Oriol Navarro Erausquin


Maybe you have heard about the Downs paradox, also known as the paradox of voting. Since the chance of exercising the pivotal vote is infimal, a rational, self-interested voter will not dedicate his precious time to vote (Downs 1957). Of course, a lot of political scientists have responded to this issue, usually attacking the thesis that we as citizens behave as pure egoistic rational individuals. But on the other hand, the perception that your vote will not make a change can, and usually does cause a lot of citizens not to leave their homes to go to vote. In fact, one of the key aspects that political parties have to do in every election is to convince the citizens that already like them to actually go to the polling stations and vote.

Luckily for politicians, it is not that hard to make people to go to vote, especially in tied elections, where competitiveness is high. If the chances, or more correctly, the perception that the chances of winning or losing are close, the citizen can be motivated to think that their vote can really make a change. But a question can be asked, a question whose answer can be of vital importance to any politician who wants to win an election. Between appearing to be barely winning or barely losing, what can get me more votes?

We will consider that it might be preferable to give an image of being barely losing, as a way to counteract the Downs paradox. Our goal is to prevent a citizen who prefers our candidacy from not going to vote because he thinks and feels that it will not make any difference. If they are not going to vote because it is very unlikely that their vote makes any difference, we have to make them believe that, on the contrary, their vote can make a change. To do so, we will first focus on the “can”. It is clear that we have to make the voter think it is possible that their vote could be the pivotal, so it is important to remark that the two candidates are very close; that every vote counts. But, why is it preferable to show an image that our candidate is losing? Now we will focus on “making a change”.

If we announce to the voter that we are winning, we are telling them that their effort to go to vote only serves to reinforce a given reality. If we proclaim that the battle is already won, the suffragist is going to feel that they are not really making any difference; that their vote will only serve to ensure the actual reality. If instead of that we tell the voter that his candidate is barely losing, it will generate an image that his vote may be the one that makes the change. What we will be telling him is that if he does not vote he will not do anything to avoid the predicted loss of their candidate, while if he actually votes he can turn the table, make a change. In this way, going to vote becomes a heroic exercise, and that is why it generates a much more attractive narration for the voter.

Todd Rogers and Don Moore, in their paper “The Motivating Power of Under-Confidence” (2014), studied the impact of political campaigns in close races. Specifically, they studied the effect produced by making the voters know their candidate could win (over-confidence) or lose (under-confidence). Two fundraising email field experiments showed that emphasizing polls that informed that a preferred candidate was barely losing raised between 55% and 60% more than emphasizing polls that showed that the same candidate was barely winning. It even augmented the proportion of people that opened the email (2,3%), the proportion of clicks on the link to donate in the email (22%), and the proportion of people that actually donated (46,6%).(2014, 18)

In fact, studying emails collected during the 2012 US Presidential election they realized that the politicians, or at least their campaign managers, seem to already know this effect: they were more likely to report that preferred candidates were almost winning than that preferred candidates were almost losing. This was even clearer with the campaign of Donald Trump, who presented himself as a candidate who was going to win the elections, but one week before the elections was asking to his voters in Miami the following: “The polls are all saying we’re going to win Florida,” “Don’t believe it, don’t believe it. Get out there and vote. Pretend we’re slightly behind.” (Mason 2016). At the end, Trump achieved a narrow victory in Florida, the most important swing-state of the US, by only a margin of 100,000 votes.

The study of Rogers and Moore effectively demonstrated that those who already support a candidate are more motivated when the  candidate appears to be barely losing, and tend  to donate or volunteer more to the campaign, and presumably, to actually go to vote. Daniel Fleitas also supports this theory in his 2016  paper, calling it the “Underdog effect”. But he also exposed a contrary effect, presented too in the study of Rogers and Moore. They detected in the uncommitted voters a preference for the candidate that was winning. In other words,  voters who were unclear to whom to vote were more attracted to the candidate that appeared to be winning. This preference of the uncommitted for the winning option has been already studied by different political scientists, and it is called the bandwagon effect. When we do not have an opinion, we tend to actuate as the majority does, like a herd instinct.

Keeping the study in mind and the fact that campaign managers usually prefer to present their candidate as barely losing, we can conclude that it is optimal to appear as almost the winner, at least in some situations. What we can for sure say is that in emails, party rallies, and other types of communication that we know will get only to people that already support our candidate, it is more effective to appear as barely losing. In other forms of promotion that get to the general public, as advertisings, if the uncommitted citizens are considerable, it can be wiser not to use this technique to avoid the bandwagon effect.




Downs, Anthony. 1957. An Economic Theory of Democracy. New York: Harper. Fleitas, D. 1971. Bandwagon and Underdog Effects in Minimal-Information Elections.

American Political Science Review, 65(2), 434-438. doi:10.2307/1954459

Mason, Melanie. 2016. “Fighting the urge to declare a certain win, Trump tells supporters,    ‘Pretend    it’s    close’”    Los    Angeles   Times,    November    2,    2016. http://www.latimes.com/nation/politics/trailguide/la-na-trailguide-updates-1478112162-htmlstory.html

LeDuc,Lawrence and Niemi, Richard. 2002. Comparing Democracies 2: New Challenges in the Study of Elections and Voting. London: Sage

Rae, Douglas 1967. The political Consequences of Electoral Laws. New Haven, CT: Yale University Press

Rogers, Todd and Moore, Don A. 2014. The Motivating Power of Under-Confidence: ‘The Race is Close But We’re Losing‘. HKS Working Paper No. RWP14-047. https://ssrn.com/abstract=2528690 or http://dx.doi.org/10.2139/ssrn.2528690