At last. Mechanism designers got their Nobel prize. But did they deserve it?
For me economics became a science sometime around 1945* with John von Neumann’s and Oskar Morgenstern’s Theory of Games and Economic Behavior. Naturally, this book was only an outline of things to come and far from finishing the edifice we call today game theory. It took the subsequent efforts of people like John Nash, John Harsanyi, Reinhard Selten, Robert Aumann and Thomas Schelling to complete this beautiful structure of clear rationality expressed through elegant mathematical formulation.
The first Nobel prizes however for the masters of this field came only in 1994 (Harsanyi, Nash, Selten). Maybe it is the traditional reluctance of the Swedes to acknowledge the importance of new fields, or other reasons, it does not matter. But it seems that since 1994 they are trying to correct their mistake by awarding prizes to all the other aforementioned researchers and also people who I consider to be related to GT like William Vickrey, Vernon Smith etc.
It seems microeconomics and especially game theory are accepted now as one of the strongest tools in the social sciences with applications from biology to sociology. Mechanism design is one of these applications and it was chosen to be honoured this year, with the Nobel prize going to Leonid Hurwicz, Eric S. Maskin and Roger B. Myerson “for having laid the foundations of mechanism design theory”.
Mechanism what?
So what is mechanism design and was this honour warranted?
Mechanism design is basically one thing: the quest for truth telling. It is giving people incentives to reveal their private information, information that is relevant to markets. A standard example is an auction: people want to acquire an object and the owner wants to know the value every potential buyer puts on the object to maximize his profits. The mechanism designer is the guy who (what else?) designs a mechanism that does exactly this: gives people the incentives to reveal their preferences through the offers they make for the object.
In general, the job of the economist as mechanism designer is to make people put their money where their mouth is, to make them tell the truth by putting them in a special setting (mechanism) that offers them appropriate incentives.
Is this so important? Well among economists mechanisms and auctions have been a big thing and many people have built their entire careers on the idea that mechanism design is important.
People over at MarginalRevolution seem to think otherwise.
Some of you might argue: “These guys have already had a big impact on real world auctions and incentive schemes.” In terms of the induced improvement in human welfare, I find that a difficult case to make. The important progress has come from recognizing much simpler truths about incentives.
My personal opinion is a bit schizophrenic. Since my doctoral thesis is on what one would call applied mechanism design, I am bound to think of MD as something interesting and useful. But is it really helping mankind? Well at first glance no. Most problems in mechanism design are close to zero sum games, what the seller wins the buyer has to lose and vice versa. This leads to the thought that economists are once more helping fat, ugly capitalists skim their victims. Well that is not entirely true, allocative efficiency, a goal of mechanism design and long time economic dream, really matters, right? If a licence (for mobile telephony, radio or other) is not awarded efficiently this means a potentially bad producer gets it and incurs higher costs than a competitor. This waste is bad for the whole economy. But using another economic idea, cost-benefit analysis, perhaps the (mostly modest) gain in efficiency from better mechanisms is not really worth the huge effort going into the design of these mechanisms, all these bright minds spending decades of research and some millions in grants. Low tech ideas, like say the Grameen Bank, would be much more useful to the world’s citizens, if directly helping the world is what academia is supposed to be doing.
Here comes the twist: directly helping the world is not the goal of academics, it’s the goal of applied science and politics, in our case it’s not the job of economics professors but the job of IMF and World Bank directors, UN employees etc. The lack of direct impact on the real world is a constant frustration for many PhDs in economics, but remains a hallmark characteristic of the ivory tower and rightly so I might add. To paraphrase an old joke, if we had academic economists trying to save the world and humanitarian workers doing economic research the result would be funny but rather unpleasant and catastrophic on the whole. Academics think, their students translate thoughts to reality.
Practical Mechanism Design
So what does MD have to offer to practical people? Well, all kinds of things, when its practitioners combine a good grasp of the theory with a practical mind.
For example in rich countries, we have recently had a nice example of this, Al Roth’s kidney exchange programme, that arguably is saving actual human lives. Another nice application is better school selection/choice systems. Economists have taken abstract mechanism design, mixed it with some common sense, knowledge of real life peculiarities and some psychology (behavioural economics in the jargon) and delivered nice results. Thus, these are occasions where a smart application of the theory greatly enhances welfare. And there is much more to be done.
Quite promising, for me, are the cases where mechanisms are tested in laboratory experiments or in the field and adjusted for the complex interplay of financial and non pecuniary incentives, social norms etc (Uri Gneezy’s kindergarten example was a funny way to show how these change things). Interestingly, even pure theorists can use the theory imaginatively. Paul Milgrom (pioneer of auction theory, whom some would have added to the Nobel prize worthies) has applied the theory in many cases with much more imagination and feeling for relevant details in real markets, than what his abstract papers would suggest.
In poor countries a smart application of mechanism design can be even more interesting. Take incompetence, idiocracy (lack of meritocracy) and corruption, possibly the most important plagues in poor countries, maybe not directly claiming lives like AIDS, but probably responsible for the suffering of billions of people. We don’t live in a perfect world and fixing these problems cannot happen by just some more polite advice or even a bit of policing. Mechanism design can offer sustainable solutions to these problems, by proposing mechanisms that can be efficient to some degree and self enforcing, with no need for costly and intrusive policing.
I would say then in conclusion, abstract mechanism design might “be going out of style” but its practical applications are only now starting to bear fruit…
————————————————–
PS As for you statistics freaks, it seems this years Nobel Laureates had enough citations. Maskin had 12 papers with more than 200 citations (in Google Scholar) , Myerson had about 10…
*OK, people like Cournot in the 18th century did great microeconomic work too. But it seems their work was only understood with the advent of formal game theory. And surely, people like Adam Smith or J.M.Keynes were no intellectual dwarfs either… I guess you could reread my phrase as “in 1945 economics acquired its best tool (from mathematicians)”.
The post is in English since it might be of general interest (not only Greek that is). Please comment in English (this goes for you Germans out there too! You know who you are, stop hiding behind that tree)…
Το κειμενο ειναι στα αγγλικα γιατι ισως ενδιαφερει και μη ελληνοφωνους, παρακαλω σχολιαστε στα αγγλικα.
It is quite interesting that the applications of mechanism design theory that you describe mixes in “common sense, knowledge of real life peculiarities and some psychology (behavioural economics in the jargon)” that eventually produces good results. Do you claim that using standard assumptions of full rationality of the theoretical models would not “cut it” in real world applications?
Understandably, the Nobel prize goes to foundational work but shouldn’t it also reward economists that take the highly abstract (or even unrealistic) and adjust it so that societies reap the benefits? So the question is, should the award go only to the theorist rather than theorists “appliers”? Did the applied economists that chose to do something useful with the abstract theory face significant challenges in turning the theory more relevant to real world applications?
You suggest that it is not generally the role of economists to “directly help the world”; but their true added value as scientists is the production of knowledge that will lead to better economic/societal outcomes. Theory just for the sake of mathematical beauty, although intellectually stimulating, lacks in added value. MD theory has and will increasingly continue to have an impact; but has the Nobel committee rewarded the “right” mix of researchers?
tricky questions maikwl :-)
well I think there are actually two problems with standard models. First there are assumptions that are useful for tractability of a theoretical paper, but are extremely restrictive. Theoretical mechanism designers dislike simulations, but I think for real world applications we should do away with the restrictive assumptions, lose the power to algebraically come to some solution (which is usually not in a closed form anyway) and use numerical methods instead.
Second, as you guessed, full rationality can only be assumed under very specific circumstances and even then it is not guaranteed it would lead to the predicted equilibrium. For real world applications we need many tests of the robustness of a given equilibrium, both theoretically (for example by examining the exact shape of the payoff functions and the cost of deviations) and empirically (using experiments in the lab and in the field).
People who actually designed real mechanisms, like the FCC auction, tried as far as I know to use all our accumulated knowledge, both theoretical and empirical.
Did they face challenges? Surely, I think it is so hard that many politicians are not even convinced yet of the necessity to hire a mechanism designer in such cases.
Do they deserve a prize? That’s a very hard question. Inventing penicillin gets you a Nobel prize, so inventing an economic mechanism that actually saves lives should get you one too, right? Well I think the problem is, it’s too hard to discern the effects of a mechanism and attribute the success or failure to a single person. On the other hand it is very easy to check if a theory is mathematically correct and useful, just check the proofs and count the number of citations!
I guess what we actually need is two prizes, one for theory and one for application. I think people like Milton Friedman would deserve more the second and game theorists the first.
hm i dont know, but I really hope that people who are applying MD now will eventually get their own prizes. And well, if they don’t, knowing they are directly making the world a better place should give some satisfaction too…
Hi S G.
As a completely ignorant layman, as regards Economics itself (and trying to learn it) I have a juicy layman’s question for you.
Your description (that follows) rings a bell in my own mind:
“Mechanism design is basically one thing: the quest for truth telling. It is giving people incentives to reveal their private information, information that is relevant to markets. A standard example is an auction: people want to acquire an object and the owner wants to know the value every potential buyer puts on the object to maximize his profits. The mechanism designer is the guy who (what else?) designs a mechanism that does exactly this: gives people the incentives to reveal their preferences through the offers they make for the object.
In general, the job of the economist as mechanism designer is to make people put their money where their mouth is, to make them tell the truth by putting them in a special setting (mechanism) that offers them appropriate incentives.”
I interpret this (perhaps too simplistically) as one of the first attempts to look deeply into people’s NEEDS and preferences, as the motive force and latent structure beneath the market, revealing a substratum of knowledge that can be as important to liberal thinking as to liberal socialism (of certain exquisit innovative varieties).
Bringing needs and preferences into the market simulation game is impressive in itself, needless to say.
well I would say economics is based on people’s needs and preferences, this is the role of the utility function that people maximise.
Mechanism design analyses mechanisms (obviously) that are manipulated in a way such that people’s preferences lead them to do something that we want them to do, reveal the truth etc. All this is of course quite simplistically explained…
Μήπως χρειαζόμαστε καινούργιο νήμα για τη συζήτηση του 2009; Πολύ ωραίο θέμα το φετινό.
Να γραψει ο μαικωλ που αποτι θυμαμαι την συμπαθει ιδιαιτερως την Οστρομ. Παντως δεν μου βγαινει απο το μυαλο οτι το εδωσαν σε αυτους για να ειναι οσο το δυνατον πιο μακρια γινεται απο χρηματοοικονομικα και μακροοικονομολογους που συσχετιζονται εστω και αμυδρα με την κριση.
Λεει μετα το Τσερνομπυλ η πυρηνικη φυσικη δεν πηρε Νομπελ για χρονια, ελπιζω να μην γινει το ιδιο και στα οικονομικα.
Χαχα, ναι, γουστάρω ιδιαίτερα το συγκεκριμένο νόμπελ. Νομίζω είναι τρία τα μυνήματα που προσπαθούν να περάσουν:
1) the importance of institutions and good governance (it’s not all about markets)
2) we need to address climate change and economists have some tools we can use
3) you don’t need fancy math to do good economics
Συμφωνώ με τον maikwl. Ιδιαίτερα στο 3. Αν και νομίζω θα είναι κρίμα να καταντήσει και το Νόμπελ Οικονομικών ένα ακόμα «πολιτικό» βραβείο που να ακολουθεί αυστηρά τα σημεία των καιρών μας. Ήδη έχουμε το Ειρήνης που έχει χαλάσει…
Πάντως το φετινό θέμα ήταν όντως πολύ ωραίο. Σημείωση: έβαλα σύνδεσμο στην εκλαϊκευμένη εξήγηση. Υπάρχει και η πιο επιστημονική στο site των Nobel.
Νομίζω ότι πρέπει να αναθέσουμε σε πολλούς από αυτούς που σχολιάζουν ή αρθρογραφούν εκεί στους roosters εντατικά μαθήματα στη δουλιά και των δύο νέων Νομπελιστών!
μαικωλ γραψε κειμενο γιατι μαλλον διαφωνω :-)
Κάτι είχα στο μυαλό μου αλλά δεν προλαβαίνω τίποτα πριν την Τετάρτη. Το ξέρω πως διαφωνείς :)
Πάρτε και μια δόση από κάποια βαρεμένα άτομα του mainstream να χτυπιούνται στο internet για την επιλογή:
http://www.econjobrumors.com/topic.php?id=5151
Υποψιάζομαι πως πολλοί από αυτούς είναι και καθηγητές :)
μπα τριτοετεις διδακτορικοι ειναι που βλεπουν τις επιλογες τους να τους οδηγουν στην ανεργια :-)
κενσάι , μετά το περσινό νόμπελ οικονομικών , υπάρχει ακόμα αμφιβολία οτι πρόκειται για κατ’εξοχήν πολιτικό βραβείο ;
Κώστα , αν σου είναι εύκολο , μπορείς να με παραπέμψεις σε δυο τρία άρθρα του ρουστερ , των οποίων οι συγγραφείς πιστεύεις οτι πρέπει να μελετήσουν την Οστρομ ;
Α , για να προλάβω “οργισμένες” αντιδράσεις , με το προηγούμενο σχόλιο δεν θέλω να πω ( απαραίτητα ) οτι Ο Κρούγκμαν πήρε άδικα το βραβείο ,ή οτι οι απόψεις του είναι λάθος , αλλά οτι η επιλογή του , τη δεδομένη χρονική στιγμή που έγινε, ήταν πολιτικά “βαμμένη”
Alan Dallon
ως αρθρογράφους θα αναφέρω με μεγάλη μου απογοήτευση τα τελευταία κείμενα του Ανδρέα Ανδριανόπουλου και μερικά από τα άτομα που σχολιάζουν όπως οι Kapitalist, Ex-European, κ.α.
Φαίνεται να προβάλλουν την άποψη ότι σε οικονομικό επίπεδο απαιτείται η πλήρης απορρύθμηση των αγορών (που σημαίνει ανυπαρξία θεσμικού πλαισίου) και το assignment of property rights, σε ηθικό μια ακραία μορφή ατομικισμού (individualism), και ο καπιταλισμός θα φροντίσει να κάνει τα υπόλοιπα. Για μένα αυτό δείχνει άγνοια της θεωρίας των επιχειρήσεων (theory of the firm) αλλά και οικονομικής ανθρωπολογίας. Οι δυο Νομπελίστες έχουν συνεισφορά σε αυτούς τους τομείς.
Found on the shitheads (aka Economics Job Market Rumors) forum:
Dixit on the award
http://media.bloomberg.com/bb/avfile/Economics/On_Economy/vVAO1.8h8Kms.mp3
Paul Romer
http://chartercities.org/blog/72/skyhooks-versus-cranes-the-nobel-prize-for-elinor-ostrom
Comment from Tom Schelling
http://www.forbes.com/2009/10/12/economics-nobel-elinor-ostrom-political-science-opinions-contributors-thomas-c-schelling.html
Comment from Venon Smith
http://www.forbes.com/2009/10/12/elinor-ostrom-commons-nobel-economics-opinions-contributors-vernon-l-smith.html
Πιο αναλυτικά στο ποστ που ανέβασα.
Κώστα , διάβασα μόνο περιληπτικά τα συμπεράσματα του έργου της Όλστρομ .Αν κατάλαβα καλά , συμπεραίνει οτι η διαχείριση κοινών πόρων μπορεί να γίνει αποδοτικότατα από κοινότητες σχηματισμένες αυθόρμητα και εθελοντικά και υπό αυστηρές προυποθέσεις ( τις αναφέρει περιληπτικά το αγλλικό κείμενο στο ποστ του μαικωλ ) . Απ τα λίγα που έχω διαβάσει κανένας (αναρχο) φιλελεύθερος / αυστριακός δεν διαφωνεί (τουλάχιστον επί της αρχής ) , εξ’ου και η θερμή υποδοχή της βράβευσης της Ολστρομ από το Reason , τον Steffan Karlsson και άλλους . Που ακριβώς φαίνεται στα άρθρα του Ρούστερ διαφωνία των τριών που αναφέρεις ; Θα με βοηθούσες αν έδινες συγκεκριμένα παραδείγματα
Alan,
δεν θέλω να ξεκινήσω αντιδικία για τους Ρούστες διότι δεν αναφέρθηκα στο σύνολό τους, ούτε στους Αυστριακούς. Μίλησα για ορισμένους, και μετά το συγκεκριμενοποίησα. Αλλά αν ενδιαφέρεσαι, δες τη συζήτηση που έχω ανοίξει με τον Πέλεριν στο κείμενο του Αθανάσιου για την τεκνοποίηση.
Η Όλστρομ μιλάει για communal ownership. Ο Πέλεριν φαίνεται να πιστεύει ότι there is no such thing, μόνο κράτος, και ότι όποιος διαφωνεί είναι κομμουνιστής και δίνει υπερφυσική οντότητα στην έννοια της κοινότητας/κοινωνίας.