Cyclists stand accused of putting themselves and others in danger on Britain's roads by constantly breaking the law. Certain newspapers seem obsessed with cyclists supposedly terrorising the streets and you see comments on social media all the time condemning the behaviour of 'Lycra louts'.
Our articles on changes to the Highway Code, including the introduction of a Hierarchy of Road Users and minimum 1.5-metre passing distance that have now been added the Code, led to a number of angry emails landing in the road.cc inbox: "Cyclist's are all now taking to using the pavements to cycle", "cyclists don't give a crap", "Cyclist must be insured to use the roads!"... there's a brief flavour of the less positive communications we received about the article.
Back in 2019, we also reported that Talksport presenter Andy Goldstein had stated on air that 95% of cyclists jump red lights (among other gripes about cyclists). This kind of claim is common, but what's the truth?
Comments on social media frequently cite the fact that many cyclists don't wear helmets or bright clothing as evidence that we're all criminals. You'll know that this is nonsense. The Highway Code advises cyclists to wear a helmet and light-coloured clothing in daylight, and reflective clothing and/or accessories in the dark but, of course, these aren't legal requirements.
You'll also see comments complaining that cyclists are flouting the rules by wearing earphones, riding two abreast, positioning themselves in the middle of the lane, and not using cycle lanes. Again, you'll know that none of these things are illegal.
Jumping red lights is perhaps the classic complaint about cyclists. It's an old favourite. This one at least gets off to a better start than most in that jumping red lights is against the law (some cyclists jump red lights because they feel safer moving into open space at signalised junctions rather than waiting for the following traffic to accelerate into that junction when the lights turn green – but the rights and wrongs or jumping red lights are a topic for another day).
How many cyclists jump red lights, then?
Back in 2007 (yes, we're going back a bit here, but there's not a lot of quantitative evidence out there), Transport for London's Road Network Performance & Research Team looked at the proportion of cyclists who jumped red lights at five sites in the capital and said, "An average of 16% violated red lights, whilst the remaining 84% obeyed the traffic signals. Therefore it can be concluded that the majority of cyclists do not ride through red lights."
More recently, in a 2013 YouGov survey 27% of London cyclists said they ignored red lights occasionally while another 8% said that they did so often. This survey relied on self-reporting and we couldn't judge its accuracy.
Many road users will tell you that the figures are higher than this. You'll sometimes hear claims of people seeing dozens – maybe hundreds – of cyclists jumping certain traffic lights every day. They might be right. How much of a safety concern is this? Let's have a look at some more stats...
From 2007-16, no pedestrians in Britain were killed by red light jumping cyclists, while around five a year were killed by red light jumping drivers. For pedestrians hit by red light jumpers, just 7.6% of those slightly injured and 5.4% of those seriously injured involved cyclists. The other 92%-95% involved motor vehicles.
The percentages relating to cyclists are higher in London, where the concentration of pedestrians, cyclists and traffic lights is particularly heavy: 16% of pedestrians injured or seriously injured by red light jumpers were hit by cyclists, the other 84% involved drivers/riders of motor vehicles.
Judged purely on the number of casualties, red light jumping motorists are a far greater problem than red light jumping cyclists.
If it sounds like this is turning into an 'Us versus Them' thing, bear in mind that almost everyone on the road.cc staff is a motorist as well as a cyclist, the vast majority of you road.cc readers are both, and about 90% of British Cycling members also drive. However, if cyclists are singled out for breaking the law it makes sense to examine the degree to which other road users stick to it as a means of comparison.
According to the Department for Transport's Vehicle Speed Compliance Statistics, 48% of car drivers exceeded the speed limit on motorways in 2017, 52% exceeded the speed limit on 30mph roads, and a massive 86% exceeded the speed limit on 20mph roads under free flow conditions.
How big a problem is exceeding the speed limit? Well, more figures from the Department for Transport say that in 2017 it was reported as a contributory factor in accidents that resulted in 220 deaths, 1,493 serious injuries and 5,855 slight injuries (travelling too fast for the conditions but within the speed limit was reported as a contributory factor in many, many more injuries and deaths). A massive problem, then.
The consequences of cyclists jumping red lights are small compared with the consequences of motorists jumping red lights, and they're tiny compared with the consequences of motorists speeding. In fact, all of those usual gripes you hear about cyclists – riding on the pavement and the like – result in very few casualties.
Despite that, it's cyclists who are routinely denounced as “a dangerous nuisance”, “a threat to safety" and the like on social media, and you'll encounter headlines like "Cyclists kill or maim two pedestrians every week, according to statistics" in The Express... ignoring the fact that between 2007 and 2016 motor vehicles were involved in 98.5% of collisions where a pedestrian was seriously injured and 99.4% of collisions in which a pedestrian died. The biggest threat to pedestrians certainly doesn't come from cyclists!
Mile for mile, pedestrians are more likely to be killed by a motor vehicle than by a bicycle. From 2012 to 2016 in Great Britain, for every one billion miles ridden overall, cycles were involved in 1.4 pedestrian fatalities. For every one billion miles they were driven, cars were involved in 1.9 pedestrian fatalities.
Cyclists breaking the law can sometimes have major – occasionally tragic – consequences, but why do so many people seem convinced that the problem is bigger than it actually is? There's a huge disparity between perception and reality here.
We'd guess – and it is a guess – it's partly because jumping a red light and riding on the pavement are so much more obvious than a motorist doing 35mph in a 30 zone. Plus, some people simply seem to believe that speeding at 85mph on a motorway in a two tonne vehicle is perfectly acceptable in a way that riding a push bike at night with a light out isn't.
Research prepared for the Department for Transport also suggests that drivers can exaggerate the misbehaviour of cyclists because of a tendency to see us as an ‘out group’, and are prone to “overgeneralise from the behaviour of individual members of an ‘out group’ to the behaviour of members of the ‘out group’ as a whole.”
We'll leave the final word to Sam Jones, Cycling UK’s senior campaigns officer.
“Sometimes it feels as though cyclists are represented in the national media as the ‘real danger’ of our roads," he said. "The reality is that we cause negligible harm, but are disproportionately at risk of suffering serious injury. While that might seem to paint a grim picture for us cyclists, it’s worth bearing in mind Britain’s roads are relatively safe, and the benefits of cycling will always far outweigh any risks.
“Just because cyclists represent a smaller risk to other road users, that doesn’t justify inconsiderate or dangerous cycling. Cycling UK would encourage everyone to cycle considerately and within the boundaries of the law.
“The main problem is the shocking injustice that victims of road traffic incidents all too often face. It’s high time the Government took this problem seriously and ended the injustice suffered by far too many families who are being let down by the system.”
Cycling UK has compiled an excellent document to counter many of the accusations most commonly fired at cyclists, by the way. Check it out here.
Add new comment
75 comments
None of this really matters, if all the deaths, incpacities, illnesses, destructions caused by motor traffic through it's various consequences socially, environmentally etc. are factored in.
Motor traffic is orders of magnitude more dangerous than cycling. Full stop.
“It is amusing at this point in the controversy to look back at the abuse with which the poor bicyclist was received for his presumption in daring to use the roads at all seeing that he was usually "not even a ratepayer"; For his crime in being the cause of alarmed horses and too timid pedestrians; for the dust that he raised and the mud that he scattered; And for the recklessness of his excessive speed. It is as recent as the close of the 19th century but it is difficult, in the beginning of the 20th, to believe that it can ever have happened! Amid this storm of criticism and complaint, the cyclist found it far from easy to get accorded equal rights with horse drawn vehicles.” Sydney Webb, “The Story of the King’s Highway”, 1913.
I'm not about to dispute the numbers, but without some reference to a base rate they are all but meaningless. The best construction that can be put on them is that cyclists, in their current numbers, are too few to have much impact on KSI figures of others. And so long as their numbers remain low they are probably not much worth bothering about.
But that leaves the way wide open for others to argue that, in proportion to their far greater numbers, drivers cause less harm. And that therefore every motorist replaced by a cyclist acts only to increase the danger to others.
Now I don't know if that is true, and it is difficult to find the appropriate base rate - is it per mile travelled, per hour travelled, or what. But certainly just comparing the raw numbers for the relatively few cyclists against the overwhelmingly greater number of drivers is not sensible.
Oh, and the last time I did have a stab at comparing the two relative to some base rate, it didn't look so clever for cyclists.
But that is false comparison. At any given set of traffic lights, only one vehicle per lane can go through the lights at a time (not including those instances where a chain of vehicles plays follow-my-leader through on red). So the greater volume of cars is managed down by the fact the opportunity to go through on red is limited to one vehicle per lane plus outlying data.
You're saying that because the opportunities for cyclists to jump red lights are greater than for motorists they should get a pass on the consequences? I'm missing something in the logic.
Regardless, my argument is less about the specific offence of RLJ, more about the impact on KSI (of others) of cyclists as compared with motorists. If the relatively low impact of cyclists rests mainly on their low number then it's not much incentive to increase their number.
Yes, you are missing something in the logic as far as it relates to the context of RLJs. It's not about giving cyclists a 'pass', but about not pro rating the numbers because there are more cars on the road. There are more cars on the road, but they are not all at the stop line. In other words, you cannot give statistical credit for stopping at a red light to the car behind the one that stops at the red light. In that microexample, there is only one car that stopped for a red light.
OK, but you quoted specifically in the context of RLJs, so I concentrated on that. Nevertheless, fortunately we have other countries to look at where there are more cycles and more cycling, and we don't see disproportionate serious injuries involving cyclists. Moreover, if we are not lookingat RLJs, we should be careful that we do not mistake KSIs involving cyclists with pedestrians as collisions caused by cyclists with pedestrians. Remove the RLJ and you remove a causative liability factor.
You seem to be appealing to some sense of fairness, that we ought to level the playing field before comparing cyclists against motorists. I don't see why we should - I'm only looking at the trajectory of the figures if cars are replaced by bicycles as a mode of travel.
So to take your example, if ten cars queueing at a red light were replaced by ten bicycles you're saying we have ten potential RLJs instead of just the one motorist at the head of the queue, and that distorts the figures. Whereas I'm saying, if that is how it is then that is how it is measured.
Nothing to do with fairness, it's to do with understanding statistics and what base lane should be used.
That isn't my example; I'm not saying that; it isn't how it is and it isn't how it's measured. But there are too many presumptions to fix in the statement.
No - they're saying that if you have 10 drivers and 1 cyclist approaching the lights you have 2 potential RLJs - one driver and one cyclist. And if you have 10 cyclists and one driver approaching the lights you have 2 potential RLJs - one driver and one cyclist.*
But in the latter case, an analysis based on modal share will dilute the danger from cycling because of the 9 irrelevant cyclists, and vice versa for the former case.
Or, to think of it another way, imagine that there are equal numbers of cars and bicycles on the roads, and you only ever get one of each approaching lights as they change. But there are then 10 times as many cars parked in a giant car park somewhere. It would make no sense to include those cars in the car park when calculating how likely a car is to be involved in killing or injuring someone due to jumping lights. The cars from number 2 in a queue backwards are similar to the cars in the car park.
[*Roughly speaking. There can of course be cases where the first car or cyclist jumps the lights, and the second follows them through, and in the case of cycling there's more of an opportunity for the second cyclist to go around the first and jump the lights, which skews things even more in favour of driving.]
It's always worth checking / computing rates but I think we're safe on "any number factor times zero pedestrians killed isn't a problem". So your last about "not worth bothering about" applies.
Being cautious of course we should also check the "externalities" e.g. is this because - like cars / car infrastructure - the cyclists are suppressing walking / scaring the pedestrians off the roads? Unlikely. Maybe three cyclists in the UK are actually a homicidal maniacs - but the casualty numbers are low as they only go out once a year? (There are certainly some wronguns on bikes and cycling doesn't make you a saint but this is likely still a no due to the much lesser potential for harm.) etc.
I take your general point - on-road cycling is still a minority activity in the UK and currently it's likely that both the small numbers and "self-selecting" effects may skew what a "cyclist" looks like, relative to the averages of the wider population.
I'm not sure what the baseline should be either and it depends on the question. Given that car use and bike use is not "like-for-like" in several ways is it even worth starting? If so then maybe a "trip-based" count?
Base rate is usually ignored by most of the cycling press.
If the above numbers are correct then on a national scale cyclists are over represented in terms of pedestrian KSIs caused by RLJ.
Which, unfortunately, partially justifies the continual complaints about cyclists jumping red lights.
Can you show your working?
The percentage of RLJ injuries caused by cyclists is greater than cyclists' modal share of traffic on a national level.
A modal share of traffic everywhere? Including all the car dominated rural highways and motorways that have either no traffic lights, no pedestrians or no combination of the two?
Can you explain why anyone would think that is an appropriate base rate to use?
Which is why I was careful to include the word "national".
Do you have any better data?
Edit: According to 2019 DfT statistics cyclists travelled 2.5bn miles on urban roads. This compares to 132.4bn miles for all traffic. Cyclists therefore represent 1.9% of all urban miles travelled.
They are still significantly overrepresented in terms of injuries to pedestrians caused by RLJ.
Base rate is usually ignored by most of the
cyclingpress.As for "justification" not really because the complaints have nothing to do with stats, numbers or indeed logic. Pretty sure that's entirely about "someone who's not in my group / a minority not following the rules" which we're very sensitive to as humans. Will people cite the research in justification? Sure - but you'll find most of the audience would be equally happy with "a cyclist nearly killed my friend the other day".
Given the billion miles rate at the mileage I cycle I'll need a lot of reincarnation before I see one of these. (Wonder if converting to "per time travelled" is any better though. Does a "trip based" measure make any sense?). Tiny numbers in absolute terms so is it worth digging deeper in the UK at all? Is it possible to learn anything more widely applicable from the number (zero killed, not many injured in some years)? Genuine question about the significance of these. If no-one died in the last x years, you make a change and one person dies over the next two can you say if this is success or failure from the stats?
Anyway I'm all for priorities and the low hanging fruit so let's talk about fixing sponges, tea cosies, socks and trousers (yeah, 2000). Or if we want to keep to "infrastructure" then stairs.
Your last paragraph is just multiple examples of why the base rate is important.
The overall number of injuries in those examples is high but the rate of injury is incredibly low. As a consequence they represent the opposite of 'low hanging fruit'.
"A cyclist nearly killed my friend the other day" or variations thereof is a common complaint. The statistics seem to suggest that the frequency of such events is not as exaggerated as we cyclists may have assumed.
I'd assumed it was really very rare. And it is. But hey - it's comparable in "per billion vehicle miles" to cars (assuming it's statistically significant etc.). That is slightly surprising to me. Just wondering again if we can actually make much out of such small absolute casualty numbers. "Narrative" in this case may not be a distraction but give you a pointer to any issues. * Also, if you made an intervention you might have to wait several years to know if you'd made any meaningful difference...
I speculate that "A cyclist nearly killed my friend the other day" is likely a "common complaint" because:
a) it is actually an uncommon complaint - unlike "my friend almost got really wet on the way home" or "my friend had to dodge a couple of cars the other day". We don't tend to pass those stories on so much. Like the converse of Richard Feynman's example "You know, the most amazing thing happened to me tonight... I saw a car with the license plate ARW 357. Can you imagine? Of all the millions of license plates in the state, what was the chance that I would see that particular one tonight? Amazing!"
b) it's also likely to be reported by the victim to someone else in the first place because of its unusual nature.
c) There are certainly legitimate reasons why cyclists might be more "startling" / "aggrevating" than car close passes for pedestrians. There's the startle effect - you may not be aware of the cyclist until they're very close (because quiet). They may also be where you don't expect anyone other than pedestrians (legally or not). They may be taller than you if sat up - we're sensitive to things looming over us. So more like a larger human running at you - maybe that's more alarming than a vehicle coming near. The distancing effect of being in a box on wheels working both ways. If you shout at them it may be you're more likely to get a reaction than from the occupants of a car.
Finally people are almost certainly overestimating the likely consequences of any collision.
* We're talking about such a small number of *incidents* so maybe a couple of councils put cycle routes down a hill next to the old folks home? (That could certainly be fixed). It's possible the effects of a poorly organised / marshalled sportive or a few wronguns on bikes could take the leaderboard from a large fraction of the road.cc mile-eaters.
The discussion is about injuries caused by Red Light Jumping.
Yes it's a small number of injuries relative to WW2 etc but cyclists are still overrepresented.
Given that a pedestrian is far less likely to be injured in a collision with a bicycle than with a car that suggests an even greater over representation in terms of collisions.
Kinetic energy is directly proportional to the mass of the object and to the square of its velocity: K.E. = 1/2 m v2. If the mass has units of kilograms and the velocity of meters per second, the kinetic energy has units of kilograms-meters squared per second squared.
This formula tells us that all other things being equal, cars will have a higher impact on a pedestrian they hit. At 20km/h a 1500kg car has 15 times the kinetic energy of a 100kg cyclist, at 40km/h it's 60 times.
Sure, I did secondary school physics too. But you can only kill a person so much, after that they are dead anyway, and the rest of the kinetic energy is just wasted, so to speak, even if it was transferred to the casualty in the first place, which is hardly a given.
All of which is besides the point - the stats stand on their own two feet.
Indeed - but the absolute numbers are very small. I'm not saying people on bikes do not collide with crossing pedestrians at lights and elsewhere. There's at least one case everyone knows of - salient because this is apparently so rare. But with such small numbers an appropriate baseline may even be "compared with numbers of people who had a heart attack while crossing the road, or collided with another pedestrian / animal, or tripped over and injured themselves..."?
Not sure about the relevance of the heart attack thing, but anyway.
Yes, with cyclists the numbers are very small. And that is a valid point when the question is about which legislative levers to pull, which will save most lives, given the current numbers of cyclists and drivers? Legislation targeting motorists has a bigger effect than legislation targeting cyclists - because of their relative numbers.
However, when it comes to arguing for replacing motor traffic with bikes, you need some measures per unit of travel. You can't advocate for "big cycling" by relying on the statistical insignificance of "small cycling".
Articles such as this which say that cyclists account for so few casualties are really only saying that there are so few cyclists - it's not a compelling reason to increase their number.
Well I think you *can* advocate for big cycling even in the face of some potential negatives. (It's all negative to some...) Of course it's the whole picture of costs and benefits as compared to those of the status quo (or where we're likely to get to otherwise). You can make the roads safer by removing the remaining cyclists and pedestrians from getting about entirely - should we? Helmets for drivers and pedestrians? No driving until you're 30 etc..
"However, when it comes to arguing for replacing motor traffic with bikes, you need some measures per unit of travel."
Agree on the rate in general. However this would be correct *if* if the cycling accident rate were somehow fixed and / or you were doing a like-for-like. So turning today's car users into today's bike riders - and somehow maintaining the conditions which keep today's bike riders riding as they do. (You could keep the car numbers static of course by "importing" cyclists via their being more riding without replacing car journeys or a modal switch from walking / the bus etc.)
I think for "big cycling" to occur (more than an extra % or two) conditions would have to change anyway so I doubt the cycling figures - and possibly the motoring ones - will stay constant. If achieved it will change other factors also.
Ah - the notes in the PACTS 2019 do a concise summary (albeit for rates in terms of injuries *experienced* by a particular mode):
Also some recent analysis by Aldred et al -
https://injuryprevention.bmj.com/content/27/1/71
Quite different figures for cycles relative to cars here (although same ballpark). Paper also looks at how that varies across different road environments.
Looks like the CUK ones are from Road traffic: Road Traffic Estimates Great Britain 2016. Tables TRA0402 (pedal cycle traffic) &
TRA0104 (motor traffic). (DfT, 2017)
It is very simplistic to think that increasing units of travel equates to an increase in incidents. In the first lock down, we saw a massive drop in motor traffic - however we didn't see proportional drop in collisions, KSIs or motoring offences. Probably because a higher proportion of those still driving were people who take risks and don't follows rules - coupled with emptier roads and therefore more opportunity to break rules (without all the safer drivers getting in their way).
In my opinion, adult cyclists largely fall into three categories:
1. Those who understand that actually with sensible precautions, the risks outweigh the dangers of cycling. Some of these may jump lights, they might ride on pavements, but only ever in a very safe manner.
2. Those who cycle rarely, and when they do rarely cycle on the road.
3. Those who think they are invincible and ride recklessly through red lights and recklessly.
If more people start driving less and cycling more - I would wager that most of their numbers would come from groups 1 and 2 - and many could move naturally or with the aid of cyclability from group 2 to group 1.
This whole piece is very much like the Daily Mail articles it decrys, just in inverse. Deflect blame to the out group and then climb a hobby horse.
For them it is cyclists jumping red lights for this publication it is car drivers speeding. This is despite the fact that speeding is not even close to being the most common factor in road accidents/deaths. Where is the vitrol towards "looked didn't see" and "failed to anticipate the path of another vehicle".
Secondly the quantity of lies/damn lies/statistics most of which can be pulled apart pretty easily. E.g. 85mph on the motorway is the upper quartile speed of cars on a non congested motorway, there is around 100 deaths per year in 60 billion passenger miles of driving on these roads, ergo 85 on the motorway is pretty safe.
Rather than being anti car cycling groups should be pro road safety and road utility. Rather than pointing fingers at speeding and red light jumping it would be a lot more productive to ask why do people do these things and fix the infrastructure to accomodate human behaviour. (example green waves where cyclists don't have to jump red lights, cycling infrastructure that simply doesn't have red lights on it or improved A and B roads with less lethal features that allow motorised traffic to get where it wants to go)
This post needs more likes
"Rather than being anti car, cycling groups should be pro road safety and road utility."
Amen to that Brother.
But you have to admit, when some gammon starts spouting off about cyclists being a danger, directing them to their local news website which will undoubtedly carry several reports of deaths, injuries, traffic jams due to collisions and all manner of motor vehicle related mayhem is a good argument about where their own road safety concerns should really be prioritised.
Pages