Last week I submitted this article to The Journal of Criminology and they rejected it immediately without comment.  But I thought you would like to read it anyway so here it is.    

In 2011 Erin Richardson and David Hemenway published a painstakingly-researched article based on 2003 data comparing gun violence in OECD countries.1  One of their findings was that the U.S. gun homicide rate was nearly 20 times higher than the overall rate for other high-income countries.   While they did not explicitly link elevated gun homicides in the U.S. to the prevalence of firearms in the civilian population, their findings have been utilized by virtually every gun control advocate to justify additional gun ownership restrictions, particularly in the wake of the Sandy Hook massacre in December, 2012.The purpose of this article is to update their data with more current information, as well as to determine whether the policy strategies being advanced to diminish gun harm aligns with the relevant data on gun violence.

The chart which follows contains updated (2010) data on national population, the number of guns in civilian hands, per capita civilian gun ownership and the gun homicide rate per 100,000.

Country Population (000’s) Civilian guns (000’s) Per capita % Hom. Rate
Australia 22,065 3,250 14.7 0.11
Austria 8,389 2,500 29.8 0.18
Canada 34,126 9,950 29.1 0.5
Czech Repub. 10,519 136 1.2 0.12
Finland 5,363 2,400 44.7 0.26
France 65,031 19,000 29.2 0.22
Germany 81,776 25,000 30.5 0.2
Hungary 10,000 560 0.05 0.13
Iceland 315 90 28.5 0
Italy 60,463 7,000 11.5 0.36
Japan 127,450 710 0.005 0
Luxembourg 506 70 13.8 0.6
Netherlands 16,615 510 3 0.2
New Zealand 4,367 1,000 22.8 0.26
Norway 4,889 1,320 26.9 0.04
Portugal 10,637 2,600 24.4 0.48
Slovakia 5,430 450 8.2 0.18
Spain 46,070 4,500 9.7 0.15
Sweden 9,378 2,800 29.8 0.19
UK 62,271 4,060 6.5 0.05
TOTAL OECD 585,660 87,906 15 0.17
USA 309,326 270,000 87.2 3.58


Notwithstanding changes in some specific values, the 2010 data shows a very similar profile to what Richardson and Hemenway discovered for 2003, namely, a correlation between gun ownership and gun homicide rates on the one hand, and a continued and significant disparity between the United States and other economically-advanced countries on the other.  Gun homicide rates per 100,000 range between null for Japan and Iceland up to .48 for Portugal, with the mean of .24 or above only being experienced by countries with a per capita gun ownership of at least 1 in 5.  There were other countries (Sweden, Norway, Iceland, France) whose per capita ownership also exceeded 1 in 5, but they were still below the mean for gun homicides.  The correlation for gun homicide and median per capita gun ownership, on the other hand, does not appear to be as strong.

Both the OECD and the U.S. gun homicide rate slipped between 2003 and 2010 (15 and 13 percent respectively) but the significant gap in gun homicide between the United States and other OECD countries remained basically unchanged.  To put this differential in a somewhat more graphic context, in the 22 countries above there were 12,070 homicide victims in 2010, of whom 11,078, or 92%, lived in the United States.  This is a remarkable statistic and there is no other form of violent death in which the disparity between the United States and its OECD cohorts displays even a fraction of this difference.3

The consistency of the data from 2003 and 2010 makes it difficult to ignore the connection between gun prevalence and gun homicide in the United States.  But the data, suggestive and comprehensive as it is, does not yield the kind of information that would allow us to align it properly with strategies designed to diminish the harm caused by guns. In particular, the evidence both for the U.S. and elsewhere is either silent or unreliable on defining the type of guns that are used in felony assaults.4  We can estimate this data from FBI-Uniform Crime Reports as well as other sources, and it appears to be the case that handguns (pistols, revolvers), as opposed to long guns (shotguns, rifles) are used in perhaps 90% of gun felony crimes.

If we deduct estimates of long gun ownership from the overall total of guns circulating amongst civilians in the United States, the per capita number for U.S. gun ownership would drop from its current 87 to somewhere below 40, placing us within the “normal” boundaries of gun ownership within the OECD.  What this simple exercise affirms is that we are not the only advanced country to allow its citizens access to small arms, but we are the only country that gives equal opportunity to acquiring both long guns and hand guns. The discussion about guns and homicide should focus on the prevalence of handguns, and not on small arms in general. Strategies to curb gun violence in the U.S. by controlling access to all types of small arms do not really catch the issue which needs to be addressed.



  1.  Erin G. Richardson & David Hemenway, “Homicide, Suicide and Unintentional Firearm Fatality: Comparing the United States with other High-Income Countries, 2003,” The Journal of Trauma, Injury, Infection, and Critical Care, Volume 70, No. 1 (January 2011), 238-243.
  2. See, for example, and
  3. The U.S. auto fatality rate per 100,000 is lower than rates recorded for many OECD countries; cf.
  4. Of the OECD countries compared in this study, less than 20% provide breakdown between long guns and hand gun ownership.  According to the ATF, 2010 was the first year since records have been kept (mid-80’s) when handguns constituted more than 50% of all guns manufactured or imported into the U.S. Cf. Firearms Commerce in the United States, Exhibits 1 and 3.
  5. Reliable estimates for 2010 in: See Table 6.