A candidate’s likability is not a useful tool for predicting election outcomes.
By Kristina Kutateli
Left: Donald Trump. (Photo: Joe Raedle/Getty Images) | Right: Hillary Clinton. (Frederic J. Brown/AFP/Getty Images)
Whether it is Madeleine Albright calling Donald Trump “crazy,” or Trump’s constant use of his “Crooked Hillary” moniker, the 2016 presidential election has been characterized by a daily onslaught of personal attacks. Voters seem to take part in the negativity too: Candidate likability, measured through favorability ratings, is at an all-time low. For Clinton and Trump, who emerged from a divisive primary season hated but (in Clinton’s case, likely) victorious, this may spell bad news — but is a candidate likability a useful tool for predicting election outcomes?
Not really, according to a 2016 paper by political scientist Martin P. Wattenberg at the University of California-Irvine. In examining 60 years of data from the American National Election Studies, Wattenberg finds that the electorate no longer focuses on character attributes as much as it did in years past. Whereas 80 percent of respondents had cited personal attributes in elections between 1952 and 1980, in the 2008 and 2012 elections that figure had dwindled to around 60 percent. The nature of those comments has differed too — the study notes that mentions of candidate traits are “now more tied to partisan identiﬁcation and have less of an independent impact on voting behavior.”
Rating a candidate’s popularity or likability is a complicated process. A candidate who earns high markings for integrity, for example, will often perform poorly in another dimension, such as competency. The study also notes that, while many believe that voters typically choose candidates with the most charisma — someone who establishes “an image of inspiring leadership” — the data shows this is not something people actually mention all that often when analyzing a candidate’s fitness for office.
Using traits such as integrity and competence, the data shows that there is no correlation between perceived honesty and election outcomes.
In a presidential race such as this one, where both Clinton and Trump have remarkably low approval ratings, the findings will likely be even more accurate. This is already evident in the primary results: Despite winning the majority of votes, Clinton and Trump have regularly maintained some of the lowest favorability ratings in election history. Clinton and Trump also poll more or less evenly in terms of honesty and trustworthiness — likely meaning that this election will play out on other terms, such as policy proposals and governing ability.
Political scientists David Holian and Charles Prysby of the University of North Carolina also used data from the ANES to determine the importance of a candidate’s personal attributes in their 2014 book, Candidate Character Traits in Presidential Elections. Their findings are similar: Using traits such as integrity and competence, the data shows that there is no correlation between perceived honesty and election outcomes.
But what does it mean for American politics that character isn’t that consequential — that Americans are choosing candidates they themselves describe as dishonest? While perhaps voters may be choosing candidates for a government that they don’t trust all that much anymore — or, as others have suggested, due to a rising authoritarian impulse — Wattenberg proffers that, for young people in particular, voters are less primed to view candidates in terms of character due to “having grown up in an era in which candidates have to make strong and distinct policy appeals to get the nomination.” Voters, above all, are choosing candidates they believe will govern effectively.
One thing is for certain: You don’t have to like someone to vote for them.