Mental health courts are a type of problem-solving court, along with drug courts, DWI courts, veterans courts, and others (Strong, Rantala, & Kyckelhahn, 2016). They are one program option to address the criminal behaviors of persons with psychiatric disorders (Skeem, Manchak, & Peterson, 2011). A study identified 337 mental health courts in the United States in 2012, which were present in 44 of 50 states, Washington DC, and United States territories (Strong et al., 2016). The effect of mental health court participation on recidivism has generally been positive, in that participation in the court reduces the likelihood of recidivism. Honegger (2015) identified 15 studies of recidivism among mental health court participants and found that 13 of 15 studies reported beneficial outcomes. In meta-analyses, Lowder, Rade, and Desmarais (2018) found that participation in mental health courts had a small effect on criminal recidivism compared to traditional courts, while Sarteschi, Vaughn, and Kim (2011) reported they had a moderate effect on reducing recidivism

Traditional Vs Mental Health Court Processes

Mental health courts differ from traditional courts in several significant ways (Redlich, Steadman, Monahan, Robbins, & Petrila, 2006). One is that mental health courts focus solely on defendants with psychiatric disorders. Also unlike traditional courts, mental health courts connect defendants to mandated community-based treatment unique to defendants’ needs. They then supervise defendants to ensure compliance with court mandates and use sanctions and rewards to influence behavior. Another difference is that participation in mental health courts is voluntary, and defendants have the option at any time in the process to return to regular court to resolve their criminal charges. In addition, unlike traditional courts that employ an adversarial system concerned with establishing guilt or innocence, mental health court have teams that function collaboratively to help defendants achieve their treatment goals while protecting public safety (Fisler, 2005; Kubiak, Comartin, Ray, & Tillander, 2018; Thompson, Osher, & Tomasini-Joshi, 2007).

Mental health court teams typically include a judge, a prosecuting attorney, a defense attorney, treatment providers, and a court administrator or case manager (Moore & Hiday, 2006; Thompson et al., 2007). Some court teams also include mental health advocates and victim advocates (Luskin, 2001; Snedker, 2018). Mental health court team meetings are typically led by the judge or the prosecuting attorney, although occasionally defense attorneys lead the meetings (Arkfeld, 2007; Snedker, 2018). Defendants are usually not present at team meetings. Although variation in duties exist across mental health courts, teams typically make referrals and admit defendants to the court, develop conditions for participation, hold case conferences prior to court hearings to review defendants’ progress since their last appearance, and recommend discharge from the court program to the judge at the appropriate time (Thompson et al., 2007; Wolff, Fabrikant, & Belenko, 2011).

Roles of Defense Attorneys in Mental Health Courts

The role of defense attorneys in traditional courts is to be a zealous advocate for the defendants they represent, ensure that defendants’ legal rights are protected; and provide high-quality legal representation (American Bar Association, 2020). Some have argued that defense attorneys play a limited role in mental health courts as a result of the collaborative team approach. For example, Casey (2004) concluded that the assistance of independent defense attorneys is relinquished when they become part of a problem-solving court team. Similarly, Hollard (2010) suggested that defense attorneys have struggled to find their place on the collaborative teams of problem-solving courts. Some empirical support exists for the limited role of defense attorneys in mental health courts. In a study of the first mental health court, researchers concluded that defense attorneys played minor roles in court hearings because communication in court is primarily between the judge and the defendant (Boothroyd, Poythress, McGaha, & Petrila, 2003). They observed that judges talked 47% of time, defendants 33%, and attorneys (defense and prosecuting attorneys) only 12%. In a study of four mental health courts, Castellano (2017) concluded that defense attorneys had diminished roles and that judges, not defense attorneys, protected the due process of defendants. Castellano (2011) also found that case managers assumed some traditional defense attorney duties such as raising reasonable doubt and objecting to court rulings, especially when judges issued negative sanctions to mental health court participants for violating court-ordered conditions. In addition, defense attorneys do not always participate in all court processes. A national study of mental health courts found that defense attorneys did not attend 18.1% of preliminary court hearings (Strong et al., 2016), while a study of seven mental health courts found that defense attorneys in two of them did not participate in case conferences held prior to court hearings (Waters, Strickland, & Gibson, 2009).

Others believe that defense attorneys can and should play important roles in mental health courts. One role of defense attorneys is to make referrals to mental health courts (McNeil & Binder, 2010; Waters et al., 2009). Studies of several mental health courts found that defense attorneys were the largest source of referrals (Snedker, Beach, & Corcoran, 2017; Wolff et al., 2011). A second role of defense attorneys is to assist referred defendants in deciding whether to participate in the mental health court program. This is particularly important because studies have identified that some mental health court participants were unaware that involvement in the court was voluntary (Poythress, Petrila, McGaha, & Boothroyd, 2002; Redlich, 2005; Redlich, Hoover, Summers, & Steadman, 2010). Consequently, assisting with the decision to enter mental health courts may be defense attorneys’ most important role (Council of State Governments, 2005; Stefan & Winick, 2005; Thompson et al., 2007). In working with defendants to decide whether to participate in the mental health court, defense attorneys help them weigh the short-term benefit of quick case resolution against the potential long-term benefits of participating in the court that will require attending court hearings, participating in treatment, and being supervised by the court to ensure compliance of court-ordered conditions (Fisler, 2005; Haimowitz, 2002; Waters et al., 2009). In helping defendants decide whether to participate, defense attorneys ensure that defendants are competent to make informed decisions; assess the strength of the prosecutor’s case; determine the potential of defendants to adhere to court-ordered conditions and successfully complete the program; get clarity on what the likely conditions of participation would be, the case outcome if defendants successfully complete the court program, and the consequences of program failure; and the implications of a guilty plea if the court requires one (Council of State Governments, 2005; Haimowitz, 2002; Keele, 2002; Kempinen, 2011; Stefan & Winick, 2005).

Another important role for defense attorneys if defendants decide to participate in the mental health court program is to attend case conferences and all court hearings (Arkfeld, 2007; Council of State Governments, 2005; Kluger, Murrell, Tauber, Zeidman, Calabrese, & Hendricks, 2002; Thompson et al., 2007). By participating on an ongoing basis, defense attorneys help to ensure that defendants attend scheduled court hearings, provide encouragement and emotional support, and share information about treatment and personal circumstances that could affect decisions made at case conferences (Kempinen, 2011). An additional aspect of ongoing support and counsel is advocating for defendants if they fail to meet court-ordered conditions and risk dismissal from the program (Council of State Governments, 2005). At this point, some argue that the collaborative process of mental health courts becomes an adversarial one because of defendants’ potential loss of liberty (Casey, 2004). Courts may use jail as a sanction (Redlich et al., 2006) or sentence defendants to jail or prison if dismissed from the program and convicted. In fulfilling this role, defense attorneys seek to ensure sanctions are reasonable and that expelled defendants face outcomes no more severe than if defendants had not participated in the mental health court (Holland, 2010; Kluger et al., 2002; McNiel & Binder, 2010; Ray, Hood, & Canada, 2015). This role is especially important because mental health court defendants should not be punished for trying the court program and failing (Snedker, 2018).

Availability of Legal Representation to Mental Health Court Defendants

Given the important roles that defense attorneys potentially can play in mental health courts, legal advocates (Arkfeld, 2007; Keele, 2002; Kempinen, 2011; Schneider, Bloom, & Heerema, 2007; Seltzer, 2005) and guidelines for operating mental health courts (Council of State Governments, 2005; Thompson et al., 2007) emphasize the importance of mental health court defendants having active legal representation throughout all court processes. However, legal and financial barriers can make it difficult for mental health courts to ensure that defendants appearing before the court have access to ongoing legal representation. Defendants in municipal mental health courts may not have the right to or access to a defense attorney. Even in mental health courts associated with state or federal courts in which defendants have a right to a defense attorney, there is no clear statutory or constitutional requirement that defense attorneys attend the preliminary mental health court hearings or participate in case conferences and court hearings until discharge from the mental health court (Kempinen, 2011).

In addition, municipal mental health court defendants may not be able to afford to hire private defense attorneys or pay their attorneys to participate in the court on an ongoing basis. Also, many mental health courts cannot afford to hire one or more defense attorneys to work with all court participants. Next, public defenders represent large numbers of criminal defendants (Harlow, 2000; Strong, 2016). However, excessive caseloads in many jurisdictions have forced public defenders to make choices about which defendants to focus their greatest amount of time (Baxter, 2012). Consequently, public defenders representing mental health court defendants may not have the time to participate in all phases of the mental health court process.

Examples exist of efforts made to provide legal representation for mental health court defendants. Dallas County Mental Health Court partnered with the local public defender’s office to assign a public defender with experience working with defendants with psychiatric disorders to the court. This person meets with each referred defendant to offer independent advice regarding whether to enroll in the mental health court and serves as an active member of the mental health court team participating in all case conferences (Carmichael, Marchbanks, Kiven, Klavensma, Durkin, & Fabelo, 2010). The public defender’s office believed relocating its resources to the mental health court was one way to serve some of its most vulnerable defendants. In the case of municipal mental health courts where defendants may not have the right to a defense attorney, court staff may be able to work with the local legal aid society or public defender’s office to provide legal representation for at least some of the individuals referred to the court. The Chicago Bar Foundation (2014) provides one example of how this can be done. Through grants and a contract with Cook County, low-income municipal court defendants can receive free legal advice from a collaboration of the Cook County’s Coordinated Advice and Referral Program for Legal Service, the Chicago Legal Clinic, and pro bono private attorneys. While not specifically designed for the county’s mental health courts, it is likely that financially eligible mental health court defendants could receive legal advice or representation through one or more of these programs.

Focus of the Current Study

To date, no empirical studies have examined associations between being represented by a defense attorney and mental health court outcomes. Studies of defense attorneys in mental health courts have been conceptual arguments about their roles and the ethical issues they face (e.g., Casey, 2004; Holland, 2010; Keele, 2002) or empirical studies on the involvement of defense attorneys in mental health courts through documentation of referral rates to these courts (Snedker et al., 2017; Wolff et al., 2011), communication in court hearings (Boothroyd et al., 2003), attendance in preliminary court hearings (Strong et al., 2016), or participation in case conferences (Waters et al., 2009).

This study examines the associations between being represented by a defense attorney and key outcomes throughout the mental health court process. It takes advantage of a municipal mental health court (MMHC) in which representation by a defense attorney is not required or guaranteed, unlike state or federal courts in which defendants have a right to counsel. The state does not require or provide legal representation in municipal courts because of the limited loss of liberty if convicted, the maximum of which is one year of community supervision or incarceration in jail. However, local municipal courts may require and provide legal representation is certain situations that they identify. Among defendants accepted to the MMHC, this study compares whether defendants with and without defense attorneys had different outcomes in four key areas: (1) whether eligible defendants chose to participate in the MMHC, (2) if defendants chose to participate, whether the MMHC resolved defendants’ criminal charges without supervision, (3) whether eligible defendants failed to appear at the initial MMHC court hearings and their case was transferred to the traditional court, and (4) whether defendants successfully completed the MMHC program. The first outcome, whether to participate in the MMHC, involves short-term involvement on the part of the attorney. The other three outcomes are associated with defense attorneys providing support and advocacy after the decision has been made to participate in the MMHC until defendants are terminated from the MMHC. This study first reports the rate at which MMHC defendants were represented by defense attorneys. Second, it identifies the characteristics of defendants with and without defense attorneys. Finally, it compares the results of the four outcomes for defendants who did or did not have a defense attorney in both bivariate and multivariate analyses.

Study Site

The site of the study is the St. Louis County MMHC located in the largest county in Missouri, which has a population of just under 1 million residents. It was created in 2001, and 2545 defendants were referred to the MMHC by the end of 2017. The MMHC receives referrals for crimes committed in unincorporated St. Louis County. In addition, cities within the county can contract with the MMHC to transfer its municipal criminal cases involving defendants with psychiatric disorders from their courts to the MMHC for a nominal fee. The four judges from the St. Louis County municipal courts serve as the MMHC judges, and each judge holds a monthly special mental health docket at the county courthouse. Referrals to the MMHC are made by municipal court judges or county counselors, who function as prosecuting attorneys in municipal courts, which involves moving defendants from municipal court to the MMHC. Referrals are also made by police officers with special training in psychiatric disorders and crisis intervention, by city courts within St. Louis County that contract with the MMHC, and by defense attorneys, probation officers, and social service agencies.

Because defendants are appearing before a municipal court in Missouri, they do not have a right to a defense attorney if they cannot afford to hire one. However, MMHC judges have helped defendants secure pro bono defense attorneys in a limited number of cases for defendants they believed could not provide adequate self-representation. The MMHC requires that defendants have a diagnosed mental illness or developmental disability. Defendants can have a co-occurring substance abuse diagnosis or personality disorder. However, these diagnoses alone do not qualify defendants for the MMHC. The MMHC does not have its own clinical staff. Consequently, defendants are required to have a mental health provider submit psychiatric diagnoses to determine eligibility. The MMHC holds case conferences prior to each docket to review all defendants appearing on the docket that day. Case conference participants include the county counselors and the two case managers assigned to the MMHC, defense attorneys (if defendants have one), and treatment providers for those defendants appearing before the MMHC that day. MMHC judges and defendants do not participate in case conferences. Conditions for participation in the MMHC are established in the case conferences and presented to the judges for their approval. Defendants’ compliance with conditions is monitored by the MMHC case managers. The MMHC does not have its own funding for mental health services. Therefore, defendants must obtain services through sources available to the general population. Case managers refer defendants to services if they cannot locate mental health providers on their own. Defendants accepted into the MMHC program typically have their criminal charges dismissed if they comply with their court-ordered conditions.

Methods

Data Sources

This study is a secondary analysis of MMHC defendant information included in a database maintained by the MMHC case managers. This database includes all defendants referred to court and selected demographic, clinical, criminal justice, and programmatic variables for each defendant. Although the MMHC started in 2001, it did not consistently collect information on whether defendants were represented by defense attorneys until 2010. As such, this study incorporated defendants admitted between 2010 and 2015, with some limited exceptions. This study omitted 36 defendants who were not eligible to participate in the MMHC, typically for not having the required psychiatric disorder; 14 defendants who died prior to discharge; 5 defendants whose cases were still open at the end of 2017; and 3 defendants whose cases were transferred out of the MMHC for administrative reasons unrelated to any actions on the part of the defendants. With these omissions, the final study sample was 1012 defendants. We did not include defendants admitted after 2015 because data were inconsistently collected on new admissions after that time related to a change in case managers who have primary responsibility for data collection and entry.

Variables

Independent Variable

The independent variable for two sets of analyses is whether or not MMHC defendants were represented by a defense attorney. The defense attorney variable had three attributes: no attorney, private attorney, and court-appointed attorney. In the bivariate and logistic regression analyses of key outcomes, we combined private and court-appointed attorneys to form a defense attorney variable with two attributes: represented by a defense attorney and not represented by a defense attorney. Whether or not defendants were represented by a defense attorney served as a dependent variable in the bivariate analysis to differentiate defendants who did and did not have attorneys.

Control Variables

Program theories of specialty courts typically include defendant factors or attributes, defendant behavior during court supervision and court reactions to it, and, in some cases, factors outside the court process (e.g., Brown, Zuelsdorff, & Gassman, 2009; Goldkamp, White, & Robinson, 2001). In this study, 3 of 4 program outcomes occur prior to defendants being supervised by the court, thus excluding defendant behavior during court supervision. Consequently, we chose to focus on defendant factors, including demographic information, psychiatric disorders, and the crimes for which defendants were referred to the MMHC (i.e., committing crimes). Demographic variables included age, measured in years at time of referral to the MMHC; sex (male and female); race (White, African American, and other); marital status at time of referral to the MMHC (never married, married, divorced, separated, and widowed); living arrangement at time of referral (independent, with parents, with extended family, other living arrangements), and per capita income in zip code of residence. We collapsed categories for race and marital status in the multivariate analyses to form the dichotomous variables of race (White and other races) and marital status (never married, yes/no). Regarding per capita income, the MMHC database included the zip code of residence but no individual income information. We identified 5-year estimates of annual per capita income by zip code from the American Community Survey, which is published by the U.S. Census Bureau (2020), for the year in which defendants were admitted to the MMHC and added that amount to the database for each defendant. Studies have found that per capita income by zip code is an adequate proxy for individual socioeconomic status (e.g., Mustard, Derksen, Berthelot, & Wolfson, 1999; Link-Gelles et al., 2016), although some researchers urge caution in its use (e.g., Geronimus, Bound, & Neidert, 1996; Soobader, LeClere, Hadden, & Maury, 2001). A reason why defendants in this study may not have had defense attorneys is that they could not afford one. Consequently, we added this measure of socioeconomic status as an important control variable.

A second set of control variables was psychiatric disorders. The MMHC database includes up to two diagnoses. Consequently, we coded these as individual dichotomous variables indicating the presence or absence of each of the following psychiatric disorders: bipolar disorder, schizophrenia, depression, anxiety disorders, developmental disorders, and other psychiatric disorders. In addition, the database included whether defendants had a history of substance abuse. Case managers indicated this history if defendants had a substance disorder diagnosis, had prior substance abuse treatment, or admitted to having a substance abuse problem even if undiagnosed.

The third set of control variables was the crime or crimes for which they were referred to the MMHC. The MMHC database listed all referral criminal charges. These were coded as individual dichotomous variables including the presence or absence of each of the following crimes: assault, resisting arrest, stealing, peace disturbance, property damage, trespassing, and other crimes. As with psychiatric disorders, defendants could have more than one committing crime.

Dependent Variables

The four dependent variables were key outcomes that could occur throughout the mental health court process. Each was coded as an individual dichotomous variable (yes/no). The first outcome is whether eligible defendants chose to participate in the MMHC. As previously stated, participation in mental health courts is voluntary. The analysis of this outcome included all 1012 defendants. The second outcome is whether defendants’ criminal charges were resolved without ongoing supervision from the MMHC. After agreeing to participate, defendants can negotiate with the MMHC prior to or at the first hearing to have their criminal charges dropped without a period of court supervision. This outcome excluded those defendants who chose not to participate, resulting in 900 cases. The third outcome is whether defendants failed to appear at the initial court hearings, and the MMHC judge transferred the case back to the regular municipal court for disposition. Even though defendants may agree to participate in the MMHC, some defendants never show up for court and a warrant is eventually issued for their arrest. Excluded from this outcome were defendants who chose not to participate and who resolved their cases without supervision, resulting in 844 cases. The fourth outcome is whether defendants supervised by the MMHC had a positive or negative termination from supervision. A positive termination typically results in criminal charges being dropped, while the cases of the defendants who had a negative termination are transferred back to the regular court for disposition. This outcome included only those defendants who were supervised by the MMHC, resulting in 616 cases.

Analytic Strategy

First, we used descriptive statistics to describe the characteristics of defendants. Second, we used chi-square and t-tests to calculate statistical differences in two sets of bivariate analyses using α < .05. To calculate effect size, we used Cramer’s V for categorical covariates and Cohen’s d for continuous covariates. We reported effect sizes when bivariate analyses were statistically significant. One set of bivariate analyses identified the characteristics of defendants who were and were not represented by defense attorneys. The second analysis used defense attorney as the independent variable and the four key outcomes as dependent variables. Third, to control for other factors in the analyses of the associations between being represented by a defense attorney and the four key outcomes, we estimated logistic regression equations that included defense attorney status as the independent variable; the demographic information, psychiatric disorders, and committing crimes as control variables, and the four key outcomes as dependent variables.

Missing Data

Some variables had missing data. Missing data can be a source of measurement error and can bias the results of analyses (Roth, 1994). In this study, missingness ranged from 0% to 29%. Acceptable levels of missingness are up to 40% (Fox-Wasylyshyn & El-Masri, 2005). Data were not missing for the four key outcomes, age, sex, race, and the committing crime variables. Missing data was found for marital status (11%), living arrangement (16.8%), per capita income by zip code (0.3%; 3 defendants were homeless), psychiatric disorders (27.1%), and history of substance abuse (29%). In most cases, data were missing for defendants who were not supervised by the court, and cases managers were unable to gather any information they may not have obtained during the initial telephone contact with defendants to explain the program. Preliminary analyses, based on the recommendations of Allison (2002), suggested that the missing data were tentatively missing at random. To account for the missing data and to reduce the resulting bias, we used STATA 16 (StataCorp, 2019) to conduct a multiple imputation to run the multivariate analyses. We imputed 10 datasets using the multiple imputation chained equations specifier. Data were combined to create final models.

Defendant Characteristics

The mean age of defendants was 36.2 years (SD = 14.2; Mdn = 33.4) and ranged from 17 to 85.8 years. While juvenile courts typically have jurisdiction over individuals under a specific age, which varies from state to state, in Missouri, the law in effect during the study period states that juvenile courts do not necessarily maintain jurisdiction over juveniles who allegedly violate a municipal ordinance prior to the age of 17 (Missouri Revised Statutes, section 211.031 [3], 2016). The mental health court in this study accepts referrals from only municipal courts. Therefore, there is no distinction needed between adults and juveniles in terms of legal representation in this study. There was no indication that the relevant juvenile court maintained concurrent jurisdiction. Defendants tended to be male (61.4%); White (64.4%) with 32.5% being African American and 3.1% other races; and never married (78.1%). About one-third of defendants (36.5%) lived independently, either alone or with a roommate or partner, while 43.2% lived with parents, 10.8% with other relatives, and 9.5% in other residential settings. Mean per capita income by zip code was $29,393 (SD = $10,938; Mdn = $27,952) and ranged from $9532 to $86,030. Three referral sources made 88% of referrals to the MMHC, including police officers trained in mental health crisis intervention (41.5%), contracts with municipals courts (23.8%), and transfers from the regular municipal court. Defense attorneys provided 3.8% of all referrals. Frequency of psychiatric disorders included 31.4% diagnosed with bipolar disorder, 26.3% with schizophrenia, 23.6% with depression, 15.3% with anxiety disorders, 16% with developmental disorders, and 13.3% with other disorders. Over half of defendants (54.9%) had a history of substance abuse. Committing crimes included 46.7% charged with assault, 19.7% with peace disturbance, 13.9% with resisting arrest, 13.7% with property damage, 11.5% with trespassing, 11.1% with stealing, and 27.5% with other crimes.

Results

Rate of Legal Representation

Most defendants in the study (79.1%) were not represented by a defense attorney. Among defendants who had legal representation, 88.6% had a private attorney and 11.4% had a court-appointed attorney.

Defendant Characteristics and Attorney Status

Several demographic, psychiatric disorder, and committing crime variables differentiated defendants who did and did not have a defense attorney, with the greatest differences being among demographic variables. The variable that had the strongest relationship with being represented by a defense attorney was per capita income by zip code. Defendants represented by a defense attorney lived in zip codes with a higher per capita income ($33,003), compared to those without a defense attorney ($28,439) (d = .423). Selected other demographic variables had moderate relationships. Defendants represented by a defense attorney were older (38.8 years) than those without a defense attorney (35.5 years) (d = .232). In addition, African American defendants were less likely to have a defense attorney than other races. African Americans constituted 17.1% of defendants represented by a defense attorney and 36.6% of defendants without a defense attorney (V = .175). Defendants who lived independently were more likely to have a defense attorney than those living in other arrangements. Defendants living independently constituted 47.6% of defendants represented by a defense attorney and 33.3% of defendants without a defense attorney (V = .128). Finally, defendants who never married were less likely to have a defense attorney than other marital statuses. Those never married constituted 70.2% of defendants represented by a defense attorney and 80.3% of defendants without a defense attorney (V = .104).

Three psychiatric disorders were associated with being represented by a defense attorney, all with moderate relationships. Defendants who had a defense attorney were more likely to be diagnosed with depression or anxiety and less likely to be diagnosed with schizophrenia. Defendants diagnosed with depression constituted 34% of defendants represented by a defense attorney and 20.7% of defendants without a defense attorney (V = .130), and defendants diagnosed with anxiety disorders constituted 25.3% of defendants represented by a defense attorney and 12.5% of defendants without a defense attorney (V = .147). In contrast, defendants diagnosed with schizophrenia constituted 14.2% of defendants represented by a defense attorney and 29.7% of defendants without a defense attorney (V = .146).

Finally, two committing crimes were associated with being represented by a defense attorney. Defendants charged with peace disturbance constituted 14.2% of defendants represented by a defense attorney and 21.1% of defendants without a defense attorney, although the relationship was weak (V = .070). Also, defendants charged with other crimes constituted 37.2% of defendants represented by a defense attorney and 24.8% of defendants without a defense attorney (V = .115). The most frequent crimes in the other category were alcohol or drug offenses (8.8%), driving or traffic violations (5.8%), and harassment (2.8%). Table 1 includes more complete information.

Table 1 Characteristics Associated with Being Represented by a Defense Attorney in the MMHC
Full size table

Attorney Status and Key Outcomes

Bivariate Analyses

In the bivariate analyses, statistical differences existed between defendants with and without defense attorneys across the four keys outcomes in the MMHC, with the strength of the relationships being moderate in 3 of 4 outcomes. Defendants represented by a defense attorney were more likely to choose not to participate in the MMHC (20.9%) than defendants without a defense attorney (8.5%) (V = .160). Also, defendants represented by a defense attorney were more likely to resolve criminal charges without being supervised by the MMHC (9.6%) than defendants without a defense attorney (5.5%), although the relationship was weak (V = .066). In addition, defendants represented by a defense attorney were less likely to have their cases transferred to the regular municipal court for failing to appear at the initial MMHC hearings (11.3%) compared to defendants without a defense attorney (30.4%) (V = .166). Finally, defendants represented by a defense attorney were less likely to have a negative termination from the MMHC (12.7%) compared to defendants without a defense attorney (24.5%) (V = .118). See Table 2 for additional information.

Table 2 Bivariate Associations between Defense Attorney Status and Key MMHC Outcomes
Full size table

Logistic Regression Analyses

The associations between defense attorney status and the four key outcomes identified in the bivariate analyses held when controlling for demographic, psychiatric disorder, and committing crime variables in the logistic regression analyses. First, being represented by a defense attorney increased the odds that defendants chose not to participate in the MMHC (OR = 1.78). The amount of variance explained by the logistic regression model was 6%. Control variables that increased the odds of not participating in the MMHC included being older (OR = 1.03), being male (OR = 1.87), and being White compared to other races (OR = 2.07). No psychiatric disorders or committing crimes were associated with not participating in the MMHC. These results are presented in Table 3.

Table 3 Logistic Regression Analysis of the Association between being Represented by a Defense Attorney and Choosing Not to Participate in the MMHC (N = 842)
Full size table

Second, being represented by a defense attorney increased the odds that defendants resolved their criminal charges without being supervised by the MMHC (OR = 2.80). The amount of variance explained by the logistic regression model was 13%. Control variables that increased the odds of resolving criminal charges without MMHC supervision included being older (OR = 1.028) and living in other arrangements compared to living independently (OR = 14.05), while variables that decreased the odds included being diagnosed with depression compared to other diagnoses (OR = 0.24) and being charged with other committing crimes compared to all other crimes (OR = 0.20). Table 4 includes those results.

Table 4 Logistic Regression Analysis of the Association between being Represented by a Defense Attorney and Resolving Criminal Charges without MMHC Supervision (N = 758)
Full size table

Third, being represented by a defense attorney decreased the odds of defendants’ cases being transferred to the regular municipal court for failure to appear at the initial court sessions (OR = 0.47). The amount of variance explained by the logistic regression model was 10%. Two control variables increased the odds of case transfers, including being older (OR = 1.02) and being charged with trespassing (OR = 3.21). No psychiatric disorders were associated with court transfers. See Table 5 for complete results.

Table 5 Logistic Regression Analysis of the Association between being Represented by a Defense Attorney and the Case Transferred for Non-Attendance (N = 705)
Full size table
Table 6 Logistic Regression Analysis of the Association between being Represented by a Defense Attorney and Negative Termination from MMHC Supervision (N = 598)
Full size table

Finally, being represented by a defense attorney decreased the odds of defendants having a negative termination from MMHC supervision (OR = 0.51). The amount of variance explained by the logistic regression model was 8%. One demographic variable, being older, decreased the odds of negative termination (OR = 0.954). No psychiatric disorders or committing crimes were associated with negative termination Table 6.

Discussion

We expected that the majority of defendants accepted into the MMHC would not have a defense attorney, as defendants do not have a right to legal representation in Missouri municipal courts unless local rule dictates otherwise. Only 20.9% of defendants were represented by a defense attorney, with most of those being private attorneys (88.6%) rather than court-appointed attorneys (11.4%). In addition, defense attorneys made only 3.8% of referrals to court, which is consistent with the low rate of representation by attorneys. National information does not exist on the percentage of defendants represented by defense attorneys in municipal courts, nor does it exist for municipal mental health courts. The website of the State Bar of Texas (2020) indicates that the “majority” of defendants appearing before Texas municipal courts are self-represented. Similarly, a review of Missouri municipal courts by the National Council for State Courts (Griller, Williams, Brown, & Hall, 2015, p. 20) reported that the “vast majority” of defendants appearing before those courts were self-represented. Also, an investigative newspaper reporter examining the Milwaukee, Wisconsin, municipal court found that only 2% of defendants appearing before the court between 2011 and 2014 were represented by a defense attorney (O’Brien, 2015). With these three examples as comparison points, a 79.1% of self-representation by MMHC defendants is within the range of the “vast majority” and 98% reported above by other municipal courts.

It also should not be surprising that this study identified income, measured by living in a zip code with higher per capita income, as have the strongest association with defendants being represented by a defense attorney. Other statistically significant demographic variables included being older, being White, living independently, and being married, all of which can be associated with higher levels of incomes. Regarding psychiatric disorders, defendants diagnosed with schizophrenia were less likely to be represented by a defense attorney, while defendants with more treatable psychiatric disorders, including anxiety disorders and depression, were more likely to be represented by a defense attorney. Schizophrenia is a highly disabling psychiatric disorder (Chaudhury, Deka, & Chetia, 2006), and persons diagnosed with this disorder have very high rates of unemployment overall, and in comparison to other psychiatric disorders (Hakulinen et al., 2019; Zivin et al., 2011). Consequently, they are less likely to be able to afford to hire a defense attorney. Committing crimes were minimally associated with being represented by a defense attorney.

Defendants referred to and participating in municipal mental health courts may not have guaranteed access to a defense attorney, as was the case in the court that was the focus of this study. The lack of a defense attorney may not be confined to just municipal mental health courts. Defendants in state or federal courts who do not qualify for public defenders may not be able to afford a private attorney or pay their attorneys beyond the initial mental health court hearing to enter the plea; consequently, some defendants will not have a defense attorney present at all case conferences and court hearings as is recommended (Council of State Governments, 2005; Kluger et al., 2002; Thompson et al., 2007).

This study found associations in both the bivariate and logistic regression analyses between being represented by a defense attorney and four key outcomes that span throughout the court process. We found that MMHC defendants represented by a defense attorney were more likely to choose not to participate in the MMHC. This is especially important given that studies have found that some mental health court participants were unaware that participation was voluntary (Poythress et al., 2002; Redlich, 2005; Redlich et al., 2010). Defense attorneys can provide mental health court defendants with thoughtful and unbiased information on the advantages and disadvantages of participation and assist them to weigh their options (Council of State Governments, 2005; Holland, 2010; Kempinen, 2011; Kluger et al., 2002; Thompson et al., 2007). Some of the reasons for defendants not participating could include concern about being able to adhere to court-ordered conditions and successfully completing the program; a weak case on part of the prosecutor that potentially could be won by the defense; not getting a substantially better mental health court case outcome if successfully completing the program compared to pleading guilty in regular court; and being charged with a minor crime that would not result in jail time or probation supervision if pleading guilty (Council of State Governments, 2005; Haimowitz, 2002; Keele, 2002; Kempinen, 2011; Stefan & Winick, 2005).

This study also found that MMHC defendants represented by a defense attorney were more likely to resolve their criminal charges without supervision. Anecdotal information from MMHC case managers suggests that judges consider two types of information when making the decision about whether to require supervision. One is the length of time between the committing crime occurring and the first court appearance and whether defendants received ongoing psychiatric treatment during this period, were currently psychiatrically stable, and had supportive and permanent living arrangements. The second is whether defendants had a significant disability with low functioning levels and were now living in highly supervised environments. Defense attorneys are in an ideal position to gather and present information such as this to mental health courts that cases should be resolved without supervision (Kempinen, 2011).

Additionally, we identified that MMHC defendants represented by a defense attorney were less likely to have their cases transferred back to the regular court as a result of failure to appear at the initial court hearings. Being represented by a defense attorney may explain this in at least two ways (Arkfeld, 2007; Council of State Governments, 2005; Kempinen, 2011; Thompson et al., 2007). One is that attorneys can ensure that defendants are notified of court dates and have transportation to court hearings. A second way is that if defendants do miss the initial court hearings, attorneys can petition the mental health court to give their defendants additional chances to appear.

Finally, this study found that MMHC defendants represented by a defense attorney were less likely to have a negative termination from the court. This is consistent with the stated role of attorneys to negotiate sanctions short of termination when defendants do not adhere to their conditions of participation (Holland, 2010; Kluger et al., 2002; McNiel & Binder, 2010; Ray et al., 2015).

The positive associations between being represented by defense attorneys with the latter three mental health court outcomes, that is, resolving charges without court supervision, attending initial courts hearings, and successfully completing the program, are consistent with the active involvement of attorneys throughout defendants’ entire participation in the court, as others have advocated (Arkfeld, 2007; Council of State Governments, 2005; Kempinen, 2011; Kluger et al., 2002; Thompson et al., 2007).

While finding associations between being represented by a defense attorney and mental health court outcomes, the amount of variance in outcomes explained in the four regressions models was low, ranging from 6% to 13%. This would indicate that, while associations exist between defense attorneys and mental health court outcomes, many other factors not included in the models are associated with mental health court outcomes. Legal representation, while possibly providing some potential benefits for mental health court defendants, appears to play a minor role in relationship to mental health court outcomes when compared to other, yet unidentified, factors.

Limitations and Future Research

This is the first empirical study to examine associations between being represented by a defense attorney and mental health court outcomes. However, given the study’s design, we cannot draw conclusions about the influence or effect of legal representation on mental health court outcomes. We identified associations between the two and provided a rationale about how and why defense attorneys could influence court outcomes, but future research, with more sophisticated designs, is needed to determine if casual connections exist.

A second limitation is that this study treated attorney status as a dichotomous variable. Rather than simply listing whether or not defendants were represented by defense attorneys, future research should include the level of participation by defense attorneys in the initial decision to participate in the mental health court, as well as in case conferences and court appearances throughout the court process. Qualitative studies of defense attorney participation could also complement quantitative studies.

Similarly, we previously noted that that judges and case managers may informally assume some of the roles traditionally held by defense attorneys in mental health court proceedings (Castellano, 2011, 2017). Future research is needed to document the roles played by case managers and judges in the mental health court process when defendants self-represent; when defense attorneys are absent from some court processes, such as initial hearings, case conferences, or ongoing court hearings; and when defense attorneys are fully present during all relevant court processes to determine the extent to which judges and case managers are informally assuming some roles of defense attorneys. It is possible, for example, that an important factor in mental health court outcomes is the level of advocacy on behalf of defendants, whether that be from judges, case managers, defense attorneys, or a combination of them.

Another study limitation is that our regression models did not include the amount, type, and quality of the services mental health court participants received while participating in the court. Mental health court defendants who were able to hire a defense attorney may have access to greater mental health and support services, although we tried to address this in part by controlling for per capita income. Future research should include some information on the type and amount of services defendants receive while participating in the court.

Next, the logistic regression models explained a low amount of variance in each of the four outcomes. Other potentially relevant factors not found in the regression models could include those previously identified, that is, the quality and level of participation by defense attorneys in the court process, the level of advocacy of judges and cases managers, and type, amount, and quality of mental health and social support services that defendants accessed, as well as defendants’ levels of psychiatric symptoms and functioning at the beginning and end of the court process, educational, vocational, and employment information, and criminal history, among others. Future research is needed to identify those factors, along with legal representation, that increase the variance explained in the four key outcomes.

Finally, these results may not be generalizable to other mental health courts. Future research on the study of defense attorneys and mental health courts is needed with courts that hear felony cases as well as misdemeanors, are housed in various geographic locations, work with diverse defendant populations, and incorporate defense attorneys in different ways. This study provides a base to build upon as research continues to identify factors related to successful mental health court outcomes and the roles that defense attorneys play in those outcomes.