Can we still trust the Gross Domestic Product (GDP) numbers or not?

Today sees or if you read this later will have seen a multitude of economic growth or Gross Domestic Product numbers released across Europe. For example we have already seen some from Spain so let me open with what was in 2014 a good news story.

The Spanish economy recorded a quarterly growth of 0.7% in the fourth quarter 2014. This rate is two higher than that recorded in the third quarter of tenths (0.5%).
– The growth compared to the same quarter last year stood at 2.0% vs. to 1.6% in the previous quarter.

These are numbers which confirm the initial estimates and if true mark a turn for the better in Spain. But my theme today which is one that regular readers will be familiar with is can we trust the numbers? To be more specific I am troubled by the deflator series which is the measure of inflation used to turn nominal GDP numbers into the real GDP figures which I have quoted above. I have pointed out before that it has been erratic and inconsistent in Spain particularly in 2011 and 12 ( the later revisions totalled some 1.1% if my memory serves me correctly) and worry that at a time of negative inflation more issues may emerge.

, the implicit deflator of the economy presents a variation rate of -0.6%, three lower than in the third quarter (-0.3%) tenths.

The UK Implied Deflator

Yesterday I attended the Public Meeting to discuss the Paul Johnson Inflation Report. There was a wide-ranging discussion partly in response to my questions but there were others of note also. Sadly Paul Johnson was unable to continue the debate as instead of there being another Question and Answer session as scheduled he declared that he had a diary clash and left at 4.34 pm. I guess that maybe we should be grateful that the soon to be Sir Paul was able to spare us that much time.

However there were issues raised about the UK Implied Deflator which are important. A major one was raised by Dr. Mark Courtney who gave a lecture on the subject of the RPI (Retail Price Index) before Christmas. The relevance here is that the usually higher RPI was replaced by the usually lower CPI in 2010. Approximately 5% already used the CPI and another 18% was added meaning that with some rounding it was 24% afterwards.

Now if you use a lower inflation measure with the same level of output hey presto you get higher real GDP and hence economic growth! In essence this is what happened from 2010 onwards. Let me use the words of Dr.Courtney as best I can recall them.

The switch changed the numbers for the consumer sector in particular and as the Implied Deflator fell below the level that using the RPI would imply there was a 0.5% per annum higher rate of economic growth recorded.

That is an interesting piece of timing as of course back then the UK was pretty much desperate for some recorded economic growth.Or as the Office for National Statistics puts it.

From Q3 2009 growth continued to be erratic, with several quarters between 2010 and 2012 recording broadly flat or declining GDP.

If we bring that forwards to today’s numbers then we could on our old measuring stick wipe this out.

UK gross domestic product (GDP) in volume terms was estimated to have increased by 0.5% between Q3 2014 and Q4 2014

It is a little unfair however to put it all in one-quarter so let me just say that the number below would now be 2.2%.

Between Q4 2013 and Q4 2014, GDP in volume terms increased by 2.7%, unrevised from the previously published estimate.

So we would be in the same ballpark as Spain rather than thinking we had a superior performance. For those who have plugged our relative outperformance this is somewhat awkward to say the least and of course it means that we need to trim a bit off this too.

In Q4 2014 GDP was estimated to have been 3.4% higher than the pre-economic downturn peak of Q1 2008.

There followed an even more technical discussion of which type of statistical index and measuring system was appropriate for a GDP Deflator. For those of you who think that this might send someone to sleep well actually it woke up the person who Paul Johnson had sent to sleep and gave us some relief from his snores! I thought that those who rang his mobile several times were a little cruel. However that discussion ended again with the point that the GDP deflator was usually the lowest inflation number these days. Just as a reference here is the latest reading and please remember all the other inflation measures were higher then.

The gross domestic product implied deflator at market prices for Q4 2014 is 1.1% above the same quarter of 2013.

Royal Bank of Scotland

This may seem an unusual issue to raise but my point is that in the pre credit crunch era it was one of the powerhouses driving the UK economy forwards into a new economic era. Hence it apparently helped push economic growth higher. Now let us move forwards to this morning.

UK state-owned bank RBS has reported £3.5bn loss for 2014, down from £9bn loss the previous year.

The results were hit by a £4bn writedown on the sale of its US business Citizens.

In addition it plans to continue to deleverage and contract over the next year.

RBS said it had reduced costs by some £1.1bn and will cut another £800m this year.

Wasn’t last year’s figure supposed to be of the “kitchen-sink” variety so that the new Chief Executive could stride powerfully into the future?How can it lose money with the favourable way banks are treated via for example the Funding for Lending Scheme of the Bank of England. But more pertinent to today’s subject is how much of the economic growth recorded prior to 2007 was a mirage or an illusion?

The rewards for failure go on as Sir Howard Davies was only recently announced as RBS Chairman. I guess his resignation from the board of my alma mater the LSE due to links with Libya and Gaddafi are a mere bagatelle to our establishment or his tenure at the FSA.

The problem that is/are rents

Paul Johnson recommended that CPIH where H equals housing costs be the UK’s premier inflation measure going forwards. A little care is needed in this as he showed a way of getting an ever lower inflation reading in his presentation and that may be along soon. However so far CPIH has been a national scandal and a disaster as my past blogs have identified and it fails this test established by the Royal Statistics Society.

Understanding and faith in published statistics are incredibly important as is the long historical record that the RPI represents.

Derek Bird who sadly for him always seems to get the job of defending the indefensible poor man presented the new improved CPIH. It looked very different which poses a challenge to those who told us all was fine and recommended it! I will leave the critique to someone from the audience who asked this question.

Did you find any owner occupiers who knew what the rental value of their property was?


There is much to consider here as we examine the issue of how inflation is measured in our national accounts. There are competing and differing views on its treatment. However if we step back for a moment and look at the fact that our GDP performance was raised by 0.5% per annum due to statistical switch from RPI to CPI there is an issue. Why not publish both? Is it because we might in Victorian terms scare the horses? Also this is far from the only statistical change as we review the effect of ESA 10 which was applied last autumn.

Average revision to level of GDP 1997 to 2009 is

Whilst we need to update our numbers from time to time there seems to be an awful lot of such “improvements” going on and always we have lower inflation and higher economic growth as a result. Surely by the laws of probability something must one day happen which reduces growth and raises inflation? As to enquiries and reports well I remember the words of Sir Humphrey Appleby.

You never interfere with a member of Her Majesty’s judiciary. What you do is appoint someone who does not need interfering with in the first place!

More problems?

Released this morning.

The Office for National Statistics has today announced that Quarterly National Accounts data consistent with the 2015 Blue and Pink Books that were due to be first published on 30 June will now be published on 30 September. The Blue and Pink Books themselves will now be published on 30 October.

Eurostat has written to all EU member states stipulating that all outstanding Gross National Income (GNI) reservations pertaining to their national accounts need to be addressed by 22 September 2015. GNI is closely involved in the calculation of member states’ contributions to the EU budget. Accordingly, ONS will work with Eurostat over the coming months to meet this deadline.

Oh and just to be clear I do not blame the ONS for this,they are merely the foot soldiers and the puppet-masters are to be found elsewhere.


36 thoughts on “Can we still trust the Gross Domestic Product (GDP) numbers or not?

  1. Did you find any owner occupiers who knew what the rental value of their property was?
    Or who gave a flying monkey’s?

    • Hi therrawbuzzin

      I particularly enjoyed that question as it highlighted the arcane world which exists in the calculation of many of these numbers. Statistically they may be pure but once you shine the light of reality on them then the theory sometimes starts to look not only misplaced but also somewhat silly.

  2. Hi Shaun

    You may be too young to remember this but the unemployment statistics in the 1980s were continually revised. As you know unemployment increased very sharply from around 1980 until 1986, the year I believe it peaked. There were, if memory serves me correctly, around 30 changes to the unemployment statistics in the early 1980s and in only around two cases was the unemployed total increased!

    Now it may be that the original series was deficient and the changes were a necessary revision but one is bound to be somewhat suspicious of this; as you imply the ONS may not be as objective as many believe.

    As far as the GDP deflator is concerned I have believed for some time that it is a figure which has been “fixed”. Personally I believe that our recent economic performance has been somewhat less impressive than billed.

    • Hi BobJ

      I do recall some of that period and it was summarised and satirised expertly in Yes Minister. There was a specific quote on unemployment statistics which I cannot find so let me offer this on official statistics overall.

      “Sir Humphrey: If local authorities don’t send us statistics, Government figures will be a nonsense.
      Hacker: Why?
      Sir Humphrey: They’ll be incomplete.
      Hacker: Government figures are a nonsense, anyway.
      Bernard: I think Sir Humphrey wants to ensure they’re a complete nonsense.”

      Pure genius…

  3. Hi Shaun, re your last point about Eurostat and GNI calculations, one can only speculate what is happening in other European capital cities. I’m thinking here particularly of the home of the Emperor Nero and the musical hobby which was occupying him whilst the fire took hold…

    I’m equally certain of a further multi-billion “unexpected increase” in the UK contribution to budget, come September. I believe the word “unexpected” is already included in your infamous Financial Lexicon of these Times?


    • Hi Andy

      Yes “unexpected” is in my Financial Lexicon of these Times in the way of the sun rose unexpectedly today! As to how the GNI recalculations will play out I am not so sure but I will keep my eyes peeled for developments elsewhere in Europe.

  4. Hi Shaun

    Great article as always.

    With regards to GDP, this is one of the problems confounding the public at the moment. If they economy is recovering, why aren’t they feeling it. The answer is there is no recovery. If you strip out imputed rent, black economy and double counting, I would imagine there is no growth at all.

    And the only place you would find this information is in your excellent blog. The MSM ignore this. Still its easier to blame the bankers eight years later, than it is the politicians.

    • That’s probably because 8 years later the blame actually rests with the people who truly wield the power in this world…
      The bankers.

      • Banks make as much money as governments allow them , blaming them is not valid – if the people don’t like RBS they will stop banking with them . Governments love deflecting the blame to bankers , the poor little government is being bullied by the banks…… I don’t think so

          • seriously mate bankers are pictured as the very source of evil but it’s false…. beyond that I don’t blame governments anymore because it’s the peoples fault , the stupid f’in people vote them in….
            Imho Banking no longer requires the intelligence it used to , and all the smart people just get up to conniving and filling their boots , as do politicians, these smart people should be teachers or artists or …. dare I say it …. capitalists
            Most people can’t be bothered to do the work required by ‘democracy’ . Shaun by going to that stats meeting and blogging it out to us, is doing his bit….. and if there was 500,000 of him (or us depending on what you do) neither bankers nor politicians would be able to act with such impunity. Misdirected reprimands only serve to delay change , in my opinion……but what do I know !!

          • Agree about the “stupid people voting them in”…most of the time.

            But seemingly even when the people vote the “right” way… Like Ireland… Greece…

            …it still ends up the banks win. So why is that?

            And why are we STILL privatising profits and socialising losses?

          • when you vote the right way , you get to vote again ,there are hordes of smart people making a nice living out of kicking the can down the street , it’s just business as usual until the limits are met , the most tragic thing is that virtually all of them are ‘good’ people.
            It’s early days for Syriza with any luck greece forces the haircuts and there is an almighty correction , with BRICs and OPEC taking over, but at this point history suggests the top dog (today it’s NATO) somehow manages to start a war for economic reasons , oh look , we’ve just sent troops to Ukraine , don’t remember war with Russia being in any manifesto. Get to know your neighbours , get some skills that don’t require computers , get some chickens , enjoy youtube while we have it and most of all get a sense of humor because the people around you will need that – all you can hope to influence is the square mile where you live – Tim
            we are beyond economical repair !

          • I already have an allotment complete with 5 lovely egg layers. 😋

            I sincerely hope Syriza grow a pair, and fast!
            I also suspect you are right about war and Ukraine, although at this stage I think it’s more about the US ensuring Europe doesn’t end up getting too cozy with the new China/Russia axis than anything at all to do with Ukraine.

          • If the Vulture funds can take a Argentine destroyer can Germany just take a Greek island ?

            Greece will need to lock in some support before officially defaulting and printing Drachmas …. they will be designing the new notes right now
            we both understand the lunacy of 21st century economics and we both have allotments and chickens…. lets hope we are only hedging !

  5. Shaun,can you explain how CPI,RPI and CPIH differ in their treatment of housing,or can you send me somewhere I can read about it.
    Where do imputed rents come in,is that in the H component of CPIH or are imputed rents part of CPI and then they add on an adjutment for the cost of actually owning a place?

    I’m confused but trying to learn.

    Thought provoking blog today

    • Hi Dutch and thank you

      Okay the easiest is CPI which ignores housing costs completely and is therefore flawed. Next comes RPI which has mortgage interest rates, house prices via depreciation and other housing costs and the total is 23.7% of the whole. CPIH includes housing costs but crucially uses rents as a valuation tool for those who own houses as well as those who rent. This is where I argued it would go wrong and so far it has.

      Confusingly the rents of CPIH are not the same as imputed rent but I was assured yesterday that they will be soon. As the soap Soap put it

      Confused you soon will be!

  6. The PP party in Spain have put their own people into most institutes and as editors of the major newspapers. Can we therefore believe what we are told about the Spanish economy? I hope for the sake of the people of Spain that growth has infact returned but Spanish politicians are some of the most duplicitous anywhere and good figures in an election year make me cautious.

    • Hi Pavlaki

      I was troubled by the behaviour of the Spanish deflator in 2011/12 and other places such as Edward Hugh’s A Fistful of Euros picked up on it. I will wait for the full final version of Spain’s GDP numbers for 2014 and take another look.

  7. Hello, Shaun. Thank you very much for attending the meeting of the Royal Statistical Society and providing a summary of it. Paul Johnson isn’t George Osborne. I would really be interested in knowing what he had on his agenda that was so important that he couldn’t stay for a question-and-answer session. It gives the appearance that he really doesn’t have much confidence in his own recommendations and didn’t relish the idea of defending them.
    You are absolutely right that most people do not have a good idea of the market rental value of their homes, excluding furniture and appliances and the cost of utilities. This question was answered poorly on the Canadian Family Expenditure Survey (FAMEX). It was one of the questions with the highest non-response rates, the most dubious responses and it also provoked a lot of complaints from respondents. This led to it being dropped from the Survey of Household Spending, which replaced FAMEX in 1997. It was rightly believed that the rental equivalence estimates available from the Canadian System of National Accounts, not derived from household budget survey estimates, were adequate for calculating an owner-occupied housing series based on the rental equivalence approach.
    Emad Mansour and I calculated the last update of the analytical Canadian OOH series based on the rental equivalence approach and other approaches, including net acquisitions, which end with August 2000. Since this series ends with the basket based on 1996 FAMEX estimates still in force, the loss of the rental equivalence question from the household budget surveys never created a problem for us. Kazi Islam calculated an updated OOH series based on rental equivalence for the subsequent 2001 basket. On my recommendation, he used a rental equivalence expenditure weight based on a projection from the 1996 FAMEX. On hindsight, I think I should have advised him to use the SNA estimate instead, since the decision to leave it out of the SHS was taken on the basis that this was an acceptable estimate.
    Starting at least with the 2005 SHS questionnaire, the rental equivalence question was restored, and an OOH series based on the 2005 basket was calculated by Faouzi Tarkhani but not published. Why this was done is hard to say. The SNA equivalent rent estimates may not have been great. It is hard to see how such estimates can ever be anything but rough. However, they were considered good enough to move a big slice of real GDP so it is hard to see why they were considered inadequate for calculating an analytical consumer price series, especially one that never got published. The response rate for the SHS is now less than 50%, and the last thing it needs is an annoying question on imputed rent in which most people see no value.

    • Hi Andrew and thank you

      The rents issue in the UK has become quite a mess and as I wrote in the post I felt sorry for Derek Bird of the UK ONS who had the job of defending/explaining it. What he did not say is that the imputed rent series has undergone quite major revisions in the UK too and up to now is separate from the rental equivalence numbers.

      Even worse the rental numbers are collected by the VOA in the UK which only passes some of the data to the ONS. It make work statistically but as the procedure was outlined I could see that in economic terms it was poor as the leads/lags are likely to be far too long. Also they cannot do any quality assessments…

      Has the treatment rent issue arisen in Canada in any similar form or has it cruised along?

      • Shaun, regrettably, there is no-one remotely like you in the Canadian blogosphere, nor is there anything resembling the RPI CPI User Group for Canadian consumer price indices. So there isn’t a lot of discussion about either the CPI rent series or a rental equivalence series for OOH here compared to the UK, but it does get discussed just the same.
        There has been real concern in Canada as in the UK that the official rent series underestimate inflation, which of course would also affect the rental equivalence series. However, unlike the UK, not very much has been done about it. John Kovar stated the main problem very well in a 1987 paper: “The rent component of the Canadian Consumer Price Index is currently based on data collected on a six month rotating basis using a Labour Force Survey Supplement. Since changes in rents generally occur on an annual basis, the effective sample size of the Canadian Labour Force Survey (LFS) design is reduced. In other words, because the dwellings remain in the LFS sample for only six months, estimates of the magnitude of change can, on average, be obtained from only half the units. Furthermore, special annual benchmarks ,obtained by reinterviewing the June sample of dwellings one year later, indicated that the rent component can suffer from various degrees of downward bias (Dolson, 1982).” John had the very good idea of keeping the same dwellings in the rent sample for 13 months, pricing in the first, fourth, seventh and 13th months. Pilot indexes were calculated for several Canadian cities in 1985-86, now almost thirty years ago, as described in his paper, but his methodology was never implemented, nor have there been any follow-up studies to show what the CPI rent series might look like with dwellings kept in sample for 13 months or more.
        There is no very vocal backer of the rental equivalence approach at the Bank of Canada to compare with, say, Martin Weale at the Bank of England. The former Deputy Governor of the Bank of Canada, Pierre Duguay, was long a member of StatCan`s Price Measurement Advisory Committee. From his own remarks, made while he was still at the BoC, he seemed to be favourably disposed to the rental equivalence approach, but he doesn’t seem to have published on that subject.
        In 2006, Professor Gregor Smith received a research fellowship from the BoC. In 2009, he published the last paper on the CPI to make a big splash in economic circles, “The Missing Links”, presumably the fruit of his work as a BoC Research Fellow. Professor Smith completely ignored the work of Eurostat in creating OOH series based on the net acquisitions approach, very well advanced at the time, although his paper does not contain any Savanarola-style denunciations of asset prices in inflation measures in the manner of Mark Carney. Professor Smith is obviously more sympathetic to an opportunity cost approach to measuring OOH than to the accounting approach currently used in the Canadian CPI, although he finally recommends that the OOH component be re-examined, but not necessarily replaced. He is not enthusiastic about the rental equivalence approach, but since an opportunity cost index for OOH can always be made to equal a rental equivalence index given the appropriate assumption about the rate of return on owner’s equity, it is easy to see how a recommendation to consider replacing the current approach with an opportunity cost approach could morph into the replacement of the existing official OOH series with an imputed rent series.
        My own belief is that there is a reservoir of sympathy for the rental equivalence approach within the BoC. In the future, it could lead to the All-items CPI being restructured somewhat along the lines of the CPIH. It would be an absolute disaster if it happened, but one can hardly exclude the possibility. Remember that as the UK is part of the EU, Canada is part of NAFTA, and Canada is the only one of the three member countries that does not use a rental equivalence approach to measuring OOH in its CPI.

  8. “…was one of the powerhouses driving the UK economy ….”

    So was oil

    I do wonder if there’s anything else left to loot …… NHS perhaps ? BBC ?


    yup popcorn is getting cheaper , must be because so much oil goes into it !

  9. the tinkering goes on , I thought the ONS was supposed to give us “best in class” for all stats but it seems to be a soppy puppy dog following it’s master/s

    And why are so many expenditures based on GDP and not tax revinues?

    GDP these days are made up Gerry Mandered figures – there’s not even error bars ! ( 3% +/- 0.1% or is that 0.5% ?? )

    and if CPI is so great , why are HMG pensions and taxes based on RPI ( because its higher? )

    and then the pollies wonder why we think they are a bunch of snivling , coniving, b’tards out to only fill their own boots with gold ……

    oh and if you want something , you can borrow , but don’t ask for a pay rise or for the price to fall ….. know you place. pleb!

    there, got that off my chest !


    • “And why are so many expenditures based on GDP and not tax revinues?”

      Because if the public saw thwe ratio of Nationaldebt to Government revenue there would be pure fear, followed by panic, followed by fury at the way they have been misled as to the real possibility of ever repaying Government debts a followed by rioting revolution and Governmental ovethrow!

    • Hi Forbin

      Don’t forget that Bank of England pensions are based on RPI. Paul Johnson said he wanted to change the rules at which point Michael Lyon of the Bank of England spoke up and pointed out that Paul Johnson would their permission to do such a thing. There was some chortling later from the pension funds who thought he was protecting the BoE pension fund!

      Paul Johnson then gave a good impression of a deckchair folding up…

  10. The internet means that we can actually beat the banks, you know.
    If only we could break through the apathetic, know-our-place ethic of the British populace.

    • The sad thing is that the British public are unbelievably ignorant of economic matters and do not appear to want to know anything outside of what they are told by the media – in conveniently short messages. I am often surprised by my friends who are well educated, hold down professional jobs (or did!) and yet have a very poor understanding of what is going on in the world of global economics. I try to impress that this has a great impact on their wealth, future lifestyle etc but eyes glaze over. To avoid being a bore I give up! It should be taught in schools.

  11. This is a really great article.

    > But more pertinent to today’s subject is how much of the economic growth recorded prior to 2007 was a mirage or an illusion?

    Yes! Any economist who blithely compares now to 2007 goes straight into the bin for me.

  12. RBS continuing to lose money, quelle surprise ?

    Just like British Leyland before them – RBS is a subsidy guzzling failure only fit for the knacker’s yard. The sooner our politicians grasp the blindingly obvious, and accept that RBS is inevitably bankrupt, the better off the country will be.

    • Hi ExpatInBG

      I like the British Leyland analogy which younger readers may have to google :). Whilst there are specific differences, spiritually they are indeed very similar situations.

      How is the BNB doing with its Euro peg?

      • The political debates are about borrowing and joining the euro.

        The peg is not currently under threat, because Denmark, Switzerland etc worried about their costs being high from strong currencies. Bulgaria is a very low cost country, and I can’t see how speculators profit betting on a rising Lev, especially while the politicians keep increasing borrowing ….

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.