SSN Key Findings

How Journalists May Inappropriately Portray People with Mental Illness Killed by Police

Policy field

Connect with the author

University of Wisconsin–Madison

Dramatic stories about people with mental illnesses killed by police have appeared in the U.S. media in the past few years. Yet not enough is known about how journalists typically report on mental illness in the context of fatal police shootings, including how journalists handle “suicide by cop” – when an individual deliberately behaves in a threatening manner with intent to provoke a lethal response from a public safety or law enforcement officer.  Because deaths of people with mental illness at the hands of police are not easy to understand, it matters how journalists report on such tragic events. To increase understanding and prevent harmful stereotyping of people with mental illness, media coverage needs to treat these events in full context and with great care.  Otherwise, reporters can find themselves offering unfair or misleading portrayals of mental illness.

Media Portrayals

My research examined news reports about individuals with mental illness who were killed by police during 2015 and 2016. I found that the individuals were often stigmatized because of either their past or present mental illness.  My study analyzed content from 301 online news reports, in which 132 reports contributed 231 examples of stigmatizing language.

Stigmatization took a variety of forms. At times, such language was used by journalists; but in other instances stigmatizing language appeared in quotes from family members, friends, community members, or police. Ordered according to how commonly they appeared, the following kinds of stigmatizing portrayals appeared in the media reports:

  • Language in 79 news references portrayed mental illness as the defining characteristic of an individual – for example, “paranoid schizophrenic,” “alcoholic,” “drug addict."
  • In 45 news references, people with mental illness were portrayed as helpless sufferers with little chance of recovery, perpetuating myths that mental illness is not treatable and recovery is not possible.
  • Derogatory language appeared in 22 news references – portraying the victims, for example, as “crazy”, “insane”, “mental”, “crazed”, “deranged”, “nut”, or prone to “drug abuse."
  • In 16 cases, news references romanticized suicide by implying that there was a single cause, such as a having relationship problems or losing a job.
  • Violence was portrayed as the norm for people with mental illness in 15 news references – typically when a journalist reported that the person was not violent despite suffering from a mental illness. Research reveals that, on average, individuals with mental illness are not violent and are more likely to be victims than perpetrators of violence.
  • Nine news references associated mental illness with supernatural possession, using terms and phrases such as “personal demons”, “haunted”, and “possessed."
  • Nine references associated suicide with criminality by using the phrase “commit suicide."
  • Four news references portrayed suicide as having no warning signs, wrongly implying that suicide is not preventable.
  • Three news references portrayed people with mental illness as “not normal” and “not mentally there” – when in reality mental illness affects many people in all social groups.

Implications for Media Organizations and Mental Illness Advocacy Groups

Guidelines used by reporters and advocates for reporting about mental illness do not cover all of the forms of stigmatizing language uncovered in my research. For example, these guidelines rarely consider how violence may, mistakenly, be portrayed as typical behavior for people with mental illness; nor do guidelines take into account many other ways people with mental illness may be portrayed as “not normal,” even described as possessed in some sense. Advocacy groups and media organizations should consider the full range of possible kinds of stigmatization that can occur as they set appropriate guidelines to counter such derogatory descriptions.

My research underscores the importance of revealing implicit forms of stigmatization, including the two types that appeared most frequently in my media analysis – using mental illness as the overriding defining characteristic of an individual; and portraying individuals with mental illness as helpless sufferers. The amount of such implicit forms of stigmatization may suggest that media and advocacy commentators are avoiding explicit stigmatizing language, such as calling someone “crazy.” Perhaps they now use, instead, subtly stigmatizing portrayals that, at first glance, do not seem offensive. Nevertheless, hints of stigma may remain and those who prepare guidelines should work to eliminate even implicit negative portrayals.

Although guides meant to encourage appropriate reporting on mental illness are widely available to journalists, a significant proportion of the news reports I analyzed still contained stigmatizing language of the sort that the guides are meant to eliminate. This suggests that consensus among advocacy organizations about appropriate language may not immediately translate into improved media reporting. Perhaps journalists do not know about available guidelines, and/or perhaps they fail to understand and apply them appropriately.  

Overall, my research underlines that media organizations and advocacy groups must take further steps to make sure journalists fairly and accurately portray the mental illnesses many Americans experience. Rare but disturbing events such as police killings of persons with mental illness must be portrayed in context and without stigma. Only then will reporting on mental illness improve and misleading stigmatization in the news decline.

Read more in Emma Frankham (2017) “Stigmatizing Media Coverage of Persons with Mental Illness Killed by Police,” Stigma and Health and J.Y. Choe, L.A. Teplin, and K.M. Abram (2008) “Perpetration of Violence, Violent Victimization, and Severe Mental Illness: Balancing Public Health Concerns,” Psychiatric Services, 59, pp.153-164.