10

Twitter user Leonydus Johnson posted the following statistics acquired from the FBI UCR Program website and the Washington Post:

For every 10,000 black people arrested for violent crime, 3 are killed

For every 10,000 white people arrested for violent crime, 4 are killed"

He claims that they show that there is not pervasive racial bias in the way African-Americans are treated by the police in the United States. Are these statistics misleading or potentially inaccurate in any way?

  • 3
    Last I checked, around 50% of victims of unjustified police homicides were black, relative to 15 percent of the US population. And, depending on where these data are taken from, the percentage may in reality be larger (or even smaller), depending on how justification was determined. – Obie 2.0 Jun 2 at 6:17
  • 8
    This question seems like a better fit for Skeptics, since it's evaluating the truth of a published statement. I'd recommend migrating it – divibisan Jun 2 at 14:05
  • @Obie2.0 The posted rates for shootings are cherry-picked. The actual study that generated those numbers shows no bias in officer shootings, true, but it finds pervasive bias at lower levels of force. – Just Me Jun 16 at 14:49
  • The author requested to migrate this question to Skeptics Stack Exchange. But unfortunately I can not do this because the question was posted too long ago. – Philipp Aug 11 at 8:15
13

First, the claim is not exactly well defined. The FBI data gives arrest rates per arrest type, while the Washington Post gives deaths per population, so it's not clear if the numbers presented are meaningful or how the author arrived at that conclusion. Additionally, it's unclear how to normalize events that can be repeated vs not; that is, a single person can be arrested multiple times in a year, while they can presumably only be killed once. And additionally, he didn't discuss the impact of race on the likeliness of being arrested in the first place (more on this below).

Second, the lead of the Washington Post's "Fatal Force" project (source of the twitter data) responded to the post on twitter saying,

Hey. You’re making this argument based on my data. Your conclusions are factually inaccurate

Now, for your question:

These seem to indicate that there is not pervasive racial bias in the US police force. Are these statistics misleading or potentially inaccurate in any way?

While the twitter post linked to the 2018 dataset, the current 2020 version, has this summary at the end

Although half of the people shot and killed by police are white, black Americans are shot at a disproportionate rate. They account for less than 13 percent of the U.S. population, but are killed by police at more than twice the rate of white Americans. Hispanic Americans are also killed by police at a disproportionate rate.

So, in absolute terms, there are more white people killed by police. This is perhaps not surprising since the dataset lists that population as almost four times larger than any other ethnicity. But as pointed out above, "The rate at which black Americans are killed by police is more than twice as high as the rate for white Americans. [em. original]"

For further reading, I would recommend this Washington Post article outlining a number of studies evaluating systemic racism related to the criminal justice system.

Of particular concern to some on the right is the term “systemic racism,” often wrongly interpreted as an accusation that everyone in the system is racist. In fact, systemic racism means almost the opposite. It means that we have systems and institutions that produce racially disparate outcomes, regardless of the intentions of the people who work within them.

There are some dissenting studies listed above, but the majority of studies point out that non-whites are disproportionately represented at each step of the criminal justice system (being pulled over; arrest; jail; bond disparity; prison sentence harshness; probation being revoked; parole).


Are these statistics misleading? Statistics are notoriously difficult to get right, and easily manipulated (lies, damn lies, and statistics). The author's intent seemed to be to challenge whether police killings are racially biased. And there was a study on the list WaPo page that said class (income) was the most important factor when determining police engagement frequency. But with all the other data, and including comments from the original author lead, I'm skeptical that the post on twitter holds merit.

| improve this answer | |
  • (+1) Regarding your last point, even for some data in this area that hasn't been contested in itself, there are some issue with drawing conclusions from it skeptics.stackexchange.com/questions/43730/… – SX welcomes ageist gossip Jun 3 at 4:15
  • 1
    Thanks for your answer. I am working through that WaPo article right now, but it's long. I agree that the original author's comments are fairly convincing that Mr. Johnson has misrepresented something, although I'm still not sure precisely how. I was previously aware, as was Mr. Johnson, of the higher shooting rate of black Americans by police. The question remains of whether this can be explained by higher crime rates by blacks, or perhaps police mistreatment of the poor, etc, as opposed to racial bias in the police force. The rest of the justice system is beyond the scope of my question. – jeremy909 Jun 3 at 5:40
  • 2
    There's an incommensurability of benchmarks that no one addresses but needs to be addressed before we start talking about bias in police killings at least. Do we use population or some modified measure of interaction rates with (noting that there's racial bias when it comes to interaction rates but there's also higher crime rates in black communities)? I'd also note that the Wesley IIRC, the WaPo person you quoted, said he didn't like the conclusion drawn because it didn't use the population benchmark, therefore it amounts to begging the question. Not a fair outlook on the situation imo. – Daniel Jun 3 at 21:45
  • 3
    Yeah, if you go by percentage of population, Black people are over-represented among police shooting victims. But if you go by percentage of arrests, then the proportion is about what's expected. – dan04 Jun 4 at 4:28
  • 1
    @jeremy909 If you want to substantially change your question (to be about something other than the twitter comment), you should open a new question on this site. If you are trying to decide if People of Color commit more crimes than Whites, go through the sources on the WaPo article. Here's a chart to get you started: hamiltonproject.org/charts/… – BurnsBA Jun 15 at 12:14
0

A question like this is actually better suited for the skeptics then politics, luckily skeptics have already covered the question here: Are African Americans victims of a disproportionate number of police killings?

The short version is that it's complicated, there are definitely a higher proportion of police shootings of black then white individuals once one adjusts for population size, but some arguments can be made that other demographics (which I'd argue all trace back to socioeconomic differences caused by racism of past and present) may explain the differences in deaths; thus making it hard to say definitively that the differences are caused specifically by racism.

Put even simpler, probably, but it's pretty hard to definitively prove racism as a cause in such a large, and difficult to create controlled tests for, a population; so there will always be other possible explanations.

| improve this answer | |

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .