Abstract
The purpose of the study was to determine how often Maryland judges agreed with the opinions of forensic evaluators in deciding whether to transfer youthful defendants to juvenile court from adult court and to investigate which factors were most important in the opinions of the evaluators and the final decisions of the judges. Data were extracted from a sample of 200 waiver evaluations, and case outcomes were determined. Factors were examined with both univariate analysis and logistic regression models, to find correlates to and predictors of judges' decisions and evaluators' opinions. The most important factor influencing the decision of the judges was the forensic evaluators' opinions. Logistic regression analysis identified three factors that were significant predictors of the evaluator's opinion: public safety risk, history of the involvement of Department of Juvenile Services, and defendant's age at the time of the offense. The judges' decisions correlated strongly with the forensic evaluators' opinions.
On the basis that the more serious and violent juvenile offenses would be better handled in adult criminal courts than in juvenile courts, nearly all states have enacted laws to allow adjudication of juveniles in adult courts, under certain conditions.1 Since the laws have been enacted, juvenile arrest rates have declined. At the same time, the number of detained and committed youths in both juvenile and adult facilities has dramatically increased.2
There are several methods by which state law determines whether juveniles are tried in adult or juvenile court. These include prosecutorial discretion, judicial waivers, and statutory exclusion. With prosecutorial discretion (also referred to as concurrent jurisdiction or direct file), the prosecutor is allowed to decide whether to file a case in juvenile or adult criminal court. Laws establish jurisdiction for certain types of offenses in both courts and permit the prosecutor alone to determine which court will try a specific case.
Transfer from juvenile court to adult court by judicial waiver may be accomplished in three ways: discretionary, presumptive, or mandatory. Typically, waiver hearings are held to determine the appropriateness of the transfers. The hearings usually involve a mental health evaluation of the juvenile (in some states, this evaluation is required by statute). Clinicians may be asked to assess the youth's level of maturity, amenability to rehabilitation, and likelihood of future violence and offense, as well as the presence and role of any mental disorder.3 Discretionary judicial waiver permits the judge to transfer the case after certain criteria have been satisfied.4 The criteria usually include consideration of rehabilitation and public safety, as established by Kent v. United States.5
Presumptive judicial transfer shifts the burden of proof from the prosecutor to the juvenile. The defense must argue that the judge should not have the case transferred to adult criminal court and that the youth would be handled best in juvenile court. Mandatory judicial waiver requires the judge to determine only whether the case meets the criteria set by law for waiver.
Statutory exclusion laws require juveniles, who by age would usually be tried in juvenile courts, to be tried initially in adult criminal courts instead when charged with certain offenses. Most often, these transfers are for serious violent offenses and will specify additional restrictions, such as age or prior offense record. In some states, reverse waiver laws then allow the adult criminal court to transfer cases back to juvenile court for adjudication or disposition. Some states have statutes that hold that if a juvenile has been transferred or waived to adult criminal court previously, all subsequent charges will be handled there: the once an adult, always an adult provision.6 Considering the fact that transfer evaluations hinge in part on a forensic assessment, it is important to know what forensic psychiatrists consider in their evaluations.
Important Factors
The definition of amenability to treatment is not clearly specified in either statutory or case law. Forensic examiners may tend to consider only traditional therapy when assessing a defendant's amenability and exclude other approaches, such as special education, work programs, foster care, advocacy, residential treatment, vocational treatment, and incarceration.7 Each clinician's personal conceptualization of delinquency etiology (e.g., psychodynamic versus social learning) may influence the treatment approaches considered and recommended. Amenability recommendations are also influenced by the services available and the time remaining to be treated in juvenile jurisdiction. Statutes require neither a specific degree of success nor an area of improvement to deem a juvenile amenable to treatment.5
The second area typically addressed by forensic examiners in transfer evaluations is the degree of future danger the juvenile poses to the public. The literature reflects the limited extent to which mental health professionals can and should predict dangerousness in adults.8 Prediction of violent behavior in juveniles may be even more difficult than in adults. However, there are risk factors that are known to correlate with future violence, such as previous violence, level of IQ, and substance abuse. These can be communicated to the court in a general risk assessment statement. Violence prediction in youths can involve an analysis of individual, family, school, peer-related, community, and neighborhood factors.9 Further complicating matters, statutes are not precise regarding specific factors to be used as criteria.10 The current basis for juvenile risk assessment is limited, and forensic examiners may have difficulty structuring this component of the transfer evaluation.
The third domain of the typical transfer evaluation is the sophistication and maturity of the juvenile. It is generally accepted that psychiatrists and psychologists possess interviewing, observational, and testing expertise for the assessment of maturity.11 Standardized, norm-referenced psychological tests are often used in legal contexts. IQ testing may be used to appreciate fully the level of functioning of the juvenile. Although all these tests can be used, there is no agreed upon standard for including them in assessments of maturity.
Although all of these factors may be considered by both evaluators and judges, the question remains as to which factors are key to the opinion of the evaluator and the decision of the judge. There are many factors that contribute to each one of the evaluating criteria, but are there certain factors that consistently influence the opinions of evaluators and judges? Grisso et al.12 identified unwillingness to accept intervention, adult-like self-reliance, and a greater offense record as the factors most associated with judicial transfer decisions. Other studies suggest that youths who were identified as dangerous, criminally sophisticated, and difficult to treat were more often transferred to adult court.13
The System in Maryland
In Maryland, a child younger than 18 years is considered to be a juvenile. A delinquent is a child who has been adjudicated by the juvenile court for an act that would be a crime if committed by an adult and who needs guidance, treatment, or rehabilitation.14 Some children may be treated as adults in adult criminal court rather than in juvenile court, depending on the severity of the crime and the child's age, prior juvenile record, and mental and physical condition, among other factors.
The Maryland Juvenile Court may have original jurisdiction but may waive jurisdiction by discretionary judicial waiver, depending on the offense and the age of the alleged delinquent. The juvenile court may not waive its jurisdiction under this section until after it has conducted a hearing, conducted solely to determine whether it should do so. The court may not waive its jurisdiction under this section unless it determines, from a preponderance of the evidence presented at the hearing, that the child is an unfit subject for juvenile rehabilitative measures.15
In Maryland, depending on the crime and the age of the juvenile, statutory exclusion gives adult criminal courts original jurisdiction. Cases can be transferred from adult court to juvenile court (reverse waiver) if the court determines, by a preponderance of the evidence, that a transfer of its jurisdiction is in the interest of the child or society. Transfers are prohibited if the child has been transferred to juvenile court and adjudicated delinquent previously, the child has been convicted in an unrelated case excluded from the jurisdiction of the juvenile court, the charge is murder in the first degree, and the accused was 16 or 17 years of age when the alleged crime was committed.16
Statutory criteria the court must consider in the evaluation of the defendant's suitability for adjudication within the juvenile justice system include the age of the child; the child's amenability to treatment in any institution, facility, or program available to delinquents; the mental and physical condition of the child; the nature of the offense and the child's alleged participation in it; and public safety. In making a determination under this section, the court may order that a study be made of the child, the child's family and environment, and of factors concerning the disposition of the case.13
In the Medical Office for the Circuit Court of Baltimore City, forensic evaluators are charged with completing reverse-waiver evaluations. They complete approximately 230 of these evaluations per year, and the number has increased over the years. In 2004, although each evaluator assessed the same factors (the age of the child, the child's amenability to treatment, the mental and physical condition of the child, the nature of the offense, and the public safety risk), the methods employed in the assessments varied. In addition, it is uncertain what factors influenced the opinions of the evaluators when assessing each area. After the evaluator's report was completed, the judge provided the final decision as to whether the case would be transferred back to juvenile court.
We decided to investigate several questions. First, we wanted to know how often the judges agreed with the opinion of the evaluators and ultimately decided to transfer juveniles to juvenile court from adult court. We also wanted to investigate which factors were most important in the opinion of the evaluator and the final decision of the judge. By conducting this study, we hoped to add clarification and begin to provide some further guidance to psychiatrists who conduct similar evaluations.
Methods
Judges in the Circuit Court for Baltimore City refer reverse-waiver cases to the Circuit Court Medical Office for evaluation. Psychiatrists and psychologists at the medical office answer the court's order for evaluation. To gather information to investigate the opinions of evaluators and decisions of the judges, we retrieved court case files that were stored in the court medical office. Two hundred randomly selected cases of defendants who participated in reverse-waiver evaluations in 2004 were reviewed. From these case files, which included completed transfer of jurisdiction evaluations, the following information was extracted:
-
Opinion on whether to transfer the case;
-
Age at time of offense and at time of evaluation;
-
Opinion on physical maturity;
-
Opinion on emotional maturity;
-
Opinion on amenability to treatment;
-
Gender;
-
Race;
-
History of employment;
-
History of mental health diagnosis (other than conduct disorder or oppositional defiant disorder);
-
History of substance use disorder diagnosis;
-
History of juvenile services involvement;
-
Opinion on risk to public safety;
-
History of previous offenses;
-
Weapon use during the crime;
-
Number of alleged perpetrators in the offense;
-
Number of children of the defendant; and
-
Living arrangements of the defendant at the time of the offense.
All factors were categorized as dichotomous variables or numericals, with some exceptions. The defendant's history of Department of Juvenile Services (DJS) involvement was subcategorized into the level of service: community, residential, or none. The public safety risk was categorized as low, moderate, and high, as indicated by the evaluator. Living arrangement was subcategorized as at home or out of the home.
To determine the ultimate judicial decision about transfer of jurisdiction, we reviewed the computerized court record system. The program provided information about the disposition of the defendants evaluated. There were three potential dispositions: transfer to juvenile court, retention in the adult criminal court, and dismissal or withdrawal of the charges.
This work involved review of public record documents for criminal defendants. No institutional review board approval or exemption was sought. No defendants can be individually identified by the data presented.
Results
Description of the Sample
A total of 200 charts of defendants referred to Medical Services of the Circuit Court for Baltimore City were reviewed. All of the defendants had been referred for evaluations concerning transfer of jurisdiction in 2004. Of the 200 cases reviewed, 39 were not analyzed because the charges against the defendants were eventually dismissed. Of the remaining 161 individuals, 138 were male and 23 were female. Of the defendants, 154 were African American and 7 were Caucasian. The defendants' ages at the time of the offense ranged from 14.10 to 18.08 years (mean = 16.81, SD = 0.70). The ages at the time of evaluation ranged from 14.58 to 19.08 years (mean = 17.26, SD = 0.74).
Crosstab Analyses: Univariate Analyses With Judge's Ruling as the Dependent Variable
Table 1 presents the variables that were found to have a significant relation to judges' decisions listed in order of the strength of the relation, as determined by chi-square analysis. While chi-square tests indicate whether associations between the variables are significant, the phi coefficient (which ranges from −1 to +1, with values closer to 1 indicating stronger relations between the variables) provides an index of the magnitude of the relation. In order from strongest to weakest, the most powerful correlations with judges' decisions were the evaluator's opinion, amenability to treatment, risk to public safety, Department of Juvenile Services (DJS) involvement, and history of offenses. Table 1 also displays the odds ratio (OR) for each of the significant relationships and the likelihood that defendants would be transferred to the juvenile system.
Univariate Analyses With Judges' Decisions as the Dependent Variable, Arranged Roughly From the Strongest Relation to the Weakest
A closer examination of the data revealed that, in 50.3 percent of cases, the judges ruled in favor of transferring defendants from adult court to juvenile court, whereas forensic examiners recommended transfer in 57.1 percent. The judges ruled against transfer in 49.7 percent of cases, and the forensic evaluators recommended retaining the defendants in the adult system in 42.9 percent. The odds ratio analysis indicated that when the evaluator recommended transfer, the odds were 12.21 times greater that the judge would decide in favor of transfer than when the evaluators recommended against it. As listed in Table 1, other significant though weaker relations were also found. The following variables did not have a significant relation to the judge's decision: Axis I diagnosis, injury to the victim, residing with family, race, and being the sole perpetrator.
Logistic Regression: Judge's Decision as the Outcome Variable
To determine the best model of predicting the judge's decision, we performed several different logistic regression analyses. The results changed, depending on how many variables were included in the model. When all variables were included, the overall size of the sample decreased to 90 because of data that were not provided in the evaluations that were reviewed. Two variables emerged as significant predictors of the judge's decision: evaluator's opinion and employment history. Because there were several variables (such as employment history) that were missing in a high number of cases, an attempt was made to create a more stable analysis model by examining variables with data found in the most cases.
If the variables missing in a high number of cases were excluded, the sample size increased and the results changed. To maximize the sample size, we removed the predictor variables one by one, until a stable model emerged with a sufficient sample size. The final model (n = 161) showed two variables to be significant predictors of the judge's decision: evaluator opinion (OR = 8.85; p < .001) and previous offense (OR = 3.67; p < .01). All other predictors with an adequate number of cases were no longer significant when these two variables were in the prediction equation. With these two predictor variables in the model, 77 percent of the judges' decisions were predictable (Nagelkerke R2 = 0.41). It should also be pointed out that amenability to treatment was a significant predictor in the univariate analysis. However, because this variable had more than 13 percent of its cases missing, the overall sample size was significantly reduced (n = 90).
The analysis was also executed without including the evaluators' recommendation. When all variables that were significant predictors were included in the univariate analyses and the loss of cases because of missing variables (n reduced to 90 cases) was ignored, amenability to treatment and public safety risk were the significant predictors.
Crosstab Analyses: Univariate Analyses With Evaluators' Recommendations as the Dependent Variable
Table 2 presents the predictor variables that were significantly related to the evaluators' recommendations. The strongest relations were found with amenability to treatment, risk to public safety, emotional maturity, DJS involvement, and history of previous offense. Variables found to be unrelated to evaluators' recommendations were injury to victim, living with family, being the sole perpetrator, employment history, and gender. Table 2 also displays the likelihood that a defendant will be recommended for transfer to the juvenile system.
Univariate Analyses With Evaluators' Recommendations as the Dependent Variable, Arranged Roughly From the Strongest Relation to the Weakest
Logistic Regression: Evaluators' Recommendations as the Outcome Variable
Similar to the analyses that were made of the judges' decisions, various predictor variables were removed because too much data were missing. The final, stable model, which had the largest sample (n = 118), showed three variables to be significant predictors of the evaluators' recommendations. All other predictors with an adequate number of cases were no longer significant when these three variables were in the prediction equation. The three variables were risk to public safety (OR = 34.83; p < .001), DJS involvement (OR = 11.04; p < .001), and age at the time of the offense (OR = 5.44; p < .006). With these three predictor variables in the model, 84.7 percent of the evaluators' recommendations could be predicted (Nagelkerke R2 = 0.67). In particular, the odds were 34.83 times greater for the evaluators to recommend against transfer when the defendants were high risks to public safety, than for those defendants who were classified as low risks. The odds were 11.04 times greater for defendants who had DJS involvement, either in the community or residential, to be recommended for retention within the adult system than for defendants who had no DJS involvement. Last, the odds were 5.41 times greater for younger defendants to be recommended at the time of the offense for transfer to the juvenile system than for older defendants to be recommended.
Discussion
Agreement Between Judges' Decisions and Evaluators' Opinion
The results of the study confirmed our hypothesis that judges' decisions correlate strongly with forensic evaluators' opinions. The current study provided basic information regarding the number of times the judge's decision varied from the forensic evaluator's. The level of agreement is encouraging, considering that judges are privy to additional information that may not be available to evaluators. In particular, judges may have more extensive details about the offense and the potential culpability of the defendant. This additional information can be highly influential when the judge decides whether to maintain jurisdiction in the adult court.
Factors Contributing to the Judge's Decision
As stated previously, the evaluator's ultimate opinion was the strongest correlation to and predictor of the judge's decision, but several other factors were found to correlate highly with it. A defendant's amenability to treatment was found to be the factor that correlated the most highly, possibly because judges may avoid allocating resources for those youths with little rehabilitative potential. Public safety risk was the second most highly correlated factor. Judges may be reluctant to transfer youths at highest public safety risk back to juvenile court because of a concern that a less secure DJS facility might be insufficient for containment.
Other factors found to correlate significantly with the judge's decision may be supporting factors of broader categories. For instance, being younger, having a limited history of receiving services in the past, and being enrolled in school may make a defendant more amenable to treatment. Having previous offenses and using drugs and weapons may increase a defendant's public safety risk.
We hypothesized that being a sole perpetrator of the crime and injuring the victim would correlate with a judge's final decision to transfer a defendant from adult to juvenile court. Our data did not support this hypothesis. This result was unexpected, because we thought that these factors would influence judges' opinions on public safety risk. It seems that a judge's decision to transfer is supported by the juvenile's pattern of offenses and the extent of services the juvenile received from DJS after prior juvenile court contacts rather than the details of one particular offense. It is also possible that other factors such as injury of the victim are too specific and are included in larger constructs (e.g., risk to public safety).
Factors Contributing to the Forensic Evaluators' Opinion
Since the most important factor influencing the judges' decisions was forensic evaluators' opinions, it follows that determining what factors contribute most significantly to the evaluators' opinions should be investigated. As with the judge's final decision, the two most important factors in the evaluator's opinion were amenability to treatment and public safety risk. More specific factors such as a history of DJS involvement, age at time of the offense, and a history of previous offenses were again identified as important correlates and predictors. A unique factor that correlated less to the judge's decision but correlated more to the evaluator's opinion was the defendant's emotional maturity. One explanation for this difference could be that emotional maturity is a characteristic that mental health professionals are specifically trained to determine. Therefore, emotional maturity may play a larger role in influencing the opinion of the evaluator.
In general, the opinion of the evaluator and the decision of the judge seem to be supported by the same factors. This similarity could be interpreted in various ways. One explanation could be that in evaluating the factors in each case, there is general agreement on which factors are most important when deciding on the jurisdiction in which the case should be handled. The alternative explanation is that judges simply defer to evaluators' opinions when making their decisions, without making their own independent analysis.
Limitations and Future Research
Although our research shed light on several aspects of the transfer evaluation, the relatively small number of cases created some limitations. The small sample was particularly a problem in performing the logistic regression models. Unlike the univariate analyses, which include all of the cases that have valid responses to the two variables of interest, logistic regression requires that all cases have valid responses to all variables. Therefore, cases are deleted if there is a missing value for any one of the variables. For example, while amenability to treatment was a significant predictor in the univariate analyses, because 13 percent of the cases had this factor missing, the overall sample size was reduced, the likely result being that the variable was not found to be a significant predictor in the logistic regression.
Another limitation was the possible overlap of some of the variables. For example, DJS involvement correlated highly with some other variables, which may explain why amenability to treatment was not found to be a significant variable in the logistic regression when that evaluator's opinion was the outcome variable. It is likely that past DJS involvement is an important factor considered by evaluators when assessing overall amenability to treatment.
In addition to strengthening the correlations that were discovered in the study, a larger sample size would allow researchers to investigate the nature of the predictor variables found in this study, such as risk to public safety and amenability to treatment. These variables are likely to be large constructs, and future research could focus on breaking them down to provide a better understanding of what evaluators are truly considering when making recommendations.
We gathered data from cases in Baltimore City in 2004. To make the results more generalizable, it would be important to include a broader range of data across jurisdictions and states. Analysis of newer cases would also be helpful in determining whether trends have changed in recent years.
A benefit of this research could be the creation of more standardized methods when clinicians conduct transfer of jurisdiction evaluations. For instance, if we can learn what evaluators regard as the most important predictor variables, then those variables could be assembled into a questionnaire to be used when performing evaluations. Such standardization may make the process of deciding jurisdiction less complicated.
Acknowledgments
The authors would like to thank Martin F. Sherman, PhD, for his assistance with the statistical design and analysis of this article.
Footnotes
-
Disclosures of financial or other potential conflicts of interest: None.
- © 2012 American Academy of Psychiatry and the Law