The upside down economy: The Hare chases the Fox

Fox & Hare_0

According to Danny Blanchflower, the US economy is back to normal because “there is no inflation and unemployment is 4 percentage points below the start level”.

It appears Janet Yellen & Associates think the same. According to Yellen:

“We believe we have seen substantial improvement in labor market conditions and while things may be uneven across regions of the country, and different industrial sectors, we see an economy that is on the path of sustainable improvement”.

Normally, it is the fox (actual output) that chases the hare (potential output). This is illustrated in the chart below, which shows how the “fox” usually pursues the “hare”.

Fox & Hare_1

Over the last eight years, however, it appears the “fox” got tired, allowing the hare to “run away”. But that cannot be, so the “fable writers” (the CBO) has given the story a twist, deciding that in the “new world” it is the hare that will pursue the “tired-out” fox!

Fox & Hare_2

DB and Yellen & Associates are correct! The economy is (almost) back to normal, a state in which actual output is very close to “potential output”.

What to target? It’s only rational to use expectations of nominal output growth

A James Alexander post

Twitter is great. It gets to the nub of issues quickly. No messing.

Every time I propose NGDP Targeting the heavyweight sophisticates immediately come back like here and here.  with “you can’t target that because of the huge revisions”. It got me thinking about what these twitterati would target.

Inflation: anyone’s guess

The inflation target is perhaps the worst of all. There is little agreement about which inflation rate to choose. Rightly so. It’s way more complex than facile economic commentators think.

  • Should the rate be the one faced by the average consumer? The average consumer is a tricky concept to start with, as we all know. Perhaps the median consumer?
  • Core inflation, excluding volatile items like food and energy? No small debate there! And lives depend on it as central banks often seem to regard high headline inflation inflation as a justification for raising rates, trumping core rates. And then low headline inflation rates are somehow to be ignored and instead a focus on core inflation, or even core inflation expectations a very long way out.
  • Housing costs are 30-40% for the average consumer. And they are very sticky. Incorporating “inflation” in those housing costs is immensely tricky, either tracking sticky prices like rent, and of then of course separating out changes to the quantity and quality of what is rented. Or, even worse where there are a lot of homeowners, you need to measure inflation in imputed rent (ie the purely inflationary element of the theoretical benefit homeowners get from owning their homes, again stripping out those pesky quality and mix changes). the UK ONS has been criticised for not measuring housing inflation properly, but it is hard to do.
  • Or should expected inflation be used? Perhaps this has the benefit of never needing to be revised. But there are numerous versions of expected inflation. Short, medium and long term. Consumers expectations, business expectations. These are very different in the UK at the moment. Or market-implied expectations? But expectations of what? CPI, RPI, Core CPI, the deflator? And here we have some genuine issues about reliability anyway, as neither the government or the central banks don’t like it they don’t really focus on helping improving the efficiency of this market. The table from the BoE August report  illustrates the “pick a number” problem.

JA Tweeterati

A better inflation measure is the deflator, but …

If we really want to improve national well-being we need more output, more wealth, so the inflation rate that really needs to be divined is the output deflator: how we move from nominal value of output to real output, stripping out inflation. Then somehow target not having too much of that sort of inflation.

That deflator is not an easy measure either. The same hedonistic challenges apply, i.e. how does the quality of output change over time. The same mix effect causes huge challenges as economies are both complex and highly dynamic. Eighty percent of most modern economies produce services, not goods. Inflation in service sector output is no easy thing to measure, e.g.

  • Housing output from existing houses (10% of a modern economy by value added)?
  • Output from the government in terms of schools, hospitals/defence and bureaucrats (about 5% each)?
  • Distribution (15% of a modern economy)?
  • To say nothing of banking, insurance, legal services, IT, pointless celebrities, etc, etc.

An excellent essay by Geoff Tily of the UK’s ONS demonstrated the challenges of measuring either service sector output and its deflator.

UK CPI is the least reliable measure to target: it’s never even gets revised!

Laughably, in the UK the Consumer Price Index is targeted by the Bank of England. This index is a political and contractual one, as it used to set welfare and pension increases, the return on index-linked bonds, for inflation swaps, and often in wage negotiations. Some clever people think it is reliable because it is never revised. This is a really foolish argument. The lack of revisions demonstrates its political and contractual importance, but also proves it is not a reliable measure of what is happening in the economy. Only economic measures that get revised are reliable as they demonstrate a proper respect to the difficulties of measuring macroeconomic variables. No respectable professional macroeconomist should use it in any model of how the economy works.

Output is the way to go, but expectations not the rear-view mirror

So what about targeting output? Well the clever people spot the problem. It is often revised. Partly due to revisions of the physical output, partly due to revisions of nominal output (by value), partly due to the deflator, partly due to mix changes, methodological changes, even mistakes. Revisions are the output number’s strength, not it’s weakness. It is constantly improved upon, unlike the joke macro number of the UK CPI.

Critics are right that targeting a number that gets revised is a problem, but that is why you should look through the problem and target expectations for the number. If there is a rationale for targeting inflation expectations, it can be applied to output too. Expectations can’t get revised, they are updated, and do influence actual behaviour. If output growth is expected to be robust people will act differently to if they expect it to be poor. Historic price numbers or output numbers are the rear-view mirror. You can’t drive by looking behind you.

There are a host of subsidiary numbers that could be targeted like wages, employment, business surveys etc., but none provides the full picture of a dynamic economy, and are all flawed for that reason. Perhaps nominal wage growth could be targeted as that is at the heart of the problem of macroeconomics, unemployment being caused by downwardly sticky wages during a turndown in aggregate demand. I feel just targeting nominal wages might be a bad thing as it could end up being gamed or rather indexed, but it is worth thinking about.

A word about unemployment targeting

Employment, or rather unemployment, could be targeted. But just look briefly at the almost endless debates about labour force participation rates, especially in the US; about the quality of the employment, especially in the  UK; or about the real number of unemployed, especially in countries with large informal economies and easy unemployment registration. Unemployment is not the answer. Perhaps, expectations of unemployment?

Derivatives of output, like output per head, or output per hour (ie productivity) could be targeted but these obviously suffer from the challenge of first measuring real output reliably. In addition, I am not sure how or who would target expectations of output per head or per hour.

Real output or nominal output?

And there is the final measurement challenge. If measuring nominal output is tough and measuring the deflator is tough, measuring a derivative of the two is more than doubly tough (RGDP=NGDP/Deflator).

I know that some chunks of RDGP are measured by measuring actual stuff produced. Officials do gather data on the oil pumped out of the wells, the cars coming off the production line, or bushels of wheat from the farms. But extraction industries, manufacturing and agriculture are a very small fraction of any economy these days – even in an industrial powerhouses like Germany.
So, which output number should be targeted, a real or a nominal one? Both.

  • Governments (or rather “society”) should target (be concerned with) real output, the creation of wealth.
  • Central banks should target expectations for nominal output (NGDP), i.e. the value of the real output measured in money terms – assuming a steady growth in nominal output is a good thing, which most people do.

The balance between nominal growth and real growth, inflation, can then be left for society to figure out.

Blanchflower Baloney

A Mark Sadowski post

In a recent post James Alexander caught Danny Blanchflower tweeting that he thought “NGDP totally impractical due to data revisions”.

This is a familiar complaint, voiced for example by Goodhart, Baker and Ashworth in January 2013.

There are numerous problems with this line of thinking.

First of all, central banks shouldn’t be targeting past values of economic variables anymore than one should attempt to drive a vehicle on a superhighway by looking in the rearview mirror. Arguably the world’s major central banks tried doing that in 2008, and we are still living with the results. Since central banks should only be targeting the expected values of economic variables, bringing up the issue of data revisions reveals a level of obtuseness that borders on the ridiculous.

And as irrelevant as the issue of data revisions is to the proper conduct of monetary policy, although NGDP levels tend to be revised, that certainly should not imply that inflation rates are not revised. In fact the personal consumption expenditure price index (PCEPI), the inflation rate of which is the official target of the Federal Reserve, often undergoes significant revisions.

There’s two main ways of measuring the size of the revisions of the components of national income and product accounts: 1) Mean Revision (MR) and 2) Mean Absolute Revision (MAR). For rate targeting MAR is the more appropriate measure, and in fact the MAR of inflation is usually smaller than the MAR of NGDP. However, for level targeting MR is more appropriate.

Interestingly, at least in the US (Page 27):

“The MRs for the price indexes for GDP and its major components are generally not smaller than those for real GDP and current-dollar GDP and its major components.”

In fact, over 1983-2009 the MR for the final revision to quarterly NGDP is 0.14, whereas over 1997-2009 the MR for the final revision to the GDP deflator and the PCEPI is 0.20 and 0.12 respectively. And over time the revisions have trended downward.

So I suspect that the MR for NGDP is smaller than the MR for PCEPI over 1997-2009.

Which means the claim you frequently hear that NGDP revisions are larger than inflation revisions is pure grade A horse manure. You will never see any evidence supporting this mindlessly repeated spurious claim, because no such evidence exists.

And, finally, inflation is a totally artificial construct requiring that we come up with an estimate of the extraordinary abstraction known as the “aggregate price level.” To see how preposterous this is imagine equating the aggregate price level between what it is now and what it was in say 14th century England. The goods and services are so different it requires the complete suspension of one’s disbelief.

In particular, PCEPI inflation is the difference between nominal PCE and real PCE, meaning PCEPI inflation is nothing more than the estimated residual between a truly nominal variable, which is relatively straightforward to measure, and a real variable, which is fundamentally an exercise in crude approximation.

It’s high time that central banks moved beyond the near medieval practice of targeting real variables and/or their residuals, and started targeting truly nominal variables, which according to the accepted tenets of monetary theory is their proper domain.

NGDP data revisions means targeting it is “totally impractical”, really?

A James Alexander post

Danny Blanchflower, the enfant terrible of the Bank of England when he was for a time on its rate-setting Monetary Policy Committee, tweeted here that NGDP Targeting was “totally impractical” because it was a figure that was so subject to data revisions.

Well, first off, revisions don’t matter to Market Monetarist advocates as the “Market” bit of MM refers to the theory that you should target expectations for NGDP Growth. Scott Sumner has argued for a market in NGDP Futures so accurate market forecasts of NGDP growth could be used. This proposal is not much of a leap for standard macro theory that argues you control inflation by targeting inflation expectations. So it’s no big deal to target forecasts, every macro-economist should get that bit.

Market Monetarists are highly sceptical of inflation expectations as measured purely in surveys. The public has little understanding of the term inflation. Although the public sees their shopping basket go up and down in cost it has only modest relevance to the central banks that like to target core inflation, i.e. excluding volatile food and energy. On saner days central banks actually target the measure of inflation that gets them from NGDP to RGDP, known as the GDP Deflator. This figure often runs well below measures of consumer inflation. A good discussion on the good reasons why this happens is here.

Consumers would also have little view on NGDP, of course, if that were targeted. They would roughly see one large element though, in aggregate wage or income growth. This is because NGDP (and therefore RGDP) can be measured by adding up all the income in the economy, via the “income method”. The two other ways to measure NGDP are the self-explanatory “expenditure method” and the very tricky “output method”. Of course, I don’t have to explain this to expert economists like Danny Blanchflower.

Second, NGDP should, in theory be more reliable to measure since RGDP is based upon NGDP. Any errors in NGDP will, de facto, be in NGDP. However, RGDP errors are then compounded by errors in the deflator. It is therefore impossible for RGDP to be more reliable than NGDP.

The question is then whether NGDP is more or less prone to error than “inflation” measures. Well, NGDP only has one definition, so that certainly helps, inflation many. The infamous switch in the UK from RPI to CPI as the main method was certainly a major revision to method and certainly sowed huge confusion. It still does cause problems as index-linked bonds are still based on RPI, as are most inflation-protected pensions and other forms of financial income.

Lastly, Blanchflower demanded that I provide some empirical research that NGDP is less prone to error than RGDP or “inflation”. Well, here is a study by the US statistical office, the Bureau for Economic Analysis that demonstrates that NGDP (“Current-dollar GDP” in the table) is at least as reliable as RGDP looking at the period 1993-2013. This can’t actually be true given RGDP’s two sources of error. The resources spent on RGDP are much larger than those spent on NGDP so that may account for the BEA’s result.

JA-D Blanchflower

Whatever, it at least shows Blanchflower’s “totally impractical” is just wrong in fact as well as theory. He should stop talking nonsense

The great benefit of NGDP Targeting is that it means the central bank concerns itself only with nominal stability and doesn’t have to get concerned with the balance between inflation and real growth, or the hard to measure productivity growth that drives the difference. Those issues can be left to real economy experts to sort out. 

In this way of thinking neither inflation nor Real GDP growth should be of any concern to a Nominal GDP Targeting central bank.