Traffic Deaths Dropped 6.7%. Thank the Scrapyard, Not the Senate.
NHTSA posted the 2025 numbers last month and took a victory lap: 36,640 traffic fatalities, down 6.7% from 2024, with the rate dropping to 1.06 deaths per 100 million VMT, the lowest since 2014.[1] The first half of the year was even more dramatic, with an 8.2% decline that NHTSA called the sharpest drop since 2008.[2] Politicians will claim credit, safety advocates will point to their campaigns, and both are wrong.
We cross-tabulated ten years of FARS model-year death data against national fleet attrition estimates, and the pattern is unambiguous.[3] Nearly two-thirds of all fatal-crash vehicles during 2014 through 2023 carried model years of 2008 or earlier. These are pre-Electronic Stability Control mandate vehicles, built before the 2012 federal requirement took effect, designed with crash structures that IIHS would politely describe as "inadequate by current standards" and that I would describe as criminal negligence frozen in sheet metal.[4]
They are disappearing, not because of regulation, but because metal fatigues, transmissions fail, and insurance companies total them out after fender benders that a newer car would shrug off. The average American vehicle lasts about 12.6 years. A 2008 model hit that expiration date in 2020, and a 2005 hit it in 2017. Each one that goes to the crusher removes a statistical death trap from the road and replaces it with nothing, or with a 2022 Civic that has ten airbags, automatic emergency braking, and a crumple zone architecture that would make a 2004 Camry weep with envy if Camrys could feel shame.
The engineering delta is staggering: vehicles from model years 2018 through 2022 produce 70.1% fewer fatalities than those from 2005 through 2009, after controlling for fleet size.[3] Body type splits the gap further: sedans involved in fatal crashes kill their occupants 64.5% of the time, while SUVs manage 52.4% and pickups 48.9%, a hierarchy that reflects mass, ride height, and a generation of structural improvements that quietly outperformed every legislative initiative of the same period. IIHS found that improved vehicle designs alone cut driver death rates by more than a third in just three model years, and nine current models achieved a zero recorded death rate in their latest analysis.[4]
About that ESC mandate. Congress required electronic stability control on all new vehicles starting with model year 2012, and NHTSA projected it would save thousands of lives annually.[5] Our FARS data shows the mandate-era cohort (2013 through 2016) killed only 1.9% fewer people than the pre-mandate cohort (2008 through 2011). That figure almost certainly understates the technology's real impact because many automakers adopted ESC voluntarily before the mandate, which means the regulatory requirement captured a smaller incremental population than the headline numbers implied. IIHS independently estimated ESC reduces single-vehicle fatal crash risk by 56% and fatal rollover risk by 77%.[6] The mandate worked. It just worked mostly before it became mandatory.
The strongest case against this thesis is behavioral, and it has real teeth. Post-COVID commute normalization reduced exposure: vehicle miles traveled grew more slowly than GDP in 2024 and 2025, meaning Americans drove proportionally less. Risk-taking behavior may have regressed toward pre-pandemic means. Economic sorting confounds the fleet-age analysis, because the people driving 2020-model-year vehicles are systematically wealthier, live in safer zip codes, and have lower baseline crash risk than the people clinging to a 2006 Altima with 190,000 miles. Our 62.3% figure spans 2014 through 2023 crash data; extrapolating it to explain a 2025 trend introduces temporal uncertainty that honest analysis cannot ignore.
But even granting all of that, the mechanical relationship between fleet composition and fatality counts is not something behavior alone can explain away, because you cannot crash a car that no longer exists, and every scrapped pre-ESC vehicle removes a specific, quantifiable risk from the denominator permanently. The policy implication is uncomfortable for anyone who wants a legislative hero in this story. Accelerated vehicle retirement programs, Cash for Clunkers descendants with safety-targeted criteria rather than emissions criteria, would likely save more lives per dollar than any ADAS mandate currently under consideration.
What this means for you: If you are driving a vehicle from 2008 or earlier, you are in the cohort that produces 62% of fatal crash deaths. Check your VIN at nhtsa.gov/recalls for any open safety recalls. If you are shopping used, prioritize 2015 or newer to ensure ESC is standard, not optional. Sedans kill their occupants at a rate 23% higher than comparably sized SUVs in fatal crashes, so if safety is your priority and budget forces you into older inventory, body type matters more than brand. And if your state legislature is debating vehicle inspection requirements, remind them that the scrapyard is currently doing more for traffic safety than any bill they have passed this decade.
Sources & References
- NHTSA, 2025 Traffic Death Estimates & 2024 FARS Annual Report. 36,640 fatalities in 2025 (6.7% decline), 1.06 per 100M VMT. nhtsa.gov
- NHTSA, Traffic Crash Deaths Early Estimates, January–June 2025. 17,140 fatalities (8.2% decline, largest since 2008). nhtsa.gov
- NHTSA, Fatality Analysis Reporting System (FARS), 2014–2023. Model-year death distributions, body-type lethality ratios, and fleet-size estimates. nhtsa.gov
- IIHS, Improved vehicle designs bring down death rates. Driver death rates reduced by >1/3 in three model years; 9 models achieved zero death rate. iihs.org
- NHTSA, Federal Motor Vehicle Safety Standard No. 126: Electronic Stability Control, Final Rule (2007). ESC required on all new light vehicles from MY 2012. govinfo.gov
- IIHS, Life-saving benefits of ESC continue to accrue. ESC reduces single-vehicle fatal crash risk by 56%, fatal rollover risk by 77%. iihs.org
Source: NHTSA FARS 2014–2023, NHTSA 2025 Early Estimates, IIHS vehicle design and ESC studies. The 62.3% old-vehicle fatality share is computed from FARS model-year death distributions across 2014–2023 crash years; extrapolation to 2025 trends assumes similar fleet-age composition. Body-type lethality ratios are ecological (per-crash, not per-VMT). The −1.9% ESC mandate figure likely understates true ESC impact due to pre-mandate voluntary adoption. Fleet attrition estimates use 12.6-year average vehicle lifespan. See methodology for caveats.