Gil had written:
Roemer doesn't have an "exploitation-as-scarcity-rent" theory per se,

me:
He would never say he did, since it would imply that his theory was
derivative from that of Henry George and George Bernard Shaw, among
others (if indeed he read their works), dressed up with fancy math and
jargon.

Gil now says:
And he doesn't *need* to say he did, since, as I noted, he doesn't. …
Your statement presumes, among other things, that somebody--Marx,
say--has successfully established a logically coherent theory of
exploitation that *isn't* "derivative from that of" one of your Georges.

Now, having read Gil's contributions in the past, a "logically
coherent theory" means a "model," that is, a formal way of expressing
a "theory." (Please tell me if this is not an accurate description of
your approach.)

Marx clearly did not present what is nowadays called a "model" (or
what Max Weber called an "ideal type" at the turn of the last
century), a purified and ideal description that emphasizes its own
internal consistency (in terms of formal logic or math or perhaps
simply simplicity, as in Weber).

Unlike the NC school of economics, I distinguish between a "theory"
and a "model." (Like most people, however, I sometimes use this
distinction in a sloppy way. If people use the word "model" in a way
that I don't embrace, I typically don't quibble.)

A non-speculative theory represents a (valid or invalid) _insight_
into the way that the (empirical) world works. For example, Marx
presents a theory of commodity fetishism or the illusions created by
competition (how the appearance of the class nature of capitalism in
"the ordinary consciousness of the agents of production themselves"
normally differs from the real-world totality of those class
relations). But he did not present a logically consistent model of
such. Nor, to my mind, did he need to do so (except that some people
don't understand anything that's not stated in model form).

For example, George Akerlof presented a model of the "market for
lemons" (adverse selection) a few decades ago. This model differs (in
philosophical terms) from the insights about concrete causation that
encourage adverse selection to exist, which existed (in incomplete
form) in Adam Smith's time and was well-known to people in the
insurance business at the time that Akerlof wrote. The model can be
seen as clarifying or even adding some theoretical insights (that's
why it's a good model) but it also adds distractions, such as the
assumption of perfect competition (without emphasizing its unreality).
Most economists would say that the benefits of Akerlof's modeling
approach exceed the costs, but those in the insurance industry may not
agree.

Similarly, the two Georges (Henry and Shaw) and Marx presented
theories of exploitation but not models. Their theories, however, were
different:

1. Henry G. presented a theory of land-rent as exploitation.

2. G. Bernard Shaw presented a generalized theory of rental income as
exploitation (as a Fabian response to Marx, BTW).

3. Marx presented as socio-economic theory of exploitation. If I had
to choose any passage from Marx to summarize it, it would be "the
worker purchases the right to work for his own livelihood only by
paying for it in surplus-value" (CAPITAL, vol. I, International Publ.
ed., p. 515).

My assertion, of course, is that Roemer presented a model formalizing
the second theory – and to a lesser extent, the first theory – but not
the third. I also assert that his model did not imply any new
theoretical insights about the world.

It did suggest that Marxian theories of exploitation, though already
valid, needed to expressed in a clearer way in order to avoid the
theoretical confusion that Roemer introduced with his
internally-consistent models. The late John Elliott and Gary Dymski,
and later Gary Dymski and myself, did this work awhile back.

Models are nice and can be aesthetically pleasing or amusing, but they
are not theories. Rather, they are ways to represent theories. And of
course, a model need not give any insight into the way that the
empirical world works. Internal logical consistency is not the same
thing as useful insight. (Euclidean geometry can prove a lot of
theorems, but those theorems need not apply in our non-Euclidean
world.)

For what it's worth, I don't think Marx--or any subsequent Marxist before
Roemer--has, but a careful demonstration of logical coherence would need to
be made on the basis of some sort of formal, i.e. essentially mathematical,
analysis, "fancy" or otherwise.

Agreed. Marx never thought that a "a logically coherent theory of
exploitation" (meaning an internally-consistent model) was necessary.
Instead of using the model-building methodology of modern NC
economics, he applied the unfashionable empirical heuristics called
"materialist dialectics."

Marx himself likens his method in CAPITAL to that of chemistry, so
maybe it applies. In chemistry, unlike NC economics, people don't
derive conclusions from axioms. It's more experimental. Now, as Marx
said, you can't use reagents on society to do experiments, so
abstraction is needed. As Bertell Ollman argued, it's a matter of
looking at the empirical world in different ways. In volume II of
CAPITAL, for example, he first looks at the circuit of money-capital
(i.e., M - C - … P … C'- M). Then, in the second chapter, he looks at
the circuit of productive capital (i.e., P … C' - M' - C … P). Then,
in the third chapter, he examines C' – M' – C … P … C', the circuit of
commodity-capital. In the end, i.e., in chapter 4, he presents a
vision of the totality of the process. It's like an experiment where
one looks at the combination of three chemicals by first looking at
the combination of two, then another two, and the third pair, before
mixing all three.

(This is not the same thing as a "gedanken experiment," though there
are some similarities.)

There's another way to explain this. In modern NC economics, the basic
logic of model-building is deductive, i.e., starting with a bunch of
axioms or assumptions and then logically (mathematically) deriving
theorems or conclusions. Sometimes, the assumptions are not made very
clear, so that it turns out later that additional assumptions are
needed. For example, the Walrasian system must assume that there
exists an all-knowing price-setter, the Auctioneer. (This assumption
did not become seen as needed until later, when economists started to
take the Walrasian model seriously after World War II.)

On the other hand, Marx engages in induction – with abstraction. He
starts with empirical reality, as he explains at the start of the
GRUNDRISSE, and then deliberately leaves out elements of perceived
reality in order to understand the complexity of the world. He starts
with a market, for example, and then ignores the seemingly random
fluctuations around the "average" price (what he called "prices of
production" in a model with no noncompetitive elements). He doesn't
assume the existence of a god-like Auctioneer, because he's dealing
with a simplified description of the market rather than presuming to
build up a "model" of the market from axioms.

So you could say that while neoclassicals have to add "abstractions"
like the Auctioneer for their idealized stories (models) to apply,
Marx's abstraction involved subtraction.

Which is why, I imagine, your own paper
introduces jargon and more-or-less "fancy" math.

I don't know if a logically-consistent model is really needed. Perhaps
not. As with a lot of academic life, it's part of the game. (Is
academic economics needed at all? Likely not.) Maybe I can get a
little respect for Marxian political economy by presenting a model
that doesn't sacrifice Marxian political economy as much as Roemer
did. I doubt that this respect will start rolling in, but as Vonnegut
says "so it goes."

My methodological attitude is summarized in the first footnote of my
paper: "It is not surprising that [the standard] utopian NC
assumptions imply non-exploitative results. But if exploitation can
exist even under such assumptions, it seems likely to exist even in
the empirical world where such ideal assumptions do not apply." That
is, if a Marxian theory of capitalist exploitation can be expressed
even in a model that makes garden-of-Eden assumptions, that's evidence
in favor of its validity. On the other hand, harmonious assumptions
generate harmonious results (garbage in/garbage out), so we shouldn't
be impressed by the nice theorems that result from perfect-world
assumptions.

Gil had written:
>>Productive assets, or the wealth advanced to
>>finance them, must be scarce at the margin in order for surplus value (and
>>thus exploitation) to arise in equilibrium, but not necessarily "scarce" in
>>the sense that yields economic rents to capitalists.

I responded:
>I don't get this distinction.

Gil now writes:
See below for an example. The general idea is that exploitation, certainly
in Marx's sense, even if capitalists earn only normal profits, and thus no
economic rent.

Yes, I agree that capitalist exploitation in Marx's sense could occur
with capitalists earning only "normal profits" without economic rents.
It could also occur with economic rents, which he interpreted as being
a redistribution from industrial capitalists (who organize the actual
production of surplus-value).

On the other hand, my assertion is that Roemer's theory is simply a
theory in which all profits represent the scarcity of productive
assets (and the monopolization of such assets within a free-market
environment). That is, his profits are totally based on the existence
of scarcity rents. (If the assets were equally distributed, then
everyone would get a piece of the pie that's assumed to exist.) There
is no actual production of surplus-value, since he does not use the
labor/labor-power distinction. He simply assumes that a surplus exists
to be taken, like ripe apples hanging from a tree.

Gil had written:
>>This point especially
>>holds for his cooperative game-theoretic generalization of the concept of
>>exploitation in Part III of his 1982 book General Theory of Exploitation
>>and Class.

me:
>I was talking about positive theory (how does exploitation work in
>practice?). The game-theoretical stuff seems to be entirely normative
>(what's wrong with exploitation?).

Gil:
No, Roemer's generalization is strictly positive: exploitation, according
to his definitions in Pt III, either obtains or it doesn't, however
one feels about its wrongness. His definitions are normatively motivated,
perhaps, but then, definitions generally are.

I didn't say anything about it being wrong simply because it was
normative. By the way, I notice now that you refer to "cooperative"
games. Those are much more idealized and unrealistic than
non-cooperative games. I don't see how they are relevant to the world.
Instead, they are relevant to an idealized world (an oxymoron!).

My impression was that Roemer used his cooperative game theory models
to justify the fact that exploitation is a bad thing. If that isn't
normative economics, I don't know what is. But I could be wrong. Given
the low benefit/cost ratio, however, I see no reason to go back and
read Roemer's toem, to see if I am wrong or not.

< ellipsis >

Gil:
>>For example, it might be that existing capitalists are worse or
>>more risk-averse entrepreneurs than existing proletarians would be if they
>>could only get access to financial wealth (which they can't, say, due to
>>imperfect credit markets). Then under Roemer's definition, workers are
>>capitalistically exploited in this scenario because they would be better
>>off with an egalitarian share of alienable productive assets, even if
>>existing capitalists don't in fact earn scarcity *rents*.

me:
>This proves my point above by making it clear that Roemer presents a
>normative definition, i.e., a comparison of (simplified) real-world
>phenomena with an ideal world (with perfect credit markets, etc.)

Gil:
No, it makes clear that Roemer's generalization of capitalist exploitation
is positive--exploitation either obtains according to the definition or it
doesn't, however one feels about it. The fact that I used a simplified
example to illustrate my point that Roemer's definition does not require
the existence of scarcity rents doesn't alter this point.

I don't see how a theory that blames exploitation on the real-world
deviation of market capitalism from the idealized world of perfect
credit markets and the like could be anything but normative. (It's
like saying that the real world differs from the Garden of Eden
because it has snakes.) The yardstick of comparison for any theory or
model should be the real (empirical) world, not some unreal world.

Gil had written:
>>In any case, I wonder if Jim would disagree that Roemer's scenario of
>>scarce, unequally distributed capital assets is at least *sufficient* for
>>the existence of capitalist exploitation.

me:
>Unfortunately, arguing that something exists in a Walrasian general
>equilibrium framework, as Roemer did, is tantamount to arguing that it
>doesn't exist at all.

Gil:
I don't see this point. It seems equivalent to saying that arguing for the
existence of surplus value in the never-obtained case of commodity exchange
at value--as Marx does-- "is tantamount to arguing that it doesn't exist at
all." The claim doesn't follow.

But Marx's explanation is not dependent on his assumption that
commodities exchange at value. Or at least he didn't think so. It's
easy to demonstrate. You can simply assume that all commodities trade
at prices of production. Why then, for an average commodity, does
profit exist? (Why is it that the slope of the intertemporal
production/consumption possibilities curve such that the interest rate
can be positive?) That is, why does the price of production of the
output per hour of labor-power hired exceed the price of production of
an hour of labor-power hired? Etc. My article in Bill Dugger's book
applies a version of this approach, as does that by John Elliott.
Marx's theory, once developed by Marx, can be presented in a lot of
different ways, as I've said before.

The problem with the assumption of price = price of production, of
course, is that it assumes profit-rate equalizing equilibrium (and we
know that the lonely hour of this equilibrium can never be attained
given various imperfections and the dynamic nature of a capitalist
economy). It also ignores the macrosocietal insight arising from the
price = value assumption. (Of course, the price of production of an
average commodity does equal its value.)

However, we should note that both Marx and Roemer made big mistakes.
While Marx danced around using the obscurantist jargon of his day
(Hegelian lingo), Roemer embraces the mystifying formulations popular
in his day (Walrasian stuff). Both undermined their cases by
communicating poorly.

me:
>He thus undermined his own case by trying to
>please and fit in with the economic orthodoxy …

Gil:
I don't think that's what he was doing.

then we disagree. I could swear he was looking for a job in Cambridge.

me:
>(It's ironic that he rejected
>"BS Marxism" by embracing neoclassical BS.)

Gil:
I think this fundamentally represents what he accomplished, and is any case
beside the point at issue: Roemer's theory, contrary to your assertion,
does not require the existence of scarcity rents.

It centers on, and is totally dependent on, the unequal distribution
of scarce nonhuman productive resources, in which the return to the
ownership on said resources is zero if they aren't scarce. That is,
Roemer's normal profits are scarcity rents.

In the model of the paper I posted on my website, these assets are
not, strictly speaking, scarce at all, since unused capacity exists.
Rather, they are _relatively_ scarce, relative to labor-power. Both
fixed assets and labor-power suffer from unemployment (since, unlike
Roemer, I reject Say's "Law"). But unlike fixed assets, people have to
live. (Smith understood this, writing that the need by labor for
capital is much stronger than vice-versa.) This means that they are
forced to pay the capitalists surplus-value in order to survive; they
do not voluntarily choose to be exploited as in Roemer, but instead
face institutional coercion.

me:
>But the existence of unequally-distributed scarce non-human productive
>assets does seem necessary to the existence of exploitation, as I've
>said before (including in the paper I uploaded to my website). The
>point, therefore, is to explain the reproduction of that social
>situation over time once it has been established (by primitive
>accumulation, something I don't discuss in my paper).

Gil:
Explaining that would require some sort of dynamic framework. Is that what
your paper does?

Because it's a "neoclassical" framework, it's not truly dynamic. (NC
economics has a very unrealistic view of time.) However, I show the
plausibility of the existence of the receipt of "normal" profits (just
for owning the means of production) in steady-state growth driven by
profit-seeking capital accumulation. In the cases of severe
Depression, a workers' rebellion, Hobbesian anarchy, or an
environmental melt-down, however, this kind of equilibrium may not
exist.

<ellipsis>

Gil:
I don't see how it's "fruitless", and I'm not talking about "the law of
value", whatever that is. I'm talking about exploitation, which I note you
don't define directly in your paper.

1) The "law of value" is Marx's term for what is usually called his
"labor theory of value." (As far as I can tell, he only used the
phrase "labor theory of value" to refer to others' ideas, e.g., those
of David Ricardo.)  I prefer the former term, because the latter is
usually read as being a labor theory of relative prices, à la Ricardo.
I think that totally misses the point of Marx's theory.

2) In the text that was replaced by second big ellipsis above, you
speculated about the different role that looking at labor intensity
(effort/hour) rather than extensity (hours/paid hours) has in terms of
value. This is what I saw as a fruitless topic for discussion, since I
do not have a good memory for quotes and do not like textual
discussion. (My text-expert tells me you're wrong in terms of what
Marx really said, but never mind.)

2) I do define exploitation, simply as "to the receipt of positive
'normal' profits merely for owning means of production." It's on page
1.

Gil:
How can you determine if exploitation
exists if you don't, for example, determine the labor embodied in the wage
bundle?

I've said several times that Marx's theory of exploitation need not be
explained in terms of his law of value. But since my model is a macro
model, however, there is a direct correspondence between value and
price categories.

<ellipsis>

Gil had written:
>>If that's the case,
>>I don't see how his paper can possibly establish a relevant basis for
>>"superceding" Roemer's theory, whatever its other merits.

me:
>I think that my paper captures Marx's idea that exploitation under
>capitalism is a social relationship that cannot be reduced to simple
>economics or the "unequally-distributed scarce non-human productive
>assets."

Gil:
The real question, it seems to me, is whether the paper *disproves*, rather
than simply refines, Roemer's claim that DOSPA is necessarily and
generically sufficient for the existence of capitalist exploitation. FWIW,
it's not obvious to me that it has.

I don't know what "generically sufficient" means.

Again, productive assets aren't absolutely scarce in my model. They
are only relatively scarce. That fits with Marx, who spoke non-Say's
Law (Keynesian) prose without knowing that he was doing so.
--
Jim Devine / "Segui il tuo corso, e lascia dir le genti." (Go your own
way and let people talk.) -- Karl, paraphrasing Dante.

Reply via email to