Several databases of potential interest to insurance companies were affected by federal policy and staffing changes in 2025. The most widely reported in insurance trades was the National Oceanic and Atmospheric Administration’s (NOAA) billion-dollar weather and climate disasters database, which NOAA announced it would no longer update beyond 2024 (although archives would remain available). AM Best warned that this could adversely impact insurers’ ability to price “secondary” perils such as severe convective storms and wildfires.
NOAA isn’t the only agency facing problems; the National Weather Service (NWS) reportedly lost more than 500 employees through layoffs or early retirement incentives, resulting in reduced overnight forecasting and fewer weather balloon launches. Amid concern around the potential for this to impact disaster preparedness and forecast accuracy, NWS began rehiring in June. Datasets most typically used for catastrophe bond triggers, such as the U.S. Geological Survey earthquake report and National Hurricane Center tropical cyclone reports, do not appear to be affected. Private firms are expected to address some of the other data-related challenges. It will take time, however, for them to build sufficiently robust datasets and to establish the trust of interested stakeholders on the level of NOAA.
It’s not just meteorological and climatological data that’s being affected; in April the Bureau of Labor Statistics reportedly pared back the extent of data collection used to produce the consumer price index (CPI) and discontinued approximately 350 producer price indices (PPI). Federal health agencies also reportedly removed several datasets and webpages following January executive orders pertaining to DEI and gender. Lawsuits helped restore access to some of the information in question. Amid the volatility, some concerned organizations — mostly nonprofits — are painstakingly laboring to mitigate potential data loss by capturing and archiving vulnerable government data via web scraping and other means.
What this means for actuaries:
When it comes to frequently used datasets, it can be easy to take for granted certain considerations in Actuarial Standard of Practice (ASOP) No. 23, such as internal consistency (3.2.b.2), sampling methods (3.2.b.7), and significant known limitations (3.2.b.5) including sufficient currentness (3.2.b.1). Because of the uncertainty surrounding the availability and continuity of government data, actuaries should brush up on ASOP No. 23 — including its guidance regarding availability of alternative data sources (3.2.b.6). This may be particularly timely for econometrics such as CPI or PPI, which likely hold value in assessing loss cost impacts of recent economic policy changes such as tariffs.
Beyond firsthand use of data, actuaries may also consider whether vendor data sources are reliant on any of the affected datasets. Although recent developments appear so far to have minimal impacts on reinsurance and catastrophe bonds, were this to change, actuaries would likely play some role in reengineering these instruments so they can continue functioning. ●
Sources:
- https://apnews.com/article/cdc-fda-websites-restored-ad1ec005261b31986ddcd3d1d17b249d.
- https://www.artemis.bm/news/parametric-cat-bond-triggers-may-face-noaa-disaster-database-wind-down-uncertainty-am-best/.
- https://www.bls.gov/ppi/notices/2025/bls-to-discontinue-selected-ppis.htm.
- https://www.cnn.com/2025/06/02/weather/nws-rehiring-meteorologists.
- https://www.insurancebusinessmag.com/reinsurance/news/breaking-news/noaa-climate-dataset-halt-raises-reinsurance-concerns-536876.aspx.
- https://www.msn.com/en-us/weather/topstories/doge-cuts-to-weather-balloon-sites-leave-u-s-without-crucial-data-meteorologists/ar-AA1Fntoz.
- https://www.nytimes.com/2025/01/31/health/trump-cdc-dei-gender.html.
- https://www.reuters.com/business/us-inflation-data-collection-hurt-by-trump-era-hiring-freeze-wsj-says-2025-06-04/.
- https://www.technologyreview.com/2025/02/07/1111328/inside-the-race-to-archive-the-us-governments-websites/.
