Gun Culture May Contribute to Suicide Rate in Rural America

[John718]

79. not a 'Harvard study', by two men who have nothing to do with Harvard

I read this paper after a pro-gun friend of mine said I should read it...so I did, and followed up many of the sources (I wonder how many people on this thread have done that?). DanTex is correct in saying that this is not a 'Harvard study' in any meaningful sense; it was clearly not peer reviewed, and neither of the writers had or have anything to do with Harvard. Both work for libertarian-leaning 'think tanks': Mauser for the Fraser Institute, and Kates for the Independent Institute. Both of these groups have also published papers 'questioning' the effects of second hand smoke (funded by tobacco companies) and human-driven climate change (funded by oil, lumber and energy companies), and are far from non-partisan. If anyone wanted to read research by actual Harvard professors, there are plenty of papers on the Harvard School of Public Health's website: http://www.hsph.harvard.edu/hicrc/ .. most of which generally directly contradict Kates and Mauser's 'research'.


Even on my first reading, I found several easily verified half-truths, cherry picked references, shoddy comparisons, and glaring inconsistencies in this paper. The stat about Luxembourg's murder rate of 9.01 for 2002 jumped out right away to me as an obvious anomaly, if nothing else. Kates and Mauser refer to this figure in two tables and a few times in the text, so it was clearly an important point for them. So I downloaded the Canadian Centre for Justice Statistics papers cited in the footnote (the only reference for this figure) from 1997-2004. This annual report is really only about Canadian homicide statistics, and each year gives one small chart of other countries. In all the years I looked into, Luxembourg is only listed in one year, 2002, and indeed there is the 9.01 figure.

One would now assume that any competent and honest researcher would see that this was, at the very least, an unusually high figure for a country like Luxembourg. One might consider checking this number against another source for global homicide data, like say, the UN, which collects and publishes crime data for most of the world's countries annually. The UN Office on Drugs and Crime report for 2002 gives Luxembourg a murder rate of .9 per 100,000 residents. This figure is fully consistent with the preceding and following years for Luxembourg and with other available data sources, and is also consistent with the general low levels of homicide in Western Europe. This also supports what DanTex suggests above, that the number referenced by Kates and Mauser may be just a decimal point error in the JURISTAT table.

Again, one would hope that any researcher would follow up on this obviously odd figure, then find conflicting figures that are consistent with the country's recent history and it's region, and would then not use this statistic! Instead, Kates and Mauser highlight this number several times, and, like DanTex says, really base a substantial part of their argument in those sections on this 'fact'. Kates and Mauser even offer a tepid preemptive concession hidden in the footnotes (a tactic they use frequently throughout the paper), saying that the cited Canadian report "gives no explanation of why it selects the various nations whose homicide statistics it covers. Also without explanation, the nations covered differ from year to year".

To any critical reader, the quoted comment from Kates and Mauser above, and an examination of the source itself and other sources for related data, immediately begs the question, "So why use these stats, when consistent international homicide figures were and are readily and easily available?" A further look at the tables which use the Luxembourg figure shows even more inconsistencies. The only reference given for Table 2, which compares murder rates and handgun ownership rates, gives only the Canadian report as a reference. But nowhere in those cited reports, in any of the years cited, is any homicide stat given for Belarus. So, where did this data come from? A figure like this without any reference (or even worse, an incorrect reference) would never make it past any real peer review.

... and it goes on from there...Kates and Mauser rely heavily on comparisons to Russia in these tables, which they call an industrialized nation in the first pages, as if Russia (particularly in 2002) is ever considered a developed economy on par with Western Europe, the US, Canada, Australia, Japan by any typical standard (IMF/ UN/ World Bank reports, GDP per capita etc), not to mention the massive upheaval in Russia and the region during the post-Soviet era of the late 80s and 90s. They say handguns are 'allowed' in Poland, without any sort of explanation or details as to what that means exactly....while in reality Poland before and after this per have had quite tight gun regulations in general, and tighter regulations on handguns. The whole premise of this table, which lists handguns as either 'banned' or 'allowed' without any decent details of what the actual laws are in those countries, is a typical example of the broad stroke, no-depth, skewed and misleading comparisons used throughout Kates and Mauser's paper. They leave out comparisons to the US whenever convenient, such as in this table, which would obviously tilt the data against the authors' intended message. There are more examples I can illustrate if you like...