Alphecca

Correlation does not equal causation. If you take the study at face value, you could just as easily conclude that people who live in violent areas feel the need to own firearms. From a common sense standpoint, it's far more likely that people who are concerned about their safety would arm themselves.

Hey isn't it funny for 12 years these gun control freaks were out of sight. I wonder why they are back everywhere now? I think gun control is a Democrat disease, along with higher taxes and communism for the masses.

And all along I thought Democrats were concerned about our freedoms, the library card and phone taps.

Now we know who really wants to take away your freedoms, Democrats.

"The survey found that approximately one in three American households reported firearm ownership."

This is one of the highest firearms ownership rates I've ever seen in print. If true, it's a big hurdle for the gun grabbers.

Only 1 in 3, eh? Must have surveyed mostly in liberal run cities.

I would estimate that this in an underreport by anywhere from 15-30% (rather than 33%, it would be more like 50-60%), if you could actually get an honest answer nationwide (you cannot, as anyone who thinks about it will decline to honestly answer the question...you never know who's calling).

Guns are one of those 'forever' products that so vex their manufacturers. Unless deliberately destroyed, or abused beyond common sense, they never go away. Which isn't all that great for sales. When one considers that there have been literally half a billion of them made, you have to ask yourself - where did they end up?

A lot of them are in my closet... ;-)

I wish i had the money to go to Harvard,so i could engauge the liars who stand behind the First Amendment. In their quest to disarm the Second Amendment and the people of the United States of America they shall fail.

Based on my personal experiance, it seems at least, if not more likely, that causation might run from high levels of violance in an area to more people deciding to be armed than the other direction. Regardless, every time I encounter one of these physicians or public health types meddling is areas far beyond their area of expertise, I like to entertain them with something I learned years ago in grad school:

"There was once a cholera epidemic in Russia. The government, in an effort to stem the disease, sent doctors to the worst-affected areas. The peasants of the province of S_____ discussed the situation and observed a very high correlation between the number of doctors in a given area and the incidence of cholera in that area (i.e. more doctors were observed in cholera areas than elsewhere). Relying on this hard fact, they rose and murdered their doctors."

Franklin M. Fisher, The Identification Problem in Econometrics,
(New York: McGraw-Hill, 1966) pp. 2-3.

A sixth-grader could spot the glaring flaws in the Miller "study."

Every criticism you make of the sudy is wide of the mark because you didn't read the study before criticising it. See here.

Here's what the study says about reverse causation:

It is possible that a
non-causal relationship explains our findings or that
the association we observe might have arisen
because individuals in states with historically high
homicide rates acquired more guns (than did
individuals in low-homicide states), as a defensive
response to actual high homicide rates in their
communities (i.e. ‘‘reverse causation’’). This broad
notion of reverse causation, while consistent with
our association between household firearms and
firearm and overall homicide, does not explain why
firearm ownership is not also significantly associated
with rates of non-firearm homicide. Furthermore,
rates of robbery and aggravated assault are not
associated with household firearm prevalence, even
after controlling for urbanization and resource
deprivation (not shown). Since individuals who
obtain firearms in an attempt to protect themselves
from violence plausibly respond to non-fatal violence
(which is far more common than fatal
violence), the lack of association between firearm
prevalence and non-lethal violent crime militates
against reverse causation as an adequate explanation
for our findings.

The fact that the press release states that the Joyce Foundation sponsors the "study" along with my previous post that Hemenway's book was sponsored by George Soros, plus the numerous press accounts, along with their track record that I've detailed, is all I need to know to realize that these "studies" are fake-phony-frauds. As Glenn says, if the NRA sponsored "research" you'd be the first one at Deltoid crying "bullshit."

Another issue worth investigating is, to what extent did the study consider that gun ownership may deter violent crime?

The nine state cited as having firearm prevalence more than one standard deviation above the mean have an average violent crime rate of 349 per 100,000 persons, according to the 2003 FBI UCR.

The nine states that had a firearm prevalence that was more than a standard deviation below the mean had a violent crime rate of 422 per 100,000.

Is there a deterrent effect operating here? Did the study consider this?

I knew Tim Lambert would show up eventually to defend another gun control advocacy "study" :-) I haven't known Tim to concede that even a single anti-gun "study" is deficient. He can be found attacking all the pro-gun studies instead. Correlation or causation? You decide.

So what the study says and did doesn't matter? Says all we need to know about you. Are you going to correct your false claims about it including suicides and relying on dividing the states into four groupo?

And I have never dismissed Lott's work because he was funded by the Olin Foundation, so you're wrong there as well.

Hey isn't it funny for 12 years these gun control freaks were out of sight. I wonder why they are back everywhere now?

Simple. The Democrats have majorities in both chambers of Congress.

When the Dems are out, the gun grabbers shut up so the pro-gun voters won't be riled up come election day. When the Dems are in, the gun grabbers revive and start trying to push their agenda... which results in the pro-gun voters showing up at the polls next election day. Lather, rinse, repeat.

The good news is that if the trend holds, we'll have another Republican president come Jan 20, 2009. Hopefully he'll be a real pro-gun R instead of one that just avoids annoying the pro-gun voters.

Jeff, you really do need to "control" for other variables in this sort of analysis. Just looking at the raw data is a cheesy way to ignore information.

I haven't read the full study, but from the blurbs, it sounds like they did regression analysis. This is a very powerful technique that allows the effect of a particular variable to be isolated from others.

There are three problems with it though:

A: It's easy to cheat and reach a desired conclusion by picking what variables to control for.

B: It's damned near impossible for anybody without a lot of statistics training to understand in detail. And no, I don't claim to have that training.

C: It's easy to get away with A because of B.

Jeff,

I love it if you could provide a .pdf format of this article that I could save to my computer, that way I could just print it out and hand it to the people who are going to throw this in my face.

The publication's text states: "States with firearm prevalence more than one standard deviation above the mean: Alabama, Arkansas, West Virginia, Mississippi, South Dakota, Idaho, Alaska, Montana, Wyoming" - which is less than 6% of the US population at the time of the surveys.

That's a pretty small 'truth' to get all strirred up about. The Harvard group's struggle is less to say something true and more to say something important. So, they add tastey text-bites in their abstract/conclusions for the news media to hungrily snap up, and bury the caveats in the paper elsewhere. That way they can leave the dirty-work of misleading the public to the professionals.

Ermmm.... how did they determine the percentage of homes with firearms?

Most gun owners I know would respond to a pollsters question about such with a short "Fuck off and die".

The question to ask is: What if there is more gun crime in areas with more guns? My answer is, so what.

I will not surrender my Constitutionally guaranteed right to bear arms regardless if guns prevent, cause or are indifferent to crime.
A case could certainly be argued that the media causes crimes to be sensationalized and that this causes copycat crimes. Would the media support restrictions on their right to publish in order to reduce these crimes? I suspect the answer would be, hell no.

I will not surrender my rights and freedoms no matter how many studies are published that say we will have a measure of safety if restrictions are passed.

The report says the study controlled for resource deprivation (poor?), urbanization (city), aggravated assault (attack with a weapon to cause great bodily injury), robbery (aggravated assault + theft), unemployment, and alcohol consumption.

I have found that the first thing to do when presented with controls is to check that the terms the study uses actually mean what the law says they do (example, the VPC included in its "Assault weapons" some guns specifically exmpted in Federal the law). Second is to make sure that all things that should be controlled for are. Last is that nothing is controlled for that should not be.

Since I can't see the full report I will skip the first part.

The second part, I understand why most of those were controlled for, what I do not understand is why things such as Gangs, Burglaries, Hate Crimes and Forcible Rapes were not controlled for as well. All four are reasons why people go out and purchase firearms. Have you heard of someone going out and purchaseing a gun because they fear gang initiations, home invasions, threats from groups like the KKK/Neo Nazis, or serial rapists? I have. Which makes me wonder why these were not included? I know some states have a much larger gang problem than others.

Third, What does alcohol consumption have to do with firearm homicides? Note that it is not a crime to consume alcohol. Did they go by gallons (amount of liquid) or proof (alcohol content) consumed? Since beer has less alcohol than wine or liquor, the quantity drank to acheive the same level of intoxication will be much higher for beer than wine or liquor, so a region that drinks beer over wine/liquor will drink more (have higher gallon consumption) but could still be equal in their actual level of intoxication. So if quantity drank is used with no regard to the proof of alcohol, regional differences between areas could skew the the results. Such as California being home to many wine vinyards with a large group of the population drinking small amounts of wine could consume less overall than a state with a large group of the population drinking beer. The quantity of liquid consumed would be more for the beer states, but the percentage of alcohol in that liquid could mean that the wine drinking areas are more intoxicated than the beer drinking ones.

Someone please explain why such a possible bias was included for firearm homicide rates because I can't think of one (other than skewing the results). I understand including crimes for public intoxication and DWI etc.. (crimes committed while intoxicated) but just not the consumption of alcohol.

"Since individuals who obtain firearms in an attempt to protect themselves from violence plausibly respond to non-fatal violence (which is far more common than fatal violence), the lack of association between firearm prevalence and non-lethal violent crime militates against reverse causation as an adequate explanation for our findings."

As I posted just above, there are other crimes that were not controlled for that are bigger reasons for protection than what was accounted for in the study. If multiple reasons for protection have been left out of the study, it means that reverse causation has not been eliminated.

> Jeff, you really do need to "control" for other variables in this sort of analysis. Just looking at the raw data is a cheesy way to ignore information.

Translation: Gee, what I see here in front of me doesn't support my preconcieved worldview. But I can't accuse Jeff of massaging the data... that's the raw take!

Oooh, oooh, I got an idea. I'll accuse him of not massaging the data enough. And then if he does, I'll just accuse him of massaging it the wrong way. Win-Win! I'll never have to re-examine my beliefs now! Take that, you dirty raw-data-shower!

All of this back and forth about statistics and studies, while thoroughly interesting, is moot.

The bottom line is, and listen up Mr. Lambert...

It Doesn't Matter!

Ok?

The foundation of free societies is the right to own property. Said property must be defended. Free people cannot be free unless they have the right to defend that freedom and their right to own something. There can be no true freedom without the right to self and property defense, only serfdom.
The 2nd ammendment is an individual right, and no amount of lawyering and pontificating will change that, regardless of what the 9nth circus court of liberal soci@list bedwetters would like us all to be duped into believing. It is not "collective", and cannot be defined as such.
All anti-gun arguments collapse in the face of this truth. That is the bottom line. Save your statistical manipulation and false definitions for those apathetic and ignorant enough to believe them.
Thatisall!

Jeff,

Just a question/point or two.

I wonder if it might be possible to include in these statistics who the governors, and maybe mayors, are as well as their lengths in office. Party affiliations, too.

Why I ask, here in Maryland - for the past 30 to 40 years as I know it - the State AG has been a Democrat that has long held the position that individual gun ownership should be rescinded/banned. His son-in-law, recently governor-elect, had been Mayor of Baltimore for some time and was heard once on WBAL radio to tell a rational, concerned small business owner he would have to move to another state because he would never be allowed a CC permit here. Despite the fact that person was doing everything by the book - the local & state police kept "losing his paperwork."

Just an observation.

Oh, and one more thing I forgot to mention.

That AG is about to be replaced with a new Democrat. But apparently holds the same (non)values.

But what I really wanted to add, was that the outgoing AG was deeply affected by his father's mugging and consequent death due to a gun-wielding criminal.

Which, to me, taints their judgement (both the outgoing AG and his son-in-law, the incoming Governor) beyond the pale.