Catalyzing new partnerships for gender equality in South Asia

“Do you agree with the statement, “Men are better suited than women to work outside the home”?”

Seema asked parents in India this question as part of a research study to understand how a school-based program designed to change attitudes about gender roles could help reduce gender bias in Haryana, India.

According to recent World Bank estimates, only 27 percent of women in India currently participate in the labor force—a steep ten percentage point decrease from 2005 rates. This large decrease is notable because, according to analysis by J-PAL affiliate Rohini Pande (Harvard) and Charity Troyer Moore, about one-third of women who work primarily in housework would like to have a job but seem to be prevented by traditional gender norms that restrict women’s mobility.

Bringing rigorous evidence to bear in understanding how gender norms affect everyone, in India and around the world, can help inform and improve policy to reduce poverty and gender inequality.

At J-PAL, the launch of a new Gender sector represents our commitment to further expand the base of policy-relevant evidence. 

The Gender sector, chaired by Seema, will address a range of issues important to citizens, policymakers, and practitioners around the world: How can we most effectively address gender disparities and inequality at scale? Are existing development programs closing the gender gap in human development? How do gender dynamics in families and society affect the impact of these programs? And how can we best measure changes in areas like agency and empowerment?

In addition to Seema's own research, J-PAL affiliated professors have investigated these and other related questions for many years, conducting more than 130 evaluations that focus on the impacts of programs and policies on women and girls or men and boys. Some examples include:

• Lori Beaman (Northwestern University) and co-author Andrew Dillon researched the consequences of targeting agricultural technology information to different parts of social networks among farmers in Mali. They found that, since different types of people are located in different parts of the network, the choice of targeting strategy may leave out the periphery, including women.

• Esther Duflo (MIT), Pascaline Dupas (Stanford) and Michael Kremer (Harvard) researched the impact of full, need-based scholarships on secondary school enrollment and the returns to secondary education in Ghana over more than ten years. Women who won full secondary school scholarships married later and delayed childbirth, particularly unwanted pregnancies, relative to women who did not win scholarships.

• Rohini Pande (Harvard), Erica Field (Duke), and Simone Schaner (Dartmouth) along with co-authors Natalia Rigol and Charity Troyer Moore, are researching the impact of offering women individual bank accounts on women’s financial control. Linking earnings from a government workfare program (that offered welfare payments in exchange for work) to women’s bank accounts led to increased employment. Women who had previously been the least involved in the labor market and whose husbands expressed the most opposition to women working experienced the largest impacts, suggesting that more control over earnings increased women's influence and agency.

With a new academic chair (Seema) and dedicated policy staff (Lucia), J-PAL’s initial sector efforts will focus on:

1. Identifying open research questions and working with funders, implementing organizations, and J-PAL affiliates to determine opportunities for new randomized evaluations with the potential to generate policy-relevant evidence that promotes gender equality.

While J-PAL’s current grantmaking initiatives already address many questions related to gender, we are exploring opportunities to develop a new grantmaking initiative that will generate new research in the sector, contribute to learning, and provide opportunities for partners and donors to collaborate on testing innovative new approaches.

2. Creating practical tools for policymakers, practitioners, and researchers to better incorporate analysis of gender dynamics and impacts into existing randomized evaluations.

To do this well, researchers need better tools for measuring concepts that are hard to quantify, like empowerment. Stay tuned for a forthcoming practical guide on measuring women’s empowerment in M&E and impact evaluation, co-authored by Rachel Glennerster, Claire Walsh, and Lucia.

3. Continuing J-PAL’s work in gender analysis and sharing these results as public resources on the J-PAL website.

We will focus on understanding the impact of policies and programs in all regions of the world that: 1) specifically seek to empower women and girls; 2) don’t necessarily have gender-specific goals but have different results by gender when impacts are disaggregated; and 3) have program impacts that are mediated by gender dynamics in society.

In the study of the attitude-change program in India, for example, Seema and co-authors found that the changes in attitudes were similar for boys and girls. Yet, behavior changes were larger among boys, perhaps because girls faced more barriers to acting on their changed attitudes. This highlights how, even when programs achieve promising results, social contexts might continue to constrain women and girls’ options and choices differently than those of men and boys.  

4. Producing topic-specific and cross-cutting policy lessons.

We already have policy lessons on topics like how to increase girls’ school enrollment and the impact of gender quotas on community decision-making. Moving forward, we will continue producing new policy lessons focused on gender equality. Broadly, this gender analysis can also help us understand the mechanisms behind how and why programs work—and for whom. This work, we hope, will inform future policies and programs. As in other J-PAL sectors, we will build partnerships with governments and implementing organizations to put this evidence into practice.  

Since women and girls make up half of the world’s population and face different preferences, constraints, and trade-offs compared to men and boys, J-PAL affiliated professors are continuing to prioritize understanding how development policies interact with gender. With the launch of the Gender sector, we will further this cross-cutting research and work to ensure that policies and programs are informed by evidence.

What are  open questions related to gender you’d like to see addressed in randomized evaluations? Let us know on Twitter (@JPAL_Global, @seema_econ) or via email at ldiazmartin[at]povertyactionlab[dot]org. 

As part of the formative research that my research team and I (Rachel) conducted for a randomized evaluation on adolescent girls’ empowerment in Bangladesh, we interviewed young women about their daily lives, aspirations, and what they wanted to do that they were prohibited from doing. We also interviewed local NGOs and community partners, asking: “What does an empowered girl do differently than a girl who is not empowered?”

Some of the most frequent answers were, “she can go where she likes,” “she can negotiate with her parents,” “she is not the last to eat at home,” and “she thinks differently about what girls can do.” Distilling these rich conversations into a set of indicators that could be measured in a survey questionnaire was not simple or straightforward.

For example, many people we spoke to said that empowered girls had the ability to travel in their communities. Yet, when we talked to girls, we quickly learned that a general question like “how far away from your home can you travel alone?” would not accurately capture how their mobility was constrained, because the answer depended on what they were doing and for whom. Adolescent girls could travel to and from school alone, but they could not travel alone to do things that only had value to them, like going to local fairs.

Since empowerment is about people’s ability to make choices that matter to them, we decided to survey young women about whether they could travel alone to a list of common locations and activities, including some that were for no one’s benefit but the girl herself. Yet, by using a question and indicator that was so locally tailored, we lost the ability to compare our findings with data on women’s mobility in other contexts.

Researchers face many such tradeoffs and challenges in measuring empowerment.

How can we measure women's power to make choices that matter to them when we rarely observe decision-making directly? How can we combine outcomes like education, health, or income with measures of agency like decision-making power or self-efficacy to better measure the process of empowerment? When should we use internationally standardized survey questions and when is it better to develop locally tailored ones? Can non-survey instruments pick up useful information that surveys can’t, and when should we think about using them?

J-PAL’s newest research resource, A Practical Guide to Measuring Women's and Girls’ Empowerment in Impact Evaluations, tackles these questions and more. Gathering insights from diverse disciplines and from the experiences of J-PAL affiliated researchers around the world, the guide offers practical tips for how to measure women's and girls’ empowerment in impact evaluations. It is designed to support the work of monitoring and evaluation practitioners, researchers, and students.

Throughout the guide, we emphasize the importance of conducting in-depth formative research to understand gender dynamics in the specific context before starting an evaluation, developing locally tailored indicators to complement internationally standardized ones, and reducing the potential for reporting bias in our instruments and data collection plan. Some highlights from the guide include:

• Seven key challenges that evaluators face in measuring women’s empowerment, along with tips for how to address them (pages 7-10).

• Five different approaches to measuring decision-making power, a key component of empowerment, when we can rarely observe decision-making directly (pages 21-24).

• Tips for conducting formative research before designing an evaluation to gain a deeper understanding of gender relations and the barriers that women face in a specific context (pages 13-17).

• Tips for pretesting and piloting survey and non-survey instruments to make sure they make sense to participants and pick up relevant information (pages 37-40).

• Tips for deciding who collects the data, when and where to interview people, and approaches and technologies that give participants options to answer privately (pages 41-44).

• A bank of hundreds of survey questions related to women’s empowerment that J-PAL affiliated researchers have used in the past (Appendix 1).

• A catalog of non-survey instruments that can be useful in measuring women’s empowerment and tips for when and how to use them (Appendix 2).

Recognizing that women's and girls’ aspirations and constraints are deeply tied to their specific context, the guide does not provide a single set of ready-to-go survey instruments for measuring empowerment. Instead, we outline a process for developing indicators appropriate to an individual study along with extensive examples.

In this way, we hope the guide can provide researchers and practitioners the tools to select or develop indicators of empowerment that are right for their impact evaluations.

Sustainable Development Goal 5 provides a rallying call to achieve equality and empowerment for all women and girls. To make progress towards this vision, we need a better understanding of effective policies to empower women and girls. And for that, we need to know how to measure empowerment well.

Access the full guide >>

See more of J-PAL's research resources and learn more about J-PAL’s work supporting gender analysis in impact evaluations.

Some text in this post is excerpted from A Practical Guide to Measuring Women's and Girls' Empowerment in Impact Evaluations. The views expressed here are those of the authors and do not necessarily reflect those of research funders or the UK Department for International Development.