Analysis of Fatal Police Shootings

Abstract

Police use of lethal force occurs rarely and is legally permissible when used appropriately. When police use fatal force, their behavior is highly scrutinized by the legal system and the public. The present study analyzed factors characteristic of 45 coronial investigations of police engagements that culminated in civilian fatalities between 1980 and 2008 in Victoria, Australia. Findings suggest that fatalities commonly occurred at arrest during unplanned police operation, and the majority of incidents were of short duration. The majority of the decedents were armed, acted aggressively, resisted arrest, and escalated the incident. A third of the incidents met the criteria for possible suicide by police. Differences in mental health, incident, and police response characteristics were found between those who met these criteria and those who did not. The findings and their implications for law enforcement, the public, and the mental health system are presented and discussed.

Keywords

police lethal force mental disorder suicide by police law-enforcement-assisted suicide suicide by cop

Police use of lethal force, when used appropriately, is legally sanctioned by both statute and case laws. It is perhaps partly the result of the many laws and regulations on police use of fatal force and partly the fact that the vast majority of citizens readily comply with police instructions that police resort to the use of fatal force very rarely (FitzGerald & Bromley, 1998; Klinger, 2005). Ultimately therefore, the decision of a police officer to discharge his or her weapon and possibly take away someone’s life is the most important decision an officer can make (Klinger, 2005). In this article, we begin with a review of the literature on lethal use of police force and of the phenomenon called suicide by police. Aiming to address the gaps and recommendations of the extant literature, we present a study of lethal use of police force. Subsequently, detailed findings and discussion of their implications are presented.

Most of the academic research on police use of lethal force occurred in the 1970s and 1980s (McElvain & Kposowa, 2008). Various factors, broadly grouped into external environment (e.g., laws, population characteristics, violent crime rates), internal environment (e.g., police subculture, training, supervision), and situational factors (e.g., officer characteristics, suspect characteristics, and incident characteristics), have been examined as to their effect on an officer’s decision to use deadly force (White, 2001, 2003). Two more recent and comprehensive studies of the effects of a range of external factors on police use of deadly force examined 982 police shootings from 1970 to 1992 in Philadelphia according to a series of external control mechanisms (White, 2001, 2003). These two studies found that although external mechanisms can affect the rates of police use of deadly force, their effects can be overridden by police internal factors. An equally comprehensive analysis of police shootings from 1853 to 1995 in Victoria, Australia, found that although external factors, such as small police–public ratio and increased level of violence in the community, were associated with increases in police shootings, internal police policies such as a ban on firing of warning shots, improved training, and decrease in police support for coercive tactics were associated with a decrease in police shootings (Ballek, 1998). White (2001), however, also found that in those shootings where the police had no option but to shoot, situational variables were the most influential.

Although some research into the effects of situational factors on police shootings has concentrated on one aspect in great depth (e.g., race of the decedents [Jacobs & O’Brien, 1998], officer perceptual distortions [Klinger & Brunson, 2009], officer personality characteristics [Girodo, 2007], officer age, gender, and education [McElvain & Kposowa, 2008]), others have tended to examine many factors at the same time to obtain a comprehensive analysis of the whole incident (e.g., Ballek, 1998; Best & Quigley, 2003; Crawford & Burns, 2008; Police Complaints Authority, 2003). Pinizzotto, Davis, and Miller’s (2007) extensive research, for instance, showed the complexities involved in fatal incidents through describing how the fluid concept they termed the “deadly mix” operates differently in different incidents. Deadly mix consists of dynamic elements specific to officers’ and offenders’ assumptions and perceptions of the situation and of the circumstances that brought them together; however, elements of each of these factors are different and changeable in terms of the influence they exert on the outcome (Pinizzotto et al., 2007).

The effects of decedents’ psychological and mental functioning at the time of the incident have only recently become a more prominent topic of research. Recent reports into police fatal use of force between 1998 and 2001 in the United Kingdom examined the effects of the decedent’s mental state at the time of the incident in more detail, looking at issues such as substance intoxication, presence of mental disorders, irrational behavior, and possible suicidal motivation (Best & Quigley, 2003; Police Complaints Authority, 2003). Findings suggest that the majority of the decedents were made vulnerable by their psychiatric conditions and consumption of substances at the time and that effective police command, which prolonged the duration of an incident, was crucial in influencing the outcome.

Therefore, studies of police use of lethal force have examined many different influencing factors; however, there is a lack of in-depth examinations of decedents’ behavioral factors, more specifically of the effects of their psychological and psychiatric functioning on the incident. One such important factor is the possible presence of the decedent’s suicidality at the time of the fatal incident. In-depth examination of this phenomenon is crucial because, if present, it could dramatically change the circumstances and the outcomes of an incident (Klinger, 2001).

Suicide by Police

Before considering the phenomenon of suicide by police, it is helpful to provide a very brief summary of the suicide literature. Although self-harm and suicidal ideation are much more common than completed suicides (De Leo, Bertolote, & Lester, 2002; De Leo, Cerin, Spathonis, & Burgis, 2005; Kessler, Berglund, Borges, Nock, & Wang, 2005), almost 1 million people throughout the world commit suicide each year (World Health Organization, 2010). Although there may be some differences in the definitions of suicide, the key element of any definition is that the person intends to die (De Leo et al., 2002). Research on suicide has mainly focused on identifying the prevalence of suicidal behavior and on examining the influence of different risk factors on suicidal behavior. Previous suicide attempts and certain types of mental disorders have been found to be the major risk factors for later suicide: affective disorders, some anxiety disorders, schizophrenia, substance abuse disorders, and personality disorders (Bostwick & Pankratz, 2000; De Leo et al., 2002; Harris & Barraclough, 1997; Kessler et al., 2005; Palmer, Pankratz, & Bostwick, 2005; Schneider, 2009). As noted by De Leo and colleagues (2002), these risk factors interact in complex and still somewhat unclear ways.

The notion that people who intend to die sometimes choose to effect this by provoking another to kill them has been reported in many different societies and historical time periods (Lindsay & Lester, 2004). An early scholarly study of the phenomenon occurred in 1959, with Wolfgang’s examination of victim-precipitated homicide. Despite this early work, further systematic scientific examination of this phenomenon has been ongoing only for the past two decades.

Despite the scant research, there are a number of different labels and definitions of the phenomenon: suicide-by-cop (Homant & Kennedy, 2000a; Kennedy, Homant, & Thomas, 1998; Lindsay & Lester, 2008; Mohandie, Meloy, & Collins, 2009), law-enforcement-assisted suicide (Lord, 2000), law-enforcement-officer-forced-assisted suicide (Hutson et al., 1998), police-involved-victim-provoked shooting (McKenzie, 2006), victim-precipitated homicide (Parent & Verdun-Jones, 1998), and suicide by police (Homant & Kennedy, 2000b), the term that is used here.

The differences between the various definitions of suicide by police seem to center on how specifically, and conversely how broadly, the phenomenon is defined. Hutson and colleagues (1998) stated that suicide by police is when

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. (p. 665)

They argued that there must be evidence of each of these elements (i.e., suicidal intent, wanting to be shot by the police, possessing a deadly weapon or what appeared to be a deadly weapon, and provoking the police into shooting him or her) for an incident to meet the criteria. However, Lord (2000) proposed a somewhat different definition, stating that suicide by police is when

those individuals who, when confronted by law enforcement officers, either verbalized their desire to be killed by law enforcement officers and/or made gestures such as pointing weapons at officers or hostages, running at officers with weapons, or throwing weapons at officers. (p. 403)

Further still, Mohandie and colleagues (2009) stated that it is “when the subject engaged in actual or apparent risk to others with the intent to precipitate the use of deadly force by law enforcement personnel” (p. 457). On the other hand, Lindsay and Lester (2004, 2008) proposed 17 criteria that are more inclusive of a number of historical and incident-level factors, claiming that a statistical cutoff of at least 60% of these (i.e., 3 of the 5 historical and 8 of the 12 incident) variables must be present for a case to meet the suicide by police criteria.

McKenzie (2006) reviewed the extant literature and criticized it for the lack of scientific rigor. He proposed that there is a difference between those individuals who may have been indifferent as to whether they live or die and those who intended to commit suicide. According to McKenzie, those people who are indifferent would exhibit behaviors such as ignoring police commands or reluctantly obeying them, whereas those who had intent to suicide would actively assist in bringing about the shooting through active hostility and threats to those present. McKenzie elaborated further that in the absence of clear and reliable evidence that the person wanted to die, a behavioral chain of events ought to be examined rather than declaring the incident as suicide by police on the basis of one or two behaviors.

Although suicide by police appears to be an international phenomenon, it is most commonly reported and studied in the United States (e.g., Lord, 2000; Lord & Sloop, 2010), Canada (e.g., Parent, 1996; Parent & Verdun-Jones, 1998), and more recently the United Kingdom (Best, Quigley, & Bailey, 2004). Suicide by police has been estimated to occur in anywhere between 10.5% and 50.0% of police shootings (Best et al., 2004; Hutson et al., 1998; Mohandie et al., 2009; Parent, 1996). Based on these studies, the following factors have been identified as significantly contributing to the occurrence of these incidents: mental disorders, presence of alcohol or drugs in decedent’s system, past suicide attempts, use of weapons, and violent behavior such as threatening the police and the public with harm (Hutson et al., 1998; Lord, 2000; Mohandie et al., 2009; Wilson, Davis, Bloom, Batten, & Kamara, 1998).

Fatal Use of Force in Victoria, Australia

The Victoria Police force is highly centralized and armed, with approximately 11,250 police officers (Victoria Police, 2009). Sworn officers must attend a 2-day training twice a year to qualify to carry firearms and less-than-lethal weapons. Sections 462A and 463B of Victoria’s Crimes Act 6231 (1958) and Section 10 of Mental Health Act 59 (1986) control their use of force. In addition, Section 6.3 of the Victoria Police manual stipulates that sworn officers can draw a firearm only if extreme danger is anticipated and can discharge a firearm only “when they reasonably believe it is necessary to protect life or prevent serious injury” (Victoria Police, 2008, p. 5).

In recent history, Victoria Police were reported to have engaged in fatal incidents at a much higher rate than those in the other Australian states (Brouwer, 2005), where in the period between 1980 and 2008 they fatally wounded 48 people (Kesic, Thomas, & Ogloff, 2010). Of the 76 fatal police shootings that occurred in Australia between 1990 and 2004, the majority (n = 29) occurred in Victoria, followed by the New South Wales (n = 18) and Queensland (n = 11; Brouwer, 2005). Following a spate of shootings in 1994, and consequent public outcry about these shootings, major changes were introduced to police policy and training under the banner of Project Beacon (Brouwer, 2005; Victoria Police, 1996). A recent study found that subsequent to the changes brought about by Project Beacon’s initiatives, the number of fatal shootings reduced by half; however, the rates of decedents with mental disorders did not fall (Kesic et al., 2010). Nevertheless, a limitation of the latter study was that analyses precluded an in-depth examination of the situational characteristics, specifically those that relate to mental disorder and behaviors exhibited and observed during the incidents.

Aims of the Current Study

There is a lack of research examining psychological and psychiatric factors in fatal force incidents, with research on suicide by police being particularly lacking. The present study aims to examine the situational environment of the fatal force encounters, as recommended in the extant literature in this field, exploring the following questions: (a) In which context did the incidents occur? (b) What were the criminal, psychiatric, and psychological backgrounds of the decedents? (c) What were the characteristics of decedents’ behavior and mental state during the incidents? (d) What was the range of tactical responses used by the police? (e) What is the extent of suicide by police? and (f) Do the characteristics of those who meet the criteria for suicide by police differ from others?

Method

Sample

Between January 1980 and February 2007, 48 people were fatally wounded by the police in Victoria, Australia (Kesic et al., 2010). The sample comprised those cases where the coroners had finalized their investigations, thereby providing a sample of 45 of the 48 cases.

Setting

Victoria is situated in the southeastern part of Australia. Although it is the smallest mainland state, comparable in size to Great Britain and closest in size to the U.S. state of Oregon, it has the second largest population of Australia, with more than 5.5 million residents (Australian Bureau of Statistics, 2010). Victoria is a highly multicultural state; its inhabitants come from more than 200 countries, and 43.6% are either first- or second-generation migrants (State Government of Victoria, 2010). More than 70% of Victorians live in the metropolitan area of the capital city Melbourne (State Government of Victoria, 2010).

Legislation

Although the Coroners Act (1985) of Victoria (hereafter termed the Act) had been repealed by the Coroners Act (2008) with some significant changes, all the coronial investigations in this sample were completed under the jurisdiction of the former Act. In Victoria the coroner has the jurisdiction to investigate all reportable deaths. These are defined as any death that occurred in Victoria; where the cause of death or the body was found in Victoria; where the death appears to have been unexpected, unnatural, violent, resulting from an accident or from an anesthetic application; where the decedent had been held in care or custody immediately prior to the death (e.g., patient under Victoria’s Mental Health Act of 1986, or under the care of the Secretary to the Department of Justice, or in police custody); and/or where the identity of the person is unknown. Following the investigation, the coroner must decide the identity of the deceased, how the death occurred, and the cause of death. Furthermore, a coroner may comment on any public safety or administration of justice issues arising from the death investigation and may make recommendations on these and any other relevant issues to any minister or public statutory authority.

Databases

Data were extracted from a series of databases: (a) the Use of Force (UoF) Register, maintained by Victoria Police; (b) the Law Enforcement Assistance Program (LEAP), also maintained by Victoria Police; and (c) the Redevelopment of Acute Psychiatric Directions (RAPID) register, also known as the Victorian Psychiatric Case Register (VPCR), which is maintained by the Department of Health. This approach is consistent with that of Kesic et al. (2010).

The UoF register came into operation in October 1995, aiming to record information on the incidents where force had been used against or by the police. The LEAP database came into operation in March 1993; it records all contacts between police and the public as witnesses of crimes, victims of crimes, alleged offenders, and offenders. For those contacts that occurred prior to 1993, information is kept on the Information Bureau of Records (IBR) cards, and, where appropriate, these records were also scrutinized for this study. The RAPID/VPCR (also in a new format of operation since 2000) records all contacts with the public mental health system, community and in-patient services, but it does not record contacts with private hospitals, private practitioners, or general practitioners for mental health issues (Wallace et al., 1998).

Materials

A three-page coding pro forma was designed. Variables coded were broadly separated into six categories, derived from a comprehensive review of the literature: (a) decedent-related background variables (i.e., gender, age, ethnicity, marital status, socioeconomic status, employment status, life circumstances, mental health history, offending history), (b) general incident variables (i.e., reason for police intervention, location, estimated duration, environmental factors), (c) decedent incident variables (i.e., violent or aggressive behavior, using violent or aggressive language, resisting arrest, exhibiting possible suicidal behavior, being substances affected, and any other pertinent information), (d) police tactical response (i.e., number of police officers, senior police present, particulars of any less-than-lethal tactical response to the incident, such as the use of communication, negotiation, cordon and containment, tactical disengagement, use of less-than-lethal tools and weaponless tactics), (e) coroner’s conclusions (i.e., was the death avoidable, who contributed to the death, implications for the police practice and training), and (f) suicide by police variables according to two different criteria (see below for details). A synopsis of each event was also recorded.¹

Procedure

Identifying data on all individuals who were fatally wounded by Victoria Police in the state of Victoria, Australia, between January 1980 and May 2008 were sourced from the UoF Database. Only their identifying information was sent to the Coroners Court of Victoria in a password-protected file to retrieve coronial reports for all closed cases retained by them. The same information was also sent to the LEAP and RAPID/VPCR databases, and mental health and offending data were returned to the researchers in a deidentified database.

In total, 28 coronial reports were received from the Coroners Court, 1 was obtained from the Coroners Court website, 5 were collected from the Monash University Library, and 11 case records of investigation were sought from the Public Records Office of Victoria. In these 11 cases, coronial reports, toxicology analysis, and IBR criminal records were sampled. We were unable to find the final report of the coronial investigation for one case, in which case we sampled all the relevant details from the evidence material (e.g., criminal history, autopsy report, counsels’ opening statements, and summary of the incident) to obtain the most comprehensive picture of the individual and of the incident. The quality, depth, and breadth of information contained in each of the reports varied widely. One of the authors (D.K.) read all the coronial reports and entered the relevant information in the corresponding coding sheet. All the variables were coded as present, absent, or unknown. The second author (S.T.) then reviewed all cases regarding their eligibility under the operationalized definition of suicide by police criteria, with consensus opinions then being reached for all 45 cases. Because of the agreement in all cases, exact interrater reliability was not calculated.

Offending history was coded as present only in instances where the person was formally charged and convicted of a criminal offense. History of interpersonal violence was coded as any incidents of violent interpersonal behavior, either where the person had been charged for interpersonal violence or where no formal charges had been made for their assaultive behavior or threats of violence (Lord, 2000). Mental health history was coded into the following categories: mental illness (all Axis I disorders excluding substance use disorders), mental disorder (all Axis I and Axis II disorders), anxiety disorders, mood disorders, substance use disorders, schizophrenia, and psychosis (including schizophrenia). Personality disorders of Axis II were also coded. This information was coded in one of three ways (a) RAPID/VPCR database diagnostic information was coded into the aforementioned categories; (b) where specified in the coroners report that the person had a history of these disorders; or (c) where there was enough evidence in the coronial report to warrant a code in one of the categories based on fitting the diagnostic criteria of the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, Text Revision (American Psychiatric Association, 2000). To these ends, 12 cases of mental health diagnoses were obtained from the RAPID/VPCR database, and in 15 cases information was obtained from the coronial reports, whereas in 1 of these cases the data were obtained from both the RAPID/VPCR and coronial reports. Most of the cases where additional information was obtained from coronial reports (13, 86.7%) contained information about substance use disorders; in 3 of these cases there was evidence for other Axis I disorders as well (e.g., a diagnosed mood disorder and schizophrenia disorder in two cases).

In consideration of the criticisms of the extant suicide by police studies regarding the need for a specific operational definition and the resulting criteria, we adapted Hutson et al.’s (1998) criteria by separating their first intent criterion into evidence of (a) gestures of suicidal intent and (b) communication of suicidal intent, with the rest of the criteria remaining the same. However, we elaborated operationalization of each criterion in line with existing studies (e.g., Lindsay & Lester, 2004; Lord, 2000; Mohandie et al., 2009). Therefore, based on these criteria our first criterion required that all of the following five be present for a case to be classed as probable suicide by police: (a) communication of suicidal intent (i.e., either before the incident through methods such as telling someone that they want to suicide, or that they specifically want police to kill them, or that if caught by the police they will not be taken alive, or writing a suicide letter, or during the incident stating that they want to kill themselves, or asking the officers to shoot them), (b) gesture of suicidal intent (i.e., pointing a weapon at the police and refusing to drop it at request, or running at the police with an edged weapon, or attacking officers with blunt weapons capable of causing fatal or serious injuries in close proximity, or self-harm, or pointing a weapon at themselves), (c) they wanted to be shot (i.e., asking the police to shoot them, or running at the police officers with a weapon, or continuing to aim a weapon at the police despite orders to drop the weapon, or telling family or friends that it is their wish to be shot by the police), (d) possession of a deadly weapon or what appeared to be a deadly weapon (i.e., any weapon capable of inflicting serious or fatal injury, such as a gun, a crossbow, or edged weapons in close proximity such as an axe, machete, or knife, or blunt instruments in close proximity, such as a metal wrench or claw hammer), and (e) evidence that they escalated the encounter by provoking the police to shoot them in self-defense or defense of another person (i.e., resisting police orders to drop the weapon and/or to stop, or attacking the police or the public instead of surrendering to orders, or threatening police or civilians with weapons).

In addition, an examination of chains of behavior during the incident was completed according to the Lindsay (2001) and Lindsay and Lester (2004, 2008) criteria. They proposed 17 variables specific to suicide by police, split into historical criteria (i.e., presence of chronic medical or mental illness, substance abuse, low socioeconomic status, past suicide attempts, and criminal history) and incident-level variables (i.e., the person initiated the incident, designed the incident to ensure police response, forced confrontation, initiated aggressive action, threatened officers with a weapon, advanced toward the officers, refused to drop the weapon, threatened citizens with harm, possessed a deadly weapon, experienced a stressor, or injured officer or citizen, and officer retreated in fear for his or her life; see Lindsay and Lester, 2004, 2008, for further details). Despite the fact that the low socioeconomic status variable was not examined because of scarcity of information present to evaluate it consistently or meaningfully, we still retained the original cutoff for the criteria to be met; thus, a minimum of three of the historical and eight of the incident variables were required for the incident to fit the criteria of suicide by police.

The study obtained ethical approval from the Monash University and Victoria Police ethics committees and organizational support from the Department of Human Services and Coroners Court of Victoria.

Analysis

Data were analyzed in SPSS 17 and 18. Simple descriptives were used to characterize the data. Associations between categorical variables were examined through chi-square tests of association, with Fischer’s exact test statistic used when the cell number fell below 5. Independent t tests were used to examine associations between continuous variables with normal distribution, and the nonparametric Mann–Whitney U test statistic was used for variables with skewed distributions. A Bonferroni correction was applied to account for multiple comparisons. All comparisons were two-tailed.

Results

The vast majority of the sample was male (n = 42, 93.3%). The mean age at death was 32.41 years (SD = 9.28, range = 19.09–50.51 years), with more than a quarter (n = 12, 26.7%) younger than 25. The vast majority of cases (n = 38, 84.4%) had a history of criminal offending. More than half (n = 26, 57.8%) had a mental disorder, whereas in 13 (28.9%) of these cases mental illness was present. The majority (n = 9, 69.2%) of those who had a mental illness also had a substance use disorder. The most common disorder was substance use disorder (n = 22, 48.9%), followed by a mood disorder (n = 8, 17.8%). Psychosis was found in 6 (13.3%) cases, of which 5 (11.1%) had a diagnosis of schizophrenia. Anxiety disorders were found in 5 (11.1%) cases; 6 (13.3%) people had a personality disorder. The vast majority (n = 36, 80.0%) had a history of interpersonal violence and a history of experiencing stressful life events (n = 44, 97.8%).

General Incident Characteristics

The majority (n = 32, 71.1%) of the incidents occurred in the metropolitan area of Victoria, and more commonly than not they were unplanned police operations (n = 26, 57.8%) as opposed to planned operations (n = 19, 42.2%). The most common reason for the involvement of the police was to apprehend the person, either for the crime he or she was committing at the time (n = 18, 40.0%) or for a crime that he or she was alleged to have committed previously (n = 10, 22.2%). In 6 (13.3%) cases police received a call regarding a person with mental health problems, in 4 (8.9%) they received a call about a possible domestic abuse situation, and in 2 (4.4%) cases they performed a traffic check. There were 3 (6.7%) cases where the reason for police involvement did not fall into any of these categories.

In more than half (n = 27, 60.0%) of the incidents general duties police primarily managed the incident, and in more than a quarter of the incidents (n = 12, 26.7%) the Special Operations Group was involved in the resolution of the incident. Roughly the same number of incidents occurred in public space (n = 24, 53.3%) as in residencies (n = 21, 46.7%), and anywhere between 1 and 14 officers were involved (M = 4.37, SD = 3.27). Senior police were present in the majority of the incidents (n = 29, 64.4%). The average estimated duration of the incidents from the time police attended the scene to when they discharged their weapon was 27.17 minutes; however, this varied enormously, with the distribution being negatively skewed (Mdn = 10, SD = 65.73, range = 0.1–360 minutes). More than a third (n = 17, 37.8%) of the incidents occurred in 1 minute or less, and 73.3% (n = 33) of the incidents occurred in 10 minutes or less.

Use of Force

As depicted in Table 1, the vast majority (n = 41, 91.1%) of decedents were behaving in an aggressive manner, had weapons (n = 41, 91.1%), refused to drop the weapons (n = 37, 82.2%), threatened officers with those weapons (n = 36, 80.0%), and resisted arrest (n = 40, 88.9%). More than half (n = 26, 57.8%) were affected by substances. Officers were assaulted in 10 (22.2%) cases; however, the exact injuries were not specified in 6 of these cases, with the injuries in the rest ranging from minor lacerations to bullet wounds.

Table 1:

Decedents’ Behavioral Characteristics in the 45 Fatal Incidents

Decedents’ Characteristics
Variable	Frequency	%	Variable	Frequency	%
Substances affected^a	26	57.8	Aggressive or violent behavior	41	91.1
Alcohol affected	14	31.1	Aggressive or violent language	19	42.2
Drugs affected	15	33.3	Resist arrest	40	88.9
Weapons^b	41	91.1	Forcing confrontation	38	84.4
Firearms	19	42.2	Refusing to drop their weapon	37	82.2
Edged weapons	20	44.4	Threatens citizens	7	15.6
Threatens officers	36	80.0

Three decedents had both alcohol and drugs in their system.

Three decedents used weapons other than edged weapons or firearms (i.e., baseball bat, hammer, and wrench).

Police used a range of methods to manage the situation (Table 2). In the first instance, police were found to have applied proper risk assessment and management in more than a half of the incidents (n = 25, 55.6%). Applying a cordon and contain approach (n = 19, 42.2%) and seeking assistance of backup teams (n = 11, 24.4%) were the most common nonlethal tactics used after communication, which was found in all of the incidents (n = 45, 100.0%). Police also requested the help of mental health practitioners in 4 cases (9.1%), and in 3 of these mental health professionals assisted the police. In a third (n = 15, 33.3%) of the incidents there was evidence of communication above that of only issuing orders (e.g., “Police. Stop. / Police. Don’t Move. / Police. Drop your weapon.”), and in 6 (13.3%) of these the police attempted to negotiate with the decedent.

Table 2:

Police Response in the 45 Incidents

Police Response
Variable	Frequency	%	Variable	Frequency	%
Risk assessment and management	25	55.6	Less-than-lethal tools^a	5	11.1
Communication	45	100.0	Baton	2	4.4
Cordon and contain	19	42.2	OC Spray or foam	2	4.4
Tactical disengagement	6	13.3	Tear gas	1	2.2
Negotiation	6	13.3	Distraction grenade	1	2.2
Backup teams	11	24.4	Firearm	45	100.0
Empty hands tactics	8	17.8

In one incident more than one less-than-lethal tool was used.

Suicide by Police

In all, 14 (31.1%) cases met the first criterion (i.e., based on Hutson et al., 1998); however, when broken down into the criterion components, results show that although 15 (33.3%) cases had evidence of communication of intent, the majority of the sample displayed evidence of gesture of intent (n = 33, 73.3%), wanted the police to shoot (n = 33, 73.3%), possessed a deadly weapon (n = 40, 88.9%), and escalated the incident where the police had to resort to shooting (n = 38, 84.4%). By contrast, 8 (17.8%) cases met the second criterion (i.e., Lindsay & Lester, 2004, 2008). There was a significant degree of overlap between the cases, where all but one case that met the second criterion also met the first criterion (this case only did not have evidence of communication of intent). The two criteria were combined together into one suicide by police variable where the cases had to meet either one of these criteria (n = 15, 33.3%), and subsequent analyses were completed on this combined variable.

All but one of these 15 decedents were male (n = 14, 93.3%). The average estimated duration of these incidents was 64.21 minutes (Mdn = 30, range = 0.1–360 minutes), and whereas more than a quarter occurred within a minute (n = 4, 26.7%), more than half (n = 8, 53.3%) lasted for more than 10 minutes. These incidents were significantly more likely to be of longer duration than those not meeting the criteria for suicide by police (U = 134.50, n ₁ = 15, n ₂ = 30, p = .027). The most likely reasons for police involvement were receiving a mental-health-related call (n = 5, 33.3%) and apprehension at the time of the decedent committing a crime (n = 5, 33.3%), followed by a domestic abuse call (n = 3, 20.0%), raid for the purpose of arrest (n = 1, 6.7%), and a traffic check (n = 1, 6.7%). Thus, the majority (n = 13, 86.7%) were unplanned police operations, found to significantly more likely occur in the suicide by police group (p = .009, Fisher’s exact test [FET], OR = 8.50, 95% CI = 1.63, 44.46).

Table 3 presents the occurrence of Lindsay and Lester’s (2004, 2008) suicide by police criteria in those cases that meet the combined criteria and compares it to those cases that do not meet the criteria. The table illustrates that chronic physical or mental illness, substance abuse, and criminal history were commonly found background factors in the suicide by police cases. Furthermore, there was a high degree of aggressive behavior, where all cases (n = 15, 100.0%) that met the suicide by police criteria evidenced (a) escalation of the incident by forcing confrontation with the officers and (b) possession of a deadly weapon, (c) with which they threatened the officers and (d) refused to drop the weapon. When we compared these characteristics between the two groups, a number of significant differences emerged: Mental or chronic physical illness, past suicide attempts, planning the incident to get the police to come to the scene, threatening officers with a weapon, advancing at officers, and refusing to drop the weapon were all significantly more likely to occur in the suicide by police group (see Table 3).

Table 3:

Comparisons between Suicide by Police and Non–Suicide by Police Cases on Lindsay and Lester’s (2004, 2008) Criteria

	SBP (n = 15)		Non-SBP (n = 30)
Variable	n	%	n	%	χ²	OR	95% CI	p
Chronic mental or physical illness	9	60.0	5	16.7	8.76	7.50	0.83, 30.73	.006
Substance use disorders	9	60.0	13	43.3	1.11	1.96	0.56, 6.92	.353
Suicide attempts	5	33.3	1	3.3		14.5	1.51, 139.53	.012^
Criminal history	12	80.0	26	86.7		0.62	0.12, 3.19	.670^
Initiated the Incident	12	80.0	17	56.7		3.06	0.71, 13.13	.189^
Planned the incident	9	60.0	5	16.7	8.76	7.50	1.83, 30.73	.006
Forcing confrontation	15	100.0	23	76.7				.077^
Initiates aggression	15	100.0	23	76.7				.077^
Threatens officer with a weapon	15	100.0	21	70.0				.020^
Advances toward officers	13	86.7	16	53.3		5.69	1.09, 29.69	.046^
Refuses to drop weapon	15	100.0	22	73.3				.038^
Threatens citizens	4	26.7	3	10.0		3.27	0.63, 17.09	.199^
Deadly weapon	15	100.0	23	76.7				.077^
Injured officer or citizen	5	33.3	8	26.7	0.22	1.38	0.36, 5.27	.732
Officers retreat in fear for life	4	26.7	4	13.3		2.37	0.50, 11.19	.410^
Recent stressor	13	86.7	11	36.7				.003^

Note. SBP = suicide by police cases; non-SPB = cases that do not meet suicide by police criteria; χ² = chi-square; OR = odds ratio; CI = confidence interval. ORs not obtained when any cell number = 0.

Fisher’s exact test (FET).

Moreover, anxiety disorders and schizophrenia were both present in more than a quarter of the suicide by police incidents (n = 4, 26.7%) and were more than 10 times more likely to occur in this group (p = .036, FET, OR = 10.55, 95% CI = 1.60, 105.03). More than a half of the suicide by police cases were affected by alcohol at the time (n = 9, 60.0%), which is more than 7 times more likely than those not meeting suicide by police criteria (χ² = 8.76, p = .006, OR = 7.50, 95% CI = 1.83, 30.73). Furthermore, personality disorders were significantly more likely to be found in the suicide by police group (n = 6, 40%, p = .001, χ² = 13.85).²

Apart from experiencing mental illness and substance use disorders, they were found to more likely have experienced a recent stressor than those who did not meet the suicide by police criteria (p = .003). They were significantly more likely to have experienced a stressor in the 24 hours leading up to the incident than those who did not meet the criteria (n = 4, 40.0% vs. n = 5, 16.7%, p < .001, FET). The most common current stressors affecting those who met the criteria for suicide by police involved interpersonal problems (n = 10, 66.7%), grief and loss (n = 5, 33.3%), legal problems (n = 4, 26.7%) and medical problems (n = 3, 20.0%). Almost half (n = 7, 46.7%) had multiple current stressors recorded in the coronial files.

In the majority of the suicide by police incidents, police applied cordon and contain (n = 10, 66.7%); they also used negotiation (n = 6, 40.0%), and police employed tactical disengagement in a third (n = 5, 33.3%) of the cases. Moreover, the police were more likely to use these less-than-lethal means of incident management in this group than in the non–suicide by police group: negotiation (p = .001, FET), tactical disengagement (p = .012, FET, OR = 14.50, 95% CI = 1.51, 139.53) and cordon and containment (χ² = 5.51, p = .027, OR = 4.67, 95% CI = 1.24, 17.60).

When we applied a Bonferroni correction to these results, only three remained statistically significant: Personality disorders and experiencing a stressor in the 24 hours prior to the incident were more likely to be found in the suicide by police group, and police were significantly more likely to use negotiation in this group. However, given the increased probability of Type II error in this study, it is important to consider all the findings significant at p < .05.

Discussion

This study examined 45 fatal force incidents that occurred in Victoria, Australia, between January 1980 and May 2008, aiming to explore the context in which these occurred, factors specific to decedents’ background and behavior during the incident, and the police response. The discussion presented hereafter should be interpreted in the context of a number of limitations. This study was primarily an exploratory investigation of a small number of cases on a topic not previously examined in Australia. Because of the small sample size, the findings should serve as possible trends and avenues of further exploration for the researchers. Furthermore, differences in geographical, population, political, and police characteristics between this context and others need to be considered first when attempting to generalize these findings elsewhere. In addition, coronial reports as a method of examining this area pose some practical and logistical challenges. Although the reports generally contained adequate information about the situational environment of the incident, the depth and the quality of this information varied. The information about decedents’ offending and psychiatric history was also garnered from the official databases to increase the reliability of the data; however, as with coronial reports, these databases also have some inherent limitations with respect to completeness (see Kesic et al., 2010).

Time, Space, and Use of Force

Research findings presented here are mainly consistent with the international literature. The vast majority of decedents were male, and the average age was 32, corresponding to those in other fatal use of force studies (Gill & Pasquale-Styles, 2009). Findings suggest that these fatal incidents are likely to be unplanned police operations effected by general duties police. They were more likely to occur in metropolitan areas, this being consistent with the state’s population distribution. In the majority of the cases police were trying to conduct an arrest, findings consistent with the latest U.S. report (Mumola, 2007).

The average duration of the incidents from police attending the scene to discharging their weapons was just less than half an hour; however, the vast majority of the incidents occurred in less than 10 minutes. The vast majority of the decedents were armed, threatened the police with the weapons, and refused to heed police commands; these findings echo those from other studies (Hutson et al., 1998; Mohandie et al., 2009; Mumola, 2007). Consistent with some other studies of police shootings (Hutson et al., 1998), police used a range of options in an attempt to manage the situation prior to using lethal force. Although they verbally communicated with the decedents in all cases, examination of these findings suggests a complex interplay of factors in the fatal shootings that need to be taken into account when examining police responses in these situations (see Figure 1). The results suggest that the extent to which the police applied appropriate risk assessment and management and used various less-than-lethal options was in the main a reflection of the type of incident, how rapidly it evolved, and the distance between the officers and the decedent.

Figure 1:

Complex Interplay of Factors Pertinent in the Management of Critical Incidents

As depicted in Figure 1, these dynamic factors (i.e., risk assessment and management, time, space, available and appropriate resources) influence each other in complex ways during any given critical encounter. Although each of the factors can directly influence a critical incident and the type of critical incident can also affect each of these factors, in addition they can interact to cumulatively influence the outcome. For instance, appropriate risk assessment can ensure that, through the use of available resources (e.g., good communication and deescalation skills), officers maintain a safe distance from the suspect while deescalating the situation, thereby prolonging the duration of the incident, allowing further reassessment of risk, and increasing the likelihood of a positive outcome. Conversely, some types of critical incidents escalate rapidly, precluding the possibility of appropriate risk assessment, where nonlethal options may not be available or appropriate because of unsafe distance, rapid escalation, or both.

In an attempt to explain the complexity of police critical incidents, Klinger (2005) applied the principles of Charles Perrow’s normal accident theory, noting that as the numbers of officers and suspects increase, and as the distance between them narrows, this leads to a highly complex and dynamic system where many things can go to error and result in a fatality. He suggested maintaining the distance between the officer and civilians and calling in the specialist teams when the appropriate skills and resources are not immediately available. As recently observed by one of the authors, the first cycle of 2010 Victoria Police training that specialized in police interaction with persons experiencing mental illness appropriately emphasized these important variables. It is imperative that this type of training continue to be a regular priority and be well incorporated within the rest of police training.

The vast majority of decedents in this study were physically aggressive, and more than half of them were affected by substances at the time. The vast majority had a criminal offending history and an interpersonal violence history. More than half had a mental disorder, most commonly substance abuse disorders; however, affective disorders, anxiety, and schizophrenia were also prevalent in the sample. Although exploration of exact mental disorders in this area is scant, similar findings regarding offending and general mental health characteristics have been commonly reported in other studies of police shootings (Hutson et al., 1998; Mohandie et al., 2009). A minority of decedents also had personality disorders, the importance of which has been established in both criminal offending and risk of suicide studies (e.g., De Leo et al., 2002; Fazel & Danesh, 2002). Findings that suggest that persons with mental disorders are commonly found in fatal force incidents have significant implications for the police (Kesic et al., 2010). Appropriate police training that encapsulates theoretical, experiential, and scenario-based methods in managing incidents where individuals appear agitated, appear under the influence of substances, or may be experiencing mental disorders, similar to that recently delivered by Victoria Police, ought to be regularly maintained to try to minimize the likelihood of a fatal incident occurring.

Suicide by Police

We applied stringent criteria to operationalize those cases identified as constituting suicide by police: (a) We adapted Hutson et al.’s (1998) criteria to clearly define that both suicidal communication and gestures were required, and (b) we used Lindsay and Lester’s (2004, 2008) criteria to examine the individual’s chains of behavior during the incident.

Despite these considerations, a few remaining issues ought to be noted. Because of the design of our study, where coronial reports of fatalities were analyzed, in many of these cases we are still unable to confidently say that the person actually intended suicide. Even when we have evidence such as decedents urging the police to shoot them and stating that they will commit suicide, along with other evidence that distinguishes them from those who seemed indifferent, evaluating intent to die versus indifference because of other factors such as high emotional arousal, impulsivity, anger, desperation, and the effect of mental disorders and substances on thinking and behavior during the incident remains problematic and subjective. Indeed, measuring intent in the suicide literature has also been problematic, and as a result most studies mainly focus on examining risk factors of suicide (Lord & Sloop, 2010). To better understand the phenomenon, those cases that meet the existing criteria for suicide by police and that are successfully averted should be studied to examine the extent of suicidal intent and to examine which police tactics are successful in averting these fatal outcomes.

A third of the fatalities met either one of the criteria used. It is important that although one in three cases met the adapted Hutson et al. (1998) criteria, one in five met the adapted Lindsay and Lester (2004, 2008) criteria. Mainly as the result of the differences in definitions, the populations studied (e.g., police shootings vs. police fatalities), and the method of study, it is difficult to make meaningful comparisons with the rest of the suicide by police studies in terms of the frequency of this phenomenon. However, our results are within the range of those found in the literature on suicide by police (Best et al., 2004; Hutson et al., 1998).

A number of significant differences emerged between those who met the criteria for suicide by police and those who did not. Before we discuss those, it is important to note that after Bonferroni correction only three variables were significant, and as such the following should be viewed as significant trends deserving of further exploration. In terms of the incident characteristics, although suicide by police incidents were found to be unplanned police operations, they also lasted significantly longer than the incidents not meeting the criteria. These findings are different from those of Hutson et al. (1998), but they seem comparable to those of Mohandie et al. (2009). The types of calls police were asked to attend were qualitatively different between these incidents and those not meeting the criteria, where in more than half of the cases calls to assist a mentally disturbed person and domestic abuse calls occurred. In addition, the police were more likely to use a number of nonlethal tactics (e.g., cordon and contain, tactical disengagement, and negotiation) in these incidents than in those not meeting the suicide by police criteria. These findings suggest that the police are perhaps identifying suicide by police incidents and responding to them differently in those cases where time and other considerations permit.

All decedents meeting our criteria for suicide by police forced confrontation with the police, initiated aggressive behavior instead of surrendering, possessed a deadly weapon, refused to drop the weapon, and advanced at the officers with the weapon. Furthermore, a number of significant differences emerged regarding the differences in mental health and psychological functioning. Namely, mental illness (specifically anxiety disorders and schizophrenia), personality disorders, and history of suicide attempts were more likely to be found in the suicide by police cases. Alcohol intoxication during the incident was also more likely to occur in this group. These findings are broadly consistent with other suicide by police studies (e.g., Hutson et al., 1998; Lord, 2000; Lord & Sloop, 2010; Mohandie et al., 2009) and general suicide studies (e.g., Bostwick & Pankratz, 2000; Kessler et al., 2005; Palmer et al., 2005). Furthermore, the majority of those who met the criteria for suicide by police were experiencing a range of stressful life events at the time, a finding also consistent with other studies (Hutson et al., 1998; Lord, 2000; Mohandie et al., 2009).

In addition, those cases that met the suicide by police criteria were found to be more likely than those who did not meet the criteria to have planned the incident to ensure police attention, threatened officers with weapons, advanced at officers, and refused to drop their weapon. The frequencies of the vast majority of variables on Lindsay and Lester’s (2008) criteria seem comparable to their recent findings; however, some observed differences may be the result of the differences in methodology between the two studies.

The findings concerning estimates of suicide by police in this sample of fatalities suggest a number of important implications. First, our study suggests that the phenomenon is as relevant in police fatal shootings outside the United Kingdom and North America. Second, as Klinger (2001) argued, the presence of suicidality in police fatal incidents affects their key defining characteristics and progression. Indeed, in those instances where the time was available and the police used a range of nonlethal options, the fatality nevertheless occurred. Third, the cases that met this strictly operationalized criteria differed significantly in their psychiatric characteristics from other cases, findings consistent with both suicide by police studies and general suicide studies. Finally, these findings suggest not only that police training on use of force must regularly accommodate issues of how to effectively engage, communicate, and manage people in acute emotional and mental crisis who may also be suicidal, but also that the mental health system ought to consider targeting the same issues through advancing effective partnerships with the police to further education, training, improved referral pathways, and information sharing. Similarly, proper recording and investigation of the circumstances surrounding these incidents, especially in the cases of attempted suicide by police is likely to increase the recognition and understanding of such incidents (see, e.g., Pinizzotto, Davis, & Miller, 2005, for an explanation of thorough investigative processes). Consistent and targeted approaches by the mental health system and the police are likely to increase positive outcomes for people who are experiencing mental disorders and who are in regular contact with the police. Lastly, research on police use of force should always consider the effects of the suspect’s mental disorders and suicidality and aim to achieve greater consensus in how both are defined and measured.

Conclusions

The current study described the situational variables most characteristic of fatal police force incidents in Victoria, Australia, between 1980 and 2008. These findings suggest that the most common reason for police involvement was for the purpose of apprehension, where the majority of the incidents escalated rapidly and were unplanned police operations. Police used a range of options to manage these situations; however, in the vast majority of the cases the decedents were armed, resisted the arrest, and threatened the police with weapons, resulting in police using lethal force to protect themselves and others.

Closer analysis of the motivations for these incidents revealed that in a third of the cases there was evidence to suggest that the person wished to commit suicide by police. These cases appeared significantly different in a number of ways than those not meeting the criteria; namely, mental disorders such as anxiety, schizophrenia, and personality disorders and past suicide attempts were overrepresented in those cases that met the criteria for suicide by police. These decedents were more likely than those not meeting the criteria to have planned the incident to get the police to come to them, to threaten officers with a weapon, to advance at the police instead of surrender, and to refuse to drop their weapon. These incidents were also more likely to last longer, with the police using negotiation, tactical disengagement, and cordon and containment of the scene significantly more here than in those cases not meeting the criteria for suicide by police.

This study presented a more complex picture of fatal police engagements. Further research is necessary to analyze other critical police incidents that are successfully resolved as those analyses are likely to validate and give a more comprehensive picture of successful police tactics in resolving critical incidents, particularly those that involve people who are in emotional distress, who may be experiencing a mental disorder, and who appear as though they wish to commit suicide. A more standardized and proactive approach to dealing with people experiencing mental disorders and who are in emotional distress is required to minimize the likelihood of these fatalities.

Footnotes

Notes

Dragana Kesic is a registered clinical forensic psychologist and a research fellow at the Centre for Forensic Behavioural Science, Monash University. Her main research area of interest is examining the issues present at the interface between police and people experiencing mental disorders.

Stuart D. M. Thomas is a senior lecturer at Monash University and deputy director of the Centre for Forensic Behavioural Science. His research interests include policing vulnerable populations, addictions and mental health, outcome measurement, and problem gambling.

James R. P. Ogloff is a professor of clinical forensic psychology at Monash University, director of the Centre for Forensic Behavioural Science, and director of psychological services at the Victorian Institute of Forensic Mental Health. He specializes in the areas of mentally ill offenders, the assessment and treatment of offenders, and medico-legal matters.

References

American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders (4th ed., text revision). Washington, DC: Author.

Australian Bureau of Statistics. (2010). Australian demographic statistics (Cat. no. 3101.0). Retrieved from http://www.abs.gov.au/ausstats/abs@.nsf/mf/3101.0/

Ballek

D. L.

(1998). Upholding the right: An historical analysis of Victoria Police involved fatal encounters 1853–1995 (Unpublished doctoral dissertation). University of Southern California, Los Angeles.

Best

Quigley

(2003). Shootings by the police: What predicts when a firearms officer in England and Wales will pull the trigger? Policing and Society, 13, 349-364.

Best

Quigley

Bailey

(2004). Police shootings as a way of self-harming: A review of the evidence for “suicide by cop” in England and Wales between 1998 and 2001. International Journal of the Sociology of Law, 32, 349-361.

Bostwick

J. M.

Pankratz

(2000). Affective disorders and suicide risk: A reexamination. American Journal of Psychiatry, 157, 1925-1932.

Brouwer

G. E.

(2005). Review of fatal shootings by Victoria Police. Melbourne, Australia: Office of Police Integrity.

10.

Crawford

Burns

(2008). Police use of force: Assessing the impact of time and space. Policing and Society, 18, 322-335.

11.

12.

De Leo

Bertolote

Lester

(2002). Self-directed violence. In Krug

E. G.

Dahlberg

L. L.

Mercy

J. A.

Zawi

A. B.

Lozano

(Eds.), World report on violence and health (pp. 185-212). Geneva, Switzerland: World Health Organization.

13.

De Leo

Cerin

Spathonis

Burgis

(2005). Lifetime risk of suicide ideation and attempts in an Australian community: Prevalence, suicidal process, and help-seeking behaviour. Journal of Affective Disorders, 86, 215-224.

14.

Fazel

Danesh

(2002). Serious mental disorder in 23000 prisoners: A systematic review of 62 surveys. Lancet, 359, 545-550.

15.

FitzGerald

S. C.

Bromley

M. L.

(1998). Surviving deadly force encounters: A case study. Journal of Police and Criminal Psychology, 13, 25-35.

16.

Gill

J. R.

Pasquale-Styles

(2009). Firearm deaths by law enforcement Journal of Forensic Sciences, 54, 185-188.

17.

Girodo

(2007). Personality and cognitive processes in life and death decision making: An exploration into the source of judgment errors by police special squads. International Journal of Psychology, 42, 418-426.

18.

Harris

Barraclough

(1997). Suicide as an outcome for mental disorders: A meta-analysis. British Journal of Psychiatry, 170, 205-228.

19.

Homant

R. J.

Kennedy

D. B.

(2000a). Effectiveness of less than lethal force in suicide-by-cop incidents. Police Quarterly, 3, 153-171.

20.

Homant

R. J.

Kennedy

D. B.

(2000b). Suicide by police: A proposed typology of law enforcement officer assisted suicide. Policing, 23, 339-355.

21.

Hutson

H. R.

Anglin

Yarabrough

Hardaway

Russell

Strote

. . . Blum

(1998). Suicide by cop. Annals of Emergency Medicine 32, 665-669.

22.

Jacobs

O’Brien

R. M.

(1998). The determinants of deadly force: A structural analysis of police violence. American Journal of Sociology, 103, 837-862.

23.

Kennedy

D. B.

Homant

R. J.

Thomas

H. R.

(1998). Suicide by cop. FBI Law Enforcement Bulletin, 67, 21-27.

24.

Kesic

Thomas

S. D. M.

Ogloff

J. R. P.

(2010). Mental illness among police fatalities in Victoria 1982–2007: Case linkage study. Australian and New Zealand Journal of Psychiatry, 44, 436-468.

25.

Kessler

R. C.

Berglund

Borges

Nock

Wang

P. S.

(2005). Trends in suicide ideation, plans, gestures, and attempts in the United States, 1990–1992 to 2001–2003. Journal of the American Medical Association, 293, 2487-2495.

26.

Klinger

D. A.

(2001). Suicidal intent in victim-precipitated homicide: Insights from the study of “suicide-by-cop.” Homicide Studies, 5, 206-226.

27.

Klinger

D. A.

(2005). Social theory and the street cop: The case of deadly force. Issues in American Policing, 7, 1-15.

28.

Klinger

Brunson

(2009). Police officers’ perceptual distortions during lethal force situations: Informing the reasonableness standard. Criminology & Public Policy, 8, 117-139.

29.

Lindsay

(2001). Identifying the dynamics of suicide by cop. In Sheehan

D. C.

Warren

L. I.

(Eds.), Suicide and law enforcement (pp. 599-606). Retrieved from http://www.ncjrs.gov/App/Publications/abstract.aspx?ID=193580

30.

Lindsay

Lester

(2004). Suicide by cop: Committing suicide by provoking the police to shoot you. Amityville, NY: Baywood.

31.

Lindsay

Lester

(2008). Criteria for suicide-by-cop incidents. Psychological Reports, 102, 603-605.

32.

Lord

V. B.

(2000). Law enforcement-assisted suicide. Criminal Justice and Behavior, 27, 401-419.

33.

Lord

V. B.

Sloop

M. W.

(2010). Suicide by cop: Police shootings as a method of self harming. Journal of Criminal Justice, 38, 889-895.

34.

McElvain

J. P.

Kposowa

A. J.

(2008). Police officer characteristics and the likelihood of using deadly force. Criminal Justice and Behavior, 35, 505-521.

35.

McKenzie

I. K.

(2006). Forcing the police to open fire: A cross-cultural/international examination of police-involved, victim-provoked shootings. Journal of Police Crisis Negotiations, 6, 5-25.

36.

37.

Mohandie

Meloy

Collins

P. I.

(2009). Suicide by cop among officer-involved shooting cases. Journal of Forensic Sciences, 54, 456-462.

38.

Mumola

C. J.

(2007). Arrest-related deaths in the United States 2003–2005 (NCJ-219534). Washington, DC: U.S. Department of Justice.

39.

Palmer

B. A.

Pankratz

Bostwick

J. M.

(2005). The lifetime risk of suicide in schizophrenia: A reexamination. Archives of General Psychiatry, 62, 247-253.

40.

Parent

R. B.

(1996). Aspects of police use of deadly force in British Columbia: The phenomenon of victim-precipitated homicide (Master’s thesis). Retrieved from http://summit.sfu.ca/system/files/iritems1/7114/b18065880.pdf

41.

Parent

R. B.

Verdun-Jones

(1998). Victim-precipitated homicide: Police use of deadly force in British Columbia. Policing, 21, 432-448.

42.

Pinizzotto

A. J.

Davis

E. F.

Miller

C. E.

III. (2005). Suicide by cop: Defining a devastating dilemma. FBI Law Enforcement Bulletin, 74, 8-18.

43.

Pinizzotto

A. J.

Davis

E. F.

Miller

C. E.

III. (2007). The deadly mix: Officers, offenders and the circumstances that bring them together. FBI Law Enforcement Bulletin, 76, 1-10.

44.

Police Complaints Authority. (2003). Review of shootings by police in England and Wales from 1998–2001. London, UK: Author.

45.

Schneider

(2009). Substance use disorders and risk for completed suicide. Archives of Suicide Research, 13, 303-316.

46.

State Government of Victoria. (2010). Facts about Victoria. Retrieved from http://www.liveinvictoria.vic.gov.au/living-in-victoria/about-victoria/facts-about-victoria

47.

Victoria Police. (1996). Project Beacon: A synopsis. Melbourne, Australia: Author.

48.

Victoria Police. (2008). Victoria Police Manual—101-3: Operational safety and other equipment. Melbourne, Australia: Author.

49.

Victoria Police. (2009). Victoria Police annual report 2008/2009. Melbourne, Australia: Author.

50.

Wallace

Mullen

Burgess

Palmer

Ruschena

Browne

(1998). Serious criminal offending and mental disorder: Case linkage study. British Journal of Psychiatry, 172, 477-484.

51.

White

M. D.

(2001). Controlling police decisions to use deadly force: Reexamining the importance of administrative policy. Crime & Delinquency, 47, 131-151.

52.

White

M. D.

(2003). Examining the impact of external influences on police use of deadly force over time. Evaluation Review, 27, 50-78.

53.

Wilson

E. F.

Davis

J. H.

Bloom

J. D.

Batten

P. J.

Kamara

S. G.

(1998). Homicide or suicide: The killing of suicidal persons by law enforcement officers. Journal of Forensic Sciences, 43, 46-52.

54.

Wolfgang

M. E.

(1959). Suicide by means of victim-precipitated homicide. Journal of Clinical & Experimental Psychopathology, 20, 335-349.

55.

World Health Organization. (2010). Suicide prevention (SUPRE). Retrieved from http://www.who.int/mental_health/prevention/suicide/suicideprevent/en/