Elsevier

World Development

Volume 34, Issue 9, September 2006, Pages 1612-1638
World Development

Microfinance in Northeast Thailand: Who benefits and how much?

https://doi.org/10.1016/j.worlddev.2006.01.006Get rights and content

Summary

This paper evaluates the outreach and impact of two microfinance programs in Thailand, controlling for endogenous self-selection and program placement. Results indicate that the wealthier villagers are significantly more likely to participate than the poor. Moreover, the wealthiest often become program committee members and borrow substantially more than rank-and-file members. However, local information on creditworthiness is also used to select members. The programs positively affect household welfare for committee members, but impact is insignificant for rank-and-file members. Policy recommendations include vigilance in targeting the poor, publicly disseminating the program rules and purpose, and introducing and enforcing eligibility criteria.

Introduction

Historically, efforts to deliver formal credit and financial services to the rural poor in developing countries have failed. Commercial banks generally do not serve the needs of the rural poor because of the perceived high risk and the high transactions costs associated with small loans and savings deposits. To fill the void, many governments have tried to deliver formal credit to rural areas by setting up special agricultural banks or directing commercial banks to loan to rural borrowers. However, these programs have almost all failed because of the political difficulty for governments to enforce loan repayment, and because the relatively wealthy and powerful, rather than the poor, received most of the loans (Adams, Adams and Vogel, 1986, World Development Report, 1989).

The recent proliferation of innovative microfinance programs, often based on group-lending methods, has been inspired largely by the belief that such programs reach the poor and have a positive impact on various measures of their welfare, including economic measures (e.g., wealth and income), social measures (e.g., educational attainment and health status), and less tangible measures such as “empowerment.” The popular press has waved the banner of microfinance as perhaps the most important recent tool to reduce poverty,1 and the 1997 Microcredit Summit called for the mobilization of $20 billion over a 10-year period to support microfinance (Microcredit Summit Report, 1997). The United Nations proclaimed 2005 as the “Year of Microcredit.” Much of this faith in microfinance is based on the highly selective anecdotal evidence of individuals who are reported to have pulled themselves and their families out of poverty with the benefit of microcredit. On the other hand, prominent dissenters to the popular view (Adams & von Pischke, 1992) have written that “debt is not an effective tool for helping most poor people enhance their economic condition—be they operators of small farms or micro entrepreneurs, or poor women.” They argue that access to credit is not a significant problem faced by small agricultural households and that factor and product prices, land tenure, technology, and risk are the factors limiting small farmer development. Yet, despite the proliferation of these programs and the outpouring of support by donors, there has been little sound empirical research that tests the hypotheses that they are reaching and benefiting the poor.2

To justify such a significant investment to reduce poverty, compared to alternative investments in other poverty alleviation programs, the proposition that microfinance reaches the poor and positively affects their welfare should be proven and not just assumed. This paper attempts to contribute to overcoming this shortcoming in the literature by examining the results of a survey of two Northeast Thailand “village bank” programs that target the poor. The survey was designed and conducted in 1995–96 with the express purpose of measuring outreach and impact on the poor, while controlling for the endogeneity biases that have plagued other studies.

The NGO programs studied in this paper targeted “the poorest of the poor” according to project documents and donor policy. The ability of any program to achieve this goal depends on the institutional context in which it is implemented, and the main premise on which microfinance programs are based is that the poor are credit constrained and have limited access to formal sector credit. In Thailand, however, the Bank for Agriculture and Agricultural Cooperatives (Bank for Agriculture & Agricultural Cooperatives, 1997) claims to serve over 80% of rural households. Hence, it is possible that the rural poor in Thailand are not credit constrained. However, the BAAC’s outreach in the Northeast, the country’s poorest region, is smaller than the rest of the country. In the fourteen villages surveyed for this study, 63% of village households were BAAC members. Moreover, as is often the case in government-led credit programs, the BAAC’s clientele is largely male; only 29.5% of BAAC members surveyed by the author were women. Hence, only 18.6% of surveyed households included women who had access to BAAC loans. On the other hand, 25.8% of surveyed households included women who were in debt to moneylenders. At the time of the surveys, BAAC’s annual interest rate varied from 3% to 12%, whereas moneylenders charged between 60% and 120% per year, and the NGO programs evaluated in this paper charged 24% per year. Hence, there is evidence that women in Northeast Thailand may be credit constrained and may benefit from access to lower-cost institutional credit.

Two main problems plague attempts to evaluate the impact of microfinance programs.3 The first is self-selection of participants. To illustrate this source of bias, consider a sample of households drawn only from villages with a village bank: some households will have selected to be village bank members, while others will have selected not to be members. It is likely that there are significant differences between self-selected village bank members and nonmembers. To the extent that such differences can be observed and measured (e.g., age, education, and wealth endowment), they can be statistically controlled for when estimating village bank impact. However, to the extent that such differences cannot be observed (e.g., entrepreneurship, risk preferences, trustworthiness, attitudes regarding the role of women in the household, and attitudes toward belonging to a program targeting the poor), direct comparison of village bank members and nonmembers will yield biased estimates of village bank impact. This bias arises because the same unobservable characteristics that lead some women to become village bank members will also affect impact measures such as income, accumulation of assets, and spending on education and health care. For example, women who are more entrepreneurial (a characteristic that is virtually impossible to measure) would be expected to have a tendency to self-select into the program, but such women would also be expected to have higher welfare measures such as income and expenditures even without the program. Uncontrolled comparisons between members and nonmembers, therefore, might incorrectly attribute such higher incomes to the village bank program. Alternatively, the relatively poor might self-select into the program if being poor is a publicly known selection criterion, and the relatively wealthy might not join to avoid any stigma related to being poor. If the program has positive impact on participants but this impact is not strong enough, then an uncontrolled comparison between poor members and rich nonmembers might lead to the erroneous conclusion that the program was impoverishing participants.

The second problem affecting attempts to measure impact is endogenous program placement and is similar to the self-selection problem. To understand this problem, it is useful to consider a commonly used sample, which includes households from villages with a village bank and households from villages without a village bank (e.g., MkNelly and Watetip, 1993, Wydick, 1999a, Wydick, 1999b). Prior to program placement, some villages may be perceived as more entrepreneurial or better organized and governed, or may have more dynamic leaders, with such leadership spilling over to affect others’ behavior in the village. These unmeasurable characteristics may cause villagers in these villages to have higher incomes, spend more on education and health, and generally have higher measures of welfare than households in other villages, even without the program. If the sponsoring NGO uses such unmeasurable characteristics to select a village for program placement, then a comparison of households in the program village and households in a nonprogram village may overestimate impact. Similarly, if the sponsoring NGO deliberately selects poorer villages because of its mission to reduce poverty, a comparison of households in the program village and households in a nonprogram village may underestimate impact. Coleman (1999), using the same data set examined in this paper, demonstrates the extent to which estimates uncorrected for member self-selection and endogenous program placement significantly overestimate average program impact.

Assessing the targeting of a village bank, for example, determining if, and to what extent, the program is reaching poor households as intended, also encounters difficulties.4 Foremost among these is that, typically, empirical studies focus on program impact, and such studies necessarily require that households be surveyed after the program has operated for some time. Because measures of poverty, such as wealth and income, have been influenced by the program, it is typically impossible to determine if a participating household was poor when it first joined the program. An ex post survey finding that a participating household is relatively wealthy could indicate either unsuccessful targeting (resulting in the poor being excluded and the rich co-opting program benefits) or successful program impact.

This paper extends and refines the methodology used in Coleman (1999). First, it exploits the unique characteristics of the survey sample to evaluate targeting, to determine if, prior to joining the program, participating households are relatively poor or not. Second, it extends the impact estimates of the earlier study to measure differential impact on different classes of participants, specifically on the relatively wealthy village bank managers and rank-and-file members who tend to be poorer. Results indicate that self-selected program participants are significantly wealthier than nonparticipants even prior to program intervention, and the wealthiest villagers are almost twice as likely to participate in the program as the poorer villagers. Moreover, some of the wealthiest villagers obtain a disproportionate share of program loan volume by virtue of holding influential positions as village bank committee members. Specifically, they do this by using multiple names to borrow more than the program’s ceiling per client. Positive impact is seen largely in this wealthier group. Impact on rank-and-file members is significantly smaller than impact on the wealthy, and is largely insignificant. Hence, similar to previous attempts to deliver low-cost credit to the poor, most of the benefit in the village banks studied is going to the wealthiest villagers.

The remainder of the paper is organized as follows. Sections 2 The NGO programs studied, 3 Survey design and data, 4 Survey area describe the design of the programs studied, the survey design and data, and the survey area. Section 5 presents results relating to participation, including member selection and borrowing, while Section 6 presents results on program impact. Section 7 concludes and discusses policy implications.

Section snippets

The NGO programs studied

The two microfinance programs studied are run by Thai NGOs: the Rural Friends Association (RFA), located in the northeast province of Surin, and the Foundation for Integrated Agricultural Management (FIAM), located in the adjacent province of Roi-Et. RFA and FIAM have been promoting microfinance since 1988 and have received financial and technical assistance from the American NGO Catholic Relief Services (CRS). Both Thai NGOs follow the “village bank” group-lending methodology pioneered by the

Survey design and data

I conducted a unique survey of 444 households in 14 villages in Northeast Thailand in 1995–96. Eight of the villages were supported by RFA, and the other six were supported by FIAM. Of the 14 villages surveyed, six had never benefited from village bank support, and did not receive any village bank loans during the survey period. These “control” villages were identified as follows. Based on their expansion plans, RFA pre-identified four villages and FIAM pre-identified two villages that they

Survey area

The provinces of Surin and Roi-Et are adjacent to each other and are located in Northeast Thailand, the country’s poorest region. Most village households engage in small-scale agriculture: 90.4% of the adult men and 91.3% of the adult women in the households surveyed listed farming as their primary or secondary occupation. In Surin, rain-fed rice is the primary crop grown, with planting in June and harvesting from November to January. During the off-season, a few households engage in

Selection of members

The raison d’être of most microfinance programs is to correct the market failure to deliver credit to the rural poor. Most microfinance programs state that their primary goal is to alleviate rural poverty by delivering credit and other financial services to poor households, especially to the women in those households. This is certainly the case for the programs studied in this paper. For example, Catholic Relief Services publishes “Eight Principles of Village Banking,” the first of which is to

Impact

To appreciate the bias potentially arising from self-selection and endogenous program placement, as discussed in Section 1, consider the following empirical specification:Bij=XijαB+VjβB+εij,Yij=XijαY+VjβY+BijδY+μij,where Bij is the amount borrowed from the village bank by household i in village j; Xij is a vector of household characteristics; Vj is a vector of village characteristics; Yij is an outcome on which we want to measure impact; αB, βB, αY, βY, and δY are parameters to be estimated;

Summary and policy conclusions

This paper has evaluated the targeting and impact of a women’s group-lending program in Northeast Thailand. To do so, it exploited a unique survey sample that included program participants from “treatment” villages that had already received program support, participants from control villages that had not yet received program support, and nonparticipants from both types of villages. The results were presented in terms of targeting (i.e., the processes of member selection and borrowing) to

Acknowledgement

I would like to thank George Akerlof, Pranab Bardhan, David Dole, Paul Gertler, Alain de Janvry, Elisabeth Sadoulet, Ken Train, Ploenpit Satsanguan, seminar participants at the University of California at Berkeley, and two anonymous reviewers for their helpful comments; the staff of CRS/Thailand, especially Yupaporn Boontid and Ruth Ellison, for their advice and support throughout the surveys; the staff of RFA/Surin and FIAM/Roi-Et for the able enumeration services of their field staff and for

References (47)

  • B. Armendáriz de Aghion et al.

    The economics of microfinance

    (2005)
  • N. Ashraf et al.

    Tying odysseus to the mast: Evidence from a commitment savings product in the Philippines

    Quarterly Journal of Economics

    (2006)
  • Bank for Agriculture and Agricultural Cooperatives (1997). Annual Report....
  • M. Chen

    Impact of Grameen Bank’s credit operations on its members: Past and future research

    (1992)
  • B. Coleman

    The impact of group lending in Northeast Thailand

    Journal of Development Economics

    (1999)
  • Coleman, B. (2005). Risk, mutual assistance, and mutual insurance among village bank members. Mimeo, Asian Development...
  • J. Conning

    Outreach, sustainability and leverage in monitored and peer-monitored lending

    Journal of Development Economics

    (1999)
  • Duflo, E., & Kremer, M. (2003). Use of randomization in the evaluation of development effectiveness. Mimeo prepared for...
  • Foundation for International Community Assistance (FINCA) (1990). Promoting and supporting village banks: Outline for a...
  • M. Grosh

    Toward quantifying the trade-off: Administrative costs and incidence in targeted programs in Latin America

  • S. Hashemi

    Those left behind: A note on targeting the hardcore poor

  • Hatch, J. (1989). A manual of village banking for community leaders and promoters. Washington,...
  • D. Hulme et al.

    Finance against poverty

    (1996)
  • Cited by (0)

    View full text