We know that as a rule the probability of a car accident increases with the distance you drive – for some background on these costs see here. There are also of course personal factors – young hormonal males, for example, are more accident prone. A useful efficiency-based reform that has appeared in insurance markets is therefore a personalised policy where premiums also depend on distance travelled. This would reduce the aggregate cost of the community insuring against property and human damages caused by driving by providing incentives for safer driving. DavidP has made me aware of a new US product that goes further than simply distance-based policy. It involves an issued insurance policy that delivers an in-vehicle unit to your automobile (a telematic device) which records not only the distance you travel but when you travel and how hard you break – the latter being a proxy perhaps for both speeding and dangerous tail-gating behaviour. The time at which you travel is highly related to your accident risk – the period midnight to 4am is a high-risk period.
If you drive safely over a 6 month period you get up to a 30% discount on your policy. Thereafter a snapshot of your driving behaviour determines your premium. Your incentives as a driver are to minimise distance travelled, drive at reasonable speeds without being continually on your brake and for driving at safe times of the day. Driving safely reduces the likely accident costs of your car travel and enables the insurance firm to offer you a discount. Accident costs then tend to get concentrated on the riskiest drivers.
It is a neat scheme although one that the private insurance industry may not readily endorse. 70% of car accidents in the US involve another vehicle so that many of the cost-saving benefits of this scheme accrue to insurance companies who do not employ it. There can be a case to regulate to help foster these forms of insurance in ways that internalise these externalities. The issue is not a minor one in Australia where the total costs of car accidents have been estimated to comprise about 4% of GDP. (1840)
One of the frustrating things here is that the pay-per kilometer insurance is not any cheaper than standard cheap insurance, and I drive less than 2K per year. I’d be happy to have that device if it meant it was actualy cheaper for me, although many companies in Aus already discount based on lack of accidents, which I imagine is not too dissimilar to just having a device in your car.
To me the biggest gains have been made thanks to the police and public campaigns. I left Melbourne for around decade before all this started, and came back after it was well established, and the difference in driving attitudes was remarkable.