Evidence quality 4.75/5
Eight-dimension review score against the quality rubric . Each dimension scored 1–5.
- D1 Source grounding
- 4/5
- D2 Source authority
- 5/5
- D3 Arithmetic
- 5/5
- D4 Uncertainty
- 4/5
- D5 Scope
- 5/5
- D6 Prose
- 5/5
- D7 Perception honesty
- 5/5
- D8 Caveat completeness
- 5/5
● your factors — click this risk ▾ to reveal
- Your factors
≈ As likely as
Perceived
Most drivers treat a 10-15 mph overage as a minor infraction, not a meaningful safety decision. Public understanding of crash physics is weak: surveys and focus groups consistently find that drivers underestimate how steeply crash severity scales with speed. Kinetic energy grows with the square of velocity, and the probability of a fatal outcome at impact grows even faster. Drivers who routinely exceed the limit by 20% tend to report that they feel in control and that their driving is only marginally riskier than keeping to the posted speed — the actual multiplier, around 2x on fatality risk, is far larger than most would guess.
Rough estimate: most drivers underestimate the risk multiplier; gut feel is often 10–20% elevated risk, not 2x
Source: editorial intuition, not polled
Actual
~1 fatal crash per ~3,850 driver-years at 20% over the speed limit
US driver travelling at 20% over posted speed limit. Baseline one-year motor-vehicle-crash death odds ≈ 1 in 7,979 (NSC lifetime odds 1 in 101 × 79.0-year life expectancy, per NSC's stated method) × 2.07 Power-Model factor → 1 in ~3,855, shown as ~3,850.
Show derivation
Step 1 — Baseline lifetime motor-vehicle-crash fatality: the National Safety Council's "Odds of Dying" (2024 data) publishes US lifetime odds of dying in a motor-vehicle crash of 1 in 101 ≈ 0.0099. (NSC derives this by dividing the one-year odds by the 79.0-year life expectancy of a person born in 2024.) Step 2 — Power Model multiplier: Nilsson (1981/2004) gives the fatality multiplier as (v₂/v₁)⁴. At v₂/v₁ = 1.2 the multiplier is 1.2⁴ = 2.0736 ≈ 2.07. Elvik's 2013 meta-analysis of 98 studies confirms this exponent; at highway speeds the empirical exponent may exceed 4, so 2.07 is a conservative lower bound. Step 3 — Lifetime estimate: A driver who consistently travels at 1.2× the limit for a significant fraction of highway miles applies roughly the full 2.07× to their baseline risk: 0.0099 × 2.07 ≈ 0.0205, rounded to 0.020 as the central estimate. The display "1 in 50" = 1 / 0.020. Uncertainty band lower bound (0.012) corresponds to a ~1.2× exposure-weighted multiplier (half of miles at limit, half at 1.2×): 0.0099 × 1.2 ≈ 0.012. Upper bound (0.023) corresponds to 2.3× (highway driving mostly at 25% over limit, exponent ~4.5): 0.0099 × 2.3 ≈ 0.023.
Caveats: The Power Model (v₂/v₁)⁴ gives a point estimate for the per-trip fatality multip…
The Power Model (v₂/v₁)⁴ gives a point estimate for the per-trip fatality multiplier at a given speed ratio. It does not account for road type (the exponent is lower on urban arterials, higher on rural two-lanes), traffic density, vehicle safety ratings, or driver skill. The 2.07× figure is a central estimate for rural/highway conditions; it understates risk on undivided rural roads and overstates it in dense urban traffic where absolute speeds are low. NHTSA's speeding classification uses "any speeding-related factor" in the crash report — it includes drivers cited for going 1 mph over as well as those going 50 mph over — so the 29% share of fatalities does not mean the average speeding driver faced a 2× elevated risk; it means speeding was a coded contributing factor across a very wide range of magnitudes. The lifetime estimate on this page applies only to drivers who regularly travel 20%+ over the limit for a meaningful fraction of their miles; it is not a population-average figure.
Related risks
Other risks on similar themes — for exploring related fears.
Video watching + driving
What are the odds of a crash from watching video on a phone while driving?
Driving on sedating meds
What are the odds of causing a fatal crash by driving while on a sedating prescription medication?
Pick challenger
The mechanism is physics before it is statistics. Kinetic energy scales with the square of velocity, and the probability that a crash is fatal scales even faster — the Power Model developed by Göran Nilsson at Sweden’s VTI research institute and confirmed by Rune Elvik’s 2004 meta-analysis of 98 studies gives the fatality multiplier as (v₂/v₁)⁴. Drive 20% over the posted limit and you are not running 20% more risk of a fatal crash; you are running roughly 2.07 times the risk of someone travelling at the limit. Drive 30% over and the multiplier climbs to 2.86×.
NHTSA’s Fatality Analysis Reporting System counted 11,775 deaths in speeding-related crashes in 2023, 29 percent of the 40,901 total US traffic fatalities that year. That share has been roughly stable for two decades, which means the absolute number tracks overall fatality trends but the behavioral pattern has not meaningfully improved. The 29% figure spans a wide range of severity: NHTSA codes a crash as speeding-related whenever an officer checks that box, whether the driver was 1 mph or 50 mph over the limit, so the share does not mean the typical speeding driver faced anything close to the 2× elevated risk estimated above. The Power Model exponent for fatal crashes on high-speed rural roads may also exceed 4, as Elvik’s 2013 re-parameterisation found the empirical exponent varies with initial speed and road type, generally running higher at highway speeds than at urban arterial speeds.
The gap between perceived and actual risk here is structural. Drivers who routinely travel 10-15 mph over the interstate limit experience no immediate consequence most of the time, which trains an intuitive model that the overage is tolerable. The physics disagrees on what happens in the relatively rare event that something goes wrong: at 75 mph versus 60 mph, a collision absorbs 56% more kinetic energy, and the margin between a serious injury and a fatality is largely determined by that energy budget. The lifetime estimate for a driver who regularly exceeds the limit by 20% on highway miles works out to roughly 1 in 50 — approximately double the US population baseline of 1 in 101 (the National Safety Council’s 2024-data lifetime odds of dying in a motor-vehicle crash) — with the uncertainty band wide enough to accommodate both cautious interpretations and higher-exponent road types.
Related tidbits
A driver who routinely runs 20%+ over the limit faces roughly ~1 in 47 lifetime odds of a fatal crash, about 1 fatal crash per ~220,000 trips at 20% over. Gut feel tends to register a 10 to 20% bump in risk rather than a doubling.
Claim ledger
Every number below is what each source reported, with the verbatim quote we relied on and how we arrived at our figure. Click any link to verify directly.
-
[1] Nilsson, G. — Swedish National Road and Transport Research Institute (VTI) / Transport Research Institute (TØI) — The Power Model of the relationship between speed and road safety
The Power Model of the relationship between speed and road safety- Statistic
Fatal crash frequency scales as (v₂/v₁)⁴ relative to a baseline speed; a 20% speed increase → 1.2⁴ ≈ 2.07× fatality risk; a 10% increase → 1.1⁴ ≈ 1.46×- Excerpt
“"The number of fatal accidents is proportional to the fourth power of the ratio between the new and old mean speed of traffic." ”
- Source data from
- 2004-01-01
- Accessed
- 2026-05-04 · archived copy
- Calculation
- Nilsson's original 1981 study established the power relationships; this 2004 VTI/TØI report is the canonical summary and reference document. The exponent of 4 for fatal crashes is the central estimate; Elvik's 2004 meta-analysis of 98 studies (460 estimates) broadly confirms this value across road types. Used here as the primary mechanism linking speed to fatality risk.
-
[2] National Highway Traffic Safety Administration (NHTSA), National Center for Statistics and Analysis — Traffic Safety Facts 2023 Data: Speeding (DOT HS 813 721)
Traffic Safety Facts 2023 Data: Speeding (DOT HS 813 721)- Statistic
11,775 fatalities in speeding-related crashes in 2023, representing 29% of all 40,901 traffic fatalities; decrease of 3% from 12,157 in 2022; estimated 332,598 people injured in speeding-related crashes- Excerpt
“"In 2023, there were 11,775 fatalities in speeding-related crashes, which represents 29 percent of the total number of traffic fatalities for the year." ”
- Source data from
- 2024-11-01
- Accessed
- 2026-05-04 · archived copy
- Calculation
- NHTSA FARS is the authoritative US count for speeding-related fatalities. The 29% share (11,775 of 40,901) establishes that speeding is the second largest behavioral factor in US traffic deaths. This source supplies the annual US fatality count and speeding share; the lifetime population baseline that the Power Model multiplier is applied to comes from NSC's stated lifetime odds (1 in 101), not from a hazard recomputed here.
-
[3] Elvik, R. — Accident Analysis and Prevention (Elsevier) — A re-parameterisation of the Power Model of the relationship between the speed of traffic and the number of accidents and accident victims
A re-parameterisation of the Power Model of the relationship between the speed of traffic and the number of accidents and accident victims- Statistic
Meta-analysis of 98 studies with 460 estimates confirms the Power Model; exponent for fatal accidents empirically estimated at approximately 4 (range ~3–6 depending on road type and initial speed)- Excerpt
“"The Power Model of the relationship between speed and road safety proposes that changes in the mean speed of traffic are associated with changes in the number of accidents and accident victims according to power functions." ”
- Source data from
- 2013-01-01
- Accessed
- 2026-05-04 · archived copy
- Calculation
- Elvik's 2013 re-parameterisation (published in Accident Analysis & Prevention) is the peer-reviewed confirmation of Nilsson's empirically derived exponent. The update shows the exponent varies with initial speed — at highway speeds the fatality exponent is closer to 4–5, strengthening the case that 20% overspeed at 60 mph is more dangerous than the simple 2.07× central estimate implies. Used here to corroborate the Power Model as the consensus scientific framework.
-
[4] National Safety Council (NSC), Injury Facts — Odds of Dying (2024 Data) — Motor-vehicle crash
Odds of Dying (2024 Data) — Motor-vehicle crash- Statistic
Lifetime odds of a US resident dying in a motor-vehicle crash: 1 in 101 (2024 data). Lifetime odds are derived by dividing one-year odds by the 79.0-year life expectancy of a person born in 2024.- Excerpt
“"Motor-vehicle crash … 1 in 101." (Lifetime odds of death for selected causes, United States, 2024.) "The lifetime odds are approximated by dividing the one-year odds by the life expectancy of a person born in 2024 (79.0 years)." ”
- Source data from
- 2026-01-01
- Accessed
- 2026-07-10 · archived copy
- Calculation
- NSC Injury Facts is the standard US reference for population-average lifetime odds of death by external cause; it computes odds from National Center for Health Statistics mortality data and Census population/life-expectancy figures. The 1-in-101 (≈0.0099) lifetime motor-vehicle-crash figure is the population baseline this entry multiplies by the Power Model factor. (An earlier NSC release, based on 2022 data, gave 1 in 93 ≈ 0.0108; the 2024-data figure is 1 in 101 and is the value used here.) Neither this baseline nor the "1 in 93" predecessor is an IIHS statistic — the prior IIHS attribution was an error.