For example, many men enjoy seeing groups of hot girls blasting heavy metal in convertibles. According to Pigou, since the girls don't take the men into account when planning their roadtrips, there are fewer groups of mobile metal girls than would be socially optimal. He would have the government subsidize the roadtrips to compensate.
Coase changed this. His insight was simple — the third person doesn't sit back and let other people make decisions that affect him. He gets involved.
The results of the Coase theorem (see David D. Friedman's explanation, or one of Ronald Coase's original papers) are that
1) when transaction costs are low, the market works optimally, and
2) when transaction costs are high, it doesn't.
According to the Coase theorem, men would pay hot women to blast heavy metal — but they can't, because there are too many men, and we can't organize the payments among ourselves.
I am going to argue in this post that high transaction costs, which are often used to justify government intervention, can be cut more efficiently by private enterprise, without abandoning the optimal outcomes of the market.
Transaction costs are simply the costs of making a transaction. (It tends to be higher when more people are involved, if only because more coordination is required.) If Bob has a good worth $5 to him, but $10 to George, and George has a good worth $5 to him but $10 to Bob, then it seems natural that they'd trade. But if it costs Bob more than $5 worth of effort (or gas, or whatever) to organize the trade, they don't trade. It wouldn't make sense; Bob would lose money. Thus as the transaction costs go up, a trade must be more and more profitable to occur.
This in itself isn't used often used to justify government intervention. The government obviously has similar transaction costs. People have to reach some sort of agreement, pay shipping costs, etc., regardless. The major transaction cost problem used to defend government is a special case — public goods.
Public goods are goods that can't be provided individually — say, national defense, a clean environment, or hot girls driving convertibles. This creates a mass prisoner's dilemma. Everyone wants the public good, but to get it, the cooperation of a large number of people is required. However, each person, individually, would be better off not cooperating. If everyone else contributes, and the good is provided, then the person gets the good regardless of whether he pays. If no one cooperates and the good is not provided, the person is better off not paying. The money would be wasted, since no one else is contributing.
Table-wise, the payoffs for each course of action look like this (the player on top gets the first payoff, the player on the right gets the second):
|Cooperate||2, 2||3, 0|
|Defect||0, 3||1, 1|
No matter the circumstance, each individual is better off not paying. And yet, at the same time, each person would prefer to have the good. This is the problem.
The key to solving the prisoners' dilemma is changing the payoffs. The ideal payoffs look like this:
|Cooperate||2, 2||0, 1|
|Defect||1, 0||0, 0|
In this case, the best choice of each individual is always to cooperate.
Many have assumed that only a government can directly alter the payoffs. However, a mass contract could accomplish the same thing neatly, if it has an enforcement mechanism. For example, a contract could be propagated with these three clauses:
1. The signer will pay a certain fraction of the cost of the public good.
2. The signer will boycott (or punish in some other way) those who don't sign the contract.
3. This contract is only valid after a sufficient amount of people (the number to be determined by the situation) have signed it.
(3. is required to make sure there are enough signers to make the threat of punishment severe enough to discourage defection.)
This kind of contract would replace the prisoners' dilemma with a situation where the most profitable choice is always to cooperate.
If one man signs the contract and no one else does, he loses nothing — the contract is only valid if enough men sign for it to take effect. If one man signs it and many other men also sign it, he gets more headbanging highway hotties. If many other men sign it, but he does not, he does not gain as much because other men boycott him.
This creates a role for a transaction company. Someone is needed to determine the best course of action, collect the payments, provide a way to discriminate between signers and non-signers (so signers know who to boycott), and make sure there is no cheating. The company could take its profit from the payments. I estimate the cost of this would be about ten thousand dollars, depending on the amount of people involved — probably less than current government solutions.
Even if this somehow is more expensive than government solutions, there is still a strong argument for adopting the private system. First, as public goods will be sold on the market, their allocation will be efficient in ways a government can't possibly imitate — supply will finally equal demand. Second, competition will give an incentive for improvements and therefore a constant lowering of transaction costs. The same cannot be said of government solutions, since they require a monopoly by definition. The governmental system allows no competition and has changed little in over two hundred years; the private system would quickly pass it up.
On top of that, the government suffers from its own public goods problems, as demonstrated by Buchanan and Tullock, 1962, Olson, 1965, and Caplan, 2007, among others. In real life, transaction costs aren't cut by government, only reorganized.
This is why there is no government subsidization of hot babes.
(Now if we could only legalize prostitution...)