Overview
The Retirement Plans Experience Committee (RPEC) of the Society of Actuaries Research Institute examines recent U.S. population mortality experience to inform mortality improvement assumptions for retirement and post-retirement benefit valuations. The 2026 report finds that mortality at ages 65 and older has generally returned to levels expected by projecting 2019 mortality rates using MP-2021.
RPEC will not release a new mortality improvement scale in 2026. The committee continues to view MP-2021 as a reasonable best-estimate scale for retirement plan actuarial valuations, especially at ages 65 and older. Additional post-pandemic experience is needed before a replacement scale can be developed.
Key Findings
- Recent mortality is slightly below MP-2021 projections. For July 2025 through June 2026, the age-standardized actual-to-expected (A/E) mortality ratio for ages 65 and older was 98.7% for males and females combined, compared with 100.8% for calendar year 2025. Expected mortality reflects 2019 population mortality projected using MP-2021. The latest result no longer indicates excess mortality relative to that benchmark.
- Results differ by sex and age. For July 2025 through June 2026, A/E ratios at ages 65 and older were 97.8% for males and 99.5% for females. Younger-age experience differs more substantially from MP-2021 projections, although younger-age mortality assumptions generally have less significance for typical pension and retiree welfare obligations.
- Evidence of a sustained change in the improvement trend remains inconclusive. Analyses excluding deaths with COVID-19 as the underlying cause produced estimated annual improvement rates at ages 65 and older of approximately 0.5% over three overlapping periods: 2010–2019, 2010–2025, and 2010–2025 excluding 2020–2023. Some shorter periods indicate higher improvement, but recent estimates are less precise. The available data do not provide clear evidence either for or against a material change in the underlying trend.
Prior Reports
RPEC 2025 Mortality Improvement Update
RPEC 2024 Mortality Improvement Update
RPEC 2023 Mortality Improvement Update
RPEC 2022 Mortality Improvement Update
FAQs
Under RPEC’s traditional methodology, at least five years of post-pandemic experience would be needed. If that period begins in 2024, experience through 2028 would be required; the reporting lag for Medicare data means an updated scale would likely not be available before fall 2030 using preliminary 2028 data. This is conditional, not a scheduled release date. RPEC is considering whether alternative methods could support an earlier update using three or four years of post-pandemic experience, and whether the period considered to be “post-pandemic” should begin in 2023 or 2024.
Actuaries may consider plan-specific adjustments, recognizing that a pension population’s experience can differ from that of the general U.S. population. The report provides A/E results by year and sex to inform consideration of mortality loads but does not conclude that modified assumptions are inherently better predictors of future U.S. population mortality. RPEC does not yet consider the post-pandemic data sufficient to develop a new scale with greater predictive power than MP-2021 for the general U.S. population ages 65 and older.
RPEC recognizes that increased use of GLP-1 drugs may have contributed to recent mortality reductions, but the available data do not yet support isolating or quantifying that contribution. The committee considers it premature to adjust long-term mortality improvement assumptions for pension and retiree welfare valuations based on GLP-1 mortality studies alone.
The analysis uses CDC WONDER death counts through June 2026 and Social Security Area population estimates from the 2026 OASDI Trustees Report. Recent death counts are adjusted for estimated reporting completeness. Results are standardized using the 2019 age and sex distribution to remove the effects of changes in population composition.
A/E ratios are estimates, not exact measurements. Population estimates are revised over time, with greater uncertainty in recent years and at advanced ages. Medicare data, which link deaths and exposures at the beneficiary level, underpin RPEC’s traditional scale development but have a reporting lag of more than two years.
The full report provides detailed results, methodological explanations, considerations for alternative assumptions, and context from the SSA and actuarial organizations in other countries. Its conclusions are those of the authors and do not represent an official position of the SOA Research Institute, the SOA, or its members.