Foster care and child welfare law (United States, 1854 to 2024)

Three federal laws (1980, 1997, 2018) set the rules for removing children, time limits and prevention funding; national counts show about 329,000 children in care in 2024 and large racial differences in who is removed.

The record

Context: the Orphan Trains

• From 1854 to 1929 the Children's Aid Society (founded by Rev. Charles Loring Brace) and the New York Foundling Hospital sent children from eastern cities to families in other states. A Virginia Commonwealth University social welfare history page gives about 200,000 children in its text and "a quarter million" in a photo caption, and does not reconcile them. It reports criticism that some farm families saw the children as "cheap labor," that some children were abused, and that there was "no formal process" and "mixed" outcomes. It calls the program a "precursor to the modern foster care system." (VCU Social Welfare History Project, Orphan Trains; opened)
• The U.S. Children's Bureau was established in 1912 (same page).
• The Indian Child Welfare Act of 1978 governs removal of Native children and is covered in Case 56.

Timeline of federal law

The 1980 law

• States must ensure "reasonable efforts will be made" both before placing a child in foster care, to prevent removal, and to make it possible for the child to return home. A judge must find that such efforts were made for foster care payments to be made. Each child must have a written case plan; status reviewed "no less frequently than once every six months"; and a dispositional hearing "no later than eighteen months after the original placement." It also created adoption assistance payments for children with "special needs," capped at what foster care in a family home would have cost. (GovInfo, 94 Stat. 500; opened)

The 1997 law (ASFA)

• The child's "health and safety shall be the paramount concern." Reasonable efforts to prevent removal and to reunify families are generally required, but not when a court finds aggravated circumstances such as abandonment, torture, chronic abuse or sexual abuse. When a child has been in foster care for 15 of the most recent 22 months, the state must file a petition to terminate parental rights, or join one, and seek an adoptive family at the same time. Exceptions: the child is with a relative (state option); the case plan documents a compelling reason that termination is not in the child's best interests; or the state has not provided services it considers necessary for safe return. The law paid states $4,000 for each foster child adoption above a base number and $2,000 for each special needs adoption above a base number, with $20 million a year authorized for fiscal 1999 to 2003. (GovInfo, 111 Stat. 2115; opened)

The 2018 law (Family First)

• Creates optional federal title IV-E funding for up to one year of mental health, substance use and in-home parenting services for children who are candidates for foster care and their parents or kin caregivers. It limits federal payment for children in group care institutions to two weeks, except for listed settings including qualified residential treatment programs (which must meet trauma-informed, family participation and at least six-month aftercare standards) and family-based residential substance use treatment. The page also reports a delay in the adoption assistance phase-in until June 30, 2024. (Child Welfare Information Gateway; opened; a government summary, not the statute)

National counts (AFCARS, the federal child welfare data system)

• The "waiting" rows use different labels in the two reports and may count different groups. We do not compare them. FFY 2022 also lists 64,561 waiting children whose parental rights had been terminated. (FFY 2022 report, copy hosted by The Center for Children; FFY 2024 AFCARS dashboard; both opened. The first is a third-party copy of a federal report. Washington and Wyoming are not included in the 2024 dashboard.)
• We did not use the FFY 2022 race and ethnicity table: our extraction of its columns was inconsistent (labels and numbers did not line up), so we rely on the labeled 2024 table below.

Racial disproportionality: documented numbers (FFY 2024, children ages 0 to 17, rates per 1,000 children)

• From these rates, the rate for Black children in care is about 2.0 times the White rate (8.06 divided by 3.98), and for American Indian/Alaska Native children about 3.7 times. For Hispanic children it is about 0.9. These ratios are our arithmetic from the dashboard. (AFCARS dashboard; opened)
• Investigations: a study by Kim, Wildeman, Johnson and Gilbert (American Journal of Public Health, 2017), reported by Washington University, found 37% of U.S. children were the subject of an investigated maltreatment report before age 18, and 53% of Black children, using national data for 2003 to 2014. (Washington University; opened. The study itself was not read.)
• Why the differences exist is contested. Drake, Jonson-Reid and coauthors (2021 book chapter, abstract only) argue that a roughly two-to-one reporting disparity for Black children "disappears or even reverses when income and other risk factors are controlled for," meaning poverty largely explains it. Other researchers argue that bias in reporting and decisions plays a role. We did not open a study taking that side. (WashU profile of the chapter; opened)

Why children come in: maltreatment data

• Child Maltreatment 2022 (Children's Bureau): 3,096,101 children received an investigation or alternative response (42.4 per 1,000); 558,899 were victims (7.7 per 1,000); 74.3% of victims experienced neglect; 70.0% of reports came from professionals; 1,990 child fatalities (national estimate). The report does not by itself say how many neglect cases involve poverty. (Child Maltreatment 2022, copy on Better Care Network; opened)

What this case shows

1. Federal law set the clock; families usually met it in a courtroom. ASFA's 15 of 22 months rule and its adoption payments were public policy that voters could debate. Parents involved had short notice of what the clock meant, and the law lists exceptions that decide who is spared. (GovInfo, 111 Stat. 2115)
2. The same law promised both services and speed. The 1980 law required reasonable efforts to keep families together; the 1997 law kept that duty but gave exceptions and added a termination deadline. We found no opened source measuring whether services were delivered. (GovInfo, 94 Stat. 500; 111 Stat. 2115)
3. The numbers show unequal exposure. In FFY 2024 the removal-and-care rate for Black children was about twice the White rate, and for American Indian/Alaska Native children about 3.7 times. Whether this reflects differences in need, in reporting, or in decisions is argued. (AFCARS dashboard; Drake et al.)
4. Prevention was a later and optional fix. Family First (2018) gave states the option to fund services before removal, and limited group-care payments. States chose whether and when to use it. (Child Welfare Information Gateway)
5. The children are the people with the least say. The Orphan Trains and today's system both moved children by adult decision. The 19th-century record shows weak oversight; the modern record shows court review but heavy dependence on caseworker and court judgment. (VCU; 1980 Act)

The strongest objections, and our answers

1. Objection: ASFA protected children who were stuck in care. The law makes safety the "paramount concern" and supporters point to children waiting in care (we opened no source documenting this for 1997). Mitigation: We state the text. The median stay was 15.5 months in FFY 2022, and 46% of exits were to reunification. We do not say the time limit was good or bad.
2. Objection: Removal is a response to real abuse, not bias. 558,899 children were confirmed victims in 2022 and 74.3% of victims suffered neglect. Mitigation: We show that data. Neglect is a legal category that the report does not separate from poverty; we leave that question open.
3. Objection: Racial gaps are poverty gaps. Drake and coauthors say so. Mitigation: We include their finding. We also show the gap in rates and the 53% investigation figure for Black children. We draw no conclusion about intent.
4. Objection: Federal money for prevention is a rule, not a result. Mitigation: Agreed. We report enactment and design, not measured effect, because we opened no evaluation.
5. Objection: The Orphan Trains are not relevant. Mitigation: We use them only as context; the VCU page calls them a precursor, and we do not link them to current outcomes.

Open questions: what would change our minds

• Whether ASFA's time limit raised adoptions or ended reunifications that would have been safe. We opened no evaluation. Unverified.
• Evidence on bias in reporting and decisions (the other side of Drake et al.) was not opened.
• Data on parents' legal representation, notice and hearing outcomes were not opened.
• State-by-state use of Family First prevention funds and any evaluation of results.
• The FFY 2022 race table, which we could not read cleanly, should be re-read from a federal source.
• The statutes were read through summaries and extracted text; a full reading of the Family First text itself was not done (we read a government summary).
• The relation between maltreatment findings and poverty: Child Maltreatment 2022 did not break this out in what we read.
• Whether AFCARS "waiting" counts between years are comparable.

Sources

Evidence status and correction history

Key claims

Correction history

Found an error or a better source? We keep every correction on the record with its date and reason. Nothing here is quietly rewritten.