Jay K. Dow

Multinomial probit and multinomial logit: a comparison of choice models for voting research

Abstract Several recent studies of voter choice in multiparty elections point to the advantages of multinomial probit (MNP) relative to multinomial/conditional logit (MNL). We compare the MNP and MNL models and argue that the simpler logit is often preferable to the more complex probit for the study of voter choice in multi-party elections. Our argument rests on three areas of comparison between MNP and MNL. First, within the limits of typical data—a small sample of revealed voter choices among a few candidates or parties—neither model will clearly appear to have generated the observed data. Second, MNP is susceptible to a number of estimation problems, the most serious of which is that the MNP is often weakly identified in application. Weak identification is difficult to diagnose and may lead to plausible, yet arbitrary or misleading inferences. Finally, the logit model is criticized because it imposes the independence of irrelevant alternatives (IIA) property on voter choice. For most applications the IIA property is neither relevant nor particularly restrictive. We illustrate our arguments using data from recent US and French presidential elections.

A comparative spatial analysis of majoritarian and proportional elections

This study estimates spatial representations of recent elections in Canada, France, The Netherlands and Israel. Its purpose is to test whether there exist systematic differences in the extent of spatial dispersion among parties and candidates in the majoritarian and proportional electoral systems. Canada and France are majoritarian systems, while The Netherlands and Israel are highly proportional. The study uses a measure of central tendency developed by Kollman; Kollman and Kollman and non-parametric statistical tests to compare the relative dispersion of parties and candidates across the maps. The analysis reveals that parties and candidates in the majoritarian systems are located significantly closer to the center of the distribution of voters than those in proportional systems. The estimated spatial maps also provide information useful for interpreting the bases of electoral politics in each country.

Party-System Extremism in Majoritarian and Proportional Electoral Systems

This study evaluates the extent of party-system extremism in thirty-one electoral democracies as a function of electoral-system proportionality. It uses data from the Comparative Studies of Electoral Systems project to estimate the extent of party-system compactness or dispersion across polities and to determine whether more proportional systems foster greater ideological divergence among parties. Electoral system characteristics most associated with party-system compactness in the ideological space are investigated. The empirics show that more proportional systems support greater ideological dispersion, while less proportional systems encourage parties to cluster nearer the centre of the electoral space. This finding is maintained in several sub-samples of national elections and does not depend on the inclusion of highly majoritarian systems (such as the United Kingdom).

Representation in Congressional Campaigns: Evidence for Discounting/Directional Voting in U.S. Senate Elections

Several recent studies suggest that voters may prefer candidates who propose policies that are similar to, but more extreme than, the voters’ sincere policy preferences. This may arise either because voters vote directionally based on the direction and intensity of candidates’ proposals or, alternatively, because voters recognize that elected officials face obstacles to implementing their policy agenda and therefore discount the candidates’ policy promises. Using data from the Pooled Senate Election Study, we evaluate the discounting/directional hypothesis versus the alternative proximity hypothesis, by conducting individual-level and aggregate-level analyses of voting in 95 Senate races held in 1988–90–92. Our results support the discounting/directional hypothesis, that voters reward candidates when they present distinctly noncentrist positions on the side of the issue (liberal or conservative) favored by their constituency. These findings have important implications for understanding voting behavior, policy representation, and candidate strategies in Senate elections.

Campaign Contributions and Legislative Voting in the California Assembly

In this article the authors test for the influence of campaign contributions from economic interest groups on legislative voting in the California Assembly. California does not restrict the size of contributions that special interests may provide to candidates for state office. Consequently, if the purpose of campaign contributions is to secure favorable voting on legislation, this should be evident in an unregulated setting. However, for both business and labor interests, neither the number nor the monetary amount of campaign contributions to incumbents are found to have a major influence on legislative roll call voting.

Directional and proximity models of voter choice in recent US presidential elections

This note evaluates relative ability of the proximity and recently proposed directional variants of the spatial model of voter choice to account for candidate evaluations in US presidential elections contested between 1980 and 1992. I do this by estimating a statistical model that represents voter preference for a candidate as a weighted average of proximity and directional components. The analysis corroborates previous studies supporting the directional model, but illustrate that these results are sensitive to statistical specification. Alternative methodological specifications favor a mixed directional-proximity model and the traditional distance representation.

The Industrial Structure of the California Assembly: Committee Assignments, Economic Interests, and Campaign Contributions

This paper examines the provision of campaign contributions made by economic interests to incumbents seeking reelection in the 1984, 1986, and 1988 California Assembly elections. The study tests whether the distribution of campaign contributions by specific industrial sectors corresponds to legislator possession of the policy property rights associated with membership on relevant Assembly standing committees. A non-parametric statistical model compares the distribution of campaign contributions from groups within a given industrial sector to members of relevant policy committees with the distribution of contributions from all other contributor classes. The empirical analysis confirms that committee assignments significantly affect the allocation of special interest resources in Assembly elections, substantiating the importance of institutional considerations in influencing special interest activity.

Voter Choice in the 1995 French Presidential Election

This study estimates a model of voter choice for the first and second ballots of the 1995 French Presidential election. Its objective is to characterize the relative importance of traditional bases of French voter choice such as social class relative to campaign issues and voter evaluations of the economy. The study also seeks to identify candidate-specific bases of voter choice. Multinomial logit analysis of voter choice among the four leading candidates reveals a stronger role for campaign issues than previous studies suggest. Voter-candidate ideological proximity is the strongest predictor of first and second ballot choice, followed by voter perceptions of candidate ability to address unemployment and political corruption. Demographic variables have limited explanatory power in both the first and second ballot models. The parameter estimates demonstrate considerable differences in the bases of voter choice across candidates, with the bases of voter choice for National Front leader Jean-Marie le Pen most distinct from those of the remaining candidates.

Political Knowledge and Electoral Choice in the 1992–2004 United States Presidential Elections: Are More and Less Informed Citizens Distinguishable?

This study estimates an individual-level model of turnout and vote choice in United States presidential elections. Extending a methodology proposed by Bartels (1996), the model allows one to assess whether the electoral choices of citizens would change if they were more knowledgeable about politics, and if such changes have implications for aggregate election results. Simulations of hypothetical electorates under different assumptions about the distribution of political knowledge show that while some citizens would change their votes if more knowledgeable, the primary effects of increasing voter knowledge is to raise turnout levels and to solidify preexisting vote tendencies. The few vote changes that result from increased political knowledge largely average out in aggregation. Increased turnout resulting from a more informed electorate, however, favors Democratic candidates in two of the four studied elections. [A] people who mean to be their own governors, must arm themselves with the power which knowledge gives.1 -James Madison