Description of Work

        The advantage incumbent members of Congress hold over their opponents in campaigns for office has steadily grown over the past five decades. While students of congressional politics have analyzed the effect of this advantage upon members' behavior in office, little is known of the campaigns that are run against them. Conventional theories of electoral competition assume that candidates' policy positions determine the number of votes these candidates receive and the chance these candidates have of winning an election, yet these models find scant support in congressional election politics. In these races, we frequently see underfinanced, obscure challengers to sitting members of the House. In this book, I describe the campaigns waged against sitting members of Congress. In these campaigns, a challenger has little to gain from an election-oriented strategy. In the majority of these generally uncompetitive races, I argue, a new type of politics is emerging - a politics of expressive campaigning, where challengers seek to use their campaigns as a platform for their own views and as a means for helping their party to achieve goals other than winning the election at hand.

        In this book I first present a rational choice theory of expressive campaigning, in which candidates and their parties assess the pre-existing advantage incumbents hold before entering a race. The strength of the incumbent determines whether candidates will seek to adopt issue positions popular with the general electorate, or whether they will adopt relatively unpopular, more overtly partisan positions in order to energize the party's core members and, perhaps, increase turnout and assist other candidates on the ballot.

        I test this theory in two ways. First, I compare incumbent and challenger positions using data on the issue positions of all 1996 candidates for the House of Representatives. These tests demonstrate that most congressional challengers fail to run campaigns aimed at appealing to the majority of voters in their districts. They demonstrate that congressional incumbents and challengers adopt strategies with a sharply different logic, but they do not demonstrate why this is the case. I thus test the theory further drawing upon a sample of interviews I conducted with nonincumbent candidates for the House of Representatives in the 1996 and 2000 elections. I analyze candidates' discussions of their campaigns to determine whether they did have accurate information about their probability of winning and of the preferences of the electorate; how this information influenced the issue positions they took and emphasized; and what role their party had in influencing these positions. I also provide case studies of several congressional races and the challengers' perspectives on these races. Results from these interviews supports my original theory by demonstrating that the issues candidates present are primarily determined by their expectations of victory - that voters are offered a true choice between electoral platforms primarily where that choice does not matter, where the outcome is preordained.

        This research makes two contributions to existing political science literature. On a theoretical level, it argues for a reconceptualization of the motives of candidates and parties in rational choice analysis. On a practical level, it seeks to enrich our understanding of the role challengers play in American elections and of why different types of challengers emerge in different types of elections. I argue that the role of challengers in the American electoral process can only be understood if we broaden our theories about rational candidate behavior.

1. Introduction: The Incumbency Advantage and Issue Competition in Congressional Elections

Many, perhaps most, American congressional elections feature a well-known incumbent running against a virtually unknown challenger. I explore the incumbency advantage in American congressional elections and the way in which that advantage discourages individuals who share the same level of political experience, knowledge, and ambition as the incumbent from challenging congressional incumbents. The incumbency advantage has three major components: legislative districts are rarely drawn to facilitate true two-party competition; members of congress have significantly greater resources, both financial and informational, than do nonincumbent candidates; and voters are rarely attentive enough to learn more than very basic information about their representative or to learn anything at all about nonincumbent candidates. These distortions, I argue, invert the logic used in our theories of how candidates run for office. Because issue positions are only infrequently major determinants of electoral success in such elections, they are best seen as a response of candidates to their perceived chances of electoral success, rather than as a major determinant of that success.

2. The Rational Candidate and the Hopeless Cause

The models we have traditionally used to explain issue competition in elections, particularly those developed in the 1950s and 1960s, are in many ways creatures of their political era. I trace the debate over political parties of the 1950s, in which normative desires for political parties to differentiate themselves from each other preceded Downs' attempt to show that such a scenario was unlikely to transpire in a two-party system. Downs' model is also a creature of the pluralist tone of 1950s political science; it presumes a relatively balanced type of political competition in which results - positions in office or other concrete benefits - are easily measurable. Pluralist theories themselves came under attack in later years, but the major assumptions of spatial models have not been contested. I draw upon the theories of critics of pluralism, as well as theories of "expressive," or noninstrumental preferences, to develop a theory of campaigning in which candidates advocate policy positions for reasons other than short-term electoral success. The utility gained by these candidates, as well as that gained by their fellow partisans and their party, can lead to candidacies that diverge significantly from the traditional median voter model, yet are fully rational. I delineate circumstances in which these types of campaigns occur and a logic for recognizing these campaigns.

3. Testing the Theory

I test the above theory using the Time/Congressional Quarterly survey of issue positions taken by candidates in the 1996 House of Representatives elections. I find that challengers and incumbents do not take positions according to the same logic, but that the data cannot further distinguish between the models. I do, however, demonstrate that strong political parties can induce challengers to take positions more strategically in competitive races that do weak parties. I discuss differences in strategic position taking by party and by level of competitiveness. Finally, I discuss limitations of the data set and ways in which open-ended interviews with a subset of the candidates covered by the data set can draw out some of the patterns I identify in the issue position data.

4. Talking With Congressional Challengers about their Campaigns

I discuss the method by which I interviewed Congressional candidates following the 1996 election, the ways in which these interviews address the limitations of the quantitative data discussed in Chapter Three, and the ways in which interview data can provide support for my theory of expressive campaigning. I describe my interview sample and provide background on the politics of the four states - Illinois, Minnesota, Ohio, and Wisconsin - in which I interviewed candidates.

5. What Plays in Peoria: Candidate Assessments of Public Opinion

The theory presented here of candidate competitions contain three major variables: candidates' probability of winning, their issue positions, and the effect of their party upon their positions. In this chapter, I present results of interviews with a sample of nonincumbent congressional candidates about their means of acquiring information about their probability of winning and about how the preferences of voters will affect their ability to win. I explore the means through which candidates acquire such information, the perspectives they have about opinion polling and other means of gathering information about voter preferences, and I assess the accuracy of the information they do possess. I conclude that disparities between candidates who use scientific methods of public opinion assessment and those who do not are practically insignificant because few candidates look upon opinion polling as a means of gathering information; instead, they look upon it as a tool for influencing the media, campaign contributors, or voters.
 

Having established that disparities in information are not a cause of divergent positions taken by candidates, I ask the same sample of candidates their reasons for taking the positions they took in their campaigns. I find support for the notion that candidates take positions dependent upon their beliefs about their chance of winning; candidates who believe they will be competitive take positions designed to appeal to the majority of voters, while those who do not believe they have a chance of running a competitive campaign refer to their ideological beliefs, not public opinion, in justifying their election positions. I also find that candidates alter their positions when they see their chances of winning increase.

7. You Don't Know Me, But Here I Am: The Role of Parties in Congressional Elections

I analyze the influence of political parties upon their candidates' positions in order to discover whether parties encourage or discourage expressive position-taking by candidates. I conclude that the level of party influence over candidate positions depends on the strength and organization of their political party in their district or state; there are both strong and weak parties in the sample. I present the candidates' reflections upon whether their party attempted to alter the positions they would have otherwise taken. To do so, I also explore candidates' perceptions of who the major forces in their party are. I find that attempts to enforce party cohesion occur primarily in regions of the country where political parties are weak organizationally. That is, party strength in an organizational sense has little to do with strength, or coherence, in an ideological sense. In such areas, party activists may prefer ideological purity over maximizing the party's vote share or chance of winning. While such activities probably did not have a significant effect upon the vote shares of individual candidates in my sample, they may do so indirectly by influencing primary results or candidate selection.

8. Expressive Campaigning in the 2000 Congressional Elections

As a validity check on both the geographical distinctiveness and the temporal distinctiveness - i.e. peculiarities of the 1996 election described in Chapters 4, 5, 6, and 7, I present comments from a second round of interviews conducted with candidates in the 2000 House elections in four different states - Delaware, Maryland, New Jersey, Pennsylvania, and Virginia.
 

I conclude by discussing the significance of candidate divergence for American democracy. Arguments for candidate convergence and arguments for candidate divergence are still a major thread in political discourse; some political observers complain that too often the two major parties are indistinguishable on policy issues, while at the same time candidates or politicians are criticized for taking positions that do not reflect the views of a majority of citizens. I argue that as long as the incumbency advantage persists at present levels, it is unlikely that candidates in congressional elections will converge on policy. Few nonincumbent candidates will follow a median voter scenario because their chances of winning will be so circumscribed before they begin their campaign. Instead, they will act in order to satisfy their own desire for political expression or the desire of party leaders to express views held by members of the party. By giving voice to minorities within their districts, however, nonincumbent candidates may be helping to maintain political activism among the politically disenfranchised and to maintain their party's organization in their district so as to capitalize on future changes in voter preferences.
 

Estimated Word Count of Project
 

10,000 words (approximately 250 trade-size pages)
35 illustrations - tables, graphs
 

Expected Audience

a) American politics scholars - elections, congress, political parties, public opinion
b) Game theory and rational choice scholars
c) Some sections would be appropriate for classroom use in advanced undergraduate or graduate courses on Congress,
        political parties, or elections
 

Related Works

This book is similar in approach to several qualitative studies on congress and elections, including

- Paul Herrnson, Congressional Elections (Congressional Quarterly, 1995)
- David Canon, Actors, Athletes, and Astronauts (University of Chicago, 1990)
- Richard Fenno, Home Style (Little, Brown, 1978)
- Jonathan Krasno, Challengers, Competition, and Re-election (Yale, 1994)
- John Kingdon, Candidates for Office: Beliefs and Strategies (Random House, 1966)
- Sandy Maisel and Walter Stone, Candidate Emergence Project (forthcoming)

In its methodological approach, it is similar to several types of rational choice works and interview-based methods of theoretical testing, including

- Joseph Schlesinger, Political Parties and the Winning of Office (Michigan, 1994)
- Susan Herbst, Reading Public Opinion (University of Chicago, 1998)
- Benjamin Page, Choices and Echoes in Presidential Elections (University of Chicago, 1978)
 

It differs from the above works because of its blend of rational choice theory and empirical work on Congressional elections. This work also will draw upon a far more extensive set of interviews with congressional candidates than has been the norm in other works, where either closed-ended, brief surveys have been used or where the number of open-ended interviews has been far smaller.
 

Schedule for Completion

All chapters currently exist in draft form. The expected completion date for the manuscript is February of 2002.