On libertarianism and public goods

I have noticed an article in Tristan Mills' shared links defending libertarianism from the objection that the free market would undersupply public goods.

If that's all Greek to you, some explanation. 

Shared links - a bloggers tool for sharing articles read elsewhere that might be interesting. My most recent shared links are listed on the right under the imperative read this too. A bit like libdig for one.

Libertarianism - an ideology that holds that a) property is more important than liberty, and b) that everything the state does is evil. At least those seem to be the priorities. If you are interested in libertarianism, I strongly recommend reading this, detailing one liberterian's escape after seeing the flaws in the ideology. In fact it is a more worthwhile read than the rest of this blog post. You can skip the stuff about the minimum wage. You may ask why I show so much interest in such a fringe ideology: it is partly based on that article that I believe libertarians are usually redeemable. (And while the arguments are specifically about the Ayn Rand flavour of libertarianism, most are applicable to saner flavours.)

Public goods  - (wikipedia) goods that we benefit from, like air, science and free software, that we don't use up by enjoying, nor can we be effectively prevented from using.

OK, so what does this article say? There's a certain amount of reference to prior arguments, so pay attention.  I should add that the author, ka1igu1a, appears to be one of the better kinds of libertarian, so I may be making some unfair assumptions.

[quoting] Underproduction of public goods is inevitable in the presence of 1) the ability to free-ride (i.e. non-excludable goods) and 2) rational self-interest.[end quote]

Knapp attacked this position by claiming "underproduction" to be subjective measure, a measure typically proclaimed by fiat by the central planner. Holtz's rejoinder to Knapp was that Knapp didn't understand "higher mathematics" and that economists have in their toolbox mathematical measures of optimality, including such measures of Pareto or Kaldor-Hicks Efficiency.

However, ... In Public Choice theory, the "Redistribution Problem" is recast as "Government Failure." Government Failure more or less holds that public goods will be systematically over-produced. It should be noted that "over-production" violates any algorithmic standard of optimality just as under-production does; and,in fact, the violations can be much worse.

I have used the ellipsis to cut out what is presumably a carefully constructed argument, which I will take for granted. Ka1igu1a's retort boils down to this. Free markets may underproduce public goods (although I'm not necessarily accepting this), but governments overproduce them. This is of course no contradiction. I think it is probably true that markets underproduce public goods and governments overproduce them. This is hardly a compelling argument for the abolition of governments, but it is one for not letting them get too big.

The classic example of a public good supposedly is "national defense." Of course, it's difficult to justify why the US needs to spend roughly a trillion dollars a year on national defense. National Defense is Exhibit A of "government failure." The government failure is not in providing the public good of "defense," but that the overproduction of this public good is orders of magnitude more inefficient than any "market failure" could ever possibly be.

A good example. I agree that the US defence budget is too large. On the other hand I'm not sure that all defence budgets around the world are always too large. There are times, aren't there, when rearmament is a necessary evil? The 30s come to mind.

If you seriously believe that the US military industrial complex, with all it's stockpiles of WMD, is somehow a legitimate consequence of meeting a "public good,"

A US focus is fair enough on a US blog - but how does this argument apply elsewhere in the world? 

Holtz seems unaware that human instrumental rationality, point (2), has been debunked by experimental economics, sociology, and game Theory for some time now.

No no no. If the problem is that it is irrational to produce public goods, it misses the point to assert that people aren't particularly rational anyway. Any rationality will lead to underproduction of public goods and any irrationality will lead to, well it could lead to anything, but there is no way you can predict it would lead to a systematic balancing overproduction of public goods. This is a very sloppy argument.

I should point out that Freedom Democrats utilizes Drupal Open Source Software. According to Holtz, I really shouldn't be even posting on this site, because Open Source Software should have long ago been killed by the non-excludable, free-rider problem. Yes, Open Source Software satisfies the definition of a "public good," however no one seriously would even dare to make a Pareto-optimality argument against it's under-production.  Holtz's "inevitably argument" is empirically debunked.

If only Holtz had said that no public goods would be freely produced, ka1igu1a would have a point, but he didn't and he doesn't. My guess is that production of more, better open source software would be closer to Pareto optimal. Not that I think the state can do much to help.

OK, so where does this leave us. Public goods will be underproduced by free markets. There is a real dearth, I suspect, of very useful information. Trivial information you can put in a magazine or something and sell, but anything really important people will hear about, and you can't copyright the truth. In some respects governments do OK here - they pay for scientific research for example. But I guess if they did the same with software, it would be a disaster.

But drawing to a conlcusion, there seems to be a desperation among libertarians to come to the right conclusions however sloppy the arguments used to get there. To me this suggests that we are not dealing with the real reasons somebody has for supporting libertarianism. It suggests an approach which claims to know the answer before the question has been asked. This is not principled or rational, but arbitrary and dogmatic.