Nominal GDP targeting would be subject to errors,revisions and likely inflation

One of the features of economic life is that there are a lot more reinventions of old ideas than there are new ones! One of these has been exhumed and shocked back to life by The Economist magazine and it is nominal GDP (Gross Domestic Product) targeting. Perhaps they hoped that everyone has forgotten! But the gist of the argument is that there were flaws in the system of targeting consumer inflation in the run-up to the credit crunch and that nominal GDP growth targeting would be a better system. Firstly let me agree that the advent of the credit crunch plainly did show up the flaws inherent in the consumer inflation targeting system of central banks. However as someone who replied to the Bean Review of UK Economic Statistics on Friday and is off to discuss the matter this evening at the Royal Statistical Society it is clear that so many issues are raised here.

What does the Economist argue?

It opens with this.

That is because the usual relationship between inflation and unemployment appears to have broken down.

Although he was looking at average earnings and unemployment Ben Broadbent of the Bank of England was not convinced of this.

Uncannily, the regression line was pretty much exactly the same as that identified by A.W.Phillips

But if we continue with the theme we are told this.

Now, by contrast, unemployment has fallen to remarkably low levels, but inflation remains anaemic.

Unlike The Economist I see this as a good thing! Fewer people are out of work and to coin an old phrase “the pound in their pocket” is going further or at least being much less affected by the ravages of inflation. To continue with the theme it is implied that this is a bad thing.

Their recoveries have instead proved nearly inflation-free

This has then proved to be a problem for central banks in setting policy. You may note the implied view is that a few central bankers are more important than the general population who are having their best economic phase for a while. So there is a bit of a straw man issue here but let us investigate the proposed solution.

Oh and perhaps no-one at The Economist has tried to buy a house in the past couple of years.

Nominal GDP targeting

Let us first check on what is meant by nominal GDP or NGDP.

the sum of all money earned in an economy each year, before accounting for inflation

This would do the following according to The Economist.

A target for NGDP growth (ie, growth in cash income) copes better with cheap imports, which boost growth, but depress prices, pulling today’s central banks in two directions at once

Does it? As it is reduced itself by lower inflation I think that they should be very careful with that argument. Let me explain with some numbers. The NGDP target has usually been for 5% with the breakdown hopefully for its supporters being 3% real GDP growth and 2% inflation. So in the UK right now we have not far of 3% real GDP growth but very little inflation, so they would stimulate the economy to hit the 5% target? That does not seem to help much! Rather than tighten central bankers would be easing into a boom, what could go wrong?

Some may think the real reason for NGDP targeting is shown blow which is why I have emphasised this sentence.

Nominal income is also more important to debtors’ economic health than either inflation or growth, because debts are fixed in cash terms.

With apologies to Meghan Trainor this sounds rather like this.

I’m all about that debt.

So many so-called reforms boil down to this which is in effect as confession that there is too much debt and we need to inflate it away except of course they are trying to do so by creating a fog where this is hidden. A bit like those who want a 4% inflation target. So in a world where in one sense there has been plenty of printed narrow we discover as we look at the wider situation that the situation is Simply Red.

Money’s too tight to mention
Oh mo-ney mo-ney mo-ney mon-ey
Mo-ney’s too tight to mention.

Debts are fixed in cash terms

Someone had better tell the UK Debt Management Office as it has some £506.6 billion of UK index-linked Gilts in issue by market value. These are indexed to Retail Price Inflation in the UK. Whilst the UK has a relatively large inflation-linked bond market it is far from alone.

The other problems with NGDP targeting

The Economist addresses the issue in one sentence.

Critics fret that NGDP is hard to measure, subject to revision, and mind-bogglingly unfamiliar to the public.

I have no idea why they are so hung up on the last bit as it could be explained. If we look at the numbers and assume a 5% target we have another problem. According to FRED (San Francisco Fed) UK NGDP growth has gone 4.7%,4.6%,4.3% and 3.4% over the past four quarters so we should be stimulating the economy?

Oh and there is another problem you see back in the second quarter of 2010 NGDP growth was running at an annual rate of 6.25%. Well actually not a problem for me as I was arguing that the UK was heading for inflation problems and needed to take action but I was rather alone, oh and correct. But for NGDP targeters there was 5%,6.25% and then 5% so happy days! What happened next?

Can we measure NGDP reliably?

Nope. GDP is often revised heavily and if we look back into UK economic history we see that the 1967 devaluation of the UK Pound £ was not necessary and that Denis Healey probably did not have to call in the IMF in 1976. More recently there was the “triple-dip” for UK GDP that in fact was not! You may be spotting something of a flaw here.

As a flow concept it measures construction of things but does not measure destruction meaning that wars for example boost it. There is a wide margin of error which invariably gets forgotten. I also think that it is poor at adjusting for technological changes -one of my points made to the Bean Review was about the measurement of the UK service-sector- and right now is a period of extraordinary technological change in many areas.

I guess some of you are thinking right now about the contribution of Volkswagen and perhaps quite a few other car manufacturers to GDP!

Another way of considering this is to look at the troubled series for US GDP in the 1st quarter of 2015. The nominal version went +0.1%.-0.9% then -0.2%. The annualisation makes this worse but there is a problem here.

Oh and there are three versions of GDP which are output expenditure and income. Most countries use output but Japan for example uses expenditure. They should come to the same answer but they do not except in the UK! Are we better at it? No, back when he was Chancellor Nigel Lawson insisted on it. Meanwhile in the United States the income and output measures give rather different answers. Or as the Bureau of Economic Affairs puts it.

For a given quarter,the estimates of GDP and GDI may differ for a variety of reasons, including the incorporation of largely independent source data.


Perhaps this was in someones diary to try again in three years as pretty much that time ago I pointed out this.

Put another way it feels at times like the episode of Star Trek where Captain Jean-Luc Picard is trapped in a repeating loop.

By escaping from the loop Captain Picard did better than the economics profession which is back to the past. Three years ago I pointed out this.

As even one of its supporters Martin Wolf admits inflation is an issue. They have no way of controlling the inflation – real growth mix  and frankly no certainty that there will be any real growth at all. Still at least they are not preparing the ground by attempting to emasculate our measures of inflation. Oh wait a minute!

This neatly fits with one of the points I intend to make tonight at the Royal Statistical Society. If you put in asset price inflation such as house prices you get a very different picture where there is disinflation in some areas but inflation in others. One thing we can be sure of is that establishment wheezes to try to hide inflation will continue.

In many ways it would suit the new economic advisory panel for the Labour party although no-one seems to have suggested it yet

15 thoughts on “Nominal GDP targeting would be subject to errors,revisions and likely inflation

  1. Great blog, Shaun , as usual. You state the case against nominal GDP targeting very well. Nominal GDP targeting seems to be like the monster in the horror movie, that is killed off again and again, but keeps springing back to life. It was studied both by the Bank of Canada and the Bank of England prior to our central banks adopting inflation-targeting regimes. Both quite properly rejected the concept then, so it really does make one despair that it seems to be making an undeserved comeback more than a quarter of a century later.
    As I say in my paper “Twenty Years of Inflation Targeting at the Bank of England” a paper by an anonymous Treasury Department official written in December 1989 outlined a number of policy alternatives for the Bank of England. One of them was inflation targeting and another a “money output framework”, which obviously related to nominal GDP targeting. The discussion of the money output framework was very cursory, and it isn’t even clear if the intention was to target the growth rate of nominal GDP or the level itself. However presumably it was the growth rate as there is no mention of the issue of not letting bygones be bygones that zealots in favour of targeting the nominal GDP level drone on about.
    The section on inflation targeting considers both the GDP deflator and the Retail Price Index as potential target inflation indicators. The discussion of the GDP deflator is somewhat naïve. There is no mention of its reliance on proxies for owner-occupied housing and financial services and so forth. However, the paper does mention that it is only available quarterly, is published with a considerable lag and is subject to multiple revisions. Since nominal GDP is a function of real GDP and the GDP deflator the same criticisms must have had some force in discouraging adoption of a nominal GDP targeting regime.

    • Hi Andrew and thank you

      The reminder that NGDP targeting has a long history and is very difficult to kill reminds me of the Alien series of films. As I mentioned in the piece I went to the meeting at the RSS this evening on the Bean review and the use of GDP statistics came up. It was Chatham House Rules so I will keep it anonymous and merely say that there was agreement that the way that minor GDP variations are recorded is pretty much madness. Imagine it if it was the central bank target too!

      Actually some argue that the GDP numbers being quarterly is an advantage, although I am not a fan of that as it is better to have a more numerous series to peruse. On that front how much detail is there in the monthly series from Canada Statistics? Do you only get quarterly deflators etc?

      Oh and as to nominal GDP growth rate the paper by Woodford suggested 5% per annum as a target which is why I referred to it. But it also looked at a catch-up of what had been lost compared to trend which means that the growth or level debate was not complete.

      • Shaun, thank you very much for your briefing on the RSS meeting.
        Regarding the monthly GDP by industry estimates for Canada, they are available at a quite detailed level. There are 273 series published at all levels of aggregation. Both chain Fisher and so-called series at constant 2007 prices are published. However, the latter series are grossly misnamed, as they are more like the chain Laspeyres estimates produced by the ONS.
        I have mixed feelings about the monthly GDP estimates by industry. The people who don’t trust the quality of the monthly estimates and believe that quarterly estimates are more reliable certainly have a point. On the other hand, at a time like this when it seems that we are recovering from a recession that ended in May, it is certainly a blessing to have monthly GDP estimates, which helps keep the public from treating the economic Cassandras here more seriously than they deserve to be. If it weren’t for monthly GDP estimates we wouldn’t know that there was growth at a 6.2% annualized rate in June, and would still be waiting for 2015Q3 estimates to come out on December 1 to know if we were back to growth. Prior to the monthly GDP estimates being available we did have a monthly index of industrial production, but it leaves out a lot of the economy. In June it showed growth at about an 18% rate, very much stronger than the growth in real GDP.

      • Yes.
        If inflation numbers are gamed.
        If “errors” tend, coincidentally to fit one side of the argument.
        If further items like childcare, which used to be done by parents but is more and more done by professionals, which make us no richer but add hugely to NGDP can be added.
        If wheezes like charging for plastic bags in supermarkets can be added…
        You know, all the usual politicoes’ machinations
        Need I go on??

  2. Perhaps its better to scrap RPI and CPI and replaced them With COLI , Cost Of Living Index.

    And make its mandate to keep inflation low ( band of 0 to 2% ) and it will be under indipendant body not subject to any government whim

    naw , it;ll never work , too much to loose and could you really say it’s free of political vice ?

    Oh, HMG could not elect these people , only the people themselves (!)


    Ps: free from gaming ? not a chance

    • Hi Forbin

      You are not to far away from the Household Inflation Index that has come out of discussions at the Royal Statistical Society.

      “a consumer price index intended as a measure of “inflation as perceived and experienced by households in their role as consumers”.”

      Oh and at the meeting of the Bean Review on UK Economic Statistics this evening there were lots of pronouncements and claims of “independence”…..

  3. Shaun,another cracker.
    Talking of crack,this does seem like a heroin addict turning to crack in the hope that it’ll change the outcome.
    The underlying disease they’re trying to deal with-boom n bust,inflationary/deflationary-isn’t a disease but the way of the world.maybe they’d like to try and moderate its excesses but the way to do that is ensure you have mechanisms that enable honest measurements.
    This blog has some serious form when it comes to exposing the sometimes laughable stats that they base govt policy on.
    I presume they’d still be including imputed rents in NGDP calcs…..I rest my case.

  4. As a non economist if someone tells me growth should be 3% and inflation 2% it all sounds rather like a set of convenient numbers. Not 3.4% and 2.1%? Where is the derivation of this? Also be interested if big macro numbers GDP, Unemployment and Inflation have not gone the way of all policy numbers – focussed till they say what we want e.g. US unemployment looks if not an outright guess to exclude anyone who can’t get a job for any length of time.

    Thus from my ignorant perspective not sure these measures should be solely targeted if at all. The rosy tint on GDP comes off when looked at per capita since 2008 say. Inflation is good but does not tally with other measures or include Rent. GDP does include drug dealers and Escorts so as my high street boards up I can feel glad this is because more money is being spent on gear and gals and not a negative sign because yes GDP rising at 3%.

    Anyway targeting less/single measures strikes me as a dumbing down of already questionable headline numbers. As well as dumb for all the reasons listed above. Targeting numbers does not work in low level 1 dimensional jobs. In many ways it’s a cop out by policy makers.

    • Hi Bedfont and welcome to Notayesmanseconomics

      In truth they wish to try to confuse people about the growth/inflation mix they are getting. Guess which they plan to give us more of and which they plan to give us less of?

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.