Women’s Economic Empowerment in Bangladesh An Evidence-Guided Toolkit for More Inclusive Policies Women’s Economic Empowerment in Bangladesh An Evidence-Guided Toolkit for More Inclusive Policies ii Table of Contents Acknowledgments iv Abbreviations and Acronyms vi Executive Summary 1 What’s New in This Report? 2 Key Messages 4 Linking Policy Priorities and the Evidence on Policy Tools 11 Introduction 13 Data Sources 17 Data Gaps 20 Key Definitions 23 Part 1 Women in the Labor Force 26 1.1 Overview 27 1.2 Gender Sectoral Segregation, Status in Employment, and Occupation 30 1.3 Trends in Women’s Wage Employment 34 1.4 The Cohort Dynamics of Women’s Wage Employment 37 1.5 Hours Worked 40 1.6 Earnings 41 1.7 The Sociodemographic Profile of Female Labor Market Participation 43 1.8 Social Norms and Women in the Labor Force 50 Part 2 Deep Dives 52 2.1 Women in Agriculture 53 2.2 Women in the Ready-Made Garment Sector 57 2.3 Women in Nonagricultural Self-Employment 61 iii Part 3 Constraints and Policy Tools 65 3.1 Overview 66 3.2 Skills and Vocational Training 69 3.3 Transport, Mobility, and Safety 74 3.4 Financial Inclusion 81 3.5 Ownership of Property and Other Assets 87 3.6 Childcare and Home Responsibilities 93 3.7 Business Ownership and Business Growth 99 3.8 Job Search 104 3.9 Digital Inclusion 109 3.10 Migration 113 Appendix A. Summary Statistics: Key Measures of Women’s Economic Empowerment 119 Appendix B. Technical Appendix 121 B.1 Aggregation of Activity Codes in the Time Use Survey 2021 121 B.2 Oaxaca Decomposition 121 B.2.1 Decomposition Results 121 B.2.2 Endowments, by Sex 122 B.3 Evidence Assessment Methodology and Expert Survey 123 B.3.1 Policy Priorities: Assessment Methodology 123 B.3.2 World Bank Expert Survey and the Ranking of Policy Priorities 124 B.4 Measurement Changes in Labor Force Survey 2022 125 B.4.1 Employment Trends among Women in the Bangladesh LFS 125 B.4.2 Causes of the Rise in Women’s Agricultural Self-Employment in LFS 2022 127 B.4.3 Conclusion 130 B.5. Changes in the ILO Definition of Employment 130 References 134 iv Acknowledgments This report was prepared by a team of authors from the South Asia Gender Innovation Lab (SAR GIL), led by Kate Vyborny and Pulkit Aggarwal, with major contributions from Sofia Amaral, Isis Gaddis, Shirleen Manzur, Isabela Salgado-Silva Pereira, Kendal Swanson, and Viet Tran. The team appreciates additional contributions from Marian Abdel Nour, Raffaella Dimastrochicco, Anusha Guha, Sayan Kundu, Diana Lopez-Avila, Osama Safeer, and Hijab Waheed. We appreciate expert insights from Bidisha Haque, Fahmida Khatun, and Shireen Haque Parvin. The report was developed with the support of the Bangladesh and Bhutan Gender Platform of the Social Development Global Practice under the leadership of Robin Mearns, Anna O’Donnell, Patricia Fernandes, Audrey Sacks, Sabah Moyeen, and Erisha Singh Suwal. v The team thanks Robert Zimmermann for editing services and Vito Raimondi for the design of the report. The authors would like to thank Sergio Daniel Olivieri, Margaret Maggie Triyana, and Abhilasha Sahay, who peer reviewed the report, as well as Cara Myers who shared detailed feedback on the draft. We gratefully acknowledge the support from the Umbrella Facility for Gender Equality, in partnership with the Gates Foundation. vi Abbreviations and Acronyms BIHS Bangladesh Integrated Household Survey DHS Demographic and Health Surveys GDP Gross Domestic Product HIES Household Income and Expenditure Survey ICLS International Conference of Labour Statisticians (ILO) ILO International Labour Organization IPV Intimate Partner Violence LFS Labor Force Survey NEET Not in Employment, Education, or Training NGO Nongovernmental Organization RMG Ready-Made Garment SAR GIL South Asia Gender Innovation Lab TVET Technical and Vocational Education and Training Credits: KM Asad / World Bank Executive Summary This report provides an examination of gender disparities in economic empowerment in Bangladesh. It aims to shed light on the current state of gender inequality by analyzing rich microdata from multiple data sources, thereby fostering insights into women’s labor market outcomes (such as employment, earnings, hours worked) and factors affecting women’s economic empowerment (such as safety, mobility, asset ownership, digital inclusion, and financial inclusion). Based on the descriptive analysis of gender gaps, the report also identifies key constraints on women’s economic empowerment in Bangladesh and presents an evidence- guided policy toolkit to inform policies that promote gender equity and support inclusive, sustainable development in Bangladesh. Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 2 What’s New in This Report? This report builds on previous analytical work conducted by the World Bank.1 It expands the analysis by drawing on the latest microdata to paint the most up-to-date picture of the state of women’s economic empowerment in Bangladesh. Most of the analysis in this report uses data produced since the Covid-19 pandemic. The resulting insights have been used to inform an evidence-based toolkit of policy instruments to guide policymakers and practitioners in developing and implementing programs that prioritize women’s economic empowerment. The report’s novel contributions are as follows: 1 Analysis of factors affecting economic empowerment In addition to measures of labor market outcomes, the report also includes an analysis of factors that constrain (and, if addressed, have the potential to boost) women’s economic empowerment, such as digital and financial inclusion, asset ownership, safety, mobility, and norms. 2 Use of rich microdata The report relies on richer and more recent rounds of microdata from 13 data sources to examine a broad range of outcomes. The combination of various kinds of data—on labor market outcomes, asset ownership, time use, safety, mobility, migration, ready-made garment (RMG) factories, and norms—paints a more complete picture of the state of women’s economic empowerment in Bangladesh, along with the related drivers and constraints. 3 Identification of data gaps The analysis of women’s economic empowerment is limited by the lack of nuanced data on various dimensions of economic empowerment. Specific data gaps that, if narrowed, can improve our understanding are highlighted. 1 This includes Farole et al. (2017); Moyeen et al. (2022); Solotaroff et al. (2019). Executive Summary 3 4 Evidence-based policy toolkit for practitioners The report presents a policy toolkit to address key constraints on women’s economic empowerment. The toolkit is based on a comprehensive review of high-quality global evidence on effective policy measures to address the gaps discussed in the report. It also includes an assessment of the available evidence from Bangladesh and South Asia, identifying areas where high-quality evidence is lacking and pointing to future priorities in research. To support decision-making, the tools have been prioritized based on their potential to enhance women’s economic empowerment in Bangladesh. This ranking has been developed using aggregated responses from a survey of World Bank experts working with Bangladesh that was conducted within the framework of this study. Credits: Mumtahina Tanni / pexels Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 4 Key Messages Key Message 1 The gender gap in labor force participation is large. The magnitude of the gap varies based on the definition of employment. As of 2022, using an inclusive statistical definition of labor force participation, the definition of the 13th International Conference of Labour Statisticians (ICLS) (ILO 1983), shows that, while 83 percent of men ages 15–64 participate in the labor force, only 45 percent of women do so. This implies that 31.0 million working-age women in Bangladesh were still out of the labor force in 2022, compared with 9.3 million working-age men. Using a more recently updated statistical definition of labor force participation (that of the 19th ICLS), 82 percent of working-age men, but only 22 percent of working- age women participate in the labor force (ILO 2013). The lower rates of labor force participation, especially among women, under this definition compared with the older definition are due to the exclusion from 2022 the definition of employment of own-use work, such as subsistence farming and helping unpaid in a household business, work that 2005 8% 6% women are significantly more likely to be involved in than men. Appendix B, section B.5, presents a more detailed comparison of the two definitions. While the rate of female labor force participation in Bangladesh is low, it still outpaces the rates in most other countries in the region. It is difficult to assess trends in women’s employment overall because of inconsistent measurement over time; but women’s wage employment only marginally increased between 2005 and 2022, increasing by 2 percentage points to 8 percent. Key Message 2 Although women’s work is concentrated in agriculture, women do not own agricultural assets. Only 29 percent of the women in agriculture are engaged in crop-related activities, compared with 82 percent of men. Women are mostly engaged in rearing of livestock, including large livestock that they do not own. Women spend more time than men on rearing large livestock. The ownership of productive assets, such as land and large livestock, is concentrated among men. 82% 29% MEN WOMEN Executive Summary 5 Key Message 3 Employed women work significantly fewer hours on average per week relative to employed men (a difference of 17 hours). A quarter of women work part time (20 hours or less) (refer to figure ES.1). This is observed across all sectors and employment statuses, suggesting that women’s work choices are often limited by the time they spend on home responsibilities. Women often engage in work that allows for shorter and more flexible hours. Figure ES.1: Hours worked in the past 7 100% days, employed individuals ages 25-64, by sex 80% Percentage (%) 60% 40% 35 27 25 20% 17 18 16 16 13 10 8 Source: Original figure for this publication 6 6 based on calculations using data of LFS 1 2 0 0 (Labor Force Survey) 2022, Bangladesh 0% Bureau of Statistics, Dhaka, Bangladesh, accessed through SARRAW (South Asia 0 1-10 11-20 21-30 31-40 41-50 51-60 61+ Raw) (database), South Asia Region Team for Statistical Development, World Bank, Hours Worked in the Past 7 days Washington, DC. Note: The figure shows hours worked by Female Male employed individuals. The share of women who are employed is larger in rural areas than in urban areas (50 percent and 25 percent, respectively), but rural women work fewer hours than urban women (29 hours vs. 41 hours, respectively). Key Message 4 –41% Women’s earnings are significantly lower than men’s earnings. On average, women earn Tk 9 thousand per month, or 41 percent less than men. This gap arises in part because women work fewer hours, but, after accounting for hours worked, women’s earnings are still 26 percent lower. The gender earnings gap is starkest in industry, where women and men work MEN WOMEN similar hours, but women earn 38 percent less per hour. Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 6 Key Message 5 Wealthier women are more likely to be more highly educated, yet less likely to be working. Those women who do work are often engaged in sectors with higher wages or better working conditions. These patterns suggest that most female employment is driven by necessity rather than choice. The relationship between education and labor force participation varies between rural and urban areas. In rural areas, increased education is associated with higher female labor force participation and a shift from agricultural and elementary occupations to professional roles in the nonagricultural sector, such as teaching and health care (refer to figure ES.2). In urban areas, female labor force participation follows a U-shaped pattern: it is highest among women with lower or higher educational attainment. This indicates that urban women with moderate educational attainment tend to leave the workforce, while more educated women work in higher status white-collar jobs. Figure ES.2: Labor force a. Rural areas participation rate, by education, region, and sex, ages 25-64 100% 95 95 93 93 92 80% 74 69 59 Percent 58 60% 47 Female 40% Male 20% 0% No Education & Primary SSC HSC HSC+ / Primary Incomplete Complete (Class 10) (Class 12) Any Tertiary b. Urban areas Source: Original figure for this publication based on calculations using data of LFS (Labor Force Survey) 2022, Bangladesh Bureau of Statistics, Dhaka, Bangladesh, accessed through SARRAW (South Asia 100% 95 95 92 91 93 Raw) (database), South Asia Region Team for Statistical Development, World Bank, 80% Washington, DC. Percent 60% Note: Labor force participation rate is 40 calculated as the labor force (employed 40% 32 24 and unemployed), divided by the 20% 14 15 total working-age population (ages 15–64). The Key Definitions sub-section in the Introduction provides a detailed 0% description of the definitions of employed No Education & Primary SSC HSC HSC+ / and unemployed individuals. The category Primary Incomplete Complete (Class 10) (Class 12) Any Tertiary ‘No Education & Primary Incomplete’ includes individuals without any education and individuals who have passed class 4 or below. The category ‘Primary Complete’ includes individuals who have passed class 5 through class 9. Executive Summary 7 Key Message 6 The ready made garments (RMG) sector is an important source of wage employment among young women in both urban and rural areas, but women do not advance to higher-level positions and higher pay with age in the sector. Furthermore, each successive birth cohort of women has been exiting the sector at an earlier age. Unlike previous cohorts, fewer young women in the youngest cohort in the data (born in the 2000s) entered the sector relative to the previous cohort, suggesting that female employment in the RMG sector may be stalling. The sector also contracted more broadly in 2024 with multiple factories shutting down and more than half a million people, mostly women, losing their jobs. This contraction is due to both global factors and domestic challenges. Key Message 7 Most employed women are self-employed (64 percent) (refer to figure ES.3). This is in contrast to employed men, who are almost equally likely to be wage employees or self-employed (48 percent and 49 percent, respectively). However, women in nonagricultural self-employment operate informal and home-based businesses. Fewer than a third of nonagricultural self-employed women in urban areas and fewer than a fifth in rural areas work outside the home, compared with almost all self-employed men. Women's businesses also tend to operate part time and earn less than men’s businesses (refer to figure ES.4). Figure ES.3: Employment status among those employed, by sex, ages 25-64 100 80 Percentage (%) 60 Self-Employed Unpaid Employee Wage Employee 40 20 Source: Original figure for this publication based on calculations using data of LFS 0 (Labor Force Survey) 2022, Bangladesh Female Male Bureau of Statistics, Dhaka, Bangladesh, accessed through SARRAW (South Asia Raw) (database), South Asia Region Team for Statistical Development, World Bank, Washington, DC. Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 8 Figure ES.4: Hours and hourly a. Average hours worked earnings of individuals self-employed outside agriculture, by region and 60 55 sex, ages 15-64 52 Avg Hours in Past 7 Days 50 40 31 Female 29 30 Male 20 10 0 Sources: Panel A: Original figure for this Rural Urban publication based on calculations using data of LFS (Labor Force Survey) 2022, Rural / Urban Bangladesh Bureau of Statistics, Dhaka, Bangladesh, accessed through SARRAW (South Asia Raw) (database), South Asia b. Average hourly earnings Region Team for Statistical Development, World Bank, Washington, DC. 150 Panel B: Original figure for this publication based on calculations using 119 data of HIES (Household Income and Hourly Earnings (in Taka) Expenditure Survey) 2022, Bangladesh 89 Bureau of Statistics, Dhaka, Bangladesh, 100 accessed through SARRAW (South Asia Raw) and SARMD (South Asia Regional 78 Micro Database) (databases), South Asia Region Team for Statistical Development, 48 World Bank, Washington, DC. 50 Note: In Panel B, earnings have been adjusted to 2024 prices using inflation data of the Bangladesh Bureau of Statistics, compiled by the World 0 Bank and accessed through the World Bank Open Data portal. Earnings are Rural Urban winsorized at 95th percentile. Key Message 8 In 2022, 19 percent of young women ages 15–24 in Bangladesh were not in employment, education, or training (NEET), nearly double the rate among young men (10 percent) (refer to figure ES.5). The gap was especially wide in urban areas, where 42 percent of young women were NEET compared with only 8 percent in rural areas, suggesting that many young women complete education or training, but do not enter the workforce. Figure ES.5: Distribution 100 of the population in employment, education or training, and NEET, by sex, 80 ages 15-24 Percentage (%) 60 NEET In Education or Training In Employment 40 Source: Original figure for this publication based on calculations using data of LFS (Labor Force Survey) 2022, Bangladesh 20 Bureau of Statistics, Dhaka, Bangladesh, accessed through SARRAW (South Asia Raw) (database), South Asia Region Team for Statistical Development, World 0 Female Male Bank, Washington, DC. Executive Summary 9 Among women, the school-to-work transition in Bangladesh is a challenge. A significant mismatch exists between the skills of graduates and the skills required by the labor market (World Bank 2024a). A survey of World Bank experts ranked the lack of technical and vocational skills as the most important constraint to women’s economic empowerment in Bangladesh. While technical and vocational education and training (TVET) programs are a common policy tool in addressing the skills gap, such programs show mixed impacts on women’s labor market outcomes, and high-quality evidence on Bangladesh is limited. Integrating features that address specific barriers faced by women, such as lower baseline confidence and limited mobility, into TVET programs would likely have high returns in women’s economic empowerment in Bangladesh. Key Message 9 64% Lack of mobility and concerns about safety in transport, public spaces, and the workplace—the second most important constraint on women’s empowerment according to the World Bank expert survey and the top constraint identified by Bangladeshi expert interviewees on women's economic empowerment—limit women’s employment. Among women working in agroindustries in Bangladesh, 64 percent experience verbal and emotional violence (World Bank 2025d). Another 11 percent experience physical or sexual violence during their commutes to work; 73 percent agree or strongly agree that an unsafe commute to the workplace adversely affects female labor force participation. About a third and a fifth of women in rural and urban areas, respectively, feel unsafe walking in their neighborhoods after dark, indicating that concerns about public safety may be a constraint on working outside the home. Policy tools, such as safe transport to the workplace, hot spot policing, and flexible working options, have the potential to improve female labor force participation by addressing concerns about mobility and safety. Key Message 10 Limited financial inclusion among women—ranked as the third most important constraint on women’s empowerment in the expert survey—are a challenge. Among women, 57 percent have no bank or mobile money account, compared with 37 percent of men. Although rural women are more likely than rural men to have taken out a loan (17 percent vs. 14 percent), conditional on having a loan, rural women tend to rely on formal savings and credit, especially nongovernmental organizations (NGOs), such as Grameen Bank, while rural men borrow more frequently from informal networks (refer to figure ES.6). This allows men to tap into larger loans and lower interest rates. Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 10 Policy tools such as extending provision of mobile money and switching to digital systems for service delivery could enhance women’s savings and control over money. Figure ES.6: Distribution of loan sources by sex, ages 15-64 Relative/Friend/ 5 Neighbor 18 Female 4 Banks Male 17 88 NGO 58 1 Money Lender 6 Source: Original figure for this publication based on calculations using data of HIES (Household Income and 1 Expenditure Survey) 2022, Bangladesh Grocery Store Bureau of Statistics, Dhaka, Bangladesh, 6 accessed through SARRAW (South Asia Raw) and SARMD (South Asia Regional Micro Database) (databases), South Asia 5 Other Region Team for Statistical Development, 8 World Bank, Washington, DC. Note: The figure shows the distribution 0% 20% 40% 60% 80% 100% of the sources of loans for individuals Percent with loans. Individuals can have more than one loan. Key Message 11 Conservative gender norms restrict both the supply of and demand for women’s labor. Traditional gender norms on the division of household labor and caregiving are reflected in time use data. Married women spend significantly more time than married men on housework and the care of dependents, reducing their opportunity for rest and education. This pattern persists among women who are working and men who are not working, which is consistent with the existence of gender norms in caregiving activities. Bangladeshi expert interviewees on women's economic empowerment ranked childcare and home responsibilities as the second most important constraint to women's economic empowerment. A majority of men (four-fifths) and women (three-fourths) also believe that men should have more access than women to scarce jobs. Bangladesh also exhibits one of the highest rates of intimate partner violence (IPV) in the world. In Bangladesh, among women who have ever been married, 70 percent have experienced some form of IPV in their lifetimes, and 54 percent of women have experienced physical or sexual violence. Evidence on the South Asia region and on lower-middle-income countries globally suggest that policies aimed at increasing women’s economic empowerment may have the unintended consequence of provoking a backlash among men through IPV if the policies challenge entrenched gender norms. Executive Summary 11 Norms affecting labor demand are also a challenge. Almost half of firms registered in the nonagricultural sectors in Bangladesh report that they do not hire women to fill management positions because of a belief that women are the source of workplace disruptions. Policy tools that might be able to change restrictive gender norms include interventions that seek to correct the beliefs of husbands and wives about social sanctions and interventions that strengthen women’s self-efficacy and aspirations through workshops, especially interventions targeted at men. Evidence on such interventions in Bangladesh is lacking. Linking Policy Priorities and the Evidence on Policy Tools Addressing gender gaps in economic empowerment amid resource constraints requires prioritizing investments and allocating budgets to the most effective policy tools. This report provides a toolkit for practitioners to promote women’s economic empowerment in Bangladesh, informed by a descriptive analysis of gender gaps and an expert survey ranking of priority issues based on the potential to improve women’s labor force participation and income generation. The approach relies on the methodology developed by the World Bank (2024g). First, high-quality global evidence on policy tools designed to enhance outcomes among women was reviewed on nine policy priorities. Each priority corresponded to a specific constraint or set of related constraints on women’s economic empowerment. This exercise provided a clear summary of the state of the evidence on each policy tool directed at each constraint. Second, a survey was conducted with experts at the World Bank who work on Bangladesh, particularly on topics that intersect with the constraints on women’s economic empowerment. The survey consisted of ranking the nine priority constraints to be targeted through policy interventions according to the potential of the interventions to increase labor force participation and incomes among women in Bangladesh. Key informant interviews with Bangladeshi experts on women's economic empowerment supplement the survey of World Bank experts. Table ES.1 presents the results of the expert survey, along with an indication of the strength of the evidence on the policy tools available to address each policy priority. Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 12 Table ES.1: Connecting policy priorities to evidence on the effectiveness of policy tools Policy priority Expert Assessment of global Evidence on ranking evidence of policy any policy tool, tools Bangladesh Skills and vocational training Growing Yes Transport, mobility, and safety* Insufficient Yes Financial inclusion Growing Yes Ownership of property and other assets Insufficient Yes Childcare and home responsibilities* Sufficient No Business ownership and business growth Growing No Job search Insufficient No Digital inclusion Growing No Migration Insufficient Yes Sources: Original table for this publication. Note: The methodology is based on World Bank (2024g). Details on the methodology can be found in Part 3 and appendix B, section B.3. The first column in this table lists the primary priorities identified through a review of the global literature. These priorities are listed in order from most important to least important, determined through a survey of World Bank experts working on topics that are relevant to women’s economic empowerment. The rank is mentioned in the second column, with darker shades indicating more importance. The third column provides an overall assessment of the available global evidence across all policy tools for the specific priority, based on the evidence review in Part 3. A priority is labeled “Sufficient” if there is a large body of evidence that conclusively shows that the policy priority is effective (or not) in enhancing women’s economic empowerment. Conversely, a priority is labeled “Insufficient” if the global evidence is scarce or lacking. “Growing” applies to priorities on which the global evidence is promising or mixed, indicating uncertainty about the effectiveness of the tools in promoting women’s economic empowerment. For a more thorough description of the toolkit and to contextualize the global evidence in Bangladesh, refer to Part 3. * Top-ranked priorities by Bangladeshi expert interviewees. Introduction Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 14 Bangladesh is often considered a success story in the promotion of gender equality in South Asia. The government and the people of Bangladesh have made strides in reducing maternal and child mortality, increasing political representation among women, broadening the access of girls and women to basic education, and raising female labor force participation (Moyeen et al. 2022), despite rigid patriarchal systems and initial extreme levels of poverty. The participation of women in the labor force has been instrumental in this progress. Higher incomes and greater educational attainment have improved women’s bargaining power within the household in decisions on marriage, reproductive health, and child health, while the visibility of women through work and community involvement has challenged cultural norms, such as purdah, the religious and social practice of sex segregation prevalent especially in South Asia. Their greater financial independence has also strengthened the role of women as caregivers of parents, reducing the preference for sons that has been common in the region (Kabeer 2024). The empowerment of women, often poor women, has been central to the growth story of Bangladesh. Policy efforts, such as promoting family planning, the expansion of education among girls and women, and the broader access of women to employment, have helped facilitate the country’s demographic transition (World Bank 2020). The agency exercised by women has enabled social progress (Kabeer 2024). Credits: Freepik Introduction 15 Despite the gains, the most recent data indicate that stark gender inequities continue to exist in labor market outcomes. Women are almost half as less likely as men to be in the labor force. They earn roughly only three-quarters of what men earn for comparable work. Women’s employment in the RMG sector, which has created employment opportunities for women and is often cited as a source of women’s empowerment, has stalled (Heath and Mobarak 2015). While involvement in the labor market can drive empowerment, the quality of employment matters greatly, too. Without job security, decent wages, or safe working conditions, it is difficult to interpret labor force participation as genuine empowerment (Kabeer 2024). In Kabeer’s (1999) framework of empowerment, a process whereby one gains the ability to make strategic life choices, these outcomes represent the achievements of women. Achievements are the observable outcomes of empowered decision- making. Rather than the outcomes driven by structural constraints, the outcomes driven by genuine choices are the ones that represent empowerment. For the process of empowerment, one must have access to two other interconnected dimensions: resources and agency. Resources are the pre-conditions – material, human, and social – that are the inputs over which women have a claim or expect to have a claim. This report finds that the gender gaps in achievements in Bangladesh partly derive from gender gaps in resources. Despite improvements in educational attainment, a large share of young, educated women in Bangladesh are unemployed or outside the labor force. Bangladeshi women also enjoy far less ownership than men over assets, such as land and livestock. Their lower access to capital and markets constrains women’s ability to grow businesses. The lack of financial inclusion, digital inclusion, and control over resources compounds these challenges – for example, the report shows that almost all women in Bangladesh use a phone, but a significantly smaller share of women own phones. Evidence shows that fostering skills among women can increase women's employment and earnings, and on-the-job training can improve women’s career prospects (Das 2021; Uckat and Woodruff 2020). This suggests that the lack of skills is another factor that may be constraining women’s employment in Bangladesh. The path from resources to achievement is made possible through agency, the ability to define one’s goals and to act on these goals. Several factors hinder women’s agency over resources in Bangladesh. Mobility constraints and concerns over public safety and safety in the workplace, drastically constrain the choices available to women. Norms that promote unequal responsibility for unpaid caregiving work and household tasks hinder the ability of women to gain education, skills, or employment. A lack of control over their own earnings—only a third of Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 16 women report that they alone are able to determine the use of their own money— reduces the negotiating power of women. Women’s agency is also inhibited by a lack of legal rights and limited public spending to support the full participation of women in the labor market. The World Bank Women, Business and the Law 2024 report (World Bank 2024e) offers a series of indicators to define the barriers to women's economic empowerment across three main categories: the barriers to women's legal rights, the barriers to the application of policy tools to facilitate the enforcement of these rights, and the barriers to the practical realization of these rights. Bangladesh's legal framework score, which indicates the existence of laws protecting women across various dimensions (for example, safety, mobility, pay), is 32.5 out of 100. This is lower than the South Asian average of 45.9. The report also highlights a significant gap in implementation. The score of Bangladesh on the existence of supportive frameworks stands at 35.0, indicating that only around one-third of the structures necessary to facilitate the implementation of gender- equal laws are in place. This is lower than the global average score of 39.5, although it is slightly higher than the South Asia score of 31.1. The latest World Bank Bangladesh Public Finance Review (PFR) estimates that compared to the high economic cost of low female labor force participation, the total public spending on programming related to girls and women in Bangladesh has been relatively low (World Bank 2024c).2 The PFR highlights the need to establish institutional structures that incentivize government entities to promote female labor force participation through public expenditure. This includes the introduction of gender-informed programming supported by evidence on what works and the strengthening of government accountability for female labor force participation in budgeting processes, such as increased transparency in identifying expenditure line items that contribute to reducing the gender gap in labor force participation (World Bank 2024c). This report examines various areas, such as accessible and affordable childcare, technical and vocational training programs, financial and digital inclusion, safety in transport, and public safety, through which the constraints on women’s economic empowerment might be relieved and that represent key areas for targeted public programs and policies. Genuine choice and agency can be difficult to measure reliably and accurately. Given the interlinked and reinforcing nature of achievements, resources, and agency, disentangling the three in the data is challenging. Parts 1 and 2 of this report quantify the gender gaps in economic achievements and resources using the latest 2 Because of limited transparency in public expenditures on human capital formation among women and on reducing the gender gap in labor force participation, budget expenditure lines on broader programs that address gender disparities are likely to underestimate the expenditures used to address the barriers to female labor force participation. Introduction 17 available microdata. They also touch on social norms, which may limit women’s agency. In Part 3, the focus shifts to identifying a set of constraints to women’s economic empowerment and the evidence on effective interventions to improve women’s access to resources and increase women’s economic empowerment in terms of economic outcomes and agency. This analysis allows the development of a policy toolkit to guide policymakers and practitioners. Bangladesh’s path to progress has been uneven, especially in women's ability to earn a fair income, the kinds of jobs and work arrangements available to women, women’s freedom of movement, and the disproportionate burden of unpaid care work experienced by women. However, progress in these areas can be realized only through a sustained policy commitment to effective interventions, alongside efforts to reform the legal and social institutions that perpetuate gender inequality. Data Sources Household Surveys → Labor Force Survey. The LFS is a nationally representative survey conducted periodically by the Bangladesh Bureau of Statistics with the support of the International Labour Organization (ILO). In this report, the LFS 2022 round is used to analyze labor market outcomes, such as labor force participation, employment status and sector of employment, occupation, and details on hours worked. Repeated cross-sections from 2005, 2010, 2013, 2015, 2016, and 2022 are used to analyze trends and explore cohort dynamics related to wage employment.3 → Household Income and Expenditure Survey. The HIES is a nationally representative survey conducted periodically by the Bangladesh Bureau of Statistics with the support of the World Bank to collect household consumption and expenditure data. In this report, HIES 2022 provides data on earnings. The HIES measures earnings among nonwage earners more comprehensively than the LFS. In addition, HIES 2022 is also used to study financial inclusion and explore migration patterns.4 → Bangladesh Integrated Household Survey. The 2018–19 round of the BIHS, conducted by the International Food Policy Research Institute, is representative of rural Bangladesh. In this report, the nationally representative sample of BIHS 3 Refer to LFS (Labor Force Survey) 2005, 2010, 2013, 2015, 2016, 2022, Bangladesh Bureau of Statistics, Dhaka, Bangladesh, accessed through SARRAW (South Asia Raw) and GLDRAW (Global Labor Database Raw) (databases) South Asia Region Team for Statistical Development, World Bank, Washington, DC. 4 Refer to HIES (Household Income and Expenditure Survey) 2022, Bangladesh Bureau of Statistics, Dhaka, Bangladesh, accessed through SARRAW (South Asia Raw) and SARMD (South Asia Regional Micro Database) (databases), South Asia Region Team for Statistical Development, World Bank, Washington, DC. Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 18 2018–19 is used to analyze detailed information on the ownership of agricultural assets and livestock, hours worked on agriculture and livestock, loans, and savings among individuals in rural Bangladesh.5 → Time Use Survey. The 2021 Time Use Survey is a nationally representative survey conducted by the Bangladesh Bureau of Statistics to collect time diary data detailing the time spent on various activities in a 24-hour period. Combined with demographic data, the Time Use Survey is used to understand the time spent on work or productive activities, care work, learning, and rest and leisure activities.6 → Global Findex Database. Findex data are collected through a periodic survey conducted by the World Bank to measure financial inclusion across multiple countries using a standardized, comparable questionnaire. Repeated cross- sections from 2017 and 2021 are used to quantify women’s access to financial and digital tools that enable involvement in the labor market.7 → Demographic and Health Surveys. The DHS are nationally representative surveys conducted periodically that have, in the past, received support from the United States Agency for International Development. In this report, DHS 2022 is used to measure homeownership and to study transitions from the labor market around key life events, such as marriage and childbirth.8 → Multiple Indicator Cluster Surveys. These are nationally representative surveys conducted periodically by the United Nations Children’s Fund. The 2019 round of the survey is used in this report to measure digital access and perceptions of safety.9 → Bangladesh Individual Consumption Study. The BICS 2024 is a household survey conducted by the World Bank Development Data Group and the South Asia Gender Innovation Lab (SAR GIL), in collaboration with the World Bank Chief Economist of the South Asia Region. The sample is an extension of the Women’s Empowerment Metric for National Statistical Systems survey conducted by the International Food Policy Research Institute in 2022. The study focuses on rural villages and informal urban settlements. High-income areas are excluded. It was conducted in Khulna Division (south), Mymensingh Division (center), and Rangpur Division (north). While the survey is not nationally representative, the sample allocation across the various parts of the country, including urban and rural localities, ensures that the results reflect Bangladesh’s geographic 5 Refer to BIHS (Bangladesh Integrated Household Survey) 2018–19, International Food Policy Research Institute, Washington, DC, https://doi.org/10.7910/DVN/ NXKLZJ, Harvard Dataverse, V2. The BIHS sample is comprised of two subsamples: a nationally representative subsample and an additional subsample for the evaluation of the Feed the Future Program. The analysis in this report uses only the nationally representative subsample. 6 Refer to TUS (Time Use Survey) 2021, Bangladesh Bureau of Statistics and UN Women Bangladesh, Dhaka, Bangladesh. 7 Refer to 2017 and 2021 versions, respectively, of Global Findex (Global Findex Database), World Bank, Washington, DC, https://doi.org/10.48529/rv7t-ng66; https://doi.org/10.48529/qda7-6z97. 8 Refer to DHS (Demographic and Health Surveys) 2022, ICF International, Fairfax, VA. 9 Refer to MICS (Multiple Indicator Cluster Surveys) 2019, Bangladesh Bureau of Statistics and United Nations Children's Fund, Dhaka, Bangladesh, https://mics. unicef.org/surveys. Introduction 19 diversity. In this report, data of the study are used to explore preferences related to childcare and childcare facilities as reported by the primary caregivers of children ages 7 or less.10 → World Values Survey. This is a nationally representative survey conducted by the World Values Survey Association across multiple countries to measure a rich set of norms and values. In this report, the 2018 round is used to quantify a broad set of supply-side norms constraining women’s labor market outcomes.11 → Dhaka Low Income Area Gender, Inclusion, and Poverty Survey. The DIGNITY 2018 survey is representative of slum and low-income nonslum areas of the two Dhaka City corporations and additional low-income locations in the Greater Dhaka Statistical Metropolitan Area based on the 2011 census. In this report, the survey is used to understand the labor market outcomes of migrants to Dhaka.12 Other Data Sources → World Bank Enterprise Surveys. These surveys cover formal private manufacturing and service firms across more than 160 economies. Each firm has at least five employees. The surveys are undertaken using a standardized questionnaire that produces comparable data. In this report, the 2013 and 2022 rounds are used to quantify demand-side constraints to women’s labor market involvement, such as the availability of on-site childcare and constraints on hiring women.13 → World Development Indicators. This World Bank database compiles internationally comparable statistics on global development. This report draws on the modeled ILO estimates imported to the database from ILOSTAT that situate the progress of Bangladesh in labor market outcomes relative to other countries.14 → Mapped in Bangladesh. The Mapped in Bangladesh Project, conducted by BRAC University between 2017 and 2021, mapped the universe of export- oriented RMG factories that are part of the Bangladesh Garment Manufacturers and Exporters Association and the Bangladesh Knitwear Manufacturers and Exporters Association. The project excludes factories in export processing zones. The data are used in this report to study the factory-level involvement of women in the RMG sector.15 10 Refer to BICS (Bangladesh Individual Consumption Study) 2024, World Bank, Washington, DC. 11 Refer to WVS (World Values Survey) Wave 7 (2017–2022), World Values Survey Association, King's College, Old Aberdeen, UK, https://doi.org/10.14281/18241.24. 12 Refer to DIGNITY (Dhaka Low Income Area Gender, Inclusion, and Poverty Survey), 2018, Microdata Library, World Bank, Washington, DC, https://doi. org/10.48529/gare-az26. 13 Refer to 2013 and 2022 versions, respectively, of WBES (World Bank Enterprise Surveys) (dashboard), World Bank, Washington, DC, https://microdata. worldbank.org/index.php/catalog/1929; https://microdata.worldbank.org/index.php/catalog/6497. 14 Refer to ILOEST (ILO Modelled Estimates Database), International Labour Organization, Geneva, https://ilostat.ilo.org/methods/concepts-and-definitions/ ilo-modelled-estimates/; WDI (World Development Indicators) (dashboard), World Bank, Washington, DC, https://datatopics.worldbank.org/world-development- indicators/. For the Stata package used to access these data (wbopendata), refer to Azevedo (2011). 15 Refer to MiB (Mapped in Bangladesh) (dashboard), Centre for Entrepreneurship Development, BRAC University, Dhaka, Bangladesh, https://ced.bracu.ac.bd/ mib-2/. Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 20 Data Gaps This report identifies the following important data gaps that present obstacles to understanding the drivers of women’s economic empowerment.16 → Comparability of women’s employment over time. The LFS 2022 survey instrument underwent methodological revisions to follow the revised ILO model questionnaires, thereby addressing the undercounting of women’s work in previous rounds. While the revisions have provided a more accurate measure of women's work, especially by capturing self-employment and women’s unpaid work in household businesses, they represent an obstacle to comparisons of women’s employment in the LFS 2022 round with earlier rounds. For a detailed discussion of the revisions in the survey instrument and the implications for the measurement of women’s employment in Bangladesh, refer to appendix B, section B.4. → Panel data tracking the work of women and men through life transitions and over the life cycle. The dearth of nationally representative household panel data in Bangladesh prevents a deeper study of trends in labor market outcomes, such as employment and transitions across sectors or occupations, over time with age and major life events, such as marriage and childbearing. The BIHS is a household panel survey, but it covers only rural areas and is, in any case, outdated. The last round was conducted in 2018–19, before the Covid-19 pandemic. Although the BIHS tracks split households, it does not track women when they move to new households. This limitation means that women in patrilocal settings (where women typically move into the homes of their husbands or of the parents of their husbands at marriage), such as Bangladesh, are almost never tracked in panel survey data. → Direct measures of work experience. Nationally representative LFSs do not measure work experience directly, requiring the use of age as a proxy for experience. However, this poses a challenge in interpreting women’s earnings over the life span because women tend to face more career interruptions and transitions relative to men. → Data on small informal firms. The World Bank Enterprise Surveys cover only registered firms with a firm size of at least five employees, leaving out more than half of all the individuals working outside agriculture (58 percent) and almost half of women working outside agriculture (48 percent) who work in small, often informal firms with fewer than five employees (based on LFS 2022 data). The lack of data on small unregistered firms prevents a thorough study of the labor demand for women’s work. 16 There are gender gaps in other areas, such as health care and leadership, that are not the focus of this report. Introduction 21 → De facto female entrepreneurship. The identification of entrepreneurs and enterprises, especially female entrepreneurs, is empirically challenging because of the lack of a standardized definition and inadequate data in lower-middle- income countries, including Bangladesh. The obstacles in identifying woman- owned enterprises, particularly in household and enterprise surveys, are well documented (Hardy, Kagy, and Jimi 2024). Most definitions of entrepreneur are centered on two concepts: (1) direct control over the activities and decisions within the enterprise and (2) ownership of the enterprise, often determined by the presence of the individual’s name on the enterprise registration documents (ILO 2018; Swaminathan et al. 2023). However, data on these characteristics are often not captured in surveys and do not exist for Bangladesh. In South Asia, it is not uncommon for de facto enterprise control to be disconnected from formal registration. Women may exercise control over enterprises that are registered in the names of others (often the husbands), or they may exercise limited control over enterprises registered formally in their own names. Such dynamics make the identification of de facto women owners more difficult. While the World Bank Enterprise Surveys collect some data on women’s ownership of firms, these data are restricted to formal firms with more than five employees. They do not collect data on the de facto control exercised within the enterprises, which may not be strongly correlated with formal ownership status in South Asia. The BIHS collects detailed data on nonagricultural enterprises, including decision-making within enterprises, but the data are restricted to rural areas, and the information is reported by the men in the survey households, potentially leading to bias in reporting. Such data can be integrated into existing surveys that already collect data on business ownership and business activities (such as the World Bank Enterprise Surveys and the HIES) or surveys that collect data on the decision-making and control over the resources in agriculture (such as the BIHS, which collects data for the construction of the women's empowerment in agriculture index), providing deeper insights into women’s entrepreneurship.17 The limited evidence that is available suggests that most microentrepreneurs in lower-middle-income countries may be entrepreneurs by necessity, not by choice. Most appear to prefer a wage job over self-employment and take up self-employment only because of labor market frictions (Jayachandran 2021). Given the importance of the ability to act on one’s choices in the process of empowerment, surveys measuring self-employment should aim also to measure systematically the motivations behind women’s self-employment (Kabeer 1999). 17 WEAI (Women’s Empowerment in Agriculture Index), (dashboard), International Food Policy Research Institute, Washington, DC, https://www.ifpri.org/project/ weai/. Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 22 → Richer data on women’s ownership of productive assets and property. Existing surveys are either geographically limited or contain coarse data. The BIHS contains rich data on land and livestock ownership, but is representative of only rural Bangladesh. Moreover, the last round was conducted in 2018, and no future rounds are planned. The DHS contains questions on landownership, but no other details. The future of the DHS is uncertain, leading to potential data gaps in the future. → More self-reported data on asset ownership. Surveys have traditionally relied on proxy respondents whereby one individual in the household reports the ownership of assets among all individuals in the household, often leading to bias in the reported ownership, especially in the case of women. The selection of the specific individual to survey in the household, usually the head of the household and, in South Asia, usually a man, can also lead to bias because the self-reported identification of the head of household is often a result of gendered social norms (Kilic and Moylan 2016). Privately interviewing multiple members of the household would lead to more accurate measurements of asset ownership (Kilic, Moylan, and Koolwal 2020). Based on these findings, ongoing methodological work by the World Bank, such as the Living Standards Measurement Study Plus Program, provides a model for measuring gender disaggregated data on labor market activities and asset ownership by conducting surveys with women. If integrated with surveys in Bangladesh, such a state-of-the-art survey methodology offers a potential path to filling the data gap. → Mobility. Indicators on women’s freedom of mobility are scarce and only measured minimally in the BIHS. However, the BIHS only covers rural areas, and, because the last round was completed in 2018–19, no recent data measure women’s mobility in Bangladesh. → Supply and demand for childcare. There was only one question each on the availability of employer-provided childcare in the World Bank Enterprise Survey 2022, which covers only formal firms with five or more employees, and in the LFS 2022. Richer data are not available on the demand for childcare and the supply of childcare facilities at the workplace or through government programs or community-based programs. To address this gap, the World Bank is mapping the availability of childcare facilities in Dhaka (Majoka and Shams 2025). In addition, ongoing work by SAR GIL and the World Bank Living Standards Measurement Study team is refining tools to measure the demand for childcare in rural Bangladesh (World Bank 2025e). → Savings. The BIHS collected detailed data on the savings held by rural households, but equivalent data on urban households are lacking. Because the final round of the BIHS has concluded, data on women’s savings behavior in either rural or urban areas will no longer be available. Introduction 23 Key Definitions → Working-Age Population: Ages 15–64 ¬ Youth population: ages 15–24 ¬ Nonyouth population: ages 25–64 → Labor Force: defined as the sum of employed individuals and unemployed individuals (ILO 2023). → Labor Force Participation Rate: defined as the working-age labor force, divided by the working-age population (ILO 2023). → Unemployment: An individual is classified as unemployed if the individual is not employed, was looking for work (that is, paid work or starting a business) in the previous month, and was available to start work in the previous week (ILO 2023). → Employment: Using the 13th ICLS classification (ILO 1983), an individual is classified as employed if the individual: ¬ Worked for wage, salary, commission, tips, or any other pay, even if for only one hour, in the previous week ¬ Ran or undertook any sort of business, farming, or other activity to generate income or profit, even if for only one hour, in the previous week ¬ Performed unpaid work in a business owned by a household member, even if for only one hour in the previous week ¬ Worked for at least one hour to produce goods and services in agriculture or fishing for own household in the previous week ¬ Had a paid job or business in the previous week, but was temporarily absent and was expected to return The employed consist of those who are (1) paid employees, (2) self-employed, or (3) unpaid family workers in the agricultural or nonagricultural sector. This report presents labor force participation rates according to both the definitions of employment of the 13th ICLS (ILO 1983) and the 19th ICLS (ILO 2013), but relies on only the definition of the 13th ICLS for the more detailed analysis of women’s labor market outcomes. The key difference between the two definitions is that the 19th ICLS definition of employment is restricted to Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 24 work performed for pay or profit, while the 13th ICLS definition includes work to produce goods for own consumption. This has implications in measurement, particularly in agriculture, where women are less likely than men to produce for sale (Gaddis et al. 2023). Appendix B, section B.5 presents a summary of the changes in the definition and their impact on the measurement of women’s employment in Bangladesh. A key reason for using the 13th ICLS definition of employment in this report is that, while the 19th ICLS definition was reflected in the 2022 Bangladesh LFS questionnaire, headline indicators still rely on the previous definition (BBS 2023). This approach also ensures consistency with other recent World Bank analyses on the region (for instance, World Bank 2024d) and with international databases. The latter include the ILO Modelled Estimates database, which does not yet apply the 19th ICLS definition of employment among countries in which it would generate a methodological break because there are not sufficient data points based on the 19th ICLS standards to produce reliable regional and global estimates (ILO 2025).18 → Wage Employee: A wage employee is an employed person who is either: ¬ A regular paid employee ¬ A day laborer → Unpaid Worker: An unpaid worker is an employed individual who works without pay in a family or household business. → Self-Employed Worker: Individuals who reported in the HIES or the LFS that they were employers, self-employed, or own-account workers. → Nonagricultural Self-Employed Workers: This report distinguishes between self-employed individuals who work alone (own-account workers) and self- employed individuals who work with at least one other individual at an enterprise (employer) (ILO 2018).19 Together, own-account workers and employers represent the individuals whom this report refers to as self-employed. → Agricultural Self-Employed Workers: The report does not distinguish between self-employed individuals who work with employees or without employees in agriculture because of the seasonal nature of agricultural work. This makes the measurement of employees challenging because many self-employed individuals may employ others only during periods of high labor requirements, 18 ILOEST Database (ILO Modelled Estimates Database), International Labour Organization, Geneva, https://ilostat.ilo.org/methods/concepts-and-definitions/ ilo-modelled-estimates/. 19 In LFS 2022, this means recategorizing the 13 percent of individuals ages 15–64 working in nonagricultural sectors who reported themselves as employers and the 25 percent who reported themselves as own-account workers, based on the reported firm size. To categorize individuals as employers, other individuals in the enterprise need not be paid employees. Introduction 25 such as during planting or harvesting. The report also does not distinguish between agricultural activities conducted for sale and for own consumption. → Not in employment, education, or training: NEET is the share of NEET youth (ages 15-24) (ILO 2022). → Main Job and Earnings: The analysis in this report is restricted to the main jobs reported by individuals in the LFS and the jobs with the highest incomes in the HIES. This covers the vast majority of employment. In LFS 2022, 1.4 percent of employed women and 6.1 percent of employed men reported that they had second jobs. In HIES 2022, 6.5 percent of employed women and 13.7 percent of employed men reported having a second job. The earnings data presented in this report correspond to earnings from the jobs with the highest monthly earnings in the HIES as calculated in the World Bank’s South Asia Regional Micro Database, which harmonizes HIES data across multiple countries and multiple data rounds.20 ¬ Wage earnings: the sum of take-home monetary income, net of deductions at the source and earnings in kind ¬ Agricultural self-employment earnings: the net of agricultural income and agricultural expenditures ¬ Nonagricultural self-employment earnings: the individual share of enterprise net revenues owned by the household based on the number of hours worked by individuals relative to the total household hours Not in Employment Education or Training N E E T 20 SARMD (South Asia Regional Micro Database) (repository), World Bank, Washington, DC, https://github.com/worldbank/SARMD?tab=readme-ov-file. 1 PART Women in the Labor Force 1 | Women in the Labor Force 27 1.1 Overview A man in Bangladesh is more likely than a woman to be economically active Despite recent progress, a large gender gap in labor force participation persists in Bangladesh. The magnitude of the gap depends on the definition of employment used. Under the definition of employment of the 13th ICLS (ILO 1983), a working-age man is almost twice as likely as a working-age woman to be economically active (83 percent vs. 45 percent). In terms of employment as a share of the population, this translates to 79.8 percent and 43.3 percent, respectively. As a share of the population, unemployment is 3.1 percent among men and 1.6 percent among women.21 In absolute numbers, this corresponds to 31 million women and 9 million men who are out of the labor force.22 Under the definition of employment of the 19th ICLS (ILO 2013), a man is almost four times as likely as a woman to be in the labor force. Among the working-age population, 82 percent of men and only 22 percent of women participated in the labor force, implying that 44 million women and 10 million men are out of the labor force in Bangladesh. Credits: Fatih Turan / pexels 21 These figures use the latest year of available Bangladesh Labor Force Survey (LFS) data, LFS 2022. The unemployment data reported are for the entire population, which translate to unemployment rates (that is, unemployment as a share of the labor force) of 3.7 percent among men and 3.6 percent among women. 21 The population level numbers are estimated using the survey weights in LFS 2022. Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 28 Use of the more recent definition leads to a significantly lower female labor force participation rate because of the exclusion of the types of work that women are significantly more likely to be involved in, specifically own-use production and unpaid work in household businesses, including subsistence farming. For a detailed discussion of the differences between the definitions under the 13th ICLS and the 19th ICLS and their implications for the measurement of women’s employment, refer to appendix B, section B.5. Women in Bangladesh have outpaced women in the rest of South Asia in labor force participation, especially since 2005.23 They have closed the gender gap relative to the regional trends. While the average female labor force participation rate fell in the rest of the region between 2005 and 2022, falling from 35 percent to 32 percent, the rate in Bangladesh rose by more than half, from 30 percent to 47 percent. Nonetheless, as of 2022, the rate in Bangladesh lags the rates among aspirational peers (China, Indonesia, and Thailand) and only outperforms the rate in India among structural peers (Cambodia, India, and Viet Nam). It also lags the rates in other lower-middle-income and upper-middle-income countries overall (refer to figure 1.1).24 Figure 1.1: Labor force participation rate, international comparison, by sex, 2022, ages 15-64 84 Bangladesh 47 Female 83 Indonesia 55 Male 82 Thailand 68 80 China 70 Source: Original figure for this publication 80 based on calculations using modeled ILO India 32 estimates from the ILOEST Database (ILO 89 Modelled Estimates Database), International Combodia 79 Labour Organization, Geneva, https://ilostat.ilo. 83 org/methods/concepts-and-definitions/ilo- Viet Nam 76 modelled-estimates/, accessed through WDI (World Development Indicators) (dashboard), 86 World Bank, Washington, DC, https://datatopics. South Asia 32 worldbank.org/world-development-indicators. Retrieved on July 25, 2025. Lower Middle 79 Income 51 Note: Data for all countries is from 2022. Comparison countries come from the Upper Middle 79 Income 61 Bangladesh Country Economic Memorandum (World Bank 2022). Bangladesh's aspirational peers are Indonesia, Thailand, and China and structural peers are India, Cambodia, and Viet 0% 20% 40% 60% 80% 100% Nam. Regional comparison is with rest of South Asia (Afghanistan, Bhutan, India, Maldives, Percentage (%) Nepal, Pakistan, Sri Lanka), and income group comparison is with Lower-Middle-Income Countries (LMIC) and Upper-Middle-Income Countries (UMIC). 23 Rest of South Asia consists of Afghanistan, Bhutan, India, Maldives, Nepal, Pakistan, and Sri Lanka. 24 The comparison countries are taken from the Bangladesh Country Economic Memorandum (World Bank 2022). These countries were identified based on their similarity to Bangladesh in total population, dependency ratio, share of rural population, gross domestic product (GDP) per capita, share of manufacturing, and the human capital index. Bangladesh’s structural peers are India, Cambodia, and Viet Nam, and aspirational peers are Indonesia, Thailand, and China. Labor force participation rates reported in figure 1.1 are based on the ILO’s modelled estimates series, which provides internationally comparable labor statistics, but reports a slightly higher female labor force participation rate for Bangladesh relative to the rate estimated using the LFS. 1 | Women in the Labor Force 29 The gender gap in labor force participation is costly Recent estimates (World Bank 2024d) suggest that raising the female labor force participation rate to the male labor force participation rate in Bangladesh could lead to long-run output gains of between 14 percent and 44 percent, equivalent to growing the country’s gross domestic product (GDP) by roughly US$64 billion to US$204 billion. The estimates of the GDP increase vary depending on the extent to which additional woman entering the labor force can be accommodated in new jobs and supported by other inputs, such as additional capital, allowing them to add to total production instead of merely displacing men in existing jobs. A fifth of young women are NEET A group of special concern among youth (ages 15–24) are those youth who are out of school and out of work, often referred to as NEET. In 2022, about a fifth of young women (19 percent) were still NEET, compared with 10 percent of young men (refer to figure ES.5). Many more young women than young men complete education or training without entering the workforce, highlighting the importance of the school-to-work transition. In urban areas, almost half (42 percent) of young women ages 15–24 are NEET, whereas, in rural areas, fewer than a 10th (8 percent) of young women are NEET. There is evidence suggesting that women’s education is one of the most desirable traits reported by husbands, which may partly explain why women complete their education, but fail to enter the labor force (Buchmann et al. 2023). This disconnect between women's education and women’s labor force participation reflects a significant underutilization of human capital. For this reason, while most of the analysis in this report focuses on the working- age population (ages 15–64), the report also examines outcomes among two subsets of the population separately whenever there are noteworthy differences between the two: first, the subset of individuals (ages 25–64) who are expected to have completed education and, second, a younger subset of individuals (ages 15–24) who might still be pursuing education. Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 30 1.2 Gender Sectoral Segregation, Status in Employment, and Occupation Women in nonagricultural work are engaged mostly in RMG, domestic work, tailoring, or education Women’s participation in nonagricultural work, that is, industry or services, is limited. Only 11 percent of women were involved in nonagricultural work in 2022, compared with 29 percent in agriculture. This is in stark contrast to men, 57 percent of whom are working outside agriculture, while only 22 percent are working in agriculture. More than 60 percent of the women and men working outside agriculture are working in services. The other third are working in industry. Among women who are in nonagricultural work, about a third are employed in manufacturing (34 percent), followed by domestic work and services (17 percent); tailoring, salons, and other services (16 percent); and education (12 percent) (refer to figure 1.2).25 A larger share of men, 27 percent, are employed in wholesale and trade. Figure 1.2: Nonagricultural employment across sub- sectors (1-digit BSIC), by sex, ages 15-64 Manufacturing 34 Female 19 Domestic Work & Services 17 Male 1 Tailoring,  Salon & Other Services 16 4 Education 12 4 Wholesale & Retail Trade 7 27 Human Health & Social Work 4 1 Accommodation & Food Services 3 4 Construction 2 12 Finance & Insurance 1 2 Source: Original figure for this publication based Public Administration & Defence 1 3 on calculations using data of LFS (Labor Force Survey) 2022, Bangladesh Bureau of Statistics, Other 3 23 Dhaka, Bangladesh, accessed through SARRAW (South Asia Raw) (database), South Asia Region 0% 10% 20% 30% 40% Team for Statistical Development, World Bank, Percent Washington, DC. Note: Sub-sectors are defined as Section codes (1-digit) in the Bangladesh Standard Industrial Classification (BSIC) 2020 used in LFS 2022. 25 The Other services category corresponds to Section S (other service activities) in the Bangladesh Standard Industrial Classification 2020 (BBS 2020). The activities among 4.2 percent of women in this category are classified as “Activities of nongovernment and nonprofit organizations (NGOs, BRAC, ASHA, Proshika, charity organization, etc.)” (Group 94930 in BBS 2020). 1 | Women in the Labor Force 31 Among young women, this sectoral split is somewhat different. Among women ages 15–24 employed outside agriculture, 50 percent work in industry, where they are concentrated in RMG manufacturing. Among young urban women in nonagricultural work, 45 percent are active in the RMG sector. This shows the relevance of the RMG sector for young women. In rural areas, almost half of young women (45 percent) and a third of older women (30 percent) in nonagricultural work are in manufacturing, with a fifth of young and older women in tailoring. Thus, rural manufacturing (not only RMG manufacturing) accounts for a substantial share (45 percent) of women’s manufacturing employment nationwide. Women work mostly self-employed, Credits: Mahmudul Hasan / Pexels while men also do wage work Almost two-thirds of employed women in Bangladesh are self-employed (64 percent). A fifth are engaged in wage work (19 percent), and fewer than a fifth (16 percent) are involved in unpaid work. This contrasts sharply with employed men, nearly half of whom are in wage employment (48 percent), while another half are self-employed (49 percent) (refer to figure ES.3).26 In agriculture, the largest sector of women’s employment, women are almost exclusively self- employed (74 percent), whereas most men are wage employees or self-employed. More than a fifth of the women (23 percent) are unpaid employees, a category that covers almost no men (4 percent) (refer to figure 1.3, panel a). This shows the significant gender disparity in employment types within the sector. 26 Fewer than 1 percent of employed individuals (0.84 percent of employed women and 0.97 percent of employed men) in LFS 2022 reported an employment status of other. These individuals are excluded from the analysis of employment status in this report. Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 32 Figure 1.3: Employment, by a. Agricultural employment employment status and sex, ages 25-64 100 80 Percentage (%) 60 Self-Employed Unpaid Employee Wage Employee 40 20 0 Female Male b. Nonagricultural employment 100 80 Percentage (%) 60 Own-Account Worker Employer Unpaid Employee Wage Employee 40 Sources: Original figure for this publication based on calculations using data of LFS (Labor Force Survey) 2022, 20 Bangladesh Bureau of Statistics, Dhaka, Bangladesh, accessed through SARRAW (South Asia Raw) (database), South Asia Region Team for Statistical Development, 0 Female Male World Bank, Washington, DC. Meanwhile in nonagricultural sectors, most women and men are wage employees (67 percent and 57 percent, respectively). The main differences in employment status between women and men derive from the greater share of men who are own-account workers and employers, while a greater share of women are unpaid employees (refer to figure 1.3, panel b). This indicates a gender gap in entrepreneurial roles. In rural areas, a similar share of men and women working outside of agriculture are wage employees or own-account workers (roughly 60 percent and 30 percent, respectively). However, 13 percent of men and 2 percent of women active in rural nonagricultural work are employers, and only 2 percent of men are unpaid employees, compared with 8 percent of women, reflecting a gender imbalance in business ownership and unpaid labor. Box 1.1 highlights the impact of changes in the LFS on understanding women’s employment. 1 | Women in the Labor Force 33 Box 1.1 Measurement of Women’s Employment in the LFS The LFS underwent important The measurement of women’s employment methodological changes in the 2022 survey in LFS 2022 is broader and captures more instrument that improved the measurement women, especially young women, working of women’s work, resulting in a substantially in informal jobs in household businesses higher number of women considered who were previously not considered employed relative to previous rounds. employed. LFS 2022 thus offers a better measure for capturing women’s work than LFS 2022 shows a stark 615 percent rise in previous rounds. However, these updates the share of young rural women ages 15– also preclude the comparison of women’s 24 in self-employment compared with LFS self-employment in LFS 2022 with previous 2016–17. Similar trends are not observed LFS rounds. among young rural men in the LFS. This change appears likely to be an artificial In light of the revisions to the survey result of changes in survey methods, not a instruments and data collection protocols true change over time. in the LFS, this report uses LFS 2022 to provide an up-to-date picture of women’s Four factors potentially explain these labor market outcomes. This is because trends. First, the LFS 2022 instrument the changes implemented in LFS 2022 was revised to follow the revised ILO improve the measurement of women’s model questionnaires and address the work, capturing a more comprehensive set undercounting in previous LFS rounds, of activities that should be considered as including the addition of recovery questions employment under the 13th ICLS definition and questions meant to probe for-profit and allowing for a richer analysis. employment. Second, the choice options in the question measuring employment However, the time trends analysis of the LFS status in LFS 2022 were revised to make is restricted to wage employment because them more descriptive relative to previous the measurement of wage employment is rounds. Third, data collection protocols not expected to be affected by instrument were improved, such as the introduction of improvements primarily in the measurement computer-assisted personal interviewing of self-employed women. and a longer enumerator training period, and this is expected to enhance the quality ––––––––––––––––––––––––––––––––––– of data. Fourth, unlike the common practice Note: For a detailed discussion of the in nationally representative household methodological changes in LFS 2022, surveys, the survey weights used in LFS along with the impact of the changes on 2016–17 were not revised based on the the measurement of women’s employment enumeration of the Population and Housing in Bangladesh, refer to appendix B, section Census 2022. B.4. Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 34 1.3 Trends in Women’s Wage Employment Wage employment has risen more slowly among women than men Wage employment provides a stable source of income among women, especially compared with self-employment, which is often more informal and precarious. In Bangladesh, a hypothetical choice experiment reveals a strong preference among both men and women for the greater stability offered by contract work. Women are willing to forgo up to a 40 percent increase in their monthly income in exchange for a long-term contract (Mahmud et al. 2021). Nonetheless, wage employment is extremely low among women in Bangladesh. The share of women in wage employment rose only marginally, from 6 percent to 8 percent, between 2005 and 2022. This represented a lag with respect to men’s wage employment, which increased from 34 percent to 40 in the same period (refer to figure 1.4, panel a). This accounted for a widening gender gap in wage employment, from 28 percentage points in 2005 to 33 percentage points in 2022. Figure 1.4: Sectoral composition of the 100% population in wage employment, ages 15-64 80% a. Wage employment, by sex Percentage (%) 60% Female population 40% Services Industry 20% Agriculture 0% 2005 2010 2013 2015 2016 2022 Year Male population 100% 80% Percentage (%) 60% 40% 20% 0% 2005 2010 2013 2015 2016 2022 Year 1 | Women in the Labor Force 35 Figure 1.4 (continued) b. Women’s wage employment, 100% by rural and urban areas 80% Rural only Percentage (%) 60% Services 40% Industry Agriculture 20% 0% 2005 2010 2013 2015 2016 2022 Year Urban only 100% 80% Percentage (%) 60% 40% 20% 0% 2005 2010 2013 2015 2016 2022 Year c. Men’s wage employment, by rural and urban areas 100% Rural only 80% Percentage (%) 60% Services Industry 40% Agriculture 20% 0% 2005 2010 2013 2015 2016 2022 Year Urban only 100% Sources: Original figure for this publication 80% based on calculations using data of LFS (Labor Percentage (%) Force Survey) 2005, 2010, 2013, 2015, 2016, 60% 2022, Bangladesh Bureau of Statistics, Dhaka, Bangladesh, accessed through SARRAW (South Asia Raw) and GLDRAW (Global Labor 40% Database Raw) (databases), South Asia Region Team for Statistical Development, World Bank, Washington, DC. 20% Note: The figure shows the share of male and female population in wage employment in each 0% sector. Employed individuals with a missing 2005 2010 2013 2015 2016 2022 sector (3.4% of women and 0.03% of men across Year all survey rounds) are treated as not employed. Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 36 In the most recent LFS data, women’s wage employment is almost three times greater in urban areas than in rural areas (14 percent vs. 5 percent, respectively). In urban areas, women’s wage employment rose from around 11 percent to 14 percent in 2005–22, while, in rural areas, it rose from about 4 percent to 5 percent (refer to figure 1.4, panel b). Hardly any women work in agricultural wage employment. The share fell from 0.9 percent to 0.6 percent in 2005–22. The marginal rise in women’s wage employment was driven by industry, which employed 3.3 percent of working-age women in 2022, up from 2.4 percent in 2005 (refer to figure 1.4, panel a). While women's wage employment in services increased slightly, from 3.0 percent to 3.7 percent, in 2005–22, the rise was smaller in industry (refer to figure 1.4, panel a). This is in line with findings in the literature that, based on data of the LFS 2013, suggests that Bangladesh has not yet experienced the transition to a service economy (ADB and ILO 2016). However, a transition to services may not raise women’s wage employment in services. Previous analysis confirms that women are restricted to the lower-paying agriculture sector (Mahmud and Bidisha 2018), while men dominate the more well paid industry and services sector Credits: Mumtahina Tanni / Pexels (Solotaroff et al. 2019). The World Bank South Asia Development Update October 2024 shows that, unlike other emerging economies, the boom in the services sector in South Asia is not correlated with increases in female labor force participation (World Bank 2024d). Further analysis is needed to understand the factors that prevent women from enjoying the expansion of service jobs that has helped women enter labor markets in other countries. 1 | Women in the Labor Force 37 1.4 The Cohort Dynamics of Women’s Wage Employment The entry of women into wage employment is now slowing An examination of wage employment across birth cohorts reveals generational shifts in women’s work behavior. Figure 1.5, panels a and b, illustrates wage employment trends among individuals born in specific decades—from the 1960s to the 2000s—as they age. Each line in the figure represents a birth cohort by decade and shows the trends in the cohort’s involvement in wage employment over the life span. This facilitates an analysis of cohort effects by comparing the colored lines and an analysis of age effects by comparing levels across points on each colored line. Figure 1.5: Share of population in wage 60% employment, by decade birth cohort and sex 50% a. Women 40% Percentage (%) 30% Born in: 20% 1960s 1970s 1980s 10% 1990s 2000s 0% 15-24 25-34 35-44 45-54 55-64 Age b. Men 60% 50% 40% Percentage (%) 30% 20% 10% 0% 15-24 25-34 35-44 45-54 55-64 Age Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 38 Figure 1.5 (continued) 60% c. Women, agriculture 50% 40% Percentage (%) Born in: 30% 1960s 1970s 20% 1980s 1990s 2000s 10% 0% 15-24 25-34 35-44 45-54 55-64 Age d. Men, agriculture 60% 50% 40% Percentage (%) 30% 20% 10% 0% 15-24 25-34 35-44 45-54 55-64 Age e. Women, nonagricultural 60% sectors 50% 40% Percentage (%) 30% 20% 10% 0% 15-24 25-34 35-44 45-54 55-64 Age f. Men, nonagricultural 60% sectors 50% 40% Percentage (%) Sources: Original figure for this publication based on calculations using data of LFS (Labor 30% Force Survey) 2005, 2010, 2013, 2015, 2016, 2022, Bangladesh Bureau of Statistics, Dhaka, Bangladesh, accessed through SARRAW 20% (South Asia Raw) and GLDRAW (Global Labor Database Raw) (databases), South Asia Region 10% Team for Statistical Development, World Bank, Washington, DC. 0% Note: Employed individuals with a missing sector 15-24 25-34 35-44 45-54 55-64 (3.4% of women and 0.03% of men across all Age survey rounds) are treated as not employed. 1 | Women in the Labor Force 39 Women born in the 1980s, now ages 30–40, are engaged in wage employment at higher rates than women born in the 1960s and 1970s when they were the same age (refer to figure 1.5, panel a). At age 15–24, women born in the 1990s were engaged in wage employment at higher rates than women born in the 1980s when they were the same age, but the rates plateau as they reach ages 25–34, converging with the lower rates among earlier generations. Among men, younger cohorts are engaged in wage employment at higher rates than older cohorts (refer to figure 1.5, panel b). Men also remain in wage employment as they age. Only the oldest observed cohort (born in the 1960s) shows decreasing rates from age 45 onward. These differences contribute to the widening in the gender gap in wage employment. Wage employment patterns differ across generations by sector. In agriculture, women's participation in wage employment is low, at less than 2% across the lifespan of all cohorts. This may be declining even further for younger cohorts as just 0.2% of women born in the 2000s were engaged in wage employment in agriculture between ages 15-24 (refer to figure 1.5, panel c). Although men are engaged in wage employment in agriculture at much higher rates across cohorts and lifespan, the trend of declining engagement for younger cohorts is also present for men. Men tend to leave agricultural wage employment as they age and enter at lower rates; 13 percent of men born in the 1960s were wage employees in agriculture at ages 35–44, compared with about 3 percent in the youngest cohort (refer to figure 1.5, panel d). This indicates a generational shift away from agricultural wage employment. In nonagricultural sectors, wage employment among women increased across cohorts until the 1990s. Women born in the 1980s show higher rates of wage employment than those born in the 1960s and 1970s (refer to figure 1.5, panel e). However, women born in the 1990s, although initially entering wage employment at a higher rate, started to exit after age 25. Women born in the 2000s cohort appear to be entering at a lower rate, aligning with the levels of the 1970s cohort. This suggests a potential reversal of the initial growth trend. Among men, the opposite is observed among all but the youngest cohort (those born in the 2000s). Each cohort except for the youngest shows higher shares of wage employment in nonagricultural sectors compared to the cohort preceding it (refer to figure 1.5, panel f). Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 40 1.5 Hours Worked Credits: Dominic Chavez / World Bank Women in wage employment and women employed in industry work the most hours On the intensive margin, employed women worked significantly fewer average hours (33 hours) than men (50 hours) during the previous week.27 Among employed women, 26 percent work part time (20 hours or less a week), compared with only 3 percent of employed men (refer to figure ES.2). The hours worked by women vary significantly with employment status. Women wage workers put in the most hours on average, working almost as much as their male counterparts (46 hours). They are followed by women employers, who work 28 hours on average. Women who are own-account workers or unpaid employees work similar hours, averaging 25 and 24 hours, respectively. Self-employed women thus work the fewest hours, which may indicate either that they face barriers to full employment or choose this sort of employment because they have limited time available for work. This contributes to the gender differences in self-employment whereby women's activities tend to be more informal and limited. Similarly, women’s work hours vary substantially by sector. Women in industry work the most hours and similar hours to men in industry. The largest gap in hours worked is observed in agriculture, where men work an average of 20 hours more per week than women, followed by services, where the difference is 16 hours. The situation is similar among youth. These patterns suggest that women enter areas of work where they can work shorter and perhaps more flexible hours, which likely reflects the constraints women face in fitting work around home responsibilities. Hypothetical choice experiments in urban and periurban areas confirm women’s preference for shorter working hours. Women in the experiment were willing to forgo a substantially higher portion of their income than men to work 10 fewer hours a week (Mahmud et al. 2021). 27 This section relies on data for the 25–64 age-group to remove the potential effect of women who may still be in education, that is, women ages 15–24. 1 | Women in the Labor Force 41 1.6 Earnings Women’s earnings are significantly lower than men’s earnings Monthly earnings are significantly lower among women than men.28 This is both because women work fewer hours a week and because women are paid less per hour than men. The gap thus narrows when accounting for hours worked because women work an average of 17 hours less a week (33 hours vs. 50 hours). Women’s hourly earnings are still only 74 percent of the hourly earnings of men. Some of the gap is also explained by sector of employment. Women’s employment is dominated by the agricultural sector, which is the least well paying sector overall. However, the largest gender wage gap is in industry where women work similar hours, but earn 37 percent less than men per hour. This may arise partly because men are more likely to hold higher positions, such as supervisory positions in garment factories. However, even within occupations, men tend to earn higher hourly wages relative to women. Other factors such as employer discrimination may also be a factor. Among both men and women, monthly earnings are almost double in urban areas relative to rural areas. A key consideration is whether women advance to more well paying roles as they gain experience in the labor market. Age is often used as a proxy for work experience because the national LFS do not directly measure experience. In Bangladesh, nonagricultural hourly earnings rise with age among wage workers only among men (refer to figure 1.6). Using age as a proxy for experience overlooks the career interruptions and transitions that women face more frequently than men, for instance, because of marriage or childbirth. The lack of data on experience or the work histories of individuals thus represents a key gap that prevents more in-depth analysis of earnings trends across the career trajectories and lives of women. 28 HIES earnings data are used. This includes earnings among wage workers and self-employed workers. Because this survey does not capture information on unpaid workers, the averages reported here are conditional on activity as a paid worker. Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 42 Figure 1.6: Hourly earnings, a. Overall, by age and sex nonagricultural wage employees, by age and sex 160 88107 103 120 103 Hourly Earnings (in Taka) Female 99 97 94 83 88 80 Male 79 75 81 69 67 74 80 66 66 54 45 40 0 15-19 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64 b. Rural areas, by age and sex 160 90 120 Hourly Earnings (in Taka) 72 90 89 85 80 73 84 82 79 78 63 66 65 68 72 80 64 59 45 50 40 0 15-19 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64 c. Urban areas, by age and sex 134 160 129 121 108 109 87 107 120 92 90 Hourly Earnings (in Taka) 83 94 83 84 75 78 71 68 80 58 59 46 Source: Original figure for this publication 40 based on calculations using data of HIES (Household Income and Expenditure Survey) 2022, Bangladesh Bureau of Statistics, Dhaka, Bangladesh, accessed through SARRAW (South Asia Raw) and SARMD (South Asia Regional Micro 0 Database) (databases), South Asia 15-19 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64 Region Team for Statistical Development, World Bank, Washington, DC. 1 | Women in the Labor Force 43 Using a Kitagawa-Oaxaca-Blinder decomposition (Blinder 1973; Kitagawa 1955; Oaxaca 1973), the earnings gender gap can be decomposed into the gap deriving from differences in observed characteristics and access to resources (that is, the endowments or the explained component) and the gap deriving from the differential returns to these characteristics and resources among men and women (that is, the unexplained component). Four broad types of observable characteristics available in the data on Bangladesh are included as endowments in this analysis: (1) demographic characteristics, including age (linear and squared), ever-married status, and relationship to the household head; (2) household characteristics, including household size, the presence of a child ages 6 or less, and the presence of an individual ages 60 or more; (3) socioeconomic indicators, including educational attainment and ownership of a cellphone in the household; and (4) region, that is, whether the household is in an urban area.29 There may also be unobservable characteristics, such as risk aversion, willingness to work in competitive environments, or discrimination, that differ between men and women and that are part of the unexplained component of the earnings gap. The decomposition analysis reveals that merely a fifth (22 percent) of the hourly earnings gender gap is explained by differences in observed endowments between men and women. Men are more educated than women. In particular, they are more likely to have attained the secondary-school certificate (class 10) and the higher- secondary certificate (class 12) (refer to appendix B, table B.3). 1.7 The Sociodemographic Profile of Female Labor Market Participation Labor force participation declines monotonically with wealth quintile among both men and women, but the fall in labor force participation is steeper among women, decreasing from 54 percent in the bottom quintile to 36 percent in the top quintile (refer to figure 1.7, panel a).30 This suggests that much female labor force participation is work of necessity which women abandon as the incomes and wealth of their households increase. 29 The decomposition analysis does not include sector of employment because sector is an outcome of many of the same factors that influence wages. The analysis is also expected to yield a lower bound because of the inclusion of household-level endowments given that men and women tend to live in similar households. 30 The wealth quintiles in LFS 2022 are estimated using survey-to-survey imputation based on data in the HIES 2022 (World Bank 2025c). Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 44 Figure 1.7: Labor force, a. Labor force participation rate employment, and NEET patterns by wealth quintile 100% and sex, ages 15-64 100% 86 85 84 82 80% 86 85 84 78 82 80% 78 60% 54 Percent 60% 50 54 Percent 45 50 42 40% 45 36 42 40% 36 20% 20% 0% 0% Bottom Q Q2 Q3 Q4 Top Q Bottom Q Q2 Q3 Q4 Top Q Female Male b. NEET 100 100 80 80 (%) (%) 60 NEET Percentage Source: Original figure for this publication 60 In Education or Training NEET Percentage based on calculations using data of LFS (Labor Force Survey) 2022, Bangladesh In Employment Education or Training Bureau of Statistics, Dhaka, Bangladesh, 40 In Employment accessed through SARRAW (South Asia 40 Raw) (database), South Asia Region Team for Statistical Development, World Bank, Washington, DC. 20 20 Note: Labor force participation is calculated as the labor force (employed 0 and unemployed) within a wealth quintile, Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 divided by the total working-age population 0 Q1 Q2 Q3 Q4 Q5 Female Q3 Q4 Q5 Q1 Q2 Male (ages 15–64) in the wealth quintile. Wealth quintiles are estimated using survey-to- Female Male survey imputation from Bangladesh HIES 2022 (World Bank, 2025c). Well-off young women are the most likely to be NEET. Among young men and women, the share in education or training increases and the share in employment declines with wealth. However, the share of NEET men falls only slightly with wealth, while the share of NEET women increases with wealth (refer to figure 1.7, panel b) (Khatun and Saadat 2020). Women in wealthier households who do work are more likely to work outside agriculture. The share of employed individuals working in industry and services rises with wealth. The increase is especially large among women. The share of women in nonagricultural work rises roughly twofold, from 17 percent in the bottom quintile to 44 percent in the top quintile. This pattern holds in urban and rural areas. So, it does not merely reflect the greater wealth in urban areas. 1 | Women in the Labor Force 45 Women in the top household quintile who work in nonagricultural sectors are also more likely to be wage employees (70 percent) and less likely to be unpaid or own- account workers (4 percent and 18 percent, respectively). However, they are still not as likely as men who work in nonagricultural sectors to be employers. Among employers, 19 percent are men in the top household quintiles, but only 5 percent are women in the top household quintile. Earnings among the employed rise with wealth quintile among both men and women. However, while men experience an increase in both agricultural and nonagricultural earnings, the rise in earnings among women is driven primarily by nonagricultural earnings, while agricultural earnings are stagnant among women across wealth quintiles. Together, these patterns are consistent with the hypothesis that wealthier women are more likely to enter the labor force only to engage in jobs with higher earnings and potentially better conditions, that is, work of choice rather than work of necessity. Nonetheless, women are still not likely to be employers, limiting the empowerment they experience through work, potentially because of discrimination or conservative social norms. Growth in the types of employment that are generally considered more desirable, particularly among women, could attract women into the workforce even as household incomes rise and women have the choice to stay at home (Gentile et al. 2023; Goldin 1995). The shape of female labor force participation in urban and rural areas The relationship between female labor force participation and educational attainment in urban areas can be best described as a U shape, as documented in the literature in South Asia and globally (Fletcher, Pande, and Moore 2018; Klasen et al. 2021; Mahmud and Bidisha 2018). The urban female labor force participation rate is highest among women with low educational attainment or high educational attainment (refer to figure ES.2, panel b). The relationship between hours worked and educational attainment shows a similar pattern. In urban areas, women with lower educational attainment are typically engaged as cleaners and helpers or in other low-skill occupations, such as woodworking, crafts, and related trades, while, at higher educational attainment, women enter more highly skilled professional occupations, particularly teaching. The share of women active in woodworking, crafts, and related trades rises with educational attainment up to the secondary-school certificate (class 10) and declines at higher Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 46 levels of education. As educational attainment rises to moderate levels, women leave the workforce and only reenter at higher educational levels to take higher status white-collar jobs. Educational attainment also significantly affects labor force participation in rural areas. A reduction in the gender gap is observed as the female labor force participation rate rises at higher educational attainment, while the male labor force participation rate remains stagnant. The rise in female labor force participation from low educational attainment to secondary education (attainment of the higher- secondary certificate [class 12]) is driven by increased agricultural employment, whereas tertiary education leads to higher nonagricultural employment (refer to figure 1.8, panel a). Figure 1.8: Labor market a. Rural areas, women, ages 25–64 participation by education, region, and sex 100 100 80 80 (%) (%) Percentage 60 Percentage Agricultural Employment 60 Nonagricultural Employment Unemployment 40 Not in Labor Force 40 20 20 0 ry ry e e 0 n n et et ) ) ) ) ia ia tio tio 10 10 12 12 e te pl pl rt rt ca ca et le ss ss ss ss m m Te Te pl p In In u u Co Co la la la la m m ny ny y yEd Ed (C (C (C (C co co y y /A /A ar ar o o SC SC C C im imN N SS SS im im + + H H SC SC ar ar Pr Pr H H Pr Pr & & b. Urban areas, women, ages 25–64 100 100 80 80 (%) (%) Percentage 60 Percentage 60 40 40 20 20 0 ry ry e e 0 n n et et ) ) ) ) ia ia tio tio 10 10 12 12 e te pl pl rt rt ca ca et le ss ss ss ss m m Te Te pl p In In u u Co Co la la la la m m ny ny y yEd Ed (C (C (C (C co co y y /A /A ar ar o o SC SC C C im imN N SS SS im im + + H H SC SC ar ar Pr Pr H H Pr Pr & & 1 | Women in the Labor Force 47 Figure 1.8 (continued) c. Rural areas, men, ages 25–64 100 80 Percentage (%) 60 Agricultural Employment Nonagricultural Employment Unemployment 40 Not in Labor Force 20 0 ry e pl n et ) ) ia m tio 10 12 e pl rt et co ca ss ss m Te In u Co la la ny y Ed (C (C y /A ar ar o SC C N SS im + H SC Pr im H Pr & d. Urban areas, men, ages 25–64 100 80 Percentage (%) 60 Source: Original figure for this publication 40 based on calculations using data of LFS (Labor Force Survey) 2022, Bangladesh Bureau of Statistics, Dhaka, Bangladesh, 20 accessed through SARRAW (South Asia Raw) (database), South Asia Region Team for Statistical Development, World Bank, Washington, DC. 0 ry e pl n et ) ) ia m tio 10 12 e pl rt et co ca ss ss m Te Note: The category 'No Education & In du Co la la ny (C (C Primary Incomplete' includes individuals E y /A ar No SC C without any education and individuals who SS im + H y SC ar Pr have passed class 4 or below. The category im H 'Primary Complete' includes individuals Pr & who have passed class 5 through class 9. Across levels of educational attainment in urban areas, men’s employment is relatively stable, though the composition of employment varies. Agricultural employment plays a larger role among men at lower educational attainment (refer to figure 1.8 panel d). Conversely, in urban areas, women’s employment shifts from low-skill occupations to high-skill professional roles with rising educational attainment. In agriculture, women move from elementary roles to skilled agricultural, forestry, and fishing occupations and shift from unpaid employment to own-account work as their educational attainment increases. Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 48 After marriage, women spend more time in housework, at the cost of rest and education Following marriage, women ages 15–64 tend to substitute time away from leisure and educational activities towards home and care responsibilities (refer to figure 1.9). Causal estimates from rural Bangladesh show that women who marry later have higher educational attainment (Field and Ambrus 2008), indicating a trade-off between marriage and education among women. Relative to married women without children, the additional time spent on housework and dependent care by married women with children comes at the cost of rest and leisure, as well as work and production. This contrasts with men who, following marriage, dedicate more time to work and perform little housework or dependent care.31 Figure 1.9: Time use by life transitions and sex, ages 15-64 24 20 Average Hours in a Day 16 12 8 4 0 Male Female Male Female Male Female Source: Original figure for this publication Unmarried Married without Child Married with Child based on calculations using data of TUS (Time Use Survey) 2021, Bangladesh Bureau of Statistics and UN Women Bangladesh, Dhaka, Bangladesh. Rest / Leisure / Self-Care Education / Learning Note: A “child” for this figure is defined as Housework / Dependent Care Work / Production an individual aged 9 or below, based on a question in TUS 2021 asking whether a child aged 9 or below was present during an activity. 31 For mapping the major division activity codes in the Time Use Survey data to the activity categories used in the figures, refer to appendix B, section B.1. 1 | Women in the Labor Force 49 Married women spend more time than married men with children during activities (an average of 11 hours vs. 7 hours a day), including work and productive activities (roughly 10 percent vs. 1 percent of the time on work and productive activities).32 This may have implications for productivity because working women may be more distracted or more likely to take an unplanned absence from work given their roles as primary caregivers. Marriage and the associated household responsibilities also play a role in the ability of women to work outside the home. Only a small share of women work outside the home before marriage. The share is larger in urban areas (13 percent) than in rural areas (7 percent). Among women working outside the home before marriage, urban women are more likely than rural women to continue working outside the home after marriage (52 percent vs. 46 percent). The low rates of work outside the home before marriage may be caused by cultural norms or beliefs that if women work before marriage, their marriage prospects are hindered. It may also simply be that women marry at an early age before they have the opportunity to work outside the home. The average age at first marriage among women ages 15–24 who have ever married is 17, on the early end of the age of first employment and perhaps too young to travel for work or to work away from home.33 Among women under 40, those who have been married are more likely than those who have never married to be in employment, the former tend to work 5–10 hours less per week than the latter, conditional on employment. This difference is driven by rural women. The difference in urban areas is minimal.34 Among married women, especially younger women, women with children work significantly fewer hours. However, this gap narrows with age as the hours worked by women without children fall. The differences in hours worked because of childbearing, unlike the differences because of marriage, are more pronounced in urban areas than in rural areas. These patterns do not necessarily describe the causal effects of marriage or childbearing on women’s choice or ability to participate in the labor force. Bussolo, Rexer, and Triyana (2024) use a matched panel approach to show that, in Bangladesh, similar to the rest of South Asia, married women are 12 percentage points less likely to be employed in the year of marriage than in the year previous to the year of marriage. The marriage penalty explains most of the decline in employment postmarriage, indicating that childbearing is not as substantial a constraint to employment as marriage among women in Bangladesh (Bussolo, Rexer, and Triyana 2024; World Bank 2024d). 32 The sample is restricted to married individuals with a child present in the household. A child is defined as any individual age 9 or less, unlike the rest of the report, where a child is defined as any individual age 6 or less. This is because the relevant Time Use Survey question refers to a child age 9 or less. 33 The age at first marriage is proxied by the age at first cohabitation, calculated using the reported year of first cohabitation and the reported year of birth, as measured in the DHS 2022. The average age at first marriage reported here is for the sample of married women ages 15–24 surveyed in DHS 2022. 34 These comparisons hold age constant. The patterns occur among women in each five-year age bin between ages 15 and 39. The sample size of never- married women ages 40 or more is too small for meaningful comparison. Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 50 1.8 Social Norms and Women in the Labor Force According to the UNDP gender social norms index, Bangladeshi men and women are more conservative than their counterparts in most peer countries.35 In Bangladesh, 93 percent of men and 91 percent of women hold views with at least two biases toward traditional gender norms. A large share of the population holds beliefs that constrain women’s labor market involvement. For example, 87 percent of men and 89 percent of women believe that the children suffer if their mother works for pay. A majority of men and women also believe that men have more right to a job than women if jobs are scarce (82 percent vs. 73 percent), that men make better executives (72 percent vs. 67 percent), that being a housewife is as fulfilling as working for pay (58 percent vs. 60 percent), and that, if a woman earns more than her husband, problems are almost certain to arise (64 percent vs. 55 percent). These beliefs are similar in urban and rural areas. Gender norms that reinforce the social dominance of men over women not only constrain women’s advancement in the labor market, but also influence how households respond to policy reforms targeting women’s economic empowerment. Growing evidence in South Asia and beyond shows that improvements in women’s economic empowerment may not always lead to greater agency and bargaining power among women within the household, particularly if existing norms or identities are threatened. Instead, such improvements may provoke a backlash among men (Bulte and Lensink 2019; Chowdhury and Bhuiya 2004; Erten and Keskin 2021; Heath 2014). For example, Daher et al. (2023) find evidence of a backlash among men after women’s right to drive was legally recognized in Saudi Arabia. While women became more likely to be employed, they also reported lower levels of financial autonomy. Several studies show that such a backlash may also take the form of IPV (Bulte and Lensink 2019; Cullen et al. 2024; Erten and Keskin 2021; Guarnieri and Rainer 2021; Heath 2014). Data from 2015 indicate that Bangladesh had the highest rates of IPV in the region and one of the highest rates in the world (WHO 2021). More recent surveys show that, while IPV rates have declined, 41 percent of ever-married women ages 15 or more had, as of 2024, experienced some form 35 The gender social norms index (UNDP 2023) is created from seven indicators in the World Values Survey data (two political, one educational, two economic, and two physical integrity dimensions). Refer to WVS (World Values Survey) (dashboard), King's College, Old Aberdeen, UK, https://www.worldvaluessurvey.org/ wvs.jsp. 1 | Women in the Labor Force 51 of violence by their partners in the previous year. A significantly larger share of women (70 percent) reported that they had experienced IPV during their lifetimes. Psychological violence is the most common type of violence. Among women, 33 percent had experienced controlling behavior, and 15 percent had faced emotional violence by their partners in the previous year. A greater share of women had experienced these forms of violence during their lifetimes, at 50 percent and 33 percent, respectively (BBS 2025a). Other cultural factors contribute to the limited mobility and opportunities among women in many developing countries, including in South Asia (Jayachandran 2015). Low female labor force participation has been attributed partly to cultural norms that enforce patrilocality (potentially restricting women’s access to inherited land or disrupting advancement in education or work), prioritize women's perceived purity, or reinforce purdah, the practice of secluding women. Such cultural institutions can persist even in a context of economic growth, suggesting that gender inequality may not be spontaneously resolved through economic development alone. Kabeer (2024) argues that variations in the ways religious norms are incorporated in practice by men and women in Bangladesh may have allowed women to navigate restrictive traditions to some degree, driving the gradual progress in women’s empowerment. The limited legal protections ensuring women’s right to work only reinforces these conservative norms (World Bank 2024e). While there are no legal restrictions on women obtaining jobs in the same way as men, there is, in Bangladesh, no legislation prohibiting gender discrimination in employment or wages, in recruitment based on marital status, parental status, or age. The law does not recognize the right of women to work in industrial jobs in the same way it recognizes the right of men to these jobs (Labor Act 2006, Sections 39, 40, 42, and 87). This unequal treatment is intensified by a lack of government mechanisms to combat gender discrimination in the workplace. In an attempt to make inroads against these obstacles, some World Bank projects have intentionally set guidelines that require employers participating in the projects to guarantee the equitable treatment of men and women, such as by requiring a certain portion of the labor hired by contractors to be women and by mandating equal pay for men and women. This is the case of the Local Government COVID-19 Response and Recovery Project, which focuses on labor-intensive public works in Bangladesh and has set a standard that could be adapted in other projects and government programs.36 36 Refer to LGCRRP (Local Government COVID-19 Response and Recovery Project) (dashboard), Local Government Engineering Department, Dhaka, Bangladesh; World Bank, Washington, DC, https://oldweb.lged.gov.bd/ProjectHome.aspx?projectID=1008. 2 PART Deep Dives 2 | Deep Dives 53 Credits: Scott Wallace / World Bank 2.1 Women in Agriculture Women work fewer hours in agriculture than those working outside agriculture Although agriculture is the largest sector of women’s employment and despite the contribution of agriculture to the expansion in women’s employment on the extensive margin, on the intensive margin women in agriculture work fewer hours per week than women in nonagricultural sectors. Women in agriculture work an average of 24 hours a week, while women in services and industry work an average of 51 hours and 37 hours a week, respectively (refer to figure 2.1). This is true no matter the status in employment (self-employment, wage employment, or unpaid work). Women also tend to spend more time in rest and leisure activities than in work in other industries. Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 54 Figure 2.1: Average hours worked in the past 7 days, by sector of employment 60 and sex, ages 25-64 51 53 53 50 Average Hours in the Past 7 Days 44 Female 40 37 Male 30 24 20 10 Source: Original figure for this publication based on calculations using data of LFS 0 (Labor Force Survey) 2022, Bangladesh Bureau of Statistics, Dhaka, Bangladesh, Agriculture Industry Services accessed through SARRAW (South Asia Raw) (database), South Asia Region Team for Statistical Development, World Bank, Washington, DC. Women seldom own productive assets in agriculture Women are extensively engaged in agriculture, but they have limited control over agricultural assets. LFS data show that fewer than a third (29 percent) of the women in agriculture are engaged in crop-related activities. By contrast, the share among men is more than four-fifths (82 percent). Richer BIHS data reveal that rural women rarely own agricultural land and do not spend much time on plots either.37 The few women landowners typically acquire land through the families of their husbands; 62 percent of the land owned by women in rural areas is inherited from the families of husbands. These patterns are reflected in the relatively low score of Bangladesh in women’s access to assets in the Women, Business, and the Law 2024 ranking (World Bank 2024e). Although the law guarantees equal administrative power and ownership rights over assets such as land, Muslim Personal Law does not grant sons and daughters equal rights to inherit the assets of their parents, nor do surviving spouses have equal inheritance rights.38 In practice, women do not typically obtain even the share of inheritance determined by law. Most of the livestock women own is poultry, while men own both poultry and large livestock (cattle). Men raise livestock primarily for income through the sale of animal products. This is less common among women. Of the livestock owned exclusively by men (one animal may have more than one owner), 28 percent are 37 The 2018–19 round of the BIHS produced rich data on livestock and land ownership in rural Bangladesh. The survey sought information on up to three primary owners of each animal or plot owned by the household. There was more than one owner in the case of 12.0 percent of the animals and 6.3 percent of the plots. The survey also captured information on the distribution of ownership across household members, entire households, and nonhousehold members. The survey responses were elicited from a man in each household. 38 Muslim Personal Law (shariat) is a set of special laws on marriage, succession, inheritance, and charities among Muslims enacted under the Raj and recognized under certain circumstances in parts of Bangladesh, India, and Pakistan. 2 | Deep Dives 55 reared solely for income rather than for own consumption or a combination of income and own consumption. By contrast, only 6 percent of the animals of which at least one woman is an owner are reared solely for income. Though women do not typically own large livestock (cattle) or small livestock (sheep or goats), they spend more time than men rearing all types of livestock (refer to figure 2.2). Figure 2.2: Animal ownership a. Average number of animals owned, by sex of owner, ages 15-64 and time input, by animal type and sex 0.2 Large Livestock 1.8 Female Male 0.3 Small Livestock 0.8 7.0 Poultry 1.9 0.5 Other 0.7 0 2 4 6 8 Animals b. Average family labor hours spent on animals in the past year, by sex of laborer 156 Large Livestock 126 58 Source: Original figure for this publication based on calculations using data of Small Livestock 26 BIHS (Bangladesh Integrated Household Survey) 2018–19, International Food Policy Research Institute, Washington, DC, https:// doi.org/10.7910/DVN/NXKLZJ, Harvard 152 Dataverse, V2. Poultry 11 Note: Panel A: Respondents could report up to three owners for each type of livestock. The figure shows the gender of first owner 4 reported (only 12% of animals had more Other than one owner). "Large Livestock" includes 2 bullock, milk cow, buffalo; "Small Livestock" includes goat, sheep, pig; "Poultry" includes duck, chicken (broyler, layer, cockrel). 0 50 100 150 200 Number of animals have been one-tail Hours winsorized at the 99th percentile. Panel B: Annual labor hours are winsorized at the 99th percentile. Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 56 Ownership rather than the time spent working is associated with the ability of an individual to make decisions about how the income generated from the activity is used. Men tend to own more large livestock, even though women spend more time working with livestock. Men also have a greater say on how the income generated from rearing large livestock is spent. Among women, 11 percent report that they have little or no say in such decisions, compared with only 4 percent of men. This pattern holds in the case of cash crops and food crops, which are tied to landownership. Respectively, 22 percent and 15 percent of women report that they have little or no say on how the income from these activities is used, compared with only 2 percent and 4 percent of men. Women own and raise more poultry, and report having more say on how the income from this activity is spent. Only 2 percent of women report that they provide little or no input in decisions on how this income is spent, compared with 42 percent of men. This suggests that supporting women’s ownership of more valuable agricultural assets could empower women in making decisions on how the associated income is spent. Despite their extensive engagement in agriculture, women are also typically not targeted by the extension services received by households in rural areas for crop agriculture, livestock, or fisheries. Credits: K M Asas / World Bank 2 | Deep Dives 57 2.2 Women in the Ready-Made Garment Sector RMGs: an important source of wage employment among young women RMGs are an important source of employment among women, particularly young urban women. Almost half (45 percent) of urban women ages 15–24 employed in nonagricultural sectors are in the RMG sector, and 98 percent of these women are wage employees. Most of these young women (85 percent) have primary education or less. The RMG sector attracts young women with lower educational attainment (up to secondary schooling), while women with high-secondary schooling or any tertiary education tend to work outside the RMG sector. Urban and larger RMG factories employ more women. Among working-age women employed in the sector, 64 percent live in urban areas. Most formal export-oriented RMG firms are in Dhaka District (37 percent) and Gazipur District (33 percent).39 This geographic concentration likely constrains women’s opportunities in the sector because of their limited migration for work. Data of the Mapped in Bangladesh Project show that the share of women workers in RMG-exporting factories rises with factory size.40 While 27 percent of small firms (firms with fewer than 20 workers) exhibit a woman worker share of more than 50 percent, the share jumps to 78 percent in large factories (more than 100 workers). Women do not obtain higher positions or more pay in RMG with age Working-age women employed in the RMG sector in urban areas earn around 25 percent an hour less than working-age men. This gap only widens as women age and seems to be associated with the employment positions available to men and women. Within RMG factories, young women mostly work as seamstresses and tailors, and about a fourth are employed in lower-level positions, such as garment helpers. Women are more likely than men in the same age range to work in these lower- level jobs. The gap widens as men and women age because men are more likely to advance to supervisory and management positions. Women’s occupations within RMG are similar across education levels, while men engage in higher-level occupations as their level of education increases (refer to figure 2.3). 39 This covers firms that are part of the Bangladesh Garment Manufacturers and Exporters Association and Bangladesh Knitwear Manufacturers and Exporters Association. 40 Refer to MiB (Mapped in Bangladesh) (dashboard), Centre for Entrepreneurship Development, BRAC University, Dhaka, Bangladesh, https://ced.bracu.ac.bd/ mib-2/. Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 58 Figure 2.3: Distribution of RMG occupations by education, age, and sex a. RMG occupations, by age and sex Ages 15-24 Ages 25-34 Ages 35-44 Ages 45-64 Tailors, Dressmakers, 40 46 44 37 Furriers and Hatters 29 27 21 13 23 14 12 17 Garment Helpers 12 9 11 10 Sewing and Embroidery 21 22 25 8 Machine Operators 25 20 15 14 Quality Checker and 4 4 2 4 Finishing Workers 8 9 6 4 Manufacturing 0 1 1 2 Supervisors/Managers 4 10 15 16 12 12 16 31 Other 23 25 33 43 0% 25% 50% 75% 0% 25% 50% 75% 0% 25% 50% 75% 0% 25% 50% 75% Percent Percent Percent Percent a. RMG occupations, by educational attainment and sex No Education & Primary Incomplete (Class 0-4) Primary Complete (Class 5-9) SSC (Class 10) & Above 45 45 36 Tailors, Dressmakers, Furriers and Hatters 36 33 15 15 18 17 Garment Helpers 10 11 8 Sewing and Embroidery 19 23 18 Machine Operators 21 22 17 Quality Checker and 2 2 13 Finishing Workers 4 6 10 Manufacturing 1 1 2 Supervisors/Managers 4 6 17 18 10 14 Other 25 22 34 0% 25% 50% 75% 0% 25% 50% 75% 0% 25% 50% 75% Percent Percent Percent Female Male Source: Original figure for this publication based on calculations using data of LFS (Labor Force Survey) 2022, Bangladesh Bureau of Statistics, Dhaka, Bangladesh, accessed through SARRAW (South Asia Raw) (database) South Asia Region Team for Statistical Development, World Bank, Washington, DC. 2 | Deep Dives 59 Women’s Employment in RMG has stalled The RMG sector has been expanding since 2005, but the growth has stagnated in recent years (according to the 2005–22 LFS). Women’s share in employment in the sector rose from 35 percent to 57 percent between 2010 and 2013. However, since 2013, the RMG employment share of men has increased, reaching 61 percent of RMG workers in 2022. Until the cohort of women born in the 2000s, each successive cohort of women was substantially more likely than its predecessor to be working in the sector at each age up to age 45. However, each successive cohort also tended to start exiting the sector at an earlier age. Although it is premature to draw conclusions about the dynamics of the life cycle of the cohort of women of the 2000s, data indicate that the cohort has engaged in the sector at a lower rate than the previous cohort (refer to figure 2.4, panel a). Figure 2.4: Share of a. Women population employed in RMG, by decade birth cohort and sex 10% 8% Percentage (%) Born in: 6% 1960s 1970s 1980s 4% 1990s 2000s 2% 0% 15-24 25-34 35-44 45-54 55-64 Age b. Men 10% 8% Percentage (%) 6% 4% Sources: Original figure for this publication based on calculations 2% using data of LFS (Labor Force Survey) 2005, 2010, 2013, 2015, 2016, 2022, Bangladesh Bureau of Statistics, Dhaka, 0% Bangladesh, accessed through SARRAW 15-24 25-34 35-44 45-54 55-64 (South Asia Raw) and GLDRAW (Global Age Labor Database Raw) (databases), South Asia Region Team for Statistical Development, World Bank, Washington, DC. Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 60 These trends are true among men as well, with one important exception: while young women born in the 1990s had reduced their involvement in the RMG sector by their mid-30s, the young men of the same birth cohort had increased their involvement in the sector by their mid- 30s, potentially indicating a fundamental shift in participation in the sector by sex (refer to figure 2.4, panel b). The RMG sector has been a source of employment and economic empowerment among women (World Bank 2016), but such reliance on one sector alone also renders any progress more vulnerable to economic shock. The sector contracted in 2024. Many factories Credits: Freepik shut down, and hundreds of thousands of women lost their livelihoods (Raihan and Bidisha 2018). The contraction was a result of global factors, such as weaker global demand and higher tariffs, and domestic challenges, such as power shortages and high interest rates. Furthermore, the sector’s biggest export markets, the European Union and the United States, which, together, accounted for roughly 70 percent of Bangladesh’s total apparel exports, led to a drop in apparel prices and to lower incomes among RMG workers in Bangladesh (Corraya 2025). The expected graduation of Bangladesh from least developed country status in 2026 will prevent the country from qualifying for the Generalized System of Preferences. This and the spread of protectionist trade policies around the globe represent additional challenges to the ability of the RMG sector to continue driving women’s economic empowerment. Expert reports on ways to improve the competitiveness of the sector after Bangladesh graduates from least developed country status typically recommend that the sector’s export markets be diversified and that the focus of the sector shift to high–value added products that offer higher profit margins, including raising internal efficiencies by leveraging digitalization and modernizing infrastructure (OECD and UNCTAD 2023; Razzaque, Islam, and Rahman 2024). 41 Refer to Export Performance (data tables), Bangladesh Garment Manufacturers and Exporters Association, Dhaka, Bangladesh, https://www.bgmea.com.bd/ page/Export_Performance. 2 | Deep Dives 61 2.2 Women in Nonagricultural Self-Employment Self-employed women seldom have employees Women in Bangladesh are less likely than men to be self-employed. There were 12.0 million self-employed men compared with only 1.5 million self-employed women in nonagricultural sectors in 2022. Women are even less likely to be employers (refer to figure 1.3, panel b). This is true in both urban and rural areas. This disparity in the likelihood of being an employer is important because estimates of average monthly earnings by status of employment show that men employers tend to earn more than men with a different employment status (refer to figure 2.5). However, more than four-fifths of the self-employed women in nonagricultural sectors in both rural and urban areas are own-account workers, that is, they do not have any employees Figure 2.5: Average monthly earnings (in thousands of taka) in nonagricultural sectors, by 28 30 employment status and sex, ages 15-64 20 Monthly Earnings (in Thousands of Taka) 25 Female 20 Male 18 17 16 15 13 13 8 10 Source: Original figure for this publication based on calculations using data of HIES (Household Income and Expenditure Survey) 2022, Bangladesh Bureau of Statistics, Dhaka, 5 Bangladesh, accessed through SARRAW (South Asia Raw) and SARMD (South Asia Regional All Employment Wage Employers Own-account Micro Database) (databases), South Asia Region Statuses Employees Workers Team for Statistical Development, World Bank, Washington, DC. Almost half (48 percent) of the women in nonagricultural self-employment operate in tailoring and other repair service subsectors. An additional 15 percent work in trade, followed by manufacturing (12 percent) and domestic work (10 percent). By contrast, among men in self-employment, almost half are in trade, followed by transport and storage (20 percent) (refer to figure 2.6). The occupations in nonagricultural self-employment are also gendered. Women are most likely to be working in crafts and related trade occupations, while about half of self-employed men are service and sale workers, followed by plant and machine operators and assemblers. These patterns are similar in both urban and rural areas. Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 62 Figure 2.6: Nonagricultural self-employment across sub- sectors (1-digit BSIC), ages Tailoring,  Salon & 48 15-64 Other Services 5 Wholesale & 15 Retail Trade 48 Manufacturing 12 Female 7 Male Domestic Work 10 & Services 1 Education 6 1 Accommodation 3 & Food Services 7 Health & 2 Social Work 1 Administrative & 1 Support Services 2 Professional, 1 Scientific & Technical 1 Transportation 1 and Storage 20 Source: Original figure for this publication based on calculations using data of LFS (Labor Force Survey) 2022, Bangladesh Bureau of Other 2 Statistics, Dhaka, Bangladesh, accessed 6 through SARRAW (South Asia Raw) (database), South Asia Region Team for Statistical 0% 20% 20% 30% 40% 50% 60% Development, World Bank, Washington, DC. Note: Sub-sectors are defined as Section codes (1-digit) in the Bangladesh Standard Industrial Classification (BSIC) 2020 used in LFS 2022. Women’s businesses operate part time and earn less than men's Self-employed women in both urban and rural areas work roughly 30 hours a week, compared with more than 50 hours among men (refer to figure ES.4). However, there is an earnings gap even after accounting for this difference in hours. The hourly earnings among the self-employed are higher among men than women (refer to figure 2.5). The largest gap is in rural areas. The earnings gap among self-employed men and women in nonagricultural sectors is driven by the self-employed in the industrial sector. The gap exists despite the fact that the educational profiles of men and women in self-employment are similar. Most have only primary education. Women’s businesses are informal and home-based Among the businesses of the self-employed in urban and rural areas, women are significantly more likely than men to operate informal enterprises. Almost all women’s businesses are unregistered, and few maintain written accounts (refer to figure 2.7, panels a and b). Women in industry and services are equally likely to operate informal enterprises, while men in services are more likely to run informal enterprises. 2 | Deep Dives 63 Figure 2.7: Formality and a. Share of self-employed individuals without business registration location of nonagricultural businesses, by region and sex of owner of enterprise, ages 100% 91 15-64 84 80% Female Percentage (%) 57 Male 60% 46 40% 20% 0% Rural Urban b. Share of self-employed individuals without written accounts 100% 77 80% 65 Percentage (%) 60% 54 40 40% 20% 0% Rural Urban c. Share of self-employed individuals working outside the home 100% 95 93 80% Percentage (%) 60% 40% 28 17 20% Source: Original figure for this publication based on calculations using data of LFS (Labor Force Survey) 2022, Bangladesh 0% Bureau of Statistics, Dhaka, Bangladesh, accessed through SARRAW (South Asia Rural Urban Raw) (database), South Asia Region Team for Statistical Development, World Bank, Washington, DC. Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 64 In addition to differences in formality, self-employed men and women also exhibit gender differences in location of work. Almost all nonagricultural self-employed men in rural and urban areas work outside the home (93 percent and 95 percent, respectively) (refer to figure 2.7, panel c). However, among nonagricultural self- employed women, less than a fifth in rural areas and less than a third in urban areas work outside the home (17 percent and 28 percent, respectively), likely reflecting mobility and childcare constraints. Moreover, evidence from Pakistan suggests that social norms also influence women’s preference for operating a business from the home, despite the associated reduction in economic benefits (Said et al. 2022). These patterns suggest high potential for interventions promoting women’s mobility and raising their market access, including through digital tools connecting self-employed women to market information, suppliers, and so on, that they cannot access easily from within the home. Part 3 addresses this and related issues, constraints, and policy solutions. 3 PART Constraints and Policy Tools Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 66 3.1 Overview Part 3 examines the various constraints to women’s economic empowerment in Bangladesh and presents policy tools aimed specifically at addressing these constraints. The discussion on each constraint starts with a presentation of findings from the latest relevant microdata (where available) and descriptive evidence in the literature. This is followed by a review of high-quality causal evidence on the impact of various policy tools. For each constraint, the main policy tools that have the potential to promote women’s economic empowerment are then surveyed, and the evidence on each policy tool are assessed. The resulting policy toolkit builds on a previous toolkit introduced by the World Bank (2024g) and follows the same methodology. The review of the literature is global and covers studies in Bangladesh, other countries in the South Asia region, and countries outside South Asia. A total of 235 high-quality studies have been included in the review. Studies evaluating the impact of similar interventions are classified together under each policy tool. While the studies within a policy tool share the same type of intervention (such as the provision of childcare), they may differ in key design features (for instance, the provision of subsidized formal childcare or childcare at the workplace). The choice of a tool should be informed by the local context and its policy considerations. The review includes only studies that meet specific inclusion criteria. If it is to be considered a high-quality impact evaluation, a study must measure the causal impact of a policy tool against a suitable comparison group. (Refer to appendix B, section B.3 for details on the methodology.) Where relevant, ongoing work by SAR GIL to fill evidence gaps is highlighted. A common theme running through all the constraints to women’s empowerment is the existence of powerful gendered social norms that restrict women’s participation in the labor market. Interventions aimed at changing such norms while addressing women’s potentially internalized beliefs fostered by these norms also show promise. For example, attendance at workshops that promote women’s generalized self-efficacy or aspirations, the provision of accurate information about the views of peers, the distribution of motivational videos on women role models, and interventions aimed at correcting the beliefs of husbands and wives about social sanctions are valuable tools in targeting gender norms in a way that can improve women’s economic empowerment (Ahmed et al. 2024; Aloud et al. 2020; Bernhardt et al. 2018; Bursztyn, González, and Yanagizawa-Drott 2020; Lowe and McKelway 2021; McKelway 2025; Orkin et al. 2023). A guide on the use of the toolkit is provided in box 3.1. Box 3.2 offers an example of the application of the toolkit. 3 | Constraints and Policy Tools 67 Box 3.1 A Guide to Using the Toolkit Step 1 / Identify the constraint or issue: On the list of policy priorities, identify the specific constraint to women’s economic empowerment to be addressed. Step 2 / Review the policy tools: → Consult the list of policy tools under the specific constraint. → Review the summary of the evidence associated with each of the relevant tools. Step 3 / Review the assessment of the evidence: The available evidence on each of the policy tools is classified according to the system in table B3.1.1 Table B3.1.1 Classification system: evidence on the policy tools Conclusive More than 10 high-quality studies evaluate the tool and reveal effects that tend in the same direction. Mixed More than 10 high-quality studies evaluate the tool and reveal effects in different directions. Emerging There are 3 to 10 high-quality studies that evaluate the tool, regardless of the direction of the effects. The amount of evidence on this tool is growing, but is not sufficient to determine conclusively the direction of the effects. Limited Two or fewer high-quality studies evaluate the tool, regardless of the direction of the effects. The evidence is limited and insufficient to determine conclusively the direction of the effects. No evidence No high-quality studies evaluate the impact of the tool, but, based on existing suggestive or nonexperimental evidence or the theoretical potential of the tool to address the constraint, the tool has the potential to close gender gaps. Step 4 / Determine if specific evidence is available on Bangladesh in the case of the particular policy tool; if not, examine the possibility of generalizing to Bangladesh from the other evidence: Refer to the list of references provided as evidence on Bangladesh. Note that the assessment of the evidence presented here refers to global evidence. The impacts of a specific policy tool may thus not translate to Bangladesh to the same extent because of contextual differences. Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 68 Box 3.2 An Example of the Application of the Toolkit Step 1 / Identify the constraint or while examples of employer-provided issue: A policymaker is interested in transport include pickup and drop off boosting female labor force participation using vans. by addressing safety concerns in public transportation (for this example, refer to Step 3 / Review the assessment of table 3.2). the evidence: The strength of the global evidence on this policy tool is indicated as Step 2 / Review the policy tools: emerging. This means that the evidence on the usefulness of the tool is growing, but is a. The policy tools aimed at improving not yet sufficient to determine conclusively women’s mobility identified in this the direction of the effects. toolkit (refer to table 3.2, panel a) are as follows: Step 4 / Determine if specific evidence is available on Bangladesh in the case → Segregated public transport and of the particular policy tool; if not, employer-provided transport examine the possibility of generalizing → General expansion of public to Bangladesh from the other evidence: transportation, without a woman- Table 3.2 indicates that evidence on the only component impact of this tool is available for both → Provision of bicycles Bangladesh and other countries in South → Gender-based violence prevention Asia. The list of references provided in and gender sensitivity training among the note to table 3.2 indicate that the service providers and commuters Bangladesh-specific study is Buchmann, Meyer, and Sullivan (2024). Studies on this Based on the summary of the global policy tool in other countries in the region evidence for each of these tools, the tool include, on Pakistan, Cheema et al. (2022); that has been shown to increase women’s Garlick, Field, and Vyborny (2025). employment by reducing safety concerns is the first one, segregated public transport Because there is limited evidence on and employer-provided transport. Bangladesh, examine how well the results of studies from other countries b. The summary of the global evidence might translate to Bangladesh given points to the potential positive impacts the similarities and differences between of this policy tool on female labor force Bangladesh and other countries. For participation through the associated a practical framework for determining reduction in harassment and an how well results from one context might increase in perceived safety. Examples translate to another context, refer to Bates of woman-only transport include buses and Glennerster (2017). and dedicated cars on the subway, 3 | Constraints and Policy Tools 69 Credits: Dominic Chavez \ World Bank 3.2 Skills and Vocational Training A lack of employable or business skills can be a key constraint on raising female labor force participation. Policy experts ranked the lack of skills and vocational training as the most important constraint to increasing female labor force participation in Bangladesh. A recent World Bank Bangladesh Development Update notes a significant disconnect between the skills of graduates and the skills demanded by the labor market in Bangladesh (World Bank 2024a; Khatun et al. 2022). Employers report substantial gaps in essential employable skills. Similar to the situation in other countries in South Asia, but unlike the case of most developing countries, the relationship between education and labor force participation is not linear in urban Bangladesh. Instead, it follows a U shape (Mahmud and Bidisha 2018). This indicates that only intermediate educational attainment is not sufficient to improve labor market outcomes among urban women. If young women are constrained by a lack of the skills required by the labor market, then TVET programs and on-the-job training programs—the two policy tools identified to address this constraint—may have positive impacts on women’s employment.42 These tools are used extensively as policy levers in developing countries, but are associated with challenges. Improving their impact could thus have particularly high returns in women’s economic empowerment in Bangladesh. 42 The evidence discussed under this policy priority excludes training among the self-employed. Such training is covered under the policy priority on business ownership and business growth. The evidence reviewed also excludes policy tools to promote foundational learning through the school system, which falls outside the scope of this report. Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 70 Technical and Vocational Education and Training Programs There has been substantial government engagement across the globe in integrating TVET programs into the formal education system to raise employment and earnings. Despite the popularity of TVET programs and their potential for workforce development in lower-middle-income countries, such programs face systemic challenges because of the many learners with weak foundational skills, the many teachers lacking industry experience, and the insufficient incentives to provide learners with industry-relevant knowledge (World Bank, UNESCO, and ILO 2023). This includes the poor alignment of formal education with technical training and the insufficient integration of programs within the workplace (Rahman, Farooq, and Salim 2021). The gender disparities in attendance, graduation, and labor market returns to TVET remain wide (World Bank, UNESCO, and ILO 2023). This demonstrates the need for gender-responsive design that addresses the specific barriers faced by women. A recent meta-analysis of evidence on the impact of vocational training programs on youth labor market outcomes among men and women in lower-middle-income countries identified design features that improve effectiveness. First, vocational training programs that combine classroom- and workplace-based components are more effective than interventions that include only one of these two. Second, shorter less-intensive programs (under 400 hours) are more effective than longer more-intensive programs. Third, interventions that involve the private sector, NGOs, or national or international organizations in design or implementation are not necessarily more effective than interventions involving only the public sector. This may be because of the weakness in lower-middle-income countries of the institutions required for public-private collaboration (Ghisletta, Kemper, and Stöterau 2021). More than half a million females are enrolled in some form of TVET in Bangladesh (30% of total TVET enrollment) (BANBEIS 2024), with the share of women in short- term formal training courses varying greatly by field of study. A tracer study of employment outcomes among TVET graduates in Bangladesh six months after graduation suggests that the success rates in finding wage employment and better salaries are higher among men than women. Employment varies widely by field of study and type of institution (Nakata, Rahman, and Rahman 2017; World Bank 2015). This is especially important in designing trade-specific policies given the gender segregation in fields such as automotive mechanics (man-dominated) and garment production (woman-dominated). 3 | Constraints and Policy Tools 71 Policy Tools and Causal Evidence on Their Impact The high-quality impact evaluations reviewed under the skills enhancement and vocational training tool cover training programs that provide vocational or soft skills at varying levels, from teaching a single skill, such as tailoring, to a few hundred women in Pakistan to national vocational training programs on a broad range of vocations to thousands in Colombia and Türkiye. The evidence shows mixed results. The impacts are often positive, but difficult to detect (Carranza and McKenzie 2024; Tripney et al. 2013). The outcomes depend greatly on program design and contextual factors. Several studies report positive effects—ranging from modest to substantial—on women’s labor force participation, productivity, and earnings (Acevedo et al. 2020; Adhvaryu, Kala, and Nyshadham 2023; Adoho et al. 2014; Da Mata, Oliveira, and Silva 2025; Ibarrarán et al. 2014; Maitra and Mani 2017). Most of these studies target specific fields that tend to attract more women, such as tailoring services, and are contextualized to facilitate improvement in women’s outcomes. This is in line with the discussion of the World Bank (2015) on the importance of the field of study. Impacts are evident in both the short and long run (Acevedo et al. 2020; Attanasio, Guarín, et al. 2017). Some programs enable women’s entry into man-dominated sectors or promote income diversification (Chun and Watanabe 2012; Croke, Goldstein, and Holla 2017). In Bangladesh, a BRAC program that targeted disadvantaged unemployed youth showed stronger employment impacts on women than on men (Das 2021). Women moved from casual work to self-employment, while men moved from casual work to wage employment. There was also a sustained increase in earnings up to almost two years after the training. Specifically, women increased the time they spent on tailoring. These impacts were driven primarily by unmarried women. TVET programs appear to be more effective among women if they integrate gender-responsive features, such as soft skills training, market linkages, and tailored mentorship. These features address the specific barriers women face in accessing and benefiting from training in the context of, for example, lower baseline confidence, limited mobility, or gendered labor market norms. For instance, soft skills training can enhance women’s employment prospects, self-esteem, and psychosocial well-being. This is clear in programs such as the Economic Empowerment of Adolescent Girls and Young Women Program in Liberia and garment sector training in India (Adhvaryu, Kala, and Nyshadham 2023; Adoho et al. 2014). Interventions on market linkages, such as a tailoring program in Pakistan (Cheema et al. 2019), help women translate skills into income and civic engagement, especially in contexts where gender norms restrict women’s mobility or access to markets. Tailored support also helps shift occupational aspirations, as demonstrated in Nigeria where women with initial biases against professional roles were more likely to enter the information and communication technology sector Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 72 after training (Croke, Goldstein, and Holla 2017). Add-ons to these programs can also be useful. Using a randomized controlled trial in Bangladesh, Nakata et al. (2021) find that a recognition of prior learning process providing certification for the lowest level of technical skills increased the probability of success in finding employment and better earnings and improved worker confidence.43 Indeed, the effect was more substantial among women than men. Program designs that are not gender responsive often fail to deliver benefits for women and sometimes even exacerbate disparities. In Argentina and Colombia, for instance, women realized limited or no gains in employment or soft skills despite their participation in training (Alzúa, Cruces, and Lopez 2016; Barrera- Osorio, Kugler, and Silliman 2023). Numerous studies reveal only limited impacts on women’s labor market outcomes if programs are gender-neutral or lack relevant complementary components (Barrera-Osorio, Kugler, and Silliman 2023; Hardy et al. 2019; Hirshleifer et al. 2016; McIntosh and Zeitlin 2022). These findings underscore that gender-responsive design is not an optional add-on, but a foundational element for achieving equitable impact in TVET. These mixed findings suggest the need for rigorous, context-specific research on TVET program design to identify best practices and enhance employment outcomes among women. Recognizing this need, especially among young women, the World Bank is collaborating with the government of Bangladesh to build marketable skills, ease the school-to-work transition among youth and polytechnic graduates, and facilitate informal apprenticeships with master craftsmen and craftswomen.44 SAR GIL is also exploring a possible collaboration with local job platforms aimed at blue-collar jobs to evaluate the impact of the platforms on job search and matches among women TVET graduates. (For a literature review by SAR GIL on skills acquisition in South Asia, refer to World Bank 2021.) 43 Recognition of prior learning is a process that assesses and validates the skills of individuals and the knowledge gained through informal or nonformal learning and matches the skills and knowledge against industry or educational standards. The process helps formalize existing capabilities through structured methods, such as interviews, portfolios, or tests, ensuring alignment with the national qualification framework. 44 Refer to ASSET (Accelerating and Strengthening Skills for Economic Transformation Project) (dashboard), World Bank, Washington, DC, https://projects. worldbank.org/en/projects-operations/project-detail/P167506; RAISE (Recovery and Advancement of Informal Sector Employment Project) (dashboard), World Bank, Washington, DC, https://projects.worldbank.org/en/projects-operations/project-detail/P174085. 3 | Constraints and Policy Tools 73 On-the-Job Training Policy Tools and Causal Evidence on Their Impact A nascent strand of literature is focused on on-the-job training. Studies evaluating on-the-job training programs in Bangladesh show promise (refer to table 3.1). An experimental study of 27 large garment factories, one-third of which had no women supervisors and among which the share of women supervisors did not exceed 20 percent, found that training in attitudes and aptitude supervisory skills may create an avenue for the promotion of women to supervisory roles (Uckat and Woodruff 2020). A study that evaluated the impact of a career promotion program in the garment industry found positive effects on women’s agency over household income (Uckat 2023). Such programs provide a potential avenue for firms and other institutions focused on the private sector, for example, International Finance Corporation, to foster retention and career advancement among women employees. Table 3.1: Policy tools for skills and vocational training Policy tool Main takeaways Evidence Evidence Assessment on from other of global Bangladesh South evidence Asian countries Skills The evidence on skill enhancement, Yes Yes Mixed enhancement including vocational skills training, and vocational soft skills training, core financial training literacy training, and apprenticeships, is mixed. While some evidence shows positive impacts on outcomes, such as employment, earnings, and formal employment, other evidence shows that the impacts of training programs with gender-neutral designs are often muted. Programs with gender-responsive elements, such as soft skills, market linkages, and awareness training, show the highest potential for improving outcomes among women. On-the-job While more research is needed, the Yes No Limited training limited evidence shows that skills enhancement targeted at career advancement and on-the-job skills training can be effective in advancing women’s economic empowerment. Source: Original table for this publication. Note: The assessment of the global evidence on the tools under this policy priority differs from the assessment in World Bank (2024g) because of the inclusion of additional literature and the reclassification of some studies. For more information on the studies on Bangladesh involved in this review, refer to Das (2021); Nakata et al. (2021); Uckat (2023); Uckat and Woodruff (2022). Other studies reviewed include Acevedo et al. (2020); Adhvaryu, Kala, and Nyshadham (2023); Adoho et al. (2014); Attanasio, Guarín, et al. (2017); Alzúa, Cruces, and Lopez (2016); Alzúa et al. (2021); Barrera-Osorio, Kugler, and Silliman (2023); Calderone et al. (2022); Chakravarty et al. (2019); Cheema et al. (2019); Chun and Watanabe (2012); Croke, Goldstein, and Holla (2017); Da Mata, Oliveira, and Silva (2025); Elsayed, Hempel, and Osman (2018); Elsayed, Namoro, and Roushdy (2021); Field et al. (2021); Hardy et al. (2019); Hirshleifer et al. (2016); Honorati (2015); Ibarrarán et al. (2014); Kugler et al. (2022); Maitra and Mani (2017); McIntosh and Zeitlin (2022); Rosas, Acevedo, and Zaldivar (2022). Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 74 3.3 Transport, Mobility, and Safety Limited mobility and concerns about safety during transport and in public areas, the workplace, and educational institutions can severely restrict women's involvement in economic and social activities. These concerns are intensified by the lack of laws in Bangladesh to protect women from gender-based violence (World Bank 2024e). The policy experts surveyed ranked mobility restrictions and safety concerns as the second most important constraint on improving labor market outcomes among women in Bangladesh. Moreover, the Bangladeshi experts interviewed consistently ranked transport, mobility and safety as the most important constraint, citing its relevance for women across all sectors and all parts of society. Transport, Mobility, and Safety during Transport Work outside the home is more limited among women than among men in Bangladesh. However, the patterns differ substantially across urban and rural areas. In urban areas, LFS data show that more than two-thirds of employed women work outside the home (69 percent). Widely accepted social norms and safety concerns can also limit mobility in rural areas, thereby negatively affecting skills acquisition and educational attainment among women (Cheema et al. 2022; Jacoby and Mansuri 2015). Most employed rural women report that they work on farms, plots, or rivers (67 percent), and a quarter work inside or next to their homes. Among rural women employed in nonagricultural sectors, a little more than half (54 percent) work outside the home. The only category of employed women who work outside the home at similar rates to men in both rural and urban areas are wage employees. The lack of safety during the commute to the workplace also constrains women’s labor market involvement. In a survey of women working in agroindustries in Bangladesh, almost two-thirds (64 percent) reported that they had experienced verbal and emotional violence during the commute to work. Another 11 percent had experienced physical or sexual violence. Almost three-fourths of the women (73 percent) also agreed or strongly agreed that an unsafe commute between home and the workplace adversely affects female labor force participation, indicating that unsafe transport is a major barrier to women’s employment (World Bank 2025d). The lack of safety during commutes also limits choices and can lead to penalties in the labor market among women. In Brazil, Sharma (2023) finds that a reduction in wages is less likely to cause women to leave their current employer if the potential new job requires an unsafe commute. 3 | Constraints and Policy Tools 75 Policy Tools and Causal Evidence on Their Impact The overall evidence base on policy tools for improving women’s mobility in public spaces, including safe transport, is limited. The review has revealed gaps that may guide future research. The report identifies five categories of tools supported by varying levels of evidence. It also highlights a category that has not yet been tested despite its recognized importance in the descriptive literature, gender- based violence prevention and gender sensitivity training for service providers and commuters (refer to table 3.2, panel a). Table 3.2: Policy tools for transport, mobility, and safety Policy tool Main takeaways Evidence Evidence Assessment on from other of global Bangladesh South evidence Asian countries a. Tools aimed at increasing mobility and safety in transport Segregated Woman-only transport (buses, dedicated Yes Yes Emerging public transport subway cars) and employer-provided and employer- transport (such as pickup and drop-off provided vans) have the potential to overcome transport mobility barriers by reducing harassment and raising perceived safety in rural and urban areas, leading to greater female labor supply. General Expansion of public transport to new No Yes Emerging expansion areas increases college enrollment and of public employment among women in those transportation, areas because of shorter commutes. without a women-only component Provision of Providing bicycles to girls to travel No Yes Emerging bicycles to school can increase school enrollment and reduce harassment and victimization. The creation of subsidies or free provision has the same impact. The evidence also points to possible negative consequences through increased early marriage and teenage pregnancy. Gender-based Pilot evidence suggests that a bystander No No No evidence violence program, along with an easier reporting prevention system for harassment and other and gender offenses, may lead to an increase in sensitivity subway rides by women. No causal training among evidence exists. service providers and commuters b. Tools aimed at increasing safety in public spaces Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 76 Table 3.2 (continued) Policy tool Main takeaways Evidence Evidence Assessment on from other of global Bangladesh South evidence Asian countries Media campaigns Information campaigns delivered No Yes Limited through targeted videos or social media can increase knowledge on women's safety issues, but do not always lead to change in practices or attitudes or more support for women. Hot spot policing Increasing the visible presence of law No Yes Limited enforcement and effective handling of harassment in public places reduce sexual harassment. Encouraging Pilot evidence shows that a bystander No No No evidence proactive program and an easier reporting system bystander lead to a reduction in sexual harassment behavior in the subway. Infrastructure Street lighting and other improvements No No No evidence improvements in in urban public spaces are correlated hot spots with improved commuter satisfaction among women. c. Tools aimed at increasing safety in the workplace and in educational institutions Flexible working Flexible working options (e.g., flexible Yes Yes Emerging options, safe working hours, working from home) and workplaces, and female-friendly amenities at work (e.g., female-friendly quality and safety of accommodation workplace facilities) have the potential to increase amenities female labor supply and increase retention of female workers by allowing women to work within social constraints of mobility and household responsibilities. In-school Training programs among teachers and No Yes Emerging antiviolence students on awareness, reporting, and training among the response to violence or harassment teachers and lead to fewer instances of violence, students increase reporting rates, and improve mental health outcomes among victims. Training that is in-person and incorporates empathy-building components is especially effective. Source: Original table for this publication. Note: For more information on the studies on Bangladesh involved in this review, refer to Buchmann, Meyer, and Sullivan (2024) (for panel a) and Boudreau (2024) (for panel c). Other relevant studies reviewed include Aguilar, Gutiérrez, and Villagrán (2021); Alba-Vivar (2024); Amaral et al. (2025); Atkin, Schoar, and Shinde (2023); Borker (2022a), (2022b); Boudreau et al. (2023); Cheema et al. (2022); Chen et al. (2024); Christensen and Osman (2021); Christia et al. (2023); Corradini, Lagos, and Sharma (2025); Dean and Jayachandran (2020); Donald and Grosset (2024); Fiala et al. (2022); Garcia-Hernandez, Prakash, and Steinert (2025); Garlick, Field, and Vyborny (2025); Gibbs, Mengel, and Siemroth (2023); Gutierrez, Molina, and Ñopo (2018); Ho, Jalota, and Karandikar (2024); Jalota and Ho (2024); Karmaliani et al. (2020); Kjelsrud, Mitra, and Moene (2024); Kondylis et al. (2025); Lei, Desai, and Vanneman (2019); Martínez and Perticará (2017); Molina and Tanaka (2023); Muralidharan and Prakash (2017); Seki and Yamada (2020); Sharma (2024); Smarrelli (2023). 3 | Constraints and Policy Tools 77 The first policy tool is segregated public transport and employer-provided transport, a temporary solution—a stop gap—to the issue of the lack of safety in transport and the limited acceptance of women’s mobility in South Asia, including in Bangladesh. The evidence on this tool is emerging. Only one study has appeared on Bangladesh so far. It evaluates the impact of employer-provided transport and finds positive impacts on female labor force participation (Buchmann, Meyer, and Sullivan 2024). Evidence on other countries in the region shows that the impacts of similar interventions are positive. In urban Pakistan, the provision of gender-segregated transport increased women’s job application rates, including among women who were not searching for work before the woman-only transportation was made available. By contrast, providing mixed-gender transport did not affect the rate of women’s job applications, indicating that safety and social acceptability, not merely convenience or cost, may be the binding constraint on women’s involvement in the labor market in this case (Garlick, Field, and Vyborny 2025). A study on rural Pakistan reveals the existence of a boundary effect, that is, women are less likely to travel outside their villages than to cover the same distance within their villages. The study finds that higher subsidies are required to induce women to overcome the boundary effect and attend skills training sessions (Cheema et al. 2019). The emerging evidence on the effects of segregated public transport and employer- provided transport on female labor market outcomes points to an important constraint faced by women in both urban and rural areas in South Asia, the lack of safety in transit and the associated low acceptance of women’s movement across public spaces. This can impose severe restrictions on women’s mobility and labor market involvement, for example, by excluding women from wage employment and restricting them to home-based work. However, the potential for unintended negative consequences to arise from segregated transport calls for caution in using this policy as a permanent tool. Kondylis et al. (2025) find that, while women riding in women-only metro cars in Brazil experience less harassment, such women-only spaces may also reinforce regressive gender norms, particularly the belief that women’s proper place is in the women-only space and that women traveling in nonsegregated spaces are more sexually open. Such beliefs may also strengthen victim-blaming norms. In Bangladesh, LFS 2022 data show that only about a 10th of women working in services (11 percent) and a 5th working in industry (22 percent) receive transport from their employers. The numbers are similar across both rural and urban areas. The emerging evidence on the impact of the second tool—general expansion of public transportation without a women-only component—indicates that enhancements in affordability and connectivity in public transit can increase women’s employment. In India, the provision of free bus transit to women has had different impacts based on the skill levels of the women. While the labor supply of skilled employed women was expanded, the labor supply of less highly skilled married women narrowed, and these women spent more time on household chores Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 78 (Chen et al. 2024). Lei, Desai, and Vanneman (2019) show that improving bus connectivity in Indian villages may have the potential to raise women’s involvement in nonfarm work. Proximity to the Delhi metro also boosted women’s employment (Seki and Yamada 2020). There is also evidence that providing nonsegregated group transport to the place of work of someone in a woman’s network increased employment in Côte d‘Ivoire, but providing mixed gender transport in Pakistan had minimal effects on job application rates among both men and women (Donald and Grosset 2024; Garlick, Field, and Vyborny 2025). Qualitative research conducted by SAR GIL suggests that bus service providers in Dhaka believe that boarding women on buses is significantly more time- consuming than boarding men. For the providers, this means men are more attractive in meeting daily passenger targets. Bus conductors who intervene in cases of harassment face penalties from drivers (who serve as the superiors of the conductors) because of delays and perceived financial losses, reducing the incentive to support victims of harassment (World Bank 2025f). Evidence indicates that the third tool, the provision of bicycles, may have the potential to raise women’s mobility to the workplace through, for example, government or employer subsidies. While the evidence on the impact of the provision of bicycles to women to facilitate the commute to the workplace is lacking, there is emerging evidence showing that the provision of bicycles to girls to travel to school has positive impacts. Providing bicycles to girls conditional on school enrollment or attendance has been shown to increase enrollment in secondary school, participation in a high-stakes examinations, and the probability of passing examinations in India (Muralidharan and Prakash 2017) and to increase enrollment and self-reported empowerment in Zambia (Fiala et al. 2022). The long- run effects of these interventions paint a more nuanced picture of the impact of such programs. In India, the program also had spillovers, enhancing various empowerment indicators for nontargeted adult women in households with girls of eligible age. While the mechanism of this increase in empowerment is unclear, it points to potential impacts on individuals beyond the direct beneficiaries of the bicycle program (Kjelsrud, Mitra, and Moene 2024). In Zambia, while the program reduced the rates of self-reported domestic violence, it also increased the rates of early marriage and teenage pregnancy five years after the intervention. These findings point to the importance of accounting for the long-run impacts of interventions, especially potential unintended consequences, and of evaluating programs for their long-run impacts before scaling up. 3 | Constraints and Policy Tools 79 Credits: Mohd.Ashabul Haque Nannu \ Pexels Safety in Public Spaces MICS 2019 data show that a fifth of urban women in Bangladesh feel unsafe walking in their neighborhoods after dark. The share jumps to 27 percent in rural areas. Women in low-income areas in Dhaka are significantly less likely than men to feel safe. This perception gap is negatively correlated with labor market outcomes. Women who feel safe are far more likely to be economically active, work beyond their neighborhoods, and pursue broader opportunities (Ahmed and Kotikula 2021). Safety in public spaces is a concern among women globally. A rise in gender-based violence has influenced women's capacity to enter the workforce or restricted their working hours in Mexico (Velásquez 2020). Borker (2021) finds that, instead of enrolling in distant, higher-quality colleges, women college students in New Delhi enroll in lower-rated colleges that are located in safer areas closer to home. This leads to substantial costs. The women students who choose the lower- quality colleges experience an estimated 17 percent penalty in postcollege lifetime earnings. The study also estimates that women are willing to incur costs of up to Rs 17,500 a year—double the average annual tuition at the University of Delhi—to travel by safer routes to receive education. Even media coverage of violence against women is linked to decreased female labor force participation in India (Siddique 2018). Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 80 Policy Tools and Causal Evidence on Their Impact Good evidence on tools to raise women’s safety in public spaces is scarce. The report identifies two categories of tools supported by any evidence and two more that have not yet been tested, but are hypothesized to be effective in improving women’s safety (refer to table 3.2, panel b). Evidence on the effectiveness of media campaigns and hot spot policing is limited. A study that examines hot spot policing in urban India finds that, although police patrols do not reduce street harassment, the presence of uniformed officers led to a 27 percent drop in severe harassment (such as forceful touching and intimidation). The patrols also reduced the likelihood that women leave high-risk areas because of harassment. The visible presence of the police officers is key. The presence of undercover officers who issued more sanctions and warnings than uniformed officers did not have an impact on the incidence of street harassment (Amaral et al. 2025). Two additional tools that could potentially increase women’s safety in public spaces are not currently supported by good evidence in developing countries. Ongoing research in other contexts and studies that find a potential role of bystanders suggest that initiatives focused on encouraging proactive bystander behavior might improve women’s safety, while infrastructure improvements in hot spots, particularly street lighting, could deter aggressive behavior (Amaral et al. 2024, 2025; Coly and Suteau 2025; Sharma 2024). Safety in the Workplace and in Educational Institutions Innovative survey designs aimed at capturing the true prevalence of harassment in the workplace reveal that harassment in garment factories in Bangladesh is widespread. Boudreau et al. (2023) find that more than half the work teams in garment factories included at least one worker who had been threatened; 38 percent included someone who had been sexually harassed; and almost a third included someone who had been physically harassed. They also provide a method to measure rates of harassment in organizations more reliably. Policy Tools and Causal Evidence on Their Impact The evidence on the tools able to improve women’s safety in the workplace and in educational institutions is growing and promising (refer to table 3.2, panel c). Tools aimed at ensuring flexible working options, safe workplaces, and female-friendly workplace amenities have the potential to boost female labor supply, including by allowing latent women workers to join the labor force (Ho, Jalota, and Karandikar 2024; Jalota and Ho 2024). Good evidence on impacts, especially the impacts in the design of flexible work options, is emerging. While flexible working conditions, such 3 | Constraints and Policy Tools 81 as the option to work from home, can enhance women’s labor market outcomes by addressing safety concerns in the workplace, they may also ease dependent care and the time constraints on women. Better legislation and policies to protect workers also improve the empowerment and well-being of women workers. For instance, studies have found that, in Bangladesh, more effective occupational safety and health committees—that is, committees representing workers by advocating for safer working conditions within an organization—improved safety in apparel factories without affecting productivity, wages, or employment (although worker satisfaction fell slightly), but only in factories with better management practices (Boudreau et al. 2023). A growing body of literature on interventions aimed at teachers and students in schools and colleges is promising. Interventions providing in-school antiviolence training among teachers and students are lowering the incidence of violence, improving reporting, and enhancing the responsible management of cases (Gutierrez, Molina, and Ñopo 2018; Karmaliani et al. 2020; Smarrelli 2023). The evidence shows that in-person sexual harassment awareness training with an empathy-building component targeted at men can be effective. In-person training in colleges in New Delhi that raised the perceived cost of engaging in harassment by making individuals aware of the social disapproval of peers and of the possibility of bystander support for the victims reduced sexual harassment and increased women’s labor market search activities without affecting mental well-being or school test scores. Unlike most sexual harassment training, which typically involves only an awareness-building component, the training that has been tested during the current study had empathy-building components. This included discussions on the impact of sexual harassment on women that relied on anonymous narratives by women and that aimed at reducing backlash effects (Sharma 2024). However, further study is needed on the integration of content on safety and gender norms in mainstream school curricula and workplace training programs. 3.4 Financial Inclusion The expert survey conducted as part of this study ranked the lack of financial inclusion as the third most important constraint. Addressing this constraint has the potential to boost female labor force participation and women’s earnings in Bangladesh. Findex 2021 data shows that there is scope for greater financial inclusion among women.45 A large share of women in Bangladesh (57 percent) have no bank or mobile money account, compared with 37 percent of men. According to HIES 2022, among the working-age population (ages 15–64), women are more 45 Refer to Global Findex (Global Findex Database), World Bank, Washington, DC, https://www.worldbank.org/en/publication/globalfindex/archive. Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 82 likely than men to have taken at least one loan (16 percent vs. 13 percent). Among the rural working-age population, women are even more likely than men to have taken at least one loan (17 percent vs. 14 percent). Conditional on having at least one loan, women borrow at higher interest rates, paying an average of 13 percent interest, while men pay 9 percent interest. A substantially larger share of men than women have loans with no interest (26 percent vs. 6 percent). The higher interest rates on loans among women are partly explained by the different sources of the loans taken by men and women. Among individuals who have taken at least one loan, 88 percent of women rely on NGOs, which are considered a formal source of credit, while only 58 percent of men do so; 18 percent of men use informal sources, such as relatives, friends, or neighbors, compared with 5 percent of women (refer to figure ES.6). Grameen Bank, an NGO, is a common savings location among rural women. Microfinance institutions have expanded the access to credit among women, but the size and terms of the available credit are limited. That women draw substantially less on informal sources of credit than men is consistent with recent evidence elsewhere in South Asia. Thus, Binzel et al. (2025) find that women’s informal borrowing capacity in rural India is half the capacity of men. The reliance of women on formal sources of credit, especially microcredit, also points to a potential source of the differences in earnings among self-employed men and women. Field et al. (2013) provide evidence on India showing that the short repayment periods imposed by microfinance institutions limit the ability of borrowers to invest in businesses or take risks that involve higher returns. In Bangladesh, BIHS 2018-19 data show that more rural women (33 percent) than rural men (20 percent) have savings and that women have more savings accounts than men.46 This is in line with the evidence indication microfinance institutions in Bangladesh target especially women in expanding access to savings, while requiring clients to maintain savings accounts if they wish to access loans (Solotaroff et al. 2019). Relative to men, women save smaller amounts. For instance, women report that they save an average of Tk 14,000 to prepare for difficult times, while men save an average of Tk 59,000. However, women save more frequently. Among women, 68 percent report that they save weekly, while 57 percent of men report that they do not save regularly. A substantially larger share of rural women maintain savings accounts for the purpose of supporting their eligibility for loans (20 percent of women’s savings accounts vs 8 percent of men’s accounts) (refer to figure 3.1). About a fifth of women (21 percent) borrow to repay other loans. 46 BIHS 2018–19 elicits responses on savings held at different locations, treating them as separate accounts. However, an individual can report multiple savings accounts at the same location (for example, at the same NGO). The data are reported on all household members by a man in the household. 3 | Constraints and Policy Tools 83 Figure 3.1: Distribution of saving “accounts" in rural areas, by purpose of saving & sex, ages 15-64 20 Get Loans 8 Female-owned Male-owned 45 Prepare for Difficult Times 39 18 Future of Children 23 Source: Original figure for this publication based 17 on calculations using data of BIHS (Bangladesh Other Integrated Household Survey) 2018–19, 29 International Food Policy Research Institute, Washington, DC, https://doi.org/10.7910/DVN/ NXKLZJ, Harvard Dataverse, V2. 0 20 40 60 80 100 Percent Note: The category 'Other' includes purposes such as buying household goods, buying agricultural implements, buying land, starting/ helping a business, education/training, marriage/dowry, building/repairing house, lending to others, sending someone abroad for a job, and no specific reason. According to HIES 2022, women borrow substantially lower average amounts per loan on average than men (Tk 70,000 vs. Tk 137,000) (refer to figure 3.2, panel b). This occurs even though men and women borrow for roughly similar purposes with the exception that men are more likely to borrow for business (refer to figure 3.2, panel a).47 The gap is larger for loans taken for economically productive activities, such as support for business enterprises and agricultural inputs. Nonetheless, several studies document that women make larger loan requests (Bangladesh Bank 2014; Singh, Asrani, and Ramaswamy 2016; Solotaroff et al. 2019). Using data from field surveys among 500 women small and medium enterprise owners across 12 districts in Bangladesh, Singh, Asrani, and Ramaswamy (2016) estimate that the finance gap, that is, the difference between the credit women business owners need to grow their businesses and the amount of credit they are able to access, is around 60 percent. Thus, 60 percent of the financing needs of woman-owned small and medium enterprises is unmet by financial institutions.48 47 Loan values are winsorized at the 99th percentile to remove any bias resulting from outlying values. 48 Given the lack of systematic gender disaggregated data, the finance gap is not a perfect measure. For instance, using the same definition, Bangladesh Bank (2014) estimates the gap at closer to 40 percent. The estimates are derived from a survey conducted by Bangladesh Bank from December 2013 to March 2014 with 450 self-employed women in 54 districts across seven divisions. Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 84 Figure 3.2: Distribution of number of loans and loan amount, by purpose of loan and sex, ages 15-64 a. Loan purpose, by sex 20 Business 27 15 Housing 14 14 Food Purchase 15 14 Agriculture 13 7 Health 7 4 Marriage 4 2 Education 2 25 Other 18 0% 20% 40% 60% 80% 100% Percent b. Loan amount, by purpose and sex 300 184 199 Amount (in Thousands of Taka) 168 200 146 120 137 91 86 100 87 74 70 72 62 58 54 52 40 36 0 All Business Housing Food Agriculture Health Marriage Education Other purposes Purchase Female Male Source: Original figure for this publication based on calculations using data of HIES (Household Income and Expenditure Survey) 2022, Bangladesh Bureau of Statistics, Dhaka, Bangladesh, accessed through SARRAW (South Asia Raw) and SARMD (South Asia Regional Micro Database) (databases), South Asia Region Team for Statistical Development, World Bank, Washington, DC. 3 | Constraints and Policy Tools 85 Women’s limited participation in financial decision-making within households may constrain the benefits of access to financial resources. DHS 2022 data show that, while 61 percent of rural and urban women report that they share decision-making with their husbands on household earnings expenditures, only 29 percent in rural areas and 33 percent in urban areas report that they decide alone. Similarly, only about 5 percent of women are the sole decision-makers on major household purchases, and about a quarter report that their husbands make these decisions alone. These dynamics suggest that, even if women are able to gain access to financial services, their autonomy over the use of financial resources is limited, potentially restricting their ability to invest, save, or borrow in ways that align with their own needs and goals. This lack of control may also disincentivize women from taking up paid employment. This underscores the importance of pairing financial inclusion interventions with initiatives that reflect an awareness of social norms and practices, that is, that function within the constraints of existing social norms or that aim to shift norms depending on opportunities in the local context (Burjorjee, El-Zoghbi, and Meyers 2017). Policy Tools and Causal Evidence on Their Impact To address the constraints, this section reviews the literature on the impacts of policy tools aimed at increasing financial inclusion (refer to table 3.3). The focus is on financial interventions that are targeted at individuals who are not in self- employment. The interventions include adoption of financial products and services, savings, and within-household decision-making and empowerment. Table 3.3: Policy tools for financial inclusion Policy tool Main takeaways Evidence Evidence Assessment on from other of global Bangladesh South evidence Asian countries Mobile money Mobile money increases women's Yes Yes Conclusive control over finances and the likelihood of savings. The effect on savings is important because women save at higher rates but lower amounts relative to men. Mobile money has raised female labor supply in a sample of rural-urban migrants in Bangladesh. Easing access Low-cost no frills accounts can provide No Yes Emerging restrictions on women with easier access to banking, banking enabling higher savings. Depositing money directly into bank accounts also allows women to save more by increasing their control over money. Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 86 Table 3.3 (continued) Policy tool Main takeaways Evidence Evidence Assessment on from other of global Bangladesh South evidence Asian countries Commitment Commitment devices, such as No No Emerging savings accounts commitment savings accounts, facilitate savings among women, especially women with stronger present bias, that is, a tendency to prefer a smaller present reward rather than wait for a larger future reward. Switching to Using biometrics for the identification Yes Yes Limited digital payment and delivery of public services can systems improve women's control over resources and financial rights. Changing default options to digital systems (for example, switching from cash to digital payroll) can also spur the adoption and use of digital accounts. Debit cards Access to debit cards reduces the costs No No Limited of accessing savings (by reducing travel time), increasing bank account usage. Debit cards also have the potential to increase formal savings among women. However, debit cards may also be captured by family members, decreasing women's control over money. Door-to-door In contexts where norms restrict No No No evidence banking services women’s mobility, door-to-door services offering products for savings and deposits can raise the financial inclusion of women. However, such services may not be the most cost-effective because of travel and labor costs. Source: Original table for this publication. Note: For more information on the studies on Bangladesh involved in this review, refer to Breza, Kanz, and Klapper (2020); Lee et al. (2021), (2022). Other studies reviewed under this policy priority include Abiona and Koppensteiner (2022); Agarwal et al. (2017); Aker et al. (2016); Ashraf, Karlan, and Yin (2006); Bachas et al. (2018), (2021); Björkegren et al. (2022); Brune et al. (2016); Callen et al. (2019); Clark et al. (2022); Delavallade, Gittard, and Vaillant (2025); de Mel et al. (2022); Dizon, Gong, and Jones (2020); Dupas and Robinson (2013); Dupas et al. (2018); Egami and Matsumoto (2020); Field et al. (2021); Gazeaud et al. (2023); Heath and Riley (2024); Karlan and Zinman (2018); Karlan et al. (2016); Karra et al. (2022); Kikulwe, Fischer, and Qaim (2014); Kipchumba and Sulaiman (2025); Muralidharan and Prakash (2017); Prina (2015); Somville and Vandewalle (2018); Suri and Jack (2016). The global evidence on the use of mobile money accounts to receive transfers and make cash deposits indicates that there are multiple benefits among women. First, mobile money reduces transaction costs, such as avoiding travel to and from ATMs or bank branches for cash. This is especially important if women face mobility constraints, including restrictive social norms or safety issues, or if they save more frequently than men (de Mel et al. 2022; Garz et al. 2020). Second, there is evidence that mobile money allows women to manage who in the household knows about any transfers they receive, ensuring their personal control over their own money(Aker et al. 2016). Third, if women already have mobile money accounts, but do not use them frequently, switching to mobile money as the 3 | Constraints and Policy Tools 87 default for payroll or loan repayments increases women’s use of mobile money for other transactions (Breza, Kanz, and Klapper 2020; Heath and Riley 2024). The evidence on Bangladesh indicates that the impacts are positive in other ways, too. Lee et al. (2021, 2022) find that assistance in the adoption of mobile money by rural-to-urban migrants increases hours of work among women. This aligns with evidence on other countries about the use of mobile money as a means to foster financial inclusion. There is good evidence in Bangladesh on the benefits of another policy tool under review, switching to digital payment systems. Breza, Kanz, and Klapper (2020) show that the switch from cash to digitalized payroll in a garment factory increased savings and transactions at similar levels among both men and women without requiring assistance. While the global evidence on this tool is limited, the availability of Bangladesh-specific evidence is promising. In Pakistan, switching from debit cards to a biometric verification system for payments in a large unconditional cash transfer program substantially increased the likelihood that women would collect and be able to use the cash themselves (Clark et al. 2022). Policymakers should be aware that women’s financial inclusion and women’s digital inclusion are closely connected. For instance, large gender gaps persist in mobile phone ownership in Bangladesh, where women typically share mobile phones with other household members. While some tools have the potential to expand financial inclusion even in the absence of digital inclusion, they are prone to capture by other household members (such as debit cards) or are expensive (such as door- to-door banking services). Moreover, rigorous empirical evidence on these tools is limited or not available. 3.5 Ownership of Property and Other Assets The ownership of productive assets, including property, was ranked as the fourth most important constraint to address in the survey of experts conducted for this study. Property Rights Religious or ethnic norms exclude women in some countries from a share in inheritance rights. In Bangladesh, the legal basis for inheritance is Muslim Personal Law, which allocates an inheritance share to women that cannot legally be willed away, although it is smaller relative to the share going to men in the same household (World Bank 2024e). Furthermore, although inheritance shares are clearly specified in the law, women often receive less in practice and may Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 88 surrender their inherited assets to brothers under familial pressure (Moyeen et al. 2022). Women in Bangladesh rarely own productive assets. DHS 2022 data show that a small share of women (7 percent) in both urban and rural areas report that they own their homes. Among the women who report ownership, more than half own the home jointly with someone else and do not have their names on the property titles. Policy Tools and Causal Evidence on Their Impact The studies covered under the policy tool, amendments in inheritance rights to increase landownership, evaluate the direct impact of legal reforms on land and property ownership among women and on related downstream outcomes, such as outcomes among children and outcomes in economic empowerment and mobility (refer to figure 3.4). An overwhelming majority of the relevant literature is focused on the amendments to the Hindu Succession Act of 1956 in India. Most of this literature finds that the reform improved labor market outcomes, access to education and health care, bargaining power, autonomy, and marriage market outcomes, including the intergenerational persistence of positive impacts on education, time use, health, and reduced violence against women (Amaral 2017; Bose and Das 2021; Deininger, Goyal, and Nagarajan 2010, 2013; Deininger et al. 2019; Heath and Tan 2020; Naaraayanan 2019; Roy 2008; Sapkal 2017; Suteau 2020; Valera et al. 2018). However, a smaller number of studies find that the reform had no effect on inheritance or on female labor force participation (Roy 2015; Suteau 2020). Credits: Freepik 3 | Constraints and Policy Tools 89 Table 3.4: Policy tools for ownership of property and other assets Policy tool Main takeaways Evidence Evidence Assessment on from other of global Bangladesh South evidence Asian countries a. Tools aimed at increasing asset ownership through improved property rights Amendments Amending inheritance rights in India to No Yes Conclusive in inheritance remove bias against women improved rights to increase labor outcomes, education, health landownership care, bargaining power, autonomy, and marriage market outcomes. However, evidence suggests the reform also exacerbated son preference and led to an increase in female child mortality and female feticide. Digitalization of There is no published evidence for No No No evidence landownership this tool. However, forthcoming SAR and boundary GIL evidence on Pakistan shows that records the digitalization and centralization of landownership records, along with biometric verification requirements, reduce the exclusion of women from land inheritance. However, such policies may have unintended consequences among women on the marriage market, along with earlier births and higher school drop-out rates. Market- Market-based land resettlement has No No Emerging oriented land shown mixed impacts on women's land redistribution rights in Malawi. These impacts may be different among women who are heads of household and women living in man- headed households. Dispute Alternatives to customary dispute No No Emerging resolution for resolution institutions (such as formal property rights mechanisms, community-based disputes mechanisms, and free legal aid) have been shown to reduce property disputes, increase women's knowledge of their rights, and boost women's satisfaction with the mediated outcomes. Joint Family law reform in Ethiopia that No No Limited administration of expanded women’s marital property marital property rights and removed restrictions on through family working outside the home increased law reform women's likelihood of working in paid and full-time occupations requiring work outside the home. Price incentives Providing price discounts for joint titling No No Limited for including increased women's landownership women in without reducing demand for titling in landownership Tanzania. Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 90 Table 3.4 (continued) Policy tool Main takeaways Evidence Evidence Assessment on from other of global Bangladesh South evidence Asian countries Community-led Community-involved mapping and No No Limited land demarcation attribution of land rights in Benin of customary encouraged increased long-term landownership investment in land parcels and raised the likelihood that woman-managed landholdings would be left fallow (an important investment in soil fertility). Land tenure Land formalization in Rwanda increased No No Limited regularization tenure security among legally married women, but decreased it among legally unmarried women. It also increased investments in land by woman-headed households by a factor of almost two relative to man-headed households. b. Tools aimed at increasing asset ownership through livelihood asset transfers One-off asset Asset transfer packages consisting of Yes Yes Emerging transfer and livestock and other assets, training, support package and cash stipends targeted at ultrapoor and microfinance households (primarily to women in the and pay-it- household) increase household welfare forward element by improving labor market outcomes, savings, consumption, and poverty rates. Graduation This tool improves women's decision- No No Limited schemes making on children's schooling and health, while it has no effect in other major areas. Source: Original table for this publication. Note: For additional information on the studies from Bangladesh included in this review, refer to Asadullah and Ara (2016); Balboni et al. (2022); Bandiera et al. (2017); Emran, Robano, and Smith (2014); Misha et al. (2019); Rahman, Bhattacharjee, and Das (2021) for Panel b. Other studies reviewed under this policy priority include Ahmed et al. (2009); Ali, Deininger, and Goldstein (2014); Angelucci et al. (2022); Banerjee, Duflo, and Sharma (2021); Bedoya et al. (2019), (2023); Bhalotra, Brulé, and Roy (2019); Blattman, Hartman, and Blair (2014); Bose and Das (2020); Bossuroy et al. (2021); Burchardi et al. (2019); Datar, Ximena, and Hoffman (2009); Deininger, Goyal, and Nagarajan (2010), (2013); Goldstein et al. (2015); Hallward-Driemeier and Gajigo (2015); Harari (2019); Heath and Tan (2019); Janzen et al. (2018); Karimli et al. (2021); La Ferrara and Milazzo (2017); Mendola and Simtowe (2015); Mishra and Sam (2016); Mueller et al. (2014), (2015); Naaraayanan (2019); Rosenblum (2015); Roy (2008); Roy (2015); Sandefur and Siddiqi (2015); Sapkal (2017); Suteau et al. (2020); Valera et al. (2018). Such interventions may also have unintended adverse consequences. For example, Bhalotra, Brulé, and Roy (2019) and Rosenblum (2015) find that the reform exacerbated son preference and led to an increase in female child mortality and female feticide, indicating that strong gender bias persisted after the reform. Bahrami-Rad (2021) finds that the likelihood of cousin marriage increased and the likelihood of working, particularly in agriculture, decreased among women who had been exposed to the 2005 reform of the Hindu Succession Act. Evidence also shows that the reform increased the rates of suicide among both men and women and the incidence of wife beating (Anderson and Genicot 2015). In Pakistan, ongoing work by SAR GIL has found that a biometric verification requirement in the inheritance mutation process increased the probability of de 3 | Constraints and Policy Tools 91 facto inheritance among women, but also shows preliminary evidence of negative impacts on marriage and education outcomes. In particular, after the reform, women married earlier, were more likely to marry at a younger age without choosing their spouse, were more likely to have less educated and lower-earning spouses, and were more likely to have children by their early 20s. In addition, women who were expected to receive large inheritances were more likely to drop out of school. Among the few studies under the market-oriented land redistribution policy tool, several evaluated the impact of Malawi’s Community Based Rural Land Development Project, which provided conditional cash grants to low-income families to help them acquire larger land plots, along with access to agricultural inputs and extension services, and offered both individual and collective land titles. The studies find mixed impacts on property rights among woman-headed households, but less favorable impacts on the rights of women within man-headed households. The limited evidence under the price incentives to include women in landownership tool shows that providing small monetary incentives for joint titling can become a path to increasing women’s landownership. The effects of such policies vary depending on household and marital status. This represents a caution against adopting universal policy recommendations or universal targeting on this issue. Evidence on the impact of customary institutions that govern land rights, although scarce, indicates that women may prefer alternative systems. For example, offering formal dispute resolution for property rights disputes as an alternative to customary dispute resolution institutions in Liberia generated greater awareness among women about their rights, greater satisfaction with the mediated outcomes, and fewer property disputes (Sandefur and Siddiqi 2013). For a review of the literature on property ownership in South Asia conducted by SAR GIL, refer to Zahra, Javed, and Muñoz-Boudet (2022). Livelihood Asset Transfer Programs Policy Tools and Causal Evidence on Their Impact Direct asset transfers targeted at ultrapoor households have demonstrated promise in lifting households out of poverty across various countries. The asset transfers include transfers of productive assets, usually livestock, and are supplemented with support packages that may include training and cash support. Such programs have been shown to raise household welfare by improving labor market outcomes, savings, consumption, and poverty rates. However, evidence specifically on the women in these households is limited. This review focuses on studies that provide evidence of changes in economic outcomes among women, such as labor force participation, hours worked, type of occupation, and so on. Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 92 In Bangladesh, asset transfer interventions such as the BRAC Ultra Poor Graduation programs that provide asset transfers to poor households, particularly the women in the households, offer some evidence of increases in the hours and days worked by women, earnings growth, and enhanced decision-making power. Evidence on long-run impacts, however, is mixed. Consideration of such evidence and impacts is crucial in designing policy tools. Some studies find long-run, positive impacts on women’s economic conditions (Bandiera et al. 2017). Others find that such effects tend to taper off in the long term (Banerjee, Duflo, and Sharma 2021). Male capture of the assets may also occur (Asadullah and Ara 2016). There is also evidence that transfer beneficiaries switch back to their initial occupations, such as work as maids or day laborers, in the long term (Misha et al. 2019). These big push interventions have been tested in several countries. The evidence on Bangladesh is emerging, but robust. Two of the studies reviewed—Angelucci, Heath, and Noble (2022) on the Democratic Republic of Congo and Karimli et al. (2021) on Burkina Faso—also measure the impact of the inclusion of gender-sensitive family coaching and the engagement of men in the graduation program. They find a positive impact on women’s decision-making and a potential reduction in IPV. Credits: Asif Iqbal Hridoy / pexels 3 | Constraints and Policy Tools 93 3.6 Childcare and Home Responsibilities Women in Bangladesh, even if they are employed, spend a large amount of time on housework and the care of dependents. Regardless of their employment status, men spend little time on these activities regardless of their employment status (refer to figure 3.3). Time use data also show that, in both rural and urban areas, young women who are not in employment or education spend most of their nonleisure time on housework and the care of dependents (8 and 7 hours a day, respectively). Among young women in employment or education, young rural women spend more time than young urban women on housework and the care of dependents (5 hours and 3 hours a day, respectively).49 The Bangladeshi experts interviewed consistently ranked childcare and home responsibilities as the second most important constraint, citing its relevance for women across all sectors and all parts of society. Figure 3.3: Time use by not in employment or education (NEE) status and sex, ages 15-64 24 20 16 Average Hours in a Day Source: Original figure for this publication 12 based on calculations using data of TUS (Time Use Survey) 2021, Bangladesh Bureau of Statistics and UN Women Bangladesh, Dhaka, Bangladesh. 8 Note: The employment status of individuals is determined using the module on working status, while time use categories are reported 4 in the time diary module. Nonzero values for work/production for individuals not in employment or education (an average of 2.0 hours among men and 0.7 hours 0 among women per day) arise because of Male Female Male Female inconsistencies in responses between these two modules by individuals; 37 percent of men In Employment or Education Not In Employment or Education and 33 percent of women not in employment or education reported nonzero time values for activities categorized as work/production. Refer to appendix B, section B.1 for the Rest / Leisure / Self-Care Education / Learning mapping of the Major Divisions activity codes of the International Classification of Activities Housework / Dependent Care Work / Production for Time-Use Statistics (UN DESA 2021) to the categories in the figure. EE =in employment or education. NEE = not in employment or education. 49 Lack of data on involvement in training in the 2021 Time Use Survey prevents the identification of individuals who are NEET. Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 94 There is reason to believe that lack of childcare availability constrains women’s labor force participation. Cross-country evidence shows that the enactment of childcare laws increased female labor force participation by 2 percent on average (compared with an average female labor force participation of 60 percent in countries that did not enact such laws). Larger impacts of 4 percent were measured five years after the enactment of the laws (Anukriti et al. 2023). In Bangladesh, LFS 2022 data show that 65 percent of women out of the labor force did not report that they lacked a desire to work, but that housework and family duties were the reason they were not looking for work.50 The substantial burden of housework and dependent care is coupled with low enrollment in childcare services. Data of the Bangladesh Individual Consumption Study, which surveyed primary caregivers of children, shows that the enrollment of children in formal childcare in Bangladesh is rare.51 Only 12 percent of children ages up to 7 were enrolled in daycare. Urban and rural children were equally likely to be enrolled. Evidence points to the role of both demand-side and supply-side constraints in limiting the use of formal childcare. Demand Preferences for Childcare Qualitative data gathered by SAR GIL and research partners as part of a World Bank–sponsored study on childcare support services for rural and semiurban NEET youth in Bangladesh indicate that childcare is considered an optional rather than an essential service and that semiurban households are more likely than rural households to report a need for and the use of childcare. The main constraints to accessing childcare reported by households include financial difficulties and the expense of daycare; concerns about child well-being, safety, and security in daycare centers; the opposition of older household members; the social stigma associated with reliance on daycare; and the distance to daycare centers (World Bank 2025e). Survey data from three divisions in Bangladesh echo these concerns. The Bangladesh Individual Consumption Study 2024 found that the quality and trustworthiness of caregivers, a preference for household or family members as caregivers, special needs accommodation, available hours, and location were the most important attributes in selecting a childcare service.52 The preference for family or household members as caregivers over daycare employees was the top reason given by respondents whose children had never attended daycare. The burden of caregiving falls disproportionately on women. Working-age women 50 Respondents could register multiple answer to the question “What is the main reason why you did not try to find a paid job or start a business in the last month?” The options discussed here are "housework/family work" and "no desire to work." 51 The Bangladesh Individual Consumption Study was led by the Development Data Group of the World Bank and SAR GIL, in collaboration with the Office of the Chief Economist of the World Bank South Asia Region. 52 Respondents could report up to four attributes that were important to them in choosing who takes care of a child in the absence of the main caregiver. 3 | Constraints and Policy Tools 95 represent 98 percent of caretakers in the survey sample. Women also spend more time than men with children during the day, including during work or productive activities. These findings together suggest that the provision of formal childcare services could play a role in increasing female labor force participation, and they highlight the need to invest in both financial support and awareness efforts to underline the benefits of childcare services. Supply of Childcare Under the Bangladesh Labor Act of 2006, employers with 40 or more women employees are legally required to provide on-site childcare services. Compliance, however, may be limited. A 2019 survey of firms that are legally mandated to support childcare found that fewer than a quarter (23 percent) of employers in Bangladesh provide any form of support to help their employees meet their childcare needs (IFC 2019). More recent data of 2022 of the World Bank Enterprise Survey similarly shows that few private formal firms in Bangladesh provide childcare on site or offer a cash or stipend benefit for childcare.53 Only 2 percent of service firms and 9 percent of firms in industry do so. This low provision of childcare services is also reflected in nationally representative household survey data. The LFS 2022 shows that, among women who have at least one child ages up to 6 at home and who are engaged in wage employment in services and industry across firms of all sizes, fewer a 10th receive childcare services from their employers (6 percent and 7 percent in services and industry, respectively). Moreover, childcare provision is less common in urban areas than in rural areas (5 percent and 8 percent, respectively). The low levels of employer support for childcare might be explained by the absence of financial support for firms that are mandated to provide childcare to employees (World Bank 2024e). Placing the financial and logistical burden of providing childcare services on firms could have the unintended effect of disincentivizing firms from hiring women of childbearing age. Conditioning employer-provided childcare on the number of women employees could have a similar discouraging effect on firms otherwise willing to hire women. Stronger public support for employer-provided childcare—for example, through some form of financial support—and wider public provision of childcare as well as Early Childhood Education might enhance the accessibility and affordability of childcare services. The government supported daycare centers through tax exemptions until 2023 (under Income Tax Ordinance No. 36 of 1984). The support was removed with the enactment of the Income-Tax Ordinance 2023, which does not contain exemptions for daycare centers. However, Bangladesh is expanding publicly provided Early Childhood Education (ECE). 53 Refer to WBES (World Bank Enterprise Surveys) (dashboard), World Bank, Washington, DC, https://www.enterprisesurveys.org/en/enterprisesurveys. Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 96 The quality of childcare facilities is also a key factor in the take-up of childcare services. Recognizing the need for high-quality daycare centers, the government enacted the Child Daycare Centre Act 2021 to regulate the quality of daycare facilities for children ages 4 months to 6 years. The law requires all daycare facilities to obtain a registration certificate to ensure compliance with standards on financial stability, infrastructure, and staff qualifications. The Department of Women Affairs also established 20 daycare centers in 2021 to support working parents. Key indicators reveal that, while Bangladesh fares better on average than the rest of the region in terms of the legal framework governing the provision of childcare, substantial gaps persist (World Bank 2024e). On the childcare legal frameworks indicator, which assesses laws governing the availability and public financing of childcare, including support for parents and nonstate providers (such as private centers and employers) and the existence of regulations to ensure service quality, Bangladesh scores 50 in 100, while the regional average score is 10 in 100 (World Bank 2024f). This indicates that only half the minimum necessary legal framework has been established in Bangladesh. The country’s score on the supportive frameworks indicator, which evaluates the availability of key instruments designed to support the implementation of the laws effectively, is only 25 in 100, meaning that only one-fourth of the minimum supportive policy framework has been established. In comparison, the average score in the region is 20 in 100 (World Bank 2024f). Supportive frameworks include publicly available registries of childcare providers, clear application procedures for financial assistance, and systems to monitor service quality through accessible quality reports. Ongoing work by the World Bank suggests that the registration requirements under the Child Daycare Centre Act 2021 may also limit the supply of childcare facilities (Majoka and Shams 2025). There is wide variation in the childcare services offered outside the home in Dhaka. Service quality often appears to be based on the demographic segment of the population served by the center. Meeting the criteria for a registration certificate may therefore be a binding constraint on opening more centers in low-income areas. Alongside stronger regulation and monitoring to ensure that quality standards are met, addressing supply-side challenges will require policies, such as incentivizing centers in low-income areas, that ease this constraint on potential childcare providers. 3 | Constraints and Policy Tools 97 Policy Tools and Causal Evidence on Their Impact The evidence on the impacts of childcare on women’s economic empowerment only involves interventions that address supply-side constraints. Three primary policy tools targeted at childcare and the division of housework have been identified (refer to table 3.5). Table 3.5: Policy tools for childcare and home responsibilities Policy tool Main takeaways Evidence Evidence Assessment on from other of global Bangladesh South evidence Asian countries Accessible Childcare provision for young children No Yes Conclusive and affordable before the start of preschool has been childcare shown to raise maternal labor force participation in many contexts. Accessible Enrollment of children in pre-school No No Conclusive and affordable allows mothers to switch to formal or preschool wage employment, even when it doesn’t induce entry into the labor force. Technological Technologies, such as the expansion of No No Limited innovations access to electricity and liquid petroleum gas, can reduce the time spent on household chores, allowing women to increase their labor force participation. Source: Original table for this publication. Note: For additional information on the studies included in this review, refer to Ajayi, Dao, and Koussoubé (2023); Anukriti et al. (2023); Attanasio, Barros, et al. (2017); Attanasio and Vera-Hernandez (2004); Barros et al. (2011); Berlinski and Galiani (2007); Berlinski, Galiani, and McEwan (2011); Berthelon, Kruger, and Oyarzún (2015); Bharati, Qian, and Yun (2020); Bjorvatn et al. (2025); Calderón (2014); Caria et al. (2023), (2025); Clark et al. (2019); Dang, Hiraga, and Nguyen (2019); Dean and Jayachandran (2020); de la Cruz Toledo (2015); Dinkelman (2011); Donald, Lowes, and Vaillant (2024); Du and Dong (2013); Garcia, Latham-Proença, and Mello (2022); Halim, Johnson, and Perova (2022); Hojman and López Bóo (2019); Kilburn and Datar (2022); Krafft and Lassassi (2024); Martínez and Perticará (2017); Medrano (2009); Nandi et al. (2020); Padilla-Romo and Cabrera-Hernández (2018); Rosero and Oosterbeek (2011); Ryu (2020); Sanfelice (2024); Talamas Marcos (2023); Zahra, Javed, and Muñoz-Boudet (2023). First, globally, numerous studies demonstrate positive impacts of the provision of accessible and affordable childcare depending on factors such as maternal marital status, sector of employment, and the ages of the children. The take-up of childcare services also varies by context, underscoring the importance of the specific local conditions faced by women (Halim, Johnson, and Perova 2022). The papers reviewed under the category of accessible and affordable childcare cover interventions such as subsidized or free formal childcare, family or community- based childcare, and the provision of formal childcare at the workplace. While the global evidence overwhelmingly shows positive impacts on maternal labor force participation in many contexts, evidence on South Asia is scant (Dean and Jayachandran 2020; Nandi et al. 2020). Good evidence on the impact of employer- provided childcare is also lacking. Requiring employers to provide childcare can disincentivize them to hiring women of childbearing age because of the higher Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 98 costs associated with supporting childcare needs. Future research should therefore investigate the differential impact of work-based versus community- based childcare, as well as the conditions under which regulations on or support provided to employers can facilitate the supply of quality childcare by firms. Second, evidence on the introduction of accessible and affordable preschool or an extension of preschool hours shows positive impacts on the labor force participation of mothers with young children, and the impacts are stronger at lower levels of baseline maternal labor force participation (Dang, Hiraga, and Nguyen 2019; Du and Dong 2013; Kilburn and Datar 2022). Third, moving from care to other forms of housework, the evidence is limited on the impact of labor-saving technological innovations, such as switching to liquid petroleum gas (a labor-saving cooking fuel) and expansion of access to electricity, that free up women’s time from household responsibilities. The effects on female labor market outcomes are positive in Indonesia and South Africa (Bharati, Qian, and Yun 2020; Dinkelman 2011). More research is needed to understand the potential of other home appliances that may be more relevant to Bangladesh, such as the use of washing machines, vacuum cleaners, and so on. Considering the cultural differences in South Asia relative to other economies, especially the importance of gender-specific social norms, the global evidence should be examined with caution. For a review of the literature on childcare in South Asia conducted by SAR GIL, refer to Zahra, Javed, and Muñoz-Boudet (2023). Credits: J R / pexels 3 | Constraints and Policy Tools 99 3.7 Business Ownership and Business Growth In Bangladesh, 64 percent of employed women are self-employed (as discussed in section 1.2). Policies that promote business ownership and growth are, therefore, central to advancing women’s economic empowerment. Agricultural and Nonagricultural Businesses Policy Tools and Causal Evidence on Their Impact The vast literature on policy tools aimed at increasing female business ownership, business growth, and business outcomes is grouped into eight categories with varying amounts of evidence (refer to table 3.6). Table 3.6: Policy tools for business ownership and business growth Policy tool Main takeaways Evidence Evidence Assessment on from other of global Bangladesh South evidence Asian countries a. Tools aimed at both agricultural and nonagricultural businesses Access to Microcredit has modest effects. The No Yes Mixed capital and effects are primarily on enterprises with finance high profits initially. Some evidence suggests that providing capital through joint-liability (that is, group) loans can lead to higher take-up and positive impacts on enterprise outcomes relative to individual- liability (that is, individual) loans. Access to In-kind transfers and the provision of No Yes Emerging capital and capital through mobile money reduce the finance and capture of capital by family members and increasing increase women’s control over how the control over money is used, resulting in higher business capital and profits. Evidence suggests that business finance outcomes are enhanced if capital is hidden from kin. Access to The evidence is mixed on the effect of the No No Emerging capital, finance, provision of training and capital together and training on the growth of women’s enterprises. Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 100 Table 3.6 (continued) Policy tool Main takeaways Evidence Evidence Assessment on from other of global Bangladesh South evidence Asian countries Business Traditional training programs have mixed No No Emerging training effects on the outcomes of women’s entrepreneurial efforts because of women’s lack of agency over decisions within enterprises. Heuristics-based business training focusing on easy-to- apply rules of thumb show promising results on sales and profits, particularly among entrepreneurs with low cognitive ability or few financial skills. Soft skills Training among women entrepreneurs No Yes Mixed training on soft skills, imagery-based learning, and aspirations shows promising, but mixed effects on business outcomes. Who entrepreneurs attend the training sessions with is also important. Attending sessions with social network peers improves the ability to apply training to business. b. Tools aimed at nonagricultural businesses Promoting Formal e-commerce platforms have the No No Limited market potential to expand the consumer base of access among women owners of small businesses who nonagricultural traditionally rely on informal means, such businesses as social media (for example, Facebook and WhatsApp). Complementing skills training interventions with support for market linkages can also enhance impacts, including by increasing economic activity, income, civic engagement, and empowerment. c. Tools aimed at agricultural businesses Agricultural Traditional training and extension No Yes Emerging extension services, paired with video messaging and services feedback channels, may have positive impacts on the productivity of women farmers. Targeting information on wives in agricultural households can empower the wives through their role in agricultural decision-making and by reducing gendered task division. The method by which information is distributed also matters. Women are more likely to receive information if the information is targeted in villages randomly or on individuals with high degree centrality (many connections), rather than targeted at individuals with high betweenness centrality (ability to act as a bridge to parts of a network). 3 | Constraints and Policy Tools 101 Table 3.6 (continued) Policy tool Main takeaways Evidence Evidence Assessment on from other of global Bangladesh South evidence Asian countries Facilitating Limited evidence points to the potentially No No Limited market linkages lower take-up of phone-based digital among farmers marketplaces by women farmers in Uganda. However, more systematic research is needed on the impacts of interventions focusing on the market linkages of women farmers. Interventions aimed at men farmers show promise. Using digital tools to create market linkages reduces transaction costs and spatial price dispersion, strengthens farmer market linkages, grows agricultural incomes, and improves supply chain efficiency. Source: Original table for this publication. Note: For additional information on the studies included in this review, refer to Alaref, Brodmann, and Premand (2020); Alhorr (2024); Angelucci, Karlan, and Zinman (2015); Arráiz, Bhanot, and Calero (2019); Ashraf et al. (2022); Attanasio, Augsburg, and De Haas (2019); Attanasio et al. (2015); Banerjee, Duflo, Glennerster, et al. (2015); Banerjee, Duflo, Goldberg, et al. (2015); Banerjee et al. (2019); Bastian et al. (2018); Batista, Sequeira, and Vicente (2022); Baul et al. (2024); Beaman and Dillon (2018); Berge, Bjorvatn, and Tungodden (2015); Bernard et al. (2023); Bjorvatn et al. (2025); Blattman, Fiala, and Martínez (2014); Brooks, Donovan, and Johnson (2018); Cheema et al. (2019); Crépon, El Komi, and Osman (2024); de Mel, McKenzie, and Woodruff (2008), (2009), (2014); Dillon, Mueller, and Salau (2011); Donald, Goldstein, and Rouanet (2022); Drexler, Fischer, and Schoar (2014); Dupas and Robinson (2013); Fafchamps et al. (2014); Ferrah et al. (2021); Fiala (2018a), (2018b); Field et al. (2016); Giné and Mansuri (2021); Heath and Riley (2024); Jack et al. (2023); Jones and Kondylis (2018); Khandelwal and Singh (2024); Lafortune, Riutort, and Tessada (2018); Lecoutere, Spielman, and van Campenhout (2023); McKenzie, Mohpal, and Yang (2022); Orkin et al. (2023); Pomeranz and Kast (2024); Riley, Shonchoy, and Osei (2025); Rojas Valdes, Wydick, and Lybbert (2022); Ubfal et al. (2022). The body of evidence on the provision of access to capital and finance to increase firm profits and sales is extensive, but mixed. This theme covers research that assesses the impacts of microcredit, grants, asset collateralization, and opening bank or savings accounts. The positive impacts of microcredit and grants on business outcomes often only accrue to previously high-profit enterprises, with minimal to no impacts on downstream outcomes, such as women’s empowerment (Banerjee, Duflo, Glennerster, et al. 2015; Crépon, El Komi, and Osman 2024). Furthermore, the evidence suggests that it may take women more time than men to improve business outcomes following access to capital or credit (Blattman, Fiala, and Martínez 2014). A recurring theme in the evidence on providing access to capital, including providing training along with capital, is the lack of agency that women have over capital and decisions within the firm. Another strand of literature is therefore categorized as the provision of access to capital and finance and increasing control over capital and finance. The studies on this category test various ways of enhancing women’s control over capital, such as by providing capital directly into mobile money accounts owned by women, providing capital in kind, and hiding capital from kin. Such interventions often produce positive impacts on firm profits and sales (Bastian et al. 2018; de Mel, McKenzie, and Woodruff 2008, 2009; Fiala 2018a; Heath and Riley 2024). Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 102 The next three strands test the impacts of the provision of training in both business skills and soft skills, including psychological interventions, to women business owners. There is limited evidence on the impact of the specific content of this sort of training, such as showing video documentaries of successful individuals from similar communities, holding workshops to encourage higher aspirations and develop personal grit, or training on positive psychology. This has mixed results on business outcomes. However, the findings indicate that women who attend the sessions accompanied by others in a shared social network increases women’s take-up and the ability to apply the content of the training to business (Field et al. 2016; Khandelwal and Singh 2024). Nonagricultural Businesses Policy Tools and Causal Evidence on Their Impact The sixth policy tool is promoting market access among nonagricultural businesses. One approach involves promoting the use of e-commerce platforms that can expand the markets available to business owners. The e-commerce market in Bangladesh has experienced massive growth since 2014. Though systematically collected gender-disaggregated data are not available, there is suggestive evidence that women constitute a major share of the country’s online business owners and that this growth is concentrated in informal e-commerce on social media platforms (such as Facebook and WhatsApp), mainly because of systemic constraints, such as complicated business registration processes and mobility issues (Moyeen et al. 2022). Descriptive evidence also shows that there has been relatively low adoption of formal e-commerce by women business owners (Azam et al. 2023). Research on efforts to encourage women business owners to adopt e-commerce platforms is scarce. Further study is therefore recommended on how formal e-commerce platforms can increase business ownership and business growth among women. To gauge the potential of such platforms and to address constraints faced by women business owners in accessing e-commerce platforms, such as the lack of digital skills, mobility constraints, safety concerns, and high entry costs, SAR GIL is exploring possible interventions in partnership with stakeholders in Bangladesh. Another approach to improving market access is to facilitate market linkages as complements to other interventions. In a study focusing on women in Pakistan, Cheema et al. (2019) investigate a market linking scheme in support of a skills training intervention among women tailors. Sales agents linked the beneficiaries to marketable designs, raw materials, and quality control mechanisms and attempted to sell the products of the beneficiaries in urban markets. This approach enhanced business outcomes above and beyond the skills training alone, leading to an expansion of the tailoring activity, earnings, knowledge of potential markets, civic engagement, and household consumption. The intervention even had a positive impact on the perceptions among men on gender roles and improved child 3 | Constraints and Policy Tools 103 nutrition outcomes. Thus, facilitating market linkages holds promise for enhancing women's empowerment. However, more research is required in Bangladesh, including a gendered approach, to identify the effect of market linkages on women in nonagricultural businesses, especially alongside the various asset transfer programs. Agricultural Businesses Women in Bangladesh generally tend to be excluded from household agricultural or livestock extension services. Only 4 percent of the households in the 2018–19 BIHS reporting that they received agricultural crop extension services had at least one woman in communication with the services. The corresponding share involving men was 98 percent. Among households receiving livestock and fisheries services, the share was a bit higher, at 11 percent and 100 percent, respectively. This highlights the need for more woman-inclusive targeting in agricultural extension service policies. Yet, in fiscal year 2020, only a miniscule portion of the Ministry of Agriculture’s budget (0.44 percent, or US$6 million) was spent on programs to increase access to extension services and training among women farmers (World Bank 2024c). Policy Tools and Causal Evidence on Their Impact The seventh tool—agricultural extension services—is widely used to enhance agricultural productivity. The literature on the effectiveness of various, related delivery methods is extensive (Suri and Udry 2024). However, only a few studies examine the impacts on women farmers, though the recent literature is beginning to fill the gap. In India, Baul et al. (2024) find that supplementing traditional training with videos that emphasize the message through repetition and add messages on common perceived challenges among women farmers raised output and yield among such farmers. In Rwanda, Jones and Kondylis (2018) find that feedback from farmers through logbooks and scorecards during these training programs encourages women to interact with field officers and improves their perceptions of them, potentially enhancing extension services. Among men, meanwhile, the feedback intensified participation in the training, but had no effect at the extensive margin. A few other studies also highlight the importance of explicitly including women. For instance, encouraging the joint participation of husbands and wives in extension services through targeted information or attendance incentives has been shown to elevate women’s roles in agricultural decision-making, reduce unilateral decisions by men, and enhance overall productivity (Donald, Goldstein, and Rouanet 2022; Lecoutere, Spielman, and van Campenhout 2023). Dillon, Mueller, and Salau (2011) demonstrate that how information is diffused through social networks affects women’s access to knowledge. Targeting individuals in a village randomly or based on their degree centrality (number of direct connections) is more effective in knowledge acquisition than targeting individuals based on their betweenness centrality (ability to act as a bridge among parts of a network). Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 104 The eighth policy tool is facilitating market linkages among farmers, such as digital platforms that match buyers and sellers. Evidence on the impact of this policy tool on women farmers is scarce. The only study that provides insights into the impact shows that women farmers in Uganda are less likely than men farmers to use a mobile phone–based marketplace that matches buyers and sellers (Bergquist, McIntosh, and Startz 2024). Studies evaluating the impact on men farmers or on both men and women farmers find promising outcomes. Interventions targeting market linkages thus reduce transaction costs and spatial price dispersion, strengthen the market linkages of farmers, boost agricultural incomes, and improve supply chain efficiency (Abate et al. 2023). However, these studies do not provide a gender disaggregated analysis of impacts on economic outcomes. Given the generally lower levels of market access among women farmers relative to men farmers, these interventions may benefit women farmers more. More research is required to measure the impact of programs focused on agricultural market linkages on women farmers. 3.8 Job Search Job search is a crucial step toward engagement in the labor market, and search frictions can often lead to suboptimal job matches. In fiscal year 2020, the Bangladesh Ministry of Labor and Employment allocated almost 24 percent of its expenditure (US$5.7 million) to developing job-matching networks and promoting women’s access to these networks. Job search frictions may arise from constraints on the supply side (job-seeker) and on the demand side (employers and intermediaries such as job platforms). Interventions aimed at alleviating supply-side constraints are discussed under the constraints in other areas (such as improving technical and vocational skills and easing childcare and time constraints). This section discusses interventions that can help employers and intermediaries—the demand side—improve women’s labor market outcomes. Demand-side factors including discrimination by employers and frictions to workplace integration are demonstrated to be important constraints in the literature, but there is limited evidence on appropriate policy tools to address them. Data of the World Bank Enterprise Survey 2013 show that the most common reason reported by private formal firms in Bangladesh for not hiring women in nonmanagerial roles is that the presence of women could cause disruptions in the workplace (reported by 45 percent of employers).54 The literature measuring and 54 WBES (World Bank Enterprise Surveys) (dashboard), World Bank, Washington, DC, https://www.enterprisesurveys.org/en/enterprisesurveys. 3 | Constraints and Policy Tools 105 Credits: Ferdous Hasan / pexels documenting discrimination is also growing. Recent evidence suggests that employers in Bangladesh discriminate against women in the hiring process and paternalistically prevent women from making their own choices (Buchmann, Meyer, and Sullivan 2024). Women supervisors in Bangladesh’s RMG sector also face discrimination from workers who have negative beliefs about women’s abilities, leading to lower productivity (Macchiavello et al. 2020). However, these gender stereotypes may not generalize to similar contexts. In Pakistan, Gentile et al. (2023) find that gender restrictions on job ads are a larger than constraint on the supply side. Another demand-side constraint is the cost of integrating women in workplaces that have smaller shares of women employees. Data of the World Bank Enterprise Survey 2013 shows that the second most common reason reported by employers in Bangladesh for not hiring women in nonmanagerial roles is that women are more expensive for the firm because they require separate working facilities and benefits (reported by 43 percent of employers). Miller, Peck, and Seflek (2022) show that, in Saudi Arabia, the costs of integrating women are primarily fixed and that firms hire more women if costs per worker have fallen because of a gender-neutral quota requirement that leads to the hiring of a large number of individuals, both men and women. Policy Tools and Causal Evidence on Their Impact The first policy tool aimed at reducing job search frictions is creating platforms for job search, which aims to improve labor market matches through mechanisms that connect job-seekers and employers, such as job fairs, recruitment services, and digital job platforms (refer to table 3.7). The evidence on the impact of job fairs and recruitment services on women’s employment is limited. In rural Philippines, facilitating matches by encouraging attendance at a job fair increased the job search effort and the likelihood of finding work in the formal sector. It also raised the perceived likelihood that women would receive offers for jobs abroad, while it reduced this perceived likelihood among men (Beam 2016). In India, recruiting services that are aimed at hiring young women for business process outsourcing jobs showed promise through positive impacts on women’s paid employment and work aspirations (Jensen 2012). Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 106 Table 3.7: Policy tools for the job search Policy tool Main takeaways Evidence Evidence Assessment on from other of global Bangladesh South evidence Asian countries Creating The use of platforms such as job fairs, No Yes Emerging platforms for job recruitment services, and digital job search platforms following current design may have limited impacts in improving women's labor market outcomes, especially by increasing women’s formal employment. The reasons for the mixed impacts include the low levels of the jobs on the platforms and the gendered structure of social networks. Integrating Jobs targeted at women tend to No Yes Emerging gender-related be associated with lower wages, information in job and information on the gender of ads supervisors and coworkers can alter women's application rates. Explicit or implicit gender preferences in job ads account for a substantial share of the employment gender gap. However, removing gender preferences in job ads may not benefit women because the evidence shows that men may enter previously woman-targeted jobs more frequently than women enter previously man-targeted jobs. Facilitating Reference letters in job applications may No No Limited reference checks increase the callbacks among women, but internal referrals under various incentive schemes may put qualified women candidates at a disadvantage. This is driven by men who choose not to refer qualified women and women who exhibit no gender preference in referrals. Reducing Reducing the monetary or psychological No Yes Emerging application costs of applying for jobs can costs (monetary, substantially raise the number of job time, and applications submitted by women, also psychological) improving the overall quality of the applicant pool. Source: Original table for this publication. Note: For more information on the studies included in this review, please refer to Abebe, Caria, and Ortiz-Ospina (2021); Abel, Burger, and Piraino (2020); Afridi et al. (2022); Archibong et al. (2025); Beam (2016); Beaman, Keleher, and Magruder (2018); Chakravorty et al. (2023); Chaturvedi, Mahajan, and Siddique (2021); Chowdhury et al. (2018); Gentile et al. (2023); Groh et al. (2016); Jensen (2012); Jones and Sen (2022); Kuhn and Shen (2023); Subramanian (2024); Vyborny et al. (2024); Wheeler et al. (2022). 3 | Constraints and Policy Tools 107 Evidence on the impact of the use of digital job platforms is emerging and shows mixed results. A program in South Africa that trained job-seekers to use LinkedIn boosted the rate of employment by up to 12 months and provided suggestive evidence that this was driven by expanding the information about employers available to job-seekers (Wheeler et al. 2022). Yet, in India and Mozambique, encouraging job-seekers to sign up to job platforms showed limited effects on women’s employment. In Mozambique, the use of a platform on formal jobs failed to raise women’s employment, but a platform on informal jobs for freelancers for services, such as plumbing and catering, increased women’s employment, hours worked, and earnings, especially among women with industrial or construction qualifications (Jones and Sen 2022). In India, encouraging the use of a job platform by the government did not raise employment among either men or women, likely because of the availability of few vacancies on the platform and the difficulty of using the platform (Chakravorty et al. 2023). Another study in India testing the impact of promoting the use of a job platform among the wife’s network had not resulted in overall employment gains among women a year after the launch of the intervention, except for an increase in home-based self-employment. Instead, the intervention increased the likelihood that the husbands would find work and increase the hours they worked and their monthly earnings (Afridi et al. 2022). The mixed impacts of the use of job platforms on women’s labor market outcomes may derive from specific design features that affect women’s employment. For example, there is emerging evidence on integrating gender-related information in job ads and how gender preferences for jobs or information about the gender of supervisors and coworkers on the platforms influence women’s engagement on the platforms and the associated labor market outcomes. In urban Pakistan, most of the gender gap in job opportunities on a jobs platform is explained by the fact that 60 percent of the job ads on the platform were closed to women. The gender restrictions imposed by the firms in the sample are the largest constraints on the number of jobs available to women with secondary educational attainment or less. This demand-side constraint is lower among women with higher educational attainment (Gentile et al. 2023). Another study in Pakistan found that, if information that a supervisor was a man was brought up during discussions about a job between women job-seekers and their families, this reduced the rate of women’s job applications by almost 60 percent. The preference for same-gender supervisors could be offset by a higher wage. Up to 70 percent of respondents reported that they would choose a job with a man supervisor if they were offered an additional PRs 10,000 in income. The preference for jobs with women coworkers is, however, much stronger and much less likely to succumb in the face of an offer of higher pay (Subramanian 2024). In India, studies have found that job ads that target women, even using only implicit language, are associated with lower wages relative to jobs targeted at men (Chaturvedi, Mahajan, and Siddique 2025; Chowdhury et al. 2018). Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 108 However, it is unclear whether removing a gender preference from job offers would necessarily benefit women. After prohibiting explicit gender requirements from a large job platform in China, men took up jobs that had been previously associated with a preference for women at greater rates than women entered jobs that had previously been linked to a preference for men (Kuhn and Shen 2023). Identifying the type of gender-related information that should be contained in job ads to improve women’s labor market outcomes is an important area for future research. SAR GIL is leading a research program to deepen the evidence on what works in job-matching platforms to improve the outcomes among women, including in Bangladesh (World Bank 2025b). SAR GIL is also partnering with an NGO in Bangladesh to improve matches in the gig economy for domestic work by providing women with access to a digital app, similar to a job platform, that supplies information on job opportunities, skills training, or both (Anukriti et al. 2024). A smaller strand of the literature tests the impact of facilitating reference checks. The evidence on this tool is limited and mixed. One study on South Africa suggests that reference letters from previous employers can increase callbacks among women applicants (Abel et al. 2020). Another study, on Malawi, suggests that internal referrals may put females at a disadvantage (Beaman, Keleher, and Magruder 2018). Good evidence on South Asia is lacking. SAR GIL is working to fill this gap by testing the impact of centralized reference verification services through a job platform in Pakistan. Evidence on the final policy tool—reducing application costs (monetary, time, and psychological)—is emerging and shows that reducing the costs of applying for jobs can improve labor market outcomes, especially among women. In Ethiopia, offering monetary incentives for submitting applications for clerical jobs attracted job-seekers from the most disadvantaged groups of job-seekers: women, the unemployed, and the less-experienced (Abebe, Caria, and Ortiz-Ospina 2021). Similarly, in Jordan, providing wage subsidies to employers who hire recent graduates boosted the employment of graduates. While the increased employment did not persist after the end of the wage subsidy, primarily because employers did not find the new hires sufficiently productive, it provided employers with information on graduate quality and young gradates with relevant experience (Groh et al. 2016). A study in Pakistan shows that reducing the psychological cost of applying for a job by calling the job-seekers to initiate the job application increased the number of applications dramatically, by 600 percent (Vyborny et al. 2024). 3 | Constraints and Policy Tools 109 Credits: K M Asad / World Bank 3.9 Digital Inclusion Mobile phones are the primary means of connecting to the internet in low-income countries (GSMA 2024). Data of the GSMA Consumer Survey 2023 show that the share of women owning a mobile phone is lower in Bangladesh (68 percent) than in its aspirational peer countries (Indonesia, at 77 percent) and structural peer countries (India, 75 percent) on which data are available (GSMA 2024). Microdata from Findex 2017 and 2021 indicate that gender gaps in digital inclusion persist across urban and rural areas and span income groups.55 Although there was some progress in reducing the gender gap in mobile phone ownership in Bangladesh in 2017–21 (falling from 26 percentage points to 19 percentage points), substantial disparity persists. This gap is evident across both rural and urban areas. Men’s mobile phone ownership was similar across urban and rural areas in 2021 (90 percent and 89 percent, respectively), but urban women showed higher rates of ownership (73 percent) than rural women (63 percent). This urban-rural divide among women underscores the uneven distribution of technological resources and opportunities by region among women. The differences between the share of women who report that they use a phone and women who report that they own a phone are stark, especially in rural areas. While 90 percent of rural women survey respondents used a phone at least once a week in the previous three months, only 69 percent of rural women own a phone, according to 2019 data of the Multiple Indicator Cluster Surveys. The corresponding gap is slightly narrower in urban areas (95 percent vs. 80 percent) (refer to figure 3.4, panel a).56 55 Refer to Global Findex (Global Findex Database), World Bank, Washington, DC, https://www.worldbank.org/en/publication/globalfindex/archive. 56 Gender gaps in phone ownership and phone use cannot be estimated because the data source, the Bangladesh Multiple Indicator Cluster Surveys, does not survey men. Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 110 Figure 3.4: Digital inclusion, a. Phone access, women, ages 15–49 by region and sex 100 95 90 80 80100 95 90 69 80 Percentage (%) 60 80 69 Percentage (%) 60 40 40 20 20 0 0 Owns Phone Uses Phone Owns Phone Uses Phone Rural Urban Rural Urban b. Internet access, by sex, ages 15–64, 2021 100% 100% Sources: Panel a: Original figure for this publication 80%80% based on calculations using data of MICS (Multiple Indicator Cluster Surveys) 2019, Bangladesh Bureau of Statistics and United Percentage (%) 60% Percentage (%) Nations Children's Fund, Dhaka, Bangladesh, 60% https://mics.unicef.org/surveys. Panel b: Original figure for this publication based on calculations using data of 38% Global Findex (Global Findex Database), 40% 38% World Bank, Washington, DC, https://doi. 40% org/10.48529/qda7-6z97. 20% Note: Phone ownership data comes from 20% 20% the question “Do you own a mobile phone? ”Phone usage is defined as having used the 20% phone at least once a week in the last three months, using the question “During the last 0% 3 months, did you use a mobile telephone Internet Access at least once a week, less than once a week 0% or not at all?” Internet access data comes from the question “Do you have access to Internet Access the internet in any way, whether on a mobile Female Male phone, a computer, or some other device?” Internet access, an increasingly important aspect of digital inclusion, is low in Bangladesh according to microdata from Findex 2021 especially among women (refer to figure 3.4, panel b). The gender gap in internet access is similar across rural and urban areas. It is also large across all income quintiles.57 The narrowest gender gaps occur among the poorest and richest quintiles (11 percent and 10 percent, respectively), and the widest gap occurs in the fourth quintile (26 percent). While 57 The Findex data cover only income quintiles, which are therefore used here. 3 | Constraints and Policy Tools 111 internet access lags among both men and women in the lowest income quintile, it rises more quickly among men than among women across the middle-income quintiles. Except for the richest quintile, the share of men who have access is about twice the share of women. Only 13 percent of the poorest women and 39 percent of the richest women report they have access. Policy Tools and Causal Evidence on Their Impact The literature review highlights the need to implement policies that effectively improve the digital inclusion of women in Bangladesh (refer to table 3.8). Enhancing women’s digital access is a promising tool for women’s economic empowerment. The evidence illuminates the heterogeneous effects across subgroups of women, underscoring the need for context-driven initiatives (Barboni et al. 2024; Klonner and Nolen 2010; Kusumawardhani et al. 2023; Viollaz and Winkler 2022). Table 3.8: Policy tools for digital inclusion Policy tool Main takeaways Evidence Evidence Assessment on from other of global Bangladesh South evidence Asian countries Mobile phone Facilitating the supply of mobile phones No Yes Emerging provision among women, along with the means and effective to realize connectivity, does not result accessibility in improved outcomes among women in India in the short or long run because men tend to take progressive control over the asset. Evidence on Africa suggests that maintaining ownership, rather than selling smartphones, boost household consumption and leads to a shift in the time women spend on farming to include more communication with clients and market trading, though this does not always translate to higher income flows. Basic phones are more effective in increasing the take-up of digital financial services among women with limited literacy. Expanding Internet coverage may improve women’s No No Emerging internet access labor force participation and full-time employment by facilitating online job search, though it may also reduce the probability of holding skilled or formal jobs. Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 112 Table 3.8 (continued) Policy tool Main takeaways Evidence Evidence Assessment on from other of global Bangladesh South evidence Asian countries Digital literacy Case studies highlight the potential of No No No evidence training programs tailored to women’s needs and preferences to be more effective at motivating women to build digital skills. Pilot evaluations suggest that group- based learning approaches can boost phone ownership, improve employment opportunities, and provide women role models and that mentors may be an effective promotion strategy for digital skill development. Targeting norms Descriptive evidence suggests that No No No evidence on women’s engaging with the gatekeepers of digital access women’s digital access—typically men household members—can mitigate backlash and encourage men to see the value in women’s access. Source: Original table for this publication. Note: For additional information on the studies included in this review, refer to Bahia et al. (2023); Barboni et al. (2024); Giulivi et al. (2023); Klonner and Nolen (2010); Kusumawardhani et al. (2023); Roessler et al. (2021); Viollaz and Winkler (2022). The policy tools to foster digital inclusion are associated with varying levels of evidence. First, the limited evidence on policies related to mobile phone provision and effective accessibility is mixed. The evidence on India shows that women often do not have control over their own phones and that the impact of offering free smartphones and connectivity to women is limited (Barboni et al. 2024). The evidence on Africa suggests that, if women are able to retain maintain control over their smartphones, the outcomes in the economic empowerment of women and households improve. Mobile phone access can help enable financial inclusion, but persistent gender gaps in the ownership and control over mobile phones can likewise undermine the effectiveness of interventions that foster financial inclusion. Policies aimed at expanding digital inclusion ought therefore also to consider intrahousehold dynamics and women's ability to use and retain control over digital devices. Second, a smaller but emerging strand in the literature demonstrates ways in which investment in expanding internet access may encourage digital inclusion. Evidence on Indonesia, Jordan, and Tanzania show gains in female labor force participation in response to the expansion of internet availability. However, these effects may be limited to certain subgroups of women, and men may maintain an 3 | Constraints and Policy Tools 113 advantage in capitalizing on labor market returns to improved information and communication technology infrastructure associated with the internet (Bahia et al. 2013; Kusumawardhani et al. 2023; Viollaz and Winkler 2022). Good evidence is not yet available on two additional tools—digital literacy training and targeting norms on women’s digital access—that may improve women’s digital inclusion. Case studies and pilot studies suggest that tailoring digital literacy training to the needs of women and girls through group training and the identification of women role models may enhance the take-up and use of information and communication technology (BIGD 2024; GSMA 2020; Mariscal et al. 2019; World Bank 2023, 2025g). While not yet rigorously tested, engagement with gatekeepers—typically men household members who circumscribe or monitor women’s digital access—shows promise as a strategy to ease the effects of social norms that restrict the digital inclusion of women (BIGD 2024; Sey and Hafkin 2019). Overall, the evidence on the effectiveness of tools to enhance the digital inclusion of women is scarce, especially in South Asia. More high-quality evidence is required to identify what works in the effort to enhance women’s ownership and women’s control over digital devices. 3.10 Migration Internal and international migration can facilitate social mobility and increase consumption. However, the use of this mechanism of economic mobility by women is constrained in South Asia. Most migrants are young men. Concerns about the constraints of distance from home, housing, travel, social networks, and safety significantly affect a woman’s ability to engage in migration independently of her spouse or of the man-headed household (Beegle, De Weerdt, and Dercon 2011). According to the migration patterns revealed by nationally representative HIES data, men migrate from rural to urban areas at far higher rates than women (refer to figure 3.5, panel a). Between 2017 and 2022, the migrants from rural to urban area included 9.7 million men, but only 1.5 million women. Two-thirds of the men were from rural households that did not own cultivable agricultural land. Such households account for 58 percent of all rural households. This suggests that most of the migration originates in poorer households and is driven by economic necessity. 58 Refer to Global Findex (Global Findex Database), World Bank, Washington, DC, https://www.worldbank.org/en/publication/globalfindex/download-data. Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 114 Figure 3.5: Rural-to-urban migration rate and reason for migration, by sex, ages 15–64 a. Share of rural residents who migrated to urban areas, by sex 4 2.7 3 Percentage(%) 2 1 0.4 0 b. Reason for migration, by sex 100% 80% 73 Percentage (%) 60% 48 40% 24 16 20% 12 8 5 6 3 4 1 1 0 1 0% Work Be Closer Migration Other Marriage to Family Live with with Family Education / Family Member Training Female Male Sources: Panel a: Original figure for this publication based on calculations using data of HIES (Household Income and Expenditure Survey) 2022, Bangladesh Bureau of Statistics, Dhaka, Bangladesh, accessed through SARRAW (South Asia Raw) and SARMD (South Asia Regional Micro Database) (databases), South Asia Region Team for Statistical Development, World Bank, Washington, DC. Panel b: Original figure for this publication based on calculations using data of DIGNITY (Dhaka Low Income Area Gender, Inclusion, and Poverty Survey), 2018, Microdata Library, World Bank, Washington, DC, https://doi.org/10.48529/gare-az26. Note: Panel a: The data are representative of the country. Panel b: The data are based on the 2011 census and are representative of slum and nonslum low-income areas of the Dhaka City Corporations and additional low- income sites in the Greater Dhaka Statistical Metropolitan Area (2018 DIGNITY survey). 3 | Constraints and Policy Tools 115 According to the Socio-Economic and Demographic Survey 2023, the main cause of internal migration in Bangladesh (42 percent) is family dependency (BBS 2024). This is true among both men (38 percent) and women (45 percent). Marriage and employment or work are also important reasons for internal migration. More women than men migrate for marriage (43.5 percent vs. 2.9 percent), while more men migrate for work (26.3 percent vs. 3.5 percent). These gender disparities in the reasons for migration are also reflected in the migration patterns of individuals in slums and other low-income areas in Dhaka in the DIGNITY data. Almost three-fourths of men migrants moved to Dhaka for work. This was true of only half the women migrants, while a quarter of the women migrants moved to Dhaka for marriage (refer to figure 3.5, panel b). Almost all men who migrated for work migrated for their own work, while more than half the women who reported moving for work, moved because of the work of their husbands. Only about a third of the women who moved for work moved for their own work. A majority of migrants to Dhaka migrate alone. Fewer than 10 percent of the migrants (men or women) moved with their families. In the same sample of low-income migrants to Dhaka, almost all men migrants (96 percent) were employed, compared with only half the women migrants. Among those migrants who were working, a substantial share, almost 80 percent, of the women migrants were engaged as temporary or permanent wage employees, compared with 60 percent of the men migrants, revealing differences in the quality of the jobs migrant men and women find. Mirroring the trends in the overall population, there was severe occupational segregation by sex among the migrants as well. Thus, 39 percent of the women migrants worked as maids, and a quarter worked in the RMG sector. Among men migrants, roughly a quarter each work in transportation, services, and business. Nationally representative Findex data show that, in both rural and urban areas, women are less likely than men to send remittances home.58 HIES data reveal that the value of the remittances sent by women is also lower, about half the value of the remittances sent by men. However, relative to men migrants, a larger share of women migrants (almost half) use mobile banking to send remittances home. This highlights the growing reliance on digital financial services among women migrants, despite the lower value of the remittances they send. Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 116 Policy Tools and Causal Evidence on Their Impact Two types of policy tools can facilitate migration by reducing the associated costs (refer to table 3.9). First, studies have shown the positive impact on men of lowering the cost of migration through credit, but the evidence on the impacts on women is limited. Providing households with grants or zero-interest credit equivalent to the cost of a rural-to-urban round trip, particularly during the lean season, increased the migration rates among men in Bangladesh. While this migration incentive was not targeted specifically on men, but was provided to households, 97 percent of the individuals who subsequently migrated were men. Additionally, while the resulting migration increased the incomes of the migrants and the consumption of the origin households, the incentive did not alter the household preference for men migrants and also did not raise the female labor force participation rate among origin households (Bryan, Chowdhury, and Mobarak 2014). Furthermore, the initial evidence on providing incentives to encourage migration was not replicated at a larger scale. Thus, an evaluation of a similar program at a larger scale did not find a similarly substantial rise in seasonal migration, household income, or food consumption (Bahl 2019). Table 3.9: Policy tools for migration Policy tool Main takeaways Evidence Evidence Assessment on from other of global Bangladesh South evidence Asian countries Lowering the Subsidizing the cost of migration Yes No Limited cost of migration through grants or zero-interest loans through credit has been shown to increase the rates of migration among men and generate positive impacts on the incomes and consumption of origin households. However, the evidence on similar programs targeted on women is limited. Lowering the Reducing the costs of sending Yes No Limited cost of migration remittances (for instance, through the by reducing use of mobile money) has the potential the cost of to increase the hours worked by women, remittances particularly in the garment sector. Lower the cost of sending remittances may also encourage migration by reducing the overall costs of migration, although there is not yet any direct evidence. Source: Original table for this publication. Note: For more information on the studies included in this review (all from Bangladesh), please refer to Lee et al. (2021), (2022); Shonchoy, Fujii, and Raihan (2018). 3 | Constraints and Policy Tools 117 Shonchoy, Fujii, and Raihan (2018) have produced the only study on this tool with a gender lens. Their study, also on Bangladesh, examines a training intervention, coupled with incentives to migrate. It finds that, while the take-up of the program was greater among both men and women if an incentive to migrate was offered, women performed below the average, partly because of low completion arising from noneconomic barriers. More research is needed on this policy tool, particularly research incorporating a gender perspective and targeted at women. A key concern in offering credit for migration is the possibility of exploitation, for example, if women risk becoming trapped in unfavorable working conditions because of their debt obligations. Policies aimed at offering migration incentives to women must account for such possible unintended consequences. Second, the mechanism of lowering the cost of migration by reducing the cost of remittances may facilitate migration to urban areas and increase the labor supply of women. Lowering the cost of remittances by, for instance, encouraging the use of mobile money may boost the amounts of remittances, raise the frequency of remittances, and reduce the time and effort required to send money back home. This may encourage more migration by increasing the returns to women who migrate. Direct evidence on this outcome is not yet available. However, some evidence indicates that encouraging the use of mobile money may raise the value of urban-to-rural remittances in Bangladesh and the number of hours worked by women migrants, especially women working in the garment industry (Lee et al. 2021). This suggests that there are potential direct impacts of reducing the cost of remittances on the labor supply of women in urban areas. Credits: Mohd.Ashabul Haque Nannu / Pexels Appendices 119 Appendix A Summary Statistics: Key Measures of Women’s Economic Empowerment Table A.1: Labor market outcomes Indicator Men, mean (SD) Women, Difference, mean (SD) men – women (SE) 0.83 0.45 0.38*** Labor force participation, % population (0.38) (0.50) (0.01) 0.80 0.43 0.36*** Employment, % population (0.40) (0.50) (0.01) 0.22 0.29 –0.06*** Agricultural employment, % population (0.42) (0.45) (0.01) 49.89 33.13 16.76*** Hours worked, conditional on employment (14.82) (17.58) (0.49) Monthly earnings (Tk 1,000s), adjusted to 15.20 8.98 6.23*** 2024 prices, Winsorized at the 95th percentile, conditional on paid employment (11.27) (9.20) (0.30) Hourly earnings (Tk), adjusted to 2024 prices, 81.96 60.88 21.07*** Winsorized at the 95th percentile, conditional on paid employment (63.93) (130.00) (2.87) Sources: Original table for this publication based on calculations using data of LFS (Labor Force Survey) 2022, Bangladesh Bureau of Statistics, Dhaka, Bangladesh, accessed through SARRAW (South Asia Raw) (database), South Asia Region Team for Statistical Development, World Bank, Washington, DC; HIES (Household Income and Expenditure Survey) 2022, Bangladesh Bureau of Statistics, Dhaka, Bangladesh, accessed through SARRAW (South Asia Raw) and SARMD (South Asia Regional Micro Database) (databases), South Asia Region Team for Statistical Development, World Bank, Washington, DC. Note: SD = standard deviation. SE = standard error. Significance level: * = 10 percent, ** = 5 percent, *** = 1 percent. Table A.2: Financial and digital inclusion Type of account owned Men, mean (SD) Women, Difference, mean (SD) men – women (SE) 0.63 0.43 0.19*** Any account (0.48) (0.50) (0.04) 0.45 0.31 0.14*** Account w/financial institution (0.50) (0.46) (0.04) 0.39 0.20 0.20*** Mobile money account (0.49) (0.40) (0.03) Source: Original table for this publication based on calculations using data of Global Findex (Global Findex Database), World Bank, Washington, DC, https://doi. org/10.48529/qda7-6z97. Note: SD = standard deviation. SE = standard error. Significance level: * = 10 percent, ** = 5 percent, *** = 1 percent. Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 120 Table A.3: Agricultural assets Assets owned Men, mean (SD) Women, Difference, mean (SD) men – women (SE) 0.48 0.06 0.42*** Agricultural land (0.49) (0.24) (0.01) 0.31 0.41 –0.09*** Any livestock (0.45) (0.48) (0.01) 0.25 0.04 0.21*** Large livestock (buffalo, bullock, milk cow) (0.42) (0.20) (0.01) Source: Original table for this publication based on calculations using data of BIHS (Bangladesh Integrated Household Survey) 2018–19, International Food Policy Research Institute, Washington, DC, https://doi.org/10.7910/DVN/NXKLZJ, Harvard Dataverse, V2. Note: SD = standard deviation. SE = standard error. Significance level: * = 10 percent, ** = 5 percent, *** = 1 percent. Table A.4: Time use, hours per day Task Men, mean (SD) Women, Difference, mean (SD) men – women (SE) 0.79 6.11 –5.32*** Chores, including dependent care (1.54) (3.37) (0.06) 3.59 6.38 –2.79*** In the company of a child (4.83) (7.15) (0.08) Source: Original table for this publication based on calculations using data of TUS (Time Use Survey) 2021, Bangladesh Bureau of Statistics and UN Women Bangladesh, Dhaka, Bangladesh. Note: SD = standard deviation. SE = standard error. Significance level: * = 10 percent, ** = 5 percent, *** = 1 percent. Table A.5: Social norms Survey question Men, mean (SD) Women, Difference, mean (SD) men – women (SE) Agree or strongly agree: if jobs are scarce, a man 0.82 0.73 0.09*** should have more right than a woman to a job (0.38) (0.44) (0.02) Agree or strongly agree: if a woman earns more 0.64 0.55 0.09*** than her husband, this is almost certain to cause problems (0.48) (0.50) (0.03) Source: Original table for this publication based on calculations using data of WVS (World Values Survey) Wave 7 (2017–2022), World Values Survey Association, King's College, Old Aberdeen, UK, https://doi.org/10.14281/18241.24. Note: SD = standard deviation. SE = standard error. Significance level: * = 10 percent, ** = 5 percent, *** = 1 percent. Appendices 121 Appendix B. Technical Appendix B.1 Aggregation of Activity Codes in the Time Use Survey 2021 Table B.1: Mapping the report’s activity categories to time use survey activity codes Report’s broad Major Division activity (UN DESA 2021) Major Division activity category (1-digit code) (UN DESA 2021) Employment and related activities 1 Work / production Production of goods for own final use 2 Unpaid volunteer, trainee, and other unpaid work 5 Unpaid domestic services for household and family 3 Housework / members dependent care Unpaid caregiving services for household and family 4 members Education Learning 6 Socializing and communication, community participation, 7 and religious practice Rest / leisure / self-care Culture, leisure, mass media, and sports practices 8 Self-care and maintenance 9 Source: Original table for this publication based on UN DESA 2021. Note: Major Division Code 5 (unpaid volunteer, trainee, and other unpaid work) refers to activities performed for individuals outside the household (UN DESA 2021). This is categorized as work/production because unpaid work is considered employment under the 13th ICLS definition (Gaddis et al. 2023; ILO 1983). B.2 Oaxaca Decomposition B.2.1 Decomposition Results Table B.2: Oaxaca decomposition of Ln (hourly earnings) Employment Region Gender gap Explained by status endowments, % Urban + rural 0.312 22 All Urban 0.397 46 Rural 0.305 0 Source: Original figure for this publication based on calculations using data of HIES (Household Income and Expenditure Survey) 2022, Bangladesh Bureau of Statistics, Dhaka, Bangladesh, accessed through SARRAW (South Asia Raw) and SARMD (South Asia Regional Micro Database) (databases), South Asia Region Team for Statistical Development, World Bank, Washington, DC. Note: The analysis accounts for the following endowments: age, age squared, household size, and indicators for ever-married status, urban region, a child age up to 6 in the household, an individual age 60 or more in the household, household ownership of a cellphone, relationship to household head, and educational attainment. Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 122 B.2.2 Endowments, by Sex Table B.3: Endowments, by sex Endowment Men, mean (SD) Women, mean Difference, (SD) men – women (SE) 38.24 36.38 1.86*** Age (12.66) (11.20) (0.30) 1,623.62 1,455.13 168.50*** Age squared (1,004.71) (859.25) (22.11) 0.32 0.38 –0.06*** Urban (0.46) (0.47) (0.02) 0.47 0.40 0.08*** Child, age 0–6 in household (0.50) (0.48) (0.01) 0.32 0.28 0.03*** Individual age 60 or more in household (0.46) (0.44) (0.01) 4.80 4.19 0.61*** Household size, members (1.90) (1.72) (0.05) 0.99 0.98 0.01*** Household owns a cellphone (0.08) (0.12) (0.00) 0.71 0.22 0.48*** Relationship: household head (0.45) (0.41) (0.01) 0.01 0.57 –0.56*** Relationship: spouse (0.08) (0.48) (0.01) 0.24 0.10 0.14*** Relationship: son or daughter (0.43) (0.29) (0.01) 0.01 0.03 –0.02*** Relationship: parent (0.08) (0.16) (0.00) 0.04 0.08 –0.04*** Relationship: other relative (0.19) (0.26) (0.01) 0.00 0.00 –0.00 Relationship: domestic worker (0.03) (0.06) (0.00) 0.00 0.00 –0.00 Relationship: other (0.02) (0.03) (0.00) 0.80 0.89 –0.09*** Ever married (0.40) (0.30) (0.01) 0.14 0.14 0.00 Education: none or some primary (0.35) (0.34) (0.01) 0.49 0.58 –0.09*** Education: completed primary (0.50) (0.48) (0.01) 0.13 0.09 0.03*** Education: some secondary, class 10 (0.34) (0.29) (0.01) 0.10 0.07 0.04*** Education: secondary , class 12 (0.31) (0.24) (0.01) 0.13 0.12 0.01 Education: secondary + any tertiary (0.34) (0.32) (0.01) Source: Original figure for this publication based on calculations using data of HIES (Household Income and Expenditure Survey) 2022, Bangladesh Bureau of Statistics, Dhaka, Bangladesh, accessed through SARRAW (South Asia Raw) and SARMD (South Asia Regional Micro Database) (databases), South Asia Region Team for Statistical Development, World Bank, Washington, DC. Note: SD = standard deviation. SE = standard error. Significance level: * = 10 percent, ** = 5 percent, *** = 1 percent. Appendices 123 B.3 Evidence Assessment Methodology and Expert Survey B.3.1 Policy Priorities: Assessment Methodology To assess the strength of the evidence associated with each identified policy priority, the analysis here follows the same methodology as the World Bank (2024g). First, a review of the existing literature on each relevant topic was conducted to identify studies examining the constraints on women’s economic empowerment and potentially relevant policy tools. This review included published articles, working papers, and literature reviews, with a total of 466 papers. Second, from this set of papers, only high-quality impact evaluations were reviewed. A study is considered a high quality impact evaluation if it assesses the impact of a policy tool using a causal methodology that compares the intervention group with a suitably identified comparison group. Such causal methodologies include randomized controlled trials, difference in differences designs (mostly two- way fixed effects methods), instrumental variables, and regression discontinuity designs. This criterion yielded 344 studies. Furthermore, studies with small sample sizes or high levels of attrition or that did not report gender disaggregated results were also excluded. This process resulted in a final set of 235 papers across all policy priorities. Next, the evidence base for each policy tool was classified based on two key dimensions, as follows: (1) the number of studies conducted and (2) the overall direction of the findings. This framework resulted in the following categorization of policy tools: → Conclusive: More than 10 high-quality studies evaluate the tool and find effects in the same direction. → Mixed: More than 10 high-quality studies evaluate the tool and find effects in different directions. → Emerging: There are 3–10 high-quality studies that evaluate the tool, regardless of the direction of effects. The evidence on the tool is growing, but is not yet sufficient to determine conclusively the direction of effects. → Limited: Two or fewer high-quality studies evaluate the tool, regardless of the direction of effects. The evidence on the tool is limited and insufficient to determine conclusively the direction of effects. → No evidence: No high-quality studies evaluate the impact of the tool, but there is potential for the tool to close gender gaps based on suggestive evidence, nonexperimental evidence, or the fact that policymakers or researchers are testing the tool because of the theoretical potential for addressing the constraint. Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 124 The classification of the evidence on policy tools did not influence the identification or categorization of policy tools or policy priorities. Summarizing the global evidence assessment presented in the third column of table ES.1 in the executive summary, each priority was assigned a qualifier based on the overall state of the available global evidence on the policy priority across all policy tools. Priorities are labeled according to the quality of the evidence. Sufficient evidence indicates that a large body of global evidence on the policy priority shows conclusively whether the use of the policy tools is effective in advancing women’s economic empowerment or not. Insufficient evidence indicates that the global evidence is either at an early stage or completely absent. Growing evidence indicates that the global evidence on the priority is either promising or mixed, but not yet sufficient to support the use of the policy tools under the priority. B.3.2 World Bank Expert Survey and the Table B.4: Summary statistics: the World Bank Experts Survey Ranking of Policy Priorities Policy priority Mean Standard To prioritize the constraints based on their significance deviation in advancing women’s economic empowerment in Skills and Bangladesh, an online survey was conducted among 6.58 2.46 vocational training World Bank experts active in operational and research Transport, mobility, teams working on topics relevant to women’s economic 6.13 2.24 and safety empowerment in Bangladesh. The objective of the survey was to measure the relative importance of Financial inclusion 5.70 2.10 alleviating various constraints on enhancing women’s Ownership of labor market outcomes. property and other 5.43 2.77 assets A convenience sample of 165 World Bank experts was Childcare used. This sample was identified through relevant and home 5.35 2.72 internal distribution lists. The sample included responsibilities researchers, country team members, and task team Business leaders involved in ongoing or planned World Bank ownership and 4.74 2.32 business growth operations in Bangladesh. The survey respondents were asked to rank the nine policy priorities in the order of their Job search 4.51 2.18 importance for reviving female labor force participation and increasing women’s incomes in Bangladesh, with a Digital inclusion 4.28 1.99 focus on low- and middle-income households. Migration 2.29 1.82 The survey was anonymous, and participants did not receive any monetary compensation. It was conducted Observations 69 by email and took approximately one minute to complete. The response rate was 42 percent. Table B.4 presents Source: Original table for this publication based on data of the World Bank Expert Survey conducted by the authors. the results of the survey. The World Bank expert survey Note: The table presents summary statistics (mean and standard deviation) from the responses of the World Bank was complemented by key informant interviews with Expert Survey. The first column lists the policy priorities in descending order. For each priority, the mean and standard selected Bangladeshi experts on the issue of women's deviation of the score attributed to each policy priority are presented. A higher score implies higher priority. The total economic empowerment. number of respondents was 69. Appendices 125 B.4 Measurement Changes in Labor Force Survey 2022 B.4.1 Employment Trends among Women in the Bangladesh LFS Data from the 2022 round of the LFS report a substantial rise in the share of women ages 15–24 engaged in self-employed work in rural areas, accompanied by a decline in unpaid work (refer to figure B.1, panel a).59 This represents a 615 percent increase in the share of self-employment in this group over a period of six years (that is, between 2016 and 2022). Meanwhile, 2022 estimates of the employment status among men do not exhibit such stark changes from previous survey rounds. Figure B.1: Rural female a. Employment status population, by employment status and sector, ages 15–24 100% 100% 80% 80% (%) (%) 60% 60% Percentage Percentage 40% 40% 20% 20% 0% 0% 2005 2010 2013 2015 2016 2022 2005 2010 2013 2015 2016 2022 Year Year Wage Employee Unpaid Employee Employer Wage Employee Unpaid Employee Employer Self-Employed Other Self-Employed Other b. Employment sector 100% 100% 80% 80% (%) (%) 60% 60% Percentage Percentage 40% 40% 20% 20% Source: Original figure for this publication based on calculations using data of LFS (Labor Force Survey) 2005, 2010, 2013, 0% 0% 2015, 2016, 2022, Bangladesh Bureau of 2005 2010 2013 2015 2016 20152016 2022 Statistics, Dhaka, Bangladesh, accessed 2005 2010 2013 2015 2016 20152016 2022 Year through SARRAW (South Asia Raw) and Year GLDRAW (Global Labor Database Raw) (databases), South Asia Region Team Agriculture Industry Services Agriculture Industry Services for Statistical Development, World Bank, Washington, DC. 59 Unlike the rest of the report where “employer” and “self-employed” individuals are re-categorized into “employer” and “own-account worker” based on the reported firm size, the figures reported in this appendix use the raw employment status categories as reported in the LFS questionnaire in order to illustrate the changes in the LFS. Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 126 The 2022 round also shows a larger share of young women employed in the agricultural sector relative to previous survey years (refer to figure B.1, panel b). Similar to the trends in employment status, the share of men working in agriculture shows no significant shift between 2016 and 2022. The stark rise in rural self-employment and decline in unpaid work among women between 2016 and 2022 is driven by women employed in agriculture (refer to figure B.2, panels a and b). This stark increase in self-employment is not observed among urban women employed in nonagricultural work (refer to figure B.2, panel c). No other major compositional changes are observed in employment status among women in agriculture. The trends among men working in agriculture display some compositional variation across survey rounds but do not exhibit a similar stark rise in self-employment and decline in unpaid work. Figure B.2: Women’s a. Rural agriculture employment, by employment status, ages 15–24 100% Wage Employee Unpaid Employee 80% Employer Percentage (%) Self-Employed 60% Other 40% 20% 0% 2005 2010 2013 20152016 2022 Year b. Rural nonagriculture 100% 80% Percentage (%) 60% 40% 20% 0% 2005 2010 2013 20152016 2022 Year Appendices 127 Figure B.2 (continued) c. Urban nonagriculture 100% Wage Employee Unpaid Employee 80% Employer Percentage (%) Self-Employed 60% Other 40% Source: Original figure for this publication 20% based on calculations using data of LFS (Labor Force Survey) 2005, 2010, 2013, 2015, 2016, 2022, Bangladesh Bureau of Statistics, 0% Dhaka, Bangladesh, accessed through 2005 2010 2013 20152016 2022 SARRAW (South Asia Raw) and GLDRAW Year (Global Labor Database Raw) (databases), South Asia Region Team for Statistical Development, World Bank, Washington, DC. B.4.2 Causes of the Rise in Women’s Agricultural Self-Employment Observed in LFS 2022 Four factors in LFS 2016 and LFS 2022 may contribute to the observed trends. First, the LFS 2022 instrument underwent revisions by adding questions to identify persons in employment. The new survey instrument follows the latest ILO model questionnaires closely.60 The 2022 Bangladesh LFS questionnaire included four questions to identify employed individuals, compared with only two questions in 2016 (refer to table B.5). There are strong reasons to believe that these changes may have boosted the number of individuals identified as employed and may thus have contributed to the substantial rise in self-employment that has been documented. Table B.5: Questions measuring employment in LFS rounds 2005–22 (1) 2005 (2) 2010 (3) 2013 (4) 2015 (5) 2016 (6) 2022 S4Q4.1: Did you S4Q4.1: Did you S5Q39: Does she/ S4Q37: In the last seven S4Q31: In the last EMP_01: During do any economic do any economic he have any kind of days did you work for at seven days did you the last week, did activity for at least activity for at business/farm, for her/ least one hour in return work for at least one you do any work one hour or more least one hour or his own or with one or of pay or profit? hour in return for for a wage, salary, during the last week more during the more partners? pay or profit? commission, tips or as paid worker or for last seven days Example: job, business, any other pay, even family gain or profit as paid worker Examples: commercial rickshaw pulling, vending if only for one hour? or for your own use or or for household farming or fishing, vegetables... consumption? gain or profit or collecting firewood or water mainly for sale, Mainly: agricultural work for own use or for selling; consumption? selling things, making things for sale, repairing Example: Cultivating things, transport rice, wheat, potatoes, business, legal or and so on. medical practice, phone shop, barbecue, shoe shining, and so on. 60 Refer to LFS (Labour Force Survey) Questionnaire Toolkit, ILOSTAT, International Labour Organization, Geneva, https://ilostat.ilo.org/resources/lfs-toolkit/. Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 128 Table B.5 (continued) (1) 2005 (2) 2010 (3) 2013 (4) 2015 (5) 2016 (6) 2022 S4Q4.2: Even if S4Q4.2: Did S5Q40: Does she/ S4Q38: In the last seven S4Q32: In the last EMP_02: During the you did not work you have any he do any work for days did you have a job seven days did last week, did you during the last week work, business, wage, salary, or in kind or business where you you have a job or run or do any kind of for some reason, or economic (excluding domestic were absent temporarily business for pay business, farming, did you have a job activity from work)? and to which you will or profit where or other activity to attachment or which you were return to work? you were absent generate income/ engagement in any absent during Examples: A regular temporarily and to profit, even if only economic activity? the last seven job, contract, casual or which you will return for one hour? days? piece work for pay, work to work? in exchange for food or housing. S4Q4.3: Please, S5Q41: Does any work S4Q39: In the last seven EMP_03: During provide the as a domestic worker for days did you work at the last week, did reason why you a wage, salary, or any least one hour producing you help unpaid in did not work payment in kind? goods and services for a business owned during the last the household? by a household seven days. (If member, even if the response Rearing animals, only for one hour? was “Leave” producing vegetables, or “Maternity and so on, on own leave”) land, plot mainly for own consumption. Conducting agricultural work, such as cultivating rice, wheat, potatoes, and so on, mainly for own use. S5Q42: Does she/he S4Q40: In the last seven Q33: In the last EMP_04: In the help, without being days did you have a job seven days did you last week, did you paid, in any kind of or business where you work for at least work for at least business run by her/his were absent temporarily one hour to produce one hour to produce household? and to which you will goods and services goods and services return to work? for your own in agriculture or household? fishing for your own household? S5Q43: Does she/he EMP_05: In general, have a job or business are the products where he was absent obtained from this temporarily and to which activity for sale/ she/he will return to barter or for family work? use? Q34: In the last EMP_06: During seven days were you the last week, did absent temporarily you have a paid job from a job in which or a business from you produce goods which you were on and services for temporary absence your own household and to which you and to which you will expect to return? return to work? Source: Original table for this publication based on questionnaires of Bangladesh LFS 2005, 2010, 2013, 2015, 2016, 2022. Specifically, the 2022 LFS questionnaire used a dedicated question (EMP_02) to capture individuals engaged in any form of employment for profit. The question asked whether the respondent was engaged in “any kind of business or other activity to generate income/profit.” The word “profit” was somewhat deemphasized. This wording was informed by extensive ILO piloting and cognitive testing. The process showed that key terms and questions routinely used to identify persons who are employed—for example, “did you work for profit?”—were often not well understood by respondents and may have led to an underestimation of employment, especially among women (Benes and Walsh 2018). In addition, the 2022 LFS questionnaire Appendices 129 included a dedicated question (EMP_03) that elicited specifically whether the individual helped, unpaid, in a business owned by a household member. In the LFS 2016, individuals were screened out of the employment module if they did not report that they had worked for pay or profit or that they had produced goods and services for their households. In contexts with strong gendered social norms and gendered roles (such as Bangladesh), women may not consider such help in household businesses as employment, leading to their exclusion from employment as measured in the 2016 LFS question (Müller and Sousa 2020). A large body of methodological evidence, including from South Asia, shows that changes, such as the ones above, can substantially increase the share of women captured in employment, especially among low-hour workers in agriculture, relative to the much smaller increases among men (Discenza et al. 2021). Second, the LFS 2022 instrument also involved revisions in the descriptions of employment status by distinguishing more explicitly between “own-account work (without regular employees in own business activity, in own agriculture activity)” and work as an “employer (with regular employees).” The instruments used in previous survey rounds provided less detailed descriptions of self-employment, thus: “self-employed,” “self-employed in agriculture,” or “self-employed in nonagriculture.” The more explicit distinctions in the new description may have affected how employment among some women was captured. As highlighted in recent methodological work, changes in question wording or categorization can lead to nontrivial differences in the magnitude of employment, and the effects vary by sex (Bardasi et al. 2011; Contreras et al. 2024). Third, changes in data collection protocols may have also contributed to the observed patterns. For example, LFS 2022 introduced computer-assisted personal interviewing to conduct data collection, and the number of days spent training enumerators rose from 7 in 2016 to 10 in 2022. Fourth, the sample frame of LFS 2016 used enumeration areas based on the Population and Housing Census 2011. It is common practice to adjust the survey weights of nationally representative household surveys based on the enumeration of the following population census to improve the accuracy of the weights. However, the weights in LFS 2016 were not adjusted based on the Population and Housing Census 2022, potentially affecting the accuracy of population-level estimates in LFS 2016. Meanwhile, the sample frame for LFS 2022 relied on the enumeration areas used in the Population and Housing Census 2022. There is little evidence of other changes that might explain these patterns, such as changes in household size or household gender composition, which might indicate a change in sampling methodology. Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 130 B.4.3 Conclusion The absence of similar trends among men suggests that the substantial changes in women’s employment patterns revealed in the LFS may stem from updates in the way employment is measured in LFS 2022. These updates, rather than true changes in the population, appear to have affected the results on women’s employment. Together, the patterns suggest that, while nonagricultural wage employment among women in rural and urban areas and nonagricultural self- employment among women in urban areas may be comparable across rounds of the Bangladesh LFS, the trends in agricultural self-employment resulting from the measurement changes in LFS 2022 preclude a comparison of women’s agricultural self-employment and agricultural unpaid employment between LFS 2022 and previous LFS rounds. B.5. Changes in the ILO Definition of Employment Compared with the 13th ICLS in 1982, the 19th ICLS in 2013 introduced a more nuanced framework for measuring labor market engagement that included a revision in the definition of employment. Under the revised standards, activities related mainly to the production of goods or to production only for household consumption are excluded from the employment classification (Sodergren and Villarreal-Fuentes 2022).61 Individuals engaged exclusively in these activities are classified as own-use production workers and are no longer considered employed under the 19th ICLS standards. This change has important implications for the measurement of employment, particularly women’s employment. First, the application of the 19th ICLS standard may lead to significantly lower estimates of employment in regions with high levels of subsistence farming (Sodergren and Villarreal-Fuentes 2022). Indeed, Gaddis et al. (2023) find that the revised standards resulted in lower employment-to- population estimates in Sub-Saharan African countries, especially in rural areas. This decline may be larger among women, who are more likely than men to produce goods and services for household consumption. Survey timing may also lead to variations in employment estimates if farmers engage in cultivating different crops and livestock products throughout the agricultural year, and these products vary in the degree to which they are intended for household consumption. Figure B.3 shows the levels of employment among men and women ages 15–64 in LFS 2022 using the two definitions. The share of men in employment changed little because of the 13th and 19th ICLS definitions, but the share of women in employment decreased significantly, dropping by 23 percentage points among 61 Activities related to the production of services for household consumption, which were already excluded from the scope of employment under the 13th ICLS, remain excluded. Appendices 131 women ages 15–64 and by 33 percentage points among women ages 15–24 (refer to figure B.3). This decline is concentrated almost entirely among rural women (refer to figures B.4 and B.5). These estimates are also consistent with those based on the 2022 population shares in employment, in the labor force, and not in the labor force, as reported using both the 13th and 19th ICLS definitions in the preliminary LFS 2024 Quarter 3 key indicators report (BBS 2025b).62 Figure B.3: Share of population a. Ages 15–64 in employment, 13th ICLS and 19th ICLS, by age and sex 100% Female 80 78 80% Male Percentage (%) 60% 43 40% 20 20% 0% 13th ICLS 19th ICLS b. Ages 15–24 100% 80% Percentage (%) 60% 46 47 46 40% 20% 13 Source: Original figure for this publication based on calculations using data of LFS 0% (Labor Force Survey) 2022, Bangladesh 13th ICLS 19th ICLS Bureau of Statistics, Dhaka, Bangladesh, accessed through SARRAW (South Asia Raw) (database), South Asia Region Team for Statistical Development, World Bank, Washington, DC. 62 Figures B.4 and B.5 refer to ages 15–64 and 15–24, while the population numbers provided in BBS (2025b) refer to ages 15+. While not reported here, the estimates on ages 15–64 calculated in the data here align with those calculated using the numbers in the key indicators report (BBS 2025b). Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 132 Figure B.4: Share of a. Rural areas population in employment, 13th ICLS and 19th ICLS, ages 15–64, by region and sex 100% 80 78 Female 80% Male Percentage (%) 60% 52 40% 19 20% 0% 13th ICLS 19th ICLS b. Urban areas 100% 79 79 80% Percentage (%) 60% 40% 23 21 20% Source: Original figure for this publication based on calculations using data of LFS (Labor Force Survey) 2022, Bangladesh 0% Bureau of Statistics, Dhaka, Bangladesh, 13th ICLS 19th ICLS accessed through SARRAW (South Asia Raw) (database), South Asia Region Team for Statistical Development, World Bank, Washington, DC. Appendices 133 Figure B.5: Share of a. Rural areas population in employment, 13th ICLS and 19th ICLS, Ages 15–24, by region and sex 100% Female 80% Male 60 Percentage (%) 60% 48 47 40% 20% 11 0% 13th ICLS 19th ICLS b. Urban areas 100% 80% Percentage (%) 60% 45 44 40% 20% 17 16 Source: Original figure for this publication based on calculations using data of LFS (Labor Force Survey) 2022, Bangladesh 0% Bureau of Statistics, Dhaka, Bangladesh, 13th ICLS 19th ICLS accessed through SARRAW (South Asia Raw) (database), South Asia Region Team for Statistical Development, World Bank, Washington, DC. Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 134 References Abate, Gashaw T., Kibrom A. Abay, Jordan Chamberlin, Aguilar, Arturo, Emilio Gutiérrez, and Paula Soto Villagrán. Yumna Kassim, David J. Spielman, and Martin Paul 2021. “Benefits and Unintended Consequences Jr Tabe-Ojong. 2023. “Digital Tools and Agricultural of Gender Segregation in Public Transportation: Market Transformation in Africa: Why Are They Not at Evidence from Mexico City’s Subway System.” Scale Yet, and What Will It Take to Get There?” Food Economic Development and Cultural Change 69 (4): Policy 116 (April), 102439. https://doi.org/10.1016/j. 1379–1410. https:/ /doi.org/10.1086/707421. foodpol.2023.102439. Ahmed, Akhter U., Mehnaz Rabbani, Munshi Sulaiman, and Abebe, Girum, Antonio Stefano Caria, and Esteban Ortiz- Narayan C. Das. 2009. “The Impact of Asset Transfer Ospina. 2021. “The Selection of Talent: Experimental on Livelihoods of the Ultra Poor in Bangladesh.” and Structural Evidence from Ethiopia.” American Research Monograph 39 (April), Research and Economic Review 111 (6): 1757–1806. Evaluation Division, BRAC Center, Dhaka, Bangladesh. Abel, Simon Martin, Roelof Petrus Burger, and Patrizio Ahmed, Hamna, Mahreen Mahmud, Farah Said, and Zunia Piraino. 2020. “The Value of Reference Letters: S. Tirmazee. 2024. “Encouraging Female Graduates Experimental Evidence from South Africa.” American to Enter the Labor Force: Evidence from a Role Model Economic Journal: Applied Economics 12 (3): 40–71. Intervention in Pakistan.” Economic Development https://doi.org/10.1257/app.20180666. and Cultural Change 72 (2): 919–57. https://doi. org/10.1086/721615. Abiona, Olukorede, and Martin Foureaux Koppensteiner. 2022. “Financial Inclusion, Shocks, and Poverty: Ahmed, Tanima, and Aphichoke Kotikula. 2021. “Women’s Evidence from the Expansion of Mobile Money in Employment and Safety Perceptions: Evidence from Tanzania.” Journal of Human Resources 57 (2): 435– Low-Income Neighborhoods of Dhaka, Bangladesh.” 64. https://doi.org/10.3368/jhr.57.2.1018-9796R1. Policy Brief (August), South Asia Gender Innovation Lab, World Bank, Washington, DC. http://documents. Acevedo, Paloma, Guillermo A. Cruces, Paul J. Gertler, worldbank.org/curated/en/629261628580364879. and Sebastián Martínez. 2020. “How Vocational Education Made Women Better Off but Left Men Ajayi, Kehinde F., Aziz Dao, and Mousson Estelle Jamel Behind.” Labour Economics 65 (August), 101824. Koussoubé. 2022. “The Effects of Childcare on Women https://doi.org/10.1016/j.labeco.2020.101824. and Children.” CGD Working Paper 628 (November), Center for Global Development, Washington, DC. ADB (Asian Development Bank) and ILO (International Labour Organization). 2016. Bangladesh: Looking Aker, Jenny C., Rachid Boumnijel, Amanda McClelland, Beyond Garments; Employment Diagnostic Study. and Niall Tierney. 2016. “Payment Mechanisms Geneva: ILO; Manilla: ADB. and Antipoverty Programs: Evidence from a Mobile Money Cash Transfer Experiment in Niger.” Economic Adhvaryu, Achyuta, Namrata Kala, and Anant Nyshadham. Development and Cultural Change 65 (1): 1–37. 2023. “Returns to On-the-Job Soft Skills Training.” https://doi.org/10.1086/687578. Journal of Political Economy 131 (8): 2165–2208. https://doi.org/10.1086/724320. Alaref, Jumana Jamal Subhi, Stefanie Brodmann, and Patrick Premand. 2020. “The Medium-Term Adoho, Franck M., Shubha Chakravarty, Dala T. Korkoyah Jr., Impact of Entrepreneurship Education on Labor Mattias K. A. Lundberg, and Afia Tasneem. 2014. “The Market Outcomes: Experimental Evidence from Impact of an Adolescent Girls Employment Program: University Graduates in Tunisia.” Labour Economics The EPAG Project in Liberia.” Policy Research Working 62 (January), 101787. https://doi.org/10.1016/j. Paper 6832, World Bank, Washington, DC. labeco.2019.101787. Afridi, Farzana, Amrita Dhillon, Sanchari Roy, and Nikita Alba-Vivar, Fabiola M. 2024. “Opportunity Bound: Sangwan. 2022. “Social Networks, Gender Norms Transport and Access to College in a Megacity.” Paper and Women’s Labor Supply: Experimental Evidence presented at the Annual World Bank Conference on Using a Job Search Platform.” IZA Discussion Paper Transport Economics, Washington, DC, March 22, DP 15767 (November), Institute of Labor Economics, 2024. https://drive.google.com/file/d/1WYmHt47- Bonn, Germany. bEhejPCjSdd9hf9XBzvmZrjm/view. Agarwal, Sumit, Shashwat Alok, Pulak Ghosh, Soumya Alhorr, Layane. 2024. “Virtual Windows through Glass Kanti Ghosh, Tomasz Piskorski, and Amit Seru. 2017. Walls? Digitalization for Low-Mobility Female “Banking the Unbanked: What Do 255 Million New Entrepreneurs.” Policy Research Working Paper Bank Accounts Reveal about Financial Access?” 10803, World Bank, Washington, DC. Columbia Business School Research Paper 17–12, Columbia Business School, Columbia University, New York. References 135 Ali, Daniel Ayalew, Klaus W. Deininger, and Markus P. Anukriti, S., Maurizio Bussolo, Jean Nahrae Lee, Siddharth Goldstein. 2014. “Environmental and Gender Impacts Sharma, and Anaise Williams. 2024. “Evaluating the of Land Tenure Regularization in Africa: Pilot Evidence Impacts of Access to a Digital Jobs Platform and from Rwanda.” Journal of Development Economics Skills Training on Female Domestic Workers.” Working 110 (September): 262–75. https:/ /doi.org/10.1016/j. paper, World Bank, Washington, DC. jdeveco.2013.12.009. Anukriti, S., Lelys Ileana Dinarte-Diaz, Marina Elefante, Aloud, Monira Essa, Sara Al-Rashood, Ina Ganguli, and Maria Montoya-Aguirre, and Alena Sakhonchik. Basit Zafar. 2020. “Information and Social Norms: 2023. “Filling the Gaps: Childcare Laws for Women’s Experimental Evidence on the Labor Market Economic Empowerment.” Policy Research Working Aspirations of Saudi Women.” NBER Working Paper Paper 10492, World Bank, Washington, DC. 26693 (December), National Bureau of Economic Research, Cambridge, MA. https:/ /doi.org/10.3386/ Archibong, Belinda, Francis Annan, Anja Benshaul- w26693. Tolonen, Oyebola Okunogbe, and Ifeatu Oliobi. 2025. “Firm Culture: How Social Norms Affect Gender Bias Alzúa, María Laura, Soyolmaa Batbekh, Altantsetseg in Hiring in Online Labor Markets.” G²LM|LIC Working Batchuluun, Bayarmaa Dalkhjav, and José Galdo. Paper 95 (May), Gender, Growth, and Labour Markets 2021. “Demand-Driven Youth Training Programs: in Low-Income Countries Programme, Foreign, Experimental Evidence from Mongolia.” World Bank Commonwealth, and Development Office, London; Economic Review 35 (3): 720–44. Institute of Labor Economics, Bonn, Germany. Alzúa, María Laura, Guillermo A. Cruces, and Carolina Arráiz, Irani, Syon P. Bhanot, and Carla Calero. 2019. “Less Lopez. 2016. “Long-Run Effects of Youth Training is More: Experimental Evidence on Heuristic-Based Programs: Experimental Evidence from Argentina.” Business Training in Ecuador.” IDB Invest Technical Economic Inquiry 54 (4): 1839–59. https://doi. Note 8 (July), Development through the Private org/10.1111/ecin.12348. Sector Series, Inter-American Development Bank, Washington, DC. Amaral, Sofia Fernando Monteiro. 2017. “Do Improved Property Rights Decrease Violence against Women Asadullah, M. Niaz, and Jinnat Ara. 2016. “Evaluating in India?” ISER Working Paper 2017–13 (October), the Long-Run Impact of an Innovative Anti-Poverty Institute for Social and Economic Research, Programme: Evidence Using Household Panel Data.” University of Essex, Colchester, UK. Applied Economics 48 (2): 107–20. https://doi.org/1 0.1080/00036846.2015.1073846. Amaral, Sofia Fernando Monteiro, Girija Borker, Nathan Fiala, Anjani Kumar, Nishith Prakash, and Maria Ashraf, Nava, Gharad Bryan, Alexia Delfino, Emily Micaela Sviatschi. 2025. “Sexual Harassment in Holmes, Leonardo Iacovone, and Ashley Pople. Public Spaces and Police Patrols: Experimental 2022. “Learning to See the World’s Opportunities: Evidence from Urban India.” Quarterly Journal of The Impact of Imagery on Entrepreneurial Success.” Economics. Published ahead of print, May 29, 2025. Innovation for Poverty Action Working Paper 1–80, https://doi.org/10.1093/qje/qjaf026. London School of Economics and Political Science, London. Amaral, Sofia Fernando Monteiro, Aixa Garcia-Ramos, Selim Gulesci, Alejandra Ramos, Sarita P. Ore- Ashraf, Nava, Dean S. Karlan, and Wesley Yin. 2006. “Tying Quispe, and Maria Micaela Sviatschi. 2024. “Gender- Odysseus to the Mast: Evidence from a Commitment Based Violence in Schools and Girls’ Education: Savings Product in the Philippines.” Quarterly Experimental Evidence from Mozambique.” NBER Journal of Economics 121 (2): 635–72. https://doi. Working Paper 33203 (November), National Bureau org/10.1162/qjec.2006.121.2.635. of Economic Research, Cambridge, MA. http:/ /www. nber.org/papers/w33203. Atkin, David Guy, Antoinette Schoar, and Sumit Shinde. 2023. “Working from Home, Worker Sorting, and Anderson, Siwan, and Garance Genicot. 2015. Development.” NBER Working Paper 31515 (July), “Suicide and Property Rights in India.” Journal of National Bureau of Economic Research, Cambridge, Development Economics 114 (May): 64–78. https:/ / MA. doi.org/10.1016/j.jdeveco.2014.11.004. Attanasio, Orazio Pietro, Britta Augsburg, and Ralph Angelucci, Manuela, Rachel Heath, and Eva Noble. De Haas. 2019. “Microcredit Contracts, Risk 2022. “Graduation Programs Targeting Women: Diversification, and Loan Take-Up.” Journal of the Evidence from the Democratic Republic of Congo.” European Economic Association 17 (6): 1797–1842. CEGA Working Paper WPS-205 (April 26), Center https://doi.org/10.1093/jeea/jvy032. for Effective Global Action, University of California, Berkeley, CA. Attanasio, Orazio Pietro, Britta Augsburg, Ralph De Haas, Emla Fitzsimons, and Heike Harmgart. 2015. “The Angelucci, Manuela, Dean S. Karlan, and Jonathan Impacts of Microfinance: Evidence from Joint- Zinman. 2015. “Microcredit Impacts: Evidence from Liability Lending in Mongolia.” American Economic a Randomized Microcredit Program Placement Journal: Applied Economics 7 (1): 90–122. https:// Experiment by Compartamos Banco.” American doi.org/10.1257/app.20130489. Economic Journal: Applied Economics 7 (1): 151–82. https://doi.org/10.1257/app.20130537. Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 136 Attanasio, Orazio Pietro, Ricardo Paes de Barros, Pedro Balboni, Clare A., Oriana Bandiera, Robin Burgess, Manuel Carneiro, David K. Evans, Lycia Lima, Pedro Maitreesh Ghatek, and Anton Heil. 2022. “Why do Olinto, and Norbert R. Schady. 2017. “Impact of Free People Stay Poor?” Quarterly Journal of Economics Availability of Public Childcare on Labour Supply and 137 (2): 785–844. https://doi.org/10.1093/qje/ Child Development in Brazil.” 3ie Impact Evaluation qjab045. Report 58 (April), International Initiative for Impact Evaluation, New Delhi. https://doi.org/10.23846/ Bandiera, Oriana, Robin Burgess, Narayan Das, Selim ow2223. Gulesci, Imran Rasul, and Munshi Sulaiman. 2017. “Labor Markets and Poverty in Village Economies.” Attanasio, Orazio Pietro, Arlen Guarín, Carlos Medina, Quarterly Journal of Economics 132 (2): 811–70. and Costas Meghir. 2017. “Vocational Training for https://doi.org/10.1093/qje/qjx003. Disadvantaged Youth in Colombia: A Long-Term Follow-Up.” American Economic Journal: Applied Banerjee, Abhijit Vinayak, Emily Breza, Esther Duflo, and Economics 9 (2): 131–43. https://doi.org/10.1257/ Cynthia Kinnan. 2019. “Can Microfinance Unlock app.20150554. a Poverty Trap for Some Entrepreneurs?” NBER Working Paper 26346 (October), National Bureau of Attanasio, Orazio Pietro, and Marcos Vera-Hernandez. Economic Research, Cambridge, MA. 2004. “Medium and Long Run Effects of Nutrition and Child Care: Evaluation of a Community Nursery Banerjee, Abhijit Vinayak, Esther Duflo, Rachel Programme in Rural Colombia.” Working Paper Glennerster, and Cynthia Kinnan. 2015. “The Miracle EWP04/06, Centre for the Evaluation of Development of Microfinance? Evidence from a Randomized Policies, Institute for Fiscal Studies, London. Evaluation.” American Economic Journal: Applied Economics 7 (1): 22–53. https://doi.org/10.1257/ Azam, Md Shah, Md Morsalin, Md Rakibul Hafiz Khan Rakib, app.20130533. and Shah Alam Kabir Pramanik. 2023. “Adoption of Electronic Commerce by Individuals in Bangladesh.” Banerjee, Abhijit Vinayak, Esther Duflo, Nathanael Information Development 39 (4): 764–86. https://doi. Goldberg, Dean S. Karlan, Robert Osei, William org/10.1177/02666669211052523. Parienté, Jeremy Shapiro, Bram Thuysbaert, and Christopher R. Udry. 2015. “A Multifaceted Program Azevedo, João Pedro. 2011. “wbopendata: Stata Module to Causes Lasting Progress for the Very Poor: Evidence Access World Bank Databases.” Statistical Software from Six Countries.” Science 348 (6236): 772–89. Components S457234, Department of Economics, https://doi.org/10.1126/science.1260799. Boston College, Chestnut Hill, MA, accessed July 25, 2025, http://ideas.repec.org/c/boc/bocode/ Banerjee, Abhijit Vinayak, Esther Duflo, and Garima s457234.html. Sharma. 2021. “Long-Term Effects of the Targeting the Ultra Poor Program.” American Economic Review: Bachas, Pierre, Paul J. Gertler, Sean Higgins, and Enrique Insights 3 (4): 471–86. https://doi.org/10.1257/ Seira. 2018. “Digital Financial Services Go a Long aeri.20200667. Way: Transaction Costs and Financial Inclusion.” AEA Papers and Proceedings 108 (May): 444–48. https:/ / BANBEIS (Bangladesh Bureau of Educational Information doi.org/10.1257/pandp.20181013. and Statistics). 2024. “Bangladesh Education Statistics 2023.” Ministry of Education, Government Bachas, Pierre, Paul J. Gertler, Sean Higgins, and Enrique of Bangladesh, Dhaka, Bangladesh. Seira. 2021. “How Debit Cards Enable the Poor to Save More.” Journal of Finance 76 (4): 1913–57. Bangladesh Bank. 2014. “A Primary Survey on Banks in https://doi.org/10.1111/jofi.13021. Promoting Women Entrepreneurship in Bangladesh.” Special Research Work 1501 (September), Monetary Bahia, Kalvin, Pau Castells, Genaro Cruz, Takaaki Masaki, Policy Department and Research Department, Carlos Rodríguez-Castelán, and Viviane Sanfelice. Bangladesh Bank, Dhaka, Bangladesh. https://www. 2023. “Mobile Broadband, Poverty, and Labor bb.org.bd/pub/research/sp_research_work/1501. Outcomes in Tanzania.” World Bank Economic pdf. Review 37 (2), 235–56. https://doi.org/10.1093/wber/ lhad003. Barboni, Giorgia, Anwesha Bhattacharya, Erica M. Field, Rohini P. Pande, Natalia Rigol, Simone Schaner, Aruj Bahl, Kanika. 2019. “We’re Shutting Down No Lean Shukla, and Charity M. Troyer Moore. 2024. “Hold Season, Our Seasonal Migration Program: Here’s the Phone: The Short- and Long-Run Impacts of Why.” Newsroom (blog), June 6, 2019. https://www. Connecting Indian Women to Digital Technology.” EGC evidenceaction.org/newsroom/were-shutting- Discussion Paper 1109, Economic Growth Center, Yale down-no-lean-season-our-seasonal-migration- University, New Haven, CT. https:/ /elischolar.library. program-heres-why. yale.edu/egcenter-discussion-paper-series. Bahrami-Rad, Duman. 2021. “Keeping It in the Family: Bardasi, Elena, Kathleen Beegle, Andrew Dillon, and Female Inheritance, Inmarriage, and the Status Pieter Serneels. 2011. “Do Labor Statistics Depend of Women.” Journal of Development Economics on How and to Whom the Questions Are Asked? 153 (November), 102714. https://doi.org/10.1016/j. Results from a Survey Experiment in Tanzania.” World jdeveco.2021.102714. Bank Economic Review 25 (3): 418–47. https:/ /doi. org/10.1093/wber/lhr022. References 137 Barrera-Osorio, Felipe, Adriana Kugler, and Mikko Silliman. Bangladesh. https://bbs.portal.gov.bd/sites/default/ 2023. “Hard and Soft Skills in Vocational Training: files/files/bbs.portal.gov.bd/page/57def76a_ Experimental Evidence from Colombia.” World aa3c_46e3_9f80_53732eb94a83/2025-03-03- Bank Economic Review 37 (3): 409–36. https:/ /doi. 05-48-f27ce30efbbcfe33a96d2c808b5908f3.pdf. org/10.1093/wber/lhad007. BBS (Bangladesh Bureau of Statistics). 2025b. “Quarterly Barros, Ricardo Paes de, Pedro Olinto, Trine Lunde, and Labor Force Survey 2024, Bangladesh (3rd Quarter, Mirela de Carvalho. 2011. “The Impact of Access to July to September)” [In Bengali]. January, BBS, Free Childcare on Women’s Labor Market Outcomes: Statistics and Informatics Division, Ministry of Evidence from a Randomized Trial in Low-income Planning, Dhaka, Bangladesh. https://bbs.portal. Neighborhoods of Rio de Janeiro.” Paper prepared gov.bd/sites/default/files/files/bbs.portal.gov.bd/ for the 2011 World Bank Economists Forum, World page/96220c5a_5763_4628_9494_950862accd Bank, Washington, DC, March. 8c/2025-01-05-10-36-7bcad4a11d26e8ee63c9ea ceb5f43b10.pdf. Bastian, Gautam G., Iacopo Bianchi, Markus P. Goldstein, and Joao Montalvao. 2018. “Short-Term Impacts of Beam, Emily A. 2016. “Do Job Fairs Matter? Experimental Improved Access to Mobile Savings, with and without Evidence on the Impact of Job-Fair Attendance.” Business Training: Experimental Evidence from Journal of Development Economics 120 (May): 32– Tanzania.” CGD Working Paper 478 (March), Center 40. https:/ /doi.org/10.1016/j.jdeveco.2015.11.004. for Global Development, Washington, DC. https:// EconPapers.repec.org/RePEc:cgd:wpaper:478. Beaman, Lori A., and Andrew Dillon. 2018. “Diffusion of Agricultural Information within Social Networks: Bates, Mary Ann, and Rachel Glennerster. 2017. “The Evidence on Gender Inequalities from Mali.” Journal Generalizability Puzzle.” Stanford Social Innovation of Development Economics 133 (July): 147–61. Review 15 (3): 50–54. https://ssir.org/articles/entry/ https://doi.org/10.1016/j.jdeveco.2018.01.009. the_generalizability_puzzle. Beaman, Lori A., Niall Keleher, and Jeremy Magruder. Batista, Catia, Sandra Sequeira, and Pedro C. Vicente. 2018. “Do Job Networks Disadvantage Women? 2022. “Closing the Gender Profit Gap?” Management Evidence from a Recruitment Experiment in Malawi.” Science 68 (12): 8553–67. https://doi.org/10.1287/ Journal of Labor Economics 36 (1): 121–57. https:// mnsc.2022.4579. doi.org/10.1086/693869. Baul, Tushi, Dean S. Karlan, Kentaro Toyama, and Kathryn Bedoya, Guadalupe, Yulia Belyakova, Aidan Coville, Thomas Vasilaky. 2024. “Improving Smallholder Agriculture Escande, Mohammad Isaqzadeh, and Aminata via Video-Based Group Extension.” Journal of Ndiaye. 2023. “The Enduring Impacts of a Big Push Development Economics 169 (June), 103267. during Multiple Crises: Experimental Evidence from https://www.sciencedirect.com/science/article/pii/ Afghanistan.” Policy Research Working Paper 10596, S0304387824000166. World Bank, Washington, DC. BBS (Bangladesh Bureau of Statistics). 2020. Bangladesh Bedoya, Guadalupe, Aidan Coville, Johannes Haushofer, Standard Industrial Classification 2020. January. Mohammad Isaqzadeh, and Jeremy P. Shapiro. 2019. Dhaka, Bangladesh: BBS, Statistics and Informatics “No Household Left Behind: Afghanistan Targeting Division, Ministry of Planning. https://bbs.portal. the Ultra Poor Impact Evaluation.” Policy Research gov.bd/sites/default/files/files/bbs.portal. Working Paper 8877, World Bank, Washington, DC. gov.bd/page/745673c8_c7ed_49bc_a4e2_ e7b05fe7a9d4/2021-06-08-08-39-6be22d0db12 Beegle, Kathleen, Joachim De Weerdt, and Stefan b4087042cca9e0ff9222e.pdf. Dercon. 2011. “Migration and Economic Mobility in Tanzania: Evidence from a Tracking Survey.” Review BBS (Bangladesh Bureau of Statistics). 2023. Labour of Economics and Statistics 93 (3): 1010–33. Force Survey (LFS) 2022, Bangladesh. October. Dhaka, Bangladesh: BBS, Statistics and Informatics Benes, Elisa M., and Kieran Walsh. 2018. “ILO LFS Pilot Division, Ministry of Planning. https://bbs.portal. Studies, Cognitive Interviewing Tests: Methodology, gov.bd/sites/default/files/files/bbs.portal.gov.bd/ Process, and Outcomes.” ILO Statistical page/b343a8b4_956b_45ca_872f_4cf9b2f1a Methodological Series 2 (May), Department of 6e0/2023-10-25-07-38-4304abd7a3f3d8799fbcb Statistics, International Labour Office, Geneva. 59ff91007b1.pdf. https://www.ilo.org/media/417251/download. BBS (Bangladesh Bureau of Statistics). 2024. Report on Berge, Lars Ivar Oppedal, Kjetil Bjorvatn, and Bertil Socio-Economic and Demographic Survey 2023. Tungodden. 2015. “Human and Financial Capital for June. Dhaka, Bangladesh: BBS, Statistics and Microenterprise Development: Evidence from a Field Informatics Division, Ministry of Planning. http:// and Lab Experiment.” Management Science 61 (4): nsds.bbs.gov.bd/storage/files/1/Publications/ 707–22. https://doi.org/10.1287/mnsc.2014.1933. others/SEDS_2023_Report.pdf. Berlinski, Samuel, and Sebastián Galiani. 2007. “The BBS (Bangladesh Bureau of Statistics). 2025a. “Key Effect of a Large Expansion of Pre-Primary School Findings of Violence against Women Survey, Facilities on Preschool Attendance and Maternal Bangladesh, 2024.” February 27, 2025, BBS, Statistics Employment.” Labour Economics 14 (3): 665–80. and Informatics Division, Ministry of Planning, Dhaka, https://doi.org/10.1016/j.labeco.2007.01.003. Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 138 Berlinski, Samuel, Sebastián Galiani, and Patrick J. Vandewalle. 2025. “Childcare, Labor Supply, and McEwan. 2011. “Preschool and Maternal Labor Market Business Development: Experimental Evidence Outcomes: Evidence from a Regression Discontinuity from Uganda.” American Economic Journal: Applied Design.” Economic Development and Cultural Economics 17 (2): 75–101. Change 59 (2): 313–44. https://doi.org/https://doi. org/10.1086/657124. Blattman, Christopher, Nathan Vincent Fiala, and Sebastián Martínez. 2014. “Generating Skilled Self- Bernard, Tanguay, Stefan Dercon, Kate Orkin, Giulio Employment in Developing Countries: Experimental Schinaia, and Alemayehu Seyoum Taffesse. 2023. Evidence from Uganda.” Quarterly Journal of “The Future in Mind: Aspirations and Long-Term Economics 129 (2): 697–752. https://doi.org/10.1093/ Outcomes in Rural Ethiopia.” CSAE Working Paper qje/qjt057. 2023-10 (October 16), Department of Economics, Centre for the Study of African Economies, Oxford Blattman, Christopher, Alexandra C. Hartman, and Robert University, Oxford, UK. A. Blair. 2014. “How to Promote Order and Property Rights under Weak Rule of Law? An Experiment in Bernhardt, Arielle, Erica M. Field, Rohini P. Pande, Natalia Changing Dispute Resolution Behavior through Rigol, Simone Schaner, and Charity M. Troyer Moore. Community Education.” American Political Science 2018. “Male Social Status and Women’s Work.” AEA Review 108 (1): 100–20. https://doi.org/10.1017/ Papers and Proceedings 108 (May): 363–67. https:/ / S0003055413000543. doi.org/10.1257/pandp.20181086. Blinder, Alan S. 1973. “Wage Discrimination: Reduced Bergquist, Lauren Falcao, Craig McIntosh, and Meredith Form and Structural Estimates.” Journal of Human Startz. 2024. “Search Costs, Intermediation, and Resources 8 (4): 436–55. Trade: Experimental Evidence from Ugandan Agricultural Markets.” NBER Working Paper 33221 Borker, Garija. 2021. “Safety First: Perceived Risk of Street (December), National Bureau of Economic Research, Harassment and Educational Choices of Women.” Cambridge, MA. Policy Research Working Paper 9731, World Bank, Washington, DC. Berthelon, Matias, Diana I. Kruger, and Melanie Alejandra Oyarzún. 2015. “The Effects of Longer School Days on Borker, Garija. 2022a. “Constraints to Women’s Use of Mothers’ Labor Force Participation.” IZA Discussion Public Transport in Developing Countries, Part I: Paper DP 9212 (July), Institute of Labor Economics, High Costs, Limited Access, and Lack of Comfort.” Bonn, Germany. Global Indicators Brief 9 (August 3), World Bank, Washington, DC. https://doi.org/10.1596/1813-9450- Bhalotra, Sonia Radhika, Rachel Brulé, and Sanchari 9731. Roy. 2020. “Women’s Inheritance Rights Reform and the Preference for Sons in India.” Journal of Borker, Garija. 2022b. “Constraints to Women’s Use of Development Economics 146 (September), 102275. Public Transport in Developing Countries, Part II: https://doi.org/10.1016/j.jdeveco.2018.08.001. Safety.” Global Indicators Brief 10 (August 3), World Bank, Washington, DC. http:/ /documents.worldbank. Bharati, Tushar, Yiwei Qian, and Jeonghwan Yun. 2020. org/curated/en/099559508012223475. “Fueling the Engines of Liberation with Cleaner Cooking Fuel.” Discussion Paper 20-03, Department Bose, Nayana, and Shreyasee Das. 2021. “Women’s of Economics, University of Western Australia, Perth, Inheritance Rights and Fertility Decisions: Evidence Australia. from India.” DETU Working Paper 21-01 (February), Department of Economics, Temple University, BIGD (BRAC Institute of Governance and Development). Philadelphia. 2024. “Women’s Economic Empowerment and Digital Connectivity: The White Paper.” Version 1.0 (February Bossuroy, Thomas, Markus P. Goldstein, Dean S. Karlan, 20), BIGD, BRAC University, Dhaka, Bangladesh. Harounan Kazianga, William Parienté, Patrick Premand, Catherine Cole Thomas, Christopher Binzel, Christine, Erica M. Field, Rohini P. Pande, and Robert Udry, Julia Vaillant, and Kelsey Ann Wright. Louise Paul-Delvaux. 2025. “Gendered Impacts 2021. “Pathways Out of Extreme Poverty: Tackling of Group-Based Formal Lending: Evidence from Psychosocial and Capital Constraints with a Multi- Indian Villages.” Working paper, Friedrich-Alexander- Faceted Social Protection Program in Niger.” Universität Erlangen-Nürnberg, Erlangen, Germany. Policy Research Working Paper 9562, World Bank, Washington, DC. Björkegren, Daniel, Joshua Evan Blumenstock, Omowunmi Folajimi-Senjobi, Jacqueline Mauro, and Suraj R. Boudreau, Laura E. 2024. “Multinational Enforcement of Nair. 2022. “Instant Loans Can Lift Subjective Well- Labor Law: Experimental Evidence on Strengthening Being: A Randomized Evaluation of Digital Credit Occupational Safety and Health Committees.” in Nigeria.” arXiv:2202.13540 [econ.GN], February Econometrica 92 (4): 1269–1308. https://doi. 28, Cornell University, Ithaca, NY. http://arxiv.org/ org/10.3982/ecta19408. abs/2202.13540. Boudreau, Laura E., Sylvain Chassang, Ada Gonzalez- Bjorvatn, Kjetil, Denise Ferris, Selim Gulesci, Arne Torres, and Rachel M. Heath. 2023. “Monitoring Nasgowitz, Vincent Emile Andre Somville, and Lore Harassment in Organizations.” NBER Working Paper 31011 (March), National Bureau of Economic References 139 Research, Cambridge, MA. https://doi.org/10.3386/ Calderón, Gabriela. 2014. “The Effects of Child Care w31011. Provision in Mexico.” Banco de México Working Paper 2014-07 (April), Banco de México, Mexico City. Breza, Emily, Martin Kanz, and Leora F. Klapper. 2020. https://doi.org/10.36095/banxico/di.2014.07. “Learning to Navigate a New Financial Technology: Evidence from Payroll Accounts.” NBER Working Calderone, Margherita, Nathan Fiala, Lemayon Lemilia Paper 28249 (December), National Bureau of Melyoki, Annekathrin Schoofs, and Rachel Steinacher. Economic Research, Cambridge, MA. 2022. “Making Intense Skills Training Work at Scale: Evidence on Business and Labor Market Outcomes in Brooks, Wyatt, Kevin Donovan, and Terrence R. Johnson. Tanzania.” Ruhr Economic Paper 950, Leibniz-Institut 2018. “Mentors or Teachers? Microenterprise Training für Wirtschaftsforschung, Essen, Germany. in Kenya.” American Economic Journal: Applied Economics 10 (4): 196–221. https://doi.org/10.1257/ Callen, Michael J., Suresh de Mel, Craig McIntosh, and app.20170042. Christopher M. Woodruff. 2019. “What Are the Headwaters of Formal Savings? Experimental Brune, Lasse, Xavier Giné, Jessica Goldberg, and Dean Evidence from Sri Lanka.” Review of Economic Yang. 2016. “Facilitating Savings for Agriculture: Studies 86 (6): 2491–2529. https://doi.org/10.1093/ Field Experimental Evidence from Malawi.” Economic restud/rdz020. Development and Cultural Change 64 (2): 187–220. Caria, Antonio Stefano, Bruno Jacques Jean Philippe Bryan, Gharad T., Shyamal K. Chowdhury, and Ahmed Crépon, Hala ElBehairy, Noha Fadlalmawla, Caroline Mushfiq Mobarak. 2014. “Underinvestment in Gould Krafft, Abdelrahman Mohamed Nagy, Lili a Profitable Technology: The Case of Seasonal Mottaghi, Nahla Ahmed Mohamed Salaheldin Zeitoun, Migration in Bangladesh.” Econometrica 82 (5): and Souraya Mahmoud Moustafa El Assiouty. 2023. 1671–1748. https://doi.org/10.3982/ECTA10489. “Child Care Subsidies, Employment Services, and Women’s Labor Market Outcomes in Egypt: First Buchmann, Nina Caroline, Erica M. Field, Rachel Midline Results.” G2LM|LIC Working Paper 73 (June), Glennerster, Shahana Nazneen, and Xiao Yu Wang. Gender, Growth, and Labour Markets in Low-Income 2023. “A Signal to End Child Marriage: Theory and Countries Programme, Foreign, Commonwealth, Experimental Evidence from Bangladesh.” American and Development Office, London; Institute of Labor Economic Review 113 (10): 2645–88. Economics, Bonn, Germany. Buchmann, Nina Caroline, Carl Meyer, and Colin D. Caria, Antonio Stefano, Bruno Jacques Jean Philippe Sullivan. 2024. “Paternalistic Discrimination.” IGC Crépon, Caroline Gould Krafft, and Abdelrahman Working Paper BGD-22021 (June), International Mohamed Nagy. 2025. “The Barriers to Female Growth Center, London School of Economics and Employment: Experimental Evidence from Egypt.” Political Science, London. G2LM|LIC Working Paper 92 (February), Gender, Growth, and Labour Markets in Low-Income Bulte, Erwin, and Robert Lensink. 2019. “Women’s Countries Programme, Foreign, Commonwealth, Empowerment and Domestic Abuse: Experimental and Development Office, London; Institute of Labor Evidence from Vietnam.” European Economic Economics, Bonn, Germany. Review 115 (June): 172–91. https://doi.org/10.1016/j. euroecorev.2019.03.003. Carranza, Eliana, and David J. McKenzie. 2024. “Job Training and Job Search Assistance Policies Burchardi, Konrad B, Selim Gulesci, Benedetta Lerva, and in Developing Countries.” Journal of Economic Munshi Sulaiman. 2019. “Moral Hazard: Experimental Perspectives 38 (1): 221–44. https://doi.org/10.1257/ Evidence from Tenancy Contracts.” Quarterly jep.38.1.221. Journal of Economics 134 (1): 281–347. https:/ /doi. org/10.1093/qje/qjy023. Chakravarty, Shubha, Mattias K. A. Lundberg, Plamen Nikolov, and Juliane Zenker. 2019. “Vocational Burjorjee, Deena M., Mayada El-Zoghbi, and Lis Meyers. Training Programs and Youth Labor Market Outcomes: 2017. “Social Norms Change for Women’s Financial Evidence from Nepal.” Journal of Development Inclusion.” With Brain Heilman. CGAP Brief (July), Economics 136 (January): 71–110. https://doi. Consultative Group to Assist the Poor, World Bank, org/10.1016/j.jdeveco.2018.09.002. Washington, DC. https://www.cgap.org/research/ publication/social-norms-change-for-womens- Chakravorty, Bhaskar, Apurav Yash Bhatiya, Clément financial-inclusion. Imbert, Maximilian Lohnert, Poonam Panda, and Roland Rathelot. 2023. “Impact of the COVID-19 Bursztyn, Leonardo, Alessandra L. González, and David Crisis on India’s Rural Youth: Evidence from a Panel Yanagizawa-Drott. 2020. “Misperceived Social Survey and an Experiment.” World Development Norms: Women Working Outside the Home in Saudi 168 (August), 106242. https://doi.org/10.1016/j. Arabia.” American Economic Review 110 (10): 2997– worlddev.2023.106242. 3029. https://doi.org/10.1257/aer.20180975. Chaturvedi, Sugat, Kanika Mahajan, and Zahra Siddique. Bussolo, Maurizio, Jonah Rexer, and Margaret Triyana. 2021. “Words Matter: Gender, Jobs, and Applicant 2024. “Education, Social Norms, and the Marriage Behavior.” IZA Discussion Paper DP 14497 (June), Penalty: Evidence from South Asia.” Policy Research Institute of Labor Economics, Bonn, Germany. Working Paper 10946, World Bank, Washington, DC. Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 140 Chaturvedi, Sugat, Kanika Mahajan, and Zahra Siddique. Demography 56 (4): 1247–72. https://doi.org/10.1007/ 2025. “Gendered Language in Job Ads and Applicant s13524-019-00793-3. Behavior: Evidence from India.” Labour Economics 96 (October), 102726. Coly, Caroline, and Margaux Suteau. 2025. “The Economic Effects of Sexual Harassment in the Workplace.” LSE Cheema, Ali, Asim I. Khwaja, Muhammad Farooq Naseer, Public Policy Review 3 (4): 1–11. https://ppr.lse.ac.uk/ and Jacob N. Shapiro. 2019. “Skills for Market (SFM articles/118/files/67d405f1023e1.pdf. 2013–14)–Market Linkage (ML 2015–16): Final Impact Evaluation Report.” Punjab Skills Development Fund, Contreras, Ivette, Lelys Ileana Dinarte-Diaz, Amparo Lahore, Pakistan. Palacios-López, Valentina Costa, and Steffanny Romero. 2024. “Closing the Gaps The Role of Cheema, Ali, Asim I. Khwaja, Muhammad Farooq Screening Questions and Self-Reporting in Naseer, and Jacob N. Shapiro. 2022. “Glass Walls: Measuring Women’s and Youths’ Employment and Experimental Evidence on Access Constraints Work.” Policy Research Working Paper 10773, World Faced by Women.” G²LM|LIC Working Paper 65 (May), Bank, Washington, DC. Gender, Growth, and Labour Markets in Low-Income Countries Programme, Foreign, Commonwealth, Corradini, Viola, Lorenzo Lagos, and Garima Sharma. 2025. and Development Office, London; Institute of Labor “Collective Bargaining for Women: How Unions Can Economics, Bonn, Germany. Create Female-Friendly Jobs.” Quarterly Journal of Economics 140 (3): 2053–2105. https://doi. Chen, Yutong, Kerem Coșar, Devaki Ghose, Shirish org/10.1093/qje/qjaf024. Mahendru, and Sheetal Sekhri. 2024. “Gender- Specific Transportation Costs and Female Time Corraya, Sumon. 2025. “Garment Industry in Crisis Use: Evidence from India’s Pink Slip Program.” with Half a Million Jobs Lost in 2024.” News (blog), Policy Research Working Paper 10701, World Bank, January 21, 2025. https://www.asianews.it/news- Washington, DC. en/Garment-industry-in-crisis-with-half-a-million- jobs-lost-in-2024-62329.html. Chowdhury, Afra Rahman, Ana Carolina Areias, Saori Imaizumi, Shinsaku Nomura, and Futoshi Yamauchi. Crépon, Bruno Jacques Jean Philippe, Mohamed El 2018. “Reflections of Employers’ Gender Preferences Komi, and Adam Osman. 2024. “Is It Who You Are or in Job Ads in India: An Analysis of Online Job Portal What You Get? Comparing the Impacts of Loans and Data.” Policy Research Working Paper 8379, World Grants for Microenterprise Development.” American Bank, Washington, DC. Economic Journal: Applied Economics 16 (1): 286– 313. https://doi.org/10.1257/app.20210683. Chowdhury, A. Mushtaque Raza, and Abbas Bhuiya. 2004. “The Wider Impacts of BRAC Poverty Alleviation Croke, Kevin, Markus P. Goldstein, and Alaka Holla. 2017. Programme in Bangladesh.” Journal of International “The Role of Skills and Gender Norms in Sector Development 16 (3): 369–86. https://doi.org/10.1002/ Switches: Experimental Evidence from a Job Training jid.1083. Program in Nigeria.” Policy Research Working Paper 8141, World Bank, Washington, DC. Christensen, Peter, and Adam Osman. 2021. “The Demand for Mobility: Evidence from an Experiment with Uber Cullen, Claire, Sarthak Joshi, Joseph Vecci, and Julia Riders.” IZA Discussion Paper DP 14179 (March), Talbot-Jones. 2024. “Female Empowerment and Institute of Labor Economics, Bonn, Germany. Male Backlash: Experimental Evidence from India.” IZA Discussion Paper DP 17450 (November), Institute Christia, Fotini, Horacio Larreguy, Elizabeth Parker- of Labor Economics, Bonn, Germany. Magyar, and Manuel Quintero. 2023. “Empowering Women Facing Gender-Based Violence amid Daher, Chaza Abou, Erica M. Field, Kendal M. Swanson, COVID-19 through Media Campaigns.” Nature Human and Kate H. Vyborny. 2023. “Drivers of Change: How Behaviour 7 (10): 1740–52. https://doi.org/10.1038/ Intra-Household Preferences Shape Employment s41562-023-01665-y. Responses to Gender Reform.” NBER Working Paper 31715 (September), National Bureau of Economic Chun, Natalie, and Makiko Watanabe. 2012. “Can Skill Research, Cambridge, MA. Diversification Improve Welfare in Rural Areas? Evidence from Bhutan.” Journal of Development Da Mata, Daniel, Rodrigo Oliveira, and Diana L. G. Silva. Effectiveness 4 (2): 214–34. https://doi.org/10.108 2025. “Who Benefits from Job Training Programs? 0/19439342.2012.674965. Evidence from a High-Dosage Program in Brazil.” Journal of Development Economics 175 (June), Clark, Julia Michal, Muhammad Haseeb, Umm E. Amen 103476. https://www.sciencedirect.com/science/ Jalal, Bilal Murtaza Siddiqi, and Kate H. Vyborny. article/pii/S0304387825000276. 2022. “Using Biometrics to Deliver Cash Payments to Women: Early Results from an Impact Evaluation in Dang, Hai-Anh H., Masako Hiraga, and Cuong Viet Nguyen. Pakistan.” ID4D Evidence Note (March), Identification 2019. “Childcare and Maternal Employment: Evidence for Development, World Bank, Washington, DC. from Vietnam.” World Development 159 (November), 106022. Clark, Shelley, Caroline W Kabiru, Sonia Laszlo, and Stella Muthuri. 2019. “The Impact of Childcare on Poor Das, Narayan. 2021. “Training the Disadvantaged Youth Urban Women’s Economic Empowerment in Africa.” and Labor Market Outcomes: Evidence from References 141 Bangladesh.” Journal of Development Economics de Mel, Suresh, David J. McKenzie, and Christopher M. 149 (March), 102585. https://doi.org/10.1016/j. Woodruff. 2014. “Business Training and Female jdeveco.2020.102585. Enterprise Start-Up, Growth, and Dynamics: Experimental Evidence from Sri Lanka.” Journal of Datar, Gayatri, Ximena Vanessa Del Carpio, and Vivian Development Economics 106 (January): 199–210. Hoffman. 2009. “Can a Market-Assisted Land https://doi.org/10.1016/j.jdeveco.2013.09.005. Redistribution Program Improve the Lives of the Poor? Evidence from Malawi.” Policy Research Dillon, Andrew, Valerie Mueller, and Sheu Salau. 2011. Working Paper 5093, World Bank, Washington, DC. “Migratory Responses to Agricultural Risk in Northern Nigeria.” American Journal of Agricultural Economics Dean, Joshua T., and Seema Jayachandran. 2020. 93 (4): 1048–61. https://doi.org/10.1093/ajae/aar033. “Attending Kindergarten Improves Cognitive Development in India, but All Kindergartens Are Dinkelman, Taryn. 2011. “The Effects of Rural Electrification Not Equal.” Working paper (July 23), Northwestern on Employment: New Evidence from South Africa.” University, Evanston, IL; Booth School of Business, American Economic Review 101 (7): 3078–3108. University of Chicago, Chicago. https://doi.org/10.1257/aer.101.7.3078. Deininger, Klaus W., Aparajita Goyal, and Hari K. Nagarajan. Discenza, Antonio Rinaldo, Isis Gaddis, Amparo Palacios- 2010. “Inheritance Law Reform and Women’s Access López, and Kieran Walsh. 2021. “Measuring Women to Capital Evidence from India’s Hindu Succession and Men’s Work: Main Findings from a Joint ILO Act.” Policy Research Working Paper 5338, World and World Bank Study in Sri Lanka.” World Bank, Bank, Washington, DC. Washington, DC. Deininger, Klaus W., Aparajita Goyal, and Hari K. Dizon, Felipe, Erick Gong, and Kelly Jones. 2020. “The Nagarajan. 2013. “Women’s Inheritance Rights and Effect of Promoting Savings on Informal Risk Sharing Intergenerational Transmission of Resources in Experimental Evidence from Vulnerable Women in India.” Journal of Human Resources 48 (1): 114–41. Kenya.” Journal of Human Resources 55 (3): 963–98. https:/ /www.jstor.org/stable/23799109. https://jhr.uwpress.org/content/55/3/963. Deininger, Klaus W., Songqing Jin, Hari K. Nagarajan, Donald, Aletheia Amalia, Markus P. Goldstein, and Léa and Fang Xia. 2019. “Inheritance Law Reform, Marie Rouanet. 2022. “Two Heads Are Better Than Empowerment, and Human Capital Accumulation: One: Agricultural Production and Investment in Côte Second-Generation Effects from India.” Journal of d’Ivoire.” Policy Research Working Paper 10047, World Development Studies 55 (12): 2549–71. https://doi. Bank, Washington, DC. org/10.1080/00220388.2018.1520218. Donald, Aletheia Amalia, and Florian Grosset. 2024. de la Cruz Toledo, Elia. 2015. “Universal Preschool and “Labor Supply Complementarities in Urban Côte Mothers’ Employment.” Working paper, Columbia d‘Ivoire: How the Ability to Commute Together Can Population Research Center, Columbia University, Boost Women’s Employment.” G2LM|LIC Policy Brief New York. https://www.sole-jole.org/assets/ 66 (January), Gender, Growth, and Labour Markets docs/15157.pdf. in Low-Income Countries Programme, Foreign, Commonwealth, and Development Office, London; Delavallade, Clara, Mélanie Gittard, and Julia Vaillant. Institute of Labor Economics, Bonn, Germany. 2025. “Women and Climate Adaptation in Rural Sub- Saharan Africa: Constraints and Research Priorities.” Donald, Aletheia Amalia, Sara Lowes, and Julia Vaillant. Policy Research Working Paper 11095, World Bank, 2024. “Experimental Evidence on Rural Childcare Washington, DC. Provision.” Paper presented at National Bureau of Economic Research, Summer Institute 2024: Gender de Mel, Suresh, Craig McIntosh, Ketki Sheth, and in the Economy, Boston, July 23–24, 2024. https:/ / Christopher M. Woodruff. 2022. “Can Mobile-Linked www.nber.org/conferences/si-2024-gender- Bank Accounts Bolster Savings? Evidence from a economy. Randomized Controlled Trial in Sri Lanka.” Review of Economics and Statistics 104 (2): 306–20. https:/ / Drexler, Alejandro, Greg Fischer, and Antoinette Schoar. doi.org/10.1162/rest_a_00956. 2014. "Keeping It Simple: Financial Literacy and Rules of Thumb." American Economic Journal: Applied de Mel, Suresh, David J. McKenzie, and Christopher Economics 6 (2): 1–31. https://doi.org/10.1257/ M. Woodruff. 2008. “Returns to Capital in app.6.2.1. Microenterprises: Evidence from a Field Experiment.” Quarterly Journal of Economics 123 (4): 1329–72. Du, Fenglian, and Xiao-yuan Dong. 2013. “Women’s https://doi.org/10.1162/qjec.2008.123.4.1329. Employment and Child Care Choices in Urban China during the Economic Transition.” Economic de Mel, Suresh, David J. McKenzie, and Christopher Development and Cultural Change 62 (1): 131–55. M. Woodruff. 2009. “Are Women More Credit https://doi.org/10.1086/671714. Constrained? Experimental Evidence on Gender and Microenterprise Returns.” American Economic Dupas, Pascaline, Dean S. Karlan, Jonathan Robinson, Journal: Applied Economics 1 (3): 1–32. https://doi. and Diego Ubfal. 2018. “Banking the Unbanked? org/10.1257/app.1.3.1. Evidence from Three Countries.” American Economic Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 142 Journal: Applied Economics 10 (2): 257–97. https:// Fiala, Nathan Vincent. 2018b. “Returns to Microcredit, doi.org/10.1257/app.20160597. Cash Grants and Training for Male and Female Entrepreneurs in Uganda.” World Development Dupas, Pascaline, and Jonathan Robinson. 2013. 105 (May): 189–200. https://doi.org/10.1016/j. “Why Don’t the Poor Save More? Evidence from worlddev.2017.12.027. Health Savings Experiments.” American Economic Review 103 (4): 1138–71. https://doi.org/10.1257/ Fiala, Nathan Vincent, Ana Garcia-Hernandez, Kritika aer.103.4.1138. Narula, and Nishith Prakash. 2022. “Wheels of Change: Transforming Girls’ Lives with Bicycles.” IZA Egami, Hiroyuki, and Tomoya Matsumoto. 2020. “Mobile Discussion Paper DP 15076 (February), Institute of Money Use and Healthcare Utilization: Evidence from Labor Economics, Bonn, Germany. Rural Uganda.” Sustainability 12 (9), 3741. https://doi. org/10.3390/su12093741. Field, Erica M., and Attila Ambrus. 2008. “Early Marriage, Age of Menarche, and Female Schooling Attainment Elsayed, Ahmed, Kevin Hempel, and Adam Osman. in Bangladesh.” Journal of Political Economy 116 (5): 2018. “Overcoming Youth Unemployment in Egypt: 881–930. Randomized Evaluations Showcase the Promise of Active Labor Market Programs.” Working paper Field, Erica M., Seema Jayachandran, Rohini P. Pande, (October), Impact Evaluation of the Emergency and Natalia Rigol. 2016. “Friendship at Work: Can Employment Investment Project, World Bank, Peer Effects Catalyze Female Entrepreneurship?” Washington, DC. American Economic Journal: Economic Policy 8 (2): 125–53. http://dx.doi.org/10.1257/pol.20140215. Elsayed, Ahmed, Soiliou Daw Namoro, and Rania Roushdy. 2021. “Empowering Women in Conservative Settings: Field, Erica M., Rohini P. Pande, John Papp, and Natalia Evidence from an Intervention in Rural Egypt.” Review Rigol. 2013. “Does the Classic Microfinance Model of Economics of the Household 20 (4): 1287–1322. Discourage Entrepreneurship among the Poor? https://doi.org/10.1007/s11150-021-09576-5. Experimental Evidence from India.” American Economic Review 103 (6): 2196–2226. https://doi. Emran, M. Shahe, Virginia Robano, and Stephen C. Smith. org/10.1257/aer.103.6.2196. 2014. “Assessing the Frontiers of Ultrapoverty Reduction: Evidence from Challenging the Frontiers Field, Erica M., Rohini P. Pande, Natalia Rigol, Simone G. of Poverty Reduction/Targeting the Ultra-poor, Schaner, and Charity M. Troyer Moore. 2021. “On Her an Innovative Program in Bangladesh.” Economic Own Account: How Strengthening Women's Financial Development and Cultural Change 62 (2): 339–80. Control Affects Labor Supply and Gender Norms.” https://doi.org/10.1086/674110. American Economic Review 111 (7): 2342–75. https:/ / doi.org/10.1257/aer.20200705. Erten, Bilge, and Pinar Keskin. 2024. “Trade-Offs? The Impact of WTO Accession on Intimate Partner Fletcher, Erin Kathleen, Rohini P. Pande, and Charity Violence in Cambodia.” Review of Economics and M. Troyer Moore. 2018. “Women and Work in India: Statistics 106 (2): 322–33. https://doi.org/10.1162/ Descriptive Evidence and a Review of Potential rest_a_01140. Policies.” HKS Faculty Research Working Paper RWP18-004 (January), Harvard Kennedy School, Fafchamps, Marcel, David J. McKenzie, Simon R. Quinn, Harvard University, Cambridge, MA. and Christopher M. Woodruff. 2014. “Microenterprise Growth and the Flypaper Effect: Evidence from a Gaddis, Isis, Gbemisola Oseni, Amparo Palacios-López, Randomized Experiment in Ghana.” Journal of and Janneke Pieters. 2023. “Who Is Employed? Development Economics 106 (January): 211–26. Evidence from Sub-Saharan Africa on Redefining https://doi.org/10.1016/j.jdeveco.2013.09.010. Employment.” Journal of African Economies 32 (2): 151–74. https://doi.org/10.1093/jae/ejac021. Farole, Thomas, Yoonyoung Cho, Laurent Bossavie, and Reyes Aterido. 2017. Jobs Diagnostic: Bangladesh, Garcia, João, Rafael Latham-Proença, and Marcela Mello. Main Report. Jobs Series 9. Washington, DC: World 2022. “Free Childcare and the Motherhood Penalty: Bank. https://documents1.worldbank.org/curated/ Evidence from São Paulo.” Paper presented at the en/834041507299981599/pdf/Main-report.pdf. North East Universities Development Consortium annual conference, Economic Growth Center, Yale Ferrah, Samih, Jules Gazeaud, Nausheen Khan, Eric University, New Haven, CT, November 5–6, 2022. Mvukiyehe, Varada Shrotri, Olivier Christian Sterck, https://jmbvgarcia.github.io/assets/pdf/Childcare_ and Samir ben Zineb. 2021. “Enhancing Female SP.pdf. Entrepreneurship through Cash Grants: Experimental Evidence from Rural Tunisia.” MNAGIL Research Garcia-Hernandez, Ana, Nishith Prakash, and Janina Policy Report (July), Middle East and North Africa Isabel Steinert. 2025. “The Empowerment Paradox? Gender Innovation Lab, World Bank, Washington, DC. The Long-Run Impact of a Cycling Program for Girls in Zambia.” IZA Discussion Paper DP 17903 (May), Fiala, Nathan Vincent. 2018a. “Business Is Tough, Institute of Labor Economics, Bonn, Germany. but Family Is Worse: Household Bargaining and Investment Decisions in Uganda.” Working Paper Garlick, Robert, Erica M. Field, and Kate H. Vyborny. 2025. 2017–05, Department of Economics, University of “Women’s Mobility and Labor Supply: Experimental Connecticut, Storrs, CT. Evidence from Pakistan.” IZA Discussion Paper DP References 143 17883 (April), Institute of Labor Economics, Bonn, Experiment in Jordan.” Review of Economics and Germany. Statistics 98 (3): 488–502. https://doi.org/10.1162/ REST_a_00584. Garz, Seth, Rachel Heath, Elijah Kipchumba, and Munshi Sulaiman. 2020. "Evidence of Digital Financial Services GSMA. 2020. “Reaching 50 Million Women with Mobile: A Impacting Women’s Economic Empowerment." BRAC Practical Guide.” March, GSMA, London. https://www. Institute of Governance & Development, White Paper. gsma.com/solutions-and-impact/connectivity- http://dx.doi.org/10.13140/RG.2.2.10430.82242. for-good/mobile-for-development/wp-content/ uploads/2020/04/Reaching-50-Million-Women- Gazeaud, Jules, Nausheen Khan, Eric Mvukiyehe, and with-Mobile-A-Practical-Guide.pdf. Olivier Christian Sterck. 2023. “With or without Him? Experimental Evidence on Cash Grants and GSMA. 2024. “The Mobile Gender Gap Report 2024.” Gender-Sensitive Trainings in Tunisia.” Journal of May, GSMA, London. https://www.gsma. Development Economics 165 (October), 103169. com/r/wp-content/uploads/2024/05/The- https://doi.org/10.1016/j.jdeveco.2023.103169. Mobile-Gender-Gap-Report-2024.pdf?utm_ source=website&utm_medium=button&utm_ Gentile, Elisabetta, Nikita Kohli, Nivedhitha Subramanian, campaign=gender-gap-2024. Zunia Tirmazee, and Kate H. Vyborny. 2023. “Barriers to Entry: Decomposing the Gender Gap in Job Search Guarnieri, Eleonora, and Helmut Rainer. 2021. in Urban Pakistan.” ADB Economics Working Paper “Colonialism and Female Empowerment: A Two- 707 (December), Asian Development Bank, Manila. Sided Legacy.” Journal of Development Economics https:/ /doi.org/10.22617/WPS230551-2. 151 (June), 102666. https://doi.org/10.1016/j. jdeveco.2021.102666. Ghisletta, Andrea, Johanna Kemper, and Jonathan Stöterau. 2021. “The Impact of Vocational Training Gutierrez, Italo A., Oswaldo Molina, and Hugo Rolando Interventions on Youth Labor Market Outcomes: A Ñopo. 2018. “Stand against Bullying: An Experimental Meta-Analysis.” Working Paper 20 (July), Chair of School Intervention.” IZA Discussion Paper DP 11623 Education Systems, Department of Management, (June), Institute of Labor Economics, Bonn, Germany. Technology, and Economics, Federal Institute of Technology, Zurich. Halim, Daniel Zefanya, Hillary C. Johnson, and Elizaveta Perova. 2022. “Preschool Availability and Women’s Gibbs, Michael, Friederike Mengel, and Christoph Siemroth. Employment: Evidence from Indonesia.” Economic 2023. “Work from Home and Productivity: Evidence Development and Cultural Change 71 (1): 39–61. from Personnel and Analytics Data on Information https://doi.org/10.1086/714439. Technology Professionals.” Journal of Political Economy Microeconomics 1 (1): 7–41. https://doi. Hallward-Driemeier, Mary, and Ousman Gajigo. 2015. org/10.1086/721803. “Strengthening Economic Rights and Women’s Occupational Choice: The Impact of Reforming Giné, Xavier, and Ghazala Mansuri. 2021. “Money or Ethiopia’s Family Law.” World Development Management? A Field Experiment on Constraints 70 (June): 260–73. https://doi.org/10.1016/j. to Entrepreneurship in Rural Pakistan.” Economic worlddev.2015.01.008. Development and Cultural Change 70 (1): 41–86. https://doi.org/10.1086/707502. Harari, Mariaflavia. 2019. “Women’s Inheritance Rights and Bargaining Power: Evidence from Kenya.” Economic Giulivi, Nicoletta, Aurélie P. Harou, Shriniwas Gautam, Development and Cultural Change 68 (1): 189–238. and Davíd Guereña. 2023. “Getting the Message https:/ /doi.org/10.1086/700630. Out: Information and Communication Technologies and Agricultural Extension.” American Journal of Hardy, Morgan L., Gisella Kagy, and Nusrat Abedin Jimi. Agricultural Economics 105 (3), 1011–45. https://doi. 2024. “Mind the Data Gaps: An Examination of org/10.1111/ajae.12348. Women-Owned Enterprise Representation.” Working paper (June), Social Science Division, New York Goldin, Claudia Dale. 1995. “The U-shaped Female Labor University, Abu Dhabi, United Arab Emirates. Force Function in Economic Development and Economic History.” In Investment in Women’s Human Hardy, Morgan L., Isaac Mulangu Mbiti, Jamie Lee Capital, edited by T. Paul Schultz, 61–90. Chicago: McCasland, and Isabelle Salcher. 2019. “The University of Chicago Press. Apprenticeship-to-Work Transition: Experimental Evidence from Ghana.” Policy Research Working Goldstein, Markus P., Kenneth Houngbedji, Florence Paper 8851, World Bank, Washington, DC. Kondylis, Michael O’Sullivan, and Harris Selod. 2015. “Formalizing Rural Land Rights in West Africa: Heath, Rachel. 2014. “Women’s Access to Labor Market Evidence from a Randomized Impact Evaluation in Opportunities, Control of Household Resources, Benin.” Policy Research Working Paper 7435, World and Domestic Violence: Evidence from Bangladesh.” Bank, Washington, DC. World Development 57 (May): 32–46. Groh, Matthew, Nandini Krishnan, David J. McKenzie, Heath, Rachel, and Ahmed Mushfiq Mobarak. 2015. and Tara Vishwanath. 2016. “Do Wage Subsidies “Manufacturing Growth and the Lives of Provide a Stepping-Stone to Employment for Recent Bangladeshi Women.” Journal of Development College Graduates? Evidence from a Randomized Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 144 Economics 115 (July): 1–15. https://doi.org/10.1016/j. Labour Underutilization, Adopted by the Nineteenth jdeveco.2015.01.006. International Conference of Labour Statisticians (October 2013).” International Labour Organization, Heath, Rachel, and Emma Alice Louise Riley. 2024. “Digital Geneva. https:/ /www.ilo.org/global/statistics-and- Financial Services and Women’s Empowerment: databases/meetings-and-events/international- Experimental Evidence from Tanzania.” CEGA Working c o n f e r e n c e - o f - l a b o u r - s t a t i s t i c i a n s / 1 9/ Paper 243 (September 25), Center for Effective WCMS_230304/lang--en/index.htm. Global Action, University of California, Berkeley, CA. ILO (International Labour Organization). 2018. “Resolution Heath, Rachel, and Xu Tan. 2020. “Intrahousehold I: Resolution Concerning Statistics on Work Bargaining, Female Autonomy, and Labor Supply: Relationships.” Report ICLS/20/2018/Resolution Theory and Evidence from India.” Journal of the I, 20th International Conference of Labour European Economic Association 18 (4): 1928–68. Statisticians, Geneva, 10–19 October 2018. Geneva: https://doi.org/10.1093/jeea/jvz026. ILO. Hirshleifer, Sarojini, David J. McKenzie, Rita Kullberg ILO (International Labour Organization). 2022. “Youth Not Almeida, and Cristobal Ridao-Cano. 2016. “The in Employment, Education, or Training in Asia and Impact of Vocational Training for the Unemployed: the Pacific: Trends and Policy Considerations.” ILO, Experimental Evidence from Türkiye.” Economic Geneva. https://www.ilo.org/publications/youth- Journal 126 (597): 2115–46. https://doi.org/10.1111/ not-employment-education-or-training-asia-and- ecoj.12211. pacific-trends-and. Ho, Lisa Y., Suhani Jalota, and Anahita Karandikar. 2024. ILO (International Labour Organization). 2023. “Resolution “Bringing Work Home: Flexible Work Arrangements II: Resolution to Amend the 19th ICLS Resolution as Gateway Jobs for Women in West Bengal.” Paper Concerning Statistics of Work, Employment and presented at the National Bureau of Economic Labour Underutilization.” Report ICLS/21/2023/ Research Conference, Summer Institute 2024: RES. II, 21st International Conference of Labour Development Economics, Conference, Cambridge, Statisticians (Geneva, 11–20 October 2023). Geneva: MA, July 22–23, 2024. https://www.povertyactionlab. ILO. org/sites/default/files/research-paper/lisaho_jmp_ hjk_bwh.pdf. ILO (International Labour Organization). 2025. “ILO Modelled Estimates: Methodological Overview.” Hojman, Andrés, and Florencia López Bóo. 2019. “Cost- May, Department of Statistics, ILO, Geneva. https:// Effective Public Daycare in a Low-Income Economy webapps.ilo.org/ilostat-files/Documents/TEM.pdf. Benefits Children and Mothers.” IDB Working Paper IDB-WP-1036 (August), Inter-American Development Jack, William, Michael R. Kremer, Joost de Laat, and Bank, Washington, DC. Tavneet Suri. 2023. “Credit Access, Selection, and Incentives in a Market for Asset-Collateralized Loans: Honorati, Maddalena. 2015. “The Impact of Private Sector Evidence From Kenya.” Review of Economic Studies Internship and Training on Urban Youth in Kenya.” 90 (6), 3153–85. https://doi.org/10.1093/restud/ Policy Research Working Paper 7404, World Bank, rdad026. Washington, DC. Jacoby, Hanan G., and Ghazala Mansuri. 2015. “Crossing Ibarrarán, Pablo, Laura Ripani, Bibiana Taboada, Juan Boundaries: How Social Hierarchy Impedes Economic Miguel Villa, and Brigida Garcia. 2014. “Life Skills, Mobility.” Journal of Economic Behavior and Employability, and Training for Disadvantaged Youth: Organization 117 (September): 135–54. https://doi. Evidence from a Randomized Evaluation Design.” IZA org/10.1016/j.jebo.2015.05.003. Journal of Labor and Development 3 (June 10), 10. http://www.izajold.com/content/3/1/10. Jalota, Suhani, and Lisa Y. Ho. 2024. “What Works for Her? How Work-from-Home Jobs Affect Female IFC (International Finance Corporation). 2019. “Tackling Labor Force Participation in Urban India.” Job Market Childcare: The Business Benefits and Challenges Paper, January 20, Stanford University, Stanford, CA. of Employer-Supported Childcare in Bangladesh.” https://doi.org/10.2139/ssrn.4739387. November, Dhaka, Bangladesh. https:/ /www.ifc.org/ en/insights-reports/2010/tackling-childcare-the- Janzen, Sarah A., Nicholas P. Magnan, Sudhindra Sharma, business-benefits-and-challenges-of-employer- and William M. Thompson. 2018. “Short-Term Impacts supported-childcare-in-bangladesh. of a Pay-It-Forward Livestock Transfer and Training Program in Nepal.” AEA Papers and Proceedings ILO (International Labour Organization). 1983. “Resolution I: 108 (May): 422–25. https://doi.org/10.1257/ Resolution Concerning Statistics of the Economically pandp.20181120. Active Population, Employment, Unemployment, and Underemployment." In The Thirteenth International Jayachandran, Seema. 2015. “The Roots of Gender Conference of Labour Statisticians, Geneva, 18–29 Inequality in Developing Countries.” Annual Review October 1982, I/1–I/9. Geneva: ILO. https:/ /www.ilo. of Economics 7 (1): 63–88. https://doi.org/10.1146/ org/public/libdoc/ilo/1982/82B09_651_engl.pdf. annurev-economics-080614-115404. ILO (International Labour Organization). 2013. “Resolution Jayachandran, Seema. 2021. “Microentrepreneurship in Concerning Statistics of Work, Employment, and Developing Countries.” In Handbook of Labor, Human References 145 Resources, and Population Economics, edited by Khandelwal, Vatsal, and Juni Singh. 2024. “Can Network Klaus F. Zimmermann, 1–31. Cham, Switzerland: Ties Help Facilitate Female Entrepreneurship?” Paper Springer International Publishing. presented at the Association française d’économie du développement’s International Conference on Jensen, Robert T. 2012. “Do Labor Market Opportunities Development Economics, ICDE 2024, Aix-Marseille Affect Young Women’s Work and Family Decisions? School of Economics, Aix-en-Provence, France, July Experimental Evidence from India.” Quarterly 2–3, 2024. Journal of Economics 127 (2): 753–92. https://doi. org/10.1093/qje/qjs002. Khatun, Fahmida, Syed Yusuf Saadat, Kashfia Ashraf, and Mohammad Abu Tayeb Taki. 2022. Skills Gap and Jones, Maria, and Florence Kondylis. 2018. “Does Feedback Youth Employment in Bangladesh: An Exploratory Matter? Evidence from Agricultural Services.” Analysis. Dhaka: Centre for Policy Dialogue. Journal of Development Economics 131 (March): 28–41. https:/ /doi.org/10.1016/j.jdeveco.2017.10.013. Khatun, Fahmida, and Syed Yusuf Saadat. 2020. Youth Employment in Bangladesh: Creating Opportunities— Jones, Sam, and Kunal Sen. 2022. “Labour Market Reaping Dividends. Singapore: Palgrave Macmillan. Effects of Digital Matching Platforms: Experimental https://doi.org/10.1007/978-981-15-1750-1. Evidence from Sub-Saharan Africa.” WIDER Working Paper 2022/69 (June), United Nations University– Kikulwe, Enoch M., Elisabeth Fischer, and Matin Qaim. World Institute for Development Economics 2014. “Mobile Money, Smallholder Farmers, and Research, Helsinki. https://doi.org/10.35188/UNU- Household Welfare in Kenya.” Edited by Ajay Mahal. WIDER/2022/200-3. PLoS One 9 (10), e109804. https://doi.org/10.1371/ journal.pone.0109804. Kabeer, Naila. 1999. “Resources, Agency, Achievements: Reflections on the Measurement of Women’s Kilburn, M. Rebecca, and Ashlesha Datar. 2002. “The Empowerment.” Development and Change 30 (3): Availability of Child Care Centers in China and Its 435–64. https:/ /doi.org/10.1111/1467-7660.00125. Impact on Child Care and Maternal Work Decisions.” Labor and Population Program Working Paper 02-12; Kabeer, Naila. 2024. Renegotiating Patriarchy: Gender, RAND Working Paper DRU-2924-NIH (December), Agency, and the Bangladesh Paradox. London: LSE RAND Corporation, Santa Monica, CA. Press. https://doi.org/10.31389/lsepress.rpg. Kilic, Talip, and Heather G. Moylan. 2016. “Methodological Karimli, Leyla, Els Lecoutere, Christine R. Wells, and Experiment on Measuring Asset Ownership from a Leyla Ismayilova. 2021. “More Assets, More Gender Perspective (MEXA).” Technical Report (April), Decision-Making Power? Mediation Model in a World Bank, Washington, DC. Cluster-Randomized Controlled Trial Evaluating the Effect of the Graduation Program on Women’s Kilic, Talip, Heather G. Moylan, and Gayatri B. Koolwal. Empowerment in Burkina Faso.” World Development 2020. “Getting the (Gender-Disaggregated) Lay of 137 (January), 105159. https://doi.org/10.1016/j. the Land: Impact of Survey Respondent Selection worlddev.2020.105159. on Measuring Land Ownership.” Policy Research Working Paper 9151, World Bank, Washington, DC. Karlan, Dean S., Margaret McConnell, Sendhil Mullainathan, and Jonathan Zinman. 2016. “Getting to the Top of Kipchumba, Elijah Kipkech, and Munshi Sulaiman. 2025. Mind: How Reminders Increase Saving.” Management “Digital Finance and Intra-Household Decision- Science 62 (12): 3393–3672. https://doi.org/10.1287/ Making: Evidence from Mobile Money Use in Kenya.” mnsc.2015.2296. Paper presented at the BRAC Institute of Governance and Development and Innovations for Poverty Karlan, Dean S., and Jonathan Zinman. 2018. “Price and Action International Conference, Access to Agency: Control Elasticities of Demand for Savings.” Journal Empowering Women through Digital Inclusion, Accra, of Development Economics 130 (January): 145–59. Ghana, July 2–3. 2025. https://bigd.bracu.ac.bd/ https://doi.org/10.1016/j.jdeveco.2017.10.004. wp-content/uploads/2025/06/Access-to-Agency- Empowering-Women-through-Digital-Inclusion.pdf. Karmaliani, Rozina, Judith McFarlane, Hussain Maqbool Ahmed Khuwaja, Yasmeen Somani, Shireen Shehzad, Kitagawa, Evelyn M. 1955. “Components of a Difference Tazeen Saeed Ali, Nargis Asad, Esnat D. Chirwa, and between Two Rates.” Journal of the American Rachel Jewkes. 2020. “Right To Play’s Intervention Statistical Association 50 (272): 1168–94. to Reduce Peer Violence among Children in Public Schools in Pakistan: A Cluster-Randomized Kjelsrud, Anders, Shabana Mitra, and Karl Moene. 2024. Controlled Trial.” Global Health Action 13 (1), 1836604. “Wheels of Power: Can Free Bicycles for Schoolgirls https://doi.org/10.1080/16549716.2020.1836604. Free Their Mothers?” Economic Development and Cultural Change 72 (2): 423–57. https://doi. Karra, Mahesh, Mindy Hernandez, Catherine Brennan, and org/10.1086/722053. Margaret McConnell. 2022. “Supply-Side Innovations to Increase Equitable Access to Digital Financial Klasen, Stephan, Tu Thi Ngoc Le, Janneke Pieters, and Services: Experimental Evidence from Mozambique.” Manuel Santos Silva. 2021. “What Drives Female IED Working Paper 374 (April 13), Institute for Labour Force Participation? Comparable Micro- Economic Development, Department of Economics, Boston University, Boston. Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 146 Level Evidence from Eight Developing and Emerging in Mobile Banking.” Journal of Economic Behavior Economies.” Journal of Development Studies 57 (3): and Organization 193 (January): 276–93. https://doi. 417–42. org/10.1016/j.jebo.2021.10.005. Klonner, Stefan, and Patrick James Nolen. 2010. Lee, Jean N., Jonathan Morduch, Saravana Ravindran, Abu “Cell Phones and Rural Labor Markets: Evidence S. Shonchoy, and Hassan Zaman. 2021. “Poverty and from South Africa.” Proceedings of the German Migration in the Digital Age: Experimental Evidence Development Economics Conference 2010 on Mobile Banking in Bangladesh.” American (Hannover) 56. Göttingen, Germany: Verein für Economic Journal: Applied Economics 13 (1): 38–71. Socialpolitik, Ausschuss für Entwicklungsländer. https://doi.org/10.1257/app.20190067. https://econpapers.repec.org/paper/zbwgdec10/56. htm. Lei, Lei, Sonalde Desai, and Reeve Vanneman. 2019. “The Impact of Transportation Infrastructure on Women’s Kondylis, Florence, Arianna Legovini, Kate H. Vyborny, Employment in India.” Feminist Economics 25 (4): Astrid Maria Theresia Zwager, and Luiza Cardoso De 94–125. https://doi.org/10.1080/13545701.2019.16 Andrade. 2025. “Demand for ‘Safe Space’: Avoiding 55162. Harassment and Complying with Norms.” Journal of Development Economics 174 (May), 103392. https:/ / Lowe, Matt, and Madeline McKelway. 2021. “Coupling doi.org/10.1016/j.jdeveco.2024.103392. Labor Supply Decisions: An Experiment in India.” CESifo Working Paper 9446 (November), Munich Krafft, Caroline Gould, and Moundir Lassassi. 2024. Society for the Promotion of Economic Research, “Public Pre-Primary and Maternal Employment in Center for Economic Studies, Ludwig Maximilian Algeria: Evidence from a Natural Experiment.” World University and Ifo Institute for Economic Research, Bank Economic Review 38 (1): 74–94. https://doi. Munich. https://doi.org/10.2139/ssrn.3979351. org/10.1093/wber/lhad024. Macchiavello, Rocco, Andreas Menzel, Atonu Rabbani, Kugler, Adriana D., Maurice David Kugler, Juan E. Saavedra, and Christopher M. Woodruff. 2020. “Challenges and Luis Omar Herrera-Prada. 2022. “Long-Term of Change: An Experiment Promoting Women to Educational Consequences of Vocational Training Managerial Roles in the Bangladeshi Garment in Colombia: Impacts on Young Trainees and Their Sector.” NBER Working Paper 27606 (July), National Relatives.” Journal of Human Resources 57 (1): 178– Bureau of Economic Research, Cambridge, MA. 216. https:/ /doi.org/10.3368/jhr.57.1.0518-9528R2. Mahmud, Minhaj, Italo A. Gutierrez, Krishna B. Kumar, and Kuhn, Peter, and Kailing Shen. 2023. “What Happens When Shanthi Nataraj. 2021. “What Aspects of Formality Do Employers Can No Longer Discriminate in Job Ads?” Workers Value? Evidence from a Choice Experiment American Economic Review 113 (4): 1013–48. https:// in Bangladesh.” World Bank Economic Review 35 (2): doi.org/10.1257/aer.20211127. 303–27. https:/ /doi.org/10.1093/wber/lhz046. Kusumawardhani, Niken, Rezanti Pramana, Nurmala Mahmud, Simeen, and Sayema Haque Bidisha. 2018. Selly Saputri, and Daniel Suryadarma. 2023. "Female Labor Market Participation in Bangladesh: “Heterogeneous Impact of Internet Availability on Structural Changes and Determinants of Labor Female Labor Market Outcomes in an Emerging Supply." In Structural Change and Dynamics of Economy: Evidence from Indonesia.” World Labor Markets in Bangladesh, edited by Selim Development 164 (April), 106182. https://doi. Raihan, 51–63. Singapore: Springer, 2018. https:// org/10.1016/j.worlddev.2022.106182. doi.org/10.1007/978-981-13-2071-2_4. La Ferrara, Eliana, and Annamaria Milazzo. 2017. Maitra, Pushkar, and Subha Mani. 2017. “Learning and “Customary Norms, Inheritance, and Human Capital: Earning: Evidence from a Randomized Evaluation in Evidence from a Reform of the Matrilineal System India.” Labour Economics 45 (April): 116–30. https:// in Ghana.” American Economic Journal: Applied doi.org/10.1016/j.labeco.2016.11.007. Economics 9 (4): 166–85. https://doi.org/10.1257/ app.20150342. Majoka, Zaineb, and Sadiq Shams. 2025. “Even in a Megapolis, It Still Takes a Village: Making Quality Lafortune, Jeanne, Julio Riutort, and José Tessada. 2018. Childcare Accessible in Dhaka.” Working paper, World “Role Models or Individual Consulting: The Impact Bank, Washington, DC. of Personalizing Micro-Entrepreneurship Training.” American Economic Journal: Applied Economics 10 Mariscal, Judith, Gloria Mayne, Urvashi Aneja, and (4): 222–45. https://doi.org/10.1257/app.20170077. Alina Sorgner. 2019. “Bridging the Gender Digital Gap.” Economics 13 (2019-9): 1–12. https://www. Lecoutere, Els, David J. Spielman, and Björn F. H. van degruyterbrill.com/document/doi/10.5018/ Campenhout. 2023. “Empowering Women through economics-ejournal.ja.2019-9/html. Targeting Information or Role Models: Evidence from an Experiment in Agricultural Extension in Uganda.” Martínez, Claudia, and Marcela Perticará. 2017. “Childcare World Development 167 (July), 106240. https:/ /doi. Effects on Maternal Employment: Evidence from org/10.1016/j.worlddev.2023.106240. Chile.” Journal of Development Economics 126 (May): 127–37. https://doi.org/10.1016/j.jdeveco.2017.01.001. Lee, Jean N., Jonathan Morduch, Saravana Ravindran, and Abu S. Shonchoy. 2022. “Narrowing the Gender Gap References 147 McIntosh, Craig T., and Andrew Zeitlin. 2022. “Skills Mueller, Valerie, Agnes Quisumbing, Hak Lim Lee, and and Liquidity Barriers to Youth Employment: Klaus Droppelmann. 2014. “Resettlement for Food Medium-Term Evidence from a Cash Benchmarking Security’s Sake: Insights from a Malawi Land Reform Experiment in Rwanda.” arXiv:2209.08574v1 [econ. Project.” Land Economics 2 (90): 222–36. GN], September 18, 2022, Cornell University, Ithaca, NY. http://arxiv.org/abs/2209.08574. Müller, Miriam, and Liliana Do Couto Sousa. 2020. “‘She Helps Me All the Time’: Underestimating Women’s McKelway, Madeline. 2025. “Women’s Self-Efficacy Economic Engagement in Rural Honduras.” and Economic Outcomes: Experimental Evidence Policy Research Working Paper 9217, World Bank, from India.” Working paper, January 3, Dartmouth Washington, DC. College, Hanover, NH. https://drive.google.com/file/ d/1Fi8Y8A0o4j-NwqOmROowAV0kLcuCC1Ai/view. Muralidharan, Karthik, and Nishith Prakash. 2017. “Cycling to School: Increasing Secondary School McKenzie, David J., Aakash Mohpal, and Dean Yang. 2022. Enrollment for Girls in India.” American Economic “Aspirations and Financial Decisions: Experimental Journal: Applied Economics 9 (3): 321–50. https:// Evidence from the Philippines.” Journal of doi.org/10.1257/app.20160004. Development Economics 156 (May), 102846. https:/ / doi.org/10.1016/j.jdeveco.2022.102846. Naaraayanan, Sankararama Lakshmi. 2019. “Women’s Inheritance Rights and Entrepreneurship Gender Medrano, Patricia. 2009. “Public Day Care and Female Gap.” Working Paper (November 12), Department Labor Force Participation: Evidence from Chile.” of Finance, Hong Kong University of Science and Working Paper SDT 306 (December), Department of Technology, Hong Kong SAR, China. Economics, University of Chile, Santiago, Chile. Nakata, Shiro, Tashmina Rahman, and Md. Mokhlesur Mendola, Mariapia, and Franklin Simtowe. 2015. “The Rahman. 2017. “Employability of Post-Secondary Welfare Impact of Land Redistribution: Evidence TVET in Bangladesh: Tracking Survey of Graduates of from a Quasi-Experimental Initiative in Malawi.” Polytechnics.” Report AUS0000063 (January 2, 2018), World Development 72 (August): 53–69. https://doi. World Bank, Dhaka, Bangladesh. https://documents. org/10.1016/j.worlddev.2015.02.010. worldbank.org/en/publication/documents-reports/ documentdetail/718501528118077327. Miller, Conrad, Jennifer Peck, and Mehmet Seflek. 2022. “Missing Women, Integration Costs, and Big Push Nakata, Shiro, Uttam Sharma, Tashmina Rahman, Md. Policies in the Saudi Labor Market.” American Mokhlesur Rahman, and Mustahsin-Ul Aziz. 2021. Economic Journal: Applied Economics 14 (2): 51–77. “Effects of Recognition of Prior Learning on Job https://doi.org/10.1257/APP.20200220. Market Outcomes: Impact Evaluation in Bangladesh.” Policy Research Working Paper 9644, World Bank, Misha, Farzana A., Wameq A. Raza, Jinnat Ara, and Ellen Washington, DC. van de Poel. 2019. “How Far Does a Big Push Really Push? Long-Term Effects of an Asset Transfer Nandi, Arijit, Parul Agarwal, Anoushaka Chandrashekar, Program on Employment Trajectories.” Economic and Sam Harper. 2020. “Access to Affordable Daycare Development and Cultural Change 68 (1): 41–62. and Women’s Economic Opportunities: Evidence https://doi.org/10.1086/700556. from a Cluster-Randomised Intervention in India.” Journal of Development Effectiveness 12 (3): 219–39. Mishra, Khushbu, and Abdoul G. Sam. 2016. “Does Women’s https:/ /doi.org/10.1080/19439342.2020.1773898. Land Ownership Promote Their Empowerment? Empirical Evidence from Nepal.” World Development Oaxaca, Ronald L. 1973. “Male-Female Wage Differentials 78 (February): 360–71. https://doi.org/10.1016/j. in Urban Labor Markets.” International Economic worlddev.2015.10.003. Review 14 (3): 693–709. Molina, Teresa, and Mari Tanaka. 2023. “Globalization and OECD (Organisation for Economic Co-operation and Female Empowerment: Evidence from Myanmar.” Development) and UNCTAD (United Nations Trade Economic Development and Cultural Change 71 (2): and Development). 2023. Production Transformation 519–65. https://doi.org/10.1086/715748. Policy Review of Bangladesh: Investing in the Future of a Trading Nation. OECD Development Moyeen, Sabah, Tara Lonnberg, Marufa Akter, Samera Pathways. Geneva: UNCTAD; Paris: OECD. https:// Chowdhury, Sabina Parvin, Jayati Sethi, Erisha doi.org/10.1787/8b925b06-en. Singh Suwal, Mohsiu Rashedin Tazrin, and Sanan Isaba Zaman. 2022. Bangladesh: Country Gender Orkin, Kate, Robert Garlick, Mahreen Mahmud, Richard Assessment 2021. Washington, DC: World Bank. Sedlmayr, Johannes Haushofer, and Stefan Dercon. http://hdl.handle.net/10986/38316. 2023. “Aspiring to a Better Future: Can a Simple Psychological Intervention Reduce Poverty?” NBER Mueller, Valerie, Lucy Billings, Tewodaj Mogues, Amber Working Paper 31735 (September), National Bureau Peterman, and Ayala Wineman. 2015. “Filling of Economic Research, Cambridge, MA. the Legal Void? Experimental Evidence from a Community-Based Legal Aid Program for Gender- Padilla-Romo, María, and Francisco Cabrera-Hernández. Equal Land Rights in Tanzania.” IFPRI Discussion 2018. “The Effect of Children’s Time in School on Paper 01434, International Food Policy Research Mothers’ Labor Supply: Evidence from Mexico’s Institute, Washington, DC. Full-Time Schools Program.” Working Paper 2018- Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 148 04 (September), Department of Economics, Haslam Development 149 (January), 105698. https://doi. College of Business, University of Tennessee, org/10.1016/j.worlddev.2021.105698. Knoxville, TN. Rosenblum, Daniel. 2015. “Unintended Consequences of Pomeranz, Dina, and Felipe Kast. 2024. “Savings Accounts Women’s Inheritance Rights on Female Mortality in to Borrow Less: Experimental Evidence from Chile.” India.” Economic Development and Cultural Change Journal of Human Resources 59 (1): 70–108. https:/ / 63 (2): 223–48. https://doi.org/10.1086/679059. doi.org/10.3368/jhr.0619-10264R3. Rosero, José, and Hessel Oosterbeek. 2011. “Trade-Offs Prina, Silvia. 2015. “Banking the Poor via Savings Accounts: between Different Early Childhood Interventions: Evidence from a Field Experiment.” Journal of Evidence from Ecuador.” Tinbergen Institute Development Economics 115 (July): 16–31. https:/ / Discussion Paper TI 2011–102/3, Faculty of doi.org/10.1016/j.jdeveco.2015.01.004. Economics and Business, University of Amsterdam, Amsterdam; Tinbergen Institute, Amsterdam. Rahman, Atiya, Anindita Bhattacharjee, and Narayan Das. 2021. “A Good Mix against Ultra-Poverty? Roy, Sanchari. 2008. “Female Empowerment through Evidence from a Randomized Controlled Trial (RCT) Inheritance Rights: Evidence from India.” Paper in Bangladesh.” Review of Development Economics presented at the Second Riccardo Faini Doctoral 25 (4): 2052–83. https://doi.org/10.1111/rode.12809. Conference on Development Economics, University of Milan, Milan, September 7–9, 2008. https://www. Rahman, Mahbubur, Mohammad Omar Farooq, and dagliano.unimi.it/media/roy.pdf. Mohammad Selim. 2021. “Mitigating Educated Youth Unemployment in Bangladesh.” Journal of Roy, Sanchari. 2015. “Empowering Women? Inheritance Developing Areas 55 (1): 185–200. https://ssrn.com/ Rights, Female Education, and Dowry Payments abstract=3780472. in India.” Journal of Development Economics 114 (May): 233–51. https://doi.org/10.1016/j. Raihan, Selim, and Sayema Haque Bidisha. 2018. "Female jdeveco.2014.12.010. employment stagnation in Bangladesh: A research paper on economic dialogue on inclusive growth Ryu, Hanbyul. 2020. “The Effect of Compulsory Preschool in Bangladesh." London: Overseas Development Education on Maternal Labour Supply.” Journal of Institute (ODI), San Francisco: The Asia Foundation, Development Studies 56 (7): 1384–1407. https://doi. and London: UKAID. org/10.1080/00220388.2019.1677890. Razzaque, Mohammad Abdur, Deen Islam, and Jillur Said, Farah, Mahreen Mahmud, Giovanna d’Adda, and Rahman. 2024. “Can Bangladesh Absorb LDC Azam Chaudhry. 2022. “Home-Based Enterprises: Graduation-Induced Tariff Hikes? Evidence Using Experimental Evidence on Female Preferences Product-Specific Price Elasticities of Demand and from Pakistan.” Economic Development and Markups for Apparel Exports to Europe.” IGC Policy Cultural Change 71 (1): 185–221. https://doi. Brief BGD-23078 (November), International Growth org/10.1086/714739. Center, London School of Economic and Political Science, London. Sandefur, Justin, and Bilal Siddiqi. 2013. “Delivering Justice to the Poor: Theory and Experimental Riley, Emma Alice Louise, Abu S. Shonchoy, and Robert Evidence from Liberia.” Paper presented at the Darko Osei. 2025. “Incentives and Endorsement for Center for Effective Global Action’s Workshop on Technology Adoption: Evidence from Mobile Banking African Political Economy, World Bank, Washington, in Ghana.” Journal of Development Economics 176 DC, May 20, 2013. https://cega.berkeley.edu/event/ (September), 103511. a-workshop-on-african-political-economy/. Roessler, Philip, Peter Carroll, Flora Myamba, Cornel Sanfelice, Viviane. 2024. “Public Childcare and Jahari, Blandina Kilama, and Daniel Nielson. 2021. Maternal Employment in Brazil.” Working paper “The Economic Impact of Mobile Phone Ownership: (September), Economics Department, University of Results from a Randomized Controlled Trial in Rochester, Rochester, NY. https://doi.org/10.2139/ Tanzania.” CSAE Working Paper WPS/2021–05 ssrn.4969594. (April), Centre for the Study of African Economies, Department of Economics, University of Oxford, Sapkal, Rahul Suresh. 2017. “From Mother to Daughter: Oxford, UK. Does Equal Inheritance Property Laws Reform Improve Female Labor Supply and Educational Rojas Valdes, Ruben Irvin, Bruce Wydick, and Travis J. Attainments in India?” Asian Journal of Law and Lybbert. 2022. “Can Hope Elevate Microfinance? Economics 8 (1): 1–36. https://doi.org/10.1515/ajle- Evidence from Oaxaca, Mexico.” Oxford Economic 2015-0028. Papers 74 (1): 236–64. https://doi.org/10.1093/oep/ gpaa039. SARTSD (South Asia Region Team for Statistical Development). 2024. DATALIBWEB: GLDRAW. Raw Rosas, Nina, Maria Cecilia Acevedo, and Samantha Data. Countries: Bangladesh 2005; 2010; 2013. World Zaldivar. 2022. “Starting Points Matter: Cash Bank, Washington DC. As of 22/08/2024. Plus Training Effects on Youth Entrepreneurship, Skills, and Resilience during an Epidemic.” World References 149 SARTSD (South Asia Region Team for Statistical Smarrelli, Gabriela. 2023. “Improving School Management Development). 2024. DATALIBWEB: SARRAW. Raw of Violence: Evidence from a Nationwide Policy in Data. Countries: Bangladesh 2015; 2016; 2022. World Peru.” CGD Working Paper 667 (November), Center Bank, Washington DC. As of 27/06/2024. for Global Development, Washington, DC. SARTSD (South Asia Region Team for Statistical Sodergren, Marie-Claire, and Mabelin Villarreal-Fuentes. Development). 2024. DATALIBWEB: SARRAW. Raw 2022. “Quick Guide to Understanding the Impact of Data. Countries: Bangladesh 2022. World Bank, the New Statistical Standards on ILOSTAT Databases.” Washington DC. As of 03/06/2024. September, Department of Statistics, International Labour Organization, Geneva. https:/ /www.ilo.org/ SARTSD (South Asia Region Team for Statistical publications/quick-guide-understanding-impact- Development). 2024. DATALIBWEB: SARMD. Ex-post new-statistical-standards-ilostat. Harmonization. Countries: Bangladesh 2022. World Bank, Washington DC. As of 03/06/2024. Solotaroff, Jennifer L., Aphichoke Kotikula, Anna Tara Lonnberg, Snigdha Ali, Rohini P. Pande, and Ferdous Seki, Mai, and Eiji Yamada. 2020. “Heterogeneous Effects Jahan. 2019. Voices to Choices: Bangladesh’s Journey of Urban Public Transportation on Employment by in Women’s Economic Empowerment. International Gender: Evidence from the Delhi Metro.” JICA-RI Development in Focus Series. Washington, DC: World Working Paper 207 (March): 1–57, JICA Research Bank. Institute, Japan International Cooperation Agency, Tokyo. https:/ /jicari.repo.nii.ac.jp/records/1046. Somville, Vincent, and Lore Vandewalle. 2018. “Saving by Default: Evidence from a Field Experiment in Sey, Araba, and Nancy Hafkin, eds. 2019. Taking Stock: Rural India.” American Economic Journal: Applied Data and Evidence on Gender Digital Equality in Economics 10 (3): 39–66. https://doi.org/10.1257/ Digital Access, Skills and Leadership. March. Macao app.20160547. SAR, China: EQUALS Global Partnership for Gender Equality in the Digital Age, United Nations University Subramanian, Nivedhitha. 2024. “Workplace Attributes Institute, and International Telecommunication and Women’s Labor Supply Decisions: Evidence Union. https://www.equalsintech.org/taking-stock. from a Randomized Experiment.” Working paper, Bates College, Lewiston, ME. https://github.com/ Sharma, Garima. 2023. “Monopsony and Gender.” Working nivedhithasubramanian/PersonalWebsite/blob/ paper (March 22), Department of Economics, main/WorkplaceAttributes_202409.pdf. Massachusetts Institute of Technology, Cambridge, MA. https://economics.mit.edu/sites/default/ Suri, Tavneet, and William Jack. 2016. “The Long-Run files/inline-files/monopsony_gender_gsharma_ Poverty and Gender Impacts of Mobile Money.” march2023.pdf. Science 354 (6317): 1288–92. https://doi.org/10.1126/ science.aah5309. Sharma, Karmini. 2024. “Tackling Sexual Harassment: Short- and Long-Run Experimental Evidence Suri, Tavneet, and Chris Udry, eds. 2024. “Agricultural from India.” CAGE Working Paper 728 (October), Technology in Africa.” VoxDevLit 5 (2): 1–42. Competitive Advantage in the Global Economy, https://voxdev.org/sites/default/files/2024-04/ CAGE Research Center, Department of Economics, Agricultural_Technology_Africa_Issue_2.pdf. University of Warwick, Coventry, UK. https://warwick. ac.uk/fac/soc/economics/research/centres/cage/ Suteau, Margaux. 2020. "Inheritance Rights and Women's manage/publications/wp728.2024.pdf. Empowerment in the Labor and Marriage Markets." THEMA Working Paper 2020-17 (December), CY Cergy Shonchoy, Abu S., Tomoki Fujii, and Selim Raihan. 2018. Paris University, Cergy, France. https://thema.u- “Barriers to Labor Migration for the Rural Poor: cergy.fr/IMG/pdf/2020-17.pdf. Experimental Evidence from a Vocational Training Program in Bangladesh.” Working paper (February), Swaminathan, Hema, Arturo Martinez Jr., Kaushal Joshi, Research Collection, School of Economics, Mildred Addawe, Christian Flora Mae Soco, and Singapore Management University, Singapore. Remedios Baes-Espineda. 2023. “Gender Gaps in https://ink.library.smu.edu.sg/cgi/viewcontent. Ownership of Nonagricultural Enterprises in Georgia, cgi?article=3797&context=soe_research. Mongolia, and the Philippines.” ADB Economics Working Paper 698 (October), Asian Development Siddique, Zahra. 2018. “Violence and Female Labor Bank, Manila. https://www.adb.org/publications/ Supply.” IZA Discussion Paper DP 11874 (October), g e n d e r- g a p s - o w n e r s h i p - n o n a g r i c u l t u r a l - Institute of Labor Economics, Bonn, Germany. enterprises. Singh, Sanjana, Radhika Asrani, and Anupama Talamas Marcos, Miguel Ángel. 2023. “Grandmothers and Ramaswamy. 2016. Study on Mapping the Market the Gender Gap in the Mexican Labor Market.” Journal Potential and Accelerating Finance for Women of Development Economics 162 (May), 103013. Entrepreneurs in Bangladesh. Washington, DC: https://doi.org/10.1016/j.jdeveco.2022.103013. International Finance Corporation. http://documents. worldbank.org/curated/en/354971490957587126. Tripney, Janice, Jorge Hombrados, Mark Newman, Kimberly Hovish, Chris Brown, Katarzyna Steinka- Fry, and Eric Wilkey. 2013. “Technical and Vocational Education and Training (TVET) Interventions to Women’s Economic Empowerment in Bangladesh | An Evidence-Guided Toolkit for More Inclusive Policies 150 Improve the Employability and Employment of Journal: Applied Economics 14 (2): 101–25. https:// Young People in Low‐ and Middle‐Income Countries: doi.org/10.1257/APP.20200025. A Systematic Review.” Campbell Systematic Reviews 9 (1): 1–171. https://doi.org/10.4073/csr.2013.9. WHO (World Health Organization). 2021. Violence against Women Prevalence Estimates, 2018: Global, Regional, Ubfal, Diego, Irani Arráiz, Diether W. Beuermann, Michael and National Prevalence Estimates for Intimate Frese, Alessandro Maffioli, and Daniel Verch. 2022. Partner Violence against Women and Global and “The Impact of Soft-Skills Training for Entrepreneurs Regional Prevalence Estimates for Non-Partner in Jamaica.” World Development 152 (April), 105787. Sexual Violence against Women. Geneva: WHO. https://doi.org/10.1016/j.worlddev.2021.105787. World Bank. 2015. “How Does the Short-Term Training Uckat, Hannah. 2023. “Leaning in at Home: Women’s Program Contribute to Skills Development in Promotions and Intra-Household Bargaining in Bangladesh? A Tracer Study of the Short-Term Bangladesh.” Policy Research Working Paper 10370, Training Graduates.” Discussion Paper, Report 82 World Bank, Washington, DC. (September), Education Global Practice, South Asia Region, World Bank, Washington, DC. Uckat, Hannah, and Christopher M. Woodruff. 2020. “Learning How to Choose or Learning How to Lead? World Bank. 2016. Stitches to riches?: Apparel employment, Experiments on Selecting and Training Female trade, and economic development in South Asia. Managers in Bangladesh’s Garment Industry.” Washington, DC: World Bank. http://documents. Working paper, World Bank, Washington, DC. https:// worldbank.org/curated/en/299231467988907350. www.hannahuckat.com/wp-content/uploads/WPT_ Paper.pdf? World Bank. 2020. “Demographic Transition: Lessons from Bangladesh’s Success Story.” Human Capital Project, UN DESA (United Nations Department of Economic and World Bank, Washington, DC. http://hdl.handle. Social Affairs). 2021. International Classification of net/10986/33672. Activities for Time-Use Statistics 2016. Statistical Paper, Series M No. 98, Document ST/ESA/STAT/ World Bank. 2021. “What Do We Know about Interventions SER.M/98. New York: Statistics Division, UN DESA, to Increase Women’s Economic Participation and United Nations. https://unstats.un.org/unsd/ Empowerment in South Asia? Skills and Training gender/timeuse/23012019%20ICATUS.pdf. Interventions.” Systematic Review on Women’s Economic Empowerment (June), South Asia Gender UNDP (United Nations Development Programme). 2023. Innovation Lab, World Bank, Washington, DC. “2023 Gender Social Norms Index; Breaking down Gender Biases: Shifting Social Norms towards Gender World Bank. 2022. Change of Fabric: Bangladesh Country Equality.” Human Development Perspectives Series, Economic Memorandum. Washington, DC: World Bank. Human Development Report Office, UNDP, New York. http://hdl.handle.net/10986/38229. Valera, Harold Glenn A., Takashi Yamano, Ranjitha Puskur, World Bank. 2023. “For Women, by Women, with Women: Prakashan Chellattan Veettil, Ishika Gupta, Phoebe Bridging the Gender Digital Divide.” Final Report, Ricarte, and Rohini Ram Mohan. 2018. “Women’s World Bank, Washington, DC. https://documents. Land Title Ownership and Empowerment: Evidence worldbank.org/en/publication/documents-reports/ from India.” ADB Economics Working Paper 559 documentdetail/099610011082285212/p17272608 (September), Asian Development Bank, Manila. e21410300976405be930707674. https://doi.org/10.22617/WPS189556-2. World Bank. 2024a. “Bangladesh Development Update, Velásquez, Andrea. 2020. “The Economic Burden of Crime: Special Focus: Creating Jobs for a Better Future.” Evidence from Mexico.” Journal of Human Resources October, World Bank, Washington, DC. 55 (4): 1287–1318. https://doi.org/10.3368/ jhr.55.4.0716-8072R2. World Bank. 2024b. “Bangladesh Development Update, Special Focus: Strengthening Domestic Viollaz, Mariana, and Hernán Jorge Winkler. 2022. “Does Resource Mobilization.” April, World Bank, Dhaka, the Internet Reduce Gender Gaps? The Case of Bangladesh. https://thedocs.worldbank.org/en/doc/ Jordan.” Journal of Development Studies 58 (3): f6dd176b982714b6e3b04d7ac672a0ab-0310012024/ 436–53. https://doi.org/10.1080/00220388.2021.1 original/Bangladesh-Development-Update-April- 965127. 2024-w-cover.pdf. Vyborny, Kate H., Robert Garlick, Nivedhitha Subramanian, World Bank. 2024c. “Bangladesh Public Finance Review: and Erica M. Field. 2024. “Why Don’t Jobseekers Fiscal Reforms for an Upper-Middle Income Economy.” Search More? Barriers and Returns to Search on a World Bank, Washington, DC. Job Matching Platform.” IZA Working Paper DP 17520 (December), Institute of Labor Economics, Bonn, World Bank. 2024d. South Asia Development Update: Germany. Women, Jobs, and Growth. October. Washington, DC: World Bank. http://hdl.handle.net/10986/42002. Wheeler, Laurel, Robert Garlick, Eric Johnson, Patrick Shaw, and Marissa Gargano. 2022. “LinkedIn(to) World Bank. 2024e. Women, Business, and the Law 2024. Job Opportunities: Experimental Evidence from Washington, DC: World Bank. http://hdl.handle. Job Readiness Training.” American Economic net/10986/41040. References 151 World Bank. 2024f. “Women, Business and the Law 2024: Bangladesh.” World Bank, Washington, DC. https://wbl. worldbank.org/content/dam/documents/wbl/2024/ pilot/WBL24-2-0-Bangladesh.pdf. World Bank. 2024g. Women’s Economic Empowerment in Pakistan: An Evidence Guided Toolkit for More Inclusive Policies . Washington, DC: South Asia Gender Innovation Lab, World Bank. http://documents. worldbank.org/curated/en/099517505282516353. World Bank. 2025a. “Bangladesh Development Update, Special Focus: Restoring Financial Sector Stability.” April, World Bank, Dhaka, Bangladesh. https://documents1.worldbank.org/curated/ en/099552504232519918/pdf/IDU-dcc82f80-885c- 492c-b4dd-432bb3d7c0f3.pdf. World Bank. 2025b. “Making Jobs Platforms Work for Women: A Guide for Practitioners and Researchers.” South Asia Gender Innovation Lab, World Bank, Washington, DC. World Bank. 2025c. “Navigating the Road to Prosperity in Bangladesh: Poverty and Equity Assessment.” World Bank, Washington, DC. World Bank. 2025d. “Opportunities for Women’s Economic Empowerment in Agro-Based Industries in Bangladesh.” Bangladesh-Bhutan Country Gender Platform, World Bank, Dhaka, Bangladesh. World Bank. 2025e. “Preferences toward Childcare: Evidence from Bangladesh.” South Asia Gender Innovation Lab, World Bank, Washington, DC. World Bank. 2025f. “Understanding the Relationship between Bus Service Providers and Female Passengers: Qualitative Evidence from Dhaka.” South Asia Gender Innovation Lab, World Bank, Washington, DC. World Bank. 2025g. What Works to Advance Women’s Digital Literacy? A Review of Good Practices and Programs. G2Px Initiative. Washington, DC: World Bank. World Bank, UNESCO (United Nations Educational, Scientific, and Cultural Organization), and ILO (International Labour Organization). 2023. Building Better Formal TVET Systems: Principles and Practice in Low- and Middle-Income Countries. Washington, DC: World Bank; Paris: UNESCO; Geneva: ILO. Zahra, Najaf, Amna Javed, and Ana María Muñoz-Boudet. 2022. “What Do We Know about Interventions to Increase Women’s Economic Participation and Empowerment in South Asia? Asset Transfers and Property Law.” Systematic Review on Women’s Economic Empowerment (September), South Asia Gender Innovation Lab, World Bank, Washington, DC. Zahra, Najaf, Amna Javed, and Ana María Muñoz-Boudet. 2023. “What Do We know About Interventions to Increase Women’s Economic Participation and Empowerment in South Asia? Care Polices and Care Services.” Systematic Review on Women’s Economic Empowerment (July), South Asia Gender Innovation Lab, World Bank, Washington, DC. Credits: Dominic Chavez / World Bank