Measuring agency: what we know and where we go from here

Good gender data is essential for tracking our progress in promoting gender equality. But how do we collect this data in a reliable way? 

Over the last decade, various initiatives have been developed to expand the collection of gender data—both within the World Bank (such as LSMS+) and outside the World Bank (like Data2x). Despite this progress, the best ways to measure many dimensions of women’s agency remain understudied. Existing measures of agency are frequently validated only for one context or population (often, a higher-income one) and lack best practices for survey implementation. This makes it difficult to measure agency precisely in the countries where this data is most urgently needed. In addition, commonly used measures—such as the ones that can more easily be embedded in large-scale country surveys— do not fully capture the multi-faceted nature of women’s agency. 

To tackle this, with colleagues from the new Measures for Advancing Gender Equality (MAGNET) initiative, we recently published a series of policy briefs. The briefs review the gaps in measuring women’s control over assets, their goal setting and decision making, and their sense of control and efficacy—and offer an action plan to bridge these gaps. The three dimensions of women’s empowerment we focus on are known both for their centrality in the policy debate on gender equality and for the challenges posed by their measurement.

 

Control over assets

Women’s capacity to exercise and expand their agency depends on the economic resources they can use, control, and own.  Across low-income settings, women largely have less access than men to physical and productive assets including land, housing, and financial assets. They also have less tenure security and power to influence decisions such as the sale or the economic use of the assets they own.

In recent years, researchers and policymakers have made remarkable progress addressing many of the challenges in collecting gendered data on assets, but knowledge gaps persist. For instance, women frequently report owning assets jointly with their spouses or other household members, but we still know little about what “jointness” means across many contexts or to what extent joint rights imply equal rights. We also do not know much about how this affects the estimation of gender gaps and intra-household inequality on wealth and asset rights— which is key, for instance, for poverty-targeting interventions. More research is also needed to understand how the survey context may introduce measurement bias. A recent paper shows that girls are less likely to report sexual activity to interviewers that hold more discriminatory gender attitudes. Could enumerator attitudes also influence asset data collection? To fill this knowledge gap, we need field experiments to assess how enumerators’ characteristics, including gender attitudes, may affect data on women’s asset rights and how data collection protocols should account for this.

 

Goal setting and decision making

For individuals to have meaningful agency in their lives, it is important that they reflect on and develop well-defined goals, which stem from their own values and preferences. But how can we best measure this ability?

Interpreting an action as an exercise of agency requires understanding the driving motivation behind it. We need new tools to understand when and how women’s actions are guided by their own values or influenced by what others demand or expect from them. How does mental bandwidth depletion—due to women’s greater care burdens, for example—affect their self-reflection on goals and decisions? 

Women’s ability to act on their goals by exercising their decision-making power is also a central component of their agency. It is most commonly measured by asking household members who make decisions over a standard set of domains (e.g., large assets purchased, children’s education). However, interpreting being a decision-maker as a proxy of empowerment is only valid if the respondent desires to be involved in such decisions. We need new decision-making questions to explore whether women are consulted and whether they feel their opinions are valued. Women could also face coercion, retaliation, and backlash, or fear when engaging in decision making so we also need to develop new tools to understand what happens to the decision-making process when there is disagreement. Lastly, pairing speech-to-text technology with text-analysis methods provides exciting opportunities to directly assess women’s agency in their own words, reducing reliance on costly transcription of manual coding of semi-structured interviews.

 

Women’s sense of control and efficacy

Measuring agency also requires understanding to what degree individuals believe they can purposely achieve their goals. A common measure in psychology is self-efficacy: the belief in one’s abilities to produce actions and achieve a goal.  Self-efficacy can be conceptualized as domain-specific or as a general personality trait. In low-income settings, domain-specific scales have focused on entrepreneurship and health displaying promising correlations with female development outcomes. Though most of the global poor earn a living from agriculture, to our knowledge there are no validated self-efficacy scales specific to agriculture in low-income countries. In addition, we need a new livelihoods self-efficacy scale, a measure that is applicable across economic activities and can be easily incorporated in household surveys.

Women’s sense of control over their lives is inexorably linked to their control over their time. But we have much more to learn about how the interplay between women’s preferences, social pressure, and internalized social norms shapes how women spend their days. A better understanding of why individuals spend time the way they do might also shed light on whether some development programs lead to unintended consequences (like decreasing women’s leisure and/or increasing their stress levels). In addition, we always worry we are not capturing well the extra burden that women bear worrying about household needs. We need better measures of the gendered patterns of cognitive labor—anticipating needs, identifying options for meeting those needs, deciding among the options, and monitoring the results—and its relationship to agency.

 

---

New Measures for Advancing Gender Equality

Tackling such a broad and complex measurement agenda requires strong a cross-institutional partnership. The Africa Gender Innovation Lab (GIL) and LSMS teams at the World Bank, IFPRI, the IRC, and researchers at Oxford University have banded together to form the Measures for Advancing Gender Equality (MAGNET) initiative. Over the coming years, we aim to broaden and deepen the measurement of women’s agency through the development of new measurement tools, rigorous testing of both new and existing methods for measuring agency across different contexts, and promoting the adoption of these measures at scale. We would love to receive your insights and suggestions, as well as hear from you on potential collaboration on the development and testing of the tools. Please check the new MAGNET page for upcoming updates!