The Coalition has punished the poor by linking benefits to CPI inflation
There are two key elements to calculating a rise in the cost of living. First, we need to get a sense of what people spend their money on – what is bought when “living”. Second, we need to decide how to bundle up the costs of individual items people spend their money on into an overall index of costs – an inflation index.
The standard measure of inflation until recently was the retail prices index (RPI). That used surveys to work out the basket of goods that people spent their money on, and then took a weighted arithmetic average of changes in the prices of those goods to construct changes in the cost of living (inflation). When we calculate average annual inflation in this way, what we find out is how much more money it takes this year to buy the same goods we were buying last year.
Another inflation measure has become popular recently – the consumer prices index (CPI). Unlike RPI, CPI is not a measure of changes in the cost of living, because large elements of the cost of living aren’t included in the CPI basket.
For example, housing costs are an important cost for households, and under normal circumstances comprise something close to one quarter of RPI, but housing costs are not included in the CPI. CPI is not a cost of living index; it is, instead, what is called a “policy index” – that is to say, it is a measure of inflation that is used to guide interest-rate setting.
Before 2003 we used to use a different policy index, the RPIX, which, again, was not a cost-of-living index (it excluded mortgage interest payments, though it did include other housing costs).
But the fact that CPI does not include large elements of the cost of living is not its only difference from RPI. CPI also bundles up the costs of individual items into the overall inflation number in a different way. Specifically, it uses “geometric” averaging. I shan’t explain the precise mathematical difference between arithmetic and geometric averaging, but the consequence is clear and intuitive.
Whereas the RPI methodology proceeds on the basis that people buy the same goods as the previous year, in the CPI methodology they are assumed to spend the same amount of money on goods as the price of them rises (ie. the amount of the goods falls in exact proportion to the price, so that the amount of money spent stays invariant).
Now, in the Emergency Budget, one of the key ways in which the Government saved money on benefits spending was by changing the way in which benefits “uprating” takes account of inflation. Whereas previously it was calculated by the RPI (and related indices), in the future it will be calculated by the CPI.
The Government argues that there are two reasons for this change: first, that the calculation method (the geometric averaging) makes it a better measure of inflation; and second that the coverage better reflects the spending of those receiving benefits. The Institute for Fiscal Studies has questioned the latter of these claims, but agrees with the first. I dispute the first as well.
I agree that the assumption that people spend the same amount is better for most people than the assumption that they buy the same amount of goods. But that is not correct of benefits recipients. Those on minimum levels of benefits are being provided with an amount of money that is just sufficient for them to get by.
It is thus intrinsic to the nature of these benefits that recipients cannot respond to price rises by buying less. The consequence of the new methodology is that, over time, those on minimum levels of benefits will be able to consume less and less – benefits will be cut at the bottom.
On the other hand, benefits taper away as incomes rise, and those towards the top of benefits thresholds are much more like the average person – they are in a better position to respond to price rises by cutting consumption.
Consequently, in my view the technically correct way to calculate uprating of these benefits would be to have the level of the benefits at the bottom rise with RPI, but to have the top of tapers rise with CPI. Then those on minimum levels of benefits would not experience cuts in their levels of consumption but those higher up the scale will be able to substitute between goods in the standard way.
Looking ahead, it would be desirable to expand the CPI basket of goods so as to cover all of those covered in RPI except mortgage interest payments – making CPI closer to a proper cost-of-living index. The only reason CPI does not include housing costs is that it is an EU-wide index.
Incorporating housing costs (except mortgage interest) into CPI could be done immediately, would improve interest rate decision-making, and has been urged by commentators from myself to Mervyn King since 2003. We should get on with it.
Andrew Lilico is the chief economist of Policy Exchange
Similar Posts:
- Pay cut Britain: how inflation outstrips wages
- Inflation could go higher if workers demand pay increases, Mervyn King warns
- US consumer inflation gains most in 15 months as food and petrol prices increase
- Bank of England turns to Google to track consumer trends
- Spencer Dale warns of ‘aggressive’ rise in rates if Bank of England loses credibility on inflation
