The Impact of MGNREGA on Agricultural Outcomes and the Rural Labour Market: A Matched DID Approach

This paper attempts to address the impact of the MGNREGA on the rural agricultural sector, focusing on cropping patterns, irrigated area, crop yields, wages and rural employment. The analysis is based on two data sources: the first is a unique district-season level panel dataset that we construct using multiple sources, and the second is unit record data from the NSS Employment Unemployment Surveys. To identify causal effects, we employ a difference-in-difference matching procedure, where districts are matched based on propensity scores; the use of propensity scores represents a novel aspect of this paper. We also examine pre-programme trends for each outcome variable to provide a check on the validity of our estimates. Our results indicate modest changes in cropping patterns that are state and period specific; however, they do not indicate any improvements in crop yields that were expected given the MGNREGA’s focus on investments in irrigation, although there is some evidence that irrigated area may have expanded after a lag. We also find that there is no systematic evidence of impact on wages and therefore no evidence that public works employment in MGNREGA crowded out casual labour in agriculture.