The reason competitive markets work is because, over time, we have stumbled on processes which force people to reveal what they really want and how much they want it.
Politicians call it a terminological inexactitude. Civil servants are economical with the truth. Most people call it a lie. When economists discuss the same thing, they describe it as a failure of incentive compatibility. That probably confirms your suspicion about economists. Still, I hope you will read on.
Many of the processes we deal with when we engage in personal or business life are not incentive compatible. In ordinary language, that means it is not sensible for us to tell the truth. When we apply for a job, we exaggerate our capabilities. Once appointed to it, we stress how difficulty it is. Then we can win the approval of our superiors when we achieve them.
Incentive compatibility, in a real sense, is the fundamental problem of economics and management. Why did central economic planning fail, in Russia and elsewhere? We say through lack of incentive, but this is clearly only a fragment of the story. Russia actually had one of the widest range of incentives that any society has ever enjoyed – from the privileges of the politburo for those who conformed most obediently to the horrors of the Gulag for those who resisted. The problem was not incentives as such, it was the inability to devise and define appropriate incentive systems. If the Central Planning Bureau could have obtained all the information it needed – about preferences, resources and production capabilities – then Stalin’s minions could have told everyone what to do. But the prospect of a visit from Stalin’s minions prevented the planners getting the necessary information in the first place. Incentive compatibility is this interlinked problem of information and incentives.
Devising incentive compatible mechanisms is not easy. There is only one copy of a valuable picture: who is to have it? The efficient solution is that it should go to the person who wants it most. But if we ask people how much they want it, everyone will say they want it a lot. We know this when we try to allocate things to our children.
We need to ration the opportunity to say what you want. If we give everyone a limited number of chips, to divide between all the things they might ask for, then they will save up their chips to use on the things they really want. So we could ask everyone to write down how much they would pay for the picture, and give it to the person who set down the highest amount. And if everyone started off with the same amount of money, this would seem to be both fair and efficient.
But incentive compatibility is not so easy. It doesn’t make sense to write down the maximum amount you are really willing to pay. If you do, you will give away all the benefit of acquiring the picture. You do not need to pay as much as you think the picture is worth. You only need to offer to pay a larger amount than anyone else.
So the thoughtful bidder will shade his or her bid down slightly. But so, of course, will everyone else. The amount you reduce your bid will reflect your expectations – not of what others think the picture is worth, but of what they themselves will decide to bid. The outcome is chaos. The picture might go to the person who wanted it most, but only by chance. We have a failure of incentive compatibility. Everyone has an incentive to manipulate the information they provide to gain a strategic advantage.
Some clever economists invented a procedure that gets round this difficulty. Everyone writes down their bid. The picture goes to the highest bidder. But the price you pay is not the amount you bid, but the amount of the second highest bid. Under this scheme, you shouldn’t bid less than the painting is worth to you. That will reduce the probability that you get it, without reducing the amount you will have to pay. But you shouldn’t bid more than the painting is worth either, because then you might end up with it and have to pay more than you really wanted. The best strategy is to tell the truth. The mechanism is incentive compatible.
You may think that this solution is ingenious, but rather arcane and theoretical. Not at all. It is, in fact, the way auctions are conducted in a saleroom. The auction continues until all bidders but one drop out. Then the remaining bidder gets the object – not for what he or she would be willing to pay, but for a fraction more than the amount the second highest bidder was willing to pay. Mr Christie and Mr Sotheby, had discovered the incentive compatible solution long before incentive compatibility had even been invented.
A triumph of practical wisdom over theory, you may think, and in a sense that is right. But a triumph that vindicates the theory, and even shows how necessary it is. The saleroom technique was found to work – the seller got a good price, the buyers were not too often dissatisfied with the outcome – and that is why it drove out most other kinds of auction mechanism.
The theory tells us why it worked. And the theory also tells us that it works only in certain situations. It is incentive compatible for problems like the assignment of the picture, where everyone agrees what the object is but people differ in how much they value it. It is not incentive compatible for the allocation of franchises, where bidders are uncertain about the real value of what they are buying: and that is why attempts to use the saleroom process for that issue have not given good results.
And the evolution of saleroom procedures is a microcosm of what has happened in economic systems more generally. The reason competitive markets work is because, over time, we have stumbled on a series of processes that achieve incentive compatibility.
