Reply #76: That wasn't gejohnston's actual claim

[Euromutt]

gejohnston said:

There are no peer reviewed research done by criminologists that supports your or the mayor's thesis.

Emphasis mine.
He didn't say anything about public health researchers, be they MDs or economists (like Hemenway), nor medical journals.

And where public health literature into firearms is concerned, peer review--or possible lack thereof--isn't the issue (I tend to subscribe to David Gorski's view that, to paraphrase Churchill, peer review is the worst system yet devised to disseminate scientific research, except for all the others). What is the issue is that a sizable segment of the medical/public health community has succumbed to groupthink where firearms are concerned: they have decided that privately owned firearms are a public health menace, and eagerly produce and publish any and all research that supports that predetermined conclusion.

There are numerous problems with the "public health menace" approach. The first is that it takes an epidemiological approach, that is, it treats firearms as a pathogen. The problem with treating firearms as if they were a pathogen is that it assumes that everybody is more or less equally at risk once they're exposed. It doesn't make any difference to a virus or a carcinoma whether you have a criminal record, are a member of a criminal organization (like a drug gang) or have a tendency toward violence coupled with poor impulse control. It most assuredly does affect the likelihood that you will handle a firearm irresponsibly, or become the target of an assault with a firearm (unlawful or otherwise). In fact, the epidemiological research data bears this out, but the researchers consistently try to sweep it under the rug with the unconvincing claim that they "controlled for other variables," and avoid mentioning it in their conclusions, abstracts and press releases.

Case in point: "Investigating the Link Between Gun Possession and Gun Assault" (Branas et al. American Journal of Public Health 2009 http://ajph.aphapublications.org/cgi/content/abstract/AJPH.2008.143099v1 ). The first warning sign was this was "science by press release", with the press release preceding print publication of the article (http://www.uphs.upenn.edu/news/News_Releases/2009/09/gun-possession-safety). From the press release:

The study estimated that people with a gun were 4.5 times more likely to be shot in an assault than those not possessing a gun.
<...>
Penn researchers investigated the link between being shot in an assault and a person’s possession of a gun at the time of the shooting. As identified by police and medical examiners, they randomly selected 677 cases of Philadelphia residents who were shot in an assault from 2003 to 2006. Six percent of these cases were in possession of a gun (such as in a holster, pocket, waistband, or vehicle) when they were shot.

You have to love econometric modeling; how else could you take a study population of shooting victims in which the ratio of non-carriers to carriers is almost 16 to 1, and come up with a result that carrying a firearm makes you over four times as likely to be shot than if you weren't carrying one?

This is the same approach that epidemiologists have historically used to establish links between such things as smoking and lung cancer or drinking and car crashes.

Yeah... Thing is, at the time Sir Richard Doll did his research, something in the order of 90% of lung cancer patients were smokers; not 6%. The hypothesis that a particular behavior is a major causal factor in a particular outcome is a lot more plausible when 9 out of 10 people who suffer that outcome engaged in that behavior than 1 out of 17.

As for establishing a link between drinking and car crashes, well, drinking by itself doesn't cause car crashes, does it? That is, you can drink enough to give yourself cirrhosis of the liver, but if you never operate a motor vehicle, you're never going to cause a car crash. It's the combination of drinking and operating a motor vehicle that is the causal factor.

And what was this "same approach"? According to the press release:

These shooting cases were matched to Philadelphia residents who acted as the study’s controls. To identify the controls, trained phone canvassers called random Philadelphians soon after a reported shooting and asked about their possession of a gun at the time of the shooting. These random Philadelphians had not been shot and had nothing to do with the shooting. This is the same approach that epidemiologists have historically used to establish links between such things as smoking and lung cancer or drinking and car crashes.

It's illegal to make unsolicited calls to a cell phone, so they had to be calling land lines.
According to the study itself:

However, compared with control participants, shooting case participants were significantly more often Hispanic, more frequently working in high-risk occupations, less educated, and had a greater frequency of prior arrest. At the time of shooting, case participants were also significantly more often involved with alcohol and drugs, outdoors, and closer to areas where more Blacks, Hispanics, and unemployed individuals resided. Case participants were also more likely to be located in areas with less income and more illicit drug trafficking (Table 1).

Which just about describes the demographic of any large city least likely to have a land line telephone. So you have a study group that contains significant numbers of individuals who are poor, members of ethnic minorities, and/or have prior histories of criminal behavior and substance abuse; who do you select as your control group? Why, people who are sufficiently settled and have enough money to have a land line telephone! Rather than try to exclude the possible confounders going in, we'll just "adjust for the variables" afterward (or at least say we did).

So might the possibility exist that, insofar as there is a causal relationship between carrying a firearm and being shot, it is that people who perceive themselves to be at greater risk of getting shot (particularly those engaged in criminal activity) are more likely to carry guns as a result? That possibility should at least merit investigation. However, according to the study itself, the researchers "did not account for the potential of reverse causation between gun possession and gun assault."
I'll repeat and emphasize that: the researchers "did not account for the potential of reverse causation between gun possession and gun assault."
Note that neither the abstract, let alone the press release, mention this fact; you have to actually dig through the article itself (that'll be $30, please) to find it.

Researchers in the medical and public health fields have an ugly tendency to imply that causation exists when all they have at best is correlation; they've even developed a whole vocabulary to serve this function: terms like "linked to," "risk factor" and "associated with." To see how this is put into practice, let's turn to "Guns in the Home and Risk of a Violent Death in the Home: Findings from a National Study" (Dahlberg et al. American Journal of Epidemiology 2004 http://aje.oxfordjournals.org/cgi/content/full/160/10/929#KWH309C20 ).
The abstract starts:

Data from a US mortality follow-back survey were analyzed to determine whether having a firearm in the home increases the risk of a violent death in the home <...>

So the authors are seeking to establish a causal relationship: is the risk of violent death increased by having a firearm in the household? However, what they conclude with is:

Results show that <...> having a gun in the home was associated with an increased risk of firearm homicide and firearm suicide in the home.

Note the phrase "associated with"; i.e. the authors established a correlation. That's all "associated with" means; having insulin in the house is "associated with" having diabetes. But correlation does not imply causation. Because the authors fail to acknowledge this difference between their stated objectives and their actual findings, they are in effect pulling a bait-and-switch. The same goes for the UPenn study, the one in which the researchers "did not account for the potential of reverse causation between gun possession and gun assault." In fact, it goes for pretty much the entirely of public health research concerning firearms.

To conclude, I'll quote an excerpt from a review looking at international comparisons of suicide rates.

A problem with international studies is the difficulty in fully accounting for the disparate cultural factors that may influence the incidence and method of suicide. The few international studies that address the gun-suicide question suggest that firearm availability affects the method of suicide and may have an influence on the total level of suicides, especially among youth. The evidence, however, is far from convincing that gun ownership levels are related to overall suicide rates for all age groups. The U.S., for example, has the highest levels of gun ownership, but its overall suicide rate is only 16th out of 26 high-income countries. One study found a statistically significant relationship between gun ownership levels and suicide rate across 14 developed nations (e.g. where survey data on gun ownership levels were available), but the association lost its statistical significance when additional countries were included.

Emphases mine. The article is "Firearm Prevalence and the Risk of Suicide: A Review" (http://www.hcs.harvard.edu/~epihc/currentissue/Fall2001/miller.htm ) by Matthew Miller and David Hemenway. Yes, the same David Hemenway.

Curiously, the "Guns and Death" page of the Harvard Injury Control Research Center sums up the findings of the very same article as:

The evidence that gun availability increases the suicide rates of adults is credible, but is currently less compelling.

Isn't it... odd how the evidence went from being "far from convincing" in the article itself, to "credible, but currently less compelling" in Hemenway's own, non-peer reviewed description of the article?