Women’s Empowerment Strategic Impact Inquiry: FY05 Methodological Note The strategic impact inquiry (SII) into CARE‘s impacts on women‘s empowerment was designed as a three-year process. FY05 was the first year and comprised seven methodologies: 1. Original field research in four project sites (Bangladesh, Yemen, India, and Ecuador); 2. Global secondary literature reviews of women‘s empowerment research; 3. Proposal analysis of a randomly selected sample of 32 CI projects from FYs 03 and 04; 4. Analysis of data from the FY04 CARE program information network (CPIN) survey; 5. Meta-evaluation of a convenience sample of 31 CARE project evaluations; 6. A gender mapping exercise conducted in Asia region; and 7. Informal harvesting of results/findings from impact research undertaken by CARE actors not participating in the SII. This package of methods and the three-year time frame were not random choices. They were based on specific design and measurement challenges that an investigation into a phenomenon like women‘s empowerment faces. Women’s Empowerment SII Research Design Challenge Conducting rigorous impact assessment on women‘s empowerment in a single project or program is a demanding chore. As numerous studies have argued (Malhotra et. al. 2002; Guijt & Shah 1998, 11-12) empowerment can have very different meanings and normative values in different socio-cultural contexts. As one of CARE‘s own evaluators wrote: It should not be assumed that western feminist definitions of women‘s roles and behaviour lead to appropriate empowerment goals, nor should it be assumed that they are sought by either women or men…. Empowerment can be valued – and measured – as process or outcome, as instrumental or intrinsic good. It can be valued – and measured – in both idiographic and nomothetic terms. As such, it is notoriously difficult to pin down, and has been for many years: Even when empowerment is clearly defined, it remains a complex problem to measure it at the project level, for the following reasons: It is difficult to measure changes in states of mind (from disempowered to empowered). Measuring different elements of empowerment, e.g. who is making a decision, particularly at the household level, can be difficult and time-consuming, requiring in- depth study and detailed qualitative analysis. Participation is a key element in empowerment, but the measurement of participation is itself complex. Definitions of 'knowledge', 'self-respect' or other elements of empowerment may be culturally specific, and therefore vary between localities and by socio-economic grouping, ethnicity and age (CIDA 1997). This 1997 description of the challenges facing those who wish to understand development program impacts on women‘s empowerment largely remains true in 2005. An element that needs to be added to the measurement complexity, however, is that ―empowerment‖ is by no means an inevitable and linear process. Individuals and groups of women can become more and less empowered through time; the multiple dimensions of empowerment do not covary in any simple synchronization. Furthermore, there is a vast ―missing middle‖ when it comes to effectively understanding (and therefore measuring) impact on women‘s empowerment: in operationalizing empowerment, there is theoretical interest but less empirical attention to aggregations that fall in the middle, especially at the community level where institutional and normative structures such as family systems, infrastructure, gender ideologies, regional or local market processes, etc. are most likely to affect women‘s empowerment. It is often precisely at these intermediary levels that normative changes occur and where programmatic or policy interventions often operate (Malhotra et. al., 2002, 15). Such variability in conceptualization, operationalization, levels of analysis, and more make global aggregation of CARE‘s impacts on women‘s empowerment a particular challenge. 1 That empowerment is context, culture, and time specific is not up for debate. That some aspects of empowerment can be generalized and operationalized—and measured either quantitatively or qualitatively—cross culturally and cross nationally is also not debatable, although measures effective at capturing empowerment changes at such a macro level tend to tell us nothing about what happens to specific women or particular groups of women in a given time, place, and social context within the larger unit being analyzed. These aspects have received their most rigorous treatment and support in Mason & Smith‘s (2003) 56-community, five-country study of empowerment at the household level, which demonstrated the critical importance of community/social context in the constitution of individual empowerment: The first point is that power within the household—the particular aspect of women‘s empowerment on which we focus—is strongly influenced by social context (national and community) because it is strongly determined by social institutions rather than by individual characteristics. The second point is that all aspects of women‘s empowerment are multidimensional and the interrelations among different dimensions depend on social context. Finally, the socioeconomic proxies for women‘s empowerment used in past studies do a good job of indexing only some aspects of women‘s empowerment and power—and only in some contexts. For this reason, use of direct measures is preferable to the use of proxies. The substantive implication of this finding is that, in the short run, providing women with schooling or other resources may do little to empower them, although the evidence suggests that these investments can have payoffs for particular aspects of empowerment (Mason & Smith 2003, 1). The FY05 Process Given the above realities, the women‘s empowerment SII deployed multiple research methods in the FY05 process, methods that could help CARE build a global mosaic – as opposed to global aggregation – of the impacts of CARE programs on women‘s empowerment. 2 In this approach we held steady a conceptual framework for operationalizing women‘s empowerment 3 (Martinez & Glenzer 2005) that included three broad dimensions and 23 sub-dimensions for empowerment but which required that specific indicators within each sub-dimension be ground-truthed locally. The global research framework – including its operationalizations of empowerment, power, gender, and gender inequity – was developed over the course of four months (September-January) in collaboration with a wide group of external researchers, theorists, consultants, and activists. The framework was improved in a global SII meeting in Cairo in early February during which site teams from Bangladesh, Yemen, Niger, and Ecuador drafted impact research designs. Once finalized, the framework informed the development of analytical tools for all seven SII methods. In this way, IMLT was able to hold steady a theoretical, conceptual, and definitional core – as well as analytical procedures – across all methods. In order to raise the quality of research designs in Bangladesh, Niger, and Ecuador, IMLT organized a global volunteer team of researchers who reviewed and provided critiques/feedback to the site teams vis-à- vis their designs. Significant time and expense went into finding high quality, experienced research consultants to lead the processes and to engage in analysis. All in-depth, original field research had as its first step the establishment of meanings and appropriate indicators of women‘s empowerment in local 1 See Williams (2005) for an intriguing meso-level solution to the aggregation challenge. Alsop & Heinsohn (2005) offer a detailed framework for measuring empowerment, quantitatively and qualitatively, at a national level, although such measures have little to do with the levels and kinds of work that CARE engages in generally. DANIDA (2005) offers a similar solution. Other strategies for the aggregation challenge include the World Economic Forum‘s (2005) calculation of the ―gender gap‖ between nations. 2 For readers with more interest in the methodological complexities IMLT has tried to negotiate in insisting on joining bottom-up, grounded-theory based approaches (Strauss 1997) with universalizing, top-down methods, refer to Schuler et. al. (1995a, 1995b) who provided us with our major lead on how and why to structure the global FY05 research process as we did. 3 A capsule summary of that framework can be found in Annex 1. context; in this process, local women themselves played the most important role. In this way, we tried to avoid imposing inappropriate northern/western concepts and measures while still maintaining that women‘s empowerment is a universal good, a fundamental component of rights-based approaches that cannot be waved away through appeals to cultural relativism. Next Methodological Steps Once FY05 findings are finalized, IMLT will review and revise the global empowerment framework and identify which methods should be redeployed in FY06. It will also determine what, if any, additional methods should be adopted in order to get a better idea of CARE‘s global impact on women‘s empowerment. Already, however, it is clear that the following improvements to the global methodology are necessary: 1. CARE needs to identify a handful of empowerment proxy indicators for empowerment that can be collected in the annual CPIN survey process; 2. More explicit requirements to investigate structural and relational dimensions of gender inequity need to be put on in-depth, original field research around women‘s empowerment. Guidance and advice on how to do this should come from the center; 3. Longer-term and deeper partnerships with external experts, activists, and researchers need to be forged in order to raise the general level of research quality across sites; 4. In-depth, participative desk reviews of ‗flagship‘ women‘s empowerment projects are needed. Such desk reviews would allow CARE to find a working compromise between the expense of formal impact research and the general lack of rigor that self-reviews tend to exhibit; and 5. More flexible, creative, and lower-cost compromises for original field research need to be deployed if the goal of CARE USA is to have some reasonable global idea of its impacts on empowerment and gender inequity more broadly. Annex 1 Capsule Overview Women’s Empowerment Global Research Framework4: Introduction The Strategic Impact Inquiry (SII) mechanism, in general, is an attempt to better answer the critical question, ―are CARE programs impacting the underlying causes of poverty and rights denial, and if so, how?‖ CARE has recognized gender inequality as a root cause of poverty across the communities we serve, in particular through its impact on the capabilities of women. Therefore, in this SII we focus on the theme of CARE’s contribution to women’s empowerment and gender equity. The SII on women‘s empowerment explores two broad questions: What impacts (positive and negative) have CARE programs had, if any, on the empowerment of women and the advancement of gender equity? What evidence (pro and con) exists regarding the link between CAREs approaches, principles, and internal dynamics (staff, structure, policies, reward systems, culture, management, etc.) and the empowerment of women and the advancement of gender equity? Both of these questions require definitional clarity, of course, but for the purposes of this document – the intent of which is to give you, the volunteer analyst of a CARE evaluation or two – we will focus on the first question. What do we mean by women‘s empowerment? How are we defining it in the context of this year‘s SII research? Defining Women’s Empowerment: Agency, Structure, Relationships One of the key underlying causes of poverty is the construction in different contexts of what it means to be a man, or a woman. Gender is, in this sense, one manifestation of a general model of power which holds that individual and group behaviors produce social structures (ideologies, rules, institutions) which, in turn, reinforce and ―normalize‖ those behaviors to the point where they are seen as common sense, as the ―normal‖ order of things. Identities, roles, and relationships are, in this view of things, socially constructed, as are the constraints and opportunities that certain actors face regarding control of, access to, and use of tangible and intangible resources. Gendered forms of power come into play in the social construction of identities, roles, relationships and distribution of resources, all of which are intimately related to women‘s human rights and the question of poverty. These gendered ―rules of the game‖ are not always perfectly obvious to women and men who live by them but can be surfaced, discussed, and challenged through personal and collective consciousness and actions. In this way, women and men contest the flow of resources, agendas and ideologies. Empowerment has been theorized from many perspectives – including those founded in a more ―zero-sum‖ notion of power and those that take a more expansive notion of power. For the purpose of this study, we focus on those discussions of empowerment that take place within a feminist, gendered perspective. Empowerment is defined broadly as ―the expansion of assets and capabilities of poor people to participate in, negotiate with, influence, control, and hold accountable the institutions that affect their lives.‖5 Notable in this definition is the recognition of empowerment as a process of building capability (and not simply the material outcomes visible in CARE‘s impact frameworks to date), and of the importance of structure as represented by the institutions affecting people‘s lives. 4 There are minor differences between this summary and the June 2005 revision of the global research framework. The global framework is a work constantly in progress: as we learn, we modify our lenses/mental models and, indeed, our very theory upon which the global research framework rests. The full framework can be found on the CARE Portal (Divisions : Program : Program Resources & Learning : Impact Measurement and Learning). 5 Deepa Narayan ed, Empowerment and Poverty Reduction: A Sourcebook. Washington, DC: World Bank, 2001. This broad conception can be further grounded in a feminist theory as ―the expansion in people‘s ability to make strategic life choices in a context where this ability was previously denied to them.‖ 6 This definition is notable in its focus on choice, which Kabeer defines as comprising three critical elements: agency (power within/to), operationalized in reference to resources (power to/over), and made visible in its resulting beneficial/valued achievements. And finally, agency is exercised, in this conception of empowerment, in opposition to a prior condition of subordination in important (strategic) arenas of life. Strategic interests, in gender and development theory, differ from ―practical gender needs,‖ in that they go beyond the basic functions/capacities which allow people to fulfill the gender roles assigned to them, and aim to open new gendered spaces of ideology, action and opportunity. In this sense, empowerment is importantly tied to impact on the structural underpinnings of women‘s subordinate status and well-being. With this conceptualization of power and social change, empowerment should be conceived of as both process and outcome that comprises three dimensions—agency, structure, and relationships. These three dimensions are intimately related, structuring and influencing one another as the graphic below implies: Agency Structure Relations AGENCY: Carrying out our own analyses, making our own decisions, and taking our own actions. Every person has agency, every person analyses, decides, and acts. Agency is a continuum, from less to more. Empowerment involves a journey through which poor women increase their agency. STRUCTURE: Routines, patterns of relationships and interaction, and conventions that lead to taken-for- granted behavior; institutions that establish agreed-upon meanings, accepted (―normal‖) forms of domination (who ―naturally‖ has power over what or whom), and agreed criteria for legitimizing the social order. Individual agents both produce and are, in important ways, produced by structure. Structures can be both tangible and intangible; they are composed of both behavioral patterns that can be observed and counted but also the ideologies that underpin why some behaviors – or thoughts – are socially acceptable (acceptable to whom?). Examples include kinship, economic markets, religion, castes and other forms of social hierarchies, educational systems, political culture, resource control/ownership dynamics, forms of organization, and many, many more. RELATIONSHIPS: Both agency and structure are mediated through relationships between and among social actors while, at the same time, forms and patterns of relationships are deeply influenced – frequently in hidden ways – by agency and structure. Empowerment, in part, consists in individual women building relationships, joint efforts, coalitions, and mutual support, in order to claim and expand agency, alter inequitable structures, and so realize rights and livelihood security. Women’s Empowerment: Sub-Dimensions Women‘s empowerment differs from culture to culture and context to context. It cannot be applied uniformly across the developing world. In all field research sites this year – Bangladesh, Yemen, and Ecuador – one of the very first steps of impact research has been to uncover local women‘s own definitions and indicators of their empowerment. But this process has been informed by a conceptual framework that asks researchers to at least consider the relevance of 23 sub-dimensions of agency, structure, and 6 Naila Kabeer, ―The Conditions and Consequences of Choice: Reflections on the Measurement of Women‘s Empowerment.‖ UNRISD Discussion Paper 108. United Nations Research Institute for Social Development, Geneva, 1999. relationships. We selected these sub-dimensions because they have, in fact, been shown to be widely relevant to women‘s empowerment across a great many studies and across numerous social, economic, cultural, historical, and political contexts. In other words, a wide variety of studies have shown an apparent positive relationship between increases/improvements in the sub-dimensions and women‘s empowerment. In asking staff to at least consider these sub-dimensions we are not pre-determining local meanings of women‘s empowerment, nor the indicators that are most relevant to decide if CARE is having an impact or not, but rather trying to inform staff of important results that already are found in the rather wide literature on women‘s empowerment so that we don‘t reinvent the wheel at every site. These 23 sub-dimensions are briefly defined below: ELEMENT Specific Sub-Dimensions Definition of Sub-Dimension Self-Image; Self Esteem Positive images of self, belief in one‘s abilities, feelings of self- efficacy Legal & Rights Knowledge of laws around issues of women‘s social positions, Awareness status, equality, etc. Information & Skills Access to information and skills that a woman deems helpful or A necessary; awareness that such information/skills even exist Education Access to and ability to deploy formal and informal forms of G education Employment/Control of Fair and equitable access to employment opportunities; fair and E Own Labor equitable working conditions; freedom to chose forms of labor Mobility in Public Space Freedom to circulate in public spaces N Decision Influence in Kinds of decisions that women can make over household resources, C Household Group Membership & processes, people, investments, etc. This sub-dimension certainly overlaps with the element below, Y Activism Relationships. Here, at the level of agency, we are looking at the degree to which women are free to join groups as a result of their own wishes to do so Material Assets Owned The kinds of material assets (land, goods, animals, crops, money) women have the power to control Body Health & Bodily Access to core health services of acceptable quality; freedom to Integrity make decisions over what happens to a woman‘s own body; a right to bodily well being and pleasure Marriage & Kinship Degree of freedom and control of marital resources; equitable Rules, Norms, Processes inheritance, divorce, and family law more generally; control of S one‘s own body T Laws and Practices of Degree of inclusiveness and equity of laws and practices around Citizenship what it means to be a citizen R Information and Access Degree to which duty bearers ensure that women have the chance to U to Services know what they‘re due, how they can access this, and what to do in C the event that they are denied information or services Access to justice Enforceability of basic human rights as well as specially designed T (enforceability of rights) laws and programs to promote gender equity U Market Accessibility Equitable access to work, credit, inputs, fair prices R Political representation Extent of women elected and appointed to public office – in the formal and informal spheres – and their degree of influence once E there State budgeting practices Allocations the state offers for important services, guarantees, and enforcement mechanisms around issues central to gender equity Civil Society The density and quality of civil society organizations that address Representation gender inequity and social exclusion R Consciousness of Self and Social connections, outreach; seeing the value of joint actions both E Others as Interdependent for self but for a larger group Negotiation & Ability and interest in engaging duty bearers, the powerful, but also L Accommodation Habits other marginalized social actors in dialogue A Alliance & Coalition Extent to which women and women‘s groups form larger alliances T Habits and coalitions and seek collective gains Pursuit & Acceptance of Skills, confidence, and knowledge to hold duty bearers and the I Accountability powerful accountable; recognition that human rights bring, also, O forms of accountability to every individual N New Social Forms Generation of new kinds of organizing, new or altered relationships, S new kinds of behaviors. H I P S To summarize, we understand impact on women‘s empowerment to be reflected in three inter-connecting aspects of social change. The first, driven by the actor-centered notion of ―agency,‖ is in the aspirations, resources, capabilities, attitudes, and achievements of women themselves. The second is in the broader social structures that are both socially produced by people but that also, once produced and ―normalized,‖ condition women‘s choices and chances. And the third is in the character of the social relationships through which women negotiate their needs and rights with other social actors. The 23 sub-dimensions written above may or may not be important in a particular social context and the concrete indicators that would show improvement along one of the sub-dimensions may well differ from place to place, era to era in the same place, or even from group to group of women in the same place and time. Nonetheless, we are interested in whether and how CARE programs purporting to focus on gender and/or women‘s empowerment are targeting these sub-dimensions as they appear so frequently in the gender and women‘s empowerment literature.