Broad indicators are essential

Focusing on narrow indicators is easy, but rarely of great value. I've blogged before about the Mickey Mouse micro-targets (the length of hospital waiting lists for example). Similarly, if a government were to target accidents to cyclists, the evidence appears to be that risk compensation would occur:

In one experiment, a British psychologist, Ian Walker of Bath University, simply got on his bicycle and monitored the behaviour of 2,300 vehicles that overtook him. When he wore a helmet, drivers were much more likely to zoom past him with little room to spare; when he was bare-headed (and indeed when he wore a female wig) the amount of space that motorists left would increase. A hazardous comparison, The Economist, 28 February

But even targeting road deaths in general wouldn't necessarily improve things:

An experiment in Munich found that the drivers of taxicabs fitted with anti-lock braking systems were involved in no fewer accidents than those without. That is because the former used those superior brakes not to practise prudence but to drive more aggressively. ... John Adams, a transport expert at University College London, has compiled data from all over the world to show that laws making drivers wear seatbelts do not make roads safer; they move deaths from inside cars to outside them because they encourage bad driving.

What about aiming to reduce the number of children killed on the roads? John Adams points out that this:

has fallen in recent years ... but mainly because they are rarely allowed out alone, so today's teenagers have less skill at navigating hazardous roads; and as a result, the number of teenagers killed in car accidents has jumped.

Sometimes the difficulty of picking appropriate indicators for targeting under a Social Policy Bond regime is pointed out to me. But exactly the same problems apply to policymaking under the current regime - except that there is rarely any transparency about what is being targeted. And that leaves the way open for expensive (and mostly futile) responses to the relatively small number of deaths caused by terrorism, while an estimated 1.2 million people are killed on the world's roads every year.