Wednesday, 18 July 2012

Does the headline rate of inflation misrepresent the impact of inflation in the economy? Taking into account past inflation as a predictor of economic behaviour

As the latest monthly inflation figures were released today, showing that the retail price index (RPI) had dropped from a high of 5.6 per cent in September 2011 to 2.8 per cent in June 2012 this got me thinking about the interplay between past inflation rates and the current rate of inflation. But first, let me put this into a little bit of context: RPI did not drop below 4 per cent for 22 consecutive months between March 2010 and January 2012, and spent the entire of 2011 barring December above 5 per cent. Why is this relevant? Well primarily because inflation has been outstripping average earnings growth for the past two years. This has meant that real wages have actually fallen significantly across the economy (by more than 10 per cent since the credit crunch). This has had a knock on effect on consumer spending, and is one of the important reasons aside from the most important (this man and this man), why economic growth has been so low recently.

(And by the way, on the topic of inflation, when this man increased VAT from 15 per cent to 20 per cent, 0.5 per cent was added to the rate of inflation and is the reason why RPI was above 5 per cent for so long, not to mention the reduction in consumer demand associated with it. How’s that for stimulating a demand-led economy recovery?) 

Back to matters at hand: my thought was, that if the rate of inflation this year is 2.8 per cent, but the rate of inflation the previous year was 5.6 per cent, the current rate of inflation is likely to have a very different feel than if the previous year’s inflation rate was only 2.4 per cent for instance. The price of goods is going to feel expensive if recent inflation has been high, even if current inflation is low. And this feeling is going to impact on the economic choices that people make. Why is this the case? Very simply, because individuals do not base their behaviour on arbitrary statistics, but rather, on their perception of factors underlying those statistics. For instance, the cost of gas will still seem expensive, if it only increases by 2 per cent this year, if in recent years there have been high price rises. 

This disconnect between the headline rate of inflation and it’s actual impact makes sense – the annual inflation figure was drawn up by economists and statisticians to give a snapshot of what is happening to the cost of goods across the economy at any one particular time. The problem is that this does not reflect how inflation impacts on the real world ie. how it impacts on the economic decisions of individuals and organisations who may also factor in past inflation rates into their behaviour. And this is important – if you want to up-rate state benefits such as the state pension, then the headline annual inflation figure will be appropriate (providing that it is up-rated consistently every year). However, if the aim is to explain and predict behaviour, which is surely one of the major purposes of economics, then it is important to use a measure of inflation which reflects its direct impact on people and organisations’ decision making.   

To draw a more personal example I judge whether or not a packet of crisps is expensive, based on what it cost me to buy a packet with my pocket money as a 12 year old in 1993 – about 18 pence – hardly surprising therefore that I eat less crisps nowadays that the price is about 55 pence, even though my spending power has increased exponentially since then. Stupid examples aside, to properly understand the impact of how inflation impacts behaviour therefore, one has to look at inflation over a longer period than just the previous 12 months. 

How might this be possible? Very simply, by factoring in previous year’s inflation into the calculation of inflation. To illustrate this I’ve put together the table below. This shows RPI going back to June 2010 as well as the compounded inflation since that date. As can be seen, in the background to the current RPI inflation rate of 2.8 per cent, there has been a 13.3 percent increase in the cost of goods since June 2009, which will be influencing the behaviour of consumers across the economy. 

To take this into account I’ve produced a weighted RPI figure that takes into account previous years’ inflation. It is called weighted RPI because I’ve weighted the previous years’ effect on the current year’s RPI, so that the current headline figure is given a 100 per cent weight, the previous year’s figure a 50 per cent weight and two year’s ago 20 per cent in the new inflation figure. This shows that while the headline annual rate is 2.8 per cent, by including previous years’ inflation (because individuals will be including previous years in their behaviour) weighted RPI is 3.71 per cent. As can be seen, weighted RPI drops below the actual RPI for June 2011 and 2010, the reason being that the RPI figure for June 2009 was -1.6 per cent. (The picture would have been different if CPI was used as it only dropped to +1.8 per cent in 2009.)

Now these weights are completely arbitrary. I’m not suggesting that a 50 per cent weighting for the previous year approximates human behaviour in any way. However, it is something that research can quantify, and if it is quantifiable, perhaps an indicator which quantifies it has more explanatory power than the indicators of inflation that we currently use.   

Annual rate of inflation (RPI)
Compound inflation (RPI)
Weighted RPI
 June 2010
 June 2011
 June 2012

No comments:

Post a Comment