Libertarianism is dead . . . vive le libertarianism!

Must one pile on Jacob Weisberg?  Why yes, one must.

Not because he is wrong that this crisis raises serious problems with market structure.  It does.  It also raises serious problems with regulatory structures.  It raises serious problems with everything we were doing.

I am not trying to sound like those dogmatic libertarians who declare, any time problems are detected in the markets, that they merely exist because we do not have "true" capitalism.  Sometimes this is true--the interstate trucking market has been indisputably improved by deregulation.  And sometimes it isn't; the market for banking services has been (to my mind) indisputably improved by the FDIC.  Markets are complicated things that rest on a mixture of law, custom, and individual action.  There is no libertarian state of nature in which human beings survived without some form of coercively enforced rules.  All libertarianism can do is maximize the scope for individual action within that framework.

To put it another way, society has multiple possible equilibria, and some of those equilibria are better than others. Removing the existing set of rules does not, of itself, guarantee or even make it particularly likely that we will arrive at a better one.

But that, of course, also holds true of government action.  Punishing the bankers, restraining them from taking certain sort of risks, or giving regulators vastly more power over them, does not in itself guarantee that we will get a better outcome.  I see many people who do not know very much about finance demanding that we reverse the much-vaunted deregulation of the 1990s.  I see very few of them proposing a coherent regulatory framework that will get us to that happy state, much less a political process that will achieve this desireable regulation rather than something else which better suits the legislators.

Nor is there any particular proposal for preventing that institution from falling prey to the same forces that grip the regulated industry.  I have said it before, but it is worth repeating:  the regulators became overconfident in the same way, and for the same reasons, that the bankers became overconfident.  Just as a long and unusually rosy period in the housing market convinced the bankers that they had gotten better at pricing credit risk, a long period without a large bank failure persuaded the regulators that they had gotten better at regulation.  They believed that their computer models, and an improved understanding of how markets and the economy worked, would allow them to see problems in time and halt them.  Obviously, they were wrong.

Now, obviously, you can tell the institution "Don't get overconfident!  Be extremely risk averse!" But for the next ten years or so, this is superfluous; the bankers themselves have learned to be extremely risk averse, a lesson that they will, at least for a while, employ in building their portfolios.  And longer than that, you cannot commit any political institution to a particular stance.  For one thing, extreme risk aversion has its own costs, and the fact that we currently wish we had been more risk averse does not mean that ultra-cautious regulatory policy is actually optimal, any more than it would be a good idea to act, permanently, on your wish that you hadn't left the house and walked into the path of that car.

Moreover, it is meaningless, in a mixed economy like ours, to attribute a massive failure like this to either "the market" or "government regulation".  The people who think that this can all be traced back to the CRA are wrong.  But so are the people who think that the only problem was too little regulation.  Just as some actions by private firms unquestionably put the economy at risk, so did some regulations.  It is indisputable that mark-to-market accounting caused potentially solvent firms to suffer ratings downgrades.  It is also not disputable that a slew of rules surrounding the credit quality of the securities that pension funds and insurance companies can invest in helped trigger the horrific cascading liquidity crises that, as much as anything, forced the government to step in and bail out the banks.  If these institutions had not been forced to keep their assets concentrated in "investment grade" securities, they would not all have had to run for the door at once as soon as there was a ratings downgrade.  Those downgrades, of course, hit the balance sheets of other institutions, some of which found themselves next in line to be trampled by the stampede.  Those regulations were good at protecting against individual firm failure.  But when there was a systemic problem, they made things worse, not better.

A doctrinaire libertarian might conclude, then, that what we need is to eliminate regulation.  And indeed, some have said just that. But the regulations are a very good idea 99% of the time when the problem is individual, not systemic.  More importantly, it's hard to think of an adequate substitute for those regulations in the face of a public that needs savings vehicles, but cannot do advanced financial accounting.  Preventing a cascade by allowing people to regularly lose their insurance coverage or pensions is not an obviously good tradeoff.  I'm hard pressed to say how it's a non-obviously good tradeoff, either.

Similarly, a doctrinaire liberal may declare, as many have, that the problem is "free markets".  But you can't pick "free markets" out of this knot any more than you can pick out "regulation".  The system went wrong as a system--it's not that there's some broken part you can swap out.  You can imagine an alternate world in which we did not have the institutional credit quality requirements, or mark-to-market accounting, in which the banks limped through a wave of mortgage defaults without systemic failures.  You can also imagine a world in which the regulators had somehow been better at regulation, and the market actors had been prevented from taking catastrophic risks.  But that's all we're doing:  imagining.  We have no proof that either state is safer than the current one. 

We can tell post-hoc stories; gold bugs favor various almost-panics in the nineteenth century, while liberals like to claim that it was the wisdom of FDR's regulation.  But much of this is post-hoc, ergo propter hoc.  What we know from behavioral and experimental economics suggests another explanation at least as plausible:  that as long as people think bubbles are very likely, you don't get bubbles.  For example, for thirty-five years, the memory of the 1929 crash was seared into the brains of the public consciousness, and the market remained relatively down to earth.  Then the generational memories faded, and you got the late-sixties bubble that led into a decade-long stagnation.

The farther I go into this crisis, the more leery I am of any neat narrative explanation of financial panics--or indeed, many other rare phenomena.  More on this later.  For now, a parting thought:  in complex systems, there can be no such thing as an individual villain.