Employer Health Costs Hit 15-Year High at $18,500 Per Worker: The Actuarial Trend-Setting Challenge

From reviewing group health renewal data across mid-size self-funded plans over the past three years, the inflection point is unmistakable: plans that left specialty pharmacy open without prior authorization saw trend acceleration of 40% to 70% above baseline. Mercer’s latest survey of over 1,700 U.S. employers confirms this observation at scale. Total health benefit cost per employee is projected to rise 6.5% in 2026, the steepest increase since 2010, pushing per-employee costs above $18,500. Without employer cost-reduction measures, the increase would have been approximately 9%. The story behind those numbers is what makes this survey significant for actuarial practice: both healthcare price and utilization are rising simultaneously, breaking the decade-long pattern where one component typically offset the other. For group benefits actuaries setting trend assumptions, constructing renewal projections, and advising on plan design, this convergence requires a fundamental reassessment of how trend models are structured and calibrated.

The 2026 Cost Landscape: Survey Consensus

Mercer’s National Survey of Employer-Sponsored Health Plans, now in its 40th year, surveyed 2,010 employers in 2025, with results weighted to represent all U.S. health plan sponsors with 50 or more employees. The preliminary findings, drawn from over 1,700 employer responses collected between June and August 2025, establish the statistical foundation for the 6.5% projection.

Mercer is not alone in flagging this acceleration. The major health benefit cost surveys released for the 2026 plan year converge on a narrow band that reinforces the severity of the trend environment:

The spread between the lowest and highest projections (6.5% to 10.0%) narrows considerably when you account for methodology. Mercer’s 6.5% figure incorporates the effect of employer plan design changes and cost-sharing shifts that other surveys report separately. Aon’s 9.5% and PwC’s 8.5% measure the underlying medical trend before employer intervention. The Segal and IFEBP figures capture employer-reported expectations, which tend to run slightly higher than actuarially adjusted projections. Across all methodologies, the consensus points to an underlying medical trend between 8.5% and 9.5%, the fourth consecutive year at elevated levels following a decade of moderate increases averaging only about 3% annually.

The Price-Utilization Convergence: Why Conventional Trend Models Break

Sunit Patel, Mercer’s U.S. Chief Actuary for Health and Benefits, identified the core actuarial challenge in the survey release: "Health benefit cost trend has two primary components: healthcare price and utilization. Right now, both are rising."

This observation carries more weight than it might appear. For most of the 2014 to 2022 period, group health actuaries could decompose medical trend into price and utilization components and rely on a rough offsetting dynamic. When provider reimbursement rates rose faster than expected, utilization often lagged due to higher out-of-pocket exposure from plan design changes, or vice versa. The standard actuarial approach was to model these components semi-independently, with an implicit assumption that the cross-product term (the interaction between price and utilization changes) was small enough to ignore or absorb into a general margin.

That framework no longer holds. The drivers pushing price higher are structurally different from those driving utilization higher, and neither shows signs of reverting to the mean in the near term:

Price accelerators. Advances in diagnostics and therapeutics, particularly in oncology immunotherapies and GLP-1 medications, produce measurably better outcomes but command substantially higher unit costs than the treatments they replace. Provider consolidation has strengthened health system negotiating leverage on reimbursement rates. And healthcare-sector wage inflation, running well above the broader CPI, has pushed facility and professional fee schedules higher. AI-assisted billing optimization platforms at large health systems are also increasing coding intensity, which actuaries see as a price effect even though no additional clinical service is delivered.

Utilization accelerators. Post-pandemic catch-up demand for deferred care continues to generate above-baseline visits for screenings, elective procedures, and chronic condition management. Inpatient behavioral health claims increased nearly 80% between January 2023 and December 2024, according to PwC’s Behind the Numbers report, reflecting both genuine demand increases and expanded access through telehealth and the Mental Health Parity and Addiction Equity Act (MHPAEA) compliance requirements. Virtual care adoption has removed geographic barriers, expanding the addressable patient population for specialty services. And the healthcare labor supply has eased, partly through AI augmentation of clinical workflows, enabling providers to see more patients per day.

For the group benefits actuary, the practical consequence is that multiplicative trend models (price trend × utilization trend) now produce materially higher projections than additive models that held during the lower-trend era. A 5% price increase combined with a 4% utilization increase does not produce a 9% total trend; it produces a 9.2% trend after the interaction term. When both components were running at 1% to 2%, the difference was negligible. At current levels, the interaction term alone adds 20 to 40 basis points to the total projection, enough to move a mid-size employer’s renewal by hundreds of thousands of dollars.

Catastrophic Claims and the Specialty Drug Spike

The Mercer survey identified managing high-cost claims as the top strategic priority for employers over the next few years. The data supports that urgency. Sun Life reported a 29% increase in claims exceeding $1 million in 2024, and a 47% surge in claims exceeding $3 million. Claims above $20 million, once considered statistical outliers, are now appearing with enough regularity to affect plan-level experience. The International Foundation of Employee Benefit Plans found that catastrophic claims (cited by 31% of respondents) and high-cost prescription drugs (23%) were the two most frequently identified cost drivers for 2026.

Specialty pharmacy is where the catastrophic claim trend and the drug cost trend converge. National GLP-1 spending rose more than 500% between 2018 and 2023, from $13.7 billion to $71.7 billion, according to HFMA analysis. GLP-1 medications are projected to account for 14% of all prescription drug spending in 2026. Mercer’s survey found that 49% of large employers now cover GLP-1s for weight management, up from 44% the previous year.

Prescription drug spending among large employers rose 9.4% on average in 2025, according to Mercer, with pharmacy trend running 2.5 percentage points above medical trend per PwC’s analysis. Plans that expanded GLP-1 coverage without implementing prior authorization protocols, step therapy requirements, or specialty pharmacy carve-outs experienced the most severe cost acceleration. From tracking renewal data across regional carriers, plans that left specialty pharmacy formulary architecture open saw year-over-year trend acceleration of 40% to 70% on the specialty line, driven overwhelmingly by GLP-1 utilization.

The actuarial challenge here is credibility. GLP-1 adoption is still on the steep portion of a utilization S-curve. Historical prescription drug trend factors based on the 2018 to 2022 period systematically underestimate the specialty pharmacy component because those years preceded the inflection in GLP-1 uptake. Actuaries setting 2027 trend assumptions need to isolate GLP-1 from base pharmacy trend at the NDC level, model the adoption curve separately using logistic growth parameters, and stress-test the interaction with stop-loss attachment points. A plan with a $250,000 specific deductible that was adequate in 2023 may be materially underpriced given current large-claim frequency.

Employer Plan Design Response: 59% Making Changes

The Mercer survey documented a sharp escalation in employer cost-shifting activity. In 2024, 44% of employers implemented plan design changes to manage cost growth. That figure rose to 48% in 2025 and is projected to reach 59% in 2026. The changes are concentrated in three categories: raising deductibles and out-of-pocket maximums, tightening network configurations, and adding utilization management requirements for high-cost services.

Among large employers (500 or more employees), 51% said they are likely or very likely to make design changes that shift costs to employees, primarily through higher deductibles, copays, or out-of-pocket maximums. Over one-third of large employers will offer non-traditional network plans in 2026, such as high-performance networks or centers-of-excellence arrangements, to steer utilization toward lower-cost, higher-quality providers.

The cost-sharing data from Mercer and KFF illustrates where the deductible frontier currently sits:

The KFF 2025 Employer Health Benefits Survey reported average annual premiums of $9,325 for single coverage and $26,993 for family coverage, with year-over-year increases of 5% and 6% respectively. Workers now contribute $6,850 annually toward family coverage. Thirty-four percent of covered workers face a deductible of $2,000 or more for single coverage, with that figure exceeding 53% at small firms (under 200 workers).

Ed Lehman, Mercer’s U.S. Health and Benefits Leader, framed the employer dilemma: employers want to minimize increases in paycheck deductions while ensuring employees can afford needed care. Two-thirds of large employers are expanding access to behavioral healthcare. Sixty-seven percent now offer three or more medical plan options, up from 60% in 2023, reflecting a shift toward plan menu diversification as a cost-management tool.

For actuaries advising employers on plan design, the critical question is how much incremental cost-sharing actually suppresses utilization at current trend levels. During the low-trend era of 2014 to 2019, deductible increases of $250 to $500 reliably produced 1 to 2 percentage points of trend mitigation. Patterns we have seen in recent renewal cycles suggest that the marginal deterrent effect of cost-sharing has weakened. When the cost drivers are catastrophic claims and specialty drugs, services that blow through deductibles and out-of-pocket maximums within a single episode, increasing the deductible from $1,500 to $2,000 has almost no effect on the claimants generating the majority of plan spend. The 1% of members responsible for up to 35% of total healthcare spend, as documented by Mercer, are largely insensitive to deductible levels.

Self-Funded Mid-Size Employers: The Unmanaged Risk

Self-funded arrangements now cover the majority of commercially insured lives in the United States, and the cost acceleration documented by Mercer carries asymmetric risk for mid-size self-funded employers (typically 200 to 2,000 employees). These employers bear direct claims risk, often with specific stop-loss deductibles in the $200,000 to $350,000 range, and many lack the actuarial staffing or consulting relationships needed to model emerging trend dynamics.

The stop-loss market is already reflecting the large-claim frequency surge. Several of the largest stop-loss writers, including Cigna, Voya, and Sun Life, have reported adverse claims experience that will flow through to higher renewal rates in 2026 and 2027. The median specific stop-loss amount for employers with 500 to 4,999 employees is $250,000, according to industry survey data. For employers with 10,000 or more employees, it is $725,000. Neither attachment point was designed for an environment where $1 million claims are reported by nearly half of plans and $3 million claims have increased 47% year over year.

Self-funded employers without explicit cost management strategies face several compounding risks at renewal: the underlying trend acceleration (which inflates expected claims), the large-claim frequency spike (which pressures stop-loss pricing), and the GLP-1 utilization ramp (which can move a plan’s pharmacy spend from a manageable percentage to a dominant cost category within a single plan year). When these three forces converge in the same renewal cycle, the aggregate financial exposure can exceed what mid-size employers have budgeted, particularly those that have not updated their trend assumptions or stop-loss attachment points in the past two years.

Milliman’s actuarial consulting practice has emphasized that managing and planning for the healthcare trend has never been more central to employer-sponsored plan viability. Employers that conduct annual actuarial reviews of their plan design, claims data, and stop-loss adequacy are positioned to anticipate and manage these cost pressures. Those that rely on broker-driven renewals without independent actuarial analysis face the highest financial risk.

Historical Context: The Decade of Suppressed Trend and Its End

To understand why the current trend environment is significant, consider the trajectory. Between roughly 2013 and 2022, employer health benefit cost growth averaged approximately 3% annually. Multiple factors contributed to this historically low period: the Affordable Care Act’s cost-containment provisions (including the Cadillac tax threat, which was never implemented but influenced plan design), aggressive adoption of high-deductible health plans, narrow network strategies, generic drug substitution, and relatively stable utilization patterns.

That era conditioned a generation of group benefits actuaries and employer CFOs to expect low-single-digit cost growth as normal. Trend assumptions in renewal projections, reserve calculations, and OPEB valuations were calibrated to a world where 3% to 5% annual increases were the baseline. The shift to 8.5% to 9.5% underlying trend, now in its fourth consecutive year at elevated levels, is not a temporary spike. It reflects structural changes in the healthcare delivery and payment system that are unlikely to revert.

The ACA individual market has seen even more dramatic cost acceleration. ACA premiums increased by 21.7% on average in 2026, according to Commonwealth Fund analysis, driven by insurer expectations of adverse selection following the expiration of enhanced premium tax credits. While this is primarily an individual market phenomenon, it provides a leading indicator of the medical cost pressures that will continue flowing into the employer-sponsored market, since the underlying provider reimbursement rates and utilization patterns affect both segments.

Actuarial Implications: Adjusting the Framework

For practicing group benefits actuaries, the Mercer data and the broader survey consensus demand several adjustments to standard practice:

Trend decomposition and multiplicative modeling. Additive trend models that sum a price component and a utilization component should be replaced with multiplicative structures that account for the interaction term. At current trend levels, the interaction adds meaningful basis points to the projection. Actuaries should also decompose the pharmacy trend into a base component and a specialty/GLP-1 component, modeling the latter separately with adoption-curve parameters rather than historical regression.

Stop-loss stress testing. Specific deductible levels set during the low-trend era are likely inadequate given current large-claim frequency. Actuaries should model the probability distribution of individual claims using updated severity parameters that reflect the Sun Life and industry data on $1M+ and $3M+ claim frequency, and test whether the current attachment point still provides meaningful risk transfer at the expected premium.

Plan design elasticity reassessment. The marginal trend-suppression effect of deductible increases should be re-estimated using recent claims data rather than historical elasticity assumptions. When catastrophic claims and specialty drugs dominate plan spend, traditional cost-sharing tools have diminished returns. Actuaries advising employers should quantify the gap between expected and actual cost-sharing savings and communicate the implications for plan design strategy.

Reserve and IBNR adequacy. For health insurers and self-funded plans that hold IBNR reserves, the trend acceleration means that development patterns calibrated to the low-trend era will understate ultimate claims. Completion factors should be reviewed against the most recent 12 to 18 months of paid claims data, with particular attention to the specialty pharmacy lag, where high-cost claims often take longer to process through prior authorization and appeals workflows.

OPEB and retiree health valuation. Actuaries valuing other post-employment benefits should update their long-term healthcare cost trend assumptions to reflect the structural shift. The assumption that trend will grade down to 4% to 5% within three to five years, common in many OPEB valuations, should be tested against the evidence that the current trend drivers are persistent rather than cyclical.

Potential Deflators: What Could Moderate the Trajectory

Not all forces point upward. Biosimilars were cited as the top cost deflator by health plan actuaries in PwC’s survey for the third consecutive year, with biosimilar adoption rates increasing significantly through 2024 and 2025. Humira biosimilars alone have saved employers an estimated $2 billion to $3 billion annually since their 2023 launch. Additional high-impact biosimilar entries expected through 2027, including for oncology and ophthalmology biologics, could compress specialty pharmacy trend by 1 to 2 percentage points.

GLP-1 pricing may also moderate. Compounded semaglutide and tirzepatide, while facing FDA enforcement actions, have demonstrated employer willingness to use lower-cost alternatives. Novo Nordisk and Eli Lilly have both introduced value-based contracting arrangements with select PBMs, tying rebate levels to clinical outcomes. If Medicare Part D price negotiation authority extends to GLP-1 categories, the reference pricing effect could pull commercial rates lower over a two- to three-year horizon.

AI-enabled care navigation and utilization management tools are showing early returns in redirecting low-acuity visits from emergency departments and specialist offices to lower-cost settings. Several large carriers have reported double-digit reductions in unnecessary imaging and specialist referrals through AI-powered prior authorization workflows, though the net cost impact after implementation expense remains difficult to isolate.

None of these deflators is expected to reverse the trend trajectory in 2026 or 2027. The actuarial consensus, reflected across all major surveys, is that employer health benefit cost growth will remain in the 7% to 9% range through at least 2028, with the magnitude of any moderation depending heavily on GLP-1 pricing dynamics and the pace of biosimilar adoption.

Why This Matters for Actuaries

The Mercer 2026 survey is not just a benefits HR story. It documents a structural inflection in the employer health cost trajectory that directly affects actuarial work across multiple practice areas. Group benefits actuaries setting trend factors for renewals and rate filings must grapple with a price-utilization convergence that breaks conventional decomposition models. Reserving actuaries must reassess IBNR completion factors calibrated to a lower-trend environment. Stop-loss actuaries face a large-claim frequency distribution that has shifted materially in two years. And consulting actuaries advising employers on plan design must recalibrate their cost-sharing elasticity assumptions when the dominant cost drivers are insensitive to deductible levels.

This continues a trend we have been tracking since the post-pandemic utilization rebound began in late 2022. Each successive survey cycle has produced higher trend projections, and each successive cycle has proven the prior year’s projections to be roughly accurate or slightly understated. The actuarial profession’s response, across pricing, reserving, and consulting practice, should be to update assumptions aggressively rather than wait for the data to revert to historical norms that may not be coming back.