Secret-shopper studies keep finding that most listed behavioral health providers are unreachable, and a 2026 wave of lawsuits, settlements, and a new federal law is forcing the issue.
Key Takeaways
- Most listed providers are not really available: Repeated secret-shopper studies find that only about 18 percent of listed behavioral health providers can actually be reached and booked. The directories describe a network that does not exist.
- Enforcement arrived in force in 2026: A major insurer paid a record state penalty, a professional association filed a class action, and Congress passed a directory-accuracy law. What was a complaint is becoming a legal liability.
- The harm is financial as well as clinical: Patients who hit dead ends delay care or get pushed to out-of-network providers and surprise bills. Inaccurate directories are now the basis of parity and fiduciary claims.
- Accuracy is becoming an operational mandate: New rules require ongoing verification, secret-shopper audits, and cost protections when directories fail. Payers and the provider data behind them face real accountability for the first time.
A mother whose teenage son just came home from three days on a psychiatric hold opens her insurer’s website, filters for a therapist who takes her plan, and starts calling down the list. The first number is disconnected. The second belongs to a clinician who left the network two years ago. The third is a valid office, but the therapist is not accepting new patients and has not been for months. The fourth never calls back.
By the tenth dead end, she has learned what researchers keep documenting in study after study: that the network her insurance card promises is substantially fictional, a directory of providers who exist on paper and cannot be reached in practice. The industry has a name for this, borrowed from the eeriness of the experience: the “ghost network.”
The phrase is not new. It surfaced in health-policy and consumer-advocacy circles in the mid-2010s, as researchers and state regulators began auditing the mental health directories of Medicaid and Medicare Advantage plans and finding that huge shares of the listed providers were unreachable, and it has been a fixture of the parity debate ever since. What is new, and what makes this the moment to pay attention, is that the complaint has finally acquired teeth. Over the course of 2026, a record financial penalty, a class-action lawsuit, a wave of state enforcement, and a new federal law have converged to turn directory accuracy from a chronic annoyance into a legal and financial liability for the payers who maintain these lists. The ghosts are being counted, and someone is finally being asked to pay for them.
What the Secret-Shopper Studies Keep Finding
The evidence has been remarkably consistent, and remarkably damning, across years and methods. When the Senate Finance Committee ran a secret-shopper study, sending staff to call providers listed in insurance directories and try to book a mental health appointment, they succeeded just 18 percent of the time. A 2024 study in JAMA, led by researchers at Weill Cornell Medicine, found the same figure from the other direction: only about 18 percent of mental health clinicians listed as in-network for Medicaid plans were actually reachable, accepting the plan, and able to offer a new-patient appointment. When New York’s attorney general deployed secret shoppers against a single large insurer’s behavioral health directory, more than 80 percent of the providers listed as accepting new patients turned out to be effectively unavailable.
That 18 percent figure repays a second look, because it inverts the entire premise of a provider network. A directory is supposed to be a map of where care can be found. When four in five of the destinations on the map are dead ends, the map is not merely imperfect; it is actively misleading, sending people in crisis on a series of failed calls at exactly the moment they have the least capacity to absorb them. And the failure is concentrated in behavioral health specifically, worse than in other specialties, which is why it has become inseparable from the broader fight over whether insurers actually deliver the mental health coverage that parity law requires them to provide.
Why Ghost Networks Persist
Ghost networks endure for reasons partly mundane and partly structural, and the structural ones are the reason the problem has proven so durable. The mundane part is data decay: provider directories are notoriously hard to keep current, clinicians move, retire, change their panel status, and close their practices to new patients, and the information rots faster than most insurers bother to update it. Keeping a directory accurate requires continuous verification work that costs money and generates no revenue, so it has chronically been nobody’s priority, an information problem adjacent to the payer-controlled credentialing and provider-data systems that increasingly gate access to care.
The structural part is the less innocent one. A thin behavioral health network is cheaper for an insurer than a full one, and a directory padded with unavailable providers lets a plan appear to meet network-adequacy requirements without actually paying to build the network those requirements are meant to guarantee. When a member cannot find an in-network provider and gives up, the plan pays nothing; when the member gives up and sees an out-of-network provider instead, the member absorbs the higher cost. The incentives, in other words, do not point toward accuracy. A directory full of ghosts is not simply a failure of data hygiene; for a plan that underpays behavioral health clinicians and struggles to keep them in network to begin with, it can be a quiet convenience, which is why external enforcement has turned out to be necessary.
How 2026 Enforcement Changed the Stakes
That convenience is becoming costly. In February 2026, New York’s attorney general reached a settlement with EmblemHealth for more than 2.5 million dollars, the largest ghost-network penalty the state had secured, after finding the insurer had overstated the availability of its behavioral health providers by as much as 80 percent. The penalty was notable less for its size than for its terms, which required ongoing secret-shopper audits with public reporting, rapid correction of directory errors, guaranteed appointment-access timeframes, and, critically, a guarantee that members misled by a bad listing into seeing an out-of-network provider would pay no more than their in-network cost. The settlement also carried a pointed history: the same insurer had promised to fix its directories in a prior settlement years earlier and had not.
State enforcement was one front among several. In January 2026, the American Psychiatric Association filed a class-action lawsuit against the same insurer on behalf of members and psychiatrists, alleging the ghost networks forced patients to forgo care or pay out of network. Courts have grown more receptive to these theories: a federal case, Hecht v. Cigna, was allowed to proceed on the argument that directory inaccuracy can constitute a fiduciary breach under federal benefits law, which opens a litigation avenue far more threatening to payers than a state fine. And Congress acted, passing the REAL Health Providers Act, which will require Medicare Advantage plans to maintain verified, accurate directories and to publish accuracy results, with the core requirements phasing in over the next few years. The combined effect is that directory accuracy is shifting from a soft expectation to a hard obligation with financial and legal consequences attached.
What Ghost-Network Enforcement Means for the Industry
For payers, the era of the costless ghost network is ending, and the operational implications are substantial. Maintaining a directory that can survive a secret-shopper audit requires continuous provider-data verification, the kind of unglamorous infrastructure investment that has been easy to defer and will now be expensive to keep deferring, and the cost-protection provisions emerging in settlements mean that inaccurate listings will increasingly convert into direct financial liability when members are pushed out of network. The provider-data vendors and verification services that can actually keep a directory current are positioned to benefit from a compliance need that just became non-optional.
For behavioral health providers, the shift is more welcome, because ghost networks have always been partly a symptom of a deeper problem the enforcement is now forcing into view, the fact that behavioral health networks are thin because clinicians will not join or stay in them at the rates and under the administrative conditions the plans offer. A directory cannot be made accurate simply by deleting the ghosts; the network has to actually contain enough reachable providers to meet the need, which eventually routes back to reimbursement and the chronic underpayment that keeps clinicians out of network in the first place. The enforcement wave will clean up the directories, which is worth doing. The harder fix is a behavioral health network deep enough that the map finally matches the territory, and the fines and lawsuits are circling that problem without yet reaching it. Cleaning up the directories is the easy part. Building the network they are supposed to describe is the part no settlement has managed to compel.
