Correlation Robust Inference When Economic Distance Is Latent

This lecture introduces Ulrich Müller and his research results about economic distances.

Since economic distance between observations is usually unknown, the authors use auxiliary variables that reflect the same underlying structure of economic distance. Based on this information, they develop new covariance matrices and a statistical test that provides reliable results even in the presence of general weak dependence.

In an application to U.S. socioeconomic data at the zip-code level, the new standard errors are substantially larger than those produced by existing methods, including methods that account for geographic distance.

Ulrich Müller is a Professor of Economics at Princeton University. He received his PhD at the University of St. Gallen in 2002 and has been at Princeton since 2003. His research is in econometric theory with interests in time series and forecasting, cluster dependence, spatial dependence, Bayesian inference and extreme value theory.