Toggle high contrast

TUC response to the joint HM Treasury and UK Statistics Authority Consultation on the reform to retail prices index methodology

7 August 2020
Report type
Consultation response
Issue date

The Trades Union Congress (TUC) exists to make the working world a better place for everyone. We bring together around 5.5 million working people who make up our 48 member unions. We support unions to grow and thrive, and we stand up for everyone who works for a living.

TUC officials have been involved since the beginning of the public debate around inflation, and once more take this opportunity to make our views known.  We have limited our response to question 6.

  1. Are there any other issues relevant to the proposal the Authority is minded to make of which the Authority or the Chancellor ought to be aware?

If the Authority chooses to proceed on the present course, it is choosing a measure of inflation that understates the changes in the cost of living facing the public. The Royal Statistical Society (RSS) state that CPI and CPIH “are an unsatisfactory measure of inflation as it affects British households”.[1] With compromises available, it is an act of vandalism to end the RPI measure with its long history and public confidence.

And UKSA will do further damage to the credibility of inflation measurement and official statistics more generally. The Hours of Lords Economic Affairs Committee has warned of a “dereliction” of public duty, words not used lightly. In their submission to the present enquiry the RSS warn the decision "has potential" to be "damaging.

The process throughout has been characterised by seeking and then disregarding the opinion of stakeholders and experts alike. A prolonged series of consultations,  public debates and the convening of two new Advisory Panels has not budged the Authority from the way forward articulated substantially by Paul Johnson of the Institute for Fiscal Studies in his ‘UK Consumer Price Statistics: A Review’. 

The TUC and trade unions have contributed to this process at every stage, rejecting the official approach and calling for a positive initiative to renew the RPI.   

This latest stage in the process was set in motion by the consultation and subsequent report of the House of Lords Committee on Economic Affairs.  We were greatly encouraged by the report that broadly agreed with critics and rejected the official line.

But entirely in keeping with the process to date, the substance of the report has been entirely disregarded by the Authority. Moreover the present consultation has been framed so that the existence of the Lords Committee Report is barely acknowledged.  Even though we have become accustomed to official machinations, the manner in which the conclusions of this highly distinguished committee has been handled is breathtaking.

The Lords Committee vindicated Trade Unions’ argument that the RPI is not fundamentally flawed: “12. We disagree with the UK Statistics Authority that RPI does not have the potential to become a good measure of inflation. With the improvements to RPI that we set out in the previous chapter … we believe RPI would be a viable candidate for the single general measure of inflation”.

Most importantly the Lords opened the way to a compromise on the formula effect. Fundamentally, they contested the official line that the Carli formula is fundamentally flawed, instead recognising that “expert opinion … differs” [recommendation 1]. They however find consensus that clothing prices should be repaired. They not only argue that this should happen, but moreover the relevant legislation “requires the Authority to attempt to fix the issue with clothing prices” (p.3,para 7, my emphasis). 

We have set out our argument on the formula effect most recently in our submission to the House of Lords, and our position has been underpinned by work carried out by Dr Mark Courtney. Operating as a private commentator (but with a background in the Government Economics Service)  Dr Courtney has recently compiled an overview of the opinions of statistical experts on the formula debate, showing on balance Carli regarded as less problematic than Jevons in practice (here). At the Royal Statistical Society debate of 21st July 2020 leading UK statistical experts were likewise unwilling to endorse Jevons.

In their response on 19 February 2019[2] the Advisory Panel on Consumer Prices (APCP) supported the key proposal of the Lords Committee: “the panel agree that the clothing issue needs to be addressed”.  However APCP were divided on most other issues in the Lords Committee Report.

The Authority rejected the panel’s advice, choosing instead the present nuclear course of disabling the RPI. Yet they choose to begin and motivate their statement of 4 September 2019 as follows: “The Advisory Panel on Consumer Prices provided advice to the National Statistician on the composition of the Retail Prices Index (RPI) in light of the House of Lords Economic Affairs Committee report Measuring Inflation, published in January 2019”.

Beyond this misrepresentation, the UKSA and ONS explanation for rejecting this advice is nonsensical, rejecting a compromise on the tautological grounds that it is not a full solution. UKSA also claim that resolving the clothing issue would still leave a number (five) of unresolved shortcomings. But as Jill Leyland said at the RSS debate, the “only one real flaw” with RPI is the use of Carli on clothing (though previously the problem was the use of Jevons for clothing in the CPI). The other issues are not complex and could be easily remedied.

Everybody knows that there is no ideal formula. As Leyland indicates, the issue with the RPI is not the formula but the way the price collection for clothing has interacted first with the Jevons formula to understate inflation and then with the Carli formula to overstate inflation. The ONS ‘fix’ to clothing prices in 2009 simply prioritised CPI over RPI, to comply quickly with European obligations. Finally from 2009 the RPI was cast aside, with officials and their allies tirelessly seeking its termination.

These European obligations are met through the Jevons formula, with the use of the Carli ruled out under EU legislation. But it should be emphasised that Carli was ruled out on grounds of comparability not wrongness. When the ONS appeal to ‘best practice’, they are appealing to the widespread use of Jevons in part because of this legislation. This is common practice, not best practice.

It is also possible that vested interest has played a role, and there has been throughout an interplay with public policy. The Boskin report of 1996 led to a number of recommendations that overall reduced the level of measured inflation, limiting the cost to government ahead of the introduction of Treasury inflation-protected securities in the US.  The recent debate in the UK coincided with austerity policies, and the government gaming uprating procedures so that over the past decade the public are paying RPI and receiving CPI. The present consultation suggests some weighing up of interests between pension funds (who win overall) and old-age pensioners (who lose).  The minutes of the January advisory panel report

“Survey data suggests approximately 20% of pension schemes may lose money as a result of the potential change (i.e. schemes with mainly CPI pension increases). Conversely, there are early indications of support for the changes from some schemes that are expecting to gain from the proposed change (i.e. schemes with mainly RPI pension increases), albeit recognising that their members would receive lower pension increases”.

A report sponsored by the BT Pension Scheme found that the change could reduce the value of RPI index-linked pension benefits by nearly 10 per cent. It also said that around two thirds of defined benefit pension schemes uprate pensions each year in line with RPI.[3] With over 5 million pensioner members and 1.1 million active members of private sector DB schemes in the UK, these changes will reduce income in retirement for millions of individuals, leading to higher levels of pensioner poverty and hardship.[4]

Notably the UKSA consultation is not weighing up the likely cost to workers if they are undercompensated for increases in the cost of living.

Finally, while the Lords Committee proposed a compromise solution, the Authority has been closed to fuller solutions.

Specifically on the formula effect even the Lords Committee feared “a perpetual debate” [3.], but this should not be the case. Those who oppose the RPI seek refuge in old theoretical arguments (paras 80-86) rather than engage with the actual evidence of using different formulae over the past 20 years. The RSS recognise that the right way forward is to consider the relevant formula on an item-by-item basis, which was the approach ONS were beginning to adopt ahead of recent controversies. This process could be led by the RSS, who have offered valuable advice throughout the process and as an institution have emerged with their credibility and reputation enhanced.

The wider developments associated with the RSS’s preferred household cost index (HCI) – which is designed explicitly as an uprating rather than macroeconomic measure – might provide an appropriate template (including, for example, more appropriate measures for owner occupiers’ housing costs and student loan repayments). However trade unions would prefer an upgrading of RPI rather than the introduction of a new measure of inflation.

[1] Letter to The Times, 21 March 2017.


[3] ‘Switch away from “flawed” RPI inflation measure in 2025 would cost BT pension scheme £1.7bn – and thousands of its workers nearly 10 per cent of their retirement income’, This is Money, 13 April 2020:; and ‘How could changes to price indices affect DB schemes?’ Pensions Policy Institute, April 2020:

[4] Occupational pension schemes survey, UK: 2018, Office for National Statistics, 20 June 2019:



Enable Two-Factor Authentication

To access the admin area, you will need to setup two-factor authentication (TFA).

Setup now