Starting that fall, and through the following spring, Todorov showed pairs of portraits to roughly a thousand people, and asked them to rate the competence of each person. Unbeknownst to the test subjects, they were looking at candidates for the House and Senate in 2000, 2002, and 2004. In study after study, participants’ responses to the question of whether someone looked competent predicted actual election outcomes at a rate much higher than chance—from sixty-six to seventy-three per cent of the time. Even looking at the faces for as little as one second, Todorov found, yielded the exact same result: a snap judgment that generally identified the winners and losers. Todorov concluded that when we make what we think of as well-reasoned voting decisions, we are actually driven in part by our initial, instinctive reactions to candidates.
Todorov’s study indicated that election results might owe themselves somewhat to what the Nobel-winning psychologist Daniel Kahneman called fast, unthinking judgment, or what the psychologist Nalini Ambady calls thin-slice judgment: the ability to make any number of social judgments from a seconds-long experience. Students can predict a professor’s end-of-semester ratings from a silent video that lasts no more than ten seconds; employers can predict interview outcomes and hiring decisions from a little more; and voters can predict the results of elections from a judgment that is made in less than a second.
Obviously, many things affect voting decisions, from political platforms to sexting scandals. But if we control for the underlying factors, the research suggests that a thin-slice judgment retains its predictive validity, and it emerges as the single strongest predictor of victory beyond external factors such as broad economic data, like the unemployment rate; personal data, like age or gender; or any other single political measure, like whether someone is an incumbent or how much has been spent on the campaign. In one study of fifty-eight gubernatorial races, the only element that outperformed the thin-slice impression was a combination of two of the political factors most closely tied to electoral success: incumbency status and campaign spending. While we are never forced to vote based on one factor alone, the apparent predictive power of competence judgements reveal how deeply that quick impression may color our evaluation of more serious considerations.
...
Those facial cues, in turn, may stem from a far more basic impulse, since we respond to those same features as children. In a 2009 study published in Science, the psychologists John Antonakis and Olaf Dalgas suggested that, when we judge a candidate as more or less competent, we do it in the same way that children do. They first asked a group of adults to rate pairs of faces, taken from the 2002 French parliamentary elections, based on how capable they seemed. When they compared the ratings to actual election results, the correspondence was seventy-two per cent. The ratings even predicted the margin of victory; the more competently-rated the face, the higher the margin. The researchers then had a group of children play a computer game, simulating a boat trip from Troy to Ithaca, in which they had to choose a captain for the voyage; their options consisted of the same 2002 election candidates. The two sets of responses were indistinguishable from each other: seventy-one per cent of the time, the children picked the election winner to pilot the boat.
