Samstag, 22. Dezember 2012

Gunownership and gun-related mortality?

Sorry, for the time being I have fallen in love with evidence-based politics & policy, statistics and economy, so this might not be the last post on this topic. I've caught the disease thanks to Nate Silver, Paul Krugman & friends, Freakonomics, et al.
Broadly speaking, the topic is relevant to this blog, which is about avoiding unnecessary death and suffering - even though biomedical research is certainly more crucial than gun laws. What good is healthy life-extension if you have been shot, right?

After the Newtown killings and post-Breivik, I would like to add to the discussion something above the evidence-free back and forth we often see in the media and elsewhere. In this case I have simply taken a very superficial look at the academic literature to see what the consensus on gun issues might be.

My working hypothesis is that societal dysfunction and lax gun regulation are both contributors to the composite endpoint of gun-related homicides, suicides and other crimes like armed robbery. No surprises here. But what are the individual contributions of each? I guess both matter although:
"There is, however, the possibility that accessibility alone does not produce more lethal violence. A society could, for example, have high firearm accessibility but due to a lack of structural deprivation its citizens simply do not use firearms either against each other or on themselves." (1)

And two weaker arguments against gun control:
"Another argument from the political debate over gun control is the argument that an equal distribution of firearms has a deterrent effect (Lott 2000). If everyone is similarly armed a thoughtful attacker will ‘‘think twice’... [but this] does [not] appear to hold empirically (Donohue 2003)...
Finally there is the argument that mere competence may reduce the lethality of weapons like firearms."
I'll let the researchers speak with all quotes from (1):

Within the USA:
"Multiple studies conducted in the United States find positive associations between firearm accessibility and suicide rates...With homicide, Newton and Zimring (1969), Fisher (1976), Kleck (1979), Seitz (1972), and Lester (1988) all reported a positive association between firearm availability and rates of firearm homicide. [and so did others to some extent]"
What about tests looking at different countries in comparison?
"Cross-national studies are needed to test the generalizability of the relationship between firearms and lethal violence...[and my reading is that such studies are weakly supportive of an effect, but] Collecting cross-national data on firearm ownership and lethal violence is a difficult project and thus the literature is plagued by small sample sizes...[and is] particularly vulnerable to outliers like the United States...U.S.–Canadian comparisons [even though though statistically weak are consistent with harmful effects of guns]"
This is broadly consistent with a neat and informal analysis you can find on do.minik.us.

... but not so fast, since
"There is a robust and strong association between social structure—inequality and poverty—
and homicide... [and apparently] suicide is explained more by cultural values"

The causality may well be "Societal dysfunction -> gun use -> homicides", and not "gun use ->  (Societal dysfunction ->)  homicides", but in the end one could intervene at any of the two steps.

Due to these problems the authors set out to perform a large cross-national study and multivariate analysis to rule out confounding. They are using the 2007 Small Arms Survey and WHO data to adjust for confounders, yielding a sample of 168 nations.
[all quotes, emphasis mine:] "The ‘‘Full Model’’ regresses homicide on firearm, population, structural, and cultural variables. In this model the effects of firearm accessibility, population structure and cultural values are reduced to zero while the economic variables remain significant and strong predictors of homicide. The Full Model accounts for 45% of the variation in homicide. The inclusion of firearm accessibility, population measures and cultural values only explains 1.2% more of the variation in homicide than the structural model alone...
The FullModel explains [only!] 26% of the variation in suicide across the nations in this study...[and  there is no effect of firearm accessibility]"
... but there are other problems with guns, particularly in unequal societies. The authors suggest that:
"Firearms certainly make violent confrontations more lethal...For example, in the United States homicide rates are higher in lower income populations but gun ownership is also ubiquitous in these populations. Thus, the combination of deprivation and access to lethal weaponry may in fact produce more death by violence
...
This suggests that access to firearms only produces more violence when the population, or parts of it, experience high deprivation and indifference from government. [but the evidence is weak and/or preliminary]"
I don't want to get into strengths and weaknesses of this paper (just read it yourself), so let us continue our look at the totality of evidence.

Is the data consistent with other reviews? (randomly sampling google scholar)
"The bulk of the evidence at this point suggests more prevalent handgun ownership engenders more widespread use of guns in crime and higher social costs of crime."
(p. 33, GUN CONTROL AFTER HELLER: THREATS AND SIDESHOWS FROM A SOCIAL WELFARE PERSPECTIVE)

Most importantly the evidence from "natural experiments" (tighten gun laws, see what happens/compare gun laws across different places & mortality):
"Strong correlations have been reported between firearm ownership and firearm related suicide, homicide, assault, and unintentional deaths.[ref. 5–13]...Dramatic reductions in overall firearm related deaths and particularly suicides by firearms were achieved in the context of the implementation of strong regulatory reform"
 Ozanne-Smith et al. 2004. Firearm related deaths: the impact of regulatory reform

And another:
"Our findings provide evidence that the introduction of restrictive firearm legislation effectively reduced the rates of firearm suicide and homicide. The decline in firearm-related deaths seems to have been mediated by the legal restriction of firearm availability. Restrictive firearm legislation should be an integral part of national suicide prevention programmes in countries with high firearm suicide rates....
Furthermore, peer-reviewed publications provide vast evidence to support the view that changes in firearm legislation have influenced the rate of firearm suicides in the USA (Ludwig & Cook, 2000), Canada (Caron, 2004), Australia (Ozanne-Smith et al, 2004 [op. cit.]) and New Zealand (Beautrais et al, 2006), whereas there is only some evidence from continental Europe and Britain to suggest the same (Hawton et al, 1998; Haw et al, 2004). Other researchers have shown that different state firearm regulations across the USA have influenced suicide (Conner & Zhong, 2003) and homicide rates (Rosengart et al, 2005)."
 Kapusta et al. 2007 (2)

Back to (1):
"Our data does not support the ‘‘more guns, less crime’’ model and instead support a less deprivation, less violence model similar to Pridemore’s (2008) findings"
Given there is weak evidence from observational studies that gun access contributes to excess mortality, "vast" evidence from quasi-experimental studies and comparisons of gun regulation, a likely risk with guns in a dysfunctional society like the United States and little or no potential downside of tight regulation, the road ahead is clear.


Solve poverty and tighten gun laws. And no you ain't telling me solving poverty is hard, it pays off.


1. Small Arms Mortality: Access to Firearms and Lethal Violence
Mark Konty, Brian Schaefer
Sociological Spectrum
Vol. 32, Iss. 6, 2012

2. Br J Psychiatry. 2007 Sep;191:253-7.
Firearm legislation reform in the European Union: impact on firearm availability, firearm suicide and homicide rates in Austria.
Kapusta ND, Etzersdorfer E, Krall C, Sonneck G.
http://bjp.rcpsych.org/content/191/3/253.full

Keine Kommentare:

Kommentar veröffentlichen