Talk:Walrasian auction

Merge French to this auction
I feel that French auction should not be merged. Dutch auction and french auction are the terms popularly used in the real world (auction world that is). Moreover the french auction goes a lil way beyond the Walrasian model, eventhough the underlying principle is same. I think the article on french auction needs a little more description. Hopefully somebody does that. I will make an attempt but not sure about the time lines.

pv

That sounds fine by me. But I don't want to do it the other way around; let's keep titles as descriptive as possible. ImperfectlyInformed | {talk - contribs} 04:14, 5 May 2008 (UTC)


 * I don't think merging is appropriate. The Walrasian auction is a conceptual device for talking about markets. The French auction (at least as the article describes) is a real-world auction mechanism. For instance, in a real auction, I'm pretty sure that the buyers don't give the auctioneer their fully-specified demand curves. So I think they shouldn't be merged; if they are the merge should be done very carefully to identify what's theory vs. reality and what the relationship is between the two. Cretog8 (talk) 08:07, 31 May 2008 (UTC)

I agree that the merge should be done very carefully so as not to spoil the contents of either article. The main Auction article obviously favours "French auction" as being the most widely accepted title for the merged article. At least for English-speaking readers "French auction" is easier to remember than either "Tâtonnement" or "Walrasian auction". But readers looking for those titles should of course be rerouted to "French auction". Also please note that several types of auctions are named after the country were the auction was first used or after the nationality of the inventor. I also strongly feel that there is a lot more to say about French auction. The article is not yet complete. Theory and practical auction design are always closely connected. Max7437 (talk) 12:27, 31 May 2008 (UTC)

I put the merge tags on because in my experience in mechanism design as studied in computer science, the term "Walrasian auction" can be used to not necessarily require full demand curves from all agents. From your comments, it sounds like "French auction" is a term used more in practice, so the Walrasian auction would be a generalization. Mike Wellman is well known in auctions and mechanism design as it pertains to AI in computer science, and has a few well-cited papers on using Walrasian auctions (search for Walras):. Halcyonhazard (talk) 02:44, 10 June 2008 (UTC)


 * Based on the updated French auction, do you still think it's a kind of Walrasian auction? It is a uniform price auction, but each bidder submits only a single price/quantity bid, rather than a schedule.Cretog8 (talk) 05:42, 26 June 2008 (UTC)


 * At this point I still don't know - it sounds like the French auction might be a subset of either/both a Walrasian auction and/or tatonnement mechanisms, or perhaps my view of one or both of the terms is too broad. Let's keep the discussing going here: Talk:French_auction Halcyonhazard (talk) 04:26, 30 June 2008 (UTC)

I got here via general equilibrium theory. If you want to represent mainstream economics you should not refer to tatonnement as any kind of auction. It is true that Walras spoke of an auctioneer but that was just a metaphor and not to be taken literally. It is meant to represent the unspecified mechanism by which markets manage to adjust when in disequilibrium, where there is most definitely noone directing them, which is against the whole decentralized spirit of perfect competition. It is what Mankiw in his introductory book refers to as the "law of supply and demand" in that if demand exceeds supply prices rise, and vice versa (though he is only speaking of one market.) Mankiw does not envision an auctioneer either.

You should probably also mention Scarf's counterexample where consumers with well behaved preferences over three goods results in an excess demand giving a unique unstable equilibrium together a stable cycle, thus establishing for the first time that a well specified tatonnement process need not result in convergence to any equilibriium whatsoever. This counterexample resulted in loss of interest in the study of tatonnement processes by much of the economic profession. The Sonnenschein Mantel Debreu theorem vastly generalizes this by showing that most any tatonnement dynamic is possible (since most any excess demand is possible). The program ennunciated by Debreu and never carried to any kind of successful conclusion is to make reasonable assumptions on the distribution of preferences and endowments that would result in well behaved excess demand where there would be a unique, stable equilibrium and perhaps useful comparative static results.

One reason to study tatonnement processes over nontatonnement ones is if you want to do the usual sort of comparative statics, since in the latter case comparative statics of the usual sort is impossible, since the final equilibrium depends on the process, which modifies the equilibrium as it goes.

I don't see why you don't just write down the basic rule for all to see: dp/dt=z(p(t)), for excess demand z, which is pretty intutitive.


 * I used to work on a topic known as market oriented programming (the optimization kind of programming) in computational game theory (often lumped in AI under computer science). That field is concerned with mechanisms and detailed rules and processes that are followed in the auction.  In that discipline, tatonnement processes referred to specific properties of mechanisms employed in auctions.  I have very little background in macroeconomics, so if you want to add a macroecon viewpoint to the article, that'd be helpful.Halcyonhazard (talk) 07:11, 2 February 2011 (UTC)

—Preceding unsigned comment added by 83.202.43.87 (talk) 19:54, 1 February 2011 (UTC)

the "auctioneer" was NOT introduced by Walras
--87.205.70.49 (talk) 12:41, 17 August 2011 (UTC)e.g. "I cannot find mention of the auctioneer in Walras’s own writings; and Donald Walker, the leading living U.S. expert on Walras, assured me (in conversation) that the auctioneer indeed does not appear in them. That prodigious figure is the invention of later theorists trying to make the theory tighter." - Leland Yeager, Is the Market a Test of Truth and Beauty?, Auburn 2011

http://mises.org/books/truth_and_beauty_yeager.pdf

Yeager is indeed (largely) correct. See Walker and van Daal's (2014) discussion (a little of which I've just added to this page). Qbt1824 (talk) 10:29, 29 November 2017 (UTC)

Introduction
I reached this page from 'Real estate appraisal'. It seems that a Walrasian auction is a theoretical one, not a real one. Shouldn't this be made clear in the introductary lines of the article? Chrisemms (talk) 20:47, 20 September 2011 (UTC)

Mistranslation
If 'crieur' -> 'auctioneer' is a mistranslation, then what is the meaning of 'crieur'? Ulaniantho (talk) 18:39, 15 August 2023 (UTC)