John Wayne with rifle

Everyone has been talking, sensibly or not, about guns since the massacre at Sandy Hook Elementary School in December. I had not waded in for a few reasons. Many experts are writing about the topic in the press; I have no particular expertise on the question of whether guns cost or save more lives; and the research literature on the subject is a morass. Moreover, studies of gun violence have been limited by federal legislation explicitly restricting health researchers from addressing the question.[1] But I’ll take this post to make a few partly-informed observations.

The morass

John Wayne with rifle

Opinions over the proper role of guns are so freighted – after all, it is a matter of life and death – that the research is often driven by the desired conclusion. American historians suffered embarrassment in a notorious case about a decade ago. They had commended and awarded a major prize to a book demonstrating that private gun ownership was rare in early America, a finding that implied that the Second Amendment had little to do with individual rights. Other scholars and a panel of inquiry discovered that the book was based on flawed, probably fraudulent, evidence. Its award was withdrawn; the author resigned his faculty position.[2] (Around the time the Constitution was written, probably a bit more than half of American households had guns.[3]) I am sure that distorted, conclusion-driven research has appeared, consciously or not, on all sides of  the gun violence debate.

Technical problems afflict statistical studies on the topic. It is difficult to disentangle cause and effect in the data. For example, communities with many guns tend to have high rates of violence. Does gun ownership promote violence or does violence promote gun ownership? Or does something else about high-gun high-violence communities – say, some features of their histories – drive both? As another example, it is hard to treat communities as really distinct cases for analysis. For instance, how can we understand the effects of gun control laws when anybody can drive any number of guns from a low-control state into a high-control state? It is a research morass.

Tentative conclusions

So, the evidence is not simple. Still, a few points seem likely to be true. If one looks cross-nationally, guns per se do not seem to drive lethal violence. For example, gun ownership is common in Switzerland, but the homicide rate there is quite low. [4] Plausibly, it is the combination of easy access to guns and a certain social context — perhaps a high level of economic deprivation, or perhaps an “honor culture” which insists that men must respond forcefully to challenges [5] — rather than guns alone that creates a lethal brew. [6]

Adding together gun deaths by crime, by suicide, and by accident, having many guns around probably raises the total death rate. For example, teens with guns in their homes are likelier than otherwise similar teens without guns at home to think about and to attempt suicide. [7]

Finally, it seems hard to argue that making military weapons with huge magazines easily accessible to the general public serves any good – unless you think that there is a reasonable chance that people will have to defend their homes against large-scale attacks. Clearly, many Americans believe just that:

A free people should be an armed people. It insures against the tyranny of the government. If they know that the biggest army is the American people, then you don’t have the tyranny that came from King George. That is why it [the 2nd Amendment] was put there.  — Congressman Louie Gohmert (R-TX).


Michael Hout sent me some sociological facts about gun ownership in the U.S., drawn from the General Social Survey. Other things being equal:

  • Men are more likely to report having a gun in the home than are women — although women who are self-employed have distinctively higher rates of ownership than other women.
  • The more money a family makes, the likelier it is to have a gun in the home.
  • Households outside metropolitan areas and especially those in rural places are likeliest to have guns.
  • Rates of gun ownership have dropped over the last 40 years. In the 1970s, about 50% of American homes had guns; in the 2000s, about 35% have.
  • Unfortunately, data such as these do not tell us much about military-style assault weapons.


    [1] Kellerman and Rivera, “Silencing the Science,” JAMA, 2012 (doi:10.1001/jama.2012.208207).

    [2] You can read about it in Wikipedia: .

    [3] Roth, “Guns,” Social Science History, 2002.

    [4] An old but solid study is Clinard, Cities with Little Crime, 1978.

    [5] E.g., Nisbett and Cohen, Culture of Honor, 1996.

    [6] E.g., Dixon and Lizotte, “Gun Ownership and the ‘Southern Subculture of Violence,’” American Journal of Sociology, 1987.

    [7] Bearman and Moody, “Suicide….,” American Journal of Public Health, 2004. See also Cutler, et al, “Explaining the Rise in Youth Suicide,” WP # W7713, National Bureau of Economic Research, 2000.

    Cross-posted from Claude Fischer’s blog, Made in America: Notes on American life from American history.