Estimating the Impact of State Policies and Institutions with Mixed-Level Data

Abstract

Researchers are often interested in the effects of state policies and institutions on individual behavior or other outcomes in sub-state-level observational units, such as election results in state legislative districts. In this article, we examine the issue of clustered data in state and local politics research and the analytical problems it can cause. Standard estimation methods applied in most regression models do not properly account for the clustering of observations within states, leading analysts to overstate the statistical significance of coefficient estimates, especially of state-level factors. We discuss the theory behind two approaches for dealing with clustering—clustered standard errors and multilevel modeling—and argue that calculating clustered standard errors is a more straightforward and practical approach, especially when working with large datasets or many cross-level interactions. We demonstrate the relevance of this topic by replicating a recent study of the effects of state post-registration laws on voter turnout (Wolfinger, Highton, and Mullin 2005).