If you’re anything like me, you’ll find this situation familiar. You’re arguing with a statist about market provision of some good or service. Your opponent points out the existence of some market failure, suggests that government can improve outcomes, shouts “QED, bitches!” and is overcome by a smug sense of satisfaction at having won the argument.
The usual libertarian response is to suggest that the market failure does not, in fact, exist. Often, this is the right approach. The market is self-correcting in the sense that every market failure provides an incentive for an entrepreneur to solve the problem. Some problems, though, have no solution. Despite enormous economic incentives, nobody has created a perpetual motion machine, for the very good reason that it’s impossible. Likewise, not all externalities can be internalized.
Markets don’t work perfectly, and there are probably even some situations in which a well-crafted piece of government intervention has the potential to improve outcomes on whatever metric of desirability you care to use. Are we, then, being mindless acolytes when we oppose government intervention even in the presence of a market failure with well-reasoned government solution?
While there are some mindless acolytes within any group, there is a very good reason to oppose government intervention even when, in theory, it could enhance freedom and welfare. When someone points to an imperfection in the market and insists that government can solve it, they are implicitly assuming that government works perfectly. By ignoring the possibility of government failure, they are committing the nirvana fallacy.
In his 1969 paper “Information and Efficiency: Another Viewpoint,” the great economist Harold Demsetz criticizes the standard approach to welfare economics, labelling it the “nirvana approach” to institutional analysis. This involves casting a harsh spotlight on the operation of one institution (the market), but taking a naïve and romantic view of the alternative (government). The only sensible response to an imperfect world is to treat both the market and the government as imperfect and ask which performs better; to engage in comparative institutional analysis.
The idea that markets and government need to be compared on an equal basis is one of the central tenets of public choice theory: instead of treating politicians as public-spirited angels, it assumes that people are motivated by roughly the same factors in every sphere of decision-making. Dealing with an imperfect world in this way is also at the core of the George Mason University approach to economics, which has influenced my own thinking a great deal (mad props to my teacher Eric Crampton, a GMU PhD, for introducing me to many of these ideas). Arnold Kling compares the “Masonomics” approach to that of the dominant traditions in economics:
The Chicago school says: “Markets work well. Use the market.”
The more common view, epitomised by MIT, says: “Markets fail. Use government.”
Masonomics says: “Markets fail. Use the market.”
While Demsetz was concerned with exposing the nirvana fallacy in academic economics, I’ve found it an extremely useful tool in general political debate. People will often describe how a perfect government would operate. I sometimes have some pretty serious moral problems with their imagined utopia, but I normally find it easier to admit for the sake of argument that their ideal government is desirable. The next question, of course, is how their government is to come about. A government is itself a spontaneous order whose behavior is determined not by some top-down design but by the interaction of the individuals comprising it. It’s all very well to imagine a perfect government and wish really hard, but, at the end of the day, the emergent government is very unlikely to resemble your blueprint.
Brad Taylor is a graduate student in Political Science at the University of Canterbury in Christchurch, New Zealand. He blogs at 
Have you looked into Austrian Economics http://www.mises.org
There may be some information there that can further your study and research in the field of economics and praxeology. I recommend their literature, but the articles are good too from time to time.
Can you elaborate on how markets don’t work perfectly? I wouldn’t call the failure of people (even under a statist system) to create a perpetual motion machine a market failure. It is more a success of physics! If you don’t mind, please cite the most blatant market failure you know of.
I wasn’t meaning that the failure to create a perpetual motion machine is a market failure, but that some externalities may be impossible to correct via voluntary interaction, just like a perpetual motion machine is impossible to build.
There are plenty of examples of insoluble market failure in the sense of an inefficient level of production of some good: anything involving pollution where transaction costs are sufficiently high to prevent Coasean bargaining. Note that market failure is really only failure relative to an idealized and impossible situation of a all-good, all-powerful, all-knowing government. Philip Booth has said:
“Failure” is a bad word, but it’s one economists are stuck with.
ahhh, thanks for the the clarification.
Bear with me on this one…
What if instead of judging the market against an inconsistent fantasy like an “all-good, all-powerful, all-knowing government,” we judged it against the very best people could do to predict the future (and act accordingly) at any given moment. Since this is what the market brings about, it cannot fail against this high, but realistic standard.
Also, I have heard a good argument by walter block which, by my understanding, says that externalities are a result of improper defense of property rights. The cost of pollution, to continue your example, should be born by the polluter as compensation to the polluted.
Having a term in economics for “markets not working perfectly” is useful. As I said, “failure” is a bad word, but it’s pretty difficult to go against such an entrenched concept. It just seems much easier to say “markets fail, but are superior to any other form of organization” than to insist they don’t really fail.
If there were perfectly-defined property rights and no costs of taking legal action, there would be no externalities. Neither condition holds in practice; and I suspect neither would hold even in a market anarchist society.
What your noting in this post is, I think, the most important idea libertarians need to hammer home with the world at large. Intellectuals throughout the world have come to accept that free markets are the most efficient way to organize a system, provided there is no public goods problem or other externality. That little proviso allows someone like Krugman, who might otherwise be a sharp fellow, to get way off the deep end in his thinking and make ridiculous arguments like government run health care would provide more cost savings than the private sector. Most intellectuals have not had enough exposure to public choice theory, and argue for government intervention, blind to the incentives government actors face every day.