Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2020 Jun 3.
Published in final edited form as: Dev Policy Rev. 2018 Apr 15;37(1):129–151. doi: 10.1111/dpr.12381

Delivering development? Evidence on self-help groups as development intermediaries in South Asia and Africa

Mary Kay Gugerty 1, Pierre Biscaye 1, C Leigh Anderson 1
PMCID: PMC7269175  NIHMSID: NIHMS1591295  PMID: 32494112

Abstract

Donors and governments increasingly seek to deliver development projects through community-based organizations such as self-help groups (SHGs), but little is known about the effectiveness of such arrangements. This article briefly summarizes hypotheses regarding the effectiveness of interventions using SHGs and presents the results of an evidence review on the impacts of interventions delivered through SHGs on health, finance, agriculture and empowerment outcomes in South Asia and sub-Saharan Africa. Though the impacts of SHG-based interventions are generally positive, the evidence base is limited and does not generally test whether alternative delivery mechanisms might be more effective.

Keywords: collective action, community-based organizations, mutual assistance, savings groups, self-help groups, women’s groups

1 |. INTRODUCTION

Self-help groups (SHGs) are a common form of community-based organization in many developing countries. Self-help groups are mutual assistance groups in which individuals undertake collective action with the goal of improving their own lives. In some contexts, delivering interventions through local SHGs may be a response to the hollowing out of the state (Hood, 1991; Milward & Provan, 2000). Such devolution is viewed by some authors as a means to enhance local ownership and control (Mansuri & Rao, 2013), and by others as a legitimization of state retrenchment under neoliberal regimes (Clarke, 2005; Fuller & Geddes, 2008). Governments, donors and non-profit organizations are increasingly delivering interventions through locally organized SHGs, in the belief that such institutional arrangements will enhance development outcomes, encourage sustainability and foster capacity in local civil society. In addition, SHGs are thought to provide economies of scale and scope, as the groups may be used to reach more people at a lower cost per person, and may also serve to deliver interventions that address multiple issues.

Using SHGs as platforms for development interventions has greatly expanded in South Asia and sub-Saharan Africa, although this growth follows different trajectories and scale in each region. In South Asia the spread of SHGs has been most notable in India, where non-governmental organizations (NGOs) began to promote village development savings groups (credit management groups) in the 1980s. In 1992, India’s National Bank for Agriculture and Rural Development (NABARD) launched its savings group linkage programme and developed a policy framework and capacity building programme for NGOs and SHGs to facilitate these linkages. By 2000, savings groups had become a central part of the Indian government’s efforts to mitigate poverty and promote rural livelihoods (Fernandez, 2006; IFAD, 2010). As of 2006, NABARD estimated that over 1,500,000 savings groups were in existence (Sinha et al., 2006).

In sub-Saharan Africa, SHGs build on longstanding forms of collective savings and labour, including rotating savings and credit associations (ROSCAs) that were widespread prior to NGO-led initiatives (Anderson & Baland, 2002; Gugerty, 2007). In recent years, many NGOs have created and disseminated similar collective savings models, such as village savings and loan associations (VSLAs) that build on the ROSCA foundation (Aniket, 2006; Odell, 2012). CARE, an international NGO, launched its first formal savings group programme in Niger in 1991, and several large NGOs, including Catholic Relief Services, Plan International, Oxfam, the Aga Khan Foundation, World Vision and Pact have since introduced savings group promotion programmes across the continent (Odell, 2012; Odell & Rippey, 2011). According to a 2011 report, the number of savings groups in Africa reported by seven NGOs across 35 countries totalled just under 200,000, reaching over 3.8 million people, however it is likely that this estimate considerably understates the true number of such groups (Odell, 2012). Some governments in sub-Saharan Africa seek to provide some support to SHGs through government extension agents. But, unlike the situation in India, most governments in these countries have not created explicit policy frameworks designed to link SHGs to financial institutions or public institutions (Sinha et al., 2006).

The original goal of many of these SHG programmes was to provide members, usually rural women, with access to credit to meet subsistence needs or invest in income-generating activities. Due to the flexibility of the savings group model, NGOs often integrate additional activities into saving group programmes, such as health, agriculture or business development (Fischer & Qaim, 2011; Greaney, Kaboski, & Van Leemput, 2013). Despite their popularity as a “platform” for the delivery of interventions, little theory exists on the mechanisms underlying SHGs, especially outside the area of microcredit, and few empirical reviews exist on whether and how implementing development projects through SHGs improves development outcomes such as health, income, access to financial services or empowerment.

We contribute to the literature through a review of 46 high-quality evaluations of development interventions in South Asia and sub-Saharan Africa implemented through SHGs that target health, financial, agricultural and empowerment outcomes. The article is organized as follows: we first offer a definition of SHGs and present some basic theoretical principles about the purpose and benefits of SHGs. Since many of these theories are not tested with existing evidence, we return to this theme at the end of the article. Next, we describe our methods for identifying the sample of studies to review, and briefly describe this evidence base. We then review evidence of the impacts of SHG-based interventions on outcomes in health, finance, agriculture and empowerment. We find that on balance SHG-based interventions are associated with positive impacts for group members, though the evidence for many outcomes is too limited to form conclusions and may under-represent null or negative effects. We conclude by proposing avenues for further inquiry, especially noting a need for additional studies testing theories of SHG effectiveness by comparing the impact of interventions when delivered through SHGs and through alternative mechanisms, as well as studies comparing the effectiveness of different forms of SHGs, such as those with and without external facilitation.

2 |. DEFINING SELF-HELP GROUPS

The term “self-help group” may refer to a wide variety of groups with different forms and institutional structures depending on the context, making these groups difficult to define. In western settings, for example, SHGs are often used to describe mental health or substance abuse support groups, whereas in India the term usually refers to financial co-operatives. A variety of community collective action groups, including peer groups, community support groups, mutual assistance groups, producer groups, asset-building communities, mutual aid groups, burial societies, savings groups and social insurance groups, could also be considered SHGs.

We define SHGs as mutual assistance organizations through which individuals undertake collective action with a primary goal of improving their own lives. Collective action implies that individuals share their time, labour, money or other assets with the group to produce both collective and individual benefits (Olson, 1965). Although such groups may provide positive social externalities and public goods (such as increased social capital or enhanced public health), such social benefits are not the primary motivation for individual participation. Private benefits to members—defined in whichever way members see fit—form the primary incentive for participation. In this way, SHGs are different from common pool resource management groups, and from local political organizations, such as village committees, that focus on providing public and common pool goods (Agarwal, 2001; Ostrom, 1990). Private benefits may not be the only motivation for or outcome of SHG participation. Many SHGs are not “single issue” organizations, and take on multiple functions and goals. For example, savings groups and women’s groups often form the basis for other kinds of collective activities (Brody et al., 2016; Greaney et al., 2013; Prost et al., 2013; Saha, Annear, & Pathak, 2013).

For the purposes of our review, we define SHGs as having the following characteristics:

  • a primary goal of improving individual member welfare

  • self-governance and member participation in decision-making

  • primary reliance on internally-raised resources which might include member contributions of time, labour, money, or other assets or knowledge

  • voluntary membership of individuals based on an affinity connection or common interest

  • regular face-to-face interactions among members (which tends to limit group size)

We can distinguish SHGs from other community-based groups on the first two dimensions above: member participation in governance, and “self-help” or a primary focus on the creation of private benefits for members. Figure 1 locates SHGs in relation to other kinds of community-based groups along these two dimensions with the level of member participation on the vertical axis and the importance of private benefits on the horizontal. Groups in the upper right-hand corner fall into our definition of SHGs with relatively higher member participation in group governance and with private benefits a primary reason for joining. Our review of SHGs excludes Grameen-style microfinance programmes for several reasons. Group-based microfinance initiatives are themselves an intervention that already include multiple components (such as training and education), whereas our goal is understanding the potential for delivering various interventions through SHGs as an alternative to other delivery mechanisms. In addition, one of our criteria is that the group rely substantially on internal resources, which microfinance groups often do not. Finally, microfinance is a well-studied intervention, with the possibility of crowding out our intent to examine SHGs in less-studied domains.

FIGURE 1.

FIGURE 1

Open in a new tab

Examples of Group-Based Approaches to Community-Based Development

Source: The authors.

3 |. EXISTING THEORY AND EVIDENCE ON SELF-HELP GROUPS

The theory and evidence on how group-based platforms can improve the effectiveness and efficiency of development interventions is surprisingly limited, given the prevalence of SHG-based interventions. In theory, implementing development interventions through SHGs could improve their cost-effectiveness, meaning lowering the cost of providing a target amount of goods and services, often through scale. Or delivery through SHGs could improve intervention efficiency, if the goal is maximizing benefits within a cost constraint.

A few studies examine whether providing development interventions through SHGs could potentially lower the per person costs of providing services through economies of scale. Manandhar et al. (2004) find that women’s group members reached an average of six other women with information on health strategies. Several studies suggest that individuals who participate in any group are more likely to be well connected and participate in multiple groups or community-based organizations, which may lead to spill over effects in the community (Anderson, Baland, & Moene, 2009; Davis & Negash, 2007). Some forms of groups may be self-replicating, with community members exposed to groups adopting practices on their own (Gillespie, 2003; Hargreaves et al., 2010).

Several studies of interventions that mobilized women’s groups to promote maternal and newborn health find that such group-based interventions are highly cost effective (Colbourn et al., 2013; Ensor et al., 2014; Lewycka et al., 2013; Manandhar et al., 2004; Prost et al., 2013; Roy et al., 2013; Tripathy et al., 2010), suggesting that provision through SHGs can provide economies of scale. Lewycka et al. (2013), however, report that the cost per year of life loss averted is greater for interventions delivered through women’s groups than interventions delivered through peer counsellors.

Group-based interventions might also provide economies of scope, including complementarities across investments or interventions, allowing providers to supply several services or products through one intervention. Several studies report that groups organized primarily to achieve health or finance outcomes are also leveraged to achieve additional outcomes (Colbourn et al., 2013; Dongre, Deshmuk, & Garg, 2007; Houweling et al., 2013; Roy et al., 2013). Few studies, however, explicitly compare the cost-effectiveness of group-based interventions to other intervention delivery mechanisms, or examine the effectiveness of specific combinations of interventions against others.

The literature also suggests that groups may generate additional benefits by providing peer pressure, commitment mechanisms or sanctions that encourage the adoption of new norms, practices or behaviours. Group participation is often hypothesized to support the development of individual and group capacity, empowerment and self-efficacy (Quisimbing & Pandofelli, 2009). Studies have found that women’s SHG participation supports greater control over household decision-making (Bhoj, Bardhan, & Kumar, 2013; Caro, Pangare, & Manfre, 2013; Desai & Joshi, 2012; Holvoet, 2005), wider participation in civil institutions and political processes (Deininger & Liu, 2009; Holvoet, 2005, and greater overall presence in society (Caro et al., 2013; Deininger & Liu, 2009; Desai & Joshi, 2012; Holvoet, 2005; Sinha et al., 2006; Tesoriero, 2006). The literature on ROSCAs suggests that group-based savings programmes create a saving commitment mechanism, helping members to stick to their savings plans and protect deposits from the claims of other family members (Gugerty, 2007; Kast, Meier, & Pomeranz, 2012).

Groups may also increase the validity and relevance of information by working through trusted peers to share experiences or provide demonstrations (Borkman, 1976; Chesler, 1991). Further, groups could act as mini-laboratories for experimentation and innovation and may be sources for the development of norms, trust and reciprocal relationships that support social capital (Putnam & Nanetti, 1993). Few studies of SHG-based interventions, however, test the mechanisms through which these groups achieve development outcomes by comparing impacts among groups with different characteristics or by comparing interventions delivered through groups versus through outreach to individuals via peers for formal institutions.

Group-based delivery mechanisms may also impede development objectives if flexibility and customization are compromised with “one size fits all” project design or delivery strategies that do not match important heterogeneity among and within groups (Anderson, Baland, & Moene, 2009; Gugerty & Kremer, 2008; Reddy & Manak, 2005; Sinha et al., 2006; Swain, 2012). Moreover, if interventions change the composition of groups as individuals seek access to benefits, such access may favour better-off individuals (Gugerty & Kremer, 2008). Differential access to groups may also exacerbate existing inequalities or tensions in a community (Mansuri & Rao, 2012) and programmes that target the poor but require self-selection into the programme may not be strongly pro-poor (Baird, McIntosh, & Özler, 2013).

The current literature does not provide clear comparative evidence on which types of interventions are most effective or on the underlying mechanisms associated with greater cost-effectiveness or efficacy. As a result, we focus on understanding whether SHG-based interventions are consistently associated with positive outcomes. We then discuss what evidence is needed to better understand the forms of SHG interventions that are likely to be effective and efficient and through what mechanisms.

4 |. METHODS AND SAMPLE

After establishing a definition of SHGs, we conducted a series of literature searches using seven academic databases (PAIS, EconLit, PubMed, the Cochrane Library, Web of Science, Scopus and Google Scholar) to identify rigorous studies of the impacts of interventions delivered through SHGs on targeted outcomes. The search period was restricted to the 10-year period 2004–2014.1 Articles were excluded if they did not meet our definition of a SHG, report on interventions delivered through these groups in either South Asia or sub-Saharan Africa or have empirical data on outcomes associated with these groups. Following the search of academic databases, we conducted a Google search for programme documents and other grey literature that provided information on interventions undertaken through SHGs. Grey literature was not included in the review of SHG impacts but did inform our understanding of SHG interventions. Following multiple searches, we narrowed our candidates for review to 210 published articles and programme documents, as summarized in Figure 2. The most common reason we excluded articles was that the intervention did not take place through SHGs as we defined them. We did exclude, however, a number of articles that were not available in English or as on-line full-text articles (many of which also appeared from the abstract to potentially not meet our screening criteria).

FIGURE 2.

FIGURE 2

Open in a new tab

Summary of Literature Screening

Note: Supplemental agriculture search string: (‘Self-Help Group’ OR ‘Self Help Group’ OR ‘Community Mobilization’ OR ‘Peer Group’ OR ‘Community Support Group’ OR ‘Mutual Assistance Group’ OR ‘Mutual Aid Group’ OR ‘Social Insurance Group’ OR ‘Savings Group’ OR ‘ROSCA’ OR ‘ASCA’ OR ‘VSLA’ OR ‘Water User Group’ OR ‘Forest User Group’ OR ‘Village Education Committee’ OR ‘Women’s Group’ OR ‘Farmer’s Group’ OR ‘Cooperative’ OR ‘co-op*’) AND (‘Developing Countr*’ OR ‘Low-Income Countr*’ OR Africa OR ‘list of individual countries in Sub-Saharan Africa and South Asia’) AND Agriculture

Source: The authors.

From this sample of 210 articles, we conducted further eligibility screening to prioritize published and peer-reviewed studies using high-quality experimental or quasi-experimental methods or that were otherwise attentive to high technical standards. Following this screening we selected 46 articles for review, each of which rated highly for relevance, evidence of impact and technical quality, and that also covered a wide variety of outcome areas and geographies in South Asia and sub-Saharan Africa. We attempted to maintain a balance between including high-quality empirical evidence and representativeness in our sample. Studies were considered local if they took place in one locality or district or two adjacent districts, regional if they involved multiple districts from one or more adjacent states or regions, national if they involved multiple states or regions dispersed throughout a country, or international if they covered interventions in multiple countries. In total the sample includes 16 local, 22 regional, four national, and four international studies. Of the 46 studies, 39 are published in peer-reviewed journals. The remaining seven include three drafts and working papers in prepublication, three programme evaluation documents and one thesis paper.

This sample of 46 studies is large in comparison to most meta-analyses or systematic reviews in this subject area, which commonly review fewer than 20 studies. Nonetheless, the sample is quite small in relation to the number of existing SHG-based interventions. Only a small fraction of SHG interventions have been studied, and an even smaller fraction studied in a systematic or rigorous manner. In addition, we found that within our regions of South Asia and sub-Saharan Africa, the literature was concentrated in particular countries. In South Asia, 24 of the 25 studies report results from India, with three reporting results from Bangladesh, two from Nepal, and one from Pakistan.2 In sub-Saharan Africa, 16 of the 22 studies report on interventions in Kenya, Malawi or Tanzania. We note that publication bias may favour positive results, though our review identified several instances of mixed, negative, or non-significant findings for particular outcomes. Our results represent the evidence supplied by a relatively small, but high quality, sample of studies of SHGs that possibly under-represent negative or null results.

We organized our review by the categories of outcomes targeted by SHG interventions, looking at the effectiveness of groups in promoting health, finance, agriculture and empowerment outcomes. Table 1 summarizes the evidence base for each of these four broad outcome areas. The majority of studies reviewed used an experimental or quasi-experimental methodology; the strength of evidence was highest for maternal, newborn and child health outcomes. Many studies report on more than one category of outcomes, with finance and empowerment outcomes the most common.

TABLE 1.

Evidence Base by Outcome Area—47 studies total

Studies Reporting on At Least 1 Indicator in this Outcome Area*
Outcome Area No. of studies Geographies covered** Scale of studies Methodology of studies
Maternal, Newborn, and Child Health (MNCH) 16 2 Bangladesh, 9 India, 4 Malawi, 2 Nepal, 1 Zambia 3 local, 10 regional, 1 national, 2 international 2 systematic review, 8 experimental, 2 quasi-experimental, 4 non-experimental
Reproductive Health and HIV 10 3 India, 2 Kenya, 1 South Africa, 2 Tanzania, 1 Zambia, 1 Zimbabwe 5 local, 4 regional, 1 national 3 experimental, 5 quasi-experimental, 2 non-experimental
Finance 23 1 Benin, 14 India, 5 Kenya, 1 Malawi, 1 South Africa, 1 Tanzania, 2 Uganda 10 local, 8 regional, 4 national, 1 international 4 experimental, 10 quasi-experimental, 9 non-experimental
Agriculture 10 4 India, 4 Kenya, 2 Tanzania, 2 Uganda 3 local, 3 regional, 3 national, 1 international 2 experimental, 5 quasi-experimental, 3 non-experimental
Empowerment 24 1 Benin, 14 India, 5 Kenya, 1 Malawi, 1 South Africa, 2 Tanzania, 1 Uganda 11 local, 8 regional, 3 national, 1 international 5 experimental, 11 quasi-experimental, 7 non-experimental
Open in a new tab
*

Note: Many studies report on several outcome areas.

**

Some studies were conducted in more than one country, so the “geographies covered” column does not always sum to the total number of studies.

We evaluate the findings for the most commonly reported outcome indicators in each outcome area according to two criteria: the strength of the evidence base and the evidence of impact. The strength of the evidence base for a given outcome is based on the number of studies reporting on that indicator, and the study methodology. A higher number of studies and a larger proportion of experimental or quasi-experimental methodologies indicate a stronger evidence base. Evidence on SHG impact is characterized as positive, mixed, negative or not significant (NS). A designation of “mixed” indicates that the results for a given outcome vary across models or sub-populations in the study. For each category of outcomes, we present a table outlining the evidence base and findings for the most common outcome indicators. We do not assign an aggregated rating for strength of the evidence base or evidence of impact in the tables for several reasons. Sample sizes for individual outcome areas are typically too small to disaggregate findings by study, types of groups or contexts. In addition, given the wide variety of indicators used for many outcomes, using typical meta-analysis techniques for aggregating results is not appropriate. Presenting disaggregated data also allows the reader to assess the relative strength of findings across areas. A full dataset of findings by study and outcome area is publicly available (Gugerty, Biscaye, Leigh Anderson, True, & Clark, 2014).

5 |. RESULTS

5.1 |. Maternal, newborn and child health (MNCH) outcomes

The outcome area of MNCH had the strongest evidence base and the most positive association between SHG delivery and outcomes, although the majority of the outcomes studied concerned the adoption of practices rather than morbidity or mortality outcomes. In our sample, 16 studies report on at least one MNCH practice or outcome.3 Most interventions were delivered through groups specifically organized for women’s health projects, however, in some cases the interventions were delivered through pre-existing savings groups. In most groups, a locally trained facilitator guided members through a process to discuss and identify key MNCH problems in the community, to select and implement relevant strategies to address them, and to assess the results. In seven of the studies, groups were trained using a common participatory learning and action (PLA) cycle; these are also the seven studies reviewed in Prost et al. (2013). Each group was encouraged to identify and implement its own combination of strategies, such as developing health education programmes, establishing vegetable gardens or purchasing bed nets to address health challenges (Colbourn et al., 2013; Prost et al., 2013). Some interventions sought also to address barriers to health care access by improving transportation (Ensor et al., 2014; Colbourn et al., 2013) or connecting groups to health care workers (Dongre et al., 2007; Lassi, Haider, & Bhutta, 2010). Only Alcock et al. (2009) and More et al. (2012) studied groups in an urban context, conducting different evaluations of the same intervention in an urban slum in Mumbai. Their results are inconclusive, raising questions about the effectiveness of SHG-based interventions in urban settings. Table 2 shows the main outcome areas and studies.

TABLE 2.

Evidence for Maternal, Newborn and Child Health Outcomes—16 Studies

Outcome No. of studies Strength of Evidence Findings
Improved maternal and newborn care practices at home 13 7 Experimental
2 Quasi-Experimental
2 Non-Experimental
2 Systematic Reviews		 11 Positive, 1 Mixed, 1 NS
Care-seeking for complications 11 5 Experimental
1 Quasi-Experimental
3 Non-Experimental
2 Systematic Reviews		 10 Positive, 1 Mixed
Institutional or skilled attendant birth delivery 10 5 Experimental
2 Quasi-Experimental
1 Non-Experimental
2 Systematic Reviews		 6 Positive, 4 NS
Infant/neonatal mortality and morbidity rates 10 8 Experimental
2 Systematic Reviews		 8 Positive, 2 NS
Maternal mortality and morbidity rates 6 4 Experimental
2 Systematic Reviews		 3 Positive, 3 NS
Open in a new tab

Improving maternal and newborn home care practices is often viewed as a critical component of improving health outcomes for women and children (Baqui et al., 2008; Bhutta et al., 2011; Fikree et al., 2005), and SHG-based interventions appear to positively impact health-seeking behaviour and practices. Of 13 studies,4 11 report positive effects on at least some maternal and newborn home care practices, such as duration of breast-feeding and wrapping of infants.

Care-seeking for complications also appears improved with SHG participation. Statistically significant positive differences between treatment and control groups in measures of care-seeking for complications are found in 11 studies,5 though four studies only measure awareness of danger signs and not actual care-seeking.6 Only Dongre, Deshmuk and Garg (2007) measure effects of women’s groups on both awareness of danger signs and care-seeking at healthcare facilities, and find positive effects on both measures. Lassi, Haider and Bhutta (2010) do not find significant impacts on healthcare seeking for maternal morbidities, but do find a positive impact on healthcare seeking for neonatal morbidities. The evidence on delivery practices (delivery in a facility or with a trained birth attendance) is more mixed, but six out of 10 studies report positive effects.7

Overall, improvements in care practices and delivery methods appear to translate into better child health outcomes. Significant decreases in neonatal and infant mortality rates among SHG members relative to non-members are reported in eight out of 10 studies.8 The evidence for effectiveness on maternal mortality outcomes is less clear. Of six studies reporting on maternal mortality,9 three report positive effects, including a meta-analysis of seven randomized controlled trials in South Asia that finds that exposure to women’s groups was associated with a 37% reduction in maternal mortality (Prost et al., 2013).10 The other three studies find no effects.

5.2 |. Reproductive health and HIV/AIDS

In our sample, 10 studies report on interventions targeting reproductive health and outcomes related to HIV/AIDS (see Table 3). These outcomes are closely related to maternal and child health outcomes, but these interventions target specific populations and tend to directly address cultural or social perceptions about sexual practices and family planning. The interventions studied were typically peer-mediated efforts to change behaviour by improving knowledge, attitudes and awareness of HIV, and to facilitate early screening and treatment for sexually transmitted infections (STIs). The results largely report a positive association between SHG participation and changes in knowledge and behaviour. Evidence on changes in health outcomes is rare; only one study reports on HIV transmission rates.

TABLE 3.

Evidence for Reproductive Health and HIV Outcomes—10 Studies

Outcome No. of studies Strength of Evidence Results
Knowledge and use of contraceptives 7 1 Experimental
4 Quasi-experimental
2 Non-experimental		 7 Positive
Risky sexual behaviour 5 2 Experimental
2 Quasi-experimental
1 Non-experimental		 4 Positive, 1 NS
Knowledge and use of family planning services and methods 2 1 Experimental
1 Quasi-experimental		 2 Positive
Transmission rates of HIV 1 1 Quasi-experimental 1 Positive
Adults seeking Voluntary Counselling and Testing (VCT) 3 2 Experimental
1 Non-experimental		 2 Positive, 1 NS
Open in a new tab

Three studies of peer-mediated behaviour change report on interventions that targeted female sex workers (Kuhlmann, Galavotti, Hastings, Narayanan, & Saggurti, 2013; Luchters et al., 2008; Odek et al., 2009). Two studies evaluate interventions that formed groups of adolescents (Carlson, Brennan, & Earls, 2012; Maro, Roberts, & Sorensen, 2009). The remaining studies describe interventions that used existing community groups as a platform to deliver peer education programmes.

Group interventions do appear to be effective for changing attitudes about and knowledge of reproductive health issues, including contraceptive use, risky sexual behaviour and knowledge of family planning services and methods, but evidence measuring actual behaviour change is limited. All seven studies reporting on knowledge and attitudes towards contraceptive use report positive effects as a result of SHG participation.11 Only two of these studies evaluate condom usage,12 but both find positive impacts of SHG participation. Four out of five studies report positive effects of SHG participation on attitudes towards risky sexual behaviour.13 Two experimental studies in India find significant positive impacts on knowledge and use of family planning services as well as methods beyond just contraceptive use.14

Evidence as to whether these attitude changes translate into health outcomes is scarce. We found limited evidence on the effectiveness of group-based interventions at combatting the spread of HIV/AIDS, though three of the four studies looking at transmission rates or the seeking of Voluntary Counselling and Testing (VCT) found positive impacts for group members.15

5.3 |. Savings and finance outcomes

We identified a relatively large evidence base for financial outcomes, although far fewer studies used experimental methods in this area (Table 4). Of the 23 studies, 19 evaluate interventions with savings groups whose primary goal was to support members’ savings groups and facilitate access to loans. In India, many interventions also involved creating linkages with formal financial services. In the other five studies reporting on financial outcomes,16 savings outcomes are reported as a byproduct of a health intervention. We note that while improved savings and access to credit are likely to have positive effects for individuals, few of these studies actually report on individual welfare. Changes in income are reported in 10 studies, but these are typically not changes in full household income or do not include all income sources. As a result, it is difficult to make assessments about the overall economic impact of SHG participation. None of the reviewed studies evaluate or otherwise report negative consequences from participation, though examples from some studies of microcredit interventions include increased debt, domestic violence or men using their wives to access credit (Karim, 2011; Schuler, Hashemi, & Badal, 1998). One systematic review of participation in economic empowerment initiatives on domestic violence finds mixed evidence; with higher levels of assets and education providing protective effects (Vyas & Watts, 2009).

TABLE 4.

Evidence on Savings and Financial Outcomes—23 Studies

Outcome No. of studies Strength of Evidence Notes
Savings 18 4 Experimental
8 Quasi-experimental
6 Non-experimental		 15 Positive, 2 Mixed, 1 Negative
Access to credit 15 2 Experimental
6 Quasi-experimental
7 Non-experimental		 12 Positive, 2 NS, 1 Mixed
Income 10 2 Experimental
6 Quasi-experimental
2 Non-experimental		 7 Positive, 3 NS
Micro-enterprise & business creation 7 1 Experimental
4 Quasi-experimental
2 Non-experimental		 7 Positive
Ownership of assets 9 1 Experimental
6 Quasi-experimental
2 Non-experimental		 8 Positive, 1 NS
Effect on the very poor within groups 6 1 Experimental
2 Quasi-experimental
3 Non-experimental		 2 Positive, 3 NS, 1 Negative
Open in a new tab

Of 18 studies analysing savings,17 15 report that savings rates increased for group members. This suggests that group-based saving helps participants either by providing a safe, accessible savings mechanism, perhaps by helping them commit to savings (Gugerty, 2007) or providing a means to resist social network demands (Dagnelie & LeMay-Boucher, 2008; Holvoet, 2005). The evidence does note the potential risk to members if peers default (Molyneux, Hutchison, Chuma, & Gilson, 2007), but the extent to which this affects individuals is not clear.

SHG participation is associated with increased access to credit, with 12 of 15 studies reporting significant positive effects on credit for group members.18 In India, a large-scale bank linkage intervention allowed groups demonstrating financial discipline to access loans from banks, which were usually significantly larger than the loans the groups could mobilize on their own (Deininger & Liu, 2009). Some participating groups, however, reported that loans were inadequate in size, took months to receive and required taking out additional loans in order to make repayments (Reddy & Manak, 2005).

The evidence suggests that SHG participation improved savings rates and access to credit that might facilitate investment in income-generating activities. Seven studies find that members of groups used funding to invest in small businesses either as a group19 or as individuals.20 Greaney et al. (2013) report that business investment and time spent on business activities was significantly greater for group members than for non-members. Seven studies report increased income for SHG members from farms or small enterprise.21 Three studies find no significant impact on income.22 Of studies reporting on the effect of group membership on asset ownership, seven out of eight have positive findings.23 Only Deininger and Liu (2009) do not find any significant impact on accumulation of assets over time, which they speculate was due to droughts and large crop failures at the time of the survey.

While the evidence on financial outcomes suggests positive effects on savings and access to credit and asset ownership, the evidence also suggests that SHGs may be less effective at reaching or changing outcomes for the very poor. Only two of the six studies reporting on impacts on the very poor report positive effects.24 Interventions in both of these studies included specific targeting mechanisms including social campaigns, in-kind contribution schemes and lending tailored to vulnerable populations. Molyneux et al. (2007) find that the poorest households and individuals are least likely to be reached through existing groups and that working only through existing groups may risk widening gaps between the poor and poorest individuals. Sinha et al. (2006) report that the barriers to group entry for poor individuals are high, and that moreover, of the women who have been a member of an Indian savings group for seven years or more, half are (still) poor, including 13% still classified as very poor. Overall, the association between SHG participation and financial outcomes appears positive, albeit with the rather large caveat that increased savings, assets and credit may not always translate into improved welfare for participants.

5.4 |. Agricultural outcomes

The impacts of SHG participation on agricultural outcomes are harder to assess: the outcome measures used (such as yield, technology adoption or income) vary widely and are often self-reported rather than objectively measured. Seven out of 10 studies25 report positive findings on a variety of agricultural outcomes, while the remaining three studies26 report mixed results. The sample for agricultural outcomes slightly favours sub-Saharan Africa, but all the African studies are located in three countries in East Africa: Kenya, Tanzania and Uganda. Either experimental or quasi-experimental designs are used by seven of the 10 studies. Five studies evaluate outcomes for farmer groups (primarily oriented around agricultural activities) and five for savings groups, and one study includes both types of groups. Table 5 presents the evidence base.

TABLE 5.

Evidence for Agriculture Outcomes—10 Studies

Outcome No. of studies Strength of Evidence Results
Technology adoption 3 1 Quasi-experimental
2 Non-experimental		 3 Positive
Access and use of inputs 5 2 Experimental
1 Quasi-experimental
2 Non-experimental		 4 Positive, 1 NS
Agricultural output and yields 5 1 Experimental
2 Quasi-experimental
2 Non-experimental		 3 Positive, 2 NS
Market involvement 4 2 Quasi-experimental
2 Non-experimental		 2 Positive, 2 Mixed
Farm income 5 3 Quasi-experimental
2 Non-experimental		 4 Positive, 1 Mixed
Open in a new tab

Three studies report positive effects of SHG participation on adoption of agricultural technology.27 Positive effects are reported by four out of five studies on access and use of inputs.28 Members negotiated for discounts on bulk purchases of pesticides and fertilizer (Caro et al., 2013) or made purchases through group credit (Sinha et al., 2006). Gugerty and Kremer (2008) note, however, that even providing groups with inputs directly does not guarantee that they will be used for activities which benefit the group as a whole, as they may be captured by particular members.

Further, assessing whether increased input use translates into increased productivity is difficult. Three of five studies report positive effects of SHG participation on agricultural productivity,29 but the measures of productivity used varied greatly and effect sizes were not included. Given the number of factors that might influence agricultural production and productivity, overall welfare effects of SHG participation are very challenging to assess.

SHG-based agricultural interventions more commonly target collective marketing to increase market access and sales rather than productivity gains. Two of four studies report positive outcomes in market involvement. Kaganzi et al. (2009) find that female SHG members were consistently able to better access high-value vegetable markets and Caro et al. (2013) find improvements in their ability to negotiate for better prices. Barham and Chitemi (2008), however, find that only 10 of the 19 farmer groups in their study report improved ability to satisfactorily market their production. Four studies30 report increased farm income, but none of these accounted for the cost of increased effort, though Kaganzi et al. (2009) do recognize these costs. One study finds higher sales prices received by SHG members, but again without accounting for marketing costs (Fisher & Qaim, 2011).

5.5 |. Empowerment

An important potential benefit of SHG participation is increased empowerment for participants, particularly for women. A relatively large number of the SHG studies we examined evaluate empowerment outcomes, but usually as a “side” benefit of the intervention rather than the primary intended outcome of the intervention. The technical evidence base in this area is weak; there is no study in which an “empowerment” intervention is randomly assigned. In addition, the consistency of empowerment measures varies widely, and empowerment is typically self-reported. Nonetheless, studies consistently show positive empowerment outcomes associated with SHG participation. SHG participation is associated with increases in self-confidence, perceptions of autonomy, knowledge of important issues, business training, negotiation skills, financial independence and mobility for members. The results on empowerment are heavily dominated by studies from India (14 of 24 studies) and by studies involving savings groups (16 out of 24) studies. Table 6 provides the evidence base for empowerment outcomes.

TABLE 6.

Evidence for Empowerment Outcomes—24 studies

Outcome No. of studies Strength of Evidence Results
Control over decision-making 13 2 Experimental
7 Quasi-experimental
4 Non-experimental		 13 Positive
Participation in other community groups or events 8 1 Experimental
3 Quasi-experimental
4 Non-experimental		 7 Positive, 1 Mixed
Participation in governance 10 3 Experimental
2 Quasi-experimental
5 Non-experimental		 9 Positive, 1 Mixed
Empowerment and self-efficacy 17 3 Experimental
7 Quasi-experimental
7 Non-experimental		 15 Positive, 2 Mixed
Open in a new tab

Positive impacts of group-based interventions on members’ control over decision-making in their households relative to non-members are found in 13 studies,31 with one study reporting that longer group membership resulted in a stronger impact (Holvoet, 2005). In general, studies of savings group interventions suggest that the savings commitment mechanisms reinforced women’s decision-making authority around savings and household finances, and four studies32 find increased participation by women in other domains of household decision-making.

SHG participation also appears associated with increased community participation and influence. Increased decision-making authority for women in community settings is found by four studies.33 Seven studies34 find that participation in groups gave women more opportunities to leave the house and become more engaged in the community, including participating in extra-household bargaining with community groups, improving community services and participating in other economic and social activities. All but one of the studies reporting on women’s participation in other activities are from India, so there is little evidence from outside this context.

Group membership also appeared to increase participation in community governance.35 Deininger and Liu (2009) report that female SHG members were more likely to attend village meetings. Sinha et al. (2006) find that in 25% of the Indian SHGs they studied a woman member ran for local political office (in the panchayat or village council), and that in 20% of Indian SHGs a woman member had been elected to office. Other studies report increases in members attending village council meetings, standing for election and participating in public decision-making bodies, but do not report magnitudes of impact of the interventions. Gugerty and Kremer (2008) report increased visits from government officials and extension agents to groups but did not report on participation in community governance by group members.

The evidence also suggests that SHG participation is associated with self-efficacy for women, with 15 studies reporting positive impacts on a variety of measures of empowerment and self-efficacy,36 including self-confidence, perceptions of autonomy, knowledge of important issues, business training, negotiation skills, financial independence and mobility. Increased financial independence in particular is highlighted in seven studies. However, two studies have mixed findings. Caro et al. (2013) note that, in spite of increased reported empowerment for female group members, division of labour within the household remained unchanged and women continued to bear a larger work burden. De Hoop, van Kempen, Linssen, and van Eerdewijk (2014) find that Indian SHG participation was associated with higher reported feelings of autonomy, but participation had a significant negative impact on subjective wellbeing in communities with more conservative gender norms, suggesting an important, but potentially consequential, unintended effect of SHG participation in such settings.

6 |. DISCUSSION

Our review found reasonably strong associations between SHG membership and improved MNCH practices in South Asia. Evidence is positive but more limited for maternal health and infant and child mortality and morbidity outcomes. Our results are consistent with reviews of other community-based health interventions, such as Lassi, Haider and Bhutta (2010) and Lassi, Kuman and Bhutta (2016), although these studies include interventions such as training and mobilizations of community health workers, in addition to group-based initiatives. These positive associations with MNCH outcomes may be the result of the relatively clearly specified and more easily measured behaviour changes targeted by these interventions, as well as by the large proportion of SHGs involved that were explicitly focused on health and formed specifically for the intervention. In addition, many of these interventions were accompanied by group facilitators, which may have strengthened programmes and further supported changes in health practices. One preliminary conclusion is that peer effects from group participation might be particularly helpful in supporting changes in norms and practices.

Participation in SHGs is positively associated with savings levels, access to credit and asset ownership, but the studies show no clear effects on income or overall welfare. Most of the SHGs in these studies had an explicit savings purpose, but none of the studies explicitly compares SHG-based savings programmes to individual savings programmes. SHG participation appears positively related to agricultural outcomes although the measurement of these agricultural outcomes remains too weak to draw strong conclusions. Only half of the studies reporting agricultural outcomes primarily evaluate farmers’ groups, raising the question of whether “layering” multiple activities onto SHGs formed for a specific goal is effective.

The evidence on the relationship between women’s participation in SHGs and empowerment is encouraging, but limited. Most studies are located in India and the measures used and methodology employed for evaluating empowerment outcomes are weak for the majority of studies. Few studies provide detail on the specific components of the intervention designed to promote empowerment, making it difficult to draw lessons for other programmes. While one study reports specific harm from SHG participation (de Hoop et al., 2014) many studies did not evaluate such potential negative effects. Theory and data collection about negative consequences should be a more explicit part of future studies.

The weak evidence base on the impact of participation in SHGs presents several challenges for generalizability and highlights a number of research gaps. First, static frameworks that enumerate the many potentially important factors predicting SHG success appear in the literature, but few of these frameworks distinguish between independent variables and co-variates in a way that makes testable hypotheses possible. A large number of factors affect individual behaviour change as well as health, agriculture, financial and empowerment outcomes. The challenge is to acknowledge this range of factors while crafting testable hypotheses and study designs that hold some of these factors constant—at least as a short-run model. Frameworks such as Michie’s behaviour wheel (2012) and Michie et al.’s taxonomy of behaviour change interventions (2013) provide helpful enumerations of key behaviour change factors to consider, but researchers will have to propose key drivers to test within these enumerations if the goal is to understand causality and propose intervention opportunities.

Second, clear and consistent definitions and measurement for many important outcomes still do not exist, particularly outside health. Outcome measurement is surprisingly uneven and context dependent, and the lack of consistency complicates the interpretation results across studies. Perhaps the trickiest measure is empowerment, considered variously an outcome, a mediating variable, or both. Empowerment might be the explicit goal of an intervention, or may be a hoped for benefit of another outcome, such as increased savings. In many of the studies reviewed, however, empowerment is an important mediating factor for behaviour change. For this specific measure, studies could more explicitly develop their model of empowerment, measuring at the beginning, middle and end of the study so that we better understand the role of empowerment as a pre-condition, a part of the intervention and an outcome.

Third, most of the evidence does not directly compare SHG-based to non-SHG-based interventions, or systematically assess the effectiveness of different intervention delivery mechanisms. Empirically, the most basic question of whether SHGs are effective cannot be answered without studies that compare SHG intervention delivery to non-SHG delivery, allowing us to assess whether it is the collective, rather than the individual, trait of SHGs that is driving outcomes. Studies could use random assignment to explicitly compare SHG-based interventions to other community mobilization strategies such as individual or peer-to-peer delivery, providing more systematic evidence on the effectiveness of the SHG platform. We found few studies that compared the effectiveness of different community mobilization models beyond three studies focused on MNCH outcomes (Colbourn et al., 2013; Lassi et al., 2010; Lewycka et al., 2013) as well as one meta-analysis (Lassi et al., 2016). These three studies find that SHG-based interventions are typically as effective as interventions delivered through health workers or clinics, but there may be additional benefits from interventions that utilize both approaches.37 For financial, agricultural and, particularly, empowerment outcomes, the evidence base could be greatly expanded by additional experimental or quasi-experimental studies. Many proponents of SHG-based approaches view the collective action and social capital creation that may result from groups as essential to behaviour change; from this perspective, running programmes without SHGs for comparison purposes makes little sense. Opportunities still exist, however, to run more systematic tests of alternative SHG programme effectiveness, as elaborated below.

Fourth, a more nuanced theory on the dynamics of SHG participation over time could help identify the most effective intervention points over the life of an SHG. Studies with a longer timeframe would also provide useful evidence about the dynamics of SHG participation. Very few studies followed group participants over time, making it difficult to assess whether positive outcomes in the short or medium run ultimately translate into longer-term improvements in participants’ lives. Repeated studies of the Ekjut trial in India (Houweling et al., 2013; Rath et al., 2010; Roy et al., 2013; Tripathy et al., 2010) and of the MaiMwana trial in Malawi (Lewycka et al., 2013; Rosato et al., 2006) indicate that positive impacts on group members persist, but the women’s groups in these interventions are formed around pregnant women and therefore do not have consistent membership. As a result, it is not clear whether positive impacts on group members are sustained after their participation in the groups.

Finally, future studies could better gather, test and report more detailed evidence on the specific characteristics of SHGs as well as on the nature of SHG interventions. Specific hypothesized predictors of group success could be tested with randomized treatment and control groups on those margins. More study designs that evaluate one type of group platform against another or that evaluate SHGs with different characteristics could greatly enhance the evidence base on the drivers of SHG effectiveness. We suggest an initial set of factors for investigation, grouping them into four categories: individual member characteristics and group composition; group structure, governance and funding; intervention characteristics; and group formation and duration.

Members’ individual characteristics such as age, gender, wealth, social status and occupation affect individual propensity to join and ability to take advantage of opportunities offered by groups. In addition, member heterogeneity within groups along these categories and pre-existing relationships and networks among members can influence group cohesion, dynamics and persistence. Future studies could move beyond average effects and gather better data on the characteristics of those who join (or are recruited) and the differential impacts of interventions on individuals and groups with different characteristics.

Group structure, governance and funding probably affect the ability of groups to engage in externally driven interventions, including whether the SHG has a history of mobilizing its own funds and resources or is externally supported. Tying development outcomes to group rules for electing leadership (as opposed to perhaps donor appointed leaders), meeting frequency, mechanisms for ensuring and sanctioning collective action through peer pressure, social sanctions, collective liability, etc. would increase our understanding of SHGs specifically and group behaviour more generally.

SHG intervention characteristics also appear to matter for outcomes. Some interventions provide an external facilitator; others train group members as facilitators. Interventions may provide training on group functioning and facilitation, others focus more narrowly on the intended outcome. Facilitation and training can be ongoing or one time; some approaches make use of particular frameworks, such as the participatory action learning cycle in India. Group linkages to public resources also vary. In India, many SHGs are linked to banks through government-supported programmes. In Africa, group formation may be a prerequisite for receiving credit or other donor assistance. Standardized and externally developed interventions could be compared to those in which groups themselves decide on what activities to pursue.

Group formation and duration also matter. SHGs may be formed as part of an intervention, or may be pre-existing. Some groups have relatively narrow and singular goals (e.g. maternal and child health), while others may have multiple concerns (savings, agricultural production and child nutrition). The extent to which group goals align directly with intervention goals may also matter. The results for MNCH and savings presented in this review suggest that groups formed for a specific purpose may be more successful in achieving this purpose. This implies that layering one intervention on top of a pre-existing group formed for another purpose may not be appropriate, but much more data are needed to assess this hypothesis. Groups may also vary in their intended duration. Savings groups typically operate on an ongoing basis, but allow member entry and exit at the end of each cycle. Other groups, such as those for pregnant women, may have a limited duration intended to support women through delivery and the early months of child-rearing, and may experience greater turnover of members.

Studying the comparative effectiveness of interventions and group characteristics presents a real challenge to researchers, requiring study designs that seek to hold certain elements of context, intervention design or group features constant. Such designs may require larger sample sizes to permit sufficient variation to test alternative delivery mechanisms. For example, studies can hold intervention design constant, and vary elements of group design, such as recruitment or member characteristics. Alternatively, groups can be formed in similar ways and in similar contexts, but elements of intervention design can vary. Though costly, this information would help policy-makers and implementers understand whether interventions that were successful in one setting are likely to translate into other settings, and to understand the mechanisms underlying successful programmes in particular contexts.

Overall our review offers some preliminary evidence to suggest that implementing development interventions through SHGs may have positive impacts on intended outcomes, but we argue that the evidence available to date is too limited to be sure. The knowledge base can usefully be built through studies that randomly assign SHG participation against treatments of interest so that effectiveness and efficiency can be compared to alternatives, and by studies that carefully describe the interventions’ theory of change and mode of implementation.

ACKNOWLEDGEMENTS

This work was supported by the Bill & Melinda Gates Foundation through a grant to the Evans School Policy Analysis & Research Group (EPAR). The findings and conclusions presented here are those of the authors and do not necessarily reflect positions or policies of the foundation. We recognize the valuable contributions of Christopher Clark and Zoe True in conducting this research.

Footnotes

1

The search strings used in our review are as follows: (‘Self-Help Group’ OR ‘Self Help Group’ OR ‘Community Mobilization’ OR ‘Peer Group’ OR ‘Community Support Group’ OR ‘Mutual Assistance Group’ OR ‘Mutual Aid Group’ OR ‘Social Insurance Group’ OR ‘Savings Group’ OR ‘ROSCA’ OR ‘ASCA’ OR ‘VSLA’ OR ‘Water User Group’ OR ‘Forest User Group’ OR ‘Village Education Committee’ OR ‘Women’s Group’) AND (‘Developing Countr*’ OR ‘Low-Income Countr*’ OR Africa OR ‘list of individual countries in Sub-Saharan Africa and South Asia’)

2

Three studies report on interventions across multiple countries.

13

Significant: Carlson et al., 2012; Hargreaves et al., 2010; Luchters et al., 2008; Maro, Roberts and Sorensen, 2009. Not significant: Fritz et al., 2011

14

Desai and Joshi, 2012; Saha, Annear and Pathak, 2013.

37

Colbourn et al. (2013), found that group interventions and improvements in health care facilities were equally effective in reducing neonatal and perinatal mortality rates, but find even greater reductions in clusters that received both interventions. Lassi et al. (2010) found that group-based interventions performed better than training health workers on reducing early neonatal mortality, but not on late neonatal or maternal mortality. Lewycka et al. (2013) found that perinatal, neonatal and infant mortality fell consistently over three years in areas given the women’s group intervention with or without volunteer peer counselling, and that the two intervention types had roughly similar impacts, varying slightly by outcome.

REFERENCES

  1. Agarwal B (2001). Participatory exclusions, community forestry, and gender: An analysis for South Asia and a conceptual framework. World Development, 29, 1623–1648. 10.1016/S0305-750X(01)00066-3 [DOI] [Google Scholar]
  2. Alcock GA, More NS, Patil S, Porel M, Vaidya L, & Osrin D (2009). Community-based health programmes: Role perceptions and experiences of female peer facilitators in Mumbai’s urban slums. Health Education Research, 24, 957–966. 10.1093/her/cyp038 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Anderson S, & Baland J-M (2002). The economics of roscas and intrahousehold resource allocation. Quarterly Journal of Economics, 117, 963–995. 10.1162/003355302760193931 [DOI] [Google Scholar]
  4. Anderson S, Baland J-M, & Moene KO (2009). Enforcement in informal saving groups. Journal of Development Economics, 90, 14–23. 10.1016/j.jdeveco.2008.09.003 [DOI] [Google Scholar]
  5. Aniket K (2006). Self Help Group Linkage Programme: A case-study. Retrieved from: http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.109.5416&rep=rep1&type=pdf
  6. Azad K, Barnett S, Banerjee B, Shaha S, Khan K, Rego, … Costello A. (2010). Effect of scaling up women’s groups on birth outcomes in three rural districts in Bangladesh: A cluster-randomised controlled trial. The Lancet, 375, 1193–1202. 10.1016/s0140-6736(10)60142-0 [DOI] [PubMed] [Google Scholar]
  7. Baird S, McIntosh C, & Ozler B (2013). The regressive demands of demand-driven development. Journal of Public Economics, 106(2013), 27–41. https://doi.Org/10.1016/j.jpubeco.2013.07.002 [Google Scholar]
  8. Baqui AH, El-Arifeen S, Darmstadt GL, Ahmed S, Williams EK, Seraji HR, … Syed U (2008). Effect of community-based newborn-care intervention package implemented through two service-delivery strategies in Sylhet district, Bangladesh: A cluster-randomised controlled trial. The Lancet, 371, 1936–1944. 10.1016/s0140-6736(08)60835-1 [DOI] [PubMed] [Google Scholar]
  9. Barham J, & Chitemi C (2009). Collective action initiatives to improve marketing performance: Lessons from farmer groups in Tanzania. Food policy, 34, 53–59. 10.1016/j.foodpol.2008.10.002 [DOI] [Google Scholar]
  10. Bhoj S, Bardhan D, & Kumar A (2013). Determinants and implications of rural women’s participation in microfinance programme: An analysis of dairy self-help groups in Uttarakhand State of India. Livestock Research for Rural Development, 25(10), Article 185 http://www.lrrd.org/lrrd25/10/bhoj25185.htm [Google Scholar]
  11. Bhutta ZA, Soofi S, Cousens S, Mohammad S, Memon ZA, Ali I, … Martines J (2011). Improvement of perinatal and newborn care in rural Pakistan through community-based strategies: A cluster-randomised effectiveness trial. The Lancet, 377(9763), 403–412. 10.1016/s0140-6736(10)62274-x [DOI] [PubMed] [Google Scholar]
  12. Borkman T (1976). Experiential knowledge: A new concept for the analysis of self-help groups. Social Service Review, 50, 445–456. 10.1086/643401 [DOI] [Google Scholar]
  13. Brody C, de Hoop T, Vojtkova M, Warnock R, Dunbar M, Murthy P, & Dworkin SL (2016). Economic self-help group programmes for improving women’s empowerment: A systematic review (3ie Systematic Review No. 23). Retrieved from International Initiative for Impact Evaluation (3ie) website: http://www.3ieimpact.org/en/evidence/systematic-reviews/details/236/ [Google Scholar]
  14. Carlson M, Brennan RT, & Earls F (2012). Enhancing adolescent self-efficacy and collective efficacy through public engagement around HIV/AIDS competence: A multilevel, cluster randomized-controlled trial. Social Science & Medicine, 75, 1078–1087. 10.1016/j.socscimed.2012.04.035 [DOI] [PubMed] [Google Scholar]
  15. Caro D, Pangare V, & Manfre C (2013). Gender impact assessment of the ASI Sunhara India Project. Retrieved from Cultural Practice website: http://www.culturalpractice.com/resources/gender-impact-assessment-of-the-asi-sunhara-india-project/
  16. Chesler MA (1991). Participatory action research with self-help groups: An alternative paradigm for inquiry and action. American Journal of Community Psychology, 19, 757–768. 10.1007/BF00938043 [DOI] [PubMed] [Google Scholar]
  17. Clarke J (2005). New Labour’s citizens: Activated, empowered, responsibilized, abandoned? Critical social policy, 25, 447–463. 10.1177/0261018305057024 [DOI] [Google Scholar]
  18. Colbourn T, Nambiar B, Bondo A, Makwenda C, Tsetekani E, Makonda-Ridley A, & Costello A (2013). Effects of quality improvement in health facilities and community mobilization through women’s groups on maternal, neonatal and perinatal mortality in three districts of Malawi: MaiKhanda, a cluster randomized controlled effectiveness trial. International health, 5, 180–195. 10.1093/inthealth/iht011 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Davis KE, & Negash M (2007). Gender, wealth, and participation in community groups in Meru Central District, Kenya (CaPRi Working Paper No. 65). Retrieved from http://www.ifpri.org/publication/gender-wealth-and-participation-community-groups-meru-central-district-kenya
  20. Dagnelie O, & LeMay-Boucher P (2008). ROSCA participation in Benin: A commitment issue (Unitat de Fonaments de l’Analisi Economica and Institut d’Analisi Economica Working Paper No. 735.08). Retrieved from http://pareto.uab.es/wp/2008/73508.pdf
  21. de Hoop T, van Kempen L, Linssen R, & van Eerdewijk A (2014). Women’s autonomy and subjective well-being: How gender norms shape the impact of self-help groups in Odisha, India. Feminist Economics, 20(3), 103–135. 10.1080/13545701.2014.893388 [DOI] [Google Scholar]
  22. Deininger K, & Liu Y (2009). Economic and social impacts of self-help groups in India (Policy Research Working Paper No. 4884). Retrieved from World Bank website: http://documents.worldbank.org/curated/en/675791468167351840/Economic-and-social-impacts-of-self-help-groups-in-India
  23. Desai RM, & Joshi S (2013). Collective action and community development: Evidence from self-help groups in rural India (Policy Research Working Paper No. WPS 6547). Retrieved from World Bank website: http://documents.world-bank.org/curated/en/178591468042542781/Collective-action-and-community-development-evidence-from-self-help-groups-in-rural-India
  24. Dongre AR, Deshmukh PR, & Garg BS (2009). A community based approach to improve health care seeking for newborn danger signs in rural Wardha, India. The Indian Journal of Pediatrics, 76, 45–50. 10.1007/s12098-009-0028-y [DOI] [PubMed] [Google Scholar]
  25. Ensor T, Green C, Quigley P, Badru AR, Kaluba D, & Kureya T (2014). Mobilizing communities to improve maternal health: Results of an intervention in rural Zambia. Bulletin of the World Health Organization, 92, 51–59. 10.2471/BLT.13.122721 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Fernandez. (2006). History and spread of the self-help affinity group movement in India: The role played by IFAD (IFAD Knowledge for Development Effectiveness Occasional Paper No. 3). Retrieved from https://www.microfinancegateway.org/library/history-and-spread-self-help-affinity-group-movement-india-role-played-ifad
  27. Fikree FF, Ali TS, Durocher JM, & Rahbar MH (2005). Newborn care practices in low socioeconomic settlements of Karachi, Pakistan. Social Science & Medicine, 60, 911–921. 10.1016/j.socscimed.2004.06.034 [DOI] [PubMed] [Google Scholar]
  28. Fischer E, & Qaim M (2012). Linking smallholders to markets: Determinants and impacts of farmer collective action in Kenya. World Development, 40, 1255–1268. 10.1016/j.worlddev.2011.11.018 [DOI] [Google Scholar]
  29. Fritz K, McFarland W, Wyrod R, Chasakara C, Makumbe K, Chirowodza A, & Woelk G (2011). Evaluation of a peer network-based sexual risk reduction intervention for men in beer halls in Zimbabwe: Results from a randomized controlled trial. AIDS and Behavior, 15, 1732–1744. 10.1007/s10461-011-9922-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Fuller C, & Geddes M (2008). Urban governance under neoliberalism: New Labour and the restructuring of state-space. Antipode, 40, 252–282. 10.1111/j.1467-8330.2008.00591.x [DOI] [Google Scholar]
  31. Gillespie S (2003). Scaling up community-driven development: A synthesis of experience (World Bank Social Development Paper No. 69). Retrieved from World Bank website: http://documents.worldbank.org/curated/en/138591468321296730/Scaling-up-community-driven-development-a-synthesis-of-experience [Google Scholar]
  32. Greaney B, Kaboski JP, & Van Leemput E (2013). Can Self-Help Groups really be “self-help”? (NBER Working Papers No. 18970). Retrieved from National Bureau of Economic Research website: http://www.nber.org/papers/w18970 [Google Scholar]
  33. Gugerty MK (2007). You can’t save alone: Commitment in rotating savings and credit associations in Kenya.Economic Development and Cultural Change, 55, 251–282. 10.1086/508716 [DOI] [Google Scholar]
  34. Gugerty MK, & Kremer M (2008). Outside funding and the dynamics of participation in community associations. American Journal of Political Science, 52, 585–602. 10.1111/j.1540-5907.2008.00331.x [DOI] [Google Scholar]
  35. Gugerty MK, Biscaye P, Leigh Anderson C, True Z, & Clark C (2014). Self-help groups in development: A review of evidence from South Asia and sub-Saharan Africa (EPAR Technical Report No. 283). Retrieved from Evans School Policy Analysis & Research Group website: https://evans.uw.edu/policy-impact/epar/research/self-help-groups-development-review-evidence-south-asia-and-sub-saharan [Google Scholar]
  36. Hargreaves J, Hatcher A, Strange V, Phetla G, Busza J, Kim J, & Bonell C (2009). Process evaluation of the Intervention with Microfinance for AIDS and Gender Equity (IMAGE) in rural South Africa. Health Education Research, 25, 27–40. 10.1093/her/cyp054 [DOI] [PubMed] [Google Scholar]
  37. Holvoet N (2005). The impact of microfinance on decision-making agency: Evidence from South India. Development and Change, 36, 75–102. 10.1111/j.0012-155X.2005.00403.x [DOI] [Google Scholar]
  38. Hood C (1991). A public management for all seasons? Public administration, 69, 3–19. 10.1111/j.1467-9299.1991.tb00779.x [DOI] [Google Scholar]
  39. Houweling TA, Tripathy P, Nair N, Rath S, Rath S, Gope R, & Prost A (2013). The equity impact of participatory women’s groups to reduce neonatal mortality in India: Secondary analysis of a cluster-randomised trial. International Journal of Epidemiology, 42, 520–532. 10.1093/ije/dyt012 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. IFAD. (2010, March). Empowering women through self-help groups: India country programme evaluation 2009 (Evaluation Insights No. 13). Retrieved from International Fund for Agricultural Development website: https://www.ifad.org/documents/38714182/39716016/india-13.pdf [Google Scholar]
  41. Kaganzi E, Ferris S, Barham J, Abenakyo A, Sanginga P, & Njuki J (2009). Sustaining linkages to high value markets through collective action in Uganda. Food Policy, 34, 23–30. 10.1016/j.foodpol.2008.10.004 [DOI] [Google Scholar]
  42. Karim L (2011). Microfinance and its discontents: Women in debt in Bangladesh. Minnesota, MN: University of Minnesota Press. [Google Scholar]
  43. Kast F, Meier S, & Pomeranz D (2012). Under-savers anonymous: Evidence on self-help groups and peer pressure as a savings commitment device (NBER Working Paper No. 18417). Retrieved from National Bureau of Economic Research website: http://www.nber.org/papers/w18417 [Google Scholar]
  44. Kuhlmann AS, Galavotti C, Hastings P, Narayanan P, & Saggurti N (2014). Investing in communities: Evaluating the added value of community mobilization on HIV prevention outcomes among FSWs in India. AIDS and Behavior, 18, 752–766. 10.1007/s10461-013-0626-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Lassi ZS, Kumar R, & Bhutta ZA (2016). Community-based care to improve maternal, newborn, and child health In Black R, Laxminarayan R, Temmerman M & Walker N (Eds.), Reproductive, maternal, newborn and child health (pp. 263–284). Washington, DC: World Bank. [PubMed] [Google Scholar]
  46. Lassi ZS, Haider BA, & Bhutta ZA (2010). Community-based intervention packages for reducing maternal and neonatal morbidity and mortality and improving neonatal outcomes. Cochrane Database of Systematic Reviews, 11 10.1002/14651858.cd007754.pub2 [DOI] [PubMed] [Google Scholar]
  47. Lewycka S, Mwansambo C, Rosato M, Kazembe P, Phiri T, Mganga A, & Costello A (2013). Effect of women’s groups and volunteer peer counselling on rates of mortality, morbidity, and health behaviours in mothers and children in rural Malawi (MaiMwana): A factorial, cluster-randomised controlled trial. The Lancet, 381, 1721–1735. 10.1016/S0140-6736(12)61959-X [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Luchters S, Chersich MF, Rinyiru A, Barasa MS, King’ola N, Mandaliya K, & Temmerman M (2008). Impact of five years of peer-mediated interventions on sexual behavior and sexually transmitted infections among female sex workers in Mombasa, Kenya. BMC Public Health, 8, 143 10.1186/1471-2458-8-143 [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Manandhar DS, Osrin D, Shrestha BP, Mesko N, Morrison J, Tumbahangphe KM, & de L Costello AM (2004). Effect of a participatory intervention with women’s groups on birth outcomes in Nepal: Cluster-randomised controlled trial. The Lancet, 364, 970–979. 10.1016/s0140-6736(04)17021-9 [DOI] [PubMed] [Google Scholar]
  50. Mansuri G,& Rao V (2012). Localizingdevelopment:Doesparticipationwork? Retrievedfrom WorldBankwebsite: http://documents.worldbank.org/curated/en/461701468150309553/Localizing-development-does-participation-work [Google Scholar]
  51. Maro CN, Roberts GC, & Sorensen M (2009). Using sport to promote HIV/AIDS education for at-risk youths: An intervention using peer coaches in football. Scandinavian Journal of Medicine and Science in Sports, 19, 129–141. 10.1111/j.1600-0838.2007.00744.x [DOI] [PubMed] [Google Scholar]
  52. Michie S (2012). The behaviour change wheel: A method for designing effective interventions. Presentation at the Implementation Science Summer School, Dublin: Retrieved from https://pdfs.semanticscholar.org/presentation/4568/e7b17682ee35238a0736791803a269c5dd87.pdf [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Michie S, Richardson M, Johnston M, Abraham C, Francis J, Hardeman, … Wood CE (2013). The behavior change technique taxonomy (v1) of 93 hierarchically clustered techniques: Building an international consensus for the reporting of behavior change interventions. Annals of Behavioral Medicine, 46, 81–95. 10.1007/s12160-013-9486-6 [DOI] [PubMed] [Google Scholar]
  54. Milward HB, & Provan KG (2000). Governing the hollow state. Journal of Public Administration Research and Theory, 10, 359–380. 10.1093/oxfordjournals.jpart.a024273 [DOI] [Google Scholar]
  55. Molyneux C, Hutchison B, Chuma J, & Gilson L (2007). The role of community-based organizations in household ability to pay for health care in Kilifi District, Kenya. Health Policy and Planning, 22, 381–392. 10.1093/heapol/czm031 [DOI] [PubMed] [Google Scholar]
  56. More NS, Bapat U, Das S, Alcock G, Patil S, Porel M, & Osrin D (2012). Community mobilization in Mumbai slums to improve perinatal care and outcomes: A cluster randomized controlled trial. PLoS medicine, 9, e1001257 10.1371/journal.pmed.1001257 [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Odek WO, Busza J, Morris CN, Cleland J, Ngugi EN, & Ferguson AG (2009). Effects of micro-enterprise services on HIV risk behaviour among female sex workers in Kenya’s urban slums. AIDS and Behavior, 13, 449–461. 10.1007/s10461-008-9485-y [DOI] [PubMed] [Google Scholar]
  58. Odell M (2012). Micro-finance in Africa: State-of-the-Sector Report 2011 Closing the gap. Retrieved from the CARE website: https://www.care.org/sites/default/files/documents/MF-2011-CARE-Access-Africa-Closing-the-Gap.pdf [Google Scholar]
  59. Odell M, & Rippey P (2011). Beyond financial services: The permanence and value of savings groups in CARE Kenya’s COSAMO Programme. Retrieved from https://www.mangotree.org/Resource/Beyond-Financial-Services-The-Permanence-and-Value-of-Savings-Groups-in-CAREs-COSAMO-Programme
  60. Olson M (1965). The logic of collective action: Public goods and the theory of groups (Harvard Economic Studies vol. 124). Cambridge, MA: Harvard University Press. [Google Scholar]
  61. Ostrom E (1990). Governing the commons: The evolution of institutions for collective action. Cambridge: Cambridge University Press. [Google Scholar]
  62. Place F, Kariuki G, Wangila J, Kristjanson P, Makauki A, & Ndubi J (2004). Calliandra groups. Assessing the factors underlying differences in achievements of farmer groups: Methodological issues and empirical findings from the highlands of Central Kenya. Agricultural Systems, 82, 257–272. 10.1016/j.agsy.2004.07.001 [DOI] [Google Scholar]
  63. Prost A, Colbourn T, Seward N, Azad K, Coomarasamy A, Copas A, … Costello A (2013). Women’s groups practicing participatory learning and action to improve maternal and newborn health in low-resource settings: A systematic review and meta-analysis. The Lancet, 381, 1736–1746. 10.1016/S0140-6736(13)60685-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
  64. Putnam R, & Nanetti R (1993). Making democracy work: Civic traditions in modern Italy. Princeton, N.J.: Princeton University Press. [Google Scholar]
  65. Quisimbing A, & Pandofelli L (2009). Promising approaches to address the needs of poor female farmers: Resources, constraints, and interventions. World Development, 38, 581–592. 10.1016/j.worlddev.2009.10.006 [DOI] [Google Scholar]
  66. Rath S, Nair N, Tripathy PK, Barnett S, Rath S, Mahapatra R, & Prost A (2010). Explaining the impact of a women’s group led community mobilisation intervention on maternal and newborn health outcomes: The Ekjut trial process evaluation. BMC International Health and Human Rights, 10, 25 10.1186/1472-698X-10-25 [DOI] [PMC free article] [PubMed] [Google Scholar]
  67. Reddy CS, & Manak S (2005). Self-help groups: A keystone of microfinance in India: Women empowerment and social security. Retrieved from https://www.microfinancegateway.org/library/self-help-groups-keystone-microfinance-india-women-empowerment-social-security
  68. Rosato M, Mwansambo CW, Kazembe PN, Phiri T, Soko QS, Lewycka S, & de L Costello AM (2006). Women’s groups’ perceptions of maternal health issues in rural Malawi. The Lancet, 368, 1180–1188. 10.1016/s0140-6736(06)69475-0 [DOI] [PubMed] [Google Scholar]
  69. Roy SS, Mahapatra R, Rath S, Bajpai A, Singh V, Rath S, & Prost A (2013). Improved neonatal survival after participatory learning and action with women’s groups: A prospective study in rural eastern India. Bulletin of the World Health Organization, 91, 426–433B. 10.2471/BLT.12.105171 [DOI] [PMC free article] [PubMed] [Google Scholar]
  70. Saha S, Annear PL, & Pathak S (2013). The effect of Self-Help Groups on access to maternal health services: Evidence from rural India. International Journal for Equity in Health, 12 10.1186/1475-9276-12-36 [DOI] [PMC free article] [PubMed] [Google Scholar]
  71. Schuler SR, Hashemi SM, & Badal SH (1998). Men’s violence against women in rural Bangladesh: Undermined or exacerbated by microcredit programmes? Development in Practice, 8, 148–157. 10.1080/09614529853774 [DOI] [PubMed] [Google Scholar]
  72. Sinha F, Tankha A, Brar A, Tirath N, Varma S, Mishra KN, … Subhasini S (2006). Self help groups in India: A study of the lights and shades. Retrieved from https://www.microfinancegateway.org/library/self-help-groups-india-study-lights-and-shades
  73. Swain R (2012). Differential impact of microfinance delivery mechanism on vulnerability. Applied Economics Letters, 19, 721–724. 10.1080/13504851.2011.597716 [DOI] [Google Scholar]
  74. Swain RB, & Varghese A (2009). Does self help group participation lead to asset creation? World Development, 37, 1674–1682. 10.1016/j.worlddev.2009.03.006 [DOI] [Google Scholar]
  75. Tesoriero F (2006). Strengthening communities through women’s self help groups in South India. Community Development Journal, 41, 321–333. 10.1093/cdj/bsi066 [DOI] [Google Scholar]
  76. Tripathy P, Nair N, Barnett S, Mahapatra R, Borghi J, Rath S, … Costello A (2010). Effect of a participatory intervention with women’s groups on birth outcomes and maternal depression in Jharkhand and Orissa, India: A cluster-randomised controlled trial. The Lancet, 375, 1182–1192. 10.1016/S0140-6736(09)62042-0 [DOI] [PubMed] [Google Scholar]
  77. Van Rompay KK, Madhivanan P, Rafiq M, Krupp K, Chakrapani V, & Selvam D (2008). Empowering the people: Development of an HIV peer education model for low literacy rural communities in India. Human Resources for Health, 6, 6 10.1186/1478-4491-6-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
  78. Vyas S, & Watts C (2009). How does economic empowerment affect women’s risk of intimate partner violence in low and middle income countries? A systematic review of published evidence. Journal of international Development, 21, 577–602. 10.1002/jid.1500 [DOI] [Google Scholar]
  79. Wenner MD (1995). Group credit: A means to improve information transfer and loan repayment performance. The Journal of Development Studies, 32, 263–281. 10.1080/00220389508422414 [DOI] [Google Scholar]

RESOURCES