Criminology & Public Policy has a really interesting article in its latest issue (Vol. 6, No. 2, May 2007--no link) on "Gangs, Guns, and Drugs: Recidivism among Serious Young Offenders." It has serious implications for guidelines systems with gun use as sentence enhancers. Looks like justifications for the enhancers really can't rely on saying guns reflect future risk to public safety anymore and will have to fall back on pure "retribution for scaring the victim" arguments. Here are parts of the abstract, but be sure to hunt down the whole article if you can as well as the intro from Scott Decker and the responses from Marie Griffin and Jean Marie McGloin emphasizing the role of offender treatment in and out of prison in future recidivism.
. . . Results from a series of proportional hazard models indicate that race, gang membership, drug dependence, and institutional behavior are critical factors in predicting the time of reconviction. Contrary to expectations, gun use was not related to postrelease involvement in the criminal justice system [emphasis mine].
. . . The findings presented here indicate that preprison weapon involvement is not significantly associated with recidivism, likely because gun use is prevalent among young, serious offenders. Although policies aimed at the incapacitation of young, violent offenders may reduce community levels of crime in the short term, the chances of recidivism are likely to increase in the long term if factors like gang membership and drug use, and the deficits that these behaviors engender for social and emotional capital, are not addressed. More broadly, the strong, significant effect of the institutional misconduct measure signals the salience of accounting for institutional behavior when making release decisions. Institutional misconduct may be an important marker of sustained gang membership, making institutional programming and appropriate aftercare services a priority for this group of offenders.