Category: Who knows

I’m guessing everyone is familiar with the Black Lives Matter group – and their contention that African Americans (and, by implication, minorities in general) are currently being treated with disproportionate and unnecessary violence by law enforcement.

Years ago, in parts (maybe much) of the country that was indeed true.  But the recent BLM allegations raise an interesting question:  is it still true today?

The conventional “wisdom” is that it is – and the media has done far more than its share to reinforce that conventional wisdom.  However, in reality it’s a surprisingly hard question to answer.

One can’t simply count arrests, or even incidents of police use of force – because in some cases police use of force is clearly warranted, and arrests (and crime) are decidedly not distributed uniformly, either geographically or among racial and ethnic groups.  So determining whether one racial/ethnic group is being treated “better” or “worse” than another with respect to police violence is not an easy problem.

. . .

Indeed, one reason that the question hasn’t been studied is that meaningful data needed to study the problem wasn’t readily available.

That’s no longer the case.  It seems as if one guy – a man named Roland G. Fryer, Jr. – decided to study the problem.  He and a group of people working with and for him have conducted a survey of data from multiple representative jurisdictions and data sources to attempt to extract such meaningful data.  They’ve also rigorously analyzed the extracted data.  The project took an aggregate of 3,000+ staff-hours of work.

The conclusions of Fryer’s study are quite interesting.  Per his study, it turns out that minorities are more likely to be on the receiving end of police violence.  That is, they are . . . until they aren’t.

I’ll explain.

The study included data from three metro areas in Texas (Houston, Dallas, and Austin); six counties in Florida (Brevard, Jacksonville, Lee, Orange, Palm Beach and Pinellas – which include the St. Petersburg, Fort Myers, Jacksonville, and Orlando metro areas as well as a good portion of Florida’s east coast north of Miami); LA County in California; and NYC.   Arrests reports and NYC “stop and frisk” data from a fifteen-year period (2000-2015) were studied; the relevant data was extracted and characterized.  This extracted data was combined with public survey results concerning citizen perception of police use of violence.  Control measures to remove bias due to officer race, gender, experience level, precinct, and location of crime were also developed.

After doing this, a rigorous statistical analysis of the resulting data set was conducted.  It sought to determine whether minorities taken in to custody by LE were more or less likely to receive varying degrees of violent treatment.  Violent treatment was defined as consisting of multiple escalating categories, beginning with being being pushed/shoved, and including being handcuffed, put forcibly on the ground, batoned, having a weapon drawn or pointed, being pepper-sprayed, and ending in being tazed or shot with a firearm by LE personnel.

The results showed that yes – minority citizens did appear statistically substantially more likely to receive nonlethal rough treatment from LE.  But then they looked at more extreme levels of violence – and a curious result emerged.

As the level of violence increased, the difference between treatment of minority subjects and the treatment of white subjects remained roughly constant.  That is, it remained roughly constant until the most extreme levels of police use of force (taser or firearms) were reached.

At that point . . . the bias against minorities vanished.  In fact, it actually appears to have reversed.

The study extensively studied the most extreme uses of police force – shootings and the use of a taser.   And when it came to being on the receiving end of a shooting, well  . . . Black Americans were statistically substantially less likely – specifically, greater than 20% less likely – to be shot during a police encounter than their white counterparts.  This seemingly aberrant result held true (though the magnitude was reduced) even after multiple corrective measures were applied to the data to correct for various sources of possible unrelated correlations leading to false indications of bias.

The same appears to be true for incidents involving use of a taser (extensive data on that form of incident from Houston, TX, was available).  However, the study doesn’t directly address that issue in its conclusions.

The result observed for raw data relating to these shootings was well outside the statistical standard error for the data on hand.  That means the result appears to be statistically significant – and thus real.

These results also held true even when type of encounter was considered as a factor – e.g., when attempts were made to categorize police shooting incidents by whether or not police use of deadly force appeared clearly justified.  Even in events where a police officer would appear to have been clearly justified in shooting, minority offenders appeared to be statistically less likely to be shot than white offenders under similar circumstances.

The study also found that both Black and white suspects involved in a police shooting were roughly equally likely to have been armed.

The same general results, with smaller magnitude, for shootings were observed in the raw data for police shootings involving  Hispanic Americans – that is, Hispanics were also observed to be somewhat less likely to be shot (about 8.5%) during a police encounter than white Americans.  However, in the case of Hispanic Americans the difference appears to have been within the statistical standard error of the data set, so the result cannot be considered statistically significant.

In short:  the study found no evidence of racial bias in police shootings.  When police were forced to shoot, race was apparently not a factor.  In fact, if anything white citizens seemed statistically a bit more likely to be the ones on the receiving end of police lead, statistically speaking.

There may well be a problem with police using rough treatment inequitably.  But any inequity doesn’t appear to extend to shootings.  Those appear to be race-neutral.

. . .

So, you ask:  just who is this Roland G. Fryer, Jr.?  Is he some John Birch Society type, or perhaps a KKK or Aryan Brotherhood member?  This study is all just bogus claptrap cobbled together by some ignorant racist – right?

Hardly.

Fryer happens to be a university professor with a doctorate degree – a professor of economics, specifically.  So he knows a thing or two about data collection, data reduction, and statistical analysis.

Fryer also isn’t some some second-rate academic teaching at some non-accredited college run by religious fanatics, or at some local community college or small, second-rate college.  The guy happens to be a tenured professor at Harvard.  Yeah, that Harvard.

Finally:   Fryer happens to be African-American.  In fact, he’s the youngest African-American professor ever to receive tenure at Harvard – as well as the first to receive recognition as the most promising US economist under the age of 40, the John Bates Clark medal.

Fryer has indicated that his result concerning shootings and minorities are “the most surprising result of my career”.  But he appears to be standing by his conclusions – though he does offer the standard caveats (i.e., incomplete data, possible data unreliability due to voluntary participation, etc . . .) one would expect regarding any such study.  He also indicates that more research is needed to reach a definitive conclusion regarding police use of violence nationwide.

Fryar has offered a theory concerning the disparity, one which is supported by his data but which he admits that though descriptive may not be correct.  His theory is that police essentially “do what works”, and that some fraction of police are indeed racially biased.  But shooting or tazing someone impose high negative consequences on a policeman even if justified, so at that point reason takes over from bias in those individuals who are racially biased and they act responsibly.

FWIW:  I think Fryar may well be on to something with his theory.  Racism does still exist, but IMO it’s hardly the pervasive conspiracy some claim.  Learning the truth – vice listening to BS from those with an agenda, or who profit from stirring up trouble – is the first step to improving the situation.  IMO Fryer has made a contribution to learning the truth here by creating the first data set suitable for detailed study, and for conducting the initial analysis.  I hope he manages to get the wherewithal to study the matter further.

. . .

You can read an article about Fryer’s study published by the NYT here.  (It’s the NYT, so if anything IMO it somewhat downplays Fryer’s conclusions.)  Or you can read the study for yourself here – and can read the appendix to the study, which gives additional details, here.  Fair warning:  there is indeed math involved.  (smile)

Unfortunately, I’m guessing Prof. Fryer is about to become persona non grata among his leftist academic colleagues.  And I’m guessing those leftist colleagues will pull out all the stops in an attempt to block publication of his study or to convince him to change it substantially (it’s currently in pre-publication draft and has yet to be published).

Sometimes the truth is indeed inconvenient.  Especially when it’s backed by hard data and doesn’t “support the agenda”.