On July 15, 2026, the New York Department of Financial Services approved a 22% reduction in workers’ compensation insurance premium rates, effective October 1, 2026 (DFS press release). DFS put the premium savings at more than $1 billion, or an average of $1,779 per policyholder; the widely quoted $1.7 billion figure adds $700 million in New York State Insurance Fund dividends and discount programs distributed over the past year. It is the sixth consecutive year of decreases, which have averaged 10.3% annually.
Read the cause carefully. DFS attributes the decline to “significant decreases in the frequency of workers’ compensation lost time claims,” helped by workplace safety efforts like the Warehouse Worker Protections Act. This is a frequency story, not a severity story. That distinction is the whole reserving problem.
Who it affects
The manual rate cut lands on standard-market buyers, but the reserving trap falls on New York self-insureds: large single-employer self-insurers, group self-insurance trusts, public entities, and captives writing New York WC exposure. These entities do not buy the manual rate, but their boards, auditors, and finance leaders see the headline. When the market rate has fallen 22% in one year and roughly 50% over six, the gravitational pull is to trim your own expected claim ratio (ECR) and case reserves “in line with the market.” That instinct is the error.
The reserve mechanism: decoupling frequency from severity
The New York Compensation Insurance Rating Board loss cost that DFS approved describes the market’s expected cost per unit of payroll, and it is falling because fewer lost-time claims are occurring. Your remaining open-claim inventory is a different animal. As frequency declines, the residual book concentrates in the claims that do not close: lifetime-medical, permanent-disability, and complex-comorbidity files whose severity is driven by medical inflation, not accident-year frequency. Those claims develop on their own schedule regardless of how many new claims the state books.
The result is a widening gap between the manual loss-cost trend and the paid development on a mature self-insured book. If you select an ECR that drifts down with the bureau rate, you under-reserve a residual book skewed toward high-severity, long-tail files. The manual rate falls faster than the severity it is supposed to describe. See Workers Compensation IBNR for Self-Insured Employers and the five core reserving methods for how ECR selection feeds the expected-loss and Bornhuetter-Ferguson calculations that this pressure distorts.
Where this shows up in your reserves
Look at the older accident years first. On your development triangle, the diagnostic is the incurred and paid movement on 2019 through 2022 New York WC, where a shrinking count of open claims should be showing stable or rising average severity even as the state’s frequency falls. If your actuary’s a priori loss ratio for recent years has stepped down in sympathy with the NYCIRB trend while your paid-loss development factors on older years have held or lengthened, that divergence is the case-reserve strengthening signal. It surfaces on Schedule P as adverse prior-year development in a period when the headline rate says costs are falling.
What this means for your next review
Put one question on the agenda: is our ECR anchored to our own paid and incurred experience, or is it drifting down with the New York manual loss-cost trend? Ask your actuary to show the severity component of your residual open claims separately from frequency, and to reconcile it against, not to, the bureau filing.
Decision-maker checklist (next 30 to 90 days)
- Confirm your reserve study selects the New York ECR from your own development, not from the NYCIRB loss-cost change.
- Ask your actuary to split the residual open-claim severity trend from frequency on 2019 to 2022 accident years.
- Flag any recent-year a priori loss ratio that stepped down more than your own paid data supports.
- Brief your board that a 22% market rate cut is a frequency signal and does not license a proportional reserve release.
- Watch NYCIRB’s next loss-cost filing for whether the medical severity component diverges from the headline rate.
Sources
- New York DFS, “Governor Hochul Announces Over $1.7 Billion in Workers’ Compensation Insurance Savings,” July 15, 2026: https://www.dfs.ny.gov/reports_and_publications/press_releases/pr20260715
- New York Compensation Insurance Rating Board loss cost filings: https://www.nycirb.org/
- NCCI 2026 State of the Line, frequency versus severity context: https://www.ncci.com/Articles/Pages/Insights-AIS2026-SOTL-Report.aspx
- Insurance Journal, “New York Approves 22% Workers’ Compensation Cost Reduction,” July 17, 2026: https://www.insurancejournal.com/news/east/2026/07/17/877978.htm