RPICPI User Group

last person joined: 3 months ago 

To foster co-operation and the exchange of information between the ONS, its advisory bodies and other users. Please see 'Aims of the RPI/CPI User Group' in the library for further details.

Today's Inflation Data

  • 1.  Today's Inflation Data

    Posted 16-01-2018 22:01
    Hi All

    I wanted to draw your attention to this from part of the releases today on UK inflation in December and in particular to this component of CPIH.

    "The OOH component annual rate is 1.3%, down from 1.5% last month."

    Is there anybody on this forum who actually believes than Owner Occupied Housing costs are only growing at 1.3% per annum?


    Shaun Richards


  • 2.  RE: Today's Inflation Data

    Posted 17-01-2018 09:31
    Certainly don't believe it.

    Has anyone seen Mark Carney's letter to the Chancellor explaining how the inflation target came to be breached last month ?


    how the US CPI actually worked to contain housing inflation


  • 3.  RE: Today's Inflation Data

    Posted 17-01-2018 18:05
    I don't believe it either.

    I think it is time that ONS abandoned this methodology and used the User Cost approach.


  • 4.  RE: Today's Inflation Data

    Posted 18-01-2018 09:25

    Better still, abandon CPI and CPIH as headline series and concentrate on the Leyland/Astin methodology.


  • 5.  RE: Today's Inflation Data

    Posted 18-01-2018 16:27
    I am of the opinion that OOH costs should include capital costs, which means Depreciation and Opportunity costs.  The closest such measure at present is that provided by the RPI, though it does have its faults.

    The only two methodologies I know of which properly do this are the Rental Equivalence (RE) and User Cost approaches.  The User Cost approach is the more direct measure, whereas the RE approach is indirect.

    I am not in principle opposed to the RE approach, but the way it has been implemented by ONS does not pass basic statistical principles and gives rise  to serious underestimation  of OOH cost inflation.  In particular:

    a) ONS does not compare like with like in quality terms.  It is unable (using its current source) to distinguish New let from renewals.  If you do not do this you will underestimate OOH inflation.

    b)  The weighting framework used for post-stratification is nowhere detailed enough for a sample which is very far from random.  I have suggested that the number of strata used for weighting should be about 100 times the number actually used.  Without detailed post-stratification, there are serious composition differences between time periods and these are almost always reflected in downward discontinuities which cause an underestimation of OOH inflation.


  • 6.  RE: Today's Inflation Data

    Posted 19-01-2018 15:20

    Dear Michael:

    I largely agree with you. For a traditional consumer price series, John and Jill's broadly defined payments approach is the best. I sent a note in that vein on January 17, or tried to, but it seems to have disappeared in a galactic worm hole, so I am resending it below. Whether it should be the headline index depends on whether you think the traditional consumer price series or the macroeconomic inflation indicator (a reformed CPI, but definitely NOT the existing CPIH) should be the headline measure. There is a lot to be said for making the macroeconomic index the headline measure, but reasonable people can agree to disagree on this.

    Best regards,


    Just to support Shaun's point, although there is no owner-occupied housing component published for the Retail Prices Index a weighted average of the OOH-related components of the RPI would indicate an inflation rate for OOH costs of 3.6%, or 2.9% if you make the adjustment for the formula effect:


    2017 wt

    Dec 2017 annual inflation rate

    Mortgage interest payments






    Repairs and maintenance charges



    Do-it-yourself materials



    Dwelling insurance and ground rent



    Estate agents' fees



    OOH components of RPI



    Note that even more than housing prices reflected in depreciation, dwelling insurance prices rose strongly. My rough estimate includes an assumed expenditure weight and an imputed price increase for estate agents' fees, part of the RPI for could be improved if it incorporated estate agents' fees, part of the RPI for fees and subscriptions. ONS staff could certainly calculate a better estimate than I have done, but it wouldn't be very different from mine. The RPI treats stamp duty as out-of-scope, and its inclusion might also bring the OOH inflation rate up.

    The biggest claim for a rental equivalence approach to OOH is its consistency with the System of National Accounts so the CPIH doesn't and logically can't include estate agents' fees and stamp duty since these are part of gross fixed capital formation in a national accounting framework. Most of the other components are supposedly covered off in the imputed rents series although obviously not in a satisfactory way if the CPIH imputed rent series is so much lower.

    In the long-term a household inflation index as recommended by John Astin  and Jill Leyland represents the best measure of the increase in OOH costs. Unfortunately, the HCIs that ONS have calculated so far don't properly reflect their vision, treating both equity payments on houses and renovation expenditures as out of scope. So for the moment, the RPI, living on life support, with its notorious formula bias, remains the only official price series designed for traditional purposes that is appropriately sensitive to housing prices.