What Happened
The Department of Labor’s Employee Benefits Security Administration (EBSA) has named MHPAEA compliance one of its two stated priority enforcement projects for 2026, alongside surprise billing. The enforcement signal is now quantified: of 210 comparative analyses reviewed under the Nonquantitative Treatment Limitation (NQTL) requirements of the Mental Health Parity and Addiction Equity Act (MHPAEA), every single one initially failed to meet compliance standards, according to DOL enforcement data reported through 2026.
EBSA’s enforcement activity, documented in the 2025 MHPAEA Report to Congress covering August 2023 through July 2025, includes 42 initial letters requesting comparative analyses for 77 NQTLs, 25 initial determination letters finding violations for 43 NQTLs, 5 final determinations of noncompliance, and corrective actions affecting more than 18 million participants across more than 39,000 group health plans. A MHPAEA final rule rewrite is expected from DOL by year-end 2026, but enforcement under existing law continues during the rulemaking period.
Who It Affects
Every self-insured employer, public entity, university, or hospital system operating a self-funded group health plan with mental health and substance use disorder (SUD) benefits is within scope. DOL enforcement authority under MHPAEA reaches self-funded ERISA plans regardless of size. The most common failures cluster around four NQTL patterns: prior authorization criteria written tighter for outpatient behavioral health than for comparable outpatient medical procedures; step therapy protocols applied only to mental health drugs; out-of-network reimbursement rates that structurally disadvantage behavioral health networks; and geographic access standards that tolerate longer travel distances for SUD treatment than for equivalent medical care.
Plan sponsors are ultimately responsible for NQTL compliance even when a TPA or benefit administrator produces the comparative analysis. A failed DOL audit lands on the plan, not the vendor.
The Reserve Mechanism
The exposure here is contingent IBNR, a category that does not appear in standard IBNR development at all. When DOL issues a corrective action order requiring retroactive payment of benefits denied in violation of MHPAEA, those denied claims were never incurred in the plan’s paid or reported claims data. They were processed as denials, not as claims. No development triangle captures them; the actuary’s chain-ladder or Bornhuetter-Ferguson estimate builds from historical experience that excludes the denied population entirely.
The liability window is open-ended. DOL corrective action orders carry no defined lookback cap. The retroactive exposure can extend years into prior plan experience, meaning the contingent IBNR for a plan with persistent prior authorization or network-adequacy NQTL failures could be material relative to annual claim budgets. Stop-loss policies typically exclude retroactive benefit obligations ordered as a plan compliance failure rather than covered claim experience; plan sponsors should verify specific policy language before assuming any reinsurance protection applies.
As noted in DOL Parity Report Flags Self-Funded Plans for NQTL Gaps, this exposure runs through the plan sponsor’s balance sheet, not the TPA’s. And as MHPAEA Rule Rewrite by Year-End Leaves Self-Funded Plans Guessing documented, enforcement will not pause while the rulemaking proceeds. For the mechanics of how self-funded health plan IBNR is built from claims data, see Self-Funded Health Plan IBNR.
What This Means for Your Next Review
At your next reserve study or interim monitoring meeting, ask two questions: Has the plan completed a written NQTL comparative analysis that would survive a DOL document request? And if DOL ordered retroactive payment of denied behavioral health claims, what dollar exposure would the plan face for the past two to four plan years? If neither question has a ready answer, the contingent IBNR line is blank when it should not be.
The mechanics of quantifying this exposure are non-standard: the starting point is a review of denied behavioral health claims against analogous medical procedures, not a development triangle. That exercise belongs to the compliance side, but its output feeds directly into whether your accrual is defensible.