I have some contributions to the discussion on the theory of money in
Marx, which I've had to delay due to other pressures, but here goes:
1. I think there is pretty strong evidence that Marx made a major change
in his views on monetary theory between the "Poverty of Philosophy" when
he followed Ricardo's views very closely, and Capital, where he puts
forward a view based on the Banking School and Tooke. The main difference
is that in the earlier period Marx accepted a version of the quantity of
money theory of money prices, while in the later period he insisted (more
consistently in my view) on rooting the money price level in the
reproduction cost of the money-commodity (gold).
2. Marx took for granted the Classical economists' distinction between
market prices, which might fluctuate due to changes in supply and demand
from day to day or even hour to hour, and natural prices, which reflected
the costs of reproduction of commodities. Most of Marx's discussion of
the determination of money prices seems to me to be focussed on the
"natural" price of commodities relative to gold, which reflect relative
production costs, and he argues, quite logically, that the cost of
production of gold relative to various commodities regulates money
prices. (I've tried to make the point in my Bergamo paper that this
process implicitly requires speculation as well.)
3. The cost of production will coincide with embodied labor coefficients
if the organic compositions of capital in all sectors are equal or the
profit rate is zero, but in general will deviate from embodied labor
coefficients, as Marx shows in his treatment of the "transformation
problem". Thus the appropriate natural prices are prices of production,
which take into account the different conditions of production of gold
and commodities. To put this another way, gold is produced like any other
commodity, and in the long run there is the same tendency for the profit
rate in gold production to equal the profit rate in other commodity
production as for any other commodity. (Of course, it might happen in
reality that gold is monopolized, or subject to a tariff, or whatever,
and that as a result the profit rate equalization is frustrated. In this
case the basic theory would have to modified to take account of these
higher-order determinations.)
4. Once the natural gold prices of commodities are established in this
way, money prices follow from the emergence of a "standard of price", a
legislated or customary identification of the monetary unit (dollar,
franc, pound) with a certain quantity of gold. Again, in the short run
the money prices might deviate from the prices of production, but they
will tend to be regulated by them. In this setting inflation and
deflation are at root reflections of relative changes in technology in
gold and commodity production. The state may also issue convertible paper
money or other tokens, but as long as they are convertible in fact, their
value will have to correspond to the underlying value of the gold they
represent.
5. In the Napoleonic War Britain suspended convertibility of pound notes
into gold. As a result the value of these notes went to discount against
gold. The size of the discount reflected speculation on Britain's
fortunes in the war and the likelihood that convertibility would be
restored at all or how long it would take. Marx treats this situation
very transparently, arguing that, except for misunderstandings and fraud,
the pound note price of commodities will reflect their gold price (still
determined by relative production costs) and the current market discount
of the notes. In the Critique of Political Economy Marx puts forward a
kind of quantity theory in this scenario, assuming that the notes will be
used only as medium of circulation, so that when the government issues
more notes than can be absorbed by circulation they will go to a discount
such that the total gold value of the notes is just equal to the amount
of gold that would have circulated in their place. (This is a bit too
neat for me, since I think speculation will also intervene.)
6. Gradually during the 20th century the convertibility of national
moneys into gold has been suspended due to military and economic crises.
This does not mean that the value of the dollar has anything to do with
the production cost of the notes themselves, in my opinion. What is going
on is that the credit of the state becomes the backing for the national
monetary circulation. This credit is sustained by the assets of the
State, including its future tax revenues. Thus the problem of the
valuation of State issued currency is really the problem of the valuation
of the State debt. I am not aware of any generally accepted consistent
treatment of this problem from this perspective. Two lines suggest
themselves: one is that the State implicitly sets relations between
commodities and the money unit when it legislates, e.g., minimum wages,
price floors, housing subsidies and the like. These links might take the
formal place of the standard of price. The other (Keynesian) line is that
the valuation of the state debt is inherently neutral in this situation,
so that the price level can drift around as the result of wage- or
profit-inflations started in particular markets, and is basically
historically determined. (I'm not sure which of these two ideas appeals
to me the most.)
The regularities Alan and Mandel point out might be an empirical clue to
understand what is going on. It is also interesting to review how stable
the gold price has been since about 1980, raising the question of whether
or not the Fed has been operating on a de facto gold standard in recent
years.
Duncan