Police Shootings & Racial Bias

The fundamental assumption underlying the protests and riots that followed the death of George Floyd is that racism in police departments across the US is responsible for the fact that a disproportionate number of Black males are shot by the police.  I say “death” and not “murder” not because I have any doubts about the facts because I suspect that the constant use of the latter term, especially by such persons as the Minneapolis chief of police, may make it harder for a court to do justice in this case. 

My purpose here is to question that assumption.  About twice as many White persons are shot and killed by police officers as Black persons, according to the Washington Post database.  That is obviously not a very useful statistic.  Non-Hispanic Whites make up about 76% of the US population; Blacks make up a little over 13%.  The WaPo helpfully explains that Black Americans “are killed by police at more than twice the rate of white Americans”.

A very good and recent summary of the statistics for fatal officer-involved shootings (FOIS) can be found in “Risk of being killed by police use of force in the United States by age, race–ethnicity, and sex.”  The authors do not address causation but confine their study to outcomes.  Here is a fascinating chart of FOIS restricted to males. 

The chart measures deaths by police per 100,000, controlling for population share, and tracks the numbers by age.  Unsurprisingly, the fatality rate rises dramatically at the teen years and drops steadily after 30 for all the age groups.  The most important fact is that the hump for Black males is much steeper than for any other defined group, and much lower than for White males. 

Does this chart point to a racial bias in FOIS nationally?  There is one anomaly that doesn’t support that explanation.  Asian Americans are proportionately less likely than Whites to be killed by police.  If the difference between the White and Black curves is evidence of racial bias against the latter, wouldn’t the low incidence for Asian Americans be evidence for a pro-Asian bias on the part of police forces across the country?  That hardly seems plausible. 

A more significant problem is that the racial bias explanation relies on a generally silent and implausible assumption: that, in the absence of racial bias, the FOIS stats for each demographic group would be perfectly proportional to its share of the total population.  That, in turn, assumes that all the demographic groups are exactly the same for all relevant characteristics.  That is also implausible. 

Another chart leads us in the right direction. 

This is about as robust a difference as one ever sees in social science.  Males are far more likely than females to be killed by police for all groups measured here.  This not, let me go out on a limb, gender bias on the part of the police.  It is a consequence of the fact that men are more likely than women to commit the violent crimes that might bring them into contact with the police and far more likely to escalate once they are exposed to the police.  Here is a bit from the Bureau of Justice Statistics (2011) tracking stats between 1980 and 2008. 

Males represented 77% of homicide victims and nearly 90% of off enders. Th e victimization rate for males (11.6 per 100,000) was 3 times higher than the rate for females (3.4 per 100,000). The offending rate for males (15.1 per 100,000) was almost 9 times higher than the rate for females (1.7 per 100,000).

Is there a connection for the Black/White differential FOIS rates and a difference in rates of violent crime?  Again from the BJS:

Blacks were disproportionately represented as both homicide victims and offenders. The victimization rate for blacks (27.8 per 100,000) was 6 times higher than the rate for whites (4.5 per 100,000). The offending rate for blacks (34.4 per 100,000) was almost 8 times higher than the rate for whites (4.5 per 100,000).

Two recent papers examine the relationship between police shootings and criminal activity among Blacks and Whites: here and here.  Here is a bit from one of them:

We first reproduce the well-known finding that Blacks are more likely to be fatally shot than Whites given population proportions… the odds were 2.5 times higher for Blacks to be killed by police compared to Whites given their population proportions.

However:

When fatal shooting data are benchmarked against the number of murder/nonnegligent manslaughter reports and arrests, the odds ratio obtained when benchmarking against population proportions flips completely. The odds were 2.7 times higher for Whites to be killed by police gunfire relative to Blacks given each group’s SRS homicide reports, 2.6 times higher for Whites given each group’s SRS homicide arrests, 2.9 times higher for Whites given each group’s NIBRS homicide reports, 3.9 times higher for Whites given each group’s NIBRS homicide arrests, and 2.5 times higher for Whites given each group’s CDC death by assault data.

In other words, given rates of homicide reports and arrests, across three databases, Whites were more likely to be shot by police than Blacks. 

The fundamental assumption underlying the current wave of civil unrest is false.  This might be an important fact.  One would hope that a responsible press would report it.  One should not expect that it will be mentioned by any major news outlet.