Yesterday I posted following the BoE Inflation Report press conference, agitated by the repeated emphasis on avoiding an overshoot of the inflation target. A private response to the post made it clear that I should write again and disentangle two separate motivations that exist for overshooting.
The first motive, Michael Woodford’s reasoning for urging that central bankers attempt to engineer a deliberate inflation target overshoot at the zero bound, is as follows. Unable to loosen the current policy instrument, the current undershoot of the inflation target and the output gap can be limited by creating the expectation of lower spot central bank rates and of higher future inflation. The inflation overshoot later on is more than compensated for by reducing the size of the undershoot now.
My private correspondent pointed out that some policy makers don’t like this advice because they fear that creating an overshoot would not be time-consistent. Later on, once the benefits of the lower interest rate expectations in terms of a higher output gap and closer to target inflation have been pocketed, it’s best to re-adjust and not follow through with lower rates. Knowing this ahead of time, observers won’t believe the announced overshooting strategy.
It’s true that this strategy may not be believed. There’s nothing to tie the central bankers’ hands. And a history of them saying they take each month’s policy meeting as it comes [reinforced somewhat by Mark Carney’s remarks to that effect in the Press Conference].
On the other hand, it may be believed, and there’s no harm in trying. Even if it isn’t, following through anyway will help bolster the overshoot strategy as a tool for escaping the zero bound the next time around. With low equilibrium real rates for the foreseeable future, we might expect that the next time will come around pretty soon.
All this said, there is a separate motive for overshooting which does not rely on expectations management, and applies even if policymakers resign themselves to behaving in a purely discretionary fashion, (a nihilist position many central bankers might find odd if it’s put to them in such stark terms).
Overshooting is justified on risk management grounds alone. Consider [as did Charles Evans and coauthors] uncertainty about future shocks hitting the economy that would perturb inflation and output away from the central bank’s central forecast.
A shock that depresses the economy and requires further loosening can’t be easily responded to at the zero bound. One that boosts inflation and output above the expected path can be dealt with easily with higher rates. Although in expectation we don’t expect any shock [that’s what our forecast is], it’s better to behave as though we think there will be a small depressing shock, as the costs of our forecast turning out too optimistic outweigh the costs of us being too pessimistic.
So, bringing this all together, we can rationalise the MPC’s desire not to overshoot as deriving from one of three sources. First, it may be because they think they have not got a hope of convincing anyone of a commitment strategy they might announce. Second, it may be that they are confident that the size of negative shocks hitting the UK will be very small, and thus won’t require much loosening to combat. Third, it may be because QE is considered a perfect substitute for interest rate policy, and there are no constraints on its use. None of these seem wholly convincing to me.