little civil society support (Goetz 1995).38 Geisler (1995) provides an interesting case study of the Ministry of Women Affairs in Zimbabwe, which was initially ambitious but gradually merged with party structures and alienated women activists in the nongovernmental organization (NGO) community. National machineries often face frequent restructuring and location shifts because, given the cross-departmental nature of gender issues, there is no one natural location for gender-focused bodies. The bureaucratic location of Ghana's National Council on Women and Development shifted six times between 1975 and 2001 (Tsikata and Kerr 2000). In many countries, machinery leadership also changes frequently. Appointees are often not familiar with gender issues because career civil servants, rather than committed gender activists, are assigned a turn in the bureaus. As well, many national women's machineries are assigned an agenda far beyond the resources given them to implement it. Goetz (1995) suggests that gender units are often expected to fill multiple formal roles: advocacy or advisory functions to integrate gender into policy or planning, policy oversight or monitoring, and implementation of programs for demonstration effects or to fill gaps. Where machineries do not enjoy constitutional powers to enforce penalties, it is almost impossible for them to effectively monitor other agencies or to hold them accountable to national or international gender commitments. One strategy commonly used to create linkages between women's machineries and other ministries and departments is the designation of gender focal points within each part of the government bureaucracy. In practice, the focal point strategy has had limited effectiveness because the often junior female staff appointed are given few extra resources or time for new responsibilities, and little training, support, and clarity about their role. Jamaica and Uganda have enjoyed some success by providing ministries and line departments with detailed gendered project guidelines, checklists, and monitoring indicators. As well, in Jamaica, methods for assessing gender-differentiated impacts have 24 Horowitz.qxd 6/11/09 1:22 PM Page 25 Getting Good Government for Women been incorporated into the training provided to civil servants (Goetz 1995, 26). Interministerial coordinating committees are another mechanism used to overcome bureaucratic resistance (Bell 2002).39 Observers also worry that gender machineries depoliticize the issue of gender inequality. Banaszak, Beckwith, and Rucht (2005) argue that "the relocation of responsibility to non-elected state bodies eventually reduces social movement influence (178). This situation is likely where national machineries replay larger government-civil society tensions or are seen as competitors of NGOs for limited donor funding or public attention. But where national machineries were set up following democratic transitions (e.g., Philippines, Chile, South Africa, Uganda), they have been relatively more successful and effective than in other places, a success that can be traced not only to a general commitment to social justice and equality but also to broad and open processes of consultation (Baden 1996). In some cases, national machineries have actually empowered national women's movements and provided them with new spaces or resources for activism (Franceschet 2003). Nevertheless, a review of six national machineries conducted in the mid-1990s implies that the relationship between these and politically active women in civil society has tended to be somewhat uneasy (Goetz 1995). Overall, the evidence suggests that gender machineries have the potential to strengthen the capacity of bureaucracies to deliver gender-sensitive administration in terms of technical skills and managerial systems for gender mainstreaming, but they frequently lack the political buy-in to be able to demand real accountability to women from the greater bureaucracy. Too often in developing countries, gender units are set up, but the structures and cultures of the broader development bureaucracies remain more or less untouched. The next section probes this topic further. 3.2. Gender in Administrative Policy and Planning If getting good government for women relies on accountability relationships, all government agencies must be accountable for equally serving women and men, not only those institutions explicitly tasked with gender mainstreaming. Gender-sensitive administration requires both the capacity to serve--which involves adequate resources, data, and skills--and the incentive to serve, which involves reward and sanction systems with the monitoring and evaluation data to enable these. With the renewed emphasis on public sector capacity building, development theorists have built an important body of research on how to elicit good performance, commitment, and responsiveness of low- and middle-level staff in poor countries through attention to incentive structures, organizational cultures, training, and management (e.g., Tendler 1998; Grindle 1997; Tendler and Freedheim 1994). This type of analysis could be essential to getting good government for women since there is very little understanding of how to craft bureaucratic incentive systems that reward the extra effort needed to work for women's advancement. Instead, as Goetz writes (2003, 69), "gender-blind and gender-biased disciplinary and reporting systems within the public administration are the greatest culprits for the many exclusions, humiliations, 25 Horowitz.qxd 6/11/09 1:22 PM Page 26 Agriculture and Rural Development and poor quality services endured by women." Performance reviews and indicators, bureaucratic communication flows, and informal professional cultures all implicitly undervalue the work necessary to tailor services specifically to the needs of women and to help women overcome barriers to accessing these. The major scholars in this field provide little gendered analysis of their topics. In Tendler's 1998 study of a set of innovative state government programs in Brazil, although clients and/or service providers in the cases are often predominantly women, she provides no explicit attention to gender issues. In Grindle's 1997 edited volume on capacity building in the public sectors of developing countries, the only discussion of gender issues is a section on building gender research capacity in the NGO sector--not even in the so- titled public administration. Others have highlighted additional promising state-centered initiatives targeted at increasing client focus, including consulting on service delivery and policy priorities, setting standards, creating new rights for service "clients," and having NGOs provide services as a way of setting a precedent for more appropriate and effective services (Goetz and Gaventa 2001). These could be tailored for gender-sensitive service delivery. Goetz (2001) suggests two main reasons for this gap in the literature. The first is the scarcity of women in public administration, especially in the senior levels of sectoral departments, which is a topic already discussed. This condition is compounded by the lack of gender sensitivity in public management education of both men and women in some regions. The second reason is that "gender biases have prevented some analysts of policy implementation processes from noting the gender-differential impact of the work of development administrators" (296). In particular, the well noted dearth of gender-disaggregated data at every point in the chain of service delivery--demand, expenditure, benefit incidence, impact, among others-- severely hinders gender-sensitive policy analysis, monitoring, and evaluation. Gender budgeting initiatives, which are discussed further in Box 3.1, may build the capacity of officials to generate gender-disaggregated expenditure data, or at least highlight the scarcity of gender-disaggregated data throughout the administrative process. Mechanisms for consulting with women would also improve the quality of information about women's needs and interests with which to inform public administration priority setting and decision making. 3.3. Anticorruption Efforts With good governance often equated with lack of corruption in the mainstream literature on the topic, no discussion of gender and governance would be complete without considering corruption. The topic has been motivated by the pair of cross-country World Bank studies discussed in Box 2.1 (Swamy et al. 2001; Dollar, Fisman, and Gatti 2001). However, gender researchers have transformed the question from whether women can help combat corruption to how governments can lessen the disproportionate effects of corruption on women. 26 Horowitz.qxd 6/11/09 1:22 PM Page 27 Getting Good Government for Women Box 3.1. Do Gender Budgets Work? Pioneered in Australia in the 1980s and in South Africa and the Philippines in the 1990s, gender-responsive budget initiatives assess the adequacy of policy and budgetary allocations for addressing the specific nature of gender inequity in a country. They have been used in over 60 countries to expose gender bias in macroeconomic policy. In Mexico, parliamentarians used a gender budget analysis to argue for a 40 percent increase in reproductive health spending. The Philippines mandated every national agency to allocate 5 percent of its budget for "gender and development" after a gender budget exercise. Often, however, the persuasive power of gender budgets does not result in real budgetary changes because those involved with such initiatives did not themselves have the power to change budgets. Budgeting is inherently a political endeavor and is ultimately as susceptible to the vagaries of gender in the political process as any other such tool. Gender-responsive budget work is most effective when it supports the direction that particular policy makers already want to take. In this case, the information provided can strengthen the policy makers' position. Such initiatives do build coalitions among members of civil society, the parliament, and the bureaucracy, and even among different bureaucratic actors of the budget planning process who may not have previously worked together. Almost all initiatives have some sort of training and capacity-building component. For citizens, this training can serve to demystify the budget, increasing their budget literacy and ability to engage with the government spending process. For parliamentarians, who frequently have little research capacity, such work can forge links with NGO researchers who are also gender advocates. They thus remain an important tool to improve the answerability of office holders for gender equity. (Budlender 2005; Budlender 2000; Valodia 1998; Elson 1998; Demery 1996) Goetz, for example, has frequently highlighted the failure of accountability institutions to respond to the rights and preferences of women (Goetz and Jenkins 2005; Goetz 2003). "For women, being at the margins of political life has translated into being, in many contexts, invisible to accountability institutions" (Goetz and Jenkins 2005, 158). This is especially harsh given that poor women are susceptible to gender-specific accountability failures on top of the general failure of bad governments to account to their citizens. These include gendered capture, such as when money destined for women's development is more easily stolen by state actors because "women tend to be less aware of their rights, and less willing than men to demand that public authorities account for missing funds" (162) and gender bias, which describes "the leakage of patriarchal norms into purportedly impartial law and policy" (163). Women may also be more susceptible to "sexual currencies" of corruption, such as demands for sexual services in lieu of money bribes. She thus argues that "the main focus of attention in efforts to increase women's political effectiveness should be on building gender sensitivity into accountability systems--improving their capacity to answer to women and to enforce sanction against public sector actors who have abused women's rights" (Goetz and Hassim 2003, 6). 27 Horowitz.qxd 6/11/09 1:22 PM Page 28 Agriculture and Rural Development Unfortunately, while there is a growing literature on ways to improve public sector accountability to citizens generally, and to poor and disadvantaged groups specifically (e.g., Mehrotra 2006; Devas and Grant 2003; Goetz and Gaventa 2001), ideas for rectifying gender-specific capture problems are limited. 3.4. Public Sector Reform and Privatization In addition, to changing the "software" of a bureaucracy--the procedures and incentive systems internal to the state--governments can change the structure of the state itself to achieve various aims. Decentralization (discussed in Chapter 4) and privatization are two reforms frequently justified as strategies to improve the quality of service delivery. In privatization, competition is introduced to a particular market for services, so that users can sanction a provider who does not provide adequate services by choosing another provider. This has been called a "choice-based" approach to accountability (UNIFEM 2009). User fees may also be implemented to introduce market incentives for providers and consumers of services. The literature suggests that market-oriented public sector reform can affect women's ability to hold service providers to account for gender-sensitive services; it also emphasizes the relevance of examining the gender sensitivity of private sector firms, who appear as service providers in developing countries in increasing numbers. The negative economic implications on women of user fees for services, especially health care, education, and water/infrastructure, have been widely noted (Lincove 2006; Coles and Wallace 2005; Nanda 2002; Beall 1997). When women or families of girl children face out-of-pocket costs for services, the costs may either become out of reach or opportunity costs may outweigh benefits. Even though many poor women may be exempt from fees in schemes that attempt to integrate efficiency and equity concerns, providers have little incentive to apply exemptions when they could instead recover costs (Nanda 2002). Where privatization leads to a failure of service delivery because neighborhoods or households can no longer pay for formerly public services such as sanitation or tap water, women often shoulder the increased labor burden of alternate provision (Hainard and Verschuur 2001). Because women often have fewer resources than men, they have a hard time exercising gender- specific purchasing power, such as rewarding providers who better tailor services or make them more accessible. Deference to male family members may also limit their capacity to exercise full choice in purchasing services for themselves (UNIFEM 2009). The "supermarket state" model of service provision can also affect women's ability to effectively use political voice to change the behavior of either service providers or those who have authorized them. As Goetz writes, "the socially fragmenting effect of markets in services means that the influence of consumers on services can be limited, particularly if they act individually" (2003, 63). Moreover, women in particular may not act effectively as a consumer lobby because their individual and collective voice "in relation to service providers may be weaker than that of consumers with better resources and greater capacity to `exit' private providers." Empirical studies to test this 28 Horowitz.qxd 6/11/09 1:22 PM Page 29 Getting Good Government for Women hypothesis, however, are lacking. Overall, "privatization may increase the number of choices, but it does not change the conditions of inequality and dependency that constrain women's access to services in the first place" (UNIFEM 2009). The internal capacity and incentives of private sector organizations to serve women also become relevant in this context, particularly because they are removed from most political accountability mechanisms. The evidence is not encouraging. In an environment where efficiency and cost recovery are privileged over equity issues, gender issues can be marginalized. Using market mechanisms to weed out poor service providers means that few have incentives to introduce new performance standards based on gender equity or new services responsive to women's needs that may weaken the balance sheet. 29 Horowitz.qxd 6/11/09 1:22 PM Page 30 Agriculture and Rural Development 4. Decentralization and Local Politics: A New Space for Accountability? As decentralization swept the development policy world in the late twentieth century, proponents frequently presented it as an important vehicle for increasing the participation of traditionally excluded or disadvantaged social groups in newly empowered local governments (Crook and Manor 1998). If a country's decentralization project implies a large measure of devolution, or "the handing over of greater power and resources to local politicians," opportunities for local voice, political compacts, and local client/citizen power may be strengthened. However, the principal argument for political decentralization-- that when decisions are made closer to the community, they will better reflect the needs and priorities of those most affected--assumes that (a) local politicians and administrators have opportunities to know the preferences of their constituents and (b) that they choose to act on them. Devolution will assist governments in providing services that better fit the particular needs of women constituents only if women have opportunities to participate in local politics and if local governments have instructions and incentives to serve them. Local government has long been seen as an especially hospitable sphere for women's participation because it is more physically accessible, may have less stringent eligibility criteria, and often deals with issues and services that directly impact women's private lives, such as health, education, or water (Beall 2005a). In some regions too, women played significant roles in local politics and administration in precolonial times and seek to reclaim this role today (Ohene-Konadu 2007). But just as a key question in the mainstream decentralization literature is the relative capture of local versus central governments (Bardhan, Mookerjee, and Torrado 2005), the extent and form of gender discrimination in local government compared to national government requires empirical research. The conclusion that emerges from the literature is that although local governments can be sites of innovation--they may be more free to experiment with service delivery methods that suit particular constituencies, if they have the resources to do so--women face significant hurdles to participate in local politics in poor communities. If local governments lack the incentives or awareness to make special efforts to include women, the assumption that decentralization can increase the voice accountability mechanism for women is likely to be a false one. 4.1. Women and Local Politics The majority of the extant empirical work on women's participation in the context of political decentralization focuses on the massive Indian experiment with women in local government.40 Close to a million women came into power in the early 1990s as a result of the 73rd and 74th amendments to the 30 Horowitz.qxd 6/11/09 1:22 PM Page 31 Getting Good Government for Women Indian constitution, which created an active panchayat system of village, block, and district councils with regular elections.41 The amendments also reserved one-third of seats in panchayat councils as well as one-third of council presidencies for women.42 Some states originally selected the chairs to be reserved for women randomly, and economists have used this identification strategy to explore causal relationships between women's representation and key governance outcomes (Beaman et al. 2008; Besley, Rao, and Pande 2005; Chattopadhyay and Duflo 2004; Duflo and Topalova 2004). A small literature also touches on the gendered aspects of decentralization in a handful of African countries, which has thus far included Uganda (Ahikire 2003; Johnson 2003; Saito 2002; Dauda 2001), South Africa (Beall 2005b; Beall 2005a; Mbatha 2003; Rangan and Gilmartin 2002), Zimbabwe (Dauda 2001), Ghana (Ohene-Konadu 2007; Ofei-Aboagye 2004), Senegal (Patterson 2002), and Nigeria (Okome 2000).43 The African literature is less systematic than that on India, and, lacking the ability to econometrically link women's representation to outcomes, most of the papers focus on either women's numerical participation or qualitative aspects of their political effectiveness.44 Although themes similar to those in the more general women and politics literature just reviewed frequently reappear in the work on women in local government, several major hurdles unique to, or more important at, the local level emerge from a review of this literature. Socioeconomic disadvantage and illiteracy hinder women's participation Gender inequities in socioeconomic status are particularly severe at the local level, and these impede women's participation in decentralized institutions. Conversations with women in urban and rural communities around the world reveal that the multiple manifestations of their social immobility severely limit their formal political participation (Vissandjee et al. 2006; Fonchingong and Ngwa 2006; McEwan 2003; Greenberg and Okani 2001). It is rarely easy for women to transfer their domestic responsibilities to their husbands, although men can leave their tasks to their wives (Meinzen-Dick and Zwarteveen 1998). Class and other nongender aspects of identity also affect women's ability to participate and the issues that motivate them. Poor women face a high opportunity cost in participation, especially if it displaces income-earning opportunities, which leads to the so-called middle-class effect documented in some places (Weinberger and Jutting 2001).45 Women with low literacy find it impossible to participate as legislators in decision-making processes that are heavily dependent on written work and the legal language of agendas, minutes, and schedules (Datta 1998). In the Indian context, "without exception, every single piece of survey research on this question cites the recognition of women representatives that they would have been better able to contribute to the proceedings and activities of the panchayats had they had the advantage of schooling" (Jayal 2006, 24). Low levels of literacy and/or language barriers, time constraints, and lack of confidence were identified as major constraints to women's participation in Ghana (Opare 2005; Ofei-Aboagye 2004) and Uganda as well (Johnson 2003). 31 Horowitz.qxd 6/11/09 1:22 PM Page 32 Agriculture and Rural Development Women in local government also require more specific training in such topics as procedure, legal literacy, and leadership (Datta 1998). Power dynamics among women are apparent here because both caste and class affect women's access to social and economic opportunities (Vijayalakshmi 2002).46 Training or support programs run by NGOs, governments, and donors have been highlighted in the literature as a response to the education gaps women community politicians face. NGO-created and -funded women's sanghas in several Indian states serve as important training grounds for women to develop and define their leadership skills. Sanghas in Karnataka have also sponsored candidates, trained them after election, and helped to create a climate of accountability and responsiveness. They more strongly link women representatives to a politically empowered female constituency (Kudva 2003; Purushothama, Kumar, and Purohit 2000). NGOs in Ghana and Uganda have also provided training for district assembly staff and others (Ofei-Aboagye 2004; Rakodi 2002).47 Many state governments provide training on local government rules and procedures, government structure, and common policy issues; some donors have also been involved with technical assistance to local governments, such as an Asian Development Bank project that trained women representatives in Bangladesh, Nepal, and Pakistan (Sultana and McCarthy 2004). Some observers of this training landscape have made a distinction between government programs that are often organized around "information dissemination" and the more "transformative" services provided by NGOs focused on skill and confidence building, role clarification, and gender awareness (Kudva 2003). Unfortunately, in India, where constituencies reserved for women rotate to new districts every five years, political parties and others may not invest in women's capacity because women who gain power through reservations most often do not remain in office after reservations rotate away (Chattopadhyay and Duflo 2004; Chathukulam and John 2000).48 This is unfortunate because women local council presidents become substantially more effective when they acquire more political experience (Ban and Rao 2006). Some have advocated for a longer reservation period (Chathukulam and John 2000). Tradition and patriarchy may be particularly strong in local politics The Indian and African literatures document women's friction with prevalent gender norms both at home and once they have entered political institutions. At home, women's domestic responsibilities mean that the opportunity cost of the time they spend involved in politics is great. Such time constraints substantively prevent many women from meaningful political engagement. In Uganda, for example, village men refuse to let their wives go out of the house after dark for local council meetings, for fear that they would neglect household chores or "misbehave" (Johnson 2003). A famous study of the Indian local government system recorded a husband's confusion when asked to allow his wife to participate in the local council. "And who will make the chapattis?" he asked (Datta 1998).49 These pressures affect women's ability or willingness to participate in community affairs not only as politicians but also as engaged citizens. Besley, 32 Horowitz.qxd 6/11/09 1:22 PM Page 33 Getting Good Government for Women Rao, and Pande (2005) used household and village survey data from southern India to describe participants in the gram sabha, or village assembly, which is responsible for approving the annual accounts of the village council. Although they conclude that it is more disadvantaged groups in general who attend the village meetings and thus improve the targeting of resources toward the neediest groups, they find that women specifically are underrepresented. Whereas a concept of structured participation at the local level has been central to South Africa's democratic transformation, black women often find this process inaccessible (McEwan 2003). Few Zimbabwean women small holders in a study by Burgess and Beilstein (1996) took action to secure the services they needed from the government. In Indonesia, Beard (2005) found that most of the opportunities for women to participate in civil society are in their roles as mothers, wives, and household managers, but not as empowered citizens. Community meetings are still not a forum for women in their current form, at least not in the places in India, Indonesia, or southern Africa that have been studied. The women who do break through the barriers into public life often face male ridicule and harassment for transcending conventional gender roles (Nandal 2003; Chathukulam and John 2000; Datta 1998). Jayal (2006) reported that in many Indian states there have been attempts to slander women who stand for election. Physical intimidation or even violence against female members is not uncommon. In Uganda, women campaigning for local councils are often required to perform "socially submissive acts" such as kneeling for the electorate at the campaign podium to conform with traditional attitudes at the local level (Ahikire 2003, 224). Men also use procedural subversion to silence women members in more subtle ways. Male representatives do not tell women members about meetings, committees, or training camps, or they schedule these when women's domestic responsibilities inhibit participation (Datta 1998; Pal 1998). Few states have a mandatory quorum of women or lower caste representatives at meetings or village assemblies, and chairpersons may conduct meetings or votes with only upper caste male members present (Jayal 2006).50 Male officials of line departments and parastatal agencies in Kerala also did not give female members due recognition (Chathukulam and John 2000). Forming associations of elected women local representatives (aided by national women's machinery) has been one solution advocated in the literature (Pal 1998). Political parties also often exert significant discipline at the local level. The female panchayat members interviewed by Chathukulam and John (2000) felt that the local party machinery exercised considerable control over them and that they were not able to function independently or without partisan bias to implement a development agenda. Dauda (2001) reports a similar resistance to integrating women within Zimbabwe's ZANU-PF party, which instead used its Women's League to intimidate families of opposition politicians at the local level. In some countries, structural tensions also exist between the recognition of traditional authorities and the empowerment of women in rural areas. Even though there are important cultural and historical reasons for recognizing 33 Horowitz.qxd 6/11/09 1:22 PM Page 34 Agriculture and Rural Development traditional governance structures, doing so can have far-reaching implications for women, especially in terms of access to land as well as to membership in leadership and decision-making bodies. This dynamic has been noted by several scholars in South Africa, where despite a celebrated commitment to gender equity at the national level, political decentralization was implemented in such a way that allowed patriarchal informal institutions to become entrenched in the formal organization of local government and law. This impeded women's rights and privileges at this scale (Beall 2005a; Beall 2005b; Mbatha 2003; Rangan and Gilmartin 2002). Beall extrapolates from this case to the broader evidence on political decentralization: "The extent to which local government and decentralization are embedded in asymmetrical social relations and informal institutional practice helps explain why local governance rarely proves to be a magic bullet for [women]" (2005b, 271). Nonlocal structures can undermine the promise of decentralization Several scholars caution against viewing "the local" in isolation from broader economic and political structures. The way that national and global trends, such as market liberalization and pressures to reduce the role of the state, intersect with local politics can serve to severely limit local policy makers' ability to enact or implement equity-enhancing measures. Beall (2005a) documents this in South Africa, where an increasingly market-oriented central government required local government to raise 90 percent of its own revenue. As she writes, "limited resources and expanded responsibilities of local government for service delivery have made cost recovery almost inevitable" (12). When local governments face pressures from prevailing political or economic doctrines to generate a large proportion of their revenue, they may have less ability to experiment with innovative, context-specific solutions to meet the heterogeneous needs of local populations. Similarly, the shape, structure, and politics of the decentralization program in individual countries will affect both male and female policy makers' ability to wield the developmental power of the state in women's interest. In some countries, decentralized structures of government may be created but given very little resources or power to actually enact an agenda defined by local citizens. This appears to be the case in Ghana (Crook and Manor 1998). Local district assemblies cannot motivate other departments to act or allocate resources, and elected representatives do not have sufficient control over spending priorities. It appears that the broader political context in which decentralization was implemented led to government failure for all citizens, rather than just for women. This situation confirms Beall's assertion that "local politics and decentralization policies are most likely to work in the context of a strong central state and robust intergovernmental coordination" (2005b, 267). Activism, while important, can be fragmented at the local level Like at the national level, it is important for women's activism to increase the presence and influence of women in local assemblies. In South Africa, women's organizations came late to the realization that local politics was an important sphere for women's empowerment, but once they did, their 34 Horowitz.qxd 6/11/09 1:22 PM Page 35 Getting Good Government for Women campaigns helped increase the numbers of women in local government (Mbatha 2003). Ghanaian women are at a much more drastic minority in their district assemblies than South Africans, but campaigns by the Ministry of Women's and Children's Affairs, as well as by development organizations and NGOs, in the 2002 local government elections nearly doubled the number of women elected as assembly members. The figure rose from 2.9 percent to nearly 5 percent (Ofei-Aboagye 2004). Goetz and Hassim suggest, however, that "the women's movement's capacity to support women in local politics and help develop gender equity policy can be fragmented by decentralization" (2003, 21). Access to quality services has also been a rallying point for women's collective action at the local level. Tinker (2006) among others has brought attention to the growing number of local protest actions organized by women around issues that directly affect their livelihoods, which subsequently have led to national political action. In South Africa, the lobbying efforts of a self- employed women workers' union encouraged Durban local governments to become more responsive to women working in the informal sector (Skinner and Valodia 2003). Indian women's efforts to collectively audit whether spending commitments for drought-relief public work schemes and the public distribution system were actually being met at the point of service delivery uncovered corruption and mismanagement (Goetz and Jenkins 2005). Once women get involved in collective action toward improving service delivery, they often continue to be more engaged with civil society and its relation to the state. Yet the challenges of adding political mobilization to the magnitude of women's production and reproduction tasks at the local level are clear. In Argentina, women seriously challenged the status quo when they united to lobby local officials about a lack of basic public services. Yet once their issue was satisfied, they returned to their daily affairs rather than continuing to build on the political space they had created because the demands on their time were too great to sustain over a long term (Hainard and Verschuur 2001). Facing barriers, women make the nonpolitical political However, even in the face of overwhelming constraints, women can be seen carving out spaces of participation using the networks, tools, and language available to them. Women around the world participate in a vibrant associational life at the community level. Women's participation in institutions such as women's associations, self-help or microlending groups, social networks, and civic and religious groups constitute the community management responsibilities that Caroline Moser (1993) famously identified as women's "third role," after production and reproduction. By strengthening women's social networks, such organizations may provide individual women with new platforms to exercise agency and may motivate women to access existing forums for political voice or create new ones (e.g., Sanyal 2008). Other forms of participation include talking through men to whom they have blood or marital relationships, and unusual modes of participation such as hissing or ululating in forums where they are formally silenced (to show opposition or approval), evocation of the supernatural, and song (Hirschmann 1991). Hirschmann advocates that researchers acknowledge "the broad array of 35 Horowitz.qxd 6/11/09 1:22 PM Page 36 Agriculture and Rural Development levels at which, and means with which, women may operate politically" and spend time and creativity to expose these at greater depth. 4.2. User Committees and Community-Driven Development In theory, community groups and associations formed to deliver participatory development projects and services such as user groups or stakeholder committees create new channels--outside local politics--for people at the grassroots to engage with government and other service delivery actors and to gain greater influence over decisions that affect them. To some analysts, such committees are key actors in the short route of accountability, in which citizens (as clients) communicate directly with frontline service providers rather than through elected representatives. In fact, donors often require the formation of user committees in their projects as part of an attempt to cushion development from politics. However, the limited evidence of such institutions in practice does not reveal a particularly emancipatory movement. User committees are usually single-purpose bodies with limited life spans related to a time-bound project or initiative. They are often assigned the fairly limited task of selecting the beneficiaries of already planned programs or projects; rarely do they have the power to influence program implementation and, more rarely still, policy design (Manor 2004).51 This does not serve women well, who need time and permanence to develop the skills and confidence to participate, as well as to benefit from the formalized rights and privileges granted by government institutions. As well, selection criteria for user committees, which are often delineated by less than democratic means, can exclude women either by definition or by process (Manor 2004; Cornwall 2003; Agarwal 2001; Guijt and Kaul 1998; Meinzen-Dick and Zwarteveen 1998; Mayoux 1995).52 User groups that deal with what are traditionally considered women's issues may include many women but may find their work siloed and less influential (Manor 2004). On the other hand, affirmative action mechanisms for women (and other disadvantaged groups, for that matter) appear to be provided more often on user committees than in elected local councils because they often follow the dictates of international organizations. Ofei-Aboagye (2004) documents the case in Ghana of projects of the United Nations Development Programme and the United Nations Children's Fund that involved the formation of local water and sanitation committees and required that one-third of available leadership positions be allocated to women. The guidelines were largely followed because noncompliance led to exclusion from benefits. Women were not, however, given positions of responsibility. Women face substantial costs of time, reputation, and effort to participate in public life. As Meinzen-Dick explains, "whether women are willing to bear these costs and face these social risks will largely depend on their assessment of the effectiveness of the organizations, and of formal participation as a means of achieving personal objectives, as compared to other means available to them" (Meinzen-Dick and Zwarteveen 1998, 342). Some suggest that institutions created as part of participatory development initiatives overemphasize consensus and mask or silence dissent (Cornwall 2003). 36 Horowitz.qxd 6/11/09 1:22 PM Page 37 Getting Good Government for Women Women may not be able to make their voices heard in such settings. As well, Manor points out that the creation of user committees often "fragments" popular participation, making it less coherent and effective. People may be drawn into user committees that lack real power, and thus "their time and energy is sapped in vain attempts to make an impact, often in single sectors" (Manor 2004, 208). They become less willing and/or more cynical about engagement in more promising areas of governance.53 In this context, women may find participation in more informal domains such as the household or unofficial networks of patronage a more effective way to access services, but "access through informal means is not as secure . . . and is more prone to be influenced by unequal power relations" (Meinzen-Dick and Zwarteveen 1998, 343). 37 Horowitz.qxd 6/11/09 1:22 PM Page 38 Agriculture and Rural Development 5. Discussion This paper has reviewed the available literature on strategies and mechanisms available for strengthening governments' responsiveness to women, especially poor women. As the bulk of literature on this topic suggests, many gender-sensitive governance initiatives focus on the demand side of good governance by seeking to increase the influence of women's voices in public decision-making processes. Legal requirements or more informal incentives may attempt to increase women's presence in the parliaments, councils, bureaucracies, and political parties that are repositories of political power, often under the assumption that women in power will more effectively represent the needs of female constituents. Other demand-side strategies create pressure for downward accountability of both male and female representatives to their female constituents so that women's needs and preferences are taken into account as criteria for public policy and resource allocation. A separate, smaller literature also discusses mechanisms designed to build the capacity of the administration to implement or "supply" a gender equity agenda. An increased sensitivity to gender in the bureaucracy is often operationalized through the creation of dedicated agencies or other institutional structures for gender. Other strategies discussed include building the administration's capacity to generate gender-disaggregated information and incentivizing responsiveness among personnel to this data through service awards or performance reviews. Box 5.1 categorizes select mechanisms discussed in this paper according to the conceptual framework outlined in the introduction. However, despite the ostensible diversity of mechanisms available to strengthen governance for women, three themes consistently emerged as important mediators of success across numerous cases. First, initiatives generally fail when gender equity is not recognized to be a political project. As multiple authors have pointed out, mechanisms such as quotas and institutionalized gender units in the bureaucracy may offer governments pressured to become more gender responsive what appear to be relatively easy institutional responses to women's disempowerment. Yet without the accompanying political will necessary to transform underlying systems of accountability, such popular mechanisms can easily devolve into technocratic fixes that increase the numbers of women present in government structures without increasing their influence over them. Bureaucratic change sincerely in service of a transformative agenda--one that is willing to challenge existing power structures in society--requires a political context supportive of women's empowerment, a state that values responsiveness to its citizens, and a vocal and effective women's movement. Women cannot change governance alone, nor by simply being granted access to unchanged patriarchal 38 Horowitz.qxd 6/11/09 1:22 PM Page 39 Getting Good Government for Women Box 5.1. Select Mechanisms for Getting Good Government for Women Increasing Political Voice · Affirmative action in electoral politics and internal party politics (quotas, reservations) · Training and support programs for local and national representatives · Women's wings in political parties · Women's manifestos · Recruitment, mentoring, and leadership development in political parties · Women-friendly institutions (timing of meetings, type of pay, safety in travel, child support, etc.) · Equal opportunity structures in civil service (e.g., antidiscrimination bureaus, merit protection agencies, equal opportunity commissions) · Party-independent bodies that provide financial and moral support to female candidates and thus link leaders to gender constituency · Organizing women and girls, promoting political awareness, leadership, and advocacy abilities · Gender quorums in community meetings Securing Administrative Accountability · Ministries/agencies of gender in national and local governments · Gender focal points in sectoral ministries and decentralized departments · Organizational gender policies and structures, such as gender working groups · Advisory councils on women's issues · Affirmative action in the public administration · Performance contracts with attention to gender · Gender-sensitive complaint mechanisms · Female field workers with discretion · Gender-responsive budgets · Gender-sensitive design and implementation of programs and projects · Gender-disaggregated and gender-sensitive monitoring indicators Reinforcing Client Power · Women's self-help groups in governance · Affirmative action in user group membership · Gender-sensitive citizen monitoring and auditing institutions. In fact, such initiatives risk a backlash of cynicism against the state. As Goetz and Jenkins write (2005, 178), "voice without accountability is a formula for disenchantment with politics." Second, women's political empowerment cannot be divorced from their economic empowerment. Because women consistently cite issues of illiteracy, language barriers, and time poverty as major hurdles in their ability to participate in public affairs, then improving their access to education and reducing work burdens are key precursors to enabling more frequent and 39 Horowitz.qxd 6/11/09 1:22 PM Page 40 Agriculture and Rural Development effective public participation. Building women's livelihoods may also allow more women to overcome the spousal resistance frequently cited as a hindrance to political engagement, since advocates suggest that ensuring women's rights to economic resources (as well as protecting these by law) is critical to women's ability to exercise influence in household decision making and beyond. Conversely, social and economic policy greatly shapes the quality of livelihoods available to women in the developing world, so political engagement can be viewed as a channel for bettering economic opportunities. Institutional biases against women's public participation, however, can lead to a vicious cycle in which unfavorable policies perpetuate women's inability to earn a living, making them less likely to exercise voice in the public arena where such policies are formulated. If women's livelihoods and the allocation of labor and resources within households are key mediators in their political empowerment, this means that social and cultural transformation also lies at the root of governance transformation. Traditional gender roles are not immutable, yet they are embedded in culture. It is not heartening that even in the most recent data, Inglehart and Norris (2003) find persistent conservatism regarding gender roles in traditional agrarian societies--a category describing much of the developing world. Yet at the same time, governance initiatives themselves-- especially those that increase women's ability to attain leadership positions-- although declawed in the short run, may accelerate the pace of long-term cultural change by creating new models of political leadership. At the same time, these arguments must acknowledge the multiple social characteristics that define women (and men), which combine with gender to affect their access, opportunities, and constraints. Interventions targeted at improving governments' response to marginalized groups must carefully disentangle disadvantage caused by gender from that caused by poverty, rural distance, ethnicity, and other factors. Finally, political, social, and institutional contexts will affect the shape and success of any single mechanism for any specific group of women across multiple places. Positive frame conditions that were highlighted in the literature include high levels of party competition, especially the prevalence of left-leaning political parties, a strong central government with constitutional commitment to gender equality to accompany decentralization reforms, and the existence of gender as a mobilizing identity within the sociopolitical sphere. More broadly, the state of democracy, the macroeconomy, the extent of clientelism, nationalist projects and the citizenship claims that accompany them, and the broader governance environment will all influence governments' ability and willingness to act on behalf of marginalized groups. 5.1. Questions for Further Study It is also clear from this review of the literature that a good deal is left to learn about how to get good government for women. The remainder of this discussion revisits the key points of the major sections of this paper to gain insight on this point. 40 Horowitz.qxd 6/11/09 1:22 PM Page 41 Getting Good Government for Women 5.1.1. Women and Politics Does women's involvement in decision making improve the provision of gender-responsive services? If so, when? Thus far, the majority of scholars examining this topic have relied on aggregate, often cross-national data sets that link the numbers of women in political office to various institutional outputs. This work shows, for example, that having a greater proportion of women in a legislature is correlated with increased support for women's- issues bills and lower levels of corruption. However, both gender and policy change are nuanced concepts, and such macroanalyses provide limited and occasionally flawed insight into microlevel preferences. In most fields very little is known about how the behavior of particular individuals (women and men) holding particular notions of and commitments to gender equity influence the actions and priorities of institutions. Following the argument forwarded by Childs and Krook (2005), this finding suggests that scholars should move away from such cross-country quantitative work in favor of qualitative and microeconomic post hoc analyses examining the behavior of individuals over time, thus using the study of "micro-level behaviors to explain macro-level results" (15), rather than the other way around. Another underexplored tactic would be to compare the background characteristics of similarly elected male and female officials to better disaggregate how important are background characteristics vis-ŕ-vis gender to a candidate's election or decisions made while in office. There is also far more work to be done on how characteristics of specific legislative contexts, such as committees, parties, institutional norms, and the political climate, constrain or enable the substantive representation of women. Expanding the type of institutionalist analysis conducted by Goetz and Hassim (2003) to other countries and regions would be an important contribution to this topic and may enable the beginning of theory building around which design features of political institutions and cultures enhance the perceived legitimacy of politicians with feminist sympathies. Additional empirical research on the nature and quality of links between feminist movements and female political elites across countries is also needed. Few scholars have provided concrete insights into how women's movements build accountability relationships with women entering politics and/or target female elites on policy issues. In fact, even though the literature on women's involvement in democratic transitions is substantial, there appears to be a significant knowledge gap on the demands and strategies of women's movements in the increasing number of relatively stable states in the global South, where they might have the opportunity to develop a strategic repertoire over time. The role of gender-based social movements in securing substantive representation for women in developing countries is an important area for future research. How domestic women's movements draw on the growing transnational women's movement for collaboration, strategy, resources, and other forms of support to increase their political effectiveness is also an increasingly relevant, yet underexplored research question. 41 Horowitz.qxd 6/11/09 1:22 PM Page 42 Agriculture and Rural Development 5.1.2. Decentralization and Local Politics Do local political structures represent a realm in which women are more effective in achieving gender equity interests compared to national politics? Although local government is the sphere of government operating in closest physical proximity to rural women and thus ostensibly the easiest for them to access, conservative attitudes and elite capture may prove intractable at this scale. However, the particular ways in which these factors play themselves out within national and local political arenas is highly context specific, and because most of the evidence on this topic is relatively new following the recent wave of decentralization in the developing world, there is no clear consensus in the literature on this question.54 What does appear to be clear is that where decentralization policies are intended to be a tool for women's empowerment, they are more likely to succeed when they are implemented and supported by a strong central state itself committed to gender equity and by good intergovernmental coordination to take advantage of spillovers and externalities. Thus the key question may not be one of either/or--national or local political structures-- but rather how the advantages inherent in the different scales of government can all be used in the service of women's empowerment, using each to amplify the strengths of the other. Cross-country research aimed at understanding why there are fewer women in local government than in national parliaments in most regions would further this realm of inquiry. Setting up a comparison among Latin America, Southeast Asia, and other developing regions where women are systematically better represented or empowered at the local level might be particularly fruitful. In addition, far more work needs to be done on understanding how governments actually work at the local level and thus how specific men and women work within any given institutional context to get things done. In particular, increased attention should be paid to how the informal rules that govern formal institutions exclude women--or create unconventional opportunities. Formal political institutions may be stronger in democracies and autocracies than in the semidemocracies that govern much of the developing world (Bjanegard 2008; Helmke and Levitsky 2004). Again, the call for more microlevel studies resonates here, specifically of the gendered political dynamics of decentralization and the political spaces it creates. As Hirschmann suggests (1991, 1679), "we [should] not overlook or close off those narrow spaces of participation which may, by default, have been allowed to women, or which women may have fashioned for themselves," but rather structure research with the time and patience necessary to discover and understand these. Gender equity policies built on an understanding of the realities of grassroots politics in specific places will inevitably be more effective and meaningful. Finally, Beckwith notes that "despite a growing body of case research, no systematic assessment of decentralized versus centralized state power has been undertaken with reference to feminist movements" (2000, 456). A systematic examination of how different state structures influence the quality 42 Horowitz.qxd 6/11/09 1:22 PM Page 43 Getting Good Government for Women of the space for women's political organizing would thus be an important contribution. 5.1.3. Public Sector Capacity and Responsiveness This review found a disconcerting silence in the public administration literature on gender. Outside of the specialized literature on state feminism or gender machinery, there is almost no empirical work on strategies to increase gender responsiveness within the bureaucracy. How to integrate gender sensitivity into the norms, incentives, and practices of mainstream agencies in the public administration of developing countries is a topic little studied, at least in the peer-reviewed literature reviewed here.55 Serious studies documenting and evaluating context-specific experiences with initiatives aiming to alter such bureaucratic incentive systems as performance reviews and indicators, communication flows, and informal professional cultures would thus be an important contribution to this literature. As well, further conceptual and empirical work on gender-specific forms of corruption and capture would be useful, followed by further elucidation of successful initiatives to combat these. In the context of widespread governance reforms in developing countries, it will be important to know whether and how public sector reforms alter women's space for political organizing and political effectiveness. As discussed, Goetz (2003) hypothesizes that reforms that model public service provision on the market have a socially fragmenting effect on poor consumers and especially on women. They may complicate efforts to lobby for change by changing the incentives for individuals to join in collective action. They also implicitly transform women's demands for services from one of a citizen of the state--enacted in the political arena--to one of a customer or a client, enacted in the market. Generating the empirical evidence to test the impact of these reforms on women's political voice is extremely important. While this paper did not look explicitly at the courts, police, or legal rights frameworks that may be the most formalized accountability mechanisms for women, the papers reviewed did raise important questions about governments' accountability to women. Specifically, a key question that remains is how to imbue existing informal accountability mechanisms such as gender-responsive budget initiatives or citizen audits with stronger repercussive influence, moving them beyond simple information provision projects. Without enforceability, the significant financial and social capital invested in such projects will continue to result in few real budget or policy changes. How to design accountability systems that make power holders answer for their performance in meeting gender equity goals is the key question in getting good government for women. Finally, the extent to which broader trends affecting governance in the developing world, such as economic globalization and liberalization, impact the ability of the public sector to play a major role in alleviating women's poverty and improving social power deserves further attention. Although this review starts from the assumption that accountability within a nation's public sector can centrally affect women's entitlements and thus their ability to 43 Horowitz.qxd 6/11/09 1:22 PM Page 44 Agriculture and Rural Development escape poverty, future research must honestly assess the degree to which any given country in the developing world has full leverage over its internal public agenda. 5.1.4. Operations and Outcomes The final knowledge gap to point out is the operational and practical implications of this body of scholarship. While the papers discussed above have established a great deal of understanding on both the barriers to and frame conditions for changes in women's political participation, they offer less guidance on the concrete actions that policy makers or development practitioners could take to improve governance for women. (Box 5.1 compiles the mechanisms that are discussed in the reviewed papers.) This bias was likely exacerbated by focusing this review on the peer-reviewed literature. Gray and sectoral literatures may hold the earliest indications of governance innovations, so detailed reviews of sectoral strategies (including operations literature) for taking women's and men's differential needs into account when designing and providing services will be a valuable future contribution to this discourse. This is extremely important given UNIFEM's finding that less than 5 percent of World Bank lending on public sector governance from 2002 to 2007 had a gender focus (UNIFEM 2009). As well, the literature, which focuses on the admittedly important topic of women's participation as a goal in its own right, does not yet provide much insight into the links between governance reforms and actual outcomes for poor women in their lives and livelihoods. Additional studies using household-level data to examine how different governance structures and/or levels of women's political enfranchisement affect public spending, service delivery, or poverty outcomes would be useful in this endeavor. Interdisciplinary partnerships between quantitative and qualitative streams within political science, economics, anthropology, geography, and others could be a powerful force to advance this inquiry. 44 Horowitz.qxd 6/11/09 1:22 PM Page 45 Getting Good Government for Women 6. Conclusion The historical absence of women's voices--as leaders, implementers, and citizens--in the management of public affairs and resources of their countries, as well as subtle bias in public institutions, has ensured the preservation of the gender inequity that hinders growth and development. And since the majority of the Millennium Development Goals hinge on household decisions conventionally controlled by women--nutrition, maternal and child health, education, drinking water, environmental protection--women's empowerment is central to achieving them. The current political climate presents unprecedented opportunities for the inclusion of women and women's issues in governance in the developing world. Women are frequenting the highest levels of politics at a greater rate than ever before, and women heads of state are becoming fairly common in Latin America, the Caribbean, East and South Asia, and Eastern Europe. Yet, as this review has shown, academic understanding of the specific mechanisms by which women can strengthen their public voice and by which elected and administrative public officials can be incentivized to meet the needs of poor women is still in early stages. Only a tiny segment of the vast contemporary literature on governance for development deals squarely with gender. Mainstream scholars and policy analysts concerned with improving public service provision and reducing problems of corruption and elite capture most often write about "citizens" or "poor people" or "policy makers" without acknowledging the different needs, opportunities, and power relationships of men and women within these populations. And while many feminist scholars are concerned with the topic of increasing women's political participation, few have traced the links between voice and pro-poor outcomes. When Amartya Sen estimated that more than 100 million more women would be alive today had women and girls around the world been provided with the same medicine, nutrition, care, and education as males, he exposed gender inequality and its disastrous ramifications as being among the more "momentous, and neglected, problems facing the world today" (1990b). Such an indictment points to the large and important research agenda ahead. 45 Horowitz.qxd 6/11/09 1:22 PM Page 46 Agriculture and Rural Development 7. References Agarwal, B. 2001. Participatory exclusions, community forestry, and gender: An analysis for South Asia and a conceptual framework. World Development 29:1623­48. Ahikire, J. 2003. Gender equity and local democracy in contemporary Uganda: Addressing the challenge of women's political effectiveness in local government. In No shortcuts to power: African women in politics and policy making, ed. A. M. Goetz and S. Hassim, 213­39. London and New York: Zed Books. Alhassan-Alolo, N. 2007. Gender and corruption: Testing the new consensus. Public Administration and Development 27:227­37. Alvarez, S. (1990). Engendering democracy in Brazil: Women's movements in transition politics. Princeton, NJ: Princeton University Press. Baden, S. 1996. Gender issues in financial liberalization and financial sector reform (Rep. No. 39). Brighton, UK: Institute of Development Studies. Baldez, L. 2001. Coalition politics and the limits of state feminism in Chile. Women & Politics 22:1­28. Ban, R., and V. Rao. 2006. Tokenism or agency? The impact of women's reservations on panchayats in South India. Washington, DC: World Bank. Banaszak, L. A., K. Beckwith, and D. Rucht. 2005. Women's movements facing the reconfigured state. Cambridge, MA: Cambridge University Press. Bardhan, P., D. Mookerjee, and M. Torrado. 2005. Impact of yeservations of panchayat pradhans on targeting in West Bengal. Bureau for Research in Economic Development Working Paper No. 104 Bauer, G. 2004. "The hand that stirs the pot can also run the country": Electing women to parliament in Namibia. The Journal of Modern African Studies 42:479­509. Bauer, G., and H. Britton. 2006. Women in African parliaments. Boulder, CO, and London: Lynne Rienner Publishers. Beall, J. 1997. Thoughts on poverty from a South Asian rubbish dump--Gender, inequality and household waste. IDS Bulletin­Institute of Development Studies 28:73­90. ------. 2005a. Decentralizing government and centralizing gender in Southern Africa: Lessons from the South African experience (Rep. No. 8). Geneva: United National Research Institute for Social Development. ------. 2005b. Decentralizing government and decentering gender: Lessons from local government reform in South Africa. Politics & Society 33:253­76. Beaman, L., R. Chattopadhyay, E. Duflo, R. Pande, and P. Topalova. 2008. Powerful women: Does exposure reduce bias? (Rep. No. 14198). Cambridge, MA: National Bureau of Economic Research. 46 Horowitz.qxd 6/11/09 1:22 PM Page 47 Getting Good Government for Women Beard, V. A. 2005. Individual determinants of participation in community development in Indonesia. Environment and Planning C: Government and Policy 23:21­39. Beck, L. J. 2003. Democratization and the hidden public--The impact of patronage networks on Senegalese women. Comparative Politics 35:147­69. Beckwith, K. 2000. Beyond compare? Women's movements in comparative perspective. European Journal of Political Research 37:431­68. ------. 2005. The comparative politics of women's movements. Perspectives on Politics 3:583­96. Bell, E. 2002. National machineries for women in development: Experiences, lessons and strategies. Brighton, UK: Institute of Development Studies. Besley, T., and M. Ghatak. 2004. Public goods and economic development. Working Paper, London School of Economics. Besley, T., V. Rao, and R. Pande. 2005. Participatory democracy in action: Survey evidence from South India. Journal of the European Economic Association 3:648­57. Bjanegard, E. 2008. Gender and corruption--Reversing the causal direction. In Annual Meeting of the American Political Science Association. Boston, MA. Boserup, E. 1970. Woman's role in economic development. New York: St. Martin's Press. Britton, H. (2002). Coalition building, election rules, and party politics: South African women's path to parliament. Africa Today 49:32­67. Brown, A. M. 2001. Democratization and the Tanzanian state: Emerging opportunities for achieving women's empowerment. Canadian Journal of African Studies 35:67­98. Budlender, D. 2000. The political economy of women's budgets in the South. World Development 28:1365­78. ------. 2005. Expectations versus realities in gender-responsive budget initiatives. New York: United Nations Research Institute for Social Development. Burgess, S. F., and J. C. Beilstein. 1996. Women's voice and agency in rural Africa: Small-scale communal farmers in Zimbabwe. Women & Politics 16:19­43. Chathukulam, J., and M. S. John. 2000. Empowerment of women panchayat members: Learning from Kerala. Asian Journal of Women Studies 6:66­101. Chattopadhyay, R., and E. Duflo. 2004. Women as policy makers: Evidence from a randomized policy experiment in India. Econometrica 72:1409­43. Childs, S., and M. L. Krook. 2005. The substantive representation of women: Rethinking the "critical mass" debate. In Annual Meeting of the American Political Science Association. ------. 2006. Should feminists give up on critical mass? A contingent yes. Politics and Gender 2:522­30. Coles, A., and T. Wallace. 2005. Gender, water, and development. Oxford: Berg Publishers. Compaore, N. 2005. Legistlative recruitment and electoral systems: Burkina Faso. In Women in parliament: Beyond numbers, 2nd ed., ed. J. Ballington and A. Karam, 93­132. Stockholm: International Institute for Democracy and Electoral Assistance. 47 Horowitz.qxd 6/11/09 1:22 PM Page 48 Agriculture and Rural Development Cornwall, A. 2003. Whose voices? Whose choices? Reflections on gender and participatory development. World Development 31:1325­42. Crook, R. C., and J. Manor. 1998. Democracy and decentralization in South Asia and West Africa: Participation, accountability, and performance. Cambridge, UK: Cambridge University Press. Dahelerup, D. 1998. Using quotas to increase women's political representation. In Women in parliament: Beyond numbers, ed. J. Lovenduski and A. Karam. Stockholm: International IDEA. Datta, B. 1998. And who will make the chapatis? A study of all-women panchayats in Maharashtra. Calcutta: Stree. Dauda, C. 2001. Preparing the grounds for a new local politics: The case of women in two African municipalities. Canadian Journal of African Studies 35:246­281. Demery, L. 1996. Gender and public social spending: Disaggregating benefit incidence. Washington, DC: World Bank. Devas, N., and U. Grant. 2003. Local government decision-making--Citizen participation and local accountability: Some evidence from Kenya and Uganda. Public Administration and Development 23:307­16. Devlin, C., and R. Elgie. 2008. The effect of increased women's representation in Parliament: The case of Rwanda. Parliamentary Affairs 61:237­54, Advance Access. Dollar, D., R. Fisman, and R. Gatti. 2001. Are women really the "fairer" sex? Corruption and women in government. Journal of Economic Behavior and Organization 46:423­29. Doss, C. R. 2001. Designing agricultural technology for African women farmers: Lessons from 25 years of experience. World Development 29:2075­92. Dovi, S. 2002. Preferable descriptive representatives: Will just any women, black, or Latino do? American Political Science Review 96:729­42. Due, J. M., F. Magayane, and A. A. Temu. 1997. Gender again--Views of female agricultural extension officers by smallholder farmers in Tanzania. World Development 25:713­25. Duflo, E., and P. Topalova. 2004. Unappreciated service: Performance, perceptions, and women leaders in India. Cambridge, MA: Massachusetts Institute of Technology Poverty Action Lab Paper. Eagly, A. H., and S. J. Karau. 2002. Role congruity theory of prejudice toward female leaders. Psychological Review 109:573­98. Elsey, H., N. Kilonzo, R. Tolhurst, and C. Molyneux. 2005. Bypassing districts? Implications of sector-wide approaches and decentralization for integrating gender equity in Uganda and Kenya. Health Policy and Planning 20:150­57. Elson, D. 1998. Integrating gender issues into national budgetary policies and procedures: Some policy options. Journal of International Development 10:929­41. Fallon, K. (2003). Transforming women's citizenship rights within an emerging democratic state. Gender and Society 17, 525­43. Fonchingong, C., and C. Ngwa. 2006. Rethinking the cost-benefit equation of women's participation in community-driven development in North-Western Cameroon. Indian Journal of Gender Studies 13:61­82. 48 Horowitz.qxd 6/11/09 1:22 PM Page 49 Getting Good Government for Women Franceschet, S. 2003. "State feminism" and women's movements: The impact of Chile's Servicio Nacional de la Mujer on Women's Activism. Latin American Research Review 38:9­40. Friedman, E. J. 2000. State-based advocacy for gender equality in the developing world: Assessing the Venezuelan National Women's Agency. Women & Politics 21:47­80. Geisler, G. 1995. Troubled ssterhood: Women and politics in southern Africa: Case studies from Zambia, Zimbabwe, and Botswana. African Affairs 94:545­78. Gideon, J. 2006. Integrating gender interests into health policy. Development and Change 37:329­52. Gladwin, C., and D. McMillan. 1989. Is a turnaround in Africa possible without helping African women to farm? Economic Development and Cultural Change 37:345­69. Goetz, A. M. 1992. Gender and administration. IDS Bulletin­Institute of Development Studies 23:6­17. ------. 1995. The politics of integrating gender to state development processes: Trends, opportunities and constraints in Bangladesh, Chile, Jamaica, Mali, Morocco, and Uganda. (Rep. No. 2). New York: United Nations Research Institute for Social Development. ------. 2001. Women development workers: Implementing rural credit programmes in Bangladesh. Dhaka: University Press Limited. ------. 2002. No shortcuts to power: Constraints on women's political effectiveness in Uganda. Journal of Modern African Studies 40:549­75. ------. 2005. Women in politics and public life. In Gender equality: Striving for justice in an unequal world, ed. United Nations Research Institute for Social Development, 145­203. New York: United Nations. ------. 2007. Political cleaners: Women as the new anti-corruption force? Development and Change 38:87­105. Goetz, A. M., and J. Gaventa. 2001. Bringing citizen voice and client focus into service delivery. (Rep. No. 138). Sussex: Institute of Development Studies. Goetz, A. M., and R. S. Gupta. 1995. Who takes the credit? Gender, power, and control over loan use in rural credit programs in Bangladesh. World Development 24:45­63. Goetz, A. M., and S. Hassim, eds. 2003. No shortcuts to power: African women in politics and policy making. London and New York: Zed Books. Goetz, A. M., and R. Jenkins. 2005. Reinventing accountability: Making democracy work for human development. Baskingstoke, UK: Palgrave Macmillan. Goldstein, M., and C. Udry. 2008. The profits of power: Land rights and agricultural investments in Ghana. Journal of Political Economy 116:981­1022. Gray, T. 2003. Electoral gender quotas: Lessons from Argentina and Chile. Bulletin of Latin American Research 22:52­78. Greenberg, M., and R. Okani. 2001. Strengthening women's participation in decision- making at the local level in Mali. Bethesda, MD: Development Alternatives, Inc. Grindle, M. 1997. Getting good government: Capacity building in the public sectors of developing countries. Cambridge, MA: Harvard University Press. 49 Horowitz.qxd 6/11/09 1:22 PM Page 50 Agriculture and Rural Development Guijt, I., and M. Kaul. 1998. The myth of community: Gender issues in participatory development. London: Earthscan. Haddad, L., J. Hoddinott, and H. Alderman. 1997. Intrahousehold resource allocation in developing countries: Models, methods, and policy. Baltimore, MD: Johns Hopkins University Press for the International Food Policy Research Institute. Hainard, F., and C. Verschuur. 2001. Filling the urban policy breach: Women's empowerment, grass-roots organizations, and urban governance. International Political Science Review 22:33­53. Hassim, S. 2006a. Paradoxes of representation: The impact of quotas for women on democratization. In Yale conference on representation and popular rule. New Haven, CT: Yale University Press. ------. 2006b. The virtuous cycle of representation: Women in African parliaments. In Women in African parliaments, ed. G. Bauer and H. Britton, 171­86. Boulder, CO, and London: Lynne Rienner Publishers. Helmke, G., and S. Levitsky. 2004. Informal institutions and comparative politics: A research agenda. Perspectives on Politics 2:725­40. Hirschmann, D. 1991. Women and political participation in Africa: Broadening the scope for research. World Development 19:1679­94. Honour, T., J. Barry, and S. Palnitkar. 1998. Gender and public service: A case study of Mumbai. International Journal of Public Sector Management 11:188­200. Htun, M. 2004. Is gender Like ethnicity? The political representation of identity groups. Perspectives on Politics 2:439­58. Htun, M., and M. Jones. 2002. Engendering the right to participate in decision making: Electoral quotas and women's leadership in Latin America. In Gender and the Politics of Rights and Democracy in Latin America, ed. N. Craske and M. Molyneux, 32­56. New York: Palgrave. Inglehart, R., and P. Norris. 2003. Gender equality and cultural change. Cambridge, UK: Cambridge University Press. International Institute for Democracy and Electoral Assistance (IDEA). 2007. Quotas for women. Stockholm: IDEA. Inter-Parliamentary Union. Women in national parliaments. Retrieved March 15, 2009 from http://www.ipu.org/wmn-e/world.htm Jahan, R. 2005. Background paper on the role of national mechanisms in promoting gender equality and the empowerment of women: Achievements, gaps, and challenges for the future. (Rep. No. EGM/National Machinery/2004/BP.1). Rome: United Nations Division for the Advancement of Women. Jaquette, J., and S. Wolchik. 1998. Women and democracy: Latin America and Central and Eastern Europe. Baltimore, MD: Johns Hopkins University Press. Jayal, N. G. 2006. Engendering local democracy: The impact of quotas for women in India's panchayats. Democratization 13:15­35. Jewkes, R., N. Abrahams, and Z. Mvo. 1998. Why do nurses abuse patients? Reflections for South African obstetric services. Social Science and Medicine 47:1781­95. Johnson, D. 2003. Women in Ugandan local government: The impact of affirmative action. Gender and Development 11:8­18. 50 Horowitz.qxd 6/11/09 1:22 PM Page 51 Getting Good Government for Women Jones, M. 1997. Legislator gender and legislator policy priorities in the Argentine Chamber of Deputies and the United States House of Representatives. Policy Studies Journal 25:613­29. ------. 1998. Gender quotas, electoral laws, and the election of women: Lessons from the Argentine provinces. Comparative Political Studies 31:3­21. ------. 2004. Quota legislation and the election of women: Learning from the Costa Rican experience. The Journal of Politics 66:1203­33. Karam, A., and J. Lovenduski. 2005. Women in parliament: Making a difference. In Women in parliament: Beyond numbers, ed. J. Ballington and A. Karam, 187­212. Stockholm: International Institute for Democracy and Electoral Assistance. Kenworthy, L., and M. Malami. 1999. Gender inequality in political representation: A worldwide comparative analysis. Social Forces 78:235­68. Krook, M. L. 2006. Reforming representation: The diffusion of candidate gender quotas worldwide. Politics and Gender 2:303­27. Kudva, N. 2003. Engineering eections: The experiences of women in Panchayati Raj in Karnataka, India. International Journal of Politics, Culture, and Society 16:445­63. Kunovich, S., and P. Paxton. 2005. Pathways to power: The role of political parties in women's national political representation. American Journal of Sociology 111:505­52. Lincove, J. A. 2006. Efficiency, equity and girls' education. Public Administration and Development 26:339­57. Lindberg, S. I. 2004. Women's empowerment and democratization: The effects of electoral systems, participation, and experience in Africa. Studies in Comparative International Development 39:28­53. Lipsky, M. 1980. Street-level bureaucracy: Dilemmas of the individual in public services. New York: Russell Sage Foundation. Lovenduski, J., and J. Hills. 1981. The politics of the second electorate: Women and public participation. London: Routledge and Keegan Paul, Ltd. Manor, J. 2004. User committees: A potentially damaging second wave of decentralization? European Journal of Development Research 16:192­213. Mansbridge, J. 1999. Should blacks represent blacks and women represent women? A contingent "yes." The Journal of Politics 61:628­57. Matland, R. 1998. Women's representation in national legislatures: Developed and developing countries. Legislative Studies Quarterly 23:109­25. Mayoux, L. 1995. Beyond naivety: Women, gender inequality and participatory development. Development and Change 26:235. Mbatha, L. 2003. Democratising local government: Problems and opportunities in the advancement of gender equality in South Africa. In No shortcuts to power: African women in politics and policy making, ed. A. M. Goetz and S. Hassim, 188­212. London and New York: Zed Books. McAllister, I., and D. Studlar. 1992. Gender and representation among legislative candidates in Australia. Comparative Political Studies 25:388­411. McEwan, C. 2000. Engendering citizenship: gendered spaces of democracy in South Africa. Political Geography, 19:627­51. 51 Horowitz.qxd 6/11/09 1:22 PM Page 52 Agriculture and Rural Development ------. 2003. "Bringing government to the people": Women, local governance and community participation in South Africa. Geoforum 34:469­81. Meena, R. 2003. The politics of quotas in Tanzania. In IDEA/EISA/SADC parliamentary forum conference. South Africa: International Institute for Democracy and Electoral Assistance. Mehrotra, S. 2006. Governance and basic social services: Ensuring accountability in service delivery through deep democratic decentralization. Journal of International Development 18:263­83. Meinzen-Dick, R., L. Brown, H. S. Feldstein, and A. Quisumbing. 1997. Gender, property rights, and natural resources. World Development 25:1303­15. Meinzen-Dick, R., and M. Zwarteveen. 1998. Gendered participation in water management: Issues and illustrations from water users associations in South Asia. Agriculture and Human Values 15:337­45. Moore, G., and G. Shackman. 1996. Gender and authority: A cross-national study. Social Science Quarterly 77:273­88. Moser, C. 1993. Gender planning and development: Theory, practice and training. London and New York: Routledge. Nanda, P. 2002. Gender dimensions of user fees: Implications for women's utilization of health care. Reproductive Health Matters 10:127­34. Nandal, S. 2003. Reflections on new partnerships between women and local self- government in India: A rural revolution? Journal of International Women's Studies 5:122­31. Nelson, B., and N. Chowdhury. 1999. Women and politics worldwide. New Haven, CT: Yale University Press. Norris, P., and J. Lovenduski. 1993. Gender and party politics. London: Sage. Norris, P., and R. Inglehart. 2001. Cultural obstacles to equal representation. Journal of Democracy 12:126­41. Ofei-Aboagye, E. 2004. Promoting gender sensitivity in local governance in Ghana. Development in Practice 14:753­60. Ohene-Konadu, G. 2007. Gender analysis and interpretation of barriers to women's participation in Ghana's decentralized local government system. Cape Town: African Gender Institute, University of Cape Town. Okeke-Ihejirika, P. E., and S. Franceschet. 2002. Democratization and state feminism: Gender politics in Africa and Latin America. Development and Change 33:439-66. Okome, M. O. 2000. Women, the state, and the travails of decentralizing the Nigerian Federation. West Africa Review 2:1. Omar, O., and V. Ogenyi. 2004. A qualitative evaluation of women as managers in the Nigerian civil service. The International Journal of Public Sector Management 17:360­73. Opare, S. 2005. Engaging women in community decision-making processes in rural Ghana: Problems and prospects. Development in Practice 15:90­99. Orock, R. 2007. Gender equality--Whose agenda? Observations from Cameroon. Development in Practice 17:93­97. 52 Horowitz.qxd 6/11/09 1:22 PM Page 53 Getting Good Government for Women Pal, M. 1998. Women in panchayat: Experience of a training camp. Economic and Political Weekly, January 24. Patterson, A. 2002. The impact of Senegal's decentralization on women in local governance. Canadian Journal of African Studies 36:490­529. Paxton, P. 1997. Women in national legislatures: A cross-national analysis. Social Science Research 26:442­64. Paxton, P., and S. Kunovich. 2003. Women's political representation: The importance of ideology. Social Forces 82:87­114. Phillips, A. 1991. Engendering democracy. University Park: Pennsylvania State University Press. Powley, E. 2006. Rwanda: The impact of women legislators on policy outcomes affecting children and families. New York: UNICEF. Purushothama, S., P. Kumar, and S. Purohit. 2000. Engendering local governance: The Karnataka experience. Bangalore: Best Practices Foundation in collaboration with the Hunger Project. Quisumbing, A. 2003. Household decisions, gender, and development: A synthesis of recent research. Washington, DC: International Food Policy Research Institute (distributed by Johns Hopkins University Press). Quisumbing, A., L. Brown, H. S. Feldstein, L. Haddad, and C. Pena. 1995. Women: The key to food security. (Rep. No. 21). Washington, DC: International Food Policy Research Institute. Quisumbing, A., E. Payaongayong, J. B. Aidoo, and K. Otsuka. 2001. Women's land rights in the transition to individualized ownership: Implications for tree-resource management in Western Ghana. Economic Development and Cultural Change 50:157­81. Rahat, G. 2007. Candidate selection: The choice before the choice. The Journal of Democracy 18:157­70. Rai, S. 2003. Mainstreaming gender, democratizing the state: Institutional mechanisms for the advancement of women. Manchester, UK: Manchester University Press. Rai, S., N. Shah, and A. Ayaz. 2007. Achieving gender equality in public offices in Pakistan. Islamabad: United Nations Development Programme Pakistan. Rakodi, C. 2002. Influence and accountability: Citizen voices, responsiveness, and accountability in service delivery. London: One World Action. Rama, M. 2002. The gender implications of public sector downsizing: The reform program of Vietnam. The World Bank Research Observer 17:167­89. Randall, V. 1987. Women and politics: An international perspective. Chicago: University of Chicago Press. Randall, V., and G. Waylen. 1998. Gender, politics and the state. New York: Routledge. Rangan, H., and M. Gilmartin. 2002. Gender, traditional authority, and the politics of rural reform in South Africa. Development and Change 33:633­58. Rao, A., and D. Kelleher. 2005. Affirmative action as a means of social inclusion: Myths and realities. In Gender and social inclusion assessment for Nepal. Washington, DC: World Bank. 53 Horowitz.qxd 6/11/09 1:22 PM Page 54 Agriculture and Rural Development Ray, R., and A. C. Korteweg. 1999. Women's movements in the Third World: Identity, mobilization and autonomy. Annual Review of Sociology 25:47­71. Reynolds, A. 1999. Women in the legislatures and executives of the world: Knocking at the highest glass ceiling. World Politics 51:547­72. Rule, W. 1987. Electoral systems, contextual factors, and women's opportunity for election to parliament in twenty-three democracies. The Western Political Quarterly 40:477­98. Saito, F. 2002. Decentralization measures and gender equalities: Experiences in Uganda. Ryukoku Journal of Economic Studies 41. Sanyal, P. 2008. From credit to collective action: Microfinance, women's group-based mobilizations and implications for social capital. Cambridge, MA: Harvard University Press. Sargent, C. 1989. Maternity, medicine, and power: Reproductive decisions in urban Benin. Berkeley: University of California Press. Sawer, M. 2005. Gender equality in the age of governing for the mainstream. (Rep. No. EGM/National Machinery/2004/EP.4). Rome: United Nations Division for the Advancement of Women. Schwindt-Bayer, L. 2004. Women's representation in Latin American legislatures: Policy attitudes and bill initiation behavior. In Annual meeting of the Midwest Political Science Association. Chicago. Seidman, G. W. 1999. Gendered citizenship--South Africa's democratic transition and the construction of a gendered state. Gender & Society 13:287­307. ------. 2003. Institutional dilemmas: Representation versus mobilization in the South African Gender Commission. Feminist Studies 29:541­63. Selolwane, O. D. 2008. From political sidecars to legislatures: Women and party politics in Southern Africa. In Governing women: Women's political rffectiveness in contexts of democratization and governance reform, ed. A. M. Goetz, 129­47. Oxford: Taylor & Francis. Sen, A. 1990a. Gender and cooperative conflicts. In Persistent inequalities: Women and world development, ed. I. Tinker, 123­49. New York: Oxford University Press. ------. 1990b. More than 100 million women are missing. The New York Review of Books Retrieved April 1, 2009 from http://www.nybooks.com/articles/3408 Sen, G. 2000. Gender mainstreaming in finance ministries. World Development 28:1379­90. Singer, W. 2007. "A constituency suitable for ladies" and other social histories of Indian elections. New York: Oxford University Press. Skinner, C., and I. Valodia. 2003. Local government support for women in the informal economy in Durban, South Africa. International Journal of Politics, Culture, and Society 16:431­44. Sultana, M., and K. McCarthy. 2004. Gender and governance issues in local government: Regional report of technical assistance in Bangladesh, Nepal, and Pakistan. Manila: Asian Development Bank. Swamy, A., S. Knack, Y. Lee, and O. Azfar. 2001. Gender and corruption. Journal of Development Economics 64:25­55. 54 Horowitz.qxd 6/11/09 1:22 PM Page 55 Getting Good Government for Women Swers, M. 1998. Are women more likely to vote for women's issue bills than their male colleagues? Legislative Studies Quarterly 23:435­48. ------. 2002. The difference women make: The policy impact of women in Congress. Chicago: University of Chicago Press. Tamale, S. 1999. When hens begin to crow: Gender and parliamentary politics in Uganda. Kampala, Uganda: Fountain Publishers. Taylor-Robinson, M., and R. Heath. 2003. Do women legislators have different policy priorities than their male colleagues? A critical case test. Women and Politics 24:77­100. Tendler, J. 1998. Good government in the tropics. Baltimore, MD: Johns Hopkins University Press. Tendler, J., and S. Freedheim. 1994. Trust in a rent-seeking world: Health and government transformed in Northeast Brazil. World Development 22:1771­92. Theobald, S., R. Tolhurst, H. Elsey, and H. Standing. 2005. Engendering the bureaucracy? Challenges and opportunities for mainstreaming gender in ministries of health under sector-wide approaches. Health Policy and Planning 20:141­49. Thomas, S. 1991. The impact of women on state legislative policies. The Journal of Politics 53:958­76. ------. 1994. How women legislate. Oxford: Oxford University Press. Tinker, I. 2004. Quotas for women in elected legislatures: Do they really empower women? Women's Studies International Forum 27:531­46. ------. 2006. Empowerment just happened: The unexpected expansion of women's organizations. In Women and gender equality in development theory and practice, ed. J. Jaquette and G. Summerfield, 268­302. Durham, NC: Duke University Press. ------. 2009. Assumption and realities: Electoral quotas for women. Georgetown Journal of International Affairs Winter/Spring:7­14. Tripp, A. 2000. Women and politics in Uganda. Oxford: James Currey. ------. 2001a. The politics of autonomy and cooptation in Africa: The case of the Ugandan women's movement. The Journal of Modern African Studies 39:101­28. ------. 2001b. Women's movements and challenges to neopatrimonial rule: Preliminary observations from Africa. Development and Change 32:33­54. Tripp, A., and A. Kang. 2008. The global impact of quotas: On the fast track to increased female legislative representation. Comparative Political Studies 41:338­61. Tsikata, D., and J. Kerr. 2000. Demanding dignity: Women confronting economic reforms in Africa. Ontario: North-South Institute. Turner, T., and P. O'Connor. 1994. Women in the Zambian civil service: A case of equal opportunities? Public Administration and Development 14:79­92. UNIFEM (United Nations Development Fund for Women). 2009. Progress of the world's women 2008/09--Gender and accountability: Who answers to women? New York: UNICEF. Valodia, I. 1998. Engendering the public sector: An example from the Women's Budget Initiative in South Africa. Journal of International Development 10:943­55. 55 Horowitz.qxd 6/11/09 1:22 PM Page 56 Agriculture and Rural Development Vengroff, R., Z. Nyiri, and M. Fugiero. 2003. Electoral system and gender representation in sub-national legislatures: Is there a national­sub-national gender gap? Political Research Quarterly 56:163­73. Vijayalakshmi, V. 2002. A report on the politics of inclusion: Adivasi women in local governance in Karnataka. Gender, Technology, and Development 6:269­83. Vissandjee, B., S. Abdool, A. Apale, and S. Dupere. 2006. Women's political participation in rural India: Discerning discrepancies through a gender lens. Indian Journal of Gender Studies 13:425­50. Vyasulu, P., and V. Vyasulu. 2000. Women in the Panchayati Raj: Grassroots democracy in India. In Women's political participation and good governance: 21st century challenges, ed. L. Hammadeh-Bannerjee, 41­49. New York: United Nations Development Programme. Walker, L., and L. Gilson. 2004. We are bitter but we are satisfied: Nurses as street- level bureaucrats in South Africa. Social Science and Medicine 59:1251­61. Waylen, G. 1994. Women and democratization: Conceptualizing gender relations in transition politics. World Politics 46:327­54. ------. 2000. Gender and democratic politics: A comparative analysis of consolidation in Argentina and Chile. Journal of Latin American Studies 32:765­93. ------. 2003. Gender and transitions: What do we know? Democratization 10:157­78. ------. 2007. Engendering transitions: Women's mobilization, institutions and gender outcomes. Oxford: Oxford University Press. Weinberger, K., and J. P. Jutting. 2001. Women's participation in local organizations: Conditions and constraints. World Development 29:1391­1404. Wilkins, V. 2006. Exploring the causal story: Gender, active representation, and bureaucratic priorities. Journal of Public Administration Research and Theory 17:77­94. Wilson, J. 1989. Bureaucracy. New York: Basic Books. World Bank. 2003. World development report 2004: Making services work for poor people. Washington, DC, and Oxford: World Bank and Oxford University Press. Yoon, M. Y. 2004. Explaining women's legislative representation in Sub-Saharan Africa. Legislative Studies Quarterly 29:447­68. Zafarullah, H. 2000. Through the brick wall, and the glass ceiling: Women in the civil service in Bangladesh. Gender, Work, and Organization 7:197­209. 56 Horowitz.qxd 6/11/09 1:22 PM Page 57 Getting Good Government for Women Endnotes 1 In work spanning more than a decade, Anne Marie Goetz (a researcher currently acting as chief advisor on governance, peace, and security to UNIFEM) has laid the conceptual foundation for an understanding of gender- sensitive governance that takes into account both women's political voice-- inside and outside of the state--and the accountability relationships that link voice to political and institutional action (Goetz and Hassim 2003). One of the earliest and most productive scholars relating issues of gender and governance, her work is frequently cited in this review. 2 Initiatives are commonly understood to be gender sensitive if they fall into one or more of the following dimensions: (1) sensitive to gender differentials, (2) gender specific, (3) empowering to women, or (4) transformative, in that they work to alter gender relations in society, often through changing prevalent attitudes and social norms. 3 The search approach used for this compilation included four principal sources: ISI Web of Knowledge, Google Scholar, other published bibliographies, and references from published articles. ISI and Google are both searchable databases that have millions of references in various fields. At the end of 2008, ISI had more than 8,500 references under poverty and gender or women. The major research terms used to cull this literature during a first cut included governance, public administration, decentralization, politics, participation, and public sector reform. 4 Sources included in the reference list but solely relating to advanced industrialized countries and those lacking an explicit gender focus are excluded from this count. 5 The so-called second wave of feminism is a period of activism in the United States and elsewhere, generally recognized to have begun in the early 1960s, that fought for the end of legalized sex discrimination and for the right of women to have equal opportunity in the workforce. It was preceded by the original feminist movement, which secured women's right to vote in the United States, and was followed by a third wave of feminism that has placed issues of race, class, and sexuality at its center. 6 For an overview of the field, see the Web site of the Women and Politics Research Section of the American Political Science Association at http://www.apsanet.org/~wpol/index.htm. 7 Although Matland (1998) points out that, although women participate in the labor force in high rates in many developing countries, many work in subsistence-level, primary sectors, which may not have the same consciousness-raising effect as employment in the professional occupations. 57 Horowitz.qxd 6/11/09 1:22 PM Page 58 Agriculture and Rural Development 8Proportional representation refers to a category of electoral systems that seek a close match between the percentage of votes that a group of candidates (e.g., a party) obtain in elections and the percentage of seats they receive in the representative body. It is contrasted with plurality voting systems, where disproportional seat distribution results from a winner-take-all rule. Common forms of proportional representation systems include (1) the party list proportional representation discussed above, where seats within the party list get allocated to each party in proportion to the number of votes the party receives, and (2) the single transferable vote or instant runoff system. 9 The public nature of candidate lists also means that voters can hold parties accountable for including women, which is not possible in systems where a single candidate is selected by party functionaries behind closed doors. 10 Kenworthy and Malami (1999) usefully point out that, although most studies implicitly treat the party list/multimember districts system of balloting and the proportional representation method of seat allocation as a single entity, the two must be disentangled for comparative study. Whereas in Western Europe, party lists and proportional representation are always used together, a number of countries in Africa and Asia use party list balloting with a winner-take-all method of seat allocation, with a similar positive effect for women. 11Indeed, Namibian women activists and politicians explicitly drew on the "collective learning of decades of experiments with alternate electoral systems" to craft a prowomen electoral system when writing the new constitution in 1990 (Bauer 2004, 501). 12Kenworthy and Malami (1999) report on the striking disparities in women's representation in countries where different electoral systems operate at different levels: in Germany, which uses a mixed system for its lower house, 13 percent of the representatives elected from single-member districts are women, compared to 42 percent of those chosen via party lists in multimember districts. Other countries that have heterogeneous systems include France, Mexico, the United Kingdom, and the Argentine provinces. 13See Rahat (2007) on the understudied issue of candidate selection, though without an explicit gender focus. 14 Compaore (2005) notes that in Burkina Faso, political parties actually require nominees to party lists to pay a substantial cash deposit, which excludes many qualified women due to their financial constraints. 15 Under proportional representation systems, women's position among party elites translates into gains for women as candidates but not necessarily into elections, because voters elect lists rather than individual candidates. Female party leaders in plurality-majority systems are less instrumental in helping women in the competitive struggle for nomination, but for women who do achieve candidacy, female party elites can provide additional financial and institutional resources to their female candidates. They increase the likelihood that female candidates get elected. 58 Horowitz.qxd 6/11/09 1:22 PM Page 59 Getting Good Government for Women 16 See also Fallon 2003; Okeke-Ihejirika and Franceschet 2002; Britton 2002; McEwan 2000; Waylen 2000; Seidman 1999; Waylen 1994; Alvarez 1990. 17 Obviously, this relationship itself is strongly conditioned by the social culture. Eagly and Karau (2002) point out that the degree of perceived incongruity between a leader's role and the female gender role depends on factors such as the definition of the leader roles, the weight given to the female gender roles, and individuals' own approval of traditional definitions of gender roles, all of which are largely "ideological" and also change over time (578). What this theory indicates for women's current real-world prospects remains unclear, however, since Inglehart and Norris (2003) find little evidence of major transformations in culture in postcommunist and developing countries as captured in the newest World Values Survey. There traditional values are prevalent among younger as well as older citizens (136). 18 Tamale (1999, 23) notes an interesting regional distinction: affirmative action policies in most western countries are implemented in the spheres of employment and education, but nonwestern countries are experimenting with it primarily in the political sphere. 19 See http://www.quotaproject.org. 20Some parties, such as South Africa's ANC, also use gender quotas for internal party positions. 21The massive Indian experiment with reservations in local government is an important exception, but in this system entire constituencies are reserved for female-only competition. 22Except for Nepal, which uses reservations at the local level but a 5 percent quota at the national level. 23Gender reservations at both the local and district level in Uganda were also implemented by expanding the number of seats in councils rather than reallocating existing seats to women, thus leaving intact established (male) competition for ward seats. At the local level, women's seats also cover a larger constituency than regular ward seats, which create ambiguities about the constituencies they are supposed to represent, and are elected with a different voting system a full two weeks after the ward elections, lending an afterthought feeling to the process. 24 This theory appears to have been borne out in Argentina, where alliances among women groups formed to push through affirmative action declined after this nominal success was achieved; female MPs have had limited impact in promoting women's issues since then because they lack a strong constituency in civil society (Gray 2003). 25This has been described in the Rwandan case, where even though women members of parliament have been seen active in advancing a gender equality agenda, they are also constrained by the partisan nature of parliamentary politics in an era when the government has become increasingly authoritarian and intolerant of dissent. "Rwandan women legislators therefore find 59 Horowitz.qxd 6/11/09 1:22 PM Page 60 Agriculture and Rural Development themselves facilitating policies that strengthen an authoritarian state and undermine individual civil liberties" (Bauer and Britton 2006, 23). 26This school of thought assumes that women hold feminist policy preferences but that they must constitute a critical mass within political bodies, usually at least 30 percent, to be able to make a difference. See the December 2006 volume of Politics and Gender for an overview and critique of the critical mass debate. 27Although Childs and Krook (2006) point out that what is defined as a woman's issue in this literature is often less than empirically defined. 28 Ban and Rao (2006), who sought to test these findings in southern India where gender ideology is quite different, did not find female leaders more likely to make decisions that favor women's concerns than men; however, their survey did not as precisely measure women's preferences. 29 In fact, Lipsky's theory of the street-level bureaucrat culminates in the intuition that the informal practices of these actors, rather than the intentions or objectives of documents and statements developed at a central level, become the actually enacted public policy as experienced by the general public. 30They would also transform the question of "what women do" to "what do specific women do in specific contexts." Such an agenda has important methodological implications. It would require political scientists to "abandon quantitative work that infers micro-level preferences from macro-level patterns in favor of qualitative studies that draw on micro-level behaviors to explain macro-level results, at least in the short term" (15). 31 Significantly, it is in this field where much of the gender/governance literature on Latin America and Eastern Europe appears, regions that are notably underrepresented in other topic areas in this paper. 32Exceptions include Tripp (2001a; 2001b) and contributions to Goetz and Hassim (2003). 33Hassim (2006b) suggests that multiparty systems multiply the leverage that feminist constituencies can have because there are more parties to play off of. 34 There is some evidence that women in politics may encourage greater political engagement by ordinary women. Chattopadhyay and Duflo (2004) find that the fraction of women participating in West Bengal gram panchayats whose presidency is reserved for a woman is significantly greater than in unreserved ones. Village women in Karnataka also told Kudva (2003) they found it easier to approach women representatives about issues they were concerned about. This aspect has not been studied outside of India. 35These are Barbados, Chile, Iceland, India, Finland, Guatemala, Malta, Puerto Rico, and Sweden. 36Although early machineries tended to be isolated structures that actually implemented welfare-oriented projects, it is now generally recognized that the machinery should act as a catalyst for gender mainstreaming in all areas of policy and administration. 60 Horowitz.qxd 6/11/09 1:22 PM Page 61 Getting Good Government for Women 37 The primary resources focused on machineries in developing countries include an edited volume on cross-country experiences (Rai 2003), a selection of country case studies (Seidman 2003; Franceschet 2003; Baldez 2001; Friedman 2000), and a handful of overview papers prepared by affiliates of the United Nations and other development research institutions that draw on qualitative research in selected countries (Jahan 2005; Sawer 2005; Bell 2002; Goetz 1995). Several case studies also exist on the experience of gender mainstreaming in various sectoral ministries of developing countries, with the majority of these focusing on the health sector (Gideon 2006; Theobald et al. 2005; Elsey et al. 2005; Sen 2000). Gray literature country case studies prepared for the UN expert meeting on "The Role of National Mechanisms in Promoting Gender Equality and the Empowerment of Women: Achievements, Gaps, and Challenges" (http://www.un.org/womenwatch/daw/egm/nationalm2004/documents. html) provide contextual detail but will not be individually reviewed here. 38Placing machinery leadership in the hands of first ladies and other wives of politicians exacerbated this problem (Jahan 2005). 39 One promising development reported since the Fourth World Conference on Women is the significant increase in the number of special offices for women/gender created in municipal, district, and provincial governments (Jahan 2005). 40The states of West Bengal, Rajasthan, Karnataka, Andhra Pradesh, MP, Tamil Nadu, and Kerala have been the most heavily researched. The only all-India study that has been attempted so far is the examination of data from the Millennial Survey by Duflo and Topalova (2004). 41 See Singer (Singer 2007) for a history of India's reservation politics. 42Reservations for women have been implemented in all states except Bihar and Uttar Pradesh (which reserves only 25 percent of seats for women). Seats and presidencies were also reserved for the two disadvantaged minorities in India--scheduled castes and scheduled tribes--in proportion to each minority's population share in the district. 43There is a surprising absence of work on gender and local governance in Latin America, given that this is the one region where women are systematically better represented at the local level (see Goetz 2005, 194). 44 There is thus a significant bias in this literature (especially the English- language literature) toward regions with high degrees of gender discrimination, which should be noted before drawing generalizations that do not apply in more egalitarian societies. 45 Whether there are also more subtle personality and leadership characteristics that differentiate women who choose to stand for and hold public office from their peers is also a neglected topic, at least in the international academic literature. 46A debate has raged in India over the socioeconomic characteristics of the women being elected under the panchayat system. Early popular rhetoric suggested that women in reserved seats were simply token representatives: 61 Horowitz.qxd 6/11/09 1:22 PM Page 62 Agriculture and Rural Development older, poorer, or illiterate women who front for more powerful elites or male relatives. (Sometimes the argument is flipped to suggest that more upper caste women serve as representatives of their powerful local families (Kudva 2003).) However, a recent household survey conducted in four southern Indian states disputes the tokenism hypothesis, finding instead that women presidents are selected from among the more politically knowledgeable and wealthier women. Although reserved presidents compare less favorably in socioeconomic indicators to the male presidents in unreserved Gram Panchayats (but not in performance), this is likely attributed more to prevalent gender inequities in society than to any selection bias among the pool of women (Ban and Rao 2006). 47 Trained women representatives in one project in Uganda became more articulate than their fellow councilors, raising the hackles of their male peers. The organization began to train men as well as women to avoid such difficulties. 48Chattopadhyay and Duflo (2004) found that in Udaipur district in Rajasthan, none of the women who had been elected on a reserved seat in 1995 were reelected in 2000 without reservation. 49Older women may be an exception to this trend. Several studies have noted a greater acceptance of older women in politics, perhaps because these women are viewed as postmotherhood and postsexual, and therefore garner less judgment regarding traditional gender roles and probity. 50 However, Goetz (2005) reminds us that in Madhya Pradesh one-third of people assembled must be women in order to reach a quorum and that in Rajasthan women must be present in the village assembly in the same proportions that they are found in the local community--50 percent (199). Still, these measures may be abused by local councilors who ensure that only their supporters among women attend. 51 Manor suggests that the degree to which governments in general, and specifically the low-level government employees with which user committees work, cede top-down control often depends on the specific sector or subfield in question. 52For example, in Agarwal's (2001) study of community forestry groups in Indian and Nepal, many groups followed rules that restricted membership to one person per household, which effectively excluded women. Even in groups that had more inclusive groups, either a lack of awareness of rules or progressive changes to them constrained women's participation. 53 Hassim (2006b) describes such a "vicious circle" in politics more broadly: if women increase their presence in public forums, whether parliament, local councils, or committees, and see little impact on policy outcomes, women citizens may exit from formal political participation or party activism (185). See also Geisler (1995). 54An apparent bias in the published literature on gender and decentralization toward those countries that have a high degree of gender segregation or discrimination also contributes to the lack of clarity on this question. The 62 Horowitz.qxd 6/11/09 1:22 PM Page 63 Getting Good Government for Women natural focus on the South Asian experiment, which offers an enormous scale as well as opportunities for econometric analysis, partially explains this bias. It is also probably exacerbated by the English-language focus of this review, which excludes some literature from regions with more gender egalitarian societies. Still, researchers should remain attentive to this skew in the literature because work on decentralization should be grounded in the context of gender relations in particular societies. 55Several experts consulted for this paper suggested that there may be a bias in the public administration literature against feminist scholars and scholarship. This could be one reason why peer-reviewed publications may not adequately represent the kind of thinking and experimentation that is going on in the field. 63 Horowitz.qxd 6/11/09 1:22 PM Page 64