SYSTEMATIC REVIEW ON WOMEN’S ECONOMIC EMPOWERMENT APRIL 2022 WHAT DO WE KNOW ABOUT INTERVENTIONS TO INCREASE WOMEN’S ECONOMIC PARTICIPATION AND EMPOWERMENT IN SOUTH ASIA? SELF-HELP GROUP PROGRAMS Amna JavedŦ, Najaf ZahraŦ, and Ana Maria Munoz BoudetŦ BACKGROUND Existing systematic reviews have evaluated the impact of The World Bank’s South Asia Region Gender Innovation SHGs within a global or multi-regional context. This brief Lab is conducting a systematic review and meta-analysis of contributes to the literature by synthesizing evidence interventions with direct or indirect effects on measures specifically for the South Asia region. South Asia has a of women’s economic empowerment. The review focuses long history of SHGs hosting some of the largest programs on changes in labor market outcomes, incomes and globally, as measured by membership and total savings savings, and other empowerment indicators. The goal is accumulated. The evolution of the group-based model to document what has and has not worked for women in in the region and the plethora of quantitative evidence the region (covering all countries: Afghanistan, Bangladesh, evaluating impacts offer an opportunity to inform research Bhutan, India, Maldives, Nepal, Pakistan, and Sri Lanka), and policy regarding effective methods and the existence of understand the types of interventions implemented, and knowledge gaps. In contrast to much of the earlier literature, identify gaps in knowledge and action. The review organizes the brief also distinguishes between the evaluated impact interventions in six categories: Skills, Assets, Credit, Labor of SHG prevalence directly on participants and indirectly market, Entrepreneurship, and Empowerment. This brief on nonparticipants, highlighting the potential for positive summarizes the main findings from the Self-Help Group spillovers from the interventions. subtheme of the Empowerment category. Self-help groups are broadly defined as groups of individuals from a community, voluntarily convening with a common WHAT IS INCLUDED? purpose (Brody et al. 2015). These groups form under the hypothesis that mutual support and action can lead to The systematic review includes experimental and quasi- individual economic benefits for even the most marginalized experimental evidence for policies and programs, individuals. Group or community-based programs can vary in implemented in any South Asian country, which directly structure, mission, and characteristics, so we follow Gugerty aimed to change women’s economic outcomes or have et al. (2019) and define SHGs as (1) involving member indirectly done so. This brief focuses on studies that organize participation in group governance, (2) relying on internally or evaluate women’s self-help groups (SHGs) as a mechanism generated resources (e.g., savings), (3) having a primary goal for achieving changes in economic outcomes. of positive individual benefits, and (4) requiring regular face- SELF-HELP GROUPS: groups of individuals from a community, voluntarily convening with a common purpose Involve member par�cipa�on Rely on internally Have primary goal of Require regular face in group governance generated resources posi�ve individual benefits to face interac�ons Ŧ World Bank to-face interactions. We do not restrict inclusion to groups researchers’ preferred specification is used in this brief. with only female membership. Eligible studies were those that: The review includes English-language studies published • Evaluated a self-help group program. between January 1990 and April 2020 across white and gray • Employed experimental or quasi-experimental literature (peer reviewed journals, working papers, program evaluation methods. or agency reports, and academic theses, among others) • Reported outcomes for women, either as the direct identified via an extensive search of multiple databases.1 target population or a subpopulation of interest. Intervention inclusion was not limited by time, duration, • Reported required outcomes, including labor market frequency, or method of exposure. Figure 1 summarizes the outcomes (such as self-employment, participation, three-stage identification process. The first stage filtered days worked), income or earnings, and empowerment select papers relevant to the region and programs that were (including, among others, agency, well-being, happiness, specifically for women or included female beneficiaries. The mobility, financial or political empowerment). second stage filtered for intervention type and the third for methodology.2 Two reviewers independently searched and The papers included in this review were drawn from the search extracted data from the list of finalized articles, including process for our larger Empowerment theme. The selection impact effects, design, and intervention components. criteria required the inclusion of those studies that evaluated Additional outcome-specific data, such as units of reporting, empowerment programs with a rigorous methodology and coefficient significance, and standard errors were also included outcomes for women. Of the 163,876,961 papers extracted. If a study reported impact estimates using more identified in the first stage of the search process, about 36 than one specification, all were recorded, but only the percent (59,016,460) remained after filtration using the Figure 1: Search Methodology Identi cation A preliminary list of Backward and forward Resources were rechecked Stage 2: Snowballing databases was searched snowballing was conducted using the World Bank Stage 1: Base Search library lists and the Stage 3: Recheck connected paper’s website Studies were stored in a The process was repeated Final studies were added snowballing repository un�l no new studies were to a thema�c database Screening iden�fied Eligibility decisions were made a�er reading the �tle and abstract of each resource. Addi�onal scoping was done to iden�fy outcomes and methodology for some ar�cles. Each study outcome was assigned to a category, either Eligibility employment, empowerment, or income. Poten�al papers were also checked for their iden�fica�on strategy. Key informa�on about programs and par�cipants was extracted for each study including type of interven�on, sample popula�on, econometric methodology, and impact details. All studies added were given unique IDs based on interven�on type, popula�on, econometric methodology/specifica�on, year, and outcomes. If a study reported impact es�mates using more than one specifica�on, it was coded using different codes for methodology/specifica�on. Key Information 1 The search included the following databases: Econlit, Web of Science, Science Direct, National Bureau of Economic Research (NBER), Google Scholar, World Bank e-Library, UNWider, Abdul Latif Jameel Poverty Action Lab (J-PAL), Institute of Labor Economics (IZA), Center for Global Development (CGD), International Growth Center (IGC), American Economic Association (AEA), AEA Registry, International Initiative for Impact Evaluation (3iE), Research Papers in Economics (RePEc), IDEAS database, and JSTOR. The Evidence Consortium on Women’s Group (ECWG) was also searched for relevant articles on SHGs. 2 Second stage search terms included experiment, political, decision, violence, domestic, bargaining power, agency, independence, income, empowerment, mobility, seclusion, aspiration, contraceptive, contraception, marriage, age of marriage, autonomy, birth spacing, fertility, family planning, norms, attitudes, network, social network, self help groups, self-help group, and social capital. Third stage search terms included comparison group, counterfactual, counter-factual, evaluation, assessment, impact, rct, randomized control trial, impact evaluation, quasi experiment, quasi-experiment, propensity score matching, psm, regression discontinuity design, rdd, and discontinuous design. 2 | SYSTEMATIC REVIEW ON WOMEN’S ECONOMIC EMPOWERMENT second stage search terms. Further refinement in the third stage reduced the pool to 2,067,751 studies. After removing INFORMATION LIVELIHOOD, repetitions and refining by title and abstract, 37 studies FINANCIAL AND ON VOCATIONAL GOVERNMENT on SHGs were compiled in a preliminary list. These papers TRAINING PROGRAMS were read for methodology and relevance, then snowballed SOCIAL ACTIVITIES backward and forward, resulting in a final list of 33 studies PEER meeting our predefined inclusion criteria. The final sample INTERACTION only includes studies from India and Bangladesh.3 GROUP EDUCATIONAL DISCUSSIONS PROGRAMS ON SOCIAL AND CLASSES SELF-HELP GROUPS ISSUES Within South Asia, SHGs first emerged in India in the mid- 1980s. In 1992, the government of India’s National Bank In practice, primary objectives and the vision for change for Agriculture and Rural Development developed the Self- may vary across programs. For example, Mahila Samakhya Help-Group Bank Linkage Program, which quickly expanded aims to empower Indian women through education. The throughout the country. By 2019, the program had a Self-Employed Women’s Association (SEWA) designs membership of 125 million households, linked to 10 million programs to achieve women’s economic independence, SHGs across the nation, with approximately 2,333 billion INR while the objective of the Targeted Rural Initiatives for in outstanding savings.4 Poverty Termination and Infrastructure (TRIPTI) in India Self-help groups offer both direct and indirect pathways to is to reduce poverty through diversified livelihood. This economic empowerment. For example, measures to improve variation in primary objectives results in the use of a women’s access to low-interest loans, assistance with variety of activities within the SHGs. Most programs will government job program applications, group monitoring to have a core financial component, consisting of regular meet saving goals, and opportunities for vocational training savings and opportunities to borrow from internal funds. offer a direct path toward employment (self or other) and As a history of savings and repayments is established, entrepreneurship. At the same time, educational programs, the SHG will link members to formal financial institutions group discussions on current and social issues, peer support, that provide either individual or group-based funds at low and classes designed to change attitudes regarding women’s interest rates. Beyond the core financial component, the agency can lead to the creation of social capital and improve groups will offer a mix of livelihood, financial and vocational bargaining power, in turn providing women tools to access training, education programs and classes, social activities, economic opportunities. information on government programs, peer interaction, and group discussions on social issues. Often, program components are demand-driven and are introduced gradually as membership and stakeholder trust direct pathways is built. In some cases, group activities begin with financial functions such as savings, credit, and training. As men and SELF-HELP ECONOMIC in-laws in the household are convinced of the benefits of GROUPS EMPOWERMENT participation for their female relatives, empowerment modules are slowly incorporated (Kandpal, Baylis, and indirect pathways Arends-Kuenning 2013). In contrast, Mahila Samakhya tailors its activities to each village and often begins with literacy or education camps (Kandpal, Baylis, and Arends- Kuenning 2013). Programs also tend to be facilitated by locals who are familiar with village conditions, languages, and norms. Within the context of the National Rural 3 There are multiple studies on women or support groups in other South Asian countries. For example, a series of papers study the effects of self-help groups on child health and Livelihoods Mission (NRLM) in India, Joshi and Rao (2018) maternal nutrition outcomes. Multiple articles in Pakistan and Bangladesh study the impact show that groups with local facilitators are not only less of microfinance in the context of community-based groups. However, these studies are not included in this review if they do not meet our inclusion restrictions for outcome relevance expensive to manage but are more likely to engage in local or methodology. politics or collective action for public services, compared to 4 “Status of Microfinance in India 2018-2019,” NABARD. Available at https://www.nabard. org/auth/writereaddata/tender/1207192354SMFI%202018-19.pdf SHGs with external facilitators. APRIL 2022 | 3 PROGRAM DESIGN REGULAR SAVINGS Saving requirements vary across programs The studies included in this note rigorously evaluate 13 and can determine the viability of programs in South Asia, including large and well-known membership for the poorest households. ones such as JEEViKA, SEWA, and Mahila Samakhya. Six Under the NRLP, members save articles examine the general impact of membership in any approximately 40 INR every month (Kochar SHG, while others investigate the impact of membership in et al. 2020). In Bihar, JEEViKA encourages the National Rural Livelihoods Program (NRLP).5 The main members to save 8 to 40 INR per month distinct features of the programs are:6 (Datta 2015; Hoffmann et al. 2021). The comparable monthly range for SEWA PROGRAMS TARGET RURAL AREAS members is 100 to 400 INR (Desai and Apart from the Safe Cities Initiative in India, Joshi 2014; Desai and Olofsgård 2019). which was administered in slums, the projects operate in rural settings, targeting regions GROUP AND MEETING STRUCTURE with low levels of development, as measured Meetings run 30 to 120 minutes per session by income, literacy, infant mortality, health and most programs organize groups of 10 to and nutrition, or infrastructure investment. 20 women (with one facilitator) who meet weekly, biweekly, or monthly. Koolwal (2007) PROGRAM SAMPLE offers an exception by studying the effect After initial decisions on which villages to of participation in mixed-gender networks enter, projects will usually recruit women of 5 or 6 individuals. In a similar vein, the from low-income or below-poverty-line Safe Cities Initiative in Madhya Pradesh and households, with a focus on vulnerable the Do Kadam Barabari ki Ore program groups and Scheduled Castes or Tribes. in Bihar, both programs designed to Although most programs target any adult promote the prevention of violence against woman, participants tend to be 34 to 36 women, include separate intervention arms years old and have 1 to 5 years of education. comprising groups of only men (Holden et In a contrasting example, the Do Kadam al. 2016; Jejeebhoy et al. 2017). All programs Barabari Ki Ore program in India targets its in our sample provide training. However, program to prevent violence against women the training components vary and usually specifically to married women and their include a subset of job, skills or livelihood husbands. training, literacy camps, bookkeeping, education on topics of empowerment or CONDITIONS FOR MEMBERSHIP gender discrimination, and training on The most common restriction for joining a addressing social problems such as dowry SHG is that no more than one member per or violence against women. household can join a single group. SEWA has an additional condition of annual dues to PROGRAM GOVERNANCE join, amounting to 5 INR. In Bangladesh, the Often, programs construct federations at Association for Social Advancement credit the village, block/mandal, district or state groups requires admission fees, attendance level. The federations implement their in mandatory meetings, and permit own interventions including links to local withdrawals only if the member leaves the government, training and livelihood activities, group (Steele, Amin, and Naved 1998). subsidies, and provisions of lending capital or seed grants. In one example, the Indhira Kranthi Patham (IKP) program in Andhra 5 The NRLP is a subset of the National Rural Livelihoods Mission, and partners with other Pradesh groups approximately 20 SHGs into programs to implement their model at the state level. JEEViKA, also known as the Bihar Rural Livelihoods Project was funded by the World Bank and executed by the Bihar Rural Livelihoods village organizations that offer members Promotion Society. It was scaled up under the National Rural Livelihoods Mission. TRIPTI or the activities such as agriculture marketing and Odisha Rural Livelihoods Project and PRADAN also implement the mission of the NRLM. 6 Table A.1 in the appendix summarizes program details, as reported by the authors. job training (Deininger and Liu 2013). 4 | SYSTEMATIC REVIEW ON WOMEN’S ECONOMIC EMPOWERMENT PROGRAM COST at the subgroup level, such as earnings attributed specifically None of the 33 studies included in this to agriculture, casual wage labor, or the Mahatma Gandhi report discuss the cost effectiveness of the National Rural Employment Guarantee Act (MGNREGA) programs they evaluate, or the average workfare program. Reported income outcomes are costs of training modules. However, studies summarized in Table A.2 of the appendix. not included in the systematic review Of the 14 unique income estimates for participants, 5 are find evidence for significant economies positive, 1 is negative, and the remaining 8 are insignificant of scale. Specifically, Siwach, Paul, and at the 95 percent confidence level. Estimated effects de Hoop (2022) find that the annual cost are heterogeneous, with coefficient magnitudes ranging per member fell from $29 to $5 once between –22.8 and 96.3 percent. Where significant impacts the JEEViKA program was scaled up and exist, the effect of SHG membership on participating women federated—increasing membership from is an average 35 percent increase across income types after 100,000 members to 10 million. about two years of membership. Interestingly, Swain and Varghese (2009) find evidence that the length of membership STUDY DESIGNS in SHGs is positively associated with individuals moving away SHG programs are often assigned to villages from income generation from agriculture to other sources. based on predetermined characteristics, Finally, evidence of spillover effects on nonparticipants or so the most common methodology used in changes at the village level is weak. our sample of studies to estimate impacts Only five studies estimate the impact of SHGs on savings, is Propensity Score Matching.7 Three but the results are clear: membership significantly increases studies use a randomized control trial (RCT) savings accumulation and the amount members can raise in design to evaluate the SEWA and Safe Cities emergencies. These results are plausible, because members Initiative programs (Desai and Joshi 2014; are encouraged to save regularly as a core component of Desai and Olofsgård 2019; Holden et al. SHG participation. Specifically, Deininger and Liu (2013) 2016). Where reported, studies evaluate show that individuals in IKP villages are 13.2 percentage program impacts after 1 to 6 years of points more likely to set aside money for themselves after SHG membership, where both the mean three years. Swain and Varghese (2009) find that older and median length of membership is about members in Andhra Pradesh (individuals who have been SHG 3 years. members for at least six months) have approximately double the savings of new members. In West Bengal, members can raise 72 percent more than nonmembers in emergencies, PROGRAM IMPACTS relative to a baseline of 2,550 INR (Dutta 2017). In Odisha, households in areas where TRIPTI is present are 2.3 percent For this review, program impacts are separated into income, more likely to save and 66 percent more likely to rely on labor market, and empowerment outcomes. The large SHGs for savings, compared to households in non-TRIPTI pool of papers allows us to gauge separate impacts on SHG areas (Joshi, Palaniswamy, and Rao 2019). participants, spillovers on nonparticipants in program villages, and intent-to-treat (ITT) effects at the village level. For ease of B. LABOR MARKET PARTICIPATION interpretation, we present estimates in percent or percentage Of the 33 studies included in the review, 14 measure impacts point changes wherever possible. In all cases, the results on various labor market outcomes, including employment, presented are estimated impacts for women, even where whether participants own a MGNREGA scheme card, interventions have separate treatment arms for men. the number of days worked under the MGNREGA, and A. INCOME AND SAVINGS the number of hours worked.8 Table A.3 in the appendix summarizes the measurement and magnitude of impacts Eight studies measure the effect of SHG programs on income. across studies. The type of income measured varies, with some articles reporting total annual income, while others provide changes 8 The MGNREGA scheme guarantees at least 100 days of paid employment in jobs requiring unskilled manual labor. The scheme is restricted to job card holders. However, the process of obtaining a job card is difficult and can require several trips outside the village and rejections 7 Propensity score matching, or other matching techniques, allow for the construction of a by program supervisors. Thus, owning a MGNREGA job card is a proxy for access to outside control group similar in characteristics to SHG members. Attributes such as income, caste, or employment, increased mobility, and access to income (Kandpal, Baylis, and Arends- household savings are examples of characteristics used to construct the control group. Kuenning 2013). APRIL 2022 | 5 Figure 2: Labor Force Participation Outcomes for SHG Members Note: The figure shows labor force participation outcomes for SHG participants. The labels indicate the outcome and the type of effect measured (direct benefits of membership, joint effect of membership and having participating peers, and the effect of being an early versus late member). The identification strategy employed by Kochar et al. (2020) compares members in villages with early versus late treatment, capturing the impact of 2.5 extra years of membership on labor market outcomes. The female labor force participation (FLFP) primary and secondary occupation variables capture the proportion of females who are active in productive activities as the primary or secondary activity status. Authorship details are in parentheses. Coefficients have been converted to percentage point changes. 95 percent confidence intervals displayed. Fourteen unique estimates of labor force participation were Pradesh were also more likely to be aware of the MGNREGA recorded for participants, 11 of which were insignificant at workfare program, and to own and have used a job card the 95 percent confidence level, 1 was positive and 2 were (Kandpal, Baylis, and Arends-Kuenning 2013; Kumar et al. negative (with an average magnitude of 2.9 and a median 2019). of 2.2 percentage points). These estimates are displayed in Several papers report village ITT effects if identification at the Figure 2. At the less-restrictive 90 percent confidence level, individual level was not possible or if the average effect of studies that found positive impacts on the likelihood that exposure to SHG programs was of interest. The results provide participants are employed outside the household, in the inconclusive evidence for the impact of SHG programs on range of 3.7 to 42 percentage points, evaluated the NRLP, the extensive margin of labor force participation.9 Of the 12 SEWA, and Mahila Samakhya programs 2 to 4 years after estimates recorded, 4 are positive and 8 are insignificant at program implementation. Interestingly, despite overlap in program modules, the stated objectives differ, with NRLP and SEWA being self-described livelihoods program and 9 Here, the labor force participation category includes the likelihood of being employed in Mahila Samakhya being an education program. Besides casual or other labor, the proportion of women in a household who work for pay, the work participation rate for females in a household in self-employment, farm, non-farm, casual, formal, employment, SHG participants in Uttarakhand and Madhya or other livelihood activities, and the agriculture or non-agriculture FLFP rate in a village. 6 | SYSTEMATIC REVIEW ON WOMEN’S ECONOMIC EMPOWERMENT the 95 percent confidence level. The evidence for impact on Fifteen papers estimate the impact of SHGs on civic inclusion. the intensive margin, as measured by the number of days Our civic inclusion category captures both changes in worked under the MGNREGA program (overall or for pay), is knowledge of civic procedures, such as where to report also inconclusive. infrastructure grievances or actual reporting of grievances, As SHG programs provide multiple treatments to members as well as political participation, which includes voting, simultaneously (through exposure to peers, saving attending village meetings, and speaking at village meetings. requirements, changes in mobility to attend meetings, or Across settings, participants were 4 to 18 percentage points training), it is difficult to discern which program components more likely to attend village council meetings compared to are driving specific changes in economic outcomes. To nonparticipants. In rural Rajasthan, Desai and Joshi (2014) show understand one possible mechanism, Desai and Joshi that after two years, SEWA not only increased participants’ (2014) show that participation in SEWA’s vocational training knowledge of where to report civic grievances, but also programs (craft making, simple product manufacturing, increased their likelihood of approaching local authorities to agriculture training) led to a significant 14 percentage report issues. For example, participants were 14.2 percentage point increase in non-farm employment. In fact, even points more likely to report water infrastructure issues, a 68 nonparticipating women in SEWA villages were 7 percentage percent increase relative to the pooled sample mean (Desai points more likely to be employed off the farm 2 years after and Joshi 2014). This result is particularly interesting because program implementation, likely because SEWA allows all women in the region are primarily responsible for collecting women in the village including nonmembers to attend its water, indicating that participants are using the information training programs. This result indicates that the provision of obtained from the program to advocate for topics of benefit to skill or vocational training can partly explain the success of themselves and their peers. In addition, Joshi and Rao (2017) programs in increasing economic participation. find that women in the NRLP groups with external facilitators are 17.2 percentage points more likely to participate in higher C. EMPOWERMENT level institutions and attend two more village meetings on We define empowerment as a process involving the average, compared to groups with internal facilitators. Finally, freedom and expansion of choices and actions available to evidence for a positive impact on nonparticipants is weak. women, and the strengthening of their voices so that they Social capital includes trust for others in the community, may exert greater control over their lives (Narayan 2005). contribution to village projects (maintenance of roads, We pool reported empowerment outcomes into several schools, or bridges), the probability of knowing or speaking subcategories: mobility, civic inclusion, violence against to randomly chosen women, and the number of people who women (experienced and attitudes, separately), social can be approached for support or credit. Where impacts capital, norms and aspirations, and decision-making. exist, SHGs positively affect the social capital of participants Fourteen studies measure changes in women’s movement and nonparticipants in the village. Of the 33 estimates for outside the household, including their ability to visit places participants, 24 are positive and 9 insignificant; Figure 3 alone or with someone else (with or without permission), displays select results. For example, Kumar et al. (2019) find a 7 indexes of the number of places they can visit, and the percent increase in the probability of knowing at least 1 out of frequency with which they traveled outside the home or 5 randomly chosen women from a village after approximately village in a certain period. Of the 29 unique mobility estimates 4.2 years of membership, from a baseline probability of 0.74. for participants, 12 are positive and 17 are insignificant at Interestingly, members of PRADAN in Madhya Pradesh were the 95 percent confidence level, suggesting heterogeneity 5.3 percentage points (or 23.7 percent) more likely to discuss across programs. For example, the IKP program in Andhra politics semi-regularly with their friends after 6 years of Pradesh increased members’ ability to visit friends, family, membership, but not more likely to discuss politics with family health clinics, agricultural fields, and community centers (Prillaman 2016). Of the 9 estimates for nonparticipants, 5 are without anyone’s permission. After 4 years of membership positive and 4 are insignificant. in the IKP livelihood program, women were 9 percentage points more likely to visit family and 6 percentage points Across studies, decision-making is measured as a woman’s more likely to go to fields outside the village for work, from ability to participate in decisions regarding her labor, farming, a baseline of 52 and 42 percent respectively (Prennushi and household purchases, healthcare, children’s education, or Gupta 2014). Both programs designed to prevent violence voting. Consistent evidence across programs for changes in against women did not improve their mobility. decision-making is weak. However, exceptions exist. From their randomized control trial, Desai and Joshi (2014) show that two years after the random introduction of SEWA to APRIL 2022 | 7 Figure 3: Select Social Capital Outcomes for SHG Members Note: The figure displays select social capital outcomes for SHG participants. The labels indicate the outcome measured. Coefficients have been converted to percentage point changes. 95 percent confidence intervals displayed. their villages, participating women were 11.9, 12.8, and experienced or attitudes about violence toward women 7.4 percentage points more likely to make independent are also not significantly affected by programs like the Safe decisions regarding children’s schooling, healthcare, and Cities Initiative, which have a primary aim of preventing such family planning. Similarly, Deshpande and Khanna (2021) find violence. Holden et al. (2016) suggest this lack of evidence that SHG membership improves the likelihood that women can be partly explained by implementation weaknesses provide some input on decisions regarding their jewelry, and recommend that positive results can be improved by migration, creditors, and the purchase of durable goods. appropriate training of field staff and a recognition of the relative importance of prescriptive versus descriptive norms There is some evidence for improvements in financial within the study context. empowerment, as measured by knowledge of the name of the closest bank, having a fixed deposit, bank, or post office account in own name, ability to sign own name or CONCLUSION read signposts. Specifically, Deshpande and Khanna (2021) Several observations emerge from this review. First, show that individuals in SHGs are 18.3 percentage points participation in self-help group programs positively affects more likely to know where the closest bank is, an increase saving amounts, civic inclusion, and measures of social of approximately 29 percent from baseline. After two years, capital. An increase in savings is a direct implication of SHG women in JEEViKA villages were 12.4 percentage points more membership; nonetheless, accumulated saving impacts tend likely to be able to sign their own name, a 33 percent increase to be economically modest to large in magnitude. Second, from the baseline probability of 0.37 (Surendra 2020). results for income and labor market outcomes are mostly No consistent evidence across studies indicates that the insignificant, with some exceptions. While most programs presence of SHGs significantly affects, either positively pool interventions making it difficult to discern drivers of or negatively, the sexual, physical, or emotional violence experienced by women at home or in public. Likewise, SHG programs do not significantly affect acceptability or attitudes 10 The norms and aspirations subcategory includes whether respondents want their daughter to work after marriage, the ideal years of education for children, or an index on gender toward reporting violence against women to local authorities, attitudes. The self-perception subcategory includes feeling confident speaking in public or norms and aspirations, or self-perceptions.10 Interestingly, participating in politics, civic skills, and feeling respected in the household. 8 | SYSTEMATIC REVIEW ON WOMEN’S ECONOMIC EMPOWERMENT successful programs, there exist some indications of which Christian, Paul, Eeshani Kandpal, Nethra Palaniswamy, and components are successful at improving women’s economic Vijayendra Rao. 2019. “Safety Nets and Natural Disaster empowerment. In particular, skills and vocational training Mitigation: Evidence from Cyclone Phailin in Odisha.” seem to have driven increases in labor force participation Climatic Change 153: 141–64. https://doi.org/10.1007/ for women. In SEWA program villages, this result also s10584-018-02364-8 holds for nonmember women who are not exposed to the credit and savings module, suggesting impacts are driven Das, Sabyasachi, Pushkar Maitra, Paromita Sanyal. 2019. primarily through the livelihood training pathway. Third, “Credit groups, women’s political engagement and public evidence for an impact of membership on experience of or goods provision.” Working Paper. attitudes toward violence against women, self-perceptions, Datta, Upamanyu. 2015. “Socio-economic Impacts of or aspirations is insufficient, possibly due to the difficulty of JEEViKA: A Large-Scale Self-Help Group Project in Bihar, India.” changing sticky norms within a short period. World Development 68: 1–18. https://doi.org/10.1016/j. Beyond the direct effects on participants, it is also interesting worlddev.2014.11.013 to note the potential for SHGs to affect the welfare of other de Hoop, Thomas, Luuk van Kempen, Rik Linssen, and Anouka community members. For example, evidence suggests that van Eerdewijk. 2014. “Women’s Autonomy and Subjective SHGs may have significant implications for village financial Well-Being: How Gender Norms Shape the Impact of Self- markets as they offer a medium for credit and link groups to Help Groups in Odisha, India.” Feminist Economics 20 (3): larger federations with controlled rates of exchange, putting 103–35. https://doi.org/10.1080/13545701.2014.893388 downward pressure on interest rates offered by other moneylenders. This is particularly true for larger programs Deininger, Klaus, and Yanyan Liu. 2013. “Evaluating Program that can afford to saturate the village. In particular, Hoffman Impacts on Mature Self-Help Groups in India.” World Bank et al. (2021) find that JEEViKA led to a 0.7 percentage point Economic Review 27 (2): 272–96. https://doi.org/10.1093/ reduction in average village interest rates, from the previous wber/lhs028 rate of 5.27 percent. Similarly, Pandey, Gupta, and Gupta (2019) find that the NRLP led to a 19 percent reduction in ——— 2013. “Economic and Social Impacts of an Innovative interest paid on outstanding loans in treated villages. Self-Help Group Model in India.” World Development 43: 149–163. http://dx.doi.org/10.1016/j.worlddev.2012.09.019 Finally, the review finds several avenues needing further exploration to allow for a better understanding, and the Demont, Timothée. 2014. “Microcredit as insurance: successful replication, of the group-based SHG model. First, Evidence from Indian Self-Help Groups.” Working Papers because programs simultaneously expose women to multiple 1410, University of Namur, Department of Economics. interventions at once, there is a need to discern specifically Desai, Raj M., and Shareen Joshi. 2014. “Collective Action which components, or a combination thereof, effectively and Community Development: Evidence from Self-Help improve women’s economic empowerment outcomes. Groups in Rural India.” World Bank Economic Review 28 (3): Second, there is a lack of understanding regarding the impact 492–524. https://doi.org/10.1093/wber/lht024 of multiple SHG program features, including group sizes, characteristics of the meeting facilitator, costs, the amount Desai, Raj M., and Anders Olofsgård. 2019. “Can the Poor of savings required per meeting, as well as the linkage of Organize? Public Goods and Self-Help Groups in Rural India.” the SHG to larger federations. Finally, causal literature on World Development 121: 33–52. https://doi.org/10.1016/j. the impact of SHGs on women’s economic empowerment worlddev.2019.04.009 in South Asian countries other than India and Bangladesh is scant, despite the existence of similar programs. Deshpande, Ashwini, and Shantanu Khanna. 2021. “Can Weak Ties Create Social Capital? Evidence from Self-Help Groups in Rural India.” World Development 146: 105534. REFERENCES https://doi.org/10.1016/j.worlddev.2021.105534 Brody, Carinne, Thomas de Hoop, Martina Vojtkova, Ruby Dutta, Souvik, Abhirup Sarkar, and Suraj Shekhar. 2017. Warnock, Megan Dunbar, Padmini Murthy, and Shari L. “Self Help Groups: Evidence from India.” IGC Working Paper Dworkin. 2015. “Economic Self—Help Group Programs for I-35106-INC-1. Improving Women’s Empowerment: A Systematic Review.” Garikipati, Supriya. 2008. “The Impact of Lending to Women Campbell Systematic Reviews 11 (1): 1–182. https://doi. on Household Vulnerability and Women’s Empowerment: org/10.4073/csr.2015.19 APRIL 2022 | 9 Evidence from India.” World Development 36 (12): 2620– Khanna, Madhulika, Nishtha Kochhar, and Nethra 2642. http://dx.doi.org/10.1016/j.worlddev.2007.11.008 Palaniswamy. 2014. “A Retrospective Impact Evaluation of the Tamil Nadu Empowerment and Poverty Alleviation Gugerty, Mary Kay, Pierre Biscaye, C. Leigh Anderson. 2019. (Pudhu Vaazhvu) Project.” The Journal of Development “Delivering development? Evidence on self-help groups Studies 51 (9): 1210–1223. https://doi.org/10.1080/002203 as development intermediaries in South Asia and Africa.” 88.2015.1028538 Development Policy Review: The Journal of the Overseas Development Institute 37 (1): 129–151. https://doi:10.1111/ Kochar, Anjini, Bidisha Barooah, Chandan Jain, Geeta Singh, dpr.12381 Nagabhushana Closepet, Raghunathan Narayanan, Ritwik Sarkar, and Rohan Shah. 2020. “Impact Evaluation of the Hoffmann, Vivian, Vijayendra Rao, Vaishnavi Surendra, and National Rural Livelihoods Project.” 3ie Impact Evaluation Upamanyu Datta. 2021. “Relief from Usury: Impact of a Report 128, 2020. Self-Help Group Lending Program in Rural India.” Journal of Development Economics 148. https://doi.org/https://doi. Koolwal, Gayatri B. 2007. “Group Diversity and the Impacts org/10.1016/j.jdeveco.2020.102567 on Female Participants: A Quasi-Experimental Study of Income Generating Networks in India” June. https:// Holden, Jenny, Macartan Humphreys, Sohail Husain, Sarah citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.1068. Khan, and Summer Lindsey. 2016. “Evaluation of the Madhya 8613&rep=rep1&type=pdf Pradesh Safe Cities Initiative 2013–2016: Endline Report.” http://www.sddirect.org.uk/media/1348/20161201-mp-ie- Kumar, Neha, Kalyani Raghunathan, Alejandra Arrieta, Amir final-endline-report_final-edits-is.pdf Jilani, and Shinjini Pandey. 2021. “The Power of the Collective Empowers Women: Evidence from Self-Help Groups in India.” Janssens, Wendy. 2010. “Women’s Empowerment and World Development 146: 105579. https://doi.org/10.1016/j. the Creation of Social Capital in Indian Villages.” World worlddev.2021.105579 Development 38 (7): 974–988. http://dx.doi.org/10.1016/j. worlddev.2009.12.004 Narayan, Deepa. 2005. “Measuring Empowerment: Cross Disciplinary Perspectives” World Bank. https://openknow Jejeebhoy, Shireen J., K.G. Santhya, Rajib Acharya, A.J. ledge.worldbank.org/handle/10986/7441 License: CC BY 3.0 Francis Zavier, Neelanjana Pandey, Santosh Kumar Singh, IGO Komal Saxena, Shilpi Rampal, Sharmistha Basu, Aparajita Gogoi, Madhu Joshi, and Sandeep Ojha. 2017. “Empowering Palaniswamy, Nethra, Ramya Parthasarathy, and Vijayendra women and addressing violence against them through self- Rao. 2019. “Unheard Voices: The Challenge of Inducing help groups (SHGs).” New Delhi: Population Council. Women’s Civic Speech.” World Development 115: 64–77. https://doi.org/10.1016/j.worlddev.2018.10.007 Joshi, Shareen, Nethra Palaniswamy, and Vijayendra Rao. 2013. “Impact Evaluation Framework and Results: Odisha Pandey, Vivek, Abhishek Gupta, and Shivani Gupta. 2019. Rural Livelihoods Project,” 1–35. “Labor and Welfare Impacts of a Large-Scale Livelihoods Program: Quasi-Experimental Evidence from India.” Policy Joshi, Shareen, and Vijayendra Rao. 2018. “Who Should Be Research Working Paper, World Bank. https://doi.org/ at the Top of Bottom-Up Development? A Case-Study of 10.1596/1813-9450-8883 the National Rural Livelihoods Mission in Rajasthan, India.” Journal of Development Studies 54 (10): 1858–77. https:// Prennushi, G., and A Gupta. 2014. “Women’s empowerment doi.org/10.1080/00220388.2017.1329526. and socio-economic outcomes: Impacts of the Andhra Pradesh rural poverty reduction program.” Policy Research Kandpal, Eeshani, Kathy Baylis, and Mary Arends-Kuenning. Working Paper Series, World Bank. http://search.proquest. 2013. “Measuring the Effect of a Community-Level Program com/docview/1531426741?accountid=14695 on Women’s Empowerment Outcomes: Evidence from India.” Policy Research Working Paper Series, World Bank. Prillaman, Soledad Artiz. 2021. “Strength in Numbers: How Women’s Groups Close India’s Political Gender Gap.” Kandpal, Eeshani, and Kathy Baylis. 2019. “The Social Lives American Journal of Political Science 1–92. https://doi. of Married Women: Peer Effects in Female Autonomy and org/10.1111/ajps.12651 Investments in Children.” Journal of Development Economics 140: 26–43. https://doi.org/10.1016/j.jdeveco.2019.05.004 10 | SYSTEMATIC REVIEW ON WOMEN’S ECONOMIC EMPOWERMENT Raghunathan, Kalyani, Samyuktha Kannan, and Agnes R. Quisumbing. 2019. “Can Women’s Self-Help Groups Improve Access to Information, Decision-Making, and Agricultural Practices? The Indian Case.” Agricultural Economics 50 (5): 567–80. https://doi.org/10.1111/agec.12510. Siwach, Garima, Sohini Paul, and Thomas de Hoop. 2022. “Economies of scale of large-scale international development interventions: Evidence from self-help groups in India.” World Development 153. https://doi.org/10.1016/j. worlddev.2022.105839. Steele, Fiona, Sajeda Amin, and Ruchira Tabassum Naved. 1998. “The impact of an integrated micro-credit programme on women’s empowerment and fertility behavior in rural Bangladesh.” https:// doi:10.31899/pgy6.1016. Surendra, Vaishanavi. 2020. “Essays on Credit Markets in Rural India.” UC Berkeley. ProQuest ID: Surendra_ berkeley_0028E_19898. Merritt ID: ark:/13030/m5616g87. Retrieved from https://escholarship.org/uc/item/4h88t7hm Swain, Ranjula Bali, and Adel Varghese. 2009. “Does Self Help Group Participation Lead to Asset Creation?” World Development 37 (10): 1674–1682. https://doi.org/10.1016/j. worlddev.2009.03.006 STAY CONNECTED We gratefully acknowledge funding from the South Asia Trade Facilitation Program (SARTFP) and the Umbrella Facility for Gender Equality (UFGE). SARTFP is a trust fund administered by the World Bank with financial contribution from the Government of Australia’s Department of Foreign Affairs and Trade. UFGE is a multi-donor trust fund administered by the World Bank to SARGENDERLAB@WORLDBANK.ORG advance gender equality and women’s empowerment through experimentation and knowledge creation aimed at helping governments and the private sector focus policies and programs on WORLDBANK.ORG/SARGENDERLAB scalable solutions with sustainable outcomes. The UFGE has received generous contributions from Australia, Canada, Denmark, Germany, Iceland, the Netherlands, Norway, the Republic of Latvia, Spain, Sweden, Switzerland, the United Kingdom, the United States, and the Bill and Melinda Gates Foundation. APPENDIX Table A.1: Included Studies and Program Details ID Title and Program Region Sample Method Program descrip�on author(s) and design 1 The power of Mix of India: Panel survey data Nearest 10–20 women per group the collec�ve Professional Madhya collected in 2015 Neighbor empowers Assistance for Pradesh, and 2017 for Matching women: Development Chha�sgarh, sample of 1,470 Evidence from Ac�on (PRADAN) Jharkhand, rural women and self-help and non-PRADAN Odisha, West 1,344 men groups in India SHG groups Bengal Sample includes Kumar, ever-married Raghunathan, women aged 15 to Arrieta, Jilani, 49 years; members Pandey (2021) average 33 years old; 14% have 1–5 years of educa�on; 98% are married; 74% work outside the home 2 Relief from JEEViKA (Bihar India: Bihar Baseline sample of ANCOVA 10–15 women per group, Usury: Impact Rural Livelihoods 8,988 households weekly mee�ngs of a self-help Project) across 333 villages; Led through curriculum group lending baseline survey in Launched by on women’s program in 2011 with follow-up government of empowerment, basic rural India in 2014 Bihar with World literacy, numeracy, self- Hoffmann, Bank funding Primary targe�ng of advocacy, and Rao, Surendra, Scheduled Castes engagement in collec�ve Da�a (2021) and Tribes from ac�on; access to savings low-income and credit households Members contribute a minimum of 2 INR ($0.04) each week to personal savings account SHGs organized into federa�ons of village organiza�ons that provide lending capital of up to 50,000 INR per SHG, about 3 months a�er crea�on; members can borrow funds at 2% per month 12 | SYSTEMATIC REVIEW ON WOMEN’S ECONOMIC EMPOWERMENT ID Title and Program Region Sample Method Program descrip�on author(s) and design 3 Can weak �es Membership in India: Sample of 9,913 Propensity While respondents are 1 The power create socialof of Mix (government any India: Maharashtra Panel survey households data across Nearest Score 10–20 not women per necessarily from group the the collec capital? � ve Professional or Madhya collected 334 villagesin 2015 Neighbor Matching, Maharashtra NRLM SHGs empowers Evidence from Assistance for nongovernment) Pradesh, and 2017 for Matching Entropy (other programs also run 23% of female women: self-help Development SHG program Chha�sgarh, sample of 1,470 Balancing SHGs in sample area), sample is a SHG Evidence groups in from rural Ac�on (PRADAN) Jharkhand, rural women and the focus of NRLM is on member; members self-help India and non-PRADAN Odisha, West 1,344 men regular savings, internal average 5 years of groups in India SHG groups Bengal lending, discussion of Sample educa includes �on and 40% Kumar, ever-married health, sanita�on, or are from Scheduled Raghunathan, women Castes oraged Tribes15 to children’s issues, links Deshpande, Arrieta, (2021) Jilani, 49 years; members with local bodies, Khanna Women Pandey (2021) average 33 years planning for sustainable ques�onnaire with livelihoods, and access to old; 14% have 1–5 adult ever-married government schemes years of educa�on; women aged 18 to 98% are married; 50 years; poor, rural 74% work outside households the home 4 Access to JEEViKA (Bihar India: Bihar 8,988 households ANCOVA 10–15 women per 2 Relief finance,from JEEViKA Rural (Bihar Livelihoods India: Bihar Baseline sample from 333 villages; of ANCOVA 10–15 weekly group; women per� mee group, ngs Usury: Impact empowerment, Rural Livelihoods Project), measure 8,988 households baseline survey in 2011 weekly mee�ngs of a self-help Project) impact across 333 villages; Curriculum of women’s and women's two-year Led through curriculum group lending Members from low- baseline survey in empowerment, basic employment: Launched by on women’s program in income 2011 withhouseholds follow-up literacy, and numeracy; Experimental government of empowerment, rural India and primarily from in 2014 members save a basic evidence from Bihar with World literacy, numeracy, Scheduled Castes or minimum of 2 INR perself- week rural Bihar Hoffmann, Bank funding Primary targe�ng of advocacy, and Tribes Rao, Surendra, Surendra Scheduled Castes SHGs are federated engagement into in collec �ve Da�a (2021) 45% own some and Tribes fromland village ac�on; organiza access to�ons and savings (2020) and more than 80% low-income then cluster level and credit have some outstanding households federa�ons; groups can debt; almost 63% of Members contribute a borrow up to 50,000 INR women (ages 15 to minimum of 2 INR from the Village 70) work in the ($0.04) each week to Organiza�on a�er 3 months market for some personal savings account of regular savings part of the year SHGs organized into 5 Impact Na�onal Rural India: Bihar, 27,257 households Difference federa �ons ofper 10–12 women village group; evalua�on of Livelihoods Chha�sgarh, in 1,052 villages in organiza � ons that average 11 members; the Na�onal Program (NRLP), Jharkhand, Differences provide weekl lending capital y mee�ngs Average educa�on of up to 50,000 INR per Rural measure 3-year Madhya 2.84 years; 63% Mobilize SHG, aboutSHGs into village 3 months Livelihoods impact on average Pradesh, from Scheduled federa �ons, enhance a�er crea�on; members Project Maharashtra, Launched under Castes or Tribes; credit and marke�ng, can borrow funds at 2% Odisha, Kochar, the Na�onal 71% earn income fi nancial, per month technical Rajasthan, Barooah, Jain, Rural Livelihoods from unskilled wage services, provide skills U�ar Singh, Closepet, Mission (MRLM) labor and livelihood training, Pradesh, Narayanan, with support deliver social and economic West Bengal Sarkar, Shah from the World support, reduce poverty (2020) Bank Save 10 INR per week APRIL 2022 | 13 Table A.1: Included Studies and Program Details ID Title and Program Region Sample Method Program descrip�on author(s) and design 6 Unheard Pudhu Vaazhvu India: Tamil Data on 3,959 audio Propensity 10–15 members per voices: The Project (PVP) Nadu speech recordings Score group challenge of from 100 Gram Matching Implemented by Project facilitated SHGs inducing Sabha mee�ngs government of with credit, livelihood women's civic support, and crea�on of Tamil Nadu About 35% of speech social capital; worked sample households Palaniswamy, in PVP villages from closely with local Parthasarathy, Scheduled Castes or government for credit Rao (2019) Tribe; 59% female access and job training literacy rate ac�vi�es PVP federates SHGs into Village Poverty Reduc�on Commi�ees (PVRC) which focus on improving access to social safety nets, and livelihoods and youth training 7 The social lives Mahila Samakhya India: 487 women from 69 Instrumental Maximum 25 and of married U�arakhand villages plus two Variable average 17 women per Funded by DFID women: Peer friends surveyed group, biweekly literacy effects in (total sample of camp and educa�on, female 1,619 women) weekly voca�onal autonomy and training, and support Average women in investments in groups early 30s with less children than 8 years of Empower women Kandpal, Baylis educa�on; husband through formal and (2019) averages mid to late informal educa�on, 30s, with high voca�onal training, school educa�on; interac�ons with 22% of par�cipants government officials and are Brahmins; employers, provision of par�cipants have informa�on on accessing 3.75 friends on social safety nets average Program rollout usually begins with literacy camp 14 | SYSTEMATIC REVIEW ON WOMEN’S ECONOMIC EMPOWERMENT ID Title and Program Region Sample Method Program descrip�on author(s) and design 8 6 Social Unheard Membership Pudhu Vaazhvu in India: India: Tamil Data collected Data in on 3,959 audio Nearest Propensity Average 10–15 women 10–15 members per networks, voices: The any SHG Project (PVP) Madhya Nadu 2015 for sample speech recordings of Neighbor Score per group; group weekly challengeand mobility, of (PRADAN or non- Pradesh, 2,744 ever-married from 100 Gram Matching Matching mee�ngs poli�cal Implemented PRADAN) in by Odisha, rural women of Project facilitated SHGs inducing Sabha mee�ngs SHGs encourage par�cipa�on: government sample area of Chha�sgarh, ages 15–49 with credit, livelihood women's civic members to save support, and crea� an on of The poten�al Tamil Nadu Jharkhand, About 35% of speech About 38% of average of 10 INR social capital; workedper for women's West Bengal sample households sample is SHG closelyfrom week, with which local Palaniswamy, self-help groups in PVP villages from member (on members government can borrow. for credit Parthasarathy, to improve Scheduled Castes or average for 4.2 access and job training Rao (2019) access and use Tribe; 59% female Members meet to years) ac�vi�es of public literacy rate discuss ma�ers of en�tlement Sample averages 32 common interest PVP federates SHGsandinto schemes in India years old, 15% with disseminate Village Poverty informa �on 1–5 years of on health, livelihoods, Reduc�on Commi�ees Kumar, educa�on, 78% and nutri �on focus on (PVRC) which Raghunathan, from Scheduled improving social access to into SHGs are federated Arrieta, Jilani, Castes or Tribes safety nets, village-level and livelihoods Chakrabar�, and youth organiza training �ons Menon, 7 Quisumbing The social lives Mahila Samakhya India: 487 women from 69 Instrumental Maximum 25 and (2019) of married U�arakhand villages plus two Variable average 17 women per Funded by DFID 9 women: Peer Safety nets and Odisha Rural India: Odisha friends surveyed Two survey rounds Triple group, biweeklytraining TRIPTI provides literacy e ffects in natural Livelihoods (total sample of 2,874 of households Difference camp on theand educa�on,of management female disaster Program called 1,619 women) from 160 villages weekly voca�lending group-based onal and mi�ga�on:and autonomy Targeted Rural surveyedwomen in 2011in training, and support links SHGs to public and Average investments Evidence fromin Ini�a�ves for and 2014 groups private sectors, which early 30s with less children cyclone Phailin Poverty then provided than 8 years of Empower women in OdishaBaylis Kandpal, Termina�on and educa�on; husband agriculture through subsidies formal and and Chris�an (2019) Infrastructure averages mid to late product market informal educa�linkages on, Kandpal (TRIPTI), interacted 30s, with high voca alsotraining, �onal TRIPTI provides Palaniswamy, Funded by the school educa�on; grants � interac to onshouseholds with Rao (2019) World Bank, 22% of par�cipants iden�fied as o government extremely fficials and implemented by are Brahmins; poor or poor employers, provision of the government par�cipants have informa�on on accessing of Odisha 3.75 friends on social safety nets 10 Labor and Na�onal Rural India: 4,316 households average Propensity Groups provided with welfare Livelihoods Jharkhand, sampled Score Program seed fundsrollout and usually linked to impacts of a Mission (NRLM), Maharashtra, Matching, begins with literacy banks or credit Sample includes camp large-scale by the Madhya Instrument Members are socially g women from poor, livelihoods overnment of Pradesh al Variable mobilized, trained in social rural households program: India and economic skills, and belonging to Quasi- provided informa�on Scheduled Castes or experimental about government Tribes; about 46% evidence from programs of households own India some land SHGs are federated into Pandey, Gupta, various levels at which Gupta (2019) training on agriculture, livestock, and business startup are provided APRIL 2022 | 15 Table A.1: Included Studies and Program Details ID Title and Program Region Sample Method Program descrip�on author(s) and design 11 6 Can women's Unheard Professional Pudhu Vaazhvu India: Tamil India: Sample Data on of 977audio 3,959 ever- Nearest Propensity 10–20members women per 10–15 per self-help voices: The Assistance Project for (PVP) Madhya Nadu married women speech recordings Neighbor Score group mee�ng regularly group; groups challenge of Development Pradesh, aged 100 from 15–49 Gramyears Matching Matching Implemented by Members Project depositSHGs facilitated savings inducing access improve Ac�on (PRADAN), Chha�sgarh, from 80 Sabha villages; mee �ngs from which loans with credit, livelihood are to informa�on, government an NGO of Jharkhand, cross-sec�onal data women's civic disbursed; support, and program crea�on of decision- Tamil Nadu Odisha, West About 35% collected inof2015 speech Also partners bundles agriculture social capital; worked and making, and Bengal sample households with the NRLM 414 women were livelihood interven closely with local � ons, Palaniswamy, agricultural in PVP villages from SHG members, 563 builds socialfor government capital, credit Parthasarathy, prac �ces? The Scheduled Castes or were nonmembers improves access andaccess to inputs, job training Rao (2019) Indian case Tribe; 59% female literacy rate �vi�es and technical markets, ac Members average Raghunathan, knowledge; works with 34 years old with 4- PVP federates SHGs into Kannan, women to bring awareness year SHG Village Poverty Quisumbing to gender equality, provide memberships; Reduc�on Commi�ees (2019) a pla�orm for shared about half have (PVRC) which focus on experiences and to ini�ate access to their own improving access to social social and poli�cal ac�ons money and more safety nets, and livelihoods than 60% are from Livelihood and interven�ons youth training Scheduled Tribes include demonstra�ng 7 The social lives Mahila Samakhya India: 487 women from 69 agricultural Instrumental Maximum 25 prac and �ces, of married U�arakhand villages plus two Variable organizing average 17 women in per Funded by DFID women: Peer friends surveyed producer group, groups, biweekly and literacy effects in (total sample of providing camp support and educa �on, for female 1,619 women) price nego weekly voca �� �ons aonal autonomy and training, and support 12 Can the poor Self-Employed India: Average women in 1,442 female Randomized About 20 women per investments in groups organize? Women’s Rajasthan early 30s with residents from less Control Trial group; mee�ng monthly children Public goods Associa�on than SEWA 8 years ofand villages Empower for average women of 30-90 min Kandpal, Baylis and self-help (SEWA), an NGO 1,763� educa on; husband from non- through formal Group ac�vi�es andinclude (2019) groups in rural averages mid to late SEWA villages informal educa financial ac�vi�es �on, (loan India 30s, with high voca�onal training, school and 2009 educa�on; interac followed�onsby with discussion of Desai, 22% of par�cipants government o mutually important and fficials topics Olofsgård About are 83% of Brahmins; employers, provision of households in SEWA (such as civic ma�ers, (2019) par �cipants have informa domes� �con on accessing abuse, health villages belong 3.75 friends on to social safety nets ma�ers, alcoholism) Scheduled Castes or average Tribes; 18% of Program rollout usually Non-SHG ac�vi�es open to sample is literate; begins with literacy all adult females in village 95% is married; and camp 77% reside in Borrowing from bank a�er kutcha (temporary) 6 months of opera�on housing Nominal annual dues of 5 INR to join; savings target of 50–100 INR (USD 1.20) per month per member, deposited in group SHG linked bank Two leaders (Agewan) are typically selected per group 16 | SYSTEMATIC REVIEW ON WOMEN’S ECONOMIC EMPOWERMENT ID Title and Program Region Sample Method Program descrip�on author(s) and design 13 Who should be Government of India: Sample of 3,852 OLS Objec�ve of NRLM is to at the top of India’s Na�onal Rajasthan households across reduce poverty and bo�om-up Rural Livelihoods 17 districts. empower women; 7–12 development? Mission (NRLM), women per group; Compare outcomes A case study of locally known as monthly mee�ngs across SHGs with the Na�onal the Rajasthan internal versus Members pool savings Rural Rural Livelihoods external facilitators and discuss issues of Livelihoods Project; mutual importance; Mission in supported by provision of community Rajasthan, World Bank investment funds and India training of women to Joshi, Rao address social problems (2018) Facilitators in some SHGs are local women while in other SHGs it is an “outsider” from other parts of the country trained in mobiliza�on 14 Self help Paschim Banga India: West 563 individuals in Propensity 10 women per group groups: Grameen Bank Bengal SHGS, 235 Score Group leader, amount of Evidence from individuals not in Matching savings per week, India SHGs across 25 frequency of mee�ngs, villages Du�a, Sarkar, loca�on of mee�ngs Shekhar (2017) Members average decided by group 39 years old; half members jointly have zero years of educa�on; SHG members tend to have higher income compared to the nonmember sample 15 Strength in Professional India: 5,371 women and Regression 10–20 women per numbers: How Assistance for Madhya 2,399 men from 376 Discon�nui group; weekly, biweekly, women's Development Pradesh villages surveyed in ty Design or monthly mee�ngs networks close Ac�on (PRADAN) 2016 Provide informal savings India's poli�cal Average 41% Rural, mostly poor, and credit, links with gender gap uptake among ever-married lending ins�tu�ons, Prillaman women in a women training on farming and (2016) village agricultural prac�ces, access to social networks, discussions of group finances, personal or community concerns Nonmembers not allowed at mee�ngs APRIL 2022 | 17 Table A.1: Included Studies and Program Details ID Title and Program Region Sample Method Program descrip�on author(s) and design 16 Evalua�on of Safe Ci�es India: Panel data collected Difference 10–15 women per Madhya Ini�a�ve Madhya from 250 slums for in group; weekly or Pradesh Safe Pradesh 7,500 respondents Differences monthly mee�ngs Ci�es Ini�a�ve (RCT) 47% of sample Three treatment arms: Holden, includes direct (1) SHG strengthening Humphreys, beneficiaries module including Husain, Khan, (members of SHGs) training in record and Lindsey (2016) and indirect bookkeeping, regular beneficiaries group meetings, basic (members of wider gender training; (2) SHG community in slum) strengthening module plus violence against Sample of ages 18– women (VAW) module 49 years; 83% from for women (training in Scheduled response to VAW, Castes/Tribes or awareness and Other Backward preven�on, community Castes mobiliza�on); (3) Life skills module with men and boys to increase awareness of causes of VAW and provide training on ac�ons against VAW at the community level 17 Credit groups, Membership in India: 17 U�lize Rural Instrument 10–20 women per women's any SHG states Economic and al Variable group; typically meet 1– poli�cal Demographic 2 �mes per month to 88% of engagement Survey, which pool savings respondents and public covers 8,600 report presence Group ac�vi�es include goods households across of SHG in their discussions on loan provision 242 villages village; 13% of requirements, Das, Maitra, women were Average age in full repayment obliga�ons, Sanyal (2016) members of SHG sample is 38 years; and general ma�ers of 76% of respondents interest; members can are married with obtain loans through 5.5 years of bank-linking programs schooling; 32% belong to Scheduled Castes or Tribes 18 Socio- JEEViKA or Bihar India: Bihar Sample includes Propensity 10–15 women per Economic Rural Livelihoods 4,000 households in Score group; mee�ng regularly Impacts of Project, measure 400 villages; female Matching Promote socio-economic JEEViKA: A 3-year impact survey for ever- inclusion Large-Scale married women Funded by the Self-Help Ac�vi�es include savings, World Bank, Poor, rural Group Project borrowing, repayments; executed by the households largely in Bihar, India access to specific and Bihar Rural from Scheduled 18 | SYSTEMATIC REVIEW ON WOMEN’S ECONOMIC EMPOWERMENT ID Title and Program Region Sample Method Program descrip�on author(s) and design 18 Socio- JEEViKA or Bihar India: Bihar Sample includes Propensity 10–15 women per Economic Rural Livelihoods 4,000 households in Score group; mee�ng regularly Impacts of Project, measure 400 villages; female Matching Promote socio-economic JEEViKA: A 3-year impact survey for ever- inclusion Large-Scale married women Funded by the Ac�vi�es include savings, Self-Help World Bank, Poor, rural borrowing, repayments; Group Project executed by the households largely access to specific and in Bihar, India Bihar Rural from Scheduled general funding, promo�on Da�a (2015) Livelihoods Castes or Tribes of collec�ve ac�on and Promo�on poli�cal par�cipa�on Society, SHGs organized into inaugurated by federa�ons at village level, the government which provide funds and of Bihar; Scaled livelihood training up under NRLM Loan sizes range from 50 to 50,000 INR Members save between 2 to 10 INR per week 19 Women's Indira Kran� India: Andhra Sample of 4,250 Propensity Facilitate forma�on of Empowerment Patham Pradesh households in five Score groups of poor women, and Socio- districts; 39% from Matching provide seed funds and Implemented by Economic “poorest,” 31% training in social plus the government Outcomes “poor,” 22% “not so economic skills, create of Andhra poor,” and 8% from links with banks to access Prennushi, Pradesh with “not poor” credit, help members Gupta (2014) support from the households access government programs World Bank Increased access to market and social programs (pension, nutri�on centers, educa�on programs) SHGs organized into federa�ons at village mandal/block, district levels 20 Microcredit as Professional India: Sample of 1 080 Difference 10–20 women per insurance: Assistance for Jharkhand households from 36 in group; weekly mee�ngs , evidence from Development villages, 428 members, Differences Group decides on minimum Indian self-help Ac�on (PRADAN) 409 nonmembers in contribu�ons per member groups PRADAN villages; 214 (usually 5 to 10 INR per from villages with no month), interest rate to be Demont (2014) SHGs charged on loans, and Members average fines for late payments 24 years old, 52% Individual loans with belong to Scheduled public repayments; Castes or Tribes, group linked to bank for and 52% are below credit access a�er a year the poverty line of regular savings APRIL 2022 | 19 Table A.1: Included Studies and Program Details ID Title and Program Region Sample Method Program descrip�on author(s) and design 21 Collec�ve Self-Employed India: 1,410 women in Randomized 10–20 women per ac�on and Women’s Rajasthan SEWA villages, Control group; meet once a community Associa�on 1,795 in control Trial, month development: (SEWA), measure villages Propensity Intermediate with formal Evidence from 2-year impact Score The average age of financial sector, provide self-help Matching sample respondents pla�orm for engagement groups in rural is 37 years; 18% are in civic affairs, service India literate, 73% from delivery (e.g., childcare), Desai, Joshi Scheduled Tribes, basic voca�onal training, (2014) and 95% are form women’s groups, married; average provide informa�on partner age is 40 about government years, 8% are schemes and help with literate applica�ons Saving target between 25–100 INR ($5–20) per member per session All ac�vi�es led by SEWA field organizers who are usually local, married women with more than 12 years of educa�on Fee of 5 INR or $0.10 to become member 22 Measuring the Mahila Samakhya India: 487 women from 69 Ordinary Maximum 25 and average effect of a U�arakhand villages surveyed Least 17 women per group; Funded by DFID community- Squares, biweekly literacy camp Respondents level program Instrumental and educa�on, weekly average 32 years on women's Variables, voca�onal training, and old with 7 years of empowerment Propensity support groups; group educa�on; average outcomes: Score group size 5–15 women. age at marriage of Evidence from Matching 19, and fer�lity of Empower women through India 1.15 children; educa�on (literacy camps, Kandpal, husbands average adult educa�on classes, Baylis, Arends- 38 years old with 10 voca�onal training); Kuenning years of educa�on provide support groups (2013) to discuss social issues, encourage par�cipa�on in village poli�cs, and resolve domes�c disputes and community conflicts Facilitators, or sahyogini, cover a cluster of 10 villages 20 | SYSTEMATIC REVIEW ON WOMEN’S ECONOMIC EMPOWERMENT ID Title and Program Region Sample Method Program descrip�on author(s) and design 23 Economic and Indhira Kranthi India: Andhra 1,964 households in Propensity 10–20 women per group; Social Impacts Patham (IKP) also Pradesh IKP districts and Score about 70% of groups meet of an known as the 3,789 households in Matching, at least monthly Innova�ve District Poverty Rural Poverty Difference Self-Help SHGs have a mix of Ini�a�ves Project Reduc�on P roject in Group Model microcredit, empowerment (DPIP), measure districts Differences in India ac�vi�es, regular 3-year impact Sample from poor discussions, and social Deininger, Liu and vulnerable rural mobiliza�on; groups (2013) households, usually maintain records for Scheduled Tribes or internal lending and access Castes bank or project loans; groups include social Program villages ac�vi�es for improving have poor female empowerment, infrastructure and reducing discriminatory low levels of female prac�ces and vulnerability, empowerment and marke�ng; about 20% of SHGs have specific social ac�vi�es SHGs organized into federa�ons at village, mandal/block, district, and state levels 24 A Retrospec�ve Pudhu Vaazhvu India: Tamil Survey of 3,692 Propensity 10–15 women per group Impact Project (PVP) Nadu households almost Score Livelihood training Evalua�on of equally divided Matching targeted to the poor; the Tamil Nadu between PVP and cash grants and credit Empowerment non-PVP areas; for socially and Poverty conducted disadvantaged groups Allevia�on December 2012 to under the poverty line. (Pudhu March 2013 Vaazhvu) PVP links SHGs to village 14% of sample from Project organiza�ons that female-headed implement project Khanna, households ac�vi�es and further link Kochhar, to local government; Palaniswamy other ac�vi�es include (2013) providing access to safety nets and skilled employment APRIL 2022 | 21 Table A.1: Included Studies and Program Details ID Title and Program Region Sample Method Program descrip�on author(s) and design 25 Evalua�ng District Poverty India: Andhra Panel survey of Propensity 10–20 women per Program Ini�a�ves Project Pradesh 2,517 households in Score group; meet at least Impacts on (DPIP), es�mate 41 mandals (10 Matching, once per month Mature Self- impact of control and 31 Difference Discuss social issues, Help Groups in addi�onal 2.5 treatment) in form women's support India years on mature surveyed in 2004 Differences groups, provide SHG groups and 2006 Deininger, Liu intermedia�on with (2013) Implemented by formal finance sector, government of iden�fy skill gaps, Andhra Pradesh insurance, service with funding delivery, and basic from the World skills/job training Bank Members can apply for internal loans once savings are accumulated; access to commercial loans established once the group has a history of savings and repayment SHGs organized into federa�ons at village, mandal, district, and state levels through v illage organiza�ons; federa�ons assist in implementa�on of government programs (such as old age or disability benefits) 26 Women's Mahila India: Bihar 718 par�cipants, OLS, Average 28 women per Empowerment Samakhya, 714 Propensity group; mee�ngs several and the measure 4.7-year nonpar�cipants, Score �mes a month Crea�on of impact on and 559 control Matching Organizes support groups, Social Capital average households rota�ng savings, credit, in Indian surveyed Average 5.4% of provides health training Villages female Target Scheduled and hygiene knowledge, Janssens popula�on in and Backward helps access government (2010) program villages Castes; members subsidies and resources, par�cipates in from lower income se�les conflicts Mahila Samakhya and socially disadvantaged Facilitators, or sahyogini, groups; households cover a cluster of 10 villages average 7 members Groups begin to set their own agenda and mee�ng �mes, with less facilita�on, a�er 6–12 months 22 | SYSTEMATIC REVIEW ON WOMEN’S ECONOMIC EMPOWERMENT ID Title and Program Region Sample Method Program descrip�on author(s) and design 27 Women's Membership in India: Odisha Survey of 400 Propensity 10–20 women per group autonomy and any government households, 124 Score subjec�ve or NGO support households with a Matching, well-being: self-help group member in an NGO- Instrumental How gender supported SHG, 129 Variable norms shape in a government- the impact of supported SHG, and self-help 147 households groups in with no SHG members Odisha, India Households below de Hoop, the poverty line, Kempen, with monthly saving Linssen, rate less than 30 INR Eerdewijk (2014) 28 Does self-help Membership in India: Andhra Sample of 604 old OLS, Tobit 10–20 women per group group several SHGs Pradesh, SHG members, 186 Groups eligible to par�cipa�on considered, Tamil Nadu, new SHG members receive loans six months lead to asset sample averages U�ar who had not a�er a savings threshold crea�on? Pradesh, received financial is reached; groups services from the 18.75 months of Orissa, decide how to manage bank, and 52 loan Swain, nonmembers membership Maharashtra Some SHGs may include Varghese 95% of sample is (2009) training on primary female; 53% has no healthcare, skills, basic educa�on; average literacy, marke�ng, or age in sample is 34 family planning years 29 The impact of Membership in India: Andhra Sample of 117 Instrumental 10–15 women per lending to any SHG with Pradesh par�cipants and Variable group; 14.7 members on women on good opera�onal 174 nonpar�cipants average household links to banks 63% of income from Average SHG loan terms vulnerability SHG forma�on agriculture-related range from 6 to 24 and women's primarily work; average months with average empowerment: facilitated by monthly net per loan for group Evidence NGOs in sample capita income in amoun�ng to 26138 INR; from India area sample is 206 INR loans usually divided Garikipa� (60% of households equally among group (2008) fall below poverty members line) Members save 1 INR per day, which is used to provide loans APRIL 2022 | 23 Table A.1: Included Studies and Program Details ID Title and Program Region Sample Method Program descrip�on author(s) and design 30 Group District Poverty India: Survey of 274 Instrumental 5–6 men or women per Diversity and Ini�a�ves Project Madhya women across 240 Variable group the Impacts on (DPIP), a World Pradesh rural households Mo�va�on is to create Female Bank ini�a�ve DPIP districts have income security, Par�cipants: A low female literacy encourage par�cipa�on Quasi- and high infant from women and Experimental mortality rates vulnerable individuals, Study of and increase Income Program focuses on accountability of district Genera�ng women and and village governments Networks in vulnerable groups India Involves group par�cipa�on around an Koolwal (2007) income-genera�ng ac�vity (e.g., raising livestock), borrowing, and saving 31 Empowering Do Kadam India: Bihar Sample consists of Randomized 10–15 women per women and Barabari ki Ore 1,686 currently Control Trial group; fortnightly addressing (Two Steps married women mee�ngs of about two violence Towards Equality) ages 18–49, who hours each (total 24 against them program were members of sessions); monthly through self- SHGs; 688 husbands sessions for husbands help groups received treatment Women were chosen (SHGs) Targeted married from SHGs to obtain Jejeebhoy, women who were group learning sessions Santhya, Acharya, members of SHGs, that covered economic Zavier, Pandey, typically from large empowerment, gender Singh, Saxena, households and discrimina�on, no�ons Rampal, Basu, socially of masculinity and VAW Gogoi, Joshi, disadvantaged and girls, among others Ojha (2017) families Project linked members with livelihood training opportuni�es and credit access Select husbands were also chosen to receive corresponding training on VAW preven�on Female sessions delivered by Sakhi Salahkars drawn from members of the SHG who underwent preprogram training 24 | SYSTEMATIC REVIEW ON WOMEN’S ECONOMIC EMPOWERMENT ID Title and Program Region Sample Method Program descrip�on author(s) and design 32 Impact Odisha Rural India: Odisha Sample includes Regression Focus is on credit, evalua�on Livelihoods 2,874 households Discon�nuity livelihoods, social framework and Project (TRIPTI) from 160 villages at Design, mobiliza�on and results: Odisha end-line, surveyed Difference improving produc�ve Implemented by rural in 2011 and 2014 in capaci�es (improved the government livelihoods Differences micro investments, of Odisha, Project targeted project private transfers, assisted by the poor households, marke�ng support). Joshi, World Bank largely from Other ac�vi�es include Palaniswamy, Scheduled Castes savings, training, Rao (2019) and Tribes; mostly nego�a�ons with service involved in providers, and agriculture and sustainable livelihood casual labor development; may include agriculture, food security, and health and nutri�on-related interven�ons SHGs aggregated into federa�ons at the Gram Panchayat and block level 33 The impact of Credit and Bangladesh: 6,456 ever-married Mul�nomial Credit groups have an integrated savings programs Brahmanbaria women interviewed Logit formal, weekly mee�ngs micro-credit by Save the in the first round, presided over by a credit programme on Children USA, in 5,696 in second officer who collects women’s partnership with round; surveys savings and deposits empowerment the Associa�on conducted in 1993 them in a government and fer�lity for Social and 1995 bank behavior in Advancement Admission fee is rural required, mee�ngs are Bangladesh mandatory, and saving Steele, Amin, withdrawals are only Naved (1998) allowed if the member leaves the group; members are jointly responsible for loan defaults Savings groups have an informal structure, set their own rules regarding group size, frequency of mee�ngs, saving amounts, and management; groups may offer other ac�vi�es such as adult literacy programs APRIL 2022 | 25 Table A.2: Reported Income Outcomes Sample group Outcome Effect (percent change) (Swain, Varghese 2009) Does self-help group par�cipa�on lead to asset crea�on? Par�cipants Total income (INR) -10.51 Agriculture income (INR) -22.81** Other income (INR) 65** (Deininger, Liu 2013) Economic and Social Impacts of an Innova�ve Self-Help Group Model in India Par�cipants Per capita income (INR)—includes crop produc�on revenue, self- -9.33 employment profits, wages, sales of livestock or by-products Par�cipants, -7.5 converted from non-DPIP SHGs Nonpar�cipants 9.44 Village ITT 11 68.69 (Desai, Joshi 2014) Collec�ve ac�on and community development: Evidence from self-help groups in rural India Par�cipants Logged cash income earned over past three months 9.31 Nonpar�cipants -20 Village ITT 12 -15.3 (Prillaman 2016) Strength in numbers: how women’s networks close India’s poli�cal gender gap Par�cipants Income sufficiency—whether the household has enough income to 1.04 meet its needs (Du�a, Sarkar, Shekhar 2017) Self-help groups: Evidence from India Par�cipants Annual income (INR) 35.03** (Pandey, Gupta, Gupta 2019) Labor and welfare impacts of a large-scale livelihoods program: Quasi-experimental evidence from India Village ITT 13 Annual cash and in-kind income from all sources (INR) 4.6 Annual cash and in-kind income from migra�on (INR) 127.8*** Annual cash and in-kind income from livestock (INR) -3.9 (Kochar et al. 2020) Impact evalua�on of the Na�onal Rural Livelihoods Project Par�cipants Total income (past 12 months) 19.23*** Agriculture income (past 12 months) 6.07 Livestock income (past 12 months) 26.25 MGNREGA earnings (past 12 months) 96.31*** Casual wage income (past 12 months) 23.15*** 11 Enterprise income (past 12 months) Any household in a DPIP village is considered treated. 10.13 12 Any women living in SEWA villages are considered treated. 13 Households in program villages where take-up is at least 50 percent are considered treated. 26 | SYSTEMATIC REVIEW ON WOMEN’S ECONOMIC EMPOWERMENT Sample group Outcome Effect (percent change) (Swain, Varghese (Kochar 2009) et al. 2020) Does Impact self-help evalua group �on of par the Na �� cipa� onal on lead Rural to asset crea Livelihoods �on? Project Par�cipants (INR) 12 months) Total income (past 19.23*** -10.51 (INR) 12 months) Agriculture income (past 6.07 -22.81** Livestock income Other income (past 12 months) (INR) 26.25 65** MGNREGA (Deininger, Liu 2013) Economic earnings and Social(past 12 months) Impacts 96.31*** of an Innova�ve Self-Help Group Model in India Par�cipants Casual wage Per capita (past 12 months) income(INR)—includes income crop produc�on revenue, self- 23.15*** -9.33 Par�cipants, employment pro fi ts, wages, Enterprise income (past 12 months) livestock or by-products sales of 10.13 -7.5 converted from (Surendra 2020) Access to finance, empowerment, and women’s employment: Experimental evidence from rural non-DPIP SHGs Bihar �cipants NonparITT Village 14 Logged real agriculture wages 9.44 12** Village ITT 11 Logged real non-agriculture wages 68.69 10 (Desai, Joshi 2014) Collec Note: *p < 0.1, **p<0.05,� ve ac ***p < on and community development: Evidence from self-help groups in rural India �0.01 Par�cipants Logged cash income earned over past three months 9.31 Nonpar�cipants -20 Village ITT 12 -15.3 (Prillaman 2016) Strength in numbers: how women’s networks close India’s poli�cal gender gap Par�cipants Income sufficiency—whether the household has enough income to 1.04 meet its needs (Du�a, Sarkar, Shekhar 2017) Self-help groups: Evidence from India Par�cipants Annual income (INR) 35.03** (Pandey, Gupta, Gupta 2019) Labor and welfare impacts of a large-scale livelihoods program: Quasi-experimental evidence from India Village ITT 13 Annual cash and in-kind income from all sources (INR) 4.6 Annual cash and in-kind income from migra�on (INR) 127.8*** Annual cash and in-kind income from livestock (INR) -3.9 (Kochar et al. 2020) Impact evalua�on of the Na�onal Rural Livelihoods Project Par�cipants Total income (past 12 months) 19.23*** Agriculture income (past 12 months) 6.07 Livestock income (past 12 months) 26.25 MGNREGA earnings (past 12 months) 96.31*** Casual wage income (past 12 months) 23.15*** Enterprise income (past 12 months) 10.13 14 Households in early rollout JEEViKA villages are considered treated. APRIL 2022 | 27 Table A.3: Reported Labor Market Outcomes Sample group Outcome Effect (percentage point change) (Chris�an et al. 2019) Safety nets and natural disaster mi�ga�on: Evidence from cyclone Phailin in Odisha Village ITT 15 Number of days worked under the MGNREGA program 37.8 a Number of days of paid work under the MGNREGA program 50.7 a (Desai, Joshi 2014) Collec�ve ac�on and community development: Evidence from women’s self-help groups in rural India Par�cipants Whether woman is employed generally (causal laborer in -5.10 agriculture) Whether woman is employed (non-farm) over the past 3 months 8.1** Nonpar�cipants Whether woman is employed generally (causal laborer in -2.4 agriculture) Whether woman is employed (non-farm) over the past 3 months 3.9 Village ITT 16 Whether woman is employed generally (causal laborer in -0.2 agriculture) Whether woman is employed (non-farm) over the past 3 months 5.10* (Garikipa� 2008) The impact of lending to women on household vulnerability and women’s empowerment: Evidence from India Par�cipants Probability of work �me alloca�on being favorable -10.29* (Hoffmann et al. 2021) Relief from usury: Impact of a self-help group lending program in rural India Village ITT 17 Propor�on of women in the household who work for income 3.02*** Heterogeneity: Propor�on of women in the household who work for income -1.32 Scheduled Castes/Tribes (Holden et al. 2016) Evalua�on of Madhya Pradesh Safe Ci�es Ini�a�ve Par�cipants SHG Strengthening module: Currently working for pay -1 SHG Strengthening + VAW module: Currently working for pay 3.4 Life skills module for men and boys: Currently working for pay 0.4 SHG strengthening + life skills module: Currently working for pay -11** SHG strengthening + VAW+ life skills module: Currently working -11.4** for pay Nonpar�cipants SHG Strengthening module: Currently working for pay 0.4 SHG Strengthening + VAW module: Currently working for pay 2.7 Life skills module for men and boys: Currently working for pay -0.1 15 Households in TRIPTI villages are considered treated, while households in non-TRIPTI villages are considered not treated. 16 Any women living in SEWA villages are considered treated. 17 Households in treatment areas, compared to households in areas which will receive delayed treatment. 28 | SYSTEMATIC REVIEW ON WOMEN’S ECONOMIC EMPOWERMENT SHG Strengthening + VAW module: Currently working for pay 3.4 Life skills module for men and boys: Currently working for pay 0.4 SHG strengthening + life skills module: Currently working for pay -11** Sample group SHG strengthening + VAW+ life skills module: Currently working Outcome -11.4** E ffect (percentage for pay point change) Nonpar �cipants (Chris�an SHG et al. 2019) Safety Strengthening nets and natural module: disaster Currently mi�ga�on:working for pay Evidence 0.4 in Odisha from cyclone Phailin SHG Strengthening + VAW module: Currently working for pay 2.7 Village ITT 15 Number of days worked under the MGNREGA program 37.8 a Life skills module for men and boys: Currently working for pay -0.1 Number + lifework of days of paid SHG strengthening skills under the module: MGNREGA Currently program working 50.7 a for pay -9.5** (Desai, Joshi 2014) Collec�ve ac SHG �on and community strengthening + VAW+ development: Evidence life skills module: from Currently women’s 5.3 self-help groups in working rural India for pay Par �cipants (Joshi, Palaniswamy, Rao Whether woman 2019) Impact is evalua employed �on framework generally (causal and results: laborer Odisha in -5.10 project rural livelihoods agriculture) Village ITT Number of days worked on job card 309.4*** a Whether woman is employed (non-farm) over the past 3 months 8.1** a Number of days woman is paid on job card 140.2** Nonpar�cipants Whether woman is employed generally (causal laborer in -2.4 (Kandpal, Baylis 2019) The social lives of married women: Peer effects in female autonomy and investments in agriculture) children Whether woman is employed (non-farm) over the past 3 months 3.9 Par�cipants Whether woman works outside the household for pay 42.8* Village ITT 16 Whether woman is employed generally (causal laborer in -0.2 Par�cipants with Whether woman works outside the household for pay agriculture) 2.84 friends who also par�cipate in program Whether woman is employed (non-farm) over the past 3 months 5.10* (Garikipa The impact � 2008)with Nonpar�cipants of lending Whether woman to women works on household outside vulnerability the household for pay and women’s empowerment: 3.93** Evidence from friends who India par�cipate in program Par �cipants Probability of work �me alloca�on being favorable -10.29* (Kandpal, (Ho ffmannBaylis, Arends-Kuenning et al. 2021) Measuring 2013)Impact Relief from usury: the effect of a self-help of a group community-level lending program Indiaon women's program in rural empowerment outcomes: Evidence from India Village ITT 17 Propor�on of women in the household who work for income 3.02*** Par�cipants Whether woman owns a MGNREGA ID card 66.5*** Heterogeneity: Propor�on of women in the household who work for income -1.32 Village ITT Scheduled 18 Whether woman owns a MGNREGA ID card 31.3*** Castes/Tribes (Kochar et al. 2020) Impact evalua�on of the Na�onal Rural Livelihoods Project (Holden et al. 2016) Evalua Par�cipants Female Madhya �on oflabor Pradesh force par�cipaSafe �on Ci �es Ini rate, �a� ages ve 20 to 60 1.6 Par�cipants SHG Strengthening Female module: labor force par Currently �cipa� working ac on rate, primary for pay status �vity -1 -0.6 SHG Strengthening Female + VAW labor force par module: �cipa Currently �on rate, working secondary for ac�vity pay status 3.4 3.7* Life module skillshours Average for men in produc and �ve boys: work, Currently primary ac�working for pay vity status 0.4 -2.275 a SHG strengthening Average + life � hours in produc skills module: ve work, Currently secondary ac� working for pay vity status -11** a 5.701 SHG (Kumar et al. 2019) Social strengthening networks, + VAW+ mobility, life and poli skills �cal parmodule: Currently �cipa�on: working The poten �al for -11.4** women's self-help for pay groups to improve access and use of public en�tlement schemes in India Nonpar �cipants Par�cipants SHG Strengthening Whether module: the household Currently is aware working of the for pay MGNREGA scheme 0.4 5.6*** SHG Strengthening Whether + VAW the household module: has Currently used the MGNREGA working schemefor pay 2.7 4.10* Life skills (Kumar et al. 2021) The power of themodule collecfor �vemen and boys: empowers Currently women: working Evidence for self-help from pay -0.1 in India groups Par�cipants Works less than 10.5 hours per day (including �me spent on -2.4 childcare) 18 Women in program villages are considered treated. Number of hours worked (�me on primary ac�vity + half �me in 0.64 a childcare) APRIL 2022 | 29 Works less than 10.5 hours on primary ac�vity -5.3* (Kochar et al. 2020) Impact evalua�on of the Na�onal Rural Livelihoods Project Par�cipants Female labor force par�cipa�on rate, ages 20 to 60 1.6 Female labor force par�cipa�on rate, primary ac�vity status -0.6 Female labor force par�cipa�on rate, secondary ac�vity status 3.7* Table A.3: Reported Labor Market Outcomes Average hours in produc�ve work, primary ac�vity status -2.275 a Sample group Outcome Effect (percentage Average hours in produc�ve work, secondary ac�vity status 5.701 a point change) (Kumar (Chris�anetetal. 2019) al. Social 2019) networks, nets and mobility, Safety and poli natural disaster ��g mi cal a�par on:�cipa �on: The Evidence frompoten �al Phailin cyclone for women's self-help in Odisha groups to improve access and use of public en�tlement schemes in India Village ITT 15 Number of days worked under the MGNREGA program 37.8 a Par�cipants Whether the household is aware of the MGNREGA scheme 5.6*** Whetherof Number days the of paid work household under has used the the MGNREGA MGNREGA program scheme 50.7 a 4.10* (KumarJoshi (Desai, 2021) Collec et al.2014) �ve ac The power �on of theand community collec development: �ve empowers Evidence from women: Evidence fromself-help women’s self-help groups groups in in India rural India Par�cipants Works less than 10.5 hours per day (including �me spent on -2.4 Par�cipants childcare) woman is employed generally (causal laborer in -5.10 Whether agriculture) Number of hours worked (�me on primary ac�vity + half �me in 0.64 a Whether woman is employed (non-farm) over the past 3 months 8.1** childcare) Nonpar�cipants Works less Works Whether than than 10.5 lesswoman ishours 10.5 employed hours on primary on ac ac��vity vity(causal laborer in generally primary -5.3* -2.4 -5.3* agriculture) Number 1.85 a of hours worked (�me on primary ac�vity) (Pandey, Gupta, Gupta Whether woman 2019) Labor andis employed (non-farm) welfare impacts of over the past 3 months livelihoods3.9 a large-scale program: Quasi- Village ITT evidence from experimental16 India woman is employed generally (causal laborer in -0.2 Whether Village ITT 19 agriculture) Work par�cipa�on rate for female household members, self- 5.8*** employed livelihood Whether woman ac�vity (non-farm) over the past 3 months 5.10* is employed Work (Garikipa� 2008) The impact ofpar �cipa� lending onwomen to rate for onfemale household household members, vulnerability and self- 0.7*empowerment: women’s Evidence from India employed non Work par�cipa�on rate for female household members, self- 5.4*** Par�cipants of work Probabilityfarm employed �me alloca livelihood �on being favorable ac�vity -10.29* (Hoffmann et al. 2021) Relief Work from parusury: �cipa�on rate of Impact a self-help for group lending female household program members, in rural casual India 1.5* Village ITT 17 livelihood farm and non-farm ac � vity Propor�on of women in the household who work for income 3.02*** Heterogeneity: Work Propor par �on �cipa �on rate of women for female in the household household members, who work formal 0.7** for income -1.32 Scheduled salaried livelihood Castes/Tribes Work par�cipa�on rate for female household members, any 5.5* livelihood ac�vity (Holden et al. 2016) Evalua�on of Madhya Pradesh Safe Ci�es Ini�a�ve Number of livelihood ac�vi�es of female household members 38.5*** a Par�cipants SHG Strengthening module: Currently working for pay -1 (Prillaman 2016) Strength in numbers: How women’s networks close India’s poli�cal gender gap SHG Strengthening + VAW module: Currently working for pay 3.4 Par�cipants Employed in past year 3.7 Life skills module for men and boys: Currently working for pay 0.4 (Surendra 2020) Access to finance, empowerment, and women’s employment: Experimental evidence from rural SHG strengthening + life skills module: Currently working for pay -11** Bihar SHG strengthening + VAW+ life skills module: Currently working -11.4** Village ITT 20 Labor force par�cipa�on rate 2.45* for pay Agriculture labor force par�cipa�on rate 1.12 Nonpar�cipants SHG Strengthening module: Currently working for pay 0.4 Non-agriculture labor force par�cipa�on rate 0.85 SHG Strengthening + VAW module: Currently working for pay 2.7 Note: *p < 0.1, **p<0.05, ***p < 0.01. a Percent changes. Life skills module for men and boys: Currently working for pay -0.1 19 Households in program villages where take-up is at least 50 percent are considered treated. 20 Households in early rollout JEEViKA villages are considered treated. 30 | SYSTEMATIC REVIEW ON WOMEN’S ECONOMIC EMPOWERMENT