Family preservation services are the bundle of short-term, intensive interventions a child welfare agency deploys when a family is on the edge of having a child removed but the agency judges that removal might be averted. The original model, Homebuilders, was developed in Washington State in the 1970s: a single caseworker with a tiny caseload, available around the clock, in the home for several hours at a time, for four to six weeks. The premise was that the moment of crisis is also the moment of openness, and that a saturating dose of practical help — groceries, a working refrigerator, a parenting strategy, a ride to a hearing — could reset the trajectory before the state stepped in with foster care.
Family preservation became federal policy in 1993 under Title IV-B Subpart 2 and was expanded by later legislation, most consequentially the Family First Prevention Services Act (FFPSA) of 2018, which for the first time allowed Title IV-E foster-care dollars to flow to prevention services delivered to children at imminent risk of removal and their parents. This was a structural shift. For decades, federal money paid for removing children but not for keeping them home; FFPSA tried to invert that incentive. Implementation has been uneven, slowed by state inertia, the requirement that services appear on an approved evidence-based clearinghouse, and the political reality that the public forgives a missed removal less easily than it forgives a generation of unnecessary ones.
The empirical record on family preservation is contested. Early enthusiasm produced studies showing that 70 to 90 percent of families served avoided removal for the child during and shortly after the service window. Skeptics, including some who did the original work, pointed out that many of those children were never going to be removed regardless — the agencies were selecting cases where the threat of removal was rhetorical, not imminent. More rigorous randomized trials in the 1990s produced muted effects: family preservation reduced placement slightly or not at all relative to standard services. The debate that followed split into two camps. One concluded the model was oversold and the field should move on. The other concluded the model worked when actually delivered to the right population — families on the genuine cusp — but had been diluted into a generic short-term service.
What both camps largely agreed on is that the alternative — foster care — is not benign. Removal traumatizes children, fractures sibling and extended-family bonds, costs more per case than almost any preservation alternative, and produces measurable downstream harm in school performance, mental health, and adult outcomes. A program that even modestly reduces removal therefore has to be evaluated against the harm it prevents, not the perfection it fails to achieve. This is the frame that has slowly returned family preservation to the center of policy discussion after a generation of dominance by the adoption-and-permanency paradigm of the ASFA era.
The collective question family preservation forces is whether the state can be trusted to help a family rather than only to police it. The caseworker who walks into a kitchen at 9 p.m. with a clipboard is also the person who can recommend the child's removal at 9 a.m. tomorrow. Many families experience preservation services as surveillance with extra steps. The services that do best are typically delivered by community-rooted providers contracted by the agency but trusted by the family, with the agency holding the legal authority in the background. The model is harder to administer, harder to evaluate, and harder to scale than a uniform clinical curriculum, but it is what the evidence keeps pointing toward.
Family preservation also collides with the political economy of child welfare. Foster care has a constituency: foster parents, group homes, residential treatment facilities, private agencies, and the staff who service them. Prevention has a more diffuse constituency: parents in crisis, kin caregivers, neighborhood-based nonprofits. When budgets tighten, prevention loses first. FFPSA tried to address this by attaching prevention to the most protected federal funding stream in child welfare. Whether the experiment holds depends on state implementation, federal review standards, and whether a future crisis case — a child killed in a family the agency was preserving — produces a reactive policy retreat. The history of child welfare is largely a history of such retreats.
A serious collective stance recognizes that the choice is not between perfect family preservation and perfect foster care. It is between two imperfect systems with different harm profiles. Family preservation accepts a small population of children whose situations are not solved by the service and who are harmed by remaining home longer. Foster care accepts a large population of children who would have been fine at home with material support and who are harmed by removal. The math of which system causes less aggregate harm has tilted, over four decades of research, toward prevention. The political math has not always followed.