Indian Child Welfare Act (United States, 1969 to 2023)

Congress passed the Indian Child Welfare Act in 1978 after surveys reported that 25 to 35 percent of Indian children had been separated from their families; the Supreme Court read the law broadly in 1989, narrowly in 2013, and upheld it against a federal challenge in 2023.

The record

What Congress found and what the law requires

• The statute's findings, 25 U.S.C. 1901, say Congress has "plenary power over" Indian affairs, that no resource is "more vital to the continued existence and integrity of" tribes than their children, and that a very large share of Indian families are broken up when nontribal agencies remove children, often without good reason. Many are placed in non-Indian foster and adoptive homes and institutions. The findings also say state bodies have often failed to recognize "the essential tribal relations" of Indian people. The findings text we opened gives no percentage. (25 U.S.C. 1901, Cornell LII; opened)
• 25 U.S.C. 1912 sets rules for involuntary state proceedings: notice to the parent, custodian and tribe (a hearing cannot occur until at least ten days after notice); proof that "active efforts" were made to prevent the breakup of the family; a "clear and convincing" standard with expert testimony for foster care placement; and proof "beyond a reasonable doubt" with expert testimony for termination of parental rights. (25 U.S.C. 1912, Cornell LII; opened)
• 25 U.S.C. 1915 sets placement preferences. For adoption: an extended family member, other members of the child's tribe, or other Indian families, "in the absence of good cause to the contrary." For foster care, a tiered list starting with extended family, with the same "good cause" exception. A tribe may set its own order by resolution. The statute also says the preferences of the child or parent are considered, and a consenting parent's wish for anonymity must be given weight. (25 U.S.C. 1915, Cornell LII; opened)

The surveys behind the law

• In Mississippi Band of Choctaw Indians v. Holyfield (1989), the Court wrote that Association on American Indian Affairs studies in 1969 and 1974, presented in Senate hearings, "showed that 25 to 35% of all Indian children had been separated from their families." It cites H.R. Rep. No. 95-1386, p. 9. The opinion also reports that about 90 percent of Indian placements were in non-Indian homes, and that Indian children were adopted at eight times the rate of non-Indian children. (Holyfield, 490 U.S. 30 (1989), Justia; opened; the last 6,000 characters of the page were not read)
• We did not open the 1969 and 1974 AAIA surveys, the Senate hearings or the House Report. The 25 to 35 percent figure is therefore confirmed only as the Court's description of them. The base group (all Indian children, or children in certain states) and the meaning of "separated" are not stated in the text we read.

Timeline

The three Supreme Court cases

| Case | Question | Result | Vote | |---|---|---| | Mississippi Band of Choctaw v. Holyfield, 490 U.S. 30 (1989) | Did a state court have power over twins born off the reservation to reservation-domiciled parents who agreed to an adoption? | No. The twins were domiciled on the reservation, so the tribal court had exclusive jurisdiction. Brennan wrote for the Court. Stevens dissented, joined by Rehnquist and Kennedy. | 6 to 3 | | Adoptive Couple v. Baby Girl, 570 U.S. 637 (2013) | Did ICWA block termination of a father's rights, and did the placement preferences apply, where he never had custody? | Sections 1912(f) and 1912(d) did not apply, and section 1915(a) preferences did not apply where no eligible party formally sought to adopt. Alito wrote for the Court. Sotomayor dissented, joined by Ginsburg and Kagan, with Scalia joining in part. | 5 to 4 | | Haaland v. Brackeen, No. 21-376 (2023) | Is ICWA beyond Congress's power, a commandeering of the states, an unlawful delegation or a denial of equal protection? | Congress's Article I power upheld; anticommandeering claims rejected; equal protection and nondelegation claims not reached because the challengers lacked standing. Barrett wrote for the Court. | 7 to 2 |

Sources for the table: Justia pages for Holyfield and Adoptive Couple; Cornell LII for Brackeen (all opened; Justia pages truncated at about 100,000 characters).

• Holyfield: the Court held "the law of domicile Congress used in the ICWA cannot be one that permits individual reservation-domiciled tribal members to defeat the tribe's exclusive jurisdiction" by giving birth off the reservation. Congress intended a uniform federal definition of domicile.
• Adoptive Couple: the child was 3/256 Cherokee. Her biological father, a Cherokee Nation member, texted that he relinquished his rights before birth and gave no support in her first four months. South Carolina courts ruled for him, and she was transferred to him at 27 months. The Court read "continued custody" in section 1912(f) to mean "custody that a parent already has (or at least had at some point in the past)." The dissent said this ignores the law's definition of termination, could reach fathers who visit and pay support, and would create a patchwork depending on state law. The child's later circumstances are not in the text we read.
• Brackeen: three sets of families and the State of Texas sued. A.L.M. was placed with the Brackeens as a foster child at about 10 months; the Navajo and Cherokee Nations opposed adoption and the Brackeens later adopted him after the Navajo family withdrew. Hernandez chose the Librettis to adopt her newborn; the Ysleta del Sur Pueblo intervened and later withdrew. The Cliffords fostered Child P., whose grandmother belongs to the White Earth Band of Ojibwe; the court denied adoption under ICWA after the Tribe changed its position on membership. The Court said Congress's power is "plenary within its sphere, but even a sizeable sphere has borders." Section 1912 applies to "any party," including private actors, so it does not commandeer states. The individual petitioners could not show redressability, and Texas could not assert equal protection rights for its citizens. Gorsuch concurred (joined in part by Sotomayor and Jackson) and discussed the history of boarding schools and removals; Kavanaugh also concurred. Thomas and Alito dissented. We did not read the dissents. A footnote of the majority says Alito's dissent criticized the "plenary but not absolute" formulation. (Cornell LII, Brackeen; opened; last 15,000 characters not read as to the dissents)

Child welfare data today

• The National Indian Child Welfare Association (NICWA) 2026 fact sheet reports that Native children are about 0.8 percent of U.S. children (Puzzanchera et al., 2023) and about 3 percent of children placed outside their homes (HHS, 2025), a disproportionality ratio of 3.75. It also reports that, where abuse was reported, Native children were about 2 times more likely to be investigated, 2 times more likely to have allegations substantiated and 4 times more likely to be placed in foster care than White children (Hill, 2007). State examples: Minnesota 13.5 (19 percent of foster care, 1.4 percent of child population), South Dakota 4.4 (55 percent and 12.4 percent). The fact sheet is a tribal advocacy source and its own text uses both "about 3 times" and 3.75. (NICWA fact sheet, 2026; opened)

What this case shows

1. Families and tribes had no seat in the decisions that separated them. The Court's account of the 1969 and 1974 surveys is that a quarter to a third of Indian children were separated from their families, about 90 percent of placements were in non-Indian homes, and state bodies often did not recognize tribal relations. The law answered with notice, a voice for the tribe, and heavier proof before removal. (Holyfield; 25 U.S.C. 1901, 1912)
2. The record shows an affected community asking Congress for a rule, and Congress giving one. The findings tie the law to evidence gathered in Senate hearings. That is a public record ordinary people could read. We did not open the hearings themselves. (Holyfield)
3. Notice and a hearing date are concrete protections. Section 1912(a) requires registered-mail notice to the parent and the tribe, with at least ten days before a hearing. Whether the people affected can use those days is a separate question. (25 U.S.C. 1912)
4. The law has been tested three times in the highest court and reshaped each time. In 1989 the Court protected tribal jurisdiction; in 2013 it limited the reach of two protections; in 2023 it left the structure in place. Each decision turned on text and structure, and the public could read each opinion.
5. The costs fell on specific children and adults. A child in Adoptive Couple moved at 27 months to a father she had not met. Foster and adoptive families in Brackeen waited through litigation. Tribes lost or kept members depending on the reading. These are real costs on every side. (Adoptive Couple; Brackeen)

The strongest objections, and our answers

1. Objection: the placement preferences classify children by ancestry and override individual best interests. A brief by the Goldwater Institute, Texas Public Policy Foundation and Cato Institute (counsel Shapiro, Burrus, Henneke, Sandefur) says the preferences make a "race- or national-origin-based distinction" and that "ICWA overrides the best interests test." Mitigation: we report this argument as the filers' own. The Court did not decide the equal protection question in Brackeen, because the challengers lacked standing, so the issue remains open. The statute itself says preferences yield for "good cause." (Amicus brief, No. 21-380; opened)
2. Objection: ICWA can leave children in limbo or move them late. Adoptive Couple is the example. Mitigation: the Court itself narrowed the law in that case, and the Brackeen families' adoptions were finalized after the tribes withdrew. We do not claim the law helps or harms children on average; we found no study of average outcomes in our fetches.
3. Objection: the 25 to 35 percent figure is old and may not describe today. Mitigation: agreed. We present it as the 1969 and 1974 survey result as quoted by the Court, not as a current rate. Current data on this page are different measures (share of foster care compared with population share).
4. Objection: overrepresentation today shows ICWA has not worked. Or the opposite: it shows ICWA is still needed. Mitigation: the ratio of 3.75 does not by itself say whether removal is justified in any case, and it is not a before and after comparison. Both uses go beyond the data. We show the figure and its source and no more.
5. Objection: tribal advocacy sources are one-sided. Mitigation: NICWA is used only for the numbers it cites to HHS and Census-based work, and the filers' brief is used only for the opposing argument. Primary sources (statute and opinions) carry the legal points.

Open questions: what would change our minds

• The 1969 and 1974 AAIA surveys, the Senate hearings and H.R. Rep. 95-1386 were not opened. What exactly the 25 to 35 percent figure measures is unverified.
• The enactment date and public law number were not opened. We use the year 1978 as the Court names the Act.
• The Brackeen dissents (Thomas, Alito) were not read; their reasoning is not summarized here.
• No study of outcomes for children placed under or outside ICWA preferences was opened. A well designed study of placement stability and well-being would change the weight of Section 3, objections 2 and 4.
• Whether the percentage of children removed has fallen since 1978, and by how much, was not found.
• The later circumstances of the child in Adoptive Couple were not in the text we read.
• Current (2023 to 2026) state practices, tribal court capacity and compliance data were not examined.

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.