Mental Illness and Criminal Justice

Criminal Justice Policy Review 2016, Vol. 27(1) 22 –45

© 2014 SAGE Publications Reprints and permissions:

sagepub.com/journalsPermissions.nav DOI: 10.1177/0887403414561255

cjp.sagepub.com

Article

Reconsidering the Criminalization Debate: An Examination of the Predictors of Arrest Among People With Major Mental Disorders

Ellen Ballard1 and Brent Teasdale1

Abstract It is widely accepted that individuals with major mental disorders are arrested at significantly higher rates when compared with non-disordered individuals. However, theoretical consensus regarding the cause of the arrest disparity still eludes researchers today. Two prevailing perspectives have dominated the debate— criminalization and criminality. Criminalization proponents argue the arrest disparity results from structural forces in society that have increasingly caused persons with mental disorders to come into contact with the justice system. Criminality proponents argue that the source of the disparity is the increased criminal behavior of persons with mental disorders. This study tests competing hypotheses drawn from these two perspectives. We analyze data from a sample of individuals recently released from psychiatric hospitals and a comparison sample of individuals from the same communities. Results of multivariate logistic regression models predicting arrest provide support for both perspectives. Implications of this research and suggestions for future research are discussed.

Keywords mental disorder, arrest, criminalization

1Georgia State University, Atlanta, USA

Corresponding Author: Ellen Ballard, Department of Criminal Justice and Criminology, Georgia State University, P.O. Box 4018, Atlanta, GA 30302-4018, USA. Email: [email protected]

561255CJPXXX10.1177/0887403414561255Criminal Justice Policy ReviewBallard and Teasdale research-article2014

Ballard and Teasdale 23

Introduction

As Abramson (1972) coined the phrase “the criminalization of mental illness,” evi- dence that individuals with major mental disorders are disproportionately involved in the criminal justice system has amassed (Hiday & Burns, 2010; Lurigio, 2012). Researchers have estimated that individuals with major mental disorders are 10% to 20% more likely to be arrested than non-disordered individuals (Lurigio, 2012; Markowitz, 2011; Skeem, Manchak, & Peterson, 2011). This susceptibility toward arrest is reflected in the jail and prison populations. According to Lamb, Weinberger, and Gross (2004), individuals with major mental disorders comprise approximately 10% to 15% of the inmate population in state and federal jails and prisons—a preva- lence rate Markowitz (2011) estimates to be 4 times greater than the prevalence of mental disorder in the general population. This disparate representation of individuals with major mental disorders in the criminal justice system is both disturbing and prob- lematic. Once incarcerated, compared with non-disordered individuals, individuals with major mental disorders experience longer jail stays (McPherson, 2008; Solomon & Draine, 1995), are more likely to be victimized by other inmates and staff (Kondo, 2000), are more likely to violate institutional rules (Ditton, 1999; Torrey, Kennard, Eslinger, Lamb, & Pavle, 2010), and have higher rates of recidivism (Teplin, Abram, & McClelland, 1997).

Two theoretical perspectives have been offered to explain the disproportionate criminal justice involvement of individuals with major mental disorders—the crim- inalization hypothesis and the criminality thesis (Hiday & Burns, 2010). The crimi- nalization perspective asserts, holding all else constant, individuals with major mental disorders are more likely to be arrested than non-disordered individuals (Lamb & Weinberger, 1998; Lamb, Weinberger, & DeCuir, 2002; Markowitz, 2011, 2006; Teplin, 1990). Researchers favoring criminalization explanations contend, once the general risk factors for arrest common among all offenders are controlled, extralegal factors unique to individuals with major mental disorders further increase their risk of arrest (Hiday & Burns, 2010; Hirschfield, Maschi, White, Traub, & Loeber, 2006). As applied to risk factors, the term extralegal refers to system, clini- cal, police officer level, and situational factors that extend beyond the authority of the law. As such, the ability of any of these factors to predict arrest in samples of individuals with major mental disorders is evidence this population has been criminalized.

Despite strong empirical evidence for the criminalization perspective, researchers favoring the criminality perspective have issued a number of theoretical challenges to the long-standing criminalization hypothesis. The criminality perspective maintains that individuals with major mental disorders are disproportionately drawn into the criminal justice system because they have a greater propensity for violence and criminal behavior, possess a greater number of general risk factors for arrest, and are more likely to behave disrespectfully and defiantly during encounters with police (Engel & Silver, 2001; Fisher et al., 2011; Hiday & Burns, 2010; Hirschfield et al., 2006; Junginger, Claypoole, Laygo, & Crisanti, 2006; Novak & Engel, 2005; Skeem et al., 2011).

24 Criminal Justice Policy Review 27(1)

Researchers favoring criminality explanations argue that the ability of extralegal factors to predict arrest disappears when legally relevant and encounter-level factors are controlled (Engel & Silver, 2001; Novak & Engel, 2005). Examples of factors legally relevant to the arrest incident include the nature and seriousness of the offense committed and prior criminal history. Encounter-level factors are situational factors that influence police officers’ perceptions or limit police officers’ discretion. Moreover, these legally relevant and encounter-level factors similarly predict arrest in non-disor- dered samples (Engel & Silver, 2001). Criminality proponents argue that much of the research evidencing criminalization has failed to control for legally relevant and encounter-level factors. As such, criminalization is an erroneous conclusion based on “spurious” findings (Novak & Engel, 2005, p. 498).

In this study, we revisit the often contentious criminalization versus criminality debate by testing competing hypotheses. Resolution of the debate is important because of the ability of these theoretical perspectives to shape and drive policy and treatment protocols. First we juxtapose the criminalization perspective’s primary assertion, extralegal factors inherent to and indicative of mental disorder will predict arrest when legally relevant factors are controlled, with the criminality perspective’s primary assertion, only factors considered legally relevant will predict arrest, when clinical factors are controlled. Next we compare the criminalization perspective’s assertion that the factors that predict arrest among individuals with major mental disorders dif- fer from the factors that predict arrest among non-disordered individuals with the criminality perspective’s assertion that the factors that predict arrest among non-disor- dered individuals will similarly predict arrest among individuals with major mental disorders.

Previous research exploring the validity of the theoretical arguments asserted by each of the theoretical perspectives has typically examined one side of the debate in isolation from the other side. The use of competing hypotheses allows for an unbiased comparison. To test these hypotheses, we use a sample of patients discharged from psychiatric hospitals and a comparison sample drawn from the communities in which the discharged patients reside. These data are well-suited to facilitate a direct compari- son of these competing perspectives (criminality and criminalization). We focus here on arrest, a critical decision as it is the gateway to criminal justice system. In the dis- cussion that follows, we review the literature from a number of disciplines highlight- ing theoretical assertions and review research perspectives to identify the factors that have been found to significantly predict arrests of individuals with mental disorders.

Literature Review

The Criminalization Perspective

Researchers favoring the criminalization perspective argue that deinstitutionalization, strict hospitalization criteria, criminal justice system net-widening, and the wars on drugs and crime have resulted in a number of unintended consequences that have dis- proportionately affected individuals with mental disorders (Hiday & Burns, 2010;

Ballard and Teasdale 25

Lamb et al., 2004; Lurigio, 2012). Each of these movements produced macro, system- level factors which increased the risk of arrest for individuals with mental disorders by impeding access to mental health care, increasing contact with police, and limiting police discretion during encounters with individuals with mental disorders.

Since deinstitutionalization, research has consistently demonstrated that strict emergency hospitalization criteria (Bittner, 1967; Borum, Williams Deane, Steadman, & Morrissey, 1998; Lamb et al., 2004; Teplin, 1984, 1985, 2000), fragmented and underfunded community mental health systems (Teplin, 1984, 1985, 2000), and ardu- ous, time-impeding hospital admission processes (Bittner, 1967; Borum et al., 1998; Teplin, 1984, 1985, 2000) restrict the available options police officers have when responding to situations involving individuals with major mental disorders. Accordingly, Bittner (1967) and Teplin (1985) found police officers are hesitant to pursue emergency mental health treatment. According to Green (1997), once police officers have taken an individual with major mental disorders into custody, whether the individual will satisfy a hospital’s strict criteria for admission becomes the factor most salient in their decision to seek emergency psychiatric hospitalization or make an arrest.

In many jurisdictions, discrepancies exist among the statutory conditions that require police to seek emergency mental health treatment, the legal criteria for invol- untary commitment, and the availability of appropriate community mental health ser- vices. Although there is some legal variation among states, typically police officers must seek emergency mental health treatment when individuals have (a) a demon- strated risk of harm to self or others, (b) a critical disability, and/or (c) an obvious ill- ness (Green, 1997; Teplin, 1984). Yet, only individuals that present an imminent risk of harm to themselves or to others are likely to meet hospital criteria for emergency hospitalization or commitment (Lamb et al., 2004; Markowitz, 2006; Teplin, 1984). To complicate matters, individuals with co-occurring substance abuse or dependence dis- orders, co-occurring personality disorders, or those with a history of violence and criminal behavior are often denied treatment, even when they meet the criteria for emergency treatment (Teplin, 1984; Hiday & Burns, 2010; Lamb et al., 2004). Therefore, arrest, by default, may be the only option available to police officers when individuals fail to meet the required dangerousness standard, have a co-occurring sub- stance or personality disorder, or have a history of violence or criminal behavior (Hiday & Burns, 2010; Teplin, 1984, 1985), an observation that led Teplin (1984, p. 800) to conclude that the criminal justice system is the only system that “cannot say no.” Criminal justice system net-widening is, therefore, the natural product of discrep- ancies within the mental health system that places the burden on police to operate as “street corner psychiatrists” (Teplin & Pruett, 1992, p. 139).

Primary support for the criminalization perspective is derived from research dem- onstrating the ability of clinical factors to significantly predict arrest. Clinical factors are observable and/or diagnosable factors of concern in the treatment of mental disor- der. Because clinical factors occur exclusively in individuals with mental disorders, they are inextricably linked to mental disorder. Therefore, these factors are inherently defined as extralegal. Clinical factors that significantly predict arrest include diagnosis

26 Criminal Justice Policy Review 27(1)

(Becker, Ross, Boaz, & Constantine, 2011), nature and severity of symptoms (Elbogen, Mustillo, van Dorn, Swanson, & Swartz, 2007; Wolff, Diamond, & Helminiak, 1997), medication noncompliance (Brekke, Prindle, Bae, & Long, 2001; Elbogen et al., 2007; Lamb & Weinberger, 2011), poor insight (Elbogen et al., 2007; Lamb & Weinberger, 2011), impaired functioning (Brekke et al., 2001; Swartz & Lurigio, 2007), impaired intellectual functioning (Thomas, Thomas, Burgason, & Wichinsky, 2014), co-occur- ring substance abuse or dependence disorders (Brekke et al., 2001; Clark, Ricketts, & McHugo, 1999; Elbogen et al., 2007; Hiday & Burns, 2010; Hodgins, Alderton, & Mak, 2007; Junginger et al., 2006; Lamb & Weinberger, 2011; Pandiani, Rosencheck, & Banks, 2003; Swartz & Lurigio, 2007; Tengstrom, Hodgins, Grann, Langstrom, & Kullgren, 2004; White, Chafetz, Collins-Bride, & Nickens, 2006), co-occurring per- sonality disorders (Hiday & Burns, 2010; Hodgins et al., 2007; Tengstrom et al., 2004), and a history of prior psychiatric hospitalization (Elbogen et al., 2007).

Criminalization proponents typically advocate for policy solutions that focus on fortifying the mental health system and developing targeted pre-adjudication criminal justice diversionary interventions. More specifically, criminalization proponents have recommended (a) increased funding of the mental health system (Teplin, 2000; Wolff, Fruah, Huening, Shi, & Epperson, 2013), (b) expanded community mental health (Lamb & Weinberger, 2011; Markowitz, 2011), (c) increased number of psychiatric beds available (Lamb & Weinberger, 2011; Markowitz, 2006), (d) repair fragmented mental health system (Grudzinskas, Clayfield, Roy-Bujnowski, Fisher, & Richardson, 2005; Teplin, 1984, 2000; Wolff et al., 2013), (e) collaboration between the mental health and criminal justice systems (DeMatteo, LaDuke, Locklair, & Heilbrun, 2013; Fisher et al., 2010; Fisher, Roy-Bujnowski et al., 2006; Grudzinskas et al., 2005; Lamb et al., 2002; Lamb et al., 2004; Lurigio & Watson, 2010; Teplin, 1990, 2000), (f) spe- cialized police response such as crisis intervention teams (CIT; Lamb et al., 2002; Lamb et al., 2004; Tucker, VanHasselt, Vecchi, & Browning, 2011; Watson & Angell, 2013), (g) specialized police training in the recognition of mental disorder and de- escalation techniques (Clark et al., 1999; Lamb & Weinberger, 2011; Lipson, Turner, & Kasper, 2010; Lurigio & Watson, 2010; Martinez, 2010; Teplin, 1990, 2000; Watson, Corrigan, & Ottari, 2004), and (h) screening for mental disorder at jail intake with immediate diversion to community mental health (Peterson, Skeem, Hart, Vidal, & Keith, 2010; Teplin, 1990). Criminalization proponents have also been strong advo- cates for alternatives to traditional criminal justice processing such as mental health courts (DeMatteo et al., 2013).

The Criminality Perspective

Although the narrative offered by the criminalization perspective has been widely accepted by many scholars and practitioners, others have been less enthusiastic, even skeptical, of its most basic assertions. Those challenging the criminalization perspec- tive acknowledge disparities in arrest rates between those with and those without major mental disorders. However, they reject system-level explanations which attri- bute arrest disparities to deinstitutionalization and declining availability of psychiatric

Ballard and Teasdale 27

hospitals and community psychiatric services while ignoring changes within the crim- inal justice system itself. For example, Frank and Glied (2006) suggested, rather than evidence of criminalization, observed increases in the arrest rate of individuals with major mental disorders post-deinstitutionalization are indicative of the explosive growth of the criminal justice system stemming from “tough on crime” policies. That is, increases in arrest rates of individuals with major mental disorders are proportion- ate to the expansion of the criminal justice system itself and these increases post- deinstitutionalization have not been uniquely experienced by individuals with major mental disorders. Rather, greater numbers of individuals, with and without major men- tal disorders, have realized increases in arrest rates post-deinstitutionalization (Frank & Glied, 2006).

Criminality proponents argue the emphasis criminalization researchers have placed on clinical factors is misguided. In police encounters involving individuals with major mental disorders, Bittner (1967, p. 279) found that police officers will only take offi- cial action when the situation represents a “serious police matter”—one in which there is a significant threat to person or property or when police action is required to main- tain or restore public order. This suggests that the most salient factors influencing officer arrest decisions are those that are legally relevant and occur at the encounter level. Furthermore, because of their legal relevance, these factors should similarly predict arrest among non-disordered individuals. Researchers favoring the criminality perspective contend the significance of a mental disorder diagnoses disappears when these legally relevant and encounter-level factors are controlled. More directly, when treated as a suspect characteristic, mental disorder fails to significantly predict arrest during police encounters, holding all else constant (Engel & Silver, 2001).

In one test of the criminality thesis, Engel and Silver (2001) find individuals with mental disorders are less likely to be arrested than non-disordered individuals once legally relevant and encounter-level factors are controlled. Furthermore, they find the factors most salient to police arrest decisions for individuals with major mental disor- ders similarly predict arrest for non-disordered individuals. Novak and Engel (2005) find that the disproportionate arrest rates of individuals with major mental disorders are a result of their intoxicated or disorderly demeanor during encounters with the police. In their study of police encounters, they find individuals with major mental disorders are more likely to threaten or use violence, be intoxicated, and behave disre- spectfully and defiantly toward police (Novak & Engel, 2005). Similarly, Fisher and his colleagues (2011) find the strongest predictor of arrest for individuals with mental disorders was assault or battery on police officers during the police encounter.

A review of the literature reveals that a number of legally relevant and encounter- level factors have been shown to similarly predict arrest among individuals with and without mental disorders. These factors include the nature and severity of offense (Engel & Silver, 2001; Green, 1997; Novak & Engel, 2005), prior criminal history (Godfredson, Ogloff, Thomas, & Luebbers, 2010; Green, 1997; Hodgins et al., 2007), criminal thinking patterns (Peterson et al., 2010), suspect demeanor (Engel & Silver, 2001; Fisher et al., 2011; Novak & Engel, 2005), substance intoxication (Engel & Silver, 2001; Novak & Engel, 2005), public location of encounter (Engel & Silver,

28 Criminal Justice Policy Review 27(1)

2001; Novak & Engel, 2005), police-initiated encounter (Engel & Silver, 2001), offender known to victim (Engel & Silver, 2001), and victim requests for arrest (Engel & Silver, 2001; Novak & Engel, 2005).

It has also been argued that individuals with major mental disorders are dispropor- tionately arrested because they disproportionately engage in violence and crime (Engel & Silver, 2001; Hiday & Burns, 2010; Hirschfield et al., 2006; Novak & Engel, 2005; Skeem et al., 2011). The criminalization hypothesis suggests that police, unfamiliar with mental disorder, often confuse disordered behavior for criminal behavior, and as a result criminality researchers erroneously conclude that individuals with major men- tal disorders commit more crime. However, criminality researchers contend the higher level of offending by individuals with major mental disorders is less attributable to symptomatology and better explained by risk factors for offending. When compared with non-disordered offenders, offenders with mental disorders possess a greater num- ber of general risk factors for violence and criminal behavior (Skeem et al., 2011). Researchers favoring the criminality perspective have attributed this increased risk to a number of factors that similarly predispose non-disordered individuals to violence and crime including criminogenic environments (Fisher, Roy-Bujnowski et al., 2006; Silver, 2000), criminogenic needs (Peterson et al., 2010), sociodemographic variables (Fisher & Drake, 2007), psychosocial factors (Skeem et al., 2011), or personality char- acteristics (Hodgins et al., 2007; Tengstrom et al., 2004).

Previous research has identified a number of factors which increase the risk of arrest among individuals with major mental disorders that similarly predict arrest among non-disordered persons. Individuals with major mental disorders who are at the greatest risk of arrest are typically males (Becker et al., 2011; Clark et al., 1999; Wolff et al., 1997), young-adults (Brekke et al., 2001; Clark et al., 1999; Pandiani et al., 2003; Wolff et al., 1997), non-White (White et al., 2006), have low socioeconomic status (Elbogen et al., 2007; Fisher, Roy-Bjnowski et al., 2006; Wolff et al., 1997), have a history of violence (Monahan et al., 2001; Swartz & Lurigio, 2007) and/or victimization (Brekke et al., 2001; Hiday & Burns, 2010; White et al., 2006; Wolff et al., 1997), are homeless or transient (Becker et al., 2011; Brekke et al., 2001; Clark et al., 1999; White et al., 2006), and live in disorganized, urban areas (Clark et al., 1999; Hiday & Burns, 2010; Wolff et al., 1997).

Criminality proponents have advocated for policies that focus on preventative pro- gramming and back-end correctional diversionary interventions that address individu- als with major mental disorders’ risk heterogeneity, co-occurring socioeconomic, psychosocial, and substance abuse problems that contribute to their increased criminal offending (Becker et al., 2011; Fisher et al., 2010; Fisher, Silver, & Wolff, 2006; Lurigio, 2012; Peterson et al., 2010; Wolff et al., 2013; Wolff, Maschi, & Bjerklie, 2004). More specifically, criminality proponents have recommended (a) mental health treatment providers incorporate screening into their treatment programs to identify general risk factors for offending (Clark et al., 1999), (b) mental health providers ame- liorate increased risk through treatment and/or referral to social services (Becker et al., 2011; Fisher et al., 2010; Fisher, Silver, & Wolff, 2006; Lurigio, 2012; Peterson et al., 2010; Wolff et al., 2013; Wolff et al., 2004), (c) back-end diversionary programs such

Ballard and Teasdale 29

as mental health courts (DeMatteo et al., 2013), (d) correctional strategies such as outpatient commitment and intensive case management (DeMatteo et al., 2013; Lamb & Weinberger, 2011; Litschge & Vaughn, 2009; Markowitz, 2006; Peterson et al., 2010), and (e) reentry strategies that address general risk factors of offending (DeMatteo et al., 2013). Criminality proponents have similarly advocated for many of the same policy recommendations advanced by criminalization proponents, such as the expansion of community mental health, collaboration between the mental health and criminal justice systems, specialized police response, and specialized police train- ing (Engel & Silver, 2001; Fisher, Silver, & Wolff, 2006; Peterson et al., 2010).

The Current Study

In sum, the criminalization perspective argues that deinstitutionalization, strict hospi- talization criteria, criminal justice system net-widening, and the wars on drugs and crime have created a significant arrest disparity between individuals with major men- tal disorders and non-disordered individuals. According to the criminalization per- spective, the factors most predictive of arrest for individuals with major mental disorders are extralegal factors indicative of and inherent to mental disorder. Furthermore, because these extralegal factors are inextricably linked to mental disor- der, the factors that predict arrest for individuals with major mental disorders will differ from those that predict arrest in the general population.

The criminality perspective argues that, when compared with non-disordered indi- viduals, individuals with major mental disorders are disproportionately arrested because they have a greater propensity for violence and criminal behavior, possess a greater number of general risk factors for arrest, and are more likely to behave disre- spectfully and defiantly during encounters with police. According to the criminality perspective, when legally relevant and encounter-level factors are controlled, extrale- gal factors fail to predict arrest among individuals with major mental disorders. Moreover, the factors that predict arrest among individuals with major mental disor- ders are the same factors that predict arrest in the general population.

Based on these assertions, we propose the following competing hypotheses:

Hypothesis 1a: Net of controls, extralegal factors (i.e., symptomatology and impaired intellectual functioning) will predict arrest. Hypothesis 1b: Net of controls, legally relevant factors (i.e., prior criminal history, violence, and drug or alcohol use) will predict arrest. Hypothesis 2a: The factors that predict arrest among individuals with major men- tal disorders differ from the factors that predict arrest in non-disordered populations. Hypothesis 2b: The factors that predict arrest in non-disordered populations will similarly predict arrest among individuals with major mental disorders.

If the criminalization hypothesis is correct, we expect extralegal factors to predict arrest when legally relevant factors are controlled. In addition, we expect the factors

30 Criminal Justice Policy Review 27(1)

that predict arrest among individuals with major mental disorders to differ from the factors that predict arrest among non-disordered individuals. If, however, the criminal- ity thesis is correct, we expect only those factors considered legally relevant to predict arrest net of controls. Moreover, the factors that predict arrest among non-disordered individuals will similarly predict arrest among individuals with major mental disorders.

Method

To assess the competing hypotheses, we analyzed data from the MacArthur Violence Risk Assessment Study (MacRisk; for a full description of the study, see Monahan et al., 2001; Silver, Mulvey, & Monahan, 1999; Steadman et al., 1998). The MacRisk is a multi-wave study of violence perpetrated by individuals recently discharged from acute psychiatric hospitals at three sites (Pittsburgh, Pennsylvania; Kansas City Missouri; and Worcester, Massachusetts). Study participants were interviewed in the hospital and, on discharge, re-interviewed at each of five follow-ups spaced approxi- mately 10 weeks apart. Additional data were collected from patient charts, collateral interviews, and official records.

To provide comparative context, the University of Pittsburgh’s Center for Social and Urban Research drew a non-patient, community sample from the neighborhoods in which patients discharged from the Pittsburgh facility resided during the 1-year follow-up period. Participants in the community sample were administered the same assessment instruments administered to participants in the patient sample. Unlike participants in the patient sample, participants in the community sample were inter- viewed only once. Questions reference behavior and events occurring in the preced- ing 10 weeks. Additional data sources included collateral interviews and official records.

Patient Sample

Patients were selected for inclusion in the MacRisk using a stratified random sample of all eligible psychiatric admissions at each of the three facilities. The sample was stratified by race, gender, and age. Eligibility requirements included (a) civil admis- sion, (b) 18 to 40 years of age, (c) English-speaking, (d) White or African American (Hispanic patients were also eligible at the Worcester facility), and (e) chart diagnosis of a psychotic disorder, major mood disorder, substance abuse or dependence disorder, or a personality disorder. Participants were continuously enrolled beginning in 1992 and ending in 1994. Follow-up interviews with later enrolled participants concluded in 1995. Of the 1,695 patients approached to participate in the study, 71% (n = 1,203) consented to participate and 67% (n = 1,136) completed baseline interviews (Appelbaum, Robbins, & Monahan, 2000). In regard to follow-up interviews, 83.7% (n = 951) completed at least one follow-up interview, 72% (n = 818) completed three or more follow-up interviews, and 60% (n = 563) completed all five follow-up inter- views (Appelbaum et al., 2000).

Ballard and Teasdale 31

Community Sample

The community sample is comprised of randomly selected individuals residing in the neighborhoods patients resided during the year following their discharge. A list of all census tracts corresponding to the locations in which patients resided during the fol- low-up period was created. Participants in the community sample were proportion- ately sampled from the patient-matched census tracts (for a detailed description of the sampling strategy used in community sample, see Steadman et al., 1998). To be eligi- ble for inclusion, participants in the community sample must have resided at their address for at least 2 months, be between the ages 18 and 40, and identify as White or African American. All interviews with participants in the community sample were conducted between March and December, 1995. In total, 519 community interviews were completed (Silver, 2000).

Measures

Arrest. Arrest information was based on participants’ self-reports. In the patient sam- ple, participants were asked if they had been arrested since the date of last contact at each follow-up (approximately 10-week intervals for 1 year post-discharge). Responses were coded as 1 (yes) and 0 (no). Participants who reported at least one arrest during any follow-up were coded as 1. In the community sample, participants were asked if they had been arrested in the previous 2 years. Responses were coded as 1 (yes) and 0 (no).

Verbal IQ. Participants in the patient and community samples were administered the vocabulary subtest of the Wechsler’s Adult Intelligence Scale–Revised (WAIS-R). The vocabulary subtest is a 35-item instrument and has been shown to strongly cor- relate with the full IQ score (DeLisi, Vaughn, Beaver, & Wright, 2010). The raw scores for each item were scaled to form a total raw score and is used in the following analy- ses to indicate Verbal IQ. The scale ranges from 1 to 70. In both samples, the sample mean was imputed for missing values. Prior to imputation, 19.1% (n = 217) of the patient sample and 28.3% (n = 174) of the community sample had a missing value for this measure.

Victimization. In both the patient and community samples, victimization was based on participants’ self-reports to the following items: (a) “Has anyone thrown some- thing at you?” (b) “Has anyone pushed, grabbed, or shoved you?” (c) “Has anyone slapped you?” (d) “Has anyone kicked, bitten, or chocked you?” (e) “Has anyone hit you with a fist or object or beaten you up?” (e) “Has anyone threatened you with a knife or gun or other lethal weapon?” and (f) “Has anyone used a knife or fired a gun at you?” Responses were collapsed into a dichotomous variable indicating the participant experienced a victimization and was coded as 1 (yes) and 0 (no). In the patient sample, participants who reported any victimization at any follow-up were coded as 1.

32 Criminal Justice Policy Review 27(1)

Principle diagnosis. In the patient sample, participants’ principle diagnosis is included in the following patient sample analyses to control for the differential effects of disor- der types. Participants were categorized into one of four broad disorder categories based on the principle diagnosis given during the baseline interview. Disorder catego- ries include psychotic disorders, bipolar disorder, major depression disorders, and sub- stance abuse and dependence disorders. Diagnoses were based on Diagnostic and Statistical Manual of Mental Disorders (3rd ed.; DSM-III; American Psychiatric Asso- ciation [APA], 1980) criteria. Dummy variables were created for each disorder type and coded as 1 (yes: principal diagnosis) and 0 (no: not principal diagnosis).

Treatment. To control for the possibility that participants in the community sample have been diagnosed with a mental disorder, information regarding current mental health and/ or substance abuse treatment is included in the following community sample analyses. Treatment is coded as 1 (currently receiving treatment) and 0 (not receiving treatment).

Threat/control override (TCO) delusions. In both the patient and community samples, delusions were assessed using the Diagnostic Interview Schedule (DIS) and the MacArthur-Maudsley Assessment of Delusions Schedule. Delusions were coded for content. Delusions that involved themes of persecution were coded as threat delusions. These included the belief that people were spying on the participant, the belief that people were following the participant, the belief that the participant was being secretly tested or experimented on, or the belief that someone was plotting against the partici- pant or trying to harm them. Delusions that involved themes of body or mind control or thought broadcasting were coded as control-override delusions. These included the belief that strange thoughts or thoughts that were not the participant’s were “being put directly into their mind,” the belief that “someone or something could take or steal thoughts” from the participant’s mind, the belief that “special messages were being sent to the subject through the television or radio,” or the belief that “strange forces were working on (the subject), as if (he or she) was being hypnotized or magic was being performed on (him or her), or (he or she) was being hit by x-rays or laser beams.” A dichotomous variable was created to indicate the presence of TCO delusions and was coded as 1 (TCO delusions reported) and 0 (TCO delusions not reported). In the patient sample, participants who reported experiencing a TCO delusion at any follow- up were coded as 1.

Hallucinations. Hallucination information is based on participants’ self-reports. As part of a structured clinical interview, participants in both the patient and community sam- ples were asked whether they had experienced any auditory hallucinations. Specifi- cally, participants were asked, “In the last 10 weeks, have you more than once had the experience of hearing things or voices other people couldn’t hear?” A dichotomous variable was created to indicate the presence of hallucinations was created and was coded as 1 (hallucination reported) and 0 (hallucination not reported). In the patient sample, participants who reported experiencing a hallucination at any follow-up were coded as 1.

Ballard and Teasdale 33

Violence. In both the patient and community samples, violence was based on partici- pants’ self-reports to the following items: (a) “Have you thrown something at any- one?” (b) “Have you pushed, grabbed, or shoved anyone?” (c) “Have you slapped anyone?” (d) “Have you kicked, bitten, or chocked anyone?” (e) “Have you hit anyone with your fist or object or beaten anyone up?” (f) “Have you threatened any- one with a knife or gun or other lethal weapon?” (g) “Have you used a knife or fired a gun at anyone?” and (h) “Have you done anything else that might be considered violent? Responses were collapsed into a dichotomous variable indicating the par- ticipant committed violence and coded as 1 (yes) and 0 (no). In the patient sample, participants who reported any violent offending at any follow-up were coded as 1.

Antisocial personality disorder. Antisocial personality disorder was assessed using the Structured Interview for Diagnostic and Statistical Manual of Mental Disorders (3rd ed., rev.; DSM-III-R; APA, 1987) Personality (SIDP-R) in both the patient and community samples. The SIDP-R is a semi-structured interview used to assess the presence of Axis II disorders (First, Spitzer, Gibbon, & Williams, 1995). Antisocial personality disorder is coded as 1 (present) and 0 (absent). Antisocial personality disorder is included as a proxy measure for suspect demeanor/behavior.

Alcohol use. Information regarding alcohol use is based on participant’s self-reports. In both the patient and community sample, participants were asked whether they have had any alcoholic drinks in the preceding 10 weeks. Responses were coded as 1 (yes) and 0 (no). In the patient sample, participants who reported consuming an alcoholic drink at any follow-up were coded as 1.

Drug use. Information regarding alcohol use is based on participant’s self-reports. In both the patient and community sample, participants were asked “have you used any street drugs, even if it was just one time” in the preceding 10 weeks. Responses were coded as 1 (yes) and 0 (no). In the patient sample, participants who reported using drugs at any follow-up were coded as 1.

Age. In both the patient and community sample, age is indicated in years.

Sex. In both the patient and community samples, sex was coded as 1 (male) and 0 (female).

Race. A dichotomous race variable was created in both the patient and community samples and coded as 1 (non-White) and 0 (White).

Data Analysis and Results

Sample Characteristics

Frequencies of responses and descriptive statistics were calculated for both the patient and community samples to examine and compare the distribution of responses across

34 Criminal Justice Policy Review 27(1)

all variables. As shown in Table 1, 29.3% (n = 279) of the 951 participants in the patient sample experienced at least one arrest during the 1 year post-discharge. Participants in the community sample experienced substantially fewer arrests. Only 7.4% (n = 37) of the 503 participants in the community sample experienced an arrest in the 2 years preceding the date of their interview. Likewise, a greater proportion of participants in the patient sample (61.0%, n = 580) reported experiencing at least one physical victimization. In contrast, only 18.3% (n = 92) of the participants in the com- munity sample reported experiencing a physical victimization. As expected, partici- pants in the patient sample more often reported experiencing hallucinations (26.2%, n = 249). Participants in the community sample reported experiencing far fewer hallucina- tions (3.8%, n = 19). Interestingly, participants in both samples similarly reported experiencing at least one TCO delusion. In the patient sample, 28.9% (n = 275) of participants reported a TCO delusion compared with 21.7% (n = 109) of participants in the community sample.

A substantial portion of the participants in the patient sample (87.5%, n = 837) were assessed as having antisocial personality disorder. Far fewer participants in the com- munity sample were assessed as having antisocial personality disorder (43.1%,n = 217). A greater number of participants in the patient sample (55.0%, n = 523) indicated they had engaged in some form of physical violence in the past year than participants in the community sample (18.9%, n = 95). As expected, a greater proportion of partici- pants in the patient sample (38.2%, n = 363) had a history of prior arrests in compari- son with participants in the community sample (16.5%, n = 83). Participants in the patient and community samples similarly reported alcohol use. In the patient sample, 80.3% (n = 764) indicated they used alcohol. In the community sample, 74.4% (n = 374) of participants indicated they used alcohol. More than twice the number of par- ticipants in the patient sample reported drug use (49.2%, n = 468) than did participants in the community sample (23.9%, n = 120).

Table 1. Comparison of Patient and Community Sample Characteristics.

Patient sample (n = 951) Community sample (n = 503)

% Arrested 29.3 7.4 % Victimized 61.0 18.3 % TCO delusions 28.9 21.7 % Hallucinations 26.2 3.8 % Violence 55.0 18.9 % Antisocial 87.5 43.1 % Prior arrest 38.2 16.5 % Used alcohol 80.3 74.4 % Used drugs 49.2 23.9 % Male 57.6 38.2 % Non-White 31.2 40.4

Note. TCO = threat/control override.

Ballard and Teasdale 35

There were substantially more males in the patient sample (57.6%, n = 548) than in the community sample (38.2%, n = 192). Proportionately, there were more non-White partici- pants in the community sample (40.4%, n = 203) than in the patient sample (31.2%, n = 297). Participants in the patient and community samples were similarly aged with partici- pants in the patient sample slightly younger on average (µ = 29.74, σ = 6.24) than partici- pants in the community sample (µ = 31.04, σ = 6.17). The average Verbal IQ score was lower among participants in the patient sample (µ = 34.45, σ = 16.48) than among partici- pants in the community sample (µ = 46.86, σ = 13.84). This was expected as mental dis- order is associated with lower IQ (Rajput et al., 2011). In the patient sample, frequencies for principal diagnosis are not shown, but were calculated and are as follows: (a) psy- chotic disorders = 21.1%, n = 201; (b) bipolar and related disorders = 12.5%, n = 119; (c) depressive disorders = 42%, n = 399; and (d) substance abuse disorders = 22.3%, n = 212. In the community sample, 10.7% (n = 53) of the participants reported receiving treatment for a current mental health and/or substance abuse issue (not shown).

Bivariate Analysis

Table 2 compares the Pearson’s correlation coefficient and statistical significance of each predictor variable for the patient and community samples. The Pearson’s correla- tion coefficients were compared as a preliminary test of the competing hypotheses and to understand each predictor variable’s unique association with the outcome variable. In the patient sample, Verbal IQ, victimization, violence, prior history of arrest, alco- hol and drug use, sex, and race were significantly associated with arrest, at the bivari- ate level. TCO delusions, hallucinations, antisocial personality disorder, and age were not significantly associated with arrest. In the community sample, Verbal IQ, TCO delusions, antisocial personality disorder, prior history of arrest, drug use, sex, and race were significantly associated with arrest. Victimization, hallucinations, violence, alcohol use, and age were not significantly associated with arrest.

Taken together, the results of this preliminary test appear consistent with both the criminalization and criminality perspectives. In support of the criminalization hypoth- esis, two extralegal factors, Verbal IQ and victimization, were significantly associated with arrest in the patient sample and victimization was not significantly associated with arrest in the community sample. However, as the criminality thesis maintains, a number of factors were similarly associated with arrest in both samples including prior history of arrest, r, and drug use.

Multivariate Models

To fully test the competing hypotheses, multivariate logistic regression models were estimated for both the patient and community samples. Standardized odds ratios are provided to facilitate interpretation and better compare the relative importance of each predictor variable (a one standard deviation change in x [predictor variable], multiply the odds of arrest [the outcome variable] by the standardized odds ratio value, holding all else constant; Menard, 2004).

36 Criminal Justice Policy Review 27(1)

Table 3 compares the patient and community models predicting arrest. In the patient model, several factors significantly predicted arrest. These include Verbal IQ, victim- ization, violence, prior history of arrest, drug use, and sex. All else controlled, a one standard deviation increase in Verbal IQ decreases an individual’s odds of arrest by 24.5%. Individuals who reported experiencing a physical victimization have approxi- mately twice the odds of experiencing an arrest than individuals who did not report experiencing a physical victimization, net of controls. Individuals who reported engag- ing in some form of physical violence have odds of arrest approximately 1.7 times the odds of individuals who did not report engaging in any physical violence, all else controlled. Likewise, individuals with a prior history of arrest have odds of arrest approximately 1½ the times of individuals with no prior arrest history, net of controls. Individuals who reported drug use have approximately 150% the odds of arrest when compared with individuals with no reported drug use, all else controlled. Net of con- trols, the odds of arrest for males are almost twice the odds of arrest for females.

Fewer factors significantly predicted arrest in the community model. The factors include violence, antisocial personality disorder, drug use, and race. In the community model, violence is inversely associated with arrest. All else controlled, individuals who reported engaging in some form of violence have 88% of the odds of those who did not report engaging in any violence for arrest. Individuals assessed as antisocial personality disorder have more than 7 times the odds of arrest when compared with individuals who were not assessed as antisocial personality disorder, net of controls. Individuals who reported drug use have more than 5 times the odds of arrest when compared with individuals reporting no drug use all else controlled. Net of controls, non-Whites have almost 3 times the odds of arrest than Whites.

Table 2. Comparison of Patient and Community Sample Bivariate Associations With Arrest.

Patient sample Community sample

r p r p

Verbal IQ −.178 .001*** −.128 .004** Victimization .245 .001*** .064 .154 TCO delusions −.014 .674 .111 .013* Hallucinations .037 .261 −.016 .722 Violence .235 .001*** −.058 .193 Antisocial .020 .530 .247 .001*** Prior arrests .197 .001*** .182 .001*** Alcohol use .127 .001*** .061 .173 Drug use .290 .001*** .307 .000*** Age −.036 .261 .011 .803 Sex −.160 .001*** .186 .001*** Race .169 .001*** .187 .001***

Note. TCO = threat/control override. *p ≤ .05. **p ≤ .01. ***p ≤ .001 (two-tailed).

Ballard and Teasdale 37

Discussion

Substantively, the results reported in Table 3 appear to lend some support to both the criminalization and criminality perspectives. Hypothesis 1b stated, net of controls, legally relevant factors such as prior criminal history, violence, and suspect demeanor will predict arrest. Criminality proponents would argue that the significance of vio- lence, prior arrests, and drug use in the patient model are indicative of individuals with major mental disorders’ greater propensity for violence and criminal behavior, result- ing in their arrests. However, the results appear to refute Hypothesis 2b which states that the factors that predict arrest in non-disordered populations will similarly predict arrest among individuals with major mental disorders. To fully support the criminality paradigm, factors that predict arrest in the patient model should similarly predict arrest in the community model. As shown, only drug use (patient model e^bStdX for drug use = 1.556; community model e^bStdX for drug use = 2.031) and gender (patient model e^bStdX for male = 1.404; in the community model e^bStdX for male = 1.451) similarly predicted arrest.

Hypothesis 1a states, net of controls and extralegal factors, factors indicative of and inherent to mental disorder, will predict arrest among individuals with major mental disorders, is partially supported. Although legally relevant factors such as violence, prior history of arrest, and drug use significantly predicted arrest, two extralegal fac- tors, verbal IQ and victimization, remained significant in the patient model. Moreover,

Table 3. Comparison of Patient and Community Logistic Regression Models Predicting Arrest.

Patient model Community model

b SE OR e^bStdX b SE OR e^bStdX

Verbal IQ −0.019 0.006 0.981*** 0.755 −0.025 0.017 0.975 0.749 Victimization 0.784 0.205 2.191*** 1.466 0.827 0.547 2.287 1.377 TCO delusion −0.376 0.194 0.686 0.843 0.068 0.455 1.070 1.028 Hallucination 0.053 0.192 1.054 0.977 −1.376 1.115 0.253 0.769 Violence 0.529 0.197 1.698** 1.301 −2.113 0.723 0.121** 0.437 Antisocial −0.068 0.248 0.934 1.029 1.995 0.563 7.349*** 2.690 Prior history 0.519 0.167 1.680** 1.287 −0.002 0.473 0.998 0.999 Alcohol se 0.018 0.243 1.018 1.008 0.226 0.558 1.254 1.104 Drug use 0.899 0.177 2.456*** 1.556 1.659 0.444 5.252*** 2.031 Age −0.005 0.013 0.995 0.969 −0.021 0.036 0.979 0.880 Male 0.688 0.171 1.990*** 1.404 0.758 0.442 2.134 1.451 Race 0.360 0.181 1.434 1.181 1.091 0.476 2.977* 1.712 Constant −2.466 0.581 0.085 — −3.845 1.450 0.021 —

Note. In a separate model (not shown), type of disorder was included in the patient model. Only substance dependence disorders significantly predicted arrest. Psychotic disorders, bipolar and related disorders, and major depressive disorders did not significantly predict arrest above and beyond characteristic symptoms of the disorders (i.e., TCO delusions, hallucinations, substance use) already included in the models shown. TCO = threat/control override. *p ≤ .05. **p ≤ .01. ***p ≤ .001.

38 Criminal Justice Policy Review 27(1)

verbal IQ appears to be the most salient predictor of arrest among individuals in the patient sample. According to criminalization proponents, the ability of verbal IQ and victimization to predict arrest is evidence that individuals with major mental disorders are criminalized. Hypothesis 2a, the factors that predict arrest among individuals with major mental disorders differ from the factors that predict arrest in non-disordered populations, appears to be fully supported by the results. Verbal IQ and victimization did not predict arrest in the community model, and antisocial personality disorder did not predict arrest in the patient model. Violence significantly predicted arrest in both patient and community models, however the association was directionally different. In the patient model, violence significantly predicted arrest, whereas violence was inversely associated with arrest in the community model.

In current form, neither perspective comprehensively addresses all of the causal factors that contribute to the disproportionate criminal justice system involvement of individuals with major mental disorders. The findings suggest that criminalization proponents should consider how and in what ways legally relevant and encounter- level factors interact and/or exacerbate the clinically significant factors that increase the risk of arrest for individuals with major mental disorders. That is, research and programming aimed at reducing the criminalization of mental disorder should simul- taneously address the underlying criminogenic needs and criminal thinking that inten- sify this population’s already elevated risk of arrest. The study also identified that arrest and disorder do not occur in isolation. That is, disordered behavior is symptom- atic of mental disorder and cannot be separated from the disorder. Research examining individuals with major mental disorders’ propensity toward offending should be mind- ful not to minimize or trivialize the effect of mental disorder on behavior. Moreover, criminality proponents should consider how the behavioral result of impaired func- tioning and cognition influences police encounters. More directly, how and in what ways individuals’ mental disorder tax their ability to cope and actively problem solve.

That victimization and impaired intellectual functioning both predicted arrest sug- gests the need for continued and perhaps increased training of police officers in how to deal with citizens in acute decompensation situations. Specifically, the association of verbal IQ and arrest is interesting here, as is the association between victimization and arrest. These two findings coupled together suggest that disordered citizens lack the ability to adequately communicate their situations to the police officers who respond to those incidents. The result of this inability to effectively communicate is the arrest of disordered victims and those with low IQ. If disordered individuals are provided with extra time to convey their situation and the calm required for them to do so, their risks of arrest may be lessened. This is an important finding for police officers to understand as they interact with disordered individuals.

Policy Implications

In light of the findings of this study, policies aimed at reducing the disproportionate arrest rate of individuals with major mental disorders should envelop both theoretical perspectives. Moreover, and as evidenced by the mixed support for both the

Ballard and Teasdale 39

criminalization hypothesis and the criminality thesis found here, it is imperative that these policies attack the disparity on multiple fronts. To address those individuals who are the focus of criminalization researchers—individuals whose criminal justice involvement is the result of symptom-driven behavior—policy should focus on front- end diversionary measures. Police administrators should consider mandating the expansion of current CIT programs, so that all police officers receive similar special- ized training. Additional training may assist officers with dealing with individuals with low verbal abilities or who are in acute trauma states. Research has demonstrated that officers who receive this specialized training (a) more often recognize the signs of mental illness, (b) more effectively resolve encounters with individuals with major mental disorders, (c) respond more compassionately to individuals with major mental disorders, (d) reduce criminal justice system involvement, (e) reduce officer/suspect injury during encounters, and (f) share a common vocabulary with mental health pro- viders (Fisher, Silver, & Wolff, 2006; Lamb & Weinberger, 2011; Lipson et al., 2010; Lurigio & Watson, 2010; Watson & Angell, 2013). In addition, mental health provid- ers have indicated more confidence in the judgment of officers who have received specialized training (Lamb et al., 2004).

Although specialized police training has proven to be effective, it is inevitable that some individuals with mental disorders will fall through the cracks. Jails can provide a safety net by screening for mental disorder at intake (Peterson et al., 2010). For individu- als whose disorder went unrecognized by the police and whose symptoms drive their behavior, intake screening provides a second line of defense. These individuals can be identified at intake and immediately diverted to a community-based alternative. It cannot be emphasized enough that even mandated specialized police training for all police offi- cers and increased mental health screening during jail intake will likely fail to reduce the disproportionate arrest of individuals with major mental disorders without the support and collaboration of the mental health system. To be effective, police must have 24-hr access to community-based alternatives to arrest. Preferably, 24-hr access should be accompa- nied by no decline agreements between the police and community mental health provid- ers. These front-end diversionary interventions may be particularly effective for individuals who have low verbal abilities and those who have been victimized and, as a result, experience difficulty communicating and interacting with police officers.

Expansion of the CIT model and enhanced jail screening coupled with 24-hr commu- nity-based alternatives to arrest will significantly reduce the arrest rate of those individu- als most likely to be criminalized. However, as evidenced by the partial support found for the criminality thesis in this study, not all arrests of individuals with major mental disorders are the result of symptom-induced behavior. Criminalization advocates must concede that mental disorder does not occur in isolation of more general risk factors associated with criminal offending. Therefore, employment of only front-end diversion- ary measures alone will not be sufficient to remedy the disproportionate criminal justice system involvement of individuals with major mental disorders.

Front-end diversionary measures such as those mentioned above must be supple- mented with preventative programming and back-end correctional diversionary inter- ventions designed to address co-occurring problems such as homelessness, substance

40 Criminal Justice Policy Review 27(1)

abuse, inadequate coping resources, criminal thinking, criminogenic needs, antisocial behavior/personality disorder, and so on. As a preventative measure, mental health treatment providers should incorporate screening to identify general risk factors for offending. On the basis of this screening, targeted treatment plans should be developed and include remediation of any identified risk factors to reduce the likelihood of offending. Likewise, specialized correctional and back-end diversionary interventions such as intensive case management, and outpatient commitment, and mental health courts, as well as reentry strategies must address general risk factors of offending. To reduce the likelihood of reoffending, correctional and back-end diversion programs should screen for general risk factors for offending and match offenders to appropriate programs and services based on their identified risks and needs service.

The experience of a mental disorder is nuanced and complex. The disproportionate criminal justice system involvement of individuals with major mental disorders is no less complicated. As we have shown here, no one theoretical perspective is so robust as to thoroughly explain this disparity. Policy directed at reducing or eliminating the disproportionate representation of individuals with major mental disorders in the crim- inal justice system must be equally variegated to effect change. That is, to be effective, policy aimed at reducing this disparity must be multifaceted and attack the disparity on multiple fronts. It is with this common goal that the criminalization and criminality perspectives might find that their differences can be embraced to realize a comprehen- sive policy and practice strategy that has a real chance at success.

Limitations

This study is not without limitations. As noted elsewhere, the time of reference of the dependent variable and several of the independent variables differs between the patient and community samples. It is unclear how this difference would affect the findings. It is possible that the longer period of recall for the community sample increased the rate of the dependent variable and therefore power to detect effects, but it is equally plau- sible that the increased error associated with a longer recall period depressed correla- tions with the IVs. It is unclear which of these two possibilities was primary or if they cancelled each other out. Second, antisocial behavior was used as a proxy for suspect demeanor. As we did not conduct an observational study of police behavior, we had the advantage of clinically validated measures of symptoms and IQ, but we lacked interactional measures that observational studies typically include. Police observa- tional studies typically lack the sophisticated measures of symptoms and IQ that we include here. Including both types of measures should be a priority for future research. Third, the data are cross-sectional and therefore we cannot infer time order. It is pos- sible that the victimization was the consequence of the arrest, rather than the cause of the arrest; however, this temporal order seems less plausible. Future research should consider temporal ordering, and prioritize the measurement issues we identify above.

Finally, the age of the data is potentially limiting. Since the conclusion of data collec- tion (1995), there have been several revisions to the DSM. However, the diagnostic cri- teria for the disorders included in this study have remained relatively stable across DSM

Ballard and Teasdale 41

revisions. In addition, increasing interest in and awareness of the disproportionate repre- sentation of individuals with major mental disorders has spurred the development of innovative programming designed to reduce arrest disparities and mitigate the damaging effects of criminal justice system involvement can have on arrest rates of this population. For example, involuntary outpatient commitment, crisis intervention team (CIT) train- ing, and mental health courts have certainly changed the experiences of individuals with major mental disorders. However, it is unknown what impact these innovations would have on the results of this study. In spite of these limitations, the current study is the first to provide a head-to-head comparison of the criminalization and criminality perspec- tives. We believe this is an important first step in bridging the divide that separates the two perspectives and impairs collaboration between research and practice.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

Abramson, M. (1972). The criminalization of mentally disordered behavior: Possible side effects of new mental health law. Hospital and Community Psychiatry, 23, 101-105.

American Psychiatric Association. (1980). Diagnostic and statistical manual of mental disor- ders (3rd ed.). Washington, DC: Author.

American Psychiatric Association. (1987). Diagnostic and statistical manual of mental disor- ders (3rd ed., rev.). Washington, DC: Author.

Appelbaum, P. S., Robbins, P. C., & Monahan, J. (2000). Violence and delusions: Data from the MacArthur Violence Risk Assessment Study. American Journal of Psychiatry, 157, 566-572.

Becker, M. A., Ross, A., Boaz, T., & Constantine, R. (2011). Gender differences and risk of arrest among offenders with serious mental illness. Journal of Behavioral Health Services & Research, 38, 16-27.

Bittner, E. (1967). Police discretion in the emergency apprehension of mentally ill persons. Social Problems, 14, 278-292.

Borum, R., Williams Deane, M., Steadman, H. J., & Morrissey, J. (1998). Police perspec- tives on responding to mentally ill people in crisis: Perceptions of program effectiveness. Behavioral Sciences & the Law, 16, 393-405.

Brekke, J. S., Prindle, C., Bae, S. W., & Long, J. D. (2001). Risks for individuals with schizo- phrenia who are living in the community. Psychiatric Services, 52, 1358-1366.

Clark, R. E., Ricketts, S. K., & McHugo, G. J. (1999). Legal system involvement and costs for persons in treatment for severe mental illness and substance use disorders. Psychiatric Services, 50, 641-647.

DeLisi, M., Vaughn, M. G., Beaver, K. M., & Wright, J. P. (2010). The Hannibal Lecter myth: Psychopathy and verbal intelligence in the MacArthur Violence Risk Assessment Study. Journal of Psychopathology and Behavioral Assessment, 32, 169-177.

42 Criminal Justice Policy Review 27(1)

DeMatteo, D., LaDuke, C., Locklair, B. R., & Heilbrun, K. (2013). Community-based alter- natives for justice-involved individuals with severe mental illness: Diversion, problem- solving courts, and reentry. Journal of Criminal Justice, 41, 64-71.

Ditton, P. M. (1999, May). Special report: Mental health treatment of inmates and probationers (NVJ 174463). Washington, DC: U.S. Department of Justice, Bureau of Justice Statistics.

Elbogen, E. B., Mustillo, S., van Dorn, R., Swanson, J. W., & Swartz, M. S. (2007). The impact of perceived need for treatment on risk of arrest and violence among people with severe mental illness. Criminal Justice & Behavior, 34, 197-210.

Engel, R. S., & Silver, E. (2001). Policing mentally disordered suspects: A reexamination of the criminalization hypothesis. Criminology, 39, 225-252.

First, M. B., Spitzer, R. L., Gibbon, M., & Williams, J. B. (1995). Structured Clinical Interview for DSM-IV Axis I Disorders, Patient Edition, January 1995 FINAL. SCID-I/P Version 2.0). New York, NY: Biometrics Research Department, New York State Psychiatric Institute.

Fisher, W. H., Banks, S. M., Roy-Bujnowski, K., Grudzinskas, A. J., Jr., Simon, L. J., & Wolff, N. (2010). Categorizing temporal patterns of arrest in a cohort of adults with serious mental illness. Journal of Behavioral Health Services & Research, 37, 477-490.

Fisher, W.H., & Drake, R.E. (2007). Forensic mental illness and other policy misadventures. Commentary on “Extending assertive community trteatment to criminal justice settings: Origins, current evidence, and future directions.” Community Mental Health Journal, 43(5), 545-548.

Fisher, W. H., Roy-Bujnowski, K. M., Grudzinskas, A. J., Jr., Clayfield, J. C., Banks, S. M., & Wolff, N. (2006). Patterns and prevalence of arrest in a statewide cohort of mental health care consumers. Psychiatric Services, 57, 1623-1628.

Fisher, W. H., Silver, E., & Wolff, N. (2006). Beyond criminalization: Toward a criminolog- ically-informed mental health policy and services research. Administration and Policy in Mental Health and Mental Health Services Research, 33, 544-557.

Fisher, W. H., Simon, L., Roy-Bujnowski, K., Grudzinskas, A., Wolff, N., Crockett, E., & Banks, S. (2011). Risk of arrest among public mental health services recipients and the general public. Psychiatric Services, 62, 67-72.

Frank, R., & Glied, S. (2006). Mental health policy in the United States since 1950: Better but not well. Baltimore, MD: The Johns Hopkins University Press.

Godfredson, J. W., Ogloff, J. R. P., Thomas, S. D. M., & Luebbers, S. (2010). Police discretion and encounters with people experiencing mental illness. Criminal Justice and Behavior, 37, 1392-1405.

Green, T. M. (1997). Police as frontline mental health workers: The decision to arrest or refer to mental health agencies. International Journal of Law and Psychiatry, 20, 469-486.

Grudzinskas, A. J., Jr., Clayfield, J. C., Roy-Bujnowski, K., Fisher, W. H., & Richardson, M. H. (2005). Integrating the criminal justice system into mental health service delivery: The Worcester diversion experience. Behavioral Sciences & the Law, 23, 277-293.

Hiday, V. A., & Burns, P. J. (2010). Mental illness and the criminal justice system. In T. L. Scheid & T. N. Brown (Eds.), A handbook for the study of mental health: Social contexts, theories, and systems (pp. 478-498). New York, NY: Cambridge University Press.

Hirschfield, P., Maschi, T., White, H. R., Traub, L. G., & Loeber, R. (2006). Mental health and juvenile arrests: Criminality, criminalization, or compassion? Criminology, 44, 593-630.

Hodgins, S., Alderton, C. J., & Mak, T. (2007). From conduct disorder to severe mental illness: Associations with aggressive behavior, crime, and victimization. Psychological Medicine, 38, 975-987.

Ballard and Teasdale 43

Junginger, J., Claypoole, K., Laygo, R., & Crisanti, A. (2006). Effects of serious mental illness and substance abuse on criminal offenses. Psychiatric Services, 57, 879-882.

Kondo, L. L. (2000). Therapeutic jurisprudence: Issues analysis and applications: Advocacy of the establishment of mental health specialty courts in the provision of therapeutic justice for mentally ill offenders. Seattle University Law Review, 48, 296-313.

Lamb, H. R., & Weinberger, L. E. (1998). Persons with severe mental illness in jails and pris- ons: A review. Psychiatric Services, 49, 483-492.

Lamb, H. R., & Weinberger, L. E. (2011). Meeting the needs of those persons with serious mental illness who are most likely to become criminalized. The Journal of the American Academy of Psychiatry and the Law, 39, 549-554.

Lamb, H. R., Weinberger, L. E., & DeCuir, W. J. (2002). The police and mental health. Psychiatric Services, 53, 1266-1271.

Lamb, H. R., Weinberger, L. E., & Gross, B. H. (2004). Mentally ill persons in the criminal justice system: Some perspectives. Psychiatric Quarterly, 75, 107-126.

Lipson, G. S., Turner, J. T., & Kasper, R. (2010). A strategic approach to police interac- tions involving persons with mental illness. Journal of Police Crisis Negotiations, 10, 30-38.

Litschge, C. M., & Vaughn, M. G. (2009). The Mentally Ill Offender Treatment and Crime Reduction Act of 2004: Problem and prospects. The Journal of Forensic Psychiatry, 20, 542-558.

Lurigio, A. J. (2012). Responding to the needs of people with mental illness in the criminal justice system: An area ripe for research and community partnerships. Journal of Crime and Justice, 35, 1-12.

Lurigio, A. J., & Watson, A. C. (2010). The police and people with mental illness: New approaches to a longstanding problem. Journal of Police Crisis Negotiations, 10, 3-14.

Markowitz, F. E. (2006). Psychiatric hospital capacity and crime and arrest rates. Criminology, 44, 45-72.

Markowitz, F. E. (2011). Mental illness, crime, and violence: Risk, context, and social control. Aggression and Violent Behavior, 16, 36-44.

Martinez, L. E. (2010). Police departments’ response in dealing with persons with mental ill- ness. Journal of Police Crisis Negotiations, 10, 166-174.

McPherson, W. (2008). Managing the mental health population at the Broward Sheriff’s office. Corrections Today, 70, 62-67.

Menard, S. (2004). Six approaches to calculating standardized logistic regression coefficients. The American Statistician, 58, 218-223.

Monahan, J., Steadman, H. J., Silver, E., Appelbaum, P. S., Clark Robbins, P., Mulvey, E. P., . . .Banks, S. (2001). Rethinking risk assessment: The MacArthur study of mental disorder and violence. New York, NY: Oxford University Press.

Novak, K. J., & Engel, R. S. (2005). Disentangling the influence of suspect’s demeanor and men- tal disorder on arrest. Policing: International Journal of Police Strategy & Management, 28, 493-512.

Pandiani, J. A., Rosencheck, R., & Banks, S. M. (2003). Elevated risk of arrest for Veteran’s Administration health services recipients in four Florida counties. Law and Human Behavior, 27, 289-298.

Peterson, J. P., Skeem, J. L., Hart, E., Vidal, S., & Keith, F. (2010). Analyzing offense patterns as a function of mental illness to test the criminalization hypothesis. Psychiatric Services, 61, 1217-1222.

44 Criminal Justice Policy Review 27(1)

Rajput, S., Hassiotis, A, Richards, M., Hatch, S.L., & Stewart, R. (2011). Associations between IQ and common mental disorders: The 2000 British National Survey of Psychiatric Morbidity. European Psychiatry, 26(6), 390-395.

Silver, E. (2000). Extending social disorganization theory: A multilevel approach to the study of violence among persons with mental illnesses. Criminology, 38, 1043-1074.

Silver, E., Mulvey, E. P., & Monahan, J. (1999). Assessing violence risk among discharged psychiatric patients: Towards an ecological approach. Law and Human Behavior, 23, 235-253.

Skeem, J. L., Manchak, S., & Peterson, J. K. (2011). Correctional policy for offenders with mental illness: Creating a new perspective for recidivism reduction. Law and Human Behavior, 35, 110-126.

Solomon, P., & Draine, J. (1995). Issues in serving the forensic client. Social Work, 40, 25-33. Steadman, H. J., Mulvey, E. P., Monahan, J., Robbins, P. C., Appelbaum, P. S., Grisso, T.,

. . .Silver, E. (1998). Violence by people discharged from acute psychiatric inpatient facilities and by others in the same neighborhoods. Archives of General Psychiatry, 55, 393-401.

Swartz, J. A., & Lurigio, A. J. (2007). Serious mental illness and arrest: The generalized mediat- ing effect of substance use. Crime & Delinquency, 53, 581-604.

Tengstrom, A., Hodgins, S., Grann, M., Langstrom, N., & Kullgren, G. (2004). Schizophrenia and criminal offending: The role of psychopathy and substance use disorders. Criminal Justice and Behavior, 31, 367-391.

Teplin, L. A. (1984). Criminalizing mental disorder: The comparative arrest rate of the mentally ill. The American Psychologist, 39, 794-803.

Teplin, L. A. (1985). The criminality of the mentally ill: A dangerous misconception. American Journal of Psychiatry, 142, 593-599.

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, 663-669.

Teplin, L. A. (2000). Keeping the peace: Police discretion and mentally ill persons. National Institute of Justice Journal, 244, 8-15.

Teplin, L. A., Abram, K. M., & McClelland, G. M. (1997). Prevalence of psychiatric disorders among incarcerated women. Archives of General Psychiatry, 53, 505-512.

Teplin, L. A., & Pruett, N. S. (1992). Police as street corner psychiatrists: Managing the men- tally ill. International Journal of Law Psychiatry, 15, 139-156.

Thomas, J. M., Thomas, S. A., Burgason, K. A., & Wichinsky, L. C. (2014). Early contact with the criminal justice system and intellectual functioning as risk factors for violent and chronic adult offending. Western Criminology Review, 15(1), 34-50.

Torrey, E. F., Kennard, A. D., Eslinger, D., Lamb, R., & Pavle, J. (2010, May). More mentally ill persons are in jails and prisons than hospitals: A survey of the states. Arlington, VA: Treatment Advocacy Center. Retrieved from http://www.treatmentadvocacycenter.org/ storage/documents/final_jails_v_hospitals_study.pdf

Tucker, A. S., VanHasselt, V. B., Vecchi, G. M., & Browning, S. L. (2011). Responding to persons with mental illness. FBI Law Enforcement Bulletin, 80(10), 1-6.

Watson, A. C., & Angell, B. (2013). The role of stigma and uncertainty in moderating the effect of procedural justice on cooperation and resistance in police encounters with persons with mental illness. Psychiatry, Public Policy, and Law, 19, 30-39.

Watson, A. C., Corrigan, P. W., & Ottari, V. (2004). Police officers’ attitudes toward and deci- sions about persons with mental illness. Psychiatric Services, 55, 49-53.

Ballard and Teasdale 45

White, M. C., Chafetz, L., Collins-Bride, G., & Nickens, J. (2006). History of arrest, incarcera- tion, and victimization in community-based severely mentally ill. Journal of Community Health, 31, 123-135.

Wolff, N., Diamond, R. J., & Helminiak, T. W. (1997). A new look at an old issue: People with mental illness and the law enforcement system. The Journal of Mental Health Administration, 24, 152-165.

Wolff, N., Fruah, B. C., Huening, J., Shi, J., & Epperson, M. W. (2013). Practice informs the next generation of behavioral health and criminal justice interventions. International Journal of Law and Psychiatry, 36, 1-10.

Wolff, N., Maschi, T., & Bjerklie, J. R. (2004). Profiling mentally disordered prison inmates: A case study in New Jersey. Journal of Correctional Care, 11(5), 5-27.

Author Biographies

Ellen Ballard has an MS in criminal justice from Boston University. She is currently a doctoral student in the Department of Criminal Justice and Criminology at Georgia State University. She is interested in research on persons with serious mental illness in the criminal justice system, substance abuse, and stigmatized populations. Her previous work has appeared in Journal of Interpersonal Violence and Substance Abuse and Rehabilitation.

Brent Teasdale is a graduate of the Pennsylvania State University’s Crime, Law & Justice program, and an associate professor in the Department of Criminal Justice and Criminology at Georgia State University, where he is also the director of graduate studies and the editor of Social Problems Forum. He has published extensively on mental health issues and substance abuse prevention. His recent work has appeared in Social Problems, Prevention Science, Criminal Justice and Behavior, and the American Journal of Criminal Justice.