GHOST VISIT INVESTIGATION — ARTICLE 3 THE COMPLIANCE CRISIS

States Failing Children — The State-by-State Evidence

Jul 01, 2026

In Article 2 of this series, we followed the money. We showed how more than $11 billion in federal child welfare spending flows overwhelmingly toward removing children from their homes and maintaining them in out-of-home placements, while allocating less than six dollars per month per child to verify that anyone is actually checking on them. We documented the perverse financial incentives created by the open-ended Title IV-E entitlement, the FMAP matching formula that rewards placement volume, the AFDC lookback that excludes nearly half of foster children from federal oversight, the tiny prevention budget that remains an afterthought, and the quiet removal of the only federal financial penalty for visit non-compliance through P.L. 118-258 effective October 2025. We showed how this architecture makes “ghost visits” — the phenomenon of children in state custody going months or years without a meaningful safety check — not just possible but the predictable, rational outcome of the incentives the federal government has created.

Now, in Article 3, we move from the financial architecture to the human and institutional reality it produces on the ground. This is the compliance crisis — documented state by state, audit by audit, Office of the Child Advocate report by Office of the Child Advocate report, and child by child. The evidence is not ambiguous. It is not a matter of interpretation, insufficient data, or isolated bad actors. It is a 25-year unbroken record of systemic, catastrophic failure by the very institutions entrusted with protecting America’s most vulnerable children — a failure so complete that no state in the history of the federal Child and Family Services Review process has ever achieved substantial conformity with all seven federally defined child welfare outcomes.

Every child in foster care is legally entitled to a monthly face-to-face visit from their assigned caseworker. This is not a bureaucratic suggestion or a best-practice recommendation. It is federal law under 42 U.S.C. § 622(b)(17), embedded in every state’s Title IV-E state plan. It is the minimum mechanism by which the state fulfills its legal responsibility — a responsibility it assumed when it took legal custody — to know whether a child it has removed from their home is safe, alive, and not being re-victimized in the very placement the state chose and is paying for with federal and state dollars.

Yet across the country, in state after state, that basic obligation is being ignored at a scale that should be considered a national emergency. The data does not show marginal shortfalls or temporary disruptions. It shows a nationwide child welfare system that has never — in a quarter century of federal oversight — demonstrated the capacity or the will to meet its most fundamental safety obligation to the children in its care.

The National Scale of the Failure: 328,947 Children, One Broken Promise

Before examining any individual state, it is essential to understand the sheer scale of the system that is failing its most basic duty.

As of September 30, 2024, the Adoption and Foster Care Analysis and Reporting System (AFCARS) reported 328,947 children in foster care across the United States. This marks the sixth consecutive year of decline from a peak of 407,332 in 2020 — a trajectory driven by demographic shifts and the Family First Prevention Services Act, not by any demonstrated improvement in system performance or child safety.

During FY2024 alone, 170,943 children entered foster care and 176,730 exited. Behind those net numbers is constant churn: hundreds of thousands of children cycling through a system that cannot manage to conduct a single meaningful safety check on them once per month.

Each of these 328,947 children is legally entitled to a monthly face-to-face visit from their assigned caseworker. At the federal 95% compliance floor — already a generous margin of error that accepts 5% of children going unseen every month — approximately 16,447 children per month would still miss their required visit. That is the acceptable failure rate built into federal policy.

But actual compliance rates in the states where we have real data suggest the true picture is far worse.

The Math of Failure — What These Numbers Actually Mean for Children

At Connecticut’s documented 57.4% compliance rate — the worst publicly available figure from a rigorous Office of the Child Advocate investigation — applied to the national population of 328,947 children in foster care as of the end of FY2024, the scale of the unseen children becomes staggering:

140,111 children per month receive no safety check whatsoever. That is more children than the entire foster care population of most individual states.

1,681,332 missed visits per year — over one and a half million opportunities for a caseworker to detect danger, assess well-being, or simply confirm that a child is still alive and in the placement the state assigned.

4,606 missed visits per day — roughly 192 children per hour whose legal entitlement to a monthly face-to-face visit is being ignored.

At even a more conservative (and unsupported) assumption of 80% national average compliance, we would still be looking at 65,789 children per month without a safety visit, 789,470 missed visits per year, and 2,163 children per day going unmonitored.

These are not abstract numbers on a spreadsheet. They represent real children in real placements — foster family homes, therapeutic foster homes, group homes, residential treatment centers, and kinship placements — whose safety the state has assumed legal responsibility for when it removed them from their birth families, and then failed to verify month after month. A seven-year-old in a foster home in rural Alabama whose caseworker position has been vacant for four months. A teenager in a group home in New York whose caseworker is carrying 35 cases and cannot possibly visit them all. A child in Connecticut who told a caseworker they felt unsafe and then died by suicide within an hour of that visit — because the visit itself failed to trigger protective action.

The federal government knows this scale of failure. States know it. The data has been accumulating for decades through CFSRs, OIG audits, state child advocate reports, and investigative journalism. And yet the system continues to operate, to receive billions in federal matching funds under Title IV-E, to place more children into the very placements it cannot monitor, and to report aggregate compliance numbers that create the illusion of a system that is mostly working while individual children disappear from oversight entirely.

Connecticut: The Worst Documented Failure — And a Child Who Died Because the Visit Failed

The most alarming recent compliance data comes from Connecticut, where the Office of the Child Advocate (OCA) has documented what it characterizes as “increasingly alarming” deficits in the Department of Children and Families (DCF).

For the period November 2024 to October 2025, only 57.4% of children in foster homes had two documented face-to-face visits per month. That is a 37.6 percentage point gap from the 95% federal standard. More than four out of every ten required visits simply did not happen.

Even more disturbing: for in-home cases involving children the state had already determined were abused or neglected but who remained in the home, only 55.8% of cases showed all children had two documented face-to-face visits per month between May and October 2025. These are children the state has confirmed are being abused — and the agency cannot manage to check on them.

Scores for contact with children, contact with adult participants, and risk and safety assessment have remained below 60% since May 2023 — more than two full years of sustained, systemic failure across every major performance metric. Supervisory support ratings dropped from 90% in early 2022 to just 55% in the first half of 2025. Institutional capacity is not improving. It is collapsing.

The Child Suicide That Exposed the System

The OCA report was issued following a child suicide that occurred within one hour of a DCF caseworker visit.

The child had explicitly stated they felt unsafe and had requested foster care placement. DCF decided to leave the child with the parent. The child died by apparent suicide within 60 minutes of the caseworker leaving.

The OCA’s finding was devastating in its clarity: The caseworker visit failed in its most fundamental purpose — detecting danger and acting on it. The child communicated the danger explicitly, and the system still failed.

This single case illustrates the chasm between compliance metrics and actual child protection. Even when a visit occurs, if the caseworker lacks the training, authority, supervisory support, or institutional backup to act on what they observe, the visit becomes a hollow formality — a checked box on a form attached to a dead child’s case file.

Connecticut’s CFSR Round 4 review, based on data from July 2024 to March 2025, confirmed the systemic nature of the failure in the starkest possible terms: 0 of 7 outcomes in substantial conformity — total failure across safety, permanency, and well-being. Only 2 of 7 systemic factors reached substantial conformity. Key deficiencies identified by the federal review included inconsistent safety practices, lack of adequate mental health services for children and families, and delays in achieving permanency and reunification. Some strengths were noted — kinship care placements, sibling placements kept together, and behavioral health assessments — but none of these were sufficient to achieve conformity in any outcome area.

The wealth paradox in Connecticut makes the failure even more indefensible. Connecticut is one of the wealthiest states in the nation, with a median household income consistently among the top five in the country. It receives the minimum 50% federal match for foster care maintenance under Title IV-E — meaning the state itself contributes more of its own dollars per child than states with higher FMAP rates. This is not a jurisdiction struggling with poverty, inadequate tax base, or rural service delivery challenges that might explain capacity gaps. Connecticut’s failure is a pure failure of institutional will, organizational capacity, leadership priorities, and accountability mechanisms. When one of the richest states in America cannot manage to visit more than 57% of its foster children monthly — and cannot even achieve that dismal rate for children it has already confirmed are being abused in their homes — the problem is not money. The problem is that safety verification has never been treated as a non-negotiable core function.

The Illusion of Near-Compliance: Delaware and the Danger of Aggregate Metrics

Not every state fails as dramatically as Connecticut on paper. Some appear close to the 95% standard — until you look closer.

In FFY2021, Delaware scored 94.38% on the monthly caseworker visit report — falling below the federal standard by just 0.62 percentage points. The state reported 4,583 visits made out of 4,856 required. That gap of 273 missed visits represents 273 children who went an entire month without a required safety check.

Delaware’s in-home visit rate was 89.53% — above the 50% federal minimum but still meaning that more than 10% of visits happened somewhere other than where the child actually lives. A caseworker who meets a child at a DCFS office, school, or neutral location cannot assess the safety of the home environment — the very thing the in-home requirement was designed to ensure.

Delaware illustrates a critical analytical point that applies nationwide: the aggregate compliance metric masks individual-level failures.

A state reporting 94.38% compliance could be:

Visiting 94.38% of children every month and missing 5.62% consistently — meaning some children never receive visits.

Visiting all children most months but missing visits sporadically — creating unpredictable gaps in oversight that coincide with periods of risk.

Conducting nominally compliant visits that lack the quality requirements specified in ACF PI-07-08 — brief encounters that check the box without actually assessing safety.

The federal reporting system cannot distinguish between these scenarios. A state can report near-95% compliance while completely abandoning oversight of a specific subset of children — and the data would not reveal it.

This is why self-reported aggregate numbers are so dangerous. They create the illusion of a system that is mostly working, when the reality for thousands of individual children is that no one is checking on them at all.

Virginia, Kentucky, Maine: Failure Stacked on Failure

Virginia’s 2023 CFSR update reported being only 64% of the way toward its Performance Improvement Plan (PIP) goals for caseworker visits with birth parents at least every two months. The PIP standard is less demanding than the federal monthly visitation requirement for children in foster care. If a state cannot manage bimonthly visits with parents, the likelihood that it is conducting meaningful monthly visits with every foster child is vanishingly small.

A PIP is a negotiated remediation plan following a failed CFSR. Achieving only 64% of the targets means Virginia accomplished roughly two-thirds of what it promised to do after already being caught failing — and those targets were already compromises, not the full standard.

Kentucky’s HHS OIG audit during the pandemic found that the state “did not document all the required monthly caseworker visits.” The primary barrier cited was “insufficient caseworker staffing levels.” The pandemic did not create Kentucky’s compliance problem — it merely exposed the pre-existing staffing crisis that had been masked by creative reporting in normal times. Those shortages persist today.

In Maine, a 2024 HHS OIG audit found that 94% of sampled child abuse and neglect reports were not compliant with one or more requirements related to immediate screening, risk assessment, and investigation. Screening and investigation are the intake processes that feed the foster care system. A state that fails at 94% on the front end cannot credibly claim to be meeting monthly visitation requirements for children already in its custody. The foundational processes are broken.

CFSR Round 4: 25 Years of Universal Failure — Zero States in Full Conformity

Perhaps the most damning evidence of systemic failure is the 25-year track record of the Child and Family Services Reviews (CFSRs).

The CFSR is the federal government’s primary mechanism for evaluating state child welfare system performance. Initiated in 2001, the review process has completed three full rounds and is well into a fourth. As of mid-2026, 40 states have completed CFSR Round 4 reviews with final reports available.

In the entire history of the CFSR process — spanning four rounds across all states, the District of Columbia, and U.S. territories — NO STATE HAS EVER BEEN FOUND IN SUBSTANTIAL CONFORMITY WITH ALL SEVEN OUTCOME MEASUREMENTS AND ALL SEVEN SYSTEMIC FACTORS.

This is not a collection of individual state failures. It is evidence of a federal system that is structurally incapable of protecting the children it takes into custody.

What the Seven Outcomes Actually Measure — And Why Universal Failure Matters

The seven outcomes the federal government uses to evaluate state child welfare systems are not arbitrary bureaucratic checkboxes. They represent the core dimensions of child protection and well-being that the entire apparatus of Title IV-E funding, state agencies, caseworkers, courts, and service providers is supposed to deliver:

Safety Outcome 1: Children are, first and foremost, protected from abuse and neglect.

Safety Outcome 2: Children are safely maintained in their homes whenever possible and appropriate.

Permanency Outcome 1: Children have permanency and stability in their living situations.

Permanency Outcome 2: The continuity of family relationships and connections is preserved for children.

Well-Being Outcome 1: Families have enhanced capacity to provide for their children’s needs.

Well-Being Outcome 2: Children receive appropriate services to meet their educational needs.

Well-Being Outcome 3: Children receive adequate services to meet their physical and mental health needs.

Zero states achieving substantial conformity in any of these seven outcomes — across safety, permanency, and well-being — after 25 years of federal reviews means the federal child welfare system is failing children in every single dimension the government itself has defined as essential to its mission. Not one state has demonstrated that it can consistently protect children from abuse and neglect once they are in care. Not one state has shown it can achieve timely permanency. Not one state has shown it can meet children’s educational and health needs at scale.

The monthly caseworker visit is the primary mechanism by which these outcomes are supposed to be monitored, assessed, and achieved in real time. When that visit does not happen — or when it happens as a hollow formality without the training, authority, or supervisory support to act on what is observed — every other protective function downstream is compromised or rendered impossible. Safety assessments become guesswork based on outdated information. Permanency planning becomes disconnected from the child’s actual circumstances. Educational needs go unidentified. Health issues go unaddressed. Trafficking indicators are missed. Children who could have been protected or connected to services are left invisible until tragedy forces the system to pay attention.

States that perform best on the seven systemic factors tend to achieve conformity in areas like “Agency Responsiveness to the Community” (stakeholder engagement) and “Quality Assurance System” (data collection and CQI processes). These are institutional capacity and bureaucratic infrastructure metrics. They measure whether the agency can produce reports and convene meetings, not whether children are safe. The outcomes that directly measure child safety, permanency, and well-being remain universally unmet — suggesting that states have become quite good at building the bureaucratic apparatus of child welfare while failing at the core human mission of protecting children.

The Caseworker Staffing Crisis: The Workforce Behind the Ghost Visits

Behind every missed visit is a caseworker who does not exist or cannot possibly do the job.

Social work turnover in child welfare is estimated between 23% and 60% annually. The National Survey of Child and Adolescent Well-Being (NSCAW III) found that job stress and worker burnout were the most frequently cited reasons for turnover by supervisors (74.9%), followed by better pay elsewhere (44.6%) and workload (41.3%).

Each caseworker departure delays permanency for children by an average of six months. A child assigned to five different caseworkers over two years — a common experience — could see their permanency delayed by 2.5 years due to staff turnover alone.

Alabama: 50% turnover rate in 2024, with more than 150 caseworker positions vacant. When half the workforce turns over annually, institutional knowledge is destroyed and case continuity becomes impossible.

Hawaii: Vacancy rate reached 40% — nearly double the rate from just one year prior. When four out of ten positions are empty, remaining workers cannot physically visit all assigned children.

Missouri: 30% turnover at the end of FY2025 — and the agency was not seeking to add staff despite the chronic shortfall. This is a deliberate institutional choice to operate below capacity.

New York: Turnover among child welfare professionals exceeds 40%, reaching 60% in residential programs. Directors described the levels as “jarring.”

The relationship between staffing and compliance is not correlation — it is mechanical causation. When vacancy rates reach 30-40%, remaining workers absorb impossible caseloads. National standards recommend 12-15 cases per worker; actual caseloads frequently reach 25-40 or higher. At these levels, monthly visits become mathematically impossible regardless of worker effort or dedication.

Missing Foster Children: 47% Never Reported to NCMEC — The Visitation Connection

The connection between missed caseworker visits and missing children is not theoretical. It is documented in one of the most disturbing federal audit findings in recent child welfare history.

The HHS Office of Inspector General audited 74,353 episodes of children missing from foster care (July 2018 – December 2020). The findings:

34,869 missing child episodes (47%) were NEVER reported to the National Center for Missing and Exploited Children (NCMEC) as required by federal law.

16,246 additional episodes were not reported within 24 hours.

Nearly all states in the audit sample (46 states) lacked a system for tracking whether missing children had been reported to NCMEC.

State agencies generally lacked adequate systems to identify whether they had reported missing children accurately and in a timely manner.

A caseworker who visits a child monthly will discover within weeks if the child has run away or gone missing. Without visits, children can vanish from the system for months before anyone notices — if they are noticed at all.

In Connecticut’s 57.4% compliance environment, a child’s absence may not be detected for 60, 90, or 120+ days. Every day of delay is a day the child is in danger without anyone searching for them.

The NCMEC reporting failure is downstream of the visitation failure. States cannot report children as missing if they do not know the children are missing. And they cannot know the children are missing if they are not visiting them.

The same OIG work found that 65% of returned foster runaways in five audited states had no evidence of trafficking screening upon return. The monthly caseworker visit is the primary opportunity to screen for trafficking indicators — unexplained injuries, new possessions, behavioral changes. When visits do not occur, trafficking screening does not occur.

Residential Care: The Monitoring Black Hole

A 2024 HHS OIG report revealed that many states lack the most basic information needed to monitor maltreatment in residential facilities — group homes, residential treatment centers, and institutional placements where some of the most vulnerable children are housed.

States are placing children in congregate care and then failing to monitor whether those children are being abused or neglected. The monthly caseworker visit requirement applies to these children as well, but residential facilities present unique challenges: higher staff turnover (reaching 60% in New York residential programs), children with no consistent adult advocate, jurisdictional confusion, and facilities that control access and information flow.

The NYC Comptroller’s audit found that ACS allowed an average of 10 months between site visits to residential care facilities — far exceeding the policy of at least one visit every six months. Over three years, 2,154 incidents of neglect and/or abuse involving 1,641 children in foster care occurred, with over 76% happening while children visited their birth families.

The children in these settings are among the most vulnerable in the system. They are also among the least likely to have anyone consistently checking on them.

The Compliance Crisis Is Not an Accident — It Is the Predictable Result of the Architecture We Documented in Article 2

In Article 2, we showed how the financial architecture — open-ended Title IV-E entitlement, FMAP incentives, AFDC lookback, tiny prevention budget, and the removal of the only penalty for visit non-compliance via P.L. 118-258 — creates powerful incentives to prioritize placement volume over safety verification.

The compliance data in this article is the predictable downstream consequence of that architecture.

When the federal government reimburses states generously for every child placed in care but allocates less than $6 per month per child to check on them — and then removes even the symbolic penalty for failing to check — states respond rationally. They staff for placement, not for visitation. They report aggregate numbers that create the appearance of compliance while individual children go unseen. They accept 25 years of CFSR failure because there is no meaningful financial or political consequence for universal non-conformity.

The children are not failing. The states are not failing in isolation. The system — designed, funded, and incentivized by federal policy — is failing by design.

Conclusions and the Path Forward — What Must Change

The evidence presented in this article is overwhelming, consistent, and damning:

328,947 children in foster care as of the end of FY2024, each one legally entitled under federal law to a monthly face-to-face safety visit from their assigned caseworker.

Up to 140,000 children per month — and potentially far more — receiving no safety check at all when we apply documented state-level compliance rates to the national population.

Zero states in 25 years of federal Child and Family Services Reviews have ever achieved substantial conformity with all seven child welfare outcomes the federal government itself defined as essential.

47% of missing foster child episodes in a major HHS OIG audit were never reported to NCMEC — a direct downstream consequence of visitation failures that leave states unaware that children have vanished from their placements.

Caseworker turnover rates of 30-60% annually in multiple states, with vacancy rates reaching 40%, making meaningful monthly visitation mathematically impossible for the remaining workforce regardless of their dedication or effort.

The only federal financial penalty that ever existed for visit non-compliance was quietly eliminated by P.L. 118-258 effective October 1, 2025 — just as states were being caught failing at scale.

This is not a resource problem in the conventional sense. The federal government spends more than $11 billion annually on child welfare. This is a priority problem, an incentive problem, and an accountability problem of the highest order.

The monthly caseworker visit is the single most basic, most fundamental mechanism of child protection in the entire foster care system. It is the primary way the state knows whether a child it has taken legal custody of is safe, whether they are being abused in the placement, whether they are at risk of running away or being trafficked, whether their educational and health needs are being met, and whether the permanency plan is on track. When that mechanism fails at this scale — when up to 140,000 children per month have no one checking on them — every other protective function downstream is compromised. Safety assessments become guesswork. Permanency planning becomes disconnected from reality. Trafficking screening never happens. Educational and health needs go unaddressed. Children who could have been protected are left invisible until tragedy strikes.

The children in America’s foster care system did not choose to be there. They are there because a court or child welfare agency determined that removal was necessary for their protection. When the state takes that step, it assumes a profound legal and moral responsibility. The absolute minimum expression of that responsibility — the floor, not the ceiling — is ensuring that someone actually sees the child every month, talks to them, observes their living conditions, and acts on any danger that is detected.

In the articles that follow in this Ghost Visit Investigation series, we will examine the human cost of this failure to the children who are unseen and to the caseworkers who are set up to fail, and we will lay out the specific, achievable policy reforms required to end the compliance crisis: real-time independent verification, meaningful financial consequences tied to Title IV-E, public dashboards, independent audits, and a fundamental rebalancing of incentives so that safety verification is no longer the cheapest, most under-resourced function in the entire system.

Every child deserves to be seen.

Every missed visit is a broken promise with consequences that can last a lifetime.

The data is clear. The failure is systemic and decades in the making. The time for excuses, for self-reported aggregate numbers that hide individual children, and for removing penalties while children remain unseen — that time is over.

The children are waiting. The question is whether the system that took custody of them will finally choose to see them.

Project Milk Carton is a 501(c)(3) nonprofit dedicated to U.S. child welfare transparency, CPS and foster care accountability, human trafficking prevention, and OSINT-driven investigations. Our CivicOps database and investigative series power this research.

Data current as of June 2026. Sources include AFCARS FY2024, CFSR Round 4 Final Reports (40 states), HHS OIG audits, Connecticut Office of the Child Advocate, state APSRs, and PMC analysis. Full methodology and replication files available at projectmilkcarton.org.

Share this investigation. Demand that your state publish real-time, independently audited visit compliance data. Every child in foster care deserves to be seen — not just counted.

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 or donate at We Built the Tools. Help Us Keep Them Running.

Project Milk Carton Inc. Your gift keeps our site alive, spotlighting high-risk children!

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.

The Constitutional Republic

Discussion about this post

Ready for more?