Future shock

Several liberal blogs are chortling over this statement I made early in the year:

Will the economy decline in 2008?

Paul Krugman is voting for doom. It's worth keeping in mind, however, that Paul Krugman has predicted eight of the last none recessions under the Bush administration.

I think it's obvious we're in a slowdown, and a recession seems likely-ish, but Britain's skirted recession for over a decade now, so I can't be too fatalistic.

This is obviously hilarious--if you have an incredibly shaky understanding of statistics, and also, no knowledge of decision science.

If you keep predicting a recession, eventually you will be right.  Every time there was the slightest downturn in the numbers, Paul Krugman predicted a recession.  Eventually he was right.  Do we give him credit for the one he got right, or the multiple ones he got wrong?  To liberals, the answer seems obvious.  Which gives credence to the conservative belief that liberals are people who cannot do basic math. 

A more interesting question is what to do about doomsayers like Nouriel Roubini, who got the magnitude of the crisis right, but has similarly been predicting a financial holocaust for five years, with changing scenarios which mostly did not come to pass.  Obviously, he was right that the global financial system was shaky.  On the other hand, his understanding of why the global financial system was shaky does not seem to have been strong enough to predict the source of the failure--the current account deficit and the dollar have been at best minor players, at least in the way that he was worried about way back in 2004. 

There are a lot of financial pundits out there.  The law of large numbers means that one of them will turn out to have predicted almost any financial event.  How do you separate true genius from monkeys throwing darts?  That task is made much harder by the tendency of pundit partisans to look only at predictive successes, while discounting the failures.  If you eliminate all of Paul Krugman's predictions that did not come to pass from the sample set, he looks like a genius.  Unfortunately, the same is true of Larry Kudlow.

The people who, it seems to me, have been truly vindicated by this are Nassim Taleb and Benoit Mandelbrot.  They nailed what to me is the core issue:  bankers pricing security risk as if it were distributed along a normal curve with thin tails.