"The Marxian variables are MONETARY variables, not REAL variables - not the
annual "OUTPUT" and its components"
I strongly agree that we should not use NIPA 'real' (constant
price) measures of output and should instead use NIPA 'current price'
measures. However I think that a second distinction is also needed,
between the monetary measure of value and the labour-time measure. I
would be unhappy with the idea that Marx's values were purely monetary
if this were counterposed to their intrinsic, labour-time measure though
I think Fred makes a vital and valuable point in insisting that these
values are not physical.
This is pertinent because in the case of NIPA the issue isn't, I think,
posed in quite the same way as with the Sraffians, for whom the word 'real'
means 'physical'. The NIPA definition of the 'real output' is not a
physical, but a monetary measure. The 'real value' of a bundle of goods
produced today is the money that the same bundle of goods would have cost,
if purchased at the prices of some base year. This is a monetary measure.
It differs from 'current' prices not in being 'physical' but in being
measured in a different money; what NIPA does is to revalue all goods at a
price different from the price they actually sold for. The difference
between NIPA 'real' and NIPA 'current' prices is not the unit of
measurement but the timing of it.
Now here is my question: how can you square your rejection of NIPA 'real
output' prices, which I welcome, with your view that Marx values goods at
'replacement' or 'current' cost as you define these terms?
The problem is not immediately clear because the word 'current' is used in
different ways by different people, but I think NIPA 'real output' is in
fact a replacement cost measure of output. At best, the difference between
NIPA 'real output' prices is a constant factor, and the rate of profit
measured at replacement cost is identical to the rate of profit measured
using NIPA 'real prices'.
This is because on the basis of the NIPA conception, the price of any goods
entering production during a given year would have to be evaluated using
the same monetary measure as the price of the results of production; in
calculating 'real' profits we must subtract, from the price which physical
output would fetch at a definite time, the money which the inputs would
fetch at the same definite time.
This distinction is the real core of the difference between us. The words
'historical', 'replacement' and 'current' are useful up to a point but the
crucial distinction is the conceptual one: you assert that the outputs and
the inputs must be evaluated at the same price, and we assert they must be
evaluated at a different price. But your calculation leads to the same
results as the NIPA 'real output' prices that you rightly reject.
To see the point, suppose first that the current year is also the base
year. Suppose for example the base year is 1998 and that goods now serving
as inputs were purchased in 1997. The NIPA measure of 'real output',
applied consistently, tells us that the 'real' monetary measure of these
goods is not the money paid for them in 1997 but their price in 1998: their
replacement cost. This is the same no matter which year they were bought
in: their 'real' price is what they would cost in the base year, 1998: that
is, their 1998 replacement cost.
I don't think this is caused by any 'trick' of choosing the current year as
the base year. If we move the base year, all we do is multimply all prices,
regardless of their period, by the same coefficient. (Thus if I move the
base year to 1990, then I multiply all prices by the ratio of the 1990
deflator to the 1998 deflator). Thus, the only difference is at most a
constant factor. In particular, the 'replacement cost' profit rate is
identical to the NIPA 'real output' profit rate.
Cheers
Alan