MHPAEA Rollback Creates Two Compliance Regimes for Health Actuaries

The MHPAEA Final Rule occupies an unusual regulatory position: federal non-enforcement of a federal standard does not affect the state insurance codes that incorporate equivalent or stricter requirements. The DOL issued its enforcement suspension on March 17, 2026, citing its intention to replace the 2024 rule by December 31, 2026. What this suspension does not do is alter the insurance code requirements in states that have adopted mental health parity protections at or above the 2024 standard, suspend state examination authority over fully-insured group health products, or reduce plan participants’ rights to sue under ERISA Section 502(a). For the 153 million workers in employer-sponsored coverage, the enforcement map is now split: self-insured ERISA plans face genuinely reduced federal enforcement pressure, while fully-insured plans in most states face regulators who are not bound by the DOL’s enforcement posture and are not pausing.

From reviewing NQTL comparative analyses for mid-market group health plans over the past three years, the compliance gap between what regulators asked for and what plans actually submitted has been wider than most actuaries acknowledge. The March 2026 non-enforcement statement does not narrow that gap for state-regulated plans. For benefit design actuaries, stop-loss pricers, and health plan rate filers, the practical question is not whether parity matters in 2026 but rather which authority is asking the question and what evidence satisfies them. The answer depends entirely on plan type and state, and those two variables point in different directions for a substantial share of the market.

153M

Workers in Employer-Sponsored Plans Facing the Two-Regime Environment

300%+

DOL Parity Enforcement Action Increase Since 2023, Before March 2026 Pause

Dec 2026

Federal Replacement Rule Deadline, Leaving 8-9 Months of Regulatory Ambiguity

What the 2024 Rule Built, and What March 2026 Suspended

The September 2024 Final Rule (89 FR 77586) completed a multi-decade effort to make MHPAEA’s parity standard outcomes-based rather than design-based. Its core additions, effective for plan years beginning January 1, 2026, required plans to collect and analyze actual outcomes data to demonstrate that nonquantitative treatment limitations (NQTLs) applied to mental health and substance use disorder (MH/SUD) benefits were not more restrictive in operation than the predominant NQTLs applied to substantially all medical and surgical (M/S) benefits in the same classification.

The provisions the DOL suspended in March 2026 include three distinct regulatory requirements that had concentrated the actuarial workload. The relevant data evaluation requirement mandated prior authorization approval and denial rates, out-of-network utilization rates by service classification, and network adequacy metrics comparing MH/SUD to M/S provider panels, all compiled into a comparative analysis that named fiduciaries must certify. The prohibition on discriminatory factors and evidentiary standards prevented plans from basing NQTL criteria on historical utilization data that itself reflected prior parity violations. The meaningful benefits standard required that plans providing any benefit for a MH/SUD condition provide meaningful coverage in every classification where M/S benefits exist, with specific attention to applied behavior analysis for autism spectrum disorder, medication-assisted treatment for opioid use disorder, and nutritional counseling for eating disorders.

These three provisions required quantitative data analysis, methodology design, and ongoing monitoring. Plans with calendar-year cycles were, at the moment of the March 2026 suspension, in their first compliance year under the full framework. The actuarial infrastructure build, collecting claims denial rates across six benefit classifications, quantifying prior authorization approval disparities, assembling network adequacy evidence, and packaging results for fiduciary certification, was underway. The federal enforcement pause halts the DOL’s role as the enforcement authority for that infrastructure, but it does not eliminate the build requirement for state-regulated plans.

The ERISA/State Split: How Two Plans at the Same Employer Face Different Rules

The core compliance division follows the ERISA preemption line. ERISA generally preempts state laws that “relate to” employee benefit plans, but the ERISA savings clause preserves state authority to regulate the “business of insurance.” The deemer clause complicates this further: it prevents states from treating self-insured plans as insurance companies, which means state insurance laws, including state mental health parity laws, do not apply to self-insured ERISA plans as a matter of insurance regulation. For fully-insured plans, the analysis reverses. The carrier issuing the insurance product is subject to state insurance regulation, including state mental health parity laws, whether or not the federal government is enforcing an equivalent standard.

The majority of states have enacted mental health parity laws that incorporate or exceed the federal MHPAEA standard. Many track the 2024 final rule’s NQTL comparative analysis requirements. A state commissioner examining a carrier’s products does not need the DOL’s enforcement authorization to apply those state standards; the state law is the authority, not the federal rule. The DOL’s March 2026 statement is an instruction to DOL employees about DOL enforcement posture. It is not a preemption of state law, and it does not bind state insurance regulators.

Plan Type	Federal Enforcement (Post-March 2026)	State Enforcement	Primary Compliance Authority
Fully-insured group health (large employer)	Suspended under DOL statement	Continues in states with equivalent parity laws	State insurance commissioner
Fully-insured individual/small group (ACA-compliant)	Suspended under DOL statement; CMS separately evaluates ACA compliance	Continues under state insurance authority	State commissioner plus CMS for ACA market
Self-insured ERISA (employer-administered)	Suspended; enforcement risk materially reduced	Generally inapplicable under ERISA preemption	ERISA litigation risk; federal replacement rule (December 2026)
Self-insured with stop-loss (small/mid employer)	Suspended for the plan itself	Limited state authority over stop-loss carrier; plan is not an insurer	Stop-loss carrier compliance indirectly; plan sponsor ERISA litigation exposure
Church plans (ERISA-exempt)	Federal MHPAEA generally inapplicable	State law governs directly if the plan is insured	State commissioner for insured church plans

This table understates the nuance for self-insured plans. Some states have enacted mental health parity laws that extend beyond pure insurance regulation to reach self-funded plans through reporting mandates or data call requirements. Several states have enacted laws requiring health plan data submissions, including behavioral health utilization data, that apply regardless of whether the plan is self-insured or fully-insured. Actuaries advising self-insured plan sponsors should conduct a state-specific audit before concluding that the DOL’s non-enforcement statement provides complete relief. The absence of federal enforcement pressure is real and material for self-insured plans; it is not a blanket release from all parity-related obligations in every jurisdiction.

Federal Enforcement Context: What the DOL Had Built Before March 2026

The March 2026 suspension came after a sharp acceleration in federal parity enforcement. DOL enforcement actions under MHPAEA increased more than 300% between 2023 and 2025, with average per-action penalties reaching the $500,000 to $2 million range for plans with systematic NQTL violations. That acceleration reflected a deliberate DOL posture: the agency had been publishing enforcement data in annual reports to Congress and had explicitly documented that every NQTL comparative analysis it reviewed through 2024 was initially found deficient. The 100% initial failure rate on DOL-reviewed analyses was both an enforcement signal and a technical acknowledgment that plans had not yet built the data infrastructure the 2024 rule required.

The acceleration had created a market response. Carriers and third-party administrators were investing in comparative analysis infrastructure. Actuaries were increasingly engaged in the data collection and methodology work. Stop-loss carriers were revising their underwriting assumptions to account for network adequacy improvements that the federal rule would eventually require. The March 2026 suspension reverses the federal enforcement signal for self-insured plans but does not recover the compliance investment already made, and it does not close the enforcement window for two alternative channels.

ERISA Section 502(a)(1)(B) allows plan participants to sue for denied benefits. An adverse benefit determination that fails the NQTL parity test, for example, a prior authorization denial for an outpatient MH/SUD service that would not require prior authorization if it were a comparable M/S service, is a potential 502(a)(1)(B) claim regardless of whether DOL is enforcing the parity standard. Participant litigation risk did not disappear when the DOL issued its non-enforcement statement. And the fiduciary certification requirement for NQTL comparative analyses, which took effect for plan years beginning January 1, 2025, creates personal liability for named fiduciaries who certified analyses that they did not adequately review or that were materially deficient. Those certifications already exist for the 2025 plan year. The named fiduciaries who signed them remain exposed regardless of federal enforcement posture.

What State Regulators Are Doing

The NAIC Mental Health Parity (MHPAEA) (B) Working Group continued its monitoring and coordination activity through Spring 2026, reviewing state enforcement patterns and collecting data on carrier compliance with NQTL comparative analysis requirements. The Working Group has no direct enforcement authority, but it coordinates state insurance commissioner positions and develops model regulation that states adopt into their codes. The Working Group’s continued activity through the federal enforcement pause reflects the structural reality: state commissioners are the direct regulatory authority over fully-insured carriers, and the Working Group exists to coordinate their approach.

Several states have explicitly stated that their mental health parity enforcement programs operate independently of the federal posture. States that adopted the NAIC Model Mental Health Parity Act (Model Act 74) or enacted their own parity legislation referencing the 2024 MHPAEA standard are enforcing those requirements through state examination programs. Those examinations review the same data elements the federal rule requires: prior authorization denial rates by service classification, out-of-network utilization comparisons between MH/SUD and M/S services, provider credentialing standard documentation, and network adequacy metrics. The examination criteria did not change when the DOL issued its non-enforcement statement.

A 2026 Commonwealth Fund analysis documented the divergence explicitly: states accounting for a majority of the fully-insured commercial market population were continuing enforcement under 2024-equivalent standards for the products they regulate. Carriers administering fully-insured business across multiple states face compliance requirements that vary by jurisdiction and are not uniformly relieved by the federal suspension. A carrier with a twenty-five-state fully-insured book of business faces twenty-five potentially different enforcement environments. Oliver Wyman’s March 2026 brief, “Why Insurers Must Act on Mental Health Parity Now,” was addressed specifically to fully-insured carriers operating in that environment: the data infrastructure investment required to demonstrate NQTL parity is long-lead, state examinations are not scheduled around federal enforcement calendars, and carriers that defer because of the DOL’s posture will be inadequately prepared when their state examination arrives.

Georgia’s January 2026 enforcement action, which imposed nearly $25 million in fines against 11 insurers for more than 6,000 parity violations, was the most visible illustration of what state enforcement looks like when the federal government is not the primary actor. Georgia used state examination authority and a 2022 state parity law to conduct the largest single mental health parity enforcement action in state regulatory history. The action was in progress and largely resolved before the DOL issued its March 2026 suspension. It demonstrates that state enforcement timelines operate on their own cadence.

Stop-Loss Pricing in the Compliance Gap

Stop-loss insurance, purchased by self-insured employers to cap catastrophic claims exposure, is state-regulated as an insurance product even though it attaches to self-insured ERISA plans. The network adequacy and prior authorization parity improvements the 2024 federal rule required, had they been fully implemented across the self-insured market, would have shifted MH/SUD utilization patterns in ways that stop-loss pricers had begun incorporating into their forward models.

The core actuarial concern is behavioral health out-of-network utilization. Milliman research documented that behavioral health inpatient services were 5.2 times more likely to be rendered out of network compared to M/S inpatient services, a ratio that had worsened from 2.8 times in 2013. Out-of-network behavioral health claims are typically reimbursed at higher levels relative to billed charges than in-network claims, and they are more likely to reach specific deductible attachment points because the billed charges are not discounted by in-network contracting. Stop-loss pricers who built 2026 renewal models assuming that network adequacy improvements from federal rule compliance would shift behavioral health utilization toward in-network providers need to revisit those assumptions for self-insured clients. Those improvements are not materializing on the federal enforcement timeline originally anticipated.

The Oliver Wyman analysis identified network adequacy as the provisions where non-enforcement creates the largest actuarial uncertainty. The 2024 rule’s ghost-network elimination requirement, which mandated that plans compare the percentage of in-network MH/SUD clinicians actively submitting claims against the equivalent M/S metric, was a direct mechanism for moving behavioral health utilization into the in-network channel. Without that enforcement pressure on self-insured plans, the ghost-network effect, directories listing providers who are technically in-network but not accepting new patients or not billing the plan, persists as a structural driver of out-of-network utilization and elevated stop-loss frequency.

For self-insured employers, the stop-loss pricing effect cuts in an additional direction. Fully-insured competitors in the same labor market must comply with state parity requirements, including network adequacy standards, for their group products. If those standards produce better in-network access to behavioral health services for fully-insured employees, self-insured employers who do not achieve equivalent access may face workforce dynamics that show up in downstream health data: delayed treatment, higher acute episode rates, and the kind of catastrophic claims that stop-loss contracts are designed to catch. That is not a regulatory argument; it is an actuarial one.

The 2027 Rate Filing Problem

The replacement rule timeline is where the actuarial reserving uncertainty concentrates. The DOL has committed to issuing a replacement proposed rule by December 31, 2026. A proposed rule, after public comment and agency revision, typically requires six to twelve months before a final rule is published. The 2027 plan year will begin January 1, 2027. Carriers filing rates for 2027 ACA and group health products, with submission windows running through mid-2026 for most state-regulated markets, are pricing benefit packages against a regulatory standard they cannot yet see.

For ACA carriers, the asymmetry is acute. The 2027 rate filings confirmed for several states in recent months show premium increase requests in the 22% to 30% range, driven primarily by post-subsidy adverse selection and pharmacy trend, not by behavioral health parity compliance costs. But the benefit design decisions embedded in those filings, particularly decisions about prior authorization requirements for MH/SUD services, network composition, and coverage of ABA therapy and MAT, are compliance decisions as well as cost decisions. A filing that assumes the replacement rule will relax parity requirements and designs benefits accordingly creates regulatory exposure if state requirements remain unchanged. A filing that maintains full 2024 compliance for state-regulated products is almost certainly correct in states that have incorporated those standards into their insurance codes, and it leaves the carrier well-positioned regardless of what the replacement rule requires.

Large group self-insured employers planning 2027 benefits face a different version of the same problem. Benefits committed for the 2027 plan year, particularly NQTL design decisions embedded in TPA contracts and plan documents, may need mid-year amendment if the replacement rule differs materially from the current posture. Self-insured plan sponsors should begin 2027 plan amendment planning with explicit contingency provisions for regulatory change. That is not a prediction about the replacement rule’s content; it is basic risk management for a plan sponsor making twelve-month binding commitments in an environment with a known regulatory change event scheduled for the final week of 2026.

The actuarial implication for reserve estimation is also direct. Health plan liabilities under ASOP No. 5 (Incurred Health and Disability Claims) require actuaries to account for changes in benefit coverage and administrative procedures that affect the liability estimate. An actuary certifying 2027 reserves for a state-regulated carrier needs to document the compliance posture assumed in the reserve calculation: whether the reserve assumes continued state enforcement of 2024 standards, whether it reflects expected compliance costs for the NQTL comparative analysis infrastructure, and how the reserve would change under different replacement rule scenarios. That documentation is not optional; it is the professional standard for reserving under regulatory uncertainty.

Actuarial Tasks That Cannot Wait for the Federal Replacement

Five specific actuarial tasks need to proceed regardless of the federal enforcement posture, and for different reasons in each case.

State-specific compliance mapping. The compliance obligation for a carrier or employer group health plan is not uniform across states. A carrier administering fully-insured products across multiple states faces multiple potentially different enforcement environments. Actuaries advising those carriers need a state-by-state map of which 2024 provisions are embedded in state law, which states have confirmed continued enforcement, and which states may be deferring to the federal posture pending the replacement rule. This mapping is not purely a legal product; it has direct actuarial consequences for reserve estimates, rate adequacy filings, and risk-transfer structure. The mapping also changes as state legislatures and regulators respond to the federal suspension, so it requires active monitoring through 2026.

NQTL comparative analysis documentation. Maintaining a current comparative analysis serves three purposes even with federal enforcement suspended. It supports state examination responses for fully-insured carriers. It provides a record of compliance posture in the event of participant litigation under ERISA Section 502(a)(1)(B), where the plan’s documented compliance reasoning is part of its defense of adverse benefit determinations. And it gives the plan a calibrated baseline from which to adjust when the replacement rule is issued, whether the new rule is more or less demanding than the 2024 standard.

Prior authorization denial rate tracking. The CMS-0057-F rules requiring public disclosure of prior authorization denial rates for Medicare Advantage plans have increased the visibility of denial rate comparisons across the market. Many states collect comparable denial rate data for commercial group health plans through market conduct data call requirements that exist independently of MHPAEA. Plans that cannot demonstrate parity in prior authorization outcomes face state examination exposure for fully-insured products and ongoing ERISA litigation risk for self-insured products. The tracking infrastructure should remain operational.

Stop-loss attachment point review. Self-insured plan sponsors who built 2026 renewal assumptions around network adequacy improvements that federal rule compliance would have accelerated should revisit those assumptions with their stop-loss writers. Where MH/SUD utilization was projected to shift toward in-network providers based on the 2024 rule’s ghost-network elimination requirements, that shift may not materialize on the federal enforcement timeline. The attachment point selection consequences are plan-specific but are a legitimate actuarial adjustment for plans whose models incorporated the compliance trajectory.

Replacement rule contingency planning. The December 31, 2026 deadline for the proposed rule gives actuaries an approximate horizon for when the next regulatory signal will arrive. Plans that begin scenario analysis now, mapping their benefit designs against plausible replacement rule structures ranging from a near-identical standard to a substantially relaxed one, will be better positioned to respond quickly when the proposed rule is published. The comment period after a proposed rule is issued typically runs 60 days. A carrier that has already stress-tested its 2027 benefit design against multiple scenarios can respond to the proposed rule substantively during that window, rather than beginning the scenario analysis after the rule is published.

Why This Matters for Health Actuaries

The MHPAEA enforcement landscape in 2026 is not a simple relaxation of requirements. For actuaries advising fully-insured carriers, it is a continuation of the 2024 compliance environment under state authority. For actuaries advising self-insured employers and stop-loss writers, it is a period of reduced federal enforcement pressure with sustained ERISA litigation exposure and a known replacement rule deadline. In neither case does the federal non-enforcement statement eliminate the actuarial work; it changes who is asking for it.

Milliman’s detailed guidance on navigating the mental health parity rules, authored by Katie Matthews, Travis Gray, and Stoddard Davenport, remains the most comprehensive published actuarial framework for this work. The Milliman methodology and the state enforcement guidance from NAIC function as the de facto practice standard for actuaries in this gap period. Actuaries performing NQTL comparative analysis work should also consider whether their work product, when structured to support fiduciary review and certification, is appropriately documented under ASOP No. 25 (Credibility Procedures) where data quality or volume is a limiting factor, and whether an actuarial opinion letter accompanying the comparative analysis provides the “assurance” that the 2025 fiduciary certification provisions reference.

The replacement rule, when it arrives by December 31, 2026, will be issued against a market that has either maintained its compliance infrastructure or let it erode. The two-regime environment makes the maintenance argument strongest for fully-insured carriers, where the continuity of state enforcement provides no option. For self-insured plans, the argument is about optionality: a plan that preserved its NQTL analysis infrastructure can calibrate its response to the replacement rule quickly; a plan that dismantled that infrastructure during the enforcement gap will spend the first months of 2027 rebuilding what it previously had. Plans do not receive advance notice of what the replacement rule will require. The actuarial best practice is to preserve the capacity to comply, regardless of which direction the replacement rule moves.

The MHPAEA Rollback Leaves Health Actuaries Navigating Two Compliance Regimes