Village Banking and Maternal and Child Health: Evidence from Ecuador and Honduras

Abstract

Many microcredit institutions bundle credit with health, education, or other programs. This paper reviews arguments on such “tie-ins” and concludes that net benefits are possible but requires empirical testing. Quasi-experimental data from Project HOPE's “health banks” and credit-only village banks are then examined. The paper presents evidence that health practices do not improve automatically with greater wealth. In Honduras, results show that health bank participation is robustly associated with reduced subsequent conditional child diarrhea probability, but in no specification does credit-only bank participation have this effect. In Ecuador, results suggest a larger effect of credit-only banks. In both countries, health bank participation significantly raises subsequent healthcare over credit-only participation, and at least reduces the tendency to switch from breast-feeding to bottle-feeding as income rises. Effects on expenditures are ambiguous. There is no clear link between tie-ins and bank performance.

Introduction

Microenterprise credit unions, including “village banking,” has become perhaps the most prominent “fashion” in development policy. Credit has consistently been found to be a binding constraint for microenterprise development. Lack of credit particularly, though certainly not exclusively, affects women borrowers, for reasons including lack of collateral and property rights. At the same time, evidence suggests that mothers' income increases are more likely to be spent on nutrition and other health-improving expenditures for children than fathers' income increases (e.g., Thomas, 1990, Thomas, 1993). Thus, relaxing credit constraints for women has been viewed as a particularly powerful development policy tool.

In recent years, there has been a growing understanding of how microfinance schemes, such as the group lending methodology used by the Grameen Bank of Bangladesh, can use peer pressure and “social collateral” to lend successfully to poor borrowers (for recent surveys, see Ghatak & Guinnane, 1999; Morduch, 1999a). Empirically, such schemes have an impressive track record, with a repayment rate typically well over 90%, and an unusually large capacity to become financially self-sustaining. Development agencies and nongovernmental organizations (NGOs) typically play a significant role in establishing village banks. While NGOs' seed capital for initial loans is usually repaid, these funds may be provided at below-market interest rates; and NGOs' time and other resources provided are generally not repaid. Thus village banks do enjoy a significant subsidy (Khandker, Khalily, & Khan, 1995; Morduch, 1999b). Nevertheless, the success of microfinance programs has led major development agencies to embrace the concept and expand funding. The Microcredit Summit of February, 1997, which concluded with a stated goal of reaching 100 million of the world's poorest families by 2005, is an important case in point (Microcredit Summit Fulfillment Campaign, 1997).

Spurred by the success of such programs, notably the Grameen Bank, practitioners in many aspects of development have recently sought to tie their work with microfinance schemes. For example, the NGO Freedom From Hunger has attempted to integrate village banking with its basic education programs in rural Thailand and elsewhere. A group of NGOs engaged in tie-ins between credit and other social development goals has begun an active “practitioners' network,” the “Credit with Education Learning Exchange” (see, e.g., Vor der Bruegge, Dickey, & Dunford, 1997). The Grameen Bank itself has been a pioneer in tying social development goals with credit.¹

Such noncredit tie-ins, usually focused on education or health goals, are established partly out of a conviction that while removing credit constraints on microenterprises represents an unusually effective development strategy, it is necessary but insufficient. Proponents argue that raising family income is not sufficient to meet development goals such as ending malnutrition, reducing infant mortality, and increasing the education levels of girls. It is also argued that many low-income microentrepreneurs, particularly women, are also lacking in basic health services, and essential knowledge about health as well as about business practices, to successfully utilize credit to raise their families out of poverty. Such observations suggest to practitioners an interdependency between income, health and education that is best dealt with through a simultaneous and integrated program strategy.²

Although in recent years, several NGOs working in the fields of education and of maternal and child health have developed programs to tie their work with microenterprise credit, there have been only the most sketchy of data to support such a diversion of resources. The first goal of this paper is to systematically review arguments in favor of, or against, integrating credit provision with other development services, to establish a framework for testing.

The second goal of the paper is to introduce the first economic evidence on the effects of integrating village banking with noncredit tie-ins. One NGO, Project HOPE, has collected data on their “Village Health Banks” program in Ecuador and Honduras. HOPE created new village banks both with and without a health education and services component, and collected data from both bank participants and nonparticipants, including some basic financial and health information. The credit-only banks operated for several loan cycles (almost two years) before a health component was added. Thus, we have a unique social science quasi-experiment to provide an empirical analysis of an emerging trend in development policy at an early stage.

The remainder of the paper is organized as follows. In Section 2, the arguments (and available evidence) for and against the use of tie-ins in microcredit institutions are briefly summarized. As noted below, a full cost-benefit analysis of bundled programs vis-à-vis credit-only programs is outside the scope of the present paper. Rather, the focus is on examining the channels through which tied-in health interventions may (or may not) improve household outcomes over credit-only banks. Data from the quasi-experiment conducted by Project HOPE, described in Section 3, are used in order to evaluate the effectiveness of this intervention. In Section 4 the effect of household participation in health banks or credit-only banks for a range of household outcomes are estimated (expenditures, incidence of child diarrhea, breast-feeding and routine annual maternal cancer screening), followed by a brief consideration of the impact on the performance of the banks themselves. Section 5 concludes the paper.