answers to P. Skott

[Jim Devine <JNDF%LMUACAD.BITNET@xxxxxxxxxxxxxxxxx>]

JD: Below are abbreviated versions of Peter Skott's (PS's) questions
and my (JD's) answers.  I have only glanced at  Paul Davidson's
(PD's) discussion of this or the rest of the thread that follows.
Let me apologize ahead of time for any repetition that results (and
for the long delay in my answering). (Before starting, I should
again stress the fact that even though I think that the
new-Keynesian idea of the difference between notional and effective
demands for labor is useful to our efforts to understand the world,
I disagree with the NKs on many other things, such as their common
assumption that we should start any analysis at full employment
equilibrium.)

part I of II.

PS: PD and JD to have given up, but maybe a consensus can be
reached.

1. Firms - whether in perfect or imperfect competition - make a
price-output decision: they pick a point on their (conjectured)
demand curve. The profit motive has not been questioned in the
debate and for simplicity one may assume that firms (try to)
maximize profits subject to production constraints and the
(conjectured) demand function.

2. It is assumed in standard short run macroeconomics ... that
short-run expectations are fulfilled. In other words, firms achieve
the anticipated price-output combination.

... JD, as I understand him, denies either 1 or 2 ... I hope that JD
is questioning the usefulness of short run equilibrium. Outside SR
equilibrium - when SR expectations fail to be satisfied - firms will
not reach the anticipated position. In the simple case of atomistic,
price-taking firms f' will deviate from w/p.

JD: I accept assumption 1, with an emphasis on *try to* maximize
profits, given the normality of uncertainty, incomplete information,
and the like (I've replaced "profit maximization" with "profit
seeking" in my vocabulary). In most cases, given incomplete
information, uncertainty, and the like, firms will fail to maximize
profits _ex post_. However, profit-max models are useful for
pointing to the direction in which firm decision-makers will go.

I do not reject the idea of SR equilibrium as defined above. Rather,
(a) the SR equilibrium condition does not apply in the process of
moving from SR equilibrium point A to SR equilibrium point B, while
more importantly, (b) I reject the idea that SR equilibrium always
coincides with the long-term neoclassical equilibrium in which
perfectly-competitive firms can sell (or buy) all that they want at
the given price (wage).

On (a), except as a first step in a more complete analysis, I reject
the standard "comparative statics" method (which was stated well by
Samuelson) of looking only at the starting-point and the end-point
and seeing this as representing the process of getting from point A
to point B. Rather, this is a process taking place in *historical
time*. The process of getting from point A to point B may affect the
nature and location of B (the hysteresis phenomenon). (Given the
normality of exogenous and endogenous shocks, perhaps no point B
will never be attained. However, there does exist a _tendency_
toward attaining SR equilibrium which might be modelled to get
provisional and useful results. This parallels my statement above
about profit-maximization.)

On (b), in a typical quantity-constrained product-market equi-
librium, as Keynes pointed out, the demand for goods does not
coincide with (and is to the left of) the notional demand curve at
the full-employment (FE) level of output.* Firms may be in SR
equilibrium in the sense that they sell according to the expected
price-quantity combination based on the effective demand for
products.

*Many Keynesians take this for granted. As Clower points out, in
neoclassical theory, the notional demand for goods should not depend
on current income as much as wealth and the potential to earn income
in the future (Friedman's "Permanent Income"). But he argues, and
most Keynesians would agree, that there are liquidity constraints
that make current income relevant, creating a difference between
effective and notional demand for products.

In typical quantity-constrained labor market, the demand for labor
does not coincide with the notional demand-for-labor curve that is
based on the assumption of FE demand for products. Since the
effective demand-for-products curve is to the left of FE level, so
is the effective demand for labor.  Firms make their hiring/wage
decisions based on this effective demand-for-labor curve and can be
in a SR equilibrium as defined above.

Unlike Keynes, I (like the whole "disequilibrium" school) do
not assume that the effective demand for labor (given the demand
curve for products) is a _single_ combination of wages and
employment, a single point, on the technologically-determined
notional demand for labor (MPP of L in a perfectly-competitive
market). The _effective_ demand-for-labor curve is to the left of
full employment, for a simple reason. For given nominal wage and
price, the unconstrained atomistic firm hires according to w/p = f'.
But (unless w/p is very high), the quantity-constrained firm cannot
sell all of the extra product produced by hiring according to this
rule at the given w/p.  So the firm will hire fewer workers than the
unconstrained firm (unless w/p is very high). In addition to the
technological constraint represented by the production function (f),
there is a sales constraint based on the effective demand for
products.  (For the simplest possible example, see p. 23 of
Jean-Pascal Benassy's MACROECONOMICS: AN INTRODUCTION TO THE
NON-WALRASIAN APPROACH.)

For the case of a firm  with some price-setting power, see the end
of this message.

PS: How do firms react if there is SR disequilibrium? JD's position
appears to be that firms choose to react primarily through quantity
adjustments. This assumption may or may not be empirically
reasonable (my reading of the evidence is that prices are in fact
quite flexible)....

JD: No. As I've said several times, my argument is _not_ that
|dq/dt| > |dp/dt| in response to downward demand shocks, though that
may be true in many situations. Rather, as I've said several times,
the argument is that prices and money wages do not adjust
instantaneously (|dp/dt| < infinity) in order to keep the economy in
notional (FE) equilibrium at all times. (The idea here is that there
exists a vertical aggregate supply curve in price-real GDP space:
instant price adjustments would keep the economy on that curve. But
these don't happen.) Given uncertainty, it is likely that firms
would try for a "diversified portfolio" of price and quantity
changes in response to the shock (though, as PD would argue, the
elasticity of expectations plays a role her).

Because prices do not adjust *instantaneously*, quantity constraints
can develop: businesses _have to_ cut back on production. This is
especially true for *large* shocks, in the face of which prices
would need to fall below historical cost, and inventories become too
expensive to carry. In a monetary economy working in historical
time, historical costs are often embodied in debt obligations or
other contracts. Cutting prices can lead to liquidity squeezes.
(Small shocks may leave the economy in Leijonhufvud's "corridor,"
near notional equilibrium. See his "Effective Demand Failures" in
his INFORMATION AND COORDINATION.)

This argument says that the price does not change instantaneously
from price at SR equilibrium point A at FE to price at SR
equilibrium point B at FE. So, the nature of B may be changed by the
process: the actual result, ex post, B' may be less than FE.

In any event, the quantity constraints in different markets
interact, so that limited (liquidity-constrained) demand for
products leads to quantity-constrained demand for labor, which in
turn limits demand for products. This process might be seen as one
version of pecuniary externalities or the multiplier effect (in
historical time).

sincerely,

Jim Devine   BITNET: jndf@lmuacad    INTERNET: jdevine@xxxxxxxxxxxxxxx
Econ. Dept., Loyola Marymount Univ., Los Angeles, CA 90045-2699 USA
310/338-2948 (off); 310/202-6546 (hm); FAX: 310/338-1950