Thanks, Michael, for your response. Yes, we are starting from a position of
(partial) agreement, and in fact perhaps more so than you suspect. In
particular, once Marx's "value-theoretic" account in Chs. 1-6 is set aside
on logical grounds, as we agree it must be, my take on Marx's systemic
analysis of capitalist exploitation becomes one of enthusiastic if
qualified endorsement: that is, I think that his stadial or
stage-historical theory of labor subsumption is clearly relevant and
anticipates mainstream theoretical developments (beginning with Knight and
Coase) by a half century.
The qualifications are, first, it's too bad he made the apparently
11th-hour decision to exclude that account from the published version of
Capital V.I, so that it had to be discovered in retrospect, much after
Coase et al. had written on related issues. Oh well. Second, as a
consequence (and possibly cause) of that editorial decision, the theory is
incomplete and needs to be fleshed out. But that's unavoidable and hardly a
knock on his theoretical contribution.
In my reading, however, there are two things that Marx's stadial theory
does not do: first, it doesn't say that capitalist control of production
is necessary for the *existence* of capitalist exploitation and surplus
value, even under the historical conditions specific to the capitalist mode
of production. To the contrary, Marx considered then-contemporary cases in
which capitalists reaped surplus value from production processes not under
their control. Similarly, while it asserts a correspondence between forms
of labor subsumption (formal and real) and forms of surplus value
extraction (absolute and relative), it doesn't offer a direct explanation
as to why this correspondence is *necessary.* Second, his theory is cloudy
at best on what *other* systemic conditions might be required for
persistent capitalist exploitation, and consequently, on the sense in which
capitalist control of production is *sufficient* for persistent capitalist
exploitation.
These lacunae are the basis, in my view, for the necessary relevance of
Roemer's analysis, however "un-Marxist" in flavor: his isomorphism theorem
speaks to the first point, suggesting why capitalist control of production
may not be necessary for the *existence* of surplus value, even if
contingently necessary for *maximal* realization of surplus value; and his
characterization of the systemic conditions necessary and generically
sufficient for the existence of surplus value (i.e., "DOSPA") speak to the
second point, if only negatively. That is: granting that Roemer does not
provide a satisfactory story for persistence of capitalist exploitation in
the face of capitalist accumulation, understanding his formal argument
makes clear that Marx's "general law of capitalist accumulation" fails in
this respect too, effectively assuming what it must prove. So at minimum,
Roemer's analysis makes clear what is still logically incomplete in Marx's
account of capitalist exploitation--just as Roemer's analysis illuminates
what is necessarily wrong with Marx's value-theoretic account in Chs. 1-6.
And this assessment ignores his important logical critique of Marx's *value
theory* (i.e. the theory that labor embodied in commodities has
explanatory, and not just normative, significance) which I don't think
anyone has successfully answered yet--as well as his *potentially* fruitful
reformulation of the concept of exploitation.
Now specifically to your comments, the following points, presented in a
bunch rather than interlinearly to keep this post short(er). Let me know
if I've missed something.
1. I agree that CAPITAL V. I is devoted explicitly to an analysis of the
"capitalist mode of production," and that this mode of production, as an
*empirical* matter, typically proceeds on the basis of the circuit of
industrial capital, characterized by commodified labor power (CLP) and
subsumption of labor under capital (SLC) in at least the formal sense. It
is also empirically true, however, that even in this historical era surplus
value has been routinely extracted via versions of the circuit of capital
in which either CLP or SLC (or both) is missing, as Marx acknowledged and
as I discussed in the previous post. *Whether or not* the latter were
true, however, it would still constitute a post hoc fallacy to conclude
that capitalist control of production is *necessary* for capitalist
exploitation, simply because it (typically) occurs on that basis.
Thus, when you assert that "exploitation with the capitalist mode of
production occurs through the sale of labour-power (and thus (a) work
according to the goals and direction of capital and (b) capital as owner of
the products of labour)", that it (a) not universally accurate, and (b),
even if it were, it does not suffice to establish your claim that there is
a systemically necessary connection between capitalist exploitation and
capitalist production. [Compare this point to Marx's at the end of Ch.
5: acknowledging that price-value disparities are universal, he
nonetheless insists that they are "incidental" to the existence of surplus
value.] Avoiding this fallacy requires a logically coherent theory of why
SLC along with CLP is *required* for capitalist exploitation to exist.
2. As far as I can tell, Marx's stadial account of labor subsumption
doesn't quite establish the necessity of capitalist production for
capitalist exploitation. First, as I noted, Marx acknowledges cases in
which what he calls capitalist exploitation occurs in the absence of CLP or
SLC. One example is his account of interest capital extended to
"cooperative factories" (worker-owned firms) in V. III, or when he notes
that piece rates render SLC "almost superfluous" [III, p. 176]. The former
case directly contradicts your assessment that, for Marx, it is a
"necessary condition...that capital has 'seized possession of production'
(such that the capitalist is the residual claimant)." But second, even if
Marx didn't acknowledge surplus value extraction on the basis of these
contrary cases, they clearly exist in capitalist reality to challenge any
claim that capitalist production is categorically *necessary* for the
existence of capitalist exploitation in this era.
Finally, even if Marx didn't acknowledge contrary cases, and even if they
didn't exist empirically, there would still be missing an explicit and
logically coherent story about *why* direct capitalist control of
production is necessary for (at least some forms of) capitalist
exploitation. This seems particularly puzzling in light of Marx's
historical account, in which he affirms that capitalists were able, prior
to the onset of the capitalist mode of production, to extract surplus value
on the basis of circuits of capital that financed the creation of new
value--e.g., usury capital provided to small commodity producers. In V.III
Ch. 36, Marx is emphatic that these circuits were capable not only of
extracting all the surplus there was to get, but even some portion of the
value that corresponds to the wage [e.g, pp. 730-732]. The question I
don't see Marx's historical account answering explicitly is this: if these
pre-capitalist circuits were able to achieve "capitalist exploitation
without the capitalist mode of production", to use Marx's characterization,
why then does it become necessary for capitalists to assert control of
production in the capitalist era? What, specifically, changed, and why did
it matter in this way?
3. This is where I think Roemer's analysis becomes relevant, by
illuminating what must be necessary conditions for a logically coherent
answer to this puzzle. For example, his analysis indicates that one need
not invoke capitalist *monopoly power* to explain the existence of surplus
value prior to the capitalist mode of production, or posit its absence in
the capitalist era to explain the need for capitalist control of
production. Basically, the two phenomena have nothing directly to do with
each other. What is needed is a specific story about contractual
imperfections, and not just their existence, but in a form strong enough to
require control of production as a strategic substitute for contractual
means of extracting surplus value. And as I noted above, one could give a
similar justification for the critical relevance of Roemer's analysis to
Marx's treatment of "the general law of capital accumulation"--but that's
another post.
I'd better get this one out--I see Jim has just come out with a 3-part
magnum opus on the topic.
Cheers,
Gil
Sorry about the delay in responding to Gil's extensive
answer to me on the labour-power question. (As I apologised to him
privately, I had to finish up a paper for a conference in Sao Paulo--
for which I leave tomorrow).
Gil stressed that his position has evolved since our
exchanges years ago. I think I've caught glimpses of that but confess
I haven't kept up with the discussions and, for that matter, journals
for the most part. However, I think our essential differences have
not changed. We have a very important agreement-- one that unites us
in contrast to many who consider this question. Neither of us thinks
that Marx's discussion in Vol. I, ch. 1-6 provides a logical basis
for the conclusion that capitalist exploitation requires the buying
and selling of labour-power. (I've expressed this by stressing that
the separation of producers from the means of production is a
necessary but not sufficient condition--- precisely because producers
could rent means production rather than sell labour-power.) For Gil,
this is a major problem ,and you can see how many times in his note
of 12 July he specifies that he is talking about the
'value-theoretic' account of chapters 1-6. I, on the other hand,
don't see this as a major problem at all since (a) the opening line
of Capital specifies that what is to be understood is a society where
'the capitalist mode of production prevails' and (b) the necessary
condition for this is that capital has 'seized possession of
production' (such that the capitalist is the residual claimant). In
commenting about two respondents to his S&S article who make this
basic point, Gil says that this is 'assuming' the conclusion (ie.,
assuming that exploitation takes place via the purchase of
labour-power rather than any other way). I disagree--- it is, rather,
specifying the subject under investigation: exploitation under
conditions of wage-labour. After all, that's what CAPITAL is about.
Sure, capitalists may exploit in other ways-- they may exploit
commodity producers, feudal lords and the like, but exploitation
within the capitalist mode of production occurs through the sale of
labour-power (and thus (a) work according to the goals and direction
of capital (b) and capital as owner of the products of labour).
So, returning for the moment to Roemer: his 'credit market
island' is precisely a case where the capital-relation does not exist
and he achieved his famous conclusion that you don't need coercion in
the sphere of production by assuming it. My comment in 'Is
"Analytical Marxism" Marxism?' was: 'In a model in which producers
wish to maximize leisure (which includes leisure in the 'pores' of
the workday), the assumption of unchanged technical coefficients in
the two cases amounts to assuming the existence of an efficient (and
costless) capitalist monitoring process--- without acknowledging it!
One can only abstract from the requirement of capitalist domination
by assuming (as Roemer does explicitly!) that the delivery of labour
for the wage is "as simple and enforceable a transaction as the
delivery of an apple for a dime".' Ie., his assumption was no
shirking and no need for surveillance and monitoring--- exactly what
he decides you can't assume when he decides to pronounce upon market
socialism.
Finally, let me note that I think Gil did not respond to the
evolution in my argument. There is a tendency to focus upon
(fetishize?) the logical moments of Marx's dialectical argument
rather than to look at the whole--- which is when Marx asks, how does
the system reproduce itself? That was how I was responding to Walt's
original question--- if the worker sells labour (and we know that it
always APPEARS as if this is what is sold), how does the system
reproduce itself? (Recall that Marx stressed that piece-work and
time-wages were forms, on the surface, of the sale of labour-power.)
The answer we will all agree is not to be found in the first 6
chapters of Capital.
cheers,
michael
