Some Methodological Problems in "Gun Ownership and Homicide in the Home - (Kellermann et. al., New England Journal of Medicine. Oct 7, 1993)

H. Taylor Buckner (26 September 1994)  E-Mail: taylor@buckner.cc

Associate Professor of Sociology, Concordia University

Kellermann and his colleagues concluded that a person who had a gun in his or her home was 2.7 times more likely to be a victim of homicide than someone who did not (1087). They further "found no evidence of a protective benefit from gun ownership...(1087)." Neither of these conclusions are justified on the basis of their research.

They conducted their research by limiting their cases to people murdered in their own homes, thus excluding any instances where intruders were killed by the homeowner. They did not ask the victim's proxy (from whom they got their data about the victim and his or her household) whether or not the victim had previously defended him/herself with a gun. Thus their conclusion that a gun provides no protective benefit flows from their failure to consider cases where it might have.

In order to provide a control group they selected another person from the neighborhood of the same sex, race, and age group as the victim, and asked them the same questions they had asked the victim's proxy. While matched on the demographic variables, the control group was stunningly different on behavioral measures. Compared to the control group the victim group was more likely to: rent rather than own, live alone, drink alcoholic beverages, have problems in the household because of drinking, have trouble at work because of drinking, be hospitalized because of drinking, use illicit drugs, have physical fights in the home during drinking, have a household member hit or hurt in a fight in the home, have a household member require medical attention because of a fight in the home, have a household member involved in a physical fight outside the home, have any household member arrested, and be arrested themselves. Thus the victim group and the control group had very different lifestyles, with the victim group living a very high-risk lifestyle.

In fact, Kellermann found that having a gun in the home ranked fifth out of six risk factors in the victims' lives. Using illicit drugs lead to a 5.7 times risk of being murdered, being a renter 4.4 times, having any household member hit or hurt in a fight 4.4 times, living alone 3.7 times, guns in the household 2.7 times, and a household member being arrested, 2.5 times.

The entire "gun in the home is risky" analysis depends on one crucial figure, the percent of the control group (35.8%) that have guns in their homes. If this figure is underreported, the findings are false. There is good reason to assume this figure is low. First, many, many surveys report that around 48% of Americans have guns in their homes. The victim group, as reported by their proxies, had 45.4% gun owners. This figure is unlikely to be false: the victim is dead, in 49.9% of the cases by gunshot wound, the proxy cannot really lie about it. The control group actually reported owning more rifles and shotguns than the victim group, (fewer handguns) but they may well have not reported having a pistol in the home because it is illegal, or nobody's business. There is considerable evidence (Kleck, 457; Newton, 6; Erskine, 456; Kennett, 253; Stinchcombe, 115) that over the past three decades an increasing fraction of survey respondents have incorrectly denied gun ownership. In a pilot study, Kellermann tested whether people reported their gun ownership correctly by asking REGISTERED gun owners if they owned a gun, and found that their replies were "generally valid (1089)." The official records of the members of the control group were not checked to see if they had "registered" guns, and likely many had guns that were not registered and not reported. Thus even the finding that gun ownership is the fifth most important risk factor is subject to considerable doubt.

My conclusion is that this study is seriously flawed, and that the conclusions were driven by ideology more than research.

-------------------------------------------------------------------------------

References

Erskine, Hazel. 1972. "The Polls: Gun Control." Public Opinion Quarterly.

36:455- 469.

Kellermann, Arthur L., M.D., M.P.H., Frederick P. Rivara, M.D., M.P.H., Norman B. Rushforth, Ph.D., Joyce G. Banton, M.S., Donald T. Reay, M.D., Jerry T. Francisco, M.D., Ana B. Locci, Ph.D., Jancice Prodzinzki, B.A., Bela B. Hackman, M.D., and Grant Somes, Ph.D. 1993 "Gun Ownership as a Risk Factor for Homicide in the Home." New England Journal of Medicine. 329:15 1084-1091. 7 Oct.

Kennett, Lee and James LaVerne Anderson. 1975. The Gun in America: The Origins of a National Dilemma. Westport, Conn.: Greenwood Press.

Kleck, Gary. 1991. Point Blank: Guns and Violence in America. New York: Aldine de Gruyter.

Newton, George D., and Frank Zimring. 1969. Firearms and Violence in American Life. A Staff Report to the National Commission on the Causes and Prevention of Violence. Washington, D.C.; U.S. Government Printing Office.

Stinchcombe, Arthur, Rebecca Adams, Carol A. Heimer, Kim Lane Schepple, Tom W. Smith, and D. Garth Taylor. 1980. Crime and Punishment - Changing Attitudes in America. San Francisco: Jossey-Bass.

| Back to Top of Article |