Police deaths over time

johnwilliamson062 · April 20, 2014, 02:41 PM

Reading another thread got me thinking and have been rolling something around in m head.
Logic can't be applied to politics, but i still have some interest in actual data or the opinions from those who may be more familiar with it.

Looking at police officer deaths over time I expected them to peak as afunction of population size somewhere between 1980 and 1990. It seems both in gross number and relative to population the peak was somewhere around 1975.
I base this off of two unreliable sources off google:
http://media.npr.org/assets/img/2010...76a-s6-c30.jpg
http://radgeek.com/gt/2010/01/23/siege-mentality/
The NPR does not break out by cause and the "radgeek" one is from source I have never heard of that appears to be far from objective.

My thoughts leading up to this search on google:
Organized crime organizations of a given level likely have access to similar firepower, so when they fight what weapons they use is somewhat irrelevant.
When attacking "civilains" it is almost certain they will do so with superior firepower, and this is only exacerbated by regulation.
The only situation in which restricting arms might have an effect would be in conflict with un-restricted police departments.
As such, I was looking to see if the 1986 full auto restriction, 1994 semi-auto restrictions, expiration of the 1994 restrictions, or any other gun restrictions seem to have an effect on officer deaths.

From this data, and similar no more reliable sources provided by google, it seems the 1968 regulations may have had an effect. None of the others seem to correlate with the data. In fact, it seems to me as if the additional regulations affect firearms generally not in use in 1975.

Librarian · April 20, 2014, 03:16 PM

Why is it you did not use FBI data?

http://www.fbi.gov/stats-services/crimestats

Quote:

Law Enforcement Officers Killed and Assaulted (LEOKA) gathers data on felonious and accidental line-of-duty deaths and assaults of law enforcement officers. Tabular presentations include weapons used, use of body armor, and circumstances surrounding murders and assaults of officers.

2012 | 2011 | 2010 | 2009 | 2008 | 2007 | 2006 | 2005 | more

.pdf format from 1996-2003

Also http://www.fbi.gov/about-us/cjis/cji...aulted-program

Quote:

The FBI has collected data regarding assaults on law enforcement officers since 1960. Beginning in 1972, detailed facts about line-of-duty deaths have been published. Recent expansion of the data collection methods are providing even more facts that can be studied by experts and officer safety trainers in order to tailor training to real world circumstances. For instance, recent LEOKA data has shown an increase in ambush situations in assaults and officer deaths. As a result, the LEOKA Program’s Officer Safety/Awareness Trainers are studying the data with the purpose of shaping future training that addresses this problem.

johnwilliamson062 · April 20, 2014, 04:42 PM

Although it indicates they have taken the data for a long time, only 1992 and later seems to be published there and the farthest back i saw data was 1987 in the 1992 report. Not too useful for my purposes without pre-86 data.

**44 AMP** · April 20, 2014, 04:52 PM

Quote:

it seems the 1968 regulations may have had an effect. None of the others seem to correlate with the data. In fact, it seems to me as if the additional regulations affect firearms generally not in use in 1975.

Correlation is not causation.

This is one of the biggest errors gun control advocates constantly make.

In an extreme example, 99%+ of convicted felons in the USA ate bread or a bread product within 30 days of committing their crime.

This is a correlation, but hardly causation.

you say the 68 laws appear to have had an effect? What else about US society changed between 68 and 75? Anything? or was the only change gun control?

it is other factors, including the general level of unrest in society that affect these things, not just gun laws. It is still questionable to me, whether gun control laws actually have any positive effect at all, let alone being a dominant factor. Yes, one can find individual cases where the law appears to have some effect, but in honesty, can one say "if there had been no law, it wouldn't have happened?" OR can one only say that with (or without) a given law, that it "would not have happened that way"?

**Tom Servo** · April 20, 2014, 05:27 PM

Quote:

This is one of the biggest errors gun control advocates constantly make.

Gun rights advocates also make the same mistakes.

As you've pointed out, many things can change in a decade. Today's police are differently equipped and differently trained. Our penal system has also seen significant changes since the 1960's.

Add innumerable societal factors to that, and it's impossible to pin one cause down.

johnwilliamson062 · April 20, 2014, 07:24 PM

Quote:

Correlation is not causation.

This is one of the biggest errors gun control advocates constantly make.

Well, yes, but I have yet to see a case of causation without correlation? Eliminate correlation and you eliminate the possibility of causation.
Yes the water is muddy on this and leaving it open to argue the effects were just overridden or canceled other effects, but the American voter has a very short attention span. Presenting one graph is stretching it.

JimPage · April 20, 2014, 09:12 PM

Any statics text will tell you that correlation does not mean causation. Causation may be related to correlation, of course, but only incidentally. Correlation does NOT tell you anything about causation.

Yes, areas with high numbers of gun control laws are usually with high levels of gun crimes. How do you tell whether the laws are result of the high levels of crime or whether the high levels of crime are the result of the high numbers of gun control laws. And that doesn't even begin to address the relation of gun control to the effect of propaganda and politics.

So you must accept the science that tells us that correlation does NOT mean causation. Nor does it mean there is no causation. It is just a simultaneous event.

In other words, correlation is meaningless. It may be, It may not be.

**JohnKSa** · April 20, 2014, 09:37 PM

Quote:

Looking at police officer deaths over time...

It's problematic to use fatalities as a metric if you're trying to track something long term. Too many other things have changed over the years that will heavily impact whether a person survives or dies even given an identical injury.

Improved medical care, improved safety equipment, the proliferation of air transport for medical emergencies and the much-improved rapidity of notification due to higher populations and much better communication capabilities means that people are much more likely to survive an injury than they would have been just a couple of decades in the past.

You'd probably get a better overall picture of the situation if you focused on shooting injuries to police officers instead of looking at fatalities.

**44 AMP** · April 21, 2014, 10:49 AM

Another thing to look at is the data about officer deaths. Does it specifically list the CAUSE of the death?

What I mean is that officers killed on duty vs off duty, in traffic accidents, shot by bad guy, shot with their own gun by bad guy, and so on.

If they are all just dumped in the same category, then using that number as any kind of metric for gun control laws is an error. Office deaths, from all causes is a valid statistic, for some uses, but by no means all uses.

And as JohnKSa mentioned, using only deaths further skews the issue, if you are looking at gun control alone.

The classic example (for me) of disingenuous statistics is from a couple decades ago, when one of the biggest anti gun groups (can no longer remember if they were Handgun Control Inc., or calling themselves some other name at the time..) began using a horrifying number of "death of a child due to a handgun". All the other groups instantly began using the same figures, and the media parroted it endlessly...until...

A "defector" from the group publically revealed what went into the number. It seems that the number of "death of a child due to a handgun" included everyone under the age of 25 who died as a result of being shot. Suicides, actual individual murders, accidents, gang shootings, and even "children" (under 25) shot by police if they died as a result.

And they didn't make any real attempt to determine if it was actually a handgun shot that caused it, either. Any gunshot resulting in a death (under 25 yrs old) was good enough to go into their data.

The OP stated his sources were "suspect". Real, verified accurate data often leads some people to "suspect" conclusions. "Suspect" data cannot lead to anything else.

Brian Pfleuger · April 22, 2014, 11:30 AM

Quote:

Originally Posted by JimPage

In other words, correlation is meaningless. It may be, It may not be.

You have taken a generalized truth and emphasized it into a falsehood.

Correlation does not MEAN Causation. this is true.

Correlation is NOT meaningless or useless.

In fact, correlation implies a possible causation that must be further studies to prove or refute.

Fact is, correlations very often DO lead to causation, it just does not IN AND OF ITSELF, singularly prove causation.

**44 AMP** · April 22, 2014, 01:01 PM

Correlation is not meaningless. Correlation is simply some kind of identifiable link between two subject.

The link may be clear and obvious, or it may be obscure and tenuous, but it has to be there, otherwise there is simply nothing to compare.

Price of apples in Hong Kong today and voting rights for women in the USA in 1919. Find a correlation. I can't. Not linked, not related, nothing to discuss.

What causes a given situation has to be something correlated with it. But everything that has some correlation is not the cause.

In its very broadest sense everything on Earth is correlated. We are all here on the same planet, after all. But when you are looking at specific things, you need specific correlations, and from them, you have to determine which is significant, and which significant one(s) might be the cause.

csmsss · April 22, 2014, 03:45 PM

Don't forget that police equipment (inlcuding firearms), training and procedures have changed dramatically over time.

**tyme** · April 24, 2014, 03:01 PM

Quote:

From this data, and similar no more reliable sources provided by google, it seems the 1968 regulations may have had an effect. None of the others seem to correlate with the data. In fact, it seems to me as if the additional regulations affect firearms generally not in use in 1975.

If that's true why did the police death rate increase fairly dramatically from 1968 to 1974?

Lead was banned in gasoline in 1970 and there's some evidence that lead has physiological effects that increase violence.

Just because sociological factors are complex and difficult to study doesn't mean you can ignore them.

Perhaps the end of the Vietnam war caused people to be less violent.

johnwilliamson062 · April 26, 2014, 03:09 PM

Quote:

If that's true why did the police death rate increase fairly dramatically from 1968 to 1974?

A sales restriction is not going to have an immediate effect on a durable good. If hammers were outlawed today nails would continue to be sold. It would take years for screws to completely replace nails. Firearms would follow a similar, but likely more extreme pattern IMO. I'm not saying I agree with the restrictions. Having followed the CCW fight in Ohio, I have learned the benefits of an incremental approach and adjusting arguments at each turn.

Logical political arguments aren't won with 50 page scientific studies. They are won with simple charts, graphs, and paragraphs. Even that is a stretch as most US political conflicts are resolved by a "For the CHILDREN..." It is better if you can back up a chart or graph with a well written paper documenting a well designed and executed study. A simple chart that contradicts your opponents emotional claim is just as useful.

The medical advances are an obvious game changer on this I'd not considered. Increased use of vests, although I remember half those who died were not wearing them according to one of those reports.

**Frank Ettin** · April 26, 2014, 03:39 PM

Quote:

Originally Posted by johnwilliamson062

...A simple chart that contradicts your opponents emotional claim is just as useful...

No, in the long run it absolutely is not.

Making claims, no matter how pretty the graphs are, that can't be solidly supported and are vulnerable to attack wind up destroying your credibility. Once lost, credibility is extremely difficult to reacquire.

johnwilliamson062 · April 26, 2014, 08:07 PM

Don't make a claim. Ask a question.

lefty60 · April 27, 2014, 12:10 AM

AND, you are overlooking the gains that had been made over that time period.
Many times, for example shootings only, those that would have died the year before would survive at this time.
How can you factor in the advances of medicine, the changes/decay of society, protective gear, more effective guns/ammo, ect.

This all just gives me a headache. I do have medical records that show that I have brain damage

Don't mean to be a "wet blanket", but given the massive amount of variables I fail to see how a meaningful "study" can be done.

Of course that could qualify it for a Federal Grant, so that it could be studied for years to come.

Stay Safe

April 20, 2014, 02:41 PM	#1
johnwilliamson062 Junior member Join Date: May 16, 2008 Posts: 9,995	Police deaths over time Reading another thread got me thinking and have been rolling something around in m head. Logic can't be applied to politics, but i still have some interest in actual data or the opinions from those who may be more familiar with it. Looking at police officer deaths over time I expected them to peak as afunction of population size somewhere between 1980 and 1990. It seems both in gross number and relative to population the peak was somewhere around 1975. I base this off of two unreliable sources off google: http://media.npr.org/assets/img/2010...76a-s6-c30.jpg http://radgeek.com/gt/2010/01/23/siege-mentality/ The NPR does not break out by cause and the "radgeek" one is from source I have never heard of that appears to be far from objective. My thoughts leading up to this search on google: Organized crime organizations of a given level likely have access to similar firepower, so when they fight what weapons they use is somewhat irrelevant. When attacking "civilains" it is almost certain they will do so with superior firepower, and this is only exacerbated by regulation. The only situation in which restricting arms might have an effect would be in conflict with un-restricted police departments. As such, I was looking to see if the 1986 full auto restriction, 1994 semi-auto restrictions, expiration of the 1994 restrictions, or any other gun restrictions seem to have an effect on officer deaths. From this data, and similar no more reliable sources provided by google, it seems the 1968 regulations may have had an effect. None of the others seem to correlate with the data. In fact, it seems to me as if the additional regulations affect firearms generally not in use in 1975.

April 20, 2014, 09:12 PM	#7
JimPage Senior Member Join Date: February 21, 2010 Location: Rome, NY Posts: 941	Any statics text will tell you that correlation does not mean causation. Causation may be related to correlation, of course, but only incidentally. Correlation does NOT tell you anything about causation. Yes, areas with high numbers of gun control laws are usually with high levels of gun crimes. How do you tell whether the laws are result of the high levels of crime or whether the high levels of crime are the result of the high numbers of gun control laws. And that doesn't even begin to address the relation of gun control to the effect of propaganda and politics. So you must accept the science that tells us that correlation does NOT mean causation. Nor does it mean there is no causation. It is just a simultaneous event. In other words, correlation is meaningless. It may be, It may not be. __________________ Jim Page Cogito, ergo armatum sum

April 21, 2014, 10:49 AM	#9
44 AMP Staff Join Date: March 11, 2006 Location: Upper US Posts: 28,820	Another thing to look at is the data about officer deaths. Does it specifically list the CAUSE of the death? What I mean is that officers killed on duty vs off duty, in traffic accidents, shot by bad guy, shot with their own gun by bad guy, and so on. If they are all just dumped in the same category, then using that number as any kind of metric for gun control laws is an error. Office deaths, from all causes is a valid statistic, for some uses, but by no means all uses. And as JohnKSa mentioned, using only deaths further skews the issue, if you are looking at gun control alone. The classic example (for me) of disingenuous statistics is from a couple decades ago, when one of the biggest anti gun groups (can no longer remember if they were Handgun Control Inc., or calling themselves some other name at the time..) began using a horrifying number of "death of a child due to a handgun". All the other groups instantly began using the same figures, and the media parroted it endlessly...until... A "defector" from the group publically revealed what went into the number. It seems that the number of "death of a child due to a handgun" included everyone under the age of 25 who died as a result of being shot. Suicides, actual individual murders, accidents, gang shootings, and even "children" (under 25) shot by police if they died as a result. And they didn't make any real attempt to determine if it was actually a handgun shot that caused it, either. Any gunshot resulting in a death (under 25 yrs old) was good enough to go into their data. The OP stated his sources were "suspect". Real, verified accurate data often leads some people to "suspect" conclusions. "Suspect" data cannot lead to anything else. __________________ All else being equal (and it almost never is) bigger bullets tend to work better.

April 22, 2014, 01:01 PM	#11
44 AMP Staff Join Date: March 11, 2006 Location: Upper US Posts: 28,820	Correlation is not meaningless. Correlation is simply some kind of identifiable link between two subject. The link may be clear and obvious, or it may be obscure and tenuous, but it has to be there, otherwise there is simply nothing to compare. Price of apples in Hong Kong today and voting rights for women in the USA in 1919. Find a correlation. I can't. Not linked, not related, nothing to discuss. What causes a given situation has to be something correlated with it. But everything that has some correlation is not the cause. In its very broadest sense everything on Earth is correlated. We are all here on the same planet, after all. But when you are looking at specific things, you need specific correlations, and from them, you have to determine which is significant, and which significant one(s) might be the cause. __________________ All else being equal (and it almost never is) bigger bullets tend to work better.

April 20, 2014, 04:42 PM	#3
johnwilliamson062 Junior member Join Date: May 16, 2008 Posts: 9,995	Although it indicates they have taken the data for a long time, only 1992 and later seems to be published there and the farthest back i saw data was 1987 in the 1992 report. Not too useful for my purposes without pre-86 data.

April 22, 2014, 03:45 PM	#12
csmsss Senior Member Join Date: October 24, 2008 Location: Orange, TX Posts: 3,078	Don't forget that police equipment (inlcuding firearms), training and procedures have changed dramatically over time.

April 26, 2014, 08:07 PM	#16
johnwilliamson062 Junior member Join Date: May 16, 2008 Posts: 9,995	Don't make a claim. Ask a question.

April 27, 2014, 12:10 AM	#17
lefty60 Member Join Date: October 13, 2008 Posts: 18	AND, you are overlooking the gains that had been made over that time period. Many times, for example shootings only, those that would have died the year before would survive at this time. How can you factor in the advances of medicine, the changes/decay of society, protective gear, more effective guns/ammo, ect. This all just gives me a headache. I do have medical records that show that I have brain damage Don't mean to be a "wet blanket", but given the massive amount of variables I fail to see how a meaningful "study" can be done. Of course that could qualify it for a Federal Grant, so that it could be studied for years to come. Stay Safe