A new study tests the idea that candidates like Cruz, Sanders, Goldwater, and McGovern were too extreme for voters—and finds it’s just not true.
By Nathan Collins
Barry Goldwater speaking at a New York City election rally in 1964. (Photo: William Lovelace/Daily Express/Hulton Archive/Getty Images)
There’s an old advertisement from the 1964 presidential campaign called “Confessions of a Republican” that’s been re-circulating around the Internet lately. In the video, a random man explains why he’s voting for Lyndon Johnson—namely, because Barry Goldwater is too extreme.
Well, that man is in the minority: According to a new study, voters are remarkably OK with politically extreme candidates.
“In the wake of the 2016 Iowa primary, the Washington Post editorial board suggested that Ted Cruz and Bernie Sanders were too extreme for general election voters,” political scientists Marty Cohen, Mary McGrath, Peter Aronow, and John Zaller write in the Annals of the American Academy of Political and Social Science. “Yet evidence that extremism has mattered in presidential elections in thin.”
Several of the most reliable predictors of presidential election outcomes have little to do with the candidates’ politics.
It’s well known (among political scientists at least) that several of the most reliable predictors of presidential election outcomes have little to do with the candidates’ politics per se. Instead, the most important factors appear to be the economy—more specifically changes in real (i.e., inflation-adjusted) disposable income (RDI)—and the simple fact that voters, in the aggregate at least, have a tendency to tire of whichever party is currently in the White House.
For Cohen and his colleagues, the question is whether voters punish, so to speak, presidential candidates who take extreme political positions. The answer, they found, is no. The researchers looked at every presidential election from 1948 to 2012 and computed relative extremism—that is, how extreme the incumbent party’s candidate was relative to his opponent—based on expert opinions, voters’ perceptions, and candidates’ positions on Congressional bills.
After taking into account changes in RDI in the two quarters prior to an election and the number of years the incumbent president had been in office, relative extremism had essentially no effect on the two major parties’ vote shares. According to their most charitable estimates, no candidate could expect to get more than a one- or two-point advantage by being more moderate. Jimmy Carter, for example, would have had at most a 1.4 percent boost at the polls against the more extreme Ronald Reagan (not that it would have helped). Indeed, it’s not just that voters don’t punish politically extreme candidates—it’s that a candidate’s political views don’t seem to affect an election outcome.
“[I]deological location seems unlikely to be of much predictive importance in elections where one candidate begins with an important advantage,” such as a thriving or tanking economy, the team writes.