[OPE-L:3583] Re: "Debunking Economics" and Marx's value theory

From: Steve Keen (stevekeen10@hotmail.com)
Date: Wed Jul 12 2000 - 12:45:10 EDT

[ show plain text ]

Dear Andrew,

Firstly, please don't class me as a "Sraffian". If you want any proof that
I'm not, then check another chapter of "Debunking" our apart from the one on


You should also read my paper in ROPE 10(1) 1998: 73-88, "Answers (and
Questions) for Sraffians (and Kaleckians)".

I know you don't explicitly call me a Sraffian in this post, but the
implication is clear enough.

Secondly, your argument that there can be "negative surpluses" of some
commodities at some times was a device to criticise the concept of the
existence of a physical surplus. I am afraid I'm not going to agree that you
can disprove the existence of the latter on the basis of your argument for
the former.

An input-output matrix is an abstraction, of course, but it is intended to
summarise the state of production at a point in time. The Sraffians, in
general, make the mistake of then viewing that as the be-all-and-end-all of
the economy, which of course it's not. Technical change does occur, new
products are created, etc., and the behaviour of the economy reflects the
dynamics of all that and of the pre-existing "input-output relations", not
the static properties of some matrix.

That said, once one chooses a particular point in time at which to construct
an input-output matrix, then whether it will have positive or negative
surplus entries for particular commodities depends on how widely you define
"commodity". If you go for a 10,000 by 10,000 matrix, then quite probably
you will find lots of negative surplus entries. But generally, analysts work
with 130 x 130 or less. At that level, while some products may be headed for
extinction, others which are classed in the same category will be expanding.
You are much less likely to get negative entries in that instance.

If you are working as a theoretician, and attempting to model economic
dynamics, then you are likely working with 4 or less sectors. Then you are
going to have all non-negative surpluses.

You also cannot make the delightfully arbitrary choice of time periods you
made in your example. If you are going to attempt to construct a meaningful
IO matrix as a snap-shot of current methods of production, then your time
dimension for the matrix will be set by the production process with the
longest time-lag. Other products in the matrix will have a faster transfer
function than the longest.

This is unlikely to be either 1/2 a day, or 500 years; I would probably
think it would be of the order of 3 years--depending again on how aggregate
your IO matrix was.

To incorporate technical change in such a process, I would prefer to apply a
separate rate of technical change to each industry, in another separate
matrix. Jumbling everything into the one bundle is hardly the way to analyse
such processes. Incidentally, one of my research projects is the development
of a multi-commodity multi-agent model of a production economy, in which
matrices are dispensed with entirely and computer simulation of technical
change undertaken at a far more realistic level than one can do with matrix

As for measuring the physical surplus, I would be quite content to measure
it in terms of labor-time. I would of course not be content to pretend that
the surplus was generated entirely by the labor input.


>From: "Andrew_Kliman" <Andrew_Kliman@email.msn.com>
>Reply-To: ope-l@galaxy.csuchico.edu
>To: <ope-l@galaxy.csuchico.edu>
>CC: "POST-KEYNESIAN THOUGHT" <pkt@csf.colorado.edu>
>Subject: [OPE-L:3578] "Debunking Economics" and Marx's value theory
>Date: Tue, 11 Jul 2000 08:28:12 -0400
>This is a response to *part* of Steve Keen's post of Monday, June 05,
>2000 (OPE-L 3432). The backlog of work that has prevented me from
>responding until now still prevents me from responding to the rest of the
>post. I hope to get back to it soon.
>I had written:
>"Although my original point had nothing to do with your own approach, it
>is indeed the case that your approach, and the "surplus approach" in
>general, is completely incoherent. There simply is no such thing as a
>physical surplus. There are positive surpluses of some things and
>negative surpluses of others. Which outweighs which? How do you know?"
>Steve has responded:
>"Andrew, you'd better read Sraffa again. If we leave technical change out
>of the picture (which is valid at this level of analysis, though it has
>to be part of a general theory of capitalism), then if there are
>"negative surpluses" of some commodities, eventually those commodities
>will cease to be part of the input-output matrix. If the IO matrix is
>dependent upon those commodities, then the "economy" will collapse."
>I've read Sraffa well enough to know I'm right about this.
>Let me clarify that, when I wrote that "There simply is no such thing as
>a physical surplus," I was talking about the REAL WORLD. I don't deny
>that "there is" a physical surplus in some imaginary worlds, e.g., corn
>models. But if we're talking about the real world, then you're not
>entitled to "leave technical change out of the picture," because
>technical change does take place in the real world. Hic Rhodus! Hic
>BUT EVEN IN THE ABSENCE OF TECHICAL CHANGE, the rest of what you say is
>NOT correct. It is NOT the case that "if there are 'negative surpluses'
>of some commodities, eventually those commodities will cease to be part
>of the input-output matrix. If the IO matrix is dependent upon those
>commodities, then the 'economy' will collapse."
>Consider an economy with only two goods, A and B, in which the production
>of each good requires 0.4 units of itself and 0.4 units of the other
>good. The input-output matrix is thus static.
>Good A is mostly produced in Australia and B is mostly produced in
>Brazil. Production levels of the two goods fluctuate during the day
>because daytime in Brazil is nighttime in Australia and v.v. Thus
>imagine that during the *first* 12 hours of each day, 5 units of A and 15
>units of B are produced, while during the *second* 12 hours of each day,
>15 units of A and 5 units of B are produced. I will assume for
>simplicity that wages aren't advanced; thus physical surpluses and net
>products are the same thing.
>The inputs and outputs are as follows:
> ======================
> input input output
> of A of B
> ----- ----- ------
>producers of A 2 2 5
>producers of B 6 6 15
>============== == ==
>Total 8 8
> =======================
> input input output
> of A of B
> ----- ----- ------
>producers of A 6 6 15
>producers of B 2 2 5
>============== == ==
>Total 8 8
> =======================
> input input output
> of A of B
> ----- ----- ------
>producers of A 8 8 20
>producers of B 8 8 20
>============== == ==
>Total 16 16
>Thus, negative surpluses of one good are ALWAYS produced. A negative
>surplus of 3 units of A is produced each "morning" and a negative surplus
>of 3 units of B is produced each "evening."
>However, the economy DOES NOT collapse. Given sufficient reserve stocks
>of A on "Day 1," production can go on indefinitely. The positive
>physical surpluses of a good produced during one half of the day more
>than offset the negative physical surpluses produced during the other
>Steve continued:
>"In other words, if a commodity is an essential part of the input-output
>system of capitalism, then it *must* be that its net output is at the
>minimum zero. If the system is one of expanded reproduction, then there
>must be a net non-zero output of that commodity."
>The above illustration shows that this isn't the case, even if we
>abstract from technical change. The net output of A is -3 during the
>first half of each day; the net output of B is -3 during the second half
>of each day. Yet the economy's daily growth rate is 25%.
>Steve again:
>"There therefore *is* such a thing as a physical surplus."
>No there isn't. First of all, although it might be rude of me to keep
>letting reality intrude into this discussion, I must point out again that
>in the real world there *is* such a thing as technical change. And when
>there's technical change, there simply *is not* such a thing as a
>physical surplus. There are continually positive surpluses of new kinds
>of things (e.g., Pentium III computers), and negative surpluses of means
>of production (e.g., 486 computers) that are used in production but not
>Second, in the illustration above -- i.e., even without any technical
>change -- there's continually a negative surplus of one good and a
>positive surplus of the other. As I have noted, the question is: which
>outweighs which? And how do you know?"
>You may be tempted to just add up the daily totals as I have done and
>conclude that a positive surplus of 4 units of each good is produced
>every day. Sounds reasonable, but the length of a "period" is arbitrary,
>so let's extend it to a millenium. During the first 500 years, a
>negative net product of A is continually produced, but the economy
>(reserve stocks that will be replenished during the second 500 years, of
>Now, DURING this 500-year period, do we have a positive surplus or a
>negative one?
>You cannot appeal to your physicalist theorems to answer the question.
>They all presuppose that no negative surpluses exist, so they just don't
>If you would like to reject 500 years as not being "long period" enough,
>and therefore outside the bounds of "science," let me make it 5000
>years, 500,000 years, etc. One need only presume that we have
>sufficiently small physical deficits of a good relative to the stocks
>of it that are in reserve.
>To say whether or not there's "a" surplus -- to do macro-theory or even
>macro-measurement -- one clearly needs a *homogenous* measure. Only a
>homogenous measure of value will allow you to say whether the physical
>deficits of one good outweigh the physical surpluses of the other, or
>Maybe money or a numeraire can serve as this homogeneous measure? No,
>not unambiguously. That is because relative prices, just like activity
>levels, can and do fluctuate. Even if we resort to simultaneous
>valuation (input prices = output prices), the results are arbitrary and
>In my illustration above, for instance, there appears to be a positive
>surplus during the "day" as a whole. But if the
>price of good A > (7/3)*(price of good B)
>during the first half "day,"
>and the
>price of good A < (3/7)*(price of good B)
>during the second half "day,"
>then profit or monetary value added is NEGATIVE during each half "day."
>It is therefore NEGATIVE thoughout the "day" as a whole.
>Andrew Kliman

Get Your Private, Free E-mail from MSN Hotmail at http://www.hotmail.com

This archive was generated by hypermail 2b29 : Mon Jul 31 2000 - 00:00:04 EDT