Wage Determination

["Alan G. Isaac" <AISAAC@xxxxxxxxxxxx>]

I have argued in earlier posts that Keynes's _GT_ does not contain
a theory of nominal wages and that it crucially needs one. I also
said that while it would be nice to have a theory determining the
level of the nominal wage, I'm willing to settle for a theory of
wage dynamics. I view this as less demanding, since if we could
tie the level of the nominal wage to the state of the economy then
we could tie changes in the nominal wage to changes in the state of
the economy, while if we can only link wage changes to the state of
the economy we cannot similarly recover the level of the nominal
wage. (For example, the wage Phillips curve has been criticized for
allowing temporary unemployment shocks to have permanent effects on
the real wage.)

My attraction to the framework of the _GT_ is that it takes seriously
the notion of involuntary unemployment as an equilibrium position.
Of course, Keynes was criticized for the apparent exogeneity of the
nominal wage in his system. Since wages don't appear to be a "jump"
variable in real economies, I am not particularly troubled by treating
the nominal wage as a predetermined endogenous variable. That is,
allow the level to not change immediately in response to economic
conditions, while   wage dynamics _do_ respond to economic conditions.
On the one hand, this appears to justify working with an exogenous
nominal wage in a model of short-period equilibrium. On the other
hand, it seems to me that Keynes undercuts this justification when
he explicitly acknowledges that changing wages should affect
effective demand. This is why it seems that the _GT_ is missing
a crucial component: a theory of nominal wage dynamics.

Specifically, if effective demand depends on nominal wage dynamics
and if nominal wage dynamics depend on the policy influenced state
of the economy, then predicting the effect of policy becomes
problematic in the absence of a theory of nominal wage dynamics.
In addition, questions about the stability of unemployment
equilibria will hinge crucially on the nominal wage dynamics.

So far, I hope I have made my basic concerns clear. Now I wish to
raise some doubts about what I have understood to be one of
Paul Davidson's claims in this arena. (And as I seem to have
misunderstood these claims before, this may  certainly just be
my current misunderstanding.) I took the claim to be that a
theory of nominal wage determination that is fully compatible
with the goals and structure of the _GT_ was already floating
around the older PK literature. Specifically, ch.6 of Weintraub's
_Income Distribution_ book was to elucidate this. However, I find
chapter 6 to offer only i. an exogenous nominal wage, or ii. a
simple Phillips curve story. (This is not to be dismissive of
wage Phillips curves, which I find a good first stab at a
theory of wage dynamics (not levels).)

p.125 "Let us now incorporate the presence of involuntary unemployment
...The simplest, and reasonably realistic, procedure would be to draw
the ... labor-supply curve as ... perfectly elastic ..."
p.126 "[I]n a sense it would also be legitimate to regard
this wage level as an exogenous datum."

And for a story with either Phillips curve or efficiency wage
type overtones, depending on how you read it, after an
expansion of aggregate demand ...

p. 127 "it is the rise in labor demand [i.e., an increase in
effective demand] and decrease in labor offerings
[due to a price induced real wage decrease] at the going wage
level which absorb some of the involuntary unemployed, tighten
the labor market, and lead causally to higher money-wage levels."

So I find Weintraub ch.6 compatible with my earlier lists of how
PKs deal with the gap in the general theory when they desire a
formal model of unemployment equilibrium:
i. ignore the problem and maintain an exogenous nominal wage
ii. add some kind of wage Phillips curve
iii. invoke some kind of efficiency wage story
iv. tell some kind of conflicting claims story
(Again, these are not intended to be all obviously disjoint nor
necessarily exhaustive.)

Again, recall that my claim was not that nobody had offered solutions
for this problem in Keynes, but rather
i. there is a serious problem, unresolved by Keynes, for those of
us interested in modeling economies with involuntary unemployment
ii. there has not yet been convergence, theoretical or empirical,
on a solution to the problem.



[][][][][][][][][][][][][][][][][][][][][][][][][][][][][][][][][][]
[]    * *      * *   *                                            []
[]   ** *     ** **  *       Alan G. Isaac, Associate Professor   []
[]  *.* *    *.* * * *       Economics, The American University   []
[] *  * *** *  * *  **       Washington, DC 20016        U.S.A.   []
[] FAX:  (202)885-3790       Internet: AISAAC@xxxxxxxxxxxx        []
[] Phone:(202)885-3785       Bitnet:   AISAAC@xxxxxxxxxxxxxxxxx   []
[][][][][][][][][][][][][][][][][][][][][][][][][][][][][][][][][][]