Unlike their formal counterparts, micro-finance organizations in Bangladesh have made their stride in delivering financial services (both savings and credit) to the poor, especially women, at a very low loan default cost. Strategies such as collateral-free group-based lending and mobilization of savings, even in small amounts, have helped them mitigate the problems of poor outreach and high loan default costs of their formal counterpart. However, they assume high transaction costs in order to keep credit discipline among borrowers through group pressure and monitoring of borrowers' behavior. The transaction cost is substantial and programs have been relying on donors for sustaining their operations. Nonetheless, the government and donors continue to support microfinance programs in Bangladesh with the expectation that society benefits from such investment.

Therefore, policymakers and program organizers are keen to learn the extent of socioeconomic impacts of micro-finance on borrowers and on society at large. At the household level, two types of impacts can be ascertained i.e. household and individual impacts of microfinance. Household level impacts, such as impacts on income, employment, and poverty are assessments without specifying the intra-household distribution of induced benefits of microfinance. Intra-household impacts are examined to learn the distribution of benefits among different members of households, especially between men and women. Since women are disadvantaged in a society such as Bangladesh and constitute the overwhelming share of microfinance membership, the policy question is: do women benefit from micro-finance and if so, how? At the societal level, the policy question is: do micro-finance programs benefit non-program participants or do they simply help redistribute income in a society?

One of the early studies of Grameen Bank shows how Grameen Bank has been supporting the poor, especially women, in terms of employment, income generation and promotion of social indicators (Hossain 1988). Other BIDS (Bangladesh Institute of Development Studies) and non-BIDS studies also indicate the beneficial aspects of micro-finance operation in Bangladesh (e.g. Rahman 1996; Hashemi, Schuler, and Riley 1996; Schuler and Hashemi 1994). These studies show the positive correlation between micro-finance programs and their accrued benefits, but do not indicate the causality, meaning whether these programs actually matter in generating such benefits to the borrowers.

The most comprehensive and rigorous micro-finance impacts studies that have established causality were carried out more recently in a joint research by BIDS and the World Bank (Khandker 1998; Pitt and Khandker 1998). This body of research provides a strong indication that the programs do help the poor in consumption smoothing, as well as in building assets. The findings also lend support to the claim that micro-finance programs promote investment in human capital (such as schooling) and contribute to increasing awareness to reproductive health (such as the use of contraceptives) among poor families. This major study also sheds lights on the role of gender-based targeting and its impact on household or individual welfare. Findings suggest that women do acquire assets of their own and exercise power in household decision-making.

The positive impact of micro-finance programs at the borrower level are thus tenable. Even then, the question arises: what are the long-run impacts of micro-finance? Are the program impacts found in 1991/92, sustainable over time? If poverty reduction is possible with microfinance at the borrower level, what is the impact of micro-finance on aggregate poverty? Earlier estimates suggest that micro-finance can contribute to consumption at a rate of 18 percent in the case of female borrowing and at 11 percent in the case of male borrowing. Of course, this is a short-term impact and, hence, may be short-lived. It is possible that the proportion of program participants enjoying the benefits of micro-finance is very small and that the impacts of their accrued benefits on the overall economy are small as well and may not be sustainable over time.

The World Bank study based on the 1991/92 household survey indicates that only about less than 5 percent of borrowers can lift themselves out of poverty each year by borrowing from a micro-finance program, even if the estimated impacts on consumption are sustained over time (Khandker 1998). Such percentage represents only about 1 percent of the population; thus, the aggregate poverty impact of micro-finance program is quite negligible. Does it imply that microfinance programs do not need to be supported?

Despite this miniscule aggregate impact, the micro-finance movement in Bangladesh has received continuous support from donors, and more recently from the World Bank. In 1996, the World Bank provided a loan of US$115 million to the country's autonomous body called Palli Karma Shahayak Foundation (PKSF). This body works as an intermediary for wholesaling microfinance. It supports the on-lending of small non-government organizations (NGOs) and a few large NGOs such as the Bangladesh Rural Advancement Committee (BRAC). However, the Grameen Bank did not seek any loan or grant from this facility. The project ended in June 2000 and the World Bank, with the request from the government, supported a follow-up project of US$ 160 million. With the flow of funds from various micro-finance agencies to borrowers, about 8 million (out of a total of approximately 30 million) households received help from micro-finance programs in 1998/99. Loan outstanding of micro-finance programs was about US$ 600 million in 1998/99. The organized NGO sector and the specialized Grameen Bank accounted for more than 86 percent of micro-finance lending, while only 14 percent came from the country's commercial banks.

Despite the large inflow of micro-credit into the rural sector of Bangladesh, the incidence of rural poverty has been stubbornly high. Rural poverty was 54 percent in 1983/84 and has been above 50 percent over the last decade (Ravallion and Sen 1995). It has declined to about 45 percent in recent years; yet the incidence of poverty has remained high. Critics argue that this reflects the limitation of micro-finance programs as an instrument in arresting poverty in a country, which has the largest micro-finance operation in the world. Is the high incidence of poverty a result of the failure of micro-finance movement? Or is it an outcome of the persistently low economic growth rate (which has been only 4 percent over the last decade) in the country? If substantial poverty reduction largely depends on sustained high economic growth, what is the net overall contribution of the micro-finance movement in Bangladesh?

The total loan outstanding of micro-finance organizations in Bangladesh was about US$ 600 million in 1998/99. This indicates a large inflow of micro-funds in the rural areas, which is expected to make an aggregate impact on the local economy. We have just seen that microfinance has sizeable effects on the welfare of borrowing households in terms of raising consumption and non-land asset as well as in reducing moderate and extreme poverty. But are these effects felt beyond the program participants? How do we account for estimating microfinance effects on non-participants?

The benefits of non-participants depend on the amount of credit obtained by all program borrowers living in a village. We find evidence of externality of microcredit programs. Male borrowing from micro-finance programs affects the food, non-food, and total per capita expenditure of non-participants, but has no effect on the non-land asset. In contrast, although female borrowing has no significant effect on consumption of non-participants, it has a substantial effect on their non-land asset. For instance, a 10-percent increase in aggregate village microcredit borrowing by female members increases household non-land asset of non-participants by as much as 1.1 percent. Micro-finance also affects the poverty incidence among non-participants. The negative spillover effect on poverty is, however, more pronounced for extreme poverty than for moderate poverty. The two-stage logit estimates indicate that the probability of reducing extreme poverty can be as much as 0.09 percentage points. Accordingly, moderate poverty for non-participants declined by 1.1 percentage points and extreme poverty by 4.8 points over the study period because of female borrowing in the village from micro-finance programs.

The results strongly support the view that micro-credit not only affects the welfare of participants and non-participants, but also the aggregate welfare at village level. Male borrowing increases average household welfare by increasing household total consumption, as well as food and non food consumption, and by reducing extreme poverty, with only a minimal effect on moderate poverty. Female credit, on the other hand, reduces extreme poverty and increases household non-land asset for an average household in a village.

Micro-finance operation in a village, therefore, reduces the incidence of extreme poverty for an average household living in a village, but it does not affect the incidence of moderate poverty. The probability of reducing aggregate extreme poverty is approximately .03 percentage points for female borrowing and .08 percentage points for male borrowing. However, the net contribution of female borrowing over the study period is 1.7 percentage points for moderate poverty and 5.5 percentage points for extreme poverty. Non-borrowers seem to benefit from micro-finance partly due to the externality of borrowing by program participants and partly because of externality due to program placement. Therefore, micro-finance contributes to the overall welfare of the society.

Program evaluation compares outcomes of treatment groups with those of control groups. Finding control group in a non-experimental setting is very difficult. Traditionally, resorting to instruments for identifying program effects is done with cross-section data. However, finding good instruments is equally difficult. Pitt and Khandker (1998) used quasi-experimental method relying on exogenous eligibility conditions as a way of identifying program effects. Some of the conditions are restrictive and might not be reliable, for example, the non-enforceability of landholding criterion for program participation. Results may be sensitive to methods used in impact assessment. An impact assessment was carried out using a follow-up survey to see the sensitivity of the findings (see Khandker and Pitt (2002) for details).

This paper carried out a similar exercise by estimating the effects of micro-finance on consumption, poverty and non-land assets for participants, non-participants, and an average villager, assuming that micro-finance programs have spillover (externality) effects.

The results are resounding: micro-finance matters a lot for the very poor borrowers and also for the local economy. In particular, micro-finance programs matter a lot to the poor in raising per capita consumption, mainly on non-food, as well as household non-land asset. This increases the probability that the program participants may be able to lift themselves out of poverty. The welfare impact of micro-finance is also positive for all households, including non-participants, indicating that micro-finance programs are helping the poor beyond income redistribution with contribution to local income growth. Programs have spillover effects in local economies, thereby increasing local village welfare. In particular, we find that micro-finance helps reduce extreme poverty more than moderate poverty at the village level. Yet the aggregate poverty reduction effects are not quite substantial to have a large dent on national level aggregate poverty. This concern brings to the fore the effectiveness of micro-finance as an instrument to solve the problem of poverty in Bangladesh.

To exhibit a stronger impact on poverty reduction, micro-finance should perhaps go beyond the provision of financial services. It should find ways to improve the skills of its poor borrowers to improve their productivity and income. It should also assist its borrowers in marketing and improving the quality of their products. Micro-finance is, however, only one of the many instruments of poverty reduction. Growth matters too-even more significantly than other instruments. Investment in human capital and other means to empower the poor also matter. To achieve substantial poverty reduction, the other avenues must be explored as well.

References :

1. Hossain, Mahabub. 1988. "Credit for Alleviation of Rural Poverty: The Grameen Bank in Bangladesh." IFPRI Research Report # 65. International Food Policy Research Institute, Washington, D.C.

2. Khandker, Shahidur R. 1988. Fighting Poverty with Microcredit: Experience in Bangladesh. New York, NY: Oxford University Press for the World Bank.

3. Khandker, Shahidur R. and Pitt, Mark 1998. "The Impact of Group-Based Credit Programs on Poor Households in Bangladesh: Does the Gender of Participants Matter?" Journal of Political Economy 106 (October): 958-96.

4. Rahman, Rushidan I. 1996. "Impact of Grameen Krishi Foundation on the Socioeconomic Condition of Rural Households", BIDS Working paper# 7, BIDS, Bangladesh Institute of Development Studies, Dhaka.

5. Hashemi, Syed M., Sidney R. Schuler and Ann P. Riley. 1996. "Rural Credit Programs and Women's Empowerment in Bangladesh." World Development # 24 (April):635-53.

6. Schuler, Sidney R., and Syed M. Hashemi. 1994. "Credit Programs, Women's Empowerment, and Contraceptive Use in Rural Bangladesh." Studies in Family Planning # 25 (March/April):65-76.
7. Ravallion, Martin, and Binayak Sen. 1995. When Methods Matter: Towards a Resolution of the Debate about Bangladesh's Poverty Measures. Policy working Paper #1259, World Bank, Washington, D.C.