In an article
in the Independent today, I argue that it is about time the Bank of
England changed UK interest rates. But they should go down, not up.
The essence of the argument is there remains a significant risk that
we have substantial deficient demand. Even if the probability of this
is below 50%, if it is true the costs of it persisting far outweigh
the costs of some mild inflation overshooting.
One point I do not consider in the article are the implications for
nominal GDP (NGDP) targeting. Here is the picture.
I use nominal GDP per head, because that is robust to changes in
migration flows, which for the UK have been important and variable.
The serious arguments are for a levels target, so I’ve drawn in a
reference path for 4.25% growth. That is a combination of 2% output
price inflation and 2.25% real growth per head, the latter being the
1955-2008 average rate.
If the Bank of England had adopted a NGDP target, as many have
recommended,
the MPC would be tearing their collective hair out right now trying
to stimulate the economy. There would be zero talk of interest rate
increases. So there seem to be just two possibilities. Either
NGDP targeting is nuts, or monetary policy has slowly gone off the
rails by focusing on CPI inflation alone.
Time will tell. But
if the possibility that the UK could really grow quite fast right now
without inflation getting out of control turns out to be true (and
the argument I make in the Independent is just that there is a
non-trivial possibility that it might be true), what will history
say? I suspect they will talk about Goodhart’s
law,
which says “when
a measure becomes a target, it ceases to be a good measure”.
Targeting inflation and ignoring output seemed like a good idea,
because of what is called the divine coincidence. I talked about this
in what I think is one of my better posts.
Goodhart’s law applied to this case says CPI inflation ceases to be a good indicator of
both the state of the economy and maybe also the costs of inflation
when you try using it as a target.
To what extent is current low inflation down to the collapse in commodity prices? Put another way, once that collapse has washed thru the system, what would inflation rise to, assuming all else equal?
ReplyDeleteWhat makes NGDP per head growth a preferable measure of growth rather than actual GDP per head as I see you have included a factor for price inflation in NGDP?
ReplyDeleteWhat else would be excluded?
"Goodhart’s law applied to this case says CPI inflation ceases to be a good indicator of both the state of the economy and maybe also the costs of inflation when you try using it as a target."
ReplyDeleteWhy wouldn't this also apply to an NGDP target?
Good question. The answer is that the underlying targets are the cost of inflation and the output gap, both of which are difficult to measure. NGDP is better than CPI inflation from a Goodhart Law perspective because it is closer to these two ultimate objectives. Critically, it does not depend on the divine coincidence.
DeleteHave you done a post on the divine coincidence? I remember Nick Rowe doing a series on it a while back, but I'd be interested in hearing your views.
DeleteClaudio Borrio and others at BIS doubtless would applaud raising rates; how would you respond to their concerns?
ReplyDeleteThey have a better instrument to deal with their concerns, and that is financial or macropru regulation.
DeleteThis is probably naive, but I'd go quite a bit further than your Independent article does: it would be a very good thing to overshoot the inflation target because unless we do it can never be more than a measure. Without overshoot, it's difficult to see interest rates ever rising to the point where they can be a useful macroeconomic policy tool.
ReplyDeleteThis view has other manifestations of course. Since 2010 we seem to have been hell-bent on destroying policy tools: reduce the size of the state and you abdicate macroeconomic responsibility. Maintain large and diverse public economic participants, and policy makers can actually do stuff in troubled times when other agents are constrained.
Macroeconomics needs some tools, and one of the first jobs of macroeconomics is to make sure the tools are available. ?whatever the cost?
Jonathan Peters
“Bank rate had never been below 2% in its then-315 year history. . . .UK rates are not expected to reach 2% until 2019. . . .Interest rates appear to be lower than at any time in the past 5000 years (Chart 5). “
ReplyDelete(Speech given by Andrew G Haldane, Chief Economist, Bank of England Open University, Milton Keynes 30 June 2015 ) Are you SURE keeping rates low is a good idea?
The fact that rates have been at 0.5% - an emergency rate - for years is a sign that the economy is broken.
DeleteAs a (not very good ) analogy.. if the main brakes on my car are broken and so I'm using the handbrake as an emergency measure instead, then ceasing to use the handbrake will not fix the main brakes. It just means I'll crash.
The real problem is that we need full employment, productive investment, and over-target inflation including real wage inflation before we can normalise interest rates. But government policy is somewhere between agnostic and hostile to these objectives. So we stagger on waiting for the nest crisis.
As an American I support this idea for the Fed as well. My priority is for tight labor markets where labor shares in productivity gains which in turn helps reduce inequality. Increasing inequality is bad for democracy and society.
ReplyDeleteA central bank *could* use an inflation target better than the Fed or BoE have been. The Fed has been running below target for awhile now and appears to treat it more like a ceiling than a target. They get away with because the corporate media and Congress don't give them a hard time about missing their target. They are more focused on possible inflation as are the hawks on the FOMC.
The fact that there is a target only happened recently as a change in policy which was supposed to provide "forward guidance" and help boost the economy.
But the markets no longer seem to believe in the Fed's forward guidance since it has repeatedly had to mark down its overly optimistic forecasts.
As is mentioned in the post, if the Fed set an NGDP target, and there was a NGDP futures market, they would provide blinking red lights that the Fed was failing. It would be harder for the Fed to explain away its failure and for the corporate media and politicians to ignore it.
"So there seem to be just two possibilities. Either NGDP targeting is nuts, or monetary policy has slowly gone off the rails by focusing on CPI inflation alone."
ReplyDeleteNo , there's another possibility : Using monetary policy alone for stabilization - whichever target you might choose - is nuts.
It's arguable that in the U.S. , the Fed was de facto targeting NGDP for a good part , if not the majority , of Greenspan's term. Nominal output was plenty steady , Great Moderation steady , in fact ( in the UK too , no ? ) . Also steady was the growth in the pile of economy-wide debt ( which was deposited , in a flaming bag , on Bernanke's doorstep ) .
The nice straight red line in your graph is a projection that necessarily presumes that debt doesn't matter. Is there anyone - anyone at all - who still believes that at this late date ?
http://www.sciencedirect.com/science/article/pii/S0164070412000298
http://irelandp.com/papers/evolution.pdf
Marko
Simon, you have many times made the case for fiscal stimulus near the ZLB and the need for increased public investment. Yet your Indpendent article, correct in arguing against a rate rise, mentions neither of these, giving the impression that action by the BoE to lower rates (within the narrow margin allowed by the ZLB) will suffice to raise growth towards what might be expected after a recession.
ReplyDeleteWas this your intention?
I have not changed my views about fiscal policy. But if the MPC continue with their current view, they might just use any fiscal expansion as an excuse to raise rates sooner. Indeed, if they were being consistent that is what they would do.
DeleteGood stuff. Great to read.
ReplyDeleteHowever, for NGDP Targeting to really work you have to use expectations, central banks should "target the forecast". You must work with the market or it could be too little, too late.
https://thefaintofheart.wordpress.com/2015/09/07/one-small-step-for-supply-siders-a-giant-leap-for-wren-lewis/
Simon, you´re back on the "right path":
ReplyDeletehttps://thefaintofheart.wordpress.com/2015/02/14/why-insist-on-searching-for-the-holy-grail-aka-nairu/
I never left it. The difference between my position and that of some others is that I think NGDP targeting would improve monetary policy but would not make it infallible, and I do not have this silly (ideological?) aversion to fiscal stabilisation.
DeleteOf course this website is an ideological free-zone as well demonstrated during the recent UK General Election campaign. ;-)
DeleteIt's just an ad hominem insult. Like saying "silly". We are all "ideological" otherwise we wouldn't have any ideas. Everyone else is silly, too, if they don't agree with me.
This comment has been removed by the author.
ReplyDeleteI favor an experiment with NGDP targeting, but I think the main advantage over DPI targeting is that it targets levels. A price level target might do almost as well.
ReplyDeleteI also thing that it is important to point out that NGDP targeting (or PL targeting) does not imply abandoning expansionary fiscal policy. The actions that the monetary authority will take when NGDP (or PL) is below target will lower the government's cost of borrowing. If governments follow the income maximizing rule of investing in activities with positive net present values, higher deficit finance expenditure will result -- what most people think of as expansionary fiscal policy.
So I looked at the graph displaying nominal GDP per head over time. Is there a reason as to why real GDP isn't added? Wouldn't it be easier to show the impact of inflation on actual GDP by displaying the gap between the two types?
ReplyDeleteAlso, how would nominal GDP be resistant to changes in terms of migration? Wouldn't there be several factors like quality of life and immigration that cause migration to change GDP, especially when considering inflation?