NYCIRB Files 21.9% New York Workers Comp Loss Cost Decrease for October 2026

The New York Compensation Insurance Rating Board filed a 21.9% overall decrease in workers compensation loss costs effective October 1, 2026, under Bulletin RC-2633, submitted to the New York State Department of Financial Services in June 2026. If DFS approves the filing without modification, it will mark ten consecutive annual loss cost reductions in New York, following a 13.2% decrease effective October 1, 2025 (NYCIRB, June 2026). NCCI's 2026 State of the Line frames the backdrop: a calendar-year 2025 combined ratio of 91 sits 11 points above an accident-year estimate of 102, with $14 billion in prior-year reserve redundancy doing the bridging work (NCCI, May 2026). That gap is the ratemaking question behind the headline figure.

At 21.9%, the filing is the largest single-year cut in the current downward run, nearly 9 percentage points greater than the prior year's indication. Building a ratemaking case for a decrease of that magnitude requires specific actuarial judgment calls on tail development and medical trend selection that compound each other, and the national market data provides reasons to examine both closely before adopting comparable indications from NCCI's platform in other states.

How NYCIRB Builds Loss Costs Across 600 Classification Codes

NYCIRB functions as New York's licensed rating organization, operating independently from NCCI and collecting experience data directly from member carriers under authorization granted by New York Insurance Law. The advisory loss cost for each of the board's 600-plus classification codes starts from five to seven accident years of paid and incurred loss data, organized by injury type and payroll exposure. Age-to-age development factors drawn from loss development triangles carry each accident year's losses from their current paid age to ultimate, and separate frequency and severity trend factors, applied independently for medical and indemnity components, project those ultimate losses forward to the prospective policy period. The resulting pure premium, ultimate trended losses divided by developed payroll, is the advisory loss cost filed with DFS.

Carriers do not file the advisory pure premium as their rate. Each carrier applies a loss cost multiplier that covers expenses, profit provision, and carrier-level adjustments, filing the LCM separately with DFS. The filed premium rate is the product of the advisory loss cost and the carrier's approved multiplier. This structure separates the loss estimate, which is NYCIRB's responsibility, from the expense and profit loading, which is carrier-specific. Rate adequacy depends on both components: an accurate advisory pure premium and an LCM calibrated to current expense and competitive conditions.

The 21.9% overall indication is a payroll-weighted average across all classification codes. Individual code results diverge substantially from the aggregate. Codes with sustained frequency declines and stable medical utilization receive larger decreases in the advisory filing. Construction trades and healthcare worker classifications, which carry higher medical severity exposure and emerging injury patterns, may receive smaller reductions or in certain cases increases. The aggregate conceals this variation, and pricing actuaries working specific industry segments should not apply the 21.9% uniformly across their book.

Calendar Year 91, Accident Year 102: The Development Signal

NCCI's 2026 State of the Line documents the national tension that applies with equal force to any state-level WC filing. Calendar year 2025 produced an industry-wide combined ratio of 91, ninth consecutive year below 100 and a figure that any line manager would call profitable (NCCI, May 2026). The accident year 2025 estimate, which excludes prior-year reserve releases from the calendar year result, came in at 102. NCCI chief actuary Donna Glenn noted at the 2026 Annual Issues Symposium that lost-time claim frequency declined only 2% in accident year 2025, the slowest pace in the current cycle and below the 3 to 4 percent annual average of the prior decade, citing state-specific employment mix shifts in tourism and hospitality as a contributing factor (NCCI, May 2026). Slower frequency improvement narrows the margin available to absorb a medical severity deterioration that is occurring simultaneously.

The 11-point gap between calendar-year and accident-year combined ratios is generated by $14 billion in industry reserve redundancy from prior accident years, primarily cohorts from 2013 through 2020 (NCCI, May 2026). That redundancy enters each calendar year as favorable development on prior accident years and compresses the reported combined ratio below what current-period underwriting alone would produce. The reserve cushion has been shrinking: it stood at $16 billion in 2024 and fell to $14 billion by the 2025 measurement. At that erosion rate, the development tailwind supporting calendar-year profitability has a finite lifespan, and the calendar-year combined ratio will converge upward toward the accident-year level as the cushion diminishes.

For the NYCIRB indication, the national data creates a benchmark question. Are New York's prior-year development patterns consistent with the national picture, or does New York carry additional redundancy or deficiency from its unique benefit and legislative environment that would shift the state-level accident-year result relative to the aggregate? The answer shapes whether the 21.9% overall indication reflects current accident-year adequacy or borrows against a reserve cushion that is already eroding.

Medical Severity and the Utilization Signal Above the WCWMI

Medical severity rose 4% in accident year 2025, outpacing the Workers Compensation Weighted Medical Index price change (NCCI, May 2026). The WCWMI measures fee-schedule-level prices for a basket of WC medical services; when severity growth exceeds the WCWMI, the excess reflects utilization changes rather than price. Utilization increases, particularly in inpatient services, are harder to contain and more persistent once they embed in the claim population, because they require behavioral change at the provider or adjuster level rather than regulatory intervention in a fee schedule.

WCRI's 2026 Medical Price Index for Workers Compensation quantifies the cross-state fee schedule effect across 36 study states: prices paid for professional WC services range from 28% below to 174% above the multistate median (WCRI, 2026). Non-fee-schedule states run 41% to 188% above the median. New York, which maintains a medical fee schedule, operates in the better-controlled portion of the price distribution, which means the above-WCWMI severity signal in 2025 is more likely to reflect utilization than pure price inflation. That distinction matters for trend selection: price-driven inflation has a natural ceiling in fee schedule updates, while utilization-driven growth requires a different set of management tools and tends to persist longer in the loss data before reversing.

Florida's January 2025 fee schedule overhaul, which moved from a uniform 110% of Medicare basis to a range of 110% to 175% of Medicare depending on provider type, produced a 41% jump in professional services prices in the filing year (WCRI, 2026). New York's fee schedule structure avoids that kind of step-change volatility, but the Florida data illustrates how directly fee schedule architecture shapes the medical severity trend and how quickly the WCWMI can diverge from actual paid loss experience when the underlying price basis shifts. A New York-specific medical trend selection should account for both the current fee schedule basis and any pending DFS updates that would alter it during the prospective period.

Selecting the forward medical trend for the NYCIRB indication requires separating the 2023 and 2025 data points. In 2023, medical severity ran near flat, reflecting post-pandemic normalization and deferred utilization catching up to baseline. The 2025 print shows 4% growth with a utilization signal above the WCWMI. A standard least-squares trend through five years of data blends both regimes and produces a selected trend below the most recent print. Whether that suppression is appropriate depends on whether the 2023 deceleration was a genuine structural improvement or a transitory dip, and two years of above-WCWMI severity growth argues for weighting the recent data more heavily in the selection.

Tail Factor and Medical Trend: A Three-Year Sensitivity Example

A simplified three-accident-year example shows how the overall indicated change responds to two common actuarial judgment calls: the tail factor applied to the most recent accident year and the annual medical severity trend rate selected for the prospective period.

Consider a New York WC loss base with accident years 2022, 2023, and 2024 showing losses developed to their current paid ages of $100 million, $115 million, and $125 million respectively. At a tail development factor of 1.025, the ultimates are $102.5 million, $117.9 million, and $128.1 million, for an aggregate of $348.5 million. Moving the tail factor to 1.040, a 1.5-point increase within the defensible range given that NCCI's February 2026 study found 59% of large WC claims now emerge within 24 months, shifts the ultimates to $104.0 million, $119.6 million, and $130.0 million, totaling $353.6 million. That 1.5% increase in aggregate ultimate reduces the indicated loss cost change by 1.5 percentage points before any trend adjustment is applied.

Medical trend selection compounds the effect across the projection period. At a 4.0% annual medical severity trend applied over a five-year average prospective period, the trend factor is (1.04)^5, or 1.217. At 4.5%, it is (1.045)^5, or 1.246. The difference is 2.4% applied to the projected medical loss base. If medical losses represent 55% of total WC costs, consistent with the NCCI industry composite, that 2.4% spread shifts the indicated overall change by approximately 1.3 percentage points on the total loss base.

The two adjustments combined, a 1.5-point tail factor move and a 0.5-percentage-point upward shift in the annual medical trend rate, reduce the indicated decrease by roughly 2.8 percentage points. Starting from 21.9%, that produces an indicated change closer to 19%. Neither adjustment is extreme; both fall within the range of defensible actuarial selection. The exercise is not a critique of the NYCIRB indication specifically, but a demonstration that pricing actuaries using NCCI's national platform as a benchmark in their own state filings should run analogous sensitivity analyses rather than importing the aggregate indication without adjustment. A 2.8-percentage-point swing is material in a filing cycle where rate competition has already pushed LCMs toward the lower end of historically observed ranges.

LCM Compression and the Rate Adequacy Double Exposure

The loss cost multiplier adds a second layer of rate adequacy risk that compounds the loss cost indication. In a prolonged soft market, carriers compete by filing lower LCMs against the advisory loss cost, passing margin from the loss cost indication directly to policyholders as premium reductions. After nine consecutive years of advisory loss cost decreases in New York, DFS's April 2026 release of current carrier LCMs shows the filed multiplier distribution has drifted toward the lower end of the historically observed range as the market has priced in continued favorable trends.

When the advisory loss cost decreases and the LCM compresses simultaneously, the filed premium rate declines on both dimensions at once. If the advisory loss cost subsequently proves inadequate because of a tail factor or trend selection that understated forward losses, the carrier's rate exposure is compounded: the pure premium component is deficient, and the carrier discounted that deficient base further through a compressed multiplier. Nine consecutive years of profitable calendar-year results have trained the market to expect continued rate reductions, creating institutional pressure that pushes LCMs lower even when the accident-year indicators argue for caution. These are not independent risks.

The October 2026 Filing in the Broader WC Pricing Context

For pricing actuaries reviewing WC rate filings or preparing state-level indications with the NYCIRB filing as a reference, three specific examinations address the risks identified above. First: run the medical trend selection excluding the 2020 and 2021 experience years and compare it to the full five-year fitted trend. If the recent-years trend materially exceeds the COVID-inclusive average, the filed trend may be anchored to a suppressed period that no longer characterizes the forward medical cost environment. Second: document the tail factor sensitivity across the range of defensible selections. The 1.5-percentage-point tail factor move in the example above translates to more than a one-point swing in the aggregate indication; filing actuarial support should quantify this range explicitly rather than presenting only the selected point. Third: cross-check the carrier's LCM against the accident-year adequacy picture before committing to a market-competitive filing. A calendar-year combined ratio of 91 does not support further LCM compression when the national accident-year reading is 102 and the reserve cushion driving the difference has shrunk by $2 billion in a single measurement year.

Nevada filed a 21.6% workers compensation loss cost increase for the same October 2026 filing cycle, the increase driven by SB 317's payroll cap expansion and a structurally different benefit and litigation environment (NVCIRB, June 2026). New York at negative 21.9% and Nevada at positive 21.6% represent opposite ends of a 43-point range for the same filing cycle. National aggregate profitability at a 91 combined ratio conceals that variation entirely. The actuarial work of separating genuine loss cost improvement from prior-year reserve releases and LCM compression is what distinguishes a filing that holds through the next accident year from one that requires correction by 2028.