Keywords

A growing number of students enter college each year with mental health issues (Gallagher, 2012). A recent survey of campus counseling centers revealed that 73 % of psychological crises among students required immediate responses; of the 39 % presenting with severe psychological problems, 6 % were so severe that the students could not remain in school without extensive psychiatric help (Gallagher, 2012). Developmentally, the transition from adolescence to early adulthood is a time when many mental health problems emerge (English & Park, 2012). The expansion of legal rights for persons with mental illness and the development of better treatments options enable more of these college-age individuals to enroll than in years past (Mowbray et al., 2006). Consequently, college and university campuses sometimes seem like “ground zero” in the debate over how to balance individual liberties of persons with mental illness with public safety. That this debate often occurs under the gaze of media scrutiny only heightens the tension, as such scrutiny may drive—and not simply reflect—the community’s perception of risk, dangerousness, and “safety-enhancing” responses. Thusly, incidents of campus-connected violence may be informed less by careful research or individualized attention, but rather may be manipulated to serve expedient, politicized ends.

This chapter addresses the interrelated dynamics among mental health, public safety, media attention, and community sentiment, and specifically, the law’s response (and at times effect on) this interplay of issues. For purposes of this discussion, law itself is seen as an intervention that has effects on behaviors, attitudes, perceptions, and outcomes—positive or less so, intended or not (Campbell, 2010). To ground discussion of the law’s role within this context, case examples drawing on recent episodes of campus-based or campus-connected “mass killings” are featured, and reference is made to related legal developments. Current legal mechanisms for addressing (often) community-fueled requests for action are compared with a potential alternative framing mechanism—therapeutic jurisprudence (“TJ”; see also chapters 12 and 13 for more on therapeutic jurisprudence and sentiment). TJ “seeks to sensitize legal policy makers to a frequently ignored aspect of … policy analysis—the therapeutic impact of legal rules and procedures” (Wexler & Winick, 1991a, p. 981). Proponents of TJ argue that “[l]egal decisionmaking should consider not only the economic factors, public safety, and the protection of patients’ rights;… [but also] the therapeutic implications of a rule and its alternatives” (Wexler & Winick, 1991a, p. 982). This chapter contends that use of a therapeutic frame to tease out the therapeutic and anti-therapeutic drivers and consequences of our legal mechanisms offers a potentially more effective response to campus safety concerns and community emotions, while keeping with an evidence-based approach.

Campus-Connected Violence: The Media and the Numbers

The past 10 years have seen campuses become ground zero in the debate over mental illness and risk of violence and if/how policy—whether driven by concerns for public safety or mental health—can mitigate future risk. Four of the most noteworthy cases are discussed below: two cases are campus-based and two have campus connections (see also chapters 1 and 2, this volume, for more on the link between sentiment and media).

Virginia Polytechnic Institute and State University (Virginia Tech)

On April 16, 2007, Seung-Hui Cho, a senior at Virginia Tech, killed 32 individuals and injured 17 before killing himself. Reportedly, there were no “outward signs of his deteriorating mental state” (Virginia Tech Review Panel, 2007, p. 52), but during his time at Virginia Tech, Cho was ordered into temporary detention at a psychiatric facility. However, follow-up outpatient appointments were not required, and Cho “disappeared” again. Much of this history came out when news outlets reported on government investigations and did their own investigative reporting. In fact, Cho himself mailed a package to NBC News the day of the shooting that included images of himself armed—images NBC News decided to air and which arguably impacted public perceptions of the event and Cho (Friedman, 2009; Kluger, 2007; NBC News, 2007; Vargas & Gardner, 2008).

Northern Illinois University (NIU)

With Virginia Tech still fresh in many minds, less than a year later Steven Kazmierczak opened fire in an NIU classroom, killing 5 and wounding 21 before killing himself. Immediate reports indicated that the former NIU undergraduate and graduate student showed no warning signs of a “path [that] diverged into madness” (Heinzman, Smith, & Zorn, 2008; Northern Illinois University, 2010, p. xvi). However, it eventually was reported that he had a history of suicide attempts, multiple psychiatric hospitalizations, and a discharge from the army due to his mental health history (Boudreau & Zamost, 2009). Prior to the shooting, he was on antidepressant, antianxiety, and sleeping medications, but had stopped taking the antidepressant 3 weeks prior to the shooting (Boudreau & Zamost, 2008).

Tucson, Arizona

On January 8, 2011, Jared Lee Loughner, 22, a former Pima Community College student, killed 6 and injured 13 at a Tucson, Arizona, shopping plaza. His erratic behavior while at Pima resulted in him ultimately being suspended 3 months prior to the shooting and being told not to return without mental health certification that he was no longer dangerous (Anglen, 2011; Billeaud, 2011). Said a campus spokesman, “[W]e dealt with it [Loughner’s behavior] in a way that protected our students and our employees” (Sulzberger & Gabriel, 2011). Post-tragedy reports featured the shooter’s mental health history: depression since 2006 with signs of schizophrenia since 2008 (Anglen, 2011).

Aurora, CO

In another mass shooting, on July 20, 2012, James Holmes, 24, killed 12 individuals and injured 58 at a movie theater. Holmes had been enrolled in a graduate program at a nearby medical campus until the month prior to the shooting (K-ABC TV, 2012). Media accounts now unearth experience at the University of Colorado Denver-Anschutz Medical Campus (“Medical Campus”), from which he was barred after threatening the psychiatrist he had seen on campus (Fantz, 2012). Most notable were the images of Holmes that hit the airwaves and papers once he showed up in court: a flame-haired young man who seemed visibly “out of it” (Pearson, 2012).

These cases have some commonalities. In all instances, media-driven messages affected by community sentiment, fear, and anger seemingly paint the picture that untreated mental health issues, especially among loners (often pictured looking “crazed”), on college campuses can lead to mass violence (Billeaud, 2011; Boudreau & Zamost, 2009; Dewan & Santora, 2007).

These high-profile incidents have clear impact on campuses. Over half of campus counseling center directors reported that campus tragedies related to students with mental troubles put them “under increasing pressure to share concerns about troubled students who might pose a risk to others even though the threat was not to a specific person” (Gallagher, 2012, p. 7).

State of Community Sentiment

Colleges may respond to these increased numbers via public health approaches to address population-level needs. Unfortunately, these efforts come up against stretched mental health services on campuses and the stigma that continues to pervade community and media depictions of mental illness. The latter is influenced by and influences community sentiment, which together impact language used to describe the numbers, and may frame the policy response. Polling of community beliefs, the overarching “community sentiment” for purposes herein, shows that the public “believes that those experiencing mental health problems pose a threat of violence towards others” (Pescosolido et al., 2000, p. 16 and Figure 4). These beliefs are more pronounced against men, and those with schizophrenia-type diagnoses (Pescosolido et al., 2000, Figure 4), and have increased since the 1950s (Martin, Pescosolido, & Tuch, 2000, p. 219). Given dangerousness data, it is unsurprising that a significant number of this same public prefers maintaining social distance (e.g., not working or living with) from persons with mental disorders (Pescosolido et al., 2000, p. 30 and Table 11; Martin et al., 2000). Social distancing is the act of separating “us” from “them” (i.e., those with mental illness). Distancing is driven in part by the negative labels (e.g., “schizophrenic”) and belief of “dangerousness” (Martin et al., 2000, p. 219–220). In sum, there is “little evidence to suggest that the stigma of mental illness has been reduced in contemporary American society” (Pescosolido et al., 2000, p. 31), and in fact, as related to beliefs of dangerousness, seems to have increased (Martin et al., 2000).

It is against this backdrop of lingering suspicion of dangerousness and disinclination for close interaction that tragic cases arise. And so today, rare but heavily covered mass killings seem to fuel public sentiment vis-à-vis violence and the “mentally ill,” namely, that there are dangerous (and deadly) individuals lurking among us (on campus or near campus). This may be described in terms of an availability heuristic whereby members of the public estimate the likelihood of events and their consequences by drawing on examples they can recall from the past (Tversky & Kahneman, 1973). Such events, in fact, may be relatively rare, but the preponderance of media attention may influence what is recalled, and thus influence a belief in the likelihood of the event (e.g., risks of violence on campus committed by students with mental health issues) or cause (e.g., “dangerous” persons on campus, i.e., those with untreated mental illness). In this way, the media can affect community sentiment. From this comes the sense that a stronger intervention/detention approach by campuses is needed to keep “us” safe.

Violence and Persons with Mental Illness

Publicity of college campus shootings has led to increased fears among college students—and their families—that they will be victims of violent crimes on campus (Kaminski, Koons-Witt, Thompson, & Weiss, 2010). Yet, the research evidence does not necessarily support this assumed link between mental illness and violence.

Violence risk factors. A new generation of research emerged in the 1990s, pointing to heightened risk of violence by those with mental illness. However, individuals evidencing the greatest “threat” represent a complex spectrum of dynamic factors—demographic, historical/dispositional, clinical, and environmental/contextual (Otto, 2000; Swanson, Borum, Swartz, & Monahan, 1996). Generally, risk factors assess: what a person “is” (e.g., age, gender, personality), what a person “has” (e.g., major mental disorder, personality disorder, substance abuse disorder), what a person “has done” (e.g., prior crime, prior violence), and what a person has experienced (e.g., pathological family environment, prior exposure to violence) (Monahan, 2006, p. 414–427). Additional studies isolate specific factors of concern among those who feel threat/control-override (i.e., inability to control violent responses to threat delusions) (Swanson et al., 1996), especially when joined by substance abuse problems (Otto, 2000; Swanson et al., 1996) and poor treatment adherence (Swartz et al., 1998). The interaction of stressful environments (including relationships), stressful events, and lack of social support can enhance and compound the risk of violence (Markowitz, 2011).

What the evidence does not say is that diagnosis (e.g., schizophrenia) equals danger or that more treatment ensures safety; nor has evidence shown a single pathway to violence or a singular type of violence risk. Rather, the presence of multiple factors implicated in and pathways to violence suggest a need for a range of analyses and targets for prevention/intervention. Further, the dynamic and multifaceted nature of risk factors suggests they represent probabilities for violence (not certainty), relative risks (not absolutes), and situational influences (not simply dispositional ones; Douglas & Skeem, 2005; Heilbrun, Dvoskin, & Heilbrun, 2009; Otto, 2000). And thus, risk assessment should be seen as a process, not an event, and as targeted prevention (aimed at reduction), not as a prediction or a silver bullet treatment (Otto, 2000; Swanson, 2008; Swartz et al., 1998).

Risk assessment approaches. If humans were simple beings, it would be possible to identify a set of characteristics based on past experience that can be used to segregate those in the population who are considered presenting the most “risk” for some given incident (i.e., violence). This certainly has appeal, as does belief in ability of clinical violence assessments to produce “binary, will-or-will not judgments” (Mossman, 2009, p. 121). Yet, there is a lack of evidence to tie a single profile to the “violent” individual, and attempts to profile can increase harm via bias, stigma, and unfair restrictions on civil liberties (Borum, Cornell, Modzeleski, & Jimerson, 2010; Mossman, 2009; Reddy et al., 2001). Relying on clinical judgment is not without some merit in violence risk assessment, but it is of questionable value in making (vs. informing) decisions (Lidz, Mulvey, & Gardner, 1993). Guides and structured clinical assessment approaches have been developed, but the use of checklists or warning signs have been questioned for their use in prevention of targeted violence (Reddy et al., 2001). Actuarial approaches (i.e., based on statistics) have similarly been questioned for limiting the value of clinical intuition, especially given the lack of consensus around targeted violence markers to plug into equations (Reddy et al., 2001).

In sum, a combination of approaches (e.g., clinical or actuarial) likely holds the most promise for prevention and early identification of those at “risk” (McNiel et al., 2004; Otto, 2000), and a deductive fact-based model focused on those who pose threats (vs. more generalized “profile” model) is likely the most appropriate for the sorts of violence contemplated here (Reddy et al., 2001). Yet, reliance on a fact-based model renders it more difficult to create a global law or policy to assess and intervene with individuals posing potential risk. Such globalization is likely over-inclusive, with the attendant risk of impeding on justice claims of incorrectly targeted individuals (explained more below); further, global policy may offer false assurances that it can help campus personnel predict who will be violent.

Applying Violence and Risk Assessment Research to the Campus

The above literature suggests that laws and policies should reflect the complexity of interactions among violence risk, mental status, and campus environments. Law and policy should also avoid oversimplified “profile” approaches or misapplication of automatic checklists. Unfortunately, much of the policy response has been driven by rare, high-profile cases, leading to poorly constructed, ineffective laws with unintended consequences (Reddy et al., 2001). (See also, chapters 15–18 in this volume for further discussion of unintended consequences of laws that often result from high-publicity cases). Certainly, campuses are affected by public perception and community sentiment that demand protective action (Heilbrun et al., 2009). But, ironically, while the public has become a bit more informed as to causes of mental illness, there has been an increase in fear of persons with mental illness—often believed to be dangerous—and increased support for social distancing (Markowitz, 2011, p. 39; Pescosolido et al., 2000, p. 30–31). Media has played a special role in shaping this public response by overemphasizing mental diagnoses, blaming mental health system gaps for violence, and over-relying on images of “crazed” shooters in pictorial accounts. These portrayals create a sense of “moral panic” within the public (Billeaud, 2011; Borum et al., 2010; Ferguson, 2008).

Incomplete or inaccurate depictions built on a limited, misinterpreted, or misapplied research base construct a metaphor of the “mentally ill” as mass killers (Borum et al., 2010). Politicized use of research can, in turn, support claims of an “epidemic” of violence and need to “quarantine” persons with mental illness (Dodge, 2008). This confluence of political, media, and research factors can also create self-fulfilling prophecies of “crazed” killers running amok, biasing polls to suggest yet more public support of profile-type, and liberty-restricting responses.

Ultimately, campuses are in a bind: They are expected to step into a parental role to take care of students entrusted to them, with liability fears further driving a “protection” focus (Bertram, 2010; Stone, 2008; Stuart, 2012). Campuses are to use evidence-based best practices in outreach to individuals on campuses with mental health issues. However, these practices might not support what the community wants or may be twisted to support political ends. At the same time, they are to foster “open environments” as they provide education. The crux of this bind is thus: How to strike the right balance among this mix of obligations—without being overly reactive or unduly privileging one set of priorities out of unfounded fear?

The Campus Response

So how have campuses responded to violence within their environs? Even before the cases described herein, there emerged an obligation for campus counselors to warn or protect identifiable third parties from becoming victims of violence perpetrated by their patients ( Tarasoff v. Regents of the University of California, 1976). This duty of protection can also extend beyond a single feared victim to an identifiable “class of persons” (Lipari v. Sears, Roebuck & Co, 1980; VandeCreek & Knapp, 2000). Also, during the 1990s, policies relaxed criteria for commitment (e.g., lesser threshold of imminence of risk) and created an outpatient commitment option (Monahan, 2006). Against this backdrop of greater tolerance for liberty restriction for individuals presenting potential risks of violence, the “lessons learned” from Virginia Tech and Northern Illinois (and as reflected in Tucson and Aurora) seemingly suggest that it’s best for “our” students if we require “them” to leave campus (and only come back with “certification” that they are no longer dangerous). Ironically, this sort of response may make the target of such response (i.e., the individual perceived to present a threat of violence) feel more isolated and aggrieved. These responses also limit colleges’ ability to keep a watchful eye on the target’s behaviors (Heilbrun et al., 2009).

Threat Assessment Approach

Perhaps one of the most defining features of campus responses post-Virginia Tech has been the development of a formal threat assessment (TA) approach to risk management. TA is a “strategy for preventing violence through identification … of individuals or groups that pose a threat to harm someone, followed by intervention designed to reduce the risk of violence” (Cornell, 2009, p. 4). This approach involves four central areas: personality traits and behaviors, family dynamics, school dynamics, and social dynamics (Borum et al., 2010; Fox & Savage, 2009). Critically, the TA approach builds on a strong research base that recognizes weaknesses of an individual profile or generalized assessment; rather, it requires particularized assessment with multiple informants covering different contexts (Fox & Savage, 2009; Heilbrun et al., 2009). The use of TA-like approaches has proliferated across campuses (Muskal, 2012). With broadened application, however, the “threat” at issue has often shifted to concern over harm to self (Wolnick, 2007) or nonlethal violence to others (Dunkle, Silverstein, & Warner, 2008). Behavioral contracts or medical withdrawals are often used to address concerns (Delworth, 1989; Dunkle et al., 2008; Eells & Rockland-Miller, 2010).

How Did We Get Here (Today) from There (Post-Virginia Tech)?

In the wake of the Virginia Tech tragedy, then-Governor Tim Kaine appointed a review panel that documented what policymakers saw as “lost opportunities” to intervene with Cho, especially perceived barriers to longer-term mandatory commitments and campus-community communication (Virginia Tech Review Panel, 2007). Recommendations from this report led to legal change, including a reformed civil commitment process that broadened standards for civil commitment, extended emergency custody and temporary detention order periods (Va. Code §37.2-808, 2010; Va. Code § 37.2-809, 2011; Va. Code § 37.2-817.1, 2010), and enhanced campus security including requiring that public colleges and universities create and use TA teams (Va. Code § 23-9.2:10, 2010). Following this, most public and private colleges within Virginia adopted new policies to encourage students to adhere to mental health treatment via voluntary medical withdrawals, mandated outpatient treatment, mandatory engagement in mental health treatment to avoid suspension/expulsion, and TA team monitoring (Monahan, Bonnie, Davis, & Flynn, 2011). Colleges, especially private ones, also adopted involuntary medical leave policies, requiring clinical verification of student treatment adherence for readmission (Monahan, Bonnie, Davis, & Flynn, 2011).

Advances in other states. This development of a TA approach informed changes in other states, e.g., Illinois, which amended existing law to require its campuses to partner with local agencies to plan and practice emergency response (Illinois Campus Security Enhancement Act, 2010). However, law did not guarantee action: Three years post enactment, there was widespread noncompliance in Illinois, in part due to lack of an enforcement mechanism and no clear line of authority for ensuring compliance (Pawlowski & Manetti, 2011). Pima and Aurora utilized TA-like teams or processes to remove Loughner and Holmes, respectively, from campus—with on-campus violence averted (although not necessarily causally linked)—yet, violence itself was not averted. Thus, while TA approaches may hold promise, they are not a magic bullet against violence.

Irrespective of a potentially more evidence-informed and less stigmatizing approach to campus-based violence risk, it proves difficult to counteract media accounts, public sentiment, and politicization of events. Campus policies have taken on a safety frame (i.e., view policy formation and implementation through the perspective of safety when facing (or frightened by the potential of) media attention, and as driven by an often-understandable community sentiment post-violence). Specifically, such campus polices may be informed by a TA team’s arsenal of recommendations. And these policies exist within a risk avoidance culture that prioritizes a “better safe than sorry” response that may more quickly lead to suspension or expulsion decisions, even if more effective violence prevention necessitates an ability to monitor at-risk students who are identifiable and remain at least somewhat “connected.”

A Therapeutic Jurisprudence Frame for Campus Response to Violence Risk

The question remains as to how the goals of an evidence-informed, public health-oriented approach to risk assessment can influence policy in a way that achieves meaningful, therapeutic, safety-enhancing, fair, and ethical results. Specifically, lawmakers must determine how to be responsive to community sentiment and its symbolic value while also cognizant of policy’s as-implemented reality and potential for harm, including less visible harms of fostering perceived hostile campus environments for those with mental health disorders. Perhaps a different frame for policymaking might help.

Defining Therapeutic Jurisprudence (“TJ”)

As a prominent TJ scholar has explained:

TJ recognizes that the law is a social force with negative and positive emotional consequences for all the people involved. …It seeks to identify those emotional consequences; assess whether they are therapeutic or counter therapeutic; and then ask whether the law can be changed, applied, interpreted, or enforced in ways that can maximize its therapeutic effects. (Daicoff, 1999, p. 813)

TJ’s early development relates to themes of this chapter: the effects of deinstitutionalization and public perceptions of dangerousness of persons with mental illness. As more and more mentally ill patients ended up in courtrooms, certain mental health lawyers developed the concept of TJ to respond to the “anti-therapeutic” effects of the legal process on these individuals (Wexler & Winick, 1991a, 1991b). TJ does not imply that therapeutic outcomes are the only—or even predominant—goals or that legal decision makers should act in deference to clinical goals (Wexler & Winick, 1991a). Critically, though, TJ urges that legal actors recognize that there may be the so-called facts upon which they act (e.g., that a person with untreated schizophrenia will likely be violent against others) that lack, and could thus benefit from, empirical support. Moreover, legal actors should also empirically gauge consequences of legal decision making, including therapeutic effects (Wexler & Winick, 1991a, p. 983). Since its formulation, TJ’s application has broadened beyond mental health law, to now include a role as frame for evidence-informed policymaking concerning a wide range of legal issues (Campbell, 2010).

Applying TJ

TJ holds promise for revising campus policy development and related state and federal legal action vis-à-vis concerns of safety on college campuses. The first necessary change is that policy itself should be viewed as an intervention (Campbell, 2010). Second, many policies, even those not directly related to health as narrowly conceived, influence individual and community well-being, physically and emotionally. When so viewed, it becomes more apparent how a study of the consequences of policy development and implementation would also include a view of its therapeutic (or not) impacts, with a natural response to enhance well-being through policy, or at the very least, in ethical terms, to “do no harm” (Brookbanks, 2001; Sharpe, 1997).

TJ can be applied as frame for policy in a variety of ways. It could help highlight therapeutic consequences and also help channel the quite natural emotions that drive and/or are driven by certain policy developments (Campbell, 2012; see also chapters 1 and 17 for more on emotions and sentiment). Consider its application to Virginia Tech:

In this environment [fear and anger], is it any wonder that policies often slant towards the coercive, punitive, or public safety expanding rather than slanting towards promotion of individual liberty or mental health? Less considered are the negative consequences of the resulting policies, and whether they best address the emotional needs of the targeted group and the public at large in an evidence-based way. (Campbell, 2012, p. 694)

Using TJ as a frame to build the evidence base. Importantly, applying TJ as frame for campus policy and broader policy development is not purely a normative exercise but an empirical one. That is, it requires pre- and post-review of agreed-to measures or tests of “therapeutic” effects to fully evaluate “success” of laws and policies. These measures might include the sort of campus environments fostered by communities, the willingness of individuals facing challenges to open up (or for their peers to come forward and report when they are worried about their friends), the feelings of respect (or lack thereof) such individuals in crisis perceive in various campus responses, etc. These sorts of evaluative questions and environmental scans pre- and post-policy intervention could be coupled with other concerns that the policies seek to address, such as fostering a sense of safety or enhancing perceived fairness in policy application.

An example to help guide such sort of empirical investigation involves the post-Virginia Tech experience. In addition to passing a series of bills to enhance mental health and campus security systems, Virginia’s legislature also commissioned a mental health study with two prongs: legal issues related to campus mental health and clinical access issues facing campus students with mental health issues. The goals for each task force were to “make recommendations for training, institutional policies and practices, and any legislative action that may be needed.” (Bonnie et al., 2011, p. 3). Importantly, their approach utilized empirical study and robust multi-stakeholder engagement to evaluate effects of its post-Virginia Tech responses across the state (Bonnie et al., 2011).

TJ and ethical concerns. Moreover, a TJ frame could also be studied for its ethical effects, including effects on confidentiality concerns among those with mental health or substance abuse troubles. Current approaches could be faulted for employing a utilitarian calculus in which public safety trumps confidentiality, with the assumption made that breaches make campuses safer (Mossman, 2009). Indeed, the current culture places a great deal of pressure on campus counseling centers, leading center directors to be more likely to break confidentiality (Gallagher, 2012). This pressure is largely a result of community sentiment.

Whether policy can effectively achieve safety without unnecessarily, unfairly, or harmfully impacting confidentiality can be empirically studied, with adolescent confidentiality studies serving as potential models (Ford, Millstein, Halpern-Felsher, & Irwin, 1997). Potential policies that are in need of evaluation include proposals to require mental health privacy waivers of incoming college students (Fox & Savage, 2009).

Further, there is ethical concern related to the use of less specific or sensitive tools with a specific population, i.e., those with mental illness. Concerns include if thus use results in high false-positive rates (i.e., detain an individual who is in fact not dangerous) or conversely, with high false-negative rates (i.e., not detain an individual who is in fact dangerous and who might benefit from treatment or supports) (Munro & Rumgay, 2000). The trick lies in identifying a threshold, above which risk level designated campus officers may seek to detain. Admittedly, this is made all the more complicated by public pressure and media scrutiny not to let another “dangerous person” slip by (Munro & Rumgay, 2000).

TJ as frame offers some assistance in addressing ethical issues by focusing attention on the psychological effects of policy—that is, the human consequences of policy as experienced in therapeutic terms vs. a focus simply on safety driven by community sentiment and/or media. Specifically, when evaluating policy effectiveness, TJ as frame necessitates consideration of factors beyond violence incidence reduction to include inclusiveness of campus environments, say, or fairness of outcome—in real and perceived terms—in application of threat assessment policies.

TJ and justice concerns. Fairness considerations point to a final, critical, area of policy impact assessment: justice implications of threat assessment policies and their kin. Individual liberty concerns arise when policies target certain behaviors—tied to certain mental health diagnoses—for punitive response or when such individuals experience disparate treatment by more “global” policies (e.g., suspensions). Applying a TJ frame can help illuminate psychological impacts on persons with mental health disorders and their families—as well as on those who may have yet to seek help. Beyond traditional liberty-based claims, there also exist other justice-related concerns. Specifically, questions can be raised as to the effects of devoting limited mental health resources to measures to avert “dangerousness” rather than to measures to enhance mental health access for all of the campus (or at least those in need—but not (yet) at the level of dangerousness) (Munro & Rumgay, 2000). Again, TJ may help in policy formation and evaluation by adding to the list of effects of resource allocation polices such policies’ effects on psychological well-being, help-seeking behaviors, and perceptions of inclusiveness.

These sorts of justice issues and demands for empirical investigation suggest that TJ, while a question-generating frame, is not divorced from the need for evidence or blind to other considerations of cost trade-offs or values beyond therapeutics, e.g., justice (Campbell, 2010). Rather, it is highly contextual and sensitive to consequences—therapeutic, emotional, or ethical. And while not the answer for policy development, “having therapeutic consequences in mind and reflecting on related evidence may be our best hope—where policy is possibly helpful or necessarily implemented [e.g., because politicians and community members demand campus responses]—of enhancing therapeutic outcomes” (Campbell, 2010, p. 291).

Future Steps

From this analysis, several steps emerge as needed. First, as just explained, building an evidence base is critical in the endeavor to revisit campus responses to violence by and upon their students. And the issue is not simply “which laws work, but which laws work best and why” (McNiel et al., 2004, p. 159). Researchers and policymakers should place more emphasis on the study of therapeutic consequences of policy responses—be they institutional, legislative, or administrative—on student behaviors (on campuses or off), as well as their justice impacts (e.g., disparate racial/ethnic effects). Safety enhancement becomes a necessary-but-not-sufficient outcome. Here, community sentiment becomes critical, inasmuch as it involves the perception of safety. Researchers can measure community sentiment not simply pre- and post-tragedy, but more proactively to assess how different sorts of media and policy responses to tragedy (actual or averted) impact perceptions of safety. Critical in this, too, is inclusion of multiple perspectives so that “community” will not remain an amorphous, or “us,” concept, but a highly contextual one inclusive of those most intimately affected by potential policy and media responses (e.g., those with serious mental illnesses on campuses). A TJ orientation can help in this process by helping maintain a focus on psychological impacts and other indicia of well-being, beyond depersonalized target goals. Also critical are considerations of how to promote therapeutically effective policies through a media-generated “atmosphere of fear” (Fox & Savage, 2009, p. 1466). For this, it will be critical to have phased-in policies with as much transparency as possible, a greater appreciation of what evidence applies, and an understanding of the limitations for application in certain policy environments (Fox & Savage, 2009).

Second, more attention needs to be paid to potential shortcomings in a system that relies on “watchful waiting” and monitoring when many of our cases may involve the “unbefriended,” that is, those who seem to slip by without friends or family supports. Yet it is difficult to monitor such isolated individuals. Monitoring and averting violence becomes even more difficult if these individuals are removed from settings where it is likely easier to accomplish at least some degree of monitoring.

Researchers and policymakers also need to examine more closely policy action “triggers.” High-profile cases trigger legal actions, often leading to laws “named” after victims (e.g., “Kendra’s Law” (NY Mental Hyg. Law § 9.60, 1999)). The particulars of one situation may not readily translate to policy action, yet a law or policy that is adopted in response to one event is expected to protect a broader class of individuals. There is a natural tendency, and can be great policy power, in seizing the moment to enact meaningful change, but fast action based on traumatic (especially rare) events may improperly apply (or ignore altogether) evidence to support potentially quite anti-therapeutic laws. Chapters 1, 17, and 18 in this volume further explore the issues associated with memorial crime legislation that sometimes results in crime control theater (CCT). CCT-type laws address the need to “do something” to address heinous crimes and appear to solve such crimes, yet have many unintended consequences and are unlikely to be successful. Such policies also risk anti-therapeutic outcomes and violation of TJ principles.

Researchers and policymakers should also be weary of “mission creep.” This refers to how a policy’s scope may be expanded, intentionally or not (and as influenced by community sentiment of fear). Policies have expanded beyond a focus on individuals whose behaviors indicate (primarily) other-directed violence to individuals with mental health challenges that are more internally directed, e.g., those with suicide risk. An example of this would be TA teams that have morphed into behavioral risk assessment teams, which support greater use of medical withdrawals to “encourage” treatment adherence. Here, a TJ reframing would require asking if these are the most therapeutic approaches and if they enhance student help-seeking behavior. Arguably, such assessments have negative therapeutic consequences; at the very least, policies as experienced should be evaluated for these potential negative consequences.

And third, rather than focus solely on the negative or areas of concern, attention should also be devoted to positive examples. There are some states, e.g., Virginia, that are incorporating evaluation into their policy agenda as a proactive response to past and potential incidents. These are efforts deserving of more analysis. Such policy agendas will broaden the research base from which others can learn policymaking best practices, such as which approaches lessen risks of violence while also balancing rights of individuals with mental health issues to privacy and to a traditional college education. Research can also tease out how contextual factors, such as public opinion, influence different policy approaches, and with what consequences. This will help confront mistaken beliefs versus perceived actual risks (in part addressing the availability heuristic).

Conclusion

In sum, tragedy begets policy response, often in an atmosphere of heightened negative, emotion-fueled community sentiment, media scrutiny, and politicization. In such environments, it is understandably difficult to foster sensitive policy development that balances the urgency of the moment with the need for thoughtful reflection and stakeholder engagement. TJ offers a mechanism to reframe policy action in therapeutic terms, and encourages therapeutic-evidence gathering and use in post-policy implementation evaluation. Recent campus-based or campus-connected tragedies provide a laboratory for investigation of what has worked (or not), as defined by whom, and with what consequences.

Evidence to date does not support simplistic policy responses that place individuals with certain mental health diagnoses in dichotomous “dangerous” or “not dangerous” categories. Campuses should not view the counseling center as a means to avert campus tragedies or see mental health treatment as the solution to violence. They should also not allow mental health counselors to be used as disciplinarians or violence risk detectors (Stone, 2008, p. 498–499). And even with data-informed, fully functioning TA teams and great communication networks between campuses and communities, campuses should not claim they are “100 % safe.” This may create a false sense of security and relieve the community of any responsibility in enhancing “safe” communities for those with and without mental disorders (Stone, 2008, p. 498).

Further, there is evidence to support considering how dynamic risk factors interact and rise to the level of “threat,” with need for greater attention on protective factors, e.g., public health approaches wherein an enhanced mental health system has as a by-product less overall violence (Mossman, 2009; Stone, 2008). This suggests, in turn, that it may be wise to let clinicians remain squarely grounded in their therapeutic role, with a focus on prevention (vs. prediction) and therapeutic aims for their patients. In so doing they may help avoid role confusion that may deter students (and others) from seeking help if they are struggling with mental health issues for fear of some bright-line safety reporting mechanism. It would also, importantly, necessitate greater discussion of, and transparency about, times when clinicians may have to make reports. Such reports should consider the therapeutic impact on those about whom reports are being made (and not simply safety of potential victims) as guidepost for such reporting protocols.

Implications for the Media and an Emotion-Driven Community

This discussion also obligates more responsible media reporting. This includes the adoption of media infrastructures that support more sensitive and contextual reporting, and less emphasis on idiosyncratic events or hyperbolic headlines and imagery (Brooks, Schiraldi, & Ziedenberg, 2000). This may be difficult in a 24/7 news environment, with greater blurring (especially through the Internet) of who qualifies as “reporter” vs. pundit vs. agitated individual commentator. In this, the community also bears responsibility for how viewing habits influence media, and in turn, policy action. That is, community members should be more thoughtful consumers of the news that each individual, by her actions, helps shape (e.g., if ratings go up for certain inflammatory coverage, that could beget yet more “frenzied” media).

In sum, if we maintain a therapeutic response that is sensitive to the context but not driven by community emotions or politics of the moment, there is hope for creating campus environments that achieve educational goals via, in part, promoting healthy development and fostering a sense of respect and fairness among all within those campuses. There may be no easy solution to avert the next campus-based or campus-affiliated tragedy. Yet, this does not mean that there is no hope for therapeutic policy response that enhances individual and public well-being overall. A more caring policy response is the very least our student bodies can expect.