Looks or Luck

It’s a safe bet that many politicians would sell their own grandmothers to learn what determines electoral success. Is it their proficiency in managing the economy? How well the country is doing relative to its peers? The image they project abroad about the “virility” of their country? Their looks? Sheer luck?

Well, any and just about all of the above, as a new study by CESifo researcher Harry Garretsen and his colleagues Janka I. Stoker, Rob Alessie and Joris Lammers shows. But, crucially, which factor influences electoral outcomes depends on the type of electoral system a country has.

Contrary to our glossy picture of ourselves, whereby we are well-informed individuals who make thorough and consistent choices based solely on the actual competence and policy stances of politicians, we are actually pretty ignorant about political issues. What actually moves us to cast our vote for one particular candidate has therefore been the topic of much research.

Most models in economics assume that voters, based on their “highest expected future utility,” will base their decisions on outcomes that can be attributed to the actual performance of politicians (their competence), and even on some factors that clearly lie beyond their control, such as the state of the world economy (i.e. “luck”).

Psychological research has also delved into the matter. Several studies show that people take decisions based on rules in order to deal with an information load, resorting to shortcuts such as religion or party affiliation. But they also resort to non-verbal cues, such as a candidate’s height and voice. The process suggested is that voters draw inferences, such as competence, from such non-verbal cues, and these inferences consequently lead to voting decisions, more or less along the lines of “the taller candidate is more dominant, and therefore the better leader.”

CESifo researcher Panu Poutvaara and colleagues (as well as other research teams) showed in 2010 that facial appearance is also a factor that influences electoral decisions. These results, and those of similar studies, are universal: despite ethnic, cultural and racial differences in response groups, the same candidates are seen as better candidates and winners of elections. Even children aged 5 to 13 years were able to predict the winner in 71% of the cases.

Mr Garretsen and his colleagues, whose paper is the first to combine and confront research streams from economics and psychology on predicting elections, point out that most of the above studies have been conducted under laboratory conditions, and, most importantly, ignore the electoral system in their analyses, concerning themselves with elections in one country, in one year, without comparing systematically elections over countries or over time. So, they created their own dataset by running an extensive experiment in which they presented 501 respondents with only the facial appearance of the incumbent and the main challenger in 196 parliamentary and national elections in 44 countries for the period 1979-1999, adding specifically the electoral system for each election.

They find that economic luck and competence matter significantly, but that luck is more relevant, with a coefficient that is twice that of the competence coefficient. These findings hold even after controlling for the degree of economic integration of a country, its level of economic development, the quality of its governance and its level of media penetration. The relevance of competence (against luck), however, is larger in countries with a higher GDP or a higher level of education.

The effects of appearance, in turn, are moderated by contextual factors, such as the characteristics of voters, time, and the electoral system. For unknowledgeable voters, such effects are much stronger, and television increases the extent to which voters rely on appearance to choose a candidate. Systems that encourage personal voting, i.e. voting for a person more than for a party, evidence stronger effects of appearance on voting decisions.

In sum, our authors find that indeed the looks of a candidate and their perceived economic competence, and to a far lesser degree, economic luck, are relevant predictors of electoral success, but that the type of electoral system plays a crucial role: looks matter in a majority system where the winner takes all, i.e. majority-based elections, whereas economic competence matters in representative elections. This holds for a number of robustness checks, including the degree of information the voters exhibit, GDP of their country, and education.

They acknowledge that their study has limitations. The small number of female incumbents and opponents forced them to limit their analysis to male politicians. Furthermore, the pictures of politicians were presented only to a group of US citizens, whereas a cross-cultural approach would be desirable.

But still, now you know it. Check out what kind of electoral system your country has, and then either head for the looks counsellor, or hone up your economic competence.

Harry Garretsen, Janka I. Stoker, Rob Alessie and Joris Lammers: Simply a Matter of Luck & Looks? Predicting Elections when Both the World Economy and the Psychology of Faces Count, CESifo Working Paper No. 4857
Other CESifo Working Papers by Harry Garretsen

This text is the responsibility of the writer and does not necessarily represent the opinion of the authors of the paper featured.