cpi rpi and pensions
What a drag – clothing prices have depressed pensions – or have they?
by Dr. Ros Altmann
(All material on this page is subject to copyright and must not be reproduced without the author’s permission.)
What has happened?
The Bank of England’s latest Inflation Report has been unwittingly dragged into the debate about inflation-linking of UK pensions. On Page 39 of its latest Report, the Bank of England explains that the data it used for consumer price index (cpi) inflation from 1997 to 2009 were incorrect, because the ONS had not measured clothing and footwear prices correctly. That problem has now been solved, but the impact of the past inaccuracies mean cpi inflation was under-reported by 0.3% a year. Therefore, if the prices had been measured more properly, inflation each year would have been 0.3% higher than actually reported.
What does this mean?
The Bank of England also states that the impact on the retail prices index (rpi) would have been even greater., but on the basis of the cpi changes alone, inflation over the period from 1997 to 2009 was under-reported by 4%. From 2010 onwards, the new method of calculation is being used, and this will increase the differential between cpi and rpi, since clothing and footwear price inflation will be higher than previously expected (unless of course clothing and footwear prices start to fall!)
What is the fuss about pensions?
UK pensions are often linked to increases in the rpi measure of inflation. All state pensions and many final salary pensions are increased each year in line with the official rpi measure. If inflation has been underestimated, this therefore means that pension increases would have been higher if the clothing and footwear price rises had been correctly recorded. In fact, if official inflation measures each year from 1997 to 2009 had been 0.3% higher, UK pensions would have been 4% higher. And as rpi is even more understated than cpi, the pension impact is even larger than 4%.
Does this matter?
Well, yes and no. The theoretical argument suggests that pensioners in the UK (and deferred members of UK pension schemes) have lower pensions than they should have because of this mistake. But in reality, this line of reasoning is a little too crude. Inflation indices are made up of dozens of goods, each of which measures hundreds or thousands of prices. Mistakes are often noted and over the years, revisions to indices and calculation methods occur. Some prices will be over-recorded and some will be under-recorded, so it is not clear what the net effect actually will be. To take the case of just one price measure, which has been consistently too low and conclude that ‘real’ inflation is higher than reported is not really fair. One would have to be sure than no other components of the indices had been measured incorrectly during that time and no other errors in the opposite direction had occurred.
In addition, pensions are only supposed to be increased in line with the official measure of rpi. Therefore, the fact that the figures may be revised after the fact is not relevant. Nobody is likely to get extra pension as a result of this at all. Equally, if the inflation rate had been over-estimated, people would not be expected to have to pay back the extra they had received because inflation had been measured too high!
What does this all highlight?
Even tiny annual changes in prices can have a huge impact over time: Firstly, that even tiny changes in inflation each year can have a seriously damaging effect on purchasing power over time. In just over ten years, a one third of one per cent change each year means losing 4% of the value of your money!
Inflation-linking of pensions is expensive: Secondly, the inflation-linking of UK pensions is very expensive and very complex, but does make a significant difference to pensioner incomes.
Impact of change in pension uprating from rpi to cpi even larger than expected: Thirdly, if in future the change in measurement of clothing and footwear in our inflation indices means a permanent upward shift in inflation, then the differential between cpi and rpi will be even larger than has been previously estimated. That could mean that the impact of tying pensions to cpi instead of rpi in future will result in even lower pensions in future than has been assumed. Will the DWP have to yet again revise its impact assessment of the effect of this change?!