From the Perfect to the Imperfect
2. The two "efficiency" theorems of welfare economics

Maria: Who cares? General equilibrium seems pure illusion to me. It's hard to imagine even a single market in equilibrium. General equilibrium is unrealistic to the point of being absurd.

McCloskey: It's not. Even if you cannot imagine a system of markets in general equilibrium, the idea of general equilibrium is very important.

Ziliak: I love G.E. theory. It's like architecture; I like to play with alternative specifications of its structure. But at the same time I sympathize with Maria. G.E. theory sounds in its abstract reaches like an existential play by Ionescu, or a game of golf with Rube Goldberg. The wizard behind the curtain is not as clever as he needs to be.

Paul: Huh-what's Rube Goldberg?

Ziliak: Reuben "Rube" Goldberg (1883-1970) was an engineer-turned-cartoonist, who gained world fame (and a Pulitzer Prize) for making sarcastic cartoons. He depicted exceedingly complex physical relations designed to achieve something trivial or impossible (depending on your perspective). Here, for example, is Goldberg's idea on "how to tee up a golf ball without bending over": But general equilibrium does not have to be so fanciful, as Edgeworth proved last century using realistic examples in box diagrams (we call them now "Edgeworth boxes").

Klamer: Yes, I agree, my skepticism towards markets notwithstanding. The idea compels you to be prepared for the intricate interdependencies and far-reaching ripple effects.

McCloskey: General equilibrium is a very important concept in economics. Ever since Adam Smith suggested that a system of markets can be self-regulating, economists have wondered whether he was right. Would markets need to be perfect, has been one of the questions. In the fifties two prominent economists, Kenneth Arrow and Gerard Debreu, showed that a general equilibrium does exist in a market system in which markets are perfect and agents have perfect information.

Ziliak: Do you in fact mean to say that Arrow and Debreu reinvented Edgeworth's idea from the 19th century, using an advanced mathematics of the 1950s?

McCloskey: Yes, I've always said so. It's the same idea.

Klamer: But note the restrictions. When markets are imperfect - and they are, I'd say - and people have imperfect information in markets - and they do - the proof of Arrow and Debreu does not apply. Who knows, in that case markets may not adjust quickly and disequilibrium situations may persist.

Ziliak: Maurice Allais (1911- ) knows. While working in the French general equilibrium tradition of Walras, Allais' researches into uncertainty and commodity surpluses have led to important but neglected insights into these and other matters, such as the interfering problems of natural monopolies and transactions costs.

McCloskey: Fair enough. General equilibrium theory circa 1957 is not divine. The research continues.

Paul: To repeat Maria's question, why would we care about the concept of general equilibrium if it is unrealistic and remains unproven?

Ziliak: Think of it as a standard by which you can assess real economies. Just as we did when we studied the conditions for maximum consumer and producer surplus in partial equilibrium analysis. Ask yourself, "What if markets were perfect and equilibrium were general?" How good would the world be then?

Rodney: Well? Happy, you mean? Ethical?

Ziliak: Those are important questions, Rodney, we all agree. But I meant "good" only in the circumscribed manner of market exchange: low prices, high output, zero economic profit. That kind of thing.

McCloskey: As you could've guessed, a market system in general equilibrium would be very good indeed. It would even be optimal from the economic point of view, that is, as efficient as possible.

Paul: But-I'm about to sound like Rodney-how just and fair would such a system be?

McCloskey: Don't know. Interestingly enough, though, we have found out that a general equilibrium in competitive markets is consistent with very different distributions of income. So it could be both fair and unfair in your eyes. These two findings go under the name of two theorems, the two Efficiency Theorems of Welfare Economics.

Maria: Welfare economics? Is that about the fact that so many people are poor?

Ziliak: Not necessarily, though sometimes, with some authors, such Amartya Sen and Maurice Allais. Welfare economics is the part of economics that is about the general happiness, or welfare of people "as a whole." So in welfare economics we try to determine how well-off people are in a particular economic situation. We'll now briefly explain both theorems.

Rodney: You said "happiness" again.

Ziliak: You're a good listener, Rodney. You keep us honest. To repeat, here we mean by the word "happiness" only what is possible to say in a utility and profit-maximizing model of human motivation. It is not the only rhetoric of motives that the authors believe to be worthy of scientific attention, but it has the merit of being the one that most economists subscribe to. In a later chapter*** we talk about alternative conceptions of "happiness" and "the good" in market societies.

Welfare economics is about general happiness or welfare of people.

The First Efficiency Theorem of Welfare Economics states that the equilibrium of a competitive economy is efficient.

"Efficiency," recall, means that no one can be made better off by making further deals, at least without making someone else worse off. In a single market the point of equilibrium of supply and demand is by definition efficient since at that point no further trades would be mutually desirable. Buyers would like to get more for a lower price but they only get such a deal at the expense of the suppliers. Accordingly, the First Theorem says an entire system of perfectly competitive markets is effective.

The result is no big surprise. A competitive market will allow the goods, whatever they are, to keep circulating until they stick to the hands who value them most. If a student does not want to go to the basketball game on Saturday she can sell her ticket to someone who values going more. Both the seller and the buyer are made better off by the trade - or else they would not have engaged in it. If trades like this are allowed to take place, then eventually no more trades will be possible. No one will be able to be made better off without making another person worse off. That means efficiency. ***insert a few lines from Ziliak's "Alternative Conceptions" piece: values.wp.

The classic application of the First Theorem is the attitude of economists towards protection against foreign competition. Economists generally do not like taxes on imports (called "tariffs"). Why? Because of the First Theorem: if the market is left alone, an efficient outcome will result.

Klamer: This is so only when markets are perfect. But they are not. The First Theorem does not apply if there are monopolies or oligopolies. Nor does it go through if the government prevents certain transactions for one reason or another. Nor if there are other imperfections.

McCloskey: Granted. The assumption of perfect markets is important. Even so, strictly taken, a tariff is inefficient because it benefits one party (the protected industry) and hurts another (consumers of the product).

Paul: So when the benefits exceed the losses a tariff would be a good thing?

Ziliak: Not necessarily. Just ask yourself the question of fairness: Is it fair that consumers should suffer to protect the jobs of those working in a protected industry?

Paul: Maybe.

Ziliak: Or maybe not. See, it's not to us, the economists, to decide on the fate of people. That's the business of politicians and their constituents.

Heterodox Box: The Allais Paradox
Maurice Allais (1911- ) has had a lifelong interest in history and experimental physics. He is one of the great economic theorists of the 20th century. In 1988 he was awarded a Nobel Prize in Economic Sciences for making seminal contributions to market-especially general equilibrium-theory. But Allais is perhaps most famous for discovering in expected utility theory a paradox, the so-called "Allais Paradox," overturning a famous result that had been forwarded by John von Neumann and Oscar Morgenstern. Human decision makers, including specialists in human decision-making such as the statistician Leonard J. Savage, choose consumption goods differently, even contradictorily, depending on how the risk and uncertainty of the outcome is presented to them.
Concept Check 2: Suppose an economist (such as McCloskey) claims that support of the poor by the government is inefficient. Use the first theorem of welfare economics to explain what she means. How could you argue against the claim?

Paul calls the attention to the distribution of income. Say a general equilibrium in the market system generates a most undesirable distribution of income with most people dirt poor and a few filthy rich. Could we correct the situation without endangering the competitive market system? The Second Efficiency Theorem tells us the theoretical possibility:

The Second Efficiency Theorem of Welfare Economics states that any distribution of income can be consistent with efficient equilibrium in competitive markets.

Accordingly, you might achieve whatever result you think is fair by way of competitive markets, by redistributing the income while allowing markets to work.

Suppose for example that the existing distribution of income yields a high income for a gangster like Al Capone and zero income for the late beloved Mother Teresa. It's a bad result on moral grounds, you say, though not inefficient if arrived at by competitive markets. Suppose you want to change it. Do you have to give up competitive markets to get wealth into the hands of Mother Teresa and to send Capone to the poorhouse? Answer: No. You can just seize Capone's property (that is, tax it) and redistribute it to Mother Teresa, then let them both back into competitive markets to trade out to the efficiency point again. The two people again exhaust the opportunities for mutually advantageous trade. But this time Capone will do badly and Mother Teresa well, though both as well as they can in view of their (new) incomes. In other words, the Second Theorem says that you can get any distribution of income (or happiness) you want between Capone and Mother Teresa within the competitive system.

The classic application of the Second Theorem is public housing, that is, the housing projects that the government has subsidized to benefit low-income families. Great idea you say? The Second Theorem tells you to think again. The reason is that the government subsidies upset the competitive market for housing. The prices of public houses are kept artificially low. Following the Second Theorem we would rather redistribute income without thwarting competitive markets. The way to do that is give the poor cash to buy (or rent) housing in the competitive market. Everyone will be better off, yet the poor will be helped just as much.

Maria: That's the same as handing over the money to the big landlords. Why would we do that? It seems quite unfair to me.

McCloskey: Your landlord will use their additional income to build more houses. See, that's how competitive markets work for the low-income families.

Rodney: But who says that housing markets are competitive? They're obviously not.

Ziliak: Quite right. In that case the Second Theorem may not help us.

McCloskey: Still, its advice is well-taken. Where and when possible, we should redistribute income without upsetting markets.

Ziliak: I can agree with that principle. In reality the advice is often difficult to follow because of political pressures. For instance, many people prefer the government to subsidize the construction of housing instead of handing out cash.

McCloskey: (Sigh) It's another case where politics stands in the way of what would the economic ideal. If only we could follow the two Theorems and respect free markets as much as possible.

Klamer: There are too many imperfections. There are monopolies, monopsonies, oligopolies and that's just for starters.

Ziliak: And then there is the 17th century insight on "unintended negative consequences." As Mandeville pointed out, the cash subsidy may cause some people to quit working altogether; the general equilibrium result may be lowered output and consumption possibilities, less competitive trade.

McCloskey: I certainly agree with that. But we've covered most of these problems already, in chapters 10 and 11. Let's now consider the problem of property rights. When property rights are undetermined or ill-defined, the two theorems do not apply. This I agree with. But the property rights problem can also be resolved.

Klamer: There will be more problems such as the problem of uncertainty.

Ziliak: And the problem of mapping utilitarian "happiness" into a notion of "the good" or even "great society." Rodney awaits!

McCloskey: Property rights first.