Mental health mitigating evidence and judicial outcomes for federal capital defendants

+

MODEL

Available online at www.sciencedirect.com

ScienceDirect International Journal of Law, Crime and Justice xx (2015) 1e25 www.elsevier.com/locate/ijlcj

Mental health mitigating evidence and judicial outcomes for federal capital defendants Shana D. Stites a,*, Katherine K. Dahlsgaard b,1 a

Einstein Center for Health Policy and Research, Einstein Healthcare Network, 5501 Old York Rd, Philadelphia, PA 19141, USA b Department of Child and Adolescent Psychiatry and Behavioral Science, The Children's Hospital of Philadelphia, 3440 Market Street Suite 200, Philadelphia, PA 19104, USA

Abstract Purpose: Limited prior empirical research has focused on the impact of mental health mitigating evidence on capital trial outcomes. Method: This study conducted a novel examination of the relationship between documented mental health mitigating illness evidence and the outcomes of the Attorney General's review and sentencing phase of trial. Archival data of capital trial proceedings from 1995 to 2000 were analyzed. Results: Statistically controlling for the severity of the underlying offense, the judicial outcomes were similar for defendants who had and did not have mitigating evidence about mental illness documented. Conclusions: The current evidentiary construct of mental health mitigation may not be sufficient to reduce judgments in the sentencing phase of trial or the U.S. Attorney General's proceedings. Further investigation is warranted to foster reliable and discernible relationships between mental health mitigating evidence and judicial outcomes. © 2015 Elsevier Ltd. All rights reserved.

Keywords: Federal; Trial outcomes; Mitigating evidence; Mental illness; Capital punishment; Death penalty

* Corresponding author. Tel.: þ1 215 456 1015. E-mail addresses: [email protected] (S.D. Stites), [email protected] (K.K. Dahlsgaard). 1 Tel.: þ1 215 590 1172; fax: þ1 215 590 7855. http://dx.doi.org/10.1016/j.ijlcj.2015.02.002 1756-0616/© 2015 Elsevier Ltd. All rights reserved.

Please cite this article in press as: Stites, S.D., Dahlsgaard, K.K., Mental health mitigating evidence and judicial outcomes for federal capital defendants, International Journal of Law, Crime and Justice (2015), http://dx.doi.org/ 10.1016/j.ijlcj.2015.02.002

+ 2

MODEL

S.D. Stites, K.K. Dahlsgaard / International Journal of Law, Crime and Justice xx (2015) 1e25

1. Introduction The death penalty is an ultimate and irrevocable punishment. According to the U.S. Supreme Court, it is “qualitatively different from a sentence of imprisonment, however long” (Woodson v. North Carolina, 1976). The death penalty thereby demands a greater degree of reliability in its application (Lockett v. Ohio, 1978). However, prior research suggests that significant problems exist with the administration of capital punishment (Wolfers, 2006; Beck and Shumsky, 1997; Bersoff, 1987; Bersoff and Ogden, 1987; Bowers, 1983) and considerable contention currently surrounds the use of mental health evidence in the system (Deitchman et al., 1991; United States v. Fields, 2007). It has been found that verdicts of guilt are not always synonymous with actual responsibility for the crime (Scheck et al., 2000; Wells et al., 2000). A particularly compelling example is the recent application of DNA technology, which has resulted in 225 postconviction determinations of actual innocence, with more than half of these involving persons sentenced to death at trial (Innocence Project Case Profiles, 2008). Death penalty prosecutions may involve persons with mental retardation and serious mental illness, such as Schizophrenia and Bipolar Disorder. It is conservatively estimated that 5e10% of death row inmates suffer from serious mental illness and about 60 people with mental illness or mental retardation have been executed (Mental Illness and the Death Penalty in the United States, 2005). Furthermore, experts estimate that the number of individuals with serious mental disorders who are prosecuted by the American Judicial System has increased as a result of deinstitutionalization, more rigid criteria for civil commitment, limited access to community treatment, and negative or dismissive social attitudes (Lamb and Weinberger, 1998). It could be reasoned that, in part as a consequence of mental illness mitigating evidence, mental health conditions would be more prevalent among inmates in the general prison population than in inmates on death row. However, the opposite has been found. Inmates on death row are disproportionately affected by mental illness, compared to both the general community population and other incarcerated groups (Cunningham and Vigen, 2002). Despite the high prevalence of mental illness diagnoses among the condemned and the gravity of the decisions in the American capital justice system, insufficient research has evaluated the effects of mental illness data on capital case mitigation, and results of these studies have been inconsistent (Barnett et al., 2004, 2007). Evidence of a defendant's mental illness is intended to decrease the likelihood that the U.S. Attorney General would recommend the prosecuting attorney seek the death penalty and decrease the chances that a jury or judge would render a sentence a death. The purpose of the current research was to examine whether or not reliable and predictable statistical associations existed between the presence of documented mental health conditions and the outcomes of both the United States (U.S.) Attorney General's review and sentencing phase of capital trials from 1995 to 2000. We describe each of these judicial processes later in more detail. 1.1. Theory of punishment and the death penalty Debates related to capital punishment have included arguments to abolish it (Groombridge, 2008) and arguments to improve consistency when such practices are applied (Furman v. Georgia, 1972). The origins of punishment are based in efforts to maintain society through the enforcement of social rules and norms. However, capital punishment is unique from other aspects of the American Judicial System because it administers the ultimate and most severe punishment, death. The laws that embody this part of the judicial system contain an explicit Please cite this article in press as: Stites, S.D., Dahlsgaard, K.K., Mental health mitigating evidence and judicial outcomes for federal capital defendants, International Journal of Law, Crime and Justice (2015), http://dx.doi.org/ 10.1016/j.ijlcj.2015.02.002

+

MODEL

S.D. Stites, K.K. Dahlsgaard / International Journal of Law, Crime and Justice xx (2015) 1e25

3

assumption of eligibility. This assumption is conveyed in the language that is used to describe a person who has violated one of the laws: he or she becomes eligible for a penalty of death. The implied assumption is that the administration of punishment against those who violate the laws thereby protects the rights of those who do not breach the laws (Foucault, 1977). As a result of the intended functions of the capital punishment system (i.e., punish those who break the laws and protect those who do not), there appear to be both individual-level and social phenomena that support the authorization of harm to others. The justifications for capital punishment typically include deterrence and retribution. These justifications serve as the underpinnings for maintaining the death penalty in the United States. They imply maintenance of the capital punishment system to: (a) facilitate public safety by communicating a threat to punish others who might commit similar acts (Archer et al., 1983; Decker and Kohfeld, 1984; Donohue and Wolfers, 2005), and (b) provide a mechanism for the victims to seek retribution (Carlsmith, 2006; Gregg v. Georgia, 1976). 1.2. Due process and mens rea: justice in the administration of punishment The perpetuation of the capital punishment system is supported by the perception that it is necessary to administer punishment, including penalties of death. However, of equal or greater importance to the American judicial system is the idea that the means by which punishment is inflicted must be just. For example, a person must be prosecuted through the accepted process, which includes such aspects as legal representation and a fair trial, and the individual must be culpable for the crime for which he or she is being punished. Justice in the American judicial system is largely contained within due process and mens rea. Mens rea is the determination of the defendant's moral culpability for the offense. For example, a person may have murdered another but may be found not morally culpable because he or she did not have the mental capabilities to understand the moral violation of the crime. Mitigation is the process for determining culpability or blameworthiness for the crimes committed. It involves a specific legal process whereby evidentiary factors are introduced into trials (U.S. Department of Justice, 2000). These mitigating factors are introduced into court proceedings as evidence intended to reduce the defendant's moral culpability for his or her actions and are intended to result in relatively less severe outcomes, included reduced charges and lesser sentences (see Acker and Lanier, 1994). The value an evidentiary fact may have in a case or how its value compares to another fact is assigned by a judge or 12-person jury. 1.3. Mental illness mitigation The intended purpose of introducing psychological data as evidence in capital trials is to reduce the severity of the potential consequences (e.g., the severity of the charge for which a defendant is found guilty or the harshness of the punishment). Findings from the limited amount of prior research have been inconclusive regarding the specific effects of psychological factors in mitigation (Baldus et al., 1983; Beck and Shumsky, 1997; Cowan et al., 1984; Garvey, 1998; Gillespie et al., 2014; Keil and Vito, 1995; McPherson, 1995; Paternoster, 1984; White, 1987). Notably, there have been no published studies that specifically examine mental health data as mitigating evidence on federal case outcomes, nor has investigation queried the impact of such evidence on the U.S. Attorney General's authorization to seek the death penalty. Please cite this article in press as: Stites, S.D., Dahlsgaard, K.K., Mental health mitigating evidence and judicial outcomes for federal capital defendants, International Journal of Law, Crime and Justice (2015), http://dx.doi.org/ 10.1016/j.ijlcj.2015.02.002

+ 4

MODEL

S.D. Stites, K.K. Dahlsgaard / International Journal of Law, Crime and Justice xx (2015) 1e25

Results from the prior research that has been conducted may be inconclusive because of the complexity of the issue. For instance, some research suggests that the actual effects of introducing psychological information may be contrary to the intended purpose (Logan, 1989; Edens et al., 2005). In a study of juror decision-making, for example, McPherson (1995) found abuse, neglect, and lack of parental guidance and support to be among the most frequent factors associated with mock jurors' recommendations of more severe penalties. Therefore, the psychological data intended to be mitigating was in actuality aggravating. Juries may be much more sympathetic to one type of psychological factor (i.e., mental retardation) than to another (i.e., mental illness; Barnett et al., 2004, 2007; Garvey et al., 2004; Gillespie et al., 2014). Stevenson et al. (2010) coded mock jurors' comments about a defendant's history of abuse during deliberations and found that jurors were more likely to either ignore the defendant's aversive history or even use it as an aggravator rather than to consider it as a mitigator. Juries have also been found to respond differently when considering the evidentiary value of substance disorders. Bjerregaard et al. (2010) reported that alcohol related mitigating factors had no impact on the outcomes of North Carolina capital murder trials, whereas drug related factors were associated with an increased risk of a death sentence. Given the high rate of comorbidity of mental health disorders among capital defendants (Salekin et al., 2010; Sevilla, 1999), the inconsistency in juror valuation of discrete types of psychological evidence may pose a particularly serious dilemma in valuating multiple pieces of psychological evidence from a given defendant. These inconsistences also raise somber policy-based concerns related to the ways in which psychological factors are operationalized in the policies and procedures that guide the adjudication process. Mitigation concerning a defendant's psychopathy is of particular controversy. Many courts have struggled with whether or not the functional impairment associated with psychopathy is sufficient to mitigate culpability (DeMatteo and Edens, 2006; Lyon and Ogloff, 2000; Walsh and Walsh, 2006). Sevilla (1999), while acknowledging the controversy, argued that antisocial personality disorder is a medically recognized condition, causing impairments in cognitive and volitional functions. In this way, it is comparable to other medically recognized diagnoses, such as mental retardation, and should not be permitted to be used as a factor justifying a death sentence. Measures of psychopathy, most notably the Hare Psychopathy Checklist e Revised (Hare, 1991, 2003), have been used to support claims by the prosecution that a defendant represents an ongoing or future threat to society. In a sample of undergraduate students, twice as many participants supported a death sentence (60%) when testimony indicated a defendant was psychopathic as compared to when testimony indicated a defendant was psychotic (30%) or not mentally disordered (38%; Edens et al., 2005). 1.4. Obtaining and presenting psychological evidence In addition to issues about how psychological data are valuated by jurors and judges, several concerns have been raised related to which, if any, information about a defendant's mental health status are introduced for consideration in these proceedings. Most saliently, an individuals' level of functioning might affect their ability to complete necessary psychological measures and hinder their willingness to participate in the evaluations (Melton et al., 2007; Podboy and Kastl, 1993). Defendants with mental illness may refuse to allow information about their conditions to be valuated as an evidentiary fact (Sandys et al., 2008). These limitations can also impede their participation in the defense strategy. Please cite this article in press as: Stites, S.D., Dahlsgaard, K.K., Mental health mitigating evidence and judicial outcomes for federal capital defendants, International Journal of Law, Crime and Justice (2015), http://dx.doi.org/ 10.1016/j.ijlcj.2015.02.002

+

MODEL

S.D. Stites, K.K. Dahlsgaard / International Journal of Law, Crime and Justice xx (2015) 1e25

5

The legal requirements for specific types of mitigating evidence are defined by statute (Federal Death Penalty Act, 1994; Woodson v. North Carolina, 1976). This is also the case for aggravating factors (Gregg v. Georgia, 1976; Godfrey v. Georgia, 1980). However, across legal jurisdictions e such as in a federal court compared to a given state's court e the requirements of a specific evidentiary factor may be different. This can lead to defendants deemed mentally retarded in one jurisdiction not meeting such criteria in another jurisdiction (Reardon et al., 2007). Even with codified requirements, establishing that a defendant meets these requirements can be a complex matter. In the case of mental retardation, these criteria include determining that a defendant has significant deficits in intellectual capabilities and adaptive behavior as well as has experienced these problems with an onset that occurred sometime from the developmental period to age 21 (APA, 2013). Confirming a diagnosis in the mild range of retardation, in which adaptive behavior deficits are less easily quantified, can be particularly challenging yet also crucial given most individuals with mental retardation fall within this range (Sandys et al., 2008). Differences in content, standardization, and floor and ceiling effects of the measures that are used to make these quantifications can all contribute to variability in the outcomes (Committee on Disability Determination for Mental Retardation, 2002). Given these measurement issues and naturally-occurring differences between individuals even between two defendants who both might meet a given evidentiary requirement the specific degrees of impairment will often differ. Inconsistencies in the quantification make it easier for opposing legal counsel to undermine the validity or/and seriousness of impairments. Additionally, apparent inconsistencies between the quantified level of functioning and the perceived level by juries can impact outcomes (Sandys et al., 2008). Even in cases where the validity or tenacity of psychological evidence is not questioned, opposing counsel may nonetheless undermine the information as mitigating evidence and argue that it is in fact an aggravating circumstance. During the sentencing phase of Cruz v. Johnson (2000), for example, the prosecutor argued that the defendant's mental retardation was an aggravating factor that warranted the death penalty: The Defense may tell you that, you know, he is not very smart. And they may try to show you that this should be some mitigation of punishment … But the main issue that you have to look at, does the fact that the defendant was intoxicated or the fact that he may not be very smart, does that make him any less dangerous? Does that make him any less of a threat to the rest of society … And I would submit to you that it doesn't make him any less dangerous. I would submit to you, it's the opposite. It makes him in fact, more dangerous. It's part of the outlook of Oliver Cruz that makes him what he is. And that's not going to change. And society cannot take the chance of having him on the streets again, or having him out in prison where there's other people that associate with him, also, for their safety. Research has found that jurors may be influenced by the appearance and behavior of the defendants while in the courtroom. For example, a defendant who appears sorry and sincere rather than bored or disengaged is less likely to be given a penalty of death (Antonio, 2006). Sandys et al. (2008) suggest that individuals with mental retardation can implicate themselves to jurors in the courtroom because they tend to appear competent (when they are not actually) and they can display inappropriate or unfounded emotional reactions. Edens et al. (2005) found undergraduate participants who perceived greater levels of psychopathy from a defendant were more likely to support a death sentence.

Please cite this article in press as: Stites, S.D., Dahlsgaard, K.K., Mental health mitigating evidence and judicial outcomes for federal capital defendants, International Journal of Law, Crime and Justice (2015), http://dx.doi.org/ 10.1016/j.ijlcj.2015.02.002

+ 6

MODEL

S.D. Stites, K.K. Dahlsgaard / International Journal of Law, Crime and Justice xx (2015) 1e25

1.5. Perceptions of dangerousness A juror's perceptions of how dangerous a defendant may be can also impact trial outcomes. These perceptions of dangerousness can be derived from observing a defendant in the courtroom but most often it is inferred from information provided by expert testimony. Voluminous prior research has shown high correlations between perceptions of dangerousness and death sentences (Barefoot v. Estelle, 1983; Cunningham, 2006; Cunningham and Reidy, 1999, 2002; Krauss et al., 2004; Marquart and Sorensen, 1989; Regnier, 2004). These findings suggest that because mental illness is often associated with dangerousness it may be an aggravator for some juries. Risk assessments and predictions of future violence are often presented by experts during sentencing to estimate the likelihood that the defendant might perpetrate violent acts in the future. This information has implications for sentence determinations and behavioral treatment planning (Cunningham and Goldstein, 2003; Cunningham and Reidy, 2002). The effectiveness of such assessments to predict future violence and risk assessment is a matter of contention (Cunningham and Reidy, 2002). As far back as the 1970s, research emerged that suggested that violence and risk assessments employed highly questionable practices. For example, Steadman and Cocozza (1978) analyzed dangerousness evaluations of 257 incompetent criminal defendants. They reported that psychiatrists appeared to make decisions about predictions of future violence based almost exclusively on the legal charges the defendant was facing rather than the defendant's overall pattern of behavior. These findings suggest that an individual being charged with a violent offense would be far more likely to be determined to have a high level of dangerousness than a defendant charged with a non-violent offense. 1.6. Additional factors in the courtroom During proceedings, factors in the courtroom other than those already discussed can also impact the decisions of jury members (Danziger et al., 2011). Jurors can be influenced by the attitudes they hold towards the type of defense that is introduced. For example, jurors have been found to hold negative attitudes towards the defense of insanity, which is a strategy sometimes used by defense counsel to argue that the defendant committed the crime but should not responsible for the consequences because of serious mental illness (Butler and Wasserman, 2006). The choice of certain defense strategies may increase the propensity of jurors to convict a defendant. The Capital Jury Project, a large consortium of university-based research studies, has also found evidence of several problems with juror decision-making, including erroneous beliefs that a death sentence is required and underestimation of alternative penalties (Bowers and Foglia, 2003). Court protocols require the judge to instruct jury members that they may not consider the race, national origin, gender, or religious beliefs of the defendant or the victim in reaching a verdict. Each juror must also certify that he or she did not consider these characteristics of the defendant or victim in reaching the determination. Nevertheless, defendant and victim race as well as other demographic characteristics have been found to influence trial outcomes (Beck and Shumsky, 1997; Bowers, 1983; Dunn et al., 2006; Keil and Vito, 1995; Paternoster, 1991; Paternoster and Kazyaka, 1988; Sorensen and Wallace, 1995; Thomson, 1997; Williams et al., 2007). Findings from many studies suggest the possibility that demographic characteristics influence trial outcomes; however, Berk et al. (2005) suggested that there have been problems in the approaches researchers have used to examine racial biases in Please cite this article in press as: Stites, S.D., Dahlsgaard, K.K., Mental health mitigating evidence and judicial outcomes for federal capital defendants, International Journal of Law, Crime and Justice (2015), http://dx.doi.org/ 10.1016/j.ijlcj.2015.02.002

+

MODEL

S.D. Stites, K.K. Dahlsgaard / International Journal of Law, Crime and Justice xx (2015) 1e25

7

capital cases and that these findings may be weak or misleading. The equivocality in the findings might also suggest that demographic factors inconsistently influence case outcomes. Research has also suggested defendants represented by court-appointed public counsel are more likely to be convicted than those who are able to retain private counsel (Baldus et al., 1990; Beck and Shumsky, 1997). These findings suggest that economic disparities may influence trial outcomes. The ability to retain private counsel is in large part contingent upon the defendant's financial resources. Individuals of lower socioeconomic statuses are more likely to be represented by court-appointed rather than privately retained counsel (Johnson and Johnson, 2001). Finally, the results of a study conducted in Maryland showed that certain counties within the state were more likely to pursue capital charges than were others (Paternoster and Brame, 2003). These results suggest that even within the same legal jurisdiction, idiosyncratic characteristics of the locality may influence the outcomes of cases. 1.7. Aspects of the judicial process relevant to the present study The laws that define capital punishment in the American Judicial System have been challenged and changed over the past several decades. In 1972, the United States Supreme Court issued a ruling that invalidated capital punishment throughout the United States (Furman v. Georgia, 1972). The federal government later reinstated capital punishment with the enactment of the Anti-Drug Abuse Act (1988) and subsequent Federal Death Penalty Act (1994) and Antiterrorism and Effective Death Penalty Act (1996); however, between 1995 and 2000, the vast majority of crimes subject to the death penalty were defined according to the Federal Death Penalty Act (1994). The Federal Death Penalty Act (1994) contains 60 offenses for which a guilty defendant could receive a death sentence. These capital-eligible offenses include crimes of murder such as kidnapping resulting in death and murder for hire. They also include certain crimes not resulting in death, such as leading a large-scale drug enterprise. The federal death penalty differs from the death penalty at the state level in that the federal death penalty encompasses a variety of crimes beyond that of first degree murder, including terrorism and large-scale drug trafficking. The federal death penalty can also be applied within any state whether or not that particular state has the death penalty. The primary intent of the federal death penalty during the time in which the current study's data were collected was, broadly, to protect federal interests (e.g., antiterrorism, international or interstate drug trafficking). In some instances, a case may have been tried in federal court rather than state court because the federal interest in the prosecution was more substantial than the interests of the state (see USDOJ, 2000). Although the policies and statutes that governed the federal death penalty remained constant during the period of the current study, contentious debate has more recently emerged about expanding the federal death penalty so that it more often supersedes state jurisdictions (see United States v. Pleau, 2012). Adjudication of federal capital-eligible cases typically consists of review by the U.S. Attorney General's committee and two-part trial. In January 1995, the process for pre-trial review of federal capital cases changed to include a full appraisal by a designated committee of the U.S. Attorney General. Prosecuting attorneys became required to submit for review all cases in which a defendant was charged with a capital-eligible offense, whether or not the prosecutor intended to seek the death penalty. The U.S. Attorney General in consultation with committee members either authorizes or denies the prosecutor's ability to seek a death sentence (United States Department of Justice (USDOJ), 2000). This pre-trial review was implemented with Please cite this article in press as: Stites, S.D., Dahlsgaard, K.K., Mental health mitigating evidence and judicial outcomes for federal capital defendants, International Journal of Law, Crime and Justice (2015), http://dx.doi.org/ 10.1016/j.ijlcj.2015.02.002

+ 8

MODEL

S.D. Stites, K.K. Dahlsgaard / International Journal of Law, Crime and Justice xx (2015) 1e25

the goal of reducing disparities in how often and for what crimes a death penalty may be sought in federal cases. The U.S. Attorney General's review of death-eligible cases is a multi-step process. The prosecutor sends a recommendation to seek or not seek the death penalty, a memo describing the basis for the recommendation, and supporting documentation to the USDOJ Capital Case Unit (CCU). The CCU staff creates a record that contains the prosecutor's information and information from the respective defense attorneys. The U.S. Attorney General's committee reviews the CCU record, hears from interested parties, and occasionally gathers additional information. The committee recommends to the U.S. Attorney General whether or not the death penalty should be authorized. The U.S. Attorney General makes the final decision, which almost always conforms to the review committee recommendation. When the U.S. Attorney General files an authorization, the death penalty is not required but rather made an option in the sentencing deliberation at the discretion of the prosecutor (USDOJ, 2000). Federal prosecutors retain discretion in terms of whom they chose to prosecute, for what crime, and what sentence they seek (USDOJ, 2000). Prosecutorial discretion has been associated with arbitrariness in sentencing (Powell and Cimino, 1995; Rabin, 1972). The U.S. Attorney General's committee-based decision whether to seek or not seek a death-penalty may reflect a more reliable measure of the prosecutorial suitability to pursue a death sentence compared to a unilateral decision by a prosecuting attorney. The degree of discretion available to federal prosecutors is similar for state prosecutors with the exception that state prosecutors have sole discretion to pursue a death penalty when available by statute and jurisdiction. 1.8. Summary Capital case adjudication is an important component of the American judicial system insofar as it determines an individual's culpability and suitability for a death sentence. Mental illness data are evidentiary factors introduced during trial to mitigate more severe determinations such as a death sentence. Prior research suggests that while mental illness evidence may be defined consistently by statute, its influence on trial outcomes may be impacted by idiosyncratic factors. Nonetheless, the conceptual framework given by the American justice system is that all mitigating evidentiary factors, independent of specific type, are intended to reduce the blameworthiness of defendants who are prosecuted for very serious offenses. However, the effectiveness of this broad legal construct has been the subject of little empirical evaluation. Minimal empirical examination has evaluated the association between mental illness evidence and federal capital case and no prior research on this relationship in the U. S. Attorney General Review. The present study had two aims: (1) explore whether mental illness data are associated with a lower likelihood that the U.S. Attorney General recommended that a death penalty be sought, and (2) examine whether a death penalty is less likely when mental illness is documented as mitigating evidence during sentencing. Due to homogeneity of the final offenses in the study data, the outcomes of the first trial phase were not examined. 2. Method 2.1. Design This research was a correlational study of the outcomes of the U.S. Attorney General's review and sentencing proceedings associated with mental illness documented as provable Please cite this article in press as: Stites, S.D., Dahlsgaard, K.K., Mental health mitigating evidence and judicial outcomes for federal capital defendants, International Journal of Law, Crime and Justice (2015), http://dx.doi.org/ 10.1016/j.ijlcj.2015.02.002

+

MODEL

S.D. Stites, K.K. Dahlsgaard / International Journal of Law, Crime and Justice xx (2015) 1e25

9

mitigating evidence in federal capital cases. Publicly available archival data originally collected by Klein and Berk (2006) were obtained from the Inter-university Consortium for Political and Social Research (ICPSR). The data included all federal capital-eligible cases submitted to the USDOJ CCU by 94 U.S. Attorney Offices beginning January 1, 1995 and decided by December 31, 2000. Klein and Berk (2006) constructed coding rules for recording data on victims, defendants, and case characteristics from the CCU's hard-copy case files. The CCU file documents abstracted by study coders included those submitted by the prosecuting attorney, the indictment, the U.S. Attorney General review committee's memoranda, the U.S. Attorney General's decision letter, and documents submitted by the defense. Five trained coders used case abstraction forms to enter information into a database. For quality assurance, about one-third of each coder's cases were validated. Any discrepancies elicited more thorough reviews of case files and coding forms. Case files often contained inconsistent information about an incident. Uncertainty in the records was managed by recording each of the multiple claims. The final database included information on up to 13 assessments of statutory and nonstatutory aggravating factors and mitigating factors separately for the prosecuting attorney, defense attorney, and U.S. Attorney General's review. Each evidentiary factor was classified by the respective legal counsel according to their provability. Only factors rated as considered provable by the prosecutor or U.S. Attorney General's committee were included in the current analyses. Case dispositions were obtained from the Public Access to Court Electronic Records (PACER) system for all but 50 of the defendants. The sentences for 38 of these 50 cases were located using data from the Federal Bureau of Prisons' SENTRY database, Executive Office for U.S. Attorneys' central system file, and the Administrative Office of the U.S. Courts' criminal master file. The remaining 12 cases without final sentencing dispositions were excluded from analysis. Further description of the data collection method is provided by Klein and Berk (2006). This research was approved by the ICPSR and the Institutional Review Board of Chestnut Hill College. 2.2. Study sample The sample consisted of 667 primary defendants. Eleven defendants charged with capitaleligible crimes during the data collection period were minors at the time of the offense. Primary defendants less than 16 years of age were excluded from the analyses because such cases undergo a different judiciary process (Roper v. Simmons, 2005; Stanford v. Kentucky, 1989). Cases where the defendant's age was unknown (n ¼ 40) were also excluded. 2.3. Mental health status Several variables related to a defendant's mental health status were documented in the CCU record. Elements included a history of head injury, a past or current substance abuse or dependence disorder, categorical ranking of Intelligence Quotient (IQ), and a past or current diagnosis of mental illness (e.g., emotional disorder or serious mental illness). Intellectual impairment was defined in the current study according to the Diagnostic and Statistical Manual of Mental Disorders, 4th edition text revision (American Psychiatric Association, 2013) as an IQ less than 70 points. However, due to limited information in the study dataset, it could not be Please cite this article in press as: Stites, S.D., Dahlsgaard, K.K., Mental health mitigating evidence and judicial outcomes for federal capital defendants, International Journal of Law, Crime and Justice (2015), http://dx.doi.org/ 10.1016/j.ijlcj.2015.02.002

+ 10

MODEL

S.D. Stites, K.K. Dahlsgaard / International Journal of Law, Crime and Justice xx (2015) 1e25

determined whether these individuals had impairments in adaptive behavior, which is also part of the diagnostic criteria for mental retardation (APA, 1980, 2013). Defendants who had documentation of one or more of these aforementioned conditions in the CCU record were categorized as having a mental health disorder or condition. The dataset did not contain information to permit analysis of other diagnostic categories, such as personality disorders or general medical conditions. Although it was not possible to discern whether a specific mental health condition was current or in the past, mental illness as an evidentiary factor is usually considered mitigating only if present at the time of the crime. The current study examines concurrent documentation of mental health disorders and mental illness related mitigating evidence. 2.4. Predictor variables and study outcomes The primary study outcomes are: (1) death penalty verdict in the sentencing phase of trial, and (2) the U.S. Attorney General's authorization to seek a death sentence. Each defendant had only one decision from each the US Attorney General's review and sentencing phase of trial. Predictor variables are, respectively, (1) documented, provable mental illness mitigating evidence during sentencing and (2) presence of provable, pre-trial mental illness data during the U.S. Attorney General's review. Of defendants in the study sample, the U. S. Attorney General authorized pursuit of the death penalty in 146 cases, and a death penalty verdict was rendered by a jury in 19 instances. The primary aim was to examine the relationship between mental illness mitigating evidence and the aforementioned judicial outcomes. Other factors expected to contribute to these trial outcomes were also examined: documented non-mental illness mitigating and aggravating evidence, and severity of the underlying offense. Each of these factors is described in turn. The Federal Death Penalty Act (1994) specifies seven mitigating and 13 aggravating factors. These evidentiary factors can be introduced during the U.S. Attorney General's review, a trial's sentencing phase, or combination thereof. The general purpose of these factors is to enter into evidence aspects of the offense or characteristics of the defendant or victim that should be weighed in the judicial verdict. Mitigating factors include any of several specific reasons listed in the statute, as well as anything else “in the defendant's background, record, or character or any other circumstance of the offense that mitigates against imposition of the death sentence” (Woodson v. North Carolina, 1976). Of the seven mitigating factors defined in the Federal Death Penalty Act (1994), three relate to the defendant's mental health status (i.e., mental illness or impairment, mental disturbance, and impaired mental capacity) and the other four relate to nonmental health mitigating factors (i.e., equally culpable defendants, lack of a prior criminal record, minor participation, and consent by the victim). Table 2 shows the odds ratios for each of the three mental illness mitigating factors relative to defendant mental health status. An aggravating factor is any circumstance that increases the culpability for a criminal act. The exact list of aggravating factors depends on the nature of the offense; however, statutory factors typically include killing multiple victims, committing a capital offense against particularly vulnerable victims or high-level public officials, committing the offense in a particularly heinous manner, or having previous convictions for other serious offenses. Aggravating factors were also defined according to a stable set of requisite requirements (Godfrey v. Georgia, 1980), which permitted unencumbered analysis of data from multiple years. Barnett (1985) examined the circumstances and verdicts of 600 death-eligible cases to devise classification rules that separated cases into homogenous subsets based on the probability of a more severe trial outcome due to greater severity of offense. All cases within the Please cite this article in press as: Stites, S.D., Dahlsgaard, K.K., Mental health mitigating evidence and judicial outcomes for federal capital defendants, International Journal of Law, Crime and Justice (2015), http://dx.doi.org/ 10.1016/j.ijlcj.2015.02.002

+

MODEL

S.D. Stites, K.K. Dahlsgaard / International Journal of Law, Crime and Justice xx (2015) 1e25

11

same class are similarly likely to incur a death penalty and the risk of incurring a death penalty increases with each rank. Using Barnett's algorithm, three measures of offense severity were scored: (a) certainty that the defendant is a deliberate killer (e.g., presence of physical evidence); (b) lethality and brutality of the killing (e.g., death of victim, use of torture); and (c) status of the victim as being either a stranger or known to the defendant. More than one indicator of greater offense severity nearly doubled the odds that a death penalty was recommended (OR ¼ 1.85, p ¼ .004). Additionally, defendants with all three of these factors were more than five times as likely to be sentenced to death compared to defendants with two or fewer of these factors (OR ¼ 5.23, p ¼ .005). Severity of offense (Barnett's scale) was included as a predictor variable in the multivariable analyses. 2.5. Statistical analyses Descriptive statistics were used to characterize defendants. T tests with unequal variances and Fisher's exact tests were used to compare defendant and offense characteristics for those with and without mental health disorders. Odds ratios (OR) and Fisher's exact 95% confidence intervals were calculated to compare trial method (jury vs. judge), mental illness mitigating factors, non-mental illness mitigating factors, and aggravating factors. Estimates compare the incidence of each of the aforementioned factors for two groups (i.e., with and without documented mental health disorders) (Schmidt and Kohlmann, 2008). ORs equal to one indicate that the event was similarly likely in the two groups; less than one indicate that the event was more likely in the first group; greater than one the event was more likely in the second group. All dichotomous variables were coded 0 for absent and 1 for present. Logistic regression was used to model the association between mental illness mitigating evidence and judicial outcomes while statistically controlling for other codified evidentiary factors and offense characteristics. The respective outcomes were: (1) U.S. Attorney General's authorization to seek a death sentence, and (2) death penalty rendered by jury during sentencing. Logistic regression models were used to predict differences between the study groups Allison (1999). Given the relatively small number of positive outcomes in some instances, exact logistic regression was used to ensure calculation of robust estimates (Courvoisier et al., 2011; Vittinghoff and McCulloch, 2007). This approach permitted entry of four covariates in our model of sentencing, which had 19 death sentence verdicts. In separate subanalyses, we examined the associations of each defendant race, victim race, and victim gender and the variables in main logistic regression analyses. Analyzing each characteristic separately avoided statistically over-fitting a given model. The test statistic reported from the logistic regression models is the Wald statistic, which compares the maximum likelihood estimate of the parameters of the predictor variables to the null value. The Wald statistic is the analog for the R2 statistic for linear regression. A notable limitation of the Wald is that it may underestimate the magnitude of the difference in models with small sample sizes (Menard, 2002). Phi correlations were used to assess collinearity between study variables in the multivariable models. When intercorrelation was detected between predictor variables, two statistical tests were used to assess the degree and impact of the association. The Variance Inflation Factor (VIF) was calculated to determine the magnitude of the increase in variance resulting from the addition of each predictor. VIFs for all models were less than four and well within the threshold (Belsley et al., 1980). Sensitivity analyses were used to assure the robustness of all statistical models. All statistical tests were two-sided and p values less than .05 were considered Please cite this article in press as: Stites, S.D., Dahlsgaard, K.K., Mental health mitigating evidence and judicial outcomes for federal capital defendants, International Journal of Law, Crime and Justice (2015), http://dx.doi.org/ 10.1016/j.ijlcj.2015.02.002

+ 12

MODEL

S.D. Stites, K.K. Dahlsgaard / International Journal of Law, Crime and Justice xx (2015) 1e25

statistically significant. All analyses were conducted using SAS software v 9.1.3. (SAS Institute, Cary, NC, USA). 2.6. Power calculations Sufficient statistical power existed to enable accurate and reliable judgments based on the results of the analyses. At an alpha level of .05, there was 99% power to detect differences in the probability of the U.S. Attorney General's authorization to seek a death penalty based on documentation of provable mental illness mitigating factors. At an alpha level of .05, 94% power existed to detect differences in the probability of a death sentence based on mental illness mitigating factors. 3. Results 3.1. Description of the sample About one-third (32%) of defendants with documented lifetime prevalence of mental health disorders had documented provable mental illness mitigating evidence in sentencing proceedings. As shown in Table 1, the vast majority of defendants who had mental illness evidence in the Attorney General's review, sentencing phase, or both were male (94%). About 41% of these defendants were African American, which was a smaller proportion compared to defendants without mental health conditions. Defendants with mental health conditions were more likely to be White compared to other defendants, 26% vs 15% respectively. About onequarter (27%) was Latino American, which was similar to defendants without documented mental health disorders. Defendants with mental health disorders were also slightly older than other defendants (Mean ¼ 2 years, p ¼ .01). 3.2. Description of the prosecuted cases Defendants who had mental health disorders were indicted for a similar number of death-eligible offenses (M ¼ 1.61, SD ¼ .80) compared to defendants without these conditions [(M ¼ 1.58, SD ¼ .78); t(614) ¼ .46, p ¼ .64]. Cases where the defendant had mental health disorders had a Table 1 Defendant mental health status by demographic characteristics, 1995e2000. Defendant characteristic

Mental health disorder (n ¼ 233)

No mental health disorder (n ¼ 383)

p Valuea

Age, mean (SD) Male, % (95% CI)

28 (8.1) 94 (90e97)

27 (10.0) 97 (94e98)

.01 .40

Race/ethnicity, % (95% CI) African American Hispanic/Latino American White American Other

41 27 26 6

54 30 15 2

.002 .46 .001 .09

(35e48) (21e33) (20e32) (4e10)

(49e59) (25e35) (11e19) (0e4)

Note. Percents are column percents. Hispanic/Latino American includes those defendants identified as White or Black with an ethnicity of Hispanic/Latino. The race category of Other includes: American Indian, Asian, and mixed racial heritage. Fishers CI ¼ 95% confidence interval for Fisher's exact test. a p Values based on Fisher's exact test for categorical variables and bootstrapped t tests for continuous variables.

Please cite this article in press as: Stites, S.D., Dahlsgaard, K.K., Mental health mitigating evidence and judicial outcomes for federal capital defendants, International Journal of Law, Crime and Justice (2015), http://dx.doi.org/ 10.1016/j.ijlcj.2015.02.002

+

MODEL

S.D. Stites, K.K. Dahlsgaard / International Journal of Law, Crime and Justice xx (2015) 1e25

13

similar number of co-defendants (M ¼ 1.51, SD ¼ .93), compared to cases where the defendant did not have any of these conditions [(M ¼ 1.68, SD ¼ 1.23); t(614) ¼ 1.82, p ¼ .07; Table 2]. About 52% of defendants who had mental health conditions were charged with causing the death of a person through the use of a firearm and 20% were charged with committing murder in aid of racketeering activity. The percents of defendants without mental health conditions were 50% and 27%, respectively. Additionally, 17% of defendants who had mental health conditions were charged with statute violations related to controlled substances, whereas 33% of defendants without documented mental health conditions were charged with these offenses. This pattern was similar for statute violations related to continuing criminal enterprise; 13% of defendants with mental health conditions were charged with these violations, whereas 25% of defendants without mental health conditions were charged with these statute violations. Cases where the defendant had mental a health disorder involved fewer victims (M ¼ 1.57, SD ¼ 1.66) than cases where the defendant did not have these conditions [(M ¼ 2.27, SD ¼ 2.40); t(607) ¼ 3.91, p < .0001]. More defendants who had mental health disorders were charged with offenses that included a female victim (29%), compared to other defendants (15%; p ¼ .035). Defendants with mental health disorders were similarly as likely as other defendants to be prosecuted by jury trial (OR ¼ .81, p ¼ .22. Data not shown). 3.3. Defendant mental health status and evidentiary factors Findings presented next describe associations between (a) defendant mental health status and mitigating factors, and (b) aggravating and mitigating factors. As expected, documentation of provable mental illness mitigating factors was far more likely when the defendant had a documented mental health disorders than when the defendant did not. For example, more defendants with mental health disorders had a mitigating factor related to mental illness or impairment documented in the U.S. Attorney General's proceedings than defendants without these conditions (OR ¼ 13.57, p < .0001; Table 3). 3.4. Concurrent evidentiary factors Defendants with documented aggravating evidence were no more or less likely to have mental illness evidence documented during either sentencing or the U.S. Attorney General's Table 2 Description of the prosecuted cases by defendant mental health status in U.S. Attorney General's review and federal capital sentencing proceedings, 1995e2000. Case characteristic

Mental health disorder (n ¼ 233)

No mental health disorder (n ¼ 383)

Number of death-eligible offenses, mean (SD) Number of co-defendants, mean (SD) Number of victims, mean (SD) White victims, %

1.61 (.80) 1.51 (.93) 1.57 (1.66) 29

1.58 (.78) 1.68 (1.23) 2.27 (2.40) 15

Top most frequent charges, % Causing the death of a person through use of a 52 firearm Committing murder in aid of racketeering activity20 Controlled substances 17 Continuing criminal enterprise 13

50 27 33 25

Please cite this article in press as: Stites, S.D., Dahlsgaard, K.K., Mental health mitigating evidence and judicial outcomes for federal capital defendants, International Journal of Law, Crime and Justice (2015), http://dx.doi.org/ 10.1016/j.ijlcj.2015.02.002

+ 14

MODEL

S.D. Stites, K.K. Dahlsgaard / International Journal of Law, Crime and Justice xx (2015) 1e25

review (both p-values  .57; Table 4). In contrast, aggravating factors were more likely to be absent when non-mental illness evidence was documented in the U.S. Attorney General's review (OR ¼ .59, p ¼ .003). Defendants who had aggravating evidence documented during sentencing were no more or less likely to have non-mental illness mitigating factors (OR ¼ 1.14, p ¼ .57). 3.5. Severity of offense Defendants who had mental illness evidence documented in the Attorney General's review had more factors that indicated greater severity of offense (M ¼ 2.98, SD ¼ .75) than other defendants [(M ¼ 2.80, SD ¼ .76); t(558) ¼ 1.98, p ¼ .048; Table 5]. Likewise, defendants with mental illness evidence documented during sentencing had more factors that indicated greater severity of offense (M ¼ 2.99, SD ¼ .75) than other defendants [(M ¼ 2.79, SD ¼ .77); t(594) ¼ 2.45, p ¼ .01]. Thus, defendants with mental illness evidence were more likely to have (a) evidence that increased the certainty that they committed the offense, (b) murdered or tortured the victim, and (c) committed the offense against a stranger compared to defendants who did not have mental illness. Offense severity (Barnett's scale) was added as a predictor variable in the logistic regression models. 3.6. Multivariable analyses testing study hypotheses Multivariable logistic regression models were used to test the following two a priori hypotheses: H1: A decision by the US Attorney General to seek a death penalty is less likely when mental health disorders were documented in the record. Table 3 Defendant mental health status by mental illness mitigating factor, U.S. Attorney General's review and federal capital sentencing proceedings, 1995e2000. Mental Illness mitigating factor

Mental health disorder (n ¼ 233)

No mental health disorder (n ¼ 383)

%

%

*OR

95% CI Lower

Upper

Mental illness or impairment U.S. Attorney General review Sentencing

7 7

0 1

13.57 6.38

3.09 2.09

59.64 19.50

Mental disturbance U.S. Attorney General review Sentencing

7 6

0 0

21.79 20.27

2.83 2.62

167.83 157.15

Impaired mental capacity U.S. Attorney General review Sentencing

9 16

1 1

17.31 20.02

4.00 6.04

74.84 66.32

Note. Mental Illness or Impairment includes diagnoses of addiction, mental illness, mental retardation, and those involving altered mental status. CI ¼ 95% confidence interval for odds ratio (OR). *All odds ratios statistically significant at p < .0001.

Please cite this article in press as: Stites, S.D., Dahlsgaard, K.K., Mental health mitigating evidence and judicial outcomes for federal capital defendants, International Journal of Law, Crime and Justice (2015), http://dx.doi.org/ 10.1016/j.ijlcj.2015.02.002

+

MODEL

S.D. Stites, K.K. Dahlsgaard / International Journal of Law, Crime and Justice xx (2015) 1e25

15

Table 4 Severity of offense by defendant mental health status in the U.S. Attorney General's review and federal capital sentencing proceedings, 1995e2000. No mental health disorder (n ¼ 383)

p Valuea

U.S. Attorney General's review Severity factors, mean (SD) 2.98 (.75)

2.80 (.76)

.048

Sentencing Severity factors, mean (SD)

2.79 (.77)

.01

Offense severity

Mental health disorder (n ¼ 233)

2.99 (.75)

Note. Higher values indicate greater severity. Offense severity (Barnett's scale) is measured by the presence of: (a) evidence that increased the certainty that they committed the offense, (b) murdered or tortured the victim, and (c) committed the offense against a stranger compared to defendants who did not have mental illness. a p Values based on bootstrapped unpaired t tests.

Table 5 Presence of aggravating evidence concurrent with mitigating evidence by judicial proceeding, 1995e2000. Aggravating factor

OR

95% CI

p Value

Lower

Upper

U.S. Attorney General's review Mental illness mitigating factor Non-mental illness mitigating factor

.99 .59

.60 .42

1.62 .84

.96 .003*

Sentencing Mental illness mitigating factor Non-mental illness mitigating factor

1.14 1.28

.73 .90

1.79 1.82

.57 .17

Note. Defendant mental illness condition coded: 0 ¼ Absent, 1 ¼ Present. CI ¼ 95% confidence interval for odds ratio (OR). *p < .05.

H2: A death penalty is less likely in the presence of documented provable mental illness mitigating evidence during sentencing.

3.6.1. Model 1 A four-predictor logistic model was fitted to the data to test the research hypothesis that a decision by the U.S. Attorney General to seek a death penalty is less likely when mental health data are included in the record. The results of the final model are presented in Table 6. The model was able to predict the U.S. Attorney General's authorization to seek a death penalty better than chance alone (Wald ¼ 106.42, p < .0001). Mental illness evidence was not an independent predictor of the U.S. Attorney General's authorization to seek a death penalty (OR ¼ 1.24, p ¼ .20). In contrast, the number of aggravating factors increased the odds that the U.S. Attorney General would recommend to seek a death penalty (OR ¼ 2.68, p < .0001) and non-mental illness mitigating evidence reduced the likelihood the U.S. Attorney General recommended seeking a death penalty (OR ¼ .71, p < .001). 3.6.1.1. Analyses of individual mental illness mitigators. In a subanalysis, the three specific types of mental illness mitigating evidence (i.e., mental illness or impairment, mental disturbance, and impaired mental capacity) were analyzed as separate covariates in the logistic Please cite this article in press as: Stites, S.D., Dahlsgaard, K.K., Mental health mitigating evidence and judicial outcomes for federal capital defendants, International Journal of Law, Crime and Justice (2015), http://dx.doi.org/ 10.1016/j.ijlcj.2015.02.002

+ 16

MODEL

S.D. Stites, K.K. Dahlsgaard / International Journal of Law, Crime and Justice xx (2015) 1e25

Table 6 Logistic regression models predicting U.S. Attorney General's decision to seek a death sentence and jury trial death sentence, 1995e2000. OR

95% CI Lower

p Value Upper

Model 1. Predicting U.S. Attorney General's decision to seek the death penalty (N ¼ 560) Mental illness mitigating factors 1.24 .89 1.72 Non-mental illness mitigating .71* .58 .89 factors Aggravating factors 2.68* 2.16 3.31 Offense severity (Barnett's scale) 1.15* 1.04 1.28 Model 2. Predicting a death sentence (N ¼ 281) Mental illness mitigating factors .003 Non-mental illness mitigating .65 factors Aggravating factors 1.38* Offense severity (Barnett's scale) 1.00

0 .38

100 1.11

1.10 .80

1.72 1.25

.20 .0001 <.0001 .0095

.96 .11 .005 .98

Notes. Decision to seek a penalty of death coded as 1 ¼ Seek, 0 ¼ Not seek. Death penalty coded 1 ¼ Sentenced to death, 0 ¼ Sentence other than death. CI ¼ 95% confidence interval for odds ratio (OR). *p < .05.

regression model predicting whether or not the U.S. Attorney General recommended a death penalty be sought. None of these factors were independent predictors in the model (all pvalues > .05). Documentation of impaired mental capacity trended towards an increased likelihood that seeking a death penalty would be recommended (OR ¼ 2.59, p ¼ .07). No material differences were found in the odds of authorization to seek a death sentence based on the presence vs. absence of mitigating evidence related to either mental illness or mental impairment/disturbance (both p-values > .05). 3.6.1.2. Analyses of defendant and victim characteristics. Because defendants with mental health conditions were more likely to be White compared to other defendants, defendant and victim race characteristics were coded into two dichotomous variables, where White defendants and White victims were the variable level of interest (coded 1) compared to other race categories (coded 0). We found no statistically significant interaction effects between mental illness mitigating evidence and the race of either the defendant or victim (both p > .05). Defendants who were White appeared more likely to receive a death penalty authorization from the U.S. Attorney General (Wald ¼ 13.89, p ¼ .0002). This association was no longer statistically discernible in analyses that controlled for the presence of aggravating factors (Wald ¼ .002, p ¼ .96). Defendants who were prosecuted for offenses against at least one White victim were contrasted with defendants who were prosecuted for offenses that did not include White victims. Defendants prosecuted on offenses that included White victims appeared more likely to receive a death penalty authorization from the U.S. Attorney General (Wald ¼ 8.01, p ¼ .005). Adjusting statistically for the presence of aggravating factors, victim race was not significantly associated with the U.S. Attorney General's decision (Wald ¼ .13, p ¼ .72). The U. S. Attorney General was more likely to authorize seeking the death penalty in cases were defendants were prosecuted for offenses against female victims (Wald ¼ 12.36, Please cite this article in press as: Stites, S.D., Dahlsgaard, K.K., Mental health mitigating evidence and judicial outcomes for federal capital defendants, International Journal of Law, Crime and Justice (2015), http://dx.doi.org/ 10.1016/j.ijlcj.2015.02.002

+

MODEL

S.D. Stites, K.K. Dahlsgaard / International Journal of Law, Crime and Justice xx (2015) 1e25

17

p ¼ .0004). However, controlling statistically for the severity of offense (Barnett's measure), these defendants were as likely as other defendants, who were prosecuted for offenses that did not include female victims, to have the U. S. Attorney authorize seeking a death penalty (Wald ¼ .13, p ¼ .71). 3.6.2. Model 2 A second four-predictor logistic model was fitted to the data to test the hypothesis that a death sentence would be less likely when provable mental illness mitigating evidence was documented in sentencing. The model predicted a death sentence with more certainty than chance alone (Wald ¼ 13.14, p ¼ .0015; Table 4). Mental illness mitigating evidence was not an independent predictor in the model (Wald ¼ .003, p ¼ .96). In contrast, the number of aggravating factors was the strongest predictor of whether or not a death sentence was rendered (OR ¼ 1.38, p ¼ .005). Although non-mental illness mitigating factors appeared to trend toward statistical significance ( p ¼ .11), aggravating evidence was the only predictor to reach statistical significance ( p < .05). 4. Discussion The primary comparisons in our study show mental illness mitigating factors were not associated with the outcomes of either the sentencing phase or the U.S. Attorney General's review. The findings from the current study raise an important question about mental health information as mitigating evidence. Why is it not associated with less severe penalties? We discuss a few possible answers within the context of our study's results. Findings from the current study show that federal capital cases involving defendants with mental illness evidence differed from other cases. Most saliently, cases in which the defendants had evidence of mental illness were also more likely to have physical evidence, increased heinousness of the underlying offense, and commission of underlying offenses against strangers. The association between the presence of mental illness evidence and increased number of the aforementioned factors was a consistent finding for both the sentencing phase and U.S. Attorney General's review. Compared to defendants without these conditions, this research suggests the defendants in federal capital cases who had documented mental health conditions were more likely to be White (26% and 15%), have victims who were White (39% vs. 25%) and female (29% vs. 15%). Their crimes were also more likely to include (a) evidence that increased the certainty that they committed the offense, (b) murder or torture of the victim, and (c) an offense against a stranger. Notably, in bivariate analyses, all of these characteristics were associated with increased likelihood of more severe outcomes. These characteristics are arguably better explained by the circumstances of the underlying offense rather than by the presence of a mental health condition. Consistent with apparent differences in the underlying offenses, defendants who had mental health conditions were more likely to have been prosecuted for violent offenses that did not include drug related or criminal enterprise infractions: About 52% were charged with causing the death of a person through the use of a firearm and 20% were charged with committing murder in aid of racketeering activity. These findings were similar for defendants who did not have mental health conditions (50% and 27%, respectively). However, about 17% of defendants with mental health conditions were charged with offenses related to manufacturing or dispensing controlled substances, whereas one-third of other defendants were charged with this Please cite this article in press as: Stites, S.D., Dahlsgaard, K.K., Mental health mitigating evidence and judicial outcomes for federal capital defendants, International Journal of Law, Crime and Justice (2015), http://dx.doi.org/ 10.1016/j.ijlcj.2015.02.002

+ 18

MODEL

S.D. Stites, K.K. Dahlsgaard / International Journal of Law, Crime and Justice xx (2015) 1e25

offense. About 13% were charged with statute violations that included continuing criminal enterprise whereas one-quarter of defendants without mental health conditions were charged with this offense. Although a reasonable argument can be offered that most defendants in the sample were convicted for violent crimes, defendants who had mental health conditions were more likely to be convicted for solely violent offenses rather than violent drug related offenses. The implicit difference between these categorizations of violent offense and violent drug related offense might suggest that defendants with mental health conditions were more likely to be prosecuted for statute violations that imbue a characterization of hostile aggression or impulsivity rather than instrumental aggression (as suggested by violent drug enterprise related offenses). Our results raise questions about why these significant differences existed between the type of underlying offense perpetrated by defendants with and without mental health conditions. Although the results provide descriptors and report on the presence of differences, it is beyond the scope of the current study to explicate the relationship between individuals' biopsychosocial functioning and characteristics of underlying criminal offenses. Prior studies examining mental illness and offense severity have shown, that while the overall base rates are low, people with mental illness are more at risk to both offend and commit violent offenses (Elbogen and Johnson, 2009; Fazel et al., 2009; Hodgins, 1992; Shaw et al., 2006; Steadman et al., 1998; Swanson, 1994; Swanson et al., 1990, 2002). It was not possible to determine from our study data specifically how these mental illness mitigating factors were used by legal counsel or whether they were accepted by juries, and these limitations in our data may contribute to the reasons that we did not find mental illness mitigating evidence to be associated with capital case outcomes. Our findings raise questions about whether legal counsel may be apt to pursue using mental illness mitigating evidence in cases with a given typology of mental health conditions or offense characteristics. Alternatively, the circumstances of the offense and legal counsel's choice of mental illness mitigators may be unrelated. Prior research suggests that mental illness mitigating evidence may lead to quite varied responses from jurors, depending on the specific evidentiary factor, and at times intended mitigators may in actuality be valuated as aggravators (Barnett et al., 2004, 2007; Bjerregaard et al., 2010; Edens et al., 2005; Garvey, 2004; Gillespie et al., 2014; McPherson, 1995; Stevenson et al., 2010). It is plausible that legal counsel may formulate a defense strategy with these various elements in mind when choosing to pursue a given mental illness mitigator. Given that criminal lawyers may not be proficient at predicting case outcomes (GoodmanDelahunty et al., 2010) investigation into the effectiveness of their choices and strategies concerning mental illness mitigators may be useful in identifying best practices. This area of further study may help to illuminate the reasons defense attorneys may or may not pursue mental illness mitigation as well as potentially uncover options for defense attorneys to pursue when defendants possess a mental impairment. Our inclusive but general category of mental illness mitigators may have been too broad or too narrow to be a reliable predictor of case outcomes. However, this broad construct should be inversely associated with the severity of case outcomes. This expectation is derived from the conceptual underpinnings of the capital justice system, which state this as the intended purpose of mitigating evidence. We may have failed to find empirical support for an association between the general construct of mental illness mitigating evidence and case outcomes because the specific types of evidence that comprised this category had varied effects (Barnett et al., 2004, 2007; Bjerregaard et al., 2010; Edens et al., 2005; Garvey, 2004; Gillespie et al., 2014; McPherson, 1995; Stevenson et al., 2010). The failure to find the expected associations is Please cite this article in press as: Stites, S.D., Dahlsgaard, K.K., Mental health mitigating evidence and judicial outcomes for federal capital defendants, International Journal of Law, Crime and Justice (2015), http://dx.doi.org/ 10.1016/j.ijlcj.2015.02.002

+

MODEL

S.D. Stites, K.K. Dahlsgaard / International Journal of Law, Crime and Justice xx (2015) 1e25

19

concerning. Given the complexity and comorbidity of mental health conditions among capital defendants, it may be difficult, or impossible, to separate these elements within legal proceedings for a given defendant (see Slobogin, 2000). Further investigation and multidisciplinary dialog are warranted to assure the most efficacious and appropriate use of mental illness evidence by the courts, including its presentation to the court by forensic mental health experts, the legal parameters surrounding its inclusion as evidence, and the ways in which it may be used in judge and jury decision-making. Failure to introduce mental health mitigating evidence could have contributed to the lack of empirical support for its association with capital case outcomes. Mental health mitigating evidence was only documented in approximately 21% of instances where defendants had mental health conditions. Two major reasons might account for this disparity. First, The Spangenberg Group (2007) conducted a thorough financial analysis of resources available to both prosecuting and defense counsels in capital cases. The results of their study were striking: The defense counsel in indigent capital cases had the equivalent of less than half (43%) of the financial resources that were available to the counsel for the prosecution. The legal precedent established in Powell v. Alabama (1932) recognized that the inequality in resources between defense and prosecution counsels can improperly impact litigation outcomes. The inequality in financial resources could suggest the defense counsel may be limited in the ability to obtain mental health mitigating evidence. Second, the defendants' level of functioning could affect their ability to complete necessary psychological measures and hinder their willingness to participate in the evaluations (Horberg, 2005; Melton et al., 2007; Podboy and Kastl, 1993). The defendants' mental state also influences their level of participation in the defense strategy. In some instances, the prosecutor will agree to a plea bargain because of the defendant's mental state, but the defendant will not accept such a plea and, as a result, these cases proceed to trial. Because defendants with documented mental health conditions in our sample were more likely to be White and more likely to have White victims, our results appear consistent with prior research that suggests violent offenders tend to perpetrate crimes against victims who are members of the same racial group (Harrell, 2007; Wilbanks, 1985). Contrary to prior research suggests that capital trial outcomes may be affected by the race characteristics of both the defendant and victim (Keil and Vito, 1995; Thomson, 1997; Williams et al., 2007), our findings are consistent with the findings reported by Klein et al. (2006). Our results did not provide empirical support for the notion that either defendant or victim demographic characteristics were directly associated with the outcomes of the U.S. Attorney General's review. Due to small sample size, we were unable to examine the effects of race on sentencing decisions.

4.1. U.S. Attorney General's review Based on searches of PubMed, WestLaw, and Google Scholar, this is the first study we are aware of that examines the association between mental illness mitigation and the outcomes of the U. S. Attorney General's review. The timing of this proceeding might suggest that the availability of mental illness evidence in the U.S. Attorney General's review may not have been as exhaustive compared to the sentencing phase. The U.S. Attorney General's review typically occurs early in the information gathering process, likely before all the facts of the case are gathered into evidence and before the legal team's defense strategy may have been finalized. It is plausible that the U.S. Attorney General's recommendations were made without review of pertinent mental illness information. In comparison the U.S. Attorney General's review, it may Please cite this article in press as: Stites, S.D., Dahlsgaard, K.K., Mental health mitigating evidence and judicial outcomes for federal capital defendants, International Journal of Law, Crime and Justice (2015), http://dx.doi.org/ 10.1016/j.ijlcj.2015.02.002

+ 20

MODEL

S.D. Stites, K.K. Dahlsgaard / International Journal of Law, Crime and Justice xx (2015) 1e25

be more reasonable to expect that the majority of facts were entered into evidence by the time of the sentencing verdict. The earlier timing of the U. S. Attorney General's Review compared to sentencing proceedings is particularly notable. Our results suggest that the consistency in the introduction of mental illness mitigating evidence across the two proceedings is in contrast to the two other types of evidentiary factors; aggravating and non-mental health mitigating factors were likely to be conjointly absent in the U.S. Attorney General review (OR ¼ .59, p ¼ .003) but similarly likely to be concurrently used in the sentencing phase ( p > .05). The constancy in the probability of the mental illness mitigating factors across judicial proceedings suggests that legal decisions related to use of mental health mitigating factors were likely determined early in the litigation process. These findings might suggest that mental health mitigating evidence was introduced only in cases when impairments in the defendant's mental health status were readily identifiable and clearly admissible early in litigation. The valuation of evidence in the U.S. Attorney General's review is a committee-based decision made by individuals with legal expertise. In contrast, valuation of evidence during sentencing is carried out by jurors who are vetted as average members of American society and capable of rendering a death verdict. Although both non-mental illness mitigating factors and aggravating factors were independent predictors of the U.S. Attorney General's recommendation, mental illness evidence was not an independent predictor and thus may be systematically different than other types of evidence. Barnett et al. (2004, 2007) found interpretation of mental illness evidence may be more subjective and idiosyncratic than other types of evidence when valuated by mock jurors. It is unknown whether legal experts such as members of the U.S. Attorney General's review committee valuate mental illness evidence similarly to non-expert jurors. Given that members of the committee are legal experts in the federal death penalty, it could be expected that a committee-based decision made by a group proficient in the legal content would consistent with the rules and guidelines that govern that decision. The U. S. Attorney General's review was initiated in an effort to reduce arbitrariness in the prosecution of federal capital cases. However, notably, about 30% of all inmates initially sentenced to death had their sentences commuted or found unconstitutional at a later time (Argys and Mocan, 2004). Although Argys and Mocan did not detail the reasoning behind these overturned sentences, the rate of 3 in 10 is substantial. Although the creation of a centralized review process for federal capital cases seems an important advancement, the findings from the current study and those from existing research suggest additional improvements are still needed.

4.2. Limitations The results from this study should be conservatively interpreted. Many varied ideographic factors can impact the outcomes of capital case prosecution. For instance, the ways in which legal counsel chose to use or not use evidence in the defendant's prosecution was not taken into consideration in the analyses. The current study could not account for any differences in the impact of substantiated vs. unsubstantiated allegations but rather included evidence based solely on legal counsels' notion the information was provable. These limitations may have resulted in the construct of mental illness mitigating evidence being both under- and overinclusive. Please cite this article in press as: Stites, S.D., Dahlsgaard, K.K., Mental health mitigating evidence and judicial outcomes for federal capital defendants, International Journal of Law, Crime and Justice (2015), http://dx.doi.org/ 10.1016/j.ijlcj.2015.02.002

+

MODEL

S.D. Stites, K.K. Dahlsgaard / International Journal of Law, Crime and Justice xx (2015) 1e25

21

4.3. Future research The findings show that both sentencing decisions and the U.S. Attorney General's authorization to seek a death penalty are similar in the severity of judgments independent of mental illness evidence. Thus, the broad construct of mental illness mitigating evidence may not be generally effective in reducing the blameworthiness of defendants who are prosecuted for very serious offenses. The study definition of “provable mental illness” may have been too broad or too narrow. Further investigation as to which, if any, type of mental illness has a mitigating effect on the US Attorney General's and juries' decisions is recommended. The effects of mental illness evidence have yet to be evaluated in terms of their influence on crimes of varying severity. It is plausible that such evidence may influence trial outcomes for less severe cases (e.g., attempted kidnapping) but not for more severe cases (e.g., murder). Additionally, further investigation and multidisciplinary dialog are warranted to assure the most efficacious and appropriate use of mental illness evidence by the courts, including its presentation to the court by forensic mental health experts, the legal parameters surrounding its inclusion as evidence, and the ways in which it may be used in judge and jury decisionmaking.

Acknowledgments The author gratefully acknowledges the assistance of Joseph Micucci, PhD, and Julie Kessler, MD, for their professional mentorship and support. The authors also thankfully acknowledge members of the Philadelphia VA Medical Center writers' group for their professional encouragement and editorial assistance.

References Acker, J.R., Lanier, C.S., 1994. In fairness and mercy: statutory mitigating factors in capital punishment laws. Crim. Law Bull. 30, 299e345. Allison, P., 1999. Logistic regression using the SAS system: Theory and application. SAS Institute, Cary, NC. American Psychiatric Association, 2013. Diagnostic and Statistical Manual of Mental Disorders, fifth ed. Author, Washington, DC. American Psychiatric Association, 1980. Diagnostic and Statistical Manual of Mental Disorders, third ed. Author, Washington, DC. Anti-Drug Abuse Act, 100, 1988. Antonio, M.E., 2006. Arbitrariness and the death penalty: how the defendant's appearance during trial influences capital jurors' punishment decision. Behav. Sci. Law 24 (2), 215e234. http://dx.doi.org/10.1002/bsl.673. Archer, D., Gartner, R., Beittel, M., 1983. Homicide and the death penalty: a cross-national test of a deterrence hypothesis. J. Crim. Law Criminol. 74 (3), 991e1013. http://dx.doi.org/10.1177/088740348600100404. Argys, L., Mocan, N., 2004. Who Shall Live and Who Shall Die? Analysis of Prisoners on Death Row in the United States. In: NBER Working Paper Series. Available at: http://www.nber.org/papers/w9507. Baldus, D.C., Pulaski, C., Woodworth, G., 1983. Comparative review of death sentences: an empirical study of the Georgia experience. J. Crim. Law Criminol. 74 (3), 661e753. http://dx.doi.org/10.2307/1143133. Baldus, D.C., Woodworth, G., Pulaski Jr., C.A., 1990. Equal Justice and the Death Penalty. Northeastern University Press, Boston. Barefoot v. Estelle, 463 U.S. 880, 897, 1983. Barnett, A., 1985. Some distribution patterns for the Georgia death sentence. U.C. Davis Law Rev. 18, 1327e1374. Retrieved from: http://lawreview.law.ucdavis.edu/index.php.

Please cite this article in press as: Stites, S.D., Dahlsgaard, K.K., Mental health mitigating evidence and judicial outcomes for federal capital defendants, International Journal of Law, Crime and Justice (2015), http://dx.doi.org/ 10.1016/j.ijlcj.2015.02.002

+ 22

MODEL

S.D. Stites, K.K. Dahlsgaard / International Journal of Law, Crime and Justice xx (2015) 1e25

Barnett, M.E., Brodsky, S.L., Davis, C.M., 2004. When mitigation evidence makes a difference: effects of psychological mitigating evidence on sentencing decisions in capital trials. Behav. Sci. Law 22 (6), 751e770. http://dx.doi.org/ 10.1002/bsl.591. Barnett, M.E., Brodsky, S.L., Price, R.J., 2007. Differential impact of mitigating evidence in capital case sentencing. J. Forensic Psychol. Pract. 7 (1), 39e45. http://dx.doi.org/10.1300/J158v07n01_04. Beck, J.C., Shumsky, R., 1997. A comparison of retained and appointed counsel in cases of capital murder. Law Hum. Behav. 21 (5), 525e538. http://dx.doi.org/10.1023/A:1024827807469. Belsley, D.A., Kuh, E., Welsch, R.E., 1980. Regression Diagnostics: Identifying Influential Data and Sources of Collinearity. John Wiley & Sons, New York. Berk, R., Li, A., Hickman, L.J., 2005. Statistical difficulties in determining the role of race in capital cases: a re-analysis of data from the State of Maryland. J. Quant. Criminol. 21 (4), 365e390. http://dx.doi.org/10.1007/s10940-0057354-7. Bersoff, D.N., 1987. Social science data and the Supreme Court: Lockhart as a case in point. Am. Psychol. 42 (1), 52e58. http://dx.doi.org/10.1037/0003-066X.42.1.52. Bersoff, D.N., Ogden, D.W., 1987. In the Supreme Court of the United States Lockhart v. McCree: amicus curiae brief for the American Psychological Association. Am. Psychol. 42 (1), 59e68. http://dx.doi.org/10.1037/0003066X.42.1.59. Bjerregaard, B., Smith, M.D., Fogel, S.J., Palacios, W., 2010. Alcohol and drug mitigation in capital murder trials: implications for sentencing decisions. Justice Q. 27, 517e537. http://dx.doi.org/10.1080/07418820903051623. Bowers, W.J., 1983. The pervasiveness of arbitrariness and discrimination under post-Furman capital statutes. J. Crim. Law Criminol. 74 (2), 1067e1100. http://dx.doi.org/10.1177/001112878002600409. Bowers, W.J., Foglia, W.D., 2003. Still singularly agonizing: law's failure to purge arbitrariness from capital sentencing. Crim. Law Bull. 39 (51), 51e86. Butler, B., Wasserman, A.W., 2006. The role of death qualification in venirepersons' attitudes toward the insanity defense. J. Appl. Soc. Psychol. 36 (7), 1744e1757. http://dx.doi.org/10.1080/10683160701483534. Carlsmith, K.M., 2006. The roles of retribution and utility in determining punishment. J. Exp. Soc. Psychol. 42 (4), 437e451. http://dx.doi.org/10.1016/j.jesp.2005.06.007. Committee on Disability Determination for Mental Retardation, 2002. Mental Retardation: Determining Eligibility for Social Security Benefits. National Research Council, pp. 1e335. Available at: http://www.ncbi.nlm.nih.gov/books/ NBK207536/pdf/TOC.pdf. Courvoisier, D.S., Combescure, C., Agoritsas, T., Gayet-Ageron, A., Perneger, T.V., 2011. Performance of logistic regression modeling: beyond the number of events per variable, the role of data structure. J. Clin. Epidemiol. 64, 993e1000. http://dx.doi.org/10.1016/j.jclinepi.2011.06.013. Cowan, C.L., Thompson, W., Ellsworth, P.C., 1984. The effects of death qualification on jurors' predisposition to convict and on the quality of deliberation. Law Hum. Behav. 8, 53e80. http://dx.doi.org/10.1007/BF01044351. Cruz v. Johnson, 121 S.Ct. 11, 2000. Cunningham, M.D., 2006. Dangerousness and death: a nexus in search of science and reason. Am. Psychol. 61 (8), 828e839. http://dx.doi.org/10.1037/0003-066X.61.8.828. Cunningham, M.D., Goldstein, A.M., 2003. Sentencing determinations in death penalty cases. In: Goldstein, A.M. (Ed.), Handbook of Psychology: Forensic Psychology, vol. 11. John Wiley & Sons, Inc, Hoboken, NJ, pp. 407e436. Cunningham, M.D., Reidy, T.J., 2002. Violence risk assessment at federal capital sentencing: individualization, generalization, relevance, and scientific standards. Crim. Justice Behav. 29 (5), 512e537. http://dx.doi.org/10.1177/ 009385402236731. Cunningham, M.D., Reidy, T.J., 1999. Don't confuse me with the facts: common errors in violence risk assessment at capital sentencing. Crim. Justice Behav. 26 (1), 20e43. http://dx.doi.org/10.1177/0093854899026001002. Cunningham, M.D., Vigen, M.P., 2002. Death row inmate characteristics, adjustment, and confinement: a critical review of literature. Behav. Sci. Law 20, 191e210. http://dx.doi.org/10.1002/bsl.473. Danziger, S., Levav, J., Avnaim-Pesso, L., 2011. Extraneous factors in judicial decisions. Proc. Natl. Acad. Sci. 108 (17), 6889e6892. Decker, S.H., Kohfeld, C.W., 1984. A deterrence study of the death penalty in Illinois, 1933e1980. J. Crim. Justice 12 (4), 367e377. http://dx.doi.org/10.1016/0047-2352(84)90049-7. Deitchman, M.A., Kennedy, W.A., Bekham, J.C., 1991. Self-selection factors in the participation of mental health professionals in competency for execution evaluations. Law Hum. Behav. 15 (3), 287e303. http://dx.doi.org/ 10.1007/BF01061714.

Please cite this article in press as: Stites, S.D., Dahlsgaard, K.K., Mental health mitigating evidence and judicial outcomes for federal capital defendants, International Journal of Law, Crime and Justice (2015), http://dx.doi.org/ 10.1016/j.ijlcj.2015.02.002

+

MODEL

S.D. Stites, K.K. Dahlsgaard / International Journal of Law, Crime and Justice xx (2015) 1e25

23

DeMatteo, D., Edens, J.F., 2006. The role and relevance of the psychopathy checklist-revised in court: a case law survey of U.S. courts (1991e2004). Psychol. Public Policy Law 12, 214e241. http://dx.doi.org/10.1037/10768971.12.2.214. Donohue, J.J., Wolfers, J., 2005. Uses and abuses of empirical evidence in the death penalty debate. Stanf. Law Rev. 58, 791e846. Retrieved from: http://ssrn.com/abstract¼870312. Dunn, K.F., Cowan, G., Downs, D., 2006. Effects of gender and race of perpetrator and method of killing on outcome judgments in a mock filicide case. J. Appl. Soc. Psychol. 36 (10), 2395e2416. http://dx.doi.org/10.1111/j.00219029.2006.00109.x. Edens, J.F., Colwell, L.H., Desforges, D.M., Fernandez, K., 2005. The impact of mental health evidence on support for capital punishment: are defendants labeled psychopathic considered more deserving of death? Behav. Sci. Law 23 (5), 603e625. http://dx.doi.org/10.1002/bsl.660. Elbogen, E.B., Johnson, S.C., 2009. The intricate link between violence and mental disorder results from the national epidemiologic survey on alcohol and related conditions. Arch. General Psychiatry 66 (2), 152e161. http:// dx.doi.org/10.1001/archgenpsychiatry.2008.537. Fazel, S., Niklas Långstr€om, N., Hjern, A., Grann, M., Lichtenstein, P., 2009. Schizophrenia, substance abuse, and violent crime. J. Am. Med. Assoc. 301 (19), 2016e2023. http://dx.doi.org/10.1001/jama.2009.675. Federal Death Penalty Act of 1994, Title VI, 1994. Foucault, M., 1977. Discipline and Punish: the Birth of the Prison. Pantheon, New York, NY. Furman v. Georgia, 408 U.S. 153, 1972. Garvey, S.P., 1998. Aggravation and mitigation in capital cases: what do jurors think? Columbia Law Rev. 98, 1538e1576. http://dx.doi.org/10.2307/1123305. Garvey, S.P., Hannaford-Agor, P., Hans, V.P., Mott, N.L., Munsterman, G.T., Wells, M.T., 2004. Juror First Votes in Criminal Trials. Cornell Law Faculty Publications. Paper 3. Gillespie, L.K., Smith, M.D., Bjerregaard, B., Fogel, S.J., 2014. Examining the impact of proximate culpability mitigation in capital punishment sentencing recommendations: the influence of mental health mitigators. Am. J. Crim. Justice 39 (4), 698e715. http://dx.doi.org/10.1007/s12103-014-9255-5. Godfrey v. Georgia, 446 U.S. 420, 1980. Goodman-Delahunty, J., Granhag, P.A., Hartwig, M., Loftus, E.F., 2010. Insightful or wishful: lawyer's ability to predict case outcomes. Psychol. Public Policy Law 16 (2), 133e157. http://dx.doi.org/10.1037/a0019060. Gregg v. Georgia, 428 U.S. 153, 1976. Groombridge, N., 2008. Dead wrong: violence, vengeance, and the victims of capital punishment. Howard J. Crim. Justice 47 (1), 97e98. http://dx.doi.org/10.1111/j.1468-2311.2008.00508_1.x. Hare, R.D., 1991. The Hare Psychopathy Checklist e Revised. Multi-Health Systems, Toronto, Ontario. Hare, R.D., 2003. Manual for the Revised Psychopathy Checklist, second ed. Multi-Health Systems, Toronto, ON, Canada. Harrell, E., 2007. Black Victims of Violent Crime. Bureau of Justice Statistics special report, NCJ 214258, pp. 1e12. Available at: http://www.bjs.gov/content/pub/pdf/bvvc.pdf. Hodgins, S., 1992. Mental disorder, intellectual deficiency, and crime evidence from a birth cohort. Arch. General Psychiatry 49 (6), 476e483. http://dx.doi.org/10.1001/archpsyc.1992.01820060056009. Horberg, R.S., 2005. The injustice of imposing death sentences on people with severe mental illnesses. Catholic Univ. Law Rev. 54 (4), 1152e1168. Available at: http://scholarship.law.edu/lawreview/vol54/iss4/7. Innocence Project Case Profiles, 2008. Retrieved 29.11.08, from: http://www.innocenceproject.org/know. Johnson, J.L., Johnson, C.F., 2001. Poverty and the death penalty. J. Econ. Issues 35 (2), 517e523. Retrieved from: http://www.jstor.org/stable/4227684. Keil, T.J., Vito, G.F., 1995. Race and the death penalty in Kentucky murder trials: 1976e1991. Am. J. Crim. Justice 20 (1), 17e36. http://dx.doi.org/10.1300/J222v04n03_02. Klein, S.P., Berk, R.A., 2006. Race and the Decision to Seek the Death Penalty in Federal Cases, 1995e2000 [United States] [Data file]. Documentation retrieved 22.07.08 from ICPSR. Web site: http://www.icpsr.umich.edu/. Klein, S.P., Berk, R.A., Hickman, L.J., 2006. Race and the decision to seek the death penalty in federal cases. Rand Corporation, Santa Monica, CA. Krauss, D.A., Lieberman, J.D., Olson, J., 2004. The effects of rational and experiential information processing of expert testimony in death penalty cases. Behav. Sci. Law 22, 801e822. http://dx.doi.org/10.1002/bsl.621. Lamb, H.R., Weinberger, L.E., 1998. Persons with severe mental illness in jails and prisons: a review. Psychiatr. Serv. 49, 483e492. http://dx.doi.org/10.1016/S1353-1131(98)90139-0. Lockett v. Ohio, 438 U.S. 586, 1978.

Please cite this article in press as: Stites, S.D., Dahlsgaard, K.K., Mental health mitigating evidence and judicial outcomes for federal capital defendants, International Journal of Law, Crime and Justice (2015), http://dx.doi.org/ 10.1016/j.ijlcj.2015.02.002

+ 24

MODEL

S.D. Stites, K.K. Dahlsgaard / International Journal of Law, Crime and Justice xx (2015) 1e25

Logan, D.D., 1989. Is it mitigation or aggravation? Troublesome areas of defense evidence in capital sentencing. Calif. Atty. Crim. Justice Forum 16 (5), 14. Lyon, D.R., Ogloff, J.R.P., 2000. Legal and ethical issues in psychopathy assessment. In: Gacono, C. (Ed.), The Clinical and Forensic Assessment of Psychopathy: a Practitioner's Guide. Lawrence Erlbaum Associate, Mahwah, NJ, pp. 139e173. Marquart, J.W., Sorensen, J.R., 1989. A national study of the Furman-commuted inmates: assessing the threat to society from capital offenders. Loyola Los Angel. Law Rev. 23 (1), 1e29. Available at: http://digitalcommons.lmu.edu/llr/ vol23/iss1/2. McPherson, S.B., 1995. Psychosocial Investigation in Death Penalty Mitigation: Procedures, Pitfalls, and Impact. De Gruyter, Berlin. Melton, G.B., Petrila, J., Poythress, N.G., Slobogin, C., 2007. Psychological Evaluations for the Courts: a Handbook for Mental Health Professionals and Lawyers, third ed. Guilford Press, New York, NY. Menard, S., 2002. Applied Logistic Regression Analysis: Quantitative Applications in the Social Sciences. Sage, Thousand Oaks, CA, p. 106. Mental Illness and the Death Penalty in the United States, 2005. Retrieved from: http://www.aclu.org/capital/ mentalillness/10617pub20050131.html. Paternoster, R., 1984. Prosecutorial discretion in requesting the death penalty: a case of victim-based racial discrimination. Law Soc. Rev. 18 (3), 437e478. Paternoster, R., 1991. Prosecutorial discretion and capital sentencing in North and South Carolina. In: Bohm, R.M. (Ed.), The Death Penalty in America: Current Research. Anderson, Cincinnati, OH, pp. 39e52. Paternoster, R., Brame, R., 2003. An Empirical Analysis of Maryland's Death Sentencing System with Respect to the Influence of Race and Legal Jurisdiction. Department of Criminology, University of Maryland. Paternoster, R., Kazyaka, A., 1988. Racial considerations in capital punishment: the failure of evenhanded justice. In: Haas, K.C., Inciardi, J.A. (Eds.), Challenging Capital Punishment: Legal and Social Science Approaches. Sage, Newbury Park, CA, pp. 113e148. Podboy, J.W., Kastl, A.J., 1993. The intentional misuse of standard psychological tests in complex trials. Am. Coll. Forensic Psychol. 11, 47e54. Powell, W.J., Cimino, M.T., 1995. Prosecutorial discretion under the federal sentencing guidelines: is the fox guarding the hen house? W. Va. Law Rev. 97 (373), 382e395. Powell v. Alabama, 287 U.S. 45, 1932. Rabin, R.L., 1972. Agency criminal referrals in the federal system: an empirical study of prosecutorial discretion. Stanf. Law Rev. 24 (8), 1036e1091. Regnier, T., 2004. Barefoot in Quicksand: the future of “future dangerousness” predictions in death penalty sentencing in the world of Daubert and Kumho. Akron Law Rev. 37, 469e506. https://www.uakron.edu/dotAsset/727519.pdf. Reardon, M.C., O'Neil, K.M., Levett, L.M., 2007. Deciding mental retardation and mental illness in capital cases: the effects of procedure, evidence, and attitudes. Psychol. Crime Law 13, 537e557. http://dx.doi.org/10.1080/ 10683160601160018. Roper v. Simmons, 2005 (03-633) 543 U.S. 551 (2005). 112 S. W. 3d 397. Sandys, M., Trahan, A., Pruss, H., 2008. Taking account of the “diminished capacities of the retarded”: are capital jurors up to the task? DePaul Law Rev. 57, 679e700. Available at: http://via.library.depaul.edu/cgi/viewcontent.cgi? article¼1271&context¼law-review. SAS Institute Inc, 2000e2004. SAS 9.1.3 Help and Documentation [Computer software]. SAS Institute Inc., Cary, NC. Salekin, K.L., Olley, J.G., Hedge, K.A., 2010. Offenders with intellectual disability: characteristics, prevalence, and issues in forensic assessment. J. Ment. Health Res. Intellect. Disabil. 3 (2), 97e116. http://dx.doi.org/10.1080/ 19315861003695769. Scheck, B., Neufeld, P., Dwyer, W., 2000. Actual Innocence. Harper, New York. Schmidt, C.O., Kohlmann, T., 2008. When to use the odds ratio or the relative risk? Int. J. Public Health 53 (3), 165e176. http://dx.doi.org/10.1007/s00038-008-7068-3. Sevilla, D.M., 1999. Anti-social personality disorder: justification for the death penalty? J. Contemp. Leg. Issues 10, 247e262. Shaw, J., Hunt, I.M., Flynn, S., Meehan, J., Robinson, J., Bickley, H., et al., 2006. Rates of mental disorder in people convicted of homicide: national clinical survey. Br. J. Psychiatry 188, 143e147. http://dx.doi.org/10.1192/ bjp.188.2.143. Slobogin, C., 2000. Mental Illness and the Death Penalty (2000). California Criminal Law Rev. 2 (3) http://dx.doi.org/ 10.2139/ssrn.216048.

Please cite this article in press as: Stites, S.D., Dahlsgaard, K.K., Mental health mitigating evidence and judicial outcomes for federal capital defendants, International Journal of Law, Crime and Justice (2015), http://dx.doi.org/ 10.1016/j.ijlcj.2015.02.002

+

MODEL

S.D. Stites, K.K. Dahlsgaard / International Journal of Law, Crime and Justice xx (2015) 1e25

25

Spangenberg Group, 2007. Resources of the Prosecution and Indigent Defense Functions in Tennessee. Author, West Newton, MA. Retrieved from: http://www.thejusticeproject.org/wp-content/uploads/spangenbergsummary-analysis. pdf. Sorensen, J.R., Wallace, D.H., 1995. Capital punishment in Missouri: examining the issue of racial disparity. Behav. Sci. Law 13 (1), 61e80. http://dx.doi.org/10.1002/bsl.2370130105. Stanford v. Kentucky, 492 U.S. 361, 1989. Steadman, H.J., Cocozza, J., 1978. Psychiatry, dangerousness, and the repetitively violent offender. J. Crim. Law Criminol. 69, 226. Steadman, H.J., Mulvey, E.P., Monahan, J., Appelbaum, P.S., Grissa, T., Roth, L.H., Silver, E., 1998. Violence by people discharged from acute psychiatric inpatient facilities and by others in the same neighborhoods. Arch. General Psychiatry 55, 393e401. http://dx.doi.org/10.1001/archpsyc.55.5.393. Stevenson, M.C., Bottoms, B.L., Diamond, S.S., 2010. Juror's discussions of a defendant's history of child abuse and alcohol abuse in capital sentencing deliberations. Psychol. Public Policy Law 16, 1e38. http://dx.doi.org/10.1037/ a0018404. Swanson, J.W., 1994. Mental disorder, substance abuse, and community violence: an epidemiological approach. In: Monahan, J., Steadman, H.J. (Eds.), Violence and Mental Disorder: Developments in Risk Assessment. University of Chicago Press, Chicago, pp. 101e136. Swanson, J.W., Holzer, C.E., Ganju, V.K., Jono, R.T., 1990. Violence and psychiatric disorder in the community: evidence from the epidemiologic catchment area surveys. Hosp. Community 41, 761e770. Swanson, J.W., Swartz, M.S., Essock, S.M., Osher, F.C., Wagner, H.R., Goodman, L.A., 2002. The socialeenvironmental context of violent behavior in persons treated for severe mental illness. Am. J. Public Health 92, 1523e1531. PMCID: PMC1447272. The Antiterrorism and Effective Death Penalty Act of 1996, 18 U.S.C. x 3599; x(1996). Thomson, E., 1997. Discrimination and the death penalty in Arizona. Crim. Justice Rev. 22 (1), 65e76. http:// dx.doi.org/10.1177/073401689702200106. U.S. Department of Justice, 2000. U.S. Attorneys' Manual. Author, Washington, DC. United States v. Fields, 483 F.3d. 313, 358 (5th Cir. 2007). United States v. Pleau, 680 F3d 1, 2012. Vittinghoff, E., McCulloch, C.E., 2007. Relaxing the rule of ten events per variable in logistic and Cox regression. Am. J. Epidemiol. 165 (6), 710e718. http://dx.doi.org/10.1093/aje/kwk052. Walsh, T., Walsh, Z., 2006. The evidentiary introduction of psychopathy checklist-revised assessed psychopathy in U.S. courts: extent and appropriateness. J. Law Hum. Behav. 30, 493e507. http://dx.doi.org/10.1007/s10979-006-9042-z. Wells, G., Malpass, R., Lindsay, R., Fisher, R., Turtle, J., Fulero, S., 2000. From the lab to the police station: a successful application of eyewitness research. Am. Psychol. 55, 581e594. http://dx.doi.org/10.1037/0003066X.55.6.581. White, L.T., 1987. The mental illness defense in the capital penalty hearing. Behav. Sci. Law 5 (4), 411e421. http:// dx.doi.org/10.1002/bsl.2370050405. Wilbanks, W., 1985. Is violent crime intraracial? Crime Delinq. 31 (1), 117e128. http://dx.doi.org/10.1177/ 0011128785031001007. Williams, M.R., Demuth, S., Holocomb, J.E., 2007. Understanding the influence of victim gender in death penalty cases: the importance of victim race, gender-related victimization, and jury decision making. Criminol. Interdiscip. J. 45 (4), 865e891. From: http://www.allacademic.com/meta/p202278_index.html. Wolfers, J.J., 2006. Uses and Abuses of Empirical Evidence in the Death Penalty Debate, Law and Economics Workshop UC Berkeley. Available at: http://escholarship.org/uc/item/0f13k805. Woodson v. North Carolina, 428 U.S. 280, 1976.

Please cite this article in press as: Stites, S.D., Dahlsgaard, K.K., Mental health mitigating evidence and judicial outcomes for federal capital defendants, International Journal of Law, Crime and Justice (2015), http://dx.doi.org/ 10.1016/j.ijlcj.2015.02.002