Citation

Kleck GD. Homicide Stud. 2001; 5(1): 64-77.

Copyright

(Copyright © 2001, SAGE Publishing)

DOI

10.1177/1088767901005001005

PMID

unavailable

Abstract

Using a case-control design comparing homicide victims with matched nonvictims, Kellermann et al. (New Engl J Med, 1993) concluded that keeping a gun in one's home increased the risk of being murdered by a factor of 2.7. The authors' underlying assumption was that a significant elevation in homicide risk derived from the risk of being murdered with a gun kept in the victim's home. This article shows that homicides are rarely committed with guns belonging to members of the victim's home and that such killings could be responsible for no more than a 2.4% increase in the relative risk of being murdered. Guns in one's own home have little to do with homicide risk. Scholars need to attend more closely to the mechanisms by which an alleged causal effect is supposed to operate and to consider their plausibility before concluding that an association reflects a causal effect.