Inference for Regression with Variables Generated from Unstructured Data

"Inference for Regression with Variables Generated from Unstructured Data" (joint work with Laura Battaglia, Timothy Christensen, and Szymon Sacher)

SPEAKER: Stephen Hansen (University College London)

ABSTRACT:

The leading strategy for analyzing unstructured data uses two steps. First, latent variables of economic interest are estimated with an upstream information retrieval model. Second, the estimates are treated as “data” in a downstream econometric model. We establish theoretical arguments for why this two-step strategy leads to biased inference in empirically plausible settings. More constructively, we propose a one-step strategy for valid inference that uses the upstream and downstream models jointly. The one-step strategy (i) substantially reduces bias in simulations; (ii) has quantitatively important effects in a leading application using CEO time-use data; and (iii) can be readily adapted by applied researchers.

Link to the paper: https://sekhansen.github.io/pdf_files/reg_unstruct.pdf

BIO:

Stephen Hansen is a Professor of Economics at University College London. His current research uses unstructured data to build new measures of economic activity and behavior across a variety of applications, most often related to organizational economics and monetary policy. His research is supported by an ERC Consolidator Grant.