advantage—the quid and the pro quo—up to a limiting point, or margin at which he estimates his
advantage to be a maximum. The “ marginal pair “ of the Austrian scheme hardly exemplifies the
law of marginal utility. We require to know, not so much the least price which each horse dealer will
take for his horse or stud, but how much horseflesh each individual, or at least all collectively, will
offer at each of several prices, with similarly graduated data for the would-be buyers. Granted data
of this sort, the mathematical economist need not trouble himself much about a matter which is vital
according to the Austrian scheme,—whether the “subjective valuation” of a horse is the same (or
very similar) for all the sellers, while the dispositions of the buyers are likewise identical. The case
of like dispositions does not constitute a special variety of the problem, one which is insoluble
without additional data. Far from being anomalous, that case may be normally assumed as a
harmless and convenient simplification, very proper to an introductory statement of the general
theory.68
“Neo Deus intersit, nisi dignus vindioe nodus
Inciderit “—
The case of like dispositions does not present any peculiar difficulty calling for so very mechanical
a Deus ex machina as the hypothesis that “the total amount of means of purchase must be strictly
limited and the buyers must be determined to spend the whole of this sum in purchase of the
commodities in question.” It is riding a one-horse illustration to death to put the accidents of an
exceptional sort of auction as representative of the actual transactions by which the great mass of
national income is distributed.
This criticism, it must be freely admitted, involves an issue about which legitimate
differences of opinion may exist,—what is the most appropriate conception of the process by which
value is determined through the higgling of the market? Any simple conception must involve a
considerable element of hypothesis, not admitting of decisive proof. The hypothetical character of
the inquiry will appear if we look back to that model labour market in which guides or porters were
supposed to be hired by amateur mountaineers. It was tacitly assumed that each party has certain
dispositions as to the amount of money that he is willing to give or take in exchange for a certain
amount of work,—a scale of subjective estimates69 which is supposed to be formed before the parties
come into communication, and not to be modified by the chaffering of the market. The constancy
of these dispositions being assumed, it is presumed that somehow a state of equilibrium will be
brought about, such that the party on one side cannot improve his position by entering into new
contracts with some party or parties on the other side. The better opinion is that only the position of
equilibrium is knowable, not the path by which equilibrium is reached. As Jevons says, “It is a far
more easy task to lay down the conditions under which trade is completed and interchange ceases
than to attempt to ascertain at what rate trade will go on when equilibrium is not attained.”70
68. It is so assumed in Mathematical Psychics.
69. Whether expressed by a demand-curve (or schedule, cf. Marshall, Principles Book III) or by way
of indifference curves, as Professor Pareto has suggested (Giornale degli Economisti, 1900).
70. Theory, 2nd edition, pp. 101-2. The context seems to impose an unnecessary limitation:
“Holders of commodities will be regarded not as continuously passing on these commodities in