In a speech today at the London School of Economics, MPC Member Professor David Miles asks “what is the most useful way for a central bank to provide information about the way in which it will set monetary policy?”
He begins by observing that some might see the Yiddish proverb “Mensch tract, und Gott lacht” – men plan and God laughs - as a fitting epitaph for forward guidance. “The Bank of England has certainly faced a good deal of criticism for the guidance it has recently been giving” but much of this “boils down to the criticism that... it is not saying what policy will be”. David argues that this criticism is misplaced and setting a pre-determined path for interest rates is a guaranteed recipe for instability.
David uses a simple model of the economy which encapsulates several uncertainties which are “particularly relevant for the UK right now” (uncertainty about the degree of spare capacity, the likely impact of interest rate rises, growth in the absence of monetary stimulus and trends in productivity). He shows the effects of setting a pre-determined path for future interest rates which, while on average correct, can turn out to be quite wrong, depending on how things play out over time. He shows that there is a 10% probability that after 16 months central bank rate would be at least 2.5 percentages points too low. There is an equal chance that interest rates would be at least 2 percentage points too high. Either scenario would be “very costly” and could lead to some “wildly unstable” outcomes.
That still leaves the big question of what sort of guidance – short of a commitment to a particular path for interest rates – is best.
In an idealised world, central banks would not need to give guidance – it would just need to say what its aims are. If everyone in the economy is “completely rational, has unlimited computational powers, has the same information and understand the processes driving the economic outcomes” then all a central bank needs to do is “credibly commit to pursue an explicit and completely specified objective”. People could then work out precisely what the chances are for interest rates to take on different values in the future.
So in an ideal world (and in idealised economic models) the central bank can just explain its objectives; maybe it would also need to describe its reaction function (how it might change interest rates in particular circumstances). No more is needed.
But clearly this is not the world we live in. There are great differences in the information people have about the economy and in their time and inclination to interpret data; there is also uncertainty as to what the best economic model. Unless outsiders have as good information on the economic outlook as the central bank – and can see exactly how to use that information with the central bank reaction function to work out what could happen to interest rates - then they may be left floundering in attempting to assess the prospects for interest rates.
And in any case a reasonable approximation to a central bank reaction function might be very complex. (Indeed it is very likely that it is; otherwise Monetary Policy Committee meetings at the Bank of England would be mechanical affairs lasting 5 minutes).
David draws the following conclusion “In light of this I believe that the central bank directly giving some assessment of the likelihood of interest rates following different paths is likely to be helpful, particularly in situations in which past action and communication is not enough to describe new trade-offs in an unusual environment. “
But constructing an explicit probability distribution for the central banks own policy rate that is based on the realities of how policy is made and gets across a message that can be widely understood is a major challenge.
Attempting to provide an explicit central path for interest rates with bands around it showing specific probabilities of interest rates moving to different levels may go beyond what is feasible and useful. Spurious accuracy is unhelpful.
David concludes: “Currently, it might be just as useful – and probably less misleading and possibly even more accurate – to give forms of guidance which are more qualitative, such as:
interest rate rises will probably be gradual and likely to be to a level below the old normal
That says something substantive; and most people can understand it. “