Scofflaws all?

This article was published in 2012, in Newsletter 100.

Jim Chisholm asked me to write about the public perception of law-breaking among cyclists, how this arises and persists through cognitive biases, and some consequences for campaigners.

Cyclists waiting at a red light.
Image as described adjacent

There’s a general belief in this country that cyclists are a bunch of scofflaws: that we ignore red lights, ride on pavements, endanger pedestrians, ride without lights, and generally have a propensity to commit crimes.

Where does this belief come from? People who hold it think that they got it by observation: we’ve all seen lots of cyclists riding without lights at night, haven’t we? Indeed we have, but are they a representative sample? The problem with trying to infer a general law from observation is that fallible humans like you and me are subject to confirmation bias, the tendency to notice things that confirm our beliefs, and to ignore things that don’t. If you already believe that cyclists are law-breakers, then whenever you see a cyclist without lights this confirms your belief, but cyclists with lights don’t register so strongly. Over time you’ll build up a collection of anecdotes, all of which confirm your belief. This bias is then reinforced by the media, because the stories that get published are chosen to be the most shocking and newsworthy. No newspaper is going to print a story of the form ‘vast majority of cyclists did no harm to anyone today’.

These effects are so strong that they affect members of the Cycling Campaign, whom you might think would be more resistant to bias against cyclists. In a discussion of this phenomenon on the Campaign’s e-mail list one member estimated that the proportion of cyclists who are law-breakers in general is ‘close to 95%’ and another estimated that the proportion of ‘without lights, front and back, would run at about 70%’.

I did a couple of informal surveys on 9 December 2011 to cast some light on the latter claim. On my commute home across north Cambridge, I counted 42 cyclists with two lights and 2 with none (5%); on a second trip through town later in the evening I counted 66 with two lights and 10 with one light or none (13%). This kind of informal survey is very easy to do and I encourage you to try it out to see if your own perceptions of the rate of law-breaking are accurate, and how representative my own surveys were.

For red light running, you might want to look at the Transport for London publication Proportion of Cyclists Who Violate Red Lights in London [1], which reports on a survey that observed the behaviour of 7,502 cyclists at five junctions in London in 2007. I won’t spoil the conclusion here: I suggest that you first make your own estimate of the proportion of cyclists that violated red lights, and then take a look at tables 16 and 17.

If you already believe that cyclists are law-breakers, any cyclist without lights confirms your belief, but cyclists with lights don’t register so strongly

Confirmation bias explains how incorrect beliefs persist, but not how they arise. Here I believe a second cognitive bias is responsible: the outgroup homogeneity effect [2]. This is the tendency to believe that members of an outgroup are similar to each other, much more similar than members of the ingroup. Cyclists are a tiny minority in the UK, so when someone sees a cyclist breaking the law, the most salient group to which they belong is ‘cyclists’ (as opposed to the many other groups, such as ‘men’ or ‘tall people’, to which they might belong), and then the outgroup homogeneity bias leads people to generalise from ‘this cyclist broke the law’ to ‘cyclists in general have a tendency to break the law’. You can see that there is a bias in effect because it doesn’t work for ingroups like motorists; when one driver breaks the speed limit, people don’t think that this means that drivers in general break the law.

A third form of cognitive bias that’s important here is the fundamental attribution error [3]. This is the idea that your behaviour is due to the situation in which you find yourself, but the behaviour of others is due to their character. Thus you might cycle without lights because you forgot to recharge your batteries, and it’s just this once, and you can’t be late for dinner (situational explanation). But you believe that he cycles without lights because he has no respect for the law or other peoples’ safety (dispositional explanation). It’s clear that law-breaking among cyclists is situational if we look at red light jumping. Cyclists don’t jump red lights at random, they jump them when they perceive the road environment makes it safe to do so and when waiting at the lights puts them in the path of motor vehicles. For example, I often go through the Milton Road/Kings Hedges Road junction and I never see red-light jumping there. The junction is too large, the traffic too fast, and there’s no gap in the sequence of lights. But at the Silver Street/Queen’s Road junction there’s an all-red phase for pedestrians in which it’s safe for cyclists to go (as long as they are careful to look for and give way to crossing pedestrians), and you’ll see many do so.

What does this mean for campaigners? Because people don’t get their beliefs about cyclists from accurate observation of the behaviour of cyclists, it follows that we can’t expect to change people’s beliefs by changing our behaviour, or by trying to improve the behaviour of other cyclists. Campaigns along these lines are likely to be futile if the objective is to change attitudes among the general public.

And the situational explanation for law-breaking ought to inform the strategy for dealing with it. Instead of a strategy of enforcement and punishment, can we find some proposal for changing the situation so that the law-breaking disappears by itself? (Just as, to stop wrong-way cycling on one- way streets, we campaign for the streets to be made two-way for cyclists.) For example, at the Silver Street/Queen’s Road junction, why not allow cyclists to go during the all-red phase? Where it’s safe for cyclists to turn left on red, why not put in a way for cyclists to bypass the light and do so?

Sadly, some readers – not you, of course: other readers – are likely to imagine that my arguments are intended to be self-serving, so I had better emphasise that I don’t break the law myself. There are good reasons to obey the law, but persuading the public that cyclists in general are law-abiding is not one of them.

Gareth Rees

[1] Available online at
[2] See Wikipedia:
[3] See Wikipedia: