The research centered on a $50,000 digital ad campaign during a non-partisan May 2017 Dallas municipal election where the mayoral seat wasn’t up. 

The study was conducted in Dallas at the request of the publisher of the Dallas Morning News “due to the city’ s historically low levels of municipal participation particularly among Millennials.”

“Dallas ranked last in a study of municipal turnout in large cities, with only 6% of voters participating in 2015,” according to the study. “Among Millennial voters, the numbers are even worse: only 1.7% of registered voters under age 34 participated in the 2015 Dallas mayoral election, ranking the city third from last nationally.”

The experiment covered a pool of 73,909 Millennial voters, ages 23-35, in 14 single-member districts where half were competitive.

They were exposed to pre-roll and banner ads for four weeks after being cookie-targeted by Audience Partners, the vendor who partnered with the researchers.

Some of the ads featured animated figures spelling out the word “vote,” and noting “everybody’s doing it.” 

“We wanted to give the impression that turnout was high to combat the impression by Millennials that people don’t vote,” Haenschen said. “If you want to use digital ads to increase turnout, emphasize that turnout is high. Humans are social animals so we really respond to these things. We’re hard-wired to want to go with the herd.”

Other spots promised voters could get “get trusted information on candidates and issues” if they clicked on the ad. The click-through went to a page on the Dallas Morning News where articles about the races were aggregated. 

After the Saturday Election Day, turnout was measured using public voting records. Haenschen said they increased voter turnout by 0.9 percent among people exposed to the ads.

She emphasized that the impact was greatest in races that were competitive.

“You really need to be testing these tactics when the people you’re testing them on have some sort of competitive election to vote in,” she said.

She noted that digital spots are “not the most cost-effective method” for voter turnout. That would be social pressure mailers, which can increase turnout by up to 3 percent. But digital ads, she said, are great for targeting voters for turnout in hard-to-reach areas — apartment dwellers or people in gated communities.

“This is a tool that should be used in compliment with broader campaign strategy,” she said.