Skip Navigation
UT wordmark
College of Liberal Arts wordmark
government masthead
Robert G. Moser, Chair BAT 2.116, Mailcode A1800, Austin, TX 78712 • 512-471-5121

How and When We Choose Our Presidents

Two recent books help demystify presidential elections

Posted: October 19, 2012

In Fall 2013, Christopher Wlezien will join the Department of Government as the Hogg Professor. Currently at Temple University, Wlezien recently published a book, coauthored with Robert S. Erikson of Columbia University, that seeks to explain the relevance of election campaigns to which candidate is ultimately elected president. “The Timeline of Presidential Elections: How Campaigns Do (And Do Not) Matter” answers some basic questions about presidential election campaigns, such as how much the electorate’s vote changes over the course of a campaign and when the most meaningful shifts in the vote occur.

Timeline of Presidential ElectionsWlezien and Erikson root their analysis in a comprehensive dataset including all available national polls from the 15 presidential elections between 1952 and 2008, as well as analyzing individual-level poll results. They find that the vote rarely ends up where it began, but on the whole the change is extremely gradual, and that sampling error in polls tends to mask the relative stability of the electorate’s preferences over the course of the campaign. Significant movement in the polls is often short-term change that fades quickly, but the slow evolution of vote change in response to campaign events leaves an impact that ultimately affects the Election Day outcome.

Beginning in early April, 200 days before the election, there are three major periods of vote reshuffle: the early stages of the primary season, when the nominees are being finalized; during and after the party national conventions; and during the final campaign week — during the final 60 days of the campaign very little change occurs, but even modest, short-term shifts matter a lot when they occur close to Election Day. During primary season, candidates come into focus and electoral preferences begin taking shape. During the national conventions voters are especially attentive, and many voters shift their preference following the conventions, with a decisive bump in the polls for the candidate who had the best convention. At this point, preferences are relatively hardened, but any number of wildcards could disrupt expectations in the final week of the campaign. As the campaign progresses, perceptions of economic growth favor or hurt the incumbent presidential party, and while the economy influences preferences early on, the economy’s impact grows over the course of the campaign.

Wlezien and Erikson divide their explanation of what is driving the presidential vote into two competing influences — the fundamentals and the campaign — and they seek to determine the dominant influence. What they conclude is that the two work together. Fundamentals are defined as things that cause a long-term shift in voter preferences, and they may be “internal”, such as party identification, or “external”, such as the economy, candidate positions on issues relative to voters, and the popularity of the sitting president. They argue that for many voters internal fundamentals, especially partisan predispositions, anchor voter beliefs, and therefore candidate choice is a long-standing decision difficult to disturb. Campaigns also have their greatest influence on voters with the least interest in politics. Wlezien and Erikson conclude that because the probability of who will win a presidential election often undergoes meaningful changes during a campaign, the campaign must matter, but that campaign dynamics are less fluid than most believe. The heart of the book’s argument is that while the fundamentals drive vote choice, campaigns deliver the fundamentals to the electorate.

Some of the book’s significance might be brought into sharper focus by considering another recent book, that published by Stephen Jessee, “Ideology and Spatial Voting in American Elections.” This book addresses one of the key fundamentals mentioned by Wlezien and Erikson — candidate positions on issues relative to voters. The book is a test of the spatial theory of voting, which assumes that voters and candidates place themselves along a liberal-conservative policy spectrum and that voters choose the candidate closest to their position. Using new direct measurements of voter ideology and candidate positions — which previous to this book have been measured on different scales — Jessee is able to demonstrate that voters are able to form relatively accurate pictures of candidate ideologies and that voter perceptions of their ideological proximity to candidates are quite accurate. Jessee also finds, however, that partisanship biases vote choice, so that voters who identify with a political party are more likely to vote for that party’s candidate regardless of their proximity to the candidate along the policy spectrum. Further, Jessee finds that the more informed voters are, the more they are able to offset the impact of partisanship and make more precise voting decisions based on ideological proximity, but even in this high-information context, partisanship continues impacting vote choice.

Ideology and Spatial VotingIt is not hard to see the relevance of Jessee’s study to the conclusions Wlezien and Erikson draw. Jessee shows that partisanship is an incredibly powerful influence on vote choice, and therefore it is unsurprising that campaigns have such a difficult time moving the electoral dial in any direction. But Jessee also demonstrates that, by and large, the electorate seeks to align themselves with the candidate perceived to most accurately represent their preferences. Therefore, it is likewise unsurprising that campaigns play a vital role in delivering this information to voters, and also that campaigns would have their greatest impact at moments such as the nomination stage and the party conventions, when parties and candidates are on their fullest display. Many puzzles remain, but these two books have provided state of the art answers to critical questions.

bottom border