Introduction to Market Failures
Market failures are institutional arrangements which deviate from the neoclassical model of perfect competition. Any market outcome which is undesirable can be referred to as a market failure, although most economists are often disdainful of any use of the term which does not correspond to some deviation from the model of perfect competition. Wikipedia lists seven categories of market failures: Monopolies, Public goods, Natural monopoly, Externalities, Bounded rationality, Information asymmetry and Property right as right of control. Each of these categories falls into one of my categories below.
Monopolies (natural or not) and cartels are cases of market power, which allows firms to price above marginal cost. If firms can collude to reduce the quantity produced, they can drive up the price and their producer surplus (see graph below). In a competitive environment, some firms would defect and produce more, driving the price back down, or new firms would enter the market to capture the higher profits. Thus, “unnatural” monopolies usually require government regulation to prevent either defection or entry. Often, the source of a monopoly is access to a new technology. In this case, the patent system is deliberately constructed to produce monopolies, so the source of the monopoly is not the market, but the government. The “natural” in natural monopoly means that a monopoly is the most efficient way to organize production. If a market organizes itself in the most efficient manner, monopoly or not, it hardly makes sense to call that a failure.
Externalities are the most common type of market failure. They are when a trade between two people affects a third person in a significant way. Situations where there are positive externalities, such as the production of public goods, result in underproduction of the good and situations where there are negative externalities, too much of the good is produced relative to a societal optimum. Production of public goods results in positive externalities for all of society. Consumption of common pool resources results in negative externalities. If the size of the externality can be identified, a Pigouvian tax can be used to get the quanitity produced to the socially optimal level, however, the tax will have distributional effects.
Coase and Property rights
Property rights are a critical part of the functioning of a market system. Poorly defined or enforced property rights result in reduced investment and work effort (since you can’t enjoy the fruits of your labor), and people will feel insecure. I don’t want to get into property rights too much right now, except to make a note about the Coase theorem, which states that if transactions costs are 0 and property rights are clearly defined and enforced, people can bargain to achieve the efficient outcome (internalize the externalities) by bargaining with one another. The cannonical example is two people who are in an office and one of whom is a smoker. Smoking produces negative externalities, but if the smoker can make side payments to the non-smoker, then the level of smoking will be societally optimal. Like Hardy Weinberg evolution model, Coase showed the conditions for an outcome, not that the outcome will always hold. The initial assignment of property rights can have income effects on people and thus affect behavior. So if the default property right is that the smoker is allowed to smoke, the non-smoker will have to pay the smoker to stop. If the default property right is that smoking is not allowed, the smoker will have to pay the non-smoker to smoke. If one person is poorer than the other, that will affect their ability to make a payment and thus impact the overall level of smoking. Transactions costs are never 0 and also affect behavior. The larger the group, the harder it is to organize that group, expecially if the gains from organization are small. Thus, the transactions costs may be asymmetric. One way of thinking about Coase’s theorem is that if transactions costs were low enough, government intervention would not be needed. If transactions costs are so high that such a bargain does not take place, perhaps the intervention would be too costly to be worthwhile. Coase’s theorem is widely misunderstood, but it is worthwhile to spend some time thinking about it.
The Nirvana Fallacy
Humans are neither angels nor gods. The neoclassical model assumes we are all-knowing, immortal and perfectly rational beings. Men are not lightning calculators of pleasure and pain, but are instead caught between alluring hopes and haunting fears. When it turns out that people are people, neoclassical economists cry “market failure!” and rush to replace people with the gods of government, as if somehow government employees were not human. We live in an imperfect world. Ignorance, vice, mistakes and malace surround us, and yet, we continue on. People solve problems by trial and error. Sometimes they come up with things which don’t resemble neoclassical models, but are nonetheless quite effective at getting the job done. Calling informational asymmetry and bounded rationality market failures is an error. They are types of human errors, which will be present under all types of institutions, not just markets. Information is a scarce good. Saying that “we don’t all have infinite information” is a market failure makes about as much sense as lamenting that we don’t have infinite steel, gasoline or food. It’s just a fact of reality.
In conclusion, the concept of market failure is fraught with wishful thinking and lack of consideration of opportunity costs. However, it is always useful to reconsider institutional arrangements and think of ways to improve outcomes. The market is not always the best way to organize economic activity, but it’s usually pretty darn good. Thinking clearly about the mechanisms of market exchange allows one to identify situations, such as externalities, where markets do not work well and need to be modified.