Peer Effects in Consumption in India : An Instrumental Variables Approach Using Negative Idiosyncratic Shocks

This paper examines peer effects in consumption in context of a less developed country. Specifically, the question that I seek to answer is whether consumption expenditure of a household is influenced by that of its peers in a less developed country. To examine this question, I use newly available household level data from India. I define a household's peer group as other households living in its village/neighborhood. In assessing the influences of peers in this context, there are two key empirical challenges including shared group-level unobservables, and simultaneity of peer influences. I address these issues by using an instrumental variables/fixed effects approach that compares households in the same district but different villages/neighborhoods who are thus exposed to different sets of peers. In particular, I use plausibly exogenous variation in idiosyncratic expenditure shocks - which are accidental and negative in nature - faced by peers as instruments for peers' consumption expenditure. Preferred specification suggests that a one standard deviation increase in average consumption expenditure of a household's peers causes the household's own consumption expenditure to increase by 0.42 standard deviations. Falsification tests and robustness checks support the validity of my results. My findings suggest that policies that influence a household's consumption will also affect the consumption of the household's peers through social interactions. This implies traditional analyses of consumption intervention programs that do not take into account such spillover effects will understate the total social impact of the programs, and hence lead to inaccurate evaluation of cost-effectiveness of such programs