Wednesday, November 13, 2019

The FBI Hate Crime Report For 2018. My Criticisms.


The FBI published its hate crime report for 2018.  I read several media takes on its contents, then read the contents themselves, then turned irritated and even angry.

And now I'm going to share all the reasons for my irritability with you!  (Not to worry; you will also learn interesting shit about the report.)

First,  different media outlets chose to stress different findings of the report in their headlines.  The New York Times chose to stress the percentage increase in hate crimes against Latinos, NBC News the LBTQ share among the victims of hate crimes, with gay men being the majority of the victims within that group, and so on.

More generally, some sites concentrated on percentage increases in crime (often calculated from a low initial base) while others talked about the actual levels of various hate crimes.  NPR sometimes did one, sometimes the other.  And none of the outlets I consulted looked at the hate crime numbers in proportion to the size of the underlying target populations.

Doing that matters.  To see why, note, for example, that the Jews were the target in 57% of all religiously motivated hate crimes.  Given that the Jewish population in the United States is estimated to be between 1.7% and 2.6%, the high absolute counts of anti-Jewish hate crimes become even more troubling (1).

So much for the media coverage.  What about the report itself?  It turns out to suffer from umpteen different problems.  The data set is full of gaps:

As with previous years’ data, it’s incomplete at best and misleading at worst. Of the 16,039 law enforcement agencies the FBI relies on to report hate crimes to its national database, only 2,026 ― a bit more than 12% ― actually did so.

The remaining 14,013 agencies, roughly 88% of the total, reported zero hate crimes whatsoever. Sadly, that’s not because of an absence of hate crimes, but an absence of reporting. Those 14,013 agencies police more than 100 million people across 49 states, and collectively claimed not a single hate crime occurred in any of their jurisdictions.

And the definitions used for defining a crime as a hate crime are likely to differ between various localities:

Of those that are reported, police or prosecutors may not document them as a hate crime. Different agencies have different definitions of what constitutes a hate crime, while different officers may have different biases and different standards of reporting. And even then, data recorded at the local and state levels isn’t necessarily shared with the FBI at the federal level.
Thus, what the FBI report covers should be taken with a large pinch of salt, to being with.

Given that, what else might we learn from the report that should have interested more of the media outlets which summarized the report but did not (2)?

The report actually gives us some data on those offenders who could be identified.  In particular, we are told the race and/or ethnicity of offenders in cases where those were known.  These are reported in Table 3 by the type of hate crime and in Table 5 by the motivation for a hate crime. (3)

I found it extremely odd that we can get some race/ethnicity data on a subgroup of offenders but no sex (or gender) data!  For some reason that is not part of the collected information.  But the odds are pretty good that if the race or ethnicity of the offenders could be determined so could their sex or gender.

And this brings me to the final victim category in the FBI hate crime report that I wish to address:  that of hate crimes motivated by the hatred of either men or women.  Table 4 in the report tells us that sex-based hate crimes in the 2018 report consisted of 32 anti-female and 26 anti-male incidents.  Among those, one woman was killed for being female.

Given that the Tallahassee yoga studio killings, with two female victims, were explicitly motivated by misogyny, the 2018 FBI report cannot have included them.

Indeed, my guess would be that most crimes motivated by misogyny are simply not recognized as hate crimes.  They are far too ubiquitous (4)  which makes them, paradoxically, much harder to see as hate crimes. 

To see what I mean, take the crime of intimidation which is mentioned in the report as one crime category.  When women face intimidation it is almost always flavored with a strong sense of misogyny (5), but because this is so very common those who report on hate crimes are unlikely to spot the hate nature of such incidents.

-------
(1)  Tables 4 and 7 in the FBI report give absolute counts of various types of violent crimes (divided into homicide/manslaughter, rape, aggravated assault, simple assault, intimidation and other) for each of the hate categories the report considers (race, religion, sexual orientation, disability, gender, gender identity).

Keep in mind that these are absolute counts, that the target populations vary immensely in size and that for most purposes the data needs to be interpreted in proportion to the size of the affected victim population.

Also note that the data on homicides/manslaughter are very sensitive to the effects of one or two mass murders within a given reporting period.  In 2018 the Pittsburgh synagogue killings alone accounted for eleven of the total twenty-four hate crime murder victims.  Thus, one hate crime incident can cause many deaths.

(2)  Surely the data on offenders should be of obvious interest to the media, too.  If we find, for example,  that one demographic group in the population is over-represented among the offenders, then our solutions to hate crime prevention could focus on that group and so on.

(3) You can analyze that data further by relating it to the population sizes of various ethnic and racial groups.  Remember, however, that the data is full of gaps here, too, and that the reporting itself may have had biases.

(4)   And often mixed up with various sexual violence motives.

(5)  Expletives such as "cunts," "bitches" and "sluts" are to me signs that intimidation is fueled, at least partially, by the hatred of women.  Such expletives are extremely common.