Car Alarms

Incentives work. Here are some observations and suggestions.

a silent alarm, of the sort sold by LoJack Corp. This one consists of a hidden radio transmitter that is activated (via radio signal) by the police department when you report that your car has disappeared. The transmitter sends out a homing signal that can be picked up by a squad car outfitted with a receiver. LoJack says that its transmitters have yielded 75,000 recoveries of stolen vehicles and that perhaps 1 in 5 cases results in an arrest. The deterrent effect reaches beyond LoJack-equipped cars: Thieves don't know which cars are protected.

Here we have a classic externality. Paying for LoJack installed on a car benefits the owner of the car, but it also benefits other car owners. One incentive is, thieves have to consider a higher likely cost of stealing cars, and change their behavior accordingly. They either go into another line of work, or spend extra time letting a car "cool off" – park it someplace and watch it to see if the police respond. During that time, of course, it's not in a chop shop, and can, in principle, be found even if it's not alarmed.

Insurance companies can calculate incentives, too.

In their Why Not column on page 88, professors Ian Ayres and Barry Nalebuff lament the fact that most insurance companies give only grudging discounts on theft insurance to LoJack buyers. But you shouldn't blame the insurers, who are acting rationally. Much of LoJack's benefit goes to society in the form of crime prevention, not to the owners or insurers of the few cars that have the device.

I don't know if anyone's calculated the relative benefit. Is it 50/50? 70/30 in favor of the owner? Against? If you know the numbers, you can calculate a fair discount.

But William Baldwin has another idea: a government subsidy, of all things!

Costs for the few, benefits for the many. This situation cries out for a government subsidy. But let's be smart about this. Don't pay for each LoJack installed-pay $10,000 for each thief caught. Instead of demanding the BMW back right away, the owner will wait patiently while the police stake it out.

And I suppose if $1000 is given to the police department as a "10% finder's fee", the police will decide car theft is important enough to devote the resources involved in a stake-out to. Of course, we don't want the payment to the police department so high that it encourages the police to find some way to game the system. (How would they do that? I can't think of a way offhand, but then no one's offering me $1000 to come up with one.)

Where do we get the money? From a tax on the other kind of car alarm. For false alarms at 3 a.m., $100 a minute would be about right.

Of course, whenever a tax is proposed, the question of enforcement comes up. I can tell when there's a car alarm going off, but I can't always tell who owns the alarm.

Some cities have installed networks of microphones to detect gunsfire. By comparing intensities across several mikes and triangulating, police can pin down the location of a gunshot to within a few feet. Car alarms may not be as loud, but they last longer, giving such detectors at least the same amount of effective signal.

Here, by the way, is an example of how technology installed for one purpose can wind up being used for another one. Proposed new applications have to be carefully vetted. For example, a system that can listen for car alarms might be sensitive enough to listen in on conversations.

Always trade-offs to consider...