Australian Institute of Criminology

Skip to content

Deaths in police custody and custody-related operations

Key indicators of performance

The NDICP reports on the following three high-level indicators for deaths in police custody and custody-related operations:

  • trends in number of deaths police custody and custody-related operations, and the proportion of deaths involving Indigenous persons;
  • trends in rate–ratio for Indigenous and non-Indigenous deaths in police custody and custody-related operations; and
  • trends in causes and circumstances of deaths in police custody and custody-related operations.
Table 49 Key indicators of performance—police custody deaths, 2008–09
Police
1 Indigenous 8 (22%)
Non-Indigenous 28 (78%)
2 Rate–ratio—Indigenous 1:4.5
Rate–ratio—non-Indigenous 1:1.3
3a Main cause—Indigenous Head injury—(38%; n=3)
Main cause—non-Indigenous External/multiple trauma—(54%; n=15)
3b Main circumstances—Indigenous Accidental—(50%; n=4)
Main circumstances—non-Indigenous Accidental—(54%; n=15)

Source: AIC NDICP 2008–09 [computer file]

Table 50 Key indicators of performance—police custody deaths, 2009–10
Police
1 Indigenous 6 (22%)
Non-Indigenous 21 (78%)
2 Rate–ratio—Indigenous 1:4.5
Rate–ratio—non-Indigenous 1:1.3
3a Main cause—Indigenous External/multiple trauma—(33%; n=2)
Main cause—non-Indigenous Gunshot (33%; n=7)
3b Main circumstances—Indigenous Accidental—(67%; n=4)
Main circumstances—non-Indigenous Accidental—(43%; n=9)

Source: AIC NDICP 2009–10 [computer file]

Table 51 Key indicators of performance—police custody deaths, 2010–11
Police
1 Indigenous 7 (29%)
Non-Indigenous 17 (71%)
2 Rate—Indigenous 1:3.3
Rate—non-Indigenous 1:1.4
3a Main cause—Indigenous External/multiple trauma—(57%; n=4)
Main cause—non-Indigenous Gunshot—(47%; n=8)
3b Main circumstances—Indigenous Accidental—(57%; n=4)
Main circumstances—non-Indigenous Self-inflicted, justifiable homicide—(both 29%; n=5)

Source: AIC NDICP 2010–11 [computer file]

Introduction

The NDICP collects data on deaths occurring in police institutional settings, such as cells, watchhouses or divisional vans, as well as deaths occurring in police custody-related operations, such as motor vehicle pursuits, sieges, raids and shootings. Information relating to deaths in police custody between 1979–80 and 1988–89 were collected by the RCIADIC research team. Only deaths occurring in police institutional settings were collected over this period. A decision was made by the APMC in 1994 to expand the definition of a death in police custody to include deaths occurring during custody-related operations (see Appendix C). Since 1 January 1990, data have been collected on deaths in custody-related operations. For this reason, long-term trends in police custody deaths are best interpreted from 1989–90 onwards.

To afford greater analysis of the nature and circumstances of deaths occurring in police custody, these deaths are divided into two broad categories—Category 1 deaths (institutional settings, raids, shootings) and Category 2 deaths (sieges, police pursuits and other). These two categories are distinguished by the level of contact or influence that police officers could exert over the deceased. As such, Category 1 deaths are those close-contact incidents where the deceased could be contained in a precise location, they were unable to take evasive action and/or the deceased could be seen and spoken to. Category 2 deaths are those in which the officers did not have such close contact as to be able to physically control the deceased’s behaviour, the deceased was able to take evasive action, or officers were unable to speak with the deceased.

Important consideration

When interpreting the results presented in this section, it is important to remember that policing is a very difficult and dangerous business. Police officers are required to respond to thousands of very challenging situations each year, many involving individuals who are affected by alcohol, drugs, or a combination of both. In some cases, police officers are required to respond to persons suffering from a mental illness, threatening harm to themselves or other people, or individuals experiencing significant personal crisis. In all of these situations, police officers are required to make split-second decisions in what are often very stressful and sometimes life-threatening situations. In this regard, the former NSW State Coroner John Abernethy made an important observation in one particular inquest

…hindsight is a wonderful thing but none of the actors in this tragic drama that night had the benefit of hindsight. Decisions were made which, with hindsight, might not have been made (Inquest 1887/04, Newcastle Coroners Court, Abernethy J 2006).

It should not be overlooked that there are deaths in police custody and custody-related operations where the officers involved made every possible attempt to prevent the death occurring, but unfortunately were not successful. There are instances where officers have shown exceptional bravery, even risking their lives, in attempting to prevent the death of someone in their custody. There have also been cases where police officers have faced legal action from the deceased’s family for not arresting a person threatening self-harm (Freckelton 2008). From all available evidence, it is difficult to conclude that police officers perform their duties with anything but the best intentions.

There were a total of 87 deaths occurring in police custody and custody-related operations over the reporting period, with 36 deaths recorded in 2008–09, 27 deaths in 2009–10 and 24 in the most recent financial year (see Table 52). The number recorded in 2008–09 is at the upper end of annual totals, while totals recorded in both 2009–10 and 2010–11 are in line with levels seen since 1999–2000.

Table 52 Deaths in police custody and custody-related operations by jurisdiction, 2008–09 to 2010–11 (n)
2008–09 2009–10 2010–11
New South Wales 6 5 5
Victoria 3 6 1
Queensland 8 6 7
Western Australia 8 3 6
South Australia 6 1 2
Tasmania 0 1 2
Northern Territory 5 4 10
Australian Capital Territory 0 1 1
Total 36 27 24

Source: AIC NDICP 1989–90 to 2010–11 [computer file]

When comparing recorded deaths in police custody across different states and territories, it can be seen that generally, those jurisdictions with the larger populations tend to have higher numbers of deaths in police custody and custody-related operations. However, over the reporting period, Queensland has consistently had the most deaths in police custody and custody-related operations (24%; n=21), despite having a smaller population than New South Wales and Victoria. The proportion of deaths in Queensland over the last three years is also higher when compared with historical proportions over the last 32 years (24% between 2008–09 to 2010–11 cf 17% between 1979–80 and 2010–11).

It is encouraging to note that both New South Wales and Victoria (jurisdictions that have traditionally had relatively high numbers of police custody deaths each year) have both experienced a decline in the number of deaths compared with historical totals. Recorded cases in recent years represent some of the lowest historical totals seen in these two jurisdictions.

Trends

The total number of deaths occurring each year since 1979–80 in police custody and custody-related operations (Categories 1 and 2) have fluctuated considerably, ranging from a low of five deaths in the first year, to a high of 42 deaths in 1986–87, 2001–02 and 2003–04. Following the last peak in 2003–04, there has been a modest decline in police custody deaths.

There is a noticeable difference in the long-term trends between Category 1 and Category 2 deaths since 1979–80. The numbers of deaths in police custody each year since 1979–80, disaggregated by the different categories of police custody deaths, are presented in Table 53. The concerns expressed by the RCIADIC regarding the increasing numbers of deaths in police institutional settings between 1980 and 1987 can be seen in the first column of data. It can be seen that Category 1 deaths (institutional settings, raids, shootings) have declined since 1989–90, while Category 2 deaths (sieges, police pursuits) increased considerably between 1989–90 and 2003–04, but have been consistently in decline ever since (see Figure 23). Overall, Category 2 deaths were more common than Category 1 deaths each year between 1999–2000 and 2009–10, with the number recorded in 2010–11 possibly representing a return to trends seen in the 1990s. However, the total number of deaths in police custody and custody-related operations has been in decline since 2000–01, a trend that was confirmed using third-order polynomial regression.

Analysis of the data showed that throughout the 1990s, deaths in police custody and custody-related operations were more likely to occur in close-contact settings—situations where the deceased was confined to a particular area and their behaviour could be influenced or controlled by police. Typically, such deaths include those in institutional settings or in close proximity to police, such as in raids or shootings. However, since the late 1990s, there has been a decline in such deaths while at the same time, deaths in situations where police were in the process of detaining the deceased but were not in close contact (such as in motor vehicle pursuits and sieges) have become the more common type of death in police custody.

Figure 23 Deaths in police custody and custody-related operations, 1979–80 to 2010–11 (n)

Source: AIC NDICP 1979–80 to 2010–11 [computer file]

Table 53 Deaths in police custody and custody-related operations, 1979–80 to 2010–11 (n)
Category 1 Category 2 Categories 1 and 2
1979–80 5 0 5
1980–81 12 0 12
1981–82 15 0 15
1982–83 22 0 22
1983–84 14 0 14
1984–85 20 1 21
1985–86 20 0 20
1986–87 42 0 42
1987–88 23 0 23
1988–89 29 0 29
1989–90 23 6 29
1990–91 24 2 26
1991–92 19 6 25
1992–93 20 18 38
1993–94 19 12 31
1994–95 18 12 30
1995–96 12 19 31
1996–97 13 21 34
1997–98 16 12 28
1998–99 14 7 21
1999–2000 14 22 36
2000–01 11 23 34
2001–02 13 29 42
2002–03 12 29 41
2003–04 11 31 42
2004–05 12 24 36
2005–06 6 18 24
2006–07 13 18 31
2007–08 13 21 34
2008–09 11 25 36
2009–10 11 16 27
2010–11 14 10 24
Total 521 382 903

Source: AIC NDICP 1989–90 to 2010–11 [computer file]

Demographic characteristics

Indigenous status

Reporting period: 2008–09 to 2010–11

During the reporting period, just under one in four deaths (24%; n=21) in police custody and custody-related operations was of an Indigenous person (see Table 54). Overall, the number of Indigenous deaths in this period showed an increase from the previous three years (2005–06 to 2007–08; n=15). Deaths of non-Indigenous persons exhibited a different trend over the reporting period (n=66), with annual totals being lower than those recorded in the early part of the decade starting in the year 2000.

Over the last three years, more Indigenous deaths were recorded in Western Australia (38%; n=8) than in any other jurisdiction. This finding is consistent with long-term trends, which show that of the 203 total Indigenous deaths in police custody in Australia since 1979–80, almost one in three (30%, n=61) occurred in Western Australia. However, it must be noted that Western Australia has a higher proportion of Indigenous Australians than the national proportion (3.3% cf 2.5%; ABS 2011b, 2009a).

Table 54 Deaths in police custody and custody-related operations by Indigenous status and jurisdiction, 2008–09 to 2010–11 (n)
2008–09
Indigenous Non-Indigenous Persons
n n Total n
New South Wales 0 6 6
Victoria 0 3 3
Queensland 1 7 8
Western Australia 1 7 8
South Australia 2 4 6
Tasmania 0 0 0
Northern Territory 4 1 5
Australian Capital Territory 0 0 0
Australia 8 28 36
2009–10
Indigenous Non-Indigenous Persons
n n Total n
New South Wales 2 3 5
Victoria 0 6 6
Queensland 0 6 6
Western Australia 2 1 3
South Australia 0 1 1
Tasmania 0 1 1
Northern Territory 2 2 4
Australian Capital Territory 0 1 1
Australia 6 21 27
2010–11
Indigenous Non-Indigenous Persons
n n Total n
New South Wales 1 4 5
Victoria 0 1 1
Queensland 1 6 7
Western Australia 5 1 6
South Australia 0 2 2
Tasmania 0 2 2
Northern Territory 0 0 10
Australian Capital Territory 0 1 1
Australia 7 17 24

Source: AIC NDICP 2008–09 to 2010–11 [computer file]

Trends

With the exception of 1980–81, non-Indigenous deaths have consistently outnumbered Indigenous deaths each year since 1979–80 (see Table 55); however, the proportion of deaths each year involving Indigenous persons has fluctuated, ranging from between one in three to less than one in 10 (see Figure 24). Note, due to the current absence of a reliable denominator (ie the number of police custody incidents each year), rates of deaths in police custody cannot be calculated.

Table 55 Deaths in police custody and custody-related operations by Indigenous status and year, 1979–80 to 2010–11 (n)
Indigenous Non-Indigenous Total
1979–80 1 4 5
1980–81 7 5 12
1981–82 2 13 15
1982–83 5 17 22
1983–84 3 11 14
1984–85 8 13 21
1985–86 5 15 20
1986–87 15 27 42
1987–88 6 17 23
1988–89 10 19 29
1989–90 9 20 29
1990–91 3 23 26
1991–92 5 20 25
1992–93 5 33 38
1993–94 3 28 31
1994–95 3 27 30
1995–96 6 25 31
1996–97 8 26 34
1997–98 5 23 28
1998–99 7 14 21
1999–2000 4 32 36
2000–01 7 27 34
2001–02 8 34 42
2002–03 11 30 41
2003–04 10 32 42
2004–05 11 25 36
2005–06 7 17 24
2006–07 3 28 31
2007–08 5 29 34
2008–09 8 28 36
2009–10 6 21 27
2010–11 7 17 24
Total 203 700 903

Source: AIC NDICP 1979–80 to 2010–11 [computer file]

Figure 24 Deaths in police custody and custody-related operations by Indigenous status and year, 1979–80 to 2010–11 (%)

Source: AIC NDICP 1979–80 to 2010–11 [computer file]

Sex

The majority (93%; n=841) of all deaths in police custody and custody-related operations involve males and this trend remained consistent over the reporting period, with only one female death being recorded in the last three years (see Table 56). Of the Indigenous deaths in police custody over this period (n=21), one involved a female while the rest were males.

Trends

Deaths involving males have always been higher in number each year since 1979–80 (see Table 56). Overall, deaths of females have accounted for slightly over one in 14 deaths in police custody and custody-related operations (7%; n=62). With the exception of 1988–89 (1 in 6), 1991–92 (1 in 5), 1993–94 (1 in 6) and 2001–02 (1 in 8), the proportion of deaths involving females each year has followed this overall trend.

Table 56 Deaths in police custody and custody-related operations by sex and year, 1979–80 to 2010–11 (n)
Male Female Total
1979–80 5 0 5
1980–81 10 2 12
1981–82 15 0 15
1982–83 19 3 22
1983–84 14 0 14
1984–85 21 0 21
1985–86 19 1 20
1986–87 41 1 42
1987–88 23 0 23
1988–89 24 5 29
1989–90 27 2 29
1990–91 25 1 26
1991–92 20 5 25
1992–93 34 4 38
1993–94 26 5 31
1994–95 28 2 30
1995–96 30 1 31
1996–97 34 0 34
1997–98 25 3 28
1998–99 18 3 21
1999–2000 36 0 36
2000–01 33 1 34
2001–02 37 5 42
2002–03 37 4 41
2003–04 40 2 42
2004–05 33 3 36
2005–06 22 2 24
2006–07 27 4 31
2007–08 32 2 34
2008–09 36 0 36
2009–10 26 1 27
2010–11 24 0 24
Total 841 62 903

Source: AIC NDICP 1979–80 to 2010–11 [computer file]

Age

More deaths occurred over the reporting period among persons aged 25–36 years than in any other age group (41%; n=36). Deaths of persons aged less than 25 years were the second most prevalent (32%; n=28), followed by persons aged 40–54 years (18%; n=16; see Table 57). The distribution of deaths across the jurisdictions followed this general trend. Over the reporting period, Indigenous persons were found to die at younger ages than their non-Indigenous counterparts, with the biggest difference in age seen in 2009–10 when the median age at death was almost nine years younger than for non-Indigenous persons. Over this period, only three deaths (14%) of an Indigenous person 40 years or older was recorded, compared with 20 deaths (30%) for the equivalent non-Indigenous cohort.

Table 57 Deaths in police custody and custody-related operations by age in years, 2008–09 (n)
2008–09
Less than 25 yrs 25–39 yrs 40–54 yrs 55 yrs+ Median age at death Total
New South Wales 2 1 3 0 33 6
Victoria 1 0 0 2 55 3
Queensland 3 4 1 0 32.5 8
Western Australia 2 6 0 0 25.5 8
South Australia 3 1 1 1 27 6
Tasmania 0 0 0 0 0
Northern Territory 2 3 0 0 28 5
Australian Capital Territory 0 0 0 0 0
Totals
Indigenous 6 2 0 0 23 8
Non-Indigenous 7 13 5 3 28.5 28
All persons 13 15 5 3 26.5 36
2009–10
Less than 25 yrs 25–39 yrs 40–54 yrs 55 yrs+ Median age at death Total
New South Wales 2 2 1 0 34 5
Victoria 1 3 2 0 31 6
Queensland 0 2 2 2 41.5 6
Western Australia 1 2 0 0 32 4
South Australia 0 1 0 0 32 1
Tasmania 1 0 0 0 16 1
Northern Territory 3 1 0 0 23.5 4
Australian Capital Territory 1 0 0 0 23 1
Totals
Indigenous 3 3 0 0 24.5 6
Non-Indigenous 6 8 5 2 33 21
All persons 9 11 5 2 32 27
2010–11
Less than 25 yrs 25–39 yrs 40–54 yrs 55 yrs+ Median age at death Total
New South Wales 1 2 1 1 35 5
Victoria 0 21 0 0 31 1
Queensland 1 3 2 1 34 7
Western Australia 3 1 2 0 24.5 6
South Australia 0 2 0 0 33 2
Tasmania 1 0 1 0 33 2
Northern Territory 10 0 0 0 - 0
Australian Capital Territory 0 1 0 0 27 1
Totals
Indigenous 3 1 3 0 28 7
Non-Indigenous 3 9 3 2 32 17
All persons 6 10 6 2 31.5 24

Source: AIC NDICP 2008–09 to 2010–11 [computer file]

Trends

Deaths recorded over the last three years have remained in line with long-term trends, in that since 1979–80, deaths of persons aged between 25 and 39 years each year have generally exceeded deaths of persons in all other age groups and comprise 42 percent (n=377) of all deaths (see Table 58 and Figure 25). Also of note is the proportion of deaths involving persons aged less than 25 years (29%; n=265). Conversely, deaths of persons aged 55 years and over have remained consistently low each year and represent eight percent (n=68) of all deaths that have occurred since 1979–80.

Of all deaths of Indigenous persons (n=203) in police custody and custody-related operations since 1979–80 (see Figure 25):

  • 38 percent (n=78) were aged less than 25 years;
  • 39 percent (n=79) were aged between 25 and 39 years;
  • 18 percent (n=37) were aged between 40 and 54 years; and
  • 4 percent (n=9) were aged 55 years or older.

Of all deaths of non-Indigenous persons (n=700) in police custody and custody-related operations since 1979–80 (see Figure 25):

  • 27 percent (n=187) were aged less than 25 years;
  • 43 percent (n=298) were aged between 25 and 39 years;
  • 22 percent (n=154) were aged between 40 and 54 years;
  • 8 percent (n=59) were 55 years or older; and
  • 0.3 percent (n=2) of non-Indigenous deaths have missing information about age at death.

The data presented here indicate that a greater proportion (38%) of Indigenous deaths in police custody and custody-related operations involve persons aged less than 25 years, compared with their non-Indigenous counterparts (27%). Put simply, Indigenous people who die in police custody or related operations are, on average, younger than their non-Indigenous counterparts.

The fact that greater proportions of Indigenous young people are dying in police custody than young non-Indigenous people is most likely an unfortunate product of the increased contact between Indigenous young people and the police, in addition to contact occurring at an earlier age than for non-Indigenous young people. Research from the AIC’s Drug Use Monitoring Program has shown that

Indigenous male police detainees were, on average, first arrested at a much younger age than non-Indigenous male detainees (14 years compared with 19 years) (Putt, Payne & Milner 2005: 4).

Indigenous people who end up in prison have also been shown to begin regularly offending at younger ages (Makkai & Payne 2003). Of most concern is evidence from a recent study published by the AIC that indicates that once young Indigenous people come into contact with police, they are less likely to be given the diversionary alternatives such as cautions and conferencing that are designed to keep them out of the formal criminal justice system (Allard et al 2010: 4). More specifically, this study found that

Indigenous young people were 2.9 times less likely than non-Indigenous young people to be cautioned compared to going to court, two times less likely to undergo conferencing with police compared to going to court and 1.5 times less likely to be cautioned compared to undergoing police referred conferencing’ (Allard et al. 2010: 4).

Finally, a recent AIC report using apprehension data from Western Australia and South Australia Police from 2005 showed that for violent offences, the Indigenous apprehension rate per 1,000 relevant persons over 10 years of age was 22 times (for males) and 17 times higher (for females) respectively when compared with the equivalent rate for the non-Indigenous population (Wundersitz 2010). For a comprehensive overview of Indigenous offending, Dr Troy Allard provides a useful illumination of the statistics and research on this topic (Allard 2010).

Table 58 Deaths in police custody and custody-related operations by age in years, 1979–80 to 2010–11 (n)
Less than 25 yrs 25–39 yrs 40–54 yrs 55 yrs+ Total
1979–80 1 3 1 0 5
1980–81 5 0 2 5 12
1981–82 4 4 6 1 15
1982–83 6 7 7 2 22
1983–84 4 3 3 4 14
1984–85 2 8 6 4 20
1985–86 4 10 4 2 20
1986–87 11 16 12 3 42
1987–88 6 9 5 3 23
1988–89 8 19 1 1 29
1989–90 10 10 6 3 29
1990–91 9 14 2 1 26
1991–92 6 10 6 3 25
1992–93 14 9 12 3 38
1993–94 9 13 7 2 31
1994–95 10 11 9 0 30
1995–96 13 10 5 3 31
1996–97 11 14 5 4 34
1997–98 7 16 4 1 28
1998–99 7 10 2 1 20
1999–2000 8 21 6 1 36
2000–01 13 11 7 3 34
2001–02 16 15 9 2 42
2002–03 13 17 9 2 41
2003–04 12 20 9 1 42
2004–05 10 18 7 1 36
2005–06 6 13 4 1 24
2006–07 5 17 7 2 31
2007–08 7 13 12 2 34
2008–09 13 15 5 3 36
2009–10 9 10 5 2 27
2010–11 6 17 6 2 24
Total 265 377 191 68 901

a: 2 cases have been excluded due to missing information about the deceased’s age at death

Source: AIC NDICP 1979–90 to 2010–11 [computer file]

Figure 25 Deaths in police custody and custody-related operation by age and Indigenous status, 1979–80 to 2010–11 (%)

Figure 26 Deaths in police custody and custody-related operation by cause of death and Indigenous status, 1989–90 to 2010–11 (%)

Circumstances surrounding death

Cause of death

The most common cause of death over the reporting period was external/multiple trauma (40%; n=34; see Table 59), with three-quarters (77%; n=26) resulting from motor vehicle pursuit accidents. The second most common cause of death was gunshots (31%; n=27), with 48 percent (n=13) being persons who shot themselves and 52 percent (n=14) being persons shot by police. The most encouraging finding over the reporting period was the very low number of hanging deaths that occurred in police custody (2%; n=2), which represents a marked decline from historical proportions.

Indigenous deaths in police custody and custody-related operations exhibited a slightly different distribution to that seen for all deaths. While external/multiple trauma was the most prevalent cause of death (40%; n=8). Proportionally, more Indigenous deaths resulted from natural causes (15%; n=3 cf 9%; n=6) and head injuries (20%; n=4 cf 5%; n=3) when compared with non-Indigenous persons over the same period. Non-Indigenous persons were more than twice as likely to die from gunshots as Indigenous persons (36%; n=24 cf 15%; n=3) over the last three financial years.

Table 59 Deaths in police custody and custody-related operations by cause of death, 2008–09 (n)
2008–09
Hanging Natural causes Head injury Gunshot External/multiple trauma Drug-related Other/multiple Total
New South Wales 0 1 1 3 1 0 0 6
Victoria 0 0 0 2 1 0 0 3
Queensland 0 0 0 1 6 0 1 8
Western Australia 0 0 1 1 6 0 0 8
South Australia 1 0 0 2 3 0 0 6
Tasmania 0 0 0 0 0 0 0 0
Northern Territory 0 1 2 1 0 1 0 5
Australian Capital Territory 0 0 0 0 0 0 0 0
Totals
Indigenous 1 1 3 1 2 0 0 8
Non-Indigenous 0 1 1 9 15 1 1 28
All persons 1 2 4 10 17 1 1 36
2009–10
Hanging Natural causes Head injury Gunshot External/multiple trauma Drug-related Other Total
New South Wales 0 0 1 4 0 0 0 5
Victoria 1 2 0 1 1 0 1 6
Queensland 0 0 0 2 2 1 1 6
Western Australia 0 1 0 0 2 0 0 3
South Australia 0 0 0 1 0 0 0 1
Tasmania 0 0 1 0 0 0 0 1
Northern Territory 0 0 1 0 2 0 1 4
Australian Capital Territory 0 0 0 0 1 0 0 1
Totals
Indigenous 0 1 1 1 2 0 1 6
Non-Indigenous 1 2 2 7 6 1 2 21
All persons 1 3 3 8 8 1 3 27
2010–11
Hanging Natural causes Head injury Gunshot External/multiple trauma Drug-related Other Total
New South Wales 0 1 0 3 0 1 0 5
Victoria 0 0 0 1 0 0 0 1
Queensland 0 2 0 0 4 0 0 6
Western Australia 0 1 0 1 4 0 0 6
South Australia 0 0 0 2 0 0 0 2
Tasmania 0 0 0 1 1 0 0 2
Northern Territory 0 0 0 0 0 0 0 0
Australian Capital Territory 0 0 0 1 0 0 0 1
Totals
Indigenous 0 1 0 1 4 0 0 6
Non-Indigenous 0 3 0 8 5 1 0 17
All personsa 0 4 0 9 9 1 0 24

a: 1 case has been excluded because the cause of death is still under investigation

Source: AIC NDICP 1989–90 to 2010–11 [computer file]

Trends

As noted above, information about cause of death is best interpreted from 1989–90 onwards, as this is when data on deaths in police custody-related operations were first included in the NDICP data collection. The findings from the reporting period are in line with historical trends (see Table 60). Since 1989–90, the most common cause of death in police custody and custody-related operations each year has been as a result of external/multiple injuries (32%; n=226) generally sustained in motor vehicle pursuit accidents. Deaths resulting from gunshots (28%; n=198), either self-inflicted or by police, were the second most common cause of death. Over the decade between 1989–90 and 1999–2000, deaths caused by drug and/or acute alcohol toxicity fluctuated between one and seven deaths each year. However, since 1999–2000, these deaths have become infrequent compared with all other causes, ranging between zero and two deaths each year. Deaths as a result of hanging have shown a similar trend, being more frequent in the late 1980s and early 1990s but subsequently declining, with only four such deaths being recorded in the last five years.

Table 60 Deaths in police custody and custody-related operations by cause of death, 1989–90 to 2010–11 (n)
Hanging Natural causes Head injury Gunshot External/multiple trauma Drug/alcohol toxicity Other/multiple Total
1989–90 7 4 2 2 6 6 2 29
1990–91 5 3 3 8 3 2 2 26
1991–92 4 5 3 6 2 5 0 25
1992–93 6 0 6 11 8 4 3 38
1993–94 2 4 1 16 4 2 2 31
1994–95 0 1 4 11 5 7 2 30
1995–96 2 4 0 11 12 1 1 31
1996–97 2 2 3 15 10 2 0 34
1997–98 3 2 1 7 12 3 0 28
1998–99 2 6 0 2 6 4 1 21
1999–2000 2 2 3 15 13 1 0 36
2000–01 0 4 5 7 13 2 3 34
2001–02 2 4 8 6 18 1 3 42
2002–03 5 3 5 10 17 0 1 41
2003–04 4 4 1 11 16 2 4 42
2004–05 3 2 3 11 15 0 2 36
2005–06 2 3 6 6 5 2 0 24
2006–07 1 4 0 7 14 2 2 30
2007–08 1 7 2 9 13 1 1 34
2008–09 1 2 4 10 17 1 1 36
2009–10 1 3 3 8 8 1 3 27
2010–11 0 4 0 9 19 1 0 23
Totala 55 73 63 198 226 50 33 698

a: 2 cases have been excluded due to missing information about cause of death

Source: AIC NDICP 1989–90 to 2010–11 [computer file]

Hanging deaths: Points and materials used

Data on hanging points and materials used in hangings in police custody have been collected since 1989–90. Of the 55 hanging deaths since 1989–90, hanging points have included:

  • 29 percent (n=16) involving cell bars;
  • 16 percent (n=9) involving other fittings in cells (such as vents);
  • 15 percent (n=8) involving fittings outside the cell (such as a tree);
  • 13 percent (n=7) involving some other hanging point;
  • 9 percent (n=5) involving a shower fitting (such as taps or shower rail);
  • 6 percent (n=3) occurring in a police van; and
  • 13 percent (n=7) have missing information about the hanging point.

Materials used in hangings included:

  • 27 percent (n=15) with clothing (other than shoelaces or belts);
  • 20 percent (n=11) with a rope/cord;
  • 18 percent (n=10) with sheets;
  • 11 percent (n=6) with shoelaces;
  • 9 percent (n=5) with belts;
  • 2 percent (n=1) with some other material; and
  • 13 percent (n=7) have missing information about material used in the hanging.

Manner of death

Manner of death is designed to capture the circumstances that led to the death. The most common manner of death in police custody and custody-related operations over the reporting period was accidental (46%; n=40), followed by self-inflicted deaths (24%; n=21) and those classified as justifiable homicides (15%; n=13; see Table 61). By way of explanation, a justifiable homicide refers to a situation where a person is killed in circumstances authorised by law, for example a police officer who, acting to protect life, fatally shoots a person who is threatening or in the process of harming others. Deaths resulting from a natural cause over the reporting period were infrequent, with only one in 10 deaths falling into this category (10%; n=9). Deaths in other circumstances, such as drowning or fire, were the least common over the reporting period, with only two such deaths (2%) being recorded.

Table 61 Deaths in police custody and custody-related operations by manner of death, 2008–09 to 2010–11 (n)
2008–09
Self-inflicted Natural causes Justifiable homicide Accident Other Total
New South Wales 2 1 1 1 0 5
Victoria 1 0 1 1 0 3
Queensland 2 0 0 6 0 8
Western Australia 2 0 0 5 1 8
South Australia 1 0 2 3 0 6
Tasmania 0 0 0 0 0 0
Northern Territory 1 1 0 3 0 5
Australian Capital Territory 0 0 0 0 0 0
Totals
Indigenous 2 1 0 4 1 8
Non-Indigenous 7 1 4 15 0 27
All personsa 9 2 4 19 1 35
2009–10
Self-inflicted Natural causes Justifiable homicide Accident Other Total
New South Wales 1 0 3 1 0 5
Victoria 2 2 0 2 0 6
Queensland 3 0 0 2 1 6
Western Australia 0 1 0 2 0 3
South Australia 1 0 0 0 0 1
Tasmania 0 0 0 1 0 1
Northern Territory 0 0 0 4 0 4
Australian Capital Territory 0 0 0 1 0 1
Totals
Indigenous 1 1 0 4 0 6
Non-Indigenous 6 2 3 9 1 21
All persons 7 3 3 13 1 27
2010–11
Self-inflicted Natural causes Justifiable homicide Accident Other Total
New South Wales 1 1 2 1 0 5
Victoria 0 0 1 0 0 1
Queensland 2 2 0 2 0 6
Western Australia 1 1 0 4 0 6
South Australia 1 0 1 0 0 2
Tasmania 0 0 1 1 0 2
Northern Territory 0 0 0 0 0 0
Australian Capital Territory 0 0 1 0 0 1
Totals
Indigenous 0 1 1 4 0 7
Non-Indigenous 5 3 5 4 0 17
All personsb 5 4 6 8 0 23

a: The manner of 1 non-Indigenous death is still under investigation

b: The manner of 1 Indigenous death is still under investigation

Source: AIC NDICP 2008–09 to 2010–11 [computer file]

Trends

Comparing long-term trends with totals recorded over the reporting period showed a pronounced decline in the proportion of self-inflicted deaths, with 24 percent (n=21) over the reporting period being self-inflicted compared with 31 percent (n=218) since 1989–90. By contrast, while self-inflicted deaths have been in decline, the proportion of deaths in accidental circumstances over the last three financial years was much higher than historical trends. Since 1989–90, 40 percent (n=281) of all deaths were in accidental circumstances, whereas over the reporting period, almost half of all deaths fell into this category (46%; n=40; see Table 62 and Figure 27).

Long-term trends show that deaths determined to be unlawful homicides (ie murder or manslaughter) occur infrequently (2%; n=14), while deaths in other circumstances (2%; n=12) are the least common type of police custody death.

Table 62 Deaths in police custody and custody-related operations by manner of death, 1989–90 to 2010–11 (n)
Self-inflicted Natural causes Unlawful homicide Justifiable homicide Accident Other Total
1989–90 8 4 1 1 14 1 29
1990–91 9 3 0 3 8 3 26
1991–92 9 5 0 4 7 0 25
1992–93 13 0 1 4 18 2 38
1993–94 11 4 0 9 6 1 31
1994–95 10 1 0 7 12 0 30
1995–96 12 4 0 4 11 0 31
1996–97 12 2 1 7 12 0 34
1997–98 6 2 1 5 14 0 28
1998–99 2 6 0 2 11 0 21
1999–2000 9 2 0 11 12 2 36
2000–01 8 4 1 2 19 0 34
2001–02 13 4 2 2 21 0 42
2002–03 17 3 1 4 15 1 41
2003–04 14 4 0 6 18 0 42
2004–05 12 2 3 6 12 1 36
2005–06 8 3 2 3 8 0 24
2006–07 11 4 0 3 12 0 30
2007–08 13 7 0 3 11 0 34
2008–09 9 2 1 4 19 0 35
2009–10 7 3 0 3 13 1 27
2010–11 5 4 0 6 8 0 23
Totala 218 73 14 99 281 12 697

a: 3 cases have been excluded due to missing information about manner of death

Source: AIC NDICP 1989–90 to 2010–11 [computer file]

Figure 27 Deaths in police custody and custody-related operations by manner of death and Indigenous status, 1989–90 to 2010–11 (%)

Mental health

Research has shown that a significant proportion of day-to-day policing involves officers responding to persons suffering mental health problems. Using a sample (n=44) of Victorian Police officers from St Kilda and South Melbourne police stations, and 13 Crisis Assessment and Treatment Team staff at the Alfred Hospital, a recent study estimated that

approximately 41% of police responded at least weekly to calls involving people with a mental illness, with 21% transporting people for emergency mental health care at least weekly (Hollander et al. 2011: 3).

While this study used a relatively small sample, it does provide some evidence that police officers have regular contact with persons suffering mental health problems.

It has been observed that changes to the mental health system in Australia in recent years have resulted in more frequent contact between police officers and persons suffering a mental illness. The consequence being, as one study highlights, the ‘deinstitutionalisation of consumers of the mental health system, [such that] police officers have been given increasing responsibility and power’ (Martinez et al. 2005: 21).

Martinez and colleagues (2005: 21) argued that as a consequence there has been an increase in demands on police officers to, in effect, ‘act as social workers, psychologists, caseworkers, customer service officers as well as law enforcers’.

In response to these increasing demands placed on officers, police agencies in all jurisdictions have reviewed and enhanced the training officers receive in responding to persons suffering a mental illness. In addition to theoretical training, officers also receive practical training using common scenarios that will be encountered in day-to-day policing. Some evaluations of the increased mental health training given to officers have been undertaken. For example, a recent study by Godfredson et al. (2010) examined how Victorian Police officers might respond persons suffering a mental illness. This study found that the factor that had the greatest influence over the approach taken by officers in resolving the matter was the severity of symptoms displayed by the person of interest. It is concluded by this study ‘that officers were adept at selecting outcomes that were appropriate given the mental health needs depicted in the [video]’ (Godfredson et al. 2010: 1399). Therefore, this study provides some indication that efforts by police agencies to better train their officers in how to respond to persons with mental health problems are having a positive effect.

Mental Illness and deaths in police custody and custody-related operations

The NDICP captures information about the deceased’s mental health and whether they were suffering the symptoms of a mental health disorder at the time they died. Information about mental health is obtained retrospectively from coronial findings, which in some cases can take years to be handed down. During 2011, more than 1,100 historical coronial findings were reviewed and where a deceased person was indicated as having a mental health problem, or suffering the symptoms of a mental illness immediately prior to death, the type of disorder/illness was recorded in the NDICP database. Tables 63 and 64 present the results of this process. It should be noted that there are a large number of cases where information is missing in relation to these variables.

Further, for those persons who died in police custody and custody-related operations where the available information did not indicate the presence of a mental health issue, the lack of information should not be taken as confirmation that the person had no mental health issues. It may be the case that information about the deceased’s mental health was not included in the coronial finding because it had little or no impact on the circumstances of the death. It is also possible that during this period, other people died in police custody and custody-related operations who had mental health issues but that these issues were not diagnosed and/or were not recorded in the available information. For this reason, findings presented here should be treated as a conservative estimate of the prevalence of mental illness among persons dying in police custody and custody-related operations.

The assignment of mental illness, disorders or conditions was made through judgement by NDICP research staff based on often limited available information. The information presented in Tables 63 and 64 is the best available data to indicate the prevalence of mental health problems among those persons who die in police custody and custody-related operations. For all deaths in police custody and custody-related operations since 1989–90 (n=700), there were 200 cases (29%) where the coronial finding indicated that the deceased was suffering the symptoms of, or had been diagnosed with, a mental illness. Mental illness was found to be more prevalent among non-Indigenous persons dying in police custody than Indigenous persons (31% cf 18%).

A mood disorder, primarily depression, was the most common form of mental illness affecting persons dying in police custody and custody-related operations (38%; n=76; see Table 64). This was followed by persons suffering from psychotic disorders such as schizophrenia (30%; n=59). Substance disorders, such as heroin addiction, were found in 12 percent (n=24) of cases, while in 21 deaths (11%), the deceased was reported as suffering from a mental illness; however, insufficient information was available to determine the precise type of disorder.

Table 63 Deaths in police custody and custody-related operations of persons with a mental illness by Indigenous status, as a proportion of all deaths by year, 1989–90 to 2010–11 (n)
Indigenous Non-Indigenous Total n
n % n % n %
1989–90 0 1 5 1 3
1990–91 0 3 13 3 12
1991–92 2 40 3 15 5 20
1992–93 1 20 7 21 8 21
1993–94 0 11 39 11 35
1994–95 2 67 3 11 5 17
1995–96 0 6 24 6 19
1996–97 1 13 9 35 10 29
1997–98 1 20 5 22 6 21
1998–99 0 2 14 2 10
1999–2000 0 8 25 8 22
2000–01 1 14 9 35 10 30
2001–02 3 38 17 50 20 48
2002–03 4 36 10 34 14 35
2003–04 1 10 11 34 12 29
2004–05 3 27 11 44 14 38
2005–06 1 14 7 41 8 33
2006–07 2 67 16 57 18 58
2007–08 2 40 15 52 17 50
2008–09 2 25 9 31 11 30
2009–10 0 9 43 9 33
2010–11 0 2 11 2 8
Total 26 18 174 31 200 29

Source: AIC NDICP 1989–90 to 2010–11 [computer file]

Table 64 Deaths in police custody and custody-related operations of persons with a known mental illness by type of disorder and Indigenous status, 1989–90 to 2010–11 (n)
Adjustment Anxiety Mood Personality Psychotic Sexual Somatoform Substance Developmental Intellectual disability Other/unknown Total
I N I N I N I N I N I N I N I N I N I N I N I N
1989–90 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1
1990–91 0 0 0 0 0 1 0 1 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 3
1991–92 0 0 0 1 1 0 0 1 0 1 0 0 0 0 0 0 0 0 0 0 1 0 2 3
1992–93 0 1 1 0 0 2 0 0 0 3 0 0 0 0 0 0 0 0 0 0 0 1 1 7
1993–94 0 0 0 0 0 4 0 0 0 4 0 0 0 1 0 0 0 0 0 0 0 2 0 11
1994–95 0 1 0 0 1 2 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2 3
1995–96 0 0 0 0 0 3 0 0 0 2 0 0 0 0 0 0 0 0 0 0 0 1 0 6
1996–97 0 0 0 1 1 1 0 1 0 5 0 0 0 0 0 0 0 1 0 0 0 0 1 9
1997–98 0 0 0 0 0 3 0 0 1 1 0 0 0 0 0 0 0 0 0 0 0 1 1 5
1998–99 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 1 0 2
1999–2000 0 0 0 0 0 2 0 0 0 2 0 0 0 0 0 0 0 0 0 0 0 4 0 8
2000–01 0 0 0 0 0 5 0 0 0 1 0 1 0 0 1 1 0 0 0 0 0 1 1 9
2001–02 0 0 0 0 0 12 0 0 1 3 0 0 0 0 2 2 0 0 0 0 0 0 3 17
2002–03 0 0 0 0 1 5 0 1 2 1 0 0 0 0 1 2 0 0 0 0 0 1 4 10
2003–04 0 0 0 0 1 4 0 1 0 4 0 0 0 0 0 2 0 0 0 0 0 0 1 11
2004–05 0 0 0 0 2 2 0 0 1 6 0 0 0 0 0 0 0 0 0 1 0 2 3 11
2005–06 0 0 0 1 0 3 0 0 1 1 0 0 0 0 0 2 0 0 0 0 0 0 1 7
2006–07 0 0 0 0 0 3 0 0 0 6 0 0 0 0 1 5 0 0 1 0 0 2 2 16
2007–08 0 0 0 0 1 6 1 0 0 4 0 0 0 0 0 2 0 0 0 1 0 2 2 15
2008–09 0 0 0 0 1 4 0 0 0 4 0 0 0 0 0 1 0 0 0 0 1 0 2 9
2009–10 0 0 0 0 0 3 0 0 0 2 0 0 0 0 0 2 0 0 0 0 0 2 0 9
2010–11 0 0 0 0 0 1 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 2
Total 0 2 1 3 9 67 2 5 6 53 0 1 0 1 5 19 0 1 1 2 2 20 26 174

Note: I=Indigenous; N=non-Indigenous

Source: AIC NDICP 1989–90 to 2010–11 [computer file]

Drugs and alcohol

Of the 700 deaths recorded over the monitoring period from 1989–90 to 2010–11, information regarding the consumption of alcohol and/or use of drugs was available in 489 deaths (70%; see Table 65). Overall, a total of 391 individuals who died in police custody over this period were found through toxicology screening to have either recently consumed alcohol, used drugs or a combination of both prior to death. There were 98 cases (20%) where the toxicology report showed no presence of drugs and/or alcohol.

It should be noted that it is not possible to determine whether the deceased was intoxicated or impaired by the consumption of substances based on the toxicology results alone. First, the level of consumption is not taken into account in this analysis. Second, it is widely recognised that persons who are regular consumers of alcohol and/or drugs can develop a tolerance to the effects of these substances and therefore require more of the substance to become intoxicated (eg see Swift, Hall & Teesson 2001; Coffey et al. cited in Ross 2007). Conversely, a person may consume only a small volume of alcohol and/or drugs, but may become intoxicated as a result. Finally, for the most recent years, the decline in annual totals where alcohol or drugs had been consumed is a product of the fact that many toxicology results are not yet available. For these reasons, caution should be taken when interpreting trends.

Among these three categories, alcohol (32%; n=154) was the most commonly recorded intoxicant found in the deceased’s system through toxicology reports. This was closely followed by those who were recorded as having used drugs (31%; n=150), while a total of 87 deaths (18%) were of persons with both drugs and alcohol in their system at the time of death. With regards to Indigenous deaths, the most prevalent intoxicant recorded was alcohol (42%; n=43), followed by drugs (28%; n=28) and a combination of both drugs and alcohol (16%; n=16).

Table 65 Summary of available toxicology results for deaths in police custody and custody-related operations by Indigenous status and year, 1989–90 to 2010–11 (n)
Alcohol Drugs Alcohol and drugs No alcohol or drugs Total
I N Total I N Total I N Total I N Total I N Total
1989–90 0 3 3 0 3 3 1 1 2 2 2 4 3 9 12
1990–91 1 2 3 1 2 3 0 2 2 0 3 3 2 9 11
1991–92 3 10 13 1 3 4 0 0 0 1 7 8 5 20 25
1992–93 1 6 7 3 10 13 0 3 3 1 10 11 5 29 34
1993–94 2 8 10 0 8 8 1 3 4 0 5 5 3 24 27
1994–95 1 5 6 0 8 8 1 3 4 1 5 6 3 21 24
1995–96 3 8 11 0 6 6 2 1 3 0 5 5 5 20 25
1996–97 3 5 8 2 8 10 1 3 4 0 6 6 6 22 28
1997–98 3 6 9 0 3 3 0 5 5 1 1 2 4 15 19
1998–99 1 0 1 0 3 3 0 1 1 2 2 4 3 6 9
1999–2000 0 2 2 0 8 8 0 2 2 1 2 3 1 14 15
2000–01 1 8 9 4 4 8 1 4 5 0 6 6 6 22 28
2001–02 5 11 16 2 7 9 0 1 1 1 8 9 8 27 35
2002–03 2 7 9 3 5 8 1 6 7 3 2 5 9 20 29
2003–04 3 6 9 2 8 10 2 3 5 0 4 4 7 21 28
2004–05 5 5 10 2 8 10 2 3 5 1 2 3 10 18 28
2005–06 2 2 4 2 3 5 2 4 6 0 3 3 6 12 18
2006–07 0 6 6 1 8 9 1 4 5 0 2 2 2 20 22
2007–08 2 3 5 2 5 7 0 9 9 0 2 2 4 19 23
2008–09 3 5 8 0 8 8 1 4 5 0 2 2 4 19 23
2009–10 2 2 4 2 4 6 0 4 4 0 3 3 4 13 17
2010–11 0 1 1 1 0 1 0 5 5 1 1 2 2 7 9
Total 43 111 154 28 122 150 16 71 87 15 83 98 102 387 488

Note: I=Indigenous; N=non-Indigenous

Source: AIC NDICP 1989–90 to 2010–11 [computer file]

Trends

Looking at the long-term trends over the period 1989–90 to 2010–11, analysis of available data showed that for deaths of persons with alcohol in their system, there were two distinct peaks in 1991–92 (n=13) and 2001–02 (n=16), with the totals recorded in all other years ranging between zero and 11. Deaths of non-Indigenous persons with alcohol, drugs, or a combination of both in their system, show more pronounced fluctuations than do deaths of Indigenous persons. A greater proportion of non-Indigenous persons were found to have drugs in their system at death (32%; n=122), rather than just alcohol (29%; n=111), or a combination of drugs and alcohol (18%; n=71). Indigenous persons represent over one in four (28%; n=43) persons with alcohol in their system at the time of death. Also of note is the fact that Indigenous persons were less likely to have only drugs in their system (28%; n=28) compared with only alcohol (42%; n=43). Overall, for those cases where toxicology results are available, eight in 10 deaths (79%; n=387) in police custody and custody-related operations are of persons with some form of intoxicant in their system at the time of death.

Most serious offence

The most common offences committed prior to a death in police custody or a custody-related operation over the reporting period was a traffic offence (31%; n=25), or violent offence (31%; n=25). Other offences (14%; n=11), which include weapons offences or breaching a domestic violence order, were the third most prevalent offence type (see Table 66). Drug-related offences were the least common offences prior to custody over the last three financial years (1%; n=1).

Indigenous persons were more likely to have committed a theft-related offence prior to custody (33%; n=6), followed by a good-order offence (22%; n=4). By contrast, traffic offences were more common among those non-Indigenous persons who died (37%; n=23), with violent offences the second most prevalent (35%; n=22).

Table 66 Deaths in police custody and custody-related operations by most serious offence, 2008–09 (n)
2008–09
Violent Theft-related Drug-related Traffic Good order Other Total
New South Wales 2 0 0 1 0 2 5
Victoria 1 1 0 1 0 0 3
Queensland 3 0 0 4 0 1 8
Western Australia 0 1 0 4 2 0 7
South Australia 2 1 0 2 0 0 5
Tasmania 0 0 0 0 0 0 0
Northern Territory 1 0 0 0 2 2 5
Australian Capital Territory 0 0 0 0 0 0 0
Totals
Indigenous 2 1 0 0 2 2 7
Non-Indigenous 7 2 0 12 2 3 26
All personsa 9 3 0 12 4 5 33
2009–10
Violent Theft-related Drug-related Traffic Good order Other Total
New South Wales 2 0 0 1 1 1 5
Victoria 3 0 0 1 1 1 6
Queensland 1 0 1 1 0 3 6
Western Australia 0 0 0 1 0 1 2
South Australia 1 0 0 0 0 0 1
Tasmania 0 0 0 1 0 0 1
Northern Territory 0 2 0 1 1 0 4
Australian Capital Territory 0 0 0 1 0 0 1
Totals
Indigenous 0 2 0 1 1 1 5
Non-Indigenous 7 0 1 6 2 5 21
All personsb 7 2 1 7 3 6 26
2010–11
Violent Theft-related Drug-related Traffic Good order Other Total
New South Wales 4 0 0 0 0 0 4
Victoria 1 0 0 0 0 0 1
Queensland 1 0 0 3 1 0 5
Western Australia 1 3 0 1 1 0 6
South Australia 1 0 0 1 0 0 2
Tasmania 0 0 0 1 1 0 2
Northern Territory 0 0 0 0 0 0 0
Australian Capital Territory 1 0 0 0 0 0 1
Totals
Indigenous 1 3 0 1 1 0 6
Non-Indigenous 8 0 0 5 2 0 15
All personsc 9 3 0 6 3 0 11

a: 2 cases have been excluded due to missing data about the offence committed and 1 case has been excluded because no offence was committed

b: 1 case has been excluded due to missing data about the offence committed

c: 3 cases have been excluded due to missing data about the offence committed

Source: AIC NDICP 2008–09 to 2010–11 [computer file]

Trends

Data recorded over the reporting period shows a departure from long-term trends, in that a greater proportion of traffic-related offenders died in police custody and custody-related operations than has traditionally been the case (31% cf 15%; see Table 67). Conversely, a smaller proportion of theft-related offenders died in the reporting period compared with long-term proportions (10% cf 22%).

Over the last 22 years for which data are available on offending, persons committing violent and theft-related offences prior to custody have shown the greatest fluctuation. The number of persons dying in custody having committed a violent offence peaked in 1999–00, with 18 such deaths, with subsequent peaks in both 2003–04 and 2005–06 (n=16). Over the last six years, the number of deaths among violent offenders has remained stable, ranging between seven and nine deaths each year. A similar trend can be seen with theft-related offences, with two distinct peaks in both 1992–93 and 2001–02 (n=16 and n=14 respectively). However, since the second peak, there has been a considerable decline, with numbers recorded in the last six years being at the low end of historical totals.

The other offence categories of good order, drug-related, traffic and other offences have all remained relatively constant over time, with the exception of a few minor spikes among traffic and other related offences, as well as a peak in good order offences during the early years of reporting. Most noteworthy are the recent spikes in the number of deaths of traffic offenders in 2008–09 (n=12) and persons committing other offences in 2007–08 (n=10), which represent the highest totals ever recorded in those categories.

With regards to Indigenous deaths in police custody and custody-related operations since 1989–90, where information is available, the offences leading to custody were distributed as follows (see Figure 28):

  • theft-related offences (35%; n=49);
  • good order offences (23%; n=33);
  • violent offences (21%; n=30);
  • other offences (10%; n=14);
  • traffic offences (6%; n=9);
  • drug-related offences (0.7%; n=1);
  • the offence was missing in three cases (2%); and
  • no offence was committed in two cases (1%).

There was a different distribution of offences among non-Indigenous deaths in police custody and custody-related operations since 1989–90 (see Figure 28);

  • violent offences (33%; n=186);
  • theft-related offences (19%; n=104);
  • traffic offences (17%; n=95);
  • good order offences (14%; n=76);
  • other offences (12%; n=65);
  • drug-related offences (3%; n=19);
  • the offence was missing in 10 cases (2%); and
  • no offence was committed in four cases (1%).
Table 67 Deaths in police custody and custody-related operations by most serious offence, 1989–90 to 2010–11 (n)
Violent theft-related Drug-related Traffic Good order Other Total
1989–90 4 6 0 3 15 1 29
1990–91 7 7 0 0 12 0 26
1991–92 5 2 1 2 11 4 25
1992–93 8 16 2 5 4 3 38
1993–94 14 7 0 3 4 3 31
1994–95 13 7 3 3 3 0 29
1995–96 13 9 0 4 4 1 31
1996–97 14 10 1 1 6 2 34
1997–98 6 6 0 4 9 1 26
1998–99 6 9 0 1 2 2 20
1999–2000 18 10 2 2 3 1 36
2000–01 9 7 0 7 6 5 34
2001–02 8 14 0 7 3 10 42
2002–03 8 8 1 9 4 10 40
2003–04 16 9 1 7 5 3 41
2004–05 16 8 3 5 2 1 35
2005–06 8 3 3 4 3 3 24
2006–07 9 3 1 6 0 8 27
2007–08 9 4 1 7 2 10 33
2008–09 9 3 0 12 4 5 33
2009–10 7 2 1 7 3 6 26
2010–11 9 3 0 6 3 0 21
Totala, b 216 153 20 104 109 79 681

a: 13 cases have been excluded due to missing information about the offence committed

b: 6 cases have been excluded because no offence was committed

Source: AIC NDICP 1989–90 to 2010–11 [computer file]

Figure 28 Deaths in police custody and custody-related operations by most serious offence and Indigenous status, 1989–90 to 2010–11 (%)

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. The different locations include:

  • public hospitals, which include cases where a person was transferred to a hospital from other locations and died there;
  • cells, which include all types of cells—single, shared or observation;
  • custodial settings, which include locations such as interview rooms or police vans;
  • private properties, which generally consist of private residences;
  • public places, such as streets, ovals and other outdoor general-use areas; and
  • other locations, which may include ambulances.

The majority of deaths in police custody and custody-related operations occurred in a public place (45%; n=39), which is to be expected given the fact that the majority of policing occurs in public spaces (see Table 68). The second most common place for a death to occur was a public hospital (23%; n=20), usually following transfer from an incident in a public place. Just over one in five deaths (23%; n=20) over the reporting period occurred on private property, most commonly the deceased’s residence. Deaths occurring in cells (6%; n=5), other locations (2%; n=2) or some other custody setting such as the back of a police van (1%; n=1), were the least common locations for a death to occur.

Over half of the 21 Indigenous deaths in the last three financial years occurred in a public place (52%; n=11), with seven deaths occurring in a public hospital, two in a cell and one on private property.

Table 68 Deaths in police custody and custody-related operations by location of death, 2008–09 to 2010–11 (n)
2008–09
Cell Other custodial setting Public hospital Private property Public place Other Total
New South Wales 1 0 0 2 3 0 6
Victoria 0 0 0 0 3 0 3
Queensland 0 0 1 3 4 0 8
Western Australia 0 0 3 0 4 1 8
South Australia 0 0 1 2 3 0 6
Tasmania 0 0 0 0 0 0 0
Northern Territory 1 0 3 0 1 0 5
Australian Capital Territory 0 0 0 0 0 0 0
Totals
Indigenous 1 0 4 0 3 0 8
Non-Indigenous 1 0 4 7 15 1 28
All persons 2 0 8 7 18 1 36
2009–10
Cell Other custodial setting Public hospital Private property Public place Other Total
New South Wales 0 0 3 0 2 0 5
Victoria 1 1 1 1 1 1 6
Queensland 1 0 1 3 1 0 6
Western Australia 0 0 1 0 2 0 43
South Australia 0 0 0 1 0 0 1
Tasmania 0 0 1 0 0 0 1
Northern Territory 0 0 2 0 2 0 4
Australian Capital Territory 0 0 1 0 0 0 1
Totals
Indigenous 0 0 2 0 4 0 6
Non-Indigenous 2 1 8 5 4 1 21
All persons 2 1 10 5 8 1 27
2010–11
Cell Other custodial setting Public hospital Private property Public place Other Total
New South Wales 0 0 0 4 1 0 5
Victoria 0 0 0 0 1 0 1
Queensland 0 0 2 2 3 0 7
Western Australia 1 0 0 1 4 0 6
South Australia 0 0 0 0 2 0 2
Tasmania 0 0 0 1 1 0 2
Northern Territory 0 0 0 0 0 0 0
Australian Capital Territory 0 0 0 0 1 0 1
Totals
Indigenous 1 0 1 1 4 0 7
Non-Indigenous 0 0 1 7 9 0 17
All person 1 0 2 8 13 0 24

Source: AIC NDICP 2008–09 to 2010–11 [computer file]

Trends

Findings from analysis of data collected over the reporting period are in line with long-term trends, in that more deaths occurred in a public place (45%; n=39 cf 42%; n=293) than any other location (see Table 69). Since 1989–90, public hospitals (26%; n=184) were the second most common location for a death to occur. While deaths in police cells peaked in 1989–90 with 11 deaths, over the ensuing two decades, the number of such deaths has dropped markedly, ranging between one and four deaths each year (with the exception of 1990–91 and 1997–98 when 6 cell deaths occurred). These data suggest that efforts over the last decade by police agencies in redesigning cells, removing hanging points and improving the monitoring of detainees deemed at risk of self-harm are having a positive effect on reducing the number of deaths in these settings.

Table 69 Deaths in police custody and custody-related operations by location of death and year, 1989–90 to 2010–11
Public hospital Cell Other custodial setting Private property Public place Other Total
1989–90 6 11 5 1 6 0 29
1990–91 7 6 4 5 4 0 26
1991–92 11 5 3 3 3 0 25
1992–93 13 3 2 3 15 2 38
1993–94 8 2 2 5 13 1 31
1994–95 7 2 2 5 14 0 30
1995–96 11 3 0 7 8 2 31
1996–97 12 2 2 4 12 2 34
1997–98 7 6 1 3 10 1 28
1998–99 10 3 2 2 3 1 21
1999–2000 8 1 1 9 17 0 36
2000–01 8 5 0 3 18 0 34
2001–02 9 2 2 5 24 0 42
2002–03 7 2 3 4 24 1 41
2003–04 3 4 0 8 27 0 42
2004–05 7 1 3 6 18 1 36
2005–06 5 2 1 4 12 0 24
2006–07 7 2 1 5 15 1 31
2007–08 18 1 0 4 11 0 34
2008–09 8 2 0 7 18 1 36
2009–10 10 2 1 5 8 1 27
2010–11 2 1 0 8 13 0 24
Total 184 68 35 106 293 14 700

Source: AIC NDICP 1989–90 to 2010–11 [computer file]

Circumstances of custodial episode for deaths in police custody and custody-related operations

The NDICP examines the circumstances of the custodial episode; that is, why the deceased was in custody at the time of death (see Table 69). The four categories into which deaths are classified are:

  • institutional—deaths in a police cell or watchhouse during the process of being transferred to or from this setting, deaths in a hospital following detainment in a police cell, or deaths occurring in police vans;
  • escaping—deaths that occur during the process of the person escaping or attempting to escape from police custody;
  • detaining—deaths occurring during the process of police officers attempting to detain the person, regardless of whether or not the person was under arrest; and
  • other cases—for example, the death of a person detained under a state Mental Health Act, those occurring in the process of the persons being conveyed via ambulance from a public hospital to a psychiatric institution under ‘police escort’, or any other situation that does not fit into one of the other three categories.

Of the 87 deaths over the last three financial years, more than four in every five (84%; n=73) occurred during the process of police officers detaining or attempting to place the deceased under arrest. Deaths in institutional settings, such as police cells and vans, were the second most common type (14%; n=12) of custodial episode prior to death.

A similar distribution of deaths can be seen among Indigenous deaths in police custody over the reporting period, in that most deaths occurred during the process of arrest (71%; n=15), with just under one in four occurring in an institutional setting (24%; n=5). There was one death of an Indigenous person attempting to escape from custody.

Table 70 Deaths in police custody and custody-related operations by custodial episode, 2008–09 to 2010–11 (n)
2008–09
Institutional Escaping Detaining Other Total
New South Wales 1 0 5 0 6
Victoria 0 0 3 0 3
Queensland 0 0 8 0 8
Western Australia 1 0 7 0 8
South Australia 0 0 6 0 6
Tasmania 0 0 0 0 0
Northern Territory 2 1 2 0 5
Australian Capital Territory 0 0 0 0 0
Totals
Indigenous 2 1 5 0 8
Non-Indigenous 2 0 26 0 28
All persons 4 1 31 0 36
2009–10
Institutional Escaping Detaining Other Total
New South Wales 0 0 5 0 5
Victoria 3 0 3 0 6
Queensland 1 0 5 0 6
Western Australia 1 0 2 0 3
South Australia 0 0 1 0 1
Tasmania 0 0 1 0 1
Northern Territory 0 0 3 1 4
Australian Capital Territory 0 0 1 0 1
Totals
Indigenous 1 0 5 0 6
Non-Indigenous 4 0 16 1 21
All persons 5 0 21 1 27
2010–11
Institutional Escaping Detaining Other Total
New South Wales 0 0 5 0 5
Victoria 0 0 1 0 1
Queensland 2 0 5 0 7
Western Australia 1 0 5 0 6
South Australia 0 0 2 0 2
Tasmania 0 0 2 0 2
Northern Territory 0 0 0 0 0
Australian Capital Territory 0 0 1 0 1
Totals
Indigenous 2 0 5 0 7
Non-Indigenous 1 0 16 0 17
All persons 3 0 21 0 24

Source: AIC NDICP 2008–09 to 2010–11 [computer file]

Detainment-related deaths

Due to the fact that the majority (73%; n=508) of all deaths in police custody and custody-related operations since 1989–90 occurred in the process of officers detaining or attempting to detain the deceased, these deaths are analysed separately. In such cases, the method of detainment is further classified under one of five categories—motor vehicle pursuit, other pursuit, siege, raid and other/shooting (see Table 71).

Deaths in the process of detaining over the last three financial years were the most common circumstance in which a death in police custody occurred (84%; n=73). Of these 73 deaths, those occurring in a motor vehicle pursuit (43%; n=31) were more common than those in any other situation (see Table 71). Deaths during sieges, where an offender defends an area against all others, were the second most common type of death occurring in the process of police officers attempting to place someone under arrest (23%; n=17).

With regards to the 15 Indigenous deaths in the process of being detained, motor vehicle pursuits were the most common (60%; n=9), followed by deaths in other situations, such as escorting or shootings (20%; n=3).

Table 71 Deaths in police custody and custody-related operations by method of detainment, 2008–09 to 2010–11 (n)
2008–09
Motor vehicle pursuit Other pursuit Siege Raid Other/shooting Total
New South Wales 1 1 1 1 1 5
Victoria 1 0 0 1 1 3
Queensland 4 0 2 0 2 8
Western Australia 5 0 0 0 2 7
South Australia 3 0 2 0 1 6
Tasmania 0 0 0 0 0 0
Northern Territory 1 0 0 0 1 2
Australian Capital Territory 0 0 0 0 0 0
Totals
Indigenous 2 0 0 0 3 5
Non-Indigenous 13 1 5 2 5 26
All personsa 15 1 5 2 8 31
2009–10
Motor vehicle pursuit Other pursuit Siege Raid Other/shooting Total
New South Wales 1 0 2 1 1 5
Victoria 1 0 1 0 1 3
Queensland 1 1 3 0 0 5
Western Australia 2 0 0 0 0 2
South Australia 0 0 1 0 0 1
Tasmania 1 0 0 0 0 1
Northern Territory 2 1 0 0 0 3
Australian Capital Territory 1 0 0 0 0 1
Totals
Indigenous 3 1 1 0 0 5
Non-Indigenous 6 1 6 1 2 16
All personsb 9 2 7 1 2 21
2010–11
Motor vehicle pursuit Other pursuit Siege Raid Other/shooting Total
New South Wales 0 0 2 3 0 5
Victoria 0 0 0 1 0 1
Queensland 2 0 2 0 1 5
Western Australia 4 0 1 0 0 5
South Australia 0 2 0 0 0 2
Tasmania 1 0 0 0 1 2
Northern Territory 0 0 0 0 0 0
Australian Capital Territory 0 0 0 1 0 1
Totals
Indigenous 4 0 0 1 0 5
Non-Indigenous 3 2 5 4 2 16
All personsc 7 2 5 5 2 21

a: 5 cases for which detaining was not the type of custody have been excluded

b: 6 cases for which detaining was not the type of custody have been excluded

c: 3 deaths for which detaining was not the type of custody have been excluded

Source: AIC NDICP 2008–09 to 2010–11 [computer file]

Trends

Deaths occurring in police institutional settings, such as cells and police vans, have been declining since 1989–90, with totals in recent years being some of the lowest ever recorded (see Table 72). At the same time as institutional deaths have been in decline, deaths in the process of detaining have increased, reaching an historical high in 2003–04 (n=35). Deaths in this category account for 73 percent (n=508) of all deaths in police custody and custody-related operations since 1989–90.

Only five deaths have occurred over the last 22 years where an individual was trying to escape after being placed in custody—one in 1996–97 and 2002–03, two in 2004–05 and one in 2008–09. Deaths in the other/marginal category, which encompasses persons detained under mental health legislation, occur infrequently with only two such deaths recorded in the last four financial years.

Table 72 Deaths in police custody and custody-related operations by type of custody, 1989–90 to 2010–11 (n)
Institution Escaping Detaining Other/marginal Total
1989–90 20 0 9 0 29
1990–91 16 0 10 0 26
1991–92 13 0 11 1 25
1992–93 11 0 26 1 38
1993–94 8 0 21 2 31
1994–95 7 0 23 0 30
1995–96 7 0 24 0 31
1996–97 5 1 28 0 34
1997–98 10 0 18 0 28
1998–99 9 0 10 2 21
1999–2000 4 0 32 0 36
2000–01 6 0 27 1 34
2001–02 6 0 30 6 42
2002–03 7 1 30 3 41
2003–04 5 2 35 0 42
2004–05 7 0 28 1 36
2005–06 5 0 19 0 24
2006–07 5 0 24 2 31
2007–08 4 0 30 0 34
2008–09 4 1 31 0 37
2009–10 5 0 21 1 27
2010–11 3 0 21 0 24
Total 167 5 508 20 700

Source: AIC NDICP 1989–90 to 2010–11 [computer file]

Trends in detainment-related deaths

Deaths during motor vehicle pursuits (41%; n=206) are the most common type of death occurring in the process of the deceased being detained by police officers (see Table 73). Deaths occurring during motor vehicle pursuits are also a type of death requiring careful examination and targeted analysis is presented in the following section of this report. Here, it can be seen that the annual number of pursuit deaths peaked to an historical high in 2001–02 with a total of 19 deaths, after which they began to decline. However, in 2008–09 there was another spike in pursuit deaths (n=15).

Deaths resulting from a shooting, or other type of situation where officers are attempting to detain, are the second most common type of death in this category (29%; n=148). Deaths in the process of police officers conducting a raid are the least common type of detainment-related death (5%; n=27).

Most noteworthy when looking at historical trends is the number of deaths occurring during sieges; generally a situation where an offender barricades themselves in a building or defined location and then defends it against police. Throughout the 1990s, these types of deaths did not occur very often; however, over the last 15 years, there has been a moderate increase in such deaths. The annual total number of siege deaths recorded in recent years is at the upper end of historical levels.

Attempting to resolve such situations without the loss of life is a very difficult task for police. In the majority of sieges, the offender is armed with a deadly weapon, such as a firearm or knife. There have been some sieges where police have attempted to negotiate with the offender to produce a peaceful outcome for more than 24 hours, however, the situation has still ended in death. The NDICP will closely monitor deaths in sieges over the next few years to ascertain whether this is an emerging trend for deaths in police custody-related operations, or whether recent rises are an aberration from the relatively stable annual totals seen over the last 22 years.

Table 73 Deaths in police custody and custody-related operations by method of detainment, 1989–90 to 2010–11 (n)
Motor vehicle pursuit Other pursuit Siege Raid Other/shooting Total
1989–90 5 1 0 0 3 9
1990–91 1 0 1 2 6 10
1991–92 2 2 1 0 6 11
1992–93 10 4 3 0 9 26
1993–94 5 1 4 1 10 21
1994–95 7 2 3 1 10 23
1995–96 9 0 3 1 11 24
1996–97 11 1 7 2 7 28
1997–98 11 1 2 0 4 18
1998–99 5 0 1 0 4 10
1999–2000 12 2 3 5 10 32
2000–01 12 6 6 2 1 27
2001–02 19 2 3 1 5 30
2002–03 15 4 3 1 7 30
2003–04 12 7 7 2 7 35
2004–05 11 4 2 1 10 28
2005–06 10 0 3 0 6 19
2006–07 9 2 4 0 9 24
2007–08 9 5 5 0 11 30
2008–09 15 1 5 2 8 31
2009–10 9 2 7 1 2 21
2010–11 7 2 5 5 2 21
Total 206 49 78 27 148 508

Source: AIC NDICP 1989–90 to 2010–11 [computer file]

Summary of deaths in police custody and custody-related operations

Since the definition of a death in police custody was expanded to include those deaths occurring in custody-related operations from 1989–90, the nature and circumstances of deaths in all forms of police custody has changed considerably. Throughout the early 1990s, deaths were more likely to occur in close-contact situations, such as a cell death or shooting. However, since the late 1990s, deaths in motor vehicle pursuits have become the most common type of death in police custody. Since the year 2000, deaths during sieges have also increased, but deaths occurring in motor vehicle pursuits remain the most common type of death each year.

In the reporting period from 2008–09 to 2010–11, the number of deaths decreased from 36 in 2008–09 to 27 deaths in 2009–10 and 24 in 2010–11. Over this three year period, almost one-quarter of all police custody deaths (24%; n=21) occurred in Queensland, with one in five (20%; n=17) occurring in Western Australia. However, while the proportion of all deaths made up by these two jurisdictions has increased, annual totals in recent years are in line with historical trends. The main driver for this proportional increase has been a decline in deaths occurring in both New South Wales (18%; n=16) and Victoria (12%; n=10). For 20 years, these two jurisdictions generally had the highest annual number of police custody deaths. Over the last three financial years, both New South Wales and Victoria have witnessed a considerable decline in deaths in police custody and custody-related operations, with recent annual totals being approximately half the number recorded in previous years.

The frequency of deaths involving Indigenous persons in police custody and custody-related operations has remained stable over the last 22 years, with the exception of a slight increase between 2002–03 and 2004–05. Unlike Indigenous deaths, the number of deaths involving non-Indigenous persons decreased over the reporting period, from 28 in 2008–09, to 21 in 2009–10 to 17 in 2010–11.

Deaths involving males have also remained relatively constant since monitoring began in 1989–90, with the exception of minor fluctuations in some years. Overall, deaths involving males have always accounted for the vast majority of deaths. Analysis of 22 years of data has also revealed that external/multiple trauma (32%; n=226) is the most common cause of death, closely followed by those who died as a result of a gunshot wound (28%; n=198). With regards to the cause of death among Indigenous persons, similar trends can be seen, in that the most common cause of death was external/multiple trauma (37%; n=52); however, the second most common cause of death was natural causes (21%; n=30). For all deaths, the most infrequent cause was other/multiple causes (5%; n=33), which includes situations such as drowning and fire. The least frequent cause of death among Indigenous persons however, was incidents related to drug and/or acute alcohol toxicity (3%; n=4).

Since 1989–90, deaths are more likely to occur in accidental circumstances (40%; n=281), closely followed by self-inflicted injuries (31%; n=218). The most common manner of death among Indigenous persons was accidental (47%; n=66), followed by those resulting from a natural cause process (21%; n=30).

One of the most interesting findings from analysis of deaths in police custody was that more violent offenders die than any other type of offender (31%; n=216). This was followed by offenders who had committed theft-related offences (22%; n=153) and those who were categorised as either good order (16%; n=109) or traffic offenders (15%; n=104). However, for Indigenous deaths in police custody and custody-related operations, a greater proportion were theft-related offenders (36%; n=49) than any other type of offender. Just over one in five deaths of Indigenous persons in police custody involved offences categorised as being related to good order (24%; n=33) or violence (22%; n=30).

In conclusion, there have been proportionally more deaths in recent years occurring during custody-related operations, typically involving motor vehicle pursuits, shootings and sieges, rather than deaths in the physical custody of police, such as in cells or watchhouses. For this reason, motor vehicle pursuit and police shooting deaths are examined separately in more detail.