In March, an unnerving statistic began appearing in the headlines of mainstream news reporting: Donald Trump’s 2016 campaign rallies had produced a staggering 226% increase in hate crimes.
For many liberals, a surge in hate-fueled violence probably seems self-evidently connected to Trump’s cruel remarks about specific minority groups, Mexicans, Muslims, etc. When the FBI reported late last year that hate crimes had increased 17% from 2016 to 2017, it seemed to confirm a certain narrative: Look around, and you will see that America has become a more spiteful place.
But there’s a big problem with this thinking: The numbers don’t add up. Counting them is difficult work, and many of the organizations tasked with doing so routinely inflate them. Worse, media outlets that report on the hate trackers’ findings are careless in their descriptions of the data and often don’t understand what is being reported. What results is public disinformation.
Most people would be forgiven for believing the spin from watchdog groups and the media who insist that hate is ever-increasing, and that Trump’s ugly rhetoric is partly to blame, but reality paints a messier picture. In the end, we don’t know if hate crimes are increasing, let alone whether the current political climate is driving such a surge.
Take the 226% statistic, which appeared in the Washington Post, Vox, CNN, The Hill, Business Insider, and elsewhere. This figure comes from a study by Ayal Feinberg, an assistant professor at the University of Texas A&M-Commerce, and Regina Branton and Valerie Martinez-Ebers, both professors at the University of North Texas. The professors found that counties visited by Trump during the 2016 campaign subsequently saw a 226% increase in “hate-motivated incidents.”
“Our research finds counties hosting a Trump campaign rally are significantly more likely to experience a rise in reported hate incidents in the months following the rally,” they wrote in their paper. “We argue this finding elucidates a connection Trump’s political rhetoric has on activating prejudicial attitudes and behavior, which ultimately leads to tangible increases in minority targeting.”
The finding quickly went viral and reappeared after the Washington Post re-published it last month. Many Democratic politicians, including presidential candidate Bernie Sanders, referenced the claim. Sanders posted it to his Facebook page, with a scolding note to Trump: “Your language creates a climate which emboldens violent extremists.”
It might come as a surprise to Sanders, then, that there was another major 2016 candidate whose visitations were correlated with an even greater spike in hate crimes and related incidents: Hillary Clinton.
Two Harvard University Ph.D. students in economics, Matthew Lilley and Brian Wheaton, replicated the work of Feinberg, Branton, and Martinez-Ebers, and made the startling discovery.
“Using additional data we collected, we also analyzed the effect of Hillary Clinton’s campaign rallies using the identical statistical framework,” they wrote in an article for Reason magazine (where I am an associate editor). “The ostensible finding: Clinton rallies contribute to an even greater increase in hate incidents than Trump rallies.”
Lilley and Wheaton’s purpose was not to disparage Clinton but to show that the model used to produce the 226% for Trump rallies was hopelessly, ludicrously flawed. The original study had failed to account for the glaringly obvious fact that politicians tend to hold political rallies in counties where a lot of people live since there’s little point visiting sparsely populated areas. Heavily populated areas also have more hate crimes, for equally obvious reasons. The alleged correlation between hate crimes and political rallies, then, was merely a reflection of demographics.
Lilley and Wheaton put it this way: “No one should be surprised that Orange County, California (population 3.19 million) was home to both more reported hate incidents (5) and Trump rallies (2) than Orange County, Indiana (population 19,840, which had zero of each).”
After controlling for population size, the 226% spike dropped precipitously, indeed “to become statistically indistinguishable from zero.”
It was the viral study’s most significant flaw, but it’s not the only one. In truth, the original authors had not even used proper hate crime figures to calculate their number. The study’s only source of data was the Anti-Defamation League’s 2016 report on hate, extremism, anti-Semitism, and terrorism. The ADL primarily tracks anti-Semitism, and the overwhelming majority of the 1,321 incidents that comprised their report — and thus solely informed the 226% statistic — were anti-Semitic. Many of them would not even be properly classified as hate crimes: Schoolyard bullying of an anti-Semitic bent is loathsome, but not criminal. This was the category of hate that Trump’s rhetoric had supposedly exacerbated.
It would not be the first time that ADL data was used to push a false narrative that hate crimes were spiraling out of control. Late last year, numerous media outlets reported an ADL finding that anti-Semitism had risen 57% in 2017. Both the New York Times and the Washington Post characterized the statistics as referring to anti-Semitic hate crimes. But this was misleading. The ADL report discussed anti-Semitic incidents, not all of which are crimes. These incidents were further broken down into three categories: vandalism, harassment, and assault. A great number of death threats made against Jewish institutions by a single deranged teenager in Israel was largely responsible for a rise in harassment; assaults, on the other hand, had decreased by 47%.
It’s important to understand that not every utterance of a hateful sentiment constitutes a hate crime. “Hate crime” is a legal term, applied to crimes such as assault and vandalism in cases where the motive was animus against a protected marginalized group. Yelling a racial epithet at a random stranger is not a hate crime; but if the perpetrator uses a racial epithet while robbing or attacking a stranger, then authorities may decide to file hate crime charges in addition to assault charges. In practice, the existence of federal laws against hate crimes often gives prosecutors a second chance to convict an accused person if state charges don’t stick, which is why some critics, including Gail Heriot, a professor of law and member of the U.S. Civil Rights Commission, have warned that the entire concept is an invitation to violate the double jeopardy clause.
The FBI compiles yearly statistics on the number of hate crimes committed in the U.S., and they invariably prompt concern when they appear to have risen, as they did last year. But, importantly, the FBI’s figures are based on information reported by local law enforcement agencies across the country, and reporting is not mandatory. About 1,000 additional agencies submitted data in 2017 vs. 2016, and the raw numerical increase in hate crimes was also about 1,000. It means that if each municipality submitting hate crime data for the first time in 2017 had reported just one hate crime, it would account for the entire increase. It’s also possible the FBI may have undercounted hate crimes in previous years, and any subsequent purported surge might be the result of more accurate counting.
Since most municipalities report zero hate crimes, even small increases can look dire if not properly contextualized. Baltimore County, population 830,000, reported just one hate crime in 2016. The following year, it reported ten hate crimes. This finding could be credibly described as a 1000% increase, which sounds terrifying, or it could be just as credibly described as a statistically insignificant fluctuation, which seems somewhat more boring. The devil is not in the details: It’s in how we read them, or how they are read to us by ideologically motivated trackers or sensationalized by irresponsible, contextless media coverage.
Tracking hate has become an industry, and no organization has made more money from it than the Southern Poverty Law Center. Even left- and far-left-wing writers such as Current Affairs’ Nathan Robinson have chided the SPLC for raising vast sums of money off its advocacy and then failing to channel this fortune back into activism. “The SPLC’s misuse of money is outrageous,” wrote Robinson, who noted that the organization raises $360,000 every single day. “Think of all the good it could have done with its millions and hasn’t done.”
The SPLC does not count hate crimes. It counts hate groups. You will not be surprised to learn that there seem to be more of them every year. In the year 2000, the SPLC counted just under 600 hate groups in the U.S.; in 2018, there were 1,020.
That may seem like a frightening increase in hate, but it’s not. Clicking on virtually any state on the SPLC’s map gives away the game. I selected Oklahoma, which is home to nine hate groups. Five of them, though, are black nationalist groups such as the anti-Semitic Nation of Islam. These groups are kooky, and their members say a lot of nasty things — NOI president Louis Farrakhan’s likening of Jewish people to termites comes to mind — but they represent a very low threat of violence, statistically speaking. What’s more, the SPLC provides no information on the relative size or membership of these or other groups. If one insane fringe organization consisting of a dozen or so crazy people splits in two due to internal divisions, the map would count a second hate group, even though the strength of the organization had actually decreased. Perhaps there’s value in keeping a record of these groups wherever they appear, but pretending that they reflect some ever-increasing far-right threat is ludicrous and, in the case of the SPLC, more than a little ideologically motivated.
It’s often observed that hate sells. But fear of hate sells, too. Thus, all too many watchdog group and media organizations have promoted the idea that the level of hate in our society is spiraling out of control. But neither hate nor crime is a new phenomenon, nor unique to the era of Trump. Cruelty should be criticized, and the perpetrators of hate crimes must be condemned. But a genuinely informed public would not pretend that there is some ever-worsening epidemic of hate crimes, or that the data support such a fear.
Robby Soave is an associate editor at Reason magazine and author of the recent book Panic Attack: Young Radicals in the Age of Trump.