Thursday 10 February 2011

What won’t bring about mass cycling (6) ‘Safety in Numbers’

Last week

the DfT released road accident figures for the year to September 2010, which show that reported cyclist fatalities and serious injuries rose by two per cent and total casualties, including those slightly injured, rose by three per cent.

(See all the statistics here)

News like this is always embarrassing for Britain’s vehicular cycling campaigners because they know that cycling among traffic is safe. They invest a lot of energy and commitment in promoting positive images of cycling, and headlines about dead and injured cyclists might be a bit off-putting for the general public, who already suffer from the ill-informed perception that cycling among motor traffic is dangerous.

The CTC rushed to correct the media:

Roger Geffen, CTC campaigns director commented: "Some media reports claimed an increase in cyclist casualties without mentioning the latest annual growth in cycle. In fact, the rise in the percentage of people cycling far exceeded the rise in cyclists killed."

"There is international evidence that cycling gets safer the more cyclists there are, because drivers get used to sharing the roads safely with cyclists."

"More cycling can mean fewer cyclist casualties. For instance cycling on London's main roads last decade went up by 117 per cent and cyclist casualties fell by 21 per cent. And even if cyclist casualty numbers go up, this can still mean that cycling is getting safer if many more people are cycling, as is happening in Britain."

But is it really true that many more people are cycling in Britain?

Cycling has made a little progress; average annual distance by bike for Britain's cyclists has increased from 42 miles per person in 2008 to 46 in 2009. That is more than the 1995/97 average. That means average trip length has increased from 2.4 to 2.8 miles. Hardly a rolling revolution, wouldn't you say? The survey states clearly: "Frequency of bicycle use has remained fairly stable over time since 1998/00. In 2009, 14% of respondents said they ride a bicycle at least once a week and a further 9% said they did so at least once a month. 68% said they use a bicycle less than once a year or never."

The likes of the CTC might find small victories in the increases in average journey length by bicycle, or the fact that more people in high social economic groups are now cycling, but I do not.

If you look at the overall figure, below, it is clear to see that despite campaigning efforts to the contrary, cycling as modal share is on a steady downward pattern. And whilst all journeys have decreased overall, there are more cars than ever on our roads.

That’s not how it appears to the CTC, however.

Chris Peck, policy officer at CTC, was also keen to downplay the latest casualty figures:

“The concept of 'Safety in Numbers' suggests that injury rates rise at 0.4 of the growth in cycling. Rates are falling or holding steady, even if actual numbers of injuries may be rising. Last year cycling deaths fell to their lowest level ever recorded at the same time that cycling reached the highest level in 17 years.

The CTC proves the case for ‘Safety in Numbers’ with examples like this

York, comparing 1991/3 and 1996/8: mode share for cycling rose from 15% to 18%, cyclist KSI fell 59% (from 38 to 15).

But this only applies ‘danger’ to KSI – those cyclists killed or seriously injured. Cyclists who were just knocked off their bikes and bruised or are otherwise slightly hurt are, in this conceptual framework, implicitly ‘safe’.

Of much more interest are trends – changes which are consistent year after year. Here are some trends. In 2002 the annual total of ALL cycling casualties was 17,107 and in 2009 it was 17,064. The figure for cycling casualties has remained roughly the same over that period. But for those same eight years the figures show a continuing decline for all other road user groups: pedestrians down from 38,784 to 26,887; motorcyclists down from 28,353 to 20,703; car users down from 197,425 to 143,412. These figures all indicate a consistent trend.

But setting aside arguments about the interpretation of those casualty statistics, let’s take a closer look at the concept of 'Safety in Numbers' which, it is claimed, suggests that injury rates rise at 0.4 of the growth in cycling.

That ‘scientific’ statistic derives from this article.

This concludes

A motorist is less likely to collide with a person walking and bicycling when there are more people walking or bicycling. Modeling this relationship as a power curve yields the result that at the population level, the number of motorists colliding with people walking or bicycling will increase at roughly 0.4 power of the number of people walking or bicycling.

It has been argued (ironically, by none other than John Forester) that

Correlation does not demonstrate Causation

That statement of conclusions leads one to believe that there had been a demonstration that increasing the volume of bicycling at some location had produced a less-than-proportionate increase in the number of car-bike collisions. In fact, Jacobsen does not report on any such event; I know of no such report anywhere. All that Jacobsen has investigated are the accident rates in different areas or different times with differing amounts of bicycling. The most that he can show are correlations between the two sets of data, because he makes no investigation into any causal relationship.

There’s a similar problem in ‘road safety’ where arguments about the safety of roads are based on interpretations of data which can be quite misleading. A ‘safe’ road with no record of cycling injuries may be ‘safe’ only because few cyclists use it, out of terror at the motor traffic speeds and volumes. And falling casualty rates don’t necessarily mean that Britain’s roads are becoming safer. Engineering plays a part in restraining reckless driving– road humps, speed cameras, carriageway markings and numerous other measures either designed to control driver behaviour or to ameliorate the consequences of reckless driving (if drivers keep crashing into trees at a particular site then the trees will be cut down). Add to that all kinds of improvements in medical care (beginning with the air ambulance whisking you off to hospital). Add to that cars which have much better protection for those inside them (seatbelts, rigid steel safety cages, airbags). Add to that the migration of those most at risk – pedestrians and cyclists – from ‘A’ roads and rural roads.

The Road Danger Reduction Forum critically scrutinises the DfT’s A Safer Way: Making Britain’s Roads the Safest in the World noting that

The chances of being killed or hurt using the more benign forms of transport (walking, cycling) may have increased even when overall numbers for these modes have declined, simply because of a decline in numbers using these forms of transport.

Sometimes this may have happened precisely because the road (or more specifically the nature of motor vehicle traffic and those responsible for it) has become more dangerous.

(See RDRF response here)

Low road casualty figures at a specific site do not automatically mean that site is ‘safe’ for pedestrians or cyclists. There may be other explanations.

There are also other ways of measuring cycling and danger. For example

Compared with driving, cycling now has a risk 12.7 times as large per km travelled, when it was just 9.3 times as risky in 1981. So cycling in Britain is now safer overall than it used to be, but it's less safe than it used to be relative to driving.

However, there are other difficulties about the CTC’s methodology, apart from the establishment of a questionable causal relationship between two sets of data, one of which is itself very questionable (‘growth in cycling’).

In putting the case for ‘safety in numbers’ the CTC reminds us that

The Netherlands has witnessed a 45% increase in cycling from 1980-2005 and a 58% decrease in cyclist fatalities.

Copenhagen, 1995-2006: 44% increase in cycling, 60% decrease in KSIs, with cycle to work modal share rising from 31% to 36%.

But what’s left out by the CTC is the question of infrastructure. This is hardly surprising, as the CTC has always been an organisation with a leadership and membership bitterly opposed to the principle upon which Dutch and Danish mass cycling is rooted, i.e. the separation of cyclists from motor vehicles. And as UK cycling has declined and then stagnated the CTC has kept the faith with 'the right to ride' among motor vehicles, displaying little or no interest how the Dutch and the Danes put a brake on their car dependency, and almost never referencing the huge corpus of documentation regarding infrastructure.

For the CTC to blithely omit the whole dimension of infrastructure once again illustrates how bogus is the causal relationship it purports to establish between two sets of statistical data. Cyclists in the Netherlands and Denmark are safe not primarily because there are lots of them but because they are either kept separate from motor vehicles or enjoy urban infrastructure which gives them not equality with but priority over drivers. And historically it is important to realise that the infrastructure came first, not the cyclists. And the particular piece of infrastructure which kick-started the Dutch cycling revolution was bike grids, i.e. safe, direct segregated cycle paths in towns and cities.

Or to put it another way

While Britain continues to ignore the need for decent cycling infrastructure, the situation will remain the same. The cycling rate will continue to flatline and cyclists will continue to be treated as outlaws with no legitimate place either on the road or on the pavement.

Copying Dutch infrastructure which gives cyclists direct journeys in safety is the only way of turning this around.

In other words, here is no such thing as an alchemical essence ‘safety in numbers’ which will transform a tiny quantity of vehicular cycling dross into golden mass cycling but rather ‘safe cycling infrastructure which encourages large numbers of cyclists’. Build it and they will come.

However, the CTC asserts that

The correlation between cyclists’ safety and cycle use - the ‘safety in numbers’ effect – may exist for a variety of reasons:

1. Drivers grow more aware of cyclists and become better at anticipating their behaviour.
2. Drivers are also more likely to be cyclists themselves, which means that they are more likely to understand how their driving may affect other road users.
3. More people cycling leads to greater political will to improve conditions for cyclists.

We want to halve the risk of cycling by doubling the numbers.

A bogus hypothesis is applied to the thesis that putting more cyclists on the road will reduce their exposure to risk by an amazing 50%. But none of the CTC’s three core principles work, or could ever work, for vehicular cycling as it exists in Britain. If the ‘Safety in Numbers’ theory was true and the presence of lots of cyclists generated greater driver awareness, then surely central London would be one of the safest places in Britain for cycling. Sadly, it isn’t.

Hackney possibly has more cycle trips than any other London borough (unless we believe the claims of Camden Council), yet it also has significantly more cycling fatalities than, say, Waltham Forest, which has far fewer cyclists and cycling journeys. This does not mean that Waltham Forest is safer (recent figures for total cycling casualties are not encouraging) but it does suggest that you are less likely to die cycling in this borough. But I don't myself think that 'safety' or 'danger' can be deduced simply from road casualty figures, or even the distance you cycle. Exposure to risk varies according to weather, road surface, time of day or night, or even random factors such as who you'll encounter on your journey. There are reckless drivers with atrocious records out there on the streets and whether or not you'll die by encountering them in the wrong place at the wrong time can be entirely a matter of chance.

Having proposed the great ‘Safety in Numbers’ project, how is the CTC to accomplish it?

The emphasis must now be on tackling the fears that prevent people from cycling more or not cycling at all. This can be done by: improving driver behaviour, creating more welcoming and cycle-friendly streets and giving people the confidence to cycle more.

Let’s examine those three initiatives.

Firstly, ‘improving driver behaviour’. Since the majority of British drivers never ride a bicycle and since drivers as a social group are the most lawless, anti-social, violent and dangerous in British society, appealing to their better nature or hoping to change their behaviour through raising their awareness is an optimistic strategy.

Seeking to change driver behaviour through new legislation – making cycling a part of the driving test or Strict Liability laws – is most unlikely to work, even if you could get such legislation through parliament (which under the ConDems you probably couldn’t, and I very much doubt if the Labour Party would push for it either). The problem isn’t a lack of legislation but rather that existing motoring laws are only minimally enforced, with generally risible penalties for the wilfully reckless, dangerous and criminally negligent operation of dangerous machinery (i.e. motor vehicles).

In any case, the notion that good driving is predicated on ‘awareness’ – lots of cyclists results in drivers being more aware of cyclists and therefore more cautious – strikes me as very questionable. Dutch drivers behave much better because almost all of them will also be cyclists; in Britain that isn’t the case, and even if you could double the numbers of people prepared to cycle who were also drivers that would still leave non-cycling drivers as an overwhelming majority. Moreover, not only are the conditions of vehicular cycling deeply unattractive in most parts of the UK, those conditions are, I believe, getting worse, not better.

I would argue that cycling in Britain is becoming objectively more dangerous in a variety of ways. Modern cars have the power and speed once only available only to racing cars, and while drivers may be ‘aware’ of speed limits, their machine bubble offers the temptation to speed. The massive reduction in Britain’s traffic police over the past twenty years results in the subliminal awareness that you can commit hundreds of criminal offences as a driver and reasonably expect never to get caught. In the unlikely event that you do get caught the penalties will generally be so insignificant as not to deter you from repeating the offence. Drivers are ‘aware’ that it is a criminal offence to steer with one hand while using a mobile phone with the other, but they correctly perceive that they are most unlikely to get caught, and if they are caught the consequence will be a small fine and 3 points on the driving licence. SatNavs are another new distraction. Drivers in central London cannot but be ‘aware’ that there are plenty of cyclists around but I don’t myself see any evidence that this changes their behaviour. Add to that local authority maintenance cuts, two bad winters, and an increased number of potholes, which may require cyclists to swerve round them – but don’t hold out too much hope that that driver coming up behind you will be ’aware’ that you might need to take avoiding action.

Driver awareness in a cycling environment: a driver in central London gets used to sharing the roads safely with cyclists.

Two thirds of drivers in Britain never cycle, so the idea that a critical mass can be built up sufficient to transform this state of affairs by encouraging more people to take up cycling among motor vehicles strikes me as wishful thinking of the most extreme kind. Especially on urban roads like this one:

Euston Road (A501). Cheer up, cyclists! Cycling on roads like this is now 2.9 times safer than it was in the year 2000. It's a scientific fact.
Or to put it another way, this is what happens when TfL’s number one priority is ‘smoothing traffic flow’. Cyclists are terrorised off the roads, apart from a battle-hardened minority, and non-cyclists have absolutely no wish to ‘catch up with the bicycle’ if it means having to encounter traffic conditions like these. As for pedestrians, they are kept waiting until they are grudgingly permitted to scamper to a cattle pen in the middle, where they are kept waiting until they are finally grudgingly allowed to reach the other side. Meanwhile the air pollution here breaches EU safety standards.

So what about ‘creating more welcoming and cycle-friendly streets’. Well we all want that, don’t we? The problem is one of definition. These photos, taken from the CTC and LCC websites, show what Britain’s two leading cycle campaign organisations mean by ‘cycle-friendly’ streets:

Here once again we come to the vast gulf which exists between orthodox British cycle campaign objectives and the non-cycling masses, because the CTC and the LCC are primarily committed to vehicular cycling solutions of the sort shown here, even though most non-cyclists aren’t interested. The LCC’s photo of best practice contains no cyclists at all and the CTC’s shows just one (marketing isn’t exactly one of the strong points of either organisation).

The CTC’s photo location is not identified on the organisation's website but I know it all too well. It shows Foss Bank in York, approaching the junction with Peasholme Green. To me it’s a truly awful piece of infrastructure on the hellish inner ring road, yet to the CTC this is the kind of cycling infrastructure that will get the masses cycling. (That slip road on the right, incidentally, leads to a massive multi-storey car park.)

This (below) is the same road on the other side of the junction.

Even if you accept the CTC case that it is statistically very, very safe to cycle in York, few non-cyclists are likely to find the prospect of crossing two lanes of fast traffic to access the cycle lane which runs alongside the lorry a particularly inviting one.

In both street scenes shown by the LCC and CTC there is plainly space for safe, segregated cycle paths. But neither organisation has ever shown any serious interest in that kind of infrastructure. Instead reality is turned on its head: the cyclists, it is argued, have to come before you get the infrastructure. Thus Chris Peck says

Conditions need to improve for cycling, but relying on the conditions to be perfect before trying to promote cycling gets it all wrong. We need the cyclists to be there to chase up local authorities and force them to create better facilities, to argue for lower speed limits, better traffic law enforcement and measures to reduce motor traffic.

In other words, more people have to be persuaded to become vehicular cyclists and this will generate sufficient critical mass to get ‘better facilities’. But what kind of facilities? Instead of an explicit commitment to infrastructure with a proven record of success we simply revert to our old friend ‘Hierarchy of Provision’, which might sound fine on paper but which is

intrinsically car-centric and methods of traffic reduction and speed reduction, re-allocation of road space, junction re-design and other infrastructure always translates as ‘pinch points’, converted pavements, lethal on-road cycle lanes that terminate at parking bays and so on.

The third instrument of ‘tackling the fears that prevent people from cycling more or not cycling at all’ is ‘giving people the confidence to cycle more’ – in other words, cycle training.

Finally, let’s scrutinise the notion that building a critical mass for cycling will mean that There is greater political will to improve cycling conditions.

Empirically, that’s not the case. Cambridge has the highest level of cycling in Britain and a very active local campaign group. Yet one local activist rages that we are treated as road vermin blocking motor-traffic progress. And another local activist gave up and decided to go elsewhere.

(Conversely, Norwich now has the highest number of Green councillors of any UK council, yet the city is as car-centric as they come, and the local Green Party’s transport policy, like the national one, is, as far as cycling is concerned, totally vehicular in thrust – read it here and marvel at its ignorance of the kind of cycling infrastructure that really works. Not surprisingly Norwich is a failed cycling city with a risible modal share, which is going nowhere for cyclists.)

Let’s consider the CTC's case for York.

CTC has found that cycling is safer in local authorities in England where cycling levels are high. York, the authority where cycling to work is most common, is, by our calculation, the safest place in England to cycle.

What’s more York Council itself claims that it puts pedestrians first, then cyclists, with car drivers further down the hierarchy of provision.

But if things are going so marvellously in this safest of all British cycling environments, why is it that that Roger Geffen, CTC campaigns director, is now quoting a modal share for York of 12%?

That suggests a one-third drop in modal share. Not that it’s particularly easy to get either year-by-year figures on modal share in York, or even up to date ones, either for commuter cycling or utility cycling. And when local authorities are coy about supplying data it’s often because the statistics don’t match the spin. Either that or they aren't taking cycle counts. If it’s really a ‘cycling city’ there ought to be a trend showing significant growth year after year. There are no obvious signs of one.

Could the failure of cycling to take off in York be related to the massive growth in car ownership over those years, and the reality that York has cycling infrastructure which is pretty much the same as anywhere else? You know, joke cycle lanes like this:

St Leonard’s Place, York. “The emphasis must now be on tackling the fears that prevent people from cycling more or not cycling at all.” (CTC, Safety in Numbers). Yeah, right…

All the evidence suggests that at best cycling in York is stagnating, and at worst it is declining. A very large modal share (in British terms) has not translated into political action to create the kind of cycling infrastructure which will attract non-cyclists. York has plenty of cycle lanes and ASLs, and it even has impressive new dedicated infrastructure such as the Millennium Bridge. But none of this is working.

Cycling is plainly under strain in York. This is remarkable when you realise that the city has a growing student population, and a large student population is always very good for raising cycling modal share statistics (many can’t afford to run cars and universities ban student car parking). York University was established in the 1960s with a student population of 3,000. The expansion of the student population in the 1990s and early 21st century saw this rise to the current figure of 13,908. The university also said student numbers were set to rise by another 5,400 through the current Heslington East expansion

Bear in mind also that there are other educational institutions in the city, including York St John University, which has around 5,600 students.

York is a city where cycling ought to be booming – but it isn’t. What’s happened in York gives the lie to the CTC’s theory that once a significant critical mass has been achieved then cycling will really take off. It won’t – not in conditions of vehicular cycling.

York’s credentials as a cycling and walking friendly city are, to my mind, exaggerated. It’s as car-sick and hostile to cycling as anywhere else, and the CTC isn’t doing cycling any favours by claiming otherwise. And ‘Safety in Numbers’ is just another desperate, doomed strategy to persuade non-cyclists to cycle in traffic. Even if you want to believe the CTC's interpretation of the data it won't change how people regard cycling, because people's impressions of subjective safety are most unlikely to be changed by arguments based on casualty statistics.

Like the man said, everybody’s got themselves a plan. Only problem is

None of this is going anywhere