Views of judges and potential jurors on responsibility for behavior in tort litigation in the genomic era

Abstract

The potential uses—and misuses—of psychiatric genetic evidence in litigation concerning defendants’ responsibility for behavior has, to date, mostly focused on criminal justice. Yet the introduction of psychiatric genetic evidence in tort litigation raises old and new legal and social questions that merit consideration. We conducted a vignette-based survey of state trial court judges (n = 465) and potential jurors (n = 2131) to assess how psychiatric genetic evidence may affect views on civil responsibility and related decisions. Psychiatric genetic evidence had limited impact on judicial decisions, but increased perceptions of the subject’s contractual incapabilities. Differences in judges’ and jurors’ views are highlighted, indicating tension between public sentiments and existing legal doctrine that disallows consideration of a person’s psychiatric condition in assessing civil liability. Unexpectedly, jurors’ gender impacted all case-related questions—the implications thereof are discussed. Future research can assess the role of education, legal training, and gender differences in judicial decision-making.

I. INTRODUCTION

Tort law concerns lawsuits for injuries perpetrated by one private party on another private party. Scholars have long debated the aims of tort law, whether its primary concern is economic restitution, cost distribution, deterrence, the enforcement of civil rights, or something else entirely.¹ Regardless, much of tort law is highly concerned with fault, not only to draw a clear causative line between the defendant’s conduct and the victim’s injury, but also to determine the defendant’s mindset and ability—both what the defendant intended and what the defendant understood about the consequences of his or her actions—while engaging in that conduct. By deciding whether and how to enforce recovery by injured parties, tort law shapes and regulates how citizens ‘may and may not treat one another and how they may expect to be treated by others.’²

Rapid developments in genetic knowledge and sequencing technology have led to the introduction of genetic information, including behavioral genetic data (ie specific biomarkers that link to particular behavioral traits³), as evidence in courts to establish a party’s responsibility for their behavior. To date, research on this issue has largely focused on judicial willingness to admit psychiatric genetic information in criminal cases, and the effects of such evidence on perceptions of criminal responsibility.⁴ However, similar questions of how to evaluate responsibility and fault, and whether genetic data can affect such evaluations, arise also in tort law. To date, genetic data have already been considered in tort cases involving medical malpractice,⁵ toxic torts and product liability,⁶ and negligence.⁷ Studies also suggest that judges, as the gatekeepers for scientific evidence,⁸ are generally receptive to genetic information. This judicial receptivity includes a willingness to admit, and even compel, genetic tests to prove or refute a plaintiff’s medical malpractice or toxic tort claim⁹ as well as in determining responsibility for civil damages.¹⁰ It is thus possible that psychiatric genetic data may soon be introduced also in civil cases with the aim of influencing evaluations of an individual’s responsibility for their behavior. Specifically, as we elaborate below, such evidence may be used to support a claim of reasonableness in a negligence claim (where defendants must be shown not to have acted according to ‘a reasonable standard’ of care), or that the defendant did not act with an intent to cause harm in an intentional tort claim (where defendants must be shown to have acted with intent to commit the act that caused harm).¹¹

I.A. Between Science and the Law: The Challenge for Psychiatric Genomic Evidence in Tort Litigation

Although admitting psychiatric genetic data in courts may seem like a natural progression given current practice, tort law has a specific and longstanding opposition to allowing the consideration of mental health conditions when evaluating liability,¹² which would encompass psychiatric genetics evidence. This is so even though evidence of a person’s psychiatric condition is admitted in criminal law and even other civil contexts, such as contract law. Given the growing public endorsement of biogenetic explanations for mental health conditions and knowledge surrounding the genetic underpinning of such conditions (see below), this situation may change in the near future. This anticipatory study aimed to forecast the ramifications of such changes, focusing on two primary areas of tort law: intentional torts (otherwise known as battery) and negligence torts.

Intentional torts arise when an actor ‘acts intending to cause a harmful or offensive contact,’ and such harmful or offensive results.¹³ In intentional torts, psychiatric conditions may be admissible in evaluating whether a defendant intended to commit a specific action¹⁴ but are excluded from consideration as regards to why the defendant committed the action.¹⁵ Thus, for example, in the cases of Johnson v. Insurance Co. of North America, 232 Va. 349 S.E.2d 616 (Va. 1986) and Polmatier v. Russ, 206 Conn. 229 A.2d 468 (Conn. 1988), the defendants were found not guilty of murder by reason of insanity in their criminal cases but were found liable of committing an intentional tort in their civil cases. In Polmatier, the defendant argued he could not be liable for an intentional tort because he killed his father-in-law while experiencing delusions of being persecuted; in Johnson the defendant was argued not to have acted with intent when he shot someone because he believed that God had ordered him to do so.¹⁶ In both cases, the courts found that the key question was whether the defendant understood the nature of what he was doing, not why the defendant committed those actions.¹⁷

Negligent torts arise when a defendant fails to act with ‘reasonable care.’¹⁸ For defendants with physical disabilities, the question is whether the defendant acted as ‘a reasonable person [would] with the same disability.’¹⁹ However, no such provision exists for defendants with a mental health disability; these defendants are expected to act as a reasonable person without the mental health condition would.²⁰ Thus, for example, in Bruenig v. Am. Fam. Ins. Co., 45 Wis. 2d 536, 173 N.W.2d 619 (Wis. 1970) and Ramey v. Knorr, 130 Wn. App. 672, 124 P.3d 314 (Wash. Ct. App. 2005), both involving negligence cases arising from car accidents, the courts held that the question was whether the defendants—both of whom experienced a delusion that led to the car collision—acted as a reasonable person would without that condition. In Bruenig, the defendant believed God had taken control of her steering wheel and drove directly into a truck because she thought her car would fly over it; in Ramey, the defendant believed that she was the subject of a conspiracy to attack her and caused a car accident as she tried to commit suicide.²¹ In both cases, the courts ruled that the defendant should be held to the objective standard of liability, ie that of a reasonable person without the psychiatric condition.

One of the primary stated reasons for this exclusion, dating back to at least 1616, is the fear that mental health conditions are difficult to prove and easy to feign.²² The Restatement (Third) of Torts acknowledges that ‘modern society is increasingly inclined to treat physical disabilities and mental disabilities similarly, and this inclination is supported by the recognition that many mental disabilities have organic causes’, but, significantly, goes on to cite ‘administrative’ concerns in how to identify and assess psychiatric conditions as a reason to maintain this distinction.²³ This longstanding rule has generated enduring debate, particularly when the historical dualism of the body–mind distinction has been long refuted. Indeed, while the earliest version of the Diagnostic and Statistical Manual for Mental Disorders (DSM) divided disorders into ‘organic syndromes’ and ‘functional disorders,’ subsequent revisions did away with this separation, recognizing that such a categorization could be falsely seen as implying that certain psychiatric conditions had no root in the physical body.²⁴ As we discuss below, emerging scientific understandings of psychiatric conditions may increasingly establish a physical component to these conditions by locating genetic markers for them, and thus address the Restatement’s longstanding concerns about the substantive and administrative feasibility of verifying and evaluating a party’s potential psychiatric condition.²⁵ The Restatement’s retention of a distinction between physical and mental health disabilities also raises the specter of enduring historical bias against people with psychiatric conditions—a population that remains highly stigmatized (see I.B. below).

The growing body of knowledge about behavioral, including psychiatric, genetics may increasingly pose a challenge to some of the primary reasons for the exclusion of mental health disabilities in tort law: namely, that mental health disabilities are hard to verify and link to behavior in a way that physical disabilities are not. On the one hand, previous uses of genetic testing in tort cases were mostly limited to more binary determinations of identity, paternity, or medical diagnosis²⁶—scenarios that are significantly different than genetics for common psychiatric conditions. Indeed, the genetic underpinnings of behavior, including psychiatric conditions, have been slow to emerge, not least because such conditions are genetically complex and involve heterogenous combinations of numerous genetic variants, each having small effects that interact with environments to produce an outcome.²⁷

On the other hand, research in psychiatric genetics is ongoing,²⁸ including efforts to establish psychiatric genetic risk scores that can be associated with certain behaviors.²⁹ Such advances are likely to continue and hence may soon be relevant for the process of determining responsibility for torts. Because psychiatric genetic data may establish a physical root of mental health conditions and positive psychiatric genetic results may in the future become part of clinical diagnosis (similar to other medical conditions),³⁰ tort law’s current strict distinction between physical and psychiatric disabilities may blur to the point of collapse. Defendants may subsequently seek to introduce psychiatric genetic evidence to argue for reduced liability, or against liability entirely, by demonstrating or verifying that the defendant has a biologically (ie ‘physically’) rooted cause for a particular psychiatric condition or behavior, thus rendering the defendant’s behavior involuntary or beyond control.³¹ Moreover, a growing awareness of the biogenetic explanations of psychiatric conditions among the general public may signal increasing willingness to consider this type of evidence.³²

Significantly, the possible introduction of psychiatric genetic data as evidence need not translate into an outright legal defense. That is: tortfeasors should not be able to evade liability simply by claiming ‘my genes made me do it,’ as some scholars and newspaper headlines have suggested in the context of criminal proceedings. Rather, growing scientific knowledge of psychiatric genetics challenges the Restatement’s continued distinction between physical and psychiatric disabilities. These developments call for re-centering liability on other grounds, such as the value of providing a means for redress to injured parties over debating finer-grained evaluations of responsibility—a discussion that ought to be had, in full consideration of the advancing body of knowledge about psychiatric conditions, including psychiatric genetics.

Whether or not the current doctrines regarding liability of individuals with psychiatric conditions change any time soon, how judicial actors will view psychiatric genetic evidence and the impact of such data in tort cases is an open question. To date, surveys of legal cases³³ and experimental studies³⁴ have found that behavioral genetic data have a limited impact on judicial decisions on criminal responsibility and sentencing. In particular, a flurry of research that arose from the introduction in criminal cases of a certain gene said to cause ‘aggressiveness’ found that the effect of such evidence was mixed at best.³⁵ However, the philosophical aims and legal principles of the civil system differ from those of the criminal system. The civil system does not have goals of moral condemnation or punishment in the same way that the criminal justice system does, and key standards for civil verdicts (eg reasonableness) require lower burdens of proof than the criminal standard of ‘beyond a reasonable doubt.’ Moreover, although tort law similarly concerns questions of responsibility, tort penalties most commonly involve monetary penalties, not incarceration. Community sentiments³⁶ and attitudes towards genetic attribution of psychiatric conditions may thus have greater impact in civil trials. No study to date has examined judicial responses to psychiatric genetic data in cases involving civil responsibility for torts, and how such data may influence judicial decisions.

I.B. Psychiatric Genomics and Stigma

Stigma is a major obstacle to social inclusion, equality, and effective treatment of persons with psychiatric conditions.³⁷ Reducing this obstacle is thus a primary goal of public health policymakers, an endeavor to which the growing field of psychiatric genomics may potentially greatly contribute.³⁸ Researchers in the field have hoped that attributing psychiatric conditions to genetics or brain disease will reduce stigma by conveying a message that such a disease is ‘a disease like any other,’ one that can be treated, controlled for, and cured, and reducing an individual’s perceived responsibility for the condition.³⁹ In reality, this hope has not materialized. While research has found greater public endorsement of biogenetic explanations, publicly held stigma, desire for social distance, and perception of dangerousness grew or remained constant.⁴⁰

In theory, legal standards and their implementations must be free from bias and prejudice. Yet tort law’s differing approaches to the civil liability of people with physical versus psychiatric disabilities raises concerns that such an approach both rests on and reinforces stigma against people with psychiatric conditions, holding them to a higher bar of behavioral liability than expected from others.⁴¹ The Restatement (Third) of Torts suggests that a person with a psychiatric condition generally cannot act reasonably,⁴² although most people with such conditions do (see, for example, meta-analysis relating to rational decisions about clinical care⁴³). There is also a range of interventions, management techniques, and calculations that a ‘reasonable’ person with a psychiatric condition could make to manage their symptoms (similar to a person with physical disability), such as deciding between treatment plans or balancing the side effects of particular medications. In claiming that ‘the law cannot expect the person [with psychiatric conditions to] wisely and appropriately to moderate conduct choices,’ the Restatement forecloses any possible discussion of what potential frameworks of liability for people with psychiatric conditions would look like. While a basic impulse of fairness would allow that individuals with physical disabilities are not responsible for simply having a disability, tort litigation may reflect a deeper, even if unconscious, view that individuals with psychiatric conditions are responsible for their own medical condition—a deeply embedded historical stigma.⁴⁴

Moreover, stigma about people with psychiatric conditions may influence both judges’ and jurors’ decisions in tort litigation;⁴⁵ studies in other judicial contexts have found that judges are influenced by the same types of biases as those held by the general public.⁴⁶ Understanding how psychiatric genomic evidence in tort litigation may affect judicial views on litigants is therefore important for the administration of justice. Because genetics may be understood in an essentialist manner (ie indicating the presence of conditions that are persistent, serious,⁴⁷ and dangerous⁴⁸), the mere introduction of psychiatric genetic risk data may unjustifiably impact judicial decisions. A deterministic view of genetics as fixed, and of resulting behaviors as unmodifiable,⁴⁹ may expand judicial perceptions of civil legal incapacity beyond the dispute at stake and result in broader losses of autonomy and legal personhood.⁵⁰

We report findings from a study that explored the effect of psychiatric genetic evidence on state trial court judges’ and potential jurors’ evaluations of liability in tort litigation, as well as the potential impact of such evidence on perceptions of a litigant’s capabilities in other contexts. Both judges and jurors play a crucial role in tort litigation. Although judges are vested with the primary judicial power, jurors represent the public, and the public’s beliefs affect policies and institutions.⁵¹ Additionally, litigants can choose whether their cases will be decided by a judge or jury (and most tort cases are jury trials⁵²), but judges and jurors differ in their training, legal experience, and demographics, differences that may impact verdicts.⁵³ Exploring how psychiatric genetic evidence impacts stakeholders’ views on responsibility in tort litigation can illuminate differences in perceptions of the role of genetics for human agency, and highlight areas for additional training or regulations.

I‌I. STUDY DESIGN

This study builds upon earlier research on how psychiatric genetic evidence (previously studied in criminal⁵⁴ and child custody cases⁵⁵) may affect judicial decisions specifically in the realm of tort litigation, specifically intentional (battery) and negligence cases. Because tort law maintains a strict distinction between physical and mental health disabilities, as described above, this study also explores current views of the public and judges on this historically rooted body–mind distinction, and whether the views of the public differ from those of judges.

I‌I.A. Methods

We designed and administered an anonymous, ~20-min long, online vignette-based survey to state trial court judges and potential jurors (see recruitment below and Table 1). The vignette provided to participants was modeled on tort cases where the defendant’s mental state was argued to be relevant (see above).⁵⁶ Our study scenario presented a man, John, described as a 30-year-old who, since his college graduation, had been steadily employed and making ‘a decent living to support his wife and baby daughter.’ The vignette stated that John recently ‘started to have strange feelings that important things were happening around him but didn’t think that anything was wrong,’ until the incident leading to the legal case: ‘One afternoon, driving home, John heard voices telling him that he was a supreme being with the power to save the world.’ The vignette then described John seeing an oncoming truck, his reaction to the truck (see below) and John’s subsequent collision with the truck that resulted in the truck driver’s injury. Finally, the vignette described the legal proceedings to recover damages for the accident (filed by the truck driver as either a battery or negligence case), including testimony from a psychiatrist about John’s mental health. In particular, the psychiatrist diagnosed John with schizophrenia; explained that this condition is often accompanied by delusions, hearing voices, and difficulty distinguishing between real and unreal events; and concluded that when the collision occurred, John ‘[had] a severe schizophrenic episode’ and ‘could not make a rational choice about what to do when he saw [the] truck approaching him.’ Finally, the vignettes provided concluding statements of both the defense and plaintiff’s attorneys and brief instructions to the decision-makers, tailored to the type of tort case at hand. For example, in the battery case, John’s attorney stated that John’s behavior was the result of delusional beliefs and without a rational intent, and in the negligence case, John’s attorney emphasized that John’s psychiatric condition is ‘a biological disorder that constitutes a physical disability’ and thus that his behavior needs to be assessed in comparison to ‘persons with a similar disability.’ In both instances, the attorneys of the plaintiff highlighted the standard of the Restatement whereby the mental health status should not be considered in deciding responsibility for civil harms (for a full description of the attorneys’ closing arguments and instructions, see Appendix; the description of the cases was similar for both judges and potential jurors).

The vignette included several independent variables that were randomly assigned, following a fully crossed, between-subjects factorial design. The first independent variable was type of tort (intentional/battery or negligence). Participants received either a description of an intentional tort/battery claim (‘John knew instantly [that the truck] had been sent to destroy him. He stepped on the gas in the belief that if he crashed into the truck, it would disintegrate’), or a negligence claim that John failed to act as a ‘reasonable person’ would under similar situations (‘John was distracted by the voices and looked around to see where they were coming from. He did not notice that he was about to crash into the truck’).⁵⁷ The second variable was whether John had experienced a prior experience of schizophrenia that had been treated (no/yes: five years earlier and treated with medication since). This variable was selected to correspond with legal doctrines that govern the liability of persons with physical disabilities, ie a prior experience of schizophrenia may increase liability under the assumption that ‘one should have known’ about, and thus taken steps to prevent, risk.⁵⁸ Participants who received a description of the battery claim were asked to determine whether John ‘intended to act in a way that would cause the harmful or offensive contact […or] acted knowing that the consequence was highly likely to occur.’ Participants who received the negligence claim were asked to determine whether John behaved ‘as a reasonable person would under the same circumstances.’

The third variable, considered the study’s primary independent variable, was genetic evidence. Participants were randomized to receive (or not) information about genetic test results showing that John has a combination of genes that predispose him to schizophrenia. This description reflects contemporary understandings of psychiatric conditions as complex conditions that involve multiple genes (as opposed to a single-gene condition). This description was provided as part of the psychiatrist’s testimony, stating that the increased risk due to genetics is grounded in research and referencing to a diagram that was presented to the participants. This variable was included because genetic test results may support defenses against liability by demonstrating a proclivity for a psychiatric condition or verifying a diagnosis that influenced individuals’ behavior beyond their control. Further, as explained above, psychiatric genetic data may establish the physical nature of psychiatric conditions because genes provide instructions for the body and could be considered an ‘organic cause’ of certain behaviors.⁵⁹ This is so regardless of whether one adopts a view of genetic determinism, ie that genes alone cause complex human behaviors⁶⁰ (which is rarely the case). Psychiatric genetic data thus challenges the existing legal doctrine that holds persons with physical disabilities liable by a standard of a reasonable person ‘with the same disability,’ but judges persons with psychiatric disabilities by the general standard of a ‘reasonable person’ with no qualifications regarding a person’s ‘mental or emotional disability.’⁶¹

The vignette was followed by instructions for the decision makers (see Appendix), as well as case-related questions to assess the effect of psychiatric genetic evidence on participants’ decisions about responsibility for behavior and genetic essentialist thinking. The first question—and primary dependent variable—was whether participants found John liable for damages. Those who received the negligence vignette were also asked whether John’s schizophrenia qualifies as a physical disability. Other questions assessed participants’ perceptions of dangerousness and control, and views of John’s capability to perform seven specific common life activities such as making marital decisions, opening a bank account, and making a decision to receive or refuse psychiatric treatment (two separate items). Response options were on a 1–4 point Likert scale (1 = ‘very capable’, 4=‘not at all capable’; reversed code in analysis; see further detail below). The seven capability items were developed based on disability scholarship and have been used in other relevant studies on the uses of psychiatric genetic data in non-clinical settings.⁶² We hypothesized that the introduction of psychiatric genetic evidence of psychiatric genetic evidence (yes/no) would be positively associated with finding John liable and that genetic information may exacerbate psychiatric-related stigma.

I‌I.B. Recruitment and Data Collection

We administered the survey to state trial court judges and potential jurors (starting with judges) with the goals of both understanding and comparing the impact of psychiatric genomic evidence on the judicial decision-making of these stakeholders, as well as to explore how such evidence may affect stigma against people with psychiatric conditions.

Judges sitting on state trial courts of general jurisdiction were recruited by the Center for Survey Research (CSR),⁶³ an academic survey research center, from The American Bench, a national directory of judges in the US.⁶⁴ Of the 11,366 judges who met the eligibility criteria, 3750 judges were randomly selected, weighted to reflect the general geographical locations (Northeast, South, Midwest, and West, as determined by the U.S. Census Bureau regions) of the entire judicial population. The selected judges were invited to participate through a mixed approach of email invitation and paper packet; both approaches included information about the study and a link to CSR’s secure website with the online survey; packets also included a hard copy of the survey instrument and a prepaid return envelope. Demographic information of judges was also collected in the survey. Non-respondents were sent two reminders with a week to three-week gap (depending on the mode of the initial invitation). A total of 465 judges completed the survey (overall response rate of 12.4%; 78 envelopes were returned as bad address and 64 judges affirmatively declined participation), see Fig. 1⁶⁵ for detailed recruitment process and response rate by recruitment method). Judges were not offered any incentives in exchange for their participation.

Potential jurors (n = 2131) were recruited by YouGov, a professional research firm that operates an Internet-based panel of the general public. Eligibility criteria were based on legal requirements for jury duty⁶⁶: adults (over 18) and US citizens who are registered to vote or hold driving licenses and are eligible for jury duty (self-report). In addition, the sample of potential jurors was selected to be representative of the U.S. population by characteristics of age, gender, race, ethnicity, years of education, and geographical region. YouGov collected participants’ demographic information, administered the survey, and offered participants ‘Polling Points’ redeemable for small gifts (equivalent to $1) for their time.

Both the judges and potential juror participants were anonymous to the research team (the recruitment companies did not share any identifying information). The N.Y. State Psychiatric Institute IRB approved the study.

I‌I.C. Statistical Analysis

Using genetic evidence as the primary independent variable, both samples were sufficiently powered to detect small effect size differences (0.18–0.20 standard deviations) with at least 80% power at a 5% significance level (Type I error rate). Descriptive statistics were used for survey responses and participant demographic characteristics. Dependent variables from the questionnaire were dichotomized (eg agree (agree/strongly agree) vs. disagree (disagree/strongly disagree)). Analysis for the capability scale was based on cumulative scores from the seven items; scores ranged from 1–4 with 2.5 midpoint and were subdivided based on capability score responses less than or equal to 2.5 and responses greater than 2.5 and dichotomized in the analysis. For case-related questions, multiple logistic regression was used to assess the impact of psychiatric genetic evidence on the dependent variables. Psychiatric genetic evidence was our primary independent variable. Type of tort (intentional/ negligence) and prior experience of schizophrenia that has been treated (no/yes) were controlled for in the analysis. In addition, the analysis controlled for demographic covariates: race (White/non-White); ethnicity (Hispanic or Latino/Not); gender (male/female); education level (see below); stakeholder group (judges/potential juror); and among judges: prior experience with tort cases (yes/no). The race variable was dichotomized in the analysis (White/non-White participants) due to limited racial and ethnic diversity among participants, particularly among judges (see Table 1). To allow for comparison by stakeholder group, the education variable was dichotomized for jurors (<4-year college degree and post-graduate education (≥MA); all judges but three had post-graduate education (these three participants were excluded from the analysis of the education variable). For all analyses, missing data were rare (generally <5%) and handled by using complete case analysis. P-values <0.05 were considered significant. Analysis of survey data was completed using the survey package (version 4.2–1) within R programming language (version 4.3.0). Specifically, the functions svydesign and svyglm from Survey package were used to account for our complex survey design, providing robust standard errors and valid inference for our primary comparisons.

I‌II. RESULTS

I‌II.A. Participant Demographics

Participants’ demographic characteristics are provided in Table 1. Among the 465 judicial participants, 72% identified as men and 28% identified as women. Judicial participants were mostly White (85.8%), followed by Black/African American (7.3%), Asian (3.7%) multiracial (1.7%), and ‘other’ (2.2%). 4.7% identified as Hispanic. Among the 2131 potential juror participants, 46.2% identified as men and 53.8% as women. Racial identification comprised of 82.3% White, 10.5% Black, 2.3% Asian, 1% American Indian/Alaska Native, 1.8% multiracial, and 2.1% other race. While the gender distribution among the potential juror participants was similar to that of the general U.S. adult population (~50%), our sample of female judges is lower than female judges’ national representation (28% of judicial participants vs. ~34% judges nationwide⁶⁷). In terms of education levels: all but one judge had at least a Master’s or a Law degree; amongst the juror participants, 31.5% had a high school degree or less, 56% had some form of undergraduate education, and 12.5% had post-graduate degrees.

I‌II.B. Civil Liability

Participants were asked to indicate whether they would find John civilly liable for battery or negligence, depending on the scenario received. Participants who read the negligence scenario were further asked to indicate whether John’s schizophrenia, as a biological disorder, qualifies as a physical disability. 50% of judges and 60.5% of potential juror who received a battery scenario found John liable, while 66.7% of judges and 62% of potential juror who received the negligence scenario found John liable. Judges who received the negligence scenario were significantly more likely than judges who received the battery scenario to find John liable (OR = 1.92, CI 1.31–2.81, P < 0.01). This difference did not exist among potential jurors (OR = 1.07, CI 0.90–1.27, P = 0.99).

Psychiatric genetic evidence had no statistically significant impact on participants’ determination of liability (OR = 0.96, CI 0.82–1.13, P = 0.554). Participants who received information that John had a prior experience of schizophrenia that had been treated were more likely to agree that he is civilly responsible (OR = 1.22, CI 1.01–1.48, P = 0.04), regardless of type of tort. Further analysis by stakeholder groups found that the impact of a prior experience of schizophrenia was statistically significant only among judges (OR = 1.54, CI 1.04–2.28, P = 0.03), not potential jurors (OR = 1.18, CI 0.95–1.46, P = 0.14), and only among judges who received the negligence scenario (OR = 2.27, CI 1.26,4.09, P = 0.006), not battery scenario (OR = 1.08, CI 0.63–1.86, P = 0.77).

I‌II.C. Perceptions of Schizophrenia as a Physical Disability

Among those who received the negligence scenario, 55% of judges and 70% of potential jurors thought that schizophrenia qualifies as a physical disability. Judges were significantly less likely than potential jurors to agree that schizophrenia qualifies as a physical disability (OR = 0.69, CI 0.56–0.87, P < 0.01). Genetic evidence (OR = 0.88, CI 0.75–1.03, P = 0.743) and prior experience of schizophrenia (OR = 1.08, CI 0.92–1.27, P = 0.252) had no impact on these views.

I‌II.D. Perceptions of Dangerousness and Control

Participants were asked to indicate their agreement with two statements: ‘John is dangerous to the public’ and ‘John has control over his actions’ (measured on a 4-point Likert scale, dichotomized in analysis). 67.6% of judges and 71.5% of potential jurors agreed that John was dangerous to the public, but participants who received the negligence scenario were significantly less likely to agree with this statement than those who received the battery scenario (OR = 0.43, CI 0.36–0.51, P < 0.01). This latter finding was significant for both judges (OR = 0.45, CI 0.30–0.67, P < 0.01) and jurors (OR = 0.43, CI 0.35–0.52, P < 0.01). Among both judges and potential jurors, evidence of a prior experience of schizophrenia did not have a statistical impact on perceptions of John’s dangerousness (OR = 1.20, CI 0.80–1.79, P = 0.33; OR = 1.09, CI 0.90–1.32, P = 0.44, respectively). Judges who received psychiatric genetic evidence were (borderline) more likely than judges who did not receive genetic evidence to agree that John is dangerous (OR = 1.51, CI 1.00–2.30, P = 0.05).

34.4% of judges and 30.6% of potential jurors agreed with the statement that “John has control over his actions,” regardless of the type of tort case that they read. Genetic psychiatric evidence (OR=0.96, CI 0.81-1.44, P = 0.212) or prior experience of schizophrenia (OR=1.12, CI 0.95-1.33, P = 0.295) did not affect participants responses to this statement.

I‌II.E. Assessments of John’s Capability

The overall capability score (dichotomized responses: somewhat/very capable vs. not very/not at all capable, per the midpoint score of 2.5—see Data Analysis section) were calculated as the average of the items (alpha = 0.85). Table 2 shows participants’ views on each capacity item. The overall sample score on the capability scale was 2.5 (SD = 0.40). Judges ranked John higher than potential jurors in overall capability (OR = 1.72, CI 1.28–2.33, P < 0.0001), as well as across all individual capabilities, except for decisions to receive or refuse medical treatment.

Genetic evidence did not significantly affect participants’ overall capability score for John (OR = 0.78, CI 0.60–1.01, P = 0.06) but did negatively impact their views on certain capability items (ie lower score on the capability item): John’s capability to make marriage-related decisions (OR = 1.38, CI 1.08–1.76, P < 0.01), sign a financial contract (OR = 1.42, CI 1.12–1.81, P < 0.01) and provide testimony in court proceedings (OR = 1.28, CI 1.03–1.59, P = 0.02). These differences remained when analyzed by stakeholder group (judges: range of P = 0.01–0.3; potential jurors: range of P = 0.02–0.04), except that providing testimony lost significance among potential jurors (OR = 1.22, CI 0.97–1.54, P = 0.09).

Participants who received evidence that John had a prior experience of schizophrenia that has been treated were more likely to rate John higher in overall capability than those who did not (3.13 average score given prior experience of schizophrenia vs. 3.04 average score without prior experience of schizophrenia; OR = 1.54, CI 1.20–1.98, P < 0.01), as well as on specific items: capability to handle financial activities (opening a bank account (OR = 1.51, CI 1.18–1.93, P < 0.01), sign a financial contract (OR = 1.46, CI 1.21–1.77, P < 0.01)) and provide testimony in court (OR = 1.31, CI 1.11–1.56, P = 0.02). Differences regarding John’s capability to sign a financial contract remained also when analyzed by stakeholder groups (jurors: OR = 1.35, CI 1.10–1.66, P = 0.001; judges: OR = 2.27, CI 1.31–3.94, P = 0.006), though differences regarding opening a bank account remained only among jurors (OR = 1.50, CI 1.15–1.96, P < 0.001) and differences regarding John’s capability to provide testimony remained only among judges (OR = 2.25, CI 1.21–4.20, P = 0.02).

I‌II.F. Demographic Trends and Impacts

Aside from gender, participants’ demographic characteristics (race, ethnicity, education level) and judges’ tort experience largely had limited to no impact on responses to the case-related or capability questions. Where differences existed, no observable trends were identified—including the variable of genetic evidence. White judges were more likely than non-White judges to find John civilly responsible (OR = 2.08, CI 1.20–3.58, P < 0.02; see Data Analysis section for an explanation of the dichotomization of the race variable). Further analysis found that this difference remained only in the negligence scenario (OR = 2.76, CI = 1.29–5.91, P = 0.02). There were no such differences among potential jurors (OR = 1.03, CI 0.82–1.30, P = 0.62). Both judges and jurors with post-graduate degree (≥MA) were less likely than jurors with <4-year college to find John civilly responsible for battery (OR = 0.59, CI 0.45–0.78, P < 0.001), but not for negligence (OR = 0.95, CI 0.99, 1. 13, P = 0.92). This difference was stronger among jurors with post-graduate education compared to judges (OR = 0.59, CI 0.45–0.78, P < 0.001) and borderline compared to jurors with <4-year college (OR = 0.66, CI 0.49–0.89, P = 0.06). Further analysis found that these differences remain significant only in the presence of genetic evidence. Specifically: among all participants who were presented with genetic evidence (either overall sample or juror only sample), those with post-graduate education were significantly less likely than jurors with <4-year college to find John responsible (overall sample: OR = 0.95, CI 0.57, 0.99, P = 0.046; juror sample: OR = 0.61, CI 0.39, 0.93, P = 0.022).

As shown in Table 3, participants’ gender (ie being female) in the combined cohort of judges and potential jurors impacted responses to all case-related questions. Specifically, female participants were less likely than male participants to believe that John is civilly responsible, a danger to the public, or had control over his action. Female participants were more likely to believe that John’s schizophrenia qualifies as a physical disability. Gender differences regarding the perception of schizophrenia as a physical disability and John’s control over his actions remained significant, regardless of the presence of genetic evidence (range of P = 0.01 to 0.03); differences in views about civil responsibility remained only among those who did not receive genetic evidence (OR = 0.72, CI 0.58–0.90, P = 0.0005). However, analysis by stakeholder group revealed that the impact of participants’ gender occurred principally in potential jurors; none of these differences remained among judges.

Similar but limited gender differences were found in the capability scale. Female participants were more likely than male participants to find John capable of providing testimony (OR = 1.26, CI 1.02–1.57, P = 0.04) and deciding whether to receive psychiatric treatment (OR = 1.27, CI 1.01–1.60, P = 0.04). However, the latter item lost its significance in analysis by stakeholder group. The aforementioned gender difference regarding views on John’s capability to provide testimony remained significant only among female potential jurors (OR = 1.27, CI 1.01–1.59, P = 0.04), not female judges. These gender differences were not impacted by the presence of genetic evidence.

IV. DISCUSSION

The increase in genomic knowledge has raised concerns—or, for some, hopes—about how psychiatric genetic evidence may alter the landscape of litigation, particularly in areas where intent and a defendant’s mindset are of paramount consideration. Psychiatric genetic evidence has been viewed as a development that may illuminate previously undeterminable factors, both in the present (eg whether a litigant has a specific medical condition) and in the future (whether a litigant may develop that condition).⁶⁸ With the increased sophistication of psychiatric genetic data, one of the major reasons cited in torts for the doctrinal separation of physical and psychiatric conditions—that the latter are difficult to verify and identify, compared to the former—is becoming increasingly outdated and inaccurate. While this should not translate into an automative decision to absolve tortfeasors with psychiatric conditions from responsibility, it highlights the need to reconsider the rationales behind existing doctrines and how justice in liability decisions may be viewed from the perspective of the general public. A discussion on the various possible rationales for tort responsibility in cases of tortfeasors with various disabilities is beyond the scope of this article, but such a discussion should include understanding both the growing knowledge in psychiatric genetics and judicial perception thereof. Focusing on the latter, this study built off both ongoing developments in the scientific landscape and the public’s growing awareness of psychiatric genetics to explore how psychiatric genetic information may impact the views of state trial court judges and potential jurors on liability in tort litigation.

Overall, a similar (small) majority of potential jurors found John civilly responsible in either tort scenario, while two-thirds of judges found him responsible for negligence and only half found him responsible for battery. Among participants who received a negligence scenario, most potential jurors, and a small majority of judges, thought that schizophrenia qualifies as a physical disability. In addition, most participants (both stakeholder groups) agreed that John was dangerous to the public—especially those who read the battery case—and only about a third of participants believed he had control over his action, regardless of the tort type. Evidence about a prior experience of schizophrenia and treatment had no impact on these responses, except that judges who received such information and read the negligence scenario were more likely to find John civilly responsible.

Contrary to our expectation, the introduction of psychiatric genetic evidence had no impact on whether participants found John liable, whether they viewed schizophrenia as a physical disability (relevant in negligence scenarios), or perceptions of dangerousness or control over one’s actions. The one exception is civil responsibility for battery, where, in the presence of genetic evidence, participants with post-graduate education (both judges and jurors) were significantly less likely than jurors with <4-year college to find John responsible for his behavior. Why this is the case is unclear and merits further research about the role of educational attainment in evaluating responsibility for behavior. Overall, however, our findings—applicable to both judges and potential jurors—may ease fears that the introduction of genetic psychiatric evidence in tort proceedings will lead to genetic reductionism or essentialism or absolve tortfeasors from responsibility for their behavior. ⁶⁹

Nonetheless, our findings about perceptions of capability indicate that psychiatric genetic evidence may have unintentional personal and legal consequences for people with psychiatric conditions. Although potential jurors in our study held significantly lower perceptions of capability compared to those of judges in both the overall score and across most capability items, the introduction of genetic evidence was negatively associated with both judges’ and potential jurors’ perceptions of capability to make marriage-related decisions, sign a financial contract, and provide testimony in court. While seemingly reflecting different daily contexts, these three activities could be thought of as contractual capabilities, ie activities that involve entering a relationship with another person, as opposed to more individualistic activities like owning land or opting into or out of personal medical treatment. There is therefore a risk that the introduction of psychiatric genetic evidence may increase perceptions that people with psychiatric conditions cannot pursue activities that involve other people. This finding may also have implications for other types of law, eg guardianship law or property law, where judges must make decisions regarding an individual’s capability to be self-sufficient. For example, parties seeking guardianship over an individual’s medical or financial decisions may introduce that individual’s genetic profile as evidence to claim that the individual is incapable of making such decisions.

Participants’ views of schizophrenia as a physical disability touch on a particularly significant question in tort law, because physical disability must be taken into consideration when determining tort liability, while mental health conditions are not.⁷⁰ Judges were significantly less likely than potential jurors to qualify schizophrenia as a physical disability. This difference may reflect judges’ legal training and adherence to already-existing legal doctrine; potential jurors’ lay understanding of psychiatric conditions may more readily accommodate a liability framework that considers not only physical disability but also psychiatric conditions.

However, it is worth noting that among judges who read the negligence case, almost a third found John not liable in the negligence framework and that a small majority of judges thought that schizophrenia qualifies as a physical disability, even with their legal training. These findings are surprising given the plain language of the Restatement, which explicitly distinguishes between physical and mental health disabilities for the purpose of determining liability, and in principle, should have resulted in a clear decision of liability. One possible reason for this is that, regardless of psychiatric genetics, mental health conditions are now seen as equally real as physical disabilities. Importantly, these findings suggest a tension between legal doctrine/education and public sentiment towards responsibility for behavior—and highlight the need to reconsider whether this legal principle should stand. This area should be further studied, including if and how decisions of liability in the presence of psychiatric genetic evidence may affect the views of judges and potential jurors about the monetary amount (ie damages) that a liable defendant with psychiatric conditions should pay. Such a study could also explore whether views on the damages differ compared to those awarded to plaintiffs with mental health conditions in contributory negligence.

Although not a focus of our study at the outset, participants’ gender was a consistently significant factor across all case-related questions: responsibility for behavior (liability), view of schizophrenia as a physical disability, and perceptions of dangerousness and control over actions; and to some extent, regarding capabilities to perform daily life activities. This finding is in line with previous research indicating that men and women view personal control differently, and specifically that men tend to perceive themselves as having more personal control to affect external outcomes than women.⁷¹ Significantly, however, our findings of gender differences occurred primarily within potential jurors and not judges, suggesting that perhaps legal training and education reduce such gender-based perspectives. Previous research has found certain differences between male and female jurors primarily in criminal cases, although these differences do not appear consistently and may depend on the type of crime being considered.⁷² Future research can explore the reasons for these gender-based perspectives and implications thereof for the judicial process.

IV.A. Limitations

Vignettes may be unable to capture the full context or nuances of a described situation, and it is unknown whether participants would react to a situation in real life in the same way they respond to a written scenario.⁷³ Relatedly, under current tort doctrine, the amount of evidence a jury would receive regarding a defendant’s mental health diagnosis would likely be limited, whereas our respondents received multiple pieces of evidence regarding John’s psychiatric condition, including psychiatric genetic data. However, such information was necessary because our study sought to understand how legal decisionmakers would react to such evidence, and specifically whether psychiatric genetic evidence impacted how respondents viewed his psychiatric condition. Whether any of this evidence would have ultimately made it to the jury would be a question for the judge; however, keeping the scenario simple allowed for the examination of specific variables of interest, and the vignette was drawn from existing court cases to keep it as close as possible to a real case scenario. The number of non-White judge participants was too small to assess differences by specific racial and ethnic identification. However, since the judiciary in the US is predominantly White,⁷⁴ our sample reflects existing limitations. To our knowledge, our study is the first to separately and comparatively explore the impact of psychiatric genetic evidence on judges’ and potential jurors’ views about civil responsibility for behavior; learning their views is informative for the ongoing discussion on the use of psychiatric genetic evidence in court settings.

V. CONCLUSION

Psychiatric genomic data is not currently developed enough to be used as evidence for proving the presence of common psychiatric conditions or creating an association between a genetic predisposition and a certain behavior. However, as the field of psychiatric genetics continues to develop, a genetically-based diagnosis of mental health conditions may be possible. If this happens, courts will likely see a corresponding increase in how often parties attempt to enter such information as evidence. Like other studies on the use of psychiatric genetic data in courts, this study suggests that psychiatric genetic evidence is likely to have a limited effect on the ultimate verdicts. At the same time, the results of this study further reinforce findings from a previous study⁷⁵ that psychiatric genetic evidence may affect perceptions of an individual’s capability to perform daily life tasks, a finding that may have implications for other areas of law and the lives of people with psychiatric conditions. In addition, responses of judges and potential jurors differed on a few questions—such as varying perceptions of schizophrenia as a physical disability and the impact of gender of judicial decision-making. These areas are worth further investigation given both the practicalities of how parties want their cases to be viewed and considered, and more general concerns about the fairness of the judicial system.⁷⁶

Although psychiatric genetic evidence did not have a significant impact on the case-related questions in this study, this might change as knowledge about how genetics interact with other factors and impacts the development of psychiatric conditions continues to grow and become more mainstream.⁷⁷ Once validated, psychiatric genetic information may further exacerbate the already high degree of social isolation that people with psychiatric condition experience and the presumption that they are incapable of carrying out personal and legal activities—in contrast to the hope that such data will reduce stigma and facilitate the view of psychiatric conditions as any other medical conditions.⁷⁸ As the introduction of such data may become increasingly conventional in judicial civil proceedings, it will be critical to develop measures and professional guidelines to ensure that such use does not infringe on the rights and preferences of the individual.

ACKNOWLEDGEMENTS

This work was supported by NHGRI grant K01HG008653 and in-kind support from the Center for Precision Medicine and Genomics at Columbia University.

DECLARATION OF INTEREST STATEMENT

Maya Sabatello is a member of the NIH All of Us Research Program’s Institutional Review Board. The authors declare no other conflict of interest.

Footnotes