Comparing Nominate and Ideal: Points of Difference and Monte Carlo Tests

Abstract

Empirical models of spatial voting allow legislators' locations in an abstract policy or ideological space to be inferred from their roll call votes. Over the past 25 years, these models have provided new insights about the US Congress and legislative behavior more generally (see, for example, Poole and Rosenthal, 1997). There are now a number of alternative models, estimators, and software that researchers can use to recover latent issue or ideological spaces from voting data. While these different estimators usually produce substantively similar estimates, important differences also arise. In this paper, we investigate the sources of observed differences between two leading methods, NOMINATE and IDEAL. Considering data from the 1994 to 1997 Supreme Court and the 109th Senate, we demonstrate that while some observed differences in the estimates produced by each model stem from fundamental differences in their underlying behavioral assumptions, others arise from arbitrary differences in implementation. Using Monte Carlo experiments, we find that neither model has a clear advantage over the other in the recovery of legislator locations or roll call midpoints in either large or small legislatures.

Keywords: NOMINATE, IDEAl, spatial voting, roll call votes, ideal points