Correlation between Gun Control and Gun Homicides (updated)


PDA






cbrgator
April 9, 2009, 06:27 PM
I recently made a youtube video explaining that stricter gun laws don't equate to lower gun homicide rates. It sparked me to make this chart to show that there is NO correlation between the 2. The chart illustrates that stricter laws do not indicate more OR less gun crime, clearly indicating there are other factors that make this determination. If there was a correlation, there would be a linear trend starting at the bottom left of the chart, up to the top right. As you can see, there is no such trend, albeit one outlier. Feel free to use this chart at will.

http://i37.photobucket.com/albums/e96/SMAcbr/chart-1.jpg

If you enjoyed reading about "Correlation between Gun Control and Gun Homicides (updated)" here in TheHighRoad.org archive, you'll LOVE our community. Come join TheHighRoad.org today for the full version!
Vonderek
April 9, 2009, 07:13 PM
I would think the chart would trend the other way. It doesn't make the argument for "more guns, less crime." BTW-where is Florida?

cbrgator
April 9, 2009, 07:23 PM
No, the states to the left should technically, according to the Brady campaign, have low rates, whereas the states on the right with lax laws should have high crime rates. For whatever reason, the FBI does not report Florida's statistics. OR if there was a negative correlation, that strict gun control means more crime, the left side would have a high crime rate and the right side a low crime rate. The point is to show that there is simply no correlation whatsoever.

Shung
April 9, 2009, 07:26 PM
Great graph mate !

im gonna use this thing in our incomming vote about gun control. thx

marklbucla
April 9, 2009, 07:28 PM
It doesn't matter what that chart said. Correlation does not imply causation.

JImbothefiveth
April 9, 2009, 07:47 PM
You should also have one with total murder rates. Reducing one firearm death does no good if it leads to 2 other deaths. Also, Michigan has really good CCW laws, yet you put that as most strict.


All these states have lower crime than WI.
MO ME CN RI OR WA CO MN NH NE VT WY MT ID SD ND

About 43% of the shall issue states

While about 48.5% of the shall issue states have lower crime than IL
MO ME CN RI OR WA CO MN NH NE VT WY MT ID SD ND OH KA WV

MD MA are the only may-issue states with crime less than WI, or about 25% of may issue states


NY DE MA MD are the only may-issue states with crime lower than IL, or about 50% of may-issue states

The state with the least, HI, is a tourist destination, so probably has less poverty which would explain the lower crime.

The state with the most is recovering from a Hurricane

Finally, it has been shown that CCW often lowers crime, and I've never heard of crime raising after.

cbrgator
April 9, 2009, 08:18 PM
JImbothefiveth, the states are ranked according to the Brady Campaign's own ranking system. I took the rankings right off their website. I didn't rank the states myself.

cbrgator
April 10, 2009, 10:51 AM
Just a bump. I thought this would get better positive feedback.

Southern Rebel
April 10, 2009, 11:18 AM
It is extremely difficult to make any judgment of the relationship between two components unless other "potentially higher level of effect" components are filtered out. Without that filtering process, the resultant data comparisons can be skewed either way. For example, I would think that large metro areas, economic level, minority population mixes, and illegal immigrant population could all have an effect on certain states and how they rank.

I am certainly not a statistician, but am a mathematician and have seen how easily numbers can be manipulated. That doesn't mean that I don't agree with your concepts - I just would like to see a sophisticated review of all significant factors would look. (Nope, I am not sophisticated enough to be of any help!)

Librarian
April 10, 2009, 04:40 PM
It is extremely difficult to make any judgment of the relationship between two components unless other "potentially higher level of effect" components are filtered out. Without that filtering process, the resultant data comparisons can be skewed either way. For example, I would think that large metro areas, economic level, minority population mixes, and illegal immigrant population could all have an effect on certain states and how they rank.
None of that stuff matters much, either.

See John Lott's More Guns, Less Crime and his web site johnrlott.blogspot.com/

Note that a lot of people find Lott controversial.

Cyborg
April 10, 2009, 05:10 PM
hey, cbrgator, how about a scatter diagram plotting Brady Foundation score vs violent crime rate? Or even gun crime rate. No need to plot lines between and the only coordinates on the graph would be Brady rating (numerical score not ranking) vs violent (or gun) crime rate. If I had the data (or a link to go to so I could GET the crime rate statistics) I would be willing to plot it. I have an excel spreadsheet set up and the Brady scores entered. If I had the crime rate stats I could have it ready to present in fairly short order. I figure on having a small circle with the state abbreviation for a label at each point.

Southern Rebel, if there were any significant correlation between Brady score and gun violence (and I would not limit it to homicides) I believe such a plot would show it. If, instead, the graph looked like someone had shot at a piece of graph paper with bird shot then I would say it would be a fairly persuasive araument for there being no correlation. Maybe you can think of a way to normalize the crime statistics to account for the fact that Montana doesn't have much in the way of urbanization where some states have little else. Or, failing that, it might be possible to just get the data from cities.

But all that being said, there is no way folks who believe that guns are evil are going to be swayed by graphs and statistics. For them it is a matter of faith and not knowledge anyway. In their minds guns in the hands of the average Joe/Jane is wrong. One of the security guards - armed - who works at the building where I work (there is a Credit Union in the same building) does not believe that ordinary people should be able to buy/own/carry firearms. It matters not that he is one of the least prepossessing individuals I have ever seen wearing the uniform of a commissioned security officer. He thinks that average people owning/bearing arms is a tragedy waiting to happen. Going solely on appearance he is the poster child for his own argument. His uniform is untidy, his bearing anything but professional and his rig is a joke. He doesn't even use keepers to secure his underbelt and gunbelt. He thinks the thirty hours of classroom training and bi-annual 4 hours of refresher makes him an expert in use of deadly force. He scares me.

cbrgator
April 10, 2009, 06:50 PM
Southern Rebel,

Your point is absolutely correct and extremely well taken. My point was to visually prove YOUR point. That the gun legislation is not what affects the crime rates, but the other factors such as demographics and highly populated areas.

Cyborg,
Here is a link to the data you want.

http://www.fbi.gov/ucr/cius2007/data/table_05.html

verdun59
April 10, 2009, 07:31 PM
+100 on demographics.

Jeff White
April 10, 2009, 07:49 PM
There is a 2004 National Academy of Sciences study on gun laws and crime. It found that there is no correlation between gun laws and the crime rate.

cbrgator
April 10, 2009, 08:50 PM
Jeff, that is untrue, and often we often skew the results in our favor to say that (I'm guilty of this as well on youtube). The true conclusion of that study was that we do not have accurate statistics or methods of tracking guns to make any conclusions. It's not that there is no correlation. Essentially what they said was that we simply have no way of knowing.

Ky Larry
April 11, 2009, 12:29 AM
Didn't Mark Twain say "There are lies, damn lies, and statistics." I am not educated enough to understand and compile statistical data, but I know any group of data can be made to "prove" just about anything.

Isher
April 11, 2009, 12:37 AM
"Figures lie

And liars figure"

One of my late mother's favorite aphorisms.


enough said


isher

JKimball
April 11, 2009, 12:54 AM
Nice work, cbrgator. That is a good graphic. Thanks for posting.

I wonder what it would look like using the average murder rate over the last 10 years or so.

PT1911
April 11, 2009, 12:57 AM
Oh man... I need to move....

maskedman504
April 11, 2009, 12:58 AM
Where is FL? :fire:

LaEscopeta
April 11, 2009, 10:18 AM
Correlation does not imply causation.Causation is EXCATLY what correlation implies. It dose not PROVE causation. Correlation points the direction to look for possible causation.

Also note the original post does not claim to show causation, just correlation.


All these states have lower crime than WI.
MO ME CN RI OR WA CO MN NH NE VT WY MT ID SD ND
About 43% of the shall issue statesTaking on aspect of gun control (shall issue) and looking at select states is an example of cherry-picking data, the main method of making data “prove” what you want it to prove. We should leave cherry-picking up to our opponents.

cbrgator
April 11, 2009, 01:03 PM
Where is FL?

The FBI doesn't have statistics compiled for FL. Sorry.

LaEscopeta
April 11, 2009, 01:24 PM
The FBI doesn't have statistics compiled for FL. Sorry.

It takes the FBI a while to compiles stats from all 50 states, conforming to a constant format from the 50 different ways each state keeps their crime stats. if you look back a few years (2006, 2005, etc) you will find FL crime states.

cbrgator
April 11, 2009, 02:06 PM
Yea, I just meant for 2007.

Southern Rebel
April 12, 2009, 10:32 PM
Southern Rebel,

Your point is absolutely correct and extremely well taken. My point was to visually prove YOUR point. That the gun legislation is not what affects the crime rates, but the other factors such as demographics and highly populated areas.

And your point is now understood! :banghead:

Now, if lack of enough gun laws does not cause crime....
and guns and lawful gun owners do not cause crime.......

Hmm, maybe, just maybe lack of control of criminals just might cause crime?
If that be the case, our illustrious new AG has a tough decision to make:

Do I investigate what causes criminals to be criminals?
Do I instead investigate how to control criminals?
(Either way, the focus would be off of us!)

Years ago, I was casual friends (mostly hunting and fishing) with a constable for a small town which had a reputation for some tough hoodlem types. I asked him how he managed to keep the town reasonably calm and safe. He had a theory that may not be outdated: (His words as best paraphrased as an old memory will allow)

"Well, I figure there are three groups of people in my town. There is a small group of gonna-be lawbreakers that will never be decent people. There is a fairly large group that might break the law if it was to their benefit. Then there is a fairly large group that just won't ever be a problem. I put all my effort toward the first group. I make sure that their arrests are quick, their jailtime is hard, and both are pretty much guaranteed.

"Now I ain't stupid - I know they will get out eventually and do the same old stuff again. But seeing what happens to them is what keeps MOST of the second group under control. And the third group? Well, I just try to stay out of their way so that they will vote me back in again, next election time!"

'THE SCHOOL OF EXPERIENCE IS A HARD TEACHER, BUT FOOLS WILL LEARN IN NO OTHER!'

gego
April 13, 2009, 05:44 AM
Causation is EXCATLY what correlation implies. It dose not PROVE causation. Correlation points the direction to look for possible causation.

Also note the original post does not claim to show causation, just correlation.

I have heard it said that correlation does not prove causation, but without correlation you cannot have causation. Here we have a case where there is not correlation, hence you can be sure that lack of gun control does not cause gun homicides.

If there had been correlation, then we would still not have proof, but cause and effect could have been possible. So the graph is quite meaningful. Is my logic incorrect?

qwert65
April 13, 2009, 06:15 AM
gego, It means that they could be linked or not but just bc they are correlated dosen't mean one causes the other. for ex. having a high concentration of churches is correlated with high crime rate, but one has nothing to do with the other (except that they are in cities)

HoosierQ
April 13, 2009, 09:48 AM
You would expect, as noted, if the antis were right, the the graph to be low at left, high at right and there to be a steady trend upward as we move left to right, most to least gun control. The seeming randomness off the whole thing tends to support a lack of correlation.

Correlcation is the key word. If you plotted a straight line that reflected the various means all the way across, it looks to me like that line would be nearly horizontal. LA kinds of messes things up a bit but in statistics, one often throws out the high and the low so who knows.

Dr. Tad Hussein Winslow
April 13, 2009, 10:43 AM
Louisiana is the fly in the ointment, but overall, it's clear there is no correlation. They need CRIMINAL control in LA.

rscalzo
April 13, 2009, 10:53 AM
other factors such as demographics and highly populated areas.

You took the words right out of my mouth. Statistics can be bent in just too many ways to fit the desired outcome.

mordechaianiliewicz
April 13, 2009, 11:07 AM
No one has ever been able to do more than correlation. But, the strongest correlations out there are poverty and crime. People being poor tends to correlate with higher crime more than any other factor. (Though even that isn't a direct straight line).

But, I like the story of Southern Rebel about the leo he knew. There are always criminals in any given place. Those folks in the middle who would commit crimes if they thought they could get away with them, yeah.... those guys get poor, and the incentive to do the wrong thing goes up. They commit those crimes they wouldn't have, had they had a better job, more money, not in debt, etc. and so forth.

Shoot, Mexico's problem (not to go too far off topic, but this is a good example) is based upon the fact that Mexican LEOs have very little ability to enforce laws, and cannot even keep the peace. While there are many good Mexican people who aren't going to get involved in the criminal activities, virtually every Mexican "person in the middle" who would, is. No consequences, and in some cases, coersion exists to do the wrong thing.

Louisiana has the same problem right now.... so does Illinois. Just not as dramatically.

RP88
April 13, 2009, 11:24 AM
the only correlation I see there is a correlation of crime and poverty/lower class income. Most of the states with high crime are also known for larger impoverished or low-class populations.

LaEscopeta
April 13, 2009, 12:13 PM
If there had been correlation, then we would still not have proof, but cause and effect could have been possible. So the graph is quite meaningful. Is my logic incorrect?Well, my logic matches yours. TIFWIIW

RyanM
April 13, 2009, 02:51 PM
Just for fun, here's a different take on it.

Cost of a concealed carry permit in US dollars per year for the initial issuance, versus homicide committed with a handgun, rates per 100,000 population.

http://www.thehighroad.org/attachment.php?attachmentid=96178&stc=1&d=1239648624

For may-issue states with common issuance (Alabama and Connecticut) the cost was unchanged. For may-issue states with limited issuance, $50 per year was added to the cost, up to a maxiumum of $100. No-CCW states were put at $100. That breaks up the data enough that you can compare shall-issue, may-issue, and no-issue states (though MD costs over $58 a year for all 3 of the people with CCWs there, so it got bumped all the way over to no-issue).

So much for easy CCW contributing to people murdering each other all the time, dueling in the streets, etc. Back to the drawing board for Brady.

What I'd really like to do is CCW $ vs. justifiable homicide with a handgun, but the FBI didn't publish per-state justifiable homicide data. :(

RyanM
April 13, 2009, 03:05 PM
And for even more fun, here's CCW cost per year, as above, vs. percent of murders committed with a handgun.

http://www.thehighroad.org/attachment.php?attachmentid=96179&d=1239649489

If you enjoyed reading about "Correlation between Gun Control and Gun Homicides (updated)" here in TheHighRoad.org archive, you'll LOVE our community. Come join TheHighRoad.org today for the full version!