Putting the brakes on the road toll

The NSW, Victorian and Federal governments will today announce a compulsory driver education program for P-platers. According to early reports, the scheme will require all new drivers to undergo eight one-hour training sessions. Unfortunately, international evidence suggests that it is unlikely to make any impact on the road toll.

Over the past decade, Australia has steadily reduced the fraction of young people killed in road accidents. Yet because other causes of death have fallen too, motor accidents remain the leading cause of death for those aged 15-24. Each year, 1 in 6250 youngsters is killed in a motor vehicle crash, with males three times as likely as females to perish.

Surely then, we should celebrate any attempt to reduce the problem? Sadly, no. While they may sound appealing, driver education programs have been subject to a multitude of rigorous studies over the past two decades, with little to recommend them.

Reviewing these studies last year, the Cochrane Collaboration, an international non-government organisation that systematically evaluates policy interventions, sought out all the randomised trials they could find that assessed post-licence driver education. Combining 19 separate studies, with a total of 300,000 participants, the researchers concluded that there was “no evidence that post-licence driver education is effective in preventing road traffic injuries or crashes”.

The same researchers also looked at pre-licence driver education (typically conducted in high schools), and arrived at an identical conclusion: driver education didn’t work. Indeed, the evidence on these programs is so poor that it prompted the Cochrane Collaboration researchers to write an article in the Lancet, arguing that driver education was “a waste of the scarce resources for road safety”, and publicly calling on the Blair Government to abandon its programs.

Why, then, are Australian governments so keen to try a program that has done so poorly elsewhere? Personal experience, it seems, has played a part. Ironically for a government that recently criticised Opposition Leader Mark Latham for drawing too heavily on his personal story during the election campaign, Deputy Prime Minister John Anderson justifies driver education on the basis that he thinks driver training helped him when he was a young driver. He also notes that car manufacturers support the approach: Not surprising, given that one of the mooted alternatives has been to install speed-limiters in new cars.

To give the governments some credit, the driver education program to be announced today will not be fully implemented immediately. Instead, it will be evaluated through a randomised trial, with the results to be known by 2007. But given the weight of international evidence suggesting that driver education makes no difference to accident rates, it would be surprising if this trial proved a success.

Rather than conducting yet another trial of driver education, state and federal governments would do better to experiment with other ways of cutting the youth road toll. And there is no shortage of alternatives. Several studies have shown that traffic calming devices can work to reduce fatalities, and it would be valuable to know more about what sorts of devices work best. A paper recently published in the Journal of Political Economy showed that across US states, lower speed limits save lives. And natural experiments in New South Wales, Victoria and Queensland have demonstrated that the use of speed cameras reduces accident rates considerably. If we really want to save lives on the road, then evidence, not anecdotes, should drive policy.

The NSW, Victorian and federal governments are to be applauded for using a randomised trial to test an approach for reducing the road toll. It’s just a pity that their proposed solution has failed in just about every other randomised trial so far. As every good driver knows, it doesn’t hurt to occasionally check the rear vision mirror.