AUGUST 2021 Measuring Women’s Goal Measures for Advancing Gender Equality (MAGNET) Setting and Decision-Making The Measures for Advancing Gender Equality (MAGNET) initiative aims to broaden and deepen the measurement of women’s agency, Improving women’s agency is crucial for advancing gender equality. Less than based on the development of new tools and half of women in Sub-Saharan Africa and South Asia participate in making rigorous testing and comparison of both new decisions over their own health care, major household purchases and visiting and existing methods for measuring agency, their families.1 Improving women’s intra-household bargaining power has also and promoting the adoption of these measures been shown to benefit economic development: women making decisions at scale. By increasing the availability of is linked to greater investments in female goods and wellbeing, as well as innovative meaningful measures of agency for a broad range of contexts, we hope our nutrition and human capital for children (Duflo 2003; Ashraf, Karlan, and Yin work will lead to an improved understanding 2010; Hou and Ma 2011; Doss 2013; Armand et al. 2020). In addition to of what women’s agency is, how it manifests strengthening women’s decision-making power, another priority for gender and how it can best be measured across equality is ensuring that women can enact well-defined goals that stem from contexts given the research question at hand. their own values and preferences. Existing evidence shows that women face MAGNET is a collaboration between the constraints in setting concrete goals and strategies for achieving them (e.g., World Bank’s Africa Gender Innovation Lab Johnson 2015) , and their motivation is less likely to stem from intrinsic goals and Living Standards Measurement Study compared to men (e.g., Vaz, Pratley, and Alkire 2016). (LSMS) teams, the International Food Policy Research Institute (IFPRI), the International Improving women’s ability to define goals and act on them is an important— Rescue Committee (IRC), and researchers and urgent—policy goal. Yet our understanding of how to achieve this goal at Oxford University. We plan to develop is hampered by the lack of adequate measurement tools and recognized a range of new survey tools, each tested best practices. First, existing measures of women’s agency tend to have across multiple contexts. MAGNET focuses on three dimensions of women’s agency that a narrow focus: standard survey questions do not capture the nuances have high potential for catalyzing progress of intra-household relationships or the constraints women face in defining on women’s economic empowerment, but and realizing their preferences—or how they relate to existing norms. for which the body of existing measurement Without knowing the different sources of and motivation for preferences, methods is weak or under-tested: (i) we are failing to measure agency in a way that is grounded in women’s ownership and control of assets, (ii) goal- values. And, even when measures capturing these broader dimensions setting and decision-making, and (iii) sense of control and efficacy. of women’s goal setting and decision-making do exist, they are often not 1 https://data.worldbank.org/indicator/SG.DMK.ALLD.FN.ZS http://www.worldbank.org/en/programs/africa-gender-innovation-lab validated in lower-income country contexts or adapted itself—over and above the final outcome—has important to the realities of women’s lives in these settings. This implications for households (Seymour and Peterman 2018; results in a fragmented understanding of women’s Annan et al. 2020; Ambler et al. 2021). Understanding agency, restricting the design of quality interventions and intra-household disagreements over decision-making the evaluation of their impact. requires first analyzing whether the process of decision- making is different for men and women, and why. Evidence Research is needed to broaden and deepen the shows that across many country studies, women report measurement of women’s goal setting and decision- joint decision-making more often than men (Acosta et making, both within and outside the household. This al. 2020; Ambler et al. 2021) but we know little about brief summarizes existing knowledge gaps in these two the reasons behind this, and how these dynamics may measurement areas and lays out how the Measures for vary across domains and contexts (Kishor and Subaiya Advancing Gender Equality (MAGNET) initiative plans to 2008; Seymour and Peterman 2018; Jarvis et al. 2020). tackle them. Do women and men have a different conceptualization of what a decision maker is or what joint decision-making KNOWLEDGE GAPS entails? Do they answer differently due to diverging beliefs on communal social norms or social desirability bias? What Understanding women’s control role does the local and cultural context play in gendered over decision-making interpretations of the decision-making process? And how Women’s decision-making power is most commonly do individual conflict strategies shape the intra-household measured by asking household members who usually decision-making process and decision outcomes (Rahim makes decisions over a standard set of domains (e.g., 2010)? Moreover, research has shown that whether—and large assets purchased, children’s education, health). how—spouses disagree on intra-household decision- However, interpreting being a decision maker as a proxy making arrangements matters for women’s and children’s of empowerment is only valid if the respondent desires outcomes (Annan et al. 2020; Ambler et al. 2021; Bussolo, to be involved in the decision. It is not hard to imagine Sarma, and Williams 2021). But the evidence base is situations in which not being a decision maker reflects thin and somewhat mixed on how these patterns vary an individual acting on their own desires. In addition, across contexts (Allendorf 2007; Story and Burgard 2012; recent work shows how small tweaks in decision-making Shakya et al. 2018). questions (Peterman et al. 2021) can substantially change Another knowledge gap is how best to capture whether our assessment of a women’s empowerment—as does women feel their opinions are valued—even when they accounting for individuals’ valuations over choices are not reported as main decision makers—and what (Maiorano et al. 2021). More research is needed to specific efforts they may undertake, both tacit and ex- understand in which instances individuals want to make plicit, to increase their power to shift decisions towards decisions, when women prefer to make decisions alone their preferred view. For example, Afzal et al. (2018) show instead of jointly with their partner or other household that women are more likely to demand increased agen- members, and what we can we conclude about women’s cy than men (measured as the willingness to pay for ex- agency when we observe them making certain choices. ecutive decision-making power over consumption de- For instance, when do women make decisions because cisions), but embedding such an experimental measure their husbands are unavailable or simply unwilling to in household surveys can be tricky. Moreover, women’s participate in burdensome decisions? When does women efforts to change or push back on decisions will also de- expressing a lack of interest in certain decisions reflect pend on the resistance and backlash that women would their preferences instead of internalized social norms? face if they tried (Angelucci and Heath 2020; Deschênes, Does the number of daily decisions a woman makes Dumas, and Lambert 2020). It is crucial to explore the result in bandwidth depletion and influence her desire to prevalent ways other individuals use to restrain wom- be less involved in other decisions? en’s power, including coercion (silencing, cutting access Evidence also suggests that wives and husbands have to information, violence) and discrediting (defamation, systematically different perceptions of who makes these misrepresentation). More work is needed to understand decisions and that the process of decision-making when, by whom, and the type of backlash (economic, physical, emotional) women face when they engage in to seasonal bias induced by changes in the respondent’s decision-making. In addition, there has been a strong fo- cognitive bandwidth, which may be relevant for women cus on discrepancies within the couple, but discrepancies in low-income settings. For instance, recent work in with other family members may also be worth exploring in development has documented strong impacts of poverty certain settings (Gram et al 2018; Akter and Francis-Tan on cognitive abilities (Mani 2013;  World Bank 2015; Li, 2020; Khanna and Pandey 2021; Gupta, Ksoll, and Mae- Yang, and Luo 2021).  rtens 2021). Understanding women’s goals and MAGNET WORKPLAN goals-setting capacity MAGNET will develop new measurement tools to answer these questions, including a new goal-setting capacity An individual’s ability to define goals that are in line with her values is a crucial component of agency. Without questionnaire, novel tools to measure agency (automatic knowing what an individual’s own goals and preferences cognition test and text analysis), a new measure of are, it is difficult to understand and measure agency, as motivational autonomy, and new decision-making observed choices may be consistent with multiple sets questions that more closely reflect the reality of women’s of expectations and preferences (Manski 2004; Donald lives: unpacking how they make decisions, influence them, et al. 2020). renegotiate them and value them. Most previous work exploring whether individuals’ actions are guided by their own values (motivational autonomy) Decision-making questions: derives from the Self-Determination Theory developed by MAGNET will generate a range of new decision-making psychology scholars. The theory classifies human actions survey questions, suitable for inclusion in large-scale as driven by internal (autonomous), external (coerced), household surveys. These will include: and introjected (internalized social pressures and norms) • A survey module to explore whether women are motivations (Ryan and Deci 2000; Vaz, Pratley, and consulted within their existing intra-household Alkire 2016). Existing survey tools, such as the Relative decision-making arrangements, whether they feel Autonomy Index, perform well regarding the distinction their opinion is valued—especially in cases where between external coercion and independent motivation. they do not describe themselves as the main However, they do not perform as well in capturing to decision maker—and how they decide which decisions are important to them (e.g., monetary what extent goals are based on internalized social norms. value, gendered preferences). More research is needed on the best way to capture the formation of women’s preferences and whether they are • Decision-making questions to capture the ‘resist’ driven by internalized norms: is it through tweaking existing and ‘backlash’ dimensions of women’s agency, allowing researchers and practitioners to answer motivational autonomy questions, or are tools to measure questions such as: what type of retaliation do automatic cognition developed within psychology and women face when participating in decision other cognitive sciences more reliable? making? Does backlash depend on preferences alignment with other decision makers? What It also remains unexplored to what extent standardized type of resistance women expect to face if they goal-setting questionnaires used in psychology studies, engaged in decision making? Who do women fear which have been validated and shown to be strongly resistance from? related to well-being outcomes (Donald et al. 2020), can • A module—focused on water use and allocation— be streamlined and adapted to measure goal-setting to measure interpersonal conflict strategies (e.g., capacity in development settings. Existing tools are avoidance, dominance, compromise), shedding industry-specific and meant for use in formal employment light on the dynamics that emerge in the context settings (Locke and Latham 1979; Lee et al. 1991): there of intra-household disagreement. is a need for tools that are more broadly applicable across • Decision-making questions to capture how domains and countries. An important aspect is also the different household arrangements (e.g., whether consistency of types of tools that measure constructs that women do all of half or half of all decisions) matter should be stable over time, since  answers may be subject for women’s well-being and household outcomes. Anchoring vignettes: Quantitative narrative analysis: MAGNET will develop and test a range of new anchoring MAGNET will propose a new text-analysis measure of vignettes (short descriptions of hypothetical individuals or women’s agency, drawing on text-to-speech transcripts situations meant to convey complicated concepts and of women’s answers to open-ended questions. This ensure that different respondents understand questions method will adapt the Franzosi (1994, 2004) methodology similarly across cultures and contexts). These vignettes will allow for the analysis of differences between spouses for analyzing semi-structured interviews, attributing an in the perception and interpretation of what joint decision- automated agency score to women’s narratives about making entails and how women’s actions are determined how they make decisions, how they live their life and how by social norms. this compares to their preferences. Other survey tools: Experiments and games: Drawing from psychology, sociology and other disciplines, MAGNET will field behavioral exercises to capture the MAGNET will create and validate a range of other tools malleability of women’s preferences and to what extent to deepen our measurement of women’s goal-setting and they are conditioned by social norms, among other decision-making. These include: dimensions of agency. • New bandwidth depletion measures to understand whether cognitive depletion—linked to multitasking In addition to the development and testing of these and many small daily decisions—affects women’s measurement tools, MAGNET will conduct mixed- desire to engage in other important economic choices. methods research and triangulate across these methods to uncover new layers in our understanding of women’s • A novel goal-setting tool, drawing on standardized goal-setting questionnaires used in psychology agency. For example, we plan to triangulate answers studies. from decision-making questions with outcomes of lab-in- • An automatic cognition test to capture the extent the-field experiments, direct observation, and automatic to which women have internalized gender norms cognition tests to understand what each measurement around specific activities, and sectoral and economic choices (e.g., that is more appropriate for women to approach is capturing and how they can complement work in less lucrative and competitive sectors). each other. Photo Credit: Vincent Tremeau, World Bank This work has been funded in part by the Umbrella Facility for Gender Equality (UFGE), which is a multi-donor trust fund administered by the World Bank to advance gender equality and women’s empowerment through experimentation and knowledge creation to help governments and the private sector focus policy and programs on scalable solutions with sustainable outcomes. The UFGE is supported with generous contributions from Australia, Canada, Denmark, Germany, Iceland, Latvia, the Netherlands, Norway, Spain, Sweden, Switzerland, United Kingdom, United States, and the Bill and Melinda Gates Foundation. To learn more about our partners, please visit Or contact World Bank Africa Gender Innovation Lab International Food Policy Research Institute Aletheia Donald (AFRGIL) (IFPRI) adonald@worldbank.org https://www.worldbank.org/en/ https://www.ifpri.org/ programs/africa-gender-innovation-lab Maria Hernandez-De-Benito International Rescue Committee (IRC) mhernandezdebeni@worldbank.org World Bank Living Standards Measurement https://www.rescue.org/ Study (LSMS) https://www.worldbank.org/ Oxford University en/programs/lsms https://www.ox.ac.uk/