1. Skip to navigation
  2. Skip to content
  3. Skip to sidebar
Source link: http://archive.mises.org/7525/mechanism-design-and-the-free-market/

Mechanism Design and the Free Market

December 10, 2007 by

NobelLike most of the important results in mainstream economics, the prize-winning work in mechanism design is very elegant mathematically, and offers many counterintuitive results — yet it has precious little bearing on the case for (or against) laissez-faire capitalism. Those who think otherwise are relying on a simplistic and naïve view of how the market and government actually operate. Before arguing this point, we should first review some basics.

For those who equate “the free market” with “atomistic individuals who reduce everything to money,” it is obvious why the insights of mechanism design appear to impugn pure capitalism, and to justify enlightened government tinkering with spontaneous outcomes. Yet this view relies on a false caricature of the market economy, and a naïve faith in political action. FULL ARTICLE

{ 16 comments }

Fundamentalist December 10, 2007 at 9:56 am

Thanks for the analysis. I wasn’t even aware of mechanism design. But it sounds like more central planning. No matter how much central planners are proven wrong, they don’t ever give up. I guess you could admire their tenacity. What frustrates me is that, just as Murphy points out, their mechanism only works if they make unrealistic assumptions, create a straw man out of capitalism and a super-hero out of socialism/central planning.

I’m also fed up with this Pareto efficient nonsense. As Hayek constantly reminded people, no one is smart enough to ensure Pareto efficiency, not even a computer with mechanism design equations. Pareto efficiency can only be achieved if one assumes that they know for a fact the subjective values of every person in the economy, and that is impossible.

George Gaskell December 10, 2007 at 10:38 am

Mechanism design … sounds like more central planning.

That was my first reaction as well.

As Hayek constantly reminded people, no one is smart enough to ensure Pareto efficiency, not even a computer with mechanism design equations.

So true. If the global economy is a “mechanism,” it’s so mind-bogglingly complex that no amount of design could possibly optimize it.

To optimize such a system through design would be like guessing the correct lottery numbers every time, every week, for 600 million billion years.

The economy is even more complex than the weather, and that’s saying a lot!

Niccolo December 10, 2007 at 12:23 pm

I can not really understand how one can say that the price offered in an auction is not the price determined as optimal.

It seems to neglect marginal prices in the determination of how one can truly comprehend the worth of an object in terms to their own preferences.

I write down a bid for $50, it apparently indicates that I value the chance of winning and keeping $150 over the painting and the chance of losing it.

It seems to indicate a general tendency to assume, as opposed to calculating the rational designs within the market, only indicated by actions and not simple guesses as to what one would have done under other circumstances.

I generally do not appreciate Pareto-efficiency as much as most do.

Ron December 10, 2007 at 1:05 pm

Central planning would work just fine if individuals would simply behave and think in the manner government wants them to.

Jim Berger December 10, 2007 at 1:11 pm

As human systems, economies do one thing that mechanistic models cannot do: they learn. We cannot create a model (or a modeler) smarter than the market.

Person December 10, 2007 at 1:18 pm

Jim_Berger: Actually, machines can learn. However, you are correct that a model or modeler can’t be smarter than the market, if only because the information specifying the market is spread across those other participants.

Fundamentalist December 10, 2007 at 1:57 pm

George: “The economy is even more complex than the weather, and that’s saying a lot!”

Excellent point! I often compare the economy to the weather. Look at what a mess climatologists have made with their faulty models instigating GW hysteria. The results for similar models of the economy will be similarly disastrous.

Person: “Actually, machines can learn.”

As someone who has worked with data mining and machine learning programs for several years, I think the idea that they can learn is mostly hype. Machine learning algorithms a not very different from traditional statistics. You still have to have dependent and independent variables. Even sophisticated neural networks programs, which can re-create complex patterns, often do a worse job modeling data than plain old statistical regression. Machine learning programs don’t learn any more than regression learns. If you have a dependent variable and some independent predictors, machine learning programs can tell you how the predictor variables are related to the response (dependent variable) and can handle some nonlinearity, but that’s about it. Without a dependent variable, all you can do with machine learning programs is cluster analysis; not very exciting.

Also, machine learning programs fall victim to all of the problems of regular statistics, such as misspecification, multicollinearity, but especially false correlations. In other words, just like regular statistics, machine learning models are no better than the theory behind them.

In addition, machine learning programs haven’t been developed to handle multiple equations yet, as far as I know, and that is necessary for most economic modeling.

Central planning got a huge boost with econometrics, which gave planners the false confidence that they could perfectly models economies. In many cases, machine learning programs are no improvement at all over econometrics.

Daniel M. Ryan December 10, 2007 at 4:49 pm

I’m not intending to criticize Dr. Murphy with this remark, but the reasoning that demonstrates that the second-price auction is “weakly dominant” only applies to one-shot games. An iterated game opens up the possibility of scratch-each-others’-backs maneuvers, which would be preferred by both to bidding the price corresponding to the true subjective value.

It should also be remembered that the auction with two bidders is not a two-person game but a three-person game. One of the players is the seller.

Daniel M. Ryan December 10, 2007 at 5:29 pm

An observation, on what those two economists call “socialism:” by my reading, it’s a completely free market except for ownership of capital goods, which all must be owned by the government. No egalitarianism in rewards, and no hampering of consumer demand apart from the government-ownership constraint. This model doesn’t square with what socialists describe as socialism.

No word as to whether the “auctions” are equivalent to sheriff’s auctions, come to think of it…

Abu Hatem December 10, 2007 at 8:09 pm

This website perhaps has the best daily articles of any other blog on the internet.

Thank you for your extremely thorough and scholarly analysis. It helps non-economists like me when you put everything in such facilitative language.

Another argument which could be put forth is that although “mechanism design,” would perhaps create better quality goods and services – there is no true way to implement mechanism design on a federal level without running into problems. And a free-market is already providing us with immense innovation, and so as the ol’ adage says “if it aint broke, don’t fix it.”

Abu Hatem.

TGGP December 11, 2007 at 1:23 am

Karl Smith says on Mechanism Design “Incentive Compatibility Constraints (ICC) are the theoretical nail in the coffin for socialism. When I talk with serious socialists, and yes they do exists, ICC is always the backbone of my arguement”.

The subject matter of the most recent Nobel Prize seems out there and of little relevance to daily life, but in a sense it shows just how inapt central planning is. Austrians should not reject all that is going on in mainstream economics for their is much thought and useful work to be found in it.

Anthony December 11, 2007 at 7:32 am

Many thanks for the link TGGP.

Daniel M. Ryan December 11, 2007 at 7:34 am

I actually sympathize with your take on mainstream economics, TGGP. Remember the subsidiary postulate “leisure is a desirable consumer good?” A lot of deterministically-oriented mainstream economics can be let in by adding this second subsidiary postulate:

“Acting according to habit is a desirable consumer good.”

This fact of human behavior explains why determinism sometimes makes sense, and why deterministic schema seem to work for a time before falling apart.

And while I’m at it: a valid extension of game theory might very well be the “Poverty-Trap Game,” a Prisoner’s-Dilemma-type game which encourages poor people to stay in their place through an opportunity cost for picking oneself up. I’m surprised that the game theorists haven’t made the connection before.

Nonebody December 11, 2007 at 8:19 am

Thank you for a much more reasoned discussion of mechanism design than the rather poor articles that have proceeded on Mises.org . Unfortunately, many of the writers here assume that if a) there is mathematics they don’t understand, or if b) someone incorrectly uses it to suggest the state is good, the fundamental idea must be unsound.

A similar discussion of public goods and building incentives for public goods (including that bugaboo– Intellectual Property) from an *intelligent* Austrian perspective would be much appreciated. I’ve noticed a number of authors on Mises have gone so far as to deny there are public goods (or claim everything is, therefore it doesn’t matter) or that finding mechanisms to support them is a good idea. It’s just silliness: you can not support state activity and still recognize your non-Austrian colleagues have formalized a *major* problem. Help resolve the problem and you’ll make real impact.

Anthony December 11, 2007 at 8:39 am

Do you mean Hoppe’s and Block’s articles on public goods? They merely point out all the logical fallacies and contradictions in public goods arguments as they commonly are given. Denying such stupidity is meritorious, IMO, and is part of the School’s negative task, i.e. to refute bad economics. But I agree, Austrians should be active in discussing these issues and proposing solutions, i.e. its positive task. We have the correct economic theory, we just need to apply it. And this is being done increasingly.

Fundamentalist December 11, 2007 at 10:13 am

Game theory is very interesting and offers some great mental exercises, but it’s nearly impossible to apply to situations with more than three people involved. So I think a lot of the interesting theory that comes from game theory will remain speculative for the near future.

Comments on this entry are closed.

Previous post:

Next post: