Tuesday, January 22, 2008

Is There Anything Worth Keeping In Neoclassical Microeconomics?

Bernard Guerrien has joined with Emmanuelle Benicourt in updating Guerrien's prior objections. Deirdre McCloskey chaired an URPE-sponsored session at the Allied Social Science Associations (ASSA) conference. Donald Katzner and Karl Case each defend, more or less, the textbooks.

Guerrin and Benicourt still find contradictions in the textbooks, arising from treatment of the question of how prices get changed. Examples are given from Stiglitz and Mankiw's textbooks. (Previously, Guerrin had used mistakes in McCloskey and David Friedman's textbooks.) Guerrin and Benicourt ask, "Does teaching elementary economics authorize elementary faults of logic?" The respondents put on good cases, too. I learn from Katzner that an auctioneer is not needed; all of the agents in the market can quote prices. Katzner, however, doesn't see how false trading - that is, transactions actually taking place at non-equilibrium prices - can be allowed for in the models.

A theory with logical contradictions is invalid. A theory with implications that are repeatedly falsified empirically is incorrect, no matter how useful it is empirically elsewhere. One cannot conclude, though, that a better theory is not some minor tweak of such a theory. The question of whether researchers should concentrate on such tweaks or look at radical alternatives is a matter of judgement. I think good teaching ought to point out flaws and alternatives.

28 comments:

Anonymous said...

I would recommend Steve Keen's book Debunking Economics as a comprehensive account of the logical contradictions and limitations of mainstream economics.

After reading his account of the illogical requirements necessary to construct market wide demand curves, I looked at David Friedman's introductory textbook. Friedman gave the same illogical arguments to justify it as Keen dissects in his critique.

Equally interesting, Keen gives some of the back ground to why neo-classical economics got itself into this mess. Simply put, assuming that utilities could be measured between people opened the door to providing an economic justification for taxing the rich.

By arguing that utility was totally subjective, that possibility was avoided. Unfortunately, it also meant that it was impossible to aggregate individual demand curves into a well behaved market wide one... Opps!

Of course, this change was portrayed as "scientific" and "objective" -- that it happened to benefit the property owning class was just pure co-incidence...

Iain
An Anarchist FAQ

T.Pettinger said...

As an Economics student and now teacher, I'm always puzzled by the obvious deficiencies in Classical economics. They start with unrealistic assumptions that patently don't match reality.

T.Pettinger
Economics Help.org

YouNotSneaky! said...

"I would recommend Steve Keen's book Debunking Economics as a comprehensive account..."


Nooooooooooooooooooooooooooo!
Look, there are decent critiques of neoclassical economics out there. Some of the stuff Robert post is part of that. But the Keen stuff is just plain nonsense. In fact, I will go as far as to say that part of the reason why many critics of neoclassical economics don't get the attention/respect they perhaps deserve is because they are unwilling to distance themselves from cranks like Keen. This makes it easy for (some) neoclassical defenders to just dismiss all of them wholesale. People like Herbert Gintis have already recognized this.

Anonymous said...

"But the Keen stuff is just plain nonsense."

Rubbish. Keen critique is spot on, particularly on the matter of aggregation of demand curves -- on which he builds his arguments on leading neo-classical economists who have discovered the limitations of their own ideology.

So, again, I would recommend Keen's book. It is well documented and well argued. It is hardly his fault that the believers in the faith cannot understand his arguments -- nor the basic laws of mathematics...

Iain
An Anarchist FAQ

h.e. said...

"It is hardly his fault that the believers in the faith cannot understand his arguments -- nor the basic laws of mathematics...
"

Well, the material that you refer did not originate with Keen, is better presented elsewhere and is not an external criticism (The results have been published in mainstream journals and textbooks.).

YouNotSneaky! said...

"Keen critique is spot on, particularly on the matter of aggregation of demand curves"

You're the one talking rubbish. In fact this is particularly the matter where Keen is completely confused.
He's confused between the DMS Theorem and the Gorman Representation Theorem.
He's confused between the meaning of the words "identical" and "one"
He's confused between the concept of "aggregation" and that of "uniqueness"
He's confused between necessary and sufficient conditions.
He's confused between different meanings that the "representative consumer" can have (admittedly this one other folks are confused on too)

Other areas he's confused about include the difference between a single firm maximizing its profits and all firms in an industry colluding to maximize the total industry profit.

And so on.

Seriously. I've been around these discussions for quite awhile. If there's any blind faith involved here, it ain't me. You seem very eager and ready to take Keen at his word simply because it fits in with your ideology. Again, I ask, whatever happened to 'question authority'?

Anonymous said...

"Other areas he's confused about include the difference between a single firm maximizing its profits and all firms in an industry colluding to maximize the total industry profit."

Actually, he is not. He is showing that the mathematics used is wrong in the neo-classical case.

"Seriously. I've been around these discussions for quite awhile. If there's any blind faith involved here, it ain't me. You seem very eager and ready to take Keen at his word simply because it fits in with your ideology."

Oh, I know -- it is the only economics book I have ever read... Except, of course, I have read quite a few. Hell, even my original post here indicated that I looked up Friedman's book to see if Keen was right and, guess what, Keen was.

I wonder why I bother posting, as clearly no one actually reads what I write. I assume that it is easier to simply think otherwise because it fits in with some people's ideology...

"Again, I ask, whatever happened to 'question authority'?"

Oh, I know -- I should just give up and accept everything you assert as being 100% true. Yes, you are right -- I will now "question authority" by accepting your authority on this matter and shut up.

Thank you. I know feel that I am a free thinking individualist who will never, ever, question the proclamations of neo-classical economics because, as you assert, they are true...

And in this new spirit of "question authority" I will accept your assertion that Keen is "confused" rather than ask for an explanation why.

I had no idea that "question authority" could be so easy... Thanks, I know realise where I have been going wrong!

Iain
An Anarchist FAQ

h.e. said...

"Actually, he is not. He is showing that the mathematics used is wrong in the neo-classical case."

Keen is arguing that the standard rule about maximizing profit by equating marginal revenue and marginal cost is fallacious, because firms could obtain higher profits by colluding (jointly acting as a monopolist). Of course, the definition of a competitive firm is that it takes the actions of it's competitors as given, implying that it could indeed increase profits by "cheating" and unilaterally producing more where, surprise, surprise, marginal revenue equals marginal cost. The issue is not so much one of correct maths as of what the "correct" solution concept is...

YouNotSneaky! said...

"The issue is not so much one of correct maths as of what the "correct" solution concept is..."

Obviously, for the case of collusive cartels, the correct solution concept is the collusive maximization of industry profits, and for the case of a competitive industry, the correct solution concept is maximization of profits by each individual firms taking the actions of others as given (since that's what 'competitive' means).

Keen seems to think that the collusive solution concept should apply in both cases, for some reason. In fact, Keen's maths are never wrong (at least I didn't see any mistakes). It's just that he's always proving the wrong thing.
He says - I'm gonna prove that A
And then goes on to prove B, which is at best tangentially related to A.

The most egregious example of this kind of rhetorical switcheroo was actually in regards to the demand aggregation mentioned above (and I don't think Friedman wrote much on the subject, so I don't see what he, or his book, read or unread, has to do with anything. Which one anyway?) where he claims that in order to aggregate individual demands there must be a single consumer and a single good and then proves that ....... the square root of 2 is irrational!

YouNotSneaky! said...

And speaking of not reading what someone wrote:

"I will accept your assertion that Keen is "confused" rather than ask for an explanation why. "

"He's confused between the DMS Theorem and the Gorman Representation Theorem.
He's confused between the meaning of the words "identical" and "one"
He's confused between the concept of "aggregation" and that of "uniqueness"
He's confused between necessary and sufficient conditions.
He's confused between different meanings that the "representative consumer" can have (admittedly this one other folks are confused on too)

Other areas he's confused about include the difference between a single firm maximizing its profits and all firms in an industry colluding to maximize the total industry profit."

Anonymous said...

"It's just that he's always proving the wrong thing.
He says - I'm gonna prove that A
And then goes on to prove B, which is at best tangentially related to A."


Have yous been editing the Wiki entry on Keen, by any chance?

I would say that yous have obviously mis-read what Keen is arguing. He is not saying that the firms "collude", he is arguing that when you correct a basic mathematical mistake made by neo-classical economics then that dogma no longer proves that perfect competition is different from monopoly.

And as someone noted, Keen does get his mathematics right... Which means that neo-classical economics gets its mathematics wrong, which is Keen's argument. And it does. It assumes that the individual firm in perfect competition faces a horizontal demand curve. But the demand curve for the industry is slopped. This is impossible.

Taking the horizontal demand curve, we get f'(x) = 0. Which is basic calculus. Keen then shows where this leads once the mistake is recognised. Hardly confused, unless confused means "unwilling to accept illogical neo-classical assumptions."

Still, I suppose I should "question authority" and agree with your assertions about Keen being "confused" rather than, say, rereading Keen's analysis and comparing it to neo-classical economic text books.

And sorry for the delay in replying, but I was ill.

Iain
An Anarchist FAQ

YouNotSneaky! said...

"Have yous been editing the Wiki entry on Keen, by any chance?"

Yes I have. A while ago at least.


"He is not saying that the firms "collude","

No, he is saying they 'collude' without ever using that word.

" he is arguing that when you correct a basic mathematical mistake made by neo-classical economics"

There are two kinds of mathematical mistakes. The first kind is saying that 2+2=5. The second kind is writing down E=mc^2 when you're talking about F=ma. Keen asserts that neoclassical economists are making the first kind of mistake, and he proves it by making the second kind of mistake. Hence, he doesn't prove or correct anything since what he's talking about is irrelevant.


"then that dogma no longer proves that perfect competition is different from monopoly."

Because he doesn't understand what "perfect competition" means.
If you assume that perfectly competitive firms act as oligopolies, then ya, you'll get the result that there's no difference.


"And as someone noted, Keen does get his mathematics right..."

That was me.

" Which means that neo-classical economics gets its mathematics wrong, which is Keen's argument."

No, that conclusion doesn't follow from the above premise. Again, Keen gets his mathematics right in the sense that when he adds 2 to 2 he says it equals 4. But he gets it wrong in the sense that this isn't the relevant equation.

"And it does. It assumes that the individual firm in perfect competition faces a horizontal demand curve."

Nope. It assumes that firms 'act as if' or 'perceive' their demand curves to be horizontal. In fact, this is just a pedagogical device when teaching principles micro. You don't have to talk at all about a demand that a particular firm faces. Just say that 'firms take prices as given' - the definition of perfect competition here. And that's it.
At best. At a very very very best, Keen might have an argument here that the terminology of 'demand that a particular firm faces' is confusing. But it's just terminology, not in anyway integral to the model.

" But the demand curve for the industry is slopped. This is impossible."

Nope. One way this is possible on purely mathematic grounds is if there is a continuum of firms - just like a in a continuous probability distribution, the probability of any particular value, say 2, is zero, but the probability of an interval is positive. In fact, in theoretical papers sometimes authors explicitly state "we assume a continuum of firms".

But even then this assumption is not necessary. All that is necessary is that "firms take prices as given" - treat prices as parameters when they're maximizing their profits.


"Taking the horizontal demand curve, we get f'(x) = 0. Which is basic calculus. Keen then shows where this leads once the mistake is recognised. Hardly confused, unless confused means "unwilling to accept illogical neo-classical assumptions.""

This is where Keen says "sum of infinitesimals is not zero"! right? Well, he gets his maths right on this one, since technically speaking sum of infinitesimals is not zero. It's just arbitrarily small. It's less than 0.00000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000001

Alternatively I could just say that 'price taking behavior' means 'price taking behavior'.

"Still, I suppose I should "question authority" and agree with your assertions about Keen being "confused" rather than, say, rereading Keen's analysis and comparing it to neo-classical economic text books."

We all got our authorities. You question yours and I'll question mine. But there's sometimes also a point where you gotta trust someone who knows a lot more about this than you that someon eelse doesn't know what the monkey heck they're talking about.

"And sorry for the delay in replying, but I was ill."

Get well soon.

YouNotSneaky! said...

the above message cut off the number -
there's about 150 more zeros there, than a 1.

Robert Vienneau said...

I have commented on Steve Keen's work before. Keen has a difficult challenge: to sumarize these objections in a way accessible to an introductory student. I tend not to do this and think Keen has a mistep here or there.

I agree with the overall point of Keen's book. When all the logical objections to neoclassical economics are aggregated, not much is left. I have no problem with thinking of some of these objections (e.g., the Sonnenschein Mantel Debreu results) as own goals.

I'm not too enthusiastic about Keen's original development of a logically consistent partial equilibrium theory. I don't think of my opinion as final or definitiive.

Even here, I was surprised by Keen's criticisms. Iain highlights one. If the market demand curve is downward sloping, the firm demand curve - logically - cannot be horizontal. I also like the point that the supposedly U-shaped firm average cost curve cannot contain a downward-sloping portion if the market contains an infinite number of firms (perfect competition?) and production for the market is finite.

I think YouNotSneaky finds Keen a bugbear.

YouNotSneaky! said...

Alright, make this simple. There's no such durn thing as a 'demand curve faced by an individual firms'. There's just the market demand curve for a good and competitive firms which take prices as given. End of story.

When all the misunderstandings and mistakes in Keen's book are aggregated there isn't much left to his objections.

And yes, Keen annoys me. He manages to fool otherwise reasonable people. And the unwillingness of some of the heterodox economists to distance themselves from his nonsense really dents their credibility. It's as if they do not care whether a particular argument is correct or not, as long as its critical of THEIR bugbear, neoclassical economics.

Anonymous said...

"Alright, make this simple. There's no such durn thing as a 'demand curve faced by an individual firms'. There's just the market demand curve for a good and competitive firms which take prices as given. End of story."

So, when introductions to neo-classical economics write about "a horizontal demand curve" then they are just, what, making stuff up? When they draw a diagram which has a horizontal line on it, they are just inventing something?

Look, it is pretty easy. I've read the books, I've seen the diagrams. Now you are asserting that every neo-classical economics textbook in the world is wrong?

"When all the misunderstandings and mistakes in Keen's book are aggregated there isn't much left to his objections."

So, are you seriously saying that neo-classical economic textbooks do not state that individual firms face a horizontal demand curve and present this information as a straight line on a diagram?

"And yes, Keen annoys me. He manages to fool otherwise reasonable people."

Sorry, but I have checked Keen's analysis against the text books and Keen is right. I even double checked after your initial post and I can state that he does present an accurate account of the flaws.

"And the unwillingness of some of the heterodox economists to distance themselves from his nonsense really dents their credibility."

So, you are stating that neo-classical economic textbooks do not talk about a horizontal demand curve for individual firms (which represents them being price takers)?

"It's as if they do not care whether a particular argument is correct or not, as long as its critical of THEIR bugbear, neoclassical economics."

Sorry, but I've looked it up. I've checked the introductions. Here is one on-line:

"The industry, i.e. all producers taken together, faces a negatively sloped demand curve but the individual producer faces a horizontal demand curve, i.e. the firm is a strict ‘price-taker’. It can sell as much as it likes at the market price."
(http://members.shaw.ca/elementaleconomics/mic_5_2.htm)

As Keen points out, this is impossible. He corrects the basic mathematical mistake and notes that the neo-classical case against monopolies no longer holds.

Now, I've presented a quote from an introduction to neo-classical economics which states exactly what Keen is critiquing. This quote reflects the position I have seen in textbooks. It reflects what Keen says he is critiquing.

Now, please, explain why Keen is "confused." And please explain why "There's no such durn thing as a 'demand curve faced by an individual firms'" when introductions to neo-classical economics state that there is...

Iain
An Anarchist FAQ

Anonymous said...

A few more comments:

"No, he is saying they 'collude' without ever using that word."

I reread it and he suggests no such thing. Ironically, he notes that it is the perfectly competitive model which assumes firms collude when it states that an individual firm has no effect on the market -- which is impossible...

"Because he doesn't understand what 'perfect competition' means.
If you assume that perfectly competitive firms act as oligopolies, then ya, you'll get the result that there's no difference."


Which is not what he is doing. He is showing that there is a flaw in the model which, when corrected, disproves the neo-classical case against monopoly.

"Nope. It assumes that firms 'act as if' or 'perceive' their demand curves to be horizontal. In fact, this is just a pedagogical device when teaching principles micro."

It is much more than that. The fact is firms face the demand curve for the industry as a whole. They do not and cannot face a horizontal one. The problem arises when, as Keen stresses, this approximation is taken for reality.

As he discusses in his book... He also discusses the "as if" methodology as well and shows its limitations...

"At best. At a very very very best, Keen might have an argument here that the terminology of 'demand that a particular firm faces' is confusing. But it's just terminology, not in anyway integral to the model."

Yet the assumption is used elsewhere to draw conclusions. Which Keen shows are false.

"Nope. One way this is possible on purely mathematic grounds is if there is a continuum of firms . . . In fact, in theoretical papers sometimes authors explicitly state "we assume a continuum of firms".

Yes, I know. Paul Ormerod mentions that, noting that a continuum of firms is an impossibility in reality...

"But even then this assumption is not necessary. All that is necessary is that "firms take prices as given" - treat prices as parameters when they're maximizing their profits."

So all firms take prices as given, so how do prices change? Damn, that is reality creeping back in... Back to the theory -- and this is shown as a horizontal demand curve, which is then used to draw false conclusions...

"We all got our authorities. You question yours and I'll question mine. But there's sometimes also a point where you gotta trust someone who knows a lot more about this than you that someone else doesn't know what the monkey heck they're talking about."

I know, I'm getting that feeling right now... So, let me get this right, I should "question authority" with Keen, but accept on "trust" your comments. Interesting.

But as I said, I have looked at the textbooks and they confirm Keen's book. Now, perhaps you can put me in the direction of a textbook which you consider accurate and I will study it and see who is correct.

Iain
An Anarchist FAQ

YouNotSneaky! said...

"So, when introductions to neo-classical economics write about "a horizontal demand curve" then they are just, what, making stuff up? When they draw a diagram which has a horizontal line on it, they are just inventing something?"

Like I said before, it's a pedagogical device. You don't like the graph, don't draw it. The theory doesn't hinge upon it.
All you need is that firms are price takes.

"Now you are asserting that every neo-classical economics textbook in the world is wrong?"

Nope, they use a particular pedagogical device, which if, you're a real quibbler about unimportant details, you could say uses some improper terminology.

"So, are you seriously saying that neo-classical economic textbooks do not state that individual firms face a horizontal demand curve and present this information as a straight line on a diagram?"

Pedagogical device. Not the theory itself.

"So, are you seriously saying that neo-classical economic textbooks do not state that individual firms face a horizontal demand curve and present this information as a straight line on a diagram?"

Ped. Dev.

"So, you are stating that neo-classical economic textbooks do not talk about a horizontal demand curve for individual firms (which represents them being price takers)?"

PD


"Sorry, but I've looked it up. I've checked the introductions. Here is one on-line:

"The industry, i.e. all producers taken together, faces a negatively sloped demand curve but the individual producer faces a horizontal demand curve, i.e. the firm is a strict ‘price-taker’. It can sell as much as it likes at the market price."
(http://members.shaw.ca/elementaleconomics/mic_5_2.htm)
"

What matters here is that the firm is a 'price-taker'. The line about individual producers facing a horizontal demand curve is a ... pedagogical device!

"As Keen points out, this is impossible."

If firms act as price takers than they act as price takers. You might not like that assumption but it is the assumption of the model so you cannot criticizes the model for making a 'mathematical mistake'


"He corrects the basic mathematical mistake and notes that the neo-classical case against monopolies no longer holds."

Ok, here's where Keen's crap really comes in. You don't like the talk about 'horizontal demand curves faced by an individual firm' - fine. Perhaps there's a point here that that is inaccurate terminology (maybe in the same way as if you called a monopolist's MC curve a 'supply' curve). But Keen doesn't correct this mistake, which would just involve dropping the terminology! He drops the assumption of price-taking firms and then shows that when firms do not act as price takers then you get the non-price taking, monopoly outcome! Whoopteedoo. All it shows is that Keen doesn't understand what 'competitive market' and 'price taking' mean.



"Now, please, explain why Keen is "confused.""

See above

"And please explain why "There's no such durn thing as a 'demand curve faced by an individual firms'" when introductions to neo-classical economics state that there is..."

PD

"I reread it and he suggests no such thing. Ironically, he notes that it is the perfectly competitive model which assumes firms collude when it states that an individual firm has no effect on the market -- which is impossible..."

No, it means firms perceive themselves as having no effect on market price. If they change the amount they produce in anticipation of how their production affects their price, they are no longer price-taking firms. And if they maximize according to Keen's stupid formula that they should produce where industry MC = industry MR, instead of the standard firm MC = firm MR then they are in effect colluding, whether you call it this or not.
There's a reason why he gets the monopoly result for a 'competitive market' (and Cournot) - a monopoly is the entire industry so in that case industry MC = firm MC. By making firms max by setting ind MC to ind MR he's making them act as one big monopoly - collude.

But that's not a competitive market.

"Which is not what he is doing. He is showing that there is a flaw in the model which, when corrected, disproves the neo-classical case against monopoly."

It's exactly what he's doing. Even though he obfuscates his chicanery.


"Yet the assumption is used elsewhere to draw conclusions. Which Keen shows are false."

Yes, but it's the assumption of 'price taking' not the so called 'horizontal demand curve faced by individual firms'... which is just a pedagogical device.

"Yes, I know. Paul Ormerod mentions that, noting that a continuum of firms is an impossibility in reality..."

Well, there's a deep insight by Paul Ormerod I guess. Seriously, there is a reason theory is called 'theory' and not 'reality'.

Double seriously, you could assume, say just 100 firms selling a differentiated good. Then you could solve a 100 maximization problems and equate 100 reaction functions. And in the end you'd get the result that each individual's firm impact on the market price is something like .000-hundred more zeros-0001.
Or you could just not play that stupid game and assume a continuum of firms.

Triple seriously, you can get the same result with just two firms selling a homogenous product and setting prices.

Finally. Here you actually have a real criticism of the theory - that a particular assumption is UNREALISTIC. Not "mathematically incorrect" or a "mistake" just not realistic. We can have that argument. But that's not Keen's argument.

"So all firms take prices as given, so how do prices change? Damn, that is reality creeping back in... Back to the theory -- and this is shown as a horizontal demand curve, which is then used to draw false conclusions..."

Again - now you're actually making a real criticism. A weak one, but real nonetheless. That's not what Keen does when he completely misunderstands the theory.

"I know, I'm getting that feeling right now... So, let me get this right, I should "question authority" with Keen, but accept on "trust" your comments. Interesting."

Yup.

And to end this I'd like to say that in fact it's actually a fairly useful pedagogical device which is why it's in the textbooks.

YouNotSneaky! said...

BTW

"Paul Ormerod mentions that, noting that a continuum of firms is an impossibility in reality..."

Let me note that corn being the only consumption good and labor the only production factor is an impossibility in reality... after all I'm typing this on a computer which was not produced with just corn and labor.

Can I get that insight published somewhere, or is it only silly criticism of neoclassical theory for which there is a market?

Anonymous said...

"Let me note that corn being the only consumption good and labor the only production factor is an impossibility in reality... after all I'm typing this on a computer which was not produced with just corn and labor."

Presumably a reference to Sraffa? I'm not aware that Sraffa used his model for anything but showing the flaws in neo-classical economics.

I'm not sure he made any policy recommendations based on this model -- unlike neo-classical economics. Perhaps I am wrong, though...

"Can I get that insight published somewhere, or is it only silly criticism of neoclassical theory for which there is a market?"

Opps, silly me. I quoted a respected economist who is critical of the absurd assumptions of neo-classical economics. I'll try and not do that again (I must remember that "question authority" means "do not question neo-classical economics").

Iain
An Anarchist FAQ

Anonymous said...

"Like I said before, it's a pedagogical device. You don't like the graph, don't draw it. The theory doesn't hinge upon it.
All you need is that firms are price takes."


Well, show me which text book shows that it does not hinge on it.

"Nope, they use a particular pedagogical device, which if, you're a real quibbler about unimportant details, you could say uses some improper terminology."

So this "pedagogical device" is wrong. It presents something which is not true. Wonderful form of teaching that...

"Pedagogical device. Not the theory itself."

And that is why the textbooks use it, because it is not part of the theory? Interesting. Okay, reference a textbook which explains the theory to your satisfaction.

"What matters here is that the firm is a 'price-taker'. The line about individual producers facing a horizontal demand curve is a ... pedagogical device!"

In other words, wrong. Yet if the firm is a price-taker then the price is constant, so they do face a horizontal demand curve... I can see why they use that "pedagogical device"!

"If firms act as price takers than they act as price takers. You might not like that assumption but it is the assumption of the model so you cannot criticizes the model for making a 'mathematical mistake'"

That assumption is, of course, unrealistic. And it implies that the price is constant and, therefore, a straight-line (and so f'(x) = 0). Keen then shows how this assumption leads to a contradiction in terms of how a monopoly operates.

"But Keen doesn't correct this mistake, which would just involve dropping the terminology! He drops the assumption of price-taking firms and then shows that when firms do not act as price takers then you get the non-price taking, monopoly outcome! Whoopteedoo. All it shows is that Keen doesn't understand what 'competitive market' and 'price taking' mean."

You really do not understand his argument, do you? He drops the assumption as it is "unrealistic" and impossible. A firm does not face a horizontal demand curve, but the market one. If it changes its supply, then the market supply changes and so does the demand.

"No, it means firms perceive themselves as having no effect on market price. If they change the amount they produce in anticipation of how their production affects their price, they are no longer price-taking firms."

In other words, the model has no bearing to reality as real firms can change prices and supply and this impacts on the market as a whole. It can only not do that if the other firms change their supply by the appropriate amount to counteract that change, which is unlikely.

"And if they maximize according to Keen's stupid formula that they should produce where industry MC = industry MR, instead of the standard firm MC = firm MR then they are in effect colluding, whether you call it this or not."

But his argument is not that "in effect" they are "colluding" but rather it makes sense for them to do so.

"There's a reason why he gets the monopoly result for a 'competitive market' (and Cournot) - a monopoly is the entire industry so in that case industry MC = firm MC. By making firms max by setting ind MC to ind MR he's making them act as one big monopoly - collude."

Nope, he is showing what happens when an individual firm faces the market demand curve rather than the impossible "horizontal demand curve".

"But that's not a competitive market."

And Keen's point is that the "competitive market" model is logically flawed...

"Yes, but it's the assumption of 'price taking' not the so called 'horizontal demand curve faced by individual firms'... which is just a pedagogical device."

It is a flawed device, then. But as noted above, I can see why it is used -- and how Keen shows that it is logically flawed.

"Well, there's a deep insight by Paul Ormerod I guess. Seriously, there is a reason theory is called 'theory' and not 'reality'."

Ormerod's basic point is that the theory in no way describes reality. If you seek to make reality fit into an impossible model, you are going to get into a mess... Keen makes the same point.

"Or you could just not play that stupid game and assume a continuum of firms."

Ah, right, making your models reflect reality are "stupid games". That explains a lot...

"Finally. Here you actually have a real criticism of the theory - that a particular assumption is UNREALISTIC. Not "mathematically incorrect" or a "mistake" just not realistic. We can have that argument. But that's not Keen's argument."

Keen simply corrects the mathematically incorrect horizontal demand curve and replaces it with the market demand curve. Once that is done, the model is shown to be flawed. Perhaps you can say that making the model mathematically consistent is the same as making it more realistic....

""So all firms take prices as given, so how do prices change? Damn, that is reality creeping back in... Back to the theory -- and this is shown as a horizontal demand curve, which is then used to draw false conclusions..."

Again - now you're actually making a real criticism. A weak one, but real nonetheless. That's not what Keen does when he completely misunderstands the theory."


Lovely, pointing out that the model is impossible in reality is a weak" critique! That does say a lot...

But, as I said, I've reread Keen and reread the textbooks. I will do so again.

"I know, I'm getting that feeling right now... So, let me get this right, I should "question authority" with Keen, but accept on "trust" your comments. Interesting."

Yup."


Well, sorry, but to coin a phrase, I will "question authority"

"And to end this I'd like to say that in fact it's actually a fairly useful pedagogical device which is why it's in the textbooks."

Well, I did ask for a textbook to be recommended that would show me errors of Keen's ways, but I am still waiting.

Also, it is a "fairly useful pedagogical device" precisely because it reflects the assumptions of the model. Keen analyses these assumptions, shows their flaws and indicates what happens when you correct them.

So what have we learned? Firstly, do not question neo-classical economics and take on "trust" what its supporters assert. Secondly, pointing out that a model is impossible in reality is a "weak" criticism of a model intended to describe reality. Thirdly, making your models reflect reality are "stupid games".

I guess I can see why people would be interested in Keen and critics of neo-classical economics....

Still, I will re-read Keen and re-read the textbooks and see who is correct. So far, Keen still seems to be on the ball. Perhaps I'm not looking in the right textbooks...

Iain
An Anarchist FAQ

YouNotSneaky! said...

I give up. This isn't worth it.

Bruschettaboy said...

I have to say it does rather look as if for Radek, any literally false statement made by Keen is an idiotic howler which destroys not only his credibility, but that of anyone who fails to vociferously denounce him, while any literally false statement made in a neoclassical textbook is a "pedagocical device". And for Iain, vice versa.

YouNotSneaky! said...

"literally false statement"

I don't care about literally false. In fact what bothers me is precisely that it's false in spirit. Keen pretends that he's uncovered a mistake in neoclassical theory that all those stupid people over the past hundred years or so have missed based on something that is completely inconsequential for the theory and in fact isn't even part of the theory.

And he does this over and over again; claims to have proved something by proving something completely unrelated.

You stack those instances next to each other, and yeah, it becomes an idiotic howler if not a deliberate attempt at misinformation, i.e. charlatanism.

YouNotSneaky! said...

It's worth recalling what this argument's about. It is not over whether there are perfectly competitive markets in the world. It is not over accuracy in Principles books. It's not over whether "demand facing an individual firm" is inappropriate use of terminology.

It is about this:

In neoclassical economics the outcome of an idealized perfectly competitive market is fundamentally different than the outcome of a monopolistic market. Indeed, a good part of the reason why you teach the competitive market model is so that you can contrast it with the monopoly outcome.

But Keen claims that there is a "mathematical mistake" in the neoclassical model of the perfectly competitive market - a mistake that somehow the thousands of people who have worked on this have failed to notice until the brilliant Keen came along - and that once you "correct" this mistake there is no difference between monopoly and perfect competition.

This is of course patent nonsense to anyone who's not hung up on the notion that any and all criticism of neoclassical theory must be correct simply because they are criticisms of neoclassical theory. And it is ridiculous on a number of levels, the fact that Keen doesn't know what in the heck he's talking about being just one.

Robert Vienneau said...

Keen does not claim brillance or, for most of the chapters in his book, even originality. He claims to be summarizing the literature for the introductory student. Even where he claims originality in chapter 4, he claims that he is extending an well-establish argument. The inability of some to acknowledge the weaknesses in mainstream textbooks is not Keen's fault. As for Keen's unoriginality, consider:

"...the theory of the cost conditions of the firm in a perfectly competitive industry remains mired in contradictions. On some level, increasing, constant and decreasing costs are all incompatible with a partial equilibrium analysis of perfect competition..."

"...on both short and long run levels, there is no adequate resolution of contradictions between the partial equilibrium theory of perfect competition and the empirical evidence of non-increasing costs. Instead of a constructive resolution, perfect competition continues to be justified by the unsatisfactory verbal and logical contortions, denials of empirical evidence and proofs by assertion that originated more than fifty years ago." -- Avi J. Cohen (1983). "'The Laws of Returns Under Competitive Conditions': Progress in Microeconomics Since Sraffa (1926)?" Eastern Economic Journal, V. IX, N. 3 (Jul.-Sep.)

YouNotSneaky! said...

Yes, increasing returns are incompatible with perfect competition. Decreasing returns, generally, are fine. But like I said the argument "is not over whether there are perfectly competitive markets in the world.". It is about whether Keen presents the model he seeks to criticize in an intellectually honest way. And the answer to that is, no.

"Even where he claims originality in chapter 4, he claims that he is extending an well-establish argument."

Is that where he claims that this "insight" was originally due to Stigler? Well, then this is another case of mis-representation as Stigler was talking about an oligopolistic Cournot market, not perfect competition.

walt said...

There are no mathematical errors in neoclassical economics. This is not at all clear from undergraduate textbooks, but it's pretty clear from graduate textbooks such as Mas Colell, Whinston, and Green.

Iain's argument about a continuum of firms being necessarily realistic is not that decisive. Physicists use weird approximations all the time. They derive the wave equation for a string by assuming the string is made up of a continuum of little springs. They approximate a small magnet on a big refrigerator door by assuming that the refrigerator door is infinitely big.