Why candidate divergence should be expected to be just as great (or even greater) in competitive seats as in non-competitive ones

Basic Downsian theory predicts candidate convergence toward the views of the median voter in two-candidate elections. Common journalistic wisdom, moreover, leads us to expect these centripetal pressures to be strongest when elections are expected to be close. Yet, the available evidence from the US Congress disconfirms this prediction. To explain this counterintuitive result, we develop a spatial model that allows us to understand the complex interactions of political competition, partisan loyalties, and incentives for voter turnout that can lead office-seeking candidates, especially candidates in close elections, to emphasize policy appeals to their voter base rather than courting the median voter.     

Policy-Motivated Candidates,Noisy Platforms,and Non-Robustness

This paper develops a model of a two-candidate election inwhich the candidates are mainly office-motivated but also tosome (arbitrarily small) extent policy-motivated, and theirchosen platforms are to some (arbitrarily small) extent noisy.The platforms' being noisy means that if a candidate haschosen a particular platform, the voter's perception is thatshe has, with positive probability, actually chosen some otherplatform. It is shown that (i) an equilibrium in which thecandidates play pure exists whether or not there is aCondorcet winner among the policy alternatives, and (ii) inthis equilibrium the candidates choose their own favoriteplatforms, which means that the platforms do not converge.     

Mixed equilibriums in a three-candidate spatial model with candidate valence

We study a spatial model of electoral competition among three office-motivated candidates of unequal valence (one advantaged and two equally disadvantaged candidates) under majority rule assuming that candidates are uncertain about the voters’ policy preferences and that the policy space consists of three alternatives (one at each extreme of the linear policy spectrum and one in the center) and we characterize mixed strategy Nash equilibriums of the game. Counterintuitively, we show that (a) when uncertainty about voters’ preferences is high, the advantaged candidate might choose in equilibrium a more extremist strategy than the disadvantaged candidates and that (b) when uncertainty about voters’ preferences is low, there exist equilibriums in which one of the disadvantaged candidates has a larger probability of election than the disadvantaged candidate of the equivalent two-candidate (one advantaged and one disadvantaged candidate) case.     

Policy divergence in multicandidate probabilistic spatial voting

Existing models of multicandidate spatial competition with probabilistic voting typically predict a high degree of policy convergence, yet in actual elections candidates advocate quite divergent sets of policies. What accounts for this disparity between theory and empirical observation? I introduce two variations on the basic probabilistic vote model which may account for candidate policy divergence: 1) a model which incorporates candidate-specific variables, so that candidates may enjoy nonpolicy-related electoral advantages (or disadvantages); 2) a model which allows nonzero correlations between the random terms associated with voters' candidate utilities, thereby capturing situations where voters view two or more candidates as similar on nonpolicy grounds. I report candidate equilibrium analyses for each model, which show far greater policy divergence than exists under the standard probabilistic vote model. I then analyze the strategic logic which underlies these results.

     

The elements of candidate reputation: The effect of record and credibility on optimal spatial location

We build a model of two-candidate elections in which voters judge candidates on the basis of how well their announced campaign positions correspond with their records. Given different records, the candidates will adopt different campaign positions. Two types of reputational advantage are analyzed: proximity of the candidate's record to the median voter's ideal point, and the range of campaign positions that a candidate may adopt and still retain some credibility with the voters.An earlier version of this paper was delivered at the annual meetings of the Public Choice Society, Tucson, Arizona, 17 March 1990. The authors thank William Dougan and Brian Roberts for some helpful comments in the formative stages of this paper, and express our gratitude to Jay Dow for providing admirable research assistance. We also wish to thank Henry Chappell and William Keech for their discussant comments at the Public Choice meetings.     

Candidates’ policy strategies in primary elections: does strategic voting by the primary electorate matter?

Empirical research reports conflicting conclusions about whether primary election voters strategically account for candidates’ general election prospects when casting their votes. We model the strategic calculations of office-seeking candidates facing two-stage elections beginning with a primary, and we compare candidates’ policy strategies in situations where primary voters strategically support the most viable general election candidate against candidate strategies when voters expressively support their preferred primary candidate regardless of electability. Our analyses—in which the candidates’ appeal is based on their policy positions and their campaigning skills—suggest a surprising conclusion: namely, that strategic and expressive primary voting typically support identical equilibrium configurations in candidate strategies. Our conclusions are relevant to candidates facing contested primaries, and also to political parties facing the strategic decision about whether or not to use primary elections to select their candidates—a common dilemma for Latin American (and some European) parties.     

Proportional Representation,Majoritarian Legislatures,and Coalitional Voting

Voters in elections under plurality rule face relatively straightforward incentives. In proportional representation systems, voters face more complex incentives as electoral outcomes don’t translate as directly into policy outcomes as in plurality rule elections. A common approach is to assume electoral outcomes translate into policy as a vote‐weighted average of all party platforms. However, most of the world’s legislatures are majoritarian institutions, and elections in PR systems are generally followed by a process of coalition formation. Results obtained using this assumption are not robust to the introduction of even minimal forms of majoritarianism. Incentives to engage in strategic voting depend on considerations about the coalitions that may form after the election, and the voters’ equilibrium strategies are shaped by policy balancing and the postelectoral coalition bargaining situation, including considerations about who will be appointed the formateur.     

Extreme idealism and equilibrium in the Hotelling–Downs model of political competition

In the classic Hotelling–Downs model of political competition, no pure strategy equilibrium with three or more strategic candidates exists when the distribution of voters’ preferred policies is unimodal. I study the effect of introducing two idealist candidates to the model who are non-strategic (i.e., fixed to their policy platforms), while allowing for an unlimited number of strategic candidates. Doing so, I show that equilibrium is restored for a non-degenerate set of unimodal distributions. In addition, the equilibria have the following features: (1) the left-most and right-most candidates (i.e., extremists) are idealists; (2) strategic candidates never share their policy platforms, which instead are spread out across the policy space; and (3) if more than one strategic candidate enters, the distribution of voter preferences must be asymmetric. I also show that equilibria can accommodate idealist fringes of candidates toward the extremes of the political spectrum.     

The political consequences of alienation-based and indifference-based voter abstention: Applications to Presidential Elections

We present a unified model of turnout and vote choice that incorporates two distinct motivations for citizens to abstain from voting: alienation from the candidates, and indifference between the candidates. Empirically, we find that alienation and indifference each motivated significant amounts of voter abstention in the 1980–1988 U.S. presidential elections. Using model-based computer simulations—which permit us to manipulate factors affecting turnout—we show that distinguishing between alienation and indifference illuminates three controversies in elections research. First, we find that abstention because of either alienation or indifference benefited Republican candidates, but only very modestly. Second, presidential elections involving attractive candidates motivate higher turnout, but only to the extent that abstention stems from alienation rather than from indifference. Third, paradoxically, citizens’ individual-level tendencies to abstain because of alienation are strongly affected by their evaluations of the candidates’ policies, whereas aggregate turnout rates do not depend significantly on the candidates’ policy platforms.     

Candidate Valence and Ideological Positions in U.S. House Elections

We examine the relationship between the valence qualities of candidates and the ideological positions they take in U.S. House elections based on a study of the 2006 midterm elections. Our design enables us to distinguish between campaign and character dimensions of candidate valence and to place candidates and districts on the same ideological scale. Incumbents with a personal‐character advantage are closer ideologically to their district preferences, while disadvantaged challengers take more extreme policy positions. Contrary to conventional wisdom, challengers can reap electoral rewards by taking more extreme positions relative to their districts. We explore a possible mechanism for this extremism effect by demonstrating that challengers closer to the extreme received greater financial contributions, which enhanced their chances of victory. Our results bear on theories of representation that include policy and valence, although the interactions between these two dimensions may be complex and counterintuitive.     

Candidate and Party Strategies in Two-Stage Elections Beginning with a Primary

In the United States and Latin America, candidates for national and state-level office frequently must win primary elections in order to advance to the general election. We model policy and valence issues for office-seeking candidates facing such two-stage elections. We determine a Nash equilibrium for the candidates' optimal strategies, and we find that holding a primary is likely to increase a party's chances of winning the general election, particularly in situations where valence issues that involve the candidates' campaigning skills and that are not known prior to the campaign are more salient than policy issues. Furthermore, we find that primary elections are especially likely to benefit parties that expect to be underdogs in the general election. Our conclusions are directly relevant to U.S. politics and by extension to the strategic decisions that many Latin American parties currently confront, about whether it is strategically desirable to hold primaries.     

Campaign contributions as valence

I study how the possibility of voters contributing to candidates in response to the candidates’ policy proposals affects the equilibrium policy in winner-take-all political competition. More specifically, I allow each partisan voter to contribute to her preferred candidate where the amount contributed depends on the relative positions of the two candidates. Candidates then use the donations to build valence through campaigning, which in turn affects the voting decision of impressionable voters. Since candidates are concerned with raising money as well as picking a policy position preferred by the median voter, I show that campaign contributions may lead to divergent equilibria in winner-take-all elections when politicians are policy-motivated, albeit only under stylized utility functions and donor densities. Further, under symmetric voter and donor densities, if either the donor density is single-peaked or the voter utility is concave, a unique equilibrium exists in which both candidates propose the ideal policy of the median voter.     

Spatial voting in non-partisan cities: A case study

In spatial voting theory, voters choose the candidate whose policy preferences are most like their own. This requires that (a) voters and candidates have policy preferences that can be meaningfully summarized in terms of low-dimensional “ideal points” on a left-right scale; (b) voters are able to discern, either directly or through relevant cues, the ideal points of the candidates who are running for office; and (c) voters incorporate this information into the choices they make at the ballot box. Perhaps more than in any other elections, it is not clear that any of these requirements are met in non-partisan municipal elections: policy preferences may not be ideologically structured, information may be inadequate, and voters may choose candidates for reasons other than ideology. This makes non-partisan municipal elections an especially hard test for spatial voting theory. Using novel data from both municipal candidates and eligible voters in a major non-partisan municipal election in Canada, we show that municipal policy attitudes are ideologically structured and that these municipal policy ideal points are strongly related to mayoral and council vote choice. Thus, despite the institutional and informational obstacles, spatial voting can play an important role in non-partisan municipal elections.     

Multi-candidate equilibria

We demonstrate the existence of an electoral equilibrium in a model with three or more candidates. We use the same kinds of assumptions that have previously been used to establish existence in two-candidate models and have not had to make special assumptions regarding dimensionality (e.g., that there is only one dimension) or distribution of voter preferences (e.g., symmetry).     

Policy competition in the 2002 French legislative and presidential elections

Abstract.  The French two-round system of presidential elections forces candidates to choose strategies designed to maximize their votes in two different, potentially conflicting strategic contexts: a first round contest between many candidates, and a second round between (typically) a left- and a right-oriented candidate. Following a constitutional change in 2000, furthermore, presidential elections are synchronized with legislative elections, more tightly linking presidential candidates to the policy platforms of the parties they represent. This article examines the consequences of policy positioning by presidential candidates, measuring, comparing and assessing positioning in the legislative elections and in the first and second presidential election rounds. The measures come from an expert survey taken in 2002, from content analysis of party manifestos and presidential speeches, and from the 2002 French National Election Survey. The findings provide hard empirical confirmation of two commonly perceived propositions: first, that Jospin's first-round loss resulted from strategic error in moving too close to the policy centre, and second, that Chirac's won an overwhelming second-round victory because he collected all of the voters from candidates eliminated in the first round.     

Issues and the presidential primary voter

Most agree that voting in presidential general elections is largely contingent on the evaluations of the candidates, issues, and parties. Yet inpresidential primary elections the determinants of voter choices are less clear. Partisanship is inconsequential, information about candidate personalities and policy positions is scarce, and a fourth factor, expectations, may influence voters. In this paper, we reconsider the influence ofpolitical issues in presidential primaries. We argue that past work has not adequately considered how issues matter in primary elections. Primaries are intraparty affairs, and the political issues that typically divide the parties are not very relevant in primaries. Instead, we focus on the policy issues each candidate chooses to emphasize in their quest for the nomination, which we call policy priorities. With data gathered about media coverage of the presidential contenders in the 1988 primaries, and using exit poll data from the 1988 Super Tuesday primaries, we show that issues, as policy priorities, do matter in presidential primary elections. This research also implies that primary campaigns matter, since information concerning the policy priorities of the candidates reaches the intended audience.An earlier version of this paper was presented at the Annual Meeting of the Midwest Political Science Association, Chicago, IL, April 1992.     

Presidential Coattails and Legislative Fragmentation

Considerable evidence suggests that legislative fragmentation can negatively affect the survival of democratic presidential regimes. While there is a vast literature examining the determinants of legislative fragmentation, one factor that has traditionally been overlooked is the impact of presidential elections. Do presidential elections increase or decrease legislative fragmentation? Does it matter if presidents are elected by plurality rule or by runoff? Using a new dataset that covers all democratic legislative and presidential elections between 1946 and 2000, I find that presidential coattails can reduce, increase, or have no effect on legislative fragmentation depending on the number of presidential candidates. I also find strong evidence that social heterogeneity increases the number of presidential candidates when runoff systems are employed. Taken together, these results suggest that the widespread adoption of runoffs by newly democratic presidential regimes will likely increase legislative fragmentation, thereby putting their democratic survival at increased risk.     

Blame,Responsibility, and the Tea Party in the 2010 Midterm Elections

There is a general consensus both in the news media and scholarly research that 2010 was a highly nationalized election year. Reports have indicated that anti-Obama sentiment, the Democrats’ legislative agenda, the economy, and the Tea Party were all factors contributing to Democratic losses in the congressional elections. In this paper, we use data from 2010 Cooperative Congressional Election Study to examine the individual-level dynamics that contributed to the heightened nationalization of the 2010 congressional elections. Our analysis shows that Tea Party support and the attribution of blame and responsibility by voters are essential to understanding the 2010 election outcome, beyond what we would expect from a simple referendum model of midterm elections. Not surprisingly, Tea Party supporters blamed Democrats for the state of national affairs, disapproved of the Democrats’ policy agenda, and overwhelmingly supported Republican candidates in the congressional elections. However, our analysis shows that not all voters who supported Republican candidates were driven by high levels of opposition to President Obama and the Democrats. Another key group of voters blamed both Democrats and Republicans for the nation’s problems but ultimately held Democrats responsible in the voting booth by supporting Republican congressional candidates.     

Equilibrium in multicandidate probabilistic spatial voting

This paper presents a multicandidate spatial model of probabilistic voting in which voter utility functions contain a random element specific to each candidate. The model assumes no abstentions, sincere voting, and the maximization of expected vote by each candidate. We derive a sufficient condition for concavity of the candidate expected vote function with which the existence of equilibrium is related to the degree of voter uncertainty. We show that, under concavity, convergent equilibrium exists at a “minimum-sum point” at which total distances from all voter ideal points are minimized. We then discuss the location of convergent equilibrium for various measures of distance. In our examples, computer analysis indicates that non-convergent equilibria are only locally stable and disappear as voter uncertainty increases.     

The impact of candidate name order on election outcomes in North Dakota

A number of studies have explored the possibility that the ordering of candidates' names on the ballot might influence how those candidates perform on election day. Strong evidence of an order effect comes from investigations of election returns in states that implemented quasi-random assignment of candidate name orders to voters. Although most such studies have identified benefits for earlier-listed candidates, much of the evidence comes from a limited set of elections in only a handful of states. This paper expands our understanding of order effects to 31 general elections held in North Dakota between 2000 and 2006; these include all state-wide races involving 2 candidates. A primacy effect appeared in 80% of the contests. The first ballot position reaped the largest benefits in non-partisan contests and in presidential election years. These findings are consistent with earlier studies from other states and provide evidence in line with proposals that a lack of information and ambivalence underlie candidate name order effects.