Australian Institute of Criminology

Skip to content

Deaths in police custody and custody-related operations

Deaths in police custody are divided into two categories:

Category 1: Deaths in institutional settings and deaths in police operations where officers were in close contact with the deceased.

Category 2: Deaths during custody-related police operations.

Full definitions and examples of the two categories are provided above (see Introduction Box 1). When interpreting data on police custody and custody-related operations, it is important to keep in mind that the number of police custody episodes/arrests for the reporting period is not known. Consequently, the calculation of rates of deaths in police custody and custody-related operations is not possible. Percentages have been provided in Tables; however, they should be interpreted with an understanding that they are in the context of overall low numbers and therefore, greater focus should be given to the counts provided in the number (‘n’) columns.

Trends in police custody deaths are best interpreted from 1990 onwards, as only Category 1a deaths (police institutional settings) were included in the collection prior to 1990; these are not directly comparable with the deaths recorded after 1990.

Category 1 and Category 2 deaths in police custody and custody-related operations are considered separately in the beginning of this section and in the cause of deaths section, in order to adequately address MVP deaths and police shootings. However, deaths in police custody and custody-related operations will generally be presented at an aggregate level.

2011–12 and 2012–13

During the 2011–12 and 2012–13 reporting period, there were 49 deaths in Australian police custody and custody-related operations; 31 occurred in 2011–12 and 18 in 2012–13. The marked change between the first and second year will be monitored in future reporting periods to determine if it is part of a trend or an anomaly.

Of the 49 deaths:

  • 22 (45%) were Category 1a and 1b deaths (close contact) and 27 (55%) were Category 2 (operational/detainment-related);
  • six deaths (12%) were of persons of Indigenous background and 43 (88%) were of persons of non-Indigenous background;
  • 47 (96%) were male and two (4%) were female; and
  • four (8%) were of persons less than 25 years of age, 25 (51%) were aged 25–39 years, 16 (33%) were aged 40–54 years and four (8%) were aged 55+ years.

Trend

The total number of deaths in police custody and custody-related operations, including those prior to 1990, are included in Table A13 and presented in Figure 10. When focusing on the deaths from 1989–90 onwards, there was a total of 750 deaths (46% Category 1, 54% Category 2). The frequency of deaths in each category have fluctuated over time, with Category 1 generally decreasing and Category 2 decreasing following peaks in 2003–04 and 2008–09. The two categories have begun to converge in the present reporting period, although the large decrease of deaths in 2012–13 may mean this convergence is temporary. The decrease in the number of police custody and custody-related operations in 2012–13 was largely in Category 1, where it dropped by 10 percent from 2011–12 and by 11 percent from 2010–11. This change will be monitored in the future to determine if it represents the emergence of a new trend or is an anomaly.

Table 10 Deaths in police custody and custody-related operations by jurisdiction and Indigenous status, 2011–12 and 2012–13 (n)
Indigenous Non-Indigenous Total
n % n % n %
2011–12
New South Wales 0 0 11 37.9 11 36
Victoria 0 0 3 10.3 3 10
Queensland 0 0 7 24.1 7 23
Western Australia 0 0 5 17.2 5 16
South Australia 0 0 3 10.3 3 10
Tasmania 0 0 0 0 0 0
Northern Territory 2 100 0 0 2 7
Australian Capital Territory 0 0 0 0 0 0
Total 2 100 29 100 31 100
2012–13
New South Wales 0 0 6 42.9 6 33
Victoria 0 0 2 14.3 2 11
Queensland 0 0 2 14.3 2 11
Western Australia 3 75 3 21.4 6 33
South Australia 1 25 0 0 1 6
Tasmania 0 0 1 7.1 1 6
Northern Territory 0 0 0 0 0 0
Australian Capital Territory 0 0 0 0 0 0
Total 4 100 14 100 18 100

Source: AIC NDICP 1979–2013 [computer file]

Figure 10 Deaths in police custody and custody-related operations, 1979–80 to 2012–13 (n)

Source: AIC NDICP 1979–2013 [computer file]

Demographic characteristics

Indigenous status

2011–12 and 2012–13

Of the 49 deaths that occurred in police custody and custody-related operations in 2011–12 and 2012–13, 12 percent (n=6) were of Indigenous persons (see Table 11). Two of the Indigenous deaths occurred in 2011–12 and four in 2012–13.

Trend

Of the 750 deaths in police custody and custody-related operations since 1989–90 to 2012–13, 20 percent (n=147) have been Indigenous detainees and 80 percent (n=603) non-Indigenous detainees. The number of Indigenous deaths increased between 2000–01 and 2004–05 and increased again (to a lesser extent) in 2007–08 and 2010–11.

Table 11 Deaths in police custody and custody-related operations by Indigenous status, sex, and age, 2011–12 and 2012–13 (n)
2011–12 2012–13 Total
Indigenous 2 4 6
Non-Indigenous 29 14 43
All persons 31 18 49
Sex
Male 30 17 47
Female 1 1 2
Total 31 18 49
Age
Less than 25 yrs 4 0 4
25–39 yrs 18 7 25
40–54 yrs 7 9 16
55+ yrs 2 2 4
Total 31 18 49

Source: AIC NDICP 1979–2013 [computer file]

Figure 11 Deaths in police custody and custody-related operations by Indigenous status, 1989–90 to 2012–13 (%)

Source: AIC NDICP 1979–2013 [computer file]

Sex

2011–12 and 2012–13

Of the 49 deaths in police custody and custody-related operations in 2011–12 and 2012–13, two (4%) were female and 47 (96%) were male. Of the two female deaths, one was of an Indigenous female (2%), while there were five (11%) Indigenous males of the 47 male deaths.

Trend

Male deaths in police custody and custody-related operations continue to outnumber female deaths, which is in accordance with the rate at which males are arrested compared with females. Since 1989–90, 93 percent (n=672) of deaths in custody have been male and seven percent (n=49) female. The number of female deaths in police custody and custody-related operations has always been low, reaching a maximum of five deaths on three occasions (1991–92, 1993–94 and 2001–02) and dropping to zero or one in 10 of the last 24 years. The average number of female deaths per year from 1989–90 to 2012–13 was 2.1, while the average for male deaths between these years was 29.2, which again reflects the rate at which males engage with police compared with females.

Age

2011–12 and 2012–13

In the reporting period of 2011–12 and 2012–13, the majority of deaths were within the 25–39 and 40–54 year age categories, with 25 (51%) deaths within the 25–39 year age group and 16 (33%) in the 40–54 year age group. There were four (8%) deaths of people aged younger than 25 years and four (8%) aged 55 years and older. The median age of Indigenous deaths in police custody and custody-related operations in 2011–12 and 2012–13 was 27.5 years, while the median age of non-Indigenous deaths was 41 years.

Figure 12 Deaths in police custody and custody-related operations by sex, 1989–90 to 2012–13 (%)

Source: AIC NDICP 1979–2013 [computer file]

Trend

The age group with the highest frequency of deaths in police custody and custody-related operations since reporting began in its full form in 1989–90 has generally been people aged 25–39 years. There have been fluctuations in the annual frequencies for the age group; however, this age group has accounted for 43 percent (n=324) of deaths in police custody and custody-related operations since 1989–90. The less than 25 year old age group has comprised 29 percent (n=219) of deaths since 1989–90 to present and has shown larger fluctuations in frequencies than other age groups. Since 2008–09, there has been a decrease in the number of deaths of individuals aged 25 or younger. The 40–54 year old age group accounts for 21 percent (n=154) of deaths in police custody and custody-related operations and has been gradually increasing since 2009–10. The 55 years and older age group has accounted for six percent (n=48) of deaths since 1989–90 and has remained relatively stable over time, with a small yet gradual increase occurring from 2005–06 to present.

Differences in the age distribution of Indigenous and non-Indigenous deaths in police custody and custody-related operations between 1989–91 and 2012–13 have been identified. Indigenous deaths were highest in the younger age categories, with 42 percent (n=58) aged less than 25 years old and 38 percent (n=53) aged 25–39 years. The 40–54 year old age group constituted 18 percent (n=25) of deaths, while two percent (n=3) was accounted for by the 55 years and older age group. Non-Indigenous deaths were also high in the younger age groups with 26 percent (n=152) of deaths in the less than 25 years category and 45 percent (n=261) in the 25–39 year group. However, unlike Indigenous deaths, the 40–54 year group was also high, accounting for 22 percent (n=128) of non-Indigenous deaths. The percentage of deaths in the 55 and over age group were also higher in non-Indigenous compared with Indigenous people (7%; n=42 and 2%; n=3, respectively).

Figure 13 Deaths in police custody and custody-related operations by age category, 1989–90 to 2012–13 (%)

Note: One case has been excluded due to missing data

Source: AIC NDICP 1979–2013 [computer file]

Cause of death

2011–12 and 2012–13

The largest proportion (n=21; 43%) of deaths in police custody and custody-related operations was caused by external/multiple trauma, such as MVPs and stab wounds. Gunshot wound was the next most common cause of death at 23 percent (n=11); five of the shooting deaths were police shootings and six were self-inflicted (see Shooting deaths). Multiple causes/other, which includes drowning and fire, accounted for six deaths (12%). There were three (6%) natural cause deaths and no hanging deaths. Head injury, drugs and alcohol were each identified as the cause of two deaths (4%) per category.

Of the six deaths of Indigenous persons, one resulted from a head injury, one as a result of external/multiple trauma, two from alcohol toxicity and two due to other/multiple causes.

Table 12 Deaths in police custody and custody-related operations by cause of death, 2011–12 and 2012–13 (n)
2011–12 2012–13 Total
Indigenous Non-Indigenous All persons Indigenous Non-Indigenous All persons n %
Hanging 0 0 0 0 0 0 0 0
Natural causes 0 1 1 0 2 2 3 6
Head injury 1 1 2 0 0 0 2 4
Gun shot 0 6 6 0 5 5 11 23
External/multiple trauma 0 15 15 1 5 6 21 43
Drugs 0 2 2 0 0 0 2 4
Alcohol 1 0 1 1 0 1 2 4
Other/multiple causes 0 3 3 2 1 3 6 12
Missing 0 1 1 0 1 1 2 4
Total 2 29 31 4 14 18 49 100

Source: AIC NDICP 1979–2013 [computer file]

Trend

The most common cause of death in police custody and custody-related operations since 1989–90, accounting for 33 percent (n=246) of all deaths, has been external and multiple trauma. In the early 1990s, gunshot wound was the most common cause of death in police custody and custody-related operations; however; over time, this has decreased slightly although it remains the second most common cause of death, comprising 28 percent (n=210) of deaths.

Hanging deaths have been decreasing since 1993–94 and no deaths have been attributed to this cause in the past three financial years (see Table A19). The present reporting period marks an increase in deaths due to drug and alcohol toxicity; however, it is too soon to identify it as an emerging trend. Causes of death in Indigenous people in police custody and custody-related operations have been largely due to external/multiple trauma (36%; n=53) and natural cause deaths (20%; n=30). Similarly, a high proportion of non-Indigenous deaths have been caused by external/multiple trauma (32%; n=193); however, unlike the Indigenous deaths, shooting has been a more common cause among non-Indigenous deaths (33%; n=197; see Table A19 for complete historical data).

Hanging deaths: Points and materials used

2011–12 and 2012–13

There were no hanging deaths in police custody and custody-related operations in the current reporting period.

Trend

Since 1989–90, there have been 56 hanging deaths in police custody settings, 15 (27%) of which were of Indigenous persons. Of the total 49 deaths, the following hanging points and materials were used.

Hanging points:

  • 16 (29%) involved cell bars;
  • nine (16%) involved other fittings in cells (such as vents);
  • eight (14%) involved fittings outside the cell (such as tree in yard);
  • five (9%) used shower fixtures;
  • three (5%) occurred in a police van;
  • seven (13%) were classified as ‘other’; and
  • eight (14%) were missing data.

Materials:

  • 10 (18%) used sheets;
  • six (11%) used shoelaces;
  • five (9%) used a belt;
  • 11 (20%) used rope/cord;
  • 15 (27%) used other clothing;
  • one (2%) was classified as ‘other’; and
  • eight (14%) were missing data.

Motor vehicle pursuit deaths

2011–12 and 2012–13

During the 2011–12 and 2012–13 financial years, there were 12 MVP deaths, eight in 2011–12 and four in 2012–13. This accounted for 25 percent of all police custody and custody-related deaths during the two financial years. Of the 12 deaths:

  • two were of Indigenous persons;
  • all were male;
  • three were aged less than 25 years, five were 25–39 years and four were 40–54 years;
  • two deaths were of people being pursued for violent offences and nine deaths resulted from a pursuit in relation to traffic offences;
  • median speed reached across all pursuits was 134 km/hr;
  • median pursuit time was 2.8 minutes; and
  • the leading cause of death was external/multiple trauma (n=11).

Figure 14 Motor vehicle pursuit deaths, 1989–90 to 2012–13 (n)

Source: AIC NDICP 1979–2013 [computer file]

Trend

There have been 199 deaths due to MVPs since 1989–90, which accounts for 27 percent of all deaths in police custody and custody-related operations. Of the 199 deaths:

  • 48 (24%) were of Indigenous persons;
  • 185 (93%) were male; and
  • 121 (61%) were aged less than 25 years, 63 (32%) were 25–39 years, 11 (6%) were 40–54 years and four (2%) aged 55 years and older.

In relation to most serious offences associated with MVP deaths, the majority are categorised as traffic offences (46%; n=92) and theft-related (39%; n=78). Other most serious offence categories have occurred at a much lower frequency than theft and traffic offences. Violent offences accounted for four percent (n=7), good order two percent (n=4), drug-related three percent (n=6), other two percent (n=3) and five percent of cases were excluded due to missing data (n=9).

As expected, the most frequent cause of death associated with MVPs was external/multiple trauma (n=156; 78%), followed by head injury (n=42; 21%) and other/multiple causes (n=1; 1%).

The maximum speed reached during pursuits was 220 km/hr, with a minimum of 30 kmh and an average speed of 137 kmh (SD=34), and a median speed of 134 kmh. The average pursuit time during 1989–90 to 2012–13 was 5.5 minutes (SD=9 minutes). The shortest pursuit was 0.06 minutes, while the longest pursuit was 81 minutes; the median pursuit was three minutes.

Shooting deaths

2011–12 and 2012–13

In 2011–12 and 2012–13, there were 11 deaths due to shootings, five (46%) were police shootings and six (54%) were self-inflicted gun shots (data by year can be seen in Table A19). All deaths were of males of non-Indigenous background, six (55%) of whom were aged 25–39 years, four (36%) were aged 40–54 years and one (9%) aged 55 years and older.

In relation to the offence associated with the 11 people who were shot, nine (82%) were violent offences, one (9%) was a good order offence and one (9%) was categorised as an ‘other’ offence. The deaths caused by gunshot occurred in public places (n=5; 45%), on private property (n=5; 45%) and one occurred in a custodial setting.

Trend

Since 1989–90, there have been 211 shooting deaths, of which 16 (8%) were of Indigenous persons. Self-inflicted shootings accounted for 99 (47%) of the deaths, while 110 (56%) were due to police shootings and one (1%) shooting was categorised as being carried by an ‘other’ official/person. The number of shooting deaths has fluctuated substantially since 1989–90, with police shootings having been the most common category of shooting death at some times and self-inflicted shootings being most common at other times.

Historically, the majority of offences associated with shooting deaths have been violent offences (n=139; 66%), followed by ‘other’ offences (n=22; 10%), theft-related offences (n=23; 11%), good order offences (n=17; 8%), drug-related (n=4; 2%) and traffic offences (n=3; 1%); three cases had missing data (1%).

The locations of deaths due to gunshot have largely been public places (n=84; 40%), as well as private property (n=80; 38%). Less frequent locations have been public hospitals (n=37; 18%), ‘other’ locations (n=7; 3%) and custodial settings (n=2; 1%). This pattern appears to accord with the locations at which police most frequently encounter perilous circumstances and/or individuals who are potentially violent.

Figure 15 Shooting deaths, 1989–90 to 2012–13 (n)

Source: AIC NDICP 1979–2013 [computer file]

Manner of death

2011–12 and 2012–13

Of the 49 deaths during the 2011–12 and 2012–13, 17 (35%) were self-inflicted deaths, 19 (39%) were considered accidents, three (6%) were due to natural causes, five (10%) were justifiable homicides, one (2%) was an unlawful homicide, two (4%) were classified as ‘other’ and in two (4%) cases, the manner of death was missing (this is often because a finding as to manner of death is not yet available).

Five of the six Indigenous deaths were classified as accidents and one was classified as ‘other’. Non-Indigenous deaths were largely in the self-inflicted and accident categories.

Trend

While the trend has fluctuated, since 1989–90 deaths within police custody and custody-related operations have been most commonly categorised as accidental (40%; n=296). Self-inflicted deaths have been the second most common manner of death (32%; n=238) and have been gradually decreasing over time, although the present reporting period marks a spike in frequency of this manner of death. Natural cause deaths and justifiable homicides have fluctuated since 1989–90 with no stable pattern identifiable; the categories accounting for 10 percent (n=78) and 14 percent (n=104) of deaths, respectively. The present reporting period marks a decrease in both categories, although in 2012–13, natural cause deaths increased slightly. Unlawful homicides and deaths falling within the ‘otherwise unspecified’ manner of death category have remained consistently low and account for two percent (n=15) of deaths each, with several years where deaths did not occur in these categories. For unlawful homicides (where the homicide was not legally justifiable, such as when a police officer shoots an armed offender in the line of duty), the pattern established since 1989–90 was continued in the present reporting period, while the ‘other’ category displayed a small increase.

Table 13 Deaths in police custody and custody-related operations by manner of death, 2011–12 and 2012–13 (n)
2011–12 2012–13 Total
Indigenous Non-Indigenous All persons Indigenous Non-Indigenous All persons n %
Self-inflicted 0 11 11 0 6 6 17 35
Natural causes 0 1 1 0 2 2 3 6
Justifiable homicide 0 4 4 0 1 1 5 10
Unlawful homicide 0 1 1 0 0 0 1 2
Accident 1 10 11 4 4 8 19 39
Other 1 1 2 0 0 0 2 4
Missing 0 1 1 0 1 1 2 4
Total 2 29 31 4 14 18 49 100

Source: AIC NDICP 1979–2013 [computer file]

Table 14 Deaths in police custody and custody-related operations by manner of death, 1989–90 to 2012–13 (n)
Indigenous Non-Indigenous Total
n % n % n %
Self-inflicted 28 19 210 35 238 32
Natural causes 31 21 47 8 78 10
Justifiable homicide 6 4 98 16 104 14
Unlawful homicide 8 5 7 1 15 2
Accident 69 47 227 38 296 40
Otherwise unspecified 5 3 7 1 15 2
Missing 0 0 4 0.7 5 0.7
Total 147 100 603 100 750 100

Source: AIC NDICP 1979–2013 [computer file]

Most serious offence

2011–12 and 2012–13

Of the 49 deaths in police custody and custody-related operations, the deceased was most likely to have been suspected of committing a most serious offence involving violence, with violent offences accounting for 41 percent (n=20) of cases during 2011–12 and 2012–13. Violent offences were far more prevalent than traffic offences, which was the next most common offence type (n=9, 18%), followed by ‘other’ offences (n=8, 16%). Good order offences were less frequent (n=5, 10%), as were theft-related offences (n=4, 8%). There were no drug offences associated with any deaths in police custody and custody-related operations during the reporting period.

The six Indigenous deaths in police custody and custody-related operations were associated with one of each of the following—most serious offences, good order, traffic, theft-related and ‘other’; in two cases the relevant data were missing.

Table 15 Deaths in police custody and custody-related operations by most serious offence 2011–12 and 2012–13 (n)
2011–12 2012–13 Total
Indigenous Non-Indigenous All persons Indigenous Non-Indigenous All persons n %
Violent 0 13 13 0 7 7 20 41
Theft-related 0 2 2 1 1 2 4 8
Drug-related 0 0 0 0 0 0 0 0
Good order 1 2 3 0 2 2 5 10
Traffic 1 7 8 0 1 1 9 18
Other 0 5 5 1 2 3 8 16
Missing 0 0 0 2 1 3 3 6
Total 2 29 31 4 14 18 49 100

Source: AIC NDICP 1979–2013 [computer file]

Trend

Although there has been some fluctuation over the years, since 1989–90, violent offences (32%, n=237) have constituted the largest proportion of most serious offences associated with policy custody deaths. Theft-related offences were associated with 21 percent (n=157) of all deaths, with a decrease in their frequency since 2001–02 that has continued into the 2011–12 and 2012–13 financial years.

Drug-related offences are not frequently associated with deaths in police custody and custody-related operations. Since 1989–90, they have been associated with between zero and three deaths per financial year, totalling three percent (n=20) over the entire period. There were no deaths associated with the offence category in the present reporting period.

Good order offences have been associated with 15 percent (n=114) of all deaths. In 2006–07, there was a drop in deaths associated with good order offences to zero, after which numbers increased and have remained at between two to four deaths per financial year.

Overall, traffic offences have been associated with 15 percent (n=114) of deaths since 1989–90 and have also fluctuated over time, with a bimodal distribution as frequency increased from 1999–00 to 2008–09. The frequency has remained higher from 2008–09 until 2011–12; however, there was a sharp decrease in 2012–13 down to only one death.

The ‘other’ offences category constitutes offences such as arson, as well as protective custody situations for intoxication and has been associated with 11 percent (n=86) of all deaths in police custody and custody-related operations. The frequency of deaths in this category has varied over time and as such, no clear pattern has been established.

Location of death

Location of death refers to the specific environment at the time of death; that is, the actual place where the death occurred.

2011–12 and 2012–13

Of the 49 deaths in custody during 2011–12 and 2012–13:

  • 28 (57%) occurred in public places;
  • 10 (20%) occurred on private property;
  • eight (16%) occurred in public hospitals;
Table 16 Deaths in police custody and custody-related operations by location of death, 2011–12 and 2012–13 (n)
2011–12 2012–13 Total
Indigenous Non-Indigenous All persons Indigenous Non-Indigenous All persons n %
Public hospital 0 6 6 0 2 2 8 16
Cell 1 0 1 1 0 1 2 4
Custodial setting 0 1 1 0 0 0 1 2
Private property 0 8 8 0 2 2 10 20
Public place 1 14 15 3 10 13 28 57
Other 0 0 0 0 0 0 0 0
Total 2 29 31 4 14 18 49 100

Source: AIC NDICP 1979–2013 [computer file]

  • two (4%) occurred in cells; and
  • one (2%) occurred in a custodial setting.

Of the six Indigenous deaths, two took place in police custody cells and four occurred in public places.

Trend

Of the 750 deaths in police custody and custody-related operations since 1989–90, most (n=321; 43%) have occurred in public places. Public hospitals have been the setting for approximately one-quarter (n=192; 26%) of all deaths over the last 22 years, while deaths located on private property have accounted for 15 percent (n=116). Deaths in cells and other police custody areas have been rare (see Table A21).

Circumstances of custodial period for deaths in police custody

NDICP examines the circumstances of the custodial period; that is, both the type of police custody the death occurred in, as well as the method used to detain the person. The four types of police custody and custody-related operations into which deaths are categorised are:

  • institution—death occurred in a police lock-up, prison, juvenile detention centre, or during transfer to or following transfer from one of these places to a hospital or other detention facility, or during transfer to a facility; for example, in a van etc;
  • escaping—death occurred during the process of the person escaping or attempting to escape from police or prison custody or juvenile detention;
  • detaining—death occurred during the process of police or prison officers attempting to detain the person, regardless of whether or not the person was under arrest or the officers intended to arrest the person; and
  • other/marginal cases—for example, the death of a person detained under a state Mental Health Act in the process of being conveyed via ambulance from a public hospital to a psychiatric institution under ‘police escort’.

The second aspect of custodial circumstances is method of detainment. This includes the following categories:

  • MVP;
  • other pursuit; for example, on foot;
  • sieges;
  • raids; and
  • ‘other’ methods; for example, shooting.

2011–12 and 2012–13

The majority (78%; n=38) of the 49 police custody and custody-related deaths during 2011–12 and 2012–13 occurred while police were attempting to detain an individual. Four (67%) of the six Indigenous deaths and 34 (79%) of the 43 non-Indigenous deaths were in the ‘attempting to detain’ category. The distribution of the remaining deaths in custody included seven (14%) deaths in the ‘other/marginal’ category, while four (8%) occurred in institutional settings. There were no deaths categorised as occurring during escaping.

The most frequent method of detainment resulting in a death in custody was MVPs (n=13, 27%); the raids category and the other/shooting category were each associated with nine deaths (18% each), followed by sieges (n=8, 16%) and other pursuits (n=4, 8%). Of the 49 cases, six (12%) were classified as detainment method ‘not applicable’ where the death occurred in police custody, for instance, in a police cell, rather than while police were attempting to detain the individual.

Trend

Deaths while attempting to detain have been the most common category associated with police custody and custody-related operations deaths since 1989–90, accounting for 73 percent (n=547) of deaths. While this is the most common category, numbers have generally been decreasing since 2008–09. Deaths in police institutional settings have accounted for 23 percent (n=171) of deaths and have been gradually declining since 2004–05. Deaths in the ‘other/marginal’ category and the escaping category have occurred infrequently since the 1990s. The present reporting period continues this pattern of low frequency, although the other/marginal category has displayed a small increase in 2011–12 to rise above the police institutional setting category. See Table A23 for historical data.

While the frequency with which particular methods of detainment have been associated with police custody deaths have varied year to year, MVPs have consistently been the most common over the last 22 years, constituting 23 percent (n=219) of deaths. The highest frequency of MVP deaths was 19 in 2001–02, although deaths associated with this method of detainment have been declining since then. The ‘other pursuits’ category has typically been the least frequently recorded method of detainment. See Table A22 for detailed historical data.