[OPEL:6212] Re: a consistent, temporal, national accounting system

Duncan K. Foley (dkf2@columbia.edu)
Sun, 22 Feb 1998 21:58:57 -0500 (EST)

Alan, in pursuing the question of how to account for unproductive labor,
makes a rather detailed set of proposals. This is a comment on a couple of
technical points.



>I think some confusion arises from an
>inadequate conceptualisation of the 'period'. The notion of a circuit of
>reproduction (C-M-C-P...C') separates out a distinct phase of circulation
>(C-M-C) that, conceptually, takes place at in between phases of production,
>not during them. This is an abstraction, of course, but I think much
>confusion comes from not applying the abstraction systematically. If one
>wishes to pass to a more concrete conception, then the supposition of a
>fixed period must be dropped, and for this purpose one must move from a
>discrete to a continuous formulation. There are many more steps before this
>can be done. If, therefore, one sticks with a fixed period, one must
>suppose that production *alternaties* with consumption and one cannot
>suddenly slip into a different conception, in which circulation is not
>separated in time from production.


I don't think this works technically. In real life, of course, circulation
and production are all intertwined and all going on at once. In most lines
of production there are a lot of circuits of capital in process at any
moment, and they are not all neatly synchronised. I think Alan agrees with
this. And I agree with him that the only mathematically consistent way to
approach this is in a continuous time model. But the mathematics of
production in continuous time involve integral equations, which raise a lot
of technical difficulties. So we are tempted to try to think through the
problems in the context of technically simpler periodized models.

It is possible to construct period models that will give the correct
intuition as to what will happen in continuous time, but it is also
possible to construct period models that are completely misleading. The key
issue is whether the period model in question has a well-defined limiting
continuous time model as we allow the period length to go to zero.

If we separate out the production period and the circulation period, and
make the "alternate", as Alan suggests, I don't think we'll arrive at a
meaningful limiting model. As the period gets shorter and shorter, we'll
have a more rapid alternation of production and circulation, which will not
lead to a well-defined continuous time limit.

Alan, towards the end of his post, proposes an accounting scheme based on
the period idea:

>The accounting identities 1: periodisation of income and expenditure

>Begin with the income side. The starting point for income is *receipts from
>sales*, at the end of a period, say 1997. This total money sum is what is
>available to be spent on the next period, 1998. We can however break it
>in terms of the costs of the previous period, exhibiting it as the sum
>C(97) + V(97) + S(97) as follows:
>C(97) is the total money which the capitalists spent on intermediate
> goods at the beginning of 1997, in order to produce the new
> commodities sold at the end of 1997.
>V(97) is the total money that the capitalists spent at the beginning
> of 1997 to hire workers to transform these intermediate goods into
> these new commodities;
>S(97) is everything else; that is, it is the difference between sales
> income and the above two sums.

>How do we now account for that portion of S that is spend on hiring
>unproductive workers for the purposes of securing their services in 1998?

>The hiring of workers is just that: an expenditure. It must therefore
>be accounted for in the second major set of accounts, the expenditure
>accounts. Only confusion can arise if the expenditure accounts of the
>current period are treated as relevant to production or consumption
>in the current period. The expenditure accounts of the current period
>(that is, the beginning of the current period) define the goods that
>are acquired to be consumed in the *next* period.

>The question is, out of which part of the total C(97) + V(97) + S(97)
>do these new expenditures come?
>There is nothing that I can see which stops us making the following
>accounting identities:
> C(98) = C(97) + IMP(98)
>where IMP(98) - new investment in means of production - is the additional
>sum of money for means of production that the capitalists invest over and
>above what they spent last year (accumulation), and which must be paid for
>out of the surplus value of 1997 (If IMP is negative we have decumulation,
>the conversion of capital into revenue. If positive we have accumulation,
>the conversion of revenue into capital)
> V(98) = V(97) + ILP(98)
>where ILP(98) - new investment in labour-power - is similarly the
>additional sum of money for labour power that the capitalists invest over
>and above what they spent last year (accumulation), and which must be paid
>for out of the surplus value of 1997
> S(97) = IMP(98)+ ILP(98) + U(98) + B(98)
>where B(98) is bourgeois consumption (luxury goods) for 1998 and everything
>else is defined. That is, the capitalists spend their 1997 revenue on
>accumulation (in both goods and labour), on hiring unproductive workers,
>and on luxury consumption.
>The expenditure-income accounting identity is that
> C(97) + V(97) + S(97) (the total money income of 1997)
>must equal the total expenditure out of this income, namely
> C(98) + V(98) + U(98) + B(98)

This way of thinking raises several problems.

First, it seems to impose an arbitrary velocity of money equal to one
turnover per period. If the velocity of money is higher than this, then the
expenditures of capitalists in 98 do not need to be financed entirely out
of their sales revenue from 97. Part might be financed within each period.

Second, if there is a finite finance lag, the assumption that the 98
expenditure can be financed out of 97 sales revenue is consistent only on
the assumption of simple reproduction. In expanded reproduction, a positive
lag implies that the past sales revenue will too small to finance current
outlays. (This is one of the points in my JET 1982 article and in the
discussion of the circuit of capital in _Understanding Capital_.)

Third, in real capitalist economies a lot of capitalist spending is
financed by borrowing, so shouldn't there be a term representing the net
increase in borrowing by the capitalists in these equations?


Duncan K. Foley
Department of Economics
Barnard College
New York, NY 10027
fax: (212)-854-8947
e-mail: dkf2@columbia.edu