Playing to a new set of rules in the game of inflation
By ROSS GITTENS
Next Wednesday we see the results of a minor revolution. That's when we're told by how much the Consumer Price Index rose in the September quarter. But it will be an all-new CPI.
The CPI will have a new ''principal purpose'', a new conceptual basis for the measurement of price changes and a new ''reference population''. It will measure the cost of housing quite differently; it will drop some items from the basket of goods and services whose prices it measures, while adding others.
What's the CPI for? Well, different people use it for different purposes: to measure changes in the purchasing power of household incomes, to measure changes in real income, to adjust social security pensions and benefits, to adjust some taxes (such as federal excises), to adjust fees and charges in a multitude of public and private sector contracts and to measure inflation for the purposes of managing the economy.
To the untrained eye, all those purposes may add up to pretty much the same thing. To economists and statisticians, however, most of them are quite different and, in theory, would require inflation to be measured in differing ways.
So, to a statistician, it's a question of deciding on the CPI's principal purpose, then making sure the CPI measures price changes in a way that fits this purpose.
Surprisingly few people realise it but, until now, the CPI's main purpose has been the indexation of award wages.
But in his recent review of the CPI, the Australian statistician realised that it's no longer being used for its principal purpose. Wage indexation was abandoned in 1987, we've move to enterprise bargaining and even the annual safety-net wage increases are set with only passing reference to what's happened to workers' cost of living.
These days, the main group whose incomes are indexed to the CPI are pensioners and others on government benefits. And their cost of living may not be rising at the same rate as wage earners' because they buy a different basket of goods and services.
Another development is that the Reserve Bank has become more serious about controlling inflation and attempting to lower people's expectations about inflation.
To this end, it's conducting monetary policy according to a target for inflation. But it regarded the old CPI as an unsuitable measure of inflation for macro-management purposes, and set its target in terms of Treasury's measure of ''underlying'' inflation (a measure that excludes roughly half the items in the CPI basket).
So the statistician decided to change the CPI's principal purpose from measuring changes in wage earners' cost of living to providing ''a general measure of price inflation for the household sector''.
And this change has meant altering the CPI's ''reference population''. Though few people realise it, to date the CPI has measured the change in price of a basket of goods and services purchased by the typical wage-earning household in the eight state and territory capitals.
And this means it has focused on the buying patterns of just the 29per cent of Australian households that met this definition.
The definition excluded the 9.6per cent of capital-city households which, though they were reliant on wages and salaries, either derived less that 75per cent of their income from wages and salaries or were in the top 10per cent of wage and salary earners.
It also excluded those capital-city households deriving their income from government pensions and benefits (17.3per cent), from self-employment (7.1per cent) and from superannuation (1.4per cent).
The new CPI will include all those groups, so it now covers the experience of 64.4per cent of households. The balance live outside the capitals.
Another change flowing from the CPI's new principal purpose is a change in the conceptual basis for the measurement of price changes. It's going from the ''outlays'' method to the ''acquisitions'' method.
In practice, the main change is in the way the CPI measures the cost of home ownership. When the CPI was designed to measure changes in workers' cost of living, it was appropriate to include the cost of mortgage interest payments.
But now the CPI is intended to be a general measure of inflation, it's more appropriate to include the cost of a new house (but not the land it stands on).
This is because, to an economist, interest rates are a different kind of price from the prices of ordinary goods and services. These are current prices, whereas interest rates are an ''inter-temporal'' (between time periods) price. That is, interest rates are the price of spending now rather than later.
It was the inclusion of interest rates in the CPI that was the main reason the Reserve Bank declined to use it for its inflation target.
The Reserve uses the manipulation of interest rates to control inflation. If it were to use a measure of inflation that included the direct effect of interest-rate changes, it would get muddled between cause and effect.
For this reason, the new CPI not only changes the way the cost of home ownership is measured, but also drops consumer credit charges (mainly credit-card interest rates) from its basket of goods and services.
And, as a result of the decision to exclude interest rates, the Reserve announced this week that it had changed from targeting underlying inflation to targeting the CPI.
It just so happens that, by excluding interest rates, the statistician has made the CPI a better measure of changes in the cost of living for the retired and other low-income earners.
Why? Because few of these people have mortgages. What's more, for most retired people interest rates aren't a cost but a source of income.
But while dropping interest rates from the basket, the statistician has used the opportunity to add others: home computers and software, domestic services (house cleaning, gardening and the like) and tertiary education fees.
The public will be oblivious to these changes but, to those in the know, they amount to a minor revolution.
theage.com.au |
