Re: GT and microfoundations

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

Paul Davidson wrote:

> No what I am saying -- and clearly you do not understand -- is you better
> start by being  LOGICALLY CONSISTENT.

Again, this simply evades my point that
being logically consistent is simple: it
says nothing in favor of the *usefulness* of
a model to note that it is logically consistent.
(Unless by logical consistency you mean something
more than the conclusions of the model following
from its premisses. If so, what?)
In fact, most of the action will be elsewhere.
Every model will ignore most of reality, since
reality is tremendously complex. Most of the action
is in deciding which parts of reality it is most useful
to ignore for a particular purpose.

Furthermore, logically flawed arguments may
not imply the uselessness of the model derived
from them. We can hope that analysis of the
flaws will be productive, of course.
But to agree with Asimakopoulos's
critique of Keynes's derivation of aggregate
supply is not to conclude that the GT is useless.
Or (to provoke you further), to recognize that
the world is not ergodic does *not* imply econometric
modeling is useless for policy purposes.
For example, over policy relevant time horizons there
is enough inertia in macroeconomic relationships
most of the time that we can produce reasonably good
macroeconomic forecasts.

Not always, of course. Did mainstream economists
predict the Asian crisis? No. Did PK economists? No.
Does this mean forecasting is useless? No.
It means it is imperfect, and sometimes badly so.

Finally, logic is merely a formalism that is
well recognized to be inadequate to human
reasoning. (Which is *not* a justification of
logical contradictions.) Many of the moves
made in modeling are not simply deductive
and are thus not simply a matter of logical
consistency.


> If not, of course, ex post you can
> explain anything and say that you are just being empiricially
> relevant.  You apparently don't care-- as long as one explains the
> past.

Explaining the past is of course important.
Everyone does it. (In published work, you
speak of what the ``historical record suggests''.)
The test of our explanations of the past
is how they help us anticipate the future.
(Btw, that is why your repeated points about
analyzing random number miss the boat:
out-of-sample fit is terrible for such
exercises.)


> Finally I wonder why you are so anxious to avoid the obvious -- namely that
> in an economy that is market oriented, people use nominal contracts to
> organize production and exchange activities (ubiquitously) -- and it is
> only in such economies -- independent of the degree of monopoly  or
> competition, that involuntary unemployment is a persistent
> phenomena.

Where have I suggested otherwise??


> Moreover the use of such nominal contacts would violate the
> most universal of all the axioms underlying all the pragmatic "most"
> economists' theories  that you cite -- namely that demand functions are
> linear and homogeneous of degree zero.  But how can they be if people in
> the model "know" that money is not neutral?

This is just hand waving, Paul.
No one require linearity.
The rest of the question is answered
in any competent mainstream model.
Since you harp on logical consistency,
consider the Lucas 1973 JET model,
which is admirable both for its internal
logical consistency and its uselessness
for policy purposes.


>   Empirical evidence was that if you weighed a log and a stone and
> then  threw them into a fire, the remaining ashes of the log weighed less
> than the original log while the stone weighed the same.  Therefore the loss
> in weight of the log was do to the
> 'phlogisten" ' that came out of the log in the form of light and
> heat.  Since the stone did not lose any weight, it did not contain any
> phlogisten.  This theory was not logically consistent with other things
> that we now know as explaining oxidation -- but what the heck phlogisten
> beleivers were only being "pragmatic and realistic".

No. They were the theorists.
The pragmatists gathered the data
which allowed development of the better
theory.


> What can applied monetary theorists tell us about the lack of inflation in
> the US in the 1990s  and the inflation in East Asia, less inflation
> in  Brazil, etc -- that is consistent across all cases =-- or should we
> have a different explanation  for each historical episode -- ??

They tell us that central bank behavior is crucial to
inflation outcomes, among many other things.
Are we still learning about how? Of course.
Is there an adequate mainstream explanation?
I haven't seen it.
Is there an adequate PK explanation?
I haven't seen it.



When I said:

> >What were the PK policy predictions in the
> >face of the growing budget surplus?
> >(The answer is in the archives of this list!)

You replied:

> PK should not have said anything -- except only if the private sector shows
> a lack of effective demand then one solution is for government to spend
> more  --with or without a deficit.   The existence of a deficit per se -- or
> lack of one -- is not relevant.

We agree on the deficit.
And I note with glee that you consent to **predict**,
as **we all must**, in order to talk about policy.
(Of course, you prefer to speak of ``solutions''
rather than ``predictions'', but it is still ex ante!!)

When I said:

> >we all
> >must predict all the time, and policy cannot
> >be made without prediction.

You replied:

> Speak for yourself Alan -- not for WE!

But look at what you did when you tried to talk
about policy, Paul. You may not like to call
it a prediction, but to ``pre-dict'' just means
to say ahead of time, which is just what you
did. This is unavoidable.

We cannot even speak without anticipating
the future: we pick our words based on our
anticipations about how they *will* (i.e., in
the future) be understood. To act is to act
toward an anticipated future, whether you
are crossing the street or making policy.


When I said:

> >But back to the point I raised: how do PKs
> >differ from the mainstream on the neutrality
> >of money. I believe you largely dodged this
> >by mischaracterizing the mainstream.
> >For example, I claim you simply misread Friedman.
>

You replied:

> I did not interpret Friedman ,  I QUOTED him -- there is a difference Alan!!

Be serious: you would not have offered that quote
as you did if you did not *interpret* it to support your claim
about his position.
I was simply pointing out that it did *not* support your claim
---or, if you like, that to offer it as support for your claim was
evidence of your misinterpretation of Friedman.
I challenge you to offer one quote from Friedman that cannot
be interpreted in the more reasonable way I suggested.




Finally, to something more important. When I said:

> >Do you deny that there are many circumstances
> >in which a one-time, modest, change (endog
> >or exog) in the level of the money supply has no
> >important (i.e., policy relevant) effects on consumption
> >and production decisions after a few years?
> >If you do not deny this, how does that square with
> >your insistence that money is never neutral?

You replied:

> I stopped beating my wife a long time ago

I do not understand.
Do you deny it? Or not deny it?
(I assume you do not deny that you have
stated that money is never neutral.)
Or are you unwilling for obscure
reasons to deny it or not deny it?
This is precisely where I (and I assume
others) need you to come clean in order
to understand your position.
How about a straightforward answer?


> May be you ought to read what I wrote and stop interpreting it??

Since I've read plenty of what you've written,
perhaps you should take some responsibility
for my remaining confusions.  ;-)

Alan