What To Do (and Not to Do) with Time-Series Cross-Section Data

TLDR

The study critiques time‑series cross‑section estimation methods, questioning conclusions in comparative political economy research. An alternative, panel‑corrected standard‑error estimator is proposed to address complex error structures in such models. The Parks GLS method underestimates variability by over 50%, while the proposed panel‑corrected standard errors perform well in simulations and improve inference in a reanalysis of a social democratic corporatist model.

Abstract

We examine some issues in the estimation of time-series cross-section models, calling into question the conclusions of many published studies, particularly in the field of comparative political economy. We show that the generalized least squares approach of Parks produces standard errors that lead to extreme overconfidence, often underestimating variability by 50% or more. We also provide an alternative estimator of the standard errors that is correct when the error structures show complications found in this type of model. Monte Carlo analysis shows that these “panel-corrected standard errors” perform well. The utility of our approach is demonstrated via a reanalysis of one “social democratic corporatist” model.