Welfare programs and household behaviour: the case of MGNREGS

Abstract

"The welfare effects of workfare programs like India’s Mahatma Gandhi National Rural Employment Guarantee Scheme (MGNREGS) have been identified and measured extensively in the literature. I hypothesize that both the temporal and intertemporal welfare effects are conditioned on the type of village-level assets created, and the duration of exposure to such programs. In this thesis, I provide empirical evidence on the identification of channels and pathways through which MGNREGS has improved rural household welfare. In chapter one, I state and introduce the problem and carry out an extensive survey of the extant literature, while explaining the extent to which this literature is not addressing the problem explored in this thesis. In chapter two, I estimate the exposure effects of MGNREGS on household-level consumption expenditures. I posit that the duration of village-level exposure to MGNREGS will create spillover effects for individual households as a result of asset creation. This produces village-wide income effects that can affect the budget constraints of both beneficiary and non-beneficiary households. Two types of assets are identified: standalone agricultural assets (water conservation, irrigation, land development, and renovation of water bodies), and a combination of agricultural and rural roads. I find that the creation of such assets leads to long-term increases in household food and non-food expenditures. In chapter three, I estimate the intertemporal effects of MGNREGS on household-level vulnerability. I first compute vulnerability as expected poverty (VEP) by using a panel of households obtained from the Rural Economic and Demographic Survey (REDS) 1999 and 2006 rounds, and the Socio-Economic Profile of Rural India (SEPRI) 2016 surveys. I show that household-level exposure to MGNREGS in the form of long-term ownership of job cards (cards that enable households to obtain employment for their desired duration) reduces the incidence of vulnerability. I also find evidence for the heterogenous effects of job card ownership for landless households, less-educated households, and households with a greater share of females. In chapter four, I estimate the impact of household-level exposure to MGNREGS on poverty dynamics. Can long-term ownership of job cards change the households’ probability to exit or re-enter poverty? I find that conditional on ex-ante poverty, long-term exposure to MGNREGS reduces poverty and the relative probability of entering poverty and enables exit out of poverty by acting as a safety net. The final chapter concludes and provides policy recommendations."