摘要：A highly publicized recent study by Lott and Mustard concludes that laws easing restrictions on licenses for carrying concealed firearms in public substantially reduce violent crime. Several serious flaws in the study render the authors' conclusions insupportable. These flaws include misclassification of gun-carrying laws, endogeneity of predictor variables, omission of confounding variables, and failure to control for the cyclical nature of crime trends. Most of these problems should bias results toward overestimating the crime-reducing effects of laws making it easier to carry concealed firearms in public. Lott and Mustard's statistical models produce findings inconsistent with criminological theories and well-established facts about crime, and subsequent reanalysis of their data challenges their conclusions. Public health professionals should understand the methodological issues raised in this commentary, particularly when flawed research could influence the introduction of policies with potentially deleterious consequences.
