Group-based Microcredit & Emergent Inequality in Social Capital: Why Socio-religious Composition Matters

There is usually some form of joint liability within these groups. Typically, group members share the responsibility for moral persuasion or disciplining in the case of repayment delays and defaults.

A majority of microcredit programs are focused on women and are group-based, however, some are structured differently.

These studies claim to have a higher standard of methodological rigor on the basis of being able to accomplish randomized exogenous assignment of women to groups (rather than the “natural” existing process of women self-selecting into groups).

Some studies that have examined whether microcredit empowers women have found no effects (Banerjee et al. 2010) or have focused on the negative effects, even as they acknowledge that one section of women have benefited from these loans (Karim 2011). In such studies a lot depends on how empowerment is conceptualized and operationalized, i.e., what is being observed and measured. See Kabeer (2001) on the conflict over conceptualizing empowerment.

Social network scholars (McPherson et al. 2001) have found that religious homophily fosters close ties of giving emergency help, lending money, giving advice, and therapeutic counseling.

Throughout, “sanctioning” refers to group members disciplining culprits of violence or other offences against women that diminish women’s welfare. Therefore, sometimes I have used the term “protective sanctioning.” It does not refer to within-group punishments for defaulting. Also, these are third-party sanctioning, i.e., punishments conducted by group members who have themselves not been harmed. Due to the shared features of requiring cooperation and producing public goods (the deterrence), I have often used the term “collective action” broadly to include group-led protective sanctioning.

In one group, the Hindu group leader and some Hindu group members had once saved the group’s cashier, a Muslim woman, from potential violence in a situation of domestic conflict and had subtly sanctioned her husband. In the other group, the leader, a Muslim woman, and some Hindu group members had initiated collaboration with several other adjoining microcredit groups and had jointly undertaken an anti-liquor campaign that was successful. In addition, the Muslim group leader had single-handedly conducted more than one protective sanctioning on behalf of different village women.

In this paper I am not addressing caste-composition in microcredit groups. I did not gather explicit data on caste categories because it was not centrally important to my research question and because of ethical concerns regarding caste identification in in-depth interviews. For Hindus, the last names of the interviewees revealed an overwhelming majority of OBC and SC categories; there were negligibly few “general castes.” Muslims in rural West Bengal are largely non-Ashrafs (i.e., “lower caste” Hindus who converted to Islam), therefore, representing a significant degree of caste-homogeneity. These reasons form the basis of my decision to suspend caste as an analytical category and to theoretically and methodologically constrain myself to the Hindu-Muslim distinction. Also, I am bracketing away the few heterogeneous groups and focusing on a discussion of the homogeneous groups. For those interested in caste and microcredit groups, see Stuart (2007).

In addition to the works cited before, in development economics see- Keefer and Knack (1997); Dasgupta and Serageldin (2000); Grootaert and van Bastelaer (2002). For an excellent summary on the evolution of social capital as a field of study across disciplines and for an insightful discussion of the theoretical importance of microcredit groups for studying the formation of social capital, see Woolcock (1998, 2001).

Critics have pointed out that it is conceptually amorphous; there are a multiplicity of indicators used for measuring it; unexamined assumptions are implicit in the metaphor of capital (that instrumental rationality guides individual investment in developing social ties and that instrumental calculations play a causal role in the way those social ties later shape outcomes); and the one-sided focus on the felicitous outcomes of social ties resulting in the failure to consider the pitfalls of network embeddedness. See Portes (2000), Portes and Landolt (2000), Fishman (2004) and Lichterman (2006).

For an overview of the use of social capital in Sociology, see Portes (1998).

For example, some interviewees claimed that punishing wife-beaters had a positive externality for a wider set of women by deterring domestic violence.

Conceptualizing and coining the new term ‘participatory public goods’ is intended as a contribution to the extensive public goods literature, that has evolved over time to include other new conceptualizations such as “symbolic public goods” (Rao 2008).

A situation where each person could make short-run gains by not cooperating but everyone is better off in the long-run if they all cooperate.

This literature is based not on microcredit groups but on simulated work groups or groups in general.

Even demographically dominant groups may feel precarious due to geopolitical reasons.

On why women are seen as reproducers of collective identity, see Yuval-Davis (1980, 1997). In brief, she argues that it is because women make possible the social cohesion of the collective by reproducing it in the legitimate manner, and thus by reproducing its specific character (1980, 25).

Thus, Israeli family law is under the jurisdiction of religious Rabbinical courts that have as their legal code the orthodox Jewish Halacha (Halakha), which is of ancient origin, and is presided over by orthodox male rabbis who are appointed as judges. Although these judges are government appointed civil servants, no woman can become a judge. One of the qualifying criteria is rabbinical ordination, which no woman can attain due to being a woman, an inferior being in the eyes of the ancient legal code, and, hence, she is considered unfit for religious ordination (Yuval-Davis 1980). Likewise, the religious law prevents women from appearing as witnesses in Rabbinical courts. And even though religious councils are state funded bodies and secular laws do not prohibit women’s participation, women are prevented from participating in elections of chief and municipal rabbis, from sitting on religious councils, and from participating in elections to those councils.

Feldman (2001) cites a Bangladeshi NGO’s report on the spread of fatwas against women in rural Bangladesh and of harassment and attacks on NGOs offering training, education and other interventions targeted at women. Karim (2011), in her recent book on microcredit in Bangladesh, has also written about the attacks by organized forces from mosques and madrassahs on NGOs running microcredit programs and other income-generating projects for women in rural Bangladesh and direct attacks on the women participating in these programs. One can conclude from these that the spread of microcredit in Bangladesh has not been without a struggle, turning into a pitched battle in a few cases, between NGOs and local Islamic institutions.

SHG stands for “self-help group” and comes from the social finance literature. SHG-based microcredit programs in India are financed by a variety of institutions, including the government, by charitable foundations, and by multilateral development agencies. They usually have a hierarchically nested structure with multiple groups in a village forming a cluster and several clusters forming the regional or village node. The basic saving and lending functions are similar to group-based microcredit programs everywhere. Microcredit interest rates are known to vary by country. These programs had a 12–14 % interest on non-declining balance basis in 2004.

“A causal network is a display of the most important independent and dependent variables in a field study and of the relationship among them. The plot of these relationships is directional, rather than solely correlational. It is assumed that some factors exert an influence on others” (Miles and Huberman 1994, 153).

The all – India data presented in the Sachar Committee’s report (GOI 2006) (committee set up by the Prime Minister in 2005 to investigate the social, economic, and educational status of the Muslim community in India) shows that the self-employment categories of “own account worker/employer/unpaid family worker” covers 61 % of the total all-India Muslim workforce compared to approximately 55 % of total Hindu workforce. There is a 6 % difference. But disaggregating Hindus into the constitutionally codified caste-groups reveals a more complicated picture. The Sachar Committee data on the overall Indian rural sector reveals that for Hindu “scheduled castes” (SCs) and for “scheduled tribes” (STs) their reliance on casual work is far greater than any other community and also far greater than their ability to engage in self-employment. These two groups are followed by Muslims and Hindu OBCs who seem to be roughly equally reliant on self-employment and casual work. As such, in terms of economic security and vulnerability, households depending solely on the earnings from casual daily labor are worse off than households of the self-employed because of the greater precariousness of finding waged agricultural labor or other types of paid casual work in villages. Viewed from this perspective, in Indian village societies SCs and STs are likely to be the worst off economically.

Some scholars have argued that, compared to all-India statistics, W. Bengal has the lowest proportion of Muslim girls enrolled in middle and matriculate levels of school in both urban and rural areas, has a particularly high middle-school drop-out rate among Muslim girls, and displays the highest intercommunity disparity in the country (Hasan and Menon 2005). However, female seclusion is arguably a major reason behind this disparity as these scholars attribute this to the lack of madrassa and sex – segregated schools.

Here, by way of example, I cite a study to show that years of schooling may not have the identical impact on a particular dimension of behavior across the two communities in this context. In a study conducted by community medicine researchers (Mandal et al. 2007) on the impact of religious faith and female literacy on fertility rates in one particular Bengali village, the researchers found that there was a significant difference in education and average number of pregnancies between Hindu and Muslim women (mothers), but, although years of schooling had some impact on lowering fertility among Hindu mothers, it did not have the same effect among Muslim mothers.

Pugh (2013, 50–51) has provided a very useful elaboration on the four different kinds of information that are obtainable from in-depth interviews. She calls them the honorable, the schematic, the visceral, and meta-feelings.

For extensive discussion on reflexive analysis in qualitative research, see Aunger (2004).

A particularly powerful example is Bourdieu’s cross-community study of Kabyle Berber’s in Northern Algeria, the ethnic group that formed the basis of his theorizing about “habitus” (1977). Incidentally, my simultaneous study of Hindu and Muslim borrowers of peasant backgrounds comes close to Bourdieu’s parallel ethnographic studies of the Kabyle Berber’s and Béarnaise peasants of Southwestern France from his own childhood village and sharing his own ethnic and linguistic identities (Goodman and Silverstein 2009; Wacquant 2004).

Desai and Andrist (2010) report the mean age at marriage for women in West Bengal to be 17.51 years based on survey data collected in 2005.

The theoretical debate regarding patriarchy and purdah is covered in Papanek (1973), Papanek and Minault (1982), and Feldman (2001).

When this pact is broken, for example when husbands abandon their wives, rural women turn to working as domestic servants in households in the city (Ray and Qayum 2009).

This might change though if large-scale employment of some sort were to become available to these women, like employment in the garment-manufacturing sector that opened up to women in the eighties in neighboring Bangladesh. See Feldman (2001) for an astute analysis of patriarchy and change in Bangladesh.

Based on a recent quantitative study with survey data from 2005, Desai and Andrist (2010) find that, in India overall, most women marry between 14 and 25 years of age with a tight clustering around 17–19. They found that, compared to economic factors and economic and familial empowerment, gender performance, particularly the value placed on female seclusion (indicators include: practice of purdah/ ghunghat; men and women eat separately; and women’s physical immobility) are consistently related to age at marriage, such that greater emphasis on gender performance is associated with a lowered age at marriage for women. Refer to their study to see how West Bengal compares with the other Indian states on various gender dimensions.

There is some writing on Islam and patriarchy in Bangladesh, including Feldman (2001). Because of the shared history of West Bengal and Bangladesh prior to India’s partition on the eve of its independence in 1947, till which time they were both part of the undivided Bengal, it might be instructive to turn our attention to this literature. However, we must bear in mind that Bangladesh is a Muslim majority country and India and West Bengal are not, and this might moderate the articulation between Islam and patriarchy in the two places. See Ahmed (2008) for an informative discussion on Islam and masculinity among microcredit borrowers in Bangladesh.

See, Sanyal (2009, 2014) for more elaborate accounts and additional examples of group-led collective action.

The village council is a federally mandated political institution.

Here it is important to also remember that the Islamic injunction against usury has not come in the way of women’s enrollment in microcredit programs in Bangladesh.

The institutional perspective on gender, first proposed by Lorber (1994) and then subscribed by others, views gender as an “institutionalized system of social practices for constituting people as two significantly different categories, men and women, and organizing social relations of inequality on the basis of that difference” (Ridgeway and Correll 2004, 510; Ridgeway and Smith-Lovin 1999). The performative perspective suggests that gender is primarily constituted by everyday performances in the course of social interactions in which women and men are held accountable to expectation regarding female and male behavior, i.e., women and men interact in ways that are socially associated with being men and women and this creates and sustains inequality (West and Zimmerman 1987).

See Ahmed (2008) on Islam and masculinities among Grameen Bank borrowers.

A powerful example of religion’s influence on gender performance via practices of female seclusion can be found from comparing Punjab, which has a largely Sikh population (approximately 60 %), and Haryana, which has a largely Hindu population (approximately 88 %). Desai and Andrist (2010, 79–80) in their paper on gender scripts and age at marriage in India write, “Punjab has one of the highest ages at marriage in the country at almost 20 years old, and about 10 % of women in Punjab get married before reaching age 16… In contrast, the neighboring state of Haryana has an average age of marriage that is 2 years lower, and about 30 % of women get married by age 16. These states have similar educational and economic profiles, which is not surprising given that Haryana was carved out of Punjab. However, they differ in cultural traditions. Punjab contains a large number of Sikhs, and even Hindus are influenced by Sikh culture. Haryana shares cultural traditions with the central plains, so the two states differ substantially in the way gender is articulated. Purdah is the most visible difference: only 33 % of the women in Punjab practice purdah, while 80 % of the Haryana women do.”

The Hindu Code Bills comprise the Hindu Marriage Act, Succession Act, Minority and Guardianship Act, and Adoption and Maintenance Act, all of which were meant to revamp the traditional personal laws.

1961, is the last year in which the national census gathered data on this.

One could also argue that this was not surprising because one of the main intentions behind this revamp was to create a unified Hindu identity, which was seen as the first step of postcolonial nation building as the Hindus were the majority population in India. The Congress party, under the leadership of Jawaharlal Nehru, sought to achieve this unification of communal identity by eliminating the complex heterogeneity of personal laws followed by Hindus of different sects and regions. However, the stark contrast is in the fact that the aspiration of unifying the laws applicable to the Hindus and unifying Hindu communal identity did not make the Congress call for relying on Hindu religious texts, such as the Dharmasastras, as legal guides. Rather, Nehru, undertook a secular revision that some have argued was informed by the aspiration of modernizing Hindu society and facilitating gender equity. However, in the case of Muslims, the Shariat provided a readily available vehicle that unified communal identity. This prioritization of collective identity around religious communities and communal identity based notion of national unity and secularism is a distinctive feature of the Congress party. See Williams (2006) for further analysis of the politics surrounding personal laws in postcolonial India. Many scholars have also argued that political expediency was also a motivating factor in the Congress completely capitulating to the virulent opposition from within the Muslim community to the controversial Muslim Women (Protection of Rights on Divorce) Bill (1986), whereas, its response to the opposition to the Hindu Code Bills was dramatically different. Some scholars have called this political negligence (Basu 2003).

This contrasts with progressive legal reforms to Muslim personal laws in Tunisia and Morrocco.

Apart from West Bengal and two other states and a Union territory, Muslims are a smaller minority in the other states of the country.

On the theoretical rationale for considering religion a part of ethnic identity, see Varshney (2001).

This literature has been summarized, cited, and critiqued by Lichterman (2006).