Tuesday, April 07, 2009
A defense of small government
Warren Meyer has written the best practical -- as opposed to normative -- defense of small government I have read in a long time. It all comes down to this: Do you believe that you, or anybody, can, in general and reproducibly, understand complex systems and understand the consequences of introduced changes? If you do, you probably also believe in the predictive power of climate models and government economists.
5 Comments:
By 
Robert Arvanitis, at
Tue Apr 07, 11:22:00 PM:
The Navier-Stokes equations describe fluids in motion. To this day, the model is incomplete; we cannot prove that smooth solutions to Navier-Stokes exist, much less describe turbulence. Yet any babbling brook “solves” the Navier-Stokes equations on its way to the sea.
In just that same way, no financial model can ever describe the capital markets. Only the actual play of market forces can find equilibria. Examples abound. Bear Stearns was in compliance with the Basle “capital models” right up to the moment it failed. Before that, LTCM knew that correlations held, until they didn’t.
That is a theoretical side to the vanity of efforts to regulate capital markets.
To that, add in the practical failings of human nature, the utter inability of politicians to resist abusing any power they are able to seize.
Thus from both theoretical and practical perspectives, it is impossible to regulate capital markets manually. It is only possible for government to ensure the premises of capitalism hold – free flow of information, many small independent agents, no subsidies, no monopolies and of course no “too-big-to-fail.”
Is it hard to back off the moral hazard of implied government backstop? Is it hard to wrest the prize from politicos' hands? Excruciatingly hard. But there is no alternative. There are no models, no regulations, no oversight that can solve the “Navier-Stokes” of finance.
By , at
Tue Apr 07, 11:40:00 PM:
TH: I will categorically fiat that, in the absence of a crushing (likely regulatory or taxational) force, certain markets will grow until their mere existence threatens the system, they "mismanage" it as you describe, or both. Given this, would you prefer that the largest actor have democratic obligations and transparency responsibilities, or be keeping things in off-balance-sheet vehicles? In a world where power didn't accumulate I would like the libertarian angle, but it is also true that in a world with fairy dust I would have a vacation home in Never Never Land.
By 
Robert Ghrist, at
Wed Apr 08, 05:23:00 AM:
If the argument is "Complex systems are computationally intractable; thus..." then I have to disagree. Contra Robert Arvanitis' comment,
smooth solutions to Navier-Stokes provably exist for short time intervals and often provably exist for all time --- what is unknown is whether *all* solutions have infinite-time existence (or whether, e.g., singularities develop). This present lacuna does not prevent using N-S to model the flow over an airfoil -- we design airplanes in silico that fly very well.
Likewise, air traffic control, mail delivery, and other problems of operations research would on the surface appear to be far too complex to ever solve. Indeed, one can prove computational intractability of many of these problems (cf. traveling salesman). But solve them we do.
I am quite skeptical of climate models, because I find them to be poor models, not because they are models of a complex system. It is tempting but wrong to say "complex=incomprehensible=stop".
By Robert Arvanitis, at
Wed Apr 08, 10:30:00 AM:
Fair point on the admittedly metaphorical use of Navier-Stokes. Glad someone read the comment, much less replied!
But note the following:
1. No system (short of a 1:1 "simulation") can capture enough detail. Bear Stearns was in compliance with all the Basle rules, up to the day it failed.
2. Markets are not entirely inefficient, but the equations ARE overdetermined. There are more constraints on the system than there are free variables. This guarantees there will always be inconsistencies, hence opportunities to game. (Example - the implied forward exchange rate of dollar-yen “must” balance the differential in dollar and yen interest rates. Except it doesn’t. There’s a delightful long-term chart of the gyrating exchange rate, showing the implied forward at many points. Never hit the realized rate except accidentally.)
3. This is poker, or prisoners’ dilemma, not chess. Unlike bankers and traders, the air molecules are not actively trying to game the system.
4. A person is smart, institutions are stupid. NASA put a man on the moon. Those same people overrode the evidence on the O-rings. Even with perfect equations, imperfect politicos cannot resist the temptations.
The net result is that the only good thing government can do is enforce the premises of free markets – free flow of information, many small independent agents, no subsidies, no monopolies and no “too big to fail.” The let Darwin do the rest.
By Counter Trey, at
Wed Apr 08, 01:22:00 PM:
TH
Coyote Blog gave a nice summary from an engineer's perspective of something that Austrian economists like von Mises and Hayek have been writing about for over 80 years.
A 3-31-09 letter in the WSJ written in response to Robert Reich's recent "Obamanomics" article highlights a key aspect of "Austrian School" thought:
"Mr. Reich mischaracterizes the agendas of Ronald Reagan and Barack Obama. Mr. Reich also demonstrates what Friedrich Hayek once called the "fatal conceit."
Mr. Reich argues that the primary philosophical difference between Messrs. Reagan and Obama is that Reagan tried to advance economic growth with "top-down" policies, while Mr. Obama is working through the more egalitarian "bottom-up" approach.
However, Reagan understood, as Hayek pointed out, that it is impossible for central planners to possess the necessary knowledge to efficiently allocate economic resources (high IQs and doctoral degrees notwithstanding).
Reagan knew that with less government the market would send unbiased price signals to every consumer. Individuals would take those signals and use their own specialized knowledge...to make economic decisions. Collectively, these decisions reflect extremely diverse knowledge and allocate resources for the whole economy in the best way possible. What could be more "bottom-up" than that?
In fact, it is Messrs. Obama and Reich who are attempting to manipulate from the top down by taking resources from every American and allocating it as they, the central planners, see fit...What could be more conceited than that?"
