After GQ reporter Julia Ioffe appeared on CNN and claimed Donald Trump had radicalized more people than ISIS, she quickly backtracked and apologized for exaggerating. She then offered what she thought was a more solid, factual basis for blaming President Trump for the shooting in Pittsburgh: “According to the ADL, the number of anti-Semitic attacks has jumped by nearly 60% in the first year that Donald Trump was in office.” But that’s not any more accurate than her claim about ISIS, as Robby Soave at Reason pointed out yesterday.

The ADL statistic captures anti-Semitic “incidents,” which is a much broader category of behavior than “hate crimes” or “attacks.” Incidents include things like bullying in schools—which is bad, but usually not indicative of criminal conduct…

The ADL report came up with three subcategories of anti-Semitic incidents: vandalism, harassment, and assault. An increase in vandalism accounts for much of the overall increase, but Bernstein doubts that all of the included incidents were actually examples of anti-Semitism. The harassment category also saw an increase, largely due to a series of bomb threats against Jewish institutions in the U.S. made by a disturbed Israeli teen. It’s not at all clear that these threats were motivated by anti-Semitism.

Finally, the assault category saw a 47 percent decrease.

Soave also refers to this article at the Volokh Conspiracy by David Bernstein which suggests the ADL report is intentionally misleading, at least with regard to the bomb threats against Jewish Centers (something I wrote about extensively when it was happening):

There are several problems with relying on this study for Trump-bashing, however. The first is that the study includes 193 incidents of bomb threats to Jewish institutions as anti-Semitic incidents, even though by the time the ADL published the study, it had been conclusively shown that the two perpetrators of the bomb threats were not motivated by anti-Semitism. One can only guess why the ADL chose to inflate its statistics in this way, but none of the explanations speak well of it…

Third, “college campuses saw a total of 204 incidents in 2017, compared to 108 in 2016.” How many of those incidents emanating from traditional forms of anti-Semitism that one might associate with Trumpian populism, and how many from leftist/pro-Palestinian sources? The ADL doesn’t say.

Fourth, the ADL counts ambiguous incidents as anti-Semitic incidents, so long as they were reported as such. For example, the report states, “Jewish graves or cemeteries were desecrated seven times in 2017…The problem is that desecrations of cemeteries of all faiths is not uncommon, and are often the product of either bored teenagers or vagrants. In fact, at least some of the cemetery incidents counted by the ADL were ultimately determined by police not to be anti-Semitic in origin. The desecraton of a cemetery in St. Louis got a particularly large amount of attention. The police eventually caught the perpetrator, and determined that he was just “mad and drunk,” not anti-Semitic. The ADL has not updated its study or press release to reflect such facts.

I think that should put to rest the claim that there was a major spike of anti-Semitic attacks in 2017. That’s not true if you’re talking about assaults and if you’re talking about the other categories the increase may not be anti-Semitic or may not be coming from the right.

I could wrap this up here but I’d like to point out that the ADL also publishes an annual report titled “Murder and Extremism in the United States in 20xx.” In 2016, the ADL published this striking claim which got quoted quite a few times by people on the left: “Over the past 10 years (2007-2016), domestic extremists of all kinds have killed at least 372 people in the United States. Of those deaths, approximately 74% were at the hands of right-wing extremists, about 24% of the victims were killed by domestic Islamic extremists, and the remainder were killed by left-wing extremists.”

Last year I asked ADL if they could provide the information to back up that claim because the actual data is not available on their website and wasn’t included in the 2016 report itself. Initially, they responded and agreed they would pull together some information for me. But it never arrived. I sent 2 or 3 follow-up emails over a period of months and they never responded to those at all.

The information in the annual extremism reports published by the ADL is based on a database of extremist murders which the ADL keeps to itself. Its definition of extremist murders includes murders that, by the ADL’s own admission, were not ideological in nature. The 2017 report includes an actual list of incidents cited in the report and here’s one that gets counted as a right-wing extremist murder:

Greeley, Colorado, August 16, 2017. Kelly Raisley, believed to be a member or associate of the 211 Crew white supremacist gang, was arrested on first-degree murder charges for the murder of his uncle, Randy Gene Baker. Baker’s wife and sister were similarly arrested. The motive was apparently personal.

Here’s another example where the violence seems to be personal rather than ideological:

North Judson, Indiana, March 3, 2017. Aryan Circle member Edward Blackburn allegedly shot and killed another man who was reportedly dating his ex-girlfriend.

This one also seems to take a lot of liberties with the truth:

Lafayette, Indiana, January 16, 2017. Wesley Andrew Hampton, a self-declared white supremacist, and another defendant allegedly robbed and murdered a man in a home invasion.

Here’s a Journal & Courier story about the robbery and murder in question:

Several people wearing masks pushed their way into the apartment where John Byler was killed about 6 a.m. Jan. 16, according to witnesses cited in a probable cause affidavit.

Witnesses told police the men demanded drugs and money, according to the affidavit.

Police believe Wesley Andrew Hampton, 39, and Aarron Christopher Vance, 35, were two of the people wearing masks.

What the ADL doesn’t tell you is that Aaron Christopher Vance, Hampton’s accomplice in the crime, is black. None of the reporting on this suggests it was a hate crime or even mentions the race of the victim. This appears to have been a drug robbery gone wrong. The shooter may, in fact, be a white supremacist but it’s not clear that mattered here.

These are a few examples but there are several more like this including prison murders which may or may not have a racial motive and other domestic disputes turned violent. All of these are included in the ADL report as “non-ideological murders” but still count as murders by white supremacists/right-wing extremists. Here’s an explanation of that from the ADL’s 2015 report [emphasis added]:

Ideology played a primary or substantial role in 10 of the 17 incidents in 2015, accounting for 34 of the 52 victims. The other seven incidents involved murders in which the perpetrator’s ideology seems to have played little to no role in the violence. Typically, white supremacists make up the vast majority of non‐ideological perpetrators, as white supremacists engage in a large amount of gang‐related and traditional criminal violent activity in addition to their hate‐ or ideologically‐motivated violence. In this sense, they are doubly dangerous.

That’s one way to look at it. I guess it’s fair to say white supremacists are doubly dangerous to their immediate family, but I don’t think that’s what most people have in mind when they skim a report titled “Murder and Extremism in the United States in 20xx.”

Similar to what it is doing with its handling of anti-Semitic incidents, the ADL appears to be padding the numbers. In the case of the extremism reports, the ADL never hid the fact it was including these non-ideological murders, but I suspect most people reading a quote second hand, like the one I started this with above, aren’t fully aware what is included in the bottom line.