The Crisis in Crisis Response

In early September 2020, the American public was given access to shocking video footage demonstrating how Daniel Prude, a 41-year-old Black man, died of asphyxiation almost 6 months earlier. His death occurred after Rochester, New York, police officers responding to a mental health crisis call left Prude in a “spit hood,” leading to his asphyxiation. After a summer of widespread protests and demands for police reform and abolition set off by the emergence of video evidence of the police-involved killings of other unarmed Black people (including George Floyd and Rayshard Brooks), the release of the Daniel Prude video amplified public dialogue on the need to create non-police alternatives to mental health crisis response.

Discussion about the need to reform responses to mental health crises is not new. In 1988, Joseph Dewayne Robinson was shot and killed in Memphis, Tennessee, by police who had responded to a call that he was threatening suicide with a knife. The outcry that followed this incident, in fact, led to the development of the Crisis Intervention Team (CIT) model, which started in Memphis and has come to be disseminated in over 3,000 jurisdictions worldwide (including Rochester). CIT, which involves educating police officers about mental illness and training them to use “de-escalation” strategies, has been supported by a good deal of research evidence indicating that it improves officer attitudes related to mental illness and preferences for using de-escalation strategies.

Nevertheless, the evidence is much less compelling regarding whether CIT leads to reductions in fatal use-of-force outcomes in the jurisdictions in which it is implemented. Taking a national view, there is evidence that people with mental illnesses are at significantly greater risk of being killed when they interact with police, despite the proliferation of CIT. This impact is further compounded when race and mental health crisis intersect (as they did in Prude’s case), such that Black people with mental illnesses are at the greatest risk of being killed by police in the United States.

Why are Black people experiencing mental health crises so much more likely to be killed by police officers than others? It is likely that part of the explanation is that police behavior is influenced by the intersection of negative stereotypes related to both mental illness and race. Specifically, negative stereotypes of dangerousness are associated with both mental illness and being Black (in the United States), and expectations of violence may be amplified when the two statuses are combined. In his 2010 book The Protest Psychosis, Jonathan Metzl discussed how stereotypes of danger related to both mental illness and race were so intertwined that they led schizophrenia to become “racialized” in its clinical characterization within the U.S. mental health system.

Police officers are susceptible to the same biases as general community members, but there is evidence that police officers actually endorse negative stereotypes about mental illness at higher rates than the broader community. Further, there is evidence that police officers are significantly more likely to associate Black faces with criminality than White faces. Given that police officers in the United States hold firearms and experience pressure to make split-second decisions regarding whether to use force, the effects of intersecting stereotypes about mental illness and race may be too powerful for targeted training efforts (such as CIT) to override.

As a result, there is now a growing consensus that CIT, though useful, is not sufficient and that additional reforms are needed. In a recently-released report, in collaboration with the advocacy organization Disability Rights New York, I and colleagues from John Jay College recommend that communities begin the process of developing and implementing non-police responses to mental health crises.

We further argue that reform efforts must take steps to ensure that well-intentioned alternative response models do not become coercive or punitive (a position advocated in two recent panels convened by the Institute for the Development of the Human Arts), but rather facilitate access to services consistent with the preferences of the person in crisis. While how exactly these alternatives will look can be determined by local communities, an example of a successful alternative model is the Crisis Assistance Helping Out On The Streets (or CAHOOTS) model, which has been used in Eugene, Oregon, for over 31 years. This model uses a coordinated dispatch approach to send a team consisting of a mental health professional and a “medic” (nurse or paramedic) to respond to mental health crises. Although police may be called for “backup” in instances where the CAHOOTS team believes it is needed, data indicate that police backup is requested in fewer than 1 percent of calls.

It should be noted that Eugene, Oregon, has very few Black residents (only 1 percent of the population), so it remains to be seen if the CAHOOTS approach leads to similar success in eliminating calls for police involvement if implemented in more diverse jurisdictions. Indeed, mental health workers may be prone to the same racial stereotypes as police officers, which may influence their decision-making. Jurisdictions that implement alternative models need to pay extra attention to the role that racial stereotypes can play in making the decision to include police in the crisis response process.

There is now a growing consensus that reform is needed. My colleagues and I urge Americans to seize this moment to reduce the incidence of deadly consequences when crisis response, race, and mental illness intersect.