The Enforcement Failure

34 Lawsuits, 28 States, and the Child Welfare System Courts Still Cannot Fix

For half a century, federal courts have documented the same constitutional failures inside America’s child welfare system — and the machinery still keeps running

There is a quiet national emergency hiding in court dockets most Americans will never read. It does not announce itself with sirens, congressional hearings, or the kind of televised outrage that usually follows a government scandal. It lives in consent decrees, monitoring reports, contempt orders, appellate reversals, settlement agreements, and case captions that stretch across decades. It is the record of a child welfare system that federal courts have repeatedly found broken, unconstitutional, dangerous, and incapable of protecting the very children it claims to rescue. Yet after half a century of litigation, the system not only persists. In state after state, it has learned how to absorb judicial oversight without fundamentally changing.

As of January 2025, there were 34 active child welfare class action lawsuits across 28 states, an all-time high according to the research record compiled for this investigation. That number should detonate in the public mind. Thirty-four active lawsuits. Twenty-eight states. Average duration: 11 years. Average time from settlement to exit: more than 17 years. Eleven of fifteen current consent decrees have been active for a decade or more, while four have been in place for 25 to 30 years. These are not isolated reform cases. This is not a handful of troubled jurisdictions. This is a national enforcement failure wearing the costume of state-by-state litigation.

The lawsuits stretch across the country: Alabama, Arizona, California, Colorado, Connecticut, Florida, Georgia, Iowa, Kansas, Maine, Maryland, Massachusetts, Michigan, Minnesota, Missouri, New Hampshire, New York, North Carolina, Ohio, Oregon, Pennsylvania, Rhode Island, South Carolina, Tennessee, Texas, Washington, Wisconsin, and the District of Columbia. Five states face multiple simultaneous lawsuits. Only nine jurisdictions have successfully exited court oversight in the entire 50-year history of child welfare class action litigation. When a system requires federal court supervision in more than half the country, the question is no longer whether specific agencies are failing. The question is whether the national architecture itself is designed to produce failure.

This investigation is about that architecture. It is about why courts can find child welfare systems unconstitutional, order reforms, appoint monitors, impose benchmarks, demand reports, hold agencies in contempt, and still fail to force durable change. It is about why children can die or be abused inside systems already under court supervision. It is about how states weaponize delay, how legislatures control the money, how appellate courts narrow remedies, how private foster care providers profit from volume, and how federal incentives keep pushing the same machinery courts are supposedly trying to fix. The most disturbing conclusion is not that the courts have done nothing. It is that the courts have done a great deal, and the system has still survived.

Start with the scale. The most common court-ordered reforms are not exotic demands from radical plaintiffs. They are basic features any minimally functioning child welfare system should already provide: health and education services, limits on congregate care, adequate placement quality, foster home recruitment, manageable caseloads, placement stability, prevention of maltreatment in care, quality assurance, case planning, visitation, kinship care, permanency planning, and oversight of contract providers. These are the pillars of child safety. Yet the compliance record is chronically weak. Workforce and caseload reforms are rarely met. Placement quality remains chronically low. Maltreatment in care remains a critical failure. Visitation compliance is low. Foster home recruitment is failing. If courts have to order the basics, and states still cannot implement them after years or decades, then the word “reform” begins to look like camouflage.

The case that best exposes the enforcement collapse is M.D. v. Abbott, the Texas foster care catastrophe. Filed in 2011, tried in 2014, decided in 2015, and still unresolved in 2026, it is a case study in how a state can be found to violate children’s constitutional rights and then spend years fighting the remedy harder than it fights the conditions that caused the lawsuit. After a multi-week trial, Judge Janis Graham Jack found that Texas’s foster care system left children “more damaged” than when they entered. The system inflicted harm through excessive placement moves, inappropriate congregate settings, abuse and neglect inside state care, denial of basic medical and mental health services, sibling separation, and failure to achieve permanency. This was not a disagreement over best practices. It was a federal constitutional finding that the state system itself was dangerous.

Judge Jack’s words cut through the fog of bureaucratic language. “I cannot find DFPS to be credible at any level,” she said in 2015. That is a staggering statement from a federal judge overseeing a state agency responsible for vulnerable children. It means the court did not merely find policy defects. It found a credibility crisis at the institutional level. When the agency charged with protecting children cannot be trusted by the court, the entire premise of state custody collapses. Foster care is supposed to be the safer alternative. The court record showed that in Texas, for too many children, it became another site of harm.

The timeline reads like a slow-motion constitutional disaster. The class action was filed in March 2011. The class was certified in 2013. Trial took place in 2014. The liability finding came in 2015. Special masters were appointed in 2016. Interim remedial orders followed in 2017. A full remedial plan was adopted in 2018. The Fifth Circuit affirmed in part but reversed in part, calling the injunction overly broad. In 2019, the Fifth Circuit narrowed the remedial orders again. The district court continued trying to enforce core protections, including awake-night supervision and restrictions on placement changes. Texas was held in contempt in 2020. Additional contempt proceedings followed. In April 2024, Judge Jack found 38 discrete instances of non-compliance and imposed $100,000 per day in fines. Then, in May 2024, the Fifth Circuit stayed the penalties. In October 2024, the Fifth Circuit removed Judge Jack from the case and vacated the contempt order.

Read that again slowly. A state foster care system was found unconstitutional. Years of remedial orders followed. The state repeatedly failed to comply. The court imposed fines. The appellate court removed the judge. The children remained in the system.

The Fifth Circuit framed its intervention in judicial-temperament and remedial-power language, finding that Judge Jack had displayed a highly antagonistic demeanor toward the state and had become too personally involved in the proceedings. But the message sent to every state defendant was unmistakable. If a district judge pushes hard enough to enforce child welfare reforms, the judge can become the problem. The underlying constitutional violation can remain intact while the enforcement tool is dismantled. Texas did not have to prove that the system had become safe. It only had to convince the appellate court that the enforcement had gone too far.

That is the enforcement failure in miniature. Liability survives. Remedy dies. Children wait.

Texas is not alone. Illinois has been under the shadow of B.H. v. Johnson since 1988, making it the longest-running child welfare class action in American history. The case has been under a consent decree since 1991. In 37 years, Illinois DCFS has never achieved full compliance. Thirty-seven years is not a transition period. It is not implementation. It is an institutional life sentence for children passing through a system that courts already knew was broken before many of their parents were born.

The original allegations in B.H. were horrifying. The ACLU documented children in conditions that would likely trigger criminal neglect charges if imposed by a parent: an eight-year-old girl raped by two twelve-year-old boys in a DCFS shelter, children faking suicide attempts and slitting their wrists to get caseworker attention, caseworkers carrying such heavy loads they could not competently perform their duties, children placed in institutions that caused physical and emotional damage, and failure to provide services essential to reunifying families. This was the emergency that brought the court in. The consent decree was supposed to force reform.

By 1994, a monitor reported that DCFS had complied with less than half of 93 required reforms. The foster care population peaked at more than 50,000 children in 1995. In 1997, the consent decree was restated. In 2001, DCFS was declared in substantial compliance with the original decree for the first and only time. Then the case continued. New implementation plans followed. Educational and mental health failures persisted. The ACLU won a court order in 2009 blocking hundreds of millions in DCFS program cuts. In 2012, investigator caseloads were found substantially out of compliance. In 2014 and 2015, reports documented severe mental health shortages and the need for fundamental leadership and monitoring changes. In 2017, plaintiffs moved to enforce workload limits because caseloads far exceeded decree standards. In 2018, allegations at Chicago Lakeshore Hospital triggered emergency action and appointment of a special master. By 2020, DCFS still held custody of more than 21,000 youth while losing hundreds of residential beds and thousands of foster homes. By 2026, the case remained active.

The Illinois paradox is brutal. The system got smaller, but it did not get safe. The foster care population dropped dramatically from its peak, but the children who remained faced persistent structural deficiencies: overloaded workers, placement instability, institutional abuse, and denial of mental health services. That is not reform. That is triage. The agency learned to reduce volume while preserving the dysfunction embedded in the remaining system.

High caseloads for DCFS investigators have already been associated with child deaths, according to the research record. Proposed budget cuts that would have pushed caseloads to 50-to-1 were twice what the B.H. decree allowed. This exposes one of the deepest weaknesses in consent decree enforcement: the court can order caseload limits, but the legislature controls appropriations. If lawmakers refuse to fund the staff, beds, services, and placements necessary to meet the decree, the court can scold, monitor, and hold conferences, but it cannot easily run the state budget. The decree creates obligations. The political branches control the oxygen.

Then there is LaShawn v. Bowser in Washington, D.C., originally LaShawn v. Barry, often cited as a success story because D.C. eventually exited court oversight and was described as a national model. But the word “eventually” is doing a lot of work. The case lasted 32 years. A generation of children lived and died inside the litigation before the system earned that label.

The initial conditions in D.C. were almost beyond comprehension. More than 40 percent of social worker positions were vacant. The agency did not know how many active employees it had. Social workers carried caseloads of more than 100 children each. Reports of child abuse went uninvestigated. Children stayed in “90-day emergency placements” for years. Foster homes were overcrowded and unsafe. Reunification services and specialized placements were missing. The District could not even secure federal funding properly because it could not apply for it. In 1991, Judge Thomas Hogan found that D.C. violated constitutional and statutory standards, describing children relegated to entire childhoods spent in foster care drift.

When D.C. ignored the court order, Judge Hogan held the District in contempt and appointed a receiver in 1995, stripping the government of control over its own child welfare agency. That is the nuclear option in child welfare litigation, and it worked better than ordinary oversight. Receivership ended in 2001 after improvements, and the Child and Family Services Agency became independent. But backsliding occurred. In 2008, four deceased children were discovered in one home after inadequate investigation. In 2019 and 2020, at least 11 children died while in some form of foster care in D.C., despite three decades of court oversight. The foster care population eventually dropped from roughly 2,500 to 3,000 children in 1989 to 667 in 2020. But it took 32 years, receivership, and deaths under supervision to get there.

That is the central moral problem with calling LaShawn a success. A system can be declared a model after three decades, but the children who passed through the system during those decades do not get their childhoods back. Reform by consent decree is often measured in institutional milestones. Families experience it in years lost, placements endured, siblings separated, services denied, and funerals attended.

Other cases show the same pattern. Connecticut’s Juan F. v. Lamont lasted 33 years before oversight ended in 2022. Conditions did improve. The foster care population was cut roughly in half. Out-of-state placements dropped from more than 300 to five. Kinship care rose. Caseloads dropped from 56 to 17. Those are real achievements. But the settlement also required 160 pages of instructions, including specifications as granular as classroom wastebaskets. That detail is not merely funny bureaucracy. It is evidence that when a system cannot be trusted to do the obvious, litigation turns into micromanagement. The decree becomes a replacement nervous system for an agency that cannot regulate itself.

Georgia’s Kenny A. v. Deal has been active for more than two decades. A consent decree entered in 2005 required 31 outcome benchmarks with independent monitoring every six months. As of 2025, Georgia still had not fully complied. Abuse investigation timeliness remained below benchmark. Timely victim interviews remained below benchmark. CPS history review in investigations was only 78 percent. The state filed an exit plan in 2016. Nine years later, it had not exited. If a state can announce an exit strategy and still remain under supervision nearly a decade later, the exit plan functions less like a doorway and more like a press release.

Michigan’s Dwayne B. v. Snyder, filed in 2006, is another exhibit in the national indictment. The case challenged the state’s failure to provide safe placements, adequate healthcare, and permanency. A consent decree entered in 2008, but monitoring findings remained devastating. The record identifies 155 children abused or neglected by caregivers while in state custody, 1,270 children experiencing repeat abuse or neglect, fewer than 30 percent of relative homes receiving required visits before placement, only 1.9 percent of unlicensed homes having completed required waivers, 50 children experiencing prohibited successive shelter stays, psychotropic medication consent on file in fewer than 70 percent of cases, and adequate medication information in only 25.9 percent of files. Monitors also identified persistent and significant defects in data systems affecting more than 13,000 children.

The language from Children’s Rights and the monitors was blunt. After more than ten years, Michigan still failed its most vulnerable children. The lapses raised reasonable questions about agency leadership’s intentions. That second phrase matters. When monitors question intentions, they are not merely saying the state lacks capacity. They are saying the state may lack will.

New Jersey’s Charlie and Nadine H. v. Whitman lasted 24 years. The facts that launched it were grotesque. Three-year-old Nadine’s mother tried to drown her. The state then placed Nadine and her brother Charlie with their father’s estranged wife, who beat them with curtain rods, metal rulers, and a bucket, while giving her boyfriend, a convicted child sex offender, access to the children. The case eventually ended with a post-oversight accountability mechanism signed into law by Governor Phil Murphy. That legislation itself is a confession: without continued accountability after court supervision, the state could not be trusted to maintain reforms.

Mississippi’s Olivia Y. v. Barbour, filed in 2004, showed children warehoused in emergency shelters for staggering periods. Before the consent decree was even entered, 168 children were in emergency shelters with an average stay of 124 days. In Lamar County, three children averaged 497 days, more than sixteen months, in emergency shelters. In Harrison County, eleven children averaged 311 days. Forrest County caseworkers carried caseloads of 103 children per worker. Harrison County workers carried 86. Lamar County workers carried 48.5. There was a backlog of nearly 3,000 uninvestigated child abuse reports. Children’s Rights attorney Eric Thompson described abused children being parked and forgotten in emergency shelters for months at a time because DHS was so understaffed.

These cases are not aberrations. They are symptoms of a model that treats litigation as a substitute for governance. When the political branches will not build safe child welfare systems, plaintiffs sue. Courts order reform. States settle. Monitors measure. Legislatures underfund. Agencies delay. Appellate courts narrow. Children age out. Then everyone returns to court for another status conference.

Why do consent decrees fail so consistently? The answer is structural.

The first mechanism is the process trap. Consent decrees often devolve into measuring process compliance rather than child outcomes. Did the worker visit the home? Was the form completed? Was the conference held? Was the box checked? Connecticut’s 160-page decree is the perfect symbol. Michigan measures whether visits occurred, not necessarily whether children are safe. States become skilled at performing compliance without transforming the underlying system. They learn the paperwork language of reform. Children live with the reality behind the paperwork.

The second mechanism is the budget veto. Federal courts can order reforms, but state legislatures control appropriations. If the legislature refuses to fund enough caseworkers, placements, mental health services, foster homes, kinship supports, or prevention programs, the decree becomes a mandate without fuel. Illinois illustrates this perfectly. Courts could block devastating cuts, but they could not simply create a fully funded, functioning child welfare system. Michigan lost 500 residential beds and 2,300 foster homes over a few years. A court order cannot house a child if the placement array has collapsed. A consent decree cannot reduce caseloads if the state refuses to pay enough qualified workers to stay.

The third mechanism is the appeal escape. Texas perfected it. Appeal every broad order. Challenge every fine. Argue impossibility. Claim the judge is biased. Narrow the injunction. Stay the penalties. Remove the judge. The underlying liability finding can remain while the remedy is steadily dismantled. This is how a state can lose on the Constitution and still win the war of delay.

The fourth mechanism is personnel churn. Consent decrees outlive governors, agency directors, commissioners, monitors, lawyers, judges, and political coalitions. Every new administration inherits the decree as bureaucratic furniture. The original outrage fades. The case caption remains. The children in the plaintiff class change. Officials who did not negotiate the decree have little emotional or political ownership of it. The reform becomes someone else’s promise from another era.

The fifth mechanism is the moving-target defense. States argue they have transformed enough and should exit oversight. Plaintiffs argue the benchmarks still are not met or the system will backslide without supervision. Courts get trapped between competing definitions of success. Meanwhile, children remain in systems that everyone agrees are still fragile. The fight becomes less about child safety and more about whether the scoreboard is fair.

The sixth mechanism is the sovereignty shield. The Eleventh Amendment limits damages against states. Federal courts can issue injunctive relief, but they cannot easily make states pay for decades of harm. Contempt fines are among the few coercive tools available, and as Texas showed, appellate courts may vacate or stay them. The most effective enforcement tool in the history of these cases may be receivership, as in D.C., but courts are reluctant to take that step because it means effectively stripping elected officials of control over a core state function. So courts keep ordering. States keep managing. Children keep waiting.

But even this structural analysis is incomplete without the money. Consent decrees fail because they are trying to reform a system whose financial incentives often reward the very behaviors the decrees prohibit.

The federal spending ratio identified in the research record is devastating: nearly $12 in foster care and adoption funding for every $1 in prevention and family reunification. Title IV-E is an open-ended entitlement, reimbursing eligible foster care and adoption expenses at significant rates. Title IV-B, which supports family preservation, is capped and runs out when depleted. In plain English, the federal system writes a much bigger check after children enter the foster care and adoption pipeline than it does to prevent unnecessary entry in the first place.

The Adoption and Safe Families Act intensifies that pressure through adoption incentive payments. States can receive per-child bonuses for adoptions finalized above baseline targets, with payments that can reach thousands of dollars per child depending on category. The bonus is paid on finalization, creating pressure toward speed and volume. If an adoption later disrupts, the system’s logic can permit another adoption process and another incentive opportunity. Former Los Angeles County DCFS director David Sanders captured the problem plainly: when you combine an incentive to initially remove a child with an incentive to adopt that child out, “there is a problem.”

There is also the TANF diversion problem. States have diverted billions in Temporary Assistance for Needy Families funds toward child welfare systems, even though TANF was intended to support impoverished families. By 2019, states were diverting approximately $2 billion in TANF funds into child welfare. ProPublica estimated average diversion at 8 percent of TANF funding, with Arizona diverting nearly two-thirds of its TANF allocation to the Department of Child Safety. This is more than budget trivia. It is a grotesque feedback loop. Money meant to stabilize poor families is redirected into the system that investigates, supervises, separates, and manages poor families. Poverty assistance is weakened, poverty worsens, more families become vulnerable to allegations of neglect, more children enter the system, and the system receives more funding to process them.

Private foster care providers add another layer. Many are paid per child per day. That creates an obvious volume incentive. The Cato Institute analysis cited in the research record described each foster parent as potential revenue and documented how agencies may use bonuses to incentivize workers to bring in more parents. The Mentor Network, a major for-profit provider, reportedly achieved profit margins of 31 percent in Alabama and 44 percent in Ohio while receiving nearly all of its revenue from government sources. The record also cites a death rate among children in its care 42 percent higher than the national average. Former employees described the model with brutal clarity: success was defined by how many heads were in beds at midnight, and the bottom line was a dollar, not a child’s well-being.

That phrase, “heads in beds at midnight,” should be engraved on the wall of every legislative committee pretending not to understand the problem. If the financial model rewards occupancy, volume, retention, and speed, then a court order demanding quality, stability, family preservation, and careful individualized decision-making is pushing against the current. Every consent decree is trying to push water uphill while the funding stream flows downhill toward removal, placement, and adoption finalization.

The revolving door makes oversight even weaker. The research record identifies officials moving from government child welfare or youth services positions into private agencies or related organizations. Edward Murphy and Bruce Naradella moved from Massachusetts youth services leadership into the Mentor Network. David Sanders and Jackie Contreras moved from Los Angeles County DCFS leadership into Casey Family Services. Georgia officials Ron Scroggy and Virginia Pryor moved into private agency roles. This creates a regulatory conflict that does not need to be a smoking gun to matter. Officials who expect future employment in the provider world have reason not to burn bridges with providers they oversee. Aggressive regulation can become career-limiting. Contract renewal becomes routine. Oversight becomes performative.

New York’s 2018 provider renewals illustrate the problem. The city reportedly renewed all 22 foster care provider contracts even though 21 of 22 fell short of federal maltreatment guidelines, with three providers scoring 2, 19, and 10 out of 100. In any rational safety system, those numbers would trigger emergency review, contract suspension, or takeover. In child welfare contracting, renewal can proceed because the system needs beds, providers need revenue, and government needs somewhere to put children after removal. The market for placements becomes more powerful than the mandate to protect.

States have also developed a political playbook for surviving litigation. Phase one is to contest jurisdiction. File motions to dismiss. Challenge class certification. Appeal certification orders. In Texas, the Fifth Circuit reversed initial class certification in 2012, consuming years before the case could fully proceed. Phase two is settlement. The state agrees to an ambitious consent decree and receives positive press for committing to reform, while the actual obligations are delayed, qualified, monitored, or underfunded. Phase three is implementation by underfunding. Agencies cite resource constraints. Legislatures appropriate less than needed. Monitors report non-compliance. Courts hold conferences. The system keeps moving.

Phase four is changed circumstances. After a decade or so, the state argues that conditions have changed and the decree should be modified or terminated. If the court denies modification, the state appeals. If the court grants modification, the state complies with reduced obligations and later seeks more reductions. Phase five is enforcement warfare. When contempt arrives, the state attacks the contempt order, the fines, the judge, the scope of the injunction, the constitutional basis for penalties, and the feasibility of compliance. Texas executed that playbook successfully. Other states were watching.

The result is a system where the average time from settlement to exit exceeds 17 years, but even that understates the damage because states that never comply never exit. They remain under perpetual, unenforced court orders. Functionally, that can become the same as having no court order at all, except with better paperwork and more expensive lawyers.

The most damning indictment is not the duration of the cases. It is the deaths and abuse under oversight. Children die inside systems already being monitored by federal courts. At least 11 children died while in foster care in D.C. during 2019 and 2020, decades into LaShawn oversight. Illinois investigator caseloads associated with child deaths persisted under B.H. Michigan recorded 155 children abused or neglected by caregivers in state custody and 1,270 experiencing repeat abuse while under Dwayne B. Texas children were sexually assaulted in placements under the M.D. remedial regime. Mississippi children were warehoused in emergency shelters for months or more under monitoring.

There is no comprehensive national count of children who died in foster care while their states operated under federal consent decrees. That absence is itself a scandal. Courts order reform, but the system does not even maintain a clear national accountability ledger showing how many children died under court-supervised systems. If the public cannot count the dead, the system can keep calling each death an isolated tragedy rather than evidence of enforcement collapse.

This brings us to the ASFA paradox. The lawsuits challenge placement instability, abuse in care, failure to achieve permanency, denial of services, and warehousing children in institutions. But ASFA’s financial architecture incentivizes rapid termination timelines, adoption over reunification, volume over quality, and revenue flows after removal rather than before it. In that sense, many child welfare systems are not failing because they departed from the federal design. They are failing because they followed the federal design to its logical conclusion.

Texas’s foster care system was not found unconstitutional because it ignored permanency. It was found unconstitutional because the state built a placement system incapable of safely managing the children in its custody. Once children are removed, the system must place them somewhere. If the placement array is inadequate, children move repeatedly, enter congregate care, suffer abuse, lose siblings, miss services, and age out damaged. The front-end removal incentive feeds the back-end placement crisis. Courts then order the state to fix the back end while the federal financial current continues feeding the front end.

Legal scholar DeLeith Gossett’s critique is central here: ASFA’s financial incentives can disrupt families permanently through speedy termination of parental rights without a corresponding move from foster care to adoptive homes. That is the legal orphan problem in policy form. The child’s rights are invoked to terminate parents, but no equally enforceable right ensures the child receives the promised permanent family. The state destroys one legal bond in the name of creating another, then sometimes fails to create the second bond at all.

Illinois offers a partial counter-model. In 1997, the state began paying private agencies for permanence, including both reunification and adoption, while penalizing foster care retention. The foster care population fell dramatically from more than 50,000 children to roughly 14,000 by 2017. That matters because it suggests incentives can be changed and outcomes can shift. But even Illinois did not fully solve the problem. B.H. remained active, and DCFS still struggled with compliance. Financial realignment may be necessary, but it is not sufficient. A system can reduce caseload volume and still fail the children left inside it.

The enforcement record therefore does more than show court failure. It provides evidence for a broader constitutional and legislative indictment. Fifty years of class action litigation have shown that case-by-case enforcement cannot remedy structural defects built into the national child welfare financing model. When 34 lawsuits in 28 states produce the same rhythm of settlement, monitoring, non-compliance, backsliding, contempt, appeal, and exhaustion, the problem is not operational. It is structural.

Under Monell, a municipality can be liable when an official policy or custom causes constitutional injury. The enforcement record shows widespread constitutional injury across jurisdictions. Consent decrees document persistent non-compliance. Federal incentives document the policy architecture. Together, they form a pattern-or-practice record that should force the question: why are courts trying to repair the consequences of a system whose incentives continue to generate the damage?

This is where the existence of consent decrees becomes paradoxical. States can point to ongoing court supervision as proof that problems are being addressed. But the evidence shows that consent decrees often do not fix the system. They become evidence of the system’s inability to comply with basic constitutional and safety standards without permanent judicial scaffolding. If a state needs decades of court orders to meet minimum obligations, the decree is not proof of reform. It is proof that ordinary governance has failed.

The five-prong enforcement-failure argument is now hard to avoid. Duration: reform takes too long, with average settlement-to-exit timelines stretching beyond 17 years. Universality: 34 lawsuits across 28 states prove the problem is not aberrational. Persistence: cases like B.H. and LaShawn lasted 37 and 32 years, respectively. Escalation: Texas required repeated contempt findings because compliance was not voluntary. Reversal: appellate courts can dismantle even aggressive enforcement, as the Fifth Circuit did in M.D. v. Abbott.

The conclusion is stark. The enforcement failure is not a bug. It is a feature. The system is built to resist reform because the incentives reward the same behaviors courts try to restrain. States can delay because children age out faster than litigation ends. Legislatures can underfund because courts cannot easily appropriate. Private providers can profit because the system needs beds. Appellate courts can protect sovereignty because broad structural injunctions trigger institutional discomfort. And the public rarely sees the full record because child welfare cases are fragmented across jurisdictions, sealed by confidentiality, and buried in legal procedure.

This is why reform language is no longer enough. The case for repeal or radical restructuring does not rest on abstract ideology. It rests on half a century of enforcement history. If court orders across 28 states cannot force durable compliance, then the problem is not simply that the agencies need more monitoring. If contempt orders cannot produce compliance, then the problem is not simply that judges need more patience. If children continue to die under oversight, then the problem is not simply that the metrics need refinement. If federal funding continues to reward removal, placement, and adoption-side outcomes more powerfully than prevention and reunification, then the problem is the architecture.

A serious response would begin by reversing the financial current. Prevention and family preservation must be funded at least as strongly as foster care and adoption. Reunification success must be rewarded. Kinship stability must be rewarded. Placement quality, not placement volume, must be rewarded. Providers should not profit from keeping beds full while children drift. States should not receive stronger federal support after a family breaks than before the break could have been prevented. TANF funds intended for poor families should not be redirected into the apparatus that polices those same families for poverty-related neglect.

A serious response would also require real enforcement teeth. Federal funding should be conditioned on transparent public reporting of child deaths, abuse in care, placement instability, caseloads, service delays, psychotropic medication oversight, legal orphan rates, reunification outcomes, and racial disparities. Consent decrees should include automatic financial consequences tied to specific violations, not endless negotiation. Independent monitors should report in plain public language, not only technical filings. Legislatures that underfund decrees should be forced to own that choice publicly. Private providers with safety failures should lose contracts, not receive renewals because the state has nowhere else to place children.

A serious response would confront the revolving door. Officials overseeing private providers should face cooling-off periods before taking jobs in the industry they regulated. Contract renewal should be tied to safety outcomes, not placement capacity alone. Provider performance scores should be public, comparable, and enforceable. If a provider fails maltreatment standards, the default should be intervention, not renewal.

A serious response would also stop using child welfare secrecy as a shield against public accountability. Protecting children’s identities is necessary. Protecting agency failure is not. The public can be told how many children died, how many were abused in care, how many experienced repeated placements, how many were warehoused in shelters, how many were separated from siblings, and how many became legal orphans without exposing private child details. Confidentiality should protect children from exploitation, not institutions from scrutiny.

The defenders of the current system will say these cases prove litigation works because some jurisdictions eventually improved. That argument is not entirely false, but it is morally incomplete. Connecticut improved after 33 years. D.C. improved after 32 years. New Jersey exited after 24 years. Tennessee exited after more than 15 years. Those exits matter. But children do not experience reform over institutional time horizons. They experience childhood once. A three-year-old entering foster care at the beginning of a 17-year consent decree is an adult before the average system exits. A newborn entering care in Illinois after B.H. was filed could have grown up, had children, and watched those children enter the same court-supervised system.

That is the measurement that matters. Not whether a decree eventually closes. Not whether an agency achieves substantial compliance after decades. Not whether monitors praise incremental progress. The measure is whether the child in state custody today is safe, served, placed stably, connected to family, protected from abuse, and given a real path to permanence without becoming collateral damage in a system designed around revenue, risk avoidance, and delay.

The record says the answer too often remains no.

The courts know the system is broken. Plaintiffs know. Monitors know. Agency leaders know. Legislatures know. Providers know. The children know first. The scandal is that knowledge has not produced transformation. It has produced procedure. It has produced litigation careers, monitoring contracts, compliance reports, appeals, revised implementation plans, and press releases announcing commitment to reform. But the machine keeps moving.

That machine survives because it has learned to metabolize accountability. A lawsuit becomes a decree. A decree becomes a plan. A plan becomes benchmarks. Benchmarks become reports. Reports become disputes. Disputes become appeals. Appeals become narrowed remedies. Narrowed remedies become partial compliance. Partial compliance becomes an exit motion. Then the next lawsuit begins somewhere else. The system does not defeat accountability by denying every problem. It defeats accountability by turning every problem into a process.

The enforcement failure is therefore not merely a legal story. It is a story about power. State power to remove children. Federal power to fund incentives. Legislative power to starve reforms. Provider power to profit from placements. Appellate power to narrow remedies. Bureaucratic power to delay. And family powerlessness when all of those forces converge.

If America is serious about protecting children, it must stop pretending that court-supervised dysfunction is reform. It must stop measuring success by whether a state can eventually exit a consent decree after two or three decades. It must stop rewarding the destruction of families more predictably than the preservation of families. It must stop allowing private providers to treat children as revenue units. It must stop hiding deaths and abuse inside fragmented state records. It must stop asking federal judges to fix by decree what Congress and state legislatures keep breaking by design.

The most honest conclusion from this investigation is also the most uncomfortable: fifty years of enforcement litigation have proven that the current system cannot be repaired by consent decrees alone. The courts can identify the harm. They can order remedies. They can appoint monitors. They can issue contempt findings. But they cannot, by themselves, reverse a financial and political architecture that rewards the very conditions they are trying to eliminate.

That architecture has to be confronted directly.

Until it is, the country will continue living with the same grotesque cycle. A child welfare system fails. Children are harmed. Plaintiffs sue. Courts find violations. States promise reform. Legislatures underfund it. Agencies delay. Monitors document failure. Appellate courts narrow enforcement. Children age out, die, disappear, or survive damaged. Then a new report announces that accountability is underway.

It is not underway. It is trapped.

The enforcement record is no longer a warning sign. It is the evidence. Thirty-four lawsuits. Twenty-eight states. Decades of non-compliance. Children harmed under court supervision. A financial structure that rewards the pipeline courts are trying to restrain. This is not a system waiting for reform. It is a system that has survived reform.

And that is why the next question cannot be how many more consent decrees America needs. The next question is how many more decades of failure it will tolerate before admitting that the child welfare machine is not broken because courts have ignored it. It is broken because courts have seen it clearly, ordered it to change, and still could not make the machine obey.

---

If you enjoyed this work and feel encouraged by independent investigations like this, please consider becoming a paid subscriber to The Constitutional Republic on Substack.

Your support helps us continue:

• investigative reporting

• public records research

• educational transparency projects

• community outreach

• and future investigations like this series

Most importantly, please share this article.

Independent journalism survives when ordinary people help spread information powerful institutions would prefer remain ignored, misunderstood, or buried beneath complexity.

Because silence protects systems.

Informed communities protect children.

ALL paid subscriptions are tax deductible Project Milk Carton | 501(c)(3) | EIN: 33-1323547

---

SOURCES & CITATIONS

PRIMARY COURT RECORDS

• M.D. ex rel. Stukenberg v. Abbott, 2:11-cv-00084 (S.D. Tex.) — Civil Rights Litigation Clearinghouse, Case 12435

• B.H. v. Johnson (Sheldon), 1:88-cv-05599 (N.D. Ill.) — Civil Rights Litigation Clearinghouse, Case 11161

• LaShawn A. v. Bowser (orig. v. Barry), (D.D.C. 1989) — ACLU-DC case file

• Kenny A. v. Deal (orig. v. Perdue), (N.D. Ga. 2002) — Children's Rights case file

• Dwayne B. v. Snyder, 2:06-cv-13548 (E.D. Mich.) — Civil Rights Litigation Clearinghouse, Case 12437

• Juan F. v. Lamont (orig. v. Weicker), (D. Conn. 1989) — CT Mirror reporting

• Charlie & Nadine H. v. Whitman, (D.N.J. 1999) — NJ.gov press releases

• Olivia Y. v. Barbour, 3:04-cv-00251 (S.D. Miss.) — Children's Rights case file

POLICY & SCHOLARSHIP

• Bipartisan Policy Center, 'Accountability in the Courtroom: Review of Child Welfare Litigation and Required Reforms' (Jan. 2025)

• Casey Family Programs, 'Summary of child welfare class-action litigation' (2022)

• Elizondo, 'Constitutional Catch-22: The Unvindicated Rights of Foster Children,' Stanford Law (2025)

• Pesavento, 'How Misaligned Incentives Hinder Foster Care Adoption,' Cato Journal (Winter 2021)

• NCCPR Issue Paper 12: 'Financial Incentives: You Get What You Pay For'

• NCCPR, 'The Children Wronged by Children's Rights' (2024-2025)

• Riley, 'Don't Make a Federal Case of It,' City Journal / American Enterprise Institute

Center for the Study of Social Policy, 'Child Welfare Reform Through Class Action Litigation'

• Gossett, DeLeith (Texas Tech University School of Law), ASFA financial incentive analysis

REPORTING & NEWS

• Texas Tribune: M.D. v. Abbott contempt coverage (2020-2025)

• Texas Monthly: 'A Judge Tried to Reform Texas Foster Care. The Fifth Circuit Removed Her.' (2024)

• ABA Journal: 'Judge with highly antagonistic demeanor must be removed' (Oct. 2024)

• Bloomberg Law: 'Fifth Circuit Removes Texas Judge Critical of Foster Care System' (2024)

• The Imprint: 'Washington, D.C., Exits Long Foster Care Court Case' (2021)

• Capitol News Illinois: 'Lawyers to seek enforcement of decades-old DCFS consent decree' (2021)

• ACLU-DC: 'After 32 Years, ACLU-DC Settles LaShawn v. Bowser' (2021)

• CT Mirror: 'Court ends federal oversight of child protection in CT' (2022)

• SCOTUSblog: M.D. v. Abbott cert petition tracking

• The Imprint: 'TX Foster Care System Lawsuit Appealed to Supreme Court' (May 2025)

• Children's Rights Inc.: Active case docket and monitoring reports

• ProPublica: TANF diversion analysis

SUPREME COURT & APPELLATE

• M.D. ex rel. Stukenberg v. Abbott, No. 24-1168 — Supreme Court docket

• Fifth Circuit opinion, No. 24-40248 (Oct. 11, 2024) — Judge removal & contempt vacatur

• Fifth Circuit en banc denial (Feb. 11, 2025) — 9-5 vote

• Supreme Court cert denied (Oct. 6, 2025)