A reassessment of the D.C. gun law: Some cautionary notes on the use of interrupted time series designs for policy impact assessment

Interrupted time series designs are commonly used to assess the impact of gun control legislation, as well as other legal and policy changes. Three common problems in the use of these designs - (1) selection of an appropriate control series, (2) specification of the intervention model, and (3) specification of the time series studied - raise questions about the validity of the conclusions reached in research on the impact of gun control. We illustrate these problems with a critical reassessment of Loftin et al.'s (1991) evaluation of the 1976 District of Columbia Gun Law. We then use monthly homicide data from the Federal Bureau of Investigation and the National Center for Health Statistics to illustrate how careful consideration of these three design issues results in a significantly different conclusion about the effectiveness of the District of Columbia Gun Law.