Description

A widespread threat to the validity of standard policy evaluation tools is the presence of spillovers between treated and untreated groups. Economic interactions across units of analysis—due to the flow of goods, factors, and payments to and from the government, for instance—result in bias in standard estimates of objects of interest such as the average treatment effect or the total effect of a program. In this paper, we develop a suite of approaches that can enable researchers to use theory and data about economic flows and distortions in order to overcome this bias. We apply this methodology to estimate the effects of a large earthquake that struck Chile in 2010.