Home > News > 2003 > January 2003 > Inflation contemplation
Inflation contemplation
Colin Mayger, a partner in Barnett Waddingham's London office, looks at the various inflation measures in use.
You will be used to hearing regularly on the news about the latest "headline" inflation figures or "underlying" inflation trends, but surprisingly few people know how these figures are actually arrived at. In this article I aim to demystify some of the process, and also to highlight some of the issues related to the measurement of inflation in the event that the United Kingdom were at some stage to adopt the Euro as its currency.
Headline Inflation (or RPI)
Properly known as the General Index of Retail Prices, this statistic has been produced since 1956, although it has existed in various forms since 1914. It is the index used for the purpose of defining the pension increases that must be granted by schemes that provide "Limited Price Indexation" ("LPI") on pensions in payment and also the increases provided by most statutory schemes (such as the Civil Service pension scheme). It is also used to calculate the coupon and capital payments from index-linked gilts.
It is based on a "basket of goods" chosen to represent the typical purchases made by consumers. Researchers physically go out to all parts of the country and check the prices at which the goods on the list are on display in shops (and nowadays on web-sites as well). This is time-consuming and cannot all be done on one day, so when reference is made to the "September" RPI, this is based not on prices on the first day of the month (or the last or the middle), but on prices collected throughout that month. The prices are compared to prices for the same goods in the previous month, and the index figure is adjusted appropriately.
Underlying inflation (also known as RPIX)
This is calculated in the same way as headline inflation, except that mortgage and other interest payments are "eXcluded" from the basket of goods. This tends to make it a much less volatile figure. The Bank of England's target inflation figure of 2.5% pa is defined in terms of RPIX not RPI. So when the Bank increases interest rates to try to reduce inflation, it will in the short term have the effect of increasing RPI but not RPIX!
The last few years have seen a general trend of falling interest rates which mean that RPI inflation has tended to be lower than RPIX - if interest rates rose, then that would be reversed. Because the two measures are different, it would be not be theoretically entirely correct for your actuary to use an inflation assumption of 2.5% pa for his or her actuarial valuation simply because this was the Bank of England's inflation target; it would not be comparing like with like.
Other measures
As you can see it is fairly easy given the base data to come up with other baskets of goods to represent different groups. For example a "pensioner" basket would presumably be very different from that of a "student"! In fact for some years a Pensioner RPI was published and some pension schemes adopted it as the measure they used when deciding on increases - over the long term however it seemed that the swings and roundabouts cancelled out and it did not differ from RPI enough to be worth the effort. A "Tax and Prices Index" was also published for a while to try to take account not just of prices in the shops but of the money consumers had to spend.
Problems
In recent years there has been a debate about failings in this sort of measure of inflation. I said that RPI has been measured since 1956. The typical basket of goods, however, is very different today from what it was then. For example the current basket includes mobile phones - these either did not exist then, or if they had, then they would have cost thousands of pounds! On the other hand a clothes mangle might have been a common item then, but if you wanted one at all now, you would most likely need to look in an antiques shop. So when you compare the RPI now with that in 1956, what is it really telling you?
In fact the basket of goods is revised every quarter to reflect changing patterns of consumption and in particular new technology. It is widely agreed in economic circles that this has the effect of overstating the rise in the cost of living, because old technology tends to drop markedly in price as it is overtaken by newer developments, so pensions that rise in line with RPI allow you not just to maintain the standard of living you had at retirement but also to improve your quality of life by continually upgrading to the latest model in the basket.
A second problem is that of "substitution". Say Double Gloucester cheese is in the basket and its price doubles while that of other cheeses remains unaffected. Consumers will not continue to buy Double Gloucester but will switch instead to, say Red Leicester. Has the cost of living gone up or not?
Harmonised Consumer Price Index
The National Statistical office has begun to publish a new index known as the Harmonised Consumer Price Index (HCPI) which uses a different method to try to overcome the two problems identified.
Currently HCPI inflation is running at 1.6% (11/02) compared to the RPI of 2.6% (11/02) but this is not just a short term difference which will even out in the long run. The calculation systematically comes up with lower figures - it is estimated that in the long run it will tend to run at about -1% pa compared to the RPI calculation.
The "Harmonised" part of the description refers to the fact that it has been adopted as a common method throughout Europe so that when inflation is measured in different countries like is compared with like. So what would happen if the UK were to join the Euro-zone?
Euro-inflation
Adopting the Euro in itself would not affect pensions in payment - in theory the amount you could buy with your monthly pension remain the same. What would pensions be increased in line with in the following years, though?
There would be nothing to stop the UK continuing to calculate and publish an old-style RPI figure based on prices measured in the UK. Limited Price Indexation could continue to be linked to this.
Index-linked gilts would be redenominated in Euros but would otherwise be unaffected - presumably the old-style RPI calculation would continue to be used as to move to the new HCPI would reduce the expected returns. But can we be sure that this would the case?
Presumably any new index-linked bonds would be linked to the HCPI, and presumably the prices measured would be those across the whole of the Eurozone. While individual country HCPIs would continue to be calculated, pension schemes are likely to be reluctant to promise pension increases unless they are able to invest in assets to match those liabilities - this will inevitably drive LPI to be redefined using the HCPI. As we have seen above, this would not be good news for pensioners, as it will systematically reduce the pension increases they would have received. Hopefully people will be able to see through any claim by politicians that joining the Euro has instantly reduced inflation, when all that has happened is that the measure has changed.
These are not mere technical issues that can be left to the bureaucrats to sort out later but are fundamentally important for the future of pension provision in this country.
Colin Mayger, January 2003.