Foster care is a system designed to do an impossible thing: separate children from parents in order to protect them, and then either return them, replace them, or hold them in suspended kinship until they age out. Every part of that sentence is contested. Who counts as protected, who counts as a parent, what return means, what replacement costs, and what happens to a young person who reaches eighteen having been raised by the state. The system makes families and breaks them, often in the same case file, and the ratio of making to breaking tracks closely with race, class, disability, and zip code.
The numbers tell one story. On any given day in the United States, roughly four hundred thousand children are in foster care. Black children are removed at rates far exceeding their share of the population. Native children, despite the Indian Child Welfare Act, are still disproportionately separated from their nations. Children with disabilities, children of parents with disabilities, and children whose parents are poor are overrepresented at every stage from hotline call to termination of parental rights. Dorothy Roberts argues that what looks like child protection is, viewed structurally, a system of family policing aimed at the same populations targeted by the criminal legal apparatus. The hotline, the caseworker, the family court, and the prison form a continuous surveillance architecture pointed at poor parents, especially Black mothers.
The counter-story matters too. Children are sometimes in genuine danger from the adults closest to them. Sexual abuse, severe physical violence, and untreated parental psychosis can be lethal. A child welfare system that refused to intervene anywhere would also break families, by leaving children to die or to grow up inside ongoing harm. The honest question is not whether intervention is ever justified but whether the current intervention apparatus is calibrated to the actual distribution of risk, and whether its tools, removal, foster placement, termination, adoption, are the right tools for the harms it claims to address.
Most evidence says the calibration is wrong. Vivek Sankaran and his colleagues document how removal itself is a trauma, not a neutral pause. Children placed in foster care show worse outcomes than similarly situated children left in struggling homes with services attached. Sibling separation is routine. Placement instability, three, five, ten homes in a few years, predicts later mental illness, incarceration, and homelessness more powerfully than the original maltreatment did. Aging out at eighteen with no kin produces a cohort whose life trajectories shame the country that produced them.
Foster parents themselves occupy an ambiguous position. Some are kin, grandparents and aunts absorbing children inside the family network, often without the financial support given to strangers. Some are strangers who become, over months and years, the only stable adults a child has known. Some are predators who use the system as a hunting ground. Most are ordinary people trying to do something difficult inside a structure that gives them little information, less authority, and almost no continuity. They are asked to love a child who may leave, to protect a child whose history they were not told, and to coparent with a state that treats them as interchangeable.
The making and breaking are not opposites. The same removal that breaks a child from a mother makes a relationship with a foster parent. The same reunification that restores a child to a parent breaks a bond with the foster family. The same adoption that secures permanence terminates a legal kinship that the child may grieve for decades. Law One, Unity, applies here as a discipline of attention: every family in this system is connected to every other family in it, and to the policies, racial histories, and economic conditions that produced the case. There is no clean intervention. There are only better and worse ways to hold the unavoidable damage.
A serious foster care politics would start by shrinking the funnel. Most cases that reach the system are neglect cases, and most neglect is poverty wearing a different name. Direct cash, housing, childcare, and addiction treatment prevent more removals than any hotline reform. When removal is necessary, kin should be the default, sibling groups should stay together, and the child's existing relationships, school, church, neighborhood, should be preserved wherever possible. Permanency should be measured in relationships, not in case closures. And the young people who do age out should leave with what any other eighteen-year-old gets from a family: a place to come back to, money when the car breaks, someone who answers the phone.