As a follow up to my recent post on defensive gun uses (DGUs), I think it’s appropriate to discuss international comparisons sometimes cited in support of the anti-gun rights agenda.This was prompted by correspondence from a fellow blogger, to whom I’ll refer as HH, who followed up with a post featuring some international data. I respect HH’s effort to collect the data and to present it with some eloquence, and with a little less rancor than the original correspondence. Nevertheless, the international comparisons are not as straightforward as HH would like to believe.

Let me state at the outset that I am not a big “gun guy”. I support individual liberty and a minimal state apparatus in general, along with gun rights, but I am not affiliated in any way with the NRA or any other pro-gun organization. As I told my well-armed older brother, he would not be impressed with my weaponry. I still keep a nasty, old fireplace iron under my bed. And I have a few rocks in my backyard.

HH believes that the high U.S. homicide rate relative to the handful of other developed countries he mentions (along with India) proves that “gun control works”. I differ for several reasons discussed below.

Causality and Gun Control: HH’s conclusion brings into focus two different aspects of the gun control question. The first is whether a change to more restrictive gun control leads to a reduction in homicides. That is not as obvious an outcome as HH thinks. For example, a gun ban cannot eliminate all guns, especially within limited jurisdictions. (Perhaps the federalist approach is partly why HH considers our gun laws “a mess”, but federalism is a feature of our system, not a bug, not least if it discourages local politicians from enacting ineffective rules.) Black market traffic in guns is likely to be sufficiently profitable to justify the legal risks in the presence of a ban. And the empirical evidence as to whether more stringent gun control reduces homicides is mixed at best (see here, here, here and here).

The empirical evidence presented by HH is not related to changes in gun laws (except for one or two suspect assertions about mass shootings). Instead, cross-country comparisons of homicide rates are given along with a single correlate: “gun laws”. The one data point driving the presumed direction of causality is the U.S., which has lenient gun laws and a high homicide rate relative to the four other countries (five if we include the U.K., from whence HH hails). The comparisons are made with no controls for the history of gun rights and ownership, demographics, other prohibitions, or any other confounding influences. For HH, it’s all because of guns.

Mass Shootings: HH spends some of the post discussing this phenomenon, which is rare albeit horrifying. Mass shootings account for very few of U.S. homicides, and there has been no discernible upward trend in the U.S. (see here, here and here). Moreover, multiple victim shootings are just as common in Europe as they are in the U.S. They usually prompt calls for bans on arbitrarily-defined “assault weapons”, but the bans do little to prevent such tragedies.

Historical Background: Guns owned by private individuals played an important role in the American revolution. In fact, early British attempts to confiscate weapons led to an increase in the hostilities leading up to the war. The Second Amendment of the U.S. Constitution was intended to protect individual gun rights and to protect the nation from future tyrants.

The homicide rate has declined steadily in the U.S. over the past three hundred years, from estimates of more than 30 per 100,000 people in the early 1700s to less than five today. A similar pattern occurred in other parts of the world, but after 1850, the decline in the U.S. failed to keep pace with declines in Europe.

Private guns were integral to westward expansion in the U.S. Leaving aside the tragic consequences for Native Americans, the scramble for resources and the under-developed legal system in the west undoubtedly contributed to homicides. At the same time, the need of settlers to defend life and property in an insecure environment made gun ownership (and DGUs) a necessity. This history and the generally high value placed by Americans on individual rights set the tone for today’s generally permissive attitude toward gun ownership in the U.S.

Alcohol, Drug Prohibition and Homicide: The temporary lows in the homicide rate prior to the 1910s “may have been illusory“, according to this abstract, because many homicides were reported as accidents in that time frame. More accurate reporting created the impression of a rising homicide rate during the 1910s. Alcohol prohibition began in 1920 and contributed to an increase in U.S. homicides until after repeal. Likewise, later in the twentieth century, the drug war, together with a bulge in the youth population, contributed to an even larger increase in the homicide rate. It is interesting that this increase was accompanied by an apparent decrease in the rate of spousal homicide. (A curious aside: one analyst has noted the strong correlation between homicide rates in the U.S. and fluctuations in the use of lead-based paints and leaded gasoline.)

Illegal drugs are just one area of black market activity in which the U.S. is a world leader. The connection between heavier underworld and gang activity and prevalent restrictions on victimless, individual behavior, on the one hand, and homicide rates on the other, helps explain the elevated U.S. homicide rate. The existence of this link is supported by an extremely strong concentration of homicides within specific social networks.

Demographics: The interaction of legal restrictions on behavior and weak economic circumstances is undoubtedly a factor contributing to high homicide rates. It is striking that U.S. homicides are so heavily concentrated within the African American community. The relative lack of legal economic opportunities within the African American community may be connected to greater illegal trade and homicides. Homicide rates are also somewhat elevated among U.S. Hispanics and Native Americans. Among the White and Asian segments of the U.S. population, homicide rates are comparable to those of Europe (and well under India’s rate).

Suicides: My antipathy for anti-gun arguments is probably softest with respect to gun suicides. Guns are certainly “weapons of convenience”, easily transported, fast and highly effective. Within the U.S., there is some evidence that gun ownership and total suicides are positively correlated, despite a negative correlation with non-gun suicides. However, total suicide rates in the U.S. and U.K. are similar. The rates in France and especially Japan are higher, while the rates in Denmark and India are lower. Moreover, suicide is symptomatic of larger social problems that have little to do with gun rights. Our inability as a society to deal effectively with mental health issues probably has much more to do with suicide and homicide rates than gun ownership.

Summary: There are many reasons to discount international comparisons of homicide rates and regulation of firearms. The comparisons often neglect measurement issues, but more importantly, strong conclusions about the efficacy of gun control from such top-line comparisons are often drawn without carefully addressing the question of causality between changes in gun laws and changes in homicide rates. The comparisons also fail to consider variations in the larger historical and legal context within which gun ownership occurs. For a large society like the U.S., there are vast differences in sub-groups that usually reflect other social problems, some of which are created by intrusive government itself.

I close below with some thoughts on HH’s criticism of my original post on DGUs.

DGU Denialism: HH’s objections to my post on DGUs were based on a belief that I:  1) quoted misleading statistics on gun violence in the U.S.; 2) engaged in scaremongering (apparently by quoting a wide range of estimates of DGUs); and 3) used a headline (“When Government Prohibits Self-Defense”) demonstrating a wildly paranoid view of the intent of the U.S. government.

The statistics on gun violence I cited in that post came from the U.S. Department of Justice and The Law Center To Prevent Gun Violence, which are hardly representative of the gun lobby. By providing information on gun homicides, suicides, accidents and nonfatal wounds presented in emergency rooms, I was seeking to provide a fairly comprehensive list of the “downsides” of guns in the U.S. I thought that was only fair as a way to lend perspective on estimates of DGUs. The statistics on gun violence vary from year-to-year, of course, and even the homicide numbers vary across different “official” sources for a given year (the example given at the link is total homicides). For these reasons, my initial intent was to quote ranges. However, not all of the data were available over multiple years from my original sources. Some of the figures were simply DOJ “estimates”. And apparently, my searches did not turn up the most recent data available (most of the figures I quoted were either 2010 or from 2005 – 2010). Well, mea culpa, mea culpa. My range for gun homicides of 10-12 thousand per annum was off, according to HH: it was actually 9 thousand! So, my range should have been broader in view of the continuing decline in gun homicides in the U.S., but I’m heartened to know that they were lower than I thought.

As for DGU’s, it is undeniable that they are a real phenomenon, though HH seems apoplectic that anyone would dare to discuss them. They obviously happen, though no one claims “there is always a good guy with a gun“. In fact, homicide statistics often exclude deaths from DGU’s and police shootings. (In the U.K., apparently one has to be found guilty of a murder for it to be counted as a homicide.)

Since any proposal to limit firearms would be more successful in disarming the law-abiding population than miscreants, it is reasonable to ask whether DGUs would decline more than non-justifiable homicides. Moreover, the low end of the range of DGU estimates I quoted came from DGU skeptics. In any case, I don’t think the following statements qualify me as a “scaremonger”:

“Estimates range from under 100 thousand per year to more than 2.5 million. There are reasons to doubt both of the extremes. … Given this range of estimates, it would be conservative to hedge toward the lower end. ”

Finally, the headline: Now, I like a punchy headline, and I’ll bet HH does too. I also believe that the ultimate goal of the statist anti-gun lobby is to outlaw private firearms. Again, such a policy would have the largest impact on gun possession among the law-abiding population; the headline was meant to convey the consequences of doing so.