Today we look into the differences between the Consumer Prices Index (CPI) and the Retail Prices Index (RPI). This follows hard on the heels of yesterday’s entry about the basket of goods and services used by statisticians in studying UK consumers’ household spending. This offers an excellent opportunity to analyse these two important indices which feature in monetary economic news stories.
Both the CPI and RPI use the same idea of measuring inflation by tracking the price changes of around 650 different goods and services on a month-by-month basis. Around 180,000 observations are made of the price of these items. The goods and services that are tracked are chosen annually, with changes in the composition of the typical ‘basket’ of items having been announced recently. Where the two indices differ is in the coverage of the data, the population base used, and the way the indices are calculated.
Let’s start with the RPI. This index used to be the main measure of inflation in the UK. But RPI wasn’t originally intended to serve this purpose; it was created at the time of the First World War in order to protect workers against price rises associated with the war. RPI was a compensation index. This function was maintained in the years to come, with RPI being used as a basis on which many UK pay negotiations were based.
The CPI is a more recent creation, developed in the 1990s in order to enable international comparisons to be made of countries’ inflation rates. CPI is now the UK government’s preferred measure of inflation. It is used to monitor progress in achieving a target rate of price change. The target is for prices of the typical basket of goods and services to stay within the range of + or – 2%. This forms part of the monetary policy framework of the Bank of England’s Monetary Policy Committee.
So what are the differences between RPI and CPI?
Well, in terms of coverage, RPI includes the following costs that CPI doesn’t: mortgage interest payments, council tax, house depreciation, house purchase costs, buildings insurance, TV licence and road fund licence (tax disc). It also measures car costs by using the price of used cars. CPI covers the following costs that RPI doesn’t: stockbroker fees, university accommodation fees, foreign student tuition fees and unit trust fees. It measures car costs by using the published price of new cars.
In terms of the population base used, RPI includes most private UK households, but excludes the highest earners and those on pensions dependent on state benefits. It includes spending within the UK and by individuals when they are abroad. The weights used to determine the importance of different items are based on the Office for National Statistics’ Living Costs and Food Survey. In contrast, the CPI includes all private UK households, including spending by institutions such as care homes. It only includes spending within the UK. The weights used are based on national accounts data.
The final big difference between RPI and CPI lies in the way the indices are calculated. RPI uses an arithmetic mean, whereas CPI uses a geometric mean. In practical terms they can end up with very different results from using the same data. If one price increases by 30% and another falls by 10%, then if we take the base starting point to be 100, then the first price is now 130 and the second is now 90. Using the arithmetic mean produces the following calculation:
130 + 90 = 220
220 / 2 = 110
So by RPI, the price level has increased by 10%. But for CPI, the geometric mean is used which produces the following result:
130 * 90 = 11700
Square root of 11700 = 108.17
So by CPI, the price level has increased by only 8.2%. This, of course, has all sorts of implications for people in the way their benefits are calculated, the method used to work out wage increases or changes in the national minimum wage, and the way the official rate of inflation differs from their own experience of price changes. These aspects will be the final topic of this tour of inflation measurement which I will post on this blog tomorrow.