Black and White Deaths During Arrests: Some Analysis

This blog looks at one aspect of possible racial bias in the justice system, namely, homicides during arrests.

You will be given the data sources. They are official government sources. You can check them yourselves. All analysis will be open to inspection.

This is a rough and ready analysis, using two sources that do not cover the same time period. There are no statistical tests. Still, it will be apparent that racism appearing in the form of substantially more homicides of black arrestees than white arrestees, in relation to their respective numbers of arrests, is not supported by the data.

There does exist a factor that hurts black people more than whites, and that factor is the War on Drugs. More on that in the conclusion.

The Bureau of Justice is the first source of data. It’s a government agency. This is official material.

On their opening page, the first thing they show, for whatever reason, is imprisonment rates since 2008. There are two things to note: levels and trends of these levels.

The imprisonment rate levels are falling for all races.

• “From 2008 to 2018, the imprisonment rate dropped 28% among black residents, 21% among Hispanic residents, and 13% among white residents.

• “In 2018, the imprisonment rate of black residents was the lowest since 1989.”

It’s hard to argue racism on this score.

The levels are higher for black people, much higher. The possible causes are many, and they include racial bias in the justice system. It can take many forms, and a lot of people are very, very upset over the homicide of George Floyd while being arrested. We move away from imprisonment trends to the meat of this analysis, which is arrests and homicides during arrests.

After a bit of search, I found a 16-page document with arrest-related deaths:

This covers 2003-2009, seven combined years.

There were 98 million total arrests. 4,831 of these arrests resulted in death.
2,931 of these 4,831 were homicides. Others were suicide, intoxication, accident, natural causes. 95% of arrests were male.

Over 7 years, there were 2,931 homicide cases out of 98 million arrests.

Of the 4,831 total arrest-related deaths, there were 42.1% white, 31.8% black, and 19.7% Hispanic. I calculate 2,034 white, 1,536 black and 952 Hispanic. There are 84 months in the period. Per month, this is 24 whites a month, 18 blacks and 11 Hispanics each month. I do it by month to make it easier to grasp. Compare this to 1.17 million arrests per month, on average.

If you look at Chicago, you will see that the number of homicides was 510 in 2019. That’s 42 a month. That’s only one big city, and a large part of that is blacks killing blacks.

This BJS report doesn’t provide crimes committed by race. That would influence the number of arrests and therefore number of death. However, it does show that if deaths from arrest are a problem, it’s not a huge problem in comparison to total arrests or other sources of homicides. In 2018 the U.S. had 16,000 homicides. If that held for 7 years, that’d be 112,000 homicides. The arrest homicides are 2,931, which is small in comparison.

I do not see any BJS data on the race of the people committing crimes. I turn elsewhere, to FBI data.

I round the numbers in this table for 2017 to two decimal places. The table shows numbers of arrests broken down by race.

There are 5.63 million white arrests and 2.22 million black arrests. They place Hispanic in the white category. There are 8.16 million total arrests (with other groups added in). Hence, we have whites being 69% of the total arrests and blacks being 27.2% of the total arrests.

Now we go back to the 2003-2009 data. Of the 4,831 homicides from all causes, the white + Hispanic percentage is 42.1 + 19.7 = 61.8%, not far from the 69% in 2017. The black percentage in 2003-2009 is 31.8%, not far from the 27.2% arrest percentage.

The FBI data are total arrests by race. The BJS data are homicides associated with arrest broken down by race.

We have the FBI saying 27.2 % of arrests are blacks in 2017, and we have BJS saying that deaths of all types during arrests conducted in the 2003-2007 period are 31.8% blacks.

We have different time periods here. We have other obvious data issues. Nonetheless, the numbers are quite close. My takeaway is that blacks are not being killed during arrests in numbers that are way out of line with the percentage of arrests being made of them.

The notion that police are going into black areas and shooting people down is false. The derivative notion that therefore a vast re-organization of police should be undertaken is likewise false. There is no support for these ideas in these data.

A speaker expressed these false ideas plainly in questioning Mayor Frey: “We have a yes or no question for you. Yes or no, will you commit to defunding the Minneapolis Police Department? We don’t want any more police, is that clear? We don’t want people with guns in our communities, shooting us down. Do you have an answer? It is a yes or a no.”

Deaths during arrest, such as Floyd’s, are not a racist phenomenon. Deaths during arrest are not a major problem either.

However, there is one source for black arrests, harsher sentencing, and thus longer prison populations that is the elephant in the room: THE WAR on DRUGS.

I’ve written about this before and presented a graph that shows prison population rising steeply after harsh drug laws were enacted. That is here.

The single most direct, effective and simplest way to decrease arrests and imprisonment of blacks in numbers out of proportion to their population is to end the War on Drugs, i.e., end the discrimination against drug users, transporters, sellers and makers.

The big stir about Floyd’s death has a number of sources that we should know about. One thing for sure, however, is that it is not directed at the really important problem that’s actually here with us: The War on Drugs.


7:21 pm on June 7, 2020