Mental illness and crime: relationship and implications

CIT Can Increase Performance With Adequate Support Mental Crime

While many proponents of Crisis Intervention Team (CIT) policing suggest the many benefits from such training for patrol officers and departments (Hanafi, Bahora, Demir & Compton, 2008, p. 432), the truth is, implementing such programs and documenting results are often challenging and sometimes unsustainable given resource constraints. This study discusses these challenges and emerging differences that show crisis intervention policing works well to reduce costs and improve outcomes in specific areas, particularly less serious incidents, but not for more serious arrests where mental illness is a factor. Substance abuse complicates CIT performance, and unique officer, subject and event characteristics affect outcomes. A robust consensus has failed to emerge and often-conflicting reports claim conflicting results limited by various research constraints. CIT has shown valuable potential to reduce costs and improve outcomes but development and improvement will have to continue before this innovative response can become standardized for all applications.

The problem: Lack of resources

The CIT style of criminal justice management is built on the hypothesis that CIT can reduce the time that patrol officers spend administering contact with mental ill subjects and put officers back on the street faster. If officers had more resources, in this case time spent not dealing with mental health-driven crime and crisis events, they could provide other services and reduce serious crime more effectively, and also generate other outcomes that would benefit mentally ill consumers and improve total system efficiency (Watson, Ottati, Morabito, Draine, Kerr & Angell, 2010, p. 302). This is achieved by diverting mentally ill intervention to emergency medical systems rather than the criminal justice system, where trivial offenders often end up for lack of mental health resources (Slate, 2009, p. 21). CIT training aims to improve this balance at the patrol level before less-serious violators with mental illness are booked (Watson, Ottati, Morabito, Draine, Kerr & Angell, 2010, p. 304).

The outcomes described in the research can be framed in four general categories described in Table I, adapted from Morabito (2007, p. 1585). Table I describes cost to the officer, which is a cost to the total system and thus to society, in time the officer spends not providing other services resolving a mental-health involved contact. 'High cost' equates to more officer time processing mental health crises and violations per shift.

Table I. Four general outcomes from interaction with mentally ill subjects

Contact Outcome

Time Cost to Officer / Society

Involuntary hospitalization

Most cost

Arrest

Second highest cost

Referral where system is viable

Second least cost

No action / informal outcome

Least cost

If involuntary hospitalization and arrest are the more costly outcomes, CIT training should aim for referral and informal outcomes for certain mental-health contacts. This will shift caseload to mental health services and perhaps (hopefully) prevent recidivism, thus reducing future cost to all systems and improving quality of life for the mentally ill and the general population. This points directly to the main problem measuring CIT program outcomes that runs through all the literature, that not all contacts are the same, not all subjects are the same, and not all officers act identically. Different types of contact produce different results across the CIT research. These specific outcomes are discussed at more length below, after considering scale of potential results and barriers to implementation.

Incidence and barriers to implementation

How big of a problem is this? If officer time on the street is the resource we are trying to maximize, how broad of results can CIT deliver? Most authors describe research limitations that prevent estimating exactly how much return potentially costly intervention training programs can produce. Measuring success is constrained by the very resource shortage CIT programs attempt to address. There must be adequate mental health intake services where officers can turn over mental health crisis subjects in and out of custody instead of the criminal justice system or CIT programs demonstrate limited results (Watson, Ottati, Morabito, Draine, Kerr & Angell, 2010, p. 314). CIT officers must have confidence in these linkages or performance falls off over time.

Experts agree that jails are crowded with "frequent fliers" (Slate, 2009, p. 20) from the mentally ill population but as often point out the lack of empirical evidence as to exactly how many (Watson, Ottati, Morabito, Draine, Kerr & Angell, 2010, p. 304). Consensus is elusive (Fisher, Banks, Roy-Bujnowski, Grudzinskas, Simon & Wolff, 2010, p. 487), but estimates generally range around 10% of officer contact is with mentally ill subjects and around 6% of jail populations (Watson, Ottati, Morabito, Draine, Kerr & Angell, 2010, p. 303). This is roughly the same incidence as mental illness in the general population absent substance abuse (Schwartz & Lurigio 2007, p. 598). In nearly every study we encounter disclaimers by researchers reminding us of the scarcity of conclusive research on actually how much these figures represent (Shwartz & Lurigio 2007, p. 583, for example). Agreement is more certain that these contacts consume far more than average officer time per contact when they result in involuntary hospitalization and arrest (Morabito, 2007, p. 1584) rather than diversion or 'contact only.'

Substance abuse complicates CIT performance

Offense and recidivism by mentally ill subjects is not particularly higher than in the general population (Colins, Vermeiren, Vahl, Markus, Broekaert & Doreleijers, 2011, p. 48) except where comorbid with substance abuse (Fisher, Banks, Roy-Bujnowski, Grudzinskas, Simon & Wolff, 2010, p. 488;), particularly for substance related offenses (Colins, Vermeiren, Vahl, Markus, Broekaert & Doreleijers, 2011, p. 48). Substance abuse often hides mental illness symptoms (Slate, 2009, p. 21). This high incidence of substance abuse with mental illness causes friction when treatment facilities prohibit eligibility for services where substance abuse is a contributing factor (Watson, Ottati, Morabito, Draine, Kerr & Angell, 2010, p. 303). If diversion to mental health is less costly in officer hours than arrest, but absent substance abuse mentally ill consumers commit the same amount of serious crime as the general population, overall systemic, public savings could be achieved reducing substance abuse by the mentally ill.

The bottom line seems to be that although the majority of mentally ill subjects are not violent or that violent crime and mental illness are "unlikely to occur together" (Swartz & Lurigio, 2007, p. 598), some mental illnesses, particularly anti-social personality disorder and substance abuse disorders, lead to higher arrests for serious crime and recidivism with mental illness than in the general population (Swartz & Lurigio, 2007, p.599-600). Confounding this however is evidence mental illness may contribute to the prevalence of substance abuse disorder (Swartz & Lurigio, 2007, p. 599), which then contributes to recidivism in a loop that can sometimes be broken with treatment rather than incarceration (Colins, Vermeiren, Vahl, Markus, Broekaert & Doreleijers, 2011, p. 48).

Outcomes depend on specific factors

These results are confounded by differences in subject, officer and event characteristics. Different subjects with and without mental illness and substance abuse disorders behave in different ways in different situations and different officers apply CIT practices with varying effectiveness. Crisis intervention training is often found not to reduce the decision to arrest (Watson, Ottati, Morabito, Draine, Kerr & Angell, 2010, p. 313) for several reasons. CIT officers do refer more subjects to mental health treatment than non-CIT trained officers, but that increase comes from the 'contact only / informal outcome' category rather than the more costly alternatives of arrest and involuntary hospitalization (Watson, Ottati, Morabito, Draine, Kerr & Angell, 2010, p. 313). If, however mental illness with substance abuse results in higher recidivism (Fisher, Banks, Roy-Bujnowski, Grudzinskas, Simon & Wolff, 2010, p. 593), these referrals may mitigate future costs especially if substance abuse can be reduced. This leads back to the problem where some treatment facilities reject mental health candidates with comorbid substance abuse disorders. The result is a deadlock that shovels mentally ill consumers into the justice system and incurs the costs CIT was intended to prevent (Slate, 2009, p. 22).

Situational factors mitigate this decision to arrest. Serious crime leaves less officer discretion to employ referral or informal contact, so arrest rates remain high with or without CIT training (Morabito, 2007, p. 1585). Likewise not all subjects respond the same even within less critical situations that may otherwise have allowed for informal disposition. Resistance predicts higher rates of arrest in the total population including mentally ill consumers, but resistance by mental health subjects predicts more successful diversion to treatment by officers with CIT training than without (Watson, Ottati, Morabito, Draine, Kerr & Angell, 2010, p.305). Gender and socioeconomic factors often affect results beside or independent of mental health differences (Morabito, 2007, p.1586). If reducing the rate of arrest for mentally ill crisis calls was the only goal of CIT programs, this would not convincingly demonstrate success, but there are other goals for and beneficial outcomes from CIT operations (Watson, Ottati, Morabito, Draine, Kerr & Angell, 2010, p. 315). The National Alliance on Mental Illness asserts CIT training has reduced officer injury, SWAT deployment and time spent interacting with mental health support staff (n.d., p. 3). Interestingly, lower-conflict incidents where mental illness is indicated in the presence of a weapon generates higher referral than without (Watson, Ottati, Morabito, Draine, Kerr & Angell, 2010, p.305), although the status of these events as less-serious implies the weapon was not used in resistance or the crime would be serious and result in arrest.

Other situational factors outside particular incidents also affect the rate of arrest for all officers. Officer workload itself pushes down on the rate of arrest, where busier districts report higher rates of 'no action' on minor crimes, and referral where mental health is a factor, which allows more officer time on the street pursuing serious crime and regular duties (Morabito, 2007, p. 1584). Finally, officer characteristics generate different responses in similar scenarios, where officer comfort with or stigma against mental illness affects rates of arrest or diversion to mental health intervention (Watson, Ottati, Morabito, Draine, Kerr & Angell, 2010, p. 303). The result is that a few distinct categories of contact, specifically non-resistant, less critical events (perhaps where a weapon is present but not employed) generate diversion to mental health treatment by CIT officers more than non-trained officers but the increase comes instead of 'contact only' rather than formal intervention for serious offenses.

The "criminalization hypothesis" also oversimplifies complex results