workers' saving & Marxian political economy

["Devine, James" <jdevine@xxxxxxx>]

[was: RE: [PEN-L:23548] Re: RE: Roemer and Veneziani]

A key introductory point: I am _not_ defending the specifics of Marx's
analysis per se (e.g., what he says in volume I, ch. 25, of CAPITAL).
Instead, I am defending his general method and theoretical framework, which
I see as applicable to all sorts of new questions and consistent with some
of the conclusions of more sophisticated (i.e., non-Walrasian) orthodox
economists. Thus, I don't quote from Marx below, while I see the KI/25
analysis as only valid or useful at its own (very high) level of
abstraction. (It _may_ be invalid, by the way, but I am not at all convinced
by Gil's critique of it.) While Marx's abstractions are useful, it's
important to avoid dwelling at such a high level of abstraction -- and to be
very clear what the level of abstraction is.

I wrote:>>...the possibility of workers' saving doesn't undermine Marx's
theory. Marx does discuss workers' saving in volume III... It's often
assumed that the "classical" (Smith/ Ricardo/Marx/etc.) assumption was that
workers didn't save, but that's not true for Marx.<<

Gil Skillman writes:>We're on the same page here.<

good.

>I didn't say Marx never talks about worker saving. To the contrary, I
anticipated your remark above when I said that one can find potential
responses to the issues raised by Roemer in Volume III, the Resultate, or
Theories of Surplus Value. What I did say is that Marx doesn't discuss this
possibility and its consequences in his discussion of "the general law of
capital accumulation" in Volume I, Chapter 25.<

Of course, in KI/25 Marx was talking about the accumulation of
surplus-value, while in the land of abstract capital and labor (i.e., in
volume I of _Capital_, after chapter 3) workers can't get surplus-value.

(Most or many Marxists would say, however, that some people who are formally
proletarians (i.e., have nothing to sell but their labor-power) can get a
piece of the action (i.e., of surplus-value) when we consider a lower level
of abstraction: some top executives are not capitalists themselves, while
the experience of star athletes and actors springs to mind. These folks
quickly _become_ capitalists.)

I continued:>>The thing is that, as Sweezy says in his THEORY OF CAPTIALIST
DEVELOPMENT [pp. 139-40, 140n], is that workers' saving is for use-value,
not for accumulation of power and wealth. What Sweezy described ... is
what's now described as "life-cycle saving," delaying some consumption to
retirement and the like. (Some is to buy consumer durables, such as
houses.)<<

Gil:>I don't see the significance of this point. I'm not suggesting that
worker saving is tantamount to capital accumulation *by workers*, and I
don't see why that possibility is relevant.<

It is relevant, since would-be theorists such as Roemer simply ignore the
phenomenon of workers' saving (or don't attempt to explain the economic
basis of the "workers don't save" assumption except in the most _ad hoc_
way) even though in theory at least, workers' saving can allow mass entry
into the capitalist class, abolishing not only the scarcity of capital goods
but also the differential ownership of the means of production -- and thus
the basis of profits and exploitation in his woefully inadequate theory [cf.
Devine/Dymski, 1991, pp. 255f].

(BTW, I should belatedly thank my former colleague Howard Naish for
suggesting the concept of "barriers to entry into the capitalist class" to
me, a few years before the D/D paper.)

In fact, it's also a common practice for lefty economists of all sorts to
simply assume that workers don't save, without analyzing this assumption. If
nothing else, this contradicts the personal experience of a lot of working
people. It's also sloppy. (BTW, for those who didn't read my original
missive in this thread, my main concern was with the knee-jerk application
this assumption.)

>The immediate *purpose* of saving shouldn't alter its potential impact on
the rate of capital  accumulation in market economies, so long as workers
don't put their savings in their mattresses or piggy banks [i.e., hoarding].
If they deposit it in banks, it becomes available to capitalists for
accumulation.<

On the "purpose" of workers' saving: when Sweezy talks about saving being
for use-value, he's not talking about its subjective purpose as much as its
objective effect (so his analysis differs on a basic level from Ando &
Modigliani's later life-cycle hypothesis). That is, even if some workers
save _in order to become_ capitalists, that's not the result in practice for
a significant number. (Even the Ando-Modigliani model hardly explains upward
mobility into the capitalist class if the coefficients are anywhere near
being realistic.)

But turn to Gil's main point. Consider the following very simple equations
(which ignore the role of liquidity preference (hoarding) and bank credit
creation, while simplifying in other ways, such as leaving out the
government and the foreign sector). (Assume that we have supply = demand
equilibrium.)

supply of loanable funds = capitalist saving + worker saving.
demand for loanable funds = capitalist investment + the demand for loans to
finance the purchase of consumer durables (including houses & condos) and/or
the effects of declining income due to unemployment, retirement, illness,
etc.

The contribution of workers as a class to the flow of loanable funds is
typically less than or equal to their aggregate demand for loanable funds,
especially once one considers the role of dissaving by the retired, the
unemployed, etc. So, in this simple framework, capitalist class saving is
typically greater than or equal to their demand for funds (so that some
capitalists can make profits by lending some surplus-value to workers, not
simply by acting as financial intermediaries). At least for the corporate
sector, US capitalists are almost completely self-financing as a group,
though there is sometimes leakage of funds across class boundaries. (I
haven't seen data for the capitalist class as a whole, but it's hard to
imagine that it would be much different.)

>To suggest otherwise, in the context of Marx's Chapter 25 argument, is to
presume that when capitalists can no longer accumulate out of retained
earnings, they can't borrow money to invest.  But they do systematically
rely on the credit system, as Marx also acknowledges in Volume III.<

You didn't make it clear that you were working at the KI/25 level of
abstraction. I usually don't work at that level, since my main focus is
using theory to understand empirical reality, not to exegesis of old books.

(That is, my aim is to understand and use Marx's method and theoretical
framework to help build political economy, not to defend or attack so-called
"substantive propositions" that some distill from his texts. I see Marx's
work as contributing to a living alternative "way of thinking" in economics
to replace the orthodox micro-madness, not simply as a collection of dead
books.)

But turning to the high level of abstraction of KI/25, Marx is talking about
abstract capital accumulating wealth and power. So he's talking about the
capitalist class _as a whole_ accumulating from the "retained earnings" of
the class as a whole (i.e., the aggregate surplus-value). Intra-class
borrowing is implicit in the story, though since he's abstracting from the
heterogeneity of capitalists ("many capitals") it's also irrelevant.

Volume III is another (lower) level of abstraction, where the role of the
heterogeneity of the many capitalists is relevant. (It's important to be
very careful if mixing different levels of abstraction.) And _of course_
capitalists use the credit system: but they typically borrow only from each
other (via financial intermediaries), so that as a class they are typically
net lenders.

(Though Marx never really gets to an analysis of the role of the
heterogeneity of _workers_, that shouldn't stop anyone. cf. Mike Lebowitz's
BEYOND CAPITAL.)

Gil writes:>Let me take that point a step further. Marx states ... that the
interest rate of loan capital is determined by "supply and demand" [for
loanable funds]. Thus, additional saving, *whatever* its source, should by
Marx's own analysis lower the interest rate and promote accumulation.<

This only follows in a significant way if you ignore workers' role on the
demand side of the loanable funds market.

More importantly, let's go back to Keynes, whom I abstracted from in
discussing the supply of and demand for loanable funds. As he pointed out,
interest rates aren't determined by the flow of loanable funds as much as by
the existing stock of money and the asset-demand for money (liquidity
preference). Keynes' monetary theory is better than Marx's. In the IS-LM
lingo that dominates orthodox macro, a rise in workers' saving has a very
small impact on the interest rate -- as opposed to the level of national
income -- unless the LM curve is very steep, i.e., if the demand curve for
labor is interest inelastic and the money supply is exogenously given,
independent of interest rates.

moi:>>The existence of unemployment and similar threats to the security of
workers makes it extremely hard for any but a very small number of workers
to save enough to cross the line between life-cycle and similar saving for
use-value on the one hand, and capitalist accumulation on the other.
Further, small savers face extreme barriers to becoming large savers because
of their lack of the ability to diversify. In any event, almost no-one
becomes a capitalist -- i.e., running others' lives by owning sufficient
means of production -- by saving. Usually, luck or theft plays a bigger
role.<< [interestingly, the spell-checker suggested that I replace "moi"
with "moil"! A cutting experience.]

Gil:>I agree that both of these considerations would raise obstacles to
workers accumulating capital on their own. But again, so far as I can see,
that's not the relevant issue in capitalist economies, the world Marx is
considering in Volume I Chapter 25.<

Strictly speaking, Marx is considering a very abstract picture of capitalist
economies in KI/25, since he has abstracted from the heterogeneity of the
capitalist class. He approaches "capitalist economies" (in the sense of
referring to a more empirical picture) step-by-step, still leaving us at a
very abstract level at the end of vol. III (since he doesn't seriously deal
with the heterogeneity of the working class).

>Indeed, the threat of unemployment and other sources of uncertainty is a
*spur* to saving, other things equal,  promoting the potential impact of
worker saving on the interest rate of loanable funds, described above.<

The working class' close-to-zero net saving (supply of saving minus demand
for loanable funds) is _also_ the result (as Sweezy says) of the fact that
retired or unemployed workers' dissaving cancels out middle-aged workers'
saving on aggregate. The general insecurity of working-class life -- along
with such real-world phenomena as liquidity constraints -- means that many
or most workers accumulate _debts_ or are unable to attain desired saving
levels. Insecurity encourages people to _want_ to save, but most can't do it
(or to save as much as they would prefer). There's often a big gap between
what people _want_ to do and what they are _able_ to achieve.

An aside: That helps explain the popularity of forced saving schemes. Before
the 1930s in the U.S., workers did tremendous amounts of saving by setting
up burial societies (so that they'd have enough cash to pay the
always-exorbitant cost of funerals while perhaps keeping the cost down),
often linked organizationally to labor unions. Some pension-type saving was
done this way, too, if I understand correctly. Both of these seem to fit the
rubric of self-imposed forced saving, an effort to insulate individual
workers from the normal vagaries of labor-power markets, personal disasters,
etc. In the 1930s, the state centralized this activity in the form of the
Social Security system, while inadvertently or advertently undermining the
power of independent working-class organizations. BTW, the fact that the US
Social Security system does net saving is only a result  of relatively
recent legislation and is now being cancelled out by the rest of the
government's dissaving.

...
> Two comments:

> First, this [reference to Sweezy's comment on the close-to-zero net saving
of the working class as a whole based on life-cycle considerations] ignores
Veneziani's key contribution, which is to indicate that the mere
*possibility* of [worker?] saving leads to a tendentially declining rate of
exploitation. Again assuming that workers don't "hoard"--put their savings
in their mattresses--then a similar argument should apply in Marx's
scenario. Conversely, if it were true that this reality--that dissaving
cancels out saving-- rescues Marx's account, then incorporating it also
rescues Roemer's, since he also allows for the possibility of dissaving.<

As I noted before, I am not familiar with Veneziani's article (though I have
it somewhere under all of these piles of paper). As I said above, I agree
that _all else equal_ in a simplistic Roemerian theory, the existence of
workers' net saving would speed up the demise of capitalism. In fact, that
was one of the points against Roemer in Dymski's and my article in ECON.&
PHIL [pp. 255f]. (I'd bet that Veneziani cites us on this, since he sent me
a copy of his article and our article appears in his bibliography.)

As noted above, Marx's KI/25 story is about accumulation of surplus-value,
not about saving in general. Even if we ignore this point, as we should if
we're interested in the real world rather than poring over dusty texts, it's
not Marx's story that needs to be "rescued." His socio-economic analysis of
capitalism as a whole gives us sufficient understanding to make it clear why
workers' net saving doesn't undermine the capitalist system in practice. In
fact, that analysis is what I've been presenting in this missive and the
previous one.

Specifically, getting beyond the simplistic Roemerian model, Gary and I said
that if workers' saving _did_ start undermining the power of capitalists by
raising workers' net worth, it would raise wages, thus encouraging the
mechanisms that Marx points to (in ch. 25) to operate, i.e., the cut-back in
the rate of accumulation encouraging the reserve army of labor to swell and
the speed-up of the implementation of labor-power-saving technical change.
These not only reduce the ability of workers to claim wages that squeeze
profits, but also would encourage worker dissaving and a fall in their net
worth as a class. In sum, the notion of workers sometimes doing net saving
as a class doesn't change Marx's story significantly.

It's Roemer who needs to be rescued, since he attempts to replace a
socio-economic analysis with an atomistic, static, and idealized general
equilibrium framework, with none of the real-world phenomena such as
class-specific liquidity constraints, uncertainty, etc. One of the key
points of the Dymski/Devine article is that in order to save Roemer, we need
to go back to Marxian method and theory. Indeed, we should skip the whole
detour into Walrasian and/or game-theoretic stuff as an unneeded substitute
for Marx's analysis.

>Second, even if this [dissaving cancels out saving] were valid, it's not
part of Marx's account in Volume I, Chapter 25.<

since he was discussing the accumulation of surplus-value, as noted above.

>The possibility of working class differentiation such that some workers
save and some dissave is a possible *implication* of his argument there. But
to establish this argument *in the first place* he needs to argue that
workers don't save, before any such differentiation has been justified. In
the theoretical world invoked at the beginning of the chapter, if the wage
rate rises above subsistence, it does so for all workers.<

Again, the "workers don't save" assumption makes perfect sense (as a first
approximation, for the society as a whole) given Marx's socio-economic
analysis of capitalism, as does the fact that it doesn't undermine
capitalist exploitation. It doesn't make sense from a Roemerian (i.e.,
Walrasian) perspective.

>A corollary of this point is that Marx allows the possibility that the
average wage rises above subsistence.  If this is so, the savings of workers
should also increase, nullifying his claims, *even granting* the point that
some workers might dissave.  To suggest otherwise  is to assume  that
savings are not income-elastic, which is empirically not the case.<

Again, this knocks Roemer's method down, not Marx's.

>>If workers were not constrained by the reserve army of labor and other
forms of insecurity from significant saving, then they would be able to
enter the capitalist class in droves.<<

>I don't think this alters my point.  Here's the chain of reasoning:

>1)  Marx's Chapter 25 argument *establishing* the existence of a persistent
reserve army depends on the premise that, when excess accumulation drives
wage rates up and profit rates correspondingly down, the rate of
accumulation will therefore fall.<

right, where "excess accumulation" refers to the effect of the demand for
labor-power growing too quickly to allow a rise in the rate of
surplus-value. Also, it should be remembered that Marx points to the
capitalist introduction of labor-power-saving technology as part of his
story.

>2)  This argument does not follow unless workers as a class don't save, and
more specifically don't increase their savings as a class when the wage rate
rises. Your suggestion about the effects of insecurity from the reserve army
assumes what must be proven, since if workers do  save, then the argument
for the existence of a reserve army will not have been established.  At any
rate, Marx does not discuss this consideration, so my point that the
possibility of worker saving challenges Marx's account in Volume I, Chapter
25 would seem still to be appropriate.<

It's important to remember that in CAPITAL, Marx was not presenting his
analysis in a deductive way, i.e., starting with a bunch of premises and
deriving conclusions. (Bertell Ollman presents one good description of
Marx's methods of analysis and presentation, in his ALIENATION. See also
Marx's own discussion in the "introduction" to the GRUNDRISSE. Marx engages
in _some_ deductive reasoning, but it's hardly the main theme of his work.)
It's hard to imagine a realist (materialist) such as Marx applying a
deductive approach, while he would have been silly to do so: instead, he
starts with empirical reality and tries to understand the totality
inductively. Instead of deductive reasoning, he applies the "law of value,"
but that's another subject.

If Marx had been following the deductive approach, your criticism would be
valid. But his non-deductive (institutionalist) approach turns out to be
much more fruitful than, say, Roemer's deductive approach in plugging the
(small) hole you point to in his analysis. (I don't know if Roemer would
have gotten anywhere if he hadn't Marx's works to translate into orthodox
lingo.) If you start with an institutional analysis of capitalism as a
whole, you're more likely to understand the problem than if you start with
an idealized and totally unrealistic picture of markets (the way Roemer
does).

>3)  If the effects of "insecurity" on savings were added, I think it would
reinforce my point, since other things equal insecurity about the future
raises the savings rate rather than reducing it.<

see above. Again, I wasn't referring to subjective insecurity as much as the
inability of people to live up to their plans.

>4)  Again, remember the significance of Veneziani's new result is that the
mere *possibility* of saving creates a tendentially falling rate of
exploitation, other things equal.  Assuming workers use banks, this should
effect Marx's scenario as well.<

I don't know the context of Veneziani's result, so I can't comment.

>5)  But finally, suppose the considerations you raise were somehow to
validate Marx's account.  That would mean that if we introduced "security"
considerations into worker savings patterns, then the possibility that
workers save would somehow not contradict Marx's account. But then  building
a similar consideration into Roemer's account would also rescue it from
Veneziani's critique. Nothing in the neoclassical methodology precludes
doing so.<

As Gary and I argue, Roemer's "general" theory of exploitation has _so many_
flaws (besides this one) that if all of them are fixed, the benefits of his
Walrasian framework are clearly less than its costs. Going back to the
Marxian approach -- whence Roemer got any insights he had -- is the best
idea.

>>The scarcity of fixed capital goods would evaporate, along with the basis
for Roemer's neo-Georgian theory of exploitation. (That is, for Roemer,
profits are a form of scarcity rent, a
la Henry George -- or George Bernard Shaw.) The profit rate would fall to
zero.<<

>This is worth an entirely distinct discussion concerning the relationship
between Marxian "surplus value" and "scarcity rents."<

Not really, since I wasn't equating Marxian surplus-value with scarcity
rents. To do so would be an error, since surplus-value refers to new
production. It's _Roemer's_ theory of profits that's an unconscious
generalization of the George/Shaw tradition. (I don't think he knew that he
was following their lead.)

>But for now I'll simply make the point that it is also true in *Marx's*
scenario that excessive capital accumulation--however funded--would drive
the profit rate to zero, a possibility Marx acknowledges at the beginning of
chapter 25, before spending the chapter explaining why it wouldn't happen.
But that explanation, again, is built on the premise that worker saving is
not responsive to changes in labor income.<<

No, his explanation is based on the fact that capitalists already own the
vast majority of the productive assets and control the production process,
while workers are dependent on the good graces of the capitalists for their
survival, so that the workers are almost totally dependent on capitalist
accumulation. Thus, the capitalists can cut back on accumulation and
institute labor-power-saving technical change as wages begin to squeeze
surplus-value. (The pre-existing capitalist hammerlock on the economy is
explained historically, in the last part of KI, where he discusses the
so-called "primitive accumulation.")

Put another way, Marx's argument rests on the capitalists' control of the
existing _stock_ of means of production (and thus the production process
which creates the means of subsistence), whereas your argument is about the
_flow_ of new workers' net saving, which is a small compared to the existing
stock. This gives an argument akin to the Keynesian story above.

>>Roemer's solution, if I remember correctly, is to simply assume that
workers don't save. A clear understanding of how capitalism actually works
would have helped him here. I'm afraid his general equilibrium and game
theoretic approaches undermine his ability to achieve such an
understanding.<<

> I don't see how this follows, since it's possible to generate the results
you refer to in the appropriately specified general equilibrium/game
theoretic model.  I demonstrate this in one such model of dynamic capital
markets.<

what model is that? Just because you present a model doesn't mean that its
assumptions are valid.  (I'm sure your logic and mathematics are perfect, so
the validity of your conclusions rests entirely on your premises.)

>In sum, I still don't think Veneziani's nonetheless insightful and
interesting contribution alters Roemer's fundamental point.<

what was Roemer's "fundamental point"? I interpret it as saying that modern
economic orthodoxy  (Walras, game theory) has replaced -- and should replace
-- Marx's method and theoretical framework. I don't see you arguing that
above, except for the reference to the "dynamic capital markets" model at
the end.

Alternatively, Roemer's fundamental point may be that "exploitation" is a
matter of the inequality of wealth ownership encouraging some individuals to
voluntarily submit to exploitation by others. You don't argue for that
above, either.

Again, what is his "fundamental point"?

>But if it did, I still assert it would necessarily challenge Marx's
account, at the level of abstraction pursued in V. I Chapter 25, and
probably even when the real-world considerations introduced here by you (but
not mentioned by Marx in Ch. 25) were added to the analysis.<

Of course the real-world considerations weren't mentioned in KI/25 (a
chapter I didn't know we were discussing), since it's at a very high level
of abstraction. But his general political-economic method has been proven
again and again to be more productive than the idealized static equilibrium
approach of orthodox economics.

Jim Devine