John Lott credible enough for you? I had his books from the gun club library so I don't have the precise citation I first read, but this interview repeats it.
It's not whether a person is credible or not - I like to read the studies myself.
John Lott is not - from what I have read - what I would consider a credible source, without carefully reading the studies he's reporting. But I feel that way about most people. Heck, I am learning Hebrew so I can read the Bible myself - I don't trust translators/
John Lott is more of what I would consider an
activist than a scholar(not that there is anything wring with that). He's pretty willing to play fast and furiously with the numbers. For example in the second article you cite, he says:
But Americans also used guns defensively more than 2 million times that year, and more than 90 percent of the time merely brandishing the weapon was sufficient to stop an attack.
He claims for a fact - the guns were used more than 2 million times defensively in a given year - what is not a fact, but a speculation based on on other speculations.
The truth is that no one knows how many times guns are used defensively each year - as you can imagine, most defensive uses of handguns gun do not result in any paperwork, which makes them very hard to count.
I have read about the derivation of the "2 million times" estimate before - as I recall, it's an extrapolation based on set of assumptions about the frequency of the unreported uses of handguns versus the reported uses, etc. That number is an extrapolation based on a series of assumptions.
If he states as an absolute fact something that is almost pure speculation - that raises red flags for me.
Scholars are very careful with facts, activists are not. A scholar would almost never state speculation as fact. The careful distinction between fact and speculation makes scholars very annoying.
I have no doubt that John Lott is competent economist and a scholar, but he seems to flip back and forth between scholar and activist mode without warning, and I'd have to study his numbers fairly carefully to determine whether activist or the scholar is speaking.
Do you recall anything about the methodology of the study that reported the following?
... when she's confronted by a criminal ... A woman who behaves passively is 2.5 times as likely to end up being seriously injured as a woman who has a gun.
Reasons to be skeptical:
I also note that he was pretty careful
not to say whether the women who were attacked and had guns used those guns to defend themselves during the attack.
Note that he very careful to slip in a bit of a false dichotomy: "passive" versus "has a gun".
I would be looking for evidence in the study that some other factor was
not the causative factor.
For example, maybe the women who had guns in general had more financial resources. Most women are killed by men with whom they are in a relationship and/or ending a relationship. Women who are poorer in general stay in abusive relationships longer (a conjecture), and thus may be more likely to be the victim of multiple abusive attacks. If that's the case we would expect a woman who has more financial resources gets out of the relationship before sustaining serious injury than a woman with fewer financial resources.
It could also be that the kinds of women who have guns are less likely to get involved with men who who will cause them serious bodily injury. Whether or not she "has a gun", she's 100% certain never to be seriously injured by her husband. I would expect that women who own guns are more independent and self-confident than those that don't. That may mean that they are less likely to be involved in an abusive relationship.
In general, you would control for these other factors by starting with a very large sample, comparing two sets of women who vary only by one factor "has a gun". But I would have to believe that the sample in this case is very small.
Assume:
- I would guess that far less than 1% of women own guns - other than in the very general sense that my wife has a gun because I do.
- Assume that half of violent crime in the US is directed against women - which is a very high estimate (most women who are the victims of violent crime attacked by men with whom they have a relationship, and are unlikely to report that crime).
- Assume 25% of crimes are are violent crimes (that may be high -many criminals prefer theft, larcency, auto theft, etc.).
So 450,000 crimes * 25% estimate of violent crimes is 100 - 125,000 violent crimes each year.
Half of the reported violent crimes are against women = 50,000 - 60,000 (reported) violent crimes against women.
If 1% of those women have guns in their possession at the time of the attack, then we are looking at 500 to 600 (reported) violent crimes against women who have guns in their possession at the time of the crime. That's a tiny sample for any kind of serious study, isn't it?
My own conjecture is that gun ownership among women is in general positively correlated with affluence and self esteem, and negatively correlated with being an abusive relationship. It seems like, at the very least, you'd need to find women who were equally affluent and equally likely to be involved in an abusive relationship with equally violent abusers and compare the women in that group who had guns those who didn't. Starting with a sample in the low hundreds, how did Lott correlate that?
Note I am not arguing - I am really curious about the methodology involved. Since crime against women is wildly under-reported, I can't figure out how any reliable stats could be genertated.
Could you sketch out the methodology that supports the "2.5 time as likely" stat?
Thanks,
Mike
