I would like to thank @simonr916 for raising some interesting challenges to my previous post on gun control.
To address the points raised:
“My main contention with your blog is your argument that there is no correlation between gun ownership and homicide rates. You take your analysis here from another blogger who gets their data from a range of sources including Wikipedia pages of disputed accuracy and pro-gun websites.”
If the data is wrong, then point out where it is wrong. It is not wrong simply because it appears in Wikipedia!
There are two sets of data presented in the blog:
International Data
In this case the data appearing in Wikepedia comes from very respected sources:
The primary sources for the Data On Intentional Homicide Rates By Country are the United Nations Global Study on Homicide and Homicide Statistics 2012.
The primary source for the Wikepedia data on Number of Guns per Capita is the Geneva Graduate Institute of International Studies. “Small Arms Survey 2007″ and Small Arms Survey 2007 Part 2
Neither of these sources are “of disputed accuracy”, nor are they “pro-gun websites”
USA State By State Data
The data on homicides and populations by state are sourced from the FBI Crime in the United States 2011
The data on gun ownership by state are sourced from About.com, the primary source is the 2001 Behavioral Risk Factor Surveillance System (BRFSS). The numbers as reported in the Washington Post can be seen here.
Neither of these sources are of disputed accuracy, nor are they “pro-gun websites”, although the pro-gun website USCarry.com did have the temerity to quote the report!
On this point, you are simply wrong, the data is from reputable sources.
“This analysis makes no attempt to adjust for socio-economic
factors such as population density, poverty and inequality, levels of
corruption within government and law enforcement etc.”
Unless you have meaningful data any attempt to control for other variables is simply adding unproven assumptions and error under the guise of increasing accuracy. In relation to the factors you quote:
Population density – This is not uniform over a state or a country. One state could have x,000 people per square mile packed into three large cities, another state could have the same density but spread across many rural communities. A state wide population density is meaningless.
Poverty and inequality are also not uniform over a state or a country. Absolute poverty? Poverty relative to the median in the state? The median in the country?, the mean? Inequality, the same questions, compared to the state, the country, the ethnic group?
Corruption. Measured in what units and with what degree of accuracy.
Yes of course homicides are caused by a multitude of factors, but without at least some correlation between the two variables you are interested in there is nothing to investigate!
“This article from the peer-reviewed American Journal of Public Health (http://www.ncbi.nlm.nih.gov/pm… looked at gun ownership and homicide rates and homicide across the US, made adjustments for socio-economic factors, and concludes that there is a correlation between gun ownership and homicide.”
The article includes no source data that can be examined, no regression equations and no indication of what or how other variables were corrected for. It does not use state level gun ownership data, but a metric it calls FS/S based on the ratio of suicides using guns to total suicides. Its comparison of arbitrary subsets of top and bottom data is not a valid statistical technique.
It is always possible to torture data into telling you whatever you want to hear by the use of undefined control variables, invalid assumptions and cherry picking elements of the data set.
As can be seen at the end of the paper, this research was funded by the anti-gun Joyce Foundation the anti-gun Robert Woods Johnson Foundation and George Soros’ anti-gun Open Society Institute, so your appeal to authority is even further undermined.
Your faith in peer review is touching but misplaced, as this shows.
“Summaries of other, scholarly articles with similar findings can be found here:
http://www.nejm.org/doi/full/1…
http://www.hsph.harvard.edu/hi…“
I will not go through every paper, but I will point out that most of the articles on the second link are by the same researcher who carried out the anti-gun lobby funded research in the first article you quoted. By way of balance, I will also point out some alternative scholarly articles that support the analysis in my original blog post.
Harvard Journal of Law & Public Policy
Journal of Research in Crime and Delinquency
It is clear that torturing the data can make it confess to whatever the relevant paymaster, special interest group wants. Most researchers in the field have a view that they are trying to support. So it is very difficult to know who you can trust.
However, a little lateral thinking will enable us to get to the truth:
If Gun Ownership were a significant factor in homicide rates then scholars investigating the relationship of homicide to other things, such as poverty, would need to control for it in their research. These scholars probably have no political interest in the gun control debate and will only control for variables that have a real influence.
This paper from the British Journal of Criminology reviews 47 cross national studies of the relationship between poverty and homicides. Conveniently for our purposes it also lists the variables that the researchers controlled for in each study.
Not one of them controlled for gun ownership!
If scholars without a vested interest don’t regard gun ownership as a variable worth controlling for in their own field of research, I think it is safe to assume it is not important.
“This article shows that gun controls introduced in Australia in 1996 were followed by
accelerated declines in firearm deaths:
http://www.ncbi.nlm.nih.gov/pm…“
Yet, this paper concludes the opposite, that it had no impact:
http://www.ssaa.org.au/capital-news/2008/2008-09-04_melbourne-uni-paper-Aust-gun-buyback.pdf
“You also make the argument that gun control will not prevent criminals from obtaining guns, and you draw a comparison with illegal drugs. However illegal guns are, almost without exception, legal guns that have become illegal either by being stolen from their lawful owner, by being transported across a border, or by a decommissioned weapon – sold as a replica – being re-commissioned.
As such, reducing the availability of legal weapons would have the result of decreasing the availability of illegal weapons. This reduced supply will not eradicate
availability altogether but would keep guns out of the hands of many criminals,
and would increase the cost to others.”
All drugs are illegal and still there is ample supply. I see no reason why making guns illegal would work any better than making drugs illegal. Certainly the existing supply of legal guns that suddenly become illegal would not disappear from existence and unless every country in the world made guns illegal there would be plenty of legal guns and replicas that could be imported.
“You argue availability of guns in the UK based upon a single newspaper article quoting ananonymous source ‘close to the trade in illegal weapons’. The reliability of this as evidence to support your case is highly debateable”
There is no register of illegal gun purchases and no hard data that I could find. However, the single newspaper article is certainly not alone in reaching that conclusion:
http://www.channel4.com/programmes/dispatches/articles/guns-in-britain
http://news.bbc.co.uk/1/hi/england/6180559.stm
http://www.express.co.uk/posts/view/347592/The-guns-and-grenades-of-gangland-Britain
“but, in any case, it is unquestionably more difficult to obtain an illegal weapon in the UK than it is to walk into Wal-Mart and buy one in the US. Whilst it is likely that hardened criminals with the right connections will always be able to access weapons, controls limit their availability and ensure that, by doing so, they face the risk of prosecution and imprisonment even if they do not use the weapon.”
That is certainly true, but it ignores the defensive value of legal gun ownership. The criminals may be slightly hindered in their ability to acquire guns, but the law abiding victims would be totally prevented from doing so.
“In the case of school shootings, where the killer is usually an isolated and troubled
teenager, such controls are likely to put guns completely beyond their reach.”
For the tiny minority of cases where a law abiding citizen with no criminal connections decides to commit mass murder then it would certainly make it more difficult to use a gun. However, There is a global phenomena of mass murder using motor vehicles as the weapon these would still be available to the school killer intent on mass murder.
“Your argument that school killers’ motive is murder/suicide and so they may choose to use a bomb instead, makes an assumption as to the killers’ motives without evidence. Other possible motives may be the feelings of power and control they expect to
experience, seeing others in fear, the glamourisation of gun use and so on.”
Many motives are possible, they will probably be different in each case. However the actions usually involve the mass killing of school pupils followed by the death of the perpetrator. Such actions can also be (and have been) carried out via homemade explosives or using a motor vehicle, and it is pure speculation that only the use of guns would satisfy the motivations.
“Following the Dunblane massacre in 1996, gun laws in the UK were tightened. There have been no mass school shootings in the UK since, and – notably – no school bombings either.”
That is a very weak argument, Dunblane is the only mass school shooting I can find in the UK. In the many decades from the invention of the first fire-arm up to to the ban in 1996 there was 1 incident. In the 16 years since there have been none. There were tens of 16 year periods without incident before the ban. If would require hundreds, possibly thousands more years without incident, before any valid conclusion could be drawn.
“You go on to argue that gun ownership provides protection against violence. This ignores statistical evidence that guns kept in homes are more likely to be involved in accidental or criminal shootings than self defence – and more likely still to be used in
the suicide of the gun owner or another family member (http://www.ncbi.nlm.nih.gov/pu….”
The paper deals only with guns that were fired. It says that for each time a gun was fired in self defence there were four unintentional shootings and seven criminal assaults or homicides.
What it fails to include are all the times a burgler, mugger or would be rapist was scared off by a victim showing a gun that didn’t need to be fired. Only in the, presumably quite rare, event of the assailant continuing their attack once the gun was produced would the gun need to be fired.
For some papers that look at the whole picture and conclude that significant self defence usage and value is provided, refer to:
Journal of Criminal Law and Criminology, vol. 86, issue 1, 1995.
St. Louis University Public Law Review
“And yes, some criminals may be discouraged from some crimes if they believe the intended victim may be armed. It seems likely that they will simply move onto another target instead.”
In a society with a culture of gun ownership, who do they move on to? Each potential target could be armed and capable of defending themselves.
“Equally, it may be that fear of an armed victim may encourage a
criminal to carry a gun or use a greater level of violence than would otherwise
be the case.”
Not every burglar, mugger or rapist is happy with moving up to the status of murderer. Just as likely they would decide that the payoffs from a life of crime had changed and perhaps some other form of occupation would become more attractive.
Some interesting challenges from @simonr916 but nothing that causes me to change my position


