Yes, institutions do fail, but ...

Let me respond to Matt Nolan's response to my response to his response to me. At least I think that's what I'm doing!!! :-) You can see Matt's response over at TVHE, and it is worth reading and thinking about.

He is right, all institutions do fail to some degree. As I said previously

Clearly markets don't work perfectly, if they did firms would not exist, for example.

But the important issue here is, do markets fail often enough and badly enough to justify government intervention, given that governments also fail. In general I would say no. In the above case, positive transaction costs mean that markets are not always the most efficient way to produce goods, and firms have arisen to deal with this. Could government action have done any better?

Matt says

[...] in reality market failures are also endemic.

But how true is this? If you are a Stiglitz type, then it is totally true, at all times. But by what criteria? As Philip Booth has put it

If I were to give you an engineering lecture and I were to start by saying, correctly I believe, that the maximum theoretical speed of a perfect car was the speed of light and that a car that travelled at any speed lower than that was a 'failed car' or suffered from 'car failure', you would probably think that it was a pretty useless lecture. And you would be right.

But this is what the market failure guys are doing. They say, as markets don't meet the perfect competition standard there is market failure. But perfect competition is a lot like travelling at the speed of light, it just isn't going to happen.

Most people, even those who are not engineers, understand the best way for evaluating cars is that you take two cars and look at different characteristics and evaluate which is best for a particular purpose. But when economists teach about market failure they, instead, suggest policies which governments could, in theory, use to make an imperfect market perfect regardless of whether it is possible, in practice, to improve economic welfare by adopting such policies.

Matt points out a number of reasons why, in theory, markets can fail.

1. Sellers of a product have market power,
2. Buyers and sellers without a voice in the market are adversely/positively influenced by the market,
3. Entry and exit from the market are difficult,
4. Information is asymmetric,
5. Institutional conflict pushes firms/buyers away from rational behaviour.

And he is right, markets will not meet the perfect competition standard for these reasons. But so what? Governments can not meet this standard either. Governments can not deal with these issues in such a way as to get the markets to perfect competition. So the real comparison is between imperfect markets and whatever imperfect outcome governments can create.

As I have argued, for the case of asymmetric information, market solutions to this problem deal with the issue as best as it can be handled. When you look at the empirical literature, some of which I discussed, you see that markets do adjust to the point where asymmetric information doesn't collapse the market.

The reasons that Matt points out above can be seen as necessary reasons for government intervention, but are they sufficient? What will the outcomes of government intervention be? What of the unintended consequences? The kinds of questions we should ask ourselves when thinking about markets failure are: 'given what we know about the imperfections of government, might non-intervention be better than intervening?', 'if we intervene, how do we minimise the possibility of bureaucratic capture?', 'how do we ensure that rights to produce whatever are held by the people who value them most?'. In short, can government intervention in reality, not just in theory, improve on the market solution to a "market failure".