Measurement and Causal Identification in Constitutional Law; a Reply to Niels Petersen and Konstantin Chatziathanasiou

Petersen and Chatziathanasiou discuss recent empirical trends in constitutional law scholarship in a recent article and suggest that this body of research may be all “smoke and mirrors.” They focus on two well-documented issues with empirical research using observational data—omitted variable bias and measurement error—arguing that these problems are severe enough to call into question the entire body of research. But while pointing out real obstacles, they fail to account for how these issues have been addressed by existing literature. We make four points in reply. First, we explain that the scholarship is continuously developing new approaches to address the problems raised. Second, we describe efforts to address these problems. Third, we argue that the methodological pluralism that the article advocates is already underway. Finally, we suggest that evaluating an entire literature is best done using a “systematic review,” to ensure a fair characterization of the field evaluated.