Measuring agency beyond direct decision-making

Measures of agency and empowerment are often operationalized through questions regarding decision-making.  Survey items may ask respondents whether they participate in, influence, or have the final say in household decisions. This approach focuses on the involvement in the decision regardless of whether the desired outcome is achieved. Yet, this standard approach to measuring agency may miss three key elements:

1. Choosing Not to Decide:  Individuals may exercise agency by opting not to participate in decision-making processes, particularly if their preferred outcomes are still achieved.
2. Reflecting Others' Needs: Decisions made by individuals, especially women, may reflect the needs and preferences of others rather than their own goals and values.
3. Complex Decision-making in Extended Households: Decision-making processes vary in extended family households compared to couple households or households with co-wives. Standard approaches typically focus on the primary couple, excluding other members of the household, particularly youth and the elderly. 

In this recently published working paper, we use unique mixed methods data on decision-making within rural households in Kilifi County, Kenya, to explore these three issues. Quantitative data was collected through a survey administered in early 2023, involving 613 households with 1,059 women and 808 men. Qualitative data from case studies conducted in October 2022 included key informant interviews, focus groups, and in-depth interviews.

Findings on Effective Power

Many people report not being involved in household decisions. Only 13 percent of men and 33 percent of women are directly involved in making decisions for all seven activities surveyed. Figure 1 presents the distribution of activities for which respondents are not directly involved in the decision, by number of activities in which they are not directly involved.  Women are less likely to participate in household expenditure decisions, while men are less involved in water-related decisions, except for water expenditures. 

Despite this, both women and men may have agency even when not directly involved in decision-making.  We define two types of such effective power: 

• Effective power by proxy is when individuals choose not to be directly involved in the decision, motivated by the desire to free up time or mental space, or a lack of interest in the process and outcome.  They may delegate the decision or trust it will meet their preferred outcomes.
• Effective power through influence or persuasion is when individuals influence the decision-makers’ minds, even without direct control. This may occur because the decision-makers do not allow them to participate or because the individual wants to avoid negative consequences from direct involvement.  

Figure 2 shows the percentage of those who do not participate directly in the decision but have effective power.  The patterns of effective power vary across decision types. Women are more likely to have effective power through influence or persuasion, while men can have effective power in either form, depending on their position in the household.  Overall, men are more likely than women to have effective power in many of the decisions when not a decision maker.  

This is most striking for water-related decisions, particularly with regards to the water source choice and water allocation in the household for which women are more likely to be the direct decision-makers. While men are less likely to be directly engaged in these decisions, many have effective power, especially effective power by proxy. This leads us to the second point.
 

Reflecting Others' Needs

Direct participation in decision-making does not always reflect agency.  For example, although women are making the decisions about the water source choice and allocation in a procedural sense, they may not have agency over these decisions.  We see above that although men are not directly involved in those decisions, a substantial share have effective power by proxy.  Correspondingly, the qualitative data indicates that while women are making the direct decisions about the allocation and use of water for the household, they often prioritize their husband’s and father-in-law’s needs over their own.
 

Beyond the Principal Couple

Our third point is that focusing on the principal couple misses two sets of important dynamics.  First, the decision-making between couples varies depending on who else is in the household.  In households with only the husband and wife, both are directly involved in most decisions. However, husbands are more likely to delegate or influence decisions indirectly. In households with the husband's parents, both spouses are less directly involved in household decisions, but husbands have more effective power.  In polygamous households, husbands are more likely to delegate or influence decisions indirectly, while wives are more directly involved but have less effective power when not involved.

Finally, a focus on the principal couple misses other adults in the household.  For example, we see that young unmarried adults living with their parents are the least likely to be involved in decisions directly and have relatively little effective power.
 

Measuring agency to inform the gender equality agenda

Our conceptualization of effective power recognizes methods beyond direct decision-making to obtain one’s goals. Measuring effective power when not directly involved in the decision-making process is crucial for comprehensively assessing agency, as it captures forms of power that are otherwise invisible.

Our findings yield two recommendations for measuring decision-making:

1. Identify Household Members Involved. Surveys should capture the complexities of decision-making structures in extended households by identifying all household members involved in decisions.
2. Ask Further Questions to Those Not Directly Engaged:  When individuals do not directly engage in decision-making, further questions should be asked to understand why they are not participating and whether the outcomes reflect their preferences. This approach provides a richer understanding of people’s agency.   

By implementing these recommendations, we obtain a more nuanced understanding of agency within households that can inform policies related to gender equality and empowerment.
 

This study is part of the Measures for Advancing Gender Equality (MAGNET) initiative, which aims to (i) broaden and deepen the measurement of women’s agency, based on the development of new tools and rigorous testing and comparing both new and existing measurement approaches, and (ii) promote the widespread adoption of these measures at scale.