Criminalization and Normalization: Some Thoughts About Offenders with Serious Mental Illness

Response to Professor E. Lea Johnston, Reconceptualizing Criminal Justice Reform for Offenders with Serious Mental Illness Abstract While Professor Johnston is persuasive that clinical factors such as diagnosis and treatment history are not, in most cases, predictive by themselves of criminal behavior, her concession that those clinical factors are associated with a constellation of risks and needs that are predictive of criminal system involvement complicates her efforts to maintain a clear boundary between the criminalization theory and the normalization thesis. Indeed, Professor Johnston’s article contains a brief section in which she identifies “possible justifications” for the specialized programs that are the target of her critique. These justifications deserve more attention, precisely because they suggest that the normalization thesis, while powerful, may not entirely displace the criminalization theory upon which those specialized programs rest. Moreover, even if the criminalization theory and the normalization theory are at least partially reconcilable, important questions remain regarding the proper allocation of limited resources both within the criminal legal system and the public mental health system. These questions, in turn, press focus on additional questions with respect to the underlying purposes of criminal system coercion more generally. The complex associations between serious mental illness and several of the key risk factors for criminal system involvement—most notably substance misuse—impact the design of effective programs intended to reduce reoffending. Given the practical difficulty of disentangling the treatment of mental disabilities from the provision of effective interventions to interrupt patterns of criminal behavior in offenders who have significant mental illness, some specialized attention to clinical mental health needs is warranted in correctional rehabilitation programs. These integrated programs are resource intensive, howeve