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.

APCP 2017 Annual Report to the National Statistician

  • 1.  APCP 2017 Annual Report to the National Statistician

    Posted 19-01-2018 15:42
    Dear users,

    Today we have published the Advisory Panels for Consumer Prices 2017 Annual Report to the National Statistician.

    You can find it on the new APCP landing page on the UKSA website here:

    Kind regards,

    Helen Sands (Prices Division, ONS)

  • 2.  RE: APCP 2017 Annual Report to the National Statistician

    Posted 20-01-2018 18:55

    Dear Helen Sands:


    Thank you for the annual report on the two advisory committees. I am sure you will receive many interesting comments from members of this group about what is in the reports. However, I would like to complain about something that isn't mentioned in the annual reports at all: studying a replacement methodology for package holidays and other seasonal products.


    The report makes repeated references to how the work plan evolved out of recommendations of the January 2015 Johnson report. This is certainly not the case where package holidays are concerned. The Johnson report states (see p.165):


    Package holidays collectively represent around 6 per cent of CPIH, and are amongst the most time-consuming items in the basket of goods and services to collect. The issue with holidaysarises because holidays purchased at different times of year are quite different products; a holiday in Austria in January is quite different to one in July. ONS's consumer price statistics deal with this by forming price relatives that compare the current month price for a holiday with the same holiday a year ago, rather than the price base period. The price index is then formed as a weighted average of the price relatives observed so far. This is different to the methods used elsewhere in the index and results in price change for the holiday items lagging the rest of the index. Eurostat have started to look at ONS's method; we agree it should be a priority.[Italics added.] This also applies to the accommodation services proxy, which covers some holiday accommodation in the UK.

    Mr. Johnson does not mention, and was possibly unaware, that horse racing admissions are treated in the same way as package holiday trips.

    Eurostat's verdict on the methodology, referenced by Mr. Johnson, was expressed in its July 2015 HICP Compliance Report. Unsurprisingly, it found the methodology non-compliant, as it employs an ersatz for of seasonal weighting, quite different from the 'class-confined seasonal weights index' permitted by Eurostat, which is also dysfunctional.  For package holidays the ONS should apply the agreed methodology of following the prices of the same products from month to month. In October 2016, the ONS issued a workplan that said:" In 2015, Eurostat published the UK Compliance Report 21 which stated that ONS's method for package holiday collection is non-compliant. This is because price relatives for package holidays are calculated by comparing the price in the current month with the price the same month a year ago, rather than comparing the price with January." (Actually, that's not what the Eurostat report said, although it is true it is non-compliant in that way.) The workplan provided for a compliant methodology to be introduced in 2019, i.e. just a week or so before the UK is supposed to exit the EU.

    My own view is that it would be useful for the UK to switch to an EU-compliant methodology for package holiday trips and horse racing admissions for comparability with the many countries, in and out of the EU, that calculate HICPs according to Eurostat guidelines. However, for the other UK consumer price indices, including the newly published HCIs, a reliable seasonal-basket formula should be adopted, either the Rothwell formula or the Balk formula. It is a shame that Bert Balk is no longer a member of the Technical Panel as I am sure he would be most persuasive in this regard. The same seasonal-basket approach should be extended to other categories, like gardens, plants and flowers. Can one imagine anything more dysfunctional than having fixed weights every month of the year for the basic class that includes Christmas trees!

    This is really work the advisory committees should be involved in, the sooner the better.

    Best regards,



  • 3.  RE: APCP 2017 Annual Report to the National Statistician

    Posted 29-01-2018 09:39

    Dear Andrew,

    Thank you for your comments. We are aware of the Johnson Review's comments on package holidays and are currently working on improvements to the collection and methodology. We have not yet taken anything to the panels on this as the work is not yet advanced enough, however, we do hope to take something to the panels towards the end of this year or the beginning of next year. This work is currently on our Development plan for consumer prices with an implementation date of 2020, although this may be subject to change as the work progresses.

    Also to note, Professor Bert Balk remains a member of the Advisory Panel on Consumer Prices – Technical.

    Kind regards,


  • 4.  RE: APCP 2017 Annual Report to the National Statistician

    Posted 31-01-2018 13:24

    Dear Chris:

    Thank you very much for your reply. I am delighted to read that the ONS will be looking into a better methodology for package holiday trips and that the advisory committees will be involved.  I apologize for believing that Bert was no longer on the technical panel.  I am sure his advice will be extremely valuable in determining how to handle package holiday trips in the UK consumer price indices.

    Today l'INSÉÉ updated the French agricultural price indices for December and within a few days I intend to put a document up on the Group website with a proxy index for Christmas trees based on adjusting the French series to pounds sterling. The main reason to look at this series when the UK has its own producer price index for Christmas trees is timeliness. The French series is always available at the end of January and could actually be used in the UK consumer price indices as a proxy if there were a one-month revision period. The UK pricing is only available after several months in the best circumstances. However, the main purpose of providing this series is not to encourage its use as a proxy but to stimulate interest in the ONS pricing Christmas trees itself for use in consumer price series with a monthly-basket approach to seasonal products.

    Best regards,