As Democrats debate the best way to regain power, two competing visions have emerged: Focus on changing the minds of reluctant Republican voters in 2016, or concentrate on motivating members of the party's core constituency to cast ballots.
New research strongly suggests the latter is, by far, the more promising option. It concludes political campaigns are completely ineffective in convincing people how they should vote.
"The direct persuasive effects of voter contact and advertising in general elections are essentially zero," write political scientists Joshua Kalla of the University of California–Berkeley and David Broockman of Stanford University.
"Voters in general elections appear to bring their vote choices into line with their predispositions close to Election Day," they write in a paper to be published in the American Political Science Review, "and are difficult to budge from there."
The researchers are not saying that campaigns don't matter at all. Their study finds they do matter in primary elections, where partisans are choosing who would best represent their party. And they can influence general elections in several indirect ways, including influence the media narrative.
Perhaps most importantly, they add, campaigns can "effectively stimulate voter turnout." Tom Perez, take note.
The study features a meta-analysis of 40 previously conducted field experiments, along with detailed descriptions of nine new ones, which they conducted "in partnership with a national, door-to-door canvassing operation in 2015 and 2016."
"Nearly all these studies found a zero effect on which candidates voted on election day," they report. "Campaign contact in general elections appears to have persuasive effects if it takes place many months before an election, but these effects decay before election day."
The results point to one narrow exception: "Circumstances in which candidates take unusually unpopular positions, and opposing campaigns invest heavily in identifying persuadable, cross-pressured voters whom they can inform about these positions."
Even under those fairly unusual conditions, finding those persuadable voters is a very difficult process that "can only be executed by campaigns with considerable resources and sophistication."
Two of the new experiments are worth highlighting. One was conducted during a mayoral primary in Philadelphia, "a competitive primary for an open seat." The researchers report their partner organization's door-to-door canvassing operation six weeks before the election "increased support for their candidate by approximately 11 percentage points."
The week before the election, this effect had decayed, but only slightly (to 9 percentage points). This makes the point that campaigns can and do help partisans decide who to vote for in a primary.
Another experiment was conducted during the 2016 general election in the swing state of North Carolina. "We found that the partner organization's canvass increased turnout by nearly two percentage points," the researchers report.
Kalla and Broockman conclude that candidates "may under-invest in voter turnout efforts relative to persuasive communication."
"Campaigns," they note, "clearly can influence whether voters bother to vote at all."
So mobilization appears to be the best use of a campaign's limited resources. This research suggests spending time and money attempting to persuade voters may be a fool's errand.