Three essays on the impact assessment of microcredit program participation in Bangladesh using household longitudinal survey

Abstract

The employment impact of microcredit, in the existing literature, has so far been assessed using cross sectional data stratified by target and non-target villages and households. However, the relevance of this experimental set up is weakened due to the rapid expansion of the microcredit program far and wide and growing competition among microfinance institutions to attract members since the mid-nineties. Now, there is hardly any village in Bangladesh where there is no microcredit program in operation. The implication of this lack of stringency in the division between control and treatment villages for program evaluation is that it has turned cross-sectional comparison of outcomes between participant and non-participant groups across program and control villages less convincing. This paper responds to this loss of quasi-experimental design in microcredit program evaluation by using household level longitudinal data in Bangladesh collected by the Pally Karma Sahayak Foundation (PKSF) in 1998 and 2004 for the Monitoring and Evaluation Study of Microfinance Institutions. Using panel data of over 2000 households, this paper examines the employment impact of microcredit program in Bangladesh on self-employment activity during 1998-2004. The pooled ordinary least squares regression of household annual self-employment work hours shows that participants worked significantly longer hours than non-participants. The household fixed effects model produces lower estimate of the contribution of microcredit to self-employment work hours suggesting that self-employment growth is also attributable to the time-invariant household characteristics such as, initial conditions and entrepreneurial skills. According to this estimate, microcredit program participation accounts for 245 hours of self-employment on average per household in a year which is equivalent to around 7 weeks of employment for a person annually. At the average hourly income around 14.40 Taka in the study area, the income generated from 245 hours of work is 3,528 Taka. This is equivalent to about 5% of the average annual household income of the households surveyed in this study.It is also observed that the participant households do not have as much non-labor income earning opportunities as the non-participants and the former group is more dependent on non-farm sector for generating their income than the latter group.