AIM is primarily based on the principles of deterrence theory, which specifies that sanctions need to be certain, swift, and proportionate to the severity of the behavior to deter criminal activity (Beccaria, 1974; Bentham, 1789; Gibbs, 1975). Underlying these principles is the notion that humans are rational and weigh the costs and benefits of their actions. Theorists distinguish between two types of deterrence—general and specific. General deterrence is based in the notion that individuals learn from the experiences of others, and suggests clear laws with dependable enforcement will deter would-be offenders. On the other hand, specific deterrence is premised in experiential learning, suggesting an individual’s own experiences shape assessment of risks and rewards for future behaviors. Thus, to be deterred, offenders must perceive a high likelihood of receiving a sanction (certainty). Further, the sooner a sanction is applied (celerity), the greater the effect it will have to deter future law violations, with longer response times increasing perceptions that the sanction is unfair or questionable. Finally, per the severity principle, the sanction must be strong enough to deter the offense (i.e., the cost outweighs the benefit to the offender), though using strong sanctions such as detention and incarceration inappropriately or too quickly may lessen their deterrent effects. Moreover, any potential deterrence effect is likely dependent on both individual and situational characteristics, as opposed to uniform (Piquero et al., 2011), suggesting that responses should be individualized to the extent possible.
A highly consistent finding from the body of literature aimed at empirically testing deterrence theory is that the certainty, rather than the severity, of punishment yields the greater deterrent effect (Apel, 2013; Nagin, 2013). Research on celerity has provided less concrete evidence of a deterrent effect of swiftness (Paternoster, 2010). Despite mixed evidence, many criminal justice and correctional practices have been informed by deterrence theory. Summary evidence from reviews and meta-analyses of adult and juvenile correctional interventions have concluded that those based on deterrence are not effective and may even have harmful effects (MacKenzie & Farrington, 2015; Lipsey, 2009; Lipsey & Cullen, 2007). However, more recent interventions related to graduated sanctions have focused on increased certainty and celerity, and less on increased severity, with some reported success (e.g., Hawken & Kleiman, 2009; Hamilton et al., 2016; see below).
***
This evaluation examined the implementation and outcomes of Maryland DJS’s graduated response system, AIM. Results showed that just over half (55%) of youth supervised with AIM in place received at least one sanction or incentive during their supervision, and sanctions were more commonly applied than incentives. Responses to positive and negative behaviors were applied relatively swiftly—within 3 days on average—and staff followed the structured guidance for sanctions in almost 100% of their decisions. The study included an assessment of several youth outcomes, showing some favorable results among youth who received AIM. Specifically, youth supervised with AIM in place were less likely to have a VOP filed with the court, a new committed residential placement, and a new DJS referral/arrest that resulted in an adjudication/conviction during supervision. Finally, a cost analysis yielded mixed evidence related to potential cost savings attributable to AIM.
AIM is a tool for case managers to use with youth with the goals of encouraging supervision compliance and reducing unfavorable outcomes. Training materials drew largely from deterrence theory, calling for responses to be certain, swift, and proportionate to the behavior to achieve positive results. To that end, this study could not assess whether responses were applied when they were warranted (certainty). Complicating this assessment is the fact that not every youth under supervision will require an AIM response to be successful. Indeed, our analysis showed that just over half of the youth had a behavior and response documented as part of AIM. There are three possible explanations for youth who did not have an AIM response indicated in their case record: 1) they did not exhibit any behaviors that warranted a response or demonstrate a need to be incentivized; 2) a response was warranted but the case manager did not apply it; or 3) a response was applied but it was not recorded in the data system. Regarding the latter groups, DJS has a fairly robust supervision and quality assurance model in place, but we cannot look past the possibility that some responses were not applied or not logged.