20 mph limits are a much debated topic in cycling circles. Here, York Cycle Campaign Committee member and science journalist Dr Kate Ravilious explains why the York Cycle Campaign has decided to remain neutral on the issue
Intuitively it seems to make sense. Introduce 20mph speed limits on our roads; vehicles slow down; pedestrians and cyclists are safer; more people decide to walk or cycle. At least that is the chain of consequences that I imagined would happen if more areas imposed 20mph speed limits, but being the nerd I am I thought I’d take a look at the evidence. Sadly 20mph limits don’t appear to be the magic bullet I hoped they might be.
Back in 2014 the Department for Transport commissioned a 20mph impact study. The resulting 68 page report was published in 2018, and presents extensive evidence from 12 case study areas across England, plus a number of 30mph comparison areas. Crucially this is the only study to have proper controls (comparing with similar 30mph zones) against the 20mph intervention, enabling a true comparison of the effect of the 20mph zone. Other studies have sometimes shown benefits of 20mph zones, but they fail to take account of other factors that may have influenced the outcome, such as increased traffic congestion, gentrification of the neighbourhood, planting more trees alongside the road, changes to the road layout such as adding pedestrian crossings and cycle lanes and so on. By comparing with equivalent 30mph zones over the same time period this study is able to determine whether it is the introduction of the 20mph zone that has brought about the changes, or whether other factors are at play.
So what impact do 20mph speed limits have? The first question I wanted to ask was whether 20mph limits actually made drivers slow down. My personal experience of driver behaviour in the 20mph zone outside Fishergate School (see photo) has been one of disappointment, with drivers regularly whizzing over the pedestrian crossing at speeds far in excess of 20mph. This 20mph zone was installed back in 2011, at a cost of £85,000 (equivalent to around £100,000 today), amidst huge controversy and against the advice of North Yorkshire Police, who said they wouldn’t be enforcing the speed limit. The evidence from the DfT report suggests that my Fishergate experience is far from an outlier. On average the introduction of a 20mph limit resulted in speeds falling by just 0.7mph in residential areas and 0.9mph in city centres.
A Department for Transport Statistical Analysis, published in June 2019, reveals that an eye-watering 87% of car drivers travel at over 20mph in a 20mph zone. Buses and lorries are little better, and 92% of motorcyclists seem to think the signs don’t apply to them. More than one fifth of vehicles travel at over 30mph in a 20mph zone. See pages 11 and 12 of the report for full breakdown of the results.
Disappointing. But nonetheless, even a small reduction in speed can make all the difference when it comes to accidents. The most recent estimates suggest that pedestrians have an 8% risk of being fatally injured when hit by a car travelling at 30mph, versus a 1.5% risk when the car is travelling at 20mph. So even a reduction in average speed of 1mph should make a big difference – right?
Wrong. The evidence from the 12 case studies showed no significant change in collisions and casualties. Any reductions that were observed were mirrored in the 30mph comparator areas, suggesting that it was other variables (such as greater traffic congestion) that were causing the reduction.
One location – Brighton – bucked the trend. In this case the benefit to pedestrians was clear, but interestingly the 20mph limits still didn’t lessen the risk for cyclists at all. Here is the relevant paragraph from page 46 of the report:
Brighton Phase 1 is the only case study area where we have been able to estimate a statistically significant change in collisions and casualties, relative to the 30mph comparator area. The results show a significant reduction in overall collisions (-18%), overall casualties (-19%), pedestrian casualties (-29%), and casualties aged 75 or over (-51%). However, there is no evidence to indicate a real change in casualties involving cyclists and under 16s, at this time.
The report authors speculate that the 20mph limits were more effective at reducing accidents involving pedestrians in Brighton because they were applied on roads that were busier in the first place. But for cyclists the benefits still failed to emerge.
So it seems that slower traffic brings no benefit for cyclists. Something I would concur with based upon my own experiences. Cycling along Fulford Road during rush hour is something I try to avoid. Usually the traffic is moving but slow. In these circumstances drivers seem more likely to stop in front of junctions and give way to traffic wanting to turn into the junction. This happens in front of Aldi all the time. It seems like a friendly gesture on the driver’s behalf, but for the poor cyclist pedalling up the inside of the traffic, it is potentially lethal. Meanwhile, it also seems that at slower speeds drivers think it is okay to skim past cyclists more closely than they might when driving faster. Too often I’ve nearly been knocked off my bike by a bus wing-mirror whipping past my head, or someone pulling a trailer that is wider than their car. And those pesky left turns across the cycle lane, without checking mirrors to see if anyone is cycling up the inside, seem to happen more frequently in slow-moving traffic. From my personal experience I feel safer cycling along Fulford Road when traffic is moving at 30mph, than when the road is congested and traffic is moving slowly.
The final question I wanted to ask was whether having 20mph limits in place encourages people to cycle more? In this case the data showed a very small upward tick in walking and cycling – 5% of residents self-reported that they walked more and 2% said they cycled more. But crucially it wasn’t clear that this small increase could be credited to the 20mph speed limits, or whether other factors were at play.
All in all the current evidence indicates that 20mph speed limits offer few, if any, benefits to cyclists. This certainly wasn’t what I wanted to hear, and it is disappointing. To put it another way cyclist safety is often framed in terms of willingness of a parent to allow a 12 year child to cycle independently on a cycle route. The DfT report indicates that imposing 20mph speed limits on a road like Fulford road (“A major strategic road”), would result in 46% of vehicles travelling at less than 20mph – a mere 3% more than when the road had 30mph limits. That means out of every thirty cars that pass the child, only one extra vehicle would be persuaded to travel at less than 20mph (a rise from 13 to 14). Sixteen vehicles would still whizz by at over 20mph. Hardly much of a change from the 17 vehicles travelling at over 20mph when the road had 30mph speed limits. As I parent, these statistics don’t persuade me that 20mph speed limits make it safe enough for my 12 year old to cycle unsupervised on this kind of road.
This is my take on the evidence, and it don’t necessarily reflect the views of everyone in the York Cycle Campaign. However, it is because 20mph zones have no proven benefit for cyclists that York Cycle Campaign remains neutral on 20mph speed limits. The focus instead is on campaigning on the measures that have been proven to increase cyclist safety and participation.
York Cycle Campaign 
Dear Rod,
thank you for taking the time to engage with this blog post and for pointing out concerns you have. I have tried to address the points you make below.
best wishes
Kate
To suggest that the Atkins report is compromised is a serious accusation. Can you please indicate the reference where you have seen this critique? If it is yourself, can you please explain why you have not published your critique and subjected it to proper independent review?
Your explanation of zones and speed limits is an interesting distinction. It is not something that is clear from reading the 20’s plenty briefing notes and I think most people reading the 20’s plenty website would assume that 20s Plenty campaigns for 20mph speed limits. Certainly the recent petition submitted by 20’s Plenty to City of York Council does not make this distinction and requests 20mph on a number of major roads and locations where average vehicle speeds are above 24mph.
Are you suggesting it was irresponsible to campaign for a 20mph limit outside Fishergate school in York? I believe the cost to the public purse was in the region of 100,000 pounds.
The authors of the Atkins report pose a standard hypothesis test, which is: “does the use of 20mph limit alone change key indicative variables?”. For such a test to be valid a control is required – ideally this would be same area, but of course this is not possible – so the authors have attempted a comparison with an alternative but equivalent area. The assumptions made are clear and, with a data set that is necessarily this small to enable the controlled comparison, the levels of significance will be small, but, crucially, this is included as part of the test. If 20mph made a large impact then we would still see the effect here, but we do not. This means that we can’t conclude that 20mph definitively makes no effect from this, but we can conclude that it puts a bound on the effect it makes, with the stated levels of confidence.
The detailed criticisms of the construction of the dataset required to provide a controlled comparison are intriguing, but until you are prepared to publish them in an appropriate format and subjected to rigorous statistical review they remain speculative. The study has limitations, but to accuse it of being flawed is a serious accusation.
You have not provided any alternate peer-reviewed evidence (with adequate control data) to support 20’s Plenty’s claims about 20mph zones reducing casualties and increasing uptake of cycling. Until such evidence exists we can’t draw any firm conclusions about the impact of 20mph as a single intervention.
York Cycle Campaign does not contest that 20mph limits/zones may well be effective as part of a wider set of interventions. However, disentangling the evidence to isolate causal effects is a huge statistical task. Our concern is that the evidence for 20mph limits/zones as a single intervention is poor, and specifically for cycling, it does not exist.
At a time when modal shift is becoming an essential requirement to reduce carbon dioxide emissions from transport it is potentially misguided to imply that this intervention (20mph) should be prioritised over others.
You say that you don’t claim 20mph to be a ‘magic bullet’ solution for walking and cycling but the 20s Plenty website certainly gives the indication that it will solve many of the problems we face. There you have published, and refuse to remove when challenged, the phrase “Cycling and walking levels are significantly up in 20mph areas” as an emboldened bullet point, despite this misrepresenting the findings of the Atkins report. This is compounded by elsewhere by your attempts to discredit the same report, including in this post. This is alarming cherry picking of data..
We appreciate the time you have taken to engage with the York Cycle Campaign, and very much hope that we can find common ground to work together on in future. We continue to remain neutral on the issue of 20mph limits and I hope that my response helps you to understand our perspective. We will continue to campaign for evidence-based and proven methods of making cycling safer, more convenient and accessible for all.
LikeLike
I have to say, I’m with Rod. My home town has introduced large 20mph areas and there is now more cycling and walking.
LikeLike
Hi Kate
I respect the fact that York Cycle Campaign have taken up a neutral position on the subject of 20mph.
We welcome evidence-based approaches. However, sadly, the “evidence” you have referred to is based on reports that are very much compromised or taken out of context.
Zones have a particular meaning in 20 speed limit terms and refer to a 20mph maximum speed across several roads that include at least one physical calming measure. I assume that you are referring to “20mph limits” which have no such physical calming features. I can tell from the picture included above that the Fishergate scheme is not a 20mph zone but is a 20mph limit. I can also see from a cycling perspective that the road geometry, width and facilities are not conducive to comfortable and safe cycling. The 20mph limit only seems to exist for a very short length of road which also does not help compliance.
The 68 page report that you referred to was published in 2018 and is one that I am familiar with in quite some detail. However it did not compare 20mph limit areas with equivalent 30mph limit areas. It compared a few very small 20mph limit areas totalling just 110 sq km with 30mph limited areas totalling 26,896 sq km. Small 20mph areas as little as 1 sq km in Winchester were compared to a 30mph area of 4,184 km. The case studies themselves were so small as to have little year on year significance for the casualties recorded. Walsall was just 2 casualties per annum and all but 2 case studies were smaller than 30 casualties per annum. Hardly numbers which will lead to high confidence levels in any comparison of before and after levels. In addition there were pointers that the data was compromised. In Winchester collisions reduced by 15% but casualties increased by 16%. In Bestwood, Nottingham collisions increased by 9% yet casualties reduced by 20%. In Liverpool collisions reduced by 7% but casualties reduced by 20%. All further signs that the data sets were too small to be credible on events that are as random as traffic collisions and casualties.
With regard to the comparator areas, these were measured over a nine-year period (five years plus one year buffer before, three years afterwards). If there were engineering or other changes to those roads during this time this would have changed casualty rates and would need to be filtered out.
Were A and B roads filtered out of the comparator areas? A and B roads are not similar to 20mph residential roads.
Were 30mph roads that were subject to any engineering changes filtered out? Many could have been as a result of localised casualties and hence data for them would not be indicative of a background trend.
Were 30mph roads that were changed to 20mph filtered out? If not, this would have reduced the 30mph data pool and hence arbitrarily excluded casualties in the after period. Whilst any other case study areas in the comparator areas were excluded, what about other councils setting 20mph limits and radically reducing the 30mph data pool?
Our concern is that it is really too simplistic to compare casualties on a very specific set of roads in small case study areas to casualties across whole regions without considering these factors. We note that the North West comparator area showed a whopping 35 per cent reduction in casualties, which was attributed to road congestion and the closure of police station counters. This should have been another warning bell to the authors regarding the credibility of what they were comparing. Surely in an age in which national road casualties are flatlining, if there is any reliability in this 35 per cent, then government road safety policy should now focus on increasing congestion and closing police station counters!
Our contention is that the case studies size and invalid comparisons were simply too compromised for any conclusion to be drawn on casualties from the report.
With regard to the report on compliance levels this is certainly taken out of context and has no bearing on most 20mph limits. The report is one which annually assesses speed compliance on a range of speed limits in “free-flowing” locations. These are roads with an absence of hazards and the report specifically advises on 20mph results that these “tend to be through-roads, so are not typical of all 20mph.” The report also states “Therefore the ‘free flow’ 20 mph sites in this data set will tend to be unrepresentative of 20mph limits in general, and this effect will be much greater than for other speed limits considered above.”
We conducted a Freedom of Information request and identified the locations of the 9 20mph sites the DfT had used. Most had no residential features (such as houses) at all and one was on an industrial estate. They simply are not representative of most 20mph roads.
Whilst it is correct that on these unrepresentative roads 22% of cars exceeded 30mph, on the equivalent 30mph roads 52% of cars exceeded 30mph. Hence the “evidence” even these unrepresentative 20mph roads would appear to show considerable speed reduction. In fact the report states that average speeds on the free-flowing 20mph roads were 6mph less than on the 30mph roads.
I do accept that most people would take the reports on their face value. But I would hope that with the issues above pointed out then you can understand that there is a limit to what may be concluded from reports due to their design, scale or appropriateness and that they do not constitute evidence of 20mph limits not working.
Throughout the country many councils are making good active travel plans to increase cycling. These include wide-area 20mph limits with public engagement, police enforcement and cross-party support. These are accompanied by many other complementary measures on cycling and walking provision. I do accept that in York there may not have been the same commitment or continuity of support from consecutive administrations to 20mph limits and active travel initiatives since they were implemented. Your police also seem to be endorsing non-compliance through their lack of support. You will know that many aspects of the York administration are atypical of other councils.
We have never claimed that 20mph limits are a “magic bullet” initiative for cycling but that when appropriately and competently implemented they play a major role in bettering conditions for walkers and cyclists. I would therefore hope that you can understand that the York experience is atypical and see the wider benefits of 20mph limits for many sections of the community beyond cyclists and take a supportive position.
For our part we will continue to call out the lack of police enforcement in York and the need for prioritisation of active travel initiatives to support cycling and walking.
Best wishes
Rod King
LikeLike