My never-before-seen graphs relating violent crime rate and gun owning % by state!

well.... I'm not sure I really see too much of a trend..... polynomial if anything. Was going to say logarithmic instead, but the only one I kinda see that in is the first graph.

If you assume that a trend is shown, it's interesting that Washington D.C. (I'm assuming that's what DC is) is an outlier. That could be used in defense of either side of the current de facto gun ban there.

I understand what you're trying to show.

I just don't think you can start interpretting a couple points that by themselves would show a trend, when there isn't really much of a trend when you look at the whole picture.

I don't disagree with the point you're trying to make, I just think this is one of those times where you research something and the work ends up fizzling out at the end. Don't see a trend that isn't there.

Anti-gun NJ and pro-gun TX have similar violent crime rates

same for anti-gun IL (I suppose if you want to label the whole state) vs pro-gun AZ

I see those graphs as saying more that highly populated more urban-oriented states have more crime and less guns than less populated more suburban and rural areas.

then again TX becomes a problem.

I don't know...... I could be wrong. I just don't really see that any clear inferences can be made from those graphs.

Nice work though.... I'd imagine that took a fair amount of time to do.

I see those graphs as saying more that highly populated more urban-oriented states have more crime and less guns than less populated more suburban and rural areas.

Click to expand...

Yeah looks to me that the relationship is violent crime vs. population density as much as anything else. If you could somehow relate that with the gun ownership figures (and I'm not sure how you would) there might be a more definititive trend.

Very interesting, and good work.

well.... I'm not sure I really see too much of a trend

Click to expand...

Exactly. I think that is the point being made. There might statistically be some kind of trend, but it is certainly not apparent by inspection.

Gun laws vary within states. You did not really represent that.

It would also be interesting (and more compicated) to show these statistics by county. Some of the larger states may have urban counties with low per capita gun ownership and high per capita crime (and rural counties were the opposite is true).
I would guess that there are no causation effects either way...which is good enough for me!

Let us selecting an outlying state here, Hawaii. Looks like about 9% of households have firearms, about 43% of murders are committed with firearms,and the violent crime rate per 100 is about 0.25.

Now, we take another easy to find state, Alaska (AK), we find about 57% of households have a firearm. 52% or so murders are committed by guns. The violent crime rate is about 0.62/100.

"Randomly" cherry picking these states would imply that the incidents of crimes by firearms is greater in places with them than in those with heavy gun control.


Were I to look for the median of the chart, it appears that the majority of the states are centered around 40% ownership, 0.3/100 violent rime rate, and about 65% of murders committed using a firearm.


It seems that something is missing, but what?

Careful examination shows me that generally the higher populated areas have higher crime and greater gun control. However, I recognize these places as high population from memory, not from your graphs. Perhaps a chart of violent crime per 100 vs the population of the state, and the ratio of self defense uses/criminal uses vs the population of the state will help to show your point with more clarity.

Way back when I was an undecided (and way way before I realized how important it is to to tell Uncle Sam to kiss off from time to time) I researched the amounts of self defense uses vs the amount of times a firearm was used in a crime. I used the FBI database and discovered that between 1987 and 1992, per year 82,500 victims a year used a firearm in . self defense
During the same period, 667,000 crimes per year took place where the offender had a firearm. Note, however, the FBI does not track how many times a firearm was used but not discharged in self defense. The national criminal justice reference service has estimated 1.5 million defensive users, to include those who did not discharge their weapon.

Putting those self defense uses into your chart will help to sway the lurkers towards the legitimate self defense use theory; and may dissuade those who would cherry pick your data to an erroneous conclusion.

What is interesting is that overall this data is scattered. This indicates that guns and crime are not all that related. The percentage of household that own a gun can be assumed to be mean LEGALLY own a gun. Therefore it is safe to say that guns legally owned by law abiding civilians appears to NOT affect crime. This indicates that there are other factors attributing crime rates, that are not on this graph. Perhaps posting CCW percentages versus crime would show a stronger correlation. I would bet that economic factors, availability of "serious" drugs, and concentration of the population have more of an effect on crime than guns do.

Also something worth mentioning is crime TRENDS tell more than individual data points. are these crime rates averages over a certain time period, or are they crime for one month/year/week?

I agree that population density and current laws are two important criteria here.
Two thoughts about this:

1) Everyone should realize that someone who actually looks at relevant data has already made up his mind, whether he knows it or not - so this is all pretty much for our benefit. I've never seen an anti look cold hard facts straight in the face and stay anti. Most of them simply refuse to look.

2) I'd like to run with alex_trebek's ball and state that no correlation is a good thing. The whole anti argument seems to center around the idea that 2+2=fuschia. There are plenty of fence sitters out there who have never thought about that, and if they did, they'd be closer to agreeing with us.

When you do it by state it doesn't reflect thet fact that the cities are the hotbeds of criminal violence, and the cities who have higher violence rates tend to have restrictive legal gun ownership. Overall state results mask the relationship. NYC vs NY state for example, the high crime rate low legal gun ownership in NewYorkCity is masked by the low crime rate high gun ownership in the rest of New York state state. Philadelphia and the whole state of Pa is another example.

You might try doing the chart with major cities, population over 500,000 instead.

The only thing those graphs tell me is that there is no strong correlation between gun ownership and gun crime.

I threw together a quick one, but only having open office, and not knowing how to use it very well, I had to edit it in GIMP after.

States only(no DC):


Using data from the 2005 FBI report for violent crime rates, and data from the 2000 census for the population density data.

http://www.fbi.gov/ucr/05cius/data/table_05.html
http://factfinder.census.gov/servle...T-PH1-R&ds_name=DEC_2000_SF1_U&geo_id=01000US

Spreadsheet is attached(I hope I exported it right)

I suggest you graph crime against values and culture.

Very interesting. Being that you included DC, I'd love to see one that compared large cities like DC.

I agree with what Titan6 said. In many states a lawful self defense would be considered a violent crime if it had occurred in a more "anti" state.

The graphs are interesting in that they almost show no correlation in terms of crime to guns. However what are you basing your data on? When you graph ownership, is it legal ownership? Are the gun related crimes commited with legal or illegal guns? If you are comparing legal gun ownership with illegal guns then the numbers don't correlate.
What we need is a graph show crimes comitted by legally vs. illegally owned guns

Interesting, but all academic.

Supposing that we could see a correlation between church attendence and violent crime, or newspaper readership and violent crime. Would we then call for the abolition of the first amendment?

It doesn't really matter what the statistics say, a right is a right.

In MA, Boston has very different guns laws than the rest of the state. In most western MA states, you can essentially get shall-issue CCW. In Boston, Cambridge, and contiguous towns you cannot get CCW at all.

Of course vritually *all* of the crime is in a couple of square miles of gang-infested wasteland in a couple of towns inside of Boston.

That's interesting, but not the point. For anti gunners, ALL guns are bad.

Click to expand...

it would show that gun control doesn't work.
For example in today's NY Daily News 2 people were killed by gunfire (a pregnant woman and a man in his 30's) both killed on the streets. Thereby breaking the NYC's rule of no CCW. Now if we were to find out that these guns were illegal and that a disproportionate % of crime is commited by illegally owned guns than you can compare the relative low ownership of (legal) guns to high crime rates commited with illegally owned guns. It may not prove that more guns = less crime but that gun control doesn't work

You really need to do a proper statistical analysis to learn anything useful from these data. Most of the plots show no obvious trend, except...

The only thing those graphs tell me is that there is no strong correlation between gun ownership and gun crime.

Click to expand...

The only trend I see is in plot #3, which shows a clear correlation between the violent crime rate and the fraction of murders done with a gun. But I don't think that will surprise anyone.

Mannix, that is also an interesting plot and it definitely shows a weak correlation between the violent crime rate and population density.

Just a nit - the usual crime rate is expressed in the US as crimes per 100,000. In Great Britain it seems to be per 1,000. Per 100 doesn't change the relative positions any, it's just different.

The next thing to try is to plot the crime rates and such against states by Brady 'grades'. (I've done it before - you won't be surprised by the results.)

Thank you

jlbraun,

Thanks for the time and effort you put into this analysis.