Aboriginal Children Placed in Out-of-Home Care: Pathways Through the Child Protection System

ABSTRACT

The overrepresentation of Aboriginal and Torres Strait Islander (respectfully referred to hereafter as “Aboriginal”) children in the child protection system is a concern in Australia, with Aboriginal children placed in out-of-home care at a rate 11 times that of non-Aboriginal children. This study utilised linked administrative data to determine the longitudinal child protection pathways from birth to age 10 years for a cohort of 15,815 Aboriginal children born in Western Australia between 2000 and 2006. In total, 9,269 (59%) children did not have any contact with child protection between one year prior to birth and their 11th birthday. Conversely, 6,546 (41%) Aboriginal children were involved with child protection during the study period, with 1,405 (9%) children placed in out-of-home care. Infants who had a child protection notification were more likely than other age groups to have a substantiated notification of abuse and neglect and placed in out-of-home care. More than half (56%) of all children were predominantly placed in kinship care; however, only 22% had their first placement with kinship carers. Aboriginal-led strategies together with the support and commitment of all levels of government are required to reduce entry into care and improve outcomes for Aboriginal children in care.

IMPLICATIONS

• Aboriginal-led strategies and wholistic responses can reduce the overrepresentation of Aboriginal children reported to child protection systems, and support families and communities.

• Infancy stands out as a key point of system involvement. This stage offers an opportunity where early intervention and family support strategies may prevent first time involvement with the child protection services and promote kinship placement, reunification with parents, and community-led supportive care.

KEYWORDS:

• Child Protection
• Indigenous Child Protection
• Out-Of-Home Care
• Family Support
• Community Support
• Aboriginal Children
• Aboriginal and Torres Strait Islander Children
• Kinship Placement
• Australia

Historically and culturally, Aboriginal and Torres Strait Islander (respectfully referred to hereafter as “Aboriginal”) children grow up embedded in family, and community care systems (Lewis & Burton, 2014). Kin and community networks create a protective environment for children where multiple people care for children, and child development is supported through autonomy and community socialisation (Lewis & Burton, 2014). Connection with family and community is central to strong Aboriginal identity formation for Aboriginal children (Krakouer et al., 2018). Since colonisation, Aboriginal people have experienced loss and fragmentation of family through government policies and colonisation practices including systemic violence, racism, genocide, oppression, protection and assimilation policies, and the forced removal of children from family (Anderson et al., 2017; Chamberlain et al., 2022). These actions have resulted in multiple, pervasive, intergenerational or transgenerational trauma (Anderson et al., 2017; Arney et al., 2015).

The overrepresentation of Aboriginal children in the child protection system is an ongoing concern in Australia (Atkinson, 2002; Liddle et al., 2022; SNAICC, 2017). In 2021, there were 19,500 Aboriginal children in out-of-home care (OOHC), a rate of 58 per 1,000, which is 11 times the rate for non-Aboriginal children (Australian Institute for Health and Welfare, 2022). Reducing the overrepresentation of Aboriginal children in OOHC is urgent and has recently become a key target across Australia (Commonwealth of Australia, 2020; Liddle et al., 2022; SNAICC, 2017). In 2021, a new national agreement on Closing the Gap was set to reduce the overrepresentation of Aboriginal children in OOHC by 45% by 2031 (Commonwealth of Australia, 2020). Australian jurisdictions have committed to fully implement the Aboriginal and Torres Strait Islander Child Placement Principle (Department of Social Services, 2015; SNAICC, 2017). The Principle aims to ensure that Aboriginal culture is embedded in policy and practice; to recognise the rights of Aboriginal children, their families and the community in child protection matters; to increase self-determination of Aboriginal people; and to reduce the overrepresentation of Aboriginal children in child protection (SNAICC, 2017).

National and state reports provide numbers and rates of Aboriginal children involved with the child protection system (Davis, 2019; Segal et al., 2019). However, detailed, contemporary and integrated data relating to the pathways and characteristics of children involved with child protection system are required to determine reasons for entering care, age of entry into OOHC, placement stability, family reunification, and possible reentry into care. This evidence needs to capture multiple indicators of placement experiences, tracking each child or young person from the date of initial admission to care, through each placement, until exit from care (Withington et al., 2013). The knowledge gained of placement pathways and patterns of children in care will enable evidenced-based, long-term planning, informing the development of strategies to reduce entry/reentry into care, promote family reunification, and improve support services and resourcing for Aboriginal Community Controlled Organisations.

Western Australia (WA) has the second highest overrepresentation of Aboriginal children in OOHC care in Australia, after Victoria, with Aboriginal children placed in care at 19 times the rate for non-Aboriginal children (Australian Institute for Health and Welfare, 2022). The aims of this study were to (1) determine the longitudinal child protection pathways including notifications, substantiations, and OOHC placements for Aboriginal children entering care between infancy and those aged 10; (2) quantify the characteristics of children in relation to different child protection pathways; and (3) investigate the age of entry to care, and placement types and how they relate to family reunification and reentry to care in WA.

Method

Study Population and Data Sources

This is a retrospective cohort study using WA-linked administrative data from March 1990 to March 2017. The study population included a cohort of 15,815 Aboriginal children born alive in WA between 2000 and 2006. Datasets for this study were linked by the WA Data Linkage Branch (WADLB) utilising probabilistic matching and clerical review (Holman et al., 2008). From the WA Department of Health, Birth Registry and Midwives Notification System data were used to identify births and key cohort characteristics. The Deaths Register was used to identify children who died in the follow-up period. Indigenous status was identified by the WADLB’s Derived Indigenous Status Flag (Christensen et al., 2014). Children with a parent/grandparent classified as Indigenous by the Derived Indigenous Status Flag were also categorised as Indigenous (Gibberd et al., 2017). Department of Communities Child Protection and Family Support (CPFS) data were used to identify children in contact with the child protection system.

Child Protection Outcomes

This study examined notifications of concerns about the wellbeing of children received by CPFS, substantiated notifications, and OOHC placement. Children’s child protection contacts typically start with a notification of concerns about the child safety and wellbeing, which are all reports made to CPFS alleging child abuse and neglect. Over time there have been changes in definitions of notifications that have previously been described by Bilson et al. (2017). CPFS response to referrals is aligned to the Children and Community Services Act 2004 (WA).

An investigation may take place to assess “ … (a) whether or not the child has experienced harm or is likely to experience harm as a result of the abuse; (b) the parent's capacity to protect the child from harm, and; (c) whether the child is in need of protection … ” (Department of Communities, 2021b, p. 4). As a result of the investigation some notifications become substantiated notifications of child abuse and neglect (hereafter referred to as “substantiations”). This may result in safety plans being put in place with the family or, if there are significant concerns about the safety of the child, they may be placed in OOHC (Department of Communities, 2021a).

The period of time in which children are placed in care is defined as a period of care. Within this period, a child may have more than one placement, which refers to the living arrangement where the child lives while in care. Children placed only for a period of care less than 30 days (child safety investigation period) and placements during hospitalisations were not counted as OOHC placements in this study (Department of Communities, 2021b). At the end of a period of care, children may be reunified with their family, some might be placed in a third-party guardianship order or others may stay in care until they reach 18 years old. Children’s placement outcomes are investigated in this study at the end of each period of care, as well as at the end of follow-up time.

During their care experience children can be placed in one or more different types of placements. Placement types include placement with foster carers, kinship/relative carers, or in residential care (placement in a residential unit or group home staffed by paid workers). First placement type refers to the placement type at the first ever OOHC placement, and predominant placement type is defined as the placement type in which the child spent the longest time within the follow-up period.

Child and Parental Characteristics

Child and parental characteristics included sex of the child, parental age at birth and maternal marital status at birth. The Index of Relative Social-Economic Advantage and Disadvantage 2006 (IRSAD) (Australian Bureau of Statistics, 2023) was used to identify socio-economic status at birth, dividing the population in five equal groups (quintiles). Residential remoteness at birth was determined by the Accessibility/Remoteness Index of Australia (ARIA) 2006, including major cities, inner and outer regional, remote, and very remote areas. IRSAD and ARIA 2006 were used for all children for consistency and comparability.

Analysis

Children were followed from one-year prior to birth (to capture prenatal notifications) until the age of 10 years old (defined as the age up to the day before their 11th birthday). Prenatal reporting is recorded in the data from 2000. However, formal amendments were included in the Community Services Act Amendment Regulations from 2010 (WA) and Reciprocal Child Protection Procedures were implemented with the Department of Health in 2007 (Department of Health, 2007). Numbers and percentages were used to describe the child protection pathways (notifications, substantiations, and OOHC placements) for Aboriginal children born in WA from birth to the age of 10 years. Univariable multinomial logistic regression was used to compare characteristics between those with and without contact with the child protection system through generating relative risks ratios (RRR) (see Appendix A in Supplementary material). RRR are the ratio of risk of an event in one group versus the risk of an event in another group (Ranganathan et al., 2015). Data analysis was performed in Stata 17.0 software (StataCorp, 2021). Survival analysis was performed to investigate the relationship between age of entering care and time in care for Aboriginal children who experienced a care placement between birth and the age of 10 years; results were included in Appendix B (see Supplementary material).

Results

There were 15,815 Aboriginal children born alive in WA between 2000 and 2006. A small proportion of children (1.6%, N = 252) died before reaching their 11th birthday. Half the children in the study cohort were females. The mean maternal age at birth was 25 years and the mean paternal age was 28. At the child’s birth, most mothers were married/de facto (71%) and 25% never married. There were 1,895 (12%) children born in least disadvantaged socioeconomic areas (quintiles 4 and 5), and 7,072 (45%) children born in the most disadvantaged quintile. Almost 40% (N = 6,023) of the cohort was born in major cities and 33% (N = 5,252) in remote and very remote areas.

A total of 9,269 (59%) children did not have any contact with the child protection system between one year prior to birth and their 11th birthday. Conversely, 6,546 (41%) had at least one notification, substantiation, or an OOHC placement by their 11th birthday. For 3,650 (23%) children, the highest level of child protection involvement by their 11th birthday was a notification and 9% (N = 1,491) had at least one substantiation (see Table 1). Of all children in the study cohort, 91% did not experience an OOHC placement; however, 9% (N = 1,405) had at least one placement (see Table 1 and Figure 1). There were 20 children who entered care without prior involvement with child protection. The reason for care entry for almost half of these children at their first placement was determined as “caregiver can’t care adequately”. The rest entered for other reasons as caregiver mental illness or homelessness.

Figure 1 The number and percentage of Aboriginal children born in WA between 2000 and 2006 with at least one child protection notification, substantiated notification, and OOHC placement before their 11th birthday

Note: Percentages were calculated using all Aboriginal children born between 2000 and 2006 (N = 15,815).

Table 1 Demographic Characteristics of Aboriginal Children Born Alive Between 2000 and 2006 in WA by Highest Level of Child Protection Involvement Between One Year Prior to Birth and Their 11th Birthday

		No Contact		Notifications		Substantiations		OOHC
		N	%	N	%	N	%	N	%
N		9,269	58.6	3,650	23.1	1,491	9.4	1,405	8.9
Sex
	Male	4,660	50.3	1,819	49.8	738	49.5	723	51.5
	Female	4,609	49.7	1,831	50.2	753	50.5	682	48.5
Maternal age at birth
	<20 years	1,783	19.2	969	26.5	397	26.6	319	22.7
	20–29 years	5,180	55.9	2,012	55.1	857	57.5	796	56.7
	30–39 years	2,181	23.5	640	17.5	225	15.1	279	19.9
	>39 years	125	1.3	29	0.8	12	0.8	11	0.8
Paternal age at birth
	<20 years	724	7.8	375	10.3	148	9.9	109	7.8
	20–29 years	4,160	44.9	1,426	39.1	572	38.4	496	35.3
	30–39 years	2,617	28.2	714	19.6	239	16.0	282	20.1
	>39 years	1,768	19.1	1,135	31.1	532	35.7	518	36.9
Marital status of mother
	Never married	1,987	21.4	1,145	31.4	485	32.5	489	34.8
	Married/de facto	7,040	76.0	2,377	65.1	914	61.3	827	58.9
	Widowed divorced or separated	176	1.9	84	2.3	49	3.3	58	4.1
	Unknown	66	0.7	44	1.2	43	2.9	31	2.2
Socioeconomic status (2006)
	1 (high disadvantage)	3,855	41.6	1,815	49.7	737	49.4	665	47.3
	2	2,227	24.0	847	23.2	320	21.5	327	23.3
	3	1,436	15.5	498	13.6	247	16.6	207	14.7
	4	921	9.9	235	6.4	75	5.0	114	8.1
	5 (low disadvantage)	404	4.4	77	2.1	33	2.2	36	2.6
	missing	426	4.6	178	4.9	79	5.3	56	4.0
Remoteness (2006)
	Major cities	3,397	36.6	1,322	36.2	580	38.9	724	51.5
	Inner Regional	792	8.5	335	9.2	90	6.0	105	7.5
	Outer Regional	1,532	16.5	557	15.3	222	14.9	168	12.0
	Remote	1,736	18.7	699	19.2	271	18.2	231	16.4
	Very remote	1,386	15.0	559	15.3	249	16.7	121	8.6
	missing	426	4.6	178	4.9	79	5.3	56	4.0

Note: See comparison test in Table 1A, Supplementary material, Appendix A.

Females and males were evenly distributed among those with child protection involvement. Children with child protection involvement were more commonly born from a younger mother (<20 years), an older father (>39 years), and a never married mother than those with no contact. The proportion of children from a highly disadvantaged socioeconomic area (1st quintile) was higher for those who had a contact with the child protection system, compared to those with no contact. A higher proportion of children with care placements were born in major cities; however, no difference was found with any other level of child protection involvement (see Table 1A in the Supplementary material).

Age at First Contact with Child Protection

Children who had a notification as the highest level of contact had, on average, their first notification at the age of 5 years (SD:3.13). The mean age at first notification for those with a substantiation was 4 years (SD:2.99) and OOHC placement was 2 years (SD:2.35). Children with first notifications as infants (aged 0) were overrepresented at higher levels of child protection involvement (from notification to OOHC). The proportion of infants that had a notification as the highest level of contact with child protection was similar to the other age groups (see Figure 2). However, for children with the highest level of child protection contact including substantiation and out-of-home care placement, 21% and 41% had their first notification as infants respectively. These percentages were in both cases higher than any other age group. Of all children first notified as infants, 44% (N = 565) were placed in OOHC at least once before their 11th birthday, and this proportion was higher than any other age group. A multinomial logistic regression analysis suggested that for each one-year increase in age at first notification, children were 11% less likely to have a substantiation (RRR:0.89; CI:0.87–0.91) and had 30% lower risk of being placed in care (RRR:0.70; CI:0.68–0.72), compared to only having a notification, before their 11th birthday (see Appendix A in Supplementary material).

Figure 2 Distribution of age at first notification for children born in WA between 2000 and 2006 whose highest level of contact with child protection was (1) notification (2) substantiation and (3) OOHC

Note: Denominators for notifications, substantiations or OOHC for each age group are the total children with notifications, substantiations or OOHC respectively.

Number of Notifications, Substantiations, and Placements

The average number of notifications increased as children's level of child protection involvement increased. Children who had notification as the highest level of involvement had, on average, 4 (SD:2.8; Rg:1–22; Med:2) notifications from birth until their 11th birthday. This compares to 7 (SD:4.7; Rg:1–28; Med:6) notifications on average for those who had substantiation as their highest level of contact and 8 (SD:5.4; Rg:1–30; Med:8) for those placed in OOHC. The average number of substantiations in the follow-up period were 3 (SD:1.4; Rg:1–12; Med:2) and 4 (SD:2.2; Rg:1–16; Med:2) for those with substantiations and care placements as their highest level of involvement respectively. Finally, those who had an OOHC placement had on average 5 (SD:4.1; Rg:1–39; Med:4) distinct placements, of varied duration, between birth and their 11th birthday.

Type of Maltreatment at the First Notification

For children who had any child protection contact, the most prevalent maltreatment type at the first notification was neglect at 35% (N = 2,290). This was followed by physical abuse at 16% (N = 1,065) and emotional abuse at 15% (N = 1,001). However, for children with notifications as the highest level of child protection involvement, neglect was slightly lower than the average, at 29%, compared to 36% and 48% of those with substantiations and OOHC placements as the highest level of contact respectively.

Children in Out-of-Home Care

A total of 1,405 Aboriginal children in the cohort had at least one OOHC placement between birth and their 11th birthday. Most children in OOHC had only one period of care (87%, N = 1,229) and spent on average 5 years in care (SD:3.5; Rg:0.1–11.0; Med:3.7). The remaining 13% (N = 176) had up to 4 periods of care within the study period. From entry up to their 11th birthday, 12% (N = 163) of children placed in OOHC had only one distinct placement, 59% (N = 826) had 2–5, 21% (N = 291) had 6–10, and 9% (N = 117) had over 10 placements.

At the end of their first period of care, 54% (N = 752) of children had remained in care until their 11th birthday, 25% (N = 357) were reunified with their family, and 21% (N = 296) exited care for other reasons (see Figure 3). Of the 653 children who exited care (357 reunified and 296 for other reasons), a total of 176 (27%, 13% of OOHC) children returned for a second period in care, including 121 (32% of the 375) children who were reunified with their family and 55 (19% of the 296) children who exited care for other reasons at the end of the first period care. Of the 176 children with a second period of care, 64% (N = 113, 8% of OOHC) continued in care until the end of the follow-up, and 17% (N = 30, 2% of OOHC) were reunified with their family. The majority of those with 3 and 4 periods of care remained in care until the end of the follow-up (see Figure 3).

Figure 3 Number and percentage of children at each stage of OOHC placement pathways through care for Aboriginal children born in WA between 2000 and 2006, from birth to their 11th birthday

Note: Percentages were calculated using the number of Aboriginal children placed in OOHC (N = 1,405).

Time in Care

On average, children placed in OOHC between birth and their 11th birthday spent almost 5 years (SD:3.4; Rg:0.1–11.0; Med:4.1) in care. However, children who were placed in care earlier were more likely to spend a longer period of time in care. Time-to-event analysis was conducted to assess the association of age at first entry into care and time in care, accounting for exposure time (see Appendix B in Supplementary material). The analysis suggested that for each one-year increase in age at care entry, children were 9% more likely to leave care before the age of 11 years (HR:1.09; CI:1.05–1.13).

Remoteness

The majority of children in OOHC were born in major cities (52%, N = 724), a lesser number were born in remote areas (16%, N = 231), and only 12% of children in OOHC were born in outer regional areas (N = 168). Children born in major cities were overrepresented within those who had a care placement before their 11th birthday, at 12%, compared to 9% of the overall cohort (see Figure 4). The rest of the regions showed a lower percentage than the average.

Figure 4 Percentage of Aboriginal children born in WA between 2000 and 2006 that were placed in OOHC between 0 and 10 years old by the remoteness or otherwise of the area in which they lived

Placement Type

At first placement, the most common placement type was foster care at 45% (N = 627), followed by kinship/relative care at 21% (N = 302), and residential care at 17% (N = 246). However, when the predominant placement type was examined, these proportions were 56% (N = 784) of children were predominantly placed in kinship/relative care, 26% (N = 360) were predominantly placed in foster care, and 9% (N = 131) in residential care. When analysing type of placement by age at first placement (see Figure 5) for the three most prevalent placement types (foster, kinship/relative, and residential care), there is a higher proportion of children placed in foster care at any age of entry. Foster care also appears to be more prevalent for those entering as infants, and this is true for first (65%) and predominate (42%) placement types. At first placement, children are placed in foster care at a higher rate when they enter care younger, and the percentage decreases as they enter care as older children. Even though this is also true for the predominant placement type, children are predominantly placed in kinship care at any age, and at a higher rate when they enter care at an older age. It is important to note that 86% (N = 899) of children who were ever placed in a kinship/relative placement were placed with an Aboriginal carer.

Figure 5 First and predominant placement type. The percentage of children placed in foster, kinship, and residential care by age at first placement

Placement Outcome by a Child’s 11th Birthday

By the end of the follow-up, two thirds of all Aboriginal children placed in care were still in care. Nineteen percent were reunified with their family at least once (some were reunified and reentered care, see Figure 3). Nineteen percent had left care for other reasons as outlined in Table 2.

Table 2 OOHC Placement Outcome by the Child’s 11th Birthday Among Aboriginal Children Placed in OOHC at Least Once

		N	%	Incidence rate per 1,000 children
In care at end of follow-up		878	62.5	624.9
Exited care at end of follow-up		527	37.5	375.2
	Reunified with family	263	18.7	187.2
	Legal order expired	112	8.0	79.7
	Special Guardianship Order	68	4.8	48.4
	Transfer of guardianship	48	3.4	34.2
	Service no longer required / child adopted	11	0.8	7.9
	Other	25	1.8	17.8
Total		1,405	100.0	1,000.0

Discussion

Our research investigating the longitudinal child protection pathways for 15,815 Aboriginal children born between 2000 and 2006 in WA found that 3 in 5 Aboriginal children had not had any contact with the child protection system, and 9 in 10 children were not placed in OOHC, by their 11th birthday. In total, 6,546 (41%) Aboriginal children were involved with the child protection system at least once before their 11th birthday: 3,650 (23%) only received a notification to child protection services, 2,491 (9%) received a substantiation but remained with their family, and 1,405 (9%) were removed into OOHC. By the age of 10, 62% of those who entered care were still in care, 19% were reunified with their family, and 19% left care for other reasons including guardianship.

Our finding that 41% of Aboriginal children born in WA were notified to the child protection system before their 11th birthday is consistent with findings that 40–55% of Aboriginal children born in South Australia were the subject of a notification from birth to age 18 (Segal et al., 2019). The extraordinarily high overrepresentation of Aboriginal children reported to child protection systems during their childhoods in multiple Australian jurisdictions highlights the urgency to address the Closing the Gap priority outcome to reduce this overrepresentation, which is also a priority in the National Framework for Protecting Australia’s Children, 2021–2031 (Coalition of Peaks, 2020; Commonwealth of Australia, 2021). SNAICC, the National Voice for our Children, have argued that strategies need to be Aboriginal-led and address the drivers of vulnerability for families and social determinants including intergenerational trauma, housing and homelessness, poverty, mental health, substance issues, and cultural connection (SNAICC, 2017). These factors are broader than just one government portfolio and will require partnerships and creative funding solutions to provide wholistic responses to support families and communities. According to SNAICC, place-based approaches are also needed to ensure responses are community driven, taking into account context and resources particularly in regional and remote areas (SNAICC, 2016).

Of the 9% of children placed into care, a high proportion will remain in care or returned to care following family reunification, highlighting the need to examine family reunification strategies, their cultural effectiveness, and the potential enablers and barriers to success (Fernandez & Lee, 2013). Our findings also highlight that infants who were notified were more likely than any other age group to have substantiated notifications and be placed in OOHC. For children who received a substantiation as their highest level of involvement, 21% were first notified as infants, and for children in OOHC, 41% were first notified as infants. This finding emphasises infancy as a key point of system involvement where early intervention and family support strategies may be effective in preventing first time involvement with the child protection system and early return of infants to their parents with community-led supportive care, for mothers in particular. Chamberlain et al. (2022) have outlined a framework for an early intervention response for children and families. Their key recommendations align with the Family Matters Building Blocks and the Aboriginal and Torres Strait Child Placement Principles providing an instrumental guiding framework for how a culturally responsive model of care could be provided to meet the needs of Aboriginal families.

The WA government in 2016 implemented an Early Intervention and Family Support strategy with some of the initiatives including an Intensive Family Support Service program for both Aboriginal and non-Aboriginal families whose children were at risk of removal and for families working towards reunification (Department for Child Protection and Family Support, 2016). Other initiatives have included an Aboriginal In-home Support Service that is delivered by an Aboriginal Community Controlled Organisations (SVA Consulting, 2021). More recently the government has funded an Aboriginal Family-Led Decision Making pilot for families involved in child protection that has many elements of existing models in Victoria and Queensland within it, and the evaluation of this pilot’s outcomes will be important.

Of the 9% of Aboriginal children placed into care, almost half were placed in foster care as their first placement, almost 1 in 5 were first placed in residential care, and 1 in 5 were first placed in kinship care. Among the 343 (24%) children placed in care as infants, almost two thirds were first placed in foster care and only 17% placed with kinship carers. However, the proportion of placements that were foster placements declined for children first entering care at an older age. When we investigated the predominant placement type, kinship/relative care was the predominant placement type for 56% (N = 784) of children. Again, for infants this was very pronounced with a high percentage of children having a predominant placement type of kinship care (48%) despite the fact that only 17% of infants had their first placement with kinship carers. This raises questions as to why kinship carers are not being predominantly utilised for the first placement of children, particularly infants, and reflects the need for improved monitoring and adherence to the Aboriginal Child Placement Principle and cultural planning (Krakouer et al., 2018). In terms of providing ongoing permanency of a placement and ensuring kinship connection, this is an opportunity at the first placement to prioritise the family and kin relationships for placement. It is hoped that the 2021 amendments to the Children and Community Service Amendement Act 2021 (WA) will strengthen the application of the Aboriginal Child Placement Principle and cultural support planning.

Children placed earlier in care were more likely to spend a longer period of time in care. Our analyses found that if children entered care at an older age they were more likely to leave care before the age of 11 years, therefore spending a shorter period of time in care. Our findings highlight that early involvement in the child protection system will result in a higher likelihood of entering care and remaining in the care system. Previous analysis of WA data has also indicated the early involvement of children, particularly infants, and the rise in substantiated cases of neglect and emotional abuse over time (Bilson et al., 2017). The early age of involvement with the child protection system is an opportunity for early intervention and family support strategies to reduce care entry and increase opportunities for family reunification (Department for Child Protection and Family Support, 2016). Greater evidence is required of the outcomes of strategies and programs providing early intervention and family support to ensure that effective initiatives can be scaled up and appropriately resourced. In addition, further research on programs that address poverty and housing stability and the impact it may have on reducing the risk of child protection involvement is needed.

Limitations

Administrative data allow the longitudinal tracking of child protection involvement, for very large numbers of children and is inclusive of all the population. However, the child protection data do not include information about informal care arrangements and hence it is not possible to accurately ascertain the support needs of all children and their families. In 2009–2010 there are estimates that in 16,000 Australian families, grandparents were raising grandchildren (Buckmaster, 2014). The legacy of colonisation, the stolen generation, stigma and fear of child protection involvement mean that Aboriginal families are reluctant to seek formal care arrangements through child protection (Commonwealth of Australia, 2014). In addition, this research could not determine the reasons for readmission to care following family reunification and this requires further investigation.

Conclusion

Most Aboriginal children born in WA in 2000–2006 were not notified to child protection and 9 in 10 children were not placed in OOHC before their 11th birthday. However, Australia faces an urgent and important need to reduce the overrepresentation of Aboriginal children involved in child protection, with our research finding 2 in 5 Aboriginal children being notified to child protection by their 11th birthday. This has been determined as a national priority in both the National Agreement on Closing the Gap and the Safe and Supported: National Framework for Protecting Australia’s Children. The challenge of achieving this outcome needs to ensure that Aboriginal-led strategies are implemented, governments at all levels work in partnership with communities and families, wholistic strategies are supported that go beyond one government portfolio or organisational sector, and creative funding solutions are utilised that meet the needs of families.

Acknowledgements

The authors acknowledge the partnership of the Western Australian Government Departments of Health, Communities, Education and Justice which provided data for this project. They thank the Western Australian Data Linkage Branch of the Department of Health for linking the data. The views expressed are not those of the Government Departments that have provided data for this project and any omissions are the responsibility of the authors. The authors would also like to thank the Western Australian Aboriginal Health Ethics Committee (943) and the Department of Health Human Research Ethics Committee (RGS 3496), who provided ethics approval for this research. The authors acknowledge the Traditional Custodians of the land on which this study is set, and pay their respects to Elders past, present, and emerging. The I-CaRe-WA study is an Aboriginal-led study that engages Aboriginal voices through its leadership and through study advisory, policy, and community reference groups, seeking feedback on ongoing and future research. The authors would like to thank the members of these groups for their feedback and contributions to this article.

Disclosure Statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This project was funded by the National Health and Medical Research Council (NHMRC) Project Grant number 1162968. Emily Banks is supported by the NHMRC Grant number 1136128.