More enforcement. Lower limits. Where is the road safety revolution?
Britain's roads became dramatically safer over the past four decades, with a substantial decline in people killed or seriously injured (KSI). Yet that decline slowed after around 2010. How does the trend compare with the expansion of speeding enforcement and lower speed limits?
1. Killed or seriously injured (KSI)
The complete annual unadjusted KSI series from 1980 to 2024, showing the long-term fall and subsequent slowing. Serious-injury recording methods changed in some forces from 2016, so the later values are not directly comparable without adjustment.
Road casualties killed or seriously injured · Great Britain · 1980–2024
Why might enforcement have so little effect? The fractions of road safety
The first chart shows large changes in enforcement and detected offences without a corresponding improvement in the KSI trend. The fractions below suggest one reason why: only part of the casualty problem involves speed, and enforcement choices determine which drivers and offences are targeted.
Under the former police contributory-factor system, approximately 15% of KSI casualties were involved in collisions where exceeding the speed limit or travelling too fast for conditions was recorded. That was the combined occurrence of either factor, not 15% for each. A recorded factor is not necessarily the sole or primary cause. The potential safety benefit of a particular intervention cannot be calculated by subtracting overlapping factors from 15%.
Enforcement choices shape which drivers are detected. Someone travelling at 36 mph in a new 30 limit might have driven at precisely the same speed legally when the limit was 50. Concentrating enforcement on such marginal breaches, where no speed-related casualty problem has been established, may produce many offences but relatively little casualty reduction.
Explanation 3 ↓1. Road deaths: every year from 1980 to 2025
2. Risk per mile travelled
3. All reported casualties
4. Speeding offences and penalties
Explanation 4 ↓2. Speed awareness courses
National Speed Awareness Course attendance · 2010–2025
Explanation 5 ↓Chart explanation 9 ↓3. Enforcement over three decades — together
This comparison places historical speeding fixed penalty notices alongside modern recorded offences and speed awareness courses. The lines represent different measures, not a single uninterrupted count. Historical penalties are FPNs; the new modern penalty line counts fines paid or registered. Modern recorded offences include cases resolved through a course, a penalty or court action. Courses should not be added to offences.
Speed enforcement · England and Wales (courses: UK)
4. Penalties plus speed awareness courses
This is the comparison that explains why the early-2000s penalty peak can be deceptive. A course is normally an alternative to a penalty, so looking at penalties alone understates the number of drivers dealt with today. The modern bars add fines paid or registered to National Speed Awareness Course completions. They are an approximate measure of disposals, not a count of offences detected: the datasets differ in geography and the year of completion can differ from the year of detection.
Speeding penalties and speed awareness courses · 1986–2025
Chart explanation 11 ↓Evidence and objectivity in road-safety decisions
Local campaigners and elected representatives can identify genuine concerns. But road-safety decisions affect the safety, time and money of large numbers of people, and proposals should be assessed against evidence rather than personal conviction alone.
That means examining collision histories, actual vehicle speeds, road design, vulnerable road users, expected casualty reductions, possible unintended consequences and alternative measures. Where a proposal is challenged, decision-makers should explain the evidence and reasoning rather than dismissing questions about effectiveness or attributing motives to critics.
The test of a road-safety measure is not how many offences it detects, but whether it demonstrably reduces harm.
The hidden cost to motorists
Fines and course fees are only the visible part of the cost of speed enforcement. Drivers may also spend time arranging and attending courses, handling correspondence, travelling, taking time off work and dealing with uncertainty. Some also face insurance consequences. These burdens deserve to be measured alongside any road-safety benefits.
How many years of people's time?
In 2025, the National Speed Awareness Course series above records 1,853,250 attendances. Adjust the assumptions below to see the scale of the time commitment. This is an illustration, not an official economic estimate.
The cost of reducing speed limits
A lower limit can entail consultation, traffic regulation orders, signage, publicity, engineering and enforcement. There may also be changes in journey times, vehicle emissions, noise and collision risk. A complete assessment should account for both costs and benefits, including effects on pedestrians and residents.
What does a few extra minutes add up to?
A small delay repeated across a large number of journeys can have a substantial aggregate cost. The calculator starts with an illustrative 50,000 journeys per day and 20 extra minutes each; adjust the figures to examine other scenarios. These defaults do not represent a measured delay on any particular road.
The test for councils: publish the predicted safety benefit, implementation cost and expected journey-time effect before a change, then measure what actually happened afterwards.
The opportunity cost: what else could be funded?
Money and staff time committed to one intervention cannot be used simultaneously for another. Cameras, courses, lower-limit schemes and their administration should be evaluated against alternatives such as visible roads policing, action against impaired or distracted driving, safer junction design and maintenance of dangerous road layouts.
That does not mean every fine or course fee could be redirected to police recruitment: budgets, operating costs and funding rules differ. Nor does it prove that more traffic officers would necessarily outperform cameras. The useful comparison is the cost per collision or casualty prevented, supported by evidence for each alternative.
Do speed awareness courses prevent collisions?
The Department for Transport's 2018 National Speed Awareness Course impact evaluation found evidence of reduced repeat speeding offences among course attendees. That is a different outcome from preventing collisions and injuries. The evaluation did not establish a statistically significant reduction in injury collisions attributable to attending the course.
Absence of statistically significant evidence is not proof of no benefit. Collision outcomes are comparatively rare and difficult to evaluate; later research should also be considered before claiming the 2018 study is definitive. If the public purpose is casualty reduction, the key measure should ultimately be collisions and casualties prevented, not simply attendance or reoffending.
Explanation 9 ↓What the evidence does—and doesn't—show
These charts establish long-term changes in reported KSI casualties and recorded speeding enforcement. They do not establish that enforcement caused the post-2010 slowdown, nor what would have happened without enforcement. Falling casualties also do not prove that faster acceleration is risk-free.
National counts of speed-limit reductions and a reliable historical percentage of sub-five-second cars remain to be assembled. They should not be represented as measured annual series without supporting data.
Explanations and methodology
The notes below explain the definitions, assumptions and limitations behind the charts and illustrations.
1. Supporting note
An apparent inverse relationship in parts of an aggregate time series is not evidence of a causal inverse relationship. The chart compares different geographic coverage and changing enforcement definitions; it should prompt investigation, not be treated as a controlled evaluation.
2. Fractions chart
How to read this: 100% and the rounded historical 15% are data; the three smaller coloured bars are conceptual only, not estimated percentages. The fifth stage examines a particular enforcement practice rather than a mathematically proven subset. Factors can overlap: drink or drug impairment, for example, may contribute to speeding, but that does not mean speed could not also affect collision severity or prevention.
3. Statistical definitions
4. Supporting note
5. Supporting note
6. Supporting note
2024 illustration: 995,656 paid/registered speeding fines plus 1,650,322 completed National Speed Awareness Courses = 2,645,978 outcomes in these two categories. This is not directly comparable with the 2004 fixed-penalty count of 1,924,400 because early course numbers and counting conventions differ. It also excludes court outcomes.
7. Supporting note
Based on the 2025 NSAC attendance count only; excludes any Dorset-specific courses outside the national scheme, other course types, travel, administration, fees and fines. Course attendance is not necessarily lost productive work time. Time valuation is illustrative, not a direct cash loss. Fee payments are transfers, so they should not simply be added to the economic cost of time as if all were resources destroyed.
8. Journey-time illustration
This is a sensitivity calculation, not a forecast for any actual road. It assumes every counted journey incurs the stated delay, with no changes to traffic, routes or congestion. Vehicle-hours are not necessarily person-hours: occupancy matters. Proper appraisal requires measured speeds and traffic volumes, collision data, and the applicable transport-appraisal method.
9. Supporting note
Further source verification is needed before publication for later course-effectiveness studies, unit costs of alternative interventions and any aggregate estimates of the costs of speed-limit schemes.
Chart note 1 — Road casualties killed or seriously injured · Great Britain · 1980–2024
The colour-coded shaded periods correspond to the five event cards above. Overlapping shades indicate overlapping periods, not measured policy intensity or causation. The 2020–2025 band illustrates the recent spread of lower speed limits. No complete GB-wide annual mileage series is available here, so this annotation does not quantify the scale or establish a causal relationship with KSI or enforcement. Orange overlay: historical speeding fixed penalties (1986–2009), then fines + NSAC completions (2010–2025); amber dashed portions mark estimated fines for 2010 and 2025. Right-hand scale in millions. Earlier historical penalties exclude unknown course totals; figures for 1986–1996 come from historical parliamentary answers and may differ in notice/payment definitions; the line therefore compares differently defined measures. Fixed penalties for speeding began in 1986; earlier court cases are not plotted. The orange line begins near zero because only 69,639 speeding fixed penalties are shown for 1986 on a scale reaching roughly 3.3 million; this does not mean that speeding enforcement or prosecutions were virtually absent.
Chart note 2 — Road casualties killed or seriously injured · Great Britain · 1980–2024
Source: Department for Transport STATS19 historical KSI series, as compiled by Road Crash UK. These are unadjusted recorded counts. The DfT's adjusted 2025 KSI figure is 29,918, but is deliberately not connected to this unadjusted series.
Chart note 3 — Road fatalities · Great Britain
Source: DfT RAS0101, Casualties worksheet, 1980–2025. Hover over points for exact values.
Chart note 4 — Fatalities per billion vehicle miles
Calculated from DfT RAS0101 fatalities and annual traffic estimates. This is a fatality rate, not the workbook's all-casualty rate.
Chart note 5 — Casualties of all severities · Great Britain
Source: DfT RAS0101. Not equivalent to KSI: serious-injury recording changed in many forces from 2016, making historical KSI comparisons less straightforward. Reporting of slight injuries can also vary.
Chart note 6 — Recorded speeding offences · 2011–2024
Home Office / parliamentary data: annual points 2011–2024. Figures for 2022–2023 come from Home Office 2023 statistics; publication revisions and differing exclusions may affect comparability. Check series coverage and recording definitions before interpreting trends.
Chart note 7 — Camera-detected speeding fixed penalties · 1993–2010
England and Wales. Paid camera-detected fixed penalties (1993–1999) and camera-detected speeding FPNs (2000–2010); historical definitions differ, so interpret the join cautiously. Sources: Parliament 1993–96, Parliament 1995–99, Home Office 2000–09 and Home Office 2010.
Chart note 8 — Speeding fixed penalties · 1986–2024
England and Wales. Historical series (1986–2009): speeding FPNs issued; modern series (2011–2024): speeding FPNs with fines paid or registered, calculated from Home Office FPN4, offence code 816, including Metropolitan Police. Different outcome definitions mean these are separate series, with 2010 unavailable; the gap is intentional. Cancelled notices, court referrals and courses are excluded from the modern fine series.
Chart note 9 — National Speed Awareness Course attendance · 2010–2025
UK National Speed Awareness Course (NSAC) attendance, 2010–2025. Separate NSAC20 attendees before November 2018 are not included. Sources: House of Commons 2010–14 and UKROEd 2014–25.
Chart note 10 — Speed enforcement · England and Wales (courses: UK)
● Historical speeding FPNs (1986–2009) ● Modern FPN fines paid/registered (2011–2024) ● Recorded speeding offences (2011–2024) ● Speed awareness course attendance (2010–2025). Break in definitions at 2011; absence of a line does not mean zero. Sources: Home Office bulletin HOSB 07/11, Hansard, Home Office FPN4 (2024 release), and UKROEd. Historical FPN figures rounded to nearest 100 where sourced from tables in thousands. The 1986–1990 figures are from Hansard (26 June 1997); 1990 is cross-checked against Hansard (14 July 1992).
Chart note 11 — Speeding penalties and speed awareness courses · 1986–2025
1986–2009: historical FPN totals only; course counts are unknown, not zero, so those bars are not comparable totals. 2010 and 2025: estimated fines shown in amber (2010 extrapolated from 2011–2013; 2025 extrapolated from 2022–2024); courses are actual figures. 2011–2024: Home Office paid/registered speeding fines (England and Wales) plus UKROEd National Speed Awareness Course completions (UK); 2025: estimated fine component, not an official total. Courses for 20 mph limits were separately classified before late 2018 and are excluded here. See UKROEd and the Home Office.
Sources
- Department for Transport, RAS0101: reported road collisions, casualties and traffic, 1926–2025 (2025 edition supplied as spreadsheet).
- UK Parliament: speeding offences 2011–2021.
- Home Office: roads policing, 2024.
- Home Office 2024 FPN4: speeding (816), outcomes by calendar year, 2011–2024 (spreadsheet supplied).
- Historical annual unadjusted KSI figures (DfT STATS19, Road Crash UK)