The American foster care system holds, on any given day, roughly four hundred thousand children. Over the course of a year, roughly six hundred thousand children pass through it. About one in three Black children, one in seven Latino children, and one in ten white children will experience a child protective services investigation by age eighteen. About six percent of all American children will be placed in foster care at some point. The Black-white disparity in placement is roughly two-to-one nationally and substantially higher in some jurisdictions. The system that produces these numbers calls itself child welfare. Dorothy Roberts has spent two decades arguing, with documentary support that has only grown stronger, that what the system actually performs is the regulation of poor and Black families through the threat and reality of child removal.

The data describe the system's operational shape. The largest single category of reasons for removal is "neglect," a term that is functionally synonymous with poverty in a substantial fraction of cases. Inadequate housing, inadequate food, inadequate childcare, parental substance use, parental mental illness, exposure to domestic violence — these conditions overlap with poverty so completely that the line between neglect and economic distress is often invisible to the families on the receiving end of the determination. Physical and sexual abuse, the categories that most people imagine when they think about foster care, account for a minority of placements. The system is built around the prevention of abuse and is staffed and oriented around the supervision of poverty.

The placement experience itself causes harm. Children in foster care experience, on average, multiple placement changes; the median is between two and three placements per spell, with a substantial minority experiencing five or more. Each placement change is, developmentally, a loss — a disruption of attachment, schooling, neighborhood, friendships, and routine. Sibling separation is common: roughly seventy percent of children in foster care with siblings are separated from at least one sibling. Educational outcomes for foster youth are catastrophic by any measure: graduation rates roughly half those of peers, post-secondary completion rates in the single digits, employment rates and earnings substantially depressed. The Casey Family Programs Northwest Foster Care Alumni Study found post-traumatic stress symptoms in nearly twenty-five percent of foster care alumni, a rate higher than for combat veterans.

Aging out of foster care produces a distinct cohort of young adults whose life outcomes the system has been documenting for thirty years. Within four years of leaving care, roughly twenty-five percent of former foster youth have experienced homelessness. Roughly twenty percent have been incarcerated. Roughly fifty percent are unemployed. The Midwest Evaluation of the Adult Functioning of Former Foster Youth, led by Mark Courtney over multiple waves, has produced the most detailed longitudinal data set on this transition. The aggregate picture is consistent across studies: the system that the state operated in lieu of the child's family produces, on average, worse adult outcomes than the system that operated in the families from which the children were removed.

The policy architecture that produced this system was built in phases. The 1974 Child Abuse Prevention and Treatment Act created the federal infrastructure for mandatory reporting and state child protection systems. The 1980 Adoption Assistance and Child Welfare Act required reasonable efforts to prevent removal and to reunify families, but funded foster care under Title IV-E in ways that incentivized placement over prevention. The 1997 Adoption and Safe Families Act dramatically shifted the system toward expedited termination of parental rights and adoption, imposing a fifteen-of-twenty-two-months timeline for filing termination petitions. The 2008 Fostering Connections to Success Act extended supports for older youth and incentivized kinship placement. The 2018 Family First Prevention Services Act, for the first time, allowed Title IV-E funds to be used for prevention services rather than only for out-of-home care. The arc, taken as a whole, has been a slow shift away from the heavy reliance on placement and toward prevention, but the shift has been slower than the rhetoric.

The racial demographics of the system are not incidental. Dorothy Roberts's Shattered Bonds (2002) and Torn Apart (2022) document, with quantitative and ethnographic detail, how the child welfare system functions as a parallel system to mass incarceration in Black communities. The mandatory reporting infrastructure routes Black mothers' encounters with hospitals, schools, and public benefits offices into CPS investigations at rates far exceeding their share of underlying need. The substantiation rates, the removal rates, the reunification rates, and the termination rates are all worse for Black families than for white families with similar case characteristics. Native American children have been the subject of a parallel and longer history; the Indian Child Welfare Act of 1978 was passed in response to a mid-twentieth-century practice that placed roughly a quarter to a third of all Native American children outside their families, frequently in white homes far from their tribes. ICWA's preferences for tribal and Native placement have been under constant litigation since enactment, including the 2023 Haaland v. Brackeen decision that upheld the statute.

The alternatives have been mapped in detail and implemented partially. Vivek Sankaran's work on parental representation, Martin Guggenheim's challenge to the system's foundational logic, Khiara Bridges's analysis of the surveillance of poor mothers, and Christopher Wildeman's quantitative work on the population-level effects of foster care placement converge on a common set of recommendations: invest in concrete material support to families before removal, ensure high-quality legal representation for parents at every stage, narrow the legal definition of neglect to exclude poverty-driven conditions, expand kinship care with full financial support, reduce reliance on congregate care, and aggressively rebuild family-based and community-based services that prevent the conditions producing CPS contact. None of these are speculative. All of them have been demonstrated, in pilot and at scale, to produce better outcomes than the system as currently configured. The question parenthood at the collective scale poses to the country is whether the family-policing system the country built will be replaced by the family-supporting system the research has described, or whether the existing system will continue to operate because too many institutions depend on its operation to imagine its disappearance.