ÎçÒ¹¸£Àû1000¼¯ºÏ

WHEN it comes to elections, some voters seem to be swayed by a simple rule of thumb – does the face fit?

A few years ago my colleagues and I obtained photographs of the six most recent Republican US presidents and the six most recent Democratic presidents, and created two composite images by merging the faces of each group. We then asked people to rate the degree to which the two resulting faces appeared caring, authoritarian and open-minded. Even though participants had no idea that the faces were based on presidents, large differences emerged. They rated the Republican composite as more authoritarian and the Democrat as far more caring and open-minded.

Other research suggests that looking competent is particularly advantageous when it comes to winning elections. In 2005, Alexander Todorov and colleagues at Princeton University in New Jersey showed students photographs of the winners and runners-up in elections for the US Senate and House of Representatives from 2000, 2002 and 2004. For each pair, Todorov asked the students to choose who looked more competent. Again, the participants did not know they were looking at politicians. The results showed that the more competent-looking candidate won about 70 per cent of the time ().

In the run-up to the recent UK general election, we decided to test whether British voters were swayed by the same effect. We obtained photographic portraits of candidates standing in 59 marginal seats – parliamentary constituencies considered winnable by at least two parties – and showed them to more than 200 people. Without telling the participants who they were looking at, we asked them to decide which face looked more competent. For each constituency we predicted that the candidate who looked most competent would win.

We also decided to use our data to predict the outcome of the entire election. Assuming that the 382 seats widely thought to be safe would not change hands, and that the findings from our sample of marginal seats would scale up to the other marginals, our competence ratings predicted that no party would win an overall majority, with 290 seats going to the Conservatives, 247 to Labour, 70 to the Liberal Democrats and 25 to other parties (we excluded Northern Ireland’s 18 seats).

So how did we do? Not particularly well. Our analysis shows that the competent-looking candidates were not especially likely to win: we predicted the correct result in only 27 of the 59 marginals. Our prediction for the outcome of the entire election was reasonably accurate, but this might just have been chance. We were certainly no better than conventional opinion polls.

There are various ways of explaining the difference between our findings and the American results. Perhaps British voters are more likely to vote on policy rather than appearance, or maybe they are less aware of what their candidates actually look like. Either way, in this general election candidates who happened to look especially competent didn’t have an unfair advantage when it came to winning a parliamentary seat.