Public Policy and Limits of Diversion Programs for Reducing Jail Exposure of Persons With Serious Mental Illness

An estimated two million persons with serious mental illnesses are booked into U.S. jails each year (1). It has long been argued that jail detention may be unnecessary for many of the charges lodged against such individuals and that jail detention may indeed be quite harmful (2). In light of this recognition, many jurisdictions have developed various forms of diversion mechanisms. These programs seek either to circumvent the criminal justice process by encouraging police to transport potential arrestees to a mental health facility rather than to jail (known as prebooking diversion) or to use the courts to order mental health treatment in lieu of criminal sanctions (postbooking diversion) (3). Despite the proliferation of these programs, large numbers of persons with serious mental illnesses continue to be incarcerated (1), suggesting that diversion mechanisms by themselves may not be a panacea for this problem.

One reason for the continued high prevalence of mental illness in jails may be the limited range of persons deemed eligible for diversion. Public funding policy in this area typically focuses on individuals who pose minimal risk to public safety. The New Freedom Commission, for example, in its report on mental health care transformation, called for the diversion of nonviolent adult offenders, and this thinking is reflected in federal funding guidelines for local diversion programs (4). The Mentally Ill Offender Treatment and Crime Reduction Act, which provides funding for joint initiatives among local mental health and criminal justice agencies, specifies “low-level, nonviolent mentally ill offenders” as the target population for the services the legislation aims to support. Similarly, a 2000 amendment to Title I of America's Law Enforcement and Mental Health Project, which provided funding for mental health courts, identifies “offenders with mental illness, mental retardation or co-occurring mental illness and substance abuse disorders who were charged with misdemeanors or nonviolent offenses” as the population that the funded courts should serve. An unanswered question, which we address here, is how large a segment of the population of arrestees with serious mental illnesses would potentially be eligible for existing diversion programs, based on their charges and on current policies, and, conversely, how many would continue to be processed through the justice system and potentially jailed?

Methods

Data were from a longitudinal study of arrests in a cohort comprising all persons ages 18–64 who received either inpatient case management or residential services from the Massachusetts Department of Mental Health (DMH) between July 1, 1991, and June 30, 1992, and followed until December 31, 2002. For the purposes of this study, this time period largely predates widespread implementation of diversion programs in Massachusetts, which allowed us to examine the size of potential diversion populations free of the concern that some arrestees might have been diverted and thus not be represented in the data set. Data on all Massachusetts arrests occurring among these individuals over this period were obtained from the Massachusetts Criminal Offender Record Information system. The project was approved by the institutional review boards of a medical school and the DMH Central Office Research Review Committee.

The arrestees in this sample were charged with a wide range of offenses, and as described in greater detail elsewhere (5), a set of categories subsuming these offenses was developed. For purposes of this analysis we further combined these charges into misdemeanor and felony groupings on the basis of Massachusetts criminal codes.

Results

A total of 13,816 individuals met criteria for inclusion in this cohort. Of these, 3,856 (27.9%) were arrested at least once during the roughly ten-year observation period. Charges ranged from minor offenses, such as trespassing and disorderly conduct, to homicide. Over the observation period, 2,760, or 71.6% of the arrestees, were arrested at least once on charges that included one or more felony offenses. In fact, felonies predominated among the charges in this cohort; of the 41,854 charges filed against cohort members over this period, 27,004 (64.5%) were for felonies. These included a mix of violent crimes against persons (including homicide, nonnegligent manslaughter, assault and battery, assault with a dangerous weapon, assault on a police officer, and forcible rape), property crimes (burglary, larceny of an item worth more than $500, welfare fraud, receiving stolen property, passing bad checks, breaking and entering, arson, and motor vehicle theft), and some drug offenses. Of particular concern in this policy discussion are the violent felonies, meaning the felony crimes against persons listed above. In all, 1,847 individuals—13.3% of the cohort and 47.9% of arrestees—experienced one or more arrests on at least one of these charges.

Discussion

These data indicate that a sizeable number of persons in this cohort experienced at least one arrest episode in which, under the terms of prevailing federal funding policies that support and encourage diversion, they would not have been deemed eligible for diversion. There is, of course, wide variation in the applicability of diversion criteria, both within and between jurisdictions and even across individuals. Police familiar with some potentially chargeable individuals may choose diversion over arrest or vice versa. Conversely, within a postbooking model, some persons repeatedly charged with nonviolent misdemeanors might be denied diversion if judges see them as poor risks for compliance with the terms of a diversion agreement. However, charges may be dropped or reduced to a lower level during plea bargaining, and for persons seen as good risks for a diversion program, there is a strong likelihood that the parties to these proceedings can work to make diversion possible. These variations aside, public policy on diversion, as articulated in federal statutes and funding policies, will likely fail to prevent jail exposure in a majority of arrest episodes involving persons with serious mental illnesses.

Only a small number of existing programs for offenders with mental illness likely receive funding under these programs, and these funding policies are not a sufficient proxy for actual diversion practices. Many programs have been quite successful in preventing rearrest among persons charged with felonies, including violent ones. For example, New York's Nathaniel Project (6) has for years been highly successful in managing persons with mental illnesses charged with violent felonies. In a recent overview of forensic assertive community treatment teams, which work to maintain justice-involved persons with severe mental illness in the community, Lamberti and colleagues (7) identified a number of programs across the country that maintain felony charges as an inclusion rather than an exclusion criterion. And in an overview of what have been termed “second-generation mental health courts,” Redlich and colleagues (8) found that persons facing such charges were becoming increasingly prevalent in these evolving mental health courts. These courts use postadjudication models (that is, a finding or plea of guilty had been made) with the execution of the penalty—typically a jail sentence—stayed if the defendant complies with court-ordered treatment.

Conclusions

The goal of mental health diversion programs of all kinds is to balance the mental health needs of persons with mental illness who become entangled in the justice system with the needs for public safety (9). Federal support for diversion programs currently is cautious, trending toward the second of these concerns. Evidence is mounting, however, that public safety is not compromised by meeting the treatment needs of persons with mental illnesses who commit felonies, even ones that involve acts of violence against others. This makes sense; a defendant facing a lengthy state prison sentence may be more likely to conform to conditions of diversion than would a person diverted for a misdemeanor whose sanctions for noncompliance may include only a short jail stay or small fine. Moreover, from a public cost-versus-benefit perspective, diverting the individual potentially facing a long sentence in state prison represents a greater level of savings in court and correctional resources than the diversion of an individual who would likely face minimal sanctions.

That said, these programs walk a thin line; a violent crime committed by a diverted arrestee or mental health court defendant, especially one that gets significant media exposure, could represent a significant public relations setback for the programs involved and perhaps lessen any prospects of broadening the scope of diversion in that locale or even nationally.

Policy makers should draw on the experiences of these diversion services in making decisions regarding funding for diversion and mental health court programs. If we assume that significant numbers of offenders with serious mental illnesses will continue to find their way into local jails, we should recognize in this trend an opportunity for multiple systems to converge in serving this population. As Steadman (10) noted, jails should serve as “public health outposts” that meet a range of medical, psychiatric, and social service needs for persons whose needs are not being met elsewhere. This, of course, requires the commitment of resources necessary to make the interventions delivered by those outposts maximally effective.