2014-11-26

Statistics: death at the hands of Police

The entire nation seems obsessed about a single death in Missouri this past summer.

Somehow, the death of a single man of a minority race at the hands of a Police officer of a majority race has turned into an excuse for a violent mob to break windows, take things, and blame Other People for problems that are likely caused by a complex mix of social forces. (This mix of social forces includes, but is not limited to, attitudes like 'I have a problem with Society, and I can express that problem by destroying/stealing property that belongs to others.')

In all of this, I stopped to think about the general subject of death-at-the-hands-of-Police-officers, as compared to violent death in general.



Two sets of statistics are easily available to me.

One set is the Fatal Injury Reports maintained by the Centers for Disease Control and Prevention (part of the U.S. Dept. of Health). This is apparently data collated from medical sources, coroners, and related sources close to the medical profession.

The second set of data is the Federal Bureau of Investigation's Uniform Crime Reports. This is data collated from reports issued by all Police Departments and Law Enforcement Agencies (ranging from small town Police Departments and County Sheriffs to various State and Federal agencies that employ sworn Law Enforcement Officers).

The CDC data source allows queries under "Intent or Manner of Injury". One category in that section is "Homicide". Another is "Homicide: Legal Intervention". The CDC data also allows queries by Race.

At first, I queried data for "Homicide: Legal Intervention", and created this chart. (The FBI stats for these years produce similar numbers for "Justifiable Homicide by Law Enforcement", with some variation. However, the FBI does not report Race of Victim on that table.)



Apparently, the absolute numbers of Whites killed by Police vary between 2x and 3x the values for the absolute numbers of Blacks killed by Police. (The CDC also provides rates-per-100,000-of-selected-group. Under that measure, Black death rates typically ran at between 0.29 and 0.32 per 100,000, while White death rates typically ran between 0.13 and 0.18 per 100,000.)

Noticeably, the absolute values and the per-capita rate for Blacks haven't risen as fast as the same values/per-capita-rates for Whites.

A lazy statistician could come to the conclusion that Whites have more to fear from the Police than Blacks, since more Whites die from "Legal Intervention". A less-than-lazy statistician might conclude the opposite, based on the per-capita rates.

Then I queried the data for "Homicide", and created this chart.

One really, really obvious item: for all racial categories, "Homicides" in general dwarf "Homicides: Legal Intervention" by a huge margin.

Another really obvious item: Homicides are split almost equally between White and Black by absolute numbers. However, the per-capita rates for Blacks are something like 4x the per-capita rates for Whites.

The FBI data produces similar numbers for Homicide, with Victim identified by Race.

The FBI also provides an data table for race of Victim and Offender, if both are known.
Again, we notice a vast disparity between two categories. Homicides in which victim and offender are the same race far outnumber homicides in which victim and offender are of different race. Though there is an uncomfortably large number of cases in which the race of the offender is not known/reported.

My conclusion: Black people have far more to fear from criminals than from Police.

No comments:

Post a Comment

I like thoughtful feedback; I prefer polite feedback.

I don't like screeds.

Comments older than a few days will have comments go into moderation.