Responsible Investment

Microfinance at the margin

October 31, 2012

Image ‘Bosnia-27231700016 World Bank’ by World Bank Photo Collection (CC BY-NC-ND 2.0)

The goal of the experiment was to quantify the causal impact of access to credit on small-scale entrepreneurship and poverty alleviation among those who would normally not qualify for microcredit from our partner institution. As part of the experiment, EKI — a large Bosnian microfinance institution — provided a random selection of so-called ‘marginal’ applicants with small loans. Loan officers across all branches were asked to identify potential marginal clients over a period of several months. In total 1,198 of such marginal loan applicants were gradually identified and subsequently interviewed as part of a detailed baseline survey.

All of these marginal applicants would normally have been rejected, for instance because they lacked sufficient collateral. However, during the experiment the research team in London would allocate newly interviewed applicants randomly with a 50 per cent probability to either the treatment (receiving a loan) or the control group (no loan). Successful applicants received a loan within a week. In contrast, applicants that were allocated to the control group did not receive a loan from EKI for the duration of the study (Figure 1). Fourteen months after the baseline survey, all participants — both those who received a loan and those who did not — were approached to be re-interviewed.

Causal impact

In all, this strategy allowed us to compare how the treatment and control groups developed over time and to clearly identify the causal impact of access to the credit from EKI on various entrepreneurship and poverty measures. In our study[1] we find that access caused increased levels of business activity and more self-employment. Credit availability (partially) relaxed the liquidity constraints of the treatment group and had a positive impact on business creation and business survival. About 14 months after the start of the program, marginal borrowers were 6 per cent more likely to own an enterprise compared with the control group. Borrowers with higher education levels mainly started businesses in the services sector whereas the less educated established small-scale agricultural activities. We find, however, that business creation did not translate into increased profits or higher household income in the 14 months of our observation period.

Second, we observe that those borrowing households that already had a business and those that were highly educated ran down their savings (compared to the control group) while less-educated households reduced their consumption expenditures. These findings are consistent with investments being lumpy so that households needed to crowd in additional resources to make up the difference and to implement investments that would have been unattainable without the loan.

A third key finding of our study is that households of marginal clients with low education levels reduced the school attendance of their teenagers (aged 16-19) and let them work more in the household’s business instead. On average these children work 35 hours per week more in this business compared with the control group and, not surprisingly given the substantial increase in working hours, are 19 per cent less likely to attend school. Yet, we do not find much evidence that the small-scale and often agricultural activities of lower-educated families will generate positive revenues that more than offset the potential loss in future income due to children’s lower human capital. Such unintended effects of access to microcredit need to be interpreted carefully. On the one hand, these young adults may be prevented (say through funding restrictions) from attending school by their families who feel internal labor is cheaper and who may not fully take into account the benefits of education that will accrue to the youth. On the other hand, if returns to education are very low, the new home business may provide an opportunity and working there may be a more efficient way of allocating time.

Mixed picture

Overall, our findings paint a mixed picture of the impact of microcredit. On the one hand, households did use the loans to start up new businesses, to keep existing ones afloat during the 2008-09 crisis, or to expand them. Where necessary they even cut back on consumption and used their savings to make sufficiently large investments. On the other hand, we do not find that these entrepreneurial activities had a positive impact on income. Even for households that already had an enterprise at the time of the baseline survey, and for whom the model in our paper would predict an increase in consumption, we do not find such a positive impact.

Moreover, we document that the program was not profitable and resulted in an implicit subsidy to the average marginal borrower of US$ 268. This reflects that the new marginal client group was significantly more risky than first-time regular clients. In particular, late payment was 1.5 times as high among marginal clients compared with regular first-time clients while in the end non-repayment was even three times as high.

There are various possible reasons why we do not find evidence of a positive impact of microcredit on enterprise profits, household income, or consumption, notwithstanding an increase in entrepreneurial activity. First, the period between our baseline and follow-up surveys — about 14 months — may have been too short to allow households to fully implement investments and increase firm profitability. Households that cut back consumption when they received a loan will have done so in the expectation that their investment will lead to higher future consumption. While profitability may thus still increase over time, one should also keep in mind that the businesses were mainly in the services and agricultural sectors and quite straightforward in nature. After loan disbursal, borrowers should in most cases have been able to implement investments and reap their pay-offs quite quickly.

Other constraints

An alternative explanation is that access to finance may not be the only binding constraint on successful entrepreneurial activity. Bruhn and Zia (2011)[2] use an RCT to study the impact of a business and financial literacy program on the firms of young Bosnian entrepreneurs, all of whom were borrowers from a local microfinance institution. They find that while training did not influence business start-up or survival, it significantly improved business practices, investments and loan terms for surviving firms. An interesting area for future research is therefore to uncover what combinations of credit and training can help stimulate entrepreneurship not only at the intensive but also at the extensive margin.


  1. Britta Augsburg, Ralph De Haas, Heike Harmgart, and Costas Meghir. Microfinance at the Microfinance at the Margin: Experimental Evidence from Bosnia and Herzegovina Vidence from Bosnia and Herzegovina. Working Papers No. 46. European Bank for Reconstruction and Development (EBRD), September 2012.
  2. Miriam Bruhn, and Bilal Zia. Stimulating Managerial Capital in Emerging Markets: The Impact of Business and Financial Literacy for Young Entrepreneurs. World Bank Policy Research Paper no. 5642, World Bank (2011).


This article may be reproduced according to our terms of use with attribution (and link, if online) to To be cited as: “Microfinance at the margin” Augsburg, B., R. De Haas, H. Harmgart, and C. Meghir,, October 31, 2012".

Read more

Microfinance at the margin
Research Paper


Ralph de Haas

Deputy Director of Research, ERDB

Relevant articles