【正文】
agship, the Grameen Bank of Bangladesh, have now spread around the world. While programs aim to bring social and economic benefits to clients, few attempts have been made to quantify benefits rigorously. This paper draws on a new crosssectional survey of nearly 1800 households, some of which are served by the Grameen Bank and two similar programs, and some of which have no access to programs. Households that are eligible to borrow and have access to the programs do not have notably higher consumption levels than control households, and, for the most part, their children are no more likely to be in school. Men also tend to work harder, and women less. More favorably, relative to controls, households eligible for programs have substantially (and significantly) lower variation in consumption and labor supply across seasons. The most important potential impacts are thus associated with the reduction of vulnerability, not of poverty per se. The consumptionsmoothing appears to be driven largely by inesmoothing, not by borrowing and lending. The evaluation holds lessons for studies of other programs in lowine countries. While it is mon to use fixed effects estimators to control for unobservable variables correlated with the placement of programs, using fixed effects estimators can exacerbate biases when, as here, programs target their programs to specific populations within larger munities. Key words: microfinance, project evaluation, Grameen Bank, Bangladesh 1. Introduction Microfinance has captured the imaginations of many people working to reduce poverty. The premise is simple. Rather than giving handouts to poor households, microfinance programs offer small loans to foster smallscale entrepreneurial activities. Such credit would otherwise not be available or would be only available at the very high interest rates charged by moneylenders (who often charge as much as 10% per month). Moneylenders operate with little petition since potential entrants quickly find that costs and risks are high and borrowers are usually unable to offer standard forms of collateral, if any at all (Rashid and Townsend, 1993). However, the emerging microfinance movement demonstrates institutional innovations that appear to greatly reduce the risk and cost of providing financial services to poor households. Innovations include contracts that give borrowers incentives to exclude bad credit risks and monitor other b