Does Party Matter? An Historical Test Using Senate Tariff Votes in Three Institutional Settings

This article investigates the historical impact of party and constituency preferences on tariff votes from the U.S. Senate over the period 1883--1930. We find that the estimated effect of party grows during periods in which legislative institutions favored strong parties. We conclude that party has a causal effect on policy. If party serves solely as a proxy for unmeasured components of personal ideology or constituency preferences, then the estimated effect of party on policy outcomes should not vary contemporaneously with changes in legislative institutions. But if party has an independent causal impact on policy outcomes, then changes in institutions favoring strong parties should lead to a greater effect of party on voting behavior, holding constituency preferences constant. Although our findings are limited to votes over tariffs in the 19th and early 20th centuries, they suggest that further research into the mechanism by which party affects political decision making is important. Copyright 2002, Oxford University Press.