[OPE-L:6456] Competition and equalization of profit rates - an abstract

Eduardo Maldonado Filho (eduardo@orion.ufrgs.br)
Mon, 13 Apr 1998 10:26:53 -0300

Jerry, in his exchang with Ted, wrote (re: 04/02/98)

> > Gerry, you wrote: "If, for instance, someone from a TSS
> > perspective has written about the changing forms of competition in
> > different branches of production under late capitalism (a micro
> > subject) then I would be interested in reading that work (or works)."
> > Yes, Eduardo Maldonado-Fihlo has written on
> > that subject. I will send his paper to you.
>
> I'll read it. Perhaps Eduardo could present an "abstract" on-list?

Sorry for the delayed response.

The paper mentioned by Ted is called "Competition and equalization of
industry profit rates: the evidence for the Brazilian economy, 1973-85".
This paper is related to a former paper on the transformation problem,
which I presented at the mini-conference of the IWGVT/97. As you may
remember, I that paper on the transformation problem I have investigated
the alegation that Marx's theory of value and competition is logically
inconsistent. I have shown, by using Steedman's own numerical example, that
this so-called inconsistency disappers when we adopt Marx's temporal
approach. The release and tying-up of productive capital,which is brought
about by changes in input prices, explain the apparent inconsistency in
Marx theory. In fact, the logical consistency of Marx theory has also been
shown, from a TSS perspective, by, for example, Ted and Andrew.

Have shown that Marx's theory is logically consistency, I turned my
investigation to the alegation that it is empirically inconsistency. As you
know, the
post-Marxist/post-Keynesian aproach argues that the Marxian theory of
competition was correct for the stage of competitive capitalism of the 19th
century capitalism, but the process of accumulation and centralization of
capital has led capitalism to a new stage of development - the stage of
monopoly capital. The formation of monopolized or oligopolized sectors has
gone hand in hand with decreasing inter-industry competition. Consequently,
the monopolized sectors of the economy, because they could impose barriers
to competition, are able to obtain higher prices than those that would have
happened under competitive conditions and therefore they are able to obtain
rates of profit which are higher than the average. The competitive sectors,
on the other hand, obtain lower rates of profit than that obtained by the
oligopolized sectors. Thus, the level of the rates of profit are directly
proportional to the degree of monopoly of the different sectors of the
economy.

In order to examine this issue, my first task was to clearly derived the
empirical predictions concerning profit rate differentials (PRD) from Marx
theory. According to Marx, the transference of capital-value between
industries, by changing the "normal" level of supply relative to demand,
tends to pull high or low profit rates towards the average level; that is,
the mobility of capital-value between industries creates a tendency towards
the equalization of industry rates of profit. Thus, competition of capitals
between industries creates a dynamic process which brings abnormally above
and below rates of profit towards the average level.

But, it is important to point out that:
1) the Marxian approach does not assume that the competitive process tends
to bring an economy to a state of equilibrium where industry rates of
profit are actually equalized. In other words, competition is not seen as
establishing an uniform rate of profit across industries in any actual
economy. There are several reasons to support this view. For example, in
those industries where production is carried out in
large scale and with a high proportion of fixed capital, the time required
for pulling an abnormally high or low rate of profit towards the average
level is substantially greater than in those industries where medium and
small firms are the rule. These factors are, by themselves, sufficient to
guarantee that for any period of time which is taken into consideration the
industry-average rates of profit will be, in general, different.

2) Marx does not assume that the equalization process is
necessarily a smooth one. It may happen that too much capital is invested
in some of those industries which have above-average rates of profit
thereby causing a temporary overproduction and, as a consequence, a
prolonged period of low profitability in those industries.

3) consequently, even when the analysis of the competitive process is
restricted only to its aspect of being an equilibrating mechanism, it is
plain that the working of the adjustment process, together with all other
factors which influence prices and profit rates, makes it unrealistic to
assume that in an actual economy the industry rates of profit tend to be
more or less uniform at any given point in time. Rather, according to the
Marxian approach it is expected that, at any period of time, differentials
in industry rates of profit do exist and, consequently, that the dynamic
process of equalization is permanently in operation.

Since the existence of PRD, even between long run equilibrium profit rates,
is not necessarily inconsistency with the empirical prediction derived from
Marx's theory, I have tested it against the post-Marxist/post-Keynesian
hypothesis that oligopolist industries are able to obtain, on the average,
profit rates which are persisitently above the profit rate obtained by the
competitive industries. If long run profit rates are positively correlated
with market power and entry barriers variables the Marx's theory is clearly
inconsistency with the empirical evidence.

In other to test this hypothesis, I have developed the following dynamic
model (based on the work of M. Glick):

[r(i,t) - RE] = a(i) + b(i) [r(i,t-1) - RE] + u(it)

where:
r(i,t): is the average profit rate of industry/sector i at time t;
RE: is the long run equilibrium profit rate for the manufacturing industry
as a whole (it is taken to be the general profit rate). It is the mean of
an autoregressive process of order p;
a(i): is a parameter of industry i. It is construct so that when the long
run equilibrium profit rate for industry/sector i [m(i)] is equal to RE,
then a(i) = 0. If m(i) >RE, the a(i) > 0; and when m(i) < RE, then a(i) <
0.
b(i): is the adjusment coefficient. The bigger b(i) the slower is the
adjustment process.

At the sector level (i.e., oligopolist and competitive industries) the
empirical results shows that their long run equilibrium profit rate are not
significantly diffrent from RE; b(i) is greater the zero (and b(i=o) >
b(i=c).

At industry level (45 industries) the results shows that a(i) = 0 for 38
industies. In relation of the oligopolist industies (n = 13) only one of
them presented a(i) 0,
but in that case a(i) < 0. Thus the results are not inconsistency with
empirical predictions of Marx's theory.

I also regress the estimated long run equilibrium profit rate and b(i), at
industry level, with the market power and entry barriers variables (CR,
economies of scale, absolute capital requirements and ratio of fixed
capital to total capital). The results confirmed that long run equilibrium
profitability is not associated with market structure variables, as the
post-Marxist/post-Keynesian approach postulated. In other words, the
empirical evidence for the brazilian economy, at least during the 1973-85
period, is consistent with the predictions of Marx's theory of competition.

Jerry, I hope this (long) abstract is sufficient to give the basic
structure of my paper

Eduardo