The Carr and Howard Governments have just announced a trial* driver education program for all P-platers, despite evidence from a series of randomised trials that these programs don’t have any statistically significant impact on the road toll. Indeed, some school-based US programs were found to increase the road toll, by encouraging drivers to get their licence earlier. However, it seems unlikely that this will be the effect of the current trial, which doesn’t work through schools. Indeed, a 1982 randomised trial in Melbourne found no negative effect, just a zero effect. My guess is that this will be the impact of the current program.
For more on why randomised trials are the best way to analyse policy, click here. To read a 2001 article in the Lancet, by a group of UK researchers calling for the Blair government to cease a similar program, click here.
UPDATE: I have a little piece on this in today’s Australian, dealing with this meta-analysis. To be honest, I feel a little conflicted about today’s announcement. It’s great to see governments conducting randomised trials of new policy programs, which they do so rarely. But I’m disappointed that they’re testing a program that’s done so poorly overseas, when there are plenty of other alternatives to be trialled.
Of course, if I’m wrong about the efficacy of this program, the randomised trial will show me up — in which case you can schedule me in for a big mea culpa in 2007.
* The original version of this post incorrectly said that the program was compulsory.