Andrew Sentance’s FT comment column this weekend needs a reply. So much that it’s worth running the risk that this blog starts to be seen as the Andrew Sentance rebuttal unit.
One thing Andrew says is:
“By acting as instruments of government policy, central banks are straying from their own dominion into political territory.”
I want to point out here that Everything central banks do is an exercise of a power delegated to them by government. They are always and everywhere an instrument of government policy. They don’t have ‘their own dominion’, and it’s unhealthy to suggest that they do. When that idea takes hold, it sews the seed of someone saying ‘why should central banks do inflation policy? let’s do it ourselves and have loads of it!’ If HMT instructs the Bank of England to do macro-prudential policy in statute, the BoE has no business doing anything but doing exactly what it’s told. Once safely in its list of delegated powers, an instrument is not ‘political territory’.
Andrew would be right to warn central banks from taking the initiative to undertake missions or pull policy levers that were not explicitly delegated to it. This is the German critique of the ECB. (And relies partly, in my view, on fallacious attempts to distinguish ‘monetary’ from ‘fiscal’ things, as I wrote previously). But it does not apply to the BoE, the Fed, or the BoC. All of which simply doing what they were mandated to do.
Sentance appears to include unconventional monetary policy in the list of political actions that ‘have risks’. But this is a puzzling argument for him to make, given his own MPC voting record, which persistently supported quantitative easing.
But the idea in the column that most needs contesting is the idea that we should take advantage of currently very low measured inflation and lower the target to zero. Andrew has been consistently calling for higher rates, even as inflation headed South of the target, and one wonders whether the reason for that was not that tighter policy was necessary to get 2 per cent inflation, but that 2 per cent was too much.
Andrew writes:
“The 2 per cent inflation target served policy makers well as a definition of price stability when they felt that zero inflation was not achievable. But price stability could mean what it says on the tin.”
The target for measured inflation was set at 2 not because 0 was not achievable [inflation is a monetary policy phenomenon]. But for two reasons. Reason one is that it’s well known that, like all CPI inflation measures, ours contains a large upward bias. So Andrew’s desire to lower the target to current rates would almost surely mean generating true deflation. If you like, in Andrew’s words, the inflation measure itself does not quite do ‘what it says on the tin.’ [For more, and some nice links, see a previous post].
Reason 2 is that the inflation target needs to be such that the average level of nominal rates consistent with it is high enough to leave room for responding to busts. It wasn’t that 0 was not ‘achievable’, it’s that it was not desirable. If there is no room to cut rates sufficiently, then there is a risk of not just inadequate demand, but losing control of inflation itself. Take a look at Japan. And wonder too whether persistently below target inflation in the US, US and EZ is not the harbinger of us becoming ‘like Japan’.
In this regard, lowering the target is exactly the opposite of the lesson to be learned from the crisis. The target was set at 2 when it was thought i) that average nominal rates would be about 6% and ii) that we had had our last financial crisis and large recession.
Looking ahead, surely everyone has lowered their guess at average nominal rates consistent with 2% inflation to about 3-4. And surely we have woken up to the fact that the Great Moderation was an illusion. So, we now realise that large recessions are a real possibility, and so too the necessity for large interest rate cuts to stabilise inflation; and we realise that we will have less room for these cuts than in the future.
So today’s target already implies much more time spent at the zero bound than planned, and relying on unconventional policies (which Andrew seems to dislike). The lesson then is that the target needs to be raised, not lowered. [See also previous posts].
Andrew could try to defend his suggestion by arguing that inflation is SO costly that lowering it by 1-2pp would bring benefits that would exceed the cost of giving up the ability to stabilise your economy. But, to repeat: it’s inflation that one would risk losing control of too. So lowering inflation on this count risks generating more variable inflation, which, by the same argument, must be costly. And, as an aside, there’s no support for thinking inflation is so costly in the macro literature.
This sentence by Sentance is worth quoting.
“..the very institutions that won their credibility by keeping a lid on prices now seem to be trying to create inflation, not subdue it.”
Isn’t that because inflation has been persistently below the targets set by their government owners? ‘Credibility’ is the reputation you have for doing what you say you will, not a reputation for leaning down on inflation. Fighting to create as much inflation as you promised is just as important as the old problem of resisting the temptation to create too much of it, as Japan is learning, and we may too.
Note too how the insidious ‘lid’ metaphor invites us to think of inflation that is always ready to escape upwards, out of the pot. At least once monetary policy has been successfully delegated to central banks, this has no economic foundation. So far as we understand the inflation process, we can lose control over inflation on the downside too. And some arguably have already.
longandvariable 