Contamination Bias in Linear Regressions

• 成果类型: Article
• 署名作者: Goldsmith-pinkham, Paul; Hull, Peter; Kolesar, Michal
• 署名单位: Yale University; Brown University; Princeton University
• 刊物名称: AMERICAN ECONOMIC REVIEW
• ISSN/ISSBN: 0002-8282
• DOI: 10.1257/aer.20221116
• 发表日期: 2024
• 页码: 4015-4051
• 关键词: instrumental variables; health-insurance; causal; identification; disparities; assignment; inference; education; demand; police

摘要

We study regressions with multiple treatments and a set of controls that is flexible enough to purge omitted variable bias. We show these regressions generally fail to estimate convex averages of heterogeneous treatment effects-instead, estimates of each treatment's effect are contaminated by nonconvex averages of the effects of other treatments. We discuss three estimation approaches that avoid such contamination bias, including the targeting of easiest-to-estimate weighted average effects. A reanalysis of nine empirical applications finds economically and statistically meaningful contamination bias in observational studies; contamination bias in experimental studies is more limited due to smaller variability in propensity scores. ( JEL C21, C31, C51, H75, I21, I28)