In 2013, we’ve heard a lot about the dangers of cycling. So, as the year comes to end, it’s time for a brief reminder of some things that aren’t dangerous.
Riding a bike without hi-viz clothing is not dangerous.
Riding a bike while keeping yourself dry with an umbrella is not dangerous.
Taking the dog for a walk by bike is not dangerous.
What can make cycling dangerous is the physical environment. Tellingly, all these pictures show people cycling in places where they are insulated from danger – where they are separated from motor traffic.
As Chris Boardman so often states, cycling is not an intrinsically dangerous activity. If we truly want to make it safe and attractive, we need to focus on changing the physical environment, not on the way people dress or behave. Let’s remember that in 2014.
(i) Give fuller references to the evidence.
(ii) Suggest the reason for the observed changes (particularly the apparent adverse effects on cyclist casualty rates).
(i) Look at helmet advocacy in the context of a car dominated “road safety” culture.
1. Specifically on New Zealand (and also for the similar case of Australia which also has a mandatory cycle helmet law), peer reviewed scientific analysis is provided by Dorothy Robinson in “Head Injuries and Helmet Laws in Australia and New Zealand”
2. The main collection of evidence for New Zealand is in “Mandatory bicycle helmet law in New Zealand” ,to be found on Chris Gilham’s web site “Mandatory bicycle helmet law in Western Australia” .
3. On cycle helmets generally, the main site to visit is cyclehelmets.org , which is administered by the Bicycle Helmet Research Foundation (BHRF), an incorporated body with an international membership, the object of which is: “to undertake, encourage, and spread the scientific study of the use of bicycle helmets, in the context of risk compensation and sustainable transport”. BHRF also considers the effect of the promotion and use of helmets (whether on a voluntary or mandatory basis) on the perception of cycling in terms of risk; and the contribution and potential of cycling towards maintaining physical and mental health.
4. Also take a look at the CTC’s “Cycle helmets: An overview of the evidence”
What happened in New Zealand and why?
The graph shown indicates that the mandatory cycle helmet law is associated both with a decline in cycling, and an increase in cyclist casualty rates. The evidence points to this occurring because:
1. 1. The helmet law put people off cycling, both because of the inconvenience of wearing helmet, but also by dangerising cycling – seeing it as something which is inherently hazardous.
So much for the decline in cycling. Although other factors associated in a car-dominated transport system may be at fault as well, the law was associated in reducing the numbers of people engaged in a healthy activity, and one which is also quite low risk. But also:
2. Safety in Numbers: The decline in cycling is itself associated with an increase in cyclist casualty rates. A reason for this is often known as “Safety in Numbers” (SiN), but is actually a phenomenon known about in studies of road casualties for some time. Essentially it is a version of behavioural adaptation or risk compensation. In this case, the implication is that other road users become more used to the presence of cyclists the more of them there are and take more care.
SiN does not mean that all that has to be done for the safety cyclists is to increase their numbers, nor does it imply that it will always work – there is normally a critical level at which other road users reach the increased awareness level. But it does mean that casualty rates for cyclists will increase with a decline in cycling.
3. 3. Risk compensation: Risk compensation (RC) is the big open secret of academic research on safety on the road. Everybody knows it occurs (at the very least, the idea that people do not adapt to perceived danger is highly improbable). In this case we are dealing with the adaptive behaviour of helmet wearers – RC by people when they wear a helmet compared to when they don’t. To a lesser extent RC also means the possible changes in behaviour by other road users towards those who are now considered to be less vulnerable.
4. Helmet effectiveness: In addition, there are doubts about the potential effectiveness of cycle helmets against likely impacts on the head, as well as the relative importance of other impacts on the body for which helmets are not supposed to have effectiveness.
All of the above refer specifically to the effects of a law, but 2 – 4 apply to cycle helmet wearing without legal compulsion.
In analysing helmet wearing the importance of a law, such as those in new Zealand or Australia, is that the effects of helmet wearing across most of the population (of cyclists) can be observed. There are, as always, issues about interpretation of the evidence. In particular, casualty rates have to be looked at in the context of general changes in the safety of road users of all types: possible benefits from helmet law’s may be claimed by helmet advocates when in fact other factors are at work to explain possible declines in casualty rates.
Helmet advocacy in context
Along with points 2 – 4 above, there is another fundamental problem with helmet advocacy. As well as short term risk compensation, there is the question of a longer term adaptation to the wearing of cycle helmets. This is a question of cultural change in which the issue is about the wider meanings of cycle helmet use: why is it assumed that cyclists should consider wearing them in the first place?
These are the group of questions which have been raised by sceptics of helmet advocacy – and not just about compulsion. ( See our posts here for just some). Why should cyclists be expected to wear helmets, and not car occupants, certainly for longer journeys? Wouldn’t it be more sensible and civilised to reduce danger at source, for the benefit of all road users? If cyclist responsibility is to expected, why shouldn’t this come in the form of supporting responsible cycling behaviour, particularly with regard to pedestrians? Apart from the Safety in Numbers effect, why don’t we look at countries which regard cycling as a normal everyday behaviour – done in normal clothes and rarely in helmets – and have both high levels of cycling and lower casualty rates? And why, when official “road safety” initiatives are supposed to be based on evidence, should it be pursued with minimal, zero, or even negative effects shown after at least 35 years of helmet advocacy?
These questions build up into an accusation against helmet advocacy – one which explains the bitterness of discussions about helmets. They lead to the conclusion that helmet advocacy is getting hold of the wrong end of the stick, is victim-blaming and a major red herring which draws attention away from what should be done for cyclists’ (and other road users’) safety.
Road safety theory and politics
Ultimately the importance of debate about cycle helmets is of great importance because it draws attention to major issues of safety on the road. Risk compensation suggests that iconic “road safety “initiatives and ideology have:
1. A. Transferred risk from the most dangerous to others to those least dangerous to others and most vulnerable, as with seat belts, more crashworthy cars and highway environments more forgiving to careless driving. In the first instance, this may lead to larger numbers of pedestrians – and to some extent cyclists – being hurt or killed as a consequence.
2. B. When the move of pedestrians and cyclists (particularly the young, elderly and disabled) away from the road environment and the resultant reduction of cyclist/pedestrian traffic has led to reduced casualties among these groups, this has been regarded by the “road safety” lobby as progress.
3. C. Reductions in road deaths related to levels of motorisation have occurred at different times in a variety of different societies irrespective of the kind of official “road safety” interventions that have been applied.
Naturally, “road safety” personnel – whether highway engineers, road safety officers, or doctors, police officers, etc. – are unwilling to fully accept this. Yet most accept the truth of risk compensation to some extent. The debate about helmets inevitably results in discussion about risk compensation, evidence based policy, and the responsibilities of different kinds of road users.
And I would argue that this debate should be pursued, as it leads to a greater understanding of the way safety on the road is thought of. In particular, we need to criticise the “road safety” lobby’s neutralising of the difference between the obligations of those with greatest potential to endanger others and that of those endangered by this behaviour. That is probably as, or more, important than whether individual cyclists wear helmets.
If that sounds overly political, the fact is that this is political: it deals with the relative destructive power of different groups of road user, as well as the official politics of government. It deals with ideology: however much the evidence needs to be presented, ultimately that has only some effect on what people will do – the issue is about culture and what constitutes acceptable behaviour.
And it has to be worked through, for the safety of cyclists and other road users.
Transport for London’s latest Travel in London report was released just before Christmas, and, as always, it is packed full of statistics. One of the most telling graphs comes early in the report -
The distance travelled by motor vehicle has fallen fairly consistently across London since the year 2000. The sharpest decline has been in central and inner London, but even in outer London there has been an 8% decline. It should also be noted that the ‘central London’ count does not correspond precisely to the congestion charging zone – TfL tell us that the fixed counters here are mostly outside of this zone (for instance, on the inner ring road). So the red line decline suggests motor traffic is declining even in places where you might expect it to have been displaced following the introduction of the charge. Transport for London also supply this titbit of information -
The traffic data considered in this section run only to the end of the 2012 calendar year. While it is too early to draw firm conclusions, it is interesting to note at this stage that observed traffic data for 2013 are showing increases in traffic relative to 2012. If sustained, this could signify a break with the now long-established pattern of slowly declining levels of road traffic in London.
That is, they seem to think motor traffic will start rising again in 2013. Certainly motor traffic in outer London (the green line) did rise by 0.3% in 2012, and this area accounts for 70% of all of London’s motor traffic. Time will tell.
Another graph that tells a story is this one -
What is noteworthy is the remarkable consistency of vehicle speeds in the different areas of London over the last seven years. It just hasn’t budged, despite ‘smoothing traffic flow’ and all the attempts to fiddle with signal timings. There’s been no change – average traffic speeds in central London still hover around 9mph, and around 12mph for inner London. The biggest changes are actually seasonal, with speeds increasing in the summer as demand dips.
The picture for cycling isn’t particularly pretty. Slow but steady growth in outer London (the black line) has tailed off, while cycling levels in inner and greater London (green and red) actually declined last year.
There were 582,000 cycle stages on an average day in 2012, which represents just a 1.8% increase on 2011.
The table above suggests that the growth in cycling from 2000 is tailing off; the large percentage increases (from an admittedly very small base) in the noughties have been replaced by much slower growth from 2009 onwards. This is something acknowledged in the Report itself -
The majority of indicators of cycling in London suggest a slowing in 2012 of the recent high rates of growth.
followed by a list of excuses -
There are a number of reasons for the slowing in cycling growth. In addition to the weather, which can impact on cycle flows, delivery of new cycling infrastructure slowed in 2012/13 in the run up and during the Games period, following a moratorium on new project construction. Further, the implementation of the Better Junctions cycle safety review has impacted on the pace of delivery of major new cycle programmes, including both the Barclays Cycle Superhighway and Better Junctions programmes.
Weather doesn’t seem like a particularly convincing explanation, as the Report concedes that
the overall trend in cycle flow has been upwards across the time series despite annual average temperatures remaining broadly stable. In particular, cycle flows increased considerably in summer 2011 despite average temperatures being lower than the three previous years [my emphasis]
and also that there is no clear relationship between rainfall and cycling levels.
The mode share figures for cycling across London are not impressive. In particular there seems to be a major problem in outer London, where cycling’s mode share is actually lower than that for Britain as a whole -
Some serious work is clearly needed even to attain the Mayor’s unambitious target of a 5% cycling mode share (for London as a whole) by 2026.
The final issue is safety. As @geographyjim pointed out on Twitter, the Report reveals that
Pedal cycles accounted for two per cent of daily journeys, but 22 per cent of KSI casualties in London in 2012
which is a shocking statistic for a mode of transport that is not intrinsically dangerous. The report also includes this interesting graph, showing the casualty rate per distance traveled, by mode of transport in London, broken down by age group.
By this measure cycling (represented by the green dots) is more hazardous than any other mode of transport, bar motorcycling, for all age groups. The casualty rate is particularly bad for teenagers cycling.
The number of pedestrians who were killed or seriously injured (KSI) rose 15% in 2012, up to 1,123 from 980 in 2011. And the picture is just as bad for those cycling, with KSIs in 2012 up to 671, an enormous rise of 18% on the figure for 2011 – just 571. In fact, cycling KSIs in London are now up 55% on 2009, when there were 433 cycling KSIs.
Obviously we should take into account the fact that more people are cycling, but it is now clear that the number of KSIs is outstripping the increase in trips. The risk of a cycling KSI per trip is now higher than any time since 2003, as this graph shows.
The changes of policy that are proposed in the Mayor’s Cycling Vision are coming at exactly the right time, with danger increasing, and cycling levels apparently starting to stagnate. The question is now whether they can be delivered.
Back in late 2011, I wrote a post about how the TfL policy of ‘smoothing traffic flow’ is antithetical to the creation of space for cycling. Creating ‘smooth flow’ means attempting to push as many motor vehicles through a green signal phase as possible, either through longer phases, or more stacking lanes. A ’100% efficient’ junction is one at which all the queuing motor vehicles manage to pass through the junction on a green signal; the queue disappears at each signal phase. So taking some of that space away for cycling, or allocating more time for pedestrians to cross, will inevitably mean ‘flow’ is ‘less smooth’, when flow is measured purely in terms of motor vehicles.
The context for this post was the death of Deep Lee at the King’s Cross gryatory in October of that year, and a public meeting in December at which TfL representatives tried to justify doing pretty much nothing at all to adjust the layout of the junction where she died. They argued that taking one of the two queuing lanes away and replacing it with protected space for cycling would cause ‘considerable queues’.
In the short term, this would probably have been true. Queues would lengthen around King’s Cross, as the amount of time allocated to drivers to pass through the junction would have been reduced. But this assumes a static volume of motor traffic, that can’t be adjusted.
People are not stupid. If congestion on the road increases, they will switch to other routes, or more importantly switch to other modes of transport, if those modes of transport are sufficiently attractive. By contrast, demand for driving in cities is so high that however much space and time you allocate to it, that space will become filled with vehicles.
We know from cities around the world that taking space away from driving does not result in congestion, or increased journey times. It results in better cities. The private motor car is an extremely inefficient mode of transport in built-up areas, and the space in our cities can be used far more efficiently.
But even if it were true that more congestion would result as a consequence of changes to junctions like King’s Cross, should that matter if people on foot or on bikes are being seriously injured, or dying, because of compromised layouts? Is it acceptable to trade off queueing times for drivers against the risk of death? I argued in that 2011 post that
Transport for London have chosen minimizing queueing times for motor vehicles over the safety – indeed, the lives – of vulnerable road users on their network.
That’s even more clear from the details that have emerged from this week’s inquest into Deep Lee’s death. She was stuck in traffic, unable to progress forwards, trapped just ahead of the HGV that was to kill her.
In this picture of the junction, taken while some minor changes were being made in 2012, you can see a man on a bike (just visible on the left) filtering his way forward through the two stationary lanes of traffic, like Deep Lee would have been doing.
She had the misfortune to find her progress forward blocked by two vehicles, a bus and a minicab (close to each other thanks to the narrow lanes here), precisely while she was stranded just ahead of an HGV, apparently in its blind spot. This much is clear from Andrea Casalotti’s summary of the inquest, particularly this detail -
Deep Lee did nothing wrong, nor dangerous. She was unable to reach the ASL, because there was no feeder lane, because TfL was and is still unwilling to take one lane out. The lanes were too narrow and she became bottled in. Even if she had been able to reach the ASL, it was occupied by two vehicles.
Quite obviously, she would not have found herself in such a dangerous position had there been a safe, dedicated cycling route to the head of the junction. Providing two narrow lanes for queuing motor vehicles, instead of an approach that takes the safety of people on bikes seriously, demonstrably leads to dangerous situations like the one that resulted in Deep Lee’s death.
The report in the Camden New Journal notes
TfL head of capital development Nigel Hardy told the court there was a plan to introduce cycle lanes in Pentonville Road and Caledonian Road as part of a second-phase revamp of King’s Cross expected to begin next year. Despite calls from the London Cycling Campaign, which attended the hearing, a cycle lane will not be set up at the junction where Ms Lee died.
So it seems that while there will be some adjustments elsewhere in King’s Cross, this junction will remain unchanged. However, the details for the Central London Bike Grid, released yesterday by Transport for London, suggest that the North-South Superhighway will run through precisely this location.
So there should be serious, substantial change here – the kind that is desperately needed. What form it will take, time will tell.