1999/03/01 by Tim Groseclose, Steven D. Levitt, James M. Snyder · 353 citations
Business, Management and Accounting · Mathematics · Psychology · Social Sciences · #Affect (linguistics) #Computer science #Econometrics #Economics #Electoral Systems and Political Participation #Index (typography) #Inflation (cosmology) #Judicial and Constitutional Studies #Law #Legislature #Mathematics #Nonparametric statistics #Political Influence and Corporate Strategies #Political science #Psychology #Set (abstract data type) #Statistics #Test (biology)
paper · doi:10.2307/2585759
published in American Political Science Review 93(1), 33-50 (Cambridge University Press)
openalex publication_date 1999/03/01 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/25
Interest group ratings are widely used in studies of legislative behavior. Since the set of votes used is not constant over time and across chambers, the scales underlying the scores can shift and stretch. We introduce an econometric model that corrects the problem. Specifically, we derive an index, much like an inflation index for consumer prices, that allows one to make intertemporal and interchamber comparisons of interest group ratings. The adjusted scores for the ADA show a strong liberal trend in the average member of Congress during 1947–94, followed by a conservative reversal. A nonparametric test using ADA and ACU scores demonstrates the validity of adjusted scores and the invalidity of nominal scores for intertemporal and interchamber comparisons. Using two studies (Levitt 1996; Shipan and Lowry 1997) we illustrate that the choice of adjusted versus nominal scores may greatly affect substantive conclusions of researchers.