NCCI presented two demographic sessions at the 2026 Annual Issues Symposium in Orlando on May 12 and 13, the first time the organization has devoted back-to-back sessions to workforce aging. Patrick Coate, NCCI's senior economist, opened with a blunt framing: by 2030, U.S. deaths will exceed births, and population growth will depend entirely on immigration. The largest worker groups are shifting from the mid-career cohort into the 40-55 age range, with the 65+ segment growing fastest. Paul Hendrick, NCCI's senior actuary, followed with claims data showing how these demographic forces are already reshaping frequency-severity dynamics in ways that standard actuarial trend methods do not capture.
Days later, the Travelers 2026 Injury Impact Report dropped its own dataset: 1.2 million indemnity claims from 2021 through 2025. The findings confirm that the aging workforce is not a future risk; it is a current pricing problem. Workers aged 60+ average 97 missed workdays versus an 80-day overall average. A single comorbidity doubles claim cost. Two or more comorbidities push costs to five times the baseline. For pricing actuaries filing loss cost amendments, the question is no longer whether demographic severity loading is necessary, but how to quantify and defend it in a regulatory submission.
The Demographic Shift by the Numbers
The Travelers dataset breaks injury distribution by age band with a level of granularity that NCCI's aggregate data does not typically provide at the individual-claim level. Workers aged 35 to 49 account for 31% of all injuries. Workers aged 50 to 59 account for 24%. Combined, the 50+ cohort represents 41% of injured employees across all industries. NCCI's Coate confirmed that this concentration will intensify: BLS projects that 25.2% of the national workforce will be aged 55 or older by 2028, with the 65-74 cohort growing 22.4% between 2023 and 2033 and workers over 75 growing 79%.
These are not marginal shifts. From tracking NCCI rate filings across the last decade, the declining frequency trend that powered cumulative WC rate decreases of more than 20% was built on a workforce that was younger and less concentrated in high-severity age bands. Coate's frequency-by-age analysis at AIS 2026 showed that claim frequency is lowest for workers aged 25 to 34, ticks upward after age 45, and rises most sharply in the 65+ range. His summary: "Among those who stay, frequency increases," and the cohort that stays is "an increasingly important part of the labor force."
The duration data compounds the frequency signal. Travelers found that workers aged 60+ required 97 days of recovery, 17 days longer than the 80-day overall average. In construction, the average across all ages was 114 lost workdays. Transportation averaged 94 days. These extended durations inflate both indemnity severity (through additional temporary total disability payments) and medical severity (through prolonged treatment protocols, higher complication rates, and post-surgical rehabilitation).
| Age Band | Share of Injuries | Avg. Lost Workdays | Key Mechanism |
|---|---|---|---|
| 18-24 | ~14% | Below average | Inexperience; soft tissue |
| 25-34 | 19% | Below average | Lowest frequency cohort |
| 35-49 | 31% | Near average (~80) | Peak employment band |
| 50-59 | 24% | Above average | Early comorbidity onset |
| 60+ | 16% (lost-time) | 97 | Fractures, comorbidity, extended rehab |
Comorbidity Multipliers: From Clinical Risk to Actuarial Loading
The Travelers report's most actionable finding for pricing actuaries is the comorbidity cost multiplier structure derived from 1.2 million claims. A single comorbidity, such as diabetes, hypertension, or arthritis, doubles the total cost of a workers' compensation claim. Two or more comorbidities multiply the cost fivefold. Industry research corroborates the gradient: NCCI data shows that claims involving comorbid conditions have nearly tripled since 2000, and obesity alone increases claim costs by five times the baseline. Depression triples claim costs.
The prevalence of comorbidities tracks directly with age. Hypertension affects roughly half of U.S. adults aged 55 to 64 and two-thirds of those 65 and older. Type 2 diabetes prevalence exceeds 20% in the 65+ population. Arthritis, COPD, and obesity further compound the clinical picture. For a pricing actuary constructing a severity model, the interaction between age-band distribution and comorbidity prevalence creates a multiplicative effect that simple additive trend extrapolation will miss.
Consider the arithmetic. If a class code's workforce shifts from 15% aged 60+ to 22% aged 60+ over a five-year period (plausible for healthcare and construction based on BLS projections), and if the average comorbidity prevalence in that age band implies a blended severity multiplier of 2.5x versus 1.0x for the under-40 cohort, the age-mix-driven severity loading on new claims alone could exceed 8 to 12 percentage points above what a historical log-linear trend would project. That loading is not noise; it is a structural shift in the loss-generating process.
The Frequency-Severity Divergence in NCCI's 2025 Data
NCCI's 2026 State of the Line report, presented by Chief Actuary Donna Glenn at AIS, confirmed the frequency-severity divergence that demographic trends would predict. Lost-time claim frequency declined 2% in 2025, slower than the long-term average annual decline that has powered a decade of rate decreases. Medical severity rose 4%. Indemnity severity rose 4%. Both severity measures outpaced their respective price proxies: indemnity severity is growing faster than wage inflation, and medical severity is growing faster than the WC Weighted Medical Price Index.
The 2024 accident year data, now more fully developed, showed even sharper severity: medical lost-time severity up 6%, with 4 percentage points attributable to increased utilization per claim rather than unit cost inflation. Glenn noted that "price has been consistent year over year, but utilization has been more volatile." That utilization volatility is precisely what an aging, comorbid workforce produces: more complex injuries requiring more treatment visits, longer physical therapy courses, and higher rates of surgical intervention.
The calendar year 2025 combined ratio of 91%, up from 86% in 2024, remains profitable but signals a turning point. The accident year 2025 combined ratio of 102% means that current pricing is no longer adequate on a pure accident-year basis, with the calendar-year result propped up by roughly $14 billion in prior-year reserve redundancy (down from $16 billion in 2024). For pricing actuaries, this is the leading indicator: the reserve cushion is thinning at the same time that severity trends are accelerating.
Pricing Mechanism: Incorporating Age-Mix Severity Loading
The standard NCCI severity trend selection method fits an exponential or log-linear regression to 10 accident years of developed ultimate severity, separately for indemnity and medical, then projects the selected trend forward to the midpoint of the prospective policy period. This method assumes that the loss-generating process producing historical severity is stationary, or at least that any structural changes are captured in the trend line's slope.
The demographic shift breaks this assumption. The historical severity data from 2015 through 2024 reflects a workforce that was younger and healthier than the workforce that will generate claims during the 2027 or 2028 policy periods. Extrapolating the historical trend forward understates future severity because the trend line was fit to data produced by a different age-mix distribution than the one that will actually exist.
Step 1: Construct a comorbidity-adjusted severity factor
Using the Travelers multiplier structure as a starting point:
| Comorbidity Status | Severity Multiplier | Prevalence (Age 60+) | Prevalence (Age <40) |
|---|---|---|---|
| No comorbidity | 1.0x | ~25% | ~75% |
| Single comorbidity | 2.0x | ~45% | ~20% |
| Dual-plus comorbidity | 5.0x | ~30% | ~5% |
The blended severity factor for the 60+ age band is: (0.25 x 1.0) + (0.45 x 2.0) + (0.30 x 5.0) = 2.65x. For the under-40 cohort: (0.75 x 1.0) + (0.20 x 2.0) + (0.05 x 5.0) = 1.40x. The ratio of 2.65 to 1.40 gives the relative severity differential between age bands attributable to comorbidity alone: approximately 1.89x.
Step 2: Weight by projected class code age distribution
For a specific class code, obtain the current and projected age distribution from BLS labor force projections and NCCI payroll data. Weight the comorbidity-adjusted severity factor by each age band's share. Then compare the weighted severity factor for the prospective policy period against the weighted factor implicit in the historical experience period. The difference is the age-mix severity loading that standard trend extrapolation misses.
For example, if a healthcare class code's 60+ share is projected to grow from 18% to 25% over the trend projection period, while the under-40 share declines from 30% to 24%, the blended comorbidity severity factor shifts from approximately 1.72x to 1.87x, an 8.7% increase that operates on top of the selected base severity trend.
Step 3: Adjust the experience rating split-point economics
NCCI's 2024 methodology update replaced the uniform $18,500 split point with state-specific split points ranging from $9,500 (Oregon) to $38,000 (Louisiana), calibrated to produce an average D-ratio of approximately 40% across states. As average claim severity increases due to comorbidity loading, a larger share of each claim falls above the split point into the excess layer. This mechanically reduces the experience modification factor's responsiveness to individual employer loss experience, because excess losses receive less weight in the mod calculation.
The Excess Loss Factor (ELF) tables, which allocate expected losses between primary and excess layers by hazard group, were calibrated to a historical severity distribution that reflected a younger workforce. If the average WC claim cost increases 10 to 15% purely from demographic severity loading, the D-ratio shifts downward (more losses in the excess layer), and the ELF tables understate the true excess-layer loading for aging-workforce-heavy class codes in healthcare and construction.
Step 4: Apply Buhlmann credibility blending
A carrier with its own claims database can estimate a carrier-specific comorbidity severity factor from observed claim costs stratified by claimant age and comorbidity flags. The filing-ready trend selection then blends this carrier-specific estimate with the NCCI industry-wide severity trend using Buhlmann credibility:
Selected trend = Z x (carrier comorbidity-adjusted trend) + (1 - Z) x (NCCI industry trend)
Where Z is the Buhlmann credibility weight based on the carrier's volume of claims with reliable comorbidity coding. For large national carriers with robust medical bill review data, Z may approach 0.6 to 0.8. For smaller carriers relying primarily on NCCI data, Z will be lower, but even a credibility weight of 0.2 shifts the selected trend meaningfully when the carrier-specific comorbidity loading is 8 to 12 points above the industry average.
The regulatory defense rests on the transparency of the method. Regulators are accustomed to actuaries explaining trend selection deviations from the fitted line. Citing the Travelers dataset (1.2 million claims, five-year study period), the NCCI AIS 2026 demographic findings, and BLS labor force projections provides the evidentiary foundation for a severity trend selection that sits above the historical fitted slope. State regulators in jurisdictions like California, where the SAWW increased 4.99% in 2025 and the maximum TTD rate rose to $1,764 per week for 2026, are already seeing the indemnity severity acceleration firsthand.
The Bimodal Distribution Problem
Travelers' data reveals a second structural distortion that compounds the aging-workforce effect. First-year employees generate 37% of all injury claims despite representing roughly 20% of the post-pandemic workforce. In restaurants, first-year injuries account for 51% of claims. In construction, 44%. These are predominantly younger, inexperienced workers sustaining soft-tissue injuries with relatively quick return-to-work timelines.
The result is a bimodal severity distribution: high-frequency, low-severity claims from inexperienced younger workers on one side, and low-frequency, high-severity claims from comorbid older workers on the other. A single-mode severity trend model, which fits one exponential curve through the combined data, will underestimate the tail driven by the older cohort while overweighting the central tendency pulled down by the younger cohort. For class codes with high turnover and aging incumbent workforces (healthcare, construction, manufacturing), the bimodality is most pronounced and the single-mode model is least reliable.
Pricing actuaries working with class codes exhibiting this bimodal pattern should consider separate severity trend selections for primary-layer and excess-layer losses. Primary-layer severity, dominated by the high-frequency younger-worker claims, may continue to exhibit moderate trend. Excess-layer severity, concentrated in the comorbid older-worker claims that breach the split point, will trend at a materially higher rate. NCCI's Safety National data underscores this: claims at $2 million incurred have increased 91% over six years, and claims at $10 million incurred have increased 183%.
Payroll Growth Compounds the Indemnity Exposure
NCCI reported approximately 5% payroll growth in 2025, driven almost entirely by wage increases rather than employment growth. The BLS Employment Cost Index for Q1 2026 confirmed total compensation costs up 3.4% year-over-year for private industry workers. Because temporary total disability and permanent partial disability benefits are wage-linked (calculated as a fraction of the injured worker's pre-injury average weekly wage, subject to statutory minimums and maximums), wage inflation flows directly into indemnity severity independent of any change in injury characteristics or demographics.
California illustrates the compounding. The state average weekly wage increased 4.99% in 2025, pushing the 2026 maximum TTD rate to $1,764.11 per week, up from $1,680.29. For a 60+ worker with a dual comorbidity claim averaging 97 lost workdays at the maximum TTD rate, the indemnity cost alone reaches approximately $24,400 before permanent disability, medical costs, or allocated loss adjustment expenses. In 2023, the same claim at the then-current maximum TTD of roughly $1,540 per week would have cost approximately $21,300 in TTD benefits. The combination of wage growth and extended duration produces a compounding effect that accelerates faster than either factor alone.
Why This Matters for Pricing Actuaries
Workers' compensation has been the insurance industry's most profitable line for over a decade. The calendar year 2025 combined ratio of 91% marked the twelfth consecutive year below 100. But that profitability was built on a frequency decline powered by younger, healthier workforces, workplace safety improvements, and a reserve cushion that is now shrinking. The demographic shift described at NCCI AIS 2026 and quantified in the Travelers report is a permanent repricing event, not a cyclical blip.
For pricing actuaries preparing loss cost filings, the practical takeaways are clear. First, historical severity trend selections that do not account for the age-mix shift will systematically understate prospective severity by 8 to 12 percentage points in the most exposed class codes. Second, the comorbidity multiplier data from Travelers (2x for single, 5x for dual-plus) provides a defensible empirical foundation for an explicit severity loading factor. Third, the experience rating split-point mechanics create a secondary effect where rising severity reduces mod responsiveness, potentially masking deteriorating loss experience at the individual employer level. Fourth, the bimodal severity distribution in high-turnover, aging-workforce class codes requires separate trend treatment for primary and excess layers to avoid systematic mispricing.
NCCI's net written premium declined 0.2% in 2025, and expected premium reductions from approved NCCI filings average negative 5.0% for 2025 to 2026. Those rate decreases were calibrated to historical severity trends that may no longer apply. The actuary who files the next round of loss cost amendments has a choice: continue riding the historical trend line into a workforce that no longer exists, or build a severity model that accounts for the workforce that is actually generating claims. The data from NCCI AIS 2026 and Travelers makes the second path both defensible and necessary.
Further Reading on actuary.info
- NCCI 2026 State of the Line: Workers Comp Profitability Masks a Medical Severity Pivot
- NCCI State of the Line CY 2025: Combined Ratio Hits 91% as Reserve Redundancy Thins
- WC Physician Dispensing Markups of 16,000% Distort Medical Severity Trends
- Tariffs Accelerate WC Medical Equipment Cost Inflation
- Florida "Two Clocks" Ruling Rewrites Workers' Comp Statute of Limitations
Sources
- NCCI, "2026 State of the Line Guide," Annual Issues Symposium, May 12, 2026 - ncci.com
- NCCI AIS 2026, "Demographic Forces Shaping Workplace Risk," Patrick Coate, PhD - workerscompensation.com
- NCCI AIS 2026, "From Demographics to Claims: Turning Data Into Insights," Paul Hendrick - workerscompensation.com
- Travelers, "2026 Injury Impact Report," May 8, 2026 - investor.travelers.com
- Insurance Journal, "Travelers: Aging Workforce Facing Longer Recovery Times," May 8, 2026 - insurancejournal.com
- Risk & Insurance, "Workers' Compensation Remains Profitable as Premium Dips and Severity Climbs," May 2026 - riskandinsurance.com
- Bureau of Labor Statistics, "Employment Cost Index, March 2026," April 30, 2026 - bls.gov
- California Department of Industrial Relations, "2026 Temporary Total Disability Rates," December 2025 - dir.ca.gov
- NCCI, "Experience Rating Methodology FAQs" - ncci.com
- Insurance Business, "Workplace Injuries Take Longer to Heal, Report Finds," May 2026 - insurancebusinessmag.com