With something like guns comes great responsibility and I would like to see mandatory training and possibly some kind of testing for the right to conceal carry. Maybe even a mental health exam or test. If you want to conceal carry you should prove at least some level of responsibility.

It reminds me so much of where I grew up. Not too far from the Canadahoovian border.

Yes, I know. I had noticed that but did not bother mentioning that additional flaw in the study . The earlier study, supposedly being “replicated” by this later one was across towns.

Regarding the rest of your argument, I objected to one thing you said: That crime rates in the US and Canada were identical, and that this was proven by the study you cited. They are not and it is not.

Regarding further arguments about gun use, gun laws, gun ownership, and crime, those arguments are all very interesting but the key point there is that the US and Canada are vastly different cultures (despite the tendency of Americans to think of Canada as a mere northward extension). The very fact that Canada has a pretty good national health care system and we don’t will impact on this issue, the fact that in both countries poverty is very serious among Native populations but that population is relatively much much larger in Canada than the US does interesting things to the crime data, and the fact that the US was a slavery country for so long and Canada was not is very important.

Let me try to be more clear by considering this objection:

This would be like arbitrarily dividing a region within the US by a particular meridian, looking for differences between those within miles on either side of the meridian, finding none, and stating that “being on opposite sides of 32 degrees north has NO EFFECT!!!!”

It would make a huge difference if the laws were completely different on opposite sides of that line. To take a silly example, say the speed limit on one side was 55 MPH and on the other it was 70 MPH. Other than that, the populations on either side are identical. If there were a difference in the average speeds of driving on each side, that would be an indication that speed laws affect the speed people actually drive. If there weren’t a difference, it would be an indication that it didn’t.

Which is what we see with respect to gun laws. I can strap a Glock to my hip and walk around in public in northern Montana with zero legal consequences. A few miles north in Canada, I’d go to prison. The gun laws, ownership and carry rates are very different on either side of the border. Little else is – and that includes crime rates.

In this case, I strongly suspect it is what I said it is … These cross border communities are not different communities to the extent you might think. They are one community united by a border that people have traditionally crossed, worked across, visited across, shopped across, married across, committed crimes across, and so on.

This would be like arbitrarily dividing a region within the US by a particular meridian, looking for differences between those within miles on either side of the meridian, finding none, and stating that “being on opposite sides of 32 degrees north has NO EFFECT!!!!”

Which would be really dumb and uninteresting.

“Many border towns are shared cultures with lots of people moving back and fort across the towns and sharing an economy, and we already know that local economic conditions predict crime rates to a very large degree. So, picking matched across the border crime is an utterly invalid test.”

In fact the opposite of your last sentence follows from your first. The whole point of the comparison is that when all other things are equal (shared economy, etc), gun ownership rates and gun laws don’t do squat. When other things aren’t equal, neither are crime rates. Quelle suprise, as they say in Quebec. (If I gave the impression I thought overall Canadian crime rates were lower, well, that would just be silly. The variable of interest is the rates among otherwise similar populations – we want guns to be as close to the only variable as possible.)

Think of it like a medical study. You need two similar sets of patients, and you give one the placebo and one the drug. See if there’s any statistical difference. Here there isn’t. What you’re saying is effectively “The drug works” when you’ve made sure to add in a bunch of medical basket cases to the placebo group but not the drug group.

The “more guns in Canada” statement is probably not right. There probably as many guns in the US as people, with somewhere north of 4 million a year added to the market.

Homicide rates in Canada are lower than in the US. The fact that the two have changed in a similar direction is not important and is not what you said. Many property crime rages are higher in Canada than in the US. Violent crime rates in general are higher in the US than in canada. If you look at certain crimes in both the Canada and the US, you find them increasing as one moves away from the border for a number of reasons. Many border towns are shared cultures with lots of people moving back and fort across the towns and sharing an economy, and we already know that local economic conditions predict crime rates to a very large degree. So, picking matched across the border crime is an utterly invalid test.

From the very study you cite, “It should be noted, however, that this kind of comparative analysis generally allows for a low level of statistical inference.” This is because the study is statistically utterly bogus. It does not compare US vs. Canadian crime rates.

So, when you said “Indeed the crime rate is not different at all.” and then you said “I’m always happy to be corrected” and “The overall Canadian homicide rate is substantially lower than the US as a whole” that was a nice bit of backpedaling, but when you said “but comparisons between demographically and socioeconomically similar areas reveal very little difference. ” you perhaps did not know about statistical independence. That’s basic, Matt. But it is never to late to learn.

Putting it more broadly, if you have two huge data sets, it is often possible to select several pairings of data, one from set A and one from set B and compare them, and when you don’t get the result you throw it out. Then you pick a different paring, and another, and another, until you get what you want in two or three parings. Then you chose, amomg them, the pairing that you can best make up some winged monkey expiation for. That is called cherry picking. That is another statistical flaw.

By the way, gun ownership rates in Canada are higher, not lower, than they are in the US, as I understand it. It’s just that the Canadahoovians tend to pack long guns.

Read this and report back:

http://www.google.com/url?sa=t&source=web&ct=res&cd=5&url=http%3A%2F%2Fwww.statcan.gc.ca%2Fpub%2F85-002-x%2F85-002-x2001011-eng.pdf&ei=qfDNSp7fMZHONZuu0P8C&usg=AFQjCNHRKOOrPhkRZgYjkTUSBUVjpW4m9w&sig2=BLJw-vwXZbw2AQuxVYg19g

itzac — I certainly do not subscribe to the philosophy that if everyone had guns, life would be fair… and crime would cease to exist. What I meant to convey is that the criminal activity of dealing drugs makes the possession of a firearm more desirable… AND that the legalization of drugs would lessen the desirability of firearm ownership.