Mark Carney’s arrival as Governor, and his well publicised support for the policy of ‘forward guidance’, has had markets guessing whether he will be able to sway other Monetary Policy Committee (MPC) members to pursue this in the UK.
‘Forward guidance’ is the term that has come to be used to describe the practice of central banks explaining what will happen to interest rates in the future, in particular when interest rates have hit their natural floor of zero. (The MPC previously decided that the practical floor for rates in the UK would be 0.5%, but I will leave this issue for some future post.) Unable to cut today’s interest rates because of this floor, forward guidance seeks to lower tomorrow’s interest rate, by inducing the expectation that rates will stay at that floor for longer than might previously have been guessed given how the central bank responded to inflation and real activity in the past. The idea is that by lowering forecasts of future interest rates this will reduce the rates paid on fixed interest longer-term loans, and will give those thinking of taking out a loan on variable rates more confidence that rates won’t rise shortly, encouraging more spending. Judging from their actions, the majority of the MPC are not in favour of this policy: if they were, they would obviously have implemented it already. None have deviated from the practice of their former Chair, Mervyn King, who would insist that they face ‘one ball at a time’, using a cricket analogy (which for the uninitiated, translates as not talking, even, for some, not thinking, about future rates).
I am not in favour either, with one qualification that I’ll come back to at the end. I think there are several problems with the policy.
First, it seems to me dubious that any more monetary stimulus is needed. We have had above target and broadly stable inflation for six years or so. A crude way of reading this is that there is no gap between actual output/employment and the level that monetary policy (or fiscal policy for that matter) should be seeking to do anything about. There is clearly a big gap between activity and the level that would have prevailed had the economy continued to grow on its pre-crisis trend. But that gap is plausibly a consequence of the supply side of the economy having been strangled by the dysfunctional financial sector. If this view is right – I don’t pretend there could be any certainty that it is – then the consequence of more monetary stimulus would be a short-term surge in activity and inflation, but one that would, if it were to be sustained, have to be followed by another stimulus, and another further increase in inflation (and so on). Ultimately, following this course will not lead to activity being any higher, and will probably damage the real economy as the costs of high and uncertain inflation bite. Whether more stimulus is needed is a difficult topic, which I haven’t really done justice to. However, even leaving aside this problem, there are others.
Two objections centre around the view by proponents of forward guidance of how expectations are formed in the private sector. The assumption is that expectations are rational; agents understand how the economy works, what motivates policymakers and how they view the world. For the moment let’s leave aside whether this is realistic and assume that expectations are rational. The operating assumption of forward guidance is that if the central bank announces that rates will be lower for longer (than one might have guessed from what the MPC did in the past) then this will be believed. The problem is that the policy is not, in the jargon, credible, or time consistent. That is, when the time comes for rates to be lower for longer, it won’t be in the interests of the central bank to follow through any more. At that point, the recovery will be secure, inflation will be rising, and, concerned to achieve their mandate of stabilising inflation and the real economy, the preferred policy (ignoring past promises) will be to raise rates as quickly as one would have expected given previous behaviour. (Because that previous behaviour was guided by the same concern to hit the same inflation target as today). To combat this, the Fed, for example, accompanied its latest brand of forward guidance with announced thresholds for unemployment and inflation that would have to be met before rates rose. This might make the policy credible, or it might not. If the economy picks up sufficiently quickly, there is always the chance that the MPC will be sufficiently worried about stoking up another boom (so soon after the last one) that the thresholds will be altered or put to one side. Especially since those thresholds will have been set in the context of a lot of uncertainty about their key ingredients – the natural rate of unemployment, and the risks of de-anchoring inflation expectations.
Forward guidance of the Fed variety, holding interest rates lower for longer, guided by thresholds, is the practical analogue of Michael Woodford’s analysis of the consequences of following through on a commitment to set policy optimally, made in normal times, when the shock that interest rates are responding to is large enough to take rates to the zero bound. However, there is an important difference between the original academic proposal and the practical. Lower for longer as was practiced by the Fed and the Bank of Canada was introduced when convenient (when they had run out of room to do what they normally do and simply cut today’s interest rate). One interpretation of past policy is that it ignored Woodford’s prescriptions about how to set interest rates, right until the moment when there was no other option. Given that record, many might guess that just as a leopard can’t change its spots, the central bank won’t be able to resist doing what’s convenient for it in the future either, namely, raising rates promptly when inflation and real activity start to pick up.
The circumstances surrounding Carney’s appointment can be read in ways that are not that favourable to this problem either. One reading is that the Government, suffering in the polls, wanted more monetary stimulus to boost growth, saw forward guidance as a way to get it, and appointing Carney, whose support for forward guidance was well-known, as a way to implement it. This may be stretching things a great deal, but that is beside the point. The possibility that this was behind him getting the job adds to the circumstantial evidence that monetary policy is being guided by short-term convenience, and will continue to be in the future when the time comes to decide whether or not to respect the promises made at the time of the announcement of forward guidance. And one can add to this the accusation that the Bank of Canada itself reneged on its own forward-guidance, firmly rebutted, of course, by Carney himself.
It might be argued that there is no harm in trying forward-guidance (provided you think more monetary stimulus is needed). But the cost of it not working could well be the erosion of credibility more generally, surrounding the operation of other instruments, perhaps even those controlled by the Financial Policy Committee (FPC).
So far we have gone with the assumption that expectations are rational, and questioned, if they are, whether promises made by the MPC will be believed. An entirely different objection is simply that expectations are not rational. Clearly there are some market participants, and others whose job it is to forecast the behaviour of the central bank, who do take a forward-looking and close look at what the MPC does, and will do. However, even they don’t understand how the economy works precisely, just as the MPC surely do not. We don’t know how forward guidance would work under these circumstances. And there are vast numbers of other agents whose lives may be better served by forming a more simplified view of things, perhaps just projecting forwards what they see in their own industry or town into the future. For these agents, promises about future rates would be irrelevant.
Another concern focuses on a different aspect of the models used to inform forward guidance: the way prices and wages are assumed to be sticky. This is a long and controversial subject itself, and is going to seem esoteric and technical to some. In brief: the analysis behind forward guidance is based on a model that assumes that there is a fixed probability that firms will get to change prices each period. (Some assume that in addition prices are indexed in between times when they are reset optimally). Including this in a model helps it fit the data, gives the central bank a job, (without sticky prices there would be no role for active stabilisation policy and defines the job it should do (the relative weight that the policy-maker should place on inflation and real activity is related to this degree of assumed price stickiness). However, this assumption is made for mathematical convenience. There are two reasons to be extremely sceptical that this model could provide useful quantitative guidelines about the effects of forward guidance. One is that the economy may be far from its natural resting point. And so the assumption of this fixed probability of price-resetting may be a long way off, even if it does a good job when the economy is bobbing long close to its resting point. The second, is that we are contemplating a marked shift in the way policy is conducted . Robert Lucas’ 1976 ‘critique’ was that features of economic models designed simply to fit the data, would likely break down if policy were to change significantly. This is surely one of those features.
Forward-guidance as conceived by Woodford was something to be embarked on as a large shock hit, and in part, as a preventative measure, to reduce the time spent at the zero bound. The situation we face now is rather different. We are more than four years into a zero bound episode, and with future interest rates already very low (though rising somewhat in the last few weeks) and looking for a cure. With less room to lower future rates, it might take a commitment to a very long period of flat rates to make any appreciable difference to real activity and inflation (again, supposing that this is what is wanted, which I question). This raises two problems. First, current MPC members can’t bind their successors in the way that such a commitment would imply. This aggravates the credibility problem sketched earlier. Second, the MPC may be in the dark when trying to figure out how much guidance they should do. The sticky-price models that are used to inform monetary policy behave very strangely when rates are held fixed for long periods, [something we know from Carlstrom, Fuerst and Paustian (2012), for example]. This strangeness adds to the suspicion that the model is not a good tool to study policy in such unusual times, or to study such large shifts in policy behaviour (the Lucas Critique again).
All this said, forward-guidance, if implemented, or even if simply debated, may leave a good legacy. If it were to lead to monetary policy being discussed as a plan for interest rates, (i.e. not just facing one ball at a time), this would be a good thing. The economy is a sluggish dynamic system, and controlling it involves conceiving of a plan for the instrument-settings. Once the MPC published such a plan in pursuit of forward-guidance, it would be hard to stop doing it in more normal times. This would forever discipline future MPC members to think of their policy problem in terms of a plan (it’s clear from their writings that not all policymakers understood what they did in these terms), and force them to disclose it, which help others hold them to account, and, in my view, lead to a beneficial reduction in uncertainty about monetary policy.
longandvariable 