An Institutional View of Congressional Elections: The Impact of Congressional Image on Seat Change in the House

Article excerpt

The literature on seat change in U.S. House elections abounds with explanations regarding the factors contributing to the biennial change in the partisan balance of the body. While a number of theoritically and empirically appealing models have been presented, many base their explanations around presidential politics and a variety of factors independent of Congress. In this article, I argue that in developing models of congressional seat change, it is worthwhile to consider how the public image of the institution impacts the electoral success of its members. I describe and test a model that captures the influence of the public's perception of Congress on party seat change. Encompassing tests suggest that this framework significantly improves upon existing models. The results underscores the importance of endogenous, Congress-specific factors in explaining aggregate seat change in the U.S. House of Representative.

The aggregate results of recent congressional elections have triggered a large-scale re-analysis of the viability of most of the prominent models of seat change in the House. In particular, the surprising outcomes in 1994 and f998 have led researchers to question whether fundamental aspects of the pre-existing structure of congressional elections have changed or disappeared altogether.1 This enterprise has resulted in many of the existing explanations being adapted to capture certain aspects of contemporary House elections, such as the importance of incumbency and the distribution of open seats (Gaddie 1997; Newman and Ostrom 2002), declining coattails (Campbell and Jurek 2003), and an apparent realignment among voters (Campbell 1997a). The revised models have thus enhanced the predictive success of the existing explanations of the factors contributing to party seat change in Congress.

While research has begun to address the absence of such dynamics from existing models, the motivation of this analysis is a persistent theoretical and empirical absence in the literature of the institutional role of Congress on the electoral fortunes of members. Despite the observations described above that document the varied ways in which contemporary elections are contested, there has been virtually no discussion of how the public perception of Congress impacts congressional elections. However, in an electoral environment in which members, their parties, and the institution more broadly have received increasing amounts of attention, it seems fitting to assess the potential electoral impact of Congress-centered forces. In fact, in light of the degree of consideration students of Congress have given to the role that members' behavior plays in impacting their electoral fortunes, it is somewhat surprising that such factors have not been considered in modeling seal change in the House.2

Of course, this is not to say that the importance of the exogenous factors offered in the seat change literature should be discounted. The role of extra-legislative political influences such as economic conditions and presidential vote share is well established.3 This article presents a theory and an empirical model thatseek to integrate the findings of previous research with a more fundamental understanding of the role Congress plays in its own electoral environment and how members of Congress may be affected in their quest for reelection.

INSTITUTIONAL CONSIDERATIONS

It is generally agreed that congressional elections are seasoned by factors such as the condition of the economy, the president's level of approval, and the balance of party seats in the House. At the same time, it seems plausible to expect the prevailing image of Congress at the time of an election to play a role in influencing the results. Presumably, few would argue that members of a Congress that is said to be out of touch, corrupt, or overly partisan would not be portrayed by the press and perceived by the public in a negative light and subsequently penalized at the polls. …