Lack Of Rigor In Mas-Colell, Whinston, And Green

Whether or not the government should intervene in the economy is a false choice. Government and the economy are not two separate and non-intertwined entities.

The standard introductory graduate microeconomics textbook now current was written by Mas-Colell, Whinston, and Green. This happens to be from Chapter 15, in the part on the Edgeworth box:

"We can now verify a simple but important fact: Any Walrasian equilibrium allocation ... necessarily belongs to the Pareto set... Thus, at any competitive allocation ..., there is no alternative feasible allocation that can benefit one consumer without hurting the other. The conclusion that Walrasian allocations yield Pareto optimal allocations is an expression of the first fundamental theorem of welfare economics, a result that ... holds with great generality...

The first fundamental welfare theorem provides, for competitive market economies, a formal expression of Adam Smith's 'invisible hand.' Under perfectly competitive conditions, any equilibrium allocation is a Pareto optimum, and the only possible welfare justification for intervention in the economy is the fulfillment of distributional objectives."

• Andreu Mas-Colell, Michael D. Whinston, and Jerry R. Green. Microeconomic Theory, Oxford University Press (1995): p. 524

Adam Smith was little interested in the allocation of given resources, as compared to economic growth. But put aside the dubious historical claim. I want to focus on other reasons why the above passage is nonsense.

What is an "intervention in the economy"? Here are some examples of what could be interventions:

• Refusing to permit contracts in which people sell themselves into slavery.
• Banning child labor.
• Banning the enforcement of clauses in deeds which prohibit owners, in perpetuity, from selling their property to jews, negroes, or members of some other groups.
• Requiring sellers of food in grocery stores to list the amount of the Recommended Daily Allowance of various vitamins and minerals provided by that food.
• Inventing a legal structure in which people can form corporations which limit their legal liability.
• Banning people from copying some books, computer source code, DVDs, and making them freely available or selling them.

Whether to count these as interventions could be viewed as a political question. I would like to think the first four are not current questions. Law provides a background, often taken as given, on which buying and selling can be based. What contracts will be backed up by government varies with time and place.

Some elements of this background are disputed at the moment, at least by those who can attract the attention of the owners of the means of communication. An alteration or decision on a disputed element could be defined as a matter of government intervention in the economy. But such a definition does not seem to have any place in a mathematical theorem.

What counts as property is a question closely related to what counts as an intervention. It is easy easy to write, "Let ω be a vector of endowments..." But whether or not something is an endowment also varies with time, space, and the legal background. Examples that come to my mind without much thinking include air rights in New York City, a capability of a eight-year old to supply so many hours of labor, and wombs in a society where one can contract for surrogate motherhood.

Notice that conventions on contracts and property law shape the distribution of income. The distribution of income is a subject that mainstream economists have been notoriously poorly-trained to discuss.

Mainstream economists may think they are getting a rigorous introduction to economics, what with the "maze of pretentious and unhelpful symbols" (Keynes) in their books. They are also getting, however, a replication of a confused naturalization and reification of the economy common in popular discourse.

Update (27 July 2018): Fixed deleted part of quotation that made nonsense of my point.