The invisible hand works well under certain conditions: competition, clear property rights, and buyers and sellers who bear the full costs and benefits of their choices and know what they are trading. When these conditions break down, markets can produce poor results even by their own standard of efficiency. Economists call this market failure.

Externalities

An externality is a cost or benefit that falls on people outside a transaction, who did not agree to it. A factory that pollutes a river imposes costs on people downstream that neither the factory nor its customers pay. This is a negative externality, and markets tend to produce too much of such goods, because their price leaves out part of their true cost. A positive externality is a benefit to outsiders: vaccination protects others as well as the person vaccinated. Markets tend to produce too little of these.

Remedies include regulation, taxes and subsidies. In 1920 the economist Arthur Pigou proposed a tax equal to the harm caused, now called a Pigouvian tax, so that polluters face the full cost of their choices. In 1960 Ronald Coase argued that when property rights are clear and bargaining is cheap, the affected parties can sometimes negotiate a solution themselves.

Public goods

A public good is non-rival (one person's use does not reduce what is available to others) and non-excludable (it is hard to stop people who do not pay from benefiting). National defense, flood barriers and street lighting are standard examples. Because people can benefit without paying, each has an incentive to be a free rider, and private markets tend to supply too little. That is the main reason such goods are usually funded by taxes.

Monopoly and market power

When one firm, or a few, dominate a market, they can raise prices above what competition would allow and produce less. Some industries are natural monopolies, where one network (such as water pipes) is far cheaper than several competing ones. Responses include antitrust (competition) laws that block harmful mergers or break up price-fixing cartels, and price regulation of natural monopolies.

Information problems

Markets work best when both sides know what they are trading. When one side knows more, there is asymmetric information. In a famous 1970 paper, George Akerlof showed how, in a used car market, buyers who cannot tell good cars from bad ones (known as lemons) offer only an average price, which drives owners of good cars out of the market. Related problems include adverse selection (those most likely to claim on insurance are most eager to buy it) and moral hazard (people may take more risks once insured).

Government failure

Market failure shows that markets are imperfect, not that government will do better. Government failure occurs when public action produces worse outcomes than the problem it was meant to fix. Public choice economists, led by James Buchanan, argued that politicians, officials and voters respond to incentives too: politicians seek reelection, agencies seek bigger budgets, and small organized groups can win benefits whose costs are spread thinly over everyone. George Stigler described regulatory capture, where agencies come to serve the industries they regulate. Governments also face the knowledge problem Hayek described: they may lack the information to set the right tax or rule.

The practical question is therefore comparative: in a given case, which imperfect arrangement (market, government, community, or some mix) is likely to work best?

Worked example: the mill and the fishing town

A paper mill upstream discharges waste into a river. Downstream, a town's fishing businesses lose about $2 million a year in catches. Cleaning up the waste would cost the mill $1 million a year.

The market outcome. The mill does not pay for the damage, so it has no reason to clean up. Society loses $2 million a year to avoid a $1 million cost: an inefficient result caused by a negative externality.

Option 1: regulation. The government requires the mill to treat its waste. The cleanup happens. Rigid rules, though, may cost more than necessary.

Option 2: a Pigouvian tax. Charge the mill a fee for each ton of waste, set at roughly the harm it does. The mill compares paying the tax with cleaning up, and since cleaning is cheaper, it cleans up, choosing the cheapest method itself. The difficulty is measuring the harm accurately.

Option 3: Coase bargaining. If the law clearly gives the town a right to a clean river, the mill must clean up or pay. If instead the mill has the right to discharge, the town could offer the mill up to $2 million to clean up, and the mill would accept anything over $1 million. Either way, cleanup happens; the rights decide who pays. This works only when few parties are involved and bargaining is cheap.

A government failure risk. Suppose the mill is the region's biggest employer and a major political donor. Regulators may go easy on it, an example of regulatory capture.

The economist's question is not “market or government?” but which arrangement, here, is most likely to get the river cleaned at the lowest total cost.