Whoops! Did Michael Roberts and Fred Moseley just revise Marx?

by Jehu

In this morning’s compare and contrast, we look at two different formulations of the category, socially necessary labor time, in Roberts’ essay, Consistent, realistic, verifiable; his review of a new book on labor theory by Fred Moseley:

Formulation 1: “Marxist value theory is based on the view that commodities are priced in the market according to the labour time expended on them.”

Formulation 2: “The market decides whether certain amounts of labour time expended on producing particular commodities are ‘socially necessary’.”

In labor theory of value, socially necessary labor time, of course, is the labor time required for production of a commodity, its value. However, in Roberts’ summary of the argument made by Fred Moseley SNLT is first described as a quantity of labor existing before exchange. Then it is later described by Roberts as a quantity of labor determined by exchange — by “the market”.

So which is it, Mr. Roberts? Is socially necessary labor time determined by production or exchange? There is no rush on this, Mr. Roberts. I am sure the proletarian revolution can wait patiently while you theoreticians figure this out.

However, I do want to point out that the first formulation suggests SNLT is determined in production and cannot be altered by any changes in “market demand”. While the second formulation suggests changes in “demand” in the market, can increase or decrease SNLT. The first formulation suggests no amount of Keynesian countercyclical deficit spending can ever change the quantity of value produced.While  the second formulation suggests the fascist state can increase SNLT by increasing deficit spending during depressions, as Keynes argued. The first formulation suggests changes in the quantity of money cannot increase the value of capital in circulation, while the second formulation says it can.

So, Mr. Roberts, what determines the duration of socially necessary labor time? The market? Or production?

(Here is the thing: For Moseley, both formulations have to be right; otherwise he could not argue, (as Roberts puts it) “The Marxist theory of value and his analysis of the laws of motion of capitalism is a macro-monetary theory.” This is a point to which I will return in a later post.)

Roberts, following Moseley (he says), argues capital begins and ends with money, so that the total quantity of money in circulation always equals the total value. But this is a tautological equality, since it must be true by definition, in the same way GDP = C + I + (X − M) + G. in bourgeois economics. The sum of all prices must equal the sum of all values and the sum of all profits must equal the sum of all surplus value.

What makes it a tautological statement is that it is true in year one and also true in year two; even if the currency depreciated in the interval, so that one dollar in year two now buys what 25 cents bought in year one. No matter how much inflation occurs, the sum of prices will always equal the sum of values in each year — by definition.

To put this another way, to say each of the three glasses in the picture below are full in no way tells us how much beer they each contain.

Two equal sums of prices will each equal the sums of values contained in them, but the sums of values may be unequal. At the same time, equal sums of value may be expressed in unequal sums of money prices.

This is always the case when the currency is deliberately devalued as was done by Roosevelt in 1933. In 1932 one troy ounce of gold was represented by 20.67 paper dollars; but in 1933 that same ounce was represented by 35 paper dollars. The quantity of gold did not change, but the “price” necessary to equal this quantity of gold increased.

While, by definition, the sum of prices will always equal the sum of values, Roberts and Moseley needed to demonstrate something else: namely, that unequal sum of prices will always represent proportionally unequal sums of values. This, of course, is necessary to demonstrate that we are dealing with a real relation between prices and values and not simply a tautology.

For instance, between 1932 and 1934 official US GDP rose from $58.7 billion to $66 billion. How would we know if this was because of a real increase in the value of GDP or simply an artifact of the devaluation of the currency? The only way to tell is to include the devaluation of the US dollar against gold to find out. After May 1933, Roosevelt devalued the US dollar so that it represented 40% less gold than before devaluation. If we adjust for this devaluation in the official statistics on gross domestic product, GDP did not rise between 1933 and 1934, but fell from $58.7 billion in 1932 to $39.6 billion in 1934 — a whopping 33% fall in gross domestic product.

Comparing US GDP adjusted for devaluation to the official GDP data provided by the fascist state gives us this chart:

Was GDP lower in 1934 than it had been in 1932? It all depends on whose numbers you use.

Was GDP lower in 1934 than it had been in 1932? It all depends on whose numbers you use.

If I am correct in my calculation, a larger sum of prices represented a smaller sum of values.

Of course, by definition, in each year the sum of prices equaled the sum of values. What changed is not this tautological equality, but the real relation between values and prices. Roosevelt was now buying gold at a 40% markup — an astonishing windfall for the holders of gold, much like the Bernanke Federal Reserve quantitative easing program.

The question we have to ask is whether Roosevelt’s decision to pay a 40% premium on gold in 1933 increased the socially necessary labor time required for the production of a troy ounce of gold? According to one formulation offered by Roberts, the answer is “No”. But his other formulation suggests the answer is “Yes”. If the socially necessary labor time of a commodity is determined by production, devaluation of the currency had no impact at all. If the socially necessary labor time is determined in the market, Roosevelt’s decision to pay 40% premium on gold altered gold’s SNLT.