What is Bernanke thinking?

Many good economists have expressed dismay at Bernanke's behavior during this recession. If they could have picked any man to preside over a financial crisis and zero interest rate policy at the Fed, Ben Bernanke would have been their first pick. So were they wrong?

I don't think so. While the chairman of the Federal Reserve is undoubtedly the most powerful member of the Fed, he certainly doesn't make policy decisions by himself. The Federal Reserve failed miserably in the crisis by letting NGDP fall dramatically, but that doesn't mean Bernanke hasn't done everything he can have to ease policy.

Here's a simple model of how the Fed Chairman might think about how he really affects policy. (In my examples I'm assuming the chairman wants easier policy than the median Fed voter but this could easily work in the other direction as well).

M = C – D^x

"M" is the easiness of the Federal Reserve as a whole, "C" represents the individual easiness of the chairman (Through statement wording, policy instruments and public statements) and "D" represents the number of dissents from official Federal Reserve policy. The variable "x" represents how strongly dissents serve to push policy in the opposite of the direction desired by the fed chairman.

Here's an illustration of the model:

(As Bernanke individually becomes more dovish he quickly runs into diminishing returns with regard to changing the Fed's overall stance)

The implications of this model are:

• The chairman individually promoting easier policy will ease actual Fed policy until it leads to dissents.
• At some point, the chairman attempting to ease policy will actually make policy tighter because it decreases the likelihood the chairman will have the votes in the future to continue current policy, and undermines his perceived power which decreases the likelihood he will be reappointed (and his replacement will likely be closer to the median board voter).
• If the chairman is going to ease his position enough to get a dissent, he should ease as much as possible without causing an additional dissent.
• The chairman needs to guess the value of x, although he will likely know the effect of C on D

So Bernanke is essentially keeping policy and his public statements much more hawkish than he really wants in order to prevent dissents from the board. He could ease policy, but doing so would cause board members to dissent and those dissents would undermine any additional easiness in policy. If you believe, as Bernanke does, that the expected path of monetary policy is monetary policy then this makes a lot of sense.

One question might be why there aren't more dissents than there already are. There is only one (Hoenig) and in addition to him not being a voting member this year, he is leaving the board in October.

A possible answer is that other voting members have formed a voting coalition with each other. If each member simply dissents when policy becomes too extreme according to their own individual opinion, the chairman would be left with a great deal of power to maneuver overall policy in his desired direction. If, on the other hand, they form an agreement to threaten dissent at the same time, they would ensure policy could become no tighter than their mutually agreed point of dissent. Since transactions costs are low with so few members, such negotiations seem highly plausible.

But where are the doves?A single dissent could theoretically make perceived policy easier and Bernanke wouldn't have to push past such a dissent in the same way he does a hawkish dissenter.

This is all speculation of course, but does anyone really think Bernanke isn't being constrained by other Fed board members? 

So maybe a better question than "Why hasn't Bernanke done more?" is "Why haven't Fed doves done more... by dissenting?"