Saturday, 2 March 2013

Et tu, Brute?

Intuitive paving design
Is this cycle path dangerous?

Remember the Dutch (Copenhagen) study I checked over last week? This week, I had a look at a German study about cycle infrastructure and safety. Headline is "Studie belegt steigende Unfallzahlen durch Radwege" and is quite attention-grabbing, possibly translating it to "Study proves increased accidents caused by cycle paths" gives you the flavour.

It appears to be a before-after comparison study somewhere (!) in Germany, published in 1991 looking at absolute totals only. I'd have thought given its advanced age, it would be dubious practice to use it to inform policy, yet it was quoted to me in support of "cycleways are dangerous" - so here it goes.

Really, it would be good to see the original study, in the absence of that, I make the following comments about the very limited data available. If anyone has the original I am happy to look at it.
  • the number of users is not stated. Imagine cycling had increased (as it can be expected when cycle infrastructure is offered up), then the collision data should be presented in relative terms to that increase, ie would be relatively lower.
  • the collision types are not categorised. This means that - dramatically put - loads of minor collisions (cuts and bruises) would count the same as a fatality. Without that distinction it's unequal accounting.
  • It investigates cycle paths running parallel to the road/highway but we do not know further details on the paths themselves or the junction design. As we all know you really have to look at it in detail, even down to design dimensions, corner radii, road markings etc. Missing that info makes it more difficult to comment. [We are all be experts in the UK! Experts on what crap cycling infra really really looks like. And feels like. Aaargh.]
The first two bullet points in the above list would have certainly been the next logical steps if a full (more meaningful) analysis were carried out. In the absence of that data, or even the study text, I - for one - remain convinced that cycleways are safe.

It's worrying that the website states that "this study unambiguously [eindeutig] proves that cycle paths have higher collisions rates compared to bike traffic on the road". This statement is clearly dramatising and polarising the debate unnecessarily, forgetting the wider view on the data and circumstances.

The title of my blogpost?

Well, the study is quoted on the adfc (German ctc) website - albeit a local branch of theirs - which makes me feel a tiny tad uneasy about their motives. I'd strongly advise the adfc branch to revise the webpage and tone it down, and possibly advise the national adfc to check up on the branches shenanigans.


  1. That sounds a bit like that press release issued by the Association of British Drivers, reported by that newspaper of Record, the Sun, and the BBC no less - "20mph streets increase casualties".

    Again, no context, for example if you double the number of 20mph streets, (and in the relevant timeframe, that was about the size of it)you might expect casualties on 20mph streets to - er, double? Also, no distinction made between KSIs and minor injuries.

  2. Are the ADFC _still_ quoting that stuff ? Remarkable. I do have some sympathies with their predicament because many German cycle-paths are not designed to cope with faster cyclists. i.e. ADFC members. The quality is overall simply too low and the paths are too inconvenient to use. This doesn't affect only faster cyclists, but also others. It's one reason why there is far less cycling in Germany than there is in the Netherlands. However, the ADFC appear to be determinedly looking in the wrong direction for a solution to this problem.

    I wrote about them a couple of years ago when they were pushing the same nonsensical anti-cycle-path propaganda.

    1. The argument of quality, yes. In the UK this argument (and apparently finding it hard to find answers to it) has even stagnated the slightest emergence of cycling infrastructure.