By John Goodman

Most introductory economics textbooks have a section on “market failure.” It is here that students learn that markets may fail to achieve their potential – leaving people worse off than they theoretically could be. The existent of market failure is often taken as an excuse for government intervention to do whatever markets fail to do.

In recent years, however, economists have been increasingly concerned with “government failure.” This is what happens when government policies fail to achieve what they could have achieved – again, leaving people worse off than they theoretically could have been.

What do we mean by the term “market” and what do we mean by “government”?

As a first approximation, I am going to define the market as the sphere of activity where everything is voluntary. All trade is voluntary. All transfers of income and wealth among people are voluntary. I am going to define government as the sphere of activity where everything is coercive. Government puts limits on private behavior and enforces those limits by fines, taxes, imprisonment and even death. Also government may force private citizens to do what they otherwise would not have done – again, under threat of force.

So everything that is voluntary is a market activity. Everything that is coercive is a government activity.

Ideal Markets

The very idea of “failure” implies that we have an idea of what success looks like. In the economic realm, that idea is very well defined. An ideal economic arrangement is one in which people have exhausted all opportunities for mutually beneficial exchange. Once they have done that, there is no way to make one person better off without making someone else worse off. Such a condition is called Pareto Optimality.

Market failure is also well defined. It means failure to achieve Pareto Optimality. Why would that occur? According to Hernando de Soto, if you live in the slums on the outskirts of Lima, Peru and you accumulate any assets, you have to invest resources in protecting your assets, since the Lima police are unlikely to be of much help. Also, since you don’t have access to a court system, you can’t form any long term economic relationships with anyone who you don’t really, really trust.

When government performs a night watchman role, protecting everyone’s person and property, and when it enforces contracts (making people keep their promises), it is doing something people cannot easily do on their own. Put another way, when government protects people from acts of force and fraud it is insuring that all behavior really is voluntary. In these roles, government is enabling and expanding the range of potential voluntary behavior and creating new opportunities for everyone to become better off.

Are there any other reasons why markets might fail to achieve the Pareto Optimal ideal? One is the existence of public goods – defined as a good whose benefits accrue to everyone, regardless of whether or not the consumer contributes to its production. A dam that prevents periodic flooding is an example. A lighthouse that prevents ship wrecks is another. In both cases I benefit even if I contribute nothing. The builders of the dam can’t charge me a price for flooding that doesn’t occur. The builders of the lighthouse can’t charge me for observing its emissions. So my economic incentive is to be a free rider and hope that others chip in and create the public good without my help. Because of this incentive, economics predicts that in a world of purely voluntary behavior public goods will be under produced and private goods will be over produced – relative to the ideal.

Another reason why markets may fail is the existence of externalities – by which we mean important effects that are external to the market. For example, if you and I are ship owners and I build a lighthouse to protect my ships from a dangerous reef, you are able to get an external benefit for which you never paid a price. As the previous paragraph shows, the problem of public goods is a subset of the more general problem of externalities.

There are also negative externalities. Pollution is an example. If l own a factory and emit pollutants that harm you then part of the social cost of my factory’s output is the harm I cause. But since there is not market for the pollutants and I don’t have to pay you for any harm, I will ignore that cost in deciding how much to produce and how to produce it.

In a Pareto Optimal world, we will produce things until the marginal social benefit equals the marginal social cost. But in the presence of these market failures, economics has a clear prediction: positive externalities will be under produced and negative externalities will be over produced.

It’s worth noting that for Milton Friedman, the relief of poverty was a public good — although he might have changed his mind later in life about the advisability of a role for government in this sphere. If we all care about the suffering of the poor and I give money to relieve that suffering, I am not just helping the poor; I am also helping you, the early Friedman reasoned. Without coercion, you and others like you will be tempted to be a free rider on my charity. Thus, purely private, voluntary charity results in too little help for the poor.

The Nobel Prize winning economist Ronald Coase had a way of summarizing all of this that is helpful. Coase said the only reason why there is market failure at all is because of transaction costs. As long as there are well defined property rights, the only thing that can prevent mutual beneficial arrangements is the costs of reaching agreement, including information costs, communication costs, bargaining costs, etc.

Imagine five people stranded on an island and suppose they deal with each other in a completely voluntary way. Is there any reason for them to form a government? None that I can think of. If there is a public good that needs to be produced (digging a well, for example), there is nothing to prevent them from agreeing to produce it and agreeing on how to share the cost. Public goods and externalities become problems only as the number of people grows large — because that’s when transactions costs get large and once they get large enough, we get market failure.

Ideal Government

Until recently, almost all economics textbooks treated government as though it always acted ideally. If there are public goods, we need government to produce them. If there are negative externalities, like pollution, we need government to curtail them. If there are positive externalities, like symphonies and art museums, again we need government to subsidize them. The idea that government might systematically do the wrong thing was not seriously considered.

That attitude is strange, when you stop to think about it. Looking back over the entire 20th century, what’s the worst thing markets ever did? Fail to build a dam or a lighthouse? You might point to an unsafe drug or two or an unsafe automobile and in these cases there may have been a few thousand preventable deaths. What’s the worst thing government did? For starters 170 million people were killed in the last century by their own governments – and that’s not counting the wars!

If that is not bad enough, it is almost self-evident that the principal reason why about 1 billion people in the world are living on the equivalent of $1 or $2 a day is because of bad government policies.  The most urgent need in the world right now is not the need to regulate private economic activity. It is the need to restrain bad government.

What would ideal government look like? It looks just like the ideal economic system. If markets can’t or won’t achieve the ideal, then an ideal government does it. Why do governments fail at this task? They fail for the same reasons markets fail.

Imagine everyone in the Unites States – all 320 million of us — gathering together in a big political convention to sort out what to do about all the public goods and externality problems. That would be impossibly unwieldy. Right? Or to use the language of economics, the transactions costs of getting agreement at such a convention would be almost infinitely high.

Moreover, those transactions costs don’t go down very much just because we adopt a rule like majority voting. In that case we would have to convince 160 million people to agree on something. If we choose a smaller subset of people to represent us in the political arena, the transactions costs don’t really go away. They just become less visible.  Consider the fact that in every election the victory of a candidate is a “public good” for all those who prefer her election. That means everyone on the same side has an incentive to be a free rider. We all benefit if our favorite candidate wins regardless of our contribution to the effort. So in general, we are less likely to vote, to contribute money, to engage in get-out-the vote activities, or even to learn much about what the candidates stand for than if the goal were to secure a private good of equal value.

The same principles apply to legislation. Imagine a bill to change government policy in a way that would be optimal. Such a change is like a public good. Each of us has an incentive to be a free rider, since we benefit from the change whether or not we help get it passed. However, on the other side there are likely to be special interests who will oppose the bill. Although these people also face problems of collective action, for them the defeat of the bill is more like a private good. Their individual stake in the issue is very large, whereas the stake for you and me is very small.

Thus the economic prediction: good laws will be under produced and bad laws will be over produced.

In another context  (read more), I identified the conditions that have to hold in order to get optimal public policy. People can make an effort to influence public policy in many ways. They can vote, contribute money, time and labor. Whatever they do, optimal public policy will result if (1) people have perfect knowledge about how various public policies will personally affect them, (2) candidates have perfect knowledge about what voters want and the level of support they are willing to offer, (3) the various forms of effort (voting, money, labor, etc.) can be easily substituted for each other at market prices, (4) the market for candidates is perfectly competitive and (5) the effort per dollar of expected benefit is the same for every citizen.

That last condition was not a misprint. The effort each of us makes as a percent of any benefit we expect to receive must be equal across all citizens. And if they are not equal, say, because some people are members of organized trade associations that bundle votes and campaign contributions in order to increase their influence while others do not, large welfare losses will result.

Since the conditions for optimality are not even remotely likely to ever be met, that leads to:

Goodman’s Impossibility Theorem: Optimal public policy will never occur.

Now consider the following problem.

Assigning Roles: Government vs. the Market

What sorts of things should we leave to markets and what sorts of things should we subject to government regulation? Since we know that government decision making is inherently faulty, the strong presumption is that things should be left to the market.

Entrepreneurship is the discovery of previously unexploited opportunities to make people better off.  If you like, you can regard every one of those previously unexploited opportunities as a case of market failure. They exist because of imperfect information, imperfect communication etc. Every case of market failure is a potential opportunity for an entrepreneur to get rich. No one gets rich by copying what everyone else is doing. Innovators get rich if they can figure out how to solve previously unsolved problems. And entrepreneurship is a natural force in a dynamic, capitalist economy.

There is no countervailing force in the public sector, however. There can be political entrepreneurs, of course. But there is no natural tendency for them to move us in the direction of a Pareto Optimal world. The reason? Politicians, unlike economic entrepreneurs, don’t gain by sharing in the social surplus they create for others. They gain by increasing their vote total.

Thus, market failure must be large and enduring before it is wise to supplant it with government decision-making. Global warming is one example. National defense is another. But for ordinary markets, we are almost always going to be better off to keep government at bay. And the overwhelming thrust of the economic literature on regulated markets, supports that conclusion – beginning with Gabriel Kolko’s history of the very first federal regulations, enacted in the progressive era.

The Failure to Understand Government Failure

For some reason a lot of people have blinders on when it comes to thinking about government. Here is Paul Krugman, writing in The New York Times:

The emission of carbon dioxide and other greenhouse gases is a classic negative externality — the “biggest market failure the world has ever seen,” in the words of Nicholas Stern, the author of a report on the subject for the British government.

Think about that statement. The world’s 6 billion people, in their role as private economic actors, have failed to curtail carbon emissions by voluntary action. But these same 6 billion people, in their role as public citizens represented by various governments, have also failed to curtail carbon emissions. Why isn’t that described as the “biggest government failure the world has ever seen”?

One hundred years ago many pollution problems were resolved by disputes among private citizens before common law courts. But today, private property rights in the environment are virtually nonexistent. Government has asserted so much control, that successful private action is almost impossible. Environmental problems are government problems, almost by definition. And failure on this score is government failure – in spades.

By the way, my own view on global warming is that the evidence justifies watchful waiting, not radical action – especially in the light of recent temperature readings. If people’s attitudes are revealed by their actions it appears that most people in the world agree with me. But if you disagree and believe that action is needed now, inaction must surely count as a failure of government, not a failure of markets.

Krugman, by the way, often writes about market failure. But if the Google search engine is any guide, he doesn’t seem to know that government failure even exists. This is a common phenomenon on the political left.

In another column Krugman writes that “self-proclaimed libertarians deal with the problem of market failure … by pretending that it doesn’t happen.”  Perhaps. But the delusions of libertarians are more than matched by the delusions of people on the left when it comes to the actions of government.