The UK National Statistician has made a mistake which will affect us all

Today sees the release of the latest UK inflation data for consumer,producer price and house prices. However I wish to draw you attention to something very important which is this statement issued last week by the UK National Statistician John Pullinger.

I have therefore concluded that we will make CPIH our preferred measure of consumer price inflation as I indicated earlier this year. I believe that CPIH has a number of desirable properties, most notably the inclusion of an element of owner occupiers’ housing costs. It also addresses several flaws and limitations present in alternative measures. We intend to make CPIH the preferred measure from March 2017, by which time all the planned improvements will have been implemented.

Tucked neatly into the post Trump election furore you could call it “a good time to bury bad news” to coin a phrase. But first let me point out a problem which even the official statement acknowledges

It also gives an update on the work to improve the Consumer Price Index including owner occupiers’ housing costs (CPIH) with a view to it being considered for redesignation as a National Statistic.

I will come to my critique in a moment but I would like this to sink in as you see even those pushing this cannot escape the fact that they had to de-designate it and have been unable to fix that! Or as the latest statement on this subject puts it.

In August 2014, the National Statistics status of the CPIH was discontinued after issues emerged relating to the processing of some of the administrative data sources used to estimate Owner Occupiers’ Housing costs.

My Critique

I stood pretty much alone when CPIH was suggested and found myself arguing for example with the economics editor of the Financial Times Chris Giles. Let me take you back to the 24th of September 2012.

In some ways even worse some of the Rental Equivalence data had to be estimated as the series planned to be introduced only began in 2007 in England,2009 in Wales and 2010 in Scotland. So we have no actual evidence of the long-term reliability of these numbers.

I predicted it would not work and the official assessment now tells us what?

ONS needs to take more time to strengthen its quality assurance of its private rents data sources, in order to provide reassurance to users about the quality of the CPIH.

There is another issue which the National Statistician has attempted to fudge by writing “the inclusion of an element of owner occupiers’ housing costs”. How very Sir Humphey Appleby! I have noted that many people have reported that house prices are being included but you see they are not. Instead there is a statistical swerve based on the Imputed Rent methodology where they assume house owners receive a rent and then put growth in that in the numbers. The same rental growth measurement that according to their own missives  they need to “strengthen”. Back in August 2014 they had to announce an embarrassing change.

In essence officialdom was admitting that the numbers were wrong and the estimate was of the order of 0.2% per annum. I will let readers guess which way?!

It is not fit for purpose

There is another big issue which is that we went through a housing boom which contributed to the credit crunch so there is logic in looking to adapt to that reality. But as I poiinted out as long ago as 2012 CPIH fails here as well.

Jill Leyland of the Royal Statistical Society has analysed the data comprehensively and it was quite plain that the use of house prices would have given at least some sort of signal whereas to quote her directly if we look at the period from 1988.

There is little difference between CPI and the rental equivalence version of CPIH

So we have a made up number which we struggle to measure and even if we could measure it then it is not far off useless as an economic signal! Oh did I say not far off? Actually it is worse than that because if you look at expansions in the housing market there is a long lag of around 2 years, so by the time you might learn something it is likely to be already too late.

Inconsistency

The rationale for us using CPI in the first place was to align us with Europe. We are now diverging from that as they have house prices in their version of CPIH. When you ask those in authority about their inconsistency you get what might be called the sound of silence.

Oh and after the long list of problems issues and flaws from CPIH you might reasonably think that if we put it politiely there is plenty of cheek here.

It also addresses several flaws and limitations present in alternative measures.

Today’s numbers

Let me thank all women, ladies and girls reading this. You see in addition to boosting GDP with your clothing and footwear purchases you have managed to reduce CPI inflation.

Clothing and footwear, where the downward effect came mainly from garments (in particular women’s outerwear), for which prices rose by 0.2% between September and October 2016, compared with a larger rise of 2.3% a year earlier.

You seem to have completely out manoeuvered us men and I suppose you are adding the word again to that. Actually the issue of how we measure clothing inflation has been troubled since around 2010 and if we look for another troubled series we see the university tuition fees one.

The downward contribution came principally from UK and EU student tuition fees, where the impact from the rise in the cap for tuition fees (first introduced for new students in England in 2012) was smaller this year than in 2015.

Accordingly we saw this.

The all items CPI annual rate is 0.9%, down from 1.0% in September

How did CPIH do?

The subject du jour told us this.

The all items CPIH annual rate is 1.2%, unchanged from last month

So to be fair it has added something. What about house prices though?

Average house prices in the UK have increased by 7.7% in the year to September 2016 (unchanged from 7.7% in the year to August 2016), continuing the strong growth seen since the end of 2013.

I will leave the UK establishment to explain how 1.2% is the new 7.7%.

How is our old measure doing? You know the one which according to our establishment is not up to international standards.

The annual rate for RPIX, the all items RPI excluding mortgage interest payments (MIPs) index, is 2.2%, unchanged from last month

Oh it is reflecting things more strongly…..What about the improved version called RPIJ?

After giving this further consideration I can however confirm that ONS will cease publication of RPIJ from March 2017.

So after pushing it and effectively misleading people they are now treating it like someone wearing a come on your spurs T-shirt at an Arsenal football club AGM.

Comment

We have learnt much today and there is an irony in the fact that clothing prices have wrong-footed expectations for the numbers today. Problems with that series kicked off a new phase of the debate over UK inflation measurement from around 2010. Back then the official estimate of the “formula gap” gap between CPI and RPI was 0.1%. That estimate has not been served well by the passage of time as it is now 1.1%. Our “experts” are only currently out by a factor of eleven or so.

Underlying this we see that UK consumer inflation is on an upwards path as we see that this month’s measurement has met some old “friends”. Producer prices rose by 2.1% (output) and 12.1% (input) and will feed into the system over time. Just as we see that inflation will be on the march we will change our measure. Simply breathtaking!

Me in the Guardian

I expressed my views to the Guardian newspaper who have included them here.

https://www.theguardian.com/uk-news/economics-blog/2016/nov/15/new-uk-inflation-measure-office-for-national-statistics-cpih

 

 

 

 

Advertisements

16 thoughts on “The UK National Statistician has made a mistake which will affect us all

  1. I stopped having any track with these “economic indexes” a long time ago. They are chosen only to reduce benefits or to increase the wealth of those in charge.

    • Hi Foxy

      What you find is that the establishment and their media acolytes close ranks. But if you continue to press ( the current one surrounds the weights used in CPIH and Imputed Rent) then you find that what was supposedly clear takes them 6 months to fail to explain! An emperor with no clothes.

    • Hi Forbin

      It is not clear to me that they have thought that far ahead. For example I can see the Bank of England being unhappy with this. Firstly it will draw attention to the house price inflation they have created ( if it doesn’t I will do my best to raise it!) and secondly they wouldn’t want their mandate to be based on CPIH ( although they may not have figure that out yet). Also HM Treasury may come under pressure to replace RPI weighting with CPIH on the bits which are worse for it.

      We could get a whole episode of Yes Prime Minister out of this.

      • I thought they put house prices in due to the fact they know the bubble has peaked.

        Amazing that we’ve had 2 decades where prices have gone up 10 fold in certain areas of London and trebled in most areas of the country yet they now decide to change this measure and ignore all this.

  2. ‘So we have a made up number which we struggle to measure and even if we could measure it then it is not far off useless as an economic signal! ‘

    Beautifully put Shaun.

    Thanks as ever for continuing to expose the inconsistencies that the MSM so blithely ignore.

  3. As ever, agree with your analysis on CPIH. This is a potential disaster for UK inflation stats. They key issue to me is that the H in CPIH stands for “housing”. However the elements included in it and how they are calculated do not correspond with what the public would expect from an inflation statistic including “housing”, ie
    1. It excludes house prices completely.
    2. Instead of measuring housing expenses directly (such as mortgage costs, building insurance, maintenance, etc) as people experience them, it uses a complex alternative measure based on rental equivalence.
    3. Despite arguing it is not possible to cost housing items directly, it then has just added in council tax as a complete exception to this rule, making OOH even more of a mess.
    The other issue I see that for the last two years CPIH has been about 0.3% higher than CPI. If this persisted, it is unlikely that government would want to use it either as a measure of their success in controlling inflation nor to index pensions and benefits (as it could result in higher expenditure). Furthermore, it would make it more difficult for the Bank of England to meet its 2% target so they won’t want it either.
    If the ONS wish to improve on the EU’s invention of HICP in this post-Brexit world, they might be far better to revert to the British measure of RPI. Although it too has its issues, it generally more accurately reflects inflation as experienced by consumers as it is usually higher than CPI (due to its simpler method of calculation) and it already weights in housing costs and house prices.

    • Yes but quite possibly they expect a dip in house prices, after all we have had unrealistic house price inflation for quite some time which must be due a correction or at least several years of slip sliding away.
      If so then the opposite will apply, and the government will not be liable for those pensions and benefits increases. Let us face it, they are so much cleverer than us!!

  4. Great post as always, Shaun. Congratulations on having your views reported in the Guardian. It is really a shame that you aren’t consulted as an expert more often by all the major UK papers.
    I really appreciate your shining a spotlight on the dysfunctional recommendations in the National Statistician’s letter. You and other people who read this blog may be interested in a related paper I wrote: “A Better Understanding of the Different Approaches for Measuring OOH Costs”. If you take the ONS background documents at their word, the ONS doesn’t realize that the OOH component of the RPI from 1995 forward is a user cost index, or that before that its OOH component was a payments index that is far narrower in scope than what Jill Leyland and John Astin have proposed for their household inflation index.

    • Hi Emily and welcome to my corner of the web

      There are various factors at play here. Renting usually does not offer the same permanence of living arrangements that owning/mortgaging does. Also lower mortgage rates have made monthly payments lower. But one area where renting does win is that you do not have to pay what are usually high prices nor saddle yourself with a large debt,

Leave a Reply

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

WordPress.com Logo

You are commenting using your WordPress.com 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 )

Google+ photo

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

Connecting to %s