Ruin in a Nation I
As it happens, police work did crack the ten most dangerous jobs in 2007. But two contributors to that relative have nothing to do with the absolute dangers of the job: 1. The 2002 list that Tom Knapp links breaks truckers and other drivers (salesmen and hacks, among others) into two separate occupations, while the 2007 list linked in the first sentence combines them; 2. Most occupations got statistically safer between 2002 and 2007. The tenth-most dangerous job in 2002 (truck-driving) has a death rate of 25 per 100,000. The tenth most dangerous job in 2007 (law enforcement) has a death rate of 21.4.
If you dig into the causes of death, you start to get a sense of where that danger comes from. The National Law Enforcement Officers Memorial Fund tallies 1,640 deaths in the last decade, including 72 on September 11, 2001 from terrorism. About half the remaining deaths stem from vehicle mishaps (cars, planes, motorcycles, bicycles, trains, horses, boats). Some of these doubtless come from high-speed chases; a lot will stem from the fact that, like truckers and salespeople, police spend a lot of time in and around motor vehicles. 9% of non-9/11 deaths are categorized as “job-related illness.” There are a proportionate handful of deaths whose relation to hostile action are ambiguous: falls, electrocution, drownings, strikes by falling objects – 52 by my count in ten years. Some of these may have happened as a result of chases or altercations.
That leaves hundreds of deaths in the last ten years that are obviously violence-related:
561 shooting deaths
11 lethal beatings
8 deaths from bombs
11 stabbings
1 strangulation
6 from “terrorist attacks” after 2001
——————
598 total
The Fund’s own breakout of police deaths in the line of duty for 1999-2008 is 728 felonious deaths (including 9/11), 912 accidental deaths. Nationwide, that’s 73 felonious deaths a year, about 66 per year excluding the deaths in the World-Trade-Center and Pentagon atrocities. 2007 was an unusually bad year in terms of absolute deaths, which helps explain how law-enforcement cracked the top ten most dangerous jobs. I don’t know how the fund allocates “blue-on-blue” deaths. Presumably, if a cop accidentally shoots another cop, it goes in the accident category, while if he shoots a fellow officer on purpose, it counts in the felonious column.
These statistics cast an interesting light on Brandon del Pozo’s defense of police doctrine posted to Crooked Timber this week. The statistics indicate that police work is less dangerous than many other possible occupations, and that less than half that danger stems from violent resistance by suspects – the kind of “loss of control” that defenders of Officer Crowley’s conduct during the Gates arrest point to as a major danger. (del Pozo gives you the high-toned version of the general case). What statistics alone can’t untangle is the impact of police praxis on the danger level: is police work relatively safe because police are hardcases about maintaining control, or is it relatively more dangerous because the provoke confrontations. And the ultimate question begged previously is where police safety ought to rank in the hierarchy: is it better that police feel as safe as possible or that citizens be respected.
I suspect that del Pozo has a point. But that means that, so long as police are going to have broad discretion to “maintain control of the situation,” we need to keep the number of sitiuations in their purview to an absolute minimum. Instead we do the opposite by contually expanding the scope of the criminal law.
Comment by Thoreau —
July 25, 2009 @ 8:35 pm
QFT
Comment by Praxis —
July 26, 2009 @ 8:03 am
Yes that’s right, it’s impossible to disentangle that. The question then is how come that doesn’t stop you from posting this and drawing conclusions about the danger level? Seems you have an agenda that is less than objective
Comment by Praxis —
July 26, 2009 @ 8:26 am
509 Murders in Chicago in 2008
314 Soldiers killed in Iraq in 2008
Ok Jim, extrapolate your logic to that statistic. In other words tell us how it is safer for an American in Iraq than in Chicago. Then toss in at the end how it just might have something to do training, but imply that doesn’t diminish your point because it’s ‘impossible to say’
Comment by Jim Henley —
July 26, 2009 @ 8:28 am
1. This is a blog. Of course I have an agenda that is less than objective.
2. You seem to think that, since statistics can only be a starting point, we shouldn’t look at them at all?
3. The statistics do make it clear that once you disentangle the accident risk from the felony risk – police are in vehicles, on the move; law enforcement is a “working-class” job with a lot of physical labor of various sorts to it – the kind of risks adduced to justify max-control police tactics could triple before felonious causes alone made police work a notably dangerous job. And it would have to increase by a factor of ten before police work rivaled logging and fishing for danger.
3. Do you have a program for further analyzing the risk factors of different police approaches to interacting with citizens beyond “trust the cops?” Some research to point to or avenues for inquiry?
4. What if we could compare death rates from times with less belligerent approaches or across districts with different (actually complied with) doctrines? How would you compare police death rates across countries?
Comment by Jim Henley —
July 26, 2009 @ 8:30 am
This is kind of a last-chance question for you, praxis. Do you know what a “death rate” is?
Comment by Praxis —
July 26, 2009 @ 9:54 am
Last Chance for what?
You still don’t get it, you can’t quantify danger with death rates. Police training is largely how to handle dangerous situations. The fact that they manage their death rate better than fisherman could simply mean they are more careful and better trained. It could mean that only danger seeking people want to fish, who knows? It also depends on where the cop is working, you didn’t mention that.
The home is called a dangerous place based on accident and death rates, but using that as a comparison tool to other places and activities is pointless if you can’t meaningfully normalize all the variables.
I don’t have a program to analyze all the risk factors here, but neither do you- it’s to some degree subjective. The difference is you still want to make conclusions to support your predetermined agenda.
Comment by A Squirrel —
July 26, 2009 @ 12:47 pm
We should all apparently throw up our hands and welcome the police state because the statistics are not as cut-and-dried as would be optimal.
Go over to Balko’s place for a bit and then tell me that suggesting some policy re-evaluation is mere tendentious hackery.
Comment by John Emerson —
July 26, 2009 @ 4:06 pm
I’ve personally known two minimum-wage convenience-store clerks who were murdered on the job. Anecdotal, but they and taxi drivers don’t get the credit they deserve.
The reason police get the sympathy is because they have guns. It’s just safer to love and admire them. People would laugh at convenience-store clerks who strutted around playing tough.
Ah, Praxis. You missed your last chance.
Anecdotal again, but here in Portland I know of at least three cases in ten years mentally ill people were killed by the police. In one case the victim was surrounded by police and assalted while he was on the ground. In one case the mentally ill person was kept quiet by amateurs for half an hour but was shot within a few minutes of the police’s arrival.
Comment by Praxis —
July 26, 2009 @ 5:05 pm
Comment by A Squirrel —
July 26, 2009 @ 9:33 pm
Uh, no one claimed an airtight statistical case for policy “x”. Not to put words in our host’s mouth, but the thrust of his post – the final paragraph – neither assumes nor requires one exists.
You missed your chance to stand athwart the blob yelling “stop!”.