Suicide By Cop? How Junk Science and Bad Law Undermine Accountability for Killings by Police

This Article offers the first critical examination of “suicide by cop” as a law enforcement theory that shifts the blame for excessive use of police force to victims, absolves police officers from accountability, and undermines civil rights. The term suicide by cop appeared in the 1980s as a moniker for encounters in which civilians intentionally provoke a lethal response from law enforcement, for example, when they leave behind a note expressing their wish to die at the hands of police. Since then, police and their advocates have developed suicide by cop into a malleable form of junk science that serves as a broad defense against personal and institutional liability.

Police officers can be held accountable for excessive use of force, but administrative sanctions such as suspension or termination are not often imposed, and criminal liability is rare. Civil rights lawsuits against police face a maze of challenges that protect officers from liability even if they acted unreasonably. In Boyd v. City and County of San Francisco, the U.S. Court of Appeals for the Ninth Circuit became the first in the country to allow expert testimony and character evidence about suicide by cop as an exculpatory rationale for police use of force, erecting another legal hurdle to accountability.

Through the lens of Boyd and a subsequent case in San Francisco that relied on its holding, this Article explores the origins and use of suicide by cop to defend officers and municipalities against liability for excessive use of force. The Article focuses on police training about suicide by cop, the emergence of suicide by cop as a forensic science, and the Ninth Circuit’s novel decision to permit the admission of expert testimony and character evidence about suicide by cop. In this relaxed scientific and evidentiary context, alleged suicidality introduces highly prejudicial information about victims of police violence in civil rights cases.

Suicide by cop demands greater scrutiny by police, mental health professionals, judges, and lawmakers, who should not use junk science and bad law to absolve officers of liability. Instead, police should be required to exercise a higher duty of care to protect people with mental illness and other members of the public from excessive use of force.

Table of Contents Show

Prologue

I went into that attic myself, just as he had that night, and climbed to the area where he was last alive. I saw my brother’s blood covering the floor and walls. There were holes from bullets everywhere, in the rafters and the walls. From where he would be positioned, it looked like bullets sprayed up from the bathroom below through the ceiling into the attic. A big hole was in the attic floor over the bedroom, where they must have pulled him down.

I couldn’t help but cry while I was in that place, trying to put myself in his place to find out what happened. Then the [police] chief herself changed the report twice and said the facts were not clear, they were just preliminary reports, and he was holding a cylindrical object in his hands that the officers thought was a gun. Then the report was changed again, stating he held an eyeglass case. After 14 hours of their crime scene team investigation there with my brother’s body, they found no weapon. The chief portrayed my brother like he wanted to get shot and the officers reacted appropriately.

– Kahlil Sullivan[1]

Introduction

On June 6, 2006, Officer Paulo Morgado of the San Francisco Police Department (SFPD) was dispatched to an apartment when a neighbor said the front door was open and she suspected it was being used as a drug house.[2] Officially, Officer Morgado was there on an SFPD code 910 “well-being check” when he pushed through the unlocked door without a warrant.[3] Before fully entering the apartment, he called for another police unit to conduct a walkthrough, and he was joined by SFPD Officers Michelle Alvis and John Keesor.[4]

Asa Sullivan, a twenty-five-year-old father of mixed race descent, was one of two friends staying in the apartment with the permission of the tenant.[5] They were helping their friend clean up the place in the hopes of getting his security deposit back.[6] After the police kicked down an internal door within the apartment and arrested the tenant without incident, they eventually found Asa in the apartment’s attic, where he had retreated upon their arrival.[7]

The officers on the scene offered conflicting accounts of the events that unfolded over a twelve-minute standoff and its aftermath. These accounts would later become central in a trial about their actions, though some of the facts were undisputed. When Officer Alvis first reported over police radio that Asa was in the attic, another officer responded, “Hey, why don’t we just pull back really quick, set up a perimeter and just try to get him later?”[8] Officer Alvis responded, “Cover both closets, I have him at gunpoint. He’s not going anywhere.”[9] Officer Morgado then requested a dog unit to try to help resolve the standoff.[10]

Officers Morgado and Keesor joined Officer Alvis in the attic, and with three guns drawn on Asa, Officer Morgado said that Asa was being uncooperative: “[H]e’s gonna be a 148,” the SFPD code for resisting arrest.[11] Another officer over police radio suggested using nonlethal options.[12] When those were ruled out because of obstructions in the attic, an officer on the scene said, “So why don’t we slow it down, see if we can get a hostage negotiator or something, because this guy’s not listening to us.”[13]

During the encounter, Officer Morgado also reported by radio that Asa “was trying to 801 by cop.”[14] 801 is the SFPD code for a person trying to commit suicide, suggesting that the officer thought Asa was trying to prompt the police to kill him.[15] In a civil rights lawsuit later brought by Asa’s family against the officers and the City and County of San Francisco for unreasonable use of force, the U.S. Court of Appeals for the Ninth Circuit summarized what happened next:

The officers stated that Sullivan refused to show his hands and made disturbing statements, such as “Kill me or I’ll kill you” and “Are you ready to shoot me?” Officer Alvis stated that she thought she saw something in Sullivan’s hands and when he moved his right arm that she thought he was going to shoot her. Officer Keesor stated that he saw something that looked like a gun in Sullivan’s hand, heard a pop, and began shooting at Sullivan.[16]

Asa’s alleged statements and the officers’ accounts of events were contested issues in the lawsuit. What is uncontested is that Officers Alvis and Keesor “fired their entire magazines” at close range, twenty-five shots in all, striking Asa more than a dozen times and killing him.[17]

Initially, the police officers falsely reported that Asa fired shots at them before they killed him in the attic.[18] But, as the Ninth Circuit would later note in weighing the balance between the government’s interests and Asa’s constitutional rights, Asa was not accused of a crime, did not pose a threat to public safety, and could not escape.[19] The appeals court further observed, Asa “had not initially caused this situation. He had not brandished a weapon, spoken of a weapon, or threatened to use a weapon. Sullivan, in fact, did not have a weapon.”[20] Although the department later acknowledged these facts, it nevertheless defended the officers’ actions by saying that Asa Sullivan died from “suicide by cop.”[21]

Asa’s mother, Kathleen Espinosa, sued the individual officers and the City of San Francisco for constitutional violations, seeking $10 million on behalf of Asa’s five-year-old son.[22] The lawsuit alleged that by entering the premises without permission and pointing his firearm at Asa, Officer Morgado violated Asa’s rights under the Fourth Amendment to be free from an unreasonable search and excessive force.[23] The lawsuit further alleged that Officers Alvis and Keesor deprived Asa of his Fourth Amendment right to be free from excessive force when they killed him in the attic.[24]

The family’s key claims at trial turned on whether the officers’ use of force against Asa was reasonable and necessary given what they knew at the scene.[25] To defend itself and the officers at trial, the City called psychiatrist Emily Keram to testify as an expert on suicide by cop. Dr. Keram conducted what she called a “psychological autopsy” of Asa, reviewing his police records, medical records, and social and developmental history, none of which were known to the officers at the time of the event.[26] Consistent with the official police narrative, Dr. Keram concluded that Asa spontaneously decided to commit suicide when confronted by police officers.[27]

In 2014, more than eight years after Asa’s mother filed the civil rights lawsuit, the jury returned a unanimous verdict in favor of the police officers and the City, finding that their actions were not unreasonable.[28] Although it is difficult to discern the reasons for a jury verdict, one of the family’s attorneys, seasoned civil rights lawyer John Burris, said that he thought the jury absolved the officers and the City of legal responsibility for Asa’s death because they concluded that he committed suicide by cop.[29] The attorney in Burris’ office who tried the case, Ben Nisenbaum, was devastated by the verdict: “I’ve had some heartbreaking cases, but it was the most painful trial I ever lost. Suicide by cop was used as a Trojan horse to get in all kinds of terrible evidence that would otherwise be inadmissible in order to cause the jury to turn its back on Asa.”[30]

I first learned about Asa’s story in 2015, when the Coalition for Police Accountability retained our law school clinic to provide legal support to pass a more robust police review law in Oakland, California. The goal of the new law was to increase the independence and authority of the local police review body to seek accountability for officer misconduct. Clinic students worked on the project for several years, helping draft both a ballot measure passed by voters in 2016 and the subsequent enabling ordinance adopted by the City Council in 2018.[31] The measure and ordinance established a Police Commission to oversee policies and procedures in the Oakland Police Department and a Community Police Review Agency to investigate and recommend discipline for police misconduct.[32]

As part of their work representing the coalition, students spoke with people impacted by police use of force—including Asa’s family—about their ideas for addressing the problem.[33] The family could not understand what had happened to Asa. From their perspective, the police had murdered their unarmed loved one who had not done anything wrong while he was cornered in an attic. Asa’s family asked the students to research suicide by cop, including three key questions:

1.  What is suicide by cop?

2.  Why did the police mention suicide by cop during the confrontation with Asa, when he wasn’t suicidal?

3.  Why did a judge allow a psychiatrist who never met Asa to say he was suicidal, including digging up so many things from his past?

The students prepared a memo for the family outlining what they found, much of which shocked us. First, suicide by cop emerged in the 1980s as an anecdotal theory of law enforcement to describe encounters in which civilians intentionally provoked police to kill them.[34] Soon after, researchers, many of them police psychologists or psychiatrists, gleaned cases of suicide by cop from press accounts and through other nonscientific methods to define and develop the theory. Almost since its inception, police and their advocates understood the potential of suicide by cop to serve as a potent defense to excessive use of force, as it did in Asa’s case.

Second, police invoke suicide by cop during encounters like Asa’s because they have been trained to do so.[35] As with other exculpatory defenses such as excited delirium, police are in no position to diagnose medical conditions, especially during high-stakes encounters. But diagnosis and treatment aside, police are trained that alleging suicide by cop is a professional best practice to defend against individual and institutional liability for excessive use of force.

Third, the trial court judge in Asa’s case allowed Dr. Keram to testify because the Ninth Circuit previously held that expert testimony and other evidence about suicide by cop was admissible to assess the credibility of the officer’s account of the events leading up to a shooting.[36] In that case, Boyd v. City and County of San Francisco, family members of twenty-nine-year-old Cammerin Boyd sued SFPD officers who killed him outside his car after a high-speed chase. Since Boyd, defense attorneys have increasingly asserted suicide by cop in excessive use of force cases to shield cities and their officers from liability.[37]

Making the students’ research findings public was not part of the work for the family, who just wanted to understand what happened to Asa. However, the family also poignantly asked the students what could be done to ensure that nothing like this ever happened again. This Article is an attempt to address that question, not by suggesting that suicide by cop never happens, but by demonstrating how the combination of junk science and bad law about suicide by cop undermines accountability for killings by police.

Part I of this Article provides a brief overview of police accountability mechanisms, including the many hurdles people face in obtaining justice in civil rights lawsuits against officers and municipalities. Law enforcement defenses to these cases have evolved in recent years to include junk science theories like suicide by cop.

Part II describes the history of suicide by cop in policing literature and officer and management training, including the law enforcement professionals who have developed and promoted the theory. Police officers are specifically trained to invoke suicide by cop during and after use-of-force incidents to try to avoid personal and organizational liability for their actions.

Part III reviews the forensic literature on suicide by cop as it has emerged over time, exploring key studies, methods, and findings. Though based on flawed research methods, these findings about suicide by cop are increasingly used in courtrooms in defense of police officers and to support the admission of expert testimony and related evidence.

Part IV describes the use of suicide by cop in civil rights lawsuits brought against police officers for excessive use of force. I argue that in deciding Boyd, the Ninth Circuit erred when it opened the door to a wide range of suicide by cop evidence, including allowing experts to testify about highly prejudicial victim information that was not known to officers on the scene. The court engaged in an unfaithful reading of Supreme Court precedent and circular logic to allow the admission of expert testimony about suicide by cop. Most importantly, the court erred in endorsing suicide by cop as scientifically valid, enabling highly prejudicial expert testimony and related evidence to absolve officers and cities of liability for excessive use of force.

Part V makes specific recommendations for law enforcement, mental health professionals, courts, and lawmakers to curb the misuse of suicide by cop as an exculpatory theory for excessive use of force that undermines civil rights laws.

I. Police Accountability and Junk Science

In a typical year, police officers in the United States kill more than one thousand people.[38] Roughly one-third of police killings occur after an alleged violent crime, but most occur during responses to nonviolent offenses like traffic stops, and one in eleven occurs as a result of a welfare or mental health check.[39] Police kill Black people at twice the rate (28 percent) of their percentage in the population (13 percent), and kill Black men at more than three times the rate they kill white men.[40] Despite public outcry and increased attention to racialized police violence, a 2021 study published in The Lancet found persistent and overwhelming racial disparities in police violence for Black people from 1980 to 2019.[41] Further, studies estimate that 10 percent of police calls, 17 percent of use of force incidents, and 20–25 percent of fatal police shootings involve people with mental illness.[42]

The 2021 Lancet study also found that more than half of all fatal encounters with the police were listed with a cause of death other than homicide, partly as a result of medical examiners and coroners who work within or directly for their police departments.[43] For example, even in high-profile cases like the police murder of George Floyd, the Minneapolis medical examiner’s preliminary autopsy “revealed no physical findings that support a diagnosis of traumatic asphyxiation” and opined that “potential intoxicants” and preexisting heart disease “likely contributed to his death.”[44]

The relationship between medical examiners and law enforcement reveals an important and often overlooked aspect of the challenge of police accountability, namely, science in service of impunity. Although Mr. Floyd’s killer was eventually found guilty of second-degree murder, criminal and civil liability for excessive use of police force are exceedingly rare relative to the scope of the problem. This Part describes the many obstacles to police accountability for excessive use of force and the use of junk science in defense of police misconduct.

A.     Police Liability for Excessive Use of Force

Police accountability advocates have been working for many years to end misconduct by law enforcement through a variety of legal mechanisms, including pursuing administrative sanctions, criminal liability, and civil liability.[45] Due largely to the power of police unions, the structure of the disciplinary process, and deeply entrenched institutional culture, police administrators rarely impose meaningful sanctions like suspension or termination on officers.[46] Criminal liability is even rarer. Prosecutors decline to bring criminal charges in more than 98 percent of killings by police, and notwithstanding a few notable cases like George Floyd’s, in the remaining less than 2 percent, they obtain convictions less than one-third of the time.[47]

With respect to civil liability, victims of police violence often seek money damages from officers and municipalities through federal lawsuits brought under 42 U.S.C. § 1983 (Section 1983) alleging violation of their constitutional rights.[48] Section 1983 was enacted as part of the Civil Rights Act of 1871, a Reconstruction Era law intended to protect African Americans against the racist vigilantism and violence of the Ku Klux Klan.[49] But it was not until 1961 that the U.S. Supreme Court held that civilians could sue police officers under Section 1983 for an unlawful search and seizure that violated their Fourth Amendment rights, opening the door to civil damage claims.[50]

Nevertheless, the legal hurdles for plaintiffs in Section 1983 cases are significant.[51] The Supreme Court has allowed defendant cities and officers to settle Section 1983 cases conditioned on a waiver of attorneys’ fees. This practice undermines the financial viability for lawyers to bring such cases, which in turn reduces access to counsel for victims of police violence.[52] Further, the Court has raised pleading requirements in federal courts, making it more difficult to bring cases in the first instance, especially where the government is in possession of key facts, such as in civil rights cases brought against police officers.[53]

In excessive use of force cases, qualified immunity shields many police officers from liability of any kind to the extent that their actions do not “violate clearly established statutory or constitutional rights of which a reasonable person would have known.”[54] In addition, U.S. Supreme Court holdings give courts wide latitude to determine the reasonableness of officers’ conduct and to substantially narrow the overall inquiry about police use of force.[55] Finally, even if officers are found to have violated a victim’s civil rights, they almost never incur personal liability because of formal and informal indemnification—the government typically pays all money damages, which reduces any deterrent effect on individual officers.[56]

On top of these high procedural barriers, police officers, police unions, and their advocates aggressively defend against civil rights lawsuits by invoking science to mitigate or escape liability for excessive use of force. These defense strategies are often deployed in cases involving so-called nonlethal police techniques, including the use of weapons, such as tasers and pepper spray, or the use of physical restraints, such as choke holds.[57] Over the last few decades, what critics have called a “cottage industry of exoneration” has grown, consisting of medical examiners, law enforcement officials, and lawyers who advise and train officers to defend themselves against liability for their misconduct.[58]

B.     Junk Science in Defense of Police Misconduct

In recent years, the cottage industry of exoneration for police killings has seized on the concept of “excited delirium” as the cause of death for victims in police custody independent of officer conduct.[59] Part of a larger category of so-called “custody death syndrome,” excited delirium is invoked to absolve police officers of responsibility for killing people during violent encounters.[60] Police are trained to recognize excited delirium, notwithstanding the dubious scientific basis for it. For example, one of the officers on the scene during George Floyd’s killing asked Officer Derek Chauvin, who was later convicted of murdering Mr. Floyd, whether they should roll him on his side, adding, “I just worry about the excited delirium, or whatever.”[61] Notably, Black and Brown people constitute more than half of all deaths in police custody attributed to excited delirium.[62]

Despite the lack of scientific evidence about excited delirium’s existence and its racially disparate impact, federal courts have accepted it as a valid medical condition and a factor relevant to establish whether an officer’s use of force was reasonable.[63] Critics of excited delirium call it junk science:

Excited delirium implies that there is a medical condition that predisposes certain individuals, often [B]lack men, to die in police custody. It draws upon aspects of real medical conditions such as delirium, psychosis, drug intoxication and sudden cardiac death. But it manipulates them to form a broadly applicable blanket diagnosis that serves the interests of law enforcement and absolves officers of accountability.[64]

The term junk science in this context refers to science that is manipulated to reach a desired result, especially in the courtroom, as opposed to pseudoscience or fake science, which have no basis in science. In his 2023 book-length treatment of junk science in criminal law, Chris Fabricant notes that the rapid growth of the U.S. legal system beginning in the 1970s—driven by mass incarceration (criminal cases) and personal injury lawsuits (civil cases)—resulted in an “explosive growth of the expert-witness industry in the latter part of the twentieth century.”[65] Junk science produced through the collaboration of medical professionals and law enforcement has a long history in the United States of being used by prosecutors and judges to wrongfully convict criminal defendants and impose harsher sentences.[66]

Junk science is also increasingly used to defend police officers accused of excessive use of force. Although suicide by cop has not received much critical attention in the context of police accountability, the theory was developed alongside excited delirium as another exculpatory defense for police misconduct. There are important differences between suicide by cop and excited delirium, including the increasing rejection of excited delirium by the medical and scientific communities. However, suicide by cop has similar junk science attributes.[67] Analogous to excited delirium, suicide by cop draws on real-world examples—in this context, of people who want to die at the hands of police—and manipulates the definition to encompass a much larger category of cases.

As described in the Part II, Officer Morgado’s report that Asa Sullivan was “trying an 801 [suicide] by cop” and the defense theory of suicide by cop later alleged in the civil rights case are part of an explicit strategy developed in the policing literature to characterize use of force incidents as suicide by cop.

II. Suicide by Cop in Police Training

The term “suicide by cop” has been used within law enforcement since at least 1983, when it was reportedly coined by former California police officer Karl Harris.[68] The concept first emerged in the medical literature in a 1985 public health case report about an incident “in which a person with a previous known suicide attempt clearly provoked the Philadelphia police into shooting him.”[69] The term generally refers to an encounter in which a person intentionally provokes the police to kill them. The paradigmatic case is premeditated, including documented instances in which the victim leaves a note explaining their motivation and even their regret in involving law enforcement. The definition of suicide by cop has subsequently been stretched to include a wide range of events, and according to police advocates, most cases are unplanned and spontaneous.

This Part describes suicide by cop in police training. The earliest accounts of suicide by cop in the trade literature were based on a small number of case studies, but during the 1990s and 2000s, police began describing a growing “phenomenon” of suicide by cop that was more dangerous to officers than to the public.[70] Starting in the late 1990s, suicide by cop was incorporated into police training across the country in response to this newly described threat. Although private law enforcement consultants provide training on suicide by cop around the country, the focus here is on officer training in California, including the training of law enforcement managers who are especially concerned about individual and organizational liability.[71]

Suicide by cop’s appearance in the mid-1980s was not a coincidence. The period was marked, among other things, by “wars” on drugs and crime and the culmination of several decades of deinstitutionalization of people with mental illness.[72] Excited delirium was also described for the first time in the 1980s in a single case study of “acute cocaine delirium” and a subsequent study of seven anecdotal incidents.[73] As one critic observed about the rise of suicide by cop as a policing theory during this era:

[In the 1980s], new theories of criminality and social deviance were changing the foundations of policing. One theory soon to become orthodox saw derelict neighborhoods—signified by “broken windows”—as sites of criminality demanding control. Simultaneously, mental illness was increasingly being treated by jails and prisons, rather than by other institutions. “Suicide by cop” joined a bundle of new concepts that recast the perceived social crises of the day as manifest in a group of violent people who would only understand violence in return; the rise of a militant police force could be explained, in this logic, as self-protection.[74]

Genuine cases of suicide, with or without the involvement of police, are tragedies for everyone involved.[75] But suicide by cop also became a sensationalized headline, and by the early 1990s, it had become a frequent, if anecdotal, topic in the policing literature.[76] Stories about suicide by cop mostly focused on the harm to police officers, the responsibility of the person who was harmed or killed, and the threat of liability to law enforcement. Suicide by cop pieces ran in a wide range of trade publications such as Law & Order,[77] The Police Chief,[78] Police: The Law Enforcement Magazine,[79] The Police Marksman,[80] Officer.com,[81] Police1,[82] and Americans for Effective Law Enforcement.[83]

The California Commission on Peace Officer Standards and Training (POST), which was established by the State Legislature to promulgate minimum training standards for law enforcement, has provided formal trainings on suicide by cop since at least 1998.[84] In these trainings, suicide by cop is situated within a larger police culture of shoot first and ask questions later.[85] For example, in the 1998 POST training video on suicide by cop, forensic psychiatrist Park Dietz, known for his controversial work on behalf of prosecutors and the police, said police officers had to be able “to act without hesitation with regard for their own safety regardless of whether some of the offenders are really suicidal.”[86] In a “community for cops” trade magazine, firearms and policing author R.K. Campbell cautioned officers not to hesitate to act even if they know a person is mentally ill: “It doesn’t matter if every instinct tells you that he or she is attempting a police­assisted suicide; you cannot assume that an emotionally disturbed person who is armed will not kill you in pursuit of his or her own certain destruction.”[87]

Suicide by cop trainings are grounded in common themes of the trade literature, especially that use of force incidents generate bad publicity for police and harm officers’ mental health. In fact, training on suicide by cop emphasizes that the officer is not responsible for the person’s death. Instead, police officers are taught that the suicidal individual is the “perpetrator” or “precipitator” and the officer is an unwilling executioner of a suicidal criminal.[88] In the POST suicide by cop training video, speakers tell police officers that “a suicide by cop event is a trap—you are the victim.”[89]

Finally, and most relevant here, police officers are trained that suicide by cop exposes them to liability, especially civil rights lawsuits for excessive use of force.[90] Trainers and law enforcement management specifically encourage police to invoke suicide by cop during and after violent encounters to reduce personal and organizational liability for excessive use of force.[91] As early as 2001, the Chief Psychologist of the Los Angeles County Sheriff’s Department, Audrey Honig, wrote that it was important to conduct a comprehensive investigation in use of force cases to support the theory of suicide by cop and “reduce both the officer’s and department’s vulnerability to a wrongful death lawsuit.”[92] She suggested that such investigations could inform police responses to suicide by cop attempts “given that these events can also lead to litigation and allegations of excessive use of force.”[93] Like most other law enforcement commentators on the phenomenon, Honig emphasized how important it was to identify suicide by cop cases to help “victim officers” deal with the stress and trauma of the incident.[94]

In 2006, Chief Honig more directly lamented the reputational and financial harms suicide by cop imposes on law enforcement. To “buffer and protect our agencies” and preempt them from public outcry and civil litigation resulting from police shootings, she asked whether all officer-involved shootings should be investigated as suicide by cop incidents.[95] Honig even offered specific strategies to develop evidence of suicide by cop and bypass rules excluding such evidence in court, noting that initial responders are often in the best position to assess a person’s behavior and should note suicidal motivation in the original incident report.[96]

Honig emphasized the importance of obtaining spontaneous statements—exceptions to the hearsay rule that can be admitted in court—made by family, friends, and witnesses, because “such statements about a desire to die are frequently made immediately following an incident.”[97] She concluded: “The determination that an incident constitutes an actual or attempted suicide by cop can make the difference between significant personal and organizational liability and zero liability. An ounce of prevention is truly worth a pound of cure.”[98]

Current training materials continue to reflect these core themes. According to the stated learning goals in the most recent available POST training materials on suicide by cop for California law enforcement managers, “[P]articipants will better understand the need to conduct a parallel investigation into the incident of suicide (SbC) perpetrated by the subject with the goal of reducing civil liability, assisting with the administrative adjudication, and assisting the involved officer(s) in their recovery from their part in this tragic incident.”[99]

The content of these trainings sheds important light on Asa Sullivan’s case, where the SFPD officers invoked suicide by cop both during and after the incident. Recall that Officer Morgado, the first on the scene, reportedly radioed that Asa was trying to commit suicide by cop, even though Asa had not initiated the encounter with police, was not suspected of a crime, had retreated, and was unarmed. Further, the SFPD did not classify Asa Sullivan’s death as an unlawful killing. In fact, when Asa’s younger brother Sangh went to the medical examiner’s office the day after his killing to find out what happened to him, homicide investigators refused to answer his questions and instead interrogated him for six hours.[100] Although the meeting was not recorded, the interrogating officer later testified at trial that Asa’s brother provided information during the interview suggesting that Asa was suicidal.[101]

From the perspective of law enforcement, suicide by cop is a threat to police-community relations, officers’ mental health, and their individual and institutional pocketbooks. These concerns have made their way into police training, which encourages police officers to invoke and investigate suicide by cop as an insurance policy against liability. As an extension of this strategy, policing advocates have advanced suicide by cop in court as a theory grounded in forensic science, allowing it to be admitted into evidence to explain away police use of force. Part III explores the development of suicide by cop in the scientific literature.

III. Suicide by Cop as a Forensic Science

Meaningful suicide by cop research did not begin in earnest until the late 1990s. American criminologists first described forms of “victim-precipitated homicides” between civilians in the 1950s, but until the 1985 case report from Philadelphia, the academic and scientific literature had not identified or analyzed suicide attempts in the policing context.[102] Since 1998, at least two dozen studies have been published in a variety of peer-reviewed and non-peer-reviewed journals, mostly about the phenomenon in the United States.[103]

This Part explores suicide by cop as a forensic science that experts rely on in their testimony and judges consider in deciding whether to admit evidence in court. The literature over the last twenty-five years has tried to define and measure suicide by cop by retrospectively identifying motivations or indicators that might shed light on the phenomenon. Researchers have deployed a variety of methods toward these ends, typically using convenience samples either from open sources (e.g., the media and internet) or from police or medical examiner records.[104]

In 1998, researchers published two important studies proposing criteria to identify suicide by cop, applying the criteria to a sample of officer-involved shootings to measure the prevalence of the phenomenon. Hutson et al. reviewed 437 officer-involved shootings (OIS) during a ten-year period in the Los Angeles Sheriff’s Department.[105] The authors, including a long-time homicide detective in the department, used the following criteria to screen the cases for suicide by cop: (1) evidence of suicidal intent, (2) evidence they wanted officers to shoot them, (3) evidence they possessed or appeared to possess a lethal weapon, and (4) evidence they intentionally escalated the encounter and provoked officers to shoot them.[106] Applying these criteria, researchers found that 10.5 percent of all OIS cases met the review criteria, including 12.5 percent (one in eight) of all fatal cases.[107]

In the second study, Kennedy et al. identified 240 use-of-force incidents from an electronic library of twenty-two newspapers in eighteen cities.[108] The researchers asked police officers to review the newspaper articles and classify the cases into five categories: (1) probable suicide, (2) possible suicide, (3) uncertain, (4) suicide improbable, and (5) no suicidal evidence.[109] These categories were briefly defined, and the officers agreed with each other’s classifications 74 percent of the time.[110] In the end, they concluded that 16 percent of the cases involved probable or possible suicide by cop and a total of 46 percent possessed “some evidence of probable or possible suicidal motivation.”[111]

In 2000, researchers published three more suicide by cop studies and a literature review attempting to refine the variables and develop typologies of the phenomenon. Lord reviewed sixty-four suicide by cop cases from the 1990s selected for her by thirty-two law enforcement agencies in North Carolina and applied historical, personality, behavioral, and situational factors to measure differences between people who successfully forced officers to shoot them and people who were unsuccessful.[112] Lord, a former police investigator and law enforcement trainer, found that most factors, including substance use, prior suicide attempts, stressful life events, and homicidal conversation during the incident had “weak relationships with the outcome of the incident.”[113]

Homant et al. reviewed 123 police use of force cases drawn from a LexisNexis search (40), a Canadian master’s thesis (28), case synopses from medical examiners (15), case studies from other authors (13), one coauthor’s prior research (9), cases on which the authors consulted (8), an ABC 20/20 program (5), cases found on the internet (3), one circuit court case, and one case from a local police department.[114] The study found significant danger to police in suicide by cop incidents and identified fourteen possible motives for someone to engage in suicide by cop organized under four categories: psychodynamic, social values, practical, and manipulative.[115]

Homant and Kennedy reviewed Homant et al.’s prior dataset of 123 cases plus another twenty-two nonrandom cases obtained from a further newspaper search, prosecutors, and appellate court literature on police shooting cases to generate a typology of suicide by cop.[116] To classify a case as suicide by cop, the researchers “required that the subject behaved in a way that seemed intended to provoke the police to shoot[,] . . . either deliberately exposing oneself while posing a threat to police or bystanders, or, in a standoff situation, knowingly forcing police to attack while harming or threatening to harm hostages or bystanders.”[117] The authors categorized the suicide by cop cases as: (1) “direct confrontations,” in which subjects instigated attacks on police (premeditated, 30.8 percent), (2) “disturbed interventions,” in which subjects took advantage of police intervention to try to end their own lives (opportunistic, 57.3 percent), and (3) “criminal interventions,” in which subjects preferred death to submission (spontaneous, 11.9 percent).[118]

In their 2000 literature review, Mohandie and Meloy identified a variety of “indicators” of suicide by cop.[119] Their review included the studies described above and case studies of suicide by cop in the policing literature. From these varied sources, the authors, one of whom was a psychologist at the Los Angeles Police Department (LAPD), derived and categorized indicia of suicide by cop into twelve verbal cues (such as “offers to surrender to the person in charge”), eighteen behavioral cues (such as “no demands”), ten historical or situational indicators (such as “faces an arrest or criminal justice situation perceived as serious”), and eight key life events (such as “poor socioeconomic background”).[120]

The pace of research into suicide by cop slowed since 2000, but three studies of note have been published since then. First, Best et al. reviewed twenty-two cases from England and Wales over a four-year period in which a police weapon was discharged and someone was injured.[121] The researchers were interested in determining the suicidal motivation of the people who were shot, including nine who were killed. Applying their proposed criteria to the incidents—nine questions in four categories: primary evidence of suicidal intent, secondary evidence of suicidal intent, state-based indicators of irrationality, and minimal evidence of suicidal intention—the authors concluded that eight of the twenty-two cases (36.4 percent) were probable suicides, with additional cases indicating some suicidal intent.

Second, Mohandie et al. reviewed a nonrandom sample of more than seven hundred cases of officer-involved shootings in North America from 1998 to 2006 to discern the frequency and other features of suicide by cop.[122] The researchers coded the cases along 110 variables, twenty-nine of which were deemed indicators of suicide by cop, under three broad categories: incident characteristics, subject data, and outcomes.[123] Applying these indicators, the study found that 36 percent of the cases were suicide by cop and another 5 percent evidenced some suicidality.[124] Further, the study found that more than 81 percent of the suicide by cop cases were unplanned and spontaneous, higher than in previous studies.[125]

Finally, in 2019, Jordan et al. reviewed 419 cases handled over a six-year period ending in 2015 by the LAPD’s Mental Evaluation Unit—officers specially trained to consult on calls involving mental health crises—that were coded by the unit as suicide by cop.[126] The coauthors were interested in whether this sample would yield different results than samples drawn from officer-involved shootings. Applying a modified version of Mohandie et al.’s 2009 coding variables to the LAPD sample, the researchers found that these incidents were much more often in response to calls involving suicidal or mentally ill subjects, much less likely to be spontaneous than in the Mohandie et al. study, and much less dangerous than in the Homant et al. study.

Considering the large racial disparities in killings by police, especially of Black men, it is a glaring omission that most suicide by cop studies do not collect and report data on race and ethnicity.[127] In the few studies that provide demographics, Black people are overrepresented in suicide by cop incidents, though none of the studies remark upon it.[128] In one national study, when compared to their representation in the United States population at the time, Black people were overrepresented by 27 percent.[129] In two Los Angeles studies, when compared to their representation in the local population at the time, Black people were overrepresented by almost 300 percent.[130]

Police management, trainers, and lawyers rely on these studies to supervise, educate, and defend officers, and they have been used to support a number of assertions about suicide by cop that have become widely accepted as facts within policing. First, the studies convey the idea that suicide by cop is prevalent, representing anywhere from 10 to 50 percent of all officer-involved shootings, and it is likely growing in frequency. Second, according to the literature, most suicide by cop incidents are not premeditated; they are either opportunistic or completely spontaneous. Third, researchers suggest that suicide by cop incidents are very dangerous for police officers. Fourth, even with some authors cautioning against using their research to interpret individual cases, the studies identify dozens of indicators that can be applied to diagnose suicide by cop after a killing by police. Finally, the complete silence on racial bias from suicide by cop researchers and trainers reveals shocking inattention to a central feature of all policing, but especially in the use of excessive force.

Part IV considers how core ideas about suicide by cop have made their way into the courtroom, particularly in the wake of Boyd, which opened the door to more widespread use of the concept as a defense in civil rights lawsuits.

IV. Suicide by Cop in the Courtroom

Suicide by cop is invoked in court by law enforcement as a defense against claims of excessive use of force in the context of Section 1983 civil rights lawsuits.[131] To prevail in a Section 1983 claim, the plaintiff—typically the person harmed by alleged police misconduct or a surviving family member—is required to establish by a preponderance of the evidence that an officer, while acting in their official capacity, used force not reasonably necessary under the circumstances.[132] Since the Supreme Court has held that police use of deadly force is considered a seizure under the Fourth Amendment, courts must determine whether the officer’s behavior is “reasonable[]” under the totality of the circumstances of the event in question.[133] Further, in Graham v. Connor, the Supreme Court held that courts must assess the objective reasonableness of a particular use-of-force incident from the perspective of an officer on the scene and not with the benefit of hindsight.[134]

Because the reasonableness of the officer’s behavior is central to the inquiry in Section 1983 lawsuits, the parties frequently contest the events prior to the use of force. In Boyd v. City & County of San Francisco, a Section 1983 lawsuit for excessive use of force, the police and municipality invoked suicide by cop to support the officer’s version of events leading up to the moment he killed Cammerin Boyd.[135] The trial court admitted suicide by cop evidence, including expert testimony about the theory and other details about Mr. Boyd’s background, in spite of Graham’s general prohibition against admitting evidence not known to the officers at the scene.[136] In affirming the trial court’s decision to allow the suicide by cop evidence, the U.S. Court of Appeals for the Ninth Circuit held, “In a case such as this, where what the officer perceived just prior to the use of force is in dispute, evidence that may support one version of events over another is relevant and admissible.”[137]

This Part explores suicide by cop in the courtroom through the lens of Boyd, which upheld a series of evidentiary rulings by the trial court. First, the Boyd court ruled that evidence about suicide by cop was legally relevant to the case, meaning that it shed light on the probability of one or more consequential facts in the case.[138] Next, it held that the expert testimony about suicide by cop “pass[ed] muster” as acceptable science and was admissible under current federal rules and case law.[139] Further, the court found that the suicide by cop evidence fell “within the large exception for otherwise inadmissible character evidence,” and thus was properly allowed.[140] Finally, the court said that the probative value of the suicide by cop testimony outweighed any potential prejudice, so it was not otherwise barred.[141]

The decision in Boyd to affirm the trial court’s admission of suicide by cop evidence, including expert testimony, was notable and arguably wrong in at least three important ways. First, as part of its relevance analysis, the Boyd court engaged in an unfaithful reading of Graham to skirt its general prohibition against hindsight evidence. Second, with respect both to general relevance and to character evidence, the court engaged in circular reasoning, allowing expert testimony about suicide by cop that relied on the officer’s version of events to corroborate the officer’s version of events. Finally, the court erred in endorsing suicide by cop as scientifically valid, which has opened the door in subsequent cases to expert testimony and related evidence about what is still a largely untested theory, introducing highly prejudicial information about people killed by police and absolving officers of accountability for excessive use of force.

A.     Boyd Unfaithfully Relied on a Footnote in Graham to Allow Hindsight Evidence About Suicide by Cop

In May 2004, San Francisco Police Department Officer Timothy Paine shot and killed twenty-nine-year-old Cammerin Boyd outside his vehicle following two alleged attempted kidnappings and a high-speed car chase.[142] Mr. Boyd’s family filed a Section 1983 civil rights lawsuit against the City and County of San Francisco claiming excessive use of force and seeking monetary damages.[143] To defend against the claims of excessive force, San Francisco city attorneys called a forensic psychiatrist, Dr. Emily Keram, to testify that Mr. Boyd committed suicide by cop.[144] The jury issued a verdict on behalf of the defendant City and County, and the family appealed to the Ninth Circuit, which affirmed the judgment.[145]

The parties disputed many facts leading up to Mr. Boyd’s killing. Officer Paine testified that Mr. Boyd, immediately prior to his killing, ignored police commands to surrender and instead reached inside his SUV, as if to grab a weapon, prompting the officer to shoot.[146] Witnesses said that in the moments before being shot by Officer Paine, Mr. Boyd raised his hands to surrender. Because Mr. Boyd was a double amputee with two prosthetic legs, they said, he was leaning against the car to lower himself to the ground when Officer Paine opened fire.[147]

Under Graham, the family was required to show that Officer Paine behaved unreasonably under the totality of the circumstances, and the jury was asked to determine whether his behavior was objectively reasonable from the perspective of an officer on the scene, not with the benefit of hindsight.[148] In other words, the jury could only evaluate the officer’s conduct based on circumstances the officer was aware of at the time of the incident. This protection is intended to keep officers from using later-discovered evidence to justify their actions. For example, if an officer used force against a person without knowing they were armed, the officer could not use a later-discovered weapon to justify the prior use of force.

At trial, the defense offered evidence about suicide by cop, including Dr. Keram’s testimony, not explicitly for the purpose of justifying the killing, but as evidence that supported Officer Paine’s version of events (that Mr. Boyd appeared to be reaching for a gun) over the version of other witnesses (that Mr. Boyd was trying to surrender). After performing what she described as a “psychological autopsy” to determine Mr. Boyd’s state of mind at the time of his death, Dr. Keram concluded that “Mr. Boyd did commit a suicide by cop on May 5, 2004,” thereby corroborating, according to the defense, Officer Paine’s testimony.[149]

To support Dr. Keram’s theory about suicide by cop, the trial court admitted five additional pieces of hindsight evidence proffered by the defense, which would have otherwise been excluded under Graham:

1.  Mr. Boyd’s prior criminal history and the potential prison sentence he faced if prosecuted for other alleged crimes just prior to the high-speed chase, both characteristics which Dr. Keram testified are common indicators of suicide by cop;[150]

2.  an April 1993 high-speed chase that resulted in the loss of Mr. Boyd’s legs as a teenager, because Dr. Keram testified that people can become suicidal around the anniversary of traumatic events, and Mr. Boyd was killed in May 2004, approximately eleven years later;[151]

3.  rap lyrics found in Mr. Boyd’s car several weeks before he was killed advocating the shooting of police officers, which Dr. Keram testified could have explained his animus toward police, another purported indicator of suicide by cop;[152]

4.  Mr. Boyd’s arrest in Oakland three days before he was fatally shot, because Dr. Keram testified that the arrest was a “practice run” and such rehearsals are a “telltale sign” in cases of suicide by cop;[153] and

5.  a lawsuit Mr. Boyd previously brought against the police for excessive use of force, which Dr. Keram testified helped form her opinion “that one of his reasons for choosing suicide by police was allowing his family to prevail in civil litigation.”[154]

The Ninth Circuit affirmed the trial judge’s rulings to admit all this evidence, except for the rap lyrics, which the court deemed was harmless error.[155] The court held that evidence about Mr. Boyd’s criminal history, 1993 car chase, arrest three days prior, and previous lawsuit against the police were relevant and admissible because they made it more probable that Mr. Boyd was committing suicide by cop.[156] Even the rap lyrics were relevant because “praising the murder of police officers also makes actions consistent with suicide by cop more probable,” but they were nevertheless “unfairly prejudicial in light of their offensive nature” because they also referenced and advocated for prostitution, which had no probative value in the case.[157]

Of course, these facts were not known to Officer Paine at the time he killed Mr. Boyd, so why were they admitted given Graham’s general prohibition against hindsight evidence? In deciding to affirm the admission of hindsight evidence, the Boyd court relied on a footnote in Graham, which it said allowed “a factfinder [to] consider outside evidence ‘in assessing the credibility of an officer’s account of the circumstances that prompted the use of force.’”[158] The text of footnote 12 in Graham reads: “Of course, in assessing the credibility of an officer’s account of the circumstances that prompted the use of force, a factfinder may consider, along with other factors, evidence that the officer may have harbored ill-will toward the citizen.”[159]

Whether the court’s aside is consistent with the opinion as a whole is debatable, but on its face, the footnote is written to protect the public against police by allowing evidence of officer animus against an individual for the purpose of challenging the officer’s version of events. The footnote does not say anything about allowing hindsight evidence about the individual subjected to police use of force for the purpose of supporting the officer’s account.[160] In other words, the Boyd court turned the cited footnote language in Graham on its head.

The Boyd court distinguished its reading of Graham from the Seventh Circuit’s holding in Palmquist v. Selvik, one of the earliest published cases to consider admission of evidence about suicide by cop.[161] The Palmquist court upheld a trial court’s ruling to exclude some evidence of suicide by cop, which the Boyd court says was because in that case “the events immediately prior to the use of force were undisputed.”[162] This characterization of the holding in Palmquist is questionable—Officer Selvik alleged he acted in self-defense when attacked by the decedent Mr. Palmquist, and the Palmquist court unequivocally restated Seventh Circuit caselaw with regard to Graham and hindsight evidence.[163] The Ninth Circuit, conversely, circumvented Graham through an unfaithful reading of a footnote meant to protect the public from police misconduct, not to protect the police from liability.

B.     Boyd Engaged in Circular Reasoning to Hold that Suicide by Cop is Relevant and Admissible

On appeal, Mr. Boyd’s family challenged the admission of multiple pieces of evidence, including Dr. Keram’s testimony, arguing that the trial court should have excluded the suicide by cop evidence because it wrongly suggested to the jury that Mr. Boyd was suicidal. The Ninth Circuit rejected that argument, holding that the suicide by cop evidence was not excluded by the rule barring character evidence “[t]o the extent that being shot by the police was Cammerin’s plan, intent, or motive.”[164] The family also argued that even if the evidence was not inadmissible as character evidence, per se, it nevertheless should have been excluded as unfairly prejudicial because it painted Mr. Boyd in such a negative light.[165] The Boyd court likewise rejected this assertion and deferred to the trial court’s weighing of the probative value of the evidence versus its prejudicial effect.[166]

The Federal Rules of Evidence generally prohibit the admission of evidence about a person’s character to show that they acted in conformity with that character during the event in question.[167] More specifically, the rules state that evidence of other crimes, wrongs, or acts are inadmissible to demonstrate a person’s character or that a person acted in accordance with that character.[168] The rules, however, permit certain kinds of character evidence to prove things like motive, intent, and planning.[169] Even if admissible for this purpose, the rules still require excluding evidence if its probative value is substantially outweighed by unfair prejudice.[170]

In affirming the trial court’s evidentiary ruling on suicide by cop evidence, the Boyd court noted that a substantial portion of the challenged evidence was admitted “in conjunction with its decision to allow expert testimony regarding the suicide by cop theory.”[171] As described above, the Boyd court held that Mr. Boyd’s criminal history, 1993 car chase, arrest three days prior, and previous lawsuit against the police were relevant and admissible because they made it more probable that he was committing suicide by cop.[172] At the same time, according to the court, “Cammerin’s alleged resistance to the police is also made more probable by evidence, in the form of expert testimony, that his actions were consistent with an attempt to commit suicide by cop.”[173]

To reach these conclusions, the Boyd court engaged in two concentric loops of circular reasoning. First, and more narrowly, the court admitted Dr. Keram’s testimony purportedly to help resolve disputed facts about the officer’s version of events in the moments just before he killed Mr. Boyd, but her testimony did just the opposite. Second, Dr. Keram’s testimony revealed a broader tautology of suicide by cop, suggesting that diametrically opposed actions by a suspect—asking the police to kill you or asking the police not to kill you and offering to surrender—could equally be interpreted as evidence of suicidal behavior.

With respect to the narrower form of circular reasoning, Dr. Keram’s opinion that Mr. Boyd intended to commit suicide by cop depends largely on the accuracy of Officer Paine’s version of events. That is, in rendering her opinion about suicide by cop, Dr. Keram assumed that Mr. Boyd reached inside his vehicle for a weapon as alleged by Officer Paine, but that was precisely the disputed fact that her testimony was supposed to help resolve, not rely upon. We can be confident of the centrality of this disputed fact to Dr. Keram’s opinion because she confirmed as much on cross-examination in a key exchange (the questions are from Mr. Boyd’s lawyer and the answers are from Dr. Keram):

Q. Okay. You told us that the person . . . wanting to commit suicide by cop needs to take some type of action at that moment to incite the police to use deadly force against him, correct?

A. Yes.

Q. I’m asking you standing up, leaning back with your hands up, at that moment is he doing anything to incite the police to shoot him?

A. No, if that’s what’s happening in that moment – that moment is not a moment in which he is provoking lethal force.

Q. And if the incident happened that way, hypothetically, you would not say this was suicide by cop, would you?

A. Again, prefacing it with him not giving a perception that he was moving or about to move.

Q. The hypo I gave you.

A. That he is just standing still with his hands raised.

Q. In that hypo.

A. Like I said, barring him giving any indication or perception that somebody would have that, you know, he had just gotten into that position or he was coming out of that position, I can’t think of at least at this point of, you know, something that would suggest that that was a suicide by cop if he was just shot while his hands were up.[174]

In other words, if Mr. Boyd was shot with his hands up, as described by the lay witnesses, Dr. Keram’s opinion was that he was not committing suicide by cop. And if Dr. Keram’s diagnosis of suicide by cop is dependent on Officer Payne’s version of events, the admission of all the other alleged evidence of suicide by cop is also improper, since it does not make it more probable that Mr. Boyd acted in a way consistent with Officer Payne’s testimony. While the trial and appellate courts said the expert testimony about suicide by cop was admitted to corroborate the officer’s version of events, Dr. Keram’s diagnosis of suicide by cop assumed the officer’s disputed statements were true in arriving at her opinion.[175]

Beyond the first form of circular reasoning, Dr. Keram’s testimony in Boyd also revealed a broader tautology of suicide by cop. For example, when asked about evidence that Mr. Boyd raised his hands in the air and stated, “Please don’t kill me, I don’t want to die,” Dr. Keram testified that Mr. Boyd’s statements did not demonstrate lack of suicidal intent.[176] While acknowledging that none of the literature she relied upon for her opinion included victims making statements to this effect, Dr. Keram testified that “people may be ambivalent or playing cat and mouse with the police or having other reasons for making that statement that there’s certainly very credible reasons that people could make that statement in suicide-by-police cases.”[177]

Dr. Keram testified that because of this ambivalence, the literature finds that “people who are committing suicide by police may at first appear to surrender.”[178] She said that appearing to surrender, including raising one’s hands, is performative conduct designed to help family members prevail in civil litigation because it “fix[es] in the witness’s minds that he was surrendering to police. And which then again, of course, puts in the witness’s minds the perception of this as an unfair shooting, that he’s being compliant.”[179]

Dr. Keram’s testimony has an Alice in Wonderland quality. If you ask the police to kill you, you are committing suicide by cop. If you ask the police not to kill you, it is not not suicide by cop. You could just be ambivalent or toying with the police when you are in fact trying to induce them to kill you. Likewise, if you refuse to surrender, that is suicide by cop. However, if you try to surrender, you could also just be ambivalent or trying to create the perception that this is an unfair police shooting so your family can win a civil rights lawsuit.

Ultimately, Dr. Keram’s testimony in the case did not help resolve disputed facts. Her testimony was predicated on the assumption that Officer Paine’s disputed version of events was true. And according to Dr. Keram, even if the officer’s version of events was untrue, Mr. Boyd’s attempt to surrender was no more than a ruse to help his family prevail in a civil rights lawsuit after he was killed. These layers of circular reasoning subvert Graham’s key holding with respect to prohibiting hindsight evidence and should not be the basis for admitting suicide by cop testimony and bootstrapping additional highly prejudicial information about people killed by police.

C.     Boyd Erred in Endorsing Suicide by Cop as Scientifically Valid

The Boyd court’s most consequential holding was to affirm the trial court’s evidentiary ruling to admit Dr. Keram’s testimony about suicide by cop and the other hindsight evidence admitted pursuant to that decision. Suicide by cop evidence was clearly relevant because at trial and on appeal the defense fought hard to get it in and the plaintiffs fought hard to keep it out: The parties, the trial court judge, and the appellate judges all understood that admitting suicide by cop testimony and related evidence would paint Mr. Boyd in a very negative light and benefit the defendants.[180] Further, Dr. Keram is an accomplished psychiatrist who easily qualified as an expert on suicide by cop under the federal rules.[181]

But relevance and expertise are not the end of the inquiry. The trial and appellate courts erred in endorsing the theory of suicide by cop as scientifically valid given its junk science attributes. Chris Fabricant’s description of how junk science is created and contributes to bad law is instructive with respect to suicide by cop:

Junk science relies extensively on case studies published in “peer-reviewed” journals like the JFS [Journal of Forensic Science]. The papers serve many purposes: In court, any publication is impressive; few are ever actually read by the lawyers. Within the forensic community, the case studies elevate the field and establish its leading practitioners . . . . The self-serving feedback loops are published in “the literature” and create the appearance of a robust canon of scholarship.[182]

To be admissible in court, evidence in general must assist the judge or jury to understand or determine a factual issue in the case.[183] In particular, courts must ensure that experts who testify have the requisite knowledge, skill, experience, training, or education to offer an opinion.[184] Further, the court must find that the testimony is grounded in sound data and methods that the expert faithfully applies to the facts in the case.[185] In Daubert v. Merrell Dow Pharmaceuticals, Inc., the U.S. Supreme Court set forth a nonexhaustive list of criteria for making this assessment, including whether the expert’s technique or theory is testable for reliability, has been subject to peer review and publication, has a known or potential error rate, and is subject to professional standards and generally accepted in the field.[186]

In reaching its conclusion that the lower court’s decision to admit Dr. Keram’s testimony about suicide by cop was not an abuse of discretion, the Boyd court noted that the trial judge held a hearing to examine Dr. Keram’s testimony under the Daubert factors before issuing a ruling.[187] According to the trial court ruling: (1) the theory of suicide by cop could not be tested by running a scientific experiment, but it was supported by peer-reviewed studies that applied forensic psychology methods to identify suicide by cop incidents after the fact; (2) Dr. Keram formed her opinion about Mr. Boyd’s case in relation to the literature on suicide by cop, which included more than a dozen peer-reviewed and non-peer-reviewed publications; (3) although there was no discussion of the potential error rate, Dr. Keram relied on studies that used criteria to exclude false positives; and (4) the theory appeared to be generally accepted in the relevant professional community.[188] Unfortunately, as the Boyd court also noted, the family’s lawyer did not call an expert to testify about suicide by cop at trial or offer other evidence to suggest that the theory is unreliable.[189]

Nevertheless, the trial court and Ninth Circuit should have been more skeptical of suicide by cop research and excluded Dr. Keram’s testimony and the other hindsight evidence it bootstrapped. First, suicide by cop researchers have proposed and applied different definitions of the phenomenon, making it impossible to compare and verify findings. Second, suicide by cop research is plagued by other methodological shortcomings, including the common use of nonrandom samples, which limits what we can learn from these studies. Third, the indicators of suicide by cop proposed by researchers have not been tested for validity (accuracy) and are often not tested for reliability (consistency). Finally, the act of committing suicide requires intent, which is completely missing from studies of suicide by cop.

1.     There is No Consensus in the Literature About What Constitutes Suicide by Cop.

The suicide by cop literature is still in its relative infancy, and the field is beset by a lack of consistent terminology. Suicide by cop has been variously referred to as “law-enforcement-assisted suicide,” “law-enforcement-officer-forced-assisted suicide,” “police-involved-victim-provoked shooting,” “victim-precipitated homicide,” and “suicide by police.”[190] This problem with nomenclature extends to definitional questions. Geberth first defined suicide by cop in 1993 as “incidents in which individuals, bent on self-destruction, engage in life-threatening and criminal behavior in order to force the police to kill them.”[191] In 1998, Hutson et al. defined suicide by cop incidents more narrowly as those where “a suicidal individual intentionally engages in life-threatening and criminal behavior with a lethal weapon or what appears to be a lethal weapon toward law enforcement officers or civilians to specifically provoke officers to shoot the suicidal individual in self-defense or to protect civilians.”[192] Subsequent researchers have added and removed elements of the definition, making it impossible to compare findings across studies.[193]

As McKenzie succinctly put it in a 2006 literature review of suicide by cop, “[c]urrent definitions are confused and confusing.”[194] The state of the literature has not improved since then. In a meta-study of eighteen suicide by cop publications from 1994–2014, Patton and Fremouw concluded the field was beset by inconsistent and imprecise definitions, leading to unreliable estimates: “Without a clear definition of what SbC is . . .[,] it will be difficult for future researchers to determine what SbC is not.”[195]

Some researchers have tried to find common threads among the conflicting definitions. As Lord and Sloop noted in 2010, the definitions include verbal or behavioral threats as perceived by police officers, and most include life-threatening behaviors or gestures.[196] Yet, the most recent study of suicide by cop noted an “interesting caveat” in the suicide by cop literature, namely the continued methodological limitations of inconsistent criteria, including some definitions that only include incidents in which police use lethal force, while others focus only on behavior directed toward police.[197]

Emblematic of the challenge, after critiquing the lack of consistent definitions in the suicide by cop literature, Jordan et al. had to use yet another definition to build a data sample from existing records coded by the LAPD’s Mental Evaluation Unit: “[S]ubjects must verbally express that they wish to be killed by police or behaviorally assert themselves in an aggressive manner in a way that would encourage SbC.”[198] This definitional confusion makes it impossible to compare and verify findings and casts doubt on whether suicide by cop is scientifically valid.

2.     Suicide by Cop Sampling Methods Are Deeply Flawed

In addition to important definitional inconsistencies, the sampling methods in suicide by cop research are seriously flawed in at least two important ways. First, studies of the phenomenon rely on nonrandom samples of police files and reports of police shooting incidents.[199] Convenience sampling of this type might be appropriate in qualitative studies to generate and even test theories, but it is not an appropriate basis for making generalizable statistical findings. For example, in the 1998 Kennedy et al. study of 240 police shootings gleaned from a search of newspapers and other unscientific sources, researchers found that 16 to 46 percent of the incidents involved suicide by cop.[200] The wide range of the almost three-fold estimate should have been reason enough for concern, but as later researchers noted, the data were not rigorously collected, the news stories included reporting biases, and the findings were not statistically reliable.[201]

These methodological flaws are not isolated—they have cascaded through later research, compounding errors over time. In their 1990 study proposing a typology of suicide by cop subjects and finding that spontaneous and opportunistic suicide is more common than premeditated suicide, Homant and Kennedy used their 1998 data set of 123 nonrandom cases, to which they added another twenty-two nonrandom cases, only compounding the sampling problem.[202] Despite the flawed underlying data, the 2009 Mohandie et al. study adopted the Homant and Kennedy typology to sort their own nonrandom sample of more than seven hundred cases of officer-involved shootings in North America.[203]

The second problem with the sampling methods of almost all studies is that they limit their inquiry to officer-involved shootings, which means that virtually everything we “know” about suicide by cop comes from this tip-of-the-iceberg subset of all police encounters with people who may be suicidal. This research choice means we know almost nothing about the much larger number of incidents involving people who are in psychological distress or suicidal that do not end in the use of lethal force. As a result of this bias, the literature tells us that suicide by cop incidents are very risky for police, which impacts police training and practices.[204] For example, California police departments have reportedly refused service calls from families of suicidal people because they are concerned about the danger to their own safety, undoubtedly informed by the literature and training that tells them so.[205]

Lord has explored the dynamics of less-than-lethal force cases involving suicidal people, and Jordan et al.’s 2019 study of LAPD’s Mental Evaluation Unit cases was meant to help correct the bias toward officer-involved shootings.[206] For example, of the 419 suicidal individuals identified in the LAPD study, the researchers found that only 3 percent were injured and 1 percent were killed.[207] Significantly, only a single police officer was injured—less than one-quarter of 1 percent of all encounters—a far cry from the earlier studies with samples heavily skewed toward officer-involved shootings that found suicide by cop incidents were highly dangerous for law enforcement.[208] A better understanding of suicide by cop, informed by incidents involving little or no use of force, would almost certainly lead to different police tactics and better outcomes.[209]

Even within research on officer-involved shootings, there are methodological weaknesses, as Mohandie et al. acknowledged in their 2009 study of more than seven hundred such cases:

[This study] includes cases where deadly force shootings usually occurred, and secondarily some cases where only less lethal force was deployed. The high loading of deadly force cases is suggestive of some degree of sampling bias towards those individuals who may have been more desperate, more intentional, and less ambivalent in their suicidal impulses. Those cases that were negotiated, resolved, or otherwise successfully intervened upon (without injury, loss of life, or deployment of deadly force) may represent another population of individuals or a different severity or kind of psychopathology.[210]

Beyond concerns regarding the sampling methods, retroactively classifying police shootings as suicide by cop requires researchers to rely on police reports, which are themselves subject to bias.[211] For example, in finding that suicide by cop was present in 36 to 41 percent of officer-involved shootings in a large, nonrandom sample from 1998–2006, Mohandie et al. concluded that their study “confirm[ed] the trend detected in earlier research that there was a growing incidence of SBC [suicide by cop].”[212] The researchers failed to control for, or even note, a highly relevant variable, namely that as early as 1998, officers were being trained to invoke suicide by cop during encounters and in police reports as a form of litigation defense.[213] This deep methodological flaw alone should give judges pause regarding the scientific validity of suicide by cop.

3.     Suicide by Cop Findings Are Not Reliable or Valid

The definitional and sampling problems yield questionable findings. The measurements in three of the oft-cited studies on suicide by cop are susceptible to inconsistencies (modest reliability) and/or inaccuracies (little or no validity). For example, the Kennedy et al. study, which found that 16 to 46 percent of officer-involved shootings were suicide by cop incidents, had modest inter-rater reliability (inconsistency across researchers applying the same criteria) and the cross-validity check (testing the method’s accuracy with different subsets of the data) was conducted using the same flawed sampling methods as in the primary study.[214] Homant and Kennedy’s study, which proposed a typology of suicide by cop incidents and established the idea that most are spontaneous, acknowledged the weakness of their sample: “To put it briefly, cross-validation is needed.”[215]

Mohandie and Meloy, whose study identified verbal, behavioral, history and life-event indicators in suicide by cop subjects, acknowledged that their variables were not subjected to any tests of reliability or validity, but this understates the methodological weaknesses.[216] They identified forty-eight indicators of suicide by cop from a handful of case studies, trade press articles, the Kennedy et al. study (mostly media sources), and an unpublished manuscript by one of the authors.[217] All eight “key life indicators” of suicide by cop—including “poor socioeconomic background,” “criminal record that includes assaultive behavior,” and “family member killed in shootout with police”—came from the unpublished manuscript and a single case study in 1993 by a retired NYPD commander.[218] The historical or situational indicator “faces an arrest or criminal justice situation perceived as serious” also comes from the unpublished manuscript, which is not publicly available. In fact, thirty-five of the forty-eight indicators come from the unpublished work and the 1993 case study.

The lack of reliability and validity has real-world consequences in suicide by cop cases. For example, in her testimony that Cammerin Boyd was attempting to commit suicide by cop, Dr. Keram relied on several of the Mohandie and Meloy indicators, which in turn served as the basis for admitting prejudicial hindsight evidence.[219] The idea that most suicide by cop incidents are spontaneous is also based on theories not subject to even the most basic, much less rigorous, scientific methods.

4.     Suicide by Cop Studies Largely Ignore Suicidal Intent

Suicide, whether by one’s own hand or through the police, requires intent on the part of the person who commits the suicidal act.[220] Mental health practitioners struggle to diagnose the suicidal intent of their living patients.[221] Relying on professionals who do not treat patients to diagnose suicidal intent posthumously, as expert witnesses do in court, is especially fraught.

Some suicide by cop researchers acknowledge that their work is inadequate to directly assess suicidal intent and that, outside of clearcut cases, reliably determining suicidal intent where the individual is already deceased is inherently difficult.[222] Mohandie and Meloy caution that no meaningful studies exist to help determine the extent to which suicide by cop may be a factor in a particular case, and emphasize that their indicators should not be used to draw conclusions about the probabilities of suicide by cop.[223] As McKenzie noted in his literature review, the studies are rife with definitional criteria that are inadequate or unclear for determining suicidal intent, making it impossible to “properly allow a conclusion that suicide is/was an intended outcome.”[224]

In a small number of suicide by cop cases described in the literature, victims of police shootings left notes mentioning or referencing their intent. For example, the Hutson et al. review of 437 officer-involved shootings by the Los Angeles Sheriff’s Department identified a single person in the forty-six alleged suicide by cop incidents who left behind a note (2.2 percent).[225] Of the 291 people Mohandie et al. identified as suicide by cop cases, thirty-three (11.3 percent) “left what could be characterized as a generic suicide note,” and four (1.3 percent) left a note saying they were seeking to die at the hands of police.[226] Even combined, the percentage of general and specific notes is half the rate found in research about suicide, where people leave a note approximately 25 percent of the time, raising further questions about the findings of prevalence of suicide by cop.[227]

D.    Suicide by Cop in Boyd’s Shadow

The holding in Boyd has been consequential in other Ninth Circuit cases, including Asa Sullivan’s. In 2014, more than eight years after Asa’s killing, the family finally got their day in court. Relying on the 2009 decision in Boyd, the court in Asa’s case admitted Dr. Keram’s expert testimony and other evidence of suicide by cop.[228] And, like in Boyd, the suicide by cop evidence was central to the defense’s argument that Asa’s desire to die at the hands of the police made the officers’ testimony more credible, resulting in the jury reaching the conclusion that the officers acted reasonably.[229]

In a pretrial report, Dr. Keram stated that Asa’s “demographic factors” (childhood records, employment records, police records, mental health records) showed he was at high risk of spontaneous suicide by cop.[230] She recounted that Asa spent time in foster care when he was a toddler while his mother stayed in women’s shelters, and she discussed evidence showing they were evicted when he was growing up.[231] Dr. Keram referenced phone records with social workers and depositions prepared for trial to opine that Asa had issues with his family or other loved ones, describing his relationship with his mother as “chaotic” and suggesting that he lacked meaningful support and reasons to live.[232]

At trial, Dr. Keram testified that Asa’s risk factors for suicide included his impulsivity, juvenile criminal history, criminal history, drug and alcohol history, and employment history.[233] On cross-examination, she opined that Asa would rather die than go to jail on a bench warrant for failing to complete an alternative work program related to a marijuana possession charge.[234] Presumably referring to the 2009 Mohandie et al. study of nonrandom officer-involved shootings, Dr. Keram testified that of the three types of suicide by cop, “about 80 percent are what are called ‘spontaneous suicides by cop’” where the person “develops their suicidal ideation and suicide by cop plans spontaneously on the scene when law enforcement arrives or during the course of a law enforcement contact.”[235]

Beyond Asa Sullivan’s case, Boyd has been cited hundreds of times in the Ninth Circuit since 2009, including more than one hundred times in Section 1983 cases.[236] Courts cite Boyd with regard to the relevance and admissibility of evidence that may support one version of events over another, especially in disputes involving an officer’s perception of events just before using force. More than three-quarters of the district court cases that cite to Boyd are in California and twenty-six of them involve suicide by cop.[237]

So far, Boyd’s holding has been geographically limited. Outside the Ninth Circuit, more than a dozen federal district courts have cited Boyd, mostly in Section 1983 cases, but none in suicide by cop cases. Only one federal appellate court outside the Ninth Circuit has cited Boyd regarding the relevance and admissibility of evidence to resolve a dispute over an officer’s version of events. In Cordova v. City of Albuquerque, the Tenth Circuit relied on Boyd for the proposition that admitting evidence of events prior to a police shooting was not an abuse of discretion, because “[c]ontextual evidence can be admitted to help explain later events, especially where it might make one version of events more or less likely.”[238]

The number of Section 1983 cases involving suicide by cop and citing Boyd is still relatively small, and it is difficult to generalize across cases with different facts, parties, and judges, but a few noteworthy trends appear to be emerging.

First, courts have taken a variety of approaches to applying Boyd in light of Graham, and some judges have shown skepticism about its broadest implications. While acknowledging that Boyd remains precedential in the Ninth Circuit, for example, a federal judge in Arizona noted that courts have “disputed the scope of [Boyd’s] ruling.”[239] The court cited inconsistencies among other district courts in the Ninth Circuit when admitting hindsight evidence under the Graham and Boyd standards.[240] More recently, a federal judge in Northern California said that the Ninth Circuit itself may be retreating somewhat from Boyd, citing cases in which the court held that Graham’s objectively reasonable standard still limits the admissibility of hindsight evidence.[241] The judge referenced numerous district court cases that have “limited [Boyd] to cases where a suicide by cop theory is being asserted or other similar circumstances where the suspect/defendant’s intent is clearly at issue.”[242]

Second, published cases post-Boyd suggest that the decision to admit suicide by cop evidence correlates with case outcomes. It is important to note that it is very difficult to control for other variables that may explain these results, but plaintiffs appear less likely to prevail in Section 1983 excessive force lawsuits when suicide by cop evidence is admitted.[243] Conversely, when suicide by cop evidence and testimony is excluded or is substantially limited, plaintiffs have a better chance to win or at least to settle their cases for damages.[244]

Given appellate courts’ deference to trial courts on evidentiary matters, the Boyd family faced long odds before the Ninth Circuit.[245] Boyd’s unfaithful reading of Graham to admit hindsight evidence, its circular reasoning about the indicators and diagnosis of suicide by cop, and its endorsement of the theory as valid science opened the door to increased use of suicide by cop in cases like Asa Sullivan’s and others.[246] Part V recommends steps that police, mental health professionals, courts, and legislators should take to limit the misuse of suicide by cop as junk science that makes bad law.

V. Ending the Misuse of Suicide by Cop

In 2000, James Fyfe, a leading expert on police use of force, wrote about tragic outcomes in street-level encounters between police and people with mental illness in alleged suicide by cop incidents:

After the fact, police have recently been prone to write off such tragedies as “suicide by cop,” a classification that, in my experience, is far more often a post hoc justification for sloppy police work than a valid explanation of why and how somebody died. The term “suicide by cop” should describe only situations in which even officers who adhere closely to the industry standard for dealing with [people with mental illness] are given no choice but to kill them. Unfortunately, it has become a catchy descriptor for a far larger number of cases in which officers put themselves unnecessarily into harm’s way and must then shoot their way out of it.[247]

When Fyfe wrote this, suicide by cop was purportedly a growing phenomenon at the intersection of policing and mental health.[248] The theory subsequently found its way into the courtroom, where lawyers have invoked it in defense of their clients and judges have admitted testimony about it as a forensic science.

This Part calls on police, doctors, and judges to curb the continued use of suicide by cop as a reflexive “post-hoc justification for sloppy police work.”[249] In encounters with people who are mentally ill, including people who may be suicidal, police officers should be required to exercise a higher duty of care, not given a green light to use deadly force. The mental health profession has a responsibility to its members, the people they serve, and the broader public to disentangle the tragedy of genuine cases of suicide from the development of suicide by cop as an exculpatory rationale for police use of lethal force. Judges should proceed with extreme caution before admitting expert testimony or other evidence about suicide by cop given its junk science attributes. Finally, lawmakers should step in to limit the use of suicide by cop specifically and police force more generally in ways that go beyond the minimum protections afforded to the public by the Fourth Amendment.

A.     Police

As noted in the introduction, one in ten police encounters involves people with serious mental illness, including one in six use of force incidents and one in every four or five fatal police shootings.[250] In recent years, police, mental health professionals, and lawmakers have paid greater attention to the violent and often deadly results of standard law enforcement responses to people experiencing mental health crises. While better training and tactics might mitigate the worst outcomes, more fundamental reform is required.

First, police should not be trained on suicide by cop as a threat in and of itself, nor as a shield against liability, because such training can make encounters between officers and the public even more likely and more deadly. Second, police should be trained on de-escalation techniques that reduce the risk of harm to everyone involved in such incidents, including officers. Third, and most importantly, police should not be the first or only responders in mental health crises, nor should they assume a lead role if they operate in teams with other professionals who are better equipped to resolve incidents with little or no violence.

Law enforcement trainings should not include misleading content about the nature and scope of suicide by cop.[251] Current trainings are based on a flawed literature that emphasizes frequent and extraordinary threats to police officers from people attempting to commit suicide by cop, which may lead to more excessive use of force. As Fyfe noted almost twenty-five years ago, fear-based training makes encounters with people with mental illness more violent, warning that:

[S]ome police training on this subject may actually be worse than none. Any police organization or government officials or medical professionals concerned with seeing that police do their work with the least violence necessary should not content themselves with knowing that officers are being trained to interact with [people with mental illness]; they must carefully examine such training to assure that it is not sending the wrong message.[252]

Fyfe proposed an approach to such encounters—like the one with Asa Sullivan—that would save lives while minimizing risk to public safety and reducing the danger to everyone involved, including police officers. He recommended that when faced with a person possibly suffering from mental illness, officers should keep a safe distance, centralize communication with the person so they are not receiving conflicting messages, avoid unnecessary displays of force and threats, and take as much time as needed “even if this runs into hours or days.”[253]

Basic de-escalation approaches like this have existed for decades, and there are more tools now than ever. In 2023, the U.S. Department of Justice and the Department of Health & Human Services issued “Guidance for Emergency Responses to People with Behavioral Health or Other Disabilities,” which focuses on improving training and methods for working with people with mental illness to reduce the unnecessary use of force.[254] The guidance recommends, among other things, the development and deployment of mobile crisis teams, coresponder teams, crisis intervention training, and de-escalation best practices.[255] In California, POST recently released “A First Responder’s Guide for Persons with Mental Illness or Developmental Disability” with a short section on suicide by cop and a summary of key cases.[256]

Training reform, however, will not be a panacea.[257] When it is obvious to law enforcement that people are mentally ill, including people who may be seeking to die at the hands of police, officers should be required to exercise a higher duty of care. State tort laws are one source of such a duty. For example, although police officers in California do not owe a general duty of care to the public, their actions do have to meet the reasonableness standard under state tort law, which is broader than the Fourth Amendment reasonableness standard. In Hayes v. County of San Diego, at the request of the Ninth Circuit, the California Supreme Court clarified that police officers owe a duty of care to the public for their conduct and decisions prior to a shooting.[258]

More fundamentally, activists, advocates, and academics have called for other reforms, including barring police from using lethal force in defined contexts and reducing reliance on police as the first or only responders, especially to incidents involving mental health crises.[259] The U.S. Department of Justice recently filed a statement of interest in a federal lawsuit in the District of Columbia arguing that the Americans with Disabilities Act might require public entities to dispatch “a different type of response to an emergency call when necessary to avoid discrimination based on disability. . . . For example, deploying mobile crisis teams staffed with behavioral health professionals may be a reasonable modification.”[260] Although its future is in doubt under the Trump administration, the American Rescue Plan of 2021 included Medicaid funding for mobile crisis intervention services, which more than twenty states have taken up.[261]

Many U.S. cities have developed promising programs to support nonpolice responses to mental health calls. Chicago’s Treatment Not Trauma uses public funds to address crisis intervention with a three-pronged approach that includes nonpolice crisis response systems, community mental health centers and crisis triage facilities, and mental health delivery led by community care workers.[262] Denver’s Support Team Assistance Response (STAR) program provides mobile crisis response for community members experiencing issues related to mental health, depression, poverty, homelessness, or substance abuse issues, with positive results.[263] Portland and San Francisco have created mental health teams that have avoided the need for police involvement in more than 95 percent of all program responses.[264]

In one setting from which we have data about the resolution of potential suicide by cop incidents, Jordan et al. found promising results in cases handled by the LAPD’s Mental Evaluation Unit (MEU) in which officers “are specially trained in crisis evaluation, negotiation, and mental illness and serve to consult with the patrol officers on calls involving mental health issues.”[265] In more than four in five cases of suspected suicide by cop handled by the MEU, LAPD officers did not use any force, resolving the matters through dialogue (341/419 cases).[266] Most of the other cases involved officers using tasers, hands-on (physical) force, beanbag shotguns, and/or pepper spray (71/419).[267] Fewer than 2 percent of the incidents (7/419) involved lethal force, and the vast majority of suicidal people in the encounters were hospitalized instead of being arrested or shot.[268]

A less police-centric approach to mental health incidents can also be good for law enforcement. The LAPD MEU study found that only a single officer suffered injuries (1/419), representing less than two-tenths of one percent of all police encounters with people with mentally illness.[269] In addition, reducing use of force incidents would mitigate the exposure of officers, departments, and municipalities to lawsuits. Collaborations between medical professionals and law enforcement raise hard questions about professional responsibility and independence, including whose norms are prioritized in mental health incidents outside the care setting.[270] But any progress on this issue will require decentering the role of police in crisis response and raising their duty of care to people with mental illness.

B.     Mental Health Professionals

In a recent article about police liability and psychiatry, Dr. Kenneth Weiss argues that notwithstanding the complexities and challenges of suicide by cop, psychiatrists should continue to play a role in the investigation and litigation of alleged incidents, including as expert witnesses.[271] At the same time, relative to their participation in litigation, he argues that there is “a clearer and perhaps more compelling role for mental health participation in response teams and policy,” citing examples where mental health evaluation teams intervened to deescalate crises and prevent use of force by officers.[272] Beyond this “compelling role” to help reduce police use of force, the mental health profession has a further obligation to help police, courts, and the public better understand the nature and scope of suicide by cop.

While their participation in mental health crisis teams is promising on the front end, the mental health professions should call on their members to collaborate cautiously, if at all, in the aftermath of use of force incidents where police allege suicide by cop. In particular, they should emphasize that suicide by cop, especially so-called “spontaneous” suicide by cop, cannot be diagnosed with any meaningful degree of certainty.[273] Because expert testimony about suicide by cop is used in part as a Trojan horse to introduce other highly prejudicial evidence of alleged suicidality that would otherwise be inadmissible, the profession should also educate its members about the full ramifications of testifying in court. Mental health professionals should not contribute to the further stigmatization of people with mental illness in any setting, including in a court of law.

In Boyd, the Ninth Circuit said that the family’s challenge to Dr. Keram’s testimony failed in part because their lawyer presented no evidence in the form of refuting experts or literature demonstrating that her theories were unreliable and no alternative methodologies and conclusions of their own.[274] Members of mental health professions can develop a critical literature and a panel of experts who can provide a scientific counterweight to the law enforcement bent of the existing literature and witnesses. The profession should challenge the junk science aspects of suicide by cop and its application in psychological autopsies in a sustained and rigorous way.

The American Psychological Association’s (APA) Code of Conduct says that psychologists should provide statements or conclusions about people’s psychological characteristics “only after they have conducted an examination” sufficient to support their opinions.[275] The APA code permits record review in scenarios when an examination is not practical.[276] However, the code also requires psychologists to clarify the limits of a record review on the reliability and validity of their opinions and circumscribe their conclusions accordingly.[277] Noting the profession’s core principles of beneficence and nonmaleficence—promoting welfare and preventing harm—the APA’s Ethics Director warns that “offering a diagnosis based solely on a record review raises special considerations and unique challenges with great ethical significance.”[278]

Developments with respect to excited delirium provide one example of the profession pushing back on junk science. In 2020 and 2021, the American Medical Association and the American Psychiatric Association announced their opposition to the use of excited delirium as a medical diagnosis in the policing context.[279] In 2022, Physicians for Human Rights Watch issued a critical report on excited delirium that included recommendations to professional associations, state and local governments, the federal government, and the United Nations, including the Independent Expert Mechanism on Systemic Racism in Law Enforcement.[280] In 2023, the National Association of Medical Examiners rejected excited delirium as a cause of death,[281] and the American College of Emergency Physicians (ACEP) withdrew approval of its 2009 White Paper Report on Excited Delirium Syndrome, stating that medical, public health, and law enforcement personnel should not use the term and that ACEP members should not serve “as expert witnesses testifying in relevant civil or criminal litigation.”[282]

Mental health professionals should critically examine suicide by cop, which has many of the same troubling attributes as excited delirium. The field needs more credible research to answer pressing questions about suicide by cop and to address the definitional and methodological flaws in the literature over the last twenty-five years, including from independent experts without strong ties to the law enforcement community.[283] Professionals with a vested interest in the findings can be consulted to help define research questions and interpret preliminary findings, but reliable and valid science will likely only be generated by more rigorous and disinterested study of the phenomenon.

C.     Judges

The holding in Boyd to admit suicide by cop testimony and evidence was significant enough to warrant a write-up in the legal digest of the Journal of the American Academy of Psychiatry and Law (JAAPL).[284] After noting how the court allowed Dr. Keram’s testimony based on a combination of her expertise and data from peer-reviewed studies, the JAAPL authors noted that courts should carefully review such studies and should not admit testimony ungrounded in science.[285] Further, the authors argued that in determining admissibility, courts should engage mental health experts to analyze the underlying studies, including their methodology.[286] Finally, they argue that the case shines a light on issues like suicide by cop “that may not be sufficiently addressed by the body of knowledge and literature of general psychiatry.”[287] In other words, judges should exercise greater care in admitting evidence about suicide by cop because the quality and validity of the forensic research is questionable, and it lies outside the established science of general psychiatry.

Post-Boyd, some federal district court judges have limited its holding specifically to suicide by cop cases only or have shown deference to Graham in spite of Boyd and refused to admit any suicide by cop evidence.[288] The Ninth Circuit itself has appeared to pull back from the broadest implications of Boyd, but the court should reaffirm more clearly and directly Graham’s controlling prohibition against hindsight evidence. As long as Boyd is viewed as binding precedent, some trial court judges will continue to admit evidence without conducting their own independent assessment of admissibility while others will issue conflicting rulings on this issue, creating a troubling patchwork of justice in excessive use of force cases.[289]

Finally, judges should not admit highly prejudicial evidence in the hopes of later limiting its impact through jury instructions. In Asa’s case, for example, the judge instructed the jury to consider suicide by cop evidence, including prejudicial evidence about his background and history, only when assessing damages.[290] “However,” the jury instructions continued, “when you are considering whether the Defendant Officers’ conduct was reasonable in terms of their search and/or seizure, you may only consider the information they knew at the time.”[291]

As the U.S. Supreme Court recognized more than half a century ago, limiting instructions of this kind do not protect parties from the impact of admitting highly prejudicial evidence.[292] Courts cannot expect juries to cabin negative character evidence like the expert testimony and related suicide by cop evidence admitted in Cammerin Boyd’s case and Asa Sullivan’s case. In fact, in a civil rights case decided just a couple of years after Asa’s case, the Ninth Circuit held that district courts must bifurcate the liability phase of the trial from the compensatory damages phase where inflammatory evidence about the person killed by the police was not known to the officer at the time of incident and has no relevance to whether he acted lawfully.[293]

D.    Lawmakers

Regardless of how the professions respond to the junk science aspects of suicide by cop and the bad law that enables its use in the courtroom, state legislators should intervene. Lawmakers can curb the misuse of suicide by cop in the same way they have begun to restrict the use of excited delirium. They can also clarify and strengthen laws governing police use of force that go beyond the minimum Fourth Amendment floor protecting the public from unreasonable seizures and harm. Recent developments in California are illustrative on both counts.

In 2023, the California Legislature passed Assembly Bill 360 to end the use of excited delirium by law enforcement, courts, and medical professionals.[294] Though police can still describe the characteristics of an individual’s conduct and condition, the new California law bars police officers from using the term “excited delirium” in an incident report.[295] The new law also prohibits public and private medical professionals from using excited delirium as a valid medical diagnosis or cause of death.[296] Finally, with respect to courts, the new law states that “evidence that a person suffered or experienced excited delirium is inadmissible in any civil action.”[297]

More broadly, state legislatures can adopt legal standards designed to reduce police use of force except where absolutely necessary. In 2019, the California Legislature passed Assembly Bill 392, which codified the requirement in Hayes v. County of San Diego that extended police officers’ duty of reasonable care to preshooting conduct. The focus on preshooting conduct addresses the very scenario that James Fyfe identified almost twenty years earlier—and a central issue in Asa’s case where officers said they could not wait for a canine unit because they could not retreat safely from the attic—about police unnecessarily putting themselves in danger and then shooting their way out of it. Taking into account such “officer-created jeopardy” can inform inquiries into whether the later use of force was reasonable.[298]

Legislation, too, will have its limits. Osagie Obasogie and Zachary Newman have raised important questions about the extent to which the law can curb excessive use of force.[299] Building on the work of Lauren Edelman, Obasogie and Newman show how courts defer to police departments and officers in defining what is reasonable conduct.[300] As a case in point, some law enforcement and medical professionals in California have already begun developing workarounds to the ban on the use of excited delirium by redefining the concept and maintaining their prerogatives to respond to events as they see fit.[301] While legislative action can send courts an important message about the contested nature of the science underlying expert testimony and evidence about excited delirium, all system actors will need to be persuaded that change is in their interests.

Conclusion

In rare instances, people may attempt to induce police officers to kill them, and not all incidents in which people wish to die at the hands of law enforcement will be resolved peacefully.

This Article is not about whether suicide by cop ever happens, but about how system actors have strategically manipulated and deployed the concept to shift the blame for excessive use of police force to victims, shield police officers from accountability, and undermine civil rights laws. Police, mental health professionals, courts, and lawmakers have contributed to the problem, and all must play a role in addressing it.

Mental health professionals and courts have enabled this strategy by endorsing junk science and making bad law, lending suicide by cop both credence and effect. In this relaxed scientific and evidentiary context, the mutually reinforcing dynamic introduces highly prejudicial evidence about the victims of police violence, increasing the likelihood that police conduct will go unpunished, no matter how unreasonable. Medical and legal professionals, therefore, have a special obligation to exercise their judgment independent of law enforcement and to fulfill their respective duties to care for patients and do justice for litigants.

Although the data about the phenomenon are woefully inadequate on this point, the emergence of suicide by cop also reveals the enduring scourge of racism in policing, medicine, law, and society, which is a key driver of both junk science and bad law. With respect to junk science, Osagie Obasogie, the foremost legal scholar on excited delirium, has noted that race and racism have been foundational to psychiatry and to doctors’ understanding of mental illness. He traces how Blackness has historically been pathologized through racist psychiatric theories from the antebellum South, through Jim Crow, to modern justifications for large disparities in mass criminalization and incarceration, to theories such as excited delirium.[302]

With respect to bad law, the proffered suicide by cop evidence, trial transcripts, and appellate court reasoning in Boyd are also replete with tropes about Black criminality that have pervaded the legal system for centuries. The suicide by cop evidence admitted at trial was intended to demonstrate that Mr. Boyd was less than human, including explicit racially charged evidence such as the rap lyrics found in his car several weeks prior to his killing. At oral argument in the Ninth Circuit, even the judge wondered aloud about the demonizing nature of some of the suicide by cop evidence:

[T]he concern here is that the improper or undue prejudice comes from painting Boyd as being a crazy man, yes, but a crazy man who does bad things and we’re all better off with him dead anyway, and that part’s not supposed to come in. How do you justify the balance here of letting in, in particular the rap lyrics is probably the most extreme example, but stuff that just tries to characterize Boyd as being an evil person?[303]

Nevertheless, like the trial court judge, the appeals court judges ultimately allowed evidence supporting this dehumanizing narrative and made bad law in the process, which later shielded police officers from any accountability in cases like Asa Sullivan’s.

There was no reason for police to kill Asa Sullivan in an attic and for his family to lose a son, a brother, and a father. He did not invite the attention of law enforcement, he had committed no crime, he was unarmed, and he posed no threat to officers or the public. While a lawsuit would not have brought Asa back, his family sought a measure of justice through the only legal mechanism available to them. The trial and appellate courts denied them justice by relying on junk science and bad law to absolve officers and the city of liability. In genuine cases of suicide by cop, police should exercise a higher duty of care to protect the public and reduce the use of excessive force.

Epilogue

We will never know what really happened that night my brother was murdered, although there still stands a need for accountability and the truth.

– Kahlil Sullivan[304]

Copyright © 2025 Jeffrey Selbin, Chancellor’s Clinical Professor of Law, U.C. Berkeley School of Law. Thanks to Berkeley Law Policy Advocacy Clinic students Ana Henderson-Arjona, Molly Coyne, and Murtaza Husain for their foundational research and memo on this topic as part of a police accountability project, Meredith Huang and Peyton Provenzano for their research assistance, Berkeley Law research librarians for finding court documents and obscure trade literature, and Berkeley Law Clinical Program paralegal Lauren Havey for editorial assistance. I received invaluable input from colleagues at the AALS Clinical Section Annual Meeting, American University Faculty Workshop, Berkeley Law Faculty Retreat, Decarceration Law Professors Workshop, Ellmann Clinical Theory Workshop, Law & Society Association Annual Conference, and Poverty Law Conference. I received generous feedback from Roxanna Altholz, Ty Alper, Chesa Boudin, Laurel Fletcher, Daniel Harawa, Erin Murphy, Osagie Obasogie, Jenny Roberts, Andrea Roth, Elisabeth Semel, Bob Solomon, and Drs. Harvey Weinstein and Howard Zonana. This piece is dedicated to the memory of Asa Sullivan and to his family, who shared his story and fought for justice so that his death was not in vain.

           [1].     Kahlil Sullivan, In that Attic, I Saw My Brother’s Blood Covering the Floor and Walls, S.F. Bay View Nat’l Black Newspaper (June 10, 2009), http://sfbayview.com/2009/06/in-that-attic-i-saw-my-brother’s-blood-covering-the-floor-and-walls/ [https://perma.cc/3YYU-AMJX].

           [2].     Espinosa v. City & County of San Francisco, 598 F.3d 528, 532 (9th Cir. 2010).

           [3].     Annie Laure Paradise, Militarized Policing and Resistance in the Social Factory: The Battle for Community Safety in the Silicon Valley 318 (2015) (Ph.D. dissertation, California Institute of Integral Studies) (http://wechargegenocide.org/wp-content/uploads/2016/07/paradisethesis.pdf) [https://perma.cc/739B-AZP3].

           [4].     Espinosa, 598 F.3d at 532.

           [5].     Id.

           [6].     Paradise, supra note 3, at 270.

           [7].     Id. at 274.

           [8].     Espinosa, 598 F.3d at 533.

           [9].     Id.

         [10].     Id. at 540 (Wu, J., concurring in part and dissenting in part).

         [11].     Id. at 540–41.

         [12].     Id. at 541.

         [13].     Id.

         [14].     Id.

         [15].     California Police Radio Codes, Police Codes, https://police-codes.com/united-states/california [https://perma.cc/3CHM-3DT8].

         [16].     Espinosa, 598 F.3d at 538.

         [17].     Id.

         [18].     Police Officers Identified in Fatal Shooting, S.F. Exam’r (June 9, 2006), https://www.sfexaminer.com/news/police-officers-identified-in-fatal-shooting/article_359f8bbd-0b2c-5109-a396-f2278a10de7d.html [https://perma.cc/HZN6-WDUE].

         [19].     Espinosa, 598 F.3d at 537 (balancing the government’s interest against an individual’s Fourth Amendment rights in determining whether an officer’s conduct was objectively reasonable based on the totality of the circumstances under Graham v. Connor, 490 U.S. 386, 396–97 (1989)); id. at 544 (Wu, J., concurring in part and dissenting in part) (same).

         [20].     Id. at 538 (majority opinion).

         [21].     Asa Sullivan was the fifteenth person the SFPD had killed in the last five years. Will Jarrett, Molly Oleson, Christine Delianne & David Mamaril Horowitz, Explore: Every Person Killed in SF Police Shootings Since 2000, Mission Loc. (Oct. 4, 2022), https://missionlocal.org/2022/10/police-shooting-deaths-san-francisco-2000-database/ [https://perma.cc/Q49B-HCH8].

         [22].     Espinosa, 598 F.3d at 531–32.

         [23].     Id.

         [24].     Id.

         [25].     See infra Part IV.

         [26].     Excerpt of Proceedings - Testimony of Emily Keram, M.D. at 122, Espinosa v. City & County of San Francisco, No. C 06-04686 (N.D. Cal. Oct. 20, 2014), Dkt. No. 443.

         [27].     Id. at 35.

         [28].     Special Verdict Form at 2–3, Espinosa, No. C 06-04686, Dkt. No. 437.

         [29].     Henry K. Lee, Family of Man Killed by S.F. Police Loses Lawsuit, S.F. Chron. (Oct. 9, 2014), https://www.sfchronicle.com/bayarea/article/Family-of-man-killed-by-S-F-police-loses-lawsuit-5810235.php [https://perma.cc/4X5N-XGGF].

         [30].     Telephone Interview with Ben Nisenbaum, Partner, Burris Nisenbaum Curry & Lacy (Sep. 13, 2024).

         [31].     Learn More About Measures LL and S1, City of Oakland (Oct. 7, 2021), https://www.oaklandca.gov/resources/learn-more-about-measure-ll [https://perma.cc/GVL3-8NRN].

         [32].     Id. In 2020, Oakland voters passed a second measure to expand the scope and authority of the Commission and the Agency. Oakland, Cal., Oakland City Charter § 604 (2025).

         [33].     The students were connected to Asa’s mother by the Center for Convivial Research and Autonomy, a local grassroots collective that conducts independent investigations of police violence. The collective centers families and survivors in its investigations to challenge the official version of state violence, to shine a light on over-policing in low-income communities of color, and to build community power. At the time, the collective was conducting a people’s investigation on behalf of Asa’s family.

         [34].     See infra Part I.

         [35].     See infra Part II.

         [36].     Boyd v. City & County of San Francisco, 576 F.3d 938, 945–46 (9th Cir. 2009).

         [37].     See infra Part IV.

         [38].     Police Shootings Database, Wash. Post (Dec. 31, 2024), https://www.washingtonpost.com/graphics/investigations/police-shootings-database/ [https://perma.cc/67A2-BPW9]; Steven Rich, Tim Arango & Nicholas Bogel-Burroughs, Since George Floyd’s Murder, Police Killings Keep Rising, Not Falling, N.Y. Times (May 24, 2025), https://www.nytimes.com/2025/05/24/us/police-killings-george-floyd.html?smid=nytcore-ios-share&referringSource=articleShare [https://perma.cc/QSS3-2DNG].

         [39].     Mapping Police Violence, https://mappingpoliceviolence.us/ [https://perma.cc/G7CT-PKZE].

         [40].     Id. See generally Paul Butler, Chokehold: Policing Black Men (2017) (documenting systematic racism in the use of police force in the United States).

         [41].     GBD 2019 Police Violence U.S. Subnational Collaborators, Fatal Police Violence by Race and State in the USA, 1980–2019: A Network Meta-Regression, 398 Lancet 1239, 1247 (2021).

         [42].     Dep’t of Just. & Dep’t of Health & Hum. Servs., Guidance for Emergency Responses to People with Behavioral Health or Other Disabilities 2 (2023); see also Amy C. Watson & Anjali J. Fulambarker, The Crisis Intervention Team Model of Police Response to Mental Health Crises: A Primer for Mental Health Practitioners, 8 Best Pracs. Mental Health 71, 72 (2012) (police calls). See generally Ayobami Laniyonu & Phillip Atiba Goff, Measuring Disparities in Police Use of Force and Injury Among Persons with Serious Mental Illness, BMC Psychiatry (Oct. 12, 2021), https://bmcpsychiatry.biomedcentral.com/articles/10.1186/s12888-021-03510-w [https://perma.cc/ZZX4-ECYZ] (use of force); Kimberly Kindy, Julie Tate, Jennifer Jenkins & Ted Mellnik, Fatal Police Shootings of Mentally Ill People Are 39 Percent More Likely to Take Place in Small and Midsized Areas, Wash. Post (Oct. 17, 2020), https://www.washingtonpost.com/national/police-mentally-ill-deaths/2020/10/17/8dd5bcf6-0245-11eb-b7ed-141dd88560ea_story.html [https://perma.cc/8WTU-FSWL] (fatal police shootings).

         [43].     GBD 2019 Police Violence U.S. Subnational Collaborators, supra note 41, at 1247, 1250; Maanvi Singh, How America’s Broken Autopsy System Can Mask Police Violence, Guardian (July 2, 2020), https://www.theguardian.com/us-news/2020/jul/02/autopsies-police-killings-medical-misleading [https://perma.cc/LRX6-5LJK]. In California, state law permits counties to have either medical examiners or coroners whose duties include investigating and certifying cause of death. Cal. Gov’t Code § 24010. Medical examiners are appointed licensed physicians with training in pathology, while coroners are elected or appointed officials, typically not physicians.

         [44].     Ann Crawford-Roberts, Sonya Shadravan, Jennifer Tsai, Nicolás E. Barceló, Allie Gips, Michael Mensah, Nichole Roxas, Alina Kung, Anna Darby, Naya Misa, Isabella Morton & Alice Shen, Opinion, George Floyd’s Autopsy and the Structural Gaslighting of America, Sci. Am. (June 6, 2020), https://blogs.scientificamerican.com/voices/george-floyds-autopsy-and-the-structural-gaslighting-of-america/ [https://perma.cc/5HGR-WMA3].

         [45].     More fundamentally, reform advocates have been pushing to reduce reliance on police as first responders in some settings. Alex S. Vitale, 10 Ways to Reduce Our Reliance on Policing and Make Our Communities Safer for Everyone, Appeal (July 20, 2020), https://theappeal.org/10-ways-to-reduce-our-reliance-on-policing-and-make-our-communities-safer-for-everyone/ [https://perma.cc/LAN8-4FVU]. Abolitionists call for an end to policing as we know it. See, e.g., Mariame Kaba & Andrea J. Ritchie, No More Police: A Case for Abolition (2022) (finding from large nationally representative data sets that police who commit misconduct

are rarely removed or even disciplined).

         [46].     For a recent analysis of the literature on this issue, see generally Max Schanzenbach, Policing the Police: Personnel Management and Police Misconduct, 75 Vand. L. Rev. 1523 (2022); Barbara E. Armacost, Organizational Culture and Police Misconduct, 72 Geo. Wash. L. Rev. 453 (2003) (describing the less official ways in which police culture and socialization send conflicting messages about issues like excessive use of force to street-level officers).

         [47].     Joanna Schwartz, Shielded: How the Police Became Untouchable xii (2023). Mapping Police Violence (MPV), a project that tracks office prosecution in cases where someone was killed by the police, estimates that between 2022 and 2024, only 2.97 percent of killings resulted in officer prosecution. MPV relies on news media for data collection and noted that estimates for 2024 may not fully reflect investigations that are ongoing and have not yet been reported upon. Campaign Zero Live Tracker, Mapping Police Violence, https://mappingpoliceviolence.org/ [https://perma.cc/NK6K-WTDP].

         [48].     42 U.S.C. § 1983. Though beyond the scope of this Article, plaintiffs can also bring state court tort and wrongful death claims in cases of alleged excessive use of force, with evidentiary standards that can differ from Section 1983 cases and vary among the states. Timothy P. Flynn & Robert J. Homant, “Suicide by Police” in Section 1983 Suits: Relevance of Police Tactics, 77 U. Det. Mercy L. Rev. 555, 571, 573 (2000).

         [49].     Erwin Chemerinsky, Rights Need Remedies, in Presumed Guilty: How the Supreme Court Empowered The Police and Subverted Civil Rights 131 (2021).

         [50].     See Monroe v. Pape, 365 U.S. 167 (1961), overruled on other grounds by Monell v. Dep’t of Soc. Servs. of N.Y., 436 U.S. 658 (1978).

         [51].     See generally Schwartz, supra note 47 (providing a comprehensive treatment of the development of laws and cases protecting police officers and municipalities from liability).

         [52].     Evans v. Jeff D., 475 U.S. 717, 730–31 (1986).

         [53].     Ashcroft v. Iqbal, 556 U.S. 662, 678 (2009) (heightening requirements for pleading plausible facts from antitrust cases in Bell Atlantic Corp. v. Twombly, 550 U.S. 544 (2007), to all federal cases).

         [54].     Harlow v. Fitzgerald, 457 U.S. 800, 815–18 (1982) (shielding officers who engage in excessive force from liability unless a federal court in their jurisdiction previously held that a similar action violated the Constitution). In particular, since the murder of George Floyd, there have been increased calls for state and federal steps to reform or end qualified immunity. See, e.g., George Floyd Justice in Policing Act of 2021, H.R. 1280, 117th Cong. (2021) (would have limited qualified immunity, but it stalled after passing the House). In 2002, the Montana Supreme Court held that police officers do not have qualified immunity as a defense in state courts, and since Floyd’s murder, the Nevada Supreme Court followed suit in 2022 and six states and localities have also limited the defense legislatively (Colorado, 2020; Massachusetts, 2020; New Mexico, 2021; California, 2021; Connecticut, 2021; New York City, 2021). Qualified Immunity State Reforms, Inst. for Just., https://ij.org/qualified-immunity-state-reforms/ [https://perma.cc/N8MH-U7GM]; Press Release, N.Y.C. Council, Council Votes to End Qualified Immunity and Seven Other Measures to Reform NYPD (Mar. 25, 2021), https://council.nyc.gov/press/2021/03/25/2079/ [https://perma.cc/QLV2-W8AD].

         [55].     See Tennessee v. Garner, 471 U.S. 1, 7 (1985) (using deadly force on unarmed fleeing persons suspected of a crime constitutes a seizure that federal courts must subject to a Fourth Amendment reasonableness analysis); Graham v. Connor, 490 U.S. 386, 388 (1989) (extending Garner to apply a Fourth Amendment reasonableness standard to all excessive force claims arising out of an arrest or investigatory stop, not only those that involve deadly force on unarmed fleeing persons). See generally Erwin Chemerinsky, Eroding Remedies for Police Misconduct, in Presumed Guilty, supra note 49 (documenting the role the U.S. Supreme Court has played in limiting police misconduct by eroding the exclusionary rule and officer and municipal liability); Erwin Chemerinsky, The Vanishing Remedies for Police Misconduct, in Presumed Guilty, supra note 49 (same).

         [56].     See, e.g., Schwartz, supra note 47, at 83 (finding that officers pay 0.02 percent of plaintiff’s awards).

         [57].     Jennifer Valentino-DeVries, Mike McIntire, Rebecca R. Ruiz, Julie Tate & Michael H. Keller, How Paid Experts Help Exonerate Police After Deaths in Custody, N.Y. Times (Dec. 26, 2021), https://www.nytimes.com/2021/12/26/us/police-deaths-in-custody-blame.html [https://perma.cc/WF2M-PYHV]. See generally Ingrid V. Eagly & Joanna C. Schwartz, Lexipol: The Privatization of Police Policymaking, 96 Tex. L. Rev. 891 (2018) (describing the proliferation of privatized police policymaking and the role of Lexipol LLC in creating internal regulations for law enforcement agencies across the United States).

         [58].     Valentino-DeVries et al., supra note 57.

         [59].     See generally Osagie K. Obasogie, Excited Delirium and Police Use of Force, 107 Va. L. Rev. 1545 (2021) (documenting the development by legal and medical actors of excited delirium as a pseudoscientific cause of death in cases of police use of force) [hereinafter Obasogie, Police Use of Force]; Osagie K. Obasogie, Excited Delirium, Policing, and the Law of Evidence, 138 Harv. L. Rev. 1497 (2025) (tracing the medical and scientific origins and critiques of excited delirium) [hereinafter Obasogie, Law of Evidence]; Brianna da Silva Bhatia, Michele Heisler, Joanna Naples-Mitchell & Julia Sherwin, Physicians for Hum. Rts., “Excited Delirium” and Deaths in Police Custody (2022), https://phr.org/wp-content/uploads/2022/03/PHR-Excited-Delirium-Report-March-2022.pdf [https://perma.cc/8FHZ-W572] (identifying the small cohort of research and defense experts who developed and spread the idea of excited delirium).

         [60].     See, e.g., Michael A. Brave & John G. Peters, Jr., Preparing for “Sudden Death!,” LAAW Int’l, Inc. (1994), https://web.archive.org/web/19970410001343/http://laaw.com:80/cdsldm.htm [https://perma.cc/PY7W-K2NP] (law enforcement risk management professionals describing “custody death syndrome”).

         [61].     Tom Lyden, Excited Delirium Dilemma: Explanation or Excuse for In-Custody Deaths?, Fox 9 KMSP (Nov. 15, 2020), https://www.fox9.com/news/excited-delirium-dilemma-explanation-or-excuse-for-in-custody-deaths [https://perma.cc/AQ47-XR7W].

         [62].     Obasogie, Police Use of Force, supra note 59, at 1595.

         [63].     See Obasogie, Law of Evidence, supra note 59.

         [64].     Méabh O’Hare, Joshua Budhu & Altaf Saadi, Police Keep Using “Excited Delirium” to Justify Brutality. It’s Junk Science, Wash. Post (July 17, 2020), https://www.washingtonpost.com/outlook/chokehold-police-excited-delirium/2020/07/17/fe907ec8-c6bc-11ea-b037-f9711f89ee46_story.html [https://perma.cc/F4QE-4RFA].

         [65].     M. Chris Fabricant, Junk Science and the American Criminal Justice System 41 (2022). Fabricant traces the modern rise of junk forensic science to People v. Marx, 126 Cal. Rptr. 350 (Ct. App. 1975), where the court allowed expert opinion on “bite marks” despite acknowledging there was no established science to do so. Id. at 38–39.

         [66].     Id. For the misuse of junk science in wrongful convictions, see Brandon L. Garrett, Convicting the Innocent: Where Criminal Prosecutions Go Wrong 84–117 (2011). For the problem in criminal sentencing, where admissibility standards to limit junk science do not apply, see generally Maneka Sinha, Junk Science at Sentencing, 89 Geo. Wash. L. Rev. 52 (2021).

         [67].     See infra Part IV.B.

         [68].     See, e.g., Alan Feuer, The Nation: Desperadoes; Drawing a Bead on a Baffling Endgame: Suicide by Cop, N.Y. Times (June 21, 1998), https://www.nytimes.com/1998/06/21/weekinreview/the-nation-desperadoes-drawing-a-bead-on-a-baffling-endgame-suicide-by-cop.html [https://perma.cc/9GHS-RFLT] (noting that Harris had a Ph.D. in psychology and had worked at a Los Angeles suicide hot line after leaving the police force).

         [69].     Richard N. Jenet & Robert J. Segal, Provoked Shooting by Police as a Mechanism for Suicide, 6 Am. J. Forensic Med. & Pathology 274, 274 (1985).

         [70].     Dean Scoville, Getting You to Pull the Trigger, Police (Oct. 31, 1998), https://www.policemag.com/print/content/15350070 [https://perma.cc/38JD-W7JE]; Michael G. Wargo, First Response to a Hostage/Barricade/Suicide Incident, in Patrol Response to Contemporary Problems: Enhancing Performance of First Responders Through Knowledge and Experience 165, 169 (John A. Kolman ed., 2006).

         [71].     Nationally, the Blue Line Lawyer Institute, dedicated to supporting and defending police officers, hosts an annual Force Forum, which it describes as “a training conference for public safety professionals and attorneys who represent them. . . . [It] is for those seeking best practices and cutting-edge research on investigating and defending use of force.” About the Force Forum, Blue Line Law. Inst., https://bluelinelawyer.com/events/ [https://perma.cc/9YEC-Q5D8]. The institute’s 2023 conference included extensive sessions on defending police in suicide by cop encounters. (Notes in possession of the author from an attendee). Calibre Press, Inc. is another private law enforcement training organization that offers hundreds of courses each year, including on suicide by cop. Calibre Press Training Network, https://www.calibrepress.com [https://perma.cc/5V73-7R3G]; Suicide by Cop: 15 Warning Signs that You Might be Involved, Calibre Press Training Network (Aug. 21, 2019), https://calibrepress.com/2019/08/suicide-by-cop-15-warning-signs-that-you-might-be-involved/ [https://perma.cc/3CHX-52KT]; Lessons Learned After Two Suicide By Cop Shootings, Calibre Press Training Network (Nov. 16, 2022), https://calibrepress.com/2022/11/officers-lessons-learned-after-two-suicide-by-cop-shootings/ [https://perma.cc/Q3EU-BWDA]. For a critique of Calibre Press trainings on the threat to officers and how such trainings can increase violent encounters, see James J. Fyfe, Policing the Emotionally Disturbed, 28 J. Am. Acad. Psychiatry L. 345, 346 (2000).

         [72].     See generally, e.g., David Farber, The War on Drugs (2021) (documenting how the war on drugs beginning in the 1970s contributed to mass incarceration, including undermining medical expertise by fueling moral panics not grounded in science); Daniel Yohanna, Deinstitutionalization of People with Mental Illness: Causes and Consequences, 15 Virtual Mentor 886 (2013) (recounting the history of the deinstitutionalization of people with mental illness during the second half of the twentieth century in the United States).

         [73].     David A. Fishbain & Charles V. Wetli, Cocaine Intoxication, Delirium, and Death in a Body Packer, 10 Ann. Emerg. Med. 531 (1981); Charles V. Wetli & David A. Fishbain, Cocaine-Induced Psychosis and Sudden Death in Recreational Cocaine Users, 30 J. Forensic Sci. 873 (1985).

         [74].     Isaac Scher, “Suicide by Cop”: How Police Present Killings as Unavoidable, Mother Jones (2023), https://www.motherjones.com/politics/2022/12/suicide-by-cop-how-police-present-killings-as-unavoidable/ [https://perma.cc/SN9Q-CBF2]; see also George L. Kelling & James Q. Wilson, Broken Windows: The Police and Neighborhood Safety, Atlantic (Mar. 1982) https://www.theatlantic.com/magazine/archive/1982/03/broken-windows/304465 [https://perma.cc/37RQ-2FYJ].

         [75].     A California newspaper first used the term suicide by cop in 1991. Clark Brooks, Suicide by Cop, S.D. Union-Trib., Aug. 26, 1991, at C1.

         [76].     See, e.g., Vernon Geberth, Suicide-by-Cop: Inviting Death from the Hands of a Police Officer, Law & Ord., July 1993, at 105 (“There has not been an empirical study of these events, but historically there have been many justifiable homicide events resulting from the imminent use of deadly physical force against police. . . . [T]here is a growing number of justifiable homicides in which police officers have shot and killed an apparently armed individual, who has threatened the officer or others with the immediate use of deadly physical force. The subsequent homicide investigation, assessment and review of the circumstances reveal many of these events to be victim precipitated.”).

         [77].     Id.; Anthony J. Pinizzotto & Edward F. Davis, Suicide by Cop: Implications for Law Enforcement Management, Law & Ord., Dec. 1999, at 95; Jim Weiss & Mary Dresser, Reaching Out to the Mentally Ill, Law & Ord., June 2000; George T. Williams, Death by Indifference, Law & Ord., Dec. 2003, at 66.

         [78].     Clinton R. Van Zandt, “Suicide by Cop,” Police Chief, July 1993, at 24; Richard B. Parent, Suicide by Cop: Victim-Precipitated Homicide, Police Chief, Oct. 1998, at 111; Audrey L. Honig, Police Assisted Suicide: Identification, Intervention, and Investigation, Police Chief, Oct. 2001, at 89; Bruce D. Praet, Suicide by Cop or Death by Indifference?, Police Chief, July 2002, at 14.

         [79].     Mark T. Civita, The Beat: Suicide by Cop, Police (Aug. 31, 1998), https://www.policemag.com/patrol/article/15350077/the-beat-suicide-by-cop [https://perma.cc/34MA-BWQK]; Scoville, supra note 70; R.K. Campbell, Don’t Go with Them, Police (Apr. 30, 2005), https://www.policemag.com/patrol/article/15349391/dont-go-with-them [https://perma.cc/7DF5-2A4W]; Dean Scoville, Suicide by Cop, Police (Apr. 29, 2010), https://www.policemag.com/patrol/article/15348452/suicide-by-cop [https://perma.cc/7CNX-MC8G].

         [80].     Jeff Martin, Suicide by Cop: What the Responding Officer Needs to Know, Police Marksman, Mar./Apr. 2003, at 31.

         [81].     Pamela Kulbarsh, The Untold Motives Behind Suicide-by-Cop, Officer.com (Apr. 8, 2015), https://www.officer.com/training-careers/article/12062592/the-untold-motives-behind-suicidebycop [https://perma.cc/BJ67-9LDC].

         [82].     Calibre Press, Investigating Suicide by Cop Incidents: What to Remember & What to Get, Police1 (Mar. 27, 2009), https://www.police1.com/investigations/articles/investigating-suicide-by-cop-incidents-what-to-remember-what-to-get-E5wn16W9meEM12Mb/ [https://perma.cc/Y762-RCMG].

         [83].     Suicide by Cop, 2007 AELE Monthly L.J. 101.

         [84].     California Commission on Peace Officer Standards and Training (POST), Suicide by Cop (1998) (notes on file with author) [hereinafter POST, Suicide by Cop]. Clinic students viewed these videos at the POST office in Sacramento, but the videos have since been deleted per the office’s records retention schedule and the office was unable to provide a copy upon request. 1999 POST Suicide by Cop training materials are referred to in other sources cited here. See, e.g., Kris Mohandie & J. Reid Meloy, Clinical and Forensic Indicators of “Suicide by Cop,” 45 J. Forensic Sci. 384 (2000).

         [85].     See generally Radley Balko, Rise of the Warrior Cop: The Militarization of America’s Police Forces (2013) (documenting the growing military training, tactics, equipment, and mentality in policing since the 1960s).

         [86].     POST, Suicide by Cop, supra note 84.

         [87].     Campbell, supra note 79.

         [88].     POST, Suicide by Cop, supra note 84; Scoville, supra note 79.

         [89].     POST, Suicide by Cop, supra note 84.

         [90].     At the same time, part of the training message is also not to worry too much about using deadly force. Ronald M. McCarthy, Litigation and Surviving Lawsuits and Prosecution, in Patrol Response to Contemporary Problems: Enhancing Performance of First Responders Through Knowledge and Experience 217, 221 (2006) (noting that “[t]he research tells us that rarely is an officer who uses deadly force guilty of a crime. And, almost always, the officer does not violate agency policy or rules”).

         [91].     The POST Telecourse on suicide by cop explicitly teaches police officers how to avoid litigation, recommending that officers stay at the hospital with suicide by cop victims, check in with the family over time, and let the family know that police are there to help. POST, Suicide by Cop, supra note 84.

         [92].     Honig, supra note 78, at 93.

         [93].     Id.

         [94].     Id.

         [95].     Audrey L. Honig, Suicide by Cop: Reducing Personal and Organizational Liability Through Investigation, 40 J. Cal. L. Enf’t, 5, 6 (2006).

         [96].     Id.

         [97].     Id. at 7.

         [98].     Id. at 8.

         [99].     Rick Wall Consulting & Investigations, Suicide by Cop: Overview for Managers: Extended Outline 1 (July 1, 2022) (detailing a training course provided for the California Commission on Peace Officer Standards and Training received by the author from POST in response to a Public Records Act request).

       [100].     Paradise, supra note 3, at 310–11.

       [101].     Id. at 311–13. Organizations in Los Angeles have documented law enforcement harassment of the families who lost loved ones to police violence. See generally Rebecca Brown, Maggie Gaffney & Lindsey V. Thompson, No Justice, No Peace: The Los Angeles County Sheriff’s Department’s Targeted Harassment of Grieving Families (2021), https://www.aclusocal.org/sites/default/files/final_lasd_family_harassment_report.pdf [https://perma.cc/N5YJ-V4YC].

       [102].     Martin E. Wolfgang, Victim Precipitated Criminal Homicide, 48 J. Crim L., Criminology, & Police Sci. 1, 2 (1957); see also R.K. Wright & J.H. Davis, Studies in the Epidemiology of Murder, 22 J. Forensic Sci. 464, 466 (1977) (finding approximately 10 percent of the homicides over a several decade period in Dade County, Florida were “Victim Participated,” involving any participation by the victim in the murder); Richard C. Harruff, Amy L. Llewellyn, Michael A. Clark, Dean A. Hawley & John E. Pless, Firearm Suicides During Confrontations with Police, 39 J. Forensic Sci. 402, 403 (1994) (finding approximately 1 percent of more than 1,200 suicides in Marion County, Indiana over an eight-year period involved a police suspect turning a firearm on themselves).

       [103].     For a comprehensive review of the literature at the time, see generally John M. Violanti & James J. Drylie, Research on Suicide by Cop (SbC), in Copicide: Concepts, Cases, and Controversies of Suicide by Cop 15 (2008).

       [104].     For a discussion of the serious methodological flaws in the suicide by cop literature, see infra Part IV.C.

       [105].     H. Range Hutson, Deirdre Anglin, John Yarbrough, Kimberly Hardaway, Marie Russell, Jared Strote, Michael Canter & Bennett Blum, Suicide by Cop, 32 Ann. Emerg. Med. 665, 667 (1998).

       [106].     Id.

       [107].     Id. at 667.

       [108].     Daniel B. Kennedy, Robert J. Homant & R. Thomas Hupp, Suicide by Cop, 67 FBI L. Enf’t Bull. 21, 21 (1998).

       [109].     Id. at 23–24.

       [110].     Id. at 23.

       [111].     Id. at 24–25.

       [112].     Vivian B. Lord, Law Enforcement-Assisted Suicide, 27 Crim. Just. & Behav. 401, 408 (2000).

       [113].     Id. at 401; Vivian B. Lord, Curriculum Vitae (2010), https://pages.charlotte.edu/vivian-lord/wp-content/uploads/sites/331/2012/10/CVvbl610.pdf [https://perma.cc/37BF-6TCL]; History of the Justice Academy, N.C. Just. Acad., https://ncdoj.gov/ncja/about-the-justice-academy/history-of-the-justice-academy/ [https://perma.cc/W9UF-4D83].

       [114].     Robert J. Homant, Daniel B. Kennedy & R. Thomas Hupp, Real and Perceived Danger in Police Officer Assisted Suicide, 28 J. Crim. Just. 43, 47 (2000).

       [115].     Id. at 46–47.

       [116].     See Robert J. Homant & Daniel B. Kennedy, Suicide by Police: A Proposed Typology of Law Enforcement Officer-Assisted Suicide, 23 Policing 339 (2000).

       [117].     Id. at 342.

       [118].     Id. at 344–48.

       [119].     See Mohandie & Meloy, supra note 84.

       [120].     Id. at 388–89.

       [121].     David Best, Anna Quigley & Alan Bailey, Police Shooting as a Method of Self-Harming: A Review of the Evidence for ‘Suicide by Cop’ in England and Wales between 1998 and 2001, 32 Int’l J. Socio. L. 349, 352 (2004).

       [122].     Kris Mohandie, J. Reid Meloy & Peter I. Collins, Suicide by Cop Among Officer-Involved Shooting Cases, 54 J. Forensic Sci. 456, 457 (2009).

       [123].     Id. at 461.

       [124].     Id.

       [125].     Id. at 459.

       [126].     Alejandra Jordan, Nancy R. Panza & Charles Dempsey, Suicide by Cop: A New Perspective on an Old Phenomenon, 23 Police Q. 82, 87–88 (2020).

       [127].     The studies also do not shed much light on gender. The percentage of women deemed to meet the inclusion criteria for suicide by cop incidents ranged from 2–5 percent on the low end, Hutson et al., supra note 105, at 667; see Kennedy, Homant & Hupp, supra note 108, at 24; Mohandie, Meloy & Collins, supra note 122, at 458, to 11–17 percent on the high end, Homant, Kennedy & Hupp, supra note 114, at 48–49; Jordan, Panza & Dempsey, supra note 126, at 90.

       [128].     Mohandie, Meloy & Collins, supra note 122, at 458; Hutson et al., supra note 105, at 667; Jordan, Panza & Dempsey, supra note 126, at 90.

       [129].     Mohandie, Meloy & Collins, supra note 122, at 458 (16.41 percent of the study sample in the United States where 12.97 percent of the population was Black in 2000).

       [130].     Jordan, Panza & Dempsey, supra note 126, at 90 (22.91 percent of the 2019 study sample in Los Angeles County where 7.93 percent of the population was Black in 2020); see also Jesse McKinnon, The Black Population: 2000 3 (2001) https://www2.census.gov/library/publications/decennial/2000/briefs/c2kbr01-05.pdf [https://perma.cc/74LU-H5AS]; Racial/Ethnic Composition Los Angeles County, 1990-2020 Census, L.A. Almanac, https://www.laalmanac.com/population/po13.php [https://perma.cc/ZZV3-LQ6J]. These racial disparities are mirrored in the context of excited delirium, where more than half of people who died in police custody from alleged excited delirium were Black and Latinx. Obasogie, Police Use of Force, supra note 59, at 1595.

       [131].     42 U.S.C. § 1983.

       [132].     See, e.g., Jud. Council of Cal., Civil Jury Instructions § 3020 (2025).

       [133].     Tennessee v. Garner, 471 U.S. 1, 7 (1985) (“While it is not always clear just when minimal police interference becomes a seizure . . . there can be no question that apprehension by the use of deadly force is a seizure subject to the reasonableness requirement of the Fourth Amendment.” (citation omitted)).

       [134].     490 U.S. 386, 387–88, 396 (1989) (establishing the objective reasonableness standard and “a careful balancing of the ‘nature and the quality of the intrusion on the individual’s Fourth Amendment interests’ against the countervailing government interests at stake” (quoting Garner, 471 U.S. at 8)).

       [135].     576 F.3d 938, 942–43 (9th Cir. 2009).

       [136].     Id.

       [137].     Id. at 944.

       [138].     Id. at 943–44; Fed. R. Evid. 401, 402.

       [139].     Boyd, 576 F.3d at 946; Fed. R. Evid. 702; Daubert v. Merrell Dow Pharms., Inc., 509 U.S. 579, 597 (1993).

       [140].     Boyd, 576 F.3d at 947; Fed. R. Evid. 404(b).

       [141].     Boyd, 576 F.3d at 943–45; Fed. R. Evid. 403, 703.

       [142].     Boyd, 576 F.3d at 942.

       [143].     Id.

       [144].     Id.

       [145].     Id. at 943.

       [146].     Id. at 944. Other witnesses testified that, when commanded to put his hands up and get on the ground, Mr. Boyd put his hands up. Id. at 942.

       [147].     Id. at 944.

       [148].     Graham v. Connor, 490 U.S. 386, 396 (1989).

       [149].     Transcript of Proceedings at 4427–28, Boyd v. City & County of San Francisco, C 04-5459 (N.D. Cal. Sep. 25, 2007), Dkt. No. 427 [hereinafter Keram Trial Testimony].

       [150].     Id. at 4429, 4431–32.

       [151].     Id. at 4429, 4518–19.

       [152].     Id. at 4429, 4433–35. Noting that “a substantial body of research shows a significant risk of unfair prejudice when rap lyrics are introduced into evidence,” in 2022, the California Legislature revised the Evidence Code to limit the admission of rap lyrics into evidence. Cal. Evid. Code § 352.2 (West 2025).

       [153].     Keram Trial Testimony, supra note 149, at 4444.

       [154].     Id. at 4429, 4446.

       [155].     Boyd v. City & County of San Francisco, 576 F.3d 938, 949–50 (9th Cir. 2009).

       [156].     Id. at 948–49.

       [157].     Id. at 949–50.

       [158].     Id. at 944 (citation omitted).

       [159].     Graham v. Connor, 490 U.S. 386, 399 n.12 (1989).

       [160].     The court also relied on another Ninth Circuit case that allowed the admission of witness reports and the decedent’s blood alcohol level to analyze reasonableness in an excessive force case. The admitted evidence in Billington, however, was contemporaneous to the event, and it did not involve a suicide by cop theory that relied on character evidence. See Billington v. Smith, 292 F.3d 1177 (9th Cir. 2002).

       [161].     See Boyd, 576 F.3d at 944; see also Palmquist v. Selvik, 111 F.3d 1332, 1341 (7th Cir. 1997).

       [162].     Boyd, 576 F.3d at 944; Palmquist, 111 F.3d at 1341.

       [163].     Palmquist, 111 F.3d at 1339 (“Sherrod states the law in this circuit: when considering a charge of excessive force under the Fourth Amendment, evidence outside the time frame of the shooting is irrelevant and prejudicial.” (citing Sherrod v. Berry, 856 F.2d 802, 803 (7th Cir. 1988))).

       [164].     Boyd, 576 F.3d at 947.

       [165].     Id. at 947.

       [166].     Id. at 947–48.

       [167].     Fed. R. Evid. 404(a)(1).

       [168].     Fed. R. Evid. 404(b)(1).

       [169].     Fed. R. Evid. 404(b)(2).

       [170].     Fed. R. Evid. 403.

       [171].     Boyd, 576 F.3d at 948.

       [172].     Id.

       [173].     Id. at 944.

       [174].     Keram Trial Testimony, supra note 149, at 4512–13.

       [175].     A version of this circular reasoning was raised by the family’s lawyer on cross-examination of Dr. Keram at the Daubert hearing. Transcript of Proceedings at 99–100, Boyd v. City & County of San Francisco, C 04-5459 (N.D. Cal. Sep. 25, 2007), Dkt. No. 468. The trial court judge later wondered aloud about the tautology—“it kind of reads bad”—but it did not prevent her from admitting the suicide by cop testimony and related evidence. Id. at 99.

       [176].     Keram Trial Testimony, supra note 149, at 4500–01.

       [177].     Id.

       [178].     Id. at 4456–57.

       [179].     Id. Under cross examination, Dr. Keram did not seem aware that Mr. Boyd had a pending personal injury lawsuit against the Oakland Police Department from which he would not be able to recover if he died. Id. at 4476.

       [180].     Fed. R. Evid. 702. See Boyd v. City & County of San Francisco, 576 F.3d 938 (9th Cir. 2009).

       [181].     Fed. R. Evid. 702; Boyd, 576 F.3d at 945–46.

       [182].     Fabricant, supra note 65, at 42.

       [183].     Fed. R. Evid. 702. For a history of the evolution of the rules of admissibility of expert testimony in the analogous context of excited delirium, see Obasogie, Law of Evidence, supra note 59, at 1515–26.

       [184].     Fed. R. Evid. 702.

       [185].     Id.

       [186].     509 U.S. 579, 592–94 (1993); see also Kumho Tire Co. v. Carmichael, 526 U.S. 137, 152 (1999) (holding that it is the role of the courts to “make certain that an expert, whether basing testimony upon professional studies or personal experience, employs in the courtroom the same level of intellectual rigor that characterizes the practice of an expert in the relevant field”).

       [187].     Boyd v. City & County of San Francisco, 576 F.3d 938, 943 (9th Cir. 2009); see also Gen. Elec. Co. v. Joiner, 522 U.S. 136, 139 (1997) (holding abuse of discretion as the standard for appellate review of trial courts’ Daubert decisions); Kumho Tire, 526 U.S. at 147 (extending Daubert to all expert testimony, not just from scientists).

       [188].     Boyd, 576 F.3d at 945–46; Daubert, 509 U.S. at 592–94.

       [189].     Boyd, 576 F.3d at 946.

       [190].     Dragana Kesic, Stuart D.M. Thomas & James R.P. Ogloff, Analysis of Fatal Police Shootings: Time, Space, and Suicide by Police, 39 Crim. Just. & Behav. 1107, 1109 (2012).

       [191].     Geberth, supra note 76, at 105.

       [192].     Hutson et al., supra note 105, at 665. The FBI uses the Hutson et al. definition in its Uniform Crime Reporting Program’s Law Enforcement Officers Killed and Assaulted data collection. Fed. Bureau of Investigation, U.S. Dep’t of Just., Law Enforcement Officers Killed and Assaulted (LEOKA) Data Collection: Definitions and Examples 28 (2019), https://ucr.fbi.gov/leoka/2019/resource-pages/definitions.pdf [https://perma.cc/8725-WY2S].

       [193].     See, e.g., Mohandie, Meloy & Collins, supra note 122, at 456 (quoting a 1999 California Peace Officer Association’s Standards and Training that defines suicide by cop as when a subject “engages in behavior which poses an apparent risk of serious injury or death, with the intent to precipitate the use of deadly force by law enforcement personnel toward the individual”); Homant & Kennedy, supra note 116, at 342 (“In short, in order to classify an incident as suicide by cop, we required that the subject behaved in a way that seemed intended to provoke the police to shoot. Specifically, this required either deliberately exposing oneself while posing a threat to police or bystanders or, in a standoff situation, knowingly forcing police to attack while harming or threatening to harm hostages or bystanders.”).

       [194].     Ian K. McKenzie, Forcing the Police to Open Fire: A Cross-Cultural/International Examination of Police-Involved, Victim-Provoked Shootings, 6 J. Police Crisis Negots. 5, 8 (2006) (including a methodological critique of almost every aspect of the existing literature).

       [195].     Christina L. Patton & William J. Fremouw, Examining “Suicide by Cop”: A Critical Review of the Literature, 27 Aggression & Violent Behav. 107, 118 (2016).

       [196].     Vivian B. Lord & Michael W. Sloop, Suicide by Cop: Police Shooting as a Method of Self-Harming, 38 J. Crim. Just. 889, 891 (2010).

       [197].     Jordan, Panza & Dempsey, supra note 126, at 85.

       [198].     Id. at 88.

       [199].     Hutson et al., supra note 105, at 665.

       [200].     Kennedy et al., supra note 108, at 24–25.

       [201].     Mohandie, Meloy & Collins, supra note 122, at 456; Jordan, Panza & Dempsey, supra note 126, at 83.

       [202].     See Homant & Kennedy, supra note 116, at 341.

       [203].     Mohandie, Meloy & Collins, supra note 122, at 457.

       [204].     Homant et al., supra note 114, at 50 (arguing the study showed that “suicide by cop situations are unpredictably dangerous and require at least the same caution as any other type of police intervention involving potentially violent persons”).

       [205].     Anita Chabria, Police Fear ‘Suicide by Cop’ Cases. So They’ve Stopped Responding to Some Calls, L.A. Times (Aug. 10, 2019), https://www.latimes.com/california/story/2019-08-09/suicide-calls-california-cops-stopped-responding [https://perma.cc/P89S-HQV7].

       [206].     See Jordan, Panza & Dempsey, supra note 126, at 85. See generally Vivian B. Lord, Police Responses in Officer-Involved Violent Deaths: Comparison of Suicide by Cop and Non-Suicide by Cop Incidents, 17 Police Q. 79 (2014).

       [207].     Jordan, Panza & Dempsey, supra note 126, at 100 (noting that these figures “represent[] a stark contrast to prior studies, like Mohandie et al. (2009), where the rates of subject death have been reported as high as 51%, with 40% of subjects and 16% of officers injured”).

       [208].     Id.

       [209].     Id. at 85.

       [210].     Mohandie, Meloy & Collins, supra note 122, at 462.

       [211].     The 1998 Hutson et al. study, which is otherwise one of the more rigorous methodologically, qualifies its findings by noting that “as with any retrospective study analyzing previously compiled information, inherent biases may exist.” Hutson et al., supra note 105, at 669.

       [212].     Mohandie, Meloy & Collins, supra note 122, at 456.

       [213].     See supra Part II.

       [214].     Kennedy et al., supra note 108, at 23 (the inter-rater agreements were a modest 74 percent).

       [215].     Homant & Kennedy, supra note 116, at 350, 353 (the inter-rater agreements were a modest 78 percent).

       [216].     Mohandie & Meloy, supra note 84, at 389.

       [217].     Id. at 388–89. The relevant indicators are in Tables 4, 5, 6, and 7. Id.

       [218].     Id. at 389; Geberth, supra note 76; Kris Mohandie & Michael Albanese, Advanced Crisis Negotiations (1998) (unpublished manuscript). There is no evidence that the 1998 unpublished manuscript upon which so much of the 2000 Mohandie and Meloy article relies to establish indicators of suicide by cop was ever published and therefore subject to scrutiny with respect to reliability, validity, or any other measure of empirical quality.

       [219].     See Testimony of Emily Keram, M.D., supra note 26, at 62, 69, 80–81.

       [220].     Am. Psychiatric Ass’n, Diagnostic and Statistical Manual of Mental Disorders (DSM-5) 830 (5th ed. 2013) (defining suicide as “the act of intentionally causing one’s own death”).

       [221].     See, e.g., Ronald C. Kessler, Robert M. Bossarte, Alex Luedtke, Alan M. Zaslavsky & Jose R. Zubizarreta, Suicide Prediction Models: A Critical Review of Recent Research with Recommendations for the Way Forward, 25 Molecular Psychiatry 168 (2020) (finding that suicide prediction tools in health care settings are weak and inadequate).

       [222].     See, e.g., McKenzie, supra note 194, at 9 (noting that “[t]he literature is beset with a problem of inadequate (or at least unclear) indications of the criteria used in assessing suicidal intent”); Best et al., supra note 121, at 350 (commenting that judgments regarding intention are problematic because it is not always the case that a person expressly intends to die).

       [223].     Mohandie & Meloy, supra note 84, at 389.

       [224].     McKenzie, supra note 194, at 9–10 (including a methodological critique of almost every aspect of the existing literature).

       [225].     Hutson et al., supra note 105, at 667.

       [226].     Mohandie, Meloy & Collins, supra note 122, at 459.

       [227].     Yolanda Mejías-Martín, Celia Martí-García, Yolanda Rodríguez-Mejías, Ana Alejandra Esteban-Burgos, Víctor Cruz-García & María Paz García-Caro, Understanding for Prevention: Qualitative and Quantitative Analyses of Suicide Notes and Forensic Reports, 20 Int’l J. Env’t Rsch. & Pub. Health 2281, 2282 (2023).

       [228].     Notably, the district court in Asa’s case excluded character evidence about two of the police officers’ subsequent misconduct because it ruled that the prejudicial effect outweighed its probative value. See Transcript of Proceedings at 9, Espinosa v. City & County Of San Francisco, No. C-06-4686 (N.D. Cal. Oct. 20, 2014), Dkt. No. 324. During the pendency of the lawsuit, Officer Morgado had been fired for lying in a police report—although his position was restored on appeal due to a procedural error—and Officer Alvis was indicted for stealing from the evidence locker. Will Jarrett, Some SFPD Cops Appear in Civil Court Again and Again, Mission Loc. (June 14, 2023), https://missionlocal.org/2023/06/sfpd-civil-lawsuits-settlements-police-officers-deputies-misconduct/ [https://perma.cc/X8MS-38SN]; Julian Mark, Paulo Morgado, the SFPD Officer Who Earned More than the Chief in 2018, Has Been Fired — Again, Mission Loc. (Sep. 3, 2019), https://missionlocal.org/2019/09/paulo-morgado-the-sfpd-officer-who-earned-more-than-the-chief-in-2018-has-been-fired-again/ [https://perma.cc/Q3D8-TL4P]; Officer Pleads Not Guilty in Theft, S.F. Exam’r (Apr. 16, 2008), https://www.sfexaminer.com/news/officer-pleads-not-guilty-in-theft/article_16614eee-6f81-54b6-80e6-c632e493dbfc.html [https://perma.cc/LT6R-68DM].

       [229].     See Special Verdict Form, supra note 28.

       [230].     Declaration of Emily A. Keram, M.D. in Support of Defendants’ Opposition to Plaintiffs’ Motion in Limine #1: Motion to Exclude Suicide by Cop and Testimony of Dr. Keram, Exhibit B at 14, 19, Espinosa, No. C-06-4686, Dkt. No. 299-2.

       [231].     Id. at 2, 4.

       [232].     Id. at 5, 17.

       [233].     Testimony of Emily Keram, M.D., supra note 26, at 62, 69, 80–81.

       [234].     Id. at 185–89.

       [235].     Id. at 18.

       [236].     In a Westlaw Keycite search on May 5, 2025, we found 258 cases in the Ninth Circuit and in federal district courts and state courts within the Ninth Circuit that cite to Boyd and are not noted as treating Boyd unfavorably, including 139 cases that reference Section 1983 and 26 cases that mention suicide by cop (exact phrase). Westlaw, “Boyd v. City & County of San Francisco, 576 F.3d 938, 943 (9th Cir. 2009),” 258 results (May 5, 2025) (on file with the California Law Review) (Keycite citing references search filtered by “cases,” “9th Circuit,” federal district courts: “Alaska,” “Arizona,” “Hawaii,” “Idaho,” “Montana,” “Nevada,” “Oregon,” “Washington,” “California,” state courts: “California,” “Montana,” “Nevada,” “Hide Negative”).

       [237].     Id.

       [238].     816 F.3d 645, 659 (10th Cir. 2016).

       [239].     Daily v. City of Phoenix, No. CV-14-00825-PHX-SPL, 2019 WL 5718051, at *4 (D. Ariz. Nov. 5, 2019).

       [240].     Id.

       [241].     Est. of Tindle v. Mateu, No. 18-cv-05755-YGR, 2020 WL 5760287, at *11 (N.D. Cal. Sep. 28, 2020). Since Boyd, the Ninth Circuit appears to have retreated somewhat from its holding. See Hayes v. County of San Diego, 736 F.3d 1223, 1232–33 (9th Cir. 2013) (“[W]e can only consider the circumstances of which [the officers] were aware when they employed deadly force. . . . [W]hen analyzing the objective reasonableness of the officers’ conduct under Graham, we cannot consider the fact that [the decedent] was intoxicated or that he had previously used a knife in harming himself.”); Glenn v. Washington County, 673 F.3d 864, 873 n.8 (9th Cir. 2011) (“We cannot consider evidence of which the officers were unaware—the prohibition against evaluating officers’ actions ‘with the 20/20 vision of hindsight’ cuts both ways.” (quoting Graham v. Connor, 490 U.S. 386, 396 (1989))).

       [242].     Est. of Tindle, 2020 WL 5760287, at *11 (citations omitted).

       [243].     In addition to Asa’s case, Espinosa v. City & County of San Francisco, 598 F.3d 528 (9th Cir. 2010) (admitted hindsight evidence, including mental health records, criminal records, family history, employment/financial records), see, for example, Pickard v. Holton, No. 12-cv-01489-JST, 2015 WL 576717 (N.D. Cal. Feb. 11, 2015) (admitted hindsight evidence, including criminal history, history of drug use, mental health history, prior suicide attempts, defendants prevailed).

       [244].     See, e.g., Reed v. City of Modesto, No. 1:11-CV-1083 AWI GSA, 2015 WL 1889048, at *4 (E.D. Cal. Apr. 24, 2015) (“[W]here there is no dispute of relevant facts, there is no need for an expert witness to opine on the likelihood of events in a suicide by cop situation to bolster or challenge the credibility of witnesses who provide conflicting testimonies. . . . Given these circumstances, Boyd remains inapplicable to this case.”) (plaintiffs subsequently prevailed, see generally Reed v. City of Modesto, 122 F. Supp. 3d 967 (E.D. Cal. 2015)); Dorger v. City of Napa, No. 12-cv-00440-WHO, 2013 WL 5804544 (N.D. Cal. Oct. 24, 2013) (suicide by cop expert testimony excluded, plaintiffs prevailed); Order Ruling on Evidentiary Motions Regarding Defendants’ Suicide by Cop Theory, Mueller v. Cruz, SA CV 13-01274-CJC (C.D. Cal. June 8, 2017), Dkt. 205 (suicide by cop expert testimony admitted in a very limited manner; the judge excluded some of the evidence upon which the suicide by cop theory typically relies, including mental health diagnosis, school discipline records, drug use history, toxicology screen results, case settled).

       [245].     The standard of review to overturn a trial court’s decision to admit evidence is abuse of discretion. Boyd v. City & County of San Francisco, 576 F.3d 938, 943 (9th Cir. 2009); see also Gen. Elec. Co. v. Joiner, 522 U.S. 136 (1997) (holding abuse of discretion as the standard for appellate review of trial courts’ Daubert decisions); Kumho Tire Co. v. Carmichael, 526 U.S. 137, 152 (1999) (extending Daubert to all expert testimony, not just from scientists). To meet the standard, Ninth Circuit caselaw requires the reviewing court to find that the trial court decision to admit the evidence was “beyond the pale of reasonable justification under the circumstances” and likely affected the outcome. Harman v. Apfel, 211 F.3d 1172, 1175 (9th Cir. 2000); McEuin v. Crown Equip. Corp., 328 F.3d 1028, 1032 (9th Cir. 2003).

       [246].     While outside the scope of this Article, suicide by cop has appeared in other contexts that may be troubling for families of victims of police killings. See, e.g., N. Am. Co. for Life & Health Ins. v. Caldwell, 55 F.4th 867, 871 (11th Cir. 2022) (holding that actively provoking suicide by cop is covered by the suicide exclusion in a life insurance policy, voiding its terms).

       [247].     Fyfe, supra note 71, at 346. This is an eerily apt description of Asa’s killing in 2006, where the officers testified that they couldn’t safely back out of the attic entrance and wait him out because they feared he was armed; therefore, they had no option but to shoot him. Espinosa, 598 F.3d at 549.

       [248].     See supra note 71.

       [249].     See Fyfe, supra note 71, at 346.

       [250].     See supra note 42 and accompanying text.

       [251].     For a critique of the general lack of clear police training standards in use of force cases, see generally Brandon Garrett & Seth Stoughton, A Tactical Fourth Amendment, 103 Va. L. Rev. 211 (2017).

       [252].     Fyfe, supra note 71, at 346.

       [253].     Id. at 346–47.

       [254].     See Dep’t of Just. & Dep’t of Health & Hum. Servs., Guidance for Emergency Responses to People with Behavioral Health or Other Disabilities (2023).

       [255].     Id. at 10–13.

       [256].     See Cal. Comm’n on Peace Officer Standards & Trainings, A First Responder’s Guide for Persons with Mental Illness or Developmental Disability (2023), https://post.ca.gov/Portals/0/post_docs/publications/Mental_Health.pdf [https://perma.cc/LAV4-JE4R]. Some of these recommendations have started making their way into more recent law enforcement training on suicide by cop. See, e.g., Suicide by Cop: Protocol and Training Guide, Police Exec. Rsch. F., https://www.policeforum.org/suicidebycop [https://perma.cc/HV8A-8UU6].

       [257].     Among other challenges in police training, social psychology has identified significant unconscious bias in policing at all stages of interactions with the public, including in use of force decisions. L. Song Richardson, Police Racial Violence: Lessons from Social Psychology, 83 Fordham L. Rev. 2961, 2962 (2015). See generally Seth W. Stoughton, How the Fourth Amendment Frustrates the Regulation of Police Violence, 70 Emory L.J. 521 (2020) (noting that regulatory alternatives to reduce police violence, such as state law changes and law enforcement agencies, often adopt the deeply flawed Fourth Amendment standards into their own frameworks).

       [258].     See 658 F.3d 867 (9th Cir. 2011). In Hayes, the Ninth Circuit had requested the Supreme Court’s guidance on state law standards in a federal case from San Diego in which the trial court had ruled against a family’s claim that officers were negligent in provoking a situation that led to deadly force. Id. at 868, 870. The federal trial court in San Diego, relying on two state appellate court decisions, had ruled that the officers did not owe any duty of care to the mentally ill person they killed. Id. at 870; see generally Adams v. City of Fremont, 68 Cal. Rptr. 2d 196 (Ct. App. 1998) (a suicide case where police failed to intervene); Munoz v. City of Union City, 16 Cal. Rptr. 3d 521 (Ct. App. 2004) (a police shooting case). In response to the Ninth Circuit’s request, the California Supreme Court ruled that under the state’s negligence principles, “case law has long recognized that peace officers have a duty to act reasonably when using deadly force,” including in their preshooting conduct, which “should be considered as part of the totality of circumstances surrounding the fatal shooting.” Hayes v. County of San Diego, 305 P.3d 252, 261 (Cal. 2013) (emphasis removed).

       [259].     See generally Nancy C. Marcus, From Edward to Eric Garner and Beyond: The Importance of Constitutional Limitations on Lethal Use of Force in Police Reform, 12 Duke J. Const. L. & Pub. Pol’y 53 (2016) (proposing that police training emphasize constitutional limitations on permissible lethal use of force); Vitale, supra note 45 (arguing that cities should dispatch mental health experts and social workers to intervene in mental health crises instead of police). Police abolitionists call for an end to law enforcement as we know it. See, e.g., Kaba & Ritchie, supra note 45.

       [260].     Statement of Interest of the United States at 15, Bread for the City v. District of Columbia, No. 23-01945-ACR (D.D.C. Feb. 22, 2024).

       [261].     42 U.S.C. § 1396w-6; Ctrs. for Medicare & Medicaid, Dep’t of Health & Hum. Servs., SHO # 21-008, Medicaid Guidance on the Scope of and Payments for Qualifying Community-Based Mobile Crisis Intervention Services (2021), https://www.medicaid.gov/federal-policy-guidance/downloads/sho21008.pdf [https://perma.cc/P8LQ-QZQA]. As of September 12, 2024, at least twenty states had received approval to implement the services (Alabama, Arizona, California, Colorado, District of Columbia, Indiana, Kentucky, Louisiana, Maryland, Massachusetts, Montana, Nevada, New Hampshire, New Jersey, New Mexico, New York, North Carolina, Oregon, Vermont, West Virginia, and Wisconsin). State Option to Provide Qualifying Community-Based Mobile Crisis Intervention Services, Medicaid.gov, https://www.medicaid.gov/medicaid/benefits/behavioral-health-services/state-option-provide-qualifying-community-based-mobile-crisis-intervention-services/index.html [https://perma.cc/S8BP-H79E].

       [262].     Collaborative for Cmty. Wellness, Treatment Not Trauma: A Community Care Infrastructure for Crisis Response, Mental Health, and Shared Safety 1 (2023), https://collaborativeforcommunitywellness.org/wp-content/uploads/2025/06/tnt-white-paper.pdf [https://perma.cc/E3QL-NJBP].

       [263].     See Support Team Assisted Response (STAR) Program, Denver, https://www.denvergov.org/Government/Agencies-Departments-Offices/Agencies-Departments-Offices-Directory/Public-Health-Environment/Community-Behavioral-Health/Behavioral-Health-Strategies/Support-Team-Assisted-Response-STAR-Program [https://perma.cc/QB6S-4AFZ]; Sarah Gillespie, Will Curran-Groome & Amy Rogin, Urb. Inst., Evaluating Alternative Crisis Response in Denver’s Support Team Assisted Response (STAR) Program: Interim Findings (2024), https://www.denvergov.org/files/assets/public/v/1/public-health-and-environment/documents/cbh/star/evaluating_alternative_crisis_response_in_denvers_support_team_assisted_response_program-interim_findings.pdf [https://perma.cc/FL28-QV3W].

       [264].     Collaborative for Cmty. Wellness, Treatment Not Trauma 5 (2023), https://collaborativeforcommunitywellness.org/wp-content/uploads/2025/06/tnt-white-paper.pdf [https://perma.cc/4H9M-J6EA].

       [265].     Jordan, Panza & Dempsey, supra note 126, at 88–89.

       [266].     Id. at 92.

       [267].     See id.

       [268].     Id.

       [269].     Id.

       [270].     For examples, legal scholars have also identified unresolved questions of law about the extent to which medical professionals can themselves be liable under Section 1983 for excessive use of force, for example, for administering drugs at the direction of police in cases of alleged excited delirium. Obasogie, Police Use of Force, supra note 59.

       [271].     Kenneth J. Weiss, Suicide by Cop and Civil Liability for Police, 51 J. Am. Acad. Psychiatry & L. 204, 212 (2023).

       [272].     Id.

       [273].     For a critique of the application of general scientific knowledge (“framework evidence”) to individual cases (“diagnostic evidence”) in expert testimony, see generally David L. Faigman, John Monahan & Christopher Slobogin, Group to Individual (G2i) Inference in Scientific Expert Testimony, 81 U. Chi. L. Rev. 417 (2014).

       [274].     Boyd v. City & County of San Francisco, 576 F.3d 938, 946 (9th Cir. 2009).

       [275].     Ethical Principles of Psychs. & Code of Conduct § 9.01(b) (Am. Psych. Assoc. 2016)

       [276].     Id. § 9.01(c).

       [277].     Id.

       [278].     Stephen Behnke, Diagnoses, Record Reviews and the New Ethics Code, 36 Monitor on Psych. 80 (2005), https://www.apa.org/monitor/jan05/ethics [https://perma.cc/RY42-VUNU].

       [279].     Press Release, Am. Med. Ass’n, New AMA Policy Opposes “Excited Delirium” Diagnosis (June 14, 2021), https://www.ama-assn.org/press-center/press-releases/new-ama-policy-opposes-excited-delirium-diagnosis [https://perma.cc/35Z3-HZXE]; Am. Psychiatric Ass’n, Position Statement on Concerns About Use of the Term “Excited Delirium” and Appropriate Medical Management in Out-of-Hospital Contexts (2020), https://www.psychiatry.org/getattachment/7769e617-ee6a-4a89-829f-4fc71d831ce0/Position-Use-of-Term-Excited-Delirium.pdf [https://perma.cc/J6MV-Y77F].

       [280].     da Silva Bhatia et al., supra note 59.

       [281].     Carla K. Johnson & Ryan J. Foley, Medical Examiners Group Steps Away from ‘Excited Delirium,’ Associated Press (Apr. 3, 2023), https://apnews.com/article/excited-delirium-police-custody-restraint-d75c5138fbed3c7911e0bd9bcde6c207 [https://perma.cc/V5CV-N3XX]; Nat’l Ass’n of Med. Exam’rs, Position Statement, Excited Delirium (2023), https://www.thename.org/assets/docs/PositionPaperUpdates/Excited%20Delirium%20Statement%20March%206%2C%202023.pdf [https://perma.cc/LD4Y-YAVC].

       [282].     ACEP Reaffirms Positions on Hyperactive Delirium, Am. Coll. Emergency Physicians (Oct. 12, 2023), https://www.acep.org/news/acep-newsroom-articles/aceps-position-on-hyperactive-delirium [https://perma.cc/8VAP-5UU8].

       [283].     Critics have noted the tension between the role of law enforcement and the role of scientific inquiry, including attempts by police to manipulate science to advance their interests. Paul C. Giannelli, Daubert and Forensic Science: The Pitfalls of Law Enforcement Control of Scientific Research, 2011 U. Ill. L. Rev. 53, 57.

       [284].     Omri Berger & John R. Chamberlain, Suicide by Cop Mental Health Expert Testimony Meets Daubert Standards, 39 J. Am. Acad. Psychiatry & L. 270 (2011).

       [285].     Id. at 272.

       [286].     Id.

       [287].     Id.

       [288].     See supra notes 239–242 and accompanying text.

       [289].     Scholars have critiqued the admission of expert testimony solely on the basis of judicial notice, that is, that other courts have already deemed the scientific methods at issue to meet the Daubert standards. Erin Murphy, The New Forensics: Criminal Justice, False Certainty, and the Second Generation of Scientific Evidence, 95 Calif. L. Rev. 721, 758 (2007); see also Obasogie, Law of Evidence, supra note 59, at 1523–24, 1538–42 (arguing that to avoid courts giving their blessing to unreliable methods through imitation, judges should “embrac[e] their gatekeeping function and independently assess[] the method’s merits de novo” when making decisions to admit expert testimony). But see Rahi Azizi, When Individuals Seek Death at the Hands of the Police: The Legal and Policy Implications of Suicide by Cop and Why Police Officers Should Use Nonlethal Force in Dealing with Suicidal Suspects, 41 Golden Gate U. L. Rev. 183, 201 (2011) (providing examples of how evidence of suicide by cop could be beneficial for plaintiffs in Section 1983 cases because it can garner sympathy for victims).

       [290].     Final Jury Instructions at 27, Espinosa v. City & County of San Francisco, No. C 06-04686 (N.D. Cal. Oct. 20, 2014), Dkt. No. 434.

       [291].     Id.

       [292].     See generally Bruton v. United States, 391 U.S. 123 (1968) (holding that a limiting instruction did not protect a defendant against the prejudicial impact of admitting the confession of a codefendant into evidence).

       [293].     Est. of Diaz v. City of Anaheim, 840 F.3d 592, 606 (9th Cir. 2016). According to Asa’s lawyer, the Diaz holding has subsequently been helpful in suicide by cop cases. Interview with Ben Nisenbaum, supra note 30.

       [294].     A.B. 360, 2023-2024 Leg., Reg. Sess. (Cal. 2023); Samantha Young, California Bans Controversial ‘Excited Delirium’ Diagnosis, L.A. Times (Oct. 12, 2023), https://www.latimes.com/california/story/2023-10-12/california-bans-controversial-excited-delirium-diagnosis [https://perma.cc/PYX4-26WL]. In April 2024, Colorado Governor Jared Polis signed a bill prohibiting training on and the use of the term excited delirium by law enforcement personnel, emergency medical service providers, other first responders, and coroners. H.B. 24-1103, 2024 Gen. Assemb., Reg. Sess. (Colo. 2024).

       [295].     Cal. A.B. 360.

       [296].     Id.

       [297].     Id.

       [298].     See generally Cynthia Lee, Officer-Created Jeopardy: Broadening the Time Frame for Assessing a Police Officer’s Use of Deadly Force, 89 Geo. Wash. L. Rev. 1362 (2021) (arguing that judges in criminal prosecutions of police officers for use of deadly force should allow juries to consider the preshooting conduct of the officer if it contributed to the risk of a deadly confrontation); Seth W. Stoughton, Jeffrey J. Noble & Geoffrey P. Albert, Evaluating Police Uses of Force 158 (2020).

       [299].     See generally Osagie K. Obasogie & Zachary Newman, The Endogenous Fourth Amendment: An Empirical Assessment of How Police Understandings of Excessive Force Become Constitutional Law, 104 Corn. L. Rev. 1281 (2019).

       [300].     See generally id.; Lauren B. Edelman, Working Law: Courts, Corporations, and Symbolic Civil Rights (2016).

       [301].     Obasogie, Law of Evidence, supra note 59, at 1546–51.

       [302].     Id. at 1510–11 & nn.93–97.

       [303].     Oral Argument at 17:43, Boyd v. City & County of San Francisco, 576 F.3d 938 (9th Cir. 2009) (No. 07-16993), https://www.ca9.uscourts.gov/media/audio/?20090714/07-16993/ [https://perma.cc/94MH-6R8L].

       [304].     Sullivan, supra note 1.