Entropy Balancing for Continuous Treatments

Document pages: 32 pages

Abstract: This paper introduces entropy balancing for continuous treatments (EBCT) byextending the original entropy balancing methodology of Hainmüller (2012). Inorder to estimate balancing weights, the proposed approach solves a globallyconvex constrained optimization problem. EBCT weights reliably eradicatePearson correlations between covariates and the continuous treatment variable.This is the case even when other methods based on the generalized propensityscore tend to yield insufficient balance due to strong selection into differenttreatment intensities. Moreover, the optimization procedure is more successfulin avoiding extreme weights attached to a single unit. Extensive Monte-Carlosimulations show that treatment effect estimates using EBCT display similar orlower bias and uniformly lower root mean squared error. These properties makeEBCT an attractive method for the evaluation of continuous treatments.