John Taylor originally suggested his rule as both a good guide to what central banks actually do and also one that “captures the spirit of the recent research”. It has been used ever since as a yardstick by which to measure monetary policy. However there are well understood reasons why it is likely to be a poor yardstick in a severe recession.
First some theory. In a world of certainty, when inflation expectations are equal to the inflation target, the optimal interest rate to set is one that delivers what is called the ‘natural’ real interest rate. You can describe this as the real interest rate that would achieve a level of demand and output which eliminated the output gap, and put unemployment at its natural rate. At that point, there should be no pressure from the domestic economy for inflation to rise or fall.
In this context it becomes obvious why the output gap (or deviation of unemployment from the natural rate) appears in the Taylor rule. Yet in reality our estimates of the output gap are poor, so it also makes sense to include the difference between inflation and its target in the rule. Finally the rule also contains a constant, which is an estimate of what the natural interest rate would be if inflation was at target and there was a zero output gap. As to the coefficients on inflation and output, you want those to be modest to avoid overreacting to false signals and to allow for lags between interest rate changes and their impact on inflation.
The best way to think about the Taylor rule is as a simple ‘horse for all courses’. It is designed to be a robust rule for all situations: booms as well as busts, small as well as large deviations from target, and where we have no additional reliable information.
In the current recession we know a number of additional things. First, the natural real rate of interest is likely to be a lot lower than the constant in any Taylor rule. There are a number of reasons for this. In the short term a balance sheet recession means that consumers want to save much more than they would normally, so the natural rate has to be unusually low to offset the impact of this on demand. In the longer term we have the issue of secular stagnation, which is one reason why policymakers in both the UK and US say that even when the economy recovers interest rates are likely to be lower than they have been in the past. (Austerity is another.)
There are other factors as well. At low levels of inflation, inflation appears to be less responsive to excess demand. On its own this means that the coefficients on excess inflation in a horse for all courses Taylor rule will be too low when inflation is below 2%. (The possibility of hitting the Zero Lower Bound can also imply the same thing.) If forecasts indicate that inflation will remain below target for some time, that can also suggest we can afford to react to inflation being too low by more than the Taylor rule would suggest.
If you want a practical illustration of all this, consider this post from Zsolt Darvas at Breugel. It uses a typical Taylor rule for the Euro area, and finds that interest rates set by the ECB have been below the level implied by that rule every year since about 2001! That is a clear illustration of the problem of assuming a constant long run natural real interest rate, in this case beginning with Bernanke’s savings glut. Exactly the same points arise in trying to assess whether US monetary policy was too expansionary in the mid-00s. This same rule also implies that the ECB raised rates by too little in 2010/11, which is clearly silly in the light of what subsequently happened. (Again we had more information, in this case about austerity.)
However the Breugel post is not really about how appropriate the ECB’s monetary policy is for the Eurozone as a whole. Instead it focuses on what the rule tells us monetary policy might have been in each individual Eurozone economy, if they had retained their own currency and had floated. Or to put it another way, it tells you for which countries the ECB’s policy is too tight, and for which it is too easy. Used in this way, the analysis is a handy way of combining information on inflation and unemployment diversity across the Eurozone.
Where is the ECB’s policy too tight? There are the obvious countries: Spain, Portugal, Italy and especially Greece. But there is another, which is the Netherlands. There is no mystery here: CPI inflation is currently (May) 0.8%, the harmonised rate is 0.1%, and unemployment has been over 7% this year, compared to an average of below 4% from 2000 to 2007. As the Netherlands does not have an independent monetary policy, it desperately needs a countercyclical fiscal policy, yet instead it is locked into the austerity trap imposed by the Eurozone’s fiscal rules. All of this was horribly predictable, which is why I wrote these posts: May12, Sept12, June13, Dec13.
These fiscal rules are not going to be abolished anytime soon, even though their intellectual rationale has disappeared. The best that we can hope for is that their impact can be softened or partially circumvented by allowing additional public investment spending: see Reza Moghadam from the IMF here, or Wolfgang Münchau here and here, Guntram Wolff here, or Mariana Mazzucato here. But if in the future anyone wants to see the clearest example of where these rules led to large and completely unnecessary social costs, just look at the Netherlands.