I read today that James Bullard is disaffected with the dot-plots, and considered pulling out of them unilaterally, reserving his dots for himself.
I think this would be a mistake, for several reasons.
At some level, a communication strategy has to be a collective one. It would be undermining of the monetary framework to fail to achieve a common purpose at the level of deciding how to communicate with the public.
That goes for those wanting more rather than less transparency too. In the UK, the Monetary Policy Committee have decided not to reveal what they think they will do with their instrument, nor any dot plots. A hypothetical MPC member in the UK who wanted, against the opinions of all the others, for the committee to produce interest rate forecasts, could produce their own dots without breaching contract. But would sabotage policy as a whole by doing that.
A common concern is that there is a loss of credibility as the dots move around. Maybe so. But if so, that reflects a misunderstanding of what it is to be a competent policymaker by those adjudicating on ‘credibility’. The outlook can change a lot from month to month, and so, correspondingly, should the best policy response in the future.
Naranya Kocherlakota writes about the dot plots, but comes down against replacing them with a collective interest rate forecast, my preferred solution. He is concerned – a view echoed across the central bank community – that such forecasts will be interpreted as commitments or promises.
There are lots of misguided views of monetary policy out there. That central bank forecasts are ‘bad’ if they are wrong ex post. that inflation is set by the monetary authorities. That the Fed needs ‘Auditing’. That counter-cyclical policy is corrosive and coddling. Policymakers fight against those misconceptions rather than giving into them. Likewise, they can fight against the incorrect view that the instrument forecast is a promise.
Of course, the fight is a subtle one to wage, because, the point of publishing the path, which itself is not a promise, is to communicate what the Fed is promising, namely, to do what it can to achieve its goals, and adhere to a particular reaction function – a data-contingent plan.
‘You promised us four hikes by 2017!’ commentators might cry. And the answer would be: well, ‘yes and no’. We looked at our model and our reaction function, and computed four hikes. But all we promised was to keep computing.
longandvariable 