IJLP-01027; No of Pages 7 International Journal of Law and Psychiatry xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

International Journal of Law and Psychiatry

Long-term recidivism of mental health court defendants Bradley Ray ⁎ Indiana University–Purdue University Indianapolis, School of Public & Environmental Affairs, Business/SPEA Building, 801 West Michigan Street, Indianapolis, IN 46202, USA

a r t i c l e

i n f o

Available online xxxx Keywords: Mental health courts Recidivism Court decisions Problem-solving courts Long-term follow-up

a b s t r a c t The first MHC was established in 1997 and now, over 15 years later, there are over 300 mental health courts in the United States. In a relatively short time these courts have become an established criminal justice intervention for persons with a mental illness. However, few studies have looked at the long-term outcomes of MHCs on criminal recidivism. Of the studies evaluating the impact of MHCs on criminal recidivism, most follow defendants after entry into the court during their participation, and only a few have followed defendants after court exit for periods of one or two years. This study follows MHC defendants for a minimum of five years to examine recidivism post-exit with particular attention to MHC completion's effect. Findings show that 53.9% of all MHC defendants were rearrested in the follow-up and averaged 15 months to rearrest. Defendants who completed MHC were significantly less likely to be rearrested (39.6% vs. 74.8%), and went longer before recidivating (17.15 months vs. 12.27 months) than those who did not complete. This study suggests that MHCs can reduce criminal recidivism among offenders with mental illness and that this effect is sustained for several years after defendants are no longer under the court's supervision. © 2014 Elsevier Ltd. All rights reserved.

1. Introduction Given the large numbers of persons with serious mental illness in the criminal justice system (Abram, Teplin, & McClelland, 2003; Abram, Teplin, McClelland, & Dulcan, 2003; Steadman, Osher, Robbins, Case, & Samuels, 2009; Teplin, 1990; Teplin, Abram, & McClelland, 1996; Trestman, Ford, Zhang, & Wiesbrock, 2007) and the fact that many of these individuals repeatedly cycle through the system, local US jurisdictions have implemented various diversionary programs for mentally ill offenders. One such program is the mental health court (hereafter MHC), which is a type of problem-solving court that attempts to divert persons with mental illness out of the cycle of arrest, incarceration, release and rearrest, by requiring and motivating them to connect with treatment and services and to change their behaviors (Almquist & Dodd, 2009). The MHC uses case management and enhanced judicial supervision to monitor a defendant's progress. Judges, probation officers, social workers, community corrections, and treatment service professionals work together as part of the MHC team to develop treatment plans for each defendant and monitor defendants' progress (or lack thereof) while under court supervision. Individualized treatment plans may include requirements like attending a treatment program, meeting with a mental health professional, submitting to drug screenings, complying with a medication regimen, and offering some form of restitution. Some defendants complete the court process meaning they were compliant with court mandates for a ⁎ Tel.: +1 317 274 8701. E-mail address: [email protected].

continuous period of time and received a full “dose” of the court's treatment, services, structure, supervision and encouragement (Moore & Hiday, 2006). Other defendants who are persistently noncompliant are terminated from the process, receive only a part of their individualized plans, and eventually have their charges sent back to traditional court. Some opt out, choosing to return to traditional court for processing of their cases. These two groups are the MHC noncompleters. The majority of empirical research on MHCs has focused on criminal recidivism and has found that defendants who participate in a MHC have lower rates of reoffending than before entering the MHC (Burns, Hiday, & Ray, 2013; Christy, Boothroyd, Petrila, & Poythress, 2003; Dirks-Linhorst & Linhorst, 2012; Frailing, 2010; Herinckx, Swart, Ama, Dolezal, & King, 2005; Moore & Hiday, 2006; Palermo, 2010; Steadman, Redlich, Callahan, Robbins, & Vesselinov, 2011; Trupin & Richards, 2003). When compared to defendants with a mental illness in a traditional criminal court, MHC defendants are no more likely to reoffend (Christy, Poythress, Boothroyd, Petrila, & Mehra, 2005; Cosden, Ellens, Schnell, Yamini-Diouf, & Wolfe, 2003; Dirks-Linhorst & Linhorst, 2012; Frailing, 2010; Hiday, Wales, & Ray, 2013; McNiel & Binder, 2007; Moore & Hiday, 2006; Steadman et al., 2011; Trupin & Richards, 2003). Some of these studies had comparison groups that consisted of defendants who were not referred to MHC or did not opt into MHC after referral (Dirks-Linhorst & Linhorst, 2012; Frailing, 2010; Hiday et al., 2013; McNiel & Binder, 2007; Moore & Hiday, 2006; Steadman et al., 2011; Trupin & Richards, 2003), while others had no comparison group and looked at recidivism between MHC completers and noncompleters (Burns et al., 2013; Herinckx et al., 2005; Hiday & Ray, 2010; Palermo, 2010).

http://dx.doi.org/10.1016/j.ijlp.2014.02.017 0160-2527/© 2014 Elsevier Ltd. All rights reserved.

Please cite this article as: Ray, B., Long-term recidivism of mental health court defendants, International Journal of Law and Psychiatry (2014), http://dx.doi.org/10.1016/j.ijlp.2014.02.017

2

B. Ray / International Journal of Law and Psychiatry xxx (2014) xxx–xxx

Although there are now over 300 MHCs throughout the United States (Almquist & Dodd, 2009)—and this number continues to grow— most of the studies have examined recidivism after MHC entry with follow-up over the time defendants are still in the MHC. Only a few studies have looked at the impact of the MHC on offending behavior post MHC exit (Burns et al., 2013; Dirks-Linhorst & Linhorst, 2012; Hiday & Ray, 2010; Hiday et al., 2013; McNiel & Binder, 2007). By examining offending behavior post MHC exit, researchers are able to determine whether the MHC program that is expected to impact recidivism does so for a sustained period of time when defendants are no longer under the court's monitoring and receiving its treatment and services. Moreover, many MHC teams acknowledge that they are trying to change long-standing patterns of criminal behavior and accept that defendants often make mistakes early on in the MHC process and may be re-arrested (Ray, Dollar, & Thames, 2011; Redlich et al., 2010). In such cases, the team can decide to add the additional charges to the original ones on the MHC docket. Looking at offending after MHC entry would count these arrests as recidivism when in fact technically these additional charges are disposed of along with the original charges if the defendant successfully completes the MHC. Of the post-exit studies, the longest follow-up period has been two years (Burns et al., 2013; Hiday & Ray, 2010). Longer-term follow-up studies allow researchers to examine how long MHCs reduce offending and whether they help reintegrate defendants back into the community as law-abiding citizens. Given the spread of MHCs in the United States, it is important that policy decision makers are provided with information on the longer-term criminal justice outcomes of offenders with serious mental illness when considering the effectiveness of MHCs. Until recently such evidence has been hard to assemble as MHCs have not been around long enough for evaluators to complete studies with long-term outcomes. The present study examines post MHC exit arrests for a minimum of five years of all defendants who participated in one MHC in its first six years, 2000 to 2006. In doing so the study also investigates differences between completers and noncompleters of the court. MHC studies consistently suggest that completers are less likely to recidivate than noncompleters because they receive a full “dose” of MHC supervision, treatment, case management, services, and support. Unlike MHC studies that have examined noncompleters' recidivism, this study accurately assesses the risk period of rearrest for noncompleters by considering traditional court disposition and jail time dates. Using survival analysis to examine the likelihood of criminal recidivism and the length of time post MHC exit until defendants recidivate, this study addresses whether MHC participation and completion leads to compliance with the law in the years following MHC exit. In multivariate analyses, it controls those factors shown by previous empirical research to be significant in predicting recidivism and desistance over time. 1.1. Setting This study's MHC was the first in North Carolina. It takes misdemeanors and felonies, nonviolent and violent charges, and preadjudication and post-adjudication cases of defendants with mental illness. To participate the defendant voluntarily signs an agreement to comply with the court ordered individualized treatment and behavioral mandates. During MHC, criminal charges for misdemeanants and sentences for felons are placed in abeyance. Defendants are required to attend court sessions each month for compliance audits. Compliance is determined at monthly team meetings that occur before each court session. This MHC does not have a phased completion status, instead defendants must remain in compliance with court orders for six consecutive months to have their charges disposed of positively. If a defendant is non-compliant, the judge may express disappointment, issue a warning, place the defendant in jail for a short stay, and/or order alternative treatments or services. If the non-compliance continues or the team determines that the defendant is unsuitable for MHC, the judge reassigns

him or her back to traditional court. Earlier reports of this court found that criminal offending was reduced during court monitoring (Moore & Hiday, 2006), for one year post entry (Moore & Hiday, 2006) and up to two years post-exit (Hiday & Ray, 2010). 2. Methods This study uses court administrative data for all defendants who were in the first MHC in North Carolina for its first six years, 2000 to 2006 (N = 449). Over the six years of data collection, the number of defendants who entered the MHC per year ranged from 56 to 100, averaging 74.8 (SD = 12.68) per year. This number varied with fluctuations in the ability of the court to provide case management and mental health services as funding fluctuated. Court data provided demographics, key arrest (the arrest that brought them into MHC), court dates and exit statuses (i.e., complete, ejected, opt-out) on all defendants; and statewide criminal history files provided their dates of arrest, arrest charges, dispositions and disposition dates of the charges. Cases were merged by key arrest to determine disposition for those defendants who were sent back to traditional criminal court. In order to accurately measure each defendant's risk period for recidivism, data were also collected on the timing of traditional court outcomes (i.e., court disposition and jail release date) for those who did not complete the MHC process. The dependent variables in this study are rearrest and time to first rearrest post-exit. Data collection on these variables occurred during November 2011, providing a follow-up period of over a decade for those defendants who left the court at the end of 2001 and over five years for those who left in the end of 2006. Most MHC studies have not examined the traditional court outcomes of MHC noncompleters (see Hiday et al., 2013, for exception); however, these outcomes are important to accurately determine each defendant's risk period for rearrest, that is, the time when a defendant is no longer under the supervision of the MHC or no longer incarcerated and in the community capable of recidivating. Cox regression survival analysis was used to examine recidivism post-exit. One of the strengths of survival analysis methods is how it handles censored data: left censoring occurs when data on the starting point are not available and right censoring when there are no data on the ending point, which often occurs when studying recidivism. In order to accurately measure each defendant's risk period for recidivism, data were also collected on the timing of traditional court outcomes (i.e., court disposition and jail release date) for those who did not complete the MHC process. As such, none of these data are left censored as all starting points for the risk of recidivism are known; for completers (n = 265) the starting point is the last day of the MHC and for defendants whose charges were sent back to traditional court (e.g., opt-outs and noncompleters, n = 184), the starting point is the date of release from jail or prison for those incarcerated and on the date that key arrest was disposed in traditional court for those not incarcerated. However, there is right censoring because for some individuals the event of interest, rearrest, had not yet occurred before the end of the study period. Cox regression uses the censored and uncensored (i.e., those that did recidivate) cases to calculate the probability of surviving (i.e., not recidivating) for each time point (Box-Steffensmeier & Bradford, 2004). Covariates can be added to the Cox regression equation to predict the hazard rate, which is the probability of the event occurring in a given time period given survival through prior time periods. Therefore, it is able to examine both the likelihood of and time to recidivism by court exit status (e.g., completers and noncompleters). With such a long-term riskperiod for recidivism these data can also explore how long a follow-up would be necessary to capture those who might recidivate post-exit. A methodological issue rarely discussed in the MHC literature (see Christy et al., 2005 for an exception) is that some defendants are accepted back into the MHC after exiting. Over the six years of MHC defendants examined in this study, 18.0% (n = 81) of those who left the MHC were accepted back into the court. Most (60.5%, n = 49) of these

Please cite this article as: Ray, B., Long-term recidivism of mental health court defendants, International Journal of Law and Psychiatry (2014), http://dx.doi.org/10.1016/j.ijlp.2014.02.017

B. Ray / International Journal of Law and Psychiatry xxx (2014) xxx–xxx

had previously completed MHC while 39.5% (n = 32) had returned to traditional court for adjudication of their key arrests. This study uses defendant's first instance in MHC, such that each defendant is counted only once and each is coded as having recidivated regardless of the outcome of the second MHC participation.

3

these 34.5% (n = 40) were sent to jail, 58.6% (n = 68) were placed on probation, and 6.9% (n = 8) received only a monetary fine. The average time between leaving the MHC and traditional court disposition was 4.7 months (SD = 4.59), and the average amount of time spent in jail (for those sentenced to jail) was slightly under 6 months (M = 5.85; SD = 7.23).

3. Results 3.3. Survival analysis predicting criminal recidivism 3.1. Sample description Table 1 displays descriptive statistics for the sample. Average age at MHC entry was 34.7 (SD = 11.94). More than half (60.4%) of the population was White, 37.0% was Black, 1.6% was Hispanic, and 1.1% was Asian. Over two-thirds (68.4%) of the sample was males. MHC defendants averaged about nine months in the MHC (M = 8.75; SD = 5.14). While the MHC accepts felony key arrests, the great majority in this sample are misdemeanors (87.5%). Lifetime prior arrests ranged from 0 to 49 with an average of 8.16 (SD = 9.28). Half of the sample had 4 or fewer arrests (17.4% had one arrest and 15.1% had two arrests). Over a third of the defendants had a prior felony arrest (35.2%); and over a quarter (27.8%) were rearrested while in the MHC under its supervision. 3.2. Exit status Of the full sample of MHC defendants from 2000 to 2006, 59.0% (N = 449) successfully completed the MHC process, consequently given a graduation ceremony in which the criminal charges were dismissed (Ray et al., 2011). The completion rate per year ranged from 37.1% to 63.9%. One-fourth (24.5%) of those who completed did so in six months, the minimum amount of time that a defendant could be in the study's MHC and successfully complete the program; but the average time to successful completion was 10 months (SD = 4.82). Of those who began the MHC process but did not complete it and had their criminal charges sent back to traditional court (41.0%, n = 184), the great majority (88.6%, n = 163) were terminated from the MHC process and 11.4% (n = 21) decided to opt-out. Those who opted-out left the MHC earlier than those who were terminated, averaging 4.19 months in the MHC (SD = 2.61); those who were terminated averaged 6.74 months (SD = 5.20) before exiting. Disposition data on noncompleters revealed that many criminal charges were dismissed as a result of the prosecutor deciding not to go forward with the charges or the judge deciding to dismiss the charges. Among noncompleters whose cases were adjudicated in a traditional court, 37.0% (n = 68) eventually had their criminal charges dismissed and received no further monitoring or punishment. The remaining 63.0% (n = 116) were found guilty; and of

Table 1 Sample characteristics. M (SD) Age Race White Black Hispanic Asian Sex Female Male Prior arrests Any prior felony Key arrest felony Length of time in MHC Any recidivism post-exit Any felony recidivism Total N = 449

N (%)

34.7 (11.94) 271 (60.4) 166 (37.0) 7 (1.6) 5 (1.1) 142 (31.6) 307 (68.4) 8.16 (9.28) 158 (35.2) 56 (12.5) 8.75 (5.14) 242 (53.9) 38 (15.7)

During the five to 10 years of the follow-up period, 46.1% (n = 207) remained out of the criminal justice system while slightly over half (53.9%, n = 242) of the sample recidivated. Most of those rearrested were charged with misdemeanor offenses (84.3%, n = 204) but 15.7% (n = 38) had felony charges. Survival analyses were used to examine time to rearrest, which produces a life-table that describes duration distributions for the full sample or by key variable levels. Table 2 displays an abbreviated life-table with the rate of recidivism during each follow-up year, with Year 1 indicating 0 through 365 days. The “survivors” count is the number of cases that are not right-censored (i.e., follow-up data are still available) that have not yet recidivated. The recidivism column indicates the number of cases that recidivated in that time period as well as the portion of recidivism in that period given survival through the earlier periods. One can observe that 27 defendants recidivated in year three which is 10.8% of all those who had not yet recidivated. Another way to describe the life-table is to determine the time interval at which half of the population experienced the event of interest, in this case, recidivism. Among this full sample, half of those who would eventually recidivate did so by month 37, approximately three years. Table 2 shows that the greatest proportion of the sample recidivated during year one (33.4%). Cumulatively, by the end of year two, 44.3% had recidivated and by the end of year three 50.3% had recidivated. Only 1.6% of the sample (n = 4) recidivated beyond year five. The time to first rearrest ranged from 9 days to 81 months and the average time to rearrest was 14.73 months (SD = 14.17). By the end of the follow-up period, 46.1% (n = 207) of all the MHC defendants, all of whom had at least one arrest to result in their referral to the MHC, had still not been rearrested. There was a significant difference between those defendants who completed the MHC and those who did not. Noncompleters were almost twice as likely to have recidivated (74.5% vs. 39.6%, χ2 = 59.03, p b .001) and to recidivate with a felony arrest (68.4% vs. 31.6%, χ2 = 12.93, p b .001) than those who completed the MHC. For both completers and noncompleters the most common recidivating offenses were property crimes (40.2% and 45.3% respectively) with the most common property crimes being larceny and second degree trespassing. Among noncompleters, 52.7% of those who recidivated did so in year one compared to only 20.0% of the completers. Moreover, while noncompleters' rate of recidivism consistently declined, completers show a slightly different pattern with the portion of recidivism increasing from year two to year three (10.4% to 12.1%). Noncompleters recidivated significantly sooner than completers (17.15 months and 12.27 months respectively, t = 2.67, p b .01). Table 2 shows that the majority of noncompleters recidivated in the first year post-exit but that many of those who completed the MHC did not recidivate until year two or three post-exit and suggests that completing the MHC gives a greater boost to avoid reoffending and remain law abiding. Another way to display the life-table is to graph the hazard of recidivism, that is, the chance of recidivism at a given time, conditional upon survival up until that time. Fig. 1 illustrates the hazard by months with lines representing completers and noncompleters. The hazard starts much higher for noncompleters than completers as 31% of the noncompleters had recidivated by the end of month 6 compared to only 9% of the completers. The hazard of recidivism continually reduces throughout the first year as the number of those who recidivate, relative to the number of those who have not recidivated, decreases. The

Please cite this article as: Ray, B., Long-term recidivism of mental health court defendants, International Journal of Law and Psychiatry (2014), http://dx.doi.org/10.1016/j.ijlp.2014.02.017

4

B. Ray / International Journal of Law and Psychiatry xxx (2014) xxx–xxx

Table 2 Life table of years to recidivism by exit status. Year

Total (N = 468)

0–1 1–2 2–3 3–4 4–5 5–6 6–7 7–8 8–9

Completers (N = 265) Recidivism

Surviving

Recidivism

Surviving

Recidivism

449 299 250 223 214 205 192 134 105

150 (33.4%) 49 (16.4%) 27 (10.8%) 9 (4.0%) 3 (1.4%) 1 (0.5%) 2 (1.2%) 0 (0.0%) 1 (1.1%)

265 212 190 167 163 159 149 103 82

53 (20.0%) 22 (10.4%) 23 (12.1%) 4 (2.4%) 1 (0.6%) 1 (0.6%) 0 (0.0%) 0 (0.0%) 1 (1.4%)

184 87 60 56 51 46 43 31 23

97 (52.7%) 27 (31.0%) 4 (6.7%) 5 (8.9%) 2 (4.0%) 0 (0.0%) 2 (5.3%) 0 (0.0%) 0 (0.0%)

sustained rate of recidivism for completers in year two and three described above is illustrated by months in Fig. 1 as a pattern of a nondecreasing hazard rate. Table 3 presents the results of Cox regression models predicting time to recidivism. Whether or not the event occurs is dichotomous (1 = yes; 0 = no) and time is measured in months. Covariates were added into the model chronologically. Model 1 looks at the effect of the demographic variables on recidivism and shows that both age and race are significant predictors. The hazard of recidivism is 1.3 times greater for Nonwhite defendants than for White defendants, while age at entry suggests that the hazard ratio for recidivism decreases by 1% at each additional year of age. Model 2 adds two measures of prior criminal behavior: number of prior arrests and whether any of these arrests was a felony. Each additional prior arrest is associated with 4% increase in the hazard of recidivism net of the other variables in the model; moreover, the addition of prior arrests mediates the relationship between race and recidivism from Model 1. This mediation is likely due to the association between race and prior arrests in this sample: Nonwhites averaged 10.17 prior arrests compared to 6.55 prior arrests for Whites (t = 3.96, p b .001). Model 2 illustrates that the relationship between age and recidivism remained despite the addition of prior arrests. Model 3 adds dichotomous measures of felony charge at key arrest, MHC exit by opt-out and rearrest during MHC supervision. Of these, only rearrest during the MHC was significant. Model 3 suggests that those who were rearrested during the MHC have a hazard rate of recidivism that is 2.14 times greater than those who were not net other variables in the model. Age and prior arrests remain significant in Model 3. Finally, Model 4 includes dichotomous variables that measure incarceration after traditional court and successful completion of the MHC.

0.06 Completers

0.05

Noncompleters

Hazard

0.04

0.03

0.02

0.01

0.00 0

Noncompleters (N = 184)

Surviving

25

50

75

100

125

Months Fig. 1. Hazard rate of recidivism for MHC completers and noncompleters.

Incarceration post-exit was not associated with a greater hazard of recidivism; however, net of the other variables in the model, those defendants who successfully completed the MHC had a hazard rate of recidivism that is 2.20 times (Exp b = −0.79) less than those who did not. In this final model, age at entry, number of prior arrests, and rearrest during the MHC remain significant predictors of recidivism. Fig. 2 shows the estimated cumulative probability of rearrest for MHC completers and noncompleters controlling for the variables in Model 4. The cumulative survival rate across is the proportion of defendants who had not recidivated during each of the follow-up month with the distance between the lines representing the additional effect of MHC completion at each time point. As shown in the figure, noncompleters had a significantly shorter time to rearrest. For example, at twelve months, 20.0% of the MHC completers had recidivated compared to 52.7% of the noncompleters. 4. Discussion To date, the longest follow-up of MHC defendants has been two years post-exit (Burns et al., 2013; Hiday & Ray, 2010; and for graduates only McNiel & Binder, 2007). The present study had a minimum followup of five years and a maximum of over ten years. It was found that despite this extended follow-up period, almost half (46.1%) of all the MHC defendants did not recidivate. This study's rearrest rate of 53.9% over a much longer period is comparable to the two years post exit rearrest rate reported by other studies (60.6%, Burns et al., 2013, 48.0%, Hiday & Ray, 2010, and 36.0% for graduates only, McNiel & Binder, 2007). In the present study, the majority of those who recidivated did so in the first two years (82.2%). If this had been a two year follow-up, the rearrest rate would have been 44.3% rather than 53.9%. This study's findings are consistent with extant research (Burns et al., 2013; Dirks-Linhorst & Linhorst, 2012; Hiday & Ray, 2010; Hiday et al., 2013; McNiel & Binder, 2007) in showing that those who complete (graduate from) the MHC process are less likely to recidivate than those who do not and that they have a longer time in the community before reoffending. More than half of the noncompleters who recidivated did so in the first year; while half of the completers still had not recidivated by the end of 5 year follow-up. The Bureau of Justice Statistics reports that 67.5% of all inmates in prisons across the United States are rearrested within three years of release (Langan & Levin, 2002); while studies of individual states report a slightly higher three year rearrest rate for released mentally ill inmates (70%, see Lovell, Gagliardi, & Peterson, 2002; Silver, Cohen, & Spodak, 1989). Most defendants from the MHC in the present study were misdemeanants and not likely to go to prison, making recidivism patterns of jail inmates with a mental illness a better comparison. Teplin, Abram, and McClelland (1994) reported that nearly half of jail detainees with a severe mental illness were rearrested for a violent crime during a 6 year follow-up. More recently, Wilson, Draine, Hadley, Metraux, and Evans (2011) found a 54% reincarceration rate of jail inmates during a 4 year follow-up. Closer to the present study, are studies following rearrests of misdemeanants in other types of community support

Please cite this article as: Ray, B., Long-term recidivism of mental health court defendants, International Journal of Law and Psychiatry (2014), http://dx.doi.org/10.1016/j.ijlp.2014.02.017

2685.16⁎⁎⁎

⁎⁎⁎

2713.71⁎⁎⁎

⁎⁎⁎

⁎⁎⁎

−0.01 (0.01) 0.06 (0.13) −0.10 (0.14) 0.03 (0.01) 0.17 (0.17) 0.00 (0.20) 0.51 (0.28) 0.76 (0.14)

2744.16⁎⁎⁎

⁎⁎

⁎⁎⁎

0.98 (0.97–0.99) 1.10 (0.84–1.43) 0.96 (0.73–1.26) 1.04 (1.02–1.05) 1.26 (0.91–1.74) −0.02 (0.01) 0.09 (0.13) −0.04 (0.14) 0.04 (0.01) 0.23 (0.17)

⁎ p b .05. ⁎⁎ p b .01. ⁎⁎⁎ p b .001.

Age Nonwhite Male Number of prior arrests Any prior felony Felony key arrest Opted-out of MHC Re-arrest during MHC Jail post MHC exit MHC completer −2 log likelihood χ2

−0.01 (0.01) 0.27 (0.13) 0.03 (0.14)

⁎ ⁎

2788.84⁎

B (SE)

0.99 (0.98–1.00) 1.31 (1.02–1.69) 1.03 (0.78–1.35)

Model 2

B (SE)

Exp b (95% CI) Model 1

Table 3 Cox regression predicting recidivism post exit.

⁎⁎⁎

⁎⁎⁎

0.99 (0.97–1.00) 0.95 (0.73–1.24) 0.85 (0.65–1.13) 1.03 (1.01–1.05) 1.10 (0.80–1.53) 1.02 (0.69–1.51) 1.13 (0.65–1.98) 1.85 (1.38–2.47) 0.90 (0.59–1.37) 0.45 (0.34–0.61) −0.01 (0.01) −0.05 (0.14) −0.16 (0.14) 0.03 (0.01) 0.10 (0.17) 0.02 (0.20) 0.12 (0.28) 0.61 (0.15) −0.10 (0.22) −0.79 (0.15) 0.99 (0.97–1.00) 1.06 (0.81–1.38) 0.91 (0.69–1.20) 1.03 (1.02–1.05) 1.18 (0.85–1.64) 1.00 (0.68–1.47) 1.66 (0.96–2.87) 2.14 (1.63–2.81)

B (SE) B (SE) Exp b (95% CI)

Model 3

⁎⁎

Exp b (95% CI)

Model 4

⁎⁎

Exp b (95% CI)

B. Ray / International Journal of Law and Psychiatry xxx (2014) xxx–xxx

5

programs and in other types of diversion programs. However, most of these studies have short follow-up periods (12 months or less) or report on reincarceration (rather than rearrest) or jail days (see Martin, Dorken, Wamboldt, & Wootten, 2012 for a review). One exception, which is comparable to the present study, is Ventura, Cassel, Jacoby, and Huang (1998) who reported a 72% rearrest rate for those with case management over a 3 year follow-up of persons released from a jail in Ohio (see also Draine, Solomon, and Meyerson (1994) who report on reincarceration, and Cusack, Morrissey, Cuddeback, Prins, and Williams (2010) who report on jail days). The rearrest rate of the present study's MHC completers is much lower than the rearrest rate of any of the above recidivism studies, while noncompleters' rearrest rate is similar to those who were not diverted and who had been in prison. Long-term recidivism research suggests a turning point in the risk of recidivism whereby the majority of those who recidivate are likely to do so by year three (Kitchener, Schmidt, & Glaser, 1977; Kurlychek, Brame, & Bushway, 2007) and that an offender's risk of recidivism declines to a point where it is the same as an individual with no criminal record in about six (Kurlychek et al., 2007) to ten years (Blumstein & Nakamura, 2009). A similar pattern was found in this study with the majority recidivating by year 3 and with only 1.6% of those who recidivated doing so after year five. This recidivism pattern suggests that future MHC evaluations should conduct a two or three year follow-up to fully capture the majority of those who recidivate post-exit. This study considered the traditional court outcomes of defendants who did not complete the MHC. Of those noncompleters (n = 184), 41.3% were found guilty, 37.0% had their charges dismissed and 21.7% were sent to jail. Determining the disposition and jail release dates was necessary in order to accurately calculate each defendant's risk period for recidivism. For example, the average sentence length for noncompleters who went to jail was six months. Without considering the time period when these noncompleters were not at risk and not able to recidivate, they would have appeared to have a much longer time until recidivating. The Cox regression analysis found that defendants who completed the MHC were significantly less likely to reoffend and had a significantly longer time to recidivism than noncompleters when confounders were controlled. Some of these, age at MHC entry, number of prior arrests, and rearrest during MHC, were also significant predictors of recidivism, not a surprising finding given that age and criminal history are among the best predictors of future criminal behavior among offenders with and without mental illness (Bonta, Law, & Hanson, 1998; Ulmer, 2001). It is likely that both prior arrests and rearrest during the MHC are tapping into the same underlying criminality because those who were rearrested during the MHC averaged nearly twice as many prior arrests as those who were not (11.08 vs. 5.79, t = 6.11, p b .000) and a significantly greater portion of them went on to recidivate post MHC exit (78.4% vs. 21.6%, χ2 = 41.85, p b .000). In interpreting these results, some limitations should be kept in mind. Only one MHC was examined. The components of this court's structure and process may be different from those of other MHCs, ultimately limiting the generalizability of these findings. For example, the observed MHC accepts defendants with either a felony or a misdemeanor arrest and does not require most defendants to plea guilty; it does not have a phased completion process; status hearings are monthly; and defendants must be compliant for six consecutive months in order to graduate. Despite possible differences, the observed setting has the essential elements of a MHC: a specialized voluntary court docket, judicial supervision and individualized treatment plans, regular status hearings to review compliance, incentives and sanctions, and positive legal outcomes for completing the program (Council of State Governments Justice Center, 2011). The available measures in this study did not include the defendant's psychological and social characteristics such as employment, housing, mental health, global functioning, substance abuse, deviant peer groups, social supports, or the appropriateness of the treatments and services before, during or after MHC. Moreover, it would

Please cite this article as: Ray, B., Long-term recidivism of mental health court defendants, International Journal of Law and Psychiatry (2014), http://dx.doi.org/10.1016/j.ijlp.2014.02.017

6

B. Ray / International Journal of Law and Psychiatry xxx (2014) xxx–xxx

Cumulative Survival Rate

References

Completer

Noncompleters

Months Post-Exit Fig. 2. Time to rearrest for MHC completers and noncompleters.

have been ideal to have random assignment to MHC or TCC so that possible referral bias could be eliminated or, short of a random control trial, to have a control group of defendants with mental illness who were matched on relevant variables. An earlier study of this MHC found the MHC participants to have significantly greater reductions in rearrests after entry than a matched group of defendants with mental illness in a traditional court (Moore & Hiday, 2006); however, the present study was limited to conducting a long-term study of rearrest data of only MHC defendants. Despite these limitations this study adds to the accumulating evidence on the effectiveness of MHCs in reducing recidivism among offenders with mental illness by looking at the long-term and continued impact of a MHC post-exit. Some defendants who completed the court had 20 or more arrests before entering but no arrests in the five to ten years post-exit, suggesting that the MHC experience served as a “turning point” in what might otherwise have been a lifetime career in crime (Sampson & Laub, 1993). Studies have suggested specific mechanisms of the MHC that promote compliance with mandates for treatment and behavioral change (Canada & Watson, 2013; Poythress, Petrila, McGaha, & Boothroyd, 2002; Ray et al., 2011; Wales et al., 2010). Future research should examine how these mechanisms, as well as other aspects of the MHC experience, directly or indirectly lead to long-term criminal desistance.

5. Conclusion Few studies have used a post-exit design in which the risk period for potential recidivism begins after MHC exit (Burns et al., 2013; DirksLinhorst & Linhorst, 2012; Hiday & Ray, 2010; Hiday et al., 2013; McNiel & Binder, 2007) rather than after MHC entry. By examining offending behavior post MHC exit researchers are able to determine whether the MHC's program that is expected to impact recidivism does so for a sustained period after the time when defendants are no longer under the court's monitoring and receiving its treatment and services. With a minimum of five years follow-up and a maximum of ten years, this study is the longest follow-up of a MHC to date. During this longer period, 46.1% of MHC defendants still had not been rearrested. Most of those who did recidivate did so sooner rather than later. Among those who completed the MHC, 60.4% did not recidivate in the five or more years post-exit. Noncompleters recidivated sooner, usually in the first year, while completers recidivated in the second or third year postexit. Defendant's age, criminal behavior and exit status were predictors of the time to recidivism.

Abram, K. M., Teplin, L. A., & McClelland, G. M. (2003). Comorbidity of severe psychiatric disorders and substance use disorders among women in jail. American Journal of Psychiatry, 160(5), 1007–1010. Abram, K. M., Teplin, L. A., McClelland, G. M., & Dulcan, M. K. (2003). Comorbid psychiatric disorders in youth in juvenile detention. Archives of General Psychiatry, 60(11), 1097–1108. Almquist, L., & Dodd, E. (2009). Mental health courts: A guide to research-informed policy and practice. New York: Council of State Governments Justice Center. Blumstein, A., & Nakamura, K. (2009). Redemption in the presence of widespread criminal background checks. Criminology, 47(2), 327–359. http://dx.doi.org/10.1111/j. 1745-9125.2009.00155.x. Bonta, J., Law, M., & Hanson, K. (1998). The prediction of criminal and violent recidivism among mentally disordered offenders: A meta-analysis. Psychological Bulletin, 123(2), 123–142. http://dx.doi.org/10.1037//0033-2909.123.2.123. Box-Steffensmeier, J. M., & Bradford, S. J. (2004). Event history modeling: A guide for social scientists. New York: Cambridge University Press. Burns, P. J., Hiday, V. A., & Ray, B. (2013). Effectiveness 2 years postexit of a recently established mental health court. American Behavioral Scientist, 57(2), 189–208. Canada, K. E., & Watson, A.C. (2013). ‘Cause everybody likes to be treated good’: Perceptions of procedural justice among mental health court participants. American Behavioral Scientist, 59(2), 209–230. http://dx.doi.org/10.1177/0002764212465415. Christy, A., Boothroyd, R. A., Petrila, J., & Poythress, N. (2003). The reported prevalence of mandated community treatment in two Florida samples. Behavioral Sciences & the Law, 21(4), 493–502. http://dx.doi.org/10.1002/Bsl.550. Christy, A., Poythress, N. G., Boothroyd, R. A., Petrila, J., & Mehra, S. (2005). Evaluating the efficiency and community safety goals of the Broward County mental health court. Behavioral Sciences & the Law, 23(2), 227–243. http://dx.doi.org/10.1002/Bsl.647. Cosden, M., Ellens, J. K., Schnell, J. L., Yamini-Diouf, Y., & Wolfe, M. M. (2003). Evaluation of a mental health treatment court with assertive community treatment. Behavioral Sciences & the Law, 21(4), 415–427. http://dx.doi.org/10.1002/Bsl.542. Council of State Governments Justice Center (2011). Mental health courts date. Retrieved from. www.consensusproject.org/issueareas/mental-health-courts Cusack, K. J., Morrissey, J. P., Cuddeback, G. S., Prins, A., & Williams, D.M. (2010). Criminal justice involvement, behavioral health service use, and costs of forensic assertive community treatment: A randomized trial. Community Mental Health Journal, 46(4), 356–363. http://dx.doi.org/10.1007/s10597-010-9299-z. Dirks-Linhorst, P. A., & Linhorst, D.M. (2012). Recidivism outcomes for suburban mental health court defendants. American Journal of Criminal Justice, 37, 76–91. Draine, J., Solomon, P., & Meyerson, A. (1994). Predictors of reincarceration among patients who received psychiatric-services in jail. Hospital & community psychiatry, 45(2), 163–167. Frailing, K. (2010). How mental health courts function: Outcomes and observations. International Journal of Law and Psychiatry, 33(4), 207–213. http://dx.doi.org/10. 1016/j.ijlp.2010.06.001. Herinckx, H. A., Swart, S.C., Ama, S. M., Dolezal, C. D., & King, S. (2005). Rearrest and linkage to mental health services among clients of the Clark county mental health court program. Psychiatric Services, 56(7), 853–857. Hiday, V. A., & Ray, B. (2010). Arrests two years after exiting a well-established mental health court. Psychiatric Services, 61(5), 463–468. Hiday, V. A., Wales, H. W., & Ray, B. (2013). Effectiveness of a short–term mental health court: Criminal recidivism one year postexit. Law and Human Behavior, 37(6), 401–411. http://dx.doi.org/10.1037/lhb0000030. Kitchener, H., Schmidt, A. K., & Glaser, D. (1977). How persistent is post-prison success. Federal Probation, 41(1), 9–15. Kurlychek, M. C., Brame, R., & Bushway, S. D. (2007). Enduring risk? Old criminal records and predictions of future criminal involvement. Crime and Delinquency, 53(1), 64–83. http://dx.doi.org/10.1177/001112870629443. Langan, P. A., & Levin, D. J. (2002). Recidivism of prisoners released in 1994 (report no. 193427). Washington, DC: Department of Justice, Office of Justice Programs, Bureau of Justice Statistics. http://dx.doi.org/10.1176/appi.ps.53.10.1290. Lovell, D., Gagliardi, G. J., & Peterson, P. D. (2002). Recidivism and use of services among persons with mental illness after release from prison. Psychiatric Services, 53(10), 1290–1296. http://dx.doi.org/10.1176/appi.ps.53.10.1290. Martin, M. S., Dorken, S. K., Wamboldt, A.D., & Wootten, S. E. (2012). Stopping the revolving door: A meta-analysis on the effectiveness of interventions for criminally involved individuals with major mental disorders. Law and Human Behavior, 36(1), 1–12. http://dx.doi.org/10.1037/H0093963. McNiel, D. E., & Binder, R. L. (2007). Effectiveness of a mental health court in reducing criminal recidivism and violence. American Journal of Psychiatry, 164(9), 1395–1403. http://dx.doi.org/10.1176/appi.apj.2007.06101664. Moore, M. E., & Hiday, V. A. (2006). Mental health court outcomes: A comparison of rearrest and re-arrest severity between mental health court and traditional court participants. Law and Human Behavior, 30(6), 659–674. http://dx.doi.org/10.1007/ s10979-006-9061-9. Palermo, G. B. (2010). The Nevada mental health courts. International Journal of Law and Psychiatry, 33(4), 214–219. http://dx.doi.org/10.1016/j.ijlp.2010.06.002. Poythress, N. G., Petrila, J., McGaha, A., & Boothroyd, R. (2002). Perceived coercion and procedural justice in the Broward mental health court. International Journal of Law and Psychiatry, 25(5), 517–533 (doi:S0160-2527(01)00110-8). Ray, B., Dollar, C. B., & Thames, K. M. (2011). Observations of reintegrative shaming in a mental health court. International Journal of Law and Psychiatry, 34(1), 49–55. http://dx.doi.org/10.1016/j.ijlp.2010.11.008. Redlich, A.D., Steadman, H. J., Callahan, L., Robbins, P. C., Vessilinov, R., & Ozdogru, A. A. (2010). The use of mental health court appearances in supervision.

Please cite this article as: Ray, B., Long-term recidivism of mental health court defendants, International Journal of Law and Psychiatry (2014), http://dx.doi.org/10.1016/j.ijlp.2014.02.017

B. Ray / International Journal of Law and Psychiatry xxx (2014) xxx–xxx International Journal of Law and Psychiatry, 33(4), 272–277. http://dx.doi.org/10. 1016/j.ijlp.2010.06.010. Sampson, R. J., & Laub, J. H. (1993). Crime in the making: Pathways and turning points through life. Cambridge, Mass.: Harvard University Press. Silver, S. B., Cohen, M. I., & Spodak, M. K. (1989). Follow-up after release of insanity acquittees, mentally disordered offenders, and convicted felons. The Bulletin of the American Academy of Psychiatry and the Law, 17(4), 387–400. Steadman, H. J., Osher, F. C., Robbins, P. C., Case, B., & Samuels, S. (2009). Prevalence of serious mental illness among jail inmates. Psychiatric Services, 60(6), 761–765. Steadman, H. J., Redlich, A., Callahan, L., Robbins, P. C., & Vesselinov, R. (2011). Effect of mental health courts on arrests and jail days: A multisite study. Archives of General Psychiatry, 68(2), 167–172. http://dx.doi.org/10.1001/archgenpsychiatry.2010.134. Teplin, L. A. (1990). The prevalence of severe mental disorder among male urban jail detainees: Comparison with the Epidemiologic Catchment-Area Program. American Journal of Public Health, 80(6), 663–669. Teplin, L. A., Abram, K. M., & McClelland, G. M. (1994). Does psychiatric-disorder predict violent crime among released jail detainees? A six-year longitudinal study. The American Psychologist, 49(4), 335–342. http://dx.doi.org/10.1037/0003-066x.49.4.335.

7

Teplin, L. A., Abram, K. M., & McClelland, G. M. (1996). Prevalence of psychiatric disorders among incarcerated women: Pretrial jail detainees. Archives of General Psychiatry, 53(6), 505–512. http://dx.doi.org/10.1001/archpsyc.1996.01830060047007. Trestman, R. L., Ford, J., Zhang, W., & Wiesbrock, V. (2007). Current and lifetime psychiatric illness among inmates not identified as acutely mentally ill at intake in Connecticut's jails. The journal of the American Academy of Psychiatry and the Law, 35(4), 490–500. Trupin, E., & Richards, H. (2003). Seattle's mental health courts: Early indicators of effectiveness. International Journal of Law and Psychiatry, 26(1), 33–53. Ulmer, J. T. (2001). Intermediate sanctions: A comparative analysis of the probability and severity of recidivism. Sociological Inquiry, 71, 164–193. Ventura, L. A., Cassel, C. A., Jacoby, J. E., & Huang, B. (1998). Case management and recidivism of mentally ill persons released from jail. Psychiatric Services, 49, 1330–1337. Wales, H. W., Hiday, V. A., & Ray, B. (2010). Procedural justice and the mental health court judge's role in reducing recidivism. International Journal of Law and Psychiatry, 33, 265–271. Wilson, A.B., Draine, J., Hadley, T., Metraux, S., & Evans, A. (2011). Examining the impact of mental illness and substance use on recidivism in a county jail. International Journal of Law and Psychiatry, 34, 264–268.

Please cite this article as: Ray, B., Long-term recidivism of mental health court defendants, International Journal of Law and Psychiatry (2014), http://dx.doi.org/10.1016/j.ijlp.2014.02.017