By Santadarshan Sadhu, IFMR Finance Foundation & Prof. Viswanath Pingali, IIM Ahmedabad
Microfinance has long been seen by many as a panacea for fighting poverty by providing access to financial services to low-income population considered unbankable by the traditional financial system. Over the past 20 years it has reached out to a large number of people across the world providing mostly access to credit. Its effect on improving household welfare have been debated with some studies finding positive effect, while some others finding no or even negative effects. One of the early evaluations of microfinance based on non-experimental design was carried out by researchers at the World Bank who, using a cross-sectional survey data from Bangladesh, found positive effect of microcredit program on household welfare1. However, these findings have been debated by other researchers on ground of methodological limitations (selection bias) of the evaluation. To avoid some of these methodological issues, many of the recent studies have used experimental approach (based on Randomized Control Trial) for evaluating microfinance programs. Some of these randomized evaluations assessed short-term program effect by investigating the impacts of microcredit program over a period of 18 months. Most of the randomized evaluations found no or limited evidence on the effect of microcredit2, and thus challenged the popular belief of effectiveness of microcredit as an anti-poverty tool.
However, these randomized evaluation are also criticized on the grounds that typical observation period is too short for assessing the benefit of microcredit. Proponents of microfinance argue that it can take years before such program can affect the productivity and business potential of low-income households, thereby influencing economic variables to generate observable differences in measures of family well-being. So, it is essential to assess the impact of such program over a much longer time period.
A recent study by Khandker and Samad has documented interesting results on the longer term effects of microcredit on household welfare. Utilizing a panel data of households which undertook three rounds of household survey (in 1991/92, 1998/99 and 2010/11) spanning over 20 years covering 87 villages of rural Bangladesh, this paper estimated long run effects of microcredit on household per capita income, expenditures, poverty, non-land assets, household net-worth, male and female labor supply, and schooling of children. Additionally, the paper also investigated whether multiple program membership and MFI competition result in adverse effects on household welfare.
The empirical analyses of the paper are based on 1,509 households from 1991/92 that are common in all three rounds of survey. In order to circumvent the issue of selection bias – a common problem of non-randomized evaluations – the authors used advanced econometric methods while estimating the effect of microcredit. The paper also identified separate estimates of the effect of borrowing by female and male on outcome variables and explored whether the effects of microcredit vary over time as well as whether past borrowing has a distinct effect from the current borrowing.
Results of this paper show that borrowing by men increases household per capita expenditure, men’s labor supply, non-land assets, and household net-worth. The estimates show that when borrowing by men increases by twofold (e.g. from TK. 5000 to TK. 10000) – household per capita expenditure increases by 0.4 percent, male labor supply by 1.8 percent, household non-land asset by 2.8 percent, and net-worth by 2 percent. On the other hand, when borrowing by women increases from 5000 to 10000 – monthly per capita income increases by 0.6 percent, male labor supply by 3.3 percent, female labor supply by 4.6 percent, household non-land asset by 2.5 percent, and boys’ and girls’ schooling enrollment rates by about 7 and 8 percentage points respectively. Borrowing by women also reduces the likelihood of being in extreme poverty. Thus, the results do indicate that borrowing from microcredit institutions helps improve economic conditions of the clients.
The comparison of effect of credit across different survey rounds suggests that the program effects change over time, and perhaps, decline for some outcomes over time. Similarly, the comparison of effect of current borrowing (over the 5 years before the current survey round) and past borrowing (over the 5 years preceding the past survey rounds) indicates that past borrowing may have effects independent of current borrowing. However, there is no uniform trend since such effects could be lingering (e.g., past male and female borrowing increases per-capita income per-capita expenditure, past female borrowing reduces likelihood of extreme poverty), diminishing (e.g., past male borrowing reduces non-land asset and net worth) or even non-existent depending on the outcome variables.
While exploring the effect of multiple borrowing (which is a prominent feature in Bangladesh and in other countries as well) on household welfare, the authors found that there is no negative effect of multiple borrowing when both current and past incidences of multiple borrowing are taken into account. So, this result indicates that multiple borrowing does not necessarily lead to reduction in household welfare and thus rejects the popular claim that adoption of multiple borrowing by MFI clients is harmful for the low-income population.
Finally, the paper estimated the effect of competition among the MFIs and found that current density of MFIs has a positive effect on household net-worth; though there is a negative wealth effect from past density of MFIs, which possibly hints on the diminishing economies of scale. The dynamic effects, however, indicate that households still benefit in terms of higher income from past incidence of higher village-level MFI density. This is possibly because, typically MFIs impose caps on loan size (credit demand is not met by single MFI). This necessitates clients to borrow from multiple sources in order to pool funds and gain economies of scale from income generating activities. However, the finding on diminishing economies of scale suggests that without complementary support (such as skill enhancement program and market linkage activities), access to microcredit alone may not be sufficient in achieving a sustained reduction in existing poverty levels.
Nonetheless, while one may agree or disagree with the findings here, this paper provides enough evidence to suggest that further research, both empirical and theoretical, is required to pass a conclusive judgment of microfinance, and a policy maker ought to take a note of this. After all, providing one more loan option to the poor cannot harm welfare, unless we assume borrower irrationality. The moot point is, can we really assume that?
1 – Khandker 1998; Fighting Poverty with Microcredit : Experience in Bangladesh, New York, NY: Oxford University Press.
Pitt and Khandker; 1998; The Impact of Group-based Credit Programs on Poor Households in Bangladesh: Does the Gender of Participants Matter?” Journal of Political Economy, 106 (June): 958-996.
Pitt and Khandker; 1996; “Household and Intrahousehold Impact of the Grameen Bank and Similar Targeted Credit Programs in Bangladesh.” World Bank Discussion Papers No. 320. Washington D.C.
2 – Studies include the following, among others: Karlan and Zinman 2010; de Mel, McKenzie and Woodruff 2008; Augsburg et al., 2012; Attanasio et al., 2011; Banerjee et al., 2010; Crepon et al., 2011; Karlan and Zinman 2011)