Andy Wood suggests that speed cameras might actually have their uses.

Advocates of speed cameras claim that their presence can reduce accident rates, perhaps by as much as seventy percent. I have little reason to doubt this claim, so let us assume that it is true. Some of the revenue generated by a speed camera goes to the police force which installs and operates it. Unfortunately, the incentives created by the latter fact are not consistent with the goal of minimising accident rates.

Here's why. Crasham Junction is a notorious accident blackspot. Several accidents a year occur here, most of them caused by a small number of drivers who drive too fast. Installing a speed camera would considerably reduce the accident rate. However, local drivers read the local papers and listen to local gossip. They know that they should slow down here. Furthermore, it is obvious to anyone driving through that Crasham Junction is a dangerous location. Sensible drivers slow down; only idiots speed through. A camera located here would only catch a handful of drivers, but it would be effective.

Miss Daisy Drive, on the other hand, is a straight, wide dual carriageway, with railings either side to prevent pedestrians stepping onto the road. The speed limit is only thirty miles per hour, but traffic regularly travels at forty and there has not been an accident for years. A camera on this road would catch dozens of drivers a week, but do very little to prevent accidents.

Suppose you are a police chief responsible for deciding where to locate a camera. You have a limited budget to spend, but some of that will be repaid by the speeding drivers the camera catches. Where should you put it, Crasham Junction or Miss Daisy Drive? Common sense would suggest Crasham Junction. However, a healthy income will look good when your department is audited and I do not believe that police chiefs are ever sacked for accidents caused by ordinary drivers. Miss Daisy Drive might just be too tempting. Indeed, it is widely suspected that this is precisely what many police forces are doing.

I have a solution to this problem. Let's privatise speed cameras.

Here's how it works. Existing cameras are auctioned off to private investors. Potential locations for new cameras are also auctioned off to private investors who may, if they wish, install a camera and paint calibration markings on the road. The camera owner can then operate the camera normally, but instead of demanding fines from speeding drivers, he simply compiles a database of speeding incidents. He funds his business by charging insurance companies a fee to search his database.

An insurance company will be very interested in using this database. On average, it has to pay out considerably more for drivers who speed through Crasham Junction than for those who slow down. If it can distinguish between those drivers, then it can increase its profits by charging low premiums to the careful drivers and high premiums to the careless ones. If you speed through Crasham Junction, you will be fined by your insurance company who will increase your premium next year, or it may fine you immediately by invoking a penalty clause in your policy. The crime of speeding will have effectively been abolished, although the deterence of speeding will remain.

What if an insurance company tries to fine drivers who speed along Miss Daisy Drive? Along this road, speeding drivers are no more likely to cause an accident than slow drivers. The average insurance payout is the same for fast and slow driver alike. If the insurer fines drivers for speeding on Miss Daisy Drive, it will retain the business of slow drivers, but the fast drivers will soon desert it for a rival insurer who ignores speeding incidents on Miss Daisy Drive.

One problem is that insurance companies already take into account speeding convictions when setting premiums and the only advantage of a speeding database would be that they could distinguish between drivers who speed where it is dangerous and those who only speed where it does not matter. Furthermore, the existing speeding laws presumably have some deterrent effect and their effective abolition may increase the accident rate. If so, then the privatisation of speed cameras could be accompanied by an increase in ex post fines and compensation payments to maintain the level of deterrence. Of course, those fines and compensation payments would themselves be insurable risks, thereby enabling the mechanism described above to convert ex post punishments for actual accidents into ex ante punishments for risky behaviours which may cause accidents.

An insurance company will be most interested in fining drivers who speed where it is most dangerous. Therefore they will only pay to search the databases collected by the camera at Crasham Junction. The owner of the camera on Miss Daisy Drive may find it more profitable to remove the camera and erect an advertising hoarding instead. If it is found that the existing speed limits have been badly chosen - that ten miles per hour over the limit is harmless, but twenty is dangerous - the insurer has an incentive to account for this and raise the speed limit accordingly.

It is even possible that we could see different speed limits for different people. If you could demonstrate that you are a particularly skillful driver and can drive safely at high speeds, an insurer could profitably attract your business by offering you a cheap policy which ignores speeding incidents.

I have described how speed cameras may be placed and speed limits set by market forces, with incentives better arranged to minimise accidents. This little idea is part of a much grander scheme to privatise traffic policing in its entirety, about which I hope to write more in the future.