The RPI scandal has been a slow motion car crash since 2010. The nation’s most well-used and well-known statistic has been subjected to silly mistakes in production, weak and indecisive management, too much political influence and misuse, an overbearing economist mindset, more dogma than imagination, limited innovation, and mixed messages from those supposedly in control. This sad story with resulting confusion for users reflects a fundamental failure of governance. This blog provides a one-stop shop for anyone looking to understand the governance, or lack of it. The next chapter in this saga will come later this week with the publication of the House of Lords report into the RPI.
ONS produces the monthly inflation numbers. The main prices landing page is here. Scrolling down the page leads to its controversial and very one-sided “shortcomings” paper from March 2018, “Shortcomings of the Retail Prices Index as a measure of inflation“. Methodology papers can be found here.
RPI governance has a long history starting with the RPI Advisory Committee, which was in place from 1947 to the mid-1990s. It seemed to work OK for the time, even if it would be seen to be lacking in these days of greater scrutiny. The rot began to set in around the time Gordon Brown became Chancellor as he chose never to convene the RPIAC group during his decade in No 11 from 1997 to 2007. This lack of transparency was despite the commitment in the 1997 Labour manifesto to “an independent National Statistical Service”, progress on which was slow. When proposals for “National Statistics” were published in 2000, there was a key get out clause, or “special arrangements”, for the RPI as the Treasury wished to “maintain control over the scope and definition of the (RPI) index”. In the name of independence, the Treasury did the opposite and took control.
After Brown became Prime Minister, there was the financial crash and all eyes were elsewhere. That meant that from the mid-90s until the early 2010s when the clothing change blew open the whole topic, there was no publicly accountable governance. It just so happened that during that period the ONS was remarkably silent on issues relating to inflation. Very little in the way of methodology papers was produced. Little is known about what actually was going on but silence is odd when changes to the statistics continued to be made. There is a strong suspicion that there was increased politicisation of the numbers.
The “problem”, such as it was developing, was there to be seen and was written up by the Statistics Commission in its Report No. 20 “Changes in the Calculation of the RPI and RPI Governance” (2004). Had just some of the recommendations (set out in the three page summary) been adopted by the ONS, we would certainly not be where we are now. The Commission recommended that: the RPI research programme should publish regular progress reports, the responsibilities of the Chancellor and National Statistician in respect of the RPI should be clearly set down, the role of the advisory committee must be clearly explained, and the National Statistician should be responsible for the scope and definition of the RPI, along with methodology. That report was, it appears, ignored by the ONS.
The next landmark was the arrival of the “Statistics and Registration Service Act 2007“. The RPI is unique as a statistic in the UK as the ONS has, according to the Act, to produce the numbers. Section 21 says:
“Retail prices index
(1) The Board must under section 20 — (a) compile and maintain the retail prices index, and (b) publish it every month.”
It is parts 2 and 3 of Section 21 of the 2007 Act go some way to explain the complexities of governance and why the tentacles of the Treasury and Bank now seem to dominate the RPI. They say:
“(2) Before making any change to the coverage or the basic calculation of the retail prices index, the Board must consult the Bank of England as to whether the change constitutes a fundamental change in the index which would be materially detrimental to the interests of the holders of relevant index-linked gilt-edged securities.
(3) If the Bank of England considers that the change constitutes a fundamental change in the index which would be materially detrimental to the interests of the holders of relevant index-linked gilt-edged securities, the Board may not make the change without the consent of the Chancellor of the Exchequer.”
Eventually it was conceded that some more visible governance was needed and the Consumer Prices Advisory Committee (CPAC) was in place from 2009. Sadly, it was too little and too late, and certainly too ineffective to stop the huge errors in calculation and weakness in management that unfolded from 2010. CPAC was secretive (minutes of meetings had to be forced out under FoI) and it was disbanded in 2013. The 15 years of near silence and non-governance had taken its toll on the numbers.
The new governance post-CPAC has not, taken as a whole, been for the better. The strong suspicion is that most decisions are now made by the “cross-government Inflation Tetrapartite Group” which will “meet quarterly to discuss the inflation statistics produced or managed by Office for National Statistics.” The terms of reference set out the function as to “…… provide a forum for discussing how well ONS’s range of price inflation statistics is meeting user needs, for sharing progress on current developments and for identifying areas for future development. When required, the group will consider routine, operational requirements (such as annual basket reviews and VAT changes) that require advice in a timely manner to ensure that the quality of price statistics is maintained.”
It is a highly secretive body – for reasons that have not been explained by UKSA or ONS. I requested its papers under Freedom of Information and eventually got a reply in December 2017. Large chunks of the documents given were redacted and it is unclear what was not released at all. For an example of redaction, see below page two of the December 2014 minutes.
The Smith Review of governance was published in 2014. It said that the inflation numbers ought to be governed by the Advisory Panels for Consumer Price Statistics, details of which appear on the UKSA website. Following the review, UKSA stated: “In June 2015, the UK Statistics Authority announced the formation of two independent advisory panels on consumer price statistics: a technical panel to advise the National Statistician on technical aspects of the statistics; and a stakeholder panel to provide advice on the uses and applications of price indices.” Details about both, including minutes of meetings, can be found in the left hand tool bar of this page. Those committees have been dominated by officials from the Bank and government departments and, in any case, have been trumped by the secretive Tetrapartite Group.
Beyond any such committees, but bearing in mind the importance of the topic, UKSA (in the form of the OSR) might be expected to be active in its RPI-related research. Rather amazingly given the focus on the RPI (and its relationship with the CPI), it seems (searching for RPI on the OSR web page) that the assessment part of UKSA has produced very little on the RPI. There was the report in March 2013 that stripped the RPI of its National Statistics status though it was hardly a thorough piece of work by OSR standards.
The UK Statistics Authority commissioned the Johnson Report, an “independent review”, in May 2013, “in order to ensure that these UK consumer price statistics meet current and future user needs.” It was published in January 2015. Following that, a consultation was launched in June 2015, closing in September 2015. John Pullinger, the National Statistician, wrote to the UK Statistics Authority Chair on 9 March 2016, setting out his initial views following the consultation. Subsequently, on 10 November 2016, he made his formal response to the consultation. That was a full six years after the gap between RPI and CPI began to widen.
Use of the RPI and the BT case
In the face of all this confusion, some bodies that use the RPI for uprating purposes have taken advantage of the wider gap between RPI and CPI since 2011 to switch from the former to the latter to reduce their costs. As explained in this RSS letter, the Government has engaged in so-called “arbitrage” by uprating money raising measures using the (higher) RPI and uprating expenditure using the (lower) CPI.
The same has happened in the private sector most notably in the case of pensions. BT wanted to shift those of its pensioners in one of the company pension schemes from getting their annual upratings based on RPI to CPI. This would save the company many tens of £ms every year, and deprive pensioners of the same amount of income. I was the expert witness in the case (for the pensioners) and the judge ruled (in January 2018) in favour of the pensioners. The story is told in a blog I wrote “RPI: Still fit for purpose“. In essence, the judge deemed that the RPI had not “become inappropriate” (the phrase used in the pension documents) and that BT had no grounds for moving a group of their pensioners to the CPI. The RPI is, therefore, still fit for purpose.
In reaching that conclusion, the judge considered two elements to be important. First, that the factors which “underlie all of the matters relied on by BT” were present, and known to be present, in RPI in 2002 (when the rules were drafted, albeit that the formula effect has worsened and the perception of those flaws has hardened). Second, the purpose of the pension is to provide protection for pensioners against increases in “the real cost of living to which they are likely to be subjected” which is better represented by the RPI.
On June 13 2018, the Royal Statistical Society hosted an important and well-attended meeting on ‘The Future of the RPI’, bringing together a range of speakers, including the UK National Statistician John Pullinger, representing a broad spectrum of opinions. There is a meeting report (including a video of the event), a page setting out the papers presented and a transcript of the event.
The extent of the problems caused by the statistical mismanagement in 2010 was recognised by the establishment of the RPI CPI User Group (RCUG) at the end of 2011. It joined many other statistical groups recognised by the ONS and RSS covering a wide range of UK statistics such as business, crime and health. There is an active message board from members of the groups on Sun.
Given the specification in legislation that RPI needs to be produced (see above), it is strange that some in UKSA/ONS want the RPI to be abolished or removed from production. Notably, Sir David Norgrove said (in June 2018, as reported in the FT) that he’d “abolish” it if he could.
It is just as strange – and a source of confusion – that that view is in contrast to the reported views of the National Statistician who indicated that the RPI could be reformed. (Also quoted in the FT, in July 2018.)
House of Lords
The HoL Economic Affairs committee has conducted an inquiry into “The use of the retail price index”. It took evidence (six recorded sessions) between June and October 2018, and is due to report on Thursday.