The dynamics of committee outliers: Evidence from the house of representatives, 1951–90

In this study, we enter the on-going debate on committee outliers by addressing two important points. First, we dispute the conventional wisdom that committee outliers are best identified and evaluated using traditional tests of statistical significance. Contesting the appropriateness of such tests, we contend that any observed difference between chamber behaviour and floor behaviour may be substantively significant and use raw mean differences between Americans for Democratic Action (ADA) scores to examine differences between the full House of Representatives and its committees. Second, we attempt to integrate the committee outlier literature into the broader literature on the historical development of the House by using data from 20 Congresses over four decades (1951–90). This longitudinal analysis indicates that there have been interesting dynamic changes in the prevalence of committee outliers across time and committee types in patterns that are consistent with both the principal-agent framework and the more general literature on congressional change and reform. Many of these important changes are not obvious when relying solely on traditional tests of statistical significance to examine committee outliers.