As Ted Cruz’s popularity surges with Iowa Republican primary voters – bringing his campaign newfound frontrunner gravitas — one issue about the candidate keeps coming up: He’s just not that likable a guy.
The New York Times ran a piece to this effect, poking fun at Cruz’s stiff persona and canned attempts at levity. He is also not held in very high esteem by his fellow senators, one of whom even once publicly called him a "wacko bird."
But perhaps the most stinging affront to the candidate’s personality came from his freshman college roommate, who told the Daily Beast: "I would rather have anybody else be the president of the United States. Anyone. I would rather pick somebody from the phone book."
Cruz’s campaign managers are aware of this problem, and they're apparently so worried that they've invested in an in-house team of data scientists to mine the Facebook data of tens of millions of users to help Cruz get more votes. So the question is worth asking: Is he wasting his money? What is "likability," and does it even ultimately matter?
"People will go for the unlikable candidate" if they represent your issues
A candidate’s likability has long dogged presidential campaigns on both sides of the aisle. Barack Obama famously told Hillary Clinton in 2008 that she was "likable enough" — though not, apparently, enough to secure the Democratic nomination that time around.
"Given the choice between a likable candidate who does not take your side on the issues and a really unlikable candidate who does, I think people will go for the unlikable candidate, more often than not," said David Redlawsk, a political scientist at Drake University in Iowa.
Redlawsk is basing those comments on a study he conducted in 2006 with a colleague titled, "I like him, but … : vote choice when candidate likability and closeness on issues clash." The experiment asked voters ahead of time to take a questionnaire that asked about their political ideology, and then showed them pictures of two candidates.
Here’s where the study gets a little thorny: To get at "likability," the researchers took still shots from real campaign ads of lesser-known candidates. For "likable" photos, they took snapshots of the candidates smiling and looking jovial. The "unlikable" snapshots showed the politicians scowling or making other negative expressions.
There’s nothing too scientific about that approach; the photos were simply chosen by a panel that voted them the ones that made the candidate look the least friendly. As Redlawsk put it, likability is a nebulous concept. "You know it when you see it," he said.
In the experiment, one candidate mirrored the study participant’s positions on the issues, but he looked pretty unpleasant in his photo. (The researchers also offered survey takers a list of negative personality traits to go with the picture.) The other candidate was the precise opposite: kind and approachable-looking in his photo, with views that differed markedly from the participant’s.
People naturally chose the candidates whom they thought best represented their issue positions.
But it’s certainly possible that the experiment’s simplistic design weeded out more subtle ways in which voters are influenced by likability. It may even be that since participants in the study were asked about their political views before completing the experiment, they were primed to care even more about specific partisan issues.
Likability hasn’t seemed to influence past presidential elections in any major way
Lab experiments are often criticized because in order to isolate the one thing researchers want to measure, they create an unrealistic situation that voters would never otherwise face.
Another study, conducted by Stanford political scientist Morris Fiorina, tries to solve that problem by correlating likability with election outcome. Fiorina tracked every presidential election from 1952 to 2000 and found no significant correlation between a candidate’s likability and his ability to take the White House.
Here, the study got at likability by clustering together a whole host of personal attributes like intelligence, sincerity, and charisma, on the assumption that it was probable that a voter liked a candidate she thought was intelligent or inspiring.
In cases where a candidate’s personality did seem to determine an election, the difference in ratings was stark. In 1964, for example, President Lyndon B. Johnson was the most likable Democrat in nearly half a century, according to Fiorina’s ratings. By comparison, the Republican candidate, Barry Goldwater, was rated the second-least-likable Republican in the entire set. LBJ won in a landslide.
But even those candidates we think of as so notoriously unlikable as to torpedo their campaigns may not have as much of an effect as we might think. In 2000, as Fiorina wrote in the New York Times, Al Gore "embarrassed" political forecasters who had predicted he would win between 53 and 60 percent of the two-party popular vote. Pundits attributed the loss to Gore’s apparent arrogance and otherwise wooden persona.
It’s true that Gore’s personality was rated slightly lower than that of George W. Bush in 2000, but Fiorina found much stronger reasons for his defeat — namely that he was perceived to occupy a position further to the left than Bill Clinton.
But personality might be fair game in the primary
Redlawsk says it’s possible that likability matters a great deal more when voters are just getting to know a candidate, like in this year’s Republican primary. Likability could be the factor that spurs a voter to research a candidate and become familiar with his positions on the issues.
Primary voters are still very issue-oriented — it’s what motivates them to come out and vote. But if voters find a candidate like Ted Cruz disagreeable, there are likely several others who represent their views well, who also have friendlier dispositions.
But, Redlawsk cautions, just because Republican elites — Ted Cruz’s Senate colleagues, say — find a candidate disagreeable, it doesn’t mean the voters feel that way.
And judging by the latest polls out of Iowa, Cruz’s likability problem may already be fading.