RE: [OPE] constant returns to scale

From: Michael Williams <>
Date: Thu Oct 02 2008 - 04:29:16 EDT

Hi Anders,

I don't disagree with any of this.

Dr Michael Williams, BA, MSc, PhD

Mob +447906172655
Home tel +4423 80768641
 Help save paper - Do you need to print this email?

-----Original Message-----
From: [] On Behalf Of Anders Ekeland
Sent: 02 October 2008 07:24
To: Outline on Political Economy mailing list
Subject: RE: [OPE] constant returns to scale

Dear Michael,

- I am also a revolutionary socialist/absurdist

- The problem with this thinking around MC, AC,
MES is that they - as you point out in you next
contribution - change over time. It is part of
real life competition to change them, to try to
invest to create a plant where all foreseeable
market demand is in the area where you have IRS.
This - given the uncoordinated nature of the
capitalist investment process leads to
overinvestment, overproduction - to "social
competition costs" - and I think "regulated MC,
with subsidy" will not solve that problem.

As soon as you do not have a static model with
technology given (long run or short run) you get
a complex process. And let me add there is a
great difference between products - and I cannot
see that MC pricing is applicable in important
cases. Take the case of Norwegian
hydro-electricity - as soon as the dams etc. are
in place the marginal cost is - close to 0.
Should the long term price be close to zero? The
same goes for oil in Saudi Arabia - where
extraction costs are very low - but it has taken
nature millions of years to create oil - and the
alternative energy sources in the long run are
much more costly. But the long run here might be
50-100 years. My point is that with such enormous
"externalities" for fundamental energy goods only
"political", i.e. planned prices will do, MC
leads to enormous short term (= decades) of
inefficient, wasteful use - in a long term perspective.

So it is either OPEC or a democratic institution
(world parliament) that will decide these prices - MC will not.


At 21:57 01.10.2008, you wrote:
>IRS are indeed widespread (oxymoronically we
>might add: correlated with large firms). A
>corollary is that the minimum efficient scale of
>output is large (in relation to the maximum
>sustainable size of the market). At any level of
>output < MES, MC pricing necessarily implies
>losses (since the long-run average cost curve is
>down-sloping until MES). In the extreme case of
>so called 'natural' monopoly the MES is greater
>than the maximum extent of the market. The
>efficient market structure is then monopoly.
>But, of course, we then lack competitive forces
>to ensure that the single firm does indeed
>behave efficiently. And if they did so - perhaps
>under sanction of the competition authorities -
>they would be operating at a loss (since MC<AC).
>So, effective regulation would require subsidy,
>however achieved. So to answer your question:
>regulated MC pricing, with subsidy, which might
>pragmatically come down to AC pricing.
>(btw, I am a revolutionary socialist, not a
>market socialist in any sense that I have seen
>that term used. Actually, with a promising
>career stretching out ... behind me, I am
>probably more of an absurdist than anything else :) )
>Dr Michael Williams, BA, MSc, PhD
>Mob +447906172655
>Home tel +4423 80768641
> Help save paper - Do you need to print this email?
>-----Original Message-----
>[] On Behalf Of Anders Ekeland
>Sent: 01 October 2008 19:57
>To: Outline on Political Economy mailing list
>Subject: RE: [OPE] constant returns to scale
>Let me throw the question of *increasing* returns to scale (IRS) into
>the debate. IRS is the great taboo of neo-classical economics, but
>very important in real life, especially for the real big firms. In
>the extreme Microsoft, but also IKEA or any other volume producer.
>IRS is what any rational capitalist want to have, very often have
>since there are always some fixed costs. In the "knowledge" economy -
>where research and development costs in the wide sense are getting
>more important since production is getting more and more automatized
>(ICT, global standardisation etc.) - and thinking, programming,
>design, organisation almost as labour intensive as it always was, IRS
>maybe becoming even more important than in the Fordist economy.
>Without taking IRS into account no "results" regarding market
>socialism versus market capitalism has any bearing on real life. In
>real life IRS leads to firms loosing in competition - this
>elementary fact - and the social costs/gains of this competitive
>selection process are never discussed. That makes such Mises vs.
>Lange debates "academic" in the negative meaning of that word.
>So how do you handle increasing returns to scale? What pricing
>mechanism is appropriate?
>At 17:08 01.10.2008, you wrote:
> > > Measuring marginal costs is impractical centrally or in a
> > de-centralized fashion,
> >
> > > because they vary with level of output. You can assume this away and then
> >
> > > assume constant returns to scale, so that MC = Average Cost, but
> > this then has
> >
> > > none of the efficiency advantages of MC pricing. This is quite
> > independent of the
> >
> > > incentive problems to which you refer.
> >
> >
> >
> >Hi Michael W:
> >
> >
> >
> >Yes, but don't most Marxian models also assume constant returns to scale?
> >
> >
> >
> >I suppose this would be legitimate in a static model, but surely any
> >
> >truly dynamic model of growth and capital accumulation must not assume
> >
> >this. This must be the case since technological change in means
> >
> >of production and the processes of the centralization and concentration
> >
> >of capital require (and, indeed, express) economies of scale.
> >
> >
> >
> >In solidarity, Jerry
> >
> >
> >
> >
> >
> >
> >
> >
> >_______________________________________________
> >ope mailing list
> >
> >
>ope mailing list
>No virus found in this incoming message.
>Checked by AVG -
>Version: 8.0.173 / Virus Database: 270.7.5/1702
>- Release Date: 01/10/2008 09:05
>ope mailing list

ope mailing list
No virus found in this incoming message.
Checked by AVG -
Version: 8.0.173 / Virus Database: 270.7.5/1702 - Release Date: 01/10/2008 09:05

ope mailing list
Received on Thu Oct 2 04:36:58 2008

This archive was generated by hypermail 2.1.8 : Wed Dec 03 2008 - 15:12:03 EST