LA Times publishes Lott

Kevin Drum is not pleased that the LA Times has yet again published a piece by John Lott. You generally expect some dodgy statistics from Lott and he duly delivers:

Well, more than nine months have passed [since the assault weapons ban ended] and the first crime numbers are in. Last week, the FBI announced that the number of murders nationwide fell by 3.6% last year, the first drop since 1999. The trend was consistent; murders kept on declining after the assault weapons ban ended.

Even more interesting, the seven states that have their own assault weapons bans saw a smaller drop in murders than the 43 states without such laws, suggesting that doing away with the ban actually reduced crime. (States with bans averaged a 2.4% decline in murders; in three states with bans, the number of murders rose. States without bans saw murders fall by more than 4%.)

He’s talking about crime figures for 2004, so there is just three months of post-ban data. It is ridiculous to base an argument on such a small amount of data. The differences between ban and non-ban states are tiny and I guarantee you that if they had been the other way around Lott would have dismissed them as not significant. Media Matters has more: it seems that crime figures broken down by month have not even been released yet.

Comments

#1 ben
June 30, 2005

It’s all moot anyway since the “ban” could not logically have made any difference on crime, except maybe aggravated bayoneting. The anti-gun groups even admit as much.
#2 Ian Gould
June 30, 2005

It’a also a moot point wince as JEt and others have kindly explaiend to us in the DDT discussion “ban” is a sufficiently elastic temr that one can insist, if so inclined, that there’s still a ban in place.

On a less sarcastic note: I think most informed peopel who’ve looked at the ban agree it had little practical benefit.

Then again, Lott’s claims that the end of the ban has led to a drasticd declien in the murder rate is equally suspect for simialr reasons unless one assumes that millions of Americans find the folding stocks make it easier to lug weapons around for “brandishing” purposes.

Y’know it occurred to me the other day that if American gun-owners accept Lott’s 98% claim they should be perfectly happy carrying replica guns rather than the real thign since they’d be 98% as effective in deterring criminals, cheaper and would totally removed the risk of an accident.
#3 Terry
June 30, 2005

Ian wrote, “Then again, Lott’s claims that the end of the ban has led to a drasticd declien in the murder rate “

Lott did NOT claim that. Why do you lie?
#4 ben
June 30, 2005

2 chances in 100 of needing the loaded gun are 2 too many.
#5 Ian Gould
June 30, 2005

Lott claims a 4% decline in states without a state ban a 2.4% decline in states with a state ban.

The implication is that at least a 1.6% decline is attributable to the end of the Federal ban.

He even states: “doing away with the ban actually reduced crime.”

Why do you waste people’s time?
#6 Ian Gould
June 30, 2005

No, Ben. Your chances of needing a gun are less than 100% in the first place.

Assume you have a 10% chance of being threatened (which is probably too high), in that case you have a 0.2% or one in 500 chance of needing to fire your weapon.

You probably have a higher lifetime risk (assuming Lott’s figures are correct) of dying by slipping in the bath-tub.
#7 Ian Gould
June 30, 2005

Apologies everyone for the multiple posts.
#8 SayUncle
June 30, 2005

Yeah, even us pro gunnies called BS on tying the crime rate to the awb since only 3 months expired.
#9 tim
June 30, 2005

“It is ridiculous to base an argument on such a small amount of data.”

Kinda like basing an argument that 100,000 have died in Iraq on 73 actual deaths.
#10 Tim Lambert
June 30, 2005

Dear tim, you really should steer away from any argument involving numbers.

73 is rather a lot bigger than 3, and the Lancet study’s result is actually based on 188 deaths.
#11 Terry
June 30, 2005

Ian, CAN’T YOU READ?
Lott wrote,
“Even more interesting, the seven states that have their own assault weapons bans saw a smaller drop in murders than the 43 states without such laws, suggesting that doing away with the ban actually reduced crime.”

Do you know what “suggesting” means?
You obviously don’t. You are a liar and an ignoramus. The proof is in your own words.
You fool yourself, not others dip.
#12 Ian Gould
June 30, 2005

Terry,

Well then he “suggested” rather than claimed.

Does that make it better?

I may or may not be “liar and an ignormaus” although my college transrcipts would rather tend to disprove the latter claim at least.

But neither am I a precious, hyper-sensitive hystreric.
#13 ben
July 1, 2005

Ian, I was OBVIOUSLY referring to the 2 times out of 100 when you have to “brandish” a gun. I know the odds of needing even to brandish are very small. And how about when the bad guys figure out that everyone is packing phony heat?

I also don’t see Lott claiming the reduction in crime due to the ban being lifted. All I see are stats that show that the ban did squat. What you really see are stats that show that states with more gun control don’t enjoy lower rates of violent crime or gun crime than states with less gun control. The title of the piece, flawed or not, ought to be gun control doesn’t work.
#14 Terry
July 1, 2005

Why have my two previous posts been made unreadable?
#15 Paul Crowley
July 1, 2005

2 chances in 100 of needing the loaded gun are 2 too many.

Now you’re off the deep end. Supposing there’s two chances in a million chance of “needing the loaded gun” – is that still too many? “If this bill saves one child’s life…” is a batshit way to think about public policy; you have to think about such things, and assess costs and benefits, in the aggregate.
#16 Ian Gould
July 1, 2005

What you really see are stats that show that states with more gun control don’t enjoy lower rates of violent crime or gun crime than states with less gun control.

Ben, this is actually the point I made some time ago about the AAS report on concealed-carry laws – there’s no evidence as yet that they hve any significant effect (positive or negative) on crime rates.

Absent evidence that concealed-carry laws cause harm, the classic liberal position is that peopel should be allowed to carry concealed fire-arms.

The ease of movement between US states and the large-scale market in illegal guns in states with relatively strict gun laws such as New York (which is, what, a three hour drive from Pennsylvania?) makes it extremely difficult to judge the effect of state-level gun-control laws.

I remain convinced though that it’s more than a coincidence that the developed country with the most permissive hand-gun laws also has a much higher murder rate.
#17 Tim Lambert
July 1, 2005

Ian, please don’t use << … >> for quoting because WordPress doesn’t deal with it properly.

A single > at the start of each quoted paragraph is all you need.
#18 Ian Gould
July 1, 2005

Sorry tim
#19 ben
July 1, 2005

Paul writes:

Now you’re off the deep end. Supposing there’s two chances in a million chance of “needing the loaded gun” – is that still too many? “If this bill saves one child’s life” is a batshit way to think about public policy; you have to think about such things, and assess costs and benefits, in the aggregate.

Paul, I was merely responding to Ian’s assertion that because it’s ONLY 2 out of 100 that we ought to carry replica weapons, not to any other argument about gun control. I agree with you about the aggregate assessment being more important. The most important being the principle of the matter, which is our right to keep and bear arms in defense of self, liberty and the state.

Ian says:

I remain convinced though that it’s more than a coincidence that the developed country with the most permissive hand-gun laws also has a much higher murder rate.

Then how do you explain that when comparing Caucasians only, Vancouver, British Columbia, CANADA has a slightly higher murder rate than Seattle, Washington USA (both cities being majority Caucasian)? I’ve called both cities my home over the last 15 years. It’s not the guns.
#20 z
July 1, 2005

OK, let’s be rigorously open-minded and hypothesize that law-abiding citizens who now own “assault weapons” reduced crime by 4% over three months, presumably by intimidating citizens who are not as law-abiding.

So, what is happening in the states with state assault weapons bans, that causes crime to drop by 2.4%? Citizens who are law abiding with respect to Federal laws but not state laws, who now own guns, reduced crime by 2.4%, by intimidating citizens who are even less law-abiding than they?

Perhaps we can produce a hierarchy of law-abidingness correlated with the tendency of each subpopulation to raise or lower the crime rate when allowed to own guns. That would be very useful.
#21 ben
July 1, 2005

z, legal ownership of “assault weapons” probably doesn’t have much of anything to do with crime rates, which tend to rise and fall opposite the economy. Our economy is on an upward trend, and crime is falling concurrently.
#22 Ian Gould
July 1, 2005

Then how do you explain that when comparing Caucasians only, Vancouver, British Columbia, CANADA has a slightly higher murder rate than Seattle, Washington USA (both cities being majority Caucasian)?

I believe that’s called cherry-picking.
#23 z
July 1, 2005

Facetious Yoda I was being, yes.
#24 ben
July 1, 2005

Taking two cities of similar size, with populations on the order of 1 million and within 2 hours of each other for comparison is hardly cherry picking. It’s called a textbook comparison. And it shows that there’s more to murder rates than guns.

z, my mind clouded by the dark side, it was.
#25 Tim Lambert
July 2, 2005

Well ben, if you do the comparison you suggest you find that Seattle has a significantly higher murder rate than Vancouver. However, I don’t think you draw much, if anything, from just one such comparison.
#26 ben
July 2, 2005

Not among Caucasians, Tim. That’s why it’s not the guns. Why don’t the Caucasians murder each other more if we have more guns in Seattle? It’s not the guns.
#27 Ian Gould
July 2, 2005

Ben – how many pairs of similar sized cities are there along the US-Canadian border?

Given the random distribution of murders, one could probably find at least one pair (especially if you also disagregate by ethnic group)to support any position you want.

Also, correct me if I’m wrong, but wasn’t there a serial killer in Vancouver who murdered around 20-30 women?

If, as is the normal practice, the murders were counted in the year in which the bodies were recovered, then that would significantly skew the murder rate.
#28 ben
July 2, 2005

the study was done before the serial murderer was discovered, if I recall correctly, so it would not have influenced the numbers. If they were counted for the year they were discovered, then only one or two would have been counted toward the total for that year. They were spread out over quite some time, I think.

Would’ve been nice for those women to have been armed, eh?
#29 Meyrick Kirby
July 2, 2005

Ben, all the comparison does is demonstrate that guns are not the only explanatory variable for murder. But that in itself doesn’t tell us anything we didn’t know before. It doesn’t say guns are not a factor in murder.

What can be learned from individual case studies and what can’t is an important point for researchers.
#30 Eli Rabett
July 2, 2005

Since about 90% of Canadian population is within 100 mi of the US border, you should have an easy time of making comparisons.
#31 ben
July 2, 2005

sure, and we can compare Chicago to just about any city anywhere and wonder why such a tough gun-control city has one of the worst gun-murder rates in the country. Same for DC. But fairly gun-friendly Seattle has a vastly lower murder rate than Chicago, DC, and even Boston, all gun-control havens.
#32 Meyrick Kirby
July 2, 2005

Ok Ben, I’ll soon be living in Edinburgh, Scotland, which I’m sure has a far lower murder rate than most cities in the US, and much tougher gun laws.

See, we can all cherry pick.
#33 ben
July 2, 2005

That’s swell, Meyrick. Don’t forget to register your knives

Now, do you have any cities to compare to that are geographically close to Edinburgh that have week gun-control laws? I’d be keen to see the results.

Would you feel more or less safe living in Switzerland where everyone owns real full-auto assault rifles? Lemme see… according to the numbers here, Scotland has a higher murder rate than Switzerland (2.29 to 1.32 per 100,000), but a lower firearm homiced rate (0.19 to 0.58 per 100,000). So I guess if you’d like to face a higher risk of being murdered as long as it’s not by being shot, then Scotland wins in that comparison!
#34 Ian Gould
July 2, 2005

If they were counted for the year they were discovered, then only one or two would have been counted toward the total for that year. They were spread out over quite some time, I think.

Actually, I believe they were listed as missing until their bodies were discovered.

Would’ve been nice for those women to have been armed, eh?

Yeah I guess that’s why there aren’t any serial killers in the US.
#35 Ian Gould
July 2, 2005

A correction – the original comparison between Seattle and Vancouver concluded that the murder rate was significantly lower in Vancouver.

A gun advocate then went through the data and found that if you only compared the white populations, Seattle had a marginal lower murder rate than Vancouver.

So the initial selection of the cities was not a case of “cherry-picking”.
#36 Ian Gould
July 2, 2005

…in Switzerland where everyone owns real full-auto assault rifles?

…which have to be kept in locked cabinets in the homes at all times except when their owners are doing their army reserve time and where hand-gun regulation is much stricter than in the US.

I suspect that most US gun-control advocates would be DELIGHTED to see the US adopt Swiss gun control laws.
#37 Meyrick Kirby
July 2, 2005

Now, do you have any cities to compare to that are geographically close to Edinburgh that have week gun-control laws? I’d be keen to see the results.

The number of homocides in Edinburgh was 6 in 2003: link

Since Edinburgh has a population of approximately 450000, that makes a murder rate of approximately 1.4 per 100000

Now, as I asked before, find me a city in America with a lower one.
#38 Meyrick Kirby
July 2, 2005

Now, do you have any cities to compare to that are geographically close to Edinburgh that have week gun-control laws? I’d be keen to see the results.

Hummm, yes, the murder rate in Switzerland is barely greater than Edinburgh (1.32 versus 1.4). So yes, those locked rifles don’t seem to be making much difference.
#39 Ian Gould
July 2, 2005

Looking back at Lott’s initial article he says that 2005 saw the first drop in the American murder rate since 1999.

Ummm, are we to assume that no states passed concealed-carry laws during that period?
#40 Chris Jarrett
July 2, 2005

For Meyrick Kirby -

You should take a look through the FBI UCR statistics. There are cities in the USA with 2003 murder rates lower than 1.4

For example, Cambridge-Newton-Framingham, MA (1.5 million people) shows a rate of 0.7

Boulder, CO – 1.2

Fargo, ND – 1.1

There aren’t many but there are some.
#41 Eli Rabett
July 2, 2005

The problem with doing this sort of comparison with cities, is, for example, Washington DC neighbors on VA which has right to carry laws and where gun stores are common. Seems to me that you should make such comparisons by metro area (urb+suburb).

It also seems to me that if you are going to fish in the Vancouver/Seattle comparisons, you should do so by income level not necessarily by ethnic group, or at least control for income level in any comparison.
#42 ben
July 3, 2005

So the initial selection of the cities was not a case of “cherry-picking”.

thanks Ian.
#43 ben
July 3, 2005

I suspect that most US gun-control advocates would be DELIGHTED to see the US adopt Swiss gun control laws.

Well, we wouldn’t mind having the government provide us all with full-auto assault rifles that we had to keep locked up unless needed or for practice. We could do without the handgun laws, thanks.
#44 Ian Gould
July 3, 2005

we wouldn’t mind having the government provide us all with full-auto assault rifles…

How about the mandatory military service that goes along with them?

Hmmm, a well-ordered militia – maybe these poeple are onto something.
#45 Meyrick Kirby
July 3, 2005

Oops, read the wrong city, it was 5 for Edinburgh, making roughly 1.1-1.2 per 100000. So that only leaves Cambridge-Newton-Framingham lower. Not that this means much (apart from I’m now happier about moving in September), since this is all cherry picking.

Individual comparisons (i.e. multi case studies) are useful, but generally as a theory building exercise, and for getting a closer look at the causal factors (chains of events) in action. When it comes to testing (refuting) theories it becomes a bit tougher.