It's time to post up the Theory of the Second Best (http://en.wikipedia.org/wiki/Theory_of_the_second_best http://en.wikipedia.org/wiki/Theory_of_the_second_bes

It's time to post up the Theory of the Second Best (http://en.wikipedia.org/wiki/Theory_of_the_second_best http://en.wikipedia.org/wiki/Theory_of_the_second_best) It's a radical old theory (from 1956), which is still ignored by most of the people who talk about economics. It states, in HN terms, that economic systems can have local optima; and that if the global optimum can't be reached then doing stupid things can actually be smart. In theory a free market is best. But if a market can't be made completely free (for whatever reason) then making the market freer may actually make things worse, because you're moving away from the local optimal point into a trough. When you consider any aspect of the system, you can make it more like a free market, or less like a free market. Maybe the internet libertarians are usually right, when they say "make it more like a free market". Or maybe real systems have been optimised towards that local optimum point, then driven away from it by free market advice, and the internet libertarians are usually wrong. But I don't think there's a hard and fast rule. Whatever the case, imperfect economic systems are a little bit complicated, and there's not always a simple answer. So it's not a free market. Cities subsidise buses. They subsidise cars. Payments for buses are so annoying that they may outweigh the amount of money changing hands. Unions want to protect their turf. Poor people like buses, while rich people like cars. And buses are much more useful when they are full. Its complicated, so figuring out what the locally optimal solution is isn't really easy.