Binary endogenous treatment in stochastic frontier models with an application to soil conservation in El Salvador

Abstract

Numerous programs exist to promote productivity, alleviate poverty, and enhance food security in developing countries. Stochastic frontier analysis can be helpful to assess their effectiveness. However, challenges can arise when accounting for treatment endogeneity, often intrinsic to these interventions. We study maximum likelihood estimation of stochastic frontier models when both the frontier and inefficiency depend on a potentially endogenous binary treatment. We use instrumental variables to define an assignment mechanism and explicitly model the density of the first and second-stage error terms. We provide empirical evidence using data from a soil conservation program in El Salvador.