NHTSA’s April 2026 release puts 2025 traffic fatalities at an estimated 36,640, a 6.7% decline from 2024’s 39,254 and the second-lowest fatality rate in recorded history at 1.10 per 100 million vehicle miles traveled. In any other line of coverage, that would be unambiguous good news.
It isn’t. Marathon Strategies’ 2025 nuclear verdict census counted 135 verdicts exceeding $10 million in 2024, a 52% increase over 2023. Total awards reached $31.3 billion, up 116%. The median nuclear verdict climbed to $51 million, more than double the $21 million median in 2020. “Thermonuclear” verdicts exceeding $100 million nearly doubled to 49, and five crossed the $1 billion threshold.
Who It Affects
Self-insured trucking fleets, delivery operations, public transit agencies, and any employer retaining commercial auto bodily injury risk. The frequency-severity split matters most for programs that use combined loss-rate trending, where the same trend factor covers both claim counts and average costs. Those programs will see falling frequency pulling projections downward while severity quietly fattens the right tail.
The Reserve Mechanism
This is a frequency-severity divergence problem. Crash frequency is declining. Verdict severity is accelerating. When frequency and severity moved in the same direction, chain-ladder development factors absorbed both trends implicitly. That assumption breaks when the two components move in opposite directions.
A combined loss-rate approach (incurred losses divided by exposure) will show moderate growth or even stability, because falling claim counts partially offset rising per-claim costs. The danger is that development factors calibrated on that blended history understate the tail. One $51 million verdict in a triangle of declining claim counts produces far more adverse development than the historical pattern would predict.
Marathon Strategies identified eight trucking-specific nuclear verdicts totaling $790.5 million in 2024. Swiss Re’s behavioral study adds context: 76% of U.S. survey respondents now view jury damage awards as too low or about right, up from 58% in 2016. Only 56% believe there are too many lawsuits, down from 90% a decade ago. The shift is generational; 83% of respondents under 40 hold plaintiff-friendly views on damages. Swiss Re also found that injury severity, not company size, drives verdict amounts in jury simulations, which means smaller self-insured fleets face nearly the same exposure as large carriers.
What This Means for Your Next Review
Ask your actuary whether your auto liability reserve analysis separates frequency and severity trend assumptions or relies on a combined loss-rate approach. If the latter, request a sensitivity test that applies a severity trend load reflecting the post-2020 median verdict trajectory. For programs with retained commercial auto bodily injury exposure, confirm that development factor selections for the most recent accident years reflect the current verdict environment, not the blended experience from years when frequency and severity moved together.
What to Watch Next
NHTSA’s detailed 2024 FARS release will include vehicle-type breakdowns for large trucks and commercial vehicles. Those numbers will show whether the commercial vehicle subset mirrors the overall fatality decline or diverges, a critical input for fleet-specific frequency assumptions.