When Donald Trump writes anti-Muslim tweets, anti-Muslim hate crimes spike, a new study released Monday suggests.
The study, conducted by researchers at the U.K.’s University of Warwick, found Trump’s anti-Muslim tweets were a reliable predictor of attacks on Muslims during the presidential campaign and in the months following his election. Although Trump’s tweets sometimes came in response to incidents like terror attacks that might have been used as justification for hate crimes, those crimes peaked in counties with higher-than-average Twitter use, suggesting Trump’s tweets might have prompted the violence.
“To be clear, we do not claim that Donald Trump himself causes hate crimes out of thin air,” Carlo Schwarz, a doctoral student who worked on the study, told The Daily Beast. His co-author on the study is fellow Warwick doctoral student Karsten Müller. “But what we think is interesting is that Trump’s tweets and hate crime only appear to be correlated after the start of his presidential run. It is also interesting that this correlation seems to be driven by areas with many Twitter users.”
In early December 2015, days after a terror attack in San Bernardino, California, then-candidate Trump took to Twitter with an anti-Muslim tirade. In a series of tweets, he proposed a ban on Muslim immigration and accused the U.K. of “trying hard to disguise their massive Muslim problem.” Offline, reports of anti-Muslim attacks spiked; in the days after Trump’s tweets, the FBI documented approximately 25 anti-Muslim hate crimes, up from approximately five the previous week.
Among those crimes was the beatdown of a Queens, New York, shopkeeper by a man pledging to “kill Muslims,” an attack on a Muslim man playing volleyball in a California park, and the discovery of a severed pig’s head on a Philadelphia mosque’s doorstep.
In general, “Trump’s Muslim tweets alone predict more than 20% of the variation in anti-Muslim hate crimes in the same week,” the University of Warwick researchers found. Their study compared Trump’s tweets against the FBI’s record of hate crimes through 2016, the last year for which those statistics are public.
Other sharp spikes in anti-Muslim hate crimes occurred after a terror attack in Brussels in March 2016, which Trump tweeted about, and after Trump’s election. The post-tweet spikes did not correlate with other trends in hate crimes, the study found, suggesting the incidents came from surges in specifically anti-Muslim sentiment.
“These correlations are consistent with at least two possible explanations,” the researchers wrote in the study. “One is that Donald Trump became more reactive to US-wide anti-Muslim sentiments. Another possibility is that Trump’s tweets themselves contribute to a climate that enables hate crimes.”
To test the reach of Trump’s tweets, the researchers studied the frequency of anti-Muslim hate crimes in counties whose residents were in the 90th percentile of Twitter usage. The difference between those counties, and others with less-online populations was striking.
“Counties with many Twitter users experienced 12 times more anti-Muslim hate crimes per capita than those with few Twitter users under Trump,” the researchers found.
Counties with heavy Twitter use weren’t particularly prone to hate crimes before Trump’s candidacy, the researchers said.
“In comparison to all other presidents, Trump is a striking outlier,” the researchers found. “In fact, the standardized number of per-capita hate crimes under Clinton’s two presidencies was lower in areas where many people use Twitter today. The differences were relatively muted under Bush, with a slight increase for Obama’s two terms,” they wrote, noting that Obama’s second term coincided with Trump’s candidacy.
Schwarz said the study’s findings were consistent even when they adjusted for demographic differences across the counties.
“The increase in anti-Muslim hate crimes is concentrated in counties with many Twitter users even if after we subtract the average number of hate crimes in that county,” he wrote. “Our findings also do not change if we take into account a large number of other variables, e.g. a county’s demographics or economic condition.
Prior to his candidacy, Trump’s Twitter wasn’t a good indicator of anti-Muslim crime. Although the now-president authored anti-Muslim tweets years before announcing his presidential bid, those tweets did not drive a spike in hate crimes, the researchers found.
Until Trump launched his Muslim-bashing campaign, “the number of hate crimes was more or less constant since 2009,” the researchers found. “With the start of Donald Trump’s presidential campaign on June, 16th 2015, however, we observe a disproportional increase in the number of hate crimes in those counties where many people use Twitter.
“There is no comparable increase in counties with low Twitter usage.”