Police Use of Deadly Force ISProblem!

“In 2016 (the last year for which full data is available) there were 17,250 homicides in the U.S. The implication is that over 6% of homicides annually are due to police. That’s a substantial proportion – not a small problem!”

[Ed. Note: In the following response to the response as to whether the current rate of officer-involved-shootings is a problem or not, Dr. Gregory Gelembiuk, PhD, from Madison, Wisconsin presents some data for us to consider.]


Best I can ascertain, only one officer in the past 14 years (Sault Ste. Marie Detective John Weir) could have been saved, maybe, I don’t know, by keeping greater distance and being quicker to shoot. The other officers, rest in peace, died doing the job they had to do.

Update: Based on a useful comment to this post, I also should have included blunt weapons and not just knives. Doing so brings the total number of police officers killed in the past 15 years with any relevance to the 21-foot-rule up to three police officers. It’s also worth mentioning that I’m not looking at officers just injured. But we don’t have those figures. And one can assume some relationship between fatal and non-fatal injuries.

 So, let’s put the 21-foot-rule in perspective. In this same time period since 2000, as many officers (3) have been killed by a moving train.

 Six officers have been killed by animals: one by cow, one by spider, and one by bee; the other three from horses (interestingly, none by dog, the only animal often at the receiving end of a 21-foot mindset).

 If one were truly interested in saving police lives rather than simply building police paranoia and mistrust of the public, we should look at the 515 law enforcement officers killed traffic fatalities. How many of these would have been prevented by officers wearing seat belts? And yet the same officer who won’t wear his seatbelt because he claims it gets caught on his equipment (which, speaking from experience, is bullshit) will be quick to spout the absurdity that his life is endangered by anybody within 21 feet, in optimal conditions.


  • One should never try to infer any larger trend from only a few years of data.

Between the years 2005 and 2008 the percent of residents who actually had any contact with the police in enforcement situations and experienced the use or threat of force by police dropped from 2.3% to 1.9 %.

This is from two data points – 2005 and 2008 – only three years apart. There’s always going to be stochastic fluctuation from year to year. In this case, Solar appeared to be trying to make an inference from Table 17 of this report (a report based on the Police-Public Contact Survey). If you look across a larger timespan (where 2011 is the most recent year available), you see something rather different.

  • There’s no evidence of a decreasing trend.

Some people attempt to argue that police use of force is not a problem, basically by seeking to minimize the issue. They offer a percentage (e.g. percent of police contacts that involve use of force), says it’s a smallish percentage, and therefore it isn’t a problem. That’s a poor argument and kind of sleight of hand – e.g. a small percentage of something can still be a huge absolute number (with enormous consequences – as when you’re talking about deaths). Planes only crash a small percentage of the time. However, no-one argues that it’s not a problem when planes do crash, or that we shouldn’t do absolutely everything possible to minimize the chance of a plane crash (as some appear to argue regarding officer involved shootings). Likewise for nuclear power plant failures.

  • In 2017, about 1147 people were killed by police. Different databases will give slightly different numbers, but they’re all in the same ballpark. 1147 is not a small number.
  • In 2016 – the last year for which full data is available, there were 17,250 homicides in the U.S.
  • The implication is that over 6% of homicides annually are due to police. That’s a substantial proportion – not a small problem.

And there’s tremendous variation across countries, with the U.S. far exceeding all others.

Now someone might argue: The U.S. has more guns. Police elsewhere don’t have to deal with all those guns. And this is valid, but only up to a point. Some other countries, such as Finland, do have high gun ownership (including a lot of handguns), and police everywhere have to deal with bladed and blunt weapons (where a substantial proportion of U.S. OIS involve subjects who aren’t armed with guns). Accounting for the guns in the U.S. doesn’t really resolve the discrepancy – police elsewhere are far less likely to kill someone, even accounting for circumstances, and aren’t themselves suffering injuries or fatalities at higher rates.

This basic point becomes even clearer if you look across cities in the U.S. There are enormous disparities across U.S. cities in the per capita rate with which police kill people – for example, look at the left hand panel here, based on data from January 2013 through June 2017. For example, compare NYPD (average annual police homicide rate of 1.63) versus Orlando (average annual police homicide rate of 13.99).

  • Moreover, there’s no correlation between the rate of violent crime across U.S. cities and the rate of police OIS.

Differences in the rate of OIS reflect differences in practices (due to policies and training), and cities with lower rates of OIS do not have higher rates of officer injuries or fatalities.


[Ed. Note: Sometimes facts get in the way of our feelings and vice versa. Whether or not the current rate of police use of deadly force is or is not a problem I will opt to the community served. What do they think? How much and in what situations do they want their police to use deadly force? That’s the open discussion we need to have in America!]