What is women’s empowerment and how do we measure it?

Speakers: Bidisha Barooah, senior evaluation specialist, 3ie; Nandita Bhan, research scientist, Center on Gender Equity and Health, University of California, San Diego; and Shagun Sabarwal, associate director for policy and training, J-PAL South Asia
Chair: Beryl Leach, director and head, policy and advocacy, 3ie

Highlights from the seminar
While there are growing investments in rigorous evaluations of interventions designed to empower women and girls economically and promote gender equality, there remains limited evidence of effectiveness. Speakers discussed the challenges and how they had tried to address them in their work.

Bidisha Barooah (3ie) discussed findings on women empowerment from the upcoming evidence gap map on group-based livelihood programmes in L&MICs. They focus on advancing financial, human and social capital of beneficiaries. The gaps suggest that impact evaluations measuring the effectiveness of livelihood programmes tend to measure progress only economic empowerment outcomes. Despite the disproportionate focus, economic outcomes specifically related to women and young girls are not included by researchers in evaluation design. Of the 129 impact evaluations included in the analysis, only six examined changes in women’s economic well-being and welfare, and only a handful disaggregated data by sex. Given this context, she emphasised the need to require sex and age disaggregation, include gender analysis, and account for gendered power dynamics in evaluations, which could provide more useful and accurate insights on barriers and facilitators to achieving programme outcomes.

Shagun Sabharwal (J-PAL South Asia) exhorted the audience to reflect on what we lose out on by not taking gendered inequality into account when designing interventions and impact evaluations. She asserted that a gender-transformative and equity-focused approach necessitates that impact evaluations evolve from being gender-blind (no consideration for the drivers of gendered inequality) and gender-sensitive (usually meaning the evaluation only disaggregates by sex) to being gender-transformative (the evaluation explicitly considers gender in primary questions; uses gender analysis and reports out gendered differential impacts). To discern what we want to measure and how to measure it, there needs to be conceptual clarity in defining empowerment, and an understanding of the process of empowerment itself. She shared that at J-PAL, this definition has been predicated on Amartya Sen’s capabilities approach, as well as Naila Kabeer’s resources, agency and achievements framework. She highlighted challenges in measuring empowerment, such as the way it is defined in different contexts, not being able to observe complex decision-making processes directly, gendered household and community power dynamics and reporting biases. Devising a reliable strategy for measuring empowerment includes formative research to understand the context, mapping the theory of change to outcomes and indicators using a gender lens, developing and refining data instruments, and devising a data collection plan that minimises measurement error. She shared the example of a 3ie-supported study on a gender sensitisation programme in Haryana (the Breakthrough study) that used innovative gender norms and attitude measurement strategies, such as implicit association tests and vignettes, in addition to triangulating data from household, school and adolescent surveys.

According to Nandita Bhan (Center for Gender Equity and Health at the University of California, San Diego), to integrate gender equality in global health and development, standard measures of gender equality and empowerment need to be valid, reliable and widely accepted. For empowerment, some accepted existing measures are not valid, some measures do not exist yet and others are not valid over time. Research methods may not be valid, such as thinking that questions designed for adult women are appropriate for young women or girls. Another challenge is that researchers do not report sufficiently on measurement methodologies of women’s empowerment in academic publications. She introduced the Evidence-based Measures of Empowerment for Research on Gender Equality (EMERGE), an initiative focused on measuring gender equality and empowerment for development and health programme monitoring and evaluation.

Members of the audience noted that doing better research on empowerment requires commitment and funding from donors that does not exist. There was also a strong audience and panel sentiment for 3ie or some entity to organise regular group discussions on defining and measuring women’s and girls’ empowerment in all of its dimensions. Beryl Leach (3ie) concluded the discussion by agreeing that 3ie would follow up on the call for more seminars on this topic. She encouraged the audience to communicate to donors about what it takes to conduct gender-responsive research and evaluations.