I reviewed indifference curves recently. Wikipedia states:
This conflicts with the position of Austrian-school economics, which is that it’s actually impossible for two or more things to provide the same utility at the same time. So it seems that Austrian-school economics can at most use indifference curves (and the concept of indifference itself) as imaginary constructions.
it is hard to bill yourself as a truly unique alternative to mainstream economics if you accept and use on of its most basic models.
Euclidean geometry has 23 definitions, five common notions, and five postulates. Non Euclidean geo differs in only one postulate, and yet they apply in completely different universes.
It suffices for Austrians to accept Say’s Law to overturn most of the mainstream policies [aka what really counts]. Who cares about the technical stuff nobody reads anyway?
This conflicts with the position of Austrian-school economics, which is that it’s actually impossible for two or more things to provide the same utility at the same time. So it seems that Austrian-school economics can at most use indifference curves (and the concept of indifference itself) as imaginary constructions.
If by “Austrian-school economics” you mean Rothbard and friends, then you are correct. They do claim that one cannot be indifferent between bundles of goods.
But, Rothbard is not the beginning and end of the Austrian school.
I don’t see why one cannot be Austrian and think indifference curves are a useful. In fact, it seems kind of odd to me that a group of people that claim to believe in “radical subjectivism” can make objective claims about other people’s preferences (that is what you are doing when you say people cannot be indifferent–you are making a claim about their preferences).
Anyway, this is one reason for me posting this anecdote from Hayek. He says he was attracted to indifference curve analysis and found it “most satisfactory”. Indeed, he even claims credit for introducing the model to John Hicks, who helped establish the use of the model in mainstream econ.
This is an opinion I’ve suspected for a while people within this “movement” unfortunately hold, but you’re the first person I’ve seen explicitly express it Dave. Indeed, I think the problems with neoclassical economics are precisely of this nature, though the work of both critiquing that paradigm and producing a complete alternative has not been done by Austrians yet (not withstanding the start made by Mises, Rothbard et al). The big problem is that the mistakes made with foundational assumptions lead logically to further mistakes considering the grander picture.
I guess the best I can do for now is provide some anecdotal evidence (this applies indirectly to your comment above too Student) from my own Master’s dissertation. The Walrasian theory of value and price taught at a core level to pretty much all economics undergraduates and graduates derives consumer demands following a set of constrained maximisation problems based on a given set of prices with perfect market information and “rationality”(defined in the sense each actor’s decision will be based on the optimal solution to this constrained optimisation problem based on perfect information) denoted for each actor based on the type of utility ordinalism I’ve already described above. Following this exercise the demands can be expressed as functions of prices (the parameters of the original exercise), and a steady state defined when the aggregated demand function/curve intersects a supply curve/function derived using a similar profit maximisation exercise relying on similar perfect information assumptions. The nature of the equilibrium produced is then usually denoted, with Pareto Optimiality/Welfare theorems, along with the very strong assumptions required to produce it. Then based on this caricature/model the deviations produced logically from this model of a market are used to critique and attack it in the form of models of imperfect information, competition etc thus denoting market failures. Thus these alternative market models have a strong reliance on core (perfectly competitive) Walrasian price theoryand do not dispute its underlying logic but only change underlying assumptions.
The problem for both sets of models are numerous. There is an unavoidable circularity produced as a consequence of the fact that demand has to be derived from on optimisation based on given prices while the prices themselves are thus based on demands. Hence you cannot explain or account for actual price formation with this approach or even on a basic level how markets made up of ordinary people clear, only explain the maintainence of a hypothetical equilibrium steady state already arrived at only with strong perfect prior information of this equilibrium price vector. Walras himself came up with an ad hoc device to try to account for price formation and salvage this model called the Walrasian auctioneer who “shouts” out prices at which a market can only clear when the same amount can be bought and sold by traders in the market at that price. As nonsensical as this sounds in a 2 good market (the only observed context for this type of institution with mone traded for a bought good on the French Stock Exchange), it becomes even more confounded if you try to think of it as a way of accounting for price formation with n goods using a hypothetical numeraire which is the actual field general equilbrium price theory is applied to, with no necessary tendency toward general convergence should clearing be closer in one market following an announced set of relative prices by a hypoethetical auctioneer (I think Jehle and Reny briefly mention these issues in their textbook too).
On top of that, 5 decades of results from experimental economics starting with Vernon Smith’s influential paper in 1962, fly in the face of the strong assumptions required for market clearing/equilibration required by Walrasian General Equilbrium theory. Typically double auction/posted offer auction experiments with small numbers of buyers/sellers in double auctions with no prior information of market conditions or commercial display strong convergence properties toward trading at equilibrium prices within a few periods of “ticker time.” These results on the other hand are not difficult to understand at all given an understanding of Austrian School price theory, principally, Bohm Bawerks’ theory of marginal pairs (indeed Smith has revealed he was indirectly inspired by Bohm Bawerk’s work), which provides a logical/praxeological account for why this convergence takes place. In addition, arguably even more holes have been punctured in the Walrasian account by results from the nascent agent based literature on price formation involving so called “zero-intelligence” agents with only reservation prices (Zi-C agents) or with reservation prices and simple price update rules (Zi-P agents), who yet display narrow trading around and with Zi-P agents rapid convergence toward trading at equilibrium prices. Indeed the logic of the update algorithm for Zi-P agents is very strongly reminiscent of precisely the type of thing we see in Bohm Bawerkian price theory, whereby the fact that supramarginal traders may prefer not to overbid but only at the risk of losing a trade, and otherwise simply copy bids of marginal traders allowing market equilibration and price formation to be organically achieved without perfect prior information and only actors with ordinal, unequal preferences implying reservation prices. Interestingly Zi-P agents and their more advanced simple intelligence variants were found in the early 00’s to outperform human traders and have largely replaced them as far as trade execution goes in financial markets. Their consistently efficient performance, in-spite of their sparse “rationality” cannot be accounted for by a Walrasian framework and indeed contradicts it completely if we wish to take that theory’s assumptions and implications seriously (indeed on another note, it was only until 2007 before a Game Theoretic Double Auction model with special markup assumptions could predict the result Smith and others have seen in experimental markets since 1962 and been simulated in agent based markets since the 90s).
Now Dave, I would never have realised these implications and connections with other literature had I not engaged in a study of the perhaps seemingly “tedious” technical details of Bohm Bawerk’s price theory (even Rothbard, inspite of his commendable sections in MES on production and factor pricing deals with the details of price formation too sparsely), as well much of the modern literature on price formation. Indeed there were issues and basic contradictions with Bohm Bawerk’s marginal pairs price theory only recently pointed out and resolved by Egger and Van Den Hauwe (writing in the 90s and 00s). I was able to take Bohm Bawerk’s logical account and also produce a generalised, algorithmic version of his price theory accounting for the exceptional cases they point out.
Finally at Student, I think Hayek is an interesting character as regards the popularisation of Walrasian price theory. Like in the above interview it seems he generally views it well, but I think he was a good enough economist to logically see problems with it, as for instance in his essay on “The Meaning of Competition.” I sometimes think if he had pursued his logic a bit more thoroughly, he would have been able to see issues with it. Having said that however, though he was aware of Bohm Bawerk’s price theory, perhaps part of the reason it didn’t develop into a fully fledged alternative approach was the fact that Bohm Bawerk in his treatment of both value and price was actually a utility cardinalist. It took Cuhel and Mises the Austrian ordinalist value theory (Plus Mises I don’t think ever wrote explicitly on price formation, assuming far too much characteristically of his readers). On another side note, I would add that I’ve gained the impression that a lot more people (not necessarily Pos-Keynesians), have realised problems with and do have problems with Walrasian price thoery and are no necessarily “Austrians”, so it is fair I think to recognise the issues with this framweork you are defending (though it is taught to all econ. students, and I think has unfortunate consequences when they take it too seriously, contrary to much of modern research).
In any case I’ve spent far too long writing this, I hope it makes some sense, but in any case that’s it I’m done! 
This is an opinion I’ve suspected for a while people within this “movement” unfortunately hold, but you’re the first person I’ve seen explicitly express it Dave.
I got it from Mises. Here’s the quote, emphasis mine:
…the public’s interest in the studies labeled as economic is entirely due to the expectation that one can learn something about the methods to be resorted to for the attainment of definite ends. The students attending the courses of the professors of “economics” as well as the governments appointing “economic” advisers are anxious to get information about the future,
That’s what economics is about, or is supposed to be about. How to predict the future. How to attain a definite end. Which is why I asked earlier for a link to a single correct prediction [or even a prediction that was incorrect, something], that mankind has gotten from all those phony models, directly or indirectly.
I have to hand it to Keynes. He at least made predictions and had prescriptions. Never mind that he was howlingly wrong and got everything totally backwards. At least he understood what target he was supposed to be shooting at.
But all this stuff you described at length in your previous post clearly has no connection, however remote, to those two objectives. Predict the future, and tell us what to do about it.
absk,
There is an unavoidable circularity produced as a consequence of the fact that demand has to be derived from on optimisation based on given prices while the prices themselves are thus based on demands. Hence you cannot explain or account for actual price formation with this approach…
I don’t think I would call it a “circularity” but you are right that the traditional model of perfect competition yields little insight on how (or whether) the market arrives at equilibrium prices. In the model you have consumers that take prices as given and producers that take prices as given. So, if no one sets prices, how are they determined? Supposedly “the market” does this. But in this model, the market is treated essentially as a computer that finds the price vector that solves the system of demand and supply equations.
I would agree that this is not a very satisfactory description of the real market process that determines prices. A few years ago, I would have agreed with Austrians that this was a major problem for mainstream economics.
However, my subsequent readings have only convinced me that this only MAY have been a problem 50 years ago. Today, economists have moved well beyond perfect competition and routinely use game theory and experiments to understand the actual process of price setting. That’s basically a large part of what industrial organization is about, where economists study auction theory (which you seem familiar with), price discrimination, imperfect competition in general etc.
Of course, none of this is to say that the model of perfect competition isn’t useful. It is an approximization we can use to easily discuss comparative statics. And really it is approximation that is consistant with game theoretic models that try to better describe how competition actually works. Take the Cournot model of imperfect competition for example. If you take this model and start adding more and more players, you will reach essentially the perfectly competitive outcome (P=MC). Cournot himself essentially realized this almost 160 years ago. And undergraduates are taught this today.
On top of that, 5 decades of results from experimental economics starting with Vernon Smith’s influential paper in 1962, fly in the face of the strong assumptions required for market clearing/equilibration required by Walrasian General Equilbrium theory.
Why stop in 1962? George Stigler was writing about how the strong assumptions that are required for perfectly competitive equilibrium actually are not required in 1957 (see link below). And both Vernon Smith and George Stigler have won the the Nobel Prize. Are you implying their work has been ignored?
I think it is pretty obvious that mainstream economists have realized for a LONG time that the perfect competition model isn’t…perfect. And it is basically that realization that has driven much of micro theory over the past 60 years. That’s why I’m suprised so many heterodox folks spill so much ink over the flaws of the perfect competition model. It seems kind of played to death to me.
Student,
I see what you’re saying but part of my point was that the issues of imperfect competition models are directly related to those difficulties exposed by the above literature and analytical problems associated with the perfect competition model since the former share the same logical structure but only differ in starting assumptions and essentially share the same theoretical structure (I will comment on Game Theory below). Thus the fact that the market failures predicted by this same (generalised model), when perfect information and perfect competition (in terms of competitior numbers for instance) are not materialised or observed, and do not create the difficulties predicted (as in the example context of price formation above), this is a still fundamental problem for both forms of the same essential theory (whether perfect or imperfect).
Now I am hesitant to paint Game Theory, a very broad field, in the same manner with the same brush as this would also include approaches with evolutionary, bounded and alternative “rationalities” as well as models that go beyond simply generalisations (in terms of logical development from modified assumptions, like Cournot) of Walrasian price theory. I do think as a branch of knowledge it brings to bear its own problems nobody in the Austrian School has yet properly addressed (and I’m not going to be the first unfortunately), given its also very strong concept of actor “rationality” employed in standard game theory, of which I think the Traveller’s dilemma offers an interesting reductio ad absurdum puzzle/game that I think could provide an interesting starting point from which further inadequacies with it could be elaborated (I don’t think I have the strength in Game Theory to undertake such a project myself, though I think it would make for interesting research). In any case there wasn’t even a game theoretic model that was able to reproduce the convergence observed in double auction markets until 2007 (fitted with markup assumptions to do so by Zhan and D. Friedman).
Within the narrow context of price formation (as well as other areas), I think there is a platform for a return of the Austrian School’s approach and their core price theory as a distinct alternative to Walras’ (my point above was that while Walrasianism is not taken seriously in terms of its underlying assumptions reflecting reality, it’s core status in syllabi has a bad effect in instilling the logical structure by which the market is conceptualised). I will say as well, that I have gained the impression following the completion of my research project, that there are branches of modern, non-Austrian economic research paradigms that I don’t think necessarily neatly fit within the Neoclassical paradigm as a kind of “unified front” either(it may also be unfair to paint modern neoclassical economics as a unified research paradigm either, but actually more like a conglomerate of somewhat related, somewhat conflicting segments. On another side note, I don’t think Vernon Smith for instance shares that much enthusiasm for standard neoclassical price theory, whether perfect or imperfect, as he discusses with Russ Roberts here), and actually do share some unnoticed methodological affinities with the Austrian school. For instance, the key point conveying what can result with price formation via simply zero-intelligence agents as Gode and Sunder did in their influential 1993 paper (the zero intelligence agent based implementation of a double auction market. I do concede a die hard neoclassical, Gary Becker was actually just as much an inspiration for that paper as they reveal as Vernon Smith :P) could be thought of as an example of what Mises called the Method of Imaginary constructions. On another note, I will add that I think the terms “double auction” and “posted offer auction” are a unfortunate and very misleading, since neither institution is an “auction” as is dealt with in Auction theory. “Double-Auctions” are just market institutions in which the practice of buyers and sellers is to spontaneously shout prices back and forth and buy and sell accepting offers in real time like in a stock/commodities market. “Posted offer auctions” are simply markets in which price listing/“shouting” is only undertaken by one side (usually the sellers), while buyers buy based on whether they think they can later get a better deal or a price is less than their reservation, thus characterising the vast majority of real world markets!
Okay, I read your post 3 times and I am still not exactly sure what you are saying. In my post, I was commenting directly on mainstream economists using game theoretic models to understand price determination. But it sounds like you have problems with these models as well? But that no one has yet fixed these problems you think you’ve identified?
Well my post above was not specific to answering your remarks about game theoretic attempts to model price formation. To make a short comment, there are a few reasons i’m not enthusiastic about them. They first of all typically employ agent based “rationality” and information awareness to an extent that has been observed (and the conclusions made use of) in the experimental and even more so the zero intelligence based literature as not at all necessary to explain/realise price formation and market clearing in double/posted offer auctions. Secondly, even with these misleadingly strong assumptions they have largely failed to reproduce the basic convergence to market clearing observed in these markets (apart from the exception I mentioned above). They are still part of the paradigm whereby you start off with false assumptions and justify them when you get fits to data (in this case though these largely haven’t occured), in turn justify a tractable “model” built from false assumptions as if logical derivations from them can give you real world relevant knowledge/inference.
FTFY
I’m going to assume that you’re being deliberately disingenuous here. The question I have is, why?
To take indifference curves - and the notion of indifference itself - as anything more than imaginary constructions requires assuming that utility is cardinal.
I’m going to assume that you’re being deliberately disingenuous here. The question I have is, why?
It’s all he’s got.
Come to think of it, though I did mention this earlier I’ll mention it again (not sure if it will meet your criteria but here goes). The ZI-P agents developed by Dave Cliff( not an economist if that gives him browny points in your eyes) at Hewlett Packard labs in 1997 (whose algorithm for bid updates I’ve already mentioned shared essentially the same logic Bohm Bawerk espoused in his theory og marginal pairs price formation), were tested along with a few other slightly more advanced agents against human traders by IBM in 2001 and found overall not only to reproduce the price convergence dynamics of human bidders in stock markets, but also outcompete them in terms of overall bid performance. These agents have gradually and largely replaced trading floor humans over the past five years in liquid financial and commodity markets as far as trade execution goes, and saved the firms employing them a lot of money, thus reaping financial dividends. Does profitable, practical employment of these agents mean anything to you?
Strictly speaking I’m not sure the concept of indifference employed in indifference curves could be accepted in the praxeological framework, since actual indifference is used rather to define the units of supply of a good subjectively, as they are all equally replaceable to achieve an ordinally ranked marginal end in the mind’s eye of the actor, which is why all units of the supply share the ordinal value/importance of the marginal end that could not be achieved if the number of units dropped from n to n-1. Hence for an actor “indifferent” between using a lump of margarine or lump of butter to achieve the end of a tasty spread on his slice of bread, these might be viewed as 2 units of the same supply of a good for one actor, while they would be different goods for another not indifferent betweent their perceived means-end relations. What I don’t think has yet been properly addressed yet is where there may be overlapping ends that may not be marginal at the moment of a decision but may become so, so for instance, if the above situation holds but in an actor’s view butter and margarine share diffeent values as far as means for the end of cooking and frying an egg, then whether both lumps are supplies of the same means would morph depending on which end is marginal for the actor. A perhaps better, less volatile definition might be that 2 objects could only be the same good subjectively if for an actor they could be perceived as equally useful for achieving all conceivable ends they could be used for in the mind’s eye of an actor.
They first of all typically employ agent based “rationality” and information awareness to an extent that has been observed (and the conclusions made use of) in the experimental literature.
#1. ??? I am not sure what you mean when you say these models assume “information awareness” to an extent not observed in exprimental data. Incorporating incomplete information is the entire point of Bayesian Games!!! It is worth point out that auction models are Bayesian Games (typically where each player only knows the value he places on an item but not the value other bidders place).
#2. IMO, the notion that experimental evidence differs greatly from what is predicted by economic theory is over blown. I think this working paper makes good points related to this: http://www.dklevine.com/papers/theoryandexperiment2.pdf.pdf
Secondly, even with these misleadingly strong assumptions they have largely failed to reproduce the basic convergence to market clearing observed in these markets (apart from the exception I mentioned above).
???
#1. What do you mean when you say “these models” and “converge to market clearing”??? I guess you MUST be talking about a subset of double auction models not having a Nash equilibrium??? Because otherwise there are many types of game theoretic models that are all about price setting in imperfectly competive markets that have Nash equilibrium. If you are just talking about a subset of double auctions, then I can’t really comment. Auction theory is not my field, but I have had enough of it to know that there are at least some double auction models that have equilibrium (see Gibbon’s textbook on Game Theory for Economists for a solution to a 1 buyer 1 seller double auction).
#2. And if you are just talkin about double auction models, why just focus on them??? It seems kind of crazy to say we should abandon all of mainstream economics because a subset of of one type of auction model doesn’t have an equilibrum under certain circumstances (a claim I can’t confirm or deny).
I am guessing because double auctions could be seen as relevant for evaluating the validity of tâtonnement process where you have a single Walrasian auctioneer clearing the market??? That’s great, and all, but like I keep saying there are other models of price determination. And the “Walrasian auctioneer” seems like the LEAST realistic of them all. Maybe you want to focus on the tâtonnement process because it is more relevant for understanding equilibrium stability in a Walrasian General Equilibrium model (which you mentioned earlier)??
If this is the point you’ve been driving at (and I’m having to do a lot of reading between the lines to try and divine this intent) then I guess I can just say “fair enough”. Walrasian General Equilibrium models are not my focus either. (Side note: is Walrasian GE theory anyone’s focus these days?). But I will just say there is much more to microeconomics than GE theory. Milton Friedman was suspicious of Walrasian GE theory too. So am I. So is Brad Delong too apparently:
http://delong.typepad.com/sdj/2011/05/milton-friedman-we-curtsy-to-marshall-but-we-walk-with-walras.html
If not believeing in a Walrasian auctioneer makes you heterodox, then I wonder who is mainstream.
I’m going to assume that you’re being deliberately disingenuous here. The question I have is, why?
Is this why you bugged me to reply to your post?
To take indifference curves - and the notion of indifference itself - as anything more than imaginary constructions requires assuming that utility is cardinal.
This is must be a claim unique to you. This is not why Rothbard or Block or anyone else I’ve read objected to indifference (they claim it doesn’t make sense because indifference can’t be demonstrated by action).
Here’s Rothbard:
http://mises.org/rothbard/toward.pdf
Here’s Block:
http://mises.org/journals/qjae/pdf/qjae2_4_2.pdf
I don’t really know how to respond to your claim except to provide a link to the definition of ordinal utility:
http://en.wikipedia.org/wiki/Ordinal_utility
Like I said, is this really why you bugged me to respond to your post? Is the type of convo you wanted to have? That’s alright man. I’m cool.
No. I “bugged” you to reply to my post because it seemed like you were ignoring it, and I think that’s rude.
Now I’ll repeat my question: why are you (I assume) being deliberately disingenuous?
I wasn’t aware that I had to fall in lockstep behind Rothbard and/or Block.
I don’t agree with that definition of “ordinal utility”. To assert that the utilities of two goods are equal implies underlying cardinality. As I understand it, the Austrian-school economics definition of “ordinal utility” is such that the utility of one good is either higher or lower than the utility of another good.
Although this seems like quite a red herring to me, I’ll bite (for now). What type of conversation do you think we’re having, or that I wanted to have?
At the end of my last post I was basically trying to say I don’t want to continue our conversation (my questions were rhetorical). If you really think just dropping a convo here is rude, i’ll give my reasons for wanting to disengage:
#1. I raise (what I think is) a valid point, then you assert I’m being disingenuous and ask me to explain why rather than address the point I raised? I can’t respond to that.
#2. Then you start critiquing the concept of indifference based on your own definition of ordinal and cardinal utilities (I have not seen any Austrian ever say indifference is a cardinal concept or use the terms as you are using them) and want me to participate on those idiosyncratic terms. It sounds like you are basically just redefining the meaning of cardinal utility to include indifference. I guess the goal is that if cardinal means “bad” and you can redefine cardinal to include the concept of indifference that will make indifference “bad”. But I don’t think redefining common terms really addresses anything important about indfference. So I can’t really respond to this comment either.
IOW: Idiosyncartic definitions of established terminology and “when did you stop beating your wife” type questions kinda ruin it for me. So I will let you have the last word on this if you want to respond to either point. Thanks for responding to the thread.
I see. Well I’m not a mind-reader.
Sure you can. If my assumption is in error, you can correct it, can’t you? The reason I’ve assumed you were being disingenuous is because I’m sure you can see that there’s no contradiction in asserting that people’s preferences are subjective, which is itself an objective assertion. But maybe you’re not using the same definitions of “objective” and “subjective” that Austrian-school economics uses. If that’s indeed the case, then I hope you’re not doing so intentionally, as that would constitute proof of intellectual dishonesty on your part.
What’s wrong with that? Are you saying that you’re unwilling to adopt another person’s semantics for the sake of argument?
Sure you can. In fact, I think you just have (in a way). But to talk of “the meaning of cardinal utility” is fallacious. There is no such thing as “the meaning” of “cardinal utility” or any other term. “Cardinal utility” may mean something different to you from what it means to me. Rather than simply insist that I’m trying to change “the meaning” of it (which implies that the meaning you’re using is right and the one I’m using is wrong, which is a complete non sequitur given the inherently arbitrary nature of meanings), I think it would be better if you explained what “cardinal utility” means to you versus what you think it means to me. Then maybe we can get somewhere. But maybe greater understanding isn’t your goal here.
I invite you to point out just where I’ve asked you any “when did you stop beating your wife” kind of question. If indeed I have, then that shouldn’t be difficult for you at all.
I’d much prefer it if you didn’t run away from this thread so soon, but it’s entirely up to you.