Can You Spot a Republican (or a Democrat) Just By Looking at One?

Political leanings and even electoral potential can be determined just by looking at a face.

Brian Resnick
Nov. 2, 2012, 5:05 p.m.

Enorm­ous sums of money have been spent on polls, ads, cam­paigns, and cov­er­age of this year’s dead-even White House con­test. There’s even re­search that tracks and in­ter­prets the can­did­ates’ eye blinks. But what if voter de­cisions boil down to a quick re­ac­tion to the way some­body looks?

Ima­gine an ex­per­i­ment in which a pair of pho­tos is shown to a group of chil­dren who are pre­sum­ably not versed in polit­ics. One is of Barack Obama’s face. The oth­er is of Mitt Rom­ney’s. After they’ve seen the pic­tures, the chil­dren are asked a simple ques­tion: “Who would you rather be the cap­tain of your ship?”

Psy­cho­lo­gic­al re­search sug­gests that such a ques­tion might be the most rel­ev­ant one of all. The 2008 ex­per­i­ment was per­formed by re­search­ers in Switzer­land, but the can­did­ates were French and the kids were Swiss—mak­ing them even less aware of the politi­cians. As a group, the chil­dren guessed the win­ners of real elec­tions with 71 per­cent ac­cur­acy. They even pre­dicted Obama would win the 2008 elec­tion.

Al­though you wouldn’t think it, just by look­ing at someone’s face, we im­me­di­ately know a lot about them. From bone struc­ture, we know a per­son’s gender. From their ex­pres­sion, we know their mood. We’re de­cent guess­ers of a per­son’s age. But these are the easy ones. Our abil­ity to read faces goes much deep­er than these sur­face fea­tures. Turns out, faces even re­veal what their own­ers be­lieve.

The chil­dren in the study aren’t out­liers.  A land­mark 2005 study by Al­ex­an­der To­dorov, pub­lished in the journ­al Sci­ence, found that people were able to pre­dict the out­comes of the 2004 U.S. Sen­ate and House elec­tions just by look­ing at faces of can­did­ates for 1 second. “This hap­pens without any pri­or know­ledge of the politi­cian’s policies, without any ex­pos­ure to cam­paign ad­vert­ising, and without in­form­a­tion about the politi­cian’s ex­per­i­ence,” said Kerri John­son, a re­search psy­cho­lo­gist at the Uni­versity of Cali­for­nia (Los Angeles).

What is at play is a per­cep­tion of com­pet­ence. In the To­dorov study, the people with faces deemed more com­pet­ent were more likely to get elec­ted. And this no­tion of com­pet­ence is con­veyed, with­in a second, by the struc­tures of the face. What does a com­pet­ent face look like? It’s largely mas­cu­line, with a square jaw and large eyes. Baby-faced politi­cians, be wary. The video be­low takes a com­pet­ence-neut­ral face, morphs it in­to a very com­pet­ent face, and then to a very in­com­pet­ent-look­ing one.

What To­dorov’s re­search shows is that voters’ in­stant im­pres­sion of can­did­ates can per­sist in­to the polling booths. The res­ults of the study, he wrote, “have chal­len­ging im­plic­a­tions for the ra­tion­al­ity of vot­ing pref­er­ences, adding to oth­er find­ings that con­sequen­tial de­cisions can be more ‘shal­low’ than we would like to be­lieve.” 

What’s go­ing on here is simple: It’s ste­reo­typ­ing. Col­lect­ively, we have an un­der­stand­ing of what our lead­ers should look like. In turn, those ste­reo­types be­come real­ized when we vote. And our abil­ity to make snap judg­ments of politi­cians goes even deep­er than pick­ing win­ners and losers. At a sub­con­scious level, we even can spot a par­tis­an in a crowd.

“We have a lot of gut feel­ings, and my re­search in­terest is in un­der­stand­ing what those gut feel­ings are and where they come from,” says Nich­olas Rule, a psy­cho­lo­gist at the Uni­versity of Toronto. In 2009,he pub­lished an ex­per­i­ment that showed that gut feel­ings about an in­di­vidu­al’s polit­ics carry some truth. He showed study par­ti­cipants pic­tures of Sen­ate can­did­ates’ faces and asked them a simple ques­tion: Is this a Re­pub­lic­an or a Demo­crat?

To a de­gree great­er than chance, the par­ti­cipants guessed the right an­swer. When he had his par­ti­cipants rate the polit­ic­al lean­ings of their non­politi­cian peers, the ef­fect was even stronger. “There’s tons of er­ror,” Rule says. “We’re look­ing at 60 per­cent ac­cur­acy at best. The reas­on that is stat­ist­ic­ally sig­ni­fic­ant is be­cause it is so re­li­able.”

Rule says Re­pub­lic­an faces tend to look more dom­in­ant than Demo­crat­ic ones. “If we were to trans­late that in­to fea­tures, for ex­ample, it would be cues to dom­in­ance, which would be an­gu­lar­ity of fa­cial fea­tures, jaw size, the heav­i­ness of the brow ridge. Mas­culin­ity in a way.” On the oth­er hand, Demo­crats tend to have “warm­er” faces.

UCLA’s John­son did fur­ther re­search in­to fa­cial re­cog­ni­tion of party ID, in par­tic­u­lar wo­men in elec­ted of­fice. She asked par­ti­cipants to judge wheth­er the rep­res­ent­at­ives were Demo­crats or Re­pub­lic­ans, by face alone. (Her par­ti­cipants were col­lege stu­dents with a near-zero chance of re­cog­niz­ing the 434 House mem­bers of the 111th Con­gress.) Her find­ings rep­lic­ated Rule’s—the stu­dents matched politi­cians to party to a de­gree great­er than chance. The par­ti­cipants were 98 per­cent more likely to guess cor­rectly when the politi­cian was a highly fem­in­ine Re­pub­lic­an wo­man—that is, when the politi­cian fit a ste­reo­type.

Of course, choos­ing a can­did­ate is more com­plic­ated than just look­ing at a face. But both John­son and Rule say these ste­reo­types do play a role in elec­tions, es­pe­cially with cas­u­al voters. “Where we think this has the biggest im­pacts is on the mar­gins when people are not either well-in­formed or they’re not en­gaged in the polit­ic­al pro­cess,” John­son says.

The au­thors of the study of the Swiss chil­dren put it in stark­er terms: “Un­for­tu­nately, voters are anchored in first im­pres­sions and do not ap­pro­pri­ately cor­rect ini­tial in­fer­ences; ad­di­tion­al in­form­a­tion on the can­did­ates does not change choices by much.”