Article: Are "Too Many" People Dying from Police Use of Force?

[Gadfly]

I saw this report a few months ago, so bear with me if it has an inaccuracy.

My agency currently sits around 6,300.... For 2013 (last year a full report is completed on) we had 7 reported shooting incidents. At least 1 (possibly 2) of those were suicides by Agents off duty. We are Feds, not street cops, so we don't have to deal with drunks and domestics. We deal with kilos, alien smuggling and human trafficking. But we don't have the same threat level that a beat cop has at 2am, while by himself. Street cops are far more likely to have to use force.

The point us, The numbers of shootings are incredibly low for us. But we chase/arm bar/punch/and tackle quite a bit when needed.

We have a report put out annually like the NYPD. Every A/D, every intentional shooting, every shooting of a dog or injured animal must be reported and documented. We also have to document use of OC/Baton/Taser. But those are rarely used.

I will try and find the report on line and see if it is official use only or if I can link it...

[RevolverRob]

With all due respect to everyone's personal opinion. There are quantifiable ways of statistically validating "rare events" (see: http://en.wikipedia.org/wiki/Rare_events for review).

If you had accurate (or heck even inaccurately) reported numbers for all incidents of interest it wouldn't be difficult to build an algorithm for predicting results. I, personally, would utilize a Bayesian algorithm with incidents of interest used to circumscribe the distributions of my priors. By accounting for the actual data, we would get a small number of statistically probable outcomes. From this, it's relatively easy to determine if an event is "rare". If it is "rare" it will 1) Fall outside the distribution of predictable outcomes. 2) Occur so infrequently that the frequency at which it occurs is no different than the random probability of the same event occuring (in essence these "rare events" are statistically no different than the expectation you would have of hitting a bullseye with a randomly thrown dart at a dart board). All we have to do is then compare the actual numbers to the predicted values. If the actual numbers are higher than predicted values, then the events in question are not "rare". If they are lower than predictions or essentially random, then they would be "rare".

There is really no need to guesstimate what a rare event is, if the given data are available. In genuinely sounds like some of those data might actually be available (per Voodoo_Man's posts). Hmm...I might have to start writing FOIA requests to collect departmental data.

-Rob

[Luger]

Another idea: How many are too many?

Germany has been gathering national data on the numbers of police shootings since mid of the seventies. In the past years we had about 30 to 40 shootings anually. In a country with a population about 80 million people. I guess this is pretty low, compared to the US. Yet there is still an outcry about police brutality over here...

On the other side: The murder rate in germany is about 0.8 per 100.000 people, while it is about 5 per 100.000 in the US. So you could expect about 6 times more police shootings per million people over there.

(While german police is extremely reluctant in shooting people and I know about several fellow officers who only survived because the were extremly lucky...)

[Shellback]

With all due respect to everyone's personal opinion. There are quantifiable ways of statistically validating "rare events" (see: http://en.wikipedia.org/wiki/Rare_events for review).

If you had accurate (or heck even inaccurately) reported numbers for all incidents of interest it wouldn't be difficult to build an algorithm for predicting results...

The only thing I've ever seen in the public domain are websites that compile incidents from news articles on the web.

[LSP552]

There is a national statistical database on fatal vehicle crashes and agencies are required to participate.

http://www.nhtsa.gov/FARS

It's time consuming and every state has a FARS analyst to enter data from crash reports. The National Highway Traffic Safety Administration (NHTSA) has a cooperative agreement with each state to participate in FARs. Law Enforcement agencies send their fatal crash reports to the POC, usually a DOT or State Highway Safety Office.

There are hundreds of data points from every crash but the important thing is they are all entered into a standardized format. Because all of the information is standardized and normalized, fatal crashes are very well understood in the global context.

It would be relatively simple to do the same for law enforcement use of force fatalities because, unlike fatal crashes, they isn't 30,000 a year.

[LSP552]

Besides, I know Claude and John would grab that data and write epic reports on what gunfighters need to be training on to correct deficiencies.

I'd start with reintroducing the idea of being a gunfighter. That has been lost somewhere along the way in many places. The problem is true change must come from the top and we all know how rare it is for a gunfighter to become a chief administrator in a major agency.

[KeeFus]

I'd start with reintroducing the idea of being a gunfighter. That has been lost somewhere along the way in many places. The problem is true change must come from the top and we all know how rare it is for a gunfighter to become a chief administrator in a major agency.

I said that in jest last week while on the range...the look from some people was like I had just called someone of importance a dirty name. The politics of being labeled a gunfighter would probably end a career in some agencies.

[Tiffany Johnson]

John’s a great researcher and a great voice for correcting the misconceptions about what LEOs face out there. This sort of information has value, even despite the current imperfections in data availability. However, as a non-LEO civilian whose heart breaks every time a suspect needlessly dies AND every time people wrongfully accuse good cops of being “trigger happy,” I have one tiny suggestion for my LEO friends. If there’s any interest in closing the trust gap that has sadly emerged between the public and police, please be careful about appearing too hastily defensive when that one-in-a-million aberration does regrettably occur. By all means, sing the praises of those thousands and thousands of officers who do their jobs admirably and protect lives even at their own expense. Educate the public and the media about the statistical rarity of “bad shoots.” But be careful not to look as if you’re excusing the missteps, however rare they might be. Otherwise, it’s like saying, “well, my kid has gotten an A in every class he’s ever taken all his life, so this one F in math is meaningless.” I bet the math teacher disagrees.

I try to thank LEOs and military for their service as often as I can, so I'll do that here too. Thank you, all of you, for risking your lives to keep us safe.

[cclaxton]

John,
You did a fantastic job in pulling together this data and researching this topic. Great work!

Is there any way to break out this data based on geographic location or by urban areas? While these national statistics are valuable, most crimes are local, and my guess is that we will see a lot of variance in use of force rates depending on jurisdiction. It would really be interesting to see that matched up against conviction rates and overall crime rates as well.

I know that is a complicated and monumental task, but just thought I would ask....

Thanks,
Cody

[GardoneVT]

If there’s any interest in closing the trust gap that has sadly emerged between the public and police, please be careful about appearing too hastily defensive when that one-in-a-million aberration does regrettably occur....(snip).

I understand your position .That being stated....the core problem is that the news media is a business. If viewers watch Channel A because they've aired an unsubstantiated narrarative that the police have screwed up ,versus Channel B who accurately reports that investgations on use of force incidents take time, Channel B loses. The days of Walter Cronkite are long gone.

So, everyone with a camera dogpiles on the idea that anytime the police OR a citizen shoots a person of color, its a racist event. The police become reactionary-because the media hurls accusations of malfesance before the spent brass cools , and LE /supervision is legally and morally obligated to say "we don't know yet what happened " it leaves an information vaccum. By the time the facts emerge one way or the other, the riot gear is being unlocked and #blacklivessometimesmatter are on the streets .

Bad shoot? Loot and burn the city.
Justified shooting? Loot and burn the city.

Police say it was a justified shooting? "COVERUP!!!"
Police say it was unjustified and condemn the former officer ? "ALL YALL COPS ARE JUST AS BAD AS HE/SHE IS!!!"

No win scenario either way. So I understand why most LE agencies just clam up and unlock the CS launchers.A message is useless if the recipient is uninterested.