Loading the library…
placement decisions
The operative claimWhether placing a maltreated child in foster care produces better or worse outcomes than leaving the child at home is genuinely unresolved: rigorous quasi-experiments using the same identification strategy reach opposite conclusions, effects apply only to "marginal" cases at the placement threshold, and the science supports neither a presumption for nor against removal.
Contested. For children at the margin of placement, rigorous natural experiments disagree on whether foster care helps or harms; the science supports neither a pro- nor anti-removal presumption.
Built on instrumental-variable natural experiments that exploit the quasi-random assignment of investigators. Doyle (2007, 2013) found marginal children fared BETTER at home (less delinquency, higher earnings); Gross & Baron (2020) found foster care IMPROVED safety and education; a meta-analysis found no clear difference between foster children and at-risk children who stayed home (Goemans et al., 2016). Findings are local to children at the margin of placement and do not speak to clear-cut severe-harm cases.
Doyle, J. J (2007). Child Protection and Child Outcomes: Measuring the Effects of Foster Care. American Economic Review.
Observational
Using effectively random assignment of investigators as an instrument, children on the margin of placement had BETTER long-term outcomes (less delinquency, fewer teen births, higher earnings) when they remained at home, especially older children.
Bearing on this claim: Doyle IV: marginal children fared BETTER at home (delinquency, teen birth, earnings).
doi.org/10.1257/aer.97.5.1583Doyle, J. J (2013). Causal effects of foster care: An instrumental-variables approach. Children and Youth Services Review.
Observational
For marginal cases, foster-care placement increased the likelihood of delinquency and emergency healthcare episodes; results apply to cases where investigators disagree about placement.
Bearing on this claim: Doyle IV replication: marginal placement increased delinquency and ED visits.
doi.org/10.1016/j.childyouth.2011.03.014Goemans, A., van Geel, M., & Vedder, P (2016). Developmental Outcomes of Foster Children. Child Maltreatment.
Meta-analysis
Foster children had lower functioning than the general population, but NO clear differences from at-risk children who remained with biological parents — both groups showed developmental problems.
Bearing on this claim: Meta-analysis: no clear difference between foster children and at-risk children who stayed home.
View sourceLawrence, C. R., Carlson, E. A., & Egeland, B (2006). The impact of foster care on development. Development and Psychopathology.
Longitudinal
Controlling for pre-placement adaptation and SES, foster children showed elevated behavior problems compared with maltreated children who remained home, suggesting placement effects beyond maltreatment history.
Bearing on this claim: Prospective study: elevated behavior problems beyond maltreatment/baseline — placement effects.
doi.org/10.1017/s0954579406060044Counter-authority (1). Studies an adversary may cite — surfaced deliberately so the grade is honest about its limits.
Gross, M., & Baron, E. J (2020). Temporary Stays and Persistent Gains: The Causal Effects of Foster Care. SSRN Working Paper (later AEJ: Applied Economics).
Observational
Using a comparable quasi-random investigator-assignment design in Michigan, foster care IMPROVED children safety and educational outcomes — the opposite directional finding to Doyle (2007), with gains emerging after reunification.
Bearing on this claim: Gross & Baron (same IV strategy, Michigan): foster care IMPROVED safety and education — opposite direction.
doi.org/10.2139/ssrn.35766404 supporting · 1 counter verified sources
Estimates are local average treatment effects for marginal cases and do not generalize to severe-harm cases; they depend on system quality, child age, and the counterfactual home environment; and unmeasured heterogeneity likely explains the cross-study divergence.
Last reviewed June 28, 2026