The over-representation of young Aboriginal or Torres Strait Islander people in the South Australian child system: A longitudinal analysis

https://doi.org/10.1016/j.childyouth.2010.06.011Get rights and content

Abstract

Although it well established that Aboriginal or Torres Strait Islander (ATSI) children are over-represented in the Australian child-welfare system, most conclusions are drawn from crosssectional analyses that are unable to examine the incidence of cases over time or cumulative impacts. In this paper, we summarise the longitudinal and comparative analysis of data relating to children receiving their first notification in 1991 as well as comparisons with children born in 1998 and 2002. The results show that ATSI children's involvement in the child welfare system is even higher when examined longitudinally. Over 50% of ATSI children born in 1991 had contact with this system by the age of 16 and this figure was already exceeded for 4 year olds in the 2002 cohort. ATSI children were found to be more likely to receive more serious notifications and to have their abuse substantiated. The findings are discussed in relation to recent studies of disproportionality and disparity in the treatment of minority children by child welfare systems.

Introduction

A consistent finding in international child protection research is that minority or indigenous children tend to be significantly over-represented in official child welfare statistics. This pattern is observed in the United States for children of African-American background, in first-nations populations in Canada, and in Maori and Pacific-islander populations in New Zealand (Tilbury & Thoburn, 2009). Very similar patterns have been observed amongst Aboriginal and Torres Strait Islander (ATSI)1 people in Australia, although it only more recently that this population has begun to receive significant international attention (Tilbury, 2009). ATSI people represent only 2–3% of the total population and around 4–5% of the child population, but this group is significantly more likely to attract child protection notifications, be investigated and to have abuse substantiated compared with their non-indigenous counterparts (Australian Institute of Health and Welfare (AIHW), 2009, Holzer and Bromfield, 2008). In 2007–08, the abuse substantiation rate for ATSI children was 35.3 for every 1000 indigenous children aged 0–16 compared with only 5.5 for non-indigenous children. Twenty-nine percent of all out-of-home care placements involved ATSI children (a placement rate of 29 per 1000 indigenous children in the general population vs. only 4/1000 for non-ATSI children (Australian Institute of Health and Welfare (AIHW), 2009, Berlyn and Bromfield, 2009). As with other minority groups such as African-Americans in US, ATSI children are not only more likely to enter care, but also to stay in care for longer (Benedict et al., 1987, Barber et al., 2000, Courtney and Skyles, 2003, Delfabbro et al., 2003, Harris and Courtney, 2003, Jenkins and Diamond, 1985).

These variations are considered controversial because of Australia's history of systematically removing ATSI children from their families during the early and middle parts of the 20th Century. Up until 30 years ago, paternalistic, often racist and discriminatory policies existed to remove (or solicit removals) of ATSI children from their parents to be raised and educated elsewhere (often termed “The Stolen Generations”). Very often, the purported educational and safety benefits of these policies were greatly out-weighted by the grief, trauma and dislocation brought about by the separation from family and culture (Human Rights and Equal Opportunity Commission, 1997). Not surprisingly, if large numbers of children continue to live away from their families because of child protection concerns, there are concerns that the negative effects of previous “stolen generations” policies are continuing to be replicated but merely under a different policy framework.

In most recent research, the term disproportionality is usually applied to describe the statistical over-representation of minority children in child welfare systems (Ards et al., 2003, Courtney and Skyles, 2003, Courtney et al., 1996, Cross, 2008, Drake et al., 2009, Fluke et al., 2003, Hines et al., 2004). Disproportionately is usually described as a ratio based on the relative proportions of each population (minority/non-minority) involved in the child welfare system, e.g., 16%/2% = 8.0. This term is usually differentiated from disparity (Tilbury & Thoburn, 2009) that relates more to the extent to which child welfare systems treats minority children equally or equitably. For example, if minority children with the same level of need or severity of problems are equally likely to receive the same outcomes in the system, a disparity is not said to exist, even if more minority children ultimately get notified or come into care. By contrast, if similar groups of children are investigated and minority children are more likely to enter care, a disparity is said to exist and this can be meaningfully quantified (Shaw, Putnam-Hornstein, Magruder, & Needell, 2008).

As Tilbury and Thoburn (2009) point out, the existence of disproportionality per se is not necessarily a reflection of poor policy. It may reflect genuine differences in social disadvantage. It is well established that ATSI communities in Australia have high rates of unemployment, poverty, domestic violence, substance abuse and other social problems. Indeed, as Bromfield and Higgins (2004) further argue, it is likely that official figures may understate the true extent of abuse. Abuse may often go undetected in ATSI communities because of less surveillance, a reluctance by authorities to report abuse, secrecy and shame in the community, or a lack of awareness of what constitutes abuse. Nevertheless, despite this, Tilbury (2009) believes that child protection systems often lead to disparities in the treatment of ATSI people. Services may not be culturally sensitive, parenting may be judged by western standards so that extended family structures (common in the ATSI communities) are afforded little importance, and there may be a generalised prejudicial assumption that ATSI people are less capable of providing the sort of care that is in the best interests of children. On some occasions, authorities may look for problems because they have come to expect them to exist. All of these factors may conspire to ensure that ATSI people are more likely to have their children come into care and/or be regarded as less suitable for reunification with their families. Many similar effects are thought to affect the treatment of African-American in the US, although recent examinations of these issues suggest that these issues are more complex than previously thought (Garland, Ellis-MacLeod, Landsverk, Granger, & Johnson, 1998) or are abating over time (Wulczyn, 2003).

A limitation of much of what has been written about the status of ATSI children in the Australian child welfare system is that data have typically been cross-sectional (e.g., AIHW, 2009). That is, the official statistics are based on the number of children who receive notifications, are investigated, or come into care within a specified interval (usually a financial year). Although useful, such figures do not allow one to examine the incidence of new notifications each year or how individual children might be affected by the system over time. Cross-sectional analyses merely count the cases in the system each year and this could include the same or different children. Analyses of this nature also do not indicate the age at which the system first begins to have an impact on children's lives because notification rates per individual child are typically expressed as a function of all children aged under 18 or 16 years of age rather than a consistent age-cohort (e.g., how do ATSI three year olds in contact with child welfare services compare with the rate expected in the population of three year olds?).

Investigation of the long-term and cumulative impact of the system is best achieved using a longitudinal approach in which statistics are examined for a consistent birth cohort tracked over time. Such approaches have, for example, been utilised in the United States by Sabol, Coulton, and Polousky (2004) as a way to highlight the disparities between the levels of child protection involvement in African-American as opposed to other population groups. In Sabol et al.'s study, the researchers applied a life-table approach in which they ascertained the probability of a child protection (e.g., a report or substantiation) occurring by the time a given cohort of children had researched ten years of age. Using these techniques, the researchers showed that 49% of African-American children (vs. 30% of Whites) had attracted reports because of suspected abuse or neglect by the age of 10. A third (of the whole population) had substantiated abuse or neglect (vs. 11% of Whites).

Similar techniques were used in the present project which was designed to examine the issue of disproportionality using birth cohort data in Australia. In this research, the cumulative impact of the child protection system in terms of its level of involvement is examined over time for a cohort of infants born in South Australia2 in 1991 and where data is available until they were 16 years of age. Comparative involvement (notification) rates for ATSI and non-ATSI children are presented along with analysis of differences in the type of abuse detected or substantiated. A final part of the study then examined whether the rates of child abuse notifications have changed over time due to a gradual strengthening of mandatory reporting provisions during the last decade. It was hypothesised, based upon Tilbury's (2009) recent analysis, that ATSI children would: (a) typically come into contact with the child welfare system earlier in life, (b) receive a greater number of notifications and substantiations over their life-time, (c) be notified for more serious forms of abuse, and (d) that these trends would strengthen over time due to the increasing expansion of mandatory reporting provisions.

Section snippets

Sampling procedure

The data for this study were drawn from the South Australian Families SA Client Information System in 2007. This system records details of all child protection notifications along with identifiers that allow the differentiation of individual children. Since the principal focus of the analysis was on the cumulative influence of child protection interventions during childhood and adolescence, the study focused on children born in 1991 because this allowed 17 years of longitudinal data to be

Overall notification rates

As of January 2007, 4410 children born in 1991 had been the subject of at least one child protection notification (22.5% of the birth cohort). Ten percent of these children were recorded as being Aboriginal or Torres Strait Islander (Table 1). This is an over-representation of ATSI children in the notified population of approximately three times.

Analysis was conducted modelling three different possible birth rates (3%, 3.5% and 4%) of live births being Aboriginal and Torres Strait Islander (

Discussion

The findings from this study confirm many of the conclusions recently articulated by Tilbury (2009) concerning the status of Australian Aboriginal or Torres Strait Islander people in the child welfare system. As shown in this paper, over 50% of Aboriginal or Torres Strait Islander children born in 1991 had received at least one notification by the time they were around 16 years of age and this figure is far higher than the figure of 15% (Tilbury, 2009) one typically finds if analyses are

References (25)

  • Australian Institute of Health and Welfare (AIHW)

    Child protection Australia 2007–08

    (2009)
  • J.G. Barber et al.

    Aboriginal and non-Aboriginal children in out-of-home care

    Children Australia

    (2000)
  • Cited by (36)

    • Lifetime risk of child protection system involvement in South Australia for Aboriginal and non-Aboriginal children, 1986–2017 using linked administrative data

      2019, Child Abuse and Neglect
      Citation Excerpt :

      This represented a considerable portion of children to enter OOHC in the 1990s but has become uncommon in recent cohorts. Large differences in CP involvement were identified between Aboriginal and non-Aboriginal children, across all cohorts, consistent with other Australian data (Bilson et al., 2015; Delfabbro et al., 2010; Productivity Commission, 2018) and the international literature (Barker et al., 2014). While SES explains part of this difference, it by no means explains all or even most.

    • Children returning from care: The challenging circumstances of parents in poverty

      2019, Children and Youth Services Review
      Citation Excerpt :

      For instance caucasian children were twice to four times as likely than children of colour to reunify (Harris & Courtney, 2003; Hill, 2003; McMurty & Lie, 1992; Wells & Guo, 1999). Indigenous children in Australia, who have been found to be more affected by poverty and social disadvantage in a number of studies, experience similar trends in achieving reunification (Delfabbro, Hirte, Rogers, & Wilson, 2010; Fernandez & Delfabbro, 2010). Family structure prior to entry to care has also been found to be related to reunification outcomes.

    • Adoption and child protection trends for children aged under five in England: Increasing investigations and hidden separation of children from their parents

      2019, Children and Youth Services Review
      Citation Excerpt :

      Child protection investigations have increased rapidly in Australia (42% between 2012 and 2016), Canada (74% between 1998 and 2008), the US (17% between 2007 and 2015; all rates cited in Bilson, 2016), and in England (122% between 2010 and 2018; DfE, 2018c). A large and growing proportion of the child population face the prospect of their family being investigated for suspected child abuse or neglect (37% in the US, Kim et al. 2017; 22.5% in South Australia, Delfabbro et al., 2010; 23.5% in New Zealand, Rouland & Vaithianathan, 2018). This study provides an anatomy of the investigative turn in England.

    View all citing articles on Scopus
    View full text