General Equilibrium, Market Failure, and Policy


Public Goods

The Efficiency of the Ideal Self-Regulating Market

To what extent should free market forces be relied upon to govern economic life? Even Adam Smith recognised that markets themselves required the protection of the state. Without some government activity -- to protect the nation from external enemies, to ensure the safety of persons and the security of their property, to enforce contracts --  the market mechanism could not be expected to function effectively. But is there any reason to think government might have to intervene in anything beyond some minimum level of support for the system overall? Once the area of free market control had been identified, could it be left free of all intervention by the state? 

Until the 1920s mainstream neo-classical economic theorists seemed to assume that this was so -- that most production and distribution could be entrusted to free competitive market forces with socially beneficial results. For reasons we looked at in the previous topic, Marxists and socialist theorists of course dissented. 

Even within the mainstream, however, there was some division with respect to the problem of monopoly, which was eventually to result in the development of the imperfect competition (monopolistic competition) extensions of the theory of the firm, the theory of oligopoly, and other modifications in the treatment of price and output decision-making. 

But such modifications left several issues unresolved.  Was it possible that, even with competition, the pursuit of self-interest by individual producers might not result in the optimum use of resources for society as a whole? If such a possibility could be demonstrated, the case for free markets and a minimum economic role for government would be questionable.

So far in this course,  the operation of individual markets has been studied under highly restrictive assumptions. Whenever something has been allowed to change, say, demand or the price of a good, the ceteris paribus assumption has been made (often implicitly). It has also been assumed that markets are independent of one another. Events in one market have not been allowed to influence events in other markets. Following the practice popularized by Marshall, this is called "partial equilibrium" analysis, because the equilibrium conditions and states under consideration have been confined to particular markets. The analysis of economic events becomes much more complex when the more realistic possibility is introduced whereby particular events can have repercussions outside the particular market in which they originate. A change in consumer demand for cabbage, for example, will influence not only the price of cabbage, but the incomes of cabbage producers, the demand for land suitable for growing cabbage, the demand for the kind of labour needed to produce cabbage, and on and on throughout the economy. 

In much of the practice of day-to-day economic analysis economists seldom have to concern themselves with these complexities because normally it is only the immediate impacts of changes in particular markets that warrant study. Only occasionally do events, such as a huge increase in energy prices, require a broader approach. But when they do, some serious difficulties with the standard analysis present themselves.

Studies of the system-wide impacts of economic events, called for obvious enough reasons "general equilibrium analysis", are too complex to be treated here, but it is possible to outline the broad features of how a market economy would look when in a "general equilibrium" position. 

Suppose all the individual, particular markets for goods and factors of production were in the state of equilibrium as described earlier in Topics 3 through 7. That is to say, in each of these markets the price is such that the quantities demanded exactly equal the quantities supplied. This requires imagining a situation in which every possible transaction, every possible trade of money for a good, of one good for another good, of work for leisure, of sacrifice of current consumption for the interest earned on investable funds, has been effected. All markets are in equilibrium. Note that does not imply that everyone has everything he or she may want. It does mean that given the constraints of income, given individual preferences, and given all the alternative opportunity costs, the system is in such a state that no trade or exchange can be made which would improve on what has come to pass.

If such a state of affairs can be imagined, the best possible allocation of scarce economic resources among alternative uses has been achieved and the total welfare of the community has been maximised. To look at it from another perspective, any change—such as a reduction in the output of one good in favour of another—would reduce total satisfaction. This state or condition is called Pareto optimality after the Italian engineer and theorist, Vilfredo Pareto (1848-1923), who gave the first precise statement of the concept. 

Do not confuse this optimum allocation of resources with any ethical concept of justice or fairness. Under the condition of Pareto optimality, it is impossible to improve the well-being of one person without reducing the well-being of at least one other person. 

Take an extreme example to illustrate the principle. Imagine a very small economy consisting of only two people, one of whom, a vicious tyrant, owns all the productive resources and receives virtually all the income generated from employing them. The other person, a slave, gets virtually nothing. If the United States or some other great power took an interest in this situation and required the tyrant to hand over some of her income to the slave, the tyrant would be worse off. Even if the slave is better off, the condition of Pareto optimality has been violated. The result may or may not be more "just", but it is certainly less "efficient". The strong case made for a free market system is that if all markets work as described in the preceding analysis of perfectly competitive product markets and marginal productivity factor pricing, the outcome will be Pareto optimal.

But, of course, all markets do not work as described in the perfect competition (equilibrium where average revenue equals marginal cost), marginal productivity factor payment (value of the marginal revenue product equals factor reward) cases which have been studied. Most real-world product markets are not perfectly competitive, but oligopolistic, monopolistic, or some variant thereof. Most factor markets are also less than perfectly competitive because of employer power, union power, and other deviations from the ideal. 

If such elements of monopoly exist, there can be doubts about prices being "correct"  and the correctness, from the standpoint of Pareto efficiency, of the allocation of resources among alternative uses under a market system. 

Worse still, there are some other important  reasons to suspect markets may not always operate in the ideal manner, two of which we will try to outline here. 

To Top


Externalities are costs or benefits arising from production or consumption which are not recognised by markets and which are consequently not taken into consideration by producers or consumers. Examples of externalities given in the older economics literature have a quaint quality which leaves the impression that they were thought of mainly as technical curiosities. Most famous perhaps is the example of how beekeepers provide orchard owners with free pollination services as a side effect of their own operations. In this case apiarists are producing a positive external benefit for orchardists, but the market presumably fails to recognise this benefit and so the apiarist cannot collect a price for these services from the orchard owner.

Today, because of the attention given to the issue of environmental pollution, externalities, particularly negative ones, have come to be recognised as major economic and social problems. Practical examples abound. If all the costs of manufacturing and operating motor vehicles were included in the prices actually paid by consumers to buy and operate them, far fewer motor vehicles would be demanded and supplied than is currently the case. The companies that mine the iron ore, the steel mills that process it, the assembly plants operated by the manufacturers are all allowed to dump wastes into the environment free of charge. Therefore the prices paid for cars are substantially lower than if the real costs incurred by society at large were added to the costs recognised by markets. The latter are often called "private costs" because they are borne by individual producers or consumers. The private cost of a new car is what the consumer has to pay for it. The real or social cost of a new car is that private cost plus all the costs which are not charged for, such as the damage to the environment. 

If prices fail to reflect the true costs (or benefits) of production, the signals received by producers and consumers are misleading and cause more or less than the optimum ("efficient") quantities of goods and services to be produced and consumed. The Pareto optimum general equilibrium referred to above is not attained and the efficiency of the market system is impaired, perhaps seriously. Why could such a condition come about? What is the root cause of markets failing to recognise these external costs and benefits?

This basic question was raised in the 1920s, not by a critic of market theory, but by one of its most distinguished proponents. The British economist Arthur Pigou was the successor to Marshall at Cambridge. His most famous student was John Maynard Keynes. His career and his work consequently bridge a dramatic stage in the evolution of modern economic thought. His theoretical work can be interpreted as representing either the high-point in the development of neo-classical theory (Keynes was to use his work as his target when he launched his devastating attack on neo-classical thought in the late 1930s) or as the beginning of its undoing.

The latter interpretation arises from Pigou’s work on what we today know as welfare economics. As early as 1912 Pigou had published a book, Wealth and Welfare, in which he raised the possibility that even competitive private markets may lead to outcomes which were wasteful of resources and undesirable from a social, as distinguished from a private point of view. His analysis was fully developed in The Economics of Welfare published in 1920. The basic ideas are set out in non-technical language in the following passage from one of his published lectures: sometimes happens that only a portion of the benefit or damage due to a person’s private action is reflected in the reward that that person receives; and, consequently, that he tends to carry that action less far or further than the general interest of society requires. For example, when a man under competitive conditions invests money in a concern for manufacturing and selling alcoholic drinks, it pays him to carry his investment up to the point at which the cost of providing the last unit of service balances the price that consumers are ready to pay. But as an indirect consequence of his action the Government is forced to expend more money on the police force than it would otherwise need to do. From the social point of view this is a part of the cost of his undertaking; but, since he does not pay for the extra police, from his private point of view it is not part of it. The social cost of his investment, at the margin, being thus greater than the private cost, it is not in the social interest that it should be carried so far as the unfettered pursuit of private interest tends to carry it. Again, if a person is interested in breeding rabbits, he will invest in these creatures up to the point at which marginal cost to him balances marginal return: but he will (except insofar as he is a philanthropist) ignore the damage that they do to his neighbours’ crops when they accidentally escape into their fields. From the point of view of him and his neighbours combined, this investment ought not to be carried so far as it pays him personally to carry it. On the other side there are certain sorts of investments that, at the margin, yield a return to other people in excess of what is reflected in the reward of the investor. The introduction of devices to prevent one’s factory chimney from smoking reduces the washing bills of one’s neighbours. Investment designed to promote mechanical improvements and fundamental scientific discoveries often yields a return to the world, which, even when rights in the new ideas are temporarily reserved by patents, enormously exceeds the private return to the investor. It is in the social interest that investments of this class should be pushed further than the unfettered pursuit of private interest will tend to push them. Plainly ... there is a prima facie case for State intervention. The specific form of such intervention recommended by Pigou was for the government to either impose taxes on firms in this situation so as to reduce the level of output of a good to the socially optimum level or pay subsidies to the firm so as to raise output to the desired level, depending on whether the externalities involved were negative or positive. 

Since only the government could impose such arrangements on private firms to bring about the reconciliation of private and social marginal costs and benefits, Pigou's approach implied possibly far-reaching government intervention in the economy. Pigou did not shy away from conceding this. Unlike his predecessors, Pigou took the view that the success of market economies was in large measure attributable to government activity! Adam Smith’s invisible hand worked, he contended,  because of the institutions which had been developed to ensure that individuals pursuing their personal interests would generate socially-beneficial outcomes. Consequently, the task of economics, according to Pigou, was to identify the areas in which government action was required, and to discover "some of the ways in which it now is, or eventually may become, feasible for governments to control the play of economic forces in such ways as to promote the economic welfare ...of their citizens as a whole."

Pigou gave attention to the kinds of problems which have only recently begun to receive serious public attention — pollution and the deterioration of the environment due to economic activity. He noted that uncontrolled private enterprise generated serious costs which had to be borne by the general society. Urban congestion, polluted air and water, the loss of amenities such as open space, quiet, and privacy were all problems which he believed had to be addressed through government measures which would force private investors to bear the costs of such consequences of their actions. He was aware, however, that there were many obstacles in the way which would make effective government action difficult to bring about. Pigou was highly sceptical of the motives and abilities of politicians and he suspected that much of the knowledge which was being accumulated about economics would ultimately end up being ignored.

Pigou’s reinterpretation of the role of government action in a market economy was in many ways revolutionary, yet the full implications of his approach were obscured by a far more revolutionary attack launched against the neo-classical edifice by his student, John Maynard Keynes, in The General Theory of Employment, Interest and Money in 1936. Paradoxically, it fell to Pigou as the reigning custodian of neo-classical orthodoxy to denounce Keynes for his heresy! This, however, belongs to a later topic in the course. Here we can note that Pigou’s own deviation from orthodoxy by proposing a more substantial role for government was criticised by an American economist, Frank H. Knight (1885–1972) who raised an interesting question about why it was needed.

Knight, who is best known for his book Risk, Uncertainty and Profit, published in 1921, was one of the founders of a powerful conservative strain of neo-classical economics which is still very much alive and based at the University of Chicago. In a paper published in the Quarterly Journal of Economics in 1924, Knight presented a devastating criticism of Pigou’s analysis of externalities. The reason Pigou’s examples of externalities led to the conclusion that government intervention in the form of taxes or subsidies was required to achieve a socially efficient allocation of resources under competitive conditions, Knight argued, was because government had failed to carry out one of its (limited) responsibilities in the first place. It was well accepted that one of the functions of government in a market-based, private-enterprise economy was to assign and preserve the rights of individuals to own property. If property rights are clearly defined, Knight’s analysis suggested, individuals who maximise their returns from their property will charge prices which reflect the full social costs and benefits of supplying whatever it is they are producing to their customers, just as the original neo-classical competitive model had claimed. This idea, tersely embedded in what appears to be a narrowly technical quibble in his 1924 article, was subsequently developed into a major component of modern economic theory by Ronald Coase and James M. Buchanan. Their work established the basis for an important body of policy prescriptions relating to the treatment of pollution, traffic congestion and other serious types of externalities with which we are all concerned today. Why do firms expect to get away with discharging wastes into a river which will damage the fishery downstream? Because the river is "common property", and as such there is no one to enforce a right to it. 

As we have seen Pigou’s solution would be to tax the polluter, thereby raising the polluting firm’s marginal private costs to the level of public cost.

Consider the case of an industry which produces recycled newsprint. The usual inverse relationship exists between the price of its product and the quantity it can sell. It also experiences rising unit costs as it expands its output. (These are the market costs it incurs to obtain old newspapers, grind them up, extract the ink, and make new paper.) Suppose that under given demand and supply conditions the industry is in equilibrium at a price $200 per tonne, with 150 tonnes being produced. Suppose, however, that there is a negative externality associated with this production. De-inking results in the creation of a nasty toxic sludge which is harmful to the environment. Assume that this externality could be valued at $75 dollars per tonne. If the external cost is combined with the internal costs the total costs are increased by $75 per tonne and the equilibrium output falls, say, from 150 tonnes at a price of $200 to 100 tonnes at a price of $250. Obviously it would be desirable from the standpoint of economic efficiency to force the industry to internalise the negative externality. This could be done by the government imposing a tax on the industry’s output, a tax in the amount of $75 per tonne. 

Following the argument of Knight, an alternative solution would be to vest property rights in the river to some individual, firm or other agency. Suppose the owners of the fishery are chosen to be owners. They could then force the polluter to either invest in the necessary anti-pollution equipment to control their discharges or pay appropriate compensation. The result would be the same, private costs would rise to the socially appropriate level. 

Which approach would be most feasible could depend upon the institutional structures involved. In this particular case, it is relevant that in much of western Europe private ownership of forests and even small rivers ("trout streams" in Scotland) is a familiar arrangement, but in North America it is uncommon. There, the idea that natural resources "belong to the people" is deeply entrenched. There are also practical problems arising from complex interactions of cause and effect, as one party’s actions may have repercussions far removed from any one geographic location. Over the length of a major river system this could entail possibly endless research, analysis, and litigation to determine liability for damages. Such practical problems are likely to be even more troublesome in the cases of air and noise pollution — to say nothing of the matter of ugly buildings!

I may well be, then, that if private ownership is impractical, government intervention in some form appears to be the only recourse. Public ownership is one possibility. In a sense all governments "own" the resources which have not been alienated to private owners in the areas under their jurisdiction, acting as custodians of the public’s assets. In some countries, for example, much of the forested land is considered public property. Private companies may be allowed to harvest trees on such land, but they are charged a price for this privilege. This price could, theoretically, be set at such a level as to capture the costs of fire protection, reforestation and other costs associated with the forestry operations. Similarly, if rivers were treated the same way, it is conceivable that the government could charge polluters for the right to dump waste into them. In the case of air pollution, firms emitting wastes into the atmosphere could be charged prices which would in effect compensate society for the reduction in air quality. 

Economists have led the way in trying to persuade governments to adopt such measures, proposing schemes whereby markets would be created for the buying and selling of pollution rights. Auctioning off the right to dump specified quantities of sulphur dioxide into the air would mean that the costs of such a reduction in air quality would be incorporated into the costs of producing steel or other products which require such practices. This "internalising" of costs would mean that the prices of automobiles, for example, would more accurately reflect what it costs society to produce them. More expensive cars might mean fewer cars and perhaps more bicycles would be produced.

A more specific form of public ownership may also be considered as a means of dealing with externalities. Publicly-owned enterprises might be established as alternatives to private firms on the grounds that they could more easily be required to operate in such a way as to reduce the effect of externalities. A publicly-owned electric-power generating firm might be instructed to price power at a level reflecting estimated social rather than actual private costs of production.

In some countries, notably the United States, a preferred alternative to outright public ownership in such situations has been government regulation of private firms. Such regulation can take many different forms. At the local level, it is common to find governments imposing regulations which prevent certain kinds of industries becoming established in particular areas. Zoning regulations in cities may forbid commercial enterprises in residential neighbourhoods, or may prohibit construction firms from operating noisy machinery at certain times of the day. At state or national levels regulations may be imposed to limit the quantity of air and water pollution caused by private firms. Such measures have considerable popular appeal, for they suggest that government is doing something about the problem of externalities. Most economists are less taken by the regulatory approach, however, mainly because of doubts about the real impact of regulation on the efficiency of the system and because they would prefer using methods that work within markets rather than as an alternative to them. More complete assignment of property rights, taxes and subsidies are obvious possibilities.

To Top

Public Goods 

Another theoretical problem inherent in neo-classical market theory has to do with the nature of goods themselves. To this point whenever goods have been mentioned they have been what may be called private goods. These are such familiar things as food and automobiles, which consumers buy and enjoy more or less privately. Other individuals are not significantly affected by the use of these goods by others, except in the important sense that they are thereby excluded from consuming them themselves. Consuming a can of beer renders it unavailable to anyone else. The producer of such goods can also exercise control over who enjoys them by denying would-be consumers access to the good until they have paid for it. 

There may be some goods which lack these features of exclusivity of use. Public roads and streets are, in practice, usually freely available to whoever wants to use them. Police and fire protection are generally available to all. Public health services which protect the community from disease cannot easily be provided for the benefit of only some citizens and not others. 

It is not impossible to imagine circumstances in which access to such services could be restricted, but often the cost of controlling access to certain kinds of goods is so great that there is little possibility private firms would be interested in trying to produce them. Trying to charge everyone who uses a sidewalk a price for the privilege and preventing anyone who refused to pay from using it seems difficult in the extreme. If a private firm did go into the business of building sidewalks, but had no effective way of controlling access to them, who would pay to use them? It is in the nature of public goods that the market demand for them, represented by the willingness of potential customers to offer money for them, understates the social returns which they would in fact provide. Furthermore, if ways were found to exclude people from using such a facility unless they did pay for it, total welfare would be reduced. Because it adds almost nothing to the cost to allow another person to use such a good, there is no gain and a certain loss by practicing exclusion in such cases. (This ceases to be true once the number of users of a good such as a sidewalk becomes large enough to create problems of congestion, in which case the good becomes more like a private good: a particular pedestrian can use it only by pushing someone else off.)

Overcoming the free rider problem may be so difficult in some circumstances that it makes sense to provide certain goods and services publicly, free of charge, and finance them through taxation. By utilising the coercive power of the state to force members of the community to pay a share of the cost of providing such goods and services, the free rider problem is resolved. This is one reason why services such as education, health care, policing, military, transportation, water and sewage, and justice are so often provided by government rather than private enterprise and financed by taxation rather than on a fee-for-service basis. (The other reasons have to do with the fact that provision of these kinds of goods by private firms is undesirable because they are natural monopolies (local transit systems, for example) or because production of certain goods in excess of what would be commercially feasible is a way of redistributing income. Provision of low-cost public housing could be an example.

Since the rise of neo-conservative political movements in the 1970s (notably Thatcherism in the UK and Reaganism in the US) many governments have adopted policies aimed at reducing the amount of public ownership and regulation. Privatisation, the selling off to private owners of public enterprises, and deregulation have become buzzwords of the times as efforts are made to transfer public responsibilities back to private markets. Selling publicly owned transportation companies to private investors has become common and the practice has attracted little opposition, except from organised workers and others directly affected. But more radical measures, such as contracting out policing and the operation of penitentiaries to private firms have brought to the fore most of the fundamental issues relating to public goods, externalities and other sources of market failure which have just been discussed.

What is particularly interesting about these developments from the standpoint of analytical economics (as opposed to whatever political views one might hold regarding these issues) is the possibility of modelling the political process using the same tools which have been developed for studying the operation of markets for goods and services. This brings us to our next topic, coming up.

To Top