Rep. Duncan D. Hunter (R-Calif.) walks past a protester holding a sign for his opponent, Democrat Ammar Campa-Najjar, as Hunter leaves federal court in San Diego in September. Hunter has been charged with using campaign funds for personal expenses. He was reelected this month. (Mike Blake/Reuters)

In the days leading up to the midterm elections, the Trump campaign released an ad that featured Luis Bracamontes, an undocumented Mexican immigrant who had been convicted of murdering two police officers, and blamed Democrats for letting him into the country. The ad was deemed so offensive that CNN refused to air it, and NBC and Fox News eventually pulled it off the air.

Trump’s ad was one of many in 2018 that appealed to voters’ prejudices. Antonio Delgado, a black congressional candidate in New York, was cast as a “big-city rapper.” Andrew Gillum, who was running to be Florida’s first African American governor, was described in a robo-call as a “negro” and “monkey.” In other states, mailers depicted Jewish candidates with fists full of cash. The Guardian asked: “Is this the most racist U.S. midterm campaign ever?”

Perhaps a bigger question is whether these ads actually worked. Weeks after the election, pundits and analysts debated this question. My research sheds light by examining one of the most controversial ads of 2018.

How I did my research

The ad was by Duncan D. Hunter, a Republican incumbent in the House who narrowly won reelection in California’s 50th Congressional District. Hunter’s opponent, Ammar Campa-Najjar, is half Arab and half Latino. In the ad, Campa-Najjar was described as a “security threat” working to “infiltrate Congress” with the help of the Muslim Brotherhood. The Washington Post’s Fact Checker gave the ad four Pinocchios for its false or misleading claims and its “naked anti-Muslim bias.”

My research involved an experiment conducted among a national sample of 1,010 Americans recruited through the firm Lucid and interviewed between Oct. 31 and Nov. 8. The sample was nationally representative in terms of age, gender, race and region.

The survey respondents were randomly assigned to one of three groups. The first group read a short biography of Hunter that detailed his conservative credentials. The second group read the biography and then viewed the campaign ad. The third group read the biography, watched the ad, and then read an excerpt from an article in which Republican and Democratic elites condemned the ad as a “racist and bigoted attack.”

Then everyone was asked how likely they would be to vote for Hunter if they lived in his congressional district.

What I found

Simply watching the ad did not increase or decrease Republican support for Hunter. But among Democrats, those who watched the ad were nine points less likely to support Hunter. This Democratic backlash is consistent with research showing that Democrats have become significantly more supportive of Muslims since Donald Trump’s presidential campaign.

But reading the condemnation of the ad did have consequences. Although opinions among Democrats did not change much, Republicans’ support for Hunter dropped 22 points after they read criticism of the ad.

This finding fits with past research that has shown that explicitly identifying a political message as biased against a minority group can mitigate that message’s impact. People’s own biases are less likely to be activated when they are told that the message violates societal norms. This is particularly true when condemnation of the message comes from leaders of one’s own party. The earlier research has focused mainly on racial bias, but my findings suggest that the same may apply to messages targeting Muslims, as well.

To be sure, real-world campaigns take place in a different environment than a social science experiment. Even if elites condemn appeals to prejudice, many Americans may not receive that message. This may help explain why Hunter won reelection despite bipartisan condemnation of this campaign ad.

Nevertheless, this condemnation may have made a difference. In a district in which registered Republicans outnumber Democrats almost two to one, Campa-Najjar almost pulled off an upset. The implication is clear: When political leaders push back against campaign tactics that play on voters’ racial, ethnic and religious prejudices, the backlash can outweigh any gains.

Maneesh Arora is a PhD candidate at the University of California at Irvine.