A Research Note on the Impact of Interparty Competition on Congressional Voting in a Competitive Era

Abstract

In the modern House of Representatives party competition for seats is not strongly correlated with or predictive of party voting behavior. The most plausible explanation for this weak relationship is that less than 20 per cent of House seats are competitive from one census to the next. In this study the Hasbrouck-Jones Fluidity Index was applied to House elections in the 1890—1900 period, and almost 50 per cent of House seats in this era were found to have been competitive at that time.From the 1892—1900 period the 55th and 56th U.S. Houses were selected for analysis. A varimax factor analysis was run on the party votes in these Congresses, and the resultant factor scores formed a unidimensional measure of the number of times a member voted with a party majority. A multiple regression analysis was run, with these party support scores used as the dependent variable and with party competition and occupational composition of the district used as the independent variables. The results showed that in both houses party competition was strongly correlated with and predictive of party support scores. The same analysis on the 89th House yielded much weaker correlations and showed little predictive ability.