Crime, Race, and Violence in America: Facts vs. Fiction

The headlines blare, “blacks more likely to be killed by police.” The other side of the argument lashes back with “more whites killed than blacks by police.” Frustratingly, each side claims to be supported by data, and it becomes increasingly difficult to separate reality from propaganda. We live in a time where there are so many competing agendas that it can be hard to understand the truth of a given situation. I don’t claim to have all the answers, but I hope I can illuminate what hard numbers  actually say about the situation.

In order to do this, I looked at 10 years of data gathered by the U.S. Department of Justice and U.S. Census Bureau. These tables look at the population of the United States split by racial demographic, as well as available crime statistics for that time frame, split again by race and subdivided into violent and non-violent crimes. It should be noted that some of this data doesn’t tally to 100% because for the sake of this article, we are simply looking at blacks and whites.

Let’s look at population stats first:

Population Data

The raw population data shows a general decline in the white population and a general increase in the percentage of blacks and other people of color (POC). Whites saw an overall decrease of 2.6%, while blacks saw an overall increase of .41%, and other POC saw an increase of 2.19%.

When examining crime stats, population data acts as a barometer for how much crime you would expect, all things being equal, to be committed by a given section of the population. That is to say, assuming no other outside factors, you would expect that 77.36% of the population would commit 77.36% of the crime.

That leads us to the second data set: crime by race.

Immediately, we see some interesting trends. Whites seem to commit 7-10% fewer crimes than would be expected by their total representation in the overall population, while blacks seem to commit between 10-15% more.

This interpretation of the data naturally raises some questions:

Are you saying blacks are prone to commit more crime than whites?

No. This data doesn’t have anything useful to address the disparity; all it does is state that more blacks are prosecuted than whites. An argument could be made that it’s more likely for a black person to be arrested than a white person, but in the spirit of investigation, we’re just following data.

Wait, does this say we’re trending toward less crime than in the past 10 years?

Yes. Social media is great at making us feel afraid of non-existent problems. In 2004, you had about a 1 in 29 chance of being the victim of a crime; in 2014, you had about a 1 in 37 chance. You’re actually safer today than ever, race notwithstanding.

What about the claim of black men being three times more likely to be shot by police than white men?

The Washington Post conducted a study on people shot by police in 2015, which was quite good in terms of data. What they did to get the oft-quoted statistic, though, was take the total population of blacks in 2015 and divide that by the number of blacks killed by police (258), then do the same for whites (494). Then, they compared. If you do that, you find that a black person has a 1 in 165,692 chance of being shot by a police officer versus a white person who has a 1 in 501,647 chance. If you divide 165,692 into 501,657, you get 3.03, so three times more likely.

The problem with that comparison is that it assumes a perfect correlation of total population percentage by race to crimes committed--it’s hard to get shot by police if you don’t interact with them--but as the above tables show, that assumption is empirically false. Whites, despite making up 78.61% of the population between 2004 and 2014, only committed 69.52% of the crimes, while blacks, making up 13.01% of the population, committed 28% of the crimes.

What we really need to do is look at the total number of crimes committed as our “population” to calculate against and re-run the numbers. If we do that for a singular year, we get the following.

First, take the number of blacks that committed crimes (2,427,683) and divide that by the number killed by police (258). Then do the same for whites (6,056,687 divided by 494). This shows that  blacks have a 1 in 9,409 chance of being killed during a police altercation while whites have a 1 in 12,261 chance. In short, blacks are about 30% more likely to be killed by police, not 300%.

If we want to add perspective to these numbers, though, consider that a typical person in the United States has a 1 in 9,737 chance of dying in an airplane accident--the odds of this happening are actually incredibly low.

So, black people are still more likely to be shot by police, just not 300% more likely?

Sort of. We could certainly extrapolate that, but the question is why.

The data tells us that blacks are prosecuted for a disproportionate number of crimes versus their population percentage. Whether this is because they are actually committing more crimes, because they tend to fare worse once in the system, because they tend to be poorer than whites, or some other variable is hard to say.

Should we accept that blacks are committing more crimes, then we still don’t know why they are doing so. Certainly no one could, in good conscience, believe that there is something intrinsic about being black that leads to being a criminal. Instead, it is likely that there are layers of missing information that we need to overlay on this to see what’s driving these numbers.

My guess would be that the driver is economic. It’s worth noting that 27.13% of blacks were reported as living below the poverty line in our sample year, compared to 12.7% of whites. Unfortunately, we don’t have income numbers on those committing crimes in our data set, so it’s hard to concretely paint that picture.

What about violent crimes? Don’t most shootings happen during violent altercations?

Absolutely, according to the study done by the Washington Post nearly 75% of all fatal police shootings occurred during an assault in progress, but what we don’t want to do is engage in statistical reductionism because we lack the information to finish that line of reasoning. If we look at violent crimes, it appears that whites are more likely to be shot (+37%), but that doesn’t give any statistical significance to apply to the overall question. It’s the same error The Washington Post made but in reverse.

Okay, so what can we conclude concretely?

  • Blacks are prosecuted for crimes at a disproportionately higher rate than whites versus their population percentage, but we don’t know why.

  • Blacks are 30% more likely to be shot in an altercation with police.

  • Whites are 37% more likely to be shot while committing a violent crime.

  • It is about as likely that you will die in an airplane crash as it is that you will be killed by police as a black man, and even less likely than that if you're white.

Finally, I leave you with this: Do your own research and ask questions of both sides. Remember, the beaten path is for beaten people--forge your own opinions.



*Crime data collected from the U.S. Department of Justice

*Population according to U.S. Census Bureau

*Violent crimes are offenses of murder and non-negligent manslaughter, rape, robbery, and aggravated assault.  

*Odds of dying chart

*Washington Post Study on Police Shootings