OPE-L message, [OPE-L] Stocks and flows in Marx's theory

Marx's capital stocks and flows do not exactly not correspond to the value of inputs and outputs of Leontief's or Sraffa's models.

The first reason is the the valuation of these inputs and outputs are defined in terms of a quantity of selected purchases and a quantity of selected sales respectively, whereas what concerns Marx is primarily how a stock of capital is converted into a larger stock of capital, under the condition where any of the elements into which capital happens to be lodged can change in valuation continually, a fact highly important in business competition.

The stock of capital is initially converted into a stock of commodities, which are then transformed by production into a stock of new commodities, which, upon sale, yield a larger stock of capital. That is Marx's story.

Any attempt to "theorise" this process however, can only be done with the aid of abstractions (whether in value or price terms) that assume some constants which, in a moving reality, do not in truth exist and only exist in the theory. For example, in accounting we sum prices paid across time, even although we jolly well know that the prices of the specific goods involved change across time (we might make adjustments for this a posteriori). Marx himself makes it quite clear that he thinks the value of whatever assets in which capital is lodged, can change continually (but he says this change will occur according to a specific pattern).

What Marx refers to as the "value" of inputs and outputs has nothing directly to do with a flow volume of purchases and sales in price terms, although he argues that, in aggregate, the value and price of the total stock of new commodities produced must be similar (for the sake of argument, he assumes they are identical). The "value" of inputs and outputs in Marx refers to the stock of commodity units purchased and sold.

When we face the problem of having to calculate the new value added by production during the year empirically, all we have as raw material is observed transaction data on purchases and sales. The first logical step is then to demarcate the boundaries of production, i.e. only those transactions which pertain to to the value of production are included, and we assume that total production purchases must equal total production sales (since for every purchase there must be a sale). The second logical step is to say that the total incomes generated by production (factor income) must be equal to the value added. The factor income equals the total production purchases or production sales (gross output), less the purchases or sales of intermediate items used up.

We exclude a large number of transactions from our calculation, on the ground that they have nothing to do with new production or nothing to do with value added, even although they generate incomes and expenses (this is all forgotten in Marxist theory). These include in national accounts (for example) land rents and subsoil rents, property income of business and individuals, sale of second-hand goods, transfers of wealth which do not create wealth (unilateral grants which do not imply direct economic exchange), certain interest income of individuals etc. That is to say, we erect a system of transactors which only capture transactions related to our definition of production.

In the Leontievian system or the Sraffian production system, the valuation of total inputs is assumed to be equal to the valuation of total outputs. But why is this logically the case? Because out of the total universe of transactions, we have abstracted out a subset of production transactions where, for every purchase, there must be a sale. Then it follows, that total inputs purchased must equal total outputs sold.

Prof. Steedman thus complains that for Marx, the price of a good is different depending on whether it is purchased or whether it is sold, whereas anybody knows that if a price is fetched, then this price is the same for the buyer or the seller. But this theoretical complaint rests on the assumption of input-output economics and double-entry bookkeeping. In the real world, much to the chagrin of many business owners, there is no necessity that the valuation of the stock of commodities purchased by producers, equals the valuation of the stock of commodities sold.

Prof. Kliman therefore has a point when he argues that the valuation of total inputs need not equal the valuation of the total outputs. Inputs and outputs could be "theorised" as stocks of commodities, or they could be theorised as the flow value of commodities purchased or sold across an accounting period, and they could be "theorised" as equal to capital expenditure/revenue in an accounting period or as capital stock advanced or capital stock realised.

Jurriaan