David Graeber and Robert Murphy have been debating the validity of the monetary regression theory. They seem to be talking past one another. Graeber is assuming that Austrian theory agrees with neo-classical theory in areas where it does not, and Murphy is assuming that Graeber is substantially more familiar with Austrian ideas than he seems to be. To clear up the confusion, we need to take a step back and start at the beginning.
Like all theoretical sciences, economics is concerned with the identification of invariants. Austrian theory and the more mainstream neo-classical theory agree on this much, but strongly disagree on methodology.
Standard treatments of economics begin by attempting to model human decision making and then apply techniques inherited from classical mechanics to aggregate these individual decision functions into economic relationships. As Graeber is no doubt aware, the conventional approach relies on the Von Neumann-Morgenstern utility theorem and its normative description of rational behavior. (More recent approaches use cumulative prospect theory, but the differences need not concern us here.)
Austrians have long contended that this approach is untenable and unscientific. First, any non-trivial economy is too highly dimensional and too non-linear to be analytically tractable. Conventional economic theories rely on various idealizing simplifications. Austrians contend that these idealizations render the results meaningless. At a minimum, they assume away the very questions economics ought to focus on answering. More importantly, the analogous physical theories rely on the studied system being linear time-invariant. (Thus a physicist can model motion in the absence of friction because his answer will merely need a simple correction). Real economies do not have this property; consequently, the standard idealizations do not produce approximately correct claims, they produce nonsense.
Second, Austrians do not think that economic theories should depend on decision theory. Because no tractable theory of decision making can capture all the nuances of actual human beings, the “laws” identified by standard economics are not true invariants, they depend on an underlying decision theory that is known to be false. And, again, because of the inherent complexity and non-linearities involved, we cannot even claim that these “laws” should be approximately correct, as far as we know, they can be arbitrarily wrong.
In the Austrian account, the economy cannot be explained by a series of differential equations modeling the impact of perfectly rational human decisions. The economy is a complex adaptive system that emerges from the interactions of real human beings. Barring the development of substantially better mathematical tools, we simply cannot make the kind of quantitatively precise statements made by other schools of thought. At best we can make qualitative statements about the necessary properties of economic systems.
Thus far, Austrians seem to agree with Graeber:
[Conventional economics assumes] that human beings are rational, calculating exchangers seeking material advantage, and that therefore it is possible to construct a scientific field that studies such behavior. The problem is that the real world seems to contradict this assumption at every turn.
Austrians confront this problem head-on. Human action is taken as a given; whatever the nature of the underlying decision process, Austrian claims will remain valid. Instead of making a series of ludicrous idealizations, Austrians develop their theory by a series of thought experiments or, as Ludwig von Mises calls them, imaginary constructions. Via these thought experiments, Austrians attempt to prove that certain properties of economic systems are logically necessary and thereby identify the invariants of economics.
Graeber believes that he has anthropological evidence disputing the conclusions of Austrian economics. He is wrong on several counts. First, he incorrectly assumes that economics does not account for the situations that have been observed. Second, he relies on observations from economically primitive societies to refute claims only applicable to more developed economies. Third, he assumes that by showing that a thought experiment is contrived, he has disproved the conclusions drawn from it.
For example, Graeber presents multiple examples of trade as it actually takes place in primitive societies and argues that these examples contradict the predictions of economic theory. In fact, these examples are well known and explicitly accounted for:
Ethnology and history have provided us with interesting information concerning the beginning and the primitive patterns of interpersonal exchange. Some consider the custom of mutual giving and returning of presents and stipulating a certain return present in advance as a precursory pattern of interpersonal exchange. […] However, to make presents in the expectation of being rewarded by the receiver’s return present or in order to acquire the favor of a man whose animosity could be disastrous, is already tantamount to interpersonal exchange.
Thus, far from falsifying the claims of economics, Graeber’s examples are actually cases of the very bartering that he wishes to prove did not take place. (Economically, “barter” applies to all interpersonal exchanges that do not involve money; exactly what form the barter transactions take is an empirical question and beyond the reach of economic theory.)
Moreover, Graeber seems to believe that the Austrian account of money rules out the possibility of credit-bartering. It does not. The use of spot transactions to illustrate the nature of pre-monetary price formation is merely a descriptive convenience. The arguments have nothing to say about credit because it is irrelevant to the thought experiment. All that is required is that the transactions be “direct” exchanges (as opposed to “indirect,” i.e., monetary, exchanges):
The elementary theory of value and prices employs, apart from other imaginary constructions to be dealt with later, the construction of a market in which all transactions are performed in direct exchange.
Importantly, this construct is 1) imaginary and 2) very general. A transaction in the economic sense is much broader than formal haggling between merchants; mutual gifting and other such examples are deliberately included. This imaginary construction is made in full and complete awareness of the fact that true economic calculation requires money; “money prices are the only vehicle of economic calculation.” In the Austrian account, money’s role as the enabler of economic calculation is paramount.
Turning to the theory of money in particular, economics does not claim that money is a universal phenomena. Rather,
money presupposes an economic order in which production is based on division of labor and in which private property consists not only in goods of the first order (consumption goods), but also in goods of higher orders (production goods).
Thus economics actually predicts, in complete agreement with Graeber’s evidence, that his primitive tribes will not need or develop money. Rather, the need for economic calculation, and thus money, will arise in situations like those he describes as existing in ancient Mesopotamia. A band of a dozen or so people does not need money to make reasonable decisions between alternatives, but a massive temple complex with “thousands of people engaged in agriculture, industry, fishing, and herding” needs to measure inputs and outputs — it needs money. The question is how money developed.
Once we have money, it is easy to see that it is valued because it can be used in exchange. But if money is valuable today because it can be exchanged for things people want, this seems to lead to an infinite regress. At some point there was no money, so how did we go from having no money to having a money that is valued almost exclusively because it can be used in exchange?
The Austrian explanation is that goods in a pre-monetary economy are not all equally marketable. “While there is only a limited and occasional demand for certain goods, that for others is more general and constant.” People can either exchange directly (barter) or they can exchange indirectly (by trading what they have for something more marketable with the intention of using the more marketable good to acquire what they want). Thus, even if a person doesn’t need something specific right now, he will be willing to trade his less marketable goods for goods that are more marketable because this will put him in a better position to get what he wants in the future. This process creates a positive feedback loop — as more people trade less marketable goods for more marketable ones, the value of the more marketable ones is increasingly determined by the marketability itself instead of its original use-value.
In this way an economy sufficiently complex to need money can develop it. The money will be a good that had preexisting exchange ratios with other economic goods (so that people can trade for it) and it will be a good that was initially relatively more marketable than the others (so that people will want to trade for it).
Graeber’s account of Mesopotamia supports these conclusions. The economy was sufficiently complex to need money. Silver had preexisting exchange ratios in the form of cultural understandings and credit arrangements. It was highly marketable because of the demand by temple complexes and thus it emerged as money (and, hence, the unit of account).
In fact, it would seem that this logic is stronger in an economy dominated by credit transactions instead of spot ones. Because the future is uncertain, lenders do not know for sure what they’ll need at the time of repayment. Conversely, the borrower doesn’t know exactly what he’ll have in the future. Consequently, both of them are motivated to demand that repayment be in the form of a highly marketable commodity. (Indeed, the collection rules for the Judean legal system are based on marketability.)
Thus, despite his claims to the contrary, Graeber has not disproved economics.