> My comparison was merely meant to suggest certain characteristics of scientific
> revolutions, namely:

>>> No, I mean if there is a shift from neoclassical economics to something
>>> else, whatever that may be, some neoclassical ideas may still be
>>> found useful. So if I want to claim neoclassical theory is fatally
>>> flawed somewhere or other, I do not need to claim that all ideas and
>>> theories currently thought of as neoclassical are useless.

>>> Certainly a lot of analytical tools were developed in the Cambridge
>>> Capital Controversy. I don't see why so few mainstream economists
>>> are willing to learn them. There's nothing anti-neoclassical about
>>> turnpike theorems, the golden rule of growth, or joint production.


Sraffianism is a scientific revolution? Don't think so.

[snip]

> I think you're wrong there. The Sraffian surplus arises in production, and
> value is conserved, so to speak, in exchange. I suspect you have in mind
> the claim that both parties gain in trade. This is a different concept.
> Furthermore, it doesn't show the generation of a surplus, unless one
> wants to talk about (kinetic) energy being gained when potential energy
> is lost. The metaphor is fairly exact. Neoclassical economics is
> built on utility being a potental for a conservative field, as was shown by
> Mirowski and later commentators (Wade Hands).


Mirowski didn't show a didly-squat. It was a systematic comparison of
math used in std. neoclass Econ and in some Phys. Sraffians use same
terminology as neoclass Econ but w/ different content and/or defns? Other
than that, I don't see the point in the above paragraph.

> [...]

>> My impression is that Sraffian ideas run into dead end, while
>> neoclass did not.

> My impression is that Sraffians and Neoclassicals stopped talking
> to one another in journals where mainstream economics is dominant [1].
> But Sraffianism continues in recently founded heterodox journals.


Which are?

>>> The Sraffians claim they have revived the political economy of Petty,
>>> Quesnay, Smith, Ricardo, and Marx. Even if one finds some neoclassical
>>> ideas useful, this viewpoint leads to thinking of neoclassical
>>> economics as shunting economic science down the wrong track.
>>> Furthermore, others find the revival of political economy useful -
>>> Post Keynesians more generally, Institutionalists, Evolutionary
>>> Economists more generally, Marxists, and even Austrians.


>> As well as neoclass.

> I did think of claiming old Chicago school as political economics. What

Neoclass Econ NOT EQUAL old Chicago school. There's much more to Pol Econ
than that. Some feel Pol Econ will be the next fad, like Macro was in 70s
and I.O. in 80s.

>>>>> (Markku, if one could find a dynamically stable equilibrium in a
>>>>> neoclassical model in which a shift in parameters leads to a higher
>>>>> wage being associated with a shift to a more labor-intensive method
>>>>> of production, would you be surprised?)


>>>> No. I could generate such a result in few minutes.

>>> With a given set of technological choices? That is, the parameter shifts
>>> are somewhere else, say in utility functions. And I'm not considering
>>> principal agent problems, etc. If fact, assume everybody knows all
>>> technologies, prices, utility curves, etc.


>> These restrictions make it trickier, and haven't tought about it. Why
>> would you want to assume complete and perfect info?

> To show supply and demand do not explain long run positions
> in the basic neoclassical model, before the addition of all your
> qualifications. My discussion below assumes all these restrictions.


I don't follow your flow here. Even if supply and demand do not explain
long run positions -- whar ever that means -- endowments, techonologies,
info, preferences, and choice sets do. From these you can derive supply
and demand (under some conds).

> I generally do make these assumptions in discussions of Sraffian
> economics unless otherwise noted. Inasmuch as Sraffians think
> they have a criticism of the coherence of neoclassical theory, these
> assumptions are needed to investigate the logic of that claim.


Why? Logic or coherence of neoclass models in no way rest on complete
info.

> [...]
>>> Substitution need not characterize comparisons across equilibria.


>> As I stated in previous post, in GE, it can easily happen that in one
>> equil both price and amount consumed are higher in one equil than in
>> another. If this is what you mean, Yes..

> Economists involved in the Cambridge Capital Controversy spent a lot
> of time analyzing the choice of technique in long run positions. They
> found that if one compared choices at two different rates of interest,
> the technique chosen at the lower rate need not be more capital
> intensive. Likewise, the technique chosen at a lower wage can be
> less labor intensive. Again, a technique may be chosen that produces
> more of a good when that good's price is lower.
> 
> Am I making a statement about comparisons across multiple equilibria?
> Perhaps not. The participants in the CCC generally did not explicitly
> model utility. Neoclassical economists may be inclined to think of the
> discussion of the choice of technique as half an equilibrium. They might
> be inclined to add utility functions to complete it. Very little explicit
> work has been done with this sort of complete model, as far as I'm
> aware, and some of it is misleading.


What is misleading?

> I know one can find a single equilibrium in which a shift in utility
> functions will lead to more labor being demanded at a higher wage,
> even though the technology from which firms choose is given. This
> is a statement about a single equilibrium. I'm not sure if such an
> equilibrium can be stable. Even if the equilibrium is unstable, we may
> get interesting dynamics.
> 
> For example, suppose the labor supply is constant, that is, the
> reservation demand for labor is zero. We may find that the demand
> for labor in a long run setting rises and falls multiple times with
> the real wage, thereby creating a multitude of stable and unstable
> equilibria [2]. Consider a locally stable equilibrium. Small shocks,
> say, in the quantity of labor supplied, will send the wage back to
> that equilibrium. A small temporary increase in labor supplied
> will lead to a lower wage, which will gradually increase back
> to the equilibrium wage. But if the increase is large enough,
> the economy might move towards a new equilibrium with
> lower wages. An unstable "perverse" switch may have separated
> the two equilibria [3].

> [...]


>> I still don't undestand what you mean by "scarcity indices". This is not
>> std jargon in Econ.

> I did not make it up. I'm sure you understand the claim that
> "scarce" goods have positive prices. And I bet you realize that


Yes.

> which goods are non-free cannot generally be determined from
> physical quantities; this determination is a solution property
> of the model [4]. This is true of both Arrow-Debreu and the


Yep.

> von Neumann model generalized to allow distribution to
> vary.


[Which v. Neumannn model?]

> Now, consider a Arrow-Debreu equilibrium in which, say,
> pure water is free. Suppose we gradually change the data by
> decreasing the endowment of water. Won't the price become
> positive at some point? If we continue to decrease the


Yes.

> endowment even more, will the equilibrium price continuously
> increase?


Not necessarily. If there were no income effects or if all the goods are
gross substitutes, then this will happen [allthough these might not be
sufficient conditions to guarantee monotonic comparative statics].

> In the generalized von Neumann model, endowments of
> produced goods are not data, but endogeneously determined.
> Nevertheless, we can ask if less or more of a factor is used
> at a higher price. Prices in long run models behave differently
> than in short-run models of intertemporal and temporary
> equilibria.


Why do prices in long run models behave differently than in short-run
models of intertemporal and temporary equilibria? And how?
 
> I still claim Sraffa was important in economists' understanding
> of those differences. And this understanding has led to a rejection
> of many previous ideas common among neoclassical economists.


Might be.