Continuing our notion of value originating as a way to compare things that have necessarily different physical units, we note that valuation has always occurred in terms of some sort of reference commodity: how much of this commodity is to be accepted in exchange for good J₁? how much for good J₂? etc. Money comes into existence as a transportable promise to tender some quantum of the reference commodity upon demand.

Gold has long been humanity's favored reference commodity because it best satisfies our longing for an imperishable standard of value:

Gold's historical success as a value standard owes first of all to its intrinsic interest, whether for its beauty, its incorruptibility, or whatever. Because gold is interesting, people are aware of how much they have, and thus the economic system comes to know how much there is.

Gold is at least somewhat fluid in that it can be divided and moved about as needed in order to, as Bonanza Jim Flood stated the matter, keep score in the game of commerce.

The total quantity of gold is essentially fixed in that we think comparatively little is left to be discovered, and comparatively little is consumed by gold's uses in industry, medicine, or the decorative arts.

In theory any commodity, however worthless, might serve as the reference commodity; and, when in their whimsical moods, economists can be quite dismissive of any intrinsic significance imputed to the actual commodity accepted for reference. They remind us that certain remote Pacific islanders were able to use cowry shells as their medium of exchange (that is until they endeavored to enrich themselves by pulling money instead of fish out of the ocean).

Socialists' rejection of profit has had economic thinkers from Saint-Simon to Veblen contemplating money in terms of some widely-used, industrially relevant commodity that would necessarily exist only in that quantity required by the overall economic product. A related Technocracy movement would define money as a certificate of ownership in the total quantum of energy consumed by all economic sectors — thus eliminating passive income from the calculus of the technocrats, who would then be free (per Mussolini and other national socialists of his period) to run the economy with the sole criterion of engineering efficiency. This idea has currency in the thinking of Sri Manmohan Singh, presently leading a rapid expansion in one of the world’s most important economies.

Piero Sraffa’s more abstract notion of the reference commodity moves us much closer to something serviceable to a capitalist theory. Sraffa computed value in reference to the monetary equivalent of all production that might occur if all extant capital were used to capacity. Though originating in economic heterodoxy, this idea is closely followed by SFEcon in that our reference commodity comprises the value of all physical flows when the economy is operating at the general optimum implicit in its parameters.

When expressed in terms of our hyperbolic utility tradeoffs, neoclassicism’s theory of general optimality allows us to compute the price of every commodity and every currency in terms of an absolute value unit that is no more variable in time than the short ton, the BTU, or the Troy-ounce. This is roughly the definition Keynes gave to his Bancor; but, not wishing to be presumptuous, we refer to our constant value unit as the Global Currency Unit, GCU, or G.

As with everything else about SFEcon models, this notion of reference prices has an operative definition. Operative (as opposed to merely narrative) definitions are expressed in controlled experiments of the sort performed by any of the prototype models to be downloaded from this site.

GCU prices exhibit their defining characteristic when the model responds to an exogenous change in a state variable, i.e.: in the amount of some commodity whether in an inventory or on the market. Such stimuli are called elastic because a pure neoclassical model should respond by eventually settling back down to the same physical equilibrium in which it began.

When, for example, the experiment is initiated by a sudden deposit of some commodity on the market, this excess will have to be worked-off by the ensuing transient behavior; and the model should settle back down to a state in which every sector has the same quantity of every good that it had before the stimulation. If this model ventures cardinal (as opposed to merely relative) price adjustments, then the experiment should conclude with wages and commodity prices that are all lower than their initial levels; but with prices in the same proportions at either end of the experiment.

The more primitive SFEcon models perform this venerable experiment successfully; but with the added feature of tracking our constant-value, GCU prices through their transients. In experiments using elastic stimuli, GCU prices re-acquire their precise initial magnitudes as the new equilibrium emerges. (More complex models, featuring rival household sectors, do not quite behave elastically.)

A second class of experiments is based on plastic stimuli requiring that a model adapt its state to a change in boundary conditions. If, for example, technical growth were modeled by making one sector a more efficient user of some input, then the final equilibrium would differ from the initial state in every particular. Economic principles would, however, reliably dictate that adjustments of this sort would be deflationary because more assets would be required to support the new optimum, and these assets would be demanded in quanta greater than current supply during the transient period.