However, if that were the dominant mechanism, this survey data would show:
- Tribal respondents looking similar to one another, and
- ICWA and kinship homes looking safer than other homes
They do not. Outcomes split sharply inside the Native groups. ICWA-placed participants report more adult substance use than non-ICWA foster children after ACE and instability are accounted for, have worse institutional and net-positive scores, while at the same time claiming better language fluency.
Reports of childhood sexual abuse do not differ by ICWA status, but Adverse Childhood Event (ACE) scores do.
Who is in this comparison
After culling participants who did not finish the survey, 898 U.S. adults participated in a 2019 Family Dynamics survey. 390 participants identified as having Native American heritage. This subsample was then separated: (heritage composite = 1 and foster/adopt status 0, 1, or 2). Participants self-identified whether the ICWA was a factor in their case.
- Never placed in foster or adoptive care (status 0): 229
- ICWA-coded foster/adopt (status 1): 108
- Native, placed, not ICWA-coded (status 2): 53
The original survey collected 1,351 responses. Of those, 408 claimed at least some heritage and 366 said a great-grandparent was Native American; 98 identified full heritage Native American with no other heritage. 652 of the 1,351 said they do not report their full heritage on the US census, primarily out of privacy concerns or that they do not think it matters. The data set was culled to those who completed the survey, leaving 898 participants, 390 of whom identified Native American heritage. 295 said they are enrolled tribal members. 136 specified federal enrollment and 60 state enrollment. The remainder were not specific. Models on this page use the cleaned working file.
Safety-related items
Adult substance use runs opposite the ICWA talking point. Childhood sexual abuse (item A) does not differ significantly. The combined suicide indicator (Those who either thought about or attempted suicide) is lower in the ICWA-coded group, but there is no measurement for those who succeeded at suicide, making that measure inconclusive. Further research is needed.
| Item | Never placed | ICWA | Native non-ICWA foster |
|---|---|---|---|
| Adult substance use | 5.7% | 23.1% | 3.8% |
| Childhood sexual abuse (A) | 13.1% | 17.6% | 20.8% |
| Any suicide item | 29.7% | 11.1% | 17.0% |
Adjusted logits (ICWA dummy, non-ICWA-foster dummy, ACE, instability). Substance use, ICWA vs never placed: OR 6.71, p = .004. Any suicide item: OR 0.12, p = .004. Childhood sexual abuse χ² p = .29.
Two clocks for “culture”
Self-rated cultural connection and self-rated language fluency are highest in the ICWA-coded group. Spirituality and net-positive events and feelings composites are lowest.
| Never placed | ICWA | Native non-ICWA foster | |
|---|---|---|---|
| Cultural connection index (mean) | 3.98 | 7.20 | 5.38 |
| Spirituality 0–100 (mean) | 75.7 | 53.0 | 61.6 |
| Net-positive 0–100 (mean) | 56.2 | 32.1 | 38.0 |
| Institutional score 0–100 (higher = better here) | 85.9 | 39.6 | 56.6 |
| Stated Native language fluency | 27.2% | 73.1% | 49.1% |
| No Native language understanding at all | 44.5% | 1.9% | 22.6% |
Cultural connection and spirituality ANOVA p < .001. Language × group χ²(6) = 80.8, p < .001. Self-rated emotional health was higher in the ICWA group (3.33 vs 2.46). “Institutional” score is a composite of hospital visits and law enforcement experiences, either as victim, perpetrator, or both
Homes are mixed, not two colors
Many children subject to foster care are in and out of varied homes several times in their young lives. Among placed Native respondents, ICWA-coded cases had both higher same-heritage scores AND higher White-caregiver episode counts. A kinship/on-rez tribal-relative flag was about equally common (57% vs 53%) to both foster care groups and did not separate outcomes.
Placed Native n = 173 for flags; n = 161 for the White-caregiver regression. An off-res nontribal nonrelative flag was only 22 people — too small for a public causal claim.
What this is not
- Not a national CPS entry rate.
- Not a determination of suicidal likeliness. Lower suicide-item endorsement among survivors could either be compatible with fewer thoughts - or to successful attempts causing reporters to be missing from the sample. Suicide attempts did not differ by group.
- Not proof that ICWA caused substance use or that a White home caused harm. ACE is the repeating covariate.
2019 family survey, SPSS file ICWA2019_Logit_work.sav. Native filter. Page assembled September 25, 2026. If charts do not load, the tables carry the same numbers.