No matter how we measure inflation, politics will forever trump economics

THE ECONOMIC Affairs Committee of the House of Lords has got its bovver boots on. Last week, the government was given a sound kicking.

The issue was the seemingly esoteric one of how to measure inflation. Inflation tells us how much the prices of goods and services are going up. The question is: what do we put into the basket when we are working this out?

The most general measure is the consumer price index (CPI). This takes into account literally everything which individuals in the UK buy. Something which is widely purchased, such as rail journeys, will carry more weight than, say, spending on parts for model railways. But they all count. The percentage change in the CPI is one measure of inflation.

Gathering all this information obviously takes time. In contrast, the retail price index (RPI) is quick and easy to calculate. It is, quite literally, based on a basket of products available in shops. The basket gets changed from time to time to reflect changes in spending patterns. The disadvantage of the RPI is that it is much more focused on goods than on services.

In recent years, inflation as measured by the RPI has been higher than the CPI. Between 2014 and 2018, the respective rises were 9.7 and 5.9 per cent.

These differences have important practical consequences. All sorts of things get increased each year by the “rate of inflation”.

The Lords accused the government of using the RPI for uprating stuff like rail fares and student loans, where directly or indirectly the government rakes in money. But it uses the CPI when it comes to paying out on pensions and benefits. “Index shopping” was their Lordships’ neat description of this practice.

But in top academic circles, much more fundamental attacks have been made on both these traditional metrics.

Measuring inflation faces a very difficult problem. How do you take into account changes in the quality of goods and services?

A simple example is a car. A particular model may cost exactly the same as the identical model last year. But suppose that, unlike last year’s model, this car has heated seats and parking sensors. The measured price has not changed, so inflation is zero. But you are getting more for your money.

The problem becomes acute in any area of new technology. Smart phones did not exist 30 years ago, and the internet was not yet developed for general use. How much have their prices changed since then? We have only to ask the question to see the problem that the vast advances in technology pose.

Even back in 2003, the top MIT econometrician Jerry Hausman estimated that the CPI was systematically overstating inflation by as much as two per cent each year, because of this quality issue.

Measured correctly, inflation could well have been negative in the current decade. But it will be hard to get politicians to take an interest in this. Imagine having to tell people that their pensions would be reduced because prices were falling.

Even if we could improve the measurement inflation, as the Lords demand, politics is forever likely to trump science here.