The Australian Family System

Part One: how it works now. A description of every system that can act on an Australian family — violence, police, protection orders, the family courts, child support, child protection, out-of-home care and adoption — in the order a family meets them, with the national figures and what each system does not record. Part Two sets out what should change.

1. Who governs what

Three layers operate at once.

The Commonwealth sets family law for married couples under the Family Law Act 1975, assesses and collects child support through Services Australia, pays Family Tax Benefit and parenting payments, and funds the national data collections held by the Australian Bureau of Statistics, the Australian Institute of Health and Welfare and the Australian Institute of Criminology.

The states and territories run everything else: police, protection orders, child protection departments, children’s courts, out-of-home care and carer payments, and adoption. Each has its own Act, its own terminology and its own reporting. An order that removes a person from their home is a state instrument in every jurisdiction, and no two states name it the same way.

Western Australia is the one structural exception in family law. It is the only state that did not refer its family law power to the Commonwealth, so parenting and property matters are heard in the Family Court of Western Australia, which applies the federal Act to married couples and the state’s own Family Court Act 1997 to de facto couples. Protection orders are Family Violence Restraining Orders under the Restraining Orders Act 1997. Child protection is administered by the Department of Communities. The state also holds its own royal commission power under the Royal Commissions Act 1968.

2. The household before anything happens

Fewer Australians are marrying, and those who do marry later. The crude marriage rate was 5.3 per 1,000 people aged 16 and over in 2025, down from 5.5, with every age-specific rate falling or flat and the largest falls between 25 and 34. The median age at marriage is now 32.7 for men and 31.2 for women. In 1976 it was 24.9.

There were 47,216 divorces granted in 2024, a crude rate of 2.1 per 1,000 — the lowest since the early 1970s. Couples who divorce now separate at a median of 9.3 years and divorce at 13.2. The crude divorce rate is calculated across the population aged 16 and over rather than per marriage, so it falls as the married share of that population falls. At the 2021 Census, 2.17 million people were living in de facto relationships. Those relationships are not registered when they form or when they end, and their separations appear in no national count.

The total fertility rate was 1.481 in 2024, a record low for the second consecutive year, against 1.795 a decade earlier and roughly 3.5 at the 1961 peak. The rate has been below replacement, 2.1, in every year since 1976. The median age of mothers is 32.1 and of fathers 33.9, both record highs. Completed cohort fertility — the number of children a cohort of women finishes with — moved from 2.08 to 2.03 over the same decade. The total fertility rate is a measure of one year’s births and falls when childbearing is deferred, so whether current rates represent smaller completed families cannot be determined until those cohorts finish childbearing.

At the 2021 Census, couples with children were 43.7 per cent of families, couples without children 38.8 per cent, and one-parent families 15.9 per cent. Across the longer run the largest movement has been the growth of couples without children, from about 28 per cent of families in 1976 to about 38 per cent in 2016. One-parent families rose through the 1990s and have been broadly stable since 2001.

3. Family violence

The Australian Bureau of Statistics measures this through the Personal Safety Survey. Since the age of 15, 23 per cent of women — 2.3 million — and 7.3 per cent of men — 692,600 — have experienced violence by an intimate partner or family member. Counting violence and emotional or economic abuse together, 27 per cent of women and 15 per cent of men have experienced partner violence or abuse: 4.2 million adults. Economic abuse alone affects 16 per cent of women and 7.8 per cent of men.

Of intimate partner homicides between 2010 and 2018, 77.2 per cent were a man killing a current or former female partner. Of the 20.9 per cent with a female offender, in 70.8 per cent of cases the woman had been the primary victim of violence in that relationship. In 2024–25 there were 98 domestic homicide victims nationally; 67 per cent of female domestic homicide victims were killed by an intimate partner, against 29 per cent of male victims. The intimate partner homicide rate fell from 0.27 to 0.21 per 100,000 between 2023–24 and 2024–25, and the domestic homicide rate has more than halved since 1989–90.

Death review analysis identifies actual or intended separation as a risk factor for intimate partner homicide, with heightened risk for women who had separated or intended to separate in the weeks and months before the death. The same analysis records identifiable risk factors and repeated prior abuse in most cases.

First Nations women were 33 times more likely to be hospitalised for family violence than other women in 2020–21. Nationally, 88,377 family and domestic violence offenders were proceeded against by police in 2022–23, an increase of 8 per cent. Western Australia recorded more than 42,000 family and domestic violence offences in 2024–25, about 115 a day.

4. First contact: police

From a single attendance, police can lay criminal charges, make a short-term order, apply to a court for a longer order, and notify the child protection department. Each of the four proceeds separately.

In Western Australia a police officer may make a police order lasting 24 or 72 hours without a court, to separate the parties immediately. Police-recorded offence counts are published by each jurisdiction and compiled nationally. What is not published anywhere is what those attendances led to: how many became orders, how many became charges, how many became child protection notifications, and how many households were attended again within a year.

5. Protection orders

A protection order restrains one person from contacting or approaching another, and can exclude them from a home they own or rent. It is a civil order with criminal consequences for breach, and it is the most frequently used instrument in the whole system.

An order can be made on an interim basis without the respondent present, on the applicant’s account alone. The interim power exists because a contested hearing takes time and the risk it addresses is immediate. The order takes effect on service, and the respondent may then object and have the matter heard.

In Western Australia the instrument is a Family Violence Restraining Order under the Restraining Orders Act 1997. A respondent has 21 days from service to object. If no objection is filed, the interim order becomes a final order, ordinarily for two years, without a hearing and without evidence being tested. In New South Wales the equivalent instrument is an Apprehended Domestic Violence Order. Terminology differs in every jurisdiction, and procedures differ with it.

No jurisdiction publishes the number of orders made without the respondent present, the proportion that become final by default rather than by decision, the time from an interim order to a contested hearing, breach rates, or whether the making of an order was followed by a reduction in violence.

6. Separation: four systems begin

At separation, a family with children can be subject to four processes simultaneously: protection orders in a magistrates court; parenting and property in a family court; child support through a Commonwealth agency; and child protection through a state department and children’s court. Each has its own governing law, its own evidentiary standard and its own timetable. None is required to know what the others have decided.

Where a protection order prohibits contact and a family court order provides for it, the protection order generally prevails until varied — so a decision made in a short interim hearing can suspend the effect of a decision that took a year to reach. Child protection operates to statutory response times of two or three working days depending on urgency; the protection order track has no equivalent deadline. And because child support is calculated from the percentage of care, any change to parenting arrangements changes the money, in both directions.

7. Children and parenting matters

Parenting matters are decided by a court applying a best-interests test. Interim orders are made first, sometimes within weeks; final orders follow after a contested process that commonly runs 12 to 24 months, at costs that can exceed $100,000 a party. The interim arrangement governs the child’s living and contact arrangements until final orders are made.

In Western Australia these matters are heard in the Family Court of Western Australia, which also determines de facto property matters under state law.

Courts publish their own performance: filings, clearance rates, time to trial. They do not publish outcomes for families. There is no published measure of how long a child’s contact with a parent was suspended by an interim arrangement, how often it was restored, how many matters returned after final orders, or what proportion of parties were unrepresented.

8. Child support

Child support is federal and uniform. Services Australia applies a statutory formula whose two principal inputs are each parent’s taxable income, drawn from tax records, and the percentage of care each parent provides. It is an administrative calculation, not a judicial one, and it is invisible to the courts deciding the same family’s arrangements.

Two failure modes are documented. A paying parent whose income falls remains assessed on the previous figure until the assessment is changed, and debt accrues in the interval. On the other side, the scheme is used as an instrument of continued control: research prepared for the Commonwealth found that 80 per cent of women reported an ex-partner replacing physical abuse with financial abuse after separation, including through child support.

Aggregate collections, caseload and debt are published. The time from an income change to a reassessment is not. Neither is the debt accrued during that interval, nor the proportion of cases in which payment stopped after a parenting dispute.

9. Property and economic outcomes

Property division is decided on contributions and future needs, without a fixed formula. Across the OECD in 2021, 29.3 per cent of children in single-adult households lived in poverty against 8.9 per cent in households with two or more adults; in Australia the figure for children in single-adult households is 35.7 per cent, more than three times the rate for children overall.

Of people who moved away from a violent partner, about 7 in 10 — 597,000 — left property or assets behind.

10. Child protection: the funnel

Child protection is a separate system with its own trigger. Anyone may notify; a range of professionals must. A notification is screened, may be investigated, and an investigation is either substantiated or not.

In 2023–24, about 179,000 children — 1 in 31 Australian children — came into contact with the child protection system. Of those, 39 per cent, about 121,000, were the subject of an investigation. Of the 98,300 investigations finalised, 43 per cent, 42,100, were substantiated and 57 per cent were not. About 20 per cent of substantiated children, 9,600, were placed on a care and protection order within twelve months.

The rate of children who were subjects of notifications rose from 49 to 53 per 1,000 between 2019–20 and 2023–24. The rate of children in out-of-home care over the same period remained close to 8.0 per 1,000. Exposure to family violence is recorded within emotional abuse, which is the most common primary type of substantiated maltreatment. Where a child is exposed to violence by one parent against the other, the substantiation is recorded against the household.

11. Removal and out-of-home care

At 30 June 2024, 59,900 children were on care and protection orders, a rate of 10 per 1,000, and 44,900 children were in out-of-home care, 7.7 per 1,000. Between 94 and 95 per cent of children in out-of-home care are there under an order.

Infants under one have the highest admission rate to out-of-home care, 6.5 per 1,000 against 2.0 or lower at every other age. Rates in regional areas, 12 per 1,000, and remote areas, 14 per 1,000, are more than double those in major cities, 6.0. Of children in care whose disability status is known, 21 per cent have a disability.

Aboriginal and Torres Strait Islander children are in out-of-home care at 50 per 1,000 against 4.6 per 1,000 for non-Indigenous children — 11 times the rate, and higher than the 9.6 times recorded in 2020. There are 20,000 Aboriginal children in care and 25,000 on orders. The Aboriginal and Torres Strait Islander Child Placement Principle governs where a removed child is placed, and 55 per cent are placed with relatives or kin. It does not apply to the decision to remove.

12. Where a removed child goes

Placement follows a hierarchy: relatives or kin first, then authorised foster carers, then family group homes, then residential care staffed on a roster, then independent living for older teenagers. About 89 per cent of children in care are in home-based care, meaning with kin or with foster carers.

Foster carer households numbered about 9,000 in 2021, 8,300 in 2023 and 8,000 in 2024 — falling, against a child population in care that is not. Relative and kinship carer households are roughly twice as numerous, at about 15,800. About 24,200 carer households held a placement in 2024; 60 per cent had one child and 38 per cent had two to four.

Residential care is what remains when a home-based placement cannot be found or has broken down. It costs substantially more per child and its residents skew older, though the number of younger children in residential care has been rising.

In 2023–24, 667 First Nations children were the subject of a substantiation of abuse while in out-of-home care, which the Institute of Health and Welfare reports as 45 per cent of all such substantiations — implying roughly 1,500 children abused while in the care of the state in a single year.

13. Adoption, and leaving care

There were 155 adoptions finalised in Australia in 2024–25 — 136 domestic and 19 intercountry — against 44,900 children in care. The number has fallen 41 per cent in five years and about 98 per cent from the 1971–72 peak of 9,798. Roughly five in six domestic adoptions are “known child” adoptions, in which the adopting parent already has the child: a step-parent, or a long-term foster carer. Intercountry adoption has fallen from a peak of 434 to 19, with a median wait of about four years.

Carer adoptions from out-of-home care are overwhelmingly a New South Wales practice; in one recent year New South Wales accounted for 131 of 143 nationally, following legislation establishing a permanency hierarchy that reaches open adoption. Most other states rarely use it. Whether the children in the two systems fared differently has not been published.

Australia moved away from adoption following the forced adoption era, which drew a National Apology in 2013, and the Stolen Generations. At 30 June 2024 there were 7,766 contact vetoes in place. Fewer than 130 Aboriginal children were adopted in the twenty-five years to 2020–21. Permanency is now pursued instead through long-term guardianship orders, which provide a stable placement without severing the legal relationship to the birth family.

Care ends at 18, though jurisdictions have been extending support to 21. Of children reunified with family in 2022–23, 85 per cent of First Nations children and 88 per cent of non-Indigenous children did not return to care within twelve months. National reporting on outcomes after leaving care — housing, education, employment, justice contact, early parenthood — began only recently through the Permanency Outcomes Performance Framework and the national out-of-home care indicators.

14. What is measured, and what is not

Measured well: births, marriages and divorces, as complete registry counts; police-recorded offences; homicides, through death review networks; and child protection, in national detail from notification through investigation, substantiation, orders, placement, abuse in care and reunification.

Not measured: whether a protection order was followed by less violence; how many orders are made without the respondent present, and how many become final by default; the time from an interim parenting arrangement to restoration of contact; the time from an income change to a child support reassessment; the separation of couples who never married; and what happens to any member of the family — parent or child — in the years after any of these systems acts.

Child protection was rebuilt following royal commissions and reports nationally against agreed indicators, including the Permanency Outcomes Performance Framework and the national out-of-home care indicators. The protection order system has not been through an equivalent process and reports against no national indicator set. Neither an applicant nor a respondent can be told from published data what orders of that kind have produced.

The case for a Western Australian royal commission across all four systems is set out in Memo 16. What should change is Part Two of this memo.

15. Sources

1. Australian Bureau of Statistics, Marriages and Divorces, Australia (2024 and 2025 releases) — the crude marriage rate of 5.3 per 1,000 in 2025 and 5.5 in 2024, age-specific rates, median ages at marriage of 32.7 and 31.2 against 24.9 in 1976, 47,216 divorces in 2024 at a crude rate of 2.1, and median durations to separation and divorce. Supports section 2.
2. Australian Bureau of Statistics, Births, Australia (2024, released October 2025) — the total fertility rate of 1.481 in 2024 and 1.499 in 2023 as successive record lows, 1.795 in 2014, below replacement in every year since 1976, median ages of mothers 32.1 and fathers 33.9, and completed cohort fertility of 2.03. Supports section 2.
3. Australian Bureau of Statistics, Census of Population and Housing 2021, with long-run comparisons published by the Australian Institute of Family Studies — family composition shares, one-parent families at 15.9 per cent, and 2.17 million people in de facto relationships. Supports section 2.
4. Australian Bureau of Statistics, Personal Safety Survey 2021–22 — 23 per cent of women and 7.3 per cent of men experiencing partner or family violence since the age of 15; 27 per cent and 15 per cent for partner violence or abuse of any kind; economic abuse at 16 and 7.8 per cent. Supports section 3.
5. Australian Domestic and Family Violence Death Review Network and ANROWS, Intimate partner violence homicides 2010–2018, and The “Pathways to intimate partner homicide” project — 77.2 per cent of intimate partner homicides male-perpetrated against a female partner, 70.8 per cent of female-perpetrated cases involving a woman who was the primary victim, heightened risk around actual or intended separation, and the presence of identifiable prior risk factors in most cases. Supports section 3.
6. Australian Institute of Criminology, National Homicide Monitoring Program, as reported by the Australian Institute of Health and Welfare — 98 domestic homicide victims in 2024–25, the intimate partner shares of female and male domestic homicide victims, the fall in the intimate partner homicide rate from 0.27 to 0.21 per 100,000, and the long-run halving of the domestic homicide rate. Supports section 3.
7. Australian Institute of Health and Welfare, Family, domestic and sexual violence — 88,377 family and domestic violence offenders proceeded against in 2022–23; First Nations women 33 times more likely to be hospitalised for family violence in 2020–21; and 597,000 people leaving property or assets behind when moving away from a violent partner. Supports sections 3 and 9.
8. Western Australia Police Force and Western Australian criminal justice reporting — more than 42,000 family and domestic violence offences recorded in 2024–25. Supports section 3.
9. Restraining Orders Act 1997 (WA); Family Court Act 1997 (WA); Family Law Act 1975 (Cth); Royal Commissions Act 1968 (WA) — the instruments, the 21-day objection period and the default conversion of an interim order to a final order, the Family Court of Western Australia’s jurisdiction over married and de facto matters, and the state royal commission power. Supports sections 1, 5 and 7.
10. Department of the Prime Minister and Cabinet, research on economic abuse after separation — 80 per cent of women reporting an ex-partner replacing physical abuse with financial abuse, including through the child support system. Supports section 8.
11. OECD, Society at a Glance and related child poverty reporting (2025), as compiled in Australian analysis of the economic effects of separation — 29.3 per cent of children in single-adult households in poverty across the OECD in 2021 against 8.9 per cent in households with two or more adults, and 35.7 per cent in Australia. Supports section 9.
12. Australian Institute of Health and Welfare, Child protection Australia 2023–24 and its updates (Insights, June 2025; Aboriginal and Torres Strait Islander children, September 2025; Safety of children in care, December 2025; Pathways from out-of-home care, March 2026) — 179,000 children in contact with the system, the investigation and substantiation counts, the rise in notification rates from 49 to 53 per 1,000, 59,900 children on orders and 44,900 in out-of-home care, age and remoteness admission rates, Aboriginal rates of 50 against 4.6 per 1,000, placement with kin, carer household counts, 667 First Nations children subject to substantiated abuse in care representing 45 per cent of the total, and reunification rates of 85 and 88 per cent. Supports sections 10, 11, 12 and 13.
13. Australian Institute of Health and Welfare, Adoptions Australia (2023–24 and 2024–25) — 155 adoptions in 2024–25, the domestic and intercountry split, known child adoptions as the large majority, the 1971–72 peak of 9,798, the intercountry peak of 434 and median wait of about four years, the concentration of carer adoptions in New South Wales, 7,766 contact vetoes, and adoptions of Aboriginal children. Supports section 13.