Comments from the Stakeholder Advisory Panel on Consumer Prices
The report by the House of Lords Economic Affairs Committee ‘Measuring Inflation’, published on January 17th 2019 made reference to the work of the Advisory Panels and there was an implication that the views of the Panels on the Committee’s recommendations should be made known. Consequently, I am writing on behalf of the Stakeholder Panel, following our meeting on January 25th.
At that meeting there was a good discussion of the House of Lords Report – although this was inevitably preliminary as some issues would need detailed debate. The Panel intends to consider these further, recognising that the Report covered a wide range of important issues and raised challenges to the status quo. We also look to the ONS, UKSA and Government for full responses, as appropriate. While the ONS, Bank of England and HM Treasury were represented at that meeting, they have refrained from commenting on, or endorsing, the contents of the letter, to avoid pre-empting any formal responses to the report.
It will not be a surprise to you that there were many areas of disagreement, but there were also some strong areas of agreement, and in what follows I aim to set out the balance of the arguments as fairly as possible. Much of this will be familiar reading.
The great majority of members agreed that:
- The present situation with regard to the RPI was unsatisfactory, and the quality of the RPI should not be allowed to deteriorate. Some six years after RPI lost its National Statistics status, it remains in widespread use (for example to pay many pensioners and as the uprating index in many private contracts) and this means more than ‘routine maintenance’ is needed.
- In particular, the increase in the formula effect following the change to collection practices in clothing is obviously problematic, although this was not foreseen. The ONS should have persisted with the work on clothing following the 2012 consultation to identify clearly if the new collection methods were resulting in better statistics for clothing prices in RPI or indeed CPI (where there was a major distortion prior to 2010). The measurement of clothing inflation remains challenging as retail practices evolve. Nevertheless, although the past cannot be unwound, the impact of the change of collection prices on the RPI needs to be addressed as a matter of urgency, to restore the RPI’s quality to the pre-2010 position.
There was less agreement on some of the other comments in the Economic Affairs Committee Report. Taking the summary of conclusions and recommendations in the Report and using the Committee’s numbering:
1-3 There are very different views on the Panel about the Carli formula. Some Panel members would support a wider piece of work on the merits of different formulae in different sectors of the inflation indices. However a majority of members argued that the open conclusion of the Report on the Carli did not reflect the balance of expert opinion.
4-8 The Panel agree that the clothing issue needs to be addressed – though members appreciate this may not be an easy or quick ‘fix’ and do not have a single view about how this should be done. Our request that the Technical Panel should work with the ONS on this was reiterated. The Panel agreed with the Report’s comment that the requirement to have the good statistics was more important than any adverse impact on some users of those statistics.
9-11 The Panel agree that the Government should use the most appropriate inflation measure for each purpose that is required, avoid the impression of ‘inflation shopping’, and welcome the Government’s steps in this direction. A clear majority of the Panel did not support the suggestion of moving to one inflation index. There are legitimate different purposes – an uprating index may not be the same as an inflation index used by the Bank of England for inflation targeting. Others considered that the conceptual justification for different types of indices was unclear, and that the public found the existing range of measures confusing.
12-14 This section provoked most disagreement. A significant minority of the Panel believe that the RPI is a candidate to be a good measure of inflation, and that work should be done to enable it to regain National Statistics status. Equally a majority do not and would prefer to start from CPI or CPIH – recognising the considerable consultation on CPIH in recent years.
There was also a division of opinion on the measurement of housing costs – with considerable support for rental equivalence. Those supporting it point to the very considerable work carried on for the Consumer Prices Advisory Committee, and the subsequent ONS consultation including with the two Panels, from which majority support for rental equivalence emerged. The way in which housing prices and costs are included in the RPI is also problematic. Those opposing rental equivalence point in particular to the risk that it is not understood by the ‘person in the street’ and so might reduce public confidence in the inflation statistics. It has a significant weight in the CPIH, but is an imputed series.
15-16 These points were not discussed.
17-18 The Panel supported the issuance of CPI-linked gilts, for which there is clear demand. A Debt Management Office consultation on this would be welcomed, given some concern as there is not yet a settled view on the future of the main inflation indices. This consultation should also consider whether there would be a preference for a legal framework for the CPI analogous (but probably not identical) to that for the RPI. It has been argued that CPIH in particular can be developed to suit UK specific issues – but if this were to be used for index-linked gilts care would need to be taken about future changes to any aspect of its construction.
19-21 The Panel supports the suggestion that there needs to be a long-term process to resolve the present shortcomings and uncertainties about the future of inflation measurement, while (as above) not all supporting the single inflation rate goal. They would expect HMT and the Bank of England to give their support to a major programme of work to determine how to resolve the many issues discussed in the House of Lords report. Key private sector representatives would also be needed, not least to discuss the legal questions arising.
Finally on the Report, there was disappointment with the brevity of the discussion on the Household Cost Indices, presently under development, which are supported by the Stakeholder Panel as a welcome addition.
There are two general points the great majority of the Panel wanted to air in this letter:
- In discussing the methodological change to clothing price quotes, the House of Lords Report states: ‘Who benefits? Holders of RPI-linked Government bonds’ subsequently qualified to refer to pre-2010 holders. This may give rise to a misleading impression about who are the ultimate beneficiaries. It is indeed true that owners of gilts prior to the clothing change will have received interest payments above expectations (though this would not be true for those buying gilts after the impact on RPI was understood in the market, probably by 2012, when the higher expected RPI would have been reflected in the price paid for the gilt and therefore that received by the Government). However, over 50% of the UK index-linked gilt stock is held directly by pension funds seeking to hedge their RPI-linked liabilities. We understand that a considerable proportion of the remainder is likely to be held indirectly by them (as backing for swaps and liability-driven investment) and by insurers (either held directly or by counterparties to swaps) as backing for inflation-linked pensions in bulk annuity portfolios.This implies that among the ultimate beneficiaries of the change are pensioners with RPI-linked pensions. More than 50% of members of DB schemes have some form of RPI-linked pension (according to the Pension Regulator’s 2018 DB landscape). Among the losers are taxpayers and also the shareholders in companies with large historic pension schemes to the extent these were not hedged against inflation in 2010 (this is somewhat complicated by the movements in the inflation-linked gilts market around the ONS consultation and CPAC recommendations in 2012/13).It is also the case that RPI-related contracts are pervasive, giving rise to many other possible ‘winners and losers’. For example, it is widely used in the context of utility regulation, but there are also many private sector uses. The impact of any reforms to the construction of the RPI would affect rather more constituencies than ‘holders of RPI-linked Government bonds’. Indeed a full review of the users of RPI would be helpful.
- The 2007 Statistics and Registration Act requires the approval of the Chancellor for a change to the RPI if the Bank of England deems it to be a ‘fundamental change’ with a ‘materially detrimental’ effect on index-linked gilt holders. Although this now strictly applies to only three gilt issues, it is clearly important to maintain confidence in this market (though note that those holding the gilts as hedge against RPI-linked liabilities would not be affected so much). However, there is no such provision in the event of a change which increases the RPI relative to the CPI (which, as the inflation target, should pin down the long-term rate of inflation). This creates an upward ratchet in both the RPI and the wedge between the RPI and CPI. ONS must be always on the alert for any change to methodology that risks widening the wedge as this will produce unwelcome consequences, as we have seen since 2010.
I am copying this letter to Sir David Norgrove, and the Panel would expect it to be made available on the UKSA website as soon as possible.
Dame Kate Barker
Chairman, ONS Stakeholder Advisory Panel on Consumer Prices