First study to find digital ads work -- on millennials

Katherine Haenschen, assistant professor in the department of communication at Virginia Tech said "this is first time we found that digital ads do something and what they do is they increase voter turnout among millennials in municipal elections."

According to research published in Political Communication digital ads increased voter participation in a Dallas municipal election.

Why Dallas? Less than 7 percent of residents, and under 2 percent of millennials voted in their 2015 municipal election, making it the worst major city in the United States for voter turnout.

Discouraged by that statistic civic leaders wanted change. The effort led to a collaboration with the Dallas Morning News, Jay Jennings of University of Texas, and of course Haenschen.

In the study, millennials were exposed to two or four weeks of ads in the month leading up to the election. One set of ads focused on providing information about city council and school board candidates published by Dallas Morning News. The other set of ads served as election participation reminders. Some groups were exposed to both sets of ads (information and reminders) while other groups only saw one set of ads. At most, people saw ads four times per day.

"Since many adults encounter over 2,000 ads a day," we gave people a very small amount of ads and were still able to change their behavior," Haenschen said.

In competitive districts, when millennials were exposed to all four weeks of ads, voter turnout went up. In non-competitive districts the effect was the opposite, suggesting users in uncontested districts may have chosen not to participate.

One of the more surprising findings was the effect the digital ads had on millennials who "are a notoriously difficult demographic to reach," said Haenschen. "They don't have landlines and move around a lot making them difficult targets for candidates."

That is why Haenschen and collaborators chose to use cookie-targeted digital ads for the study. "Millennials' IP address is perhaps more stable than their physical address," said Haenschen.

But the thing that excites Haenschen the most about this research was it showed a path to mobilize people who had never voted in an election.

"One thing that I think is so great about our study is that it was able to mobilize people who had never voted in a municipal election before," said Haenschen. "So you figure now these folks have been mobilized to vote for their city council member and school board one time. Now they will be on lists of people who voted in prior municipal elections, so future candidates will reach out to them. We can hope that it may have snowball effect over time, and this can be a way to systematically increase turnout among a tough demographic."