Recent US elections have illustrated the role that the way many Americans feel about race plays in shaping who they vote for. In new research,Nazita LajevardiandJan Zilinskyinvestigate how anti-Muslim and anti-Black Lives Matter sentiments shaped support for Donald Trump during his first term in office. They find that while from 2016 to 2019, attitudes towards Muslims were significant predictors of the Trump vote, by the 2020 election, following George Floyd’s murder, attitudes towards Black Lives Matter became more important in shaping support for Trump.
Time and time again, research has pointed to the important and outsized role that race plays in American politics. But do the attitudes towards certain groups which matter the most for political outcomes shift dramatically from election to election? While these group based attitudes are formed early on someone’s life, these predispositions may vary considerably in how they are primed to shape political preferences in a given election, depending on whether they are activated through communications from the media and elites like politicians.
Trump’s attacks on the Muslim community
To be sure, since Donald Trump’s rise on the political scene in 2015, his approach has been explicitly imbued with racial hostility – or racial animus – about many communities. But arguably, no group was as denigrated and attacked by Trump in the 2016 presidential campaign as Muslims. From calling Syrian refugees a “secret army” of 200,000 ISIS-linked people sent to destroy America, to arguing for shutting down mosques in America, stating that “Islam hates us,” saying that Islam is a “sickness” and Muslims “a sick people,” to falsely claiming that Muslims in New Jersey danced and cheered after the collapse of the Twin Towers on 9/11, Trump’s repeated anti-Muslim remarks on the campaign trail did not go unnoticed by many voters: research found that anti-Muslim attitudes were one of the most important predictors of Trump support in the 2016 election.
Yet, by the 2020 presidential election, the national conversation about race had changed dramatically. On May 25, 2020, George Floyd was murdered by a white police officer in Minneapolis, Minnesota. Floyd’s inhumane 9 minute and 29 second murder was filmed, posted to the Internet, and viewed world-wide. The national response was immediate. While many protests erupted across the United States in favor of the “Black Lives Matter” (BLM) movement, others galvanized to protest in favor of the police under the “Blue Lives Matter” movement. These protests were supplemented by high levels of elite and media attention, and in the months leading up to the presidential election, this issue arguably was among the most important and prominent topics in the national discourse.
“SUPPORT TRUMP MILITARY & POLICE” (Public Domain) by dankeck
How did anti-Muslim attitudes shape support for Trump in 2020?
This shift in the national conversation begs the question: did anti-Muslim attitudes continue to meaningfully shape support for Donald Trump in the 2020 presidential election, or did attitudes toward a wholly different set of groups become important and influence support? Our recent research answers this question and reveals a remarkable shift in voter psychology over this period. Using detailed survey data tracking the same voters over time, along with weekly polling throughout 2019 and 2020, we were able to pinpoint exactly when this shift occurred.
First, our analyses of panel data finds that attitudes toward Muslims were significant predictors of the Trump vote in 2016, and even through 2019, well into year three of Trump’s first administration. However, by 2020, attitudes toward Muslims were no longer strongly (nor significantly) related to the 2020 Trump vote. Our panel data analysis also assesses how evaluations of BLM and the police changed among people during this period. We find that while evaluations were poorly associated with the Trump vote in 2019, these evaluations became remarkably predictive of candidate support by 2020 (positively for the police and negatively for BLM). To put these findings into context, while attitudes towards BLM and the police weakly predicted Trump vote in 2019 (49.4 percent and 36.5 percent for the coldest (least supportive) BLM and police respondents; 31.0 percent and 45.9 percent for the warmest (most supportive) BLM and police respondents), they dramatically rose in explanatory importance by 2020 (86.1 percent and 0.6 percent for the coldest BLM and police respondents; 1.5 percent and 65.2 percent for the warmest BLM and police respondents).
Second, we supplemented the panel analyses with high frequency weekly data from Nationscape to estimate exactly when group evaluations shifted in the last two years of Trump’s presidency. This data confirms that the effect of anti-Muslim attitudes on Trump approval was large throughout 2019 and until the Floyd murder. But immediately after the murder, its average value declined substantially, and pro-police attitudes (which were always an important indicator of Trump approval) rose substantively in importance and this continued until the 2020 election. Our results also reveal just how important attitudes towards Black Lives Matter (BLM) were in shaping Trump approval. Though the BLM favorability item was only queried after the Floyd murder, we find that attitudes towards BLM throughout this period had the largest substantive effect on Trump approval compared to the other groups where we asked about favorability ratings
Together, our research highlights just how the groups and issues that matter most in elections can change rapidly, depending on the social and political climate. These findings demonstrate how fluid and context-dependent voters’ attitudes can be. While deeply held prejudices may persist, which attitudes drive voting behavior can change rapidly in response to major events and shifts in the national discourse.