On Fri, 30 Aug 1996, Gil Skillman wrote:
> In response to the following from me,
>
> >> Second, a question. Duncan, do you think that English Cambridge must
> >> necessarily deny the principle of factor substitution in production (in the
> >> absence of technical change)? If so, how could they?
>
> Duncan writes:
>
> >This is a complicated issue. My impression was that the Cambridge school
> >was very skeptical about substitution, both because of their attack on the
> >idea that profit was a measure of the "scarcity" of capital, and because
> >many of them opposed wage-cutting "solutions" to unemployment.
>
> Don't both of these arguments put the cart before the theoretical horse?
> Putting aside the "reswitching" problem (which does not challenge the
> *possibility* of factor substitution), a higher rate of interest necessarily
> increases the relative opportunity cost of capital-intensive production
> techniques, whatever is thought about what profit measures or what should be
> done about unemployment (and if unemployment is understood in the Keynesian
> sort, the notion that wage cutting is a "solution" is possibly a _non
> sequitur_), and thus induces factor substitution to the extent continuously
> feasible.
>
> These objections thus cloud the real issue, which concerns the empirical
> relevance of the assumption of variable proportions technology. So if these
> are the reasons the English Cambridge rejects factor substitution, then they
> have even less going for them than I thought.
Well, I confess I'm out of my depth on the Cambridge attitude toward
substitution. They tend to use examples with fixed-coefficients
technologies, and they criticize the neoclassicals for being so optimistic
about substitution, though this criticism is all entangled with the
production function debate. Certainly Steve Marglin took considerable
pains in his book to show that the real issues were not substitution,
which suggests that some people thought they were.
>
> > There is
> >surely nothing logically inconsistent about assuming no substitution (and
> >maybe it's quite close to the truth over short time horizons).
>
> Agreed. But factor substitution is a feature of "long-term" time horizons;
> does anybody believe that fixed proportions applies universally as a
> description of reality within such a horizon?
I think that technology at any one moment provides quite limited
opportunities for substitution, and that the "long-run" effects are
largely the result of technical change, which involves the movement from
one quasi-fixed coefficients technology to another. Over time this seems
to trace out an isoquant, which is what the neoclassicals look at, but
there is a crucial and potentially observable difference in that technical
change is irreversible, whereas movement along an isoquant by definition
is reversible.
Yours,
Duncan