��ɫApp

People still skeptical of the COVID-19 vaccine can be persuaded to get vaccinated after watching a public service-style announcement featuring former President Donald Trump and his family encouraging voters to get the shot, according to a new study that included researchers from ��ɫApp, the University of North Carolina at Chapel Hill, North Carolina State University and the University of California, Berkeley.

The researchers created an advertisement that included an interview between Trump and Fox News anchor Maria Bartiromo where Trump urged his supporters to get vaccinated. The video, which also included a Fox 13 News Utah broadcaster sharing Trump and First Lady Melania Trump’s vaccination status, was placed on more than 100,000 YouTube channels – including Fox News’ own YouTube channel, where it ran before segments with some of the network’s most prominent personalities like Laura Ingraham, Tucker Carlson, Sean Hannity and others.

The team found that the ad had a significant impact on vaccine uptake: the 1,000 low-vaccinated counties across the U.S. where they showed the ad had an average of 103 more recorded vaccinations per county compared to a group of similar counties where the researchers withheld the ad, totaling to an increase of 104,036 vaccinations overall. The findings were released April 4 .

Persuading vaccine-averse Trump supporters to get vaccinated

Vaccination uptick has faced a political divide, with Republicans lagging Democrats.  found that among the 27% of American adults who remained unvaccinated, 60% identify as Republicans, compared to only 17% as Democrats.

“While a majority of both parties are vaccinated, those who remain unvaccinated are largely Republicans, despite messaging from the CDC and medical experts about the efficacy of the COVID-19 vaccine,” said Brad Larsen, an assistant professor of economics at ��ɫApp’s  and a faculty fellow at the  (SIEPR) and one of the study’s authors. “We felt like there should be a better way to send a message that would resonate with people on the right.”

Inspired by social science research that has shown partisans follow cues from party leaders, the researchers wondered whether vaccine-hesitant Trump supporters could be swayed to get vaccinated if exposed to a potent message that Donald Trump had received and endorsed the vaccine, a fact the former president has touted on several occasions.

The research team put the question to the test in a large, randomized controlled trial at the county level.

The researchers, who also include political science professors from the University of North Carolina Tim Ryan, Marc Hetherington, and Rahsaan Maxwell, Steve Greene from North Carolina State University and economics Professor Steve Tadelis from UC Berkeley – worked with a professional video editor  that opens with a Fox 13 News Utah broadcaster saying “Donald Trump is urging all Americans to get the COVID-19 vaccine.” It quickly transitions to a phone interview between President Trump and national Fox News anchor Maria Bartiromo where Trump says, “I would recommend it, and I would recommend it to a lot of people that don’t want to get it, and a lot of those people voted for me, frankly.” Bartiromo is shown nodding in agreement.

“We felt like there should be a better way to send a message that would resonate with people on the right.” SIEPR Faculty Fellow Brad Larsen

Next, the ad shows footage of Trump with First Lady Melania Trump where the Fox 13 News anchor tells viewers that the Trumps had both received their vaccines privately in January 2021 at the White House. Then, a photo Ivanka Trump  of receiving the vaccine herself was shown, with her own endorsement: “Today I got the shot. I hope you do too.”

The ad closes with a brief statement: “Your vaccine is waiting for you.”

Finally, the ad links to the original Fox News broadcast to show viewers that the video clips were not taken out of context.

The researchers then rolled out a targeted campaign over YouTube’s advertising platform, Google Ads, where they were able to home in on counties with low vaccination rates – areas where the ads could be most effective, Larsen explained.

Between Oct. 14-31, 2021, the team assigned 1,083 counties across the country to receive the ad. The experiment also included a control group made up of 1,085 similar counties where the ad was not shown.

In total, the campaign resulted in 11.6 million impressions with 6 million unique viewers.

The researchers had no control over which channels the advertisement would be featured on YouTube. The fact that it appeared the most on Fox News’ YouTube channel was driven by Google Ads’ algorithms. The ad also appeared on thousands of other YouTube channels, such as Forbes, NBC, MSNBC, Glenn Beck and even Saturday Night Live.

To analyze the effect of the ad campaign, the researchers linked their experiment counties with data from the Centers for Disease Control and Prevention (CDC) on the number of vaccines administered in each county up through each date from Sept. 15 through Nov. 30, 2021 – one month before and after their advertising campaign.

In sum, they found that:

• Vaccinations in the average targeted county increased by about 103, totaling 104,036 across all counties.
• It cost on average about $1 or less in advertising to lead one more person to choose to receive the vaccine (about $99,000 was spent on advertising).
• More ads lead to more jabs: researchers found 1,000 additional ads resulted in 8.6 additional vaccines.
• Counties that received more ads per capita, or where viewers engaged with the ad more by watching more of it or clicking on the Fox News link, had larger vaccine increases.
• The counties that were the most responsive were those with a strong but not overly strong Trump base – counties with up to a 70% vote share for Trump. Those with the highest Trump vote shares were more obstinate and unmoved by the campaign.

Beating politics with more politics

The study reveals that, even among the most vaccine-hesitant crowd, the right message and messenger can change attitudes.

“As many have observed, a tragedy of the COVID-19 pandemic is the extent to which protective measures became tangled in Americans’ political identities, which led to deaths and suffering that could have been avoided,” Larsen and his co-authors wrote in the paper. “If politics characterizes one aspect of the problem, it might also point to part of a solution.”

The study received financial support from the Vaccine Confidence Fund. The experiment described here was approved under IRB.