Thanks Andy for getting two of my big early thoughts about making sure we compare per capita and looking at police killing against a societal propensity to use violence. It does make some sense to consider police propensity to kill against their society’s use of deadly force. Societal use of deadly force seems like a reasonable proxy for how often the police are justified in using it themselves when dealing with their citizens. I’m going to use overall homicide rates instead of just gun homicides. There’s some value judgement there on whether certain kinds of deadly force being used against you are more appropriate to respond to with deadly force. That argument is more appropriate for another forum IMO. As a proxy both have issues. Especially when we are comparing across cultures where citizen gun ownership is vastly different. I chose one to not discount non-gun forms of threatening law enforcement with deadly force.
Let’s consider Canada. It was in the OP and is the least violent of the US’s two bordering nations… In 2014 they had 15 police homicides in a population of 35.3 million - .043 police homicides per 100k. The US had a population of 318.8 million and …various estimates of police homicides. At the 1000 in the OP it works out to .314 police homicides per 100k. Using the economist 458 and the upthread 1,400 as more extreme estimates we get .144- .439 police homicides per 100k. At the OPs’ 1000 estimate the rate is .314 / 100k. Comparing the rates you are between 3.3 - 10.2 times as likely to be killed by police in the US as you are by Canadian police. The OP 1000 estimate gives a 7.3 times greater likelihood.
We still haven’t looked at our proxy for justification though. Let’s useUN Office on Drugs and Crime (UNODC) data on homicides published in 2013. Using 2007 - 2012 (the six years listed) Canada has an homicide rate of 1.7 per 100k against the US 5.0 per 100k in the six year average. That’s a justification proxy multiplier of 2.94 to the Canadian rates for a direct comparison that includes both population and the society propensity to use deadly force. I’ll call it the “trigger happiness score.”
Trigger happiness scores for comparison:
Canada- .126
US range of .144 - .439
UK: .008 (Math omitted; based on 1 police homicide in 2014 and a 5 year 2007-2011 rolling homicide rate since UNODC doesn’t include 2012 data. )
The US officer is 14- 248% more likely to kill than the Canadian officer controlling for similar situations through the homicide rate. It’s either a big difference or a pretty small one. A UK officer is only likely to kill 6% of the time that a Canadian officer does with the adjustment for circumstances. Colloquially Duddley Do-right is much, much closer to Barney Fife than Bobby.
A quick review ofGNI per capita(at PPP) lists for the top 50 countries and comparing their most recent homicide rates to Canada’s most recent gives us some potential candidates to consider. That’s if someone else wants to rundown their police homicide data. Most recent societal homicide annual rate in parentheses and bolded items have a higher homicide rate that the US:
The Bahamas (26.1 !!!)
Norway (2.2)
Bermuda (10.8)(British overseas territory who’s GNI/homicide numbers aren’t included in the UK but it’s completely independent either)
Belgium ( 1.6)
Estonia (5.0)
Lithuania (6.7)
Finland (1.6)
Greece (1.7)
Malta (2.8)
Greenland (10.4 and a different but similar partially autonomous status from Denmark as the UK/Bermuda accounting split.)
Obviously a good screen to prioritize looking in to them would be whether the police are routinely equipped for deadly force.