Abstract
Microfinance reduces vulnerability. We investigate if this impact varies by the delivery mechanism used. Correcting for the membership selection bias using Propensity Score Matching (PSM), the household's vulnerability is estimated using the Self-Help Group (SHG) microfinance programme data in India. The results show that the reduction in vulnerability is greater for villages with better infrastructure and for SHGs that are formed by NGOs and linked by banks (linkage model 2).