Click here to return to the “Suicide by Cop” Civil Liability Case Summaries



FBI Law Enforcement Bulletin

 August, 1998


Suicide by Cop


By Daniel B. Kennedy, Ph.D.

Robert J. Homant, Ph.D.

R. Thomas Hupp, M.A.


Upon hearing the term "suicide by cop," the average person would probably think of police officers who take their own lives. However, to law enforcement officers, this phrase refers to an individual who wishes to die and uses the police to effect that goal. The following case serves as an example of this phenomenon.


A terrified woman called police because her ex-boyfriend was breaking into her home. Upon arrival, police heard screams coming from the basement. They stopped halfway down the stairs and found the ex-boyfriend pointing a rifle at the floor. Officers observed a strange look on the subject's face as he slowly raised the rifle in their direction. Both officers fired their weapons, killing the suspect. The rifle was not loaded.


Assuming the incident happened as described, suicidal intent by the ex-boyfriend appeared to be the most reasonable explanation for the shooting. However, critics of police shootings may consider it far-fetched that the ex-boyfriend would threaten police with an empty gun. They may speculate that he wanted to surrender, or that "trigger-happy" police merely assumed the subject might aim the gun at them. Although the term suicide by cop has been discussed in social-science literature,1 in police training material, and in newspaper accounts of fatal shootings by police, much remains to be learned.


Questions arise concerning whether a shooting is necessary or avoidable in any police shooting even when it is attributed to suicide by cop. Analyzing such incidents by considering possible suicidal motivations would prove beneficial for police training, for police-community relations, for helping officers deal with postshooting stress, and for determining civil liability.


Previously, administrative reviews of police shootings often focused too narrowly on use-of-force issues, in the sense that they tried to determine only if police selected the appropriate level of force needed to subdue the subject. In many cases, however, it seemed the subject actively encouraged or challenged the officer to use deadly force. While the truth of such situations can seldom be known with certainty, a more systematic exploration of their frequency may shed some light on these complex situations. As a result, the authors found a need to further study the phenomenon of suicide by cop.




Police may confront shootings motivated by suicidal subjects more often than reports indicate. In one case, an adult male drove his car onto the front lawn of police headquarters in downtown Detroit. He exited his vehicle, took out a handgun, and began shooting at the building. Several police officers returned fire until they killed the subject.


In another case, Philadelphia police responded to a burglary-in- progress call at a local school. Upon arrival, the suspect fired twice at the police. A subsequent chase through the school corridors followed. A police dog eventually cornered the subject, and as the officers approached, they found the subject crouched and pointing a gun at them. Police fired, killing the subject. Police later found that the subject's gun was a starter pistol, incapable of firing live rounds. Furthermore, family members later identified the subject's voice on police tapes as the person who placed the initial burglary call to police. Finally, police learned that the subject had been hospitalized as the result of a suicide attempt.2


Other occasions exist in which the suicidal intent of a subject is clearly evident, but due to particularly patient and attentive police work, a shooting does not occur and a death is avoided. Such an example occurs in the following case.


An officer patrolling a hotel parking lot observed a man pushing a woman onto the floor of a vehicle. The woman was nude and bloody from the waist down. The officer approached the vehicle and noticed that the man's blue jeans were covered with blood. The man began walking toward him yelling profanities along with "Go ahead.... kill me." As the officer drew his weapon and pointed it at the subject, he ordered the man to the ground. The subject kept walking toward him saying, "kill me, you chicken. Shoot me in the head, kill me...." The officer backed up, trying to keep a safe distance, as the subject kept putting his hands in his pockets and behind his back. Backup officers surprised the subject from behind and subsequently subdued him.3  Each of the above cases reasonably implies that the suspects acted in such a manner to ensure that police officers would shoot them.




The concept of hidden suicide illustrates the complexities of suicide by cop. Many deaths appear to result from natural or accidental causes that might better be classified as forms of suicide. Some researchers have long suspected that many single-occupant car crashes' especially those that occur under excellent driving conditions, involve suicidal motivations.4 Some researchers also attribute certain airplane crashes,5 parachute fatalities,6 and workplace fatalities7 to suicidal motivations. Furthermore, several opponents of the death penalty argue that the existence of capital punishment may induce suicidal people to commit murder in order to die by execution.8 Similarly, individuals who suddenly experience a homicidal rage may be described as suicidal because their actions often result in their own deaths.9 This can occur either at the hands of intended victims who are defending themselves or due to the intervention of authorities. The ambiguity involved in determining whether a death is suicidal has given rise to the concept of a psychological autopsy and the investigative specialty of equivocal death analysis.10


Most relevant to the concept of suicide by cop is the phenomenon known as victim-precipitated homicide. By initiating an assault or otherwise provoking someone, suicide-prone individuals achieve their goals without losing self-esteem. Furthermore, police officers frequently have to deal with individuals who display a good deal of impulsivity and self-destructive behavior.11




In any modern democracy, the use of force by law enforcement officers is subject to a great deal of scrutiny. Even so, the lack of an effective centralized reporting system and other methodological problems have left researchers unable to determine the exact number of justifiable homicides by police each year.


In 1989, the FBI began to track the number of justifiable homicides by police. The average per year from 1988 to 1993 was 389.12 While data prior to 1988 are more problematic, experts generally agree that a significant reduction in the use of deadly force by police officers has occurred. Whether as a result of increased professionalism, more restrictive policies, or civil litigation, police officers exercise far more restraint than in the past. Nevertheless, police use of force remains the subject of intense national scrutiny.




The authors conducted research to determine whether suicide by cop constitutes some meaningful percentage of police shootings. Additionally, the authors attempted to determine whether any particular circumstances distinguished suicide by cop from other police shootings. In order to obtain a representative sample of police shootings, the authors reviewed an electronic library, containing full-text newspaper articles, to obtain a broad sample of accounts of police shootings in which potential cases of suicide by cop could be found. The electronic library contained the full text from 22 newspapers, representing 18 metropolitan areas. A keyword search using the words "police," "shoot," and "citizen" produced 887 articles from January 1980 through June 1995. By eliminating duplicates and those articles that did not specifically describe a police shooting incident, the authors found 240 articles suitable for analysis.


Two experienced police officers with master's degrees in criminal justice rated the 240 incidents independently of each another. They cataloged the 240 incidents into one of five categories. Most of their ratings closely agreed. In fact, in 74 percent of the cases, their ratings were exactly the same.


Although no way exists to prove that a particular incident definitely involved suicidal impulse, these five categories and the news stories that represent the cases help to illustrate the validity of the rating process.


1) Probable Suicide: The subjects show clear suicidal motivation, either by word or gesture or they confront the police with a dangerous weapon despite having no way to escape, virtually forcing the officers to shoot.


This category illustrates itself in the case of a Philadelphia man who brandished a gun and threatened to take his own life inside a police department. When officers tried to convince the man to drop the gun, he started saying, "Shoot me, shoot me." A police officer shot him, after the subject barricaded himself and pointed his gun at the officer."


2) Possible Suicide: Subjects appear disturbed or otherwise act as if they do not care whether officers kill them; they may make a futile or hopeless escape attempt.14


The following case depicts this category. A man had an argument with his mother and sister, and they threw him out of the house. Five hours later he got into a confrontation with Ocoee, Florida, police. Officers first found the 33-year-old construction worker sleeping in his car. The subject ignored orders to exit the vehicle.


As an officer radioed for backup, the man climbed into the driver's seat, started the car, and accelerated directly toward another officer. The subject ran over one officer's foot, then accelerated toward a second officer. The police shot and killed the subject.


3) Uncertain: Either too little contradictory information is given. Subjects may or may not have had some suicidal motivation. This category could be represented by a robbery suspect who gets shot after turning toward police officers with a weapon.15


4) Suicide Improbable: Subjects' behavior give no overt indication of suicidal or self-destructive intent, and the behavior can easily be accounted for without assuming such motivation. The possibility of underlying suicidal intent cannot be ruled out.


This concept presents itself, in the shooting death of a man taking part in a drug deal foiled by police. When the police confronted the group of men, one pulled a pistol and aimed at the officers, forcing the officers to fire.16


5) No Suicidal Evidence: Subjects clearly attempt to avoid being shot. If the situation involves an attempt to flee, a reasonable hope of success should appear.17


In one incident, the police shot a man in the leg after he pulled a knife and tried to drag an officer from an unmarked police car. While struggling with the robber, the officer managed to draw and fire his weapon. The police spokesman described the shooting as straightforward; the person robbing the officer was armed with a knife.18







The news articles occasionally included demographics for the civilians involved. The person was typically male (97 percent) and between the ages of 16 and 35 (68 percent). Homelessness or mental illness was identified in 14 cases (5 percent). In addition, the incidents usually involved uniformed, on-duty police officers. Occasionally, off-duty officers (13 percent) and plainclothes officers (12 percent) were involved. Additionally, a few cases involved narcotics officers, SWAT team members, members of special surveillance teams, detectives, and some special task force officers.


Suicidal Motivation


The authors found evidence of probable or possible suicidal motivation in 16 percent of the 240 incidents. They classified the vast majority of the incidents as indeterminate, due to the lack of pertinent information in the new articles and a lack of follow-up articles. It is uncertain how many of these indeterminate articles might have indicated suicidal motivation if the articles provided more details. Of the 80 incidents that provided enough detail to classify, a surprisingly large 46 percent contained some evidence of probable or possible suicidal motivation.


The researchers used three variables to assess each of the 240 incidents. Those three variables included: lethality, circumstances, and precipitating events.




With regard to lethality, 69 percent resulted in fatalities, 17 percent proved nonfatal, and in 14 percent of the cases, the outcome was unclear." Nonfatalities showed less suicidal motivation, but more evidence is needed to confirm this.




In categorizing the 240 incidents according to the crime category or reason for the officer's intervention, at least 34 types of situations occurred. Some examples of those situations include impounding an animal, investigating a prowler, responding to complaints about loud music, and responding to an armed robbery -- the single most common category.


As expected, researchers found a slight trend for suicidal incidents to involve the cluster of a general disturbance, domestic disturbance, and person with a weapon calls. A high number of suicidal incidents, however, also stemmed from armed robberies, and many of the nonsuicidal shootings began with traffic stops. Speculation holds that armed robbery often signifies a desperate crime in which offenders, while much preferring to get away, would rather be killed than captured.


Precipitating Event


Researchers identified numerous different reasons why police officers fired their weapons. Pointing or firing a gun at an officer represented the most common precipitating event by far. A cluster of events that can be construed as challenging the officer (e.g., pointing a gun at the officer, firing at the officer, reaching for a weapon, etc.) accounted for 89 percent of the suicide by cop incidents, compared to only 49 percent of the nonsuicidal incidents. The nonsuicidal incidents were more likely to involve accidental or vague circumstances.


Cross-Validation Study


The authors conducted a follow-up study because of the vague initial finding that suicidal motivation could be implicated in anywhere from 16 to 46 percent of the incidents. The new sample of incidents, taken from the Detroit Free Press files from 1992 to 1993, produced 33 usable incidents. Once again, the authors classified many of the incidents (42 percent) as indeterminate; of the 19 incidents with sufficient detail, the authors classified 9 (47 percent) as having possible suicidal motivation. Although this still leaves the range of possible suicidal incidents vague, it does support the conclusion that suicide by cop is not a rare occurrence.




Perhaps the difficulty of identifying and dealing with suicide by cop is best illustrated by two incidents found in the cross-validation study, neither of which qualified as police shootings. In one incident, a man fired a shotgun inside a home, wounding three people and killing a 2-year-old girl. When confronted behind the home by officers, the man shouted, "I'm sorry! I'm sorry! Don't shoot me! I'm gonna shoot myself after which the man fired into his chest.20


The second case involved a 17-year-old youth with no criminal record who was chased by police one night when he failed to pay for gas. He managed to elude police but later killed himself with one of many handguns found in his possession.21


One of these cases stemmed from a very serious incident; the other, from a comparatively trivial one. Both resulted in actual suicides but could have easily resulted in suicides by cop or in the killing of police officers. In both cases the suicide made the subjects' intentions obvious. Had the incidents resulted in shootings by the police, the evidence of the subjects' suicidal intent may or may not have been detectable.


Although other methodological issues could be taken into consideration, the study establishes suicidal motivation as a significant factor in many police shootings. The percentage of nationally reported police shootings involving probable or possible suicidal motivation ranges between 16 and 46 percent, and the cross-validation study found a similar range of 27 to 47 percent. These results indicate that this phenomenon warrants more careful study. At least four repercussions may develop as policy makers and citizens realize that the cause for many shooting deaths may arise more from a death wish on the part of the subject than from the officer's discretion in a shoot/don't shoot situation.


First, obvious implications exist for police-community relations. Citizens remain concerned about police use of deadly force, and citizens' attitudes play a key role in determining police effectiveness. The extent to which police shootings may be victim precipitated constitutes a variable that merits inclusion in the shaping of those attitudes.


A second important issue concerns dealing with police stress. While the concept of critical incidents for police now covers a broad spectrum, postshooting stress remains a major problem for many officers. The fatal shooting of a suicidal person, who perhaps has a mental illness, may be more or less stressful than the shooting of a dangerous felon. An awareness of this type of shooting situation remains critical for officers and police psychologists to understand more about the frequency and circumstances of suicide by cop.


A third implication involves civil litigation instigated against police officers for use of force in wrongful death actions. Admittedly, for police to say that a civilian engaged in suicide by cop may sound like a self-serving attempt to excuse the shooting, or at least to divert any negative community backlash. Yet, even an unambiguous case of suicide by cop does not necessarily exonerate the officer involved; officers still must make reasonable attempts to avoid having to use deadly force. An understanding of the dynamics of suicide by cop may help juries determine the practicality of alternative actions officers may be expected to take.


Finally, an appreciation of the extent of suicide by cop may have widespread training implications. Officers who recognize the suicidal intentions often motivating the actions of disturbed persons may use a different approach in those calls involving domestic violence offenders and barricaded subjects, as well as a variety of other calls that involve police-citizen encounters. At the academy level, instructors devote only about 9 percent of basic training time to interpersonal skills.22 The percentage of time allotted to such training may need increasing or readjusting to deal appropriately with the suicide-by-cop phenomenon.




Police officers often have suspected that many police shootings are the ultimate result of suicidal intentions on the part of the subjects themselves. Whether explained as a form of victim- precipitated homicide, a consequence of impulsivity, or a result of various personality disorders, more must be learned about the phenomenon of suicide by cop. The results reported here suggest that the phenomenon plays a significant factor in police shootings.


Further research into this topic could have a significant impact on police-community relations by illustrating the role of many shooting suspects in causing their own deaths. Police officers themselves could better adjust to the trauma of shootings by gaining an appreciation of the suicidal nature of many subjects. The ability to curb litigation also would occur as juries more appropriately assess the culpability of all parties to a shooting. Finally, management could adjust police training and tactical operations to more appropriately respond to the phenomenon of suicide by cop.


In an ideal world, no police officer would ever have to shoot a suspect; peaceful resolution would occur. Yet, every day, officers become involved in dangerous situations where this does not hold true. Properly trained officers who understand the motivations of subjects with suicidal impulses and know how to deal with them will be better prepared to avert these tragedies.




1. Geller and M. Scott, Deadly Force: What We Know (Washington DC: Police Executive Research Forum, 1992).


2. Richard N. Jenet and Robert J. Segal, "Provoked Shooting by Police as a Mechanism for Suicide," The American Journal of Forensic Medicine and Pathology 6 (March 1985): 274-75.


3. Although the hotel parking lot incident involved an armed uniformed security officer rather than a sworn police officer, the authors believe that the expressive career felon would have exhibited the same behavior. Ironically the subject was eventually sentenced to death for the murder of the woman's husband, which had occurred only minutes before the security officer arrived on the scene.


4. See, for example, M. Seltzer and C. Payne, "Automobile Accidents, Suicide and Unconscious Motivation," American Journal of Psychiatry 119 (1962): 237-40.


5. D. Phillips, "Airplane Accident Fatalities Increase Just After Newspaper Stories about Murder and Suicide," Science, 201 1978, 748- 750.


6. D. Lester, Questions and Answers About Suicide (Philadelphia: Charles Press, 1989).


7. J. Kinny, Preventing Violence at Work (Englewood Cliffs, NJ: Prentice Hall, 1995).


8. K. Wormer. "Execution-Inspired Murder A Form of Suicide?" Journal of Offender Rehabilitation 22 (1995): 1-10.


9. D. Lester, The Death Penalty (Springfield, IL: Walter C. Thomas. 1987).


10. For autoerotic asphyxiation. see R. Hazelwood. P. Dietz, and A. Burgess, "Sexual Fatalities: Behavioral Reconstruction in Equivocal Cases," Journal of Forensic, Sciences 27 (1982): 763-773. For a discussion of equivocal death analysis in general and the problems of psychological autopsies see N. Poythress et al., "APA's Expert Panel in the Congressional Review of the USS Iowa Incident", American Psychologist 48 (1993): 8-15.


11. G. Gabbard, Treatment of Psychiatric Disorders, vols. 1-2, 2d ed. (Washington DC, American Psychiatric Press, 1995). W. Wilbanks, "Fatal Accidents, Suicide and Homicide: Are They Related?" Victimology 7 (1982): 213-217; and C. Williams, J. Davidson, and I. Montgomery, "Impulsive Suicidal Behavior," Journal of Clinical Psychology 36 (1980): 90-94.


12. A. Karmen, Crime Victims, 3d ed. (Belmont, CA: Wadsworth, 1996), 12-17.


13. "Lee Brown's Sermon: Cops, Citizens Unite," Philadelphia Inquirer, March 9, 1994, sec. B 20.


14. Lauren Ritchie and Karen Samsock, "Ocoee Police Kill Driver Who Chased Them with Car" Orlando Sentinel, November 15, 1989, sec. D 1.


15. Mitch Gelman, Untitled. New York Newsday, February 5, 1990. sec. News. p. 7.


16. "Killed in Gun Battle," Philadelphia Daily News, December


17. 1993, sec. Local. p-12 17. The concept of a "reasonable hope of success" is problematic. With adrenaline flowing, a shootout with police at long odds may seem reasonable, especially to avoid life imprisonment or a possible death sentence. Nevertheless, the fact that raters can agree indicates, at least, that these categories can be applied consistently.


18. The Miami Herald staff, "Miami Robber Picks Undercover Cop, Police Say," Herald, July 24, 1991, sec. 2B.


19. The "unclear" category was often the result of the actual outcome being indeterminate at the time the story was written.


20. Jim Schaefer, "Gunman Kills Toddler, Hurts Four, Then Shoots Self," Detroit Free Press, March 24, 1992, sec. 3A.


21. "Youth Stuck Up For His Friends," Detroit Free Press, May 4, 1992, sec. 2B.


22. R. Langworthy, et al., Law Enforcement Recruitment, Selection and Training: A Survey of Major Police Departments in the U.S. (Highland Heights, KY: ACJS, 1995: 15).


© 1998 by the Federal Bureau of Investigation, U.S. Dept. of Justice



Click here to return to the “Suicide by Cop” Civil Liability Case Summaries