This release is supported by our methodology article detailing how the new HCIs have been calculated and showing how the all-households index contrasts with the all-households CPIH.
Users are invited to comment on future priorities regarding the HCIs. An optional questionnaire is provided in Annex A of the methodology article. Any feedback should be sent to email@example.com by Friday 23 February 2018.
We take this opportunity to thank you all for your continued support. In particular we are grateful to John Astin and Jill Leyland for their continued support and contributions to the development of this work.
The User Group will hold its next public meeting on 22 February 2018 (3.30 to 5 pm). It is free to attend, but please register at - https://www.eventbrite.co.uk/e/rpi-cpi-user-group-meeting-introducing-the-household-cost-indices-tickets-42528949232
Introducing the Household Cost Indices, 22 February 2018 (3.30 to 5 pm)
The user group has long been calling for a measure of inflation that better captures households true expenditures, that it believes could be used for uprating pay and benefits.
ONS recently published its first estimates of Household Costs indices, which aim to capture households experience of changing prices and costs. The paper is on the ONS website at:
The RPI CPI User Group is therefore hosting a public meeting to discuss the new indices and look to provide feedback to the ONS on user views .The meeting would include:
The meetingwill be held at the Royal Statistical Society offices (the nearest Tube stations are Moorgate, Barbican and Liverpool Street). Please pre-register if you plan to attend the meeting.
For those who may not be able to attend or would like to submit a question in advance please email firstname.lastname@example.org with the comments you would wish us to raise.
The meeting is open to all so please circulate to all interested parties/organisations.
RPI CPI User Group Committee
Helen Sands deserves congratulations for putting together the first version of the HCI but arguably there remains a considerable amount of work to be done.
I attach my completed questionnaire to the Household Costs Indices: methodology Annex A: Questionnaire. There is little that is new and the response is largely a reminder of points contributors to SUN have made in the past. The response is 5 or so pages long a large part of which are extracts from various documents to support the points made.
In summary the 4 points are –
Note: the recent ONS article – "Shortcomings of the Retail Prices Index as a measure of inflation" – which in parts is selective and factually incorrect has raised potential issues with the methodologies that ONS may allow for the development of the HCI. It is to be hoped that the following statement by the National Statistician in the forward indicates that the development of the HCI will not be restricted by the at best questionable views set out in the above article –
"The course on which ONS has embarked has therefore been two-fold:
All the best.
Interesting presentations and discussion on the HCI yesterday.
However, there does seem to be a need to clear up a couple of potential misunderstandings.
First, EU regulations do not ban Carli but specifically allow it to be used for the HICP under certain restrictive criteria – see extract from the regulations below.
Second, the ILO Manual also allows Carli – see extract below.
Finally, chain drift is not a consequence of the choice of elementary aggregate formulae like Carli but of the choice of chain linking algorithm. See SUN post below and its associated thread –
COMMISSION REGULATION (EC) No 1749/96 of 9 September 1996 on initial implementing measures for Council Regulation (EC) No 2494/95 concerning harmonized indices of consumer prices
Price indices for elementary aggregates
HICPs shall be compiled using either of the two formulae given in paragraph 1 of Annex II to this Regulation or an alternative comparable formula which does not result in an index which differs systematically from an index compiled by either of the given formulae by more than one tenth of one percentage point on average over one year against the previous year.
Formulae to be used in compiling elementary aggregates
Extract from Consumer price index manual – ILO 2004 (wcms_331153.pdf ) page 159, see final sentence in italics.
Chain versus direct indices for elementary aggregates
9.42 When a replacement item has to be included in a direct index, it will often be necessary to estimate the price of the new item in the price reference period, which may be some time in the past. The same happens if, as a result of an update of the sample, new items have to be linked into the index. Assuming that no information exists on the price of the replacement item in the price reference period, it will be necessary to estimate it using price ratios calculated for the items that remain in the elementary aggregate, a subset of these items or some other indicator. However, the direct approach should only be used for a limited period of time. Otherwise, most of the reference prices would end up being imputed, which would be an undesirable outcome. This effectively rules out the use of the Carli index over a long period of time, as the Carli can only be used in its direct form anyway, being unacceptable when chained. This implies that, in practice, the direct Carli may be used only if the overall index is chain linked annually, or at intervals of two or three years.
Note: the qualification "only" in the final sentence in italics above does not hold as the chain linking algorithm can be changed – see final point above.
Of course you are right in what you say about the Carli formula. The ILO Manual doesn't forbid its use and, to my mind, is even somewhat unfair in its comments on the rival Jevons formula regarding its alleged sensitivity to outliers. However, it isn't the only international guide out there. The Practical Guide to Producing Consumer Price Indices says that the Carli formula "is positively discouraged as it is particularly associated with some bad characteristics", which is probably going too far.
The Eurostat guidelines for the use of the Carli formula are unreasonable. The Carli and Jevons indices will differ from each other more the greater the variance of the component price relatives. It doesn't really seem to make sense to allow the use of the Carli in a context of annual chaining where there will be upward chain drift just because the variance of the price relatives is small and so the damage done is small. In a document in our library "Spoken Comments on Peter Levell's paper Is the Carli Index Flawed?" I mentioned where I thought the Carli formula should be legitimate, the most obvious being the one where pps samplng based on expenditures generated the priced items. In this context, the restriction imposed by Eurostat that the Carli index number couldn't differ greatly from the Jevons would have no merit.
In the future, the Carli formula could well be used in the HCIs but I doubt it will be used very much. It will be much closer to the RPIJ than the RPI.
There may be other situations where the use of the Carli formula would be appropriate. Until I started communicating with you and other people in the RPI CPI User Group about it I really didn't think there were any at all. It shows that these exchanges of views are useful in helping us (or me at any rate) change our views for the better.
The name CPI was adopted specifically to hide the fact that it was an EU compromise measure rather than a home-grown one. However, the Government tried to have its cake and eat it. It argued that because the CPI was an internationally agreed standard, it could not be manipulated to favour the UK Government (the implication of course being that RPI and RPIX could be). Of course, the switch from RPIX to CPI for inflation targeting, and then the switch from RPI to CPI for uprating (except when it was advantageous to retain RPI, e.g. for student loan repayment), were huge pieces of manipulation.
The HICP was designed as a macroeconomic index for use by a central bank as a target inflation indicator, so in that way it made sense to switch to it from RPIX. The big mistake was to switch before an OOH component based on a net acquisitions approach was added to the HICP. However at the time it seems everyone believed that this would be quickly done. An FT article by Scheherezade Daneshku said that: "Pilot studies completed this year will soon be extended. If they prove successful, the index would be modified to include house prices in 2006." Of course, it didn't happen. In fact I don't think the second round of pilot studies, expanding the number of countries included from five to 12, even started until 2006. One does wonder why Eurostat has proceeded at such a slow pace.
I agree with Gareth's suggestion to switch the name of the CPI back to the HICP. It also seems obvious now, more than it probably did in 2003, that there are serious defects in the HICP and in the Eurostat proposal for an OOH component, so that the ONS should probably calculate its own macroeconomic index, perhaps called HICP 2.0, which would be a more appropriate target inflation indicator for the Bank of England. One could, and should, continue to calculate an HICP according to Eurostat guidelines for comparability with all the countries to do so.
It would make more sense to call the HCI, if it were closer to the original proposal for a Household Inflation Index, the UK CPI. It could potentially be the model for CPIs all over the world.