Investors in index-linked government bonds were spared losses from future changes to the calculation of inflation on Thursday after a new review recommended leaving the “flawed” retail prices index unchanged, at the same time as urging government to stop using it.

The independent review by the Institute for Fiscal Studies confirmed that the RPI overstates inflation through the use of an out-of-date mathematical formula, and this error costs taxpayers £2bn a year at a time of severe pressure on public services.

Paul Johnson, director of the IFS, who led the review, concluded “it was difficult to unpick” a decision taken by the last national statistician two years ago not to update the RPI in future regardless of its how much it overstates true price rises.

Inflation measured by the RPI tends to be around 1 percentage point higher than when measured by the consumer price index (CPI), with an official expectation the gap will increase to 1.6 percentage points by the end of the decade.

In November, the CPI stood at 1 per cent while the RPI was 2 per cent.

Instead of bringing the RPI up to date, Mr Johnson urged all public sector bodies to stop using the index “as soon as practicable” in setting pay, elements of the tax system and regulated prices such as rail fares. He also suggested the Treasury and the Debt Management Office should use an index based on the CPI for the issuance of new government bonds.

The review recommended that the UK Statistics Authority should “do more to speak out” if government departments continued ignoring this advice from its own statisticians.

Immediately after the review’s publication, however, the DMO gave a non-committal response. It would keep the matter “under review”, it said, “[but] would need to be satisfied that any demand for gilts linked to another measure such as CPI or CPIH would be sufficiently strong and sustainable”.

Commentators on the £470bn index-linked gilt market were happy that the unexpectedly high RPI-linked returns would not be diminished by changing the calculation of the index.

Sam Hill of RBC Capital Markets said the review was “a clear acknowledgment that existing RPI-linked gilts shouldn’t be altered”.

Roger Beale spinoff cartoon

Michael Saunders, economist at Citi, said it was a “well-known and longstanding” issue that RPI was an inferior index to CPI, but that any decision to change the basis on which gilts were sold was a matter for the chancellor rather than the statistics authority and unlikely to happen quickly.

While the review reinforced the taxpayer subsidy to index-linked bond holders, in other areas, Mr Johnson was keen to see the RPI gradually phased out. “It is time for the UK government to take the next, logical step and stop using RPI in any element of the tax, benefit and regulatory systems,” he said.

The consequences of this are uncertain. Social security benefits would be unaffected as they were linked to the CPI in 2010 and increases have since been limited to a flat rate of 1 per cent, regardless of inflation. The effect on rail fares and other prices linked to RPI will depend on what the authorities decide to replace RPI with.

But Mr Johnson said there was some urgency in the change to stop the proliferation of different measures of inflation and potential “inflation shopping”, by which the government increases its income from duties linked to RPI but lowers its expenditure by linking benefits to the CPI.

Mr Johnson’s review said the CPI measure, which the Bank of England uses in monetary policy, was “a well constructed measure of inflation” but the ONS should move towards making the headline measure of inflation the CPIH, which includes the costs of owner-occupied housing.

The Johnson review rejected calls for a “household index” measure of inflation, which attempted to measure the out-of-pocket costs of changing insurance premiums and house prices, on the grounds that any measure would not be relevant to most people.

Instead, it urged the ONS to publish an annual report on the rates of inflation faced by different groups. In recent years lower income households have been hit by sharper rises in the cost of living because they spend more of their incomes on food and energy, which until recently had risen rapidly in price.

The UK Statistics Authority said it would respond to the review later this year.

Q&A: What price victory in battle of the indices?

The Johnson review of inflation measures said government should link to a version of the consumer price index and nor the old retail price index.

Why should I care?

Because it matters for your pocket. At present, the RPI suggests there is 2 per cent inflation, while the CPI inflation rate is 1 per cent. You will care which measure is used if your income or any prices that affect you are linked to an inflation index.

Which is better?

The CPI is better. At its core, the RPI contains a deeply flawed formula called “Carli” which is not used in any other advanced economy. One effect of the Carli formula is that if prices fall and then rise again to the original level, it will show a positive rate of inflation. The upward bias in the RPI is estimated to be around 0.8 percentage points a year and it rose very sharply in 2011 when the method of collecting clothing prices was changed.

Why didn’t government statisticians simply correct the flaws in the RPI?

A good question, to which there is no satisfactory answer. Mr Johnson talked of potential political and legal questions, but the strength of this objection has never been tested. The upshot is that ordinary taxpayers pay about £2bn more of interest every year on index-linked government bonds than would be the case if the formula had been corrected.

What is the legal case?

Index-linked gilts issued before 2002 (only four issues) contained a clause that allowed investors to redeem their bonds if a materially detrimental change were made to the inflation measure. With the bonds all trading above their par value, few would want to redeem, but huge opposition from index-linked bond holders persuaded the Office for National Statistics not to test whether the legal threats were credible. This let off George Osborne, chancellor, from making a decision on the issue.

Will the RPI now die?

Not likely, if the past two years are any guide. With £470bn in RPI-linked gilts already in issuance, people will still care about the RPI regardless of whether statisticians think it deserves to die. Inertia will be reinforced by the UK Statistics Authority’s decision that it was OK not to improve the RPI and to keep publishing it knowing it was flawed.

If government ditches the RPI for regulated prices, who gains?

This is not clear. Regulated rail fares, for example, are linked to RPI and normally do not rise by more than RPI plus 1 per cent. If the measure moved to CPI, there is nothing to suggest the new cap would not be CPI plus 2 per cent, leaving commuters no better off.

Get alerts on UK inflation when a new story is published

Copyright The Financial Times Limited 2019. All rights reserved.
Reuse this content (opens in new window)

Follow the topics in this article