Sunday, 18 November 2012

My ride to work - and why the cars resemble little hurricanes

Very few of the runners, stroller-pushing parents, drunks and others who wander into my path each day on the cycle lane up New York’s Hudson River Greenway strike me as especially brilliant people. If they were, they might spot the signs telling them they’re not allowed on the path and stick to the rather pleasant waterside walkway that’s been provided for them.
The runner on the left has, remarkably, chosen the footpath on
the Hudson River Greenway. One can only hope she made
the kind of solid risk assessment the Invisible Visible Man advocates

But, disturbingly, the cycle lane obstructors (CLOs) seem to be better at prioritising the risks facing pedestrians than the head of the New York Police Department’s traffic squad. The CLOs have decided that, if they’re going to obstruct traffic, it’s many, many times safer to get in cyclists’ way than cars’. Having so far seen hundreds – possibly thousands – of CLOs in four months’ cycling in New York, I’ve yet to see one choosing to take his or her chances running down the fast-moving, multi-lane West Side Highway next to the cycle path.

Yet Brian McCarthy, a deputy chief of the NYPD with responsibility for traffic policing, explained in October to WNYC, the public radio station, why its ironically-named Operation Cycle Safe – the programme that focuses police resources on fining cyclists, rather than the road users that cause traffic deaths – was now mainly tackling riding on the sidewalk (pavement, British readers). This behaviour, he reasoned, posed particular dangers to pedestrians. But, while I can find no recent record of a cyclist’s killing a pedestrian on a New York City sidewalk, it is depressingly regular for motorists to do so through excessive speed or lack of attention. The NYPD barely ever prosecutes the motoring offences that lead to such deaths. Its prioritisation is so skewed that, for it to be correct, I should probably be spotting runners every morning taking their chances with speeding SUVs on the highway, away from the terrifying bicycles.

The contrast between the CLOs and Deputy Chief McCarthy illustrates something profound about the vagaries of human beings’ efforts to assess risks. Most humans are reasonably adept at spotting and assessing the most immediate dangers. It’s pretty clear even to a stupid, inconsiderate runner unconcerned about inconveniencing others that it’s less foolish to take the chance of a collision with a cyclist riding at 20mph than a motorist driving at 50mph (the West Side Highway speed limit is 35mph – fun fact that most motorists entirely ignore). Very few people, however, are good at assessing risks in an abstract context such as a decision about how to allocate police resources. The NYPD, as I pointed out in a previous post, hands out 5 per cent of its traffic violation tickets to cyclists, way out of proportion to cyclists’ roughly 1 per cent share of city traffic. But a cyclist last killed someone in New York City in March 2009 – and even that incident wasn’t on the sidewalk. Around 1,000 people have died in the time since at the hands of motor vehicles. A sober risk assessment aimed at bringing down the number of deaths in the city would target an entirely different set of behaviours.

The problem appears to be a widespread one. In the UK, newspaper reporting about the crash that injured Bradley Wiggins – one of my great heroes – ended up producing the same, predictable commentary claiming cyclists endanger pedestrians and, somehow, cars. That’s even though the Tour de France winner’s accident involved his colliding with a car that seems to have pulled out into his path from a filling station without looking. In other words, people started complaining on the basis of the kind of accident that happens quite a lot (one where a negligent motorist injures a cyclist) about things that hardly ever happen – accidents where cyclists hurt other road users.

The question is whether it’s possible to get better to get better at understanding the risks the roads and other places pose - and to start reacting to them more rationally.

A subway station closed for Hurricane Sandy. New York's subway
turned out to be better at assessing risk than many New Yorkers.
My mind’s been particularly drawn to risk assessment in the last three weeks because of events in my new home city. At the end of the week when superstorm Sandy hit New York, I took a trip to Staten Island to report on some of the damage there. Person after person told me they’d ignored the evacuation order for the area on the grounds that a previous hurricane last year – Hurricane Irene – had mostly turned out less damaging than expected. As the sea swept in and inundated their street, they found themselves cowering in their homes’ upper storeys, with waves lapping at the windows even there. A person died in the basement of one of the houses in the street I visited, electrocuted when he stepped into flood water that had electric current flowing through it. The storm was on an entirely different scale from Hurricane Irene.

That miscalculation, it seems to me, is of a piece with people’s miscalculations about road use. In both situations, people rely far too much on personal experience and the evidence of their own senses. In the run-up to the storm, it was clear that the city was very windy but not immediately obvious that the storm would cause historic levels of damage. To realise that, one had to pay attention to something one couldn’t see – the vast storm surge that forecasters were predicting that was preparing to push its way into New York Bay to flood large tracts of the city.

Many people assessing the risks of cycling, meanwhile, look at cyclists in traffic and conclude that the slender, unprotected machines among the big metal boxes are more vulnerable than most figures actually suggest them to be. They see a cyclist on a sidewalk and assume that his greater speed compared with the pedestrians makes him a significant risk to them. It only adds, it seems to me, to people’s irritation with cyclists that bikes are quiet and people tend to notice them only at the last minute. This seems certain to trigger the kind of last-minute, fight-or-flight response that must have been useful for vulnerable cavemen. It’s a far less reliable indicator than people think of the risks around in a complex, modern urban streetscape.

Most people have particular lacunae when it comes to rare events that pose catastrophic risks. A hurricane is precisely such an event. There are fine, hard-to-discern differences between a hurricane that will do little damage – as Hurricane Irene did – kill scores of people and do billions of dollars of damage – as superstorm Sandy did – and one that will kill many hundreds – as Hurricane Katrina did. It is far beyond a normal person’s gut instinct to discern which storms need special attention, yet ordinary people continue to use their gut instincts to assess how they should react.
The motorists are blocking the bike lane. The cyclist's too close to the cars.
And they'll probably all get away with the risks they're taking.

In a road context, the catastrophic events are crashes involving cars, buses and trucks. It’s so common to see motorists driving while ‘phoning, speeding, giving cyclists too little room and so forth that it’s easy to conclude that these behaviours are trivial and pose little risk. Many people compare them with the alarming experience of finding a cycle messenger rush past their nose on a pedestrian crossing and conclude that it’s the cyclist who poses the real danger. The truth is that each time a motorist drives too fast, drives while distracted or turns without looking for cyclists he or she is involved in profoundly risky behaviour – and it’s only the luck of the particular circumstances that divides the outcome from nothing at all and a multiple-fatality crash.

People are still more blind to their actions’ long-term consequences. Almost no-one sees a cyclist on a busy street and remembers that he or she is far less likely to die of heart disease or diabetes in several decades than the neighbouring motorist cursing him or her from the exercise-free cocoon of his or her motor car. The global warming that may be making hurricanes more frequent is a still more remote such risk.

None of this is to excuse stupid behaviour by cyclists. It’s worth stopping for red lights, giving pedestrians plenty of space and going the right way up one-way streets. It shows an example of good practice and avoids annoying one’s fellow citizens. I do my best to obey the rules.

But industries prone to catastrophic but rare risks – the nuclear power industry, for example, or railways – tackle them by looking for the near-misses and minor accidents that suggest people are indulging in risky behaviour. Police forces that hold off warning motorists about dangerous behaviour and minor accidents explicitly miss the chance to follow such a policy.

I at least can do my part. I accept that runners will bound into my path on the Hudson River Greenway and cycle at a speed and on a line that means I should avoid them. I know that cars sometimes sweep across even some lightly-used crossings on the route from angles that are hard to see. So I stop for those red lights as other cyclists and runners speed past me, no doubt thinking me an over-cautious worrier. I do my best to assess when I’ve done something foolish and to avoid repeating the mistake. I can only hope that, as New York and other cities start to grapple with a future made more complicated by extreme weather, far more of those around me start to give the matter the same consideration too.


  1. It is called "the safety pyramid."

    1. Steve A,

      Thank you for your comment. I was bearing in mind as I wrote some of your excellent past posts on risk assessment and the safety pyramid. Do you agree current arrangements fail utterly to capture the risk precurors on roads?


  2. Invisible, while we can pray and work for others to show more care and consideration, on a daily basis, we can usually only change our own actions. As for road risks, my own observation is that falls and injuries occur far more often per mile on paths. OTOH, my last fall on a road was pretty serious. I counsel not riding into a freshly dug trench in predawn darkness...

    1. The trench in the pre-dawn darkness sounds nasty. I hope you've got better.

      I generally agree that paths are a poor idea - numerous things can go wrong, both with the surface and at intersections. Along the Hudson River, however, the bike path has an excellent surface and West Street is essentially an urban freeway. So the path is much the best means of getting up Manhattan. The path is also the only way to cycle over any East River bridge. So around five-and-a-half of my daily nine miles to the office are on segregated paths.


    2. I cycle daily in Copenhagen - a bike path paradise. I know for a fact that bike paths are a lot safer than riding among the cars - given, of cvourse, that there are not unguarded (and unlit) trenches on the bike paths (ouch! :(). The safety is on average on a 4:1 ratio here compared to the USA and GB. In the Netherlands, it's 5:1. American and British cyclist of the VC kind tend to overestimate the dangers of intersections in comparison to being run down from behind. Probably for much the same reasons as described by Mr. Invisible above.

    3. I beg to differ with "Anonymous." American and British cyclists of the VC kind tend to base their estimates of relative dangers on actual data. Available data clearly show that only a minuscule percentage of car-bike crashes are hits from behind. The vast majority happen where cars' and bikes' paths cross, i.e. at intersections, and it should be easy to understand why this is so.

      Regarding safety ratios (assuming that means fatalities per km traveled): Yes, Copenhagen and Netherlands are safer than the U.S. But that is due to far more than simply bike paths. There are immense differences in culture, terrain, liability laws, licensing requirements, city density, motoring costs, etc. Furthermore, even if U.S. cycling is more dangerous than Copenhagen cycling, it is far from dangerous in any absolute sense. American cyclists average perhaps 10 million miles between fatalities, and competent American cyclists ride over 30,000 miles between crashes of consequence, most of which are simple falls.

      Cycling, including in the U.S., is NOT very dangerous. It does us no good to pretend it is.

    4. Frank,

      Thank you so much for your comment. I'll get onto the substance - but I have to mention first that I was moved nearly to tears to have a commenter (a) correctly recognise the word "data" is a plural and (b) actually use proper statistics to make his point. You are, sir, a rare gem for both of those things.

      As for the substance, you are, I think, largely right about Copenhagen. As I've said before on here (, the head of Copenhagen city traffic once told me they put in segregated cycle lanes because people liked them. They didn't reduce accidents, merely rearranged them so that they happened at intersections rather than spaced evenly along the street. That fits with my experience of riding on segregated paths. They are very vulnerable to sideswipe accidents at intersections.

      As I said above, however, my view has changed a little since I moved to New York. Many of the main roads here are very, very hostile places to cycle. I positively seek out those with segregated paths if only because they give motorists at least some sense that they should respect people cycling in them.

      Have a great evening with your excellent grammar,


  3. Thanks for this post. It was great. I've been trying to post something for a while about bike riding and risk management, but you've said it much better than me.

    For what its worth, I've been trying to work out why as a pedestrian, I'm completely happy to ignore the rules, I never break them as a motorist, and I obey them most of the time as a cyclist. For me the answer is about the risk to other people, not to myself. I (foolishly perhaps) think I can work out my own risks as a pedestrian, much better than any road designer can.

    As a pedestrian, I have been knocked over by a bike (in a pedestrian only area), and it wasn't fun. But it was probably just as bad for the bike rider as me - so the risk assessment seems moderately neutral.

    1. Jennifer,

      Thank you so much for your comment. I've been knocked off my bike by another bike. It hurt - but here are two things about it. First of all, the guy who knocked me off also got hurt. It's nothing like as asymmetric as with a car. Secondly, it hurt a lot less being hit by a cyclist going full pelt (through a red light, as it happens) than suffering a glancing blow from a car. Moral? Simple physics works.


    2. I read a comment recently that may help to understand why traffic planners in the US and UK have a difficult time with working out the cyclist-pedestrian-car infrastructure. They typically start from the premise that bicycles are simply 'slow cars' while in Holland and other countries with built out cycling infrastructure they look at cyclists as 'fast pedestrians'.

      This observation manifests itself in many ways, from how we protect cyclists on busy streets to the way pedestrians react to cyclists riding through crosswalks. It even goes to the helmet debate - do you wear a helmet while walking? But you do have safety equipment in a car...


Please feel free to leave civilised comments - positive or negative - here. I'll try to reply too.

Abusive comments will be moderated out and won't appear.