Bounding Omitted Variable Bias Using Auxiliary Data with an Application to Estimate Neighborhood Effects

Abstract

This paper proposes a new estimator that bounds omitted variable bias using proxies for omitted variables with an asymptotically valid bootstrap procedure. The proxies do not need to appear in the same dataset as the outcome variable and the estimator is robust to measurement errors in proxies. I provide the open-source software to implement the estimator and its confidence interval. Next, I illustrate the application in the context of estimating neighborhood effect on intergenerational cultural transmission and find that growing up in an ethnic enclave could change some adulthood outcomes for second-generation immigrants that reflect their extent of cultural assimilation.

Keywords: Omitted Variable Bias, Two-Sample Least Squares, Data Combination, Data Fusion, Auxiliary Data, Proxy Variable, Intergenerational Cultural Transmission, Cultural Assimilation, Neighborhood Effect

JEL Classification: C01, C80, J01, J15, J16, J61, Z10, Z13