Inactivity, cost-benefit analysis, risk compensation (more John Adams)

In my previous post I outlined a well-known argument about hypermobility presented by the geographer John Adams. He’s a very readable academic whose work has been in areas that interest me and when I first came across him I read through pretty much all of the
and letters available on his website.

Walking-and-cycling advocates will probably already know of Adams because he was the co-author (with Mayer Hillman and John Whitelegg) of one false move; a study of children’s independent mobility. This substantial bit of research is the origin of the figure that, in 1971 80 per cent of seven and eight year old children were allowed to go to school without adult supervision but by 1990, this figure had fallen to 9 per cent.

But he goes waaaayyy back beyond this. One of the earliest (1970) pieces in his archive discusses ‘london’s third airport’. Yes, this airport capacity business has been going on that long (I was only around eleven at the time but I remember that it was all over the newspapers). Adams produced a tongue-in-cheek suggestion that, never mind Foulness or Cublington, the ideal location – the right size and with links already in place – is Hyde Park. Parody is always a bit of a gamble – even prior to the internet. Adams’ reductio ad absurdam of planning priorities was written about in The Times and the newspaper then received and printed a genuine letter from a retired air-vice marshall which heartily applauded such a logical location for a London airport and pointed out that he himself had made the very same recommendation in 1946. Adams relates this in a later essay about cost benefit analysis which is worth reading because this kind of issue is very much still with us. (He takes another, shorter swing at COBA in the article Vogon economics and the hyperspatial bypass)

None of the above is particularly contentious to anyone of a vaguely green disposition, but when he gets on to road safety things start to heat up because he is a seatbelt sceptic. Now that just sounds like a bit of wilful and ridiculous contrarianism, but hang on because this is interesting.

To begin, let’s be clear about what exactly it is he’s doubting. If you are in a car crash, wearing a seatbelt will increase your chance of surviving. Nobody is sceptical of that – it is supported both by evidence and common sense. So if everything else remained equal – same behaviour, same number of accidents, same kind of accidents – then seatbelt legislation really would save lives. Adams is suggesting that everything else did not (and could not) remain equal; daft as it initially sounds, some safety interventions might increase the likelihood of accidents even as they decrease the severity of their consequences.

The argument has two strands. The first is that the aggregate evidence does not support the almost universal assertion that the introduction of compulsory seatbelts has ‘saved lives’. This is quite astonishing because this claim is frequently made, usually quantified and has itself become part of accepted common sense. This part of the Adams’ argument is fairly detailed and before dismissing it you should perhaps read the full account or a more recent and shorter version, because the statistics are persuasive. Broadly, seatbelt legislation was enacted in almost all countries against a background trend of already decreasing road accidents (see fig. one here).

Now I know what you’re thinking, and to be honest I’m feeling slightly embarrassed about the previous paragraph. I might easily be mistaken for one of those shouties who leave comments on blogs that go: “Evolution/Global warming/Environment. It’s all a scam !!! As proved by this long list of badly informed links!!!”. Or worse, a clever-clever contrarian who will present smart-arse arguments against decent things just for the intellectual fun of it. I’ve just waved my hands and said ‘read this’. There’s no reason you should do so. I mean, what do I know? Not very much I’m happy to admit, but a point to note is that Adams’ argument about seatbelt efficacy is not rocket science and only needs some basic experience of statistical thinking to follow. Anyway, let’s argue about this some other time shall we? If we just provisionally assume that Adams has made a case worth considering, how does he account for this?

The second strand of the argument concerns risk compensation. This is the idea that safety benefits have a tendency to be consumed as performance benefits. The full arguments runs:

  • We all have a preferred level of risk – life has uncertainties built in, some people tolerate or even enjoy a higher risk than others and all risks are taken because of a perceived benefit – even a tiny one such as the slightly greater convenience of crossing the road away from an offical pedestrian crossing.
  • Interventions which are intended to make certain activities safer, can sometimes change how we perceive the riskiness of that activity
  • In order to maintain our preferred level of risk we will change our beahviour – for example it feels safer to drive faster with a seatbelt.

Adams calls this the ‘risk thermostat’. There seems to be gradually accumulating experimental evidence in its faviour.

The overall effect of the seatbelt law, claims Adams, is not to reduce danger but to redistribute it:

“The evidence from Britain, which has been singled out as the only jurisdiction in the world in which it is possible to measure fatality changes directly attributable a seat belt law, suggests that the law produced no net saving of lives, but redistributed the burden of risk from those who were already the best protected inside vehicles to those who were the most vulnerable outside vehicles”

The downward trend in accidents which was already present in all countries before the enactment of a law is interesting for a second reason. Almost by definition, ‘developing’ countries came late to the motorization party, so that when their citizens started to acquire cars, the cars which they bought were advanced – developing countries didn’t have to start afresh with the model-T – and came with various technological safety features. So if it were a question of technology alone, these newcomers should start with an accident rate which is close to the already-developed countries. This doesn’t seem to happen, which suggests that there is an element of learning. To be exact, pedestrians and cyclists have to learn to keep out of the way.

“the long-term decrease in death rates .. is much more plausibly attributable to myriad behavioural adjustments in response to the increases in the threat of traffic”

That suggestion is by no means unique to John Adams, but it is a hard one to swallow. Over the years a great deal of effort and ingenuity have gone into safety devices, and it seems almost insulting to say that they haven’t worked as well as hoped. Not only that, but that ‘road safety’ has increased the amount of unfairness in the world. (And btw, if you’re still reeling at the thought of anyone being a seatbelt sceptic, I will stress that the argument is NOT that you personally will be safer without a seatbelt in a context in which everyone else wears one).

It is this type of argument that sits behind the way that some transport thinkers refer to ‘the road safety industry’. I agree that this seems a very shocking way to talk if you don’t know what lies behind it. It’s not that these people are cavalier about making the roads safer, and the term road danger reduction has been coined (by Robert Davis) to capture the idea that taming and reducing motor traffic is the way to do this. In contrast, ‘road safety’ regards motor traffic as a sort of force of nature that our public policy should accomodate and defer to.

Anyway, dear imaginary reader, make of this what you will, but I thought you might find it interesting because this is not the sort of thing you usually get to hear about – even on BBC4. I have a bit more to say about John Adams but I’ll leave that for the next post.