The Mind and the Ballot: The Role of Psychology in Elections

Alex Kazberouk
By Alex Kazberouk November 22, 2008 21:09

With another round of national elections this year, candidates do their best to sway voters with cogent debates, massive rallies, expensive advertisements, and, of course, mandatory baby-kissing. Ideally, this and other elections would be decided by an educated public rationally choosing the candidate with whom it most agrees, based on the candidate’s potential for holding office.

In reality, however, we are not rational beings: it is often subconscious psychological thought, rather than objective reasoning, that dictates which name we mark on the ballot. This fact is no secret to candidates and their advisors, who attempt to harness this knowledge to gain votes. The field of political psychology studies the underlying factors that influence elections, two examples of which are described here.

Ballot Order Effect

Students in introductory psychology classes learn a phenomenon known as primacy effect: our tendency to remember items at the top of a list better than those in the middle or near the bottom. Having read the first items, we pay less attention to those that come later either because we are bored or because we are processing the information we saw at the top of the list.

This effect was classically demonstrated by psychologist Solomon Asch in 1946 when he presented subjects with the same list of positive and negative adjectives describing a person, listing either the positive ones (intelligent, industrious, etc.) or the negative ones (envious, stubborn, etc.) first. When asked to rate the person, those who were given positive adjectives earlier in the list consistently rated the person higher than those who were given negative ones first.

This same effect applies during elections, as the order of candidates on ballots can favor those listed first. For example, statistical analysis of over 20 years of elections in California shows that the so-called “ballot order effect” may have changed the winner in up to 12% of the primaries in the state.

The effect is much more significant for non-partisan elections or for small local elections where voters know little about candidates and are not divided among party lines. The effect is less pronounced in large, well-advertised elections where voters are more likely to have selected a candidate before coming to the polls (1).

Jon Krosnick, a professor of political science and psychology at Stanford University, asserts that the effect likely played a role in the 2008 New Hampshire Democratic primary. Pre-election surveys in that state showed candidate Barack Obama leading candidate Hilary Clinton by as much as 7%. However, Clinton won the crucial early state. Krosnick writes: “I’ll bet that Clinton got at least 3 percent more votes than Obama simply because she was listed close to the top” (2).

While the order of candidates on ballots varies from state to state, most states currently use a single randomized ballot that, as studies point out, actually penalizes the candidates who are randomly listed last. Other states place the current office holder first on the ballot, thus potentially making him or her even more likely to win.

Awareness of the ballot order effect has prompted psychologists to lobby for multiple ballots with different random candidate listings, and increasing numbers of states are adopting this system.

Push Polls

The ballot order effect is independent of the candidates themselves and thus cannot be used as a tactic per se by candidates seeking an unfair edge. However, candidates can – often on questionable moral grounds – use psychology as a weapon against competitors.

One very powerful example is that of push polls, brief surveys designed to give voters disinformation about a candidate via a hypothetical question. Data from these surveys is not collected; the polls serve to either remind voters of negative characteristics of candidates or to fabricate negative traits that then remain in the voters’ minds.

A famous example of push polling involved the 2000 South Carolina Republic primary, in which voters received phone calls asking if they would be more or less likely to vote for candidate John McCain if he hypothetically had fathered an illegitimate black child. The tactic was especially effective as McCain was campaigning in the state with his adopted Bangladeshi daughter. The smear campaign was done anonymously, and no candidate has admitted responsibility. Despite initially leading in the polls, McCain in fact lost to George W. Bush, 42% to 53%.

The effects of push polls were confirmed by a study published in the Journal of Consumer Research that asked undergraduates to vote for one of two potential candidates for office. Candidate A was constructed to be the more likeable candidate, with 81% to 89% of the students voting for him over his opponent after viewing both candidates’ campaign websites.

However, when students viewed the same websites but were also asked a hypothetical highly negative question about the favored candidate, only 38.9% voted for Candidate A, despite the fact that the question was worded as a hypothesis and not a fact. Students who were distracted when the question was asked voted for Candidate A 55% of the time, while students told to focus on the candidate voted for him only 16% of the time, showing that the question was indeed responsible for the drastic opinion shift (3).

Thus, push polls through hypothetical questions present effective means of influencing public opinion. As these can be conducted by anonymous groups over the phone, they are more difficult to trace and may often be done by overzealous supporters without the candidate’s approval or knowledge.

When faced with a negative push poll, a candidate may either choose to ignore it, potentially letting a rumor spread, or to address it, further engraining the slander in the electorate and distancing the campaign from real issues. Due to the staying power of false accusations, even a push poll that is proven false will subconsciously continue to taint the candidate in the eyes of the electorate.

Attempting to eliminate push polls by pressuring candidates is futile, as the polls may not necessarily be linked to the candidates themselves. An educated electorate that understands what a push poll is and when one is being conducted is thus the best defense against them.


In addition to push polls and ballot order effects, there are hundreds of examples of psychological factors affecting elections and of candidates’ exploitation of them to garner voter support. Understanding the psychology of elections thus becomes crucial not only for the candidates, but for voters who want their votes to be based on objective thinking.

Next time at a ballot box, at least appreciate the effort of the psychology experts your candidate hired to win your vote. And then think twice before voting for the candidate first on the list.

1. Imai, Kosuke. and Ho, Daniel. “Shaken, Not Stirred: Evidence on Ballot Order Effects from the California Alphabet Lottery” Paper presented at the annual meeting of the The Midwest Political Science Association, Palmer House Hilton, Chicago, Illinois, Apr 15, 2004

2. Jon A. Krosnick. “Ballot Changes Cited in Vote’s Discrepancy With Polls. story?id=4107883

3. Davis, Richard H. “The anatomy of a smear campaign.” The Boston Globe, March 21, 2004 politics/president/articles/2004/03/21/the_anatomy_of_a_ smear_campaign/

4. Fitzsimons, Gavan J. and Baba Shiv (2001), “Nonconscious and Contaminative Effects of Hypothetical Questions on Subsequent Decision Making,” Journal of Consumer Research, 28 (September), 224–38.

Alex Kazberouk
By Alex Kazberouk November 22, 2008 21:09