Monday, August 24, 2009

The living cost index that won't index living costs

"A living cost index is intended to be used to assess changes over time in the purchasing power of the after-tax incomes of households. It is therefore concerned with measuring the impact of changes in prices on the out-of-pocket expenses incurred by households to gain access to consumer goods and services." - ABS

A new cost-of-living index built specifically for pensioners to be unveilled today has been labeled a waste of up to $18 million by a critic of its design.

Promised in the Budget and recommended by the Harmer Pensions Review, the pensioner and beneficiaries index will be based on different a different mix of costs to the mainstream Consumer Price Index and will be used to increase pensions at times when it grows faster than wages.

But retired Canberra public servant Tom Hayes, a former head of the Department of Industry, says the index will hugely understate increases in the true cost of living faced by pensioners because it will measure only so-called "pure" price increases, rather than actual price movements.

"Washing machines, for example, increase in price," he told The Herald. "But because they get better all the time the Bureau of Statistics discounts the new higher prices to compare machines of equal quality. But in real life, lower-priced machines of the old quality will almost certainly no longer be available"...

"All that pensioners can find in the stores is an 'improved' model with a shelf-price not matched by the ABS estimate. That's how people dependent on on pensions indexed to the CPI find their purchasing power gradually eroded."

Mr Hayes says its hard to work out exactly how the Australian Bureau of Statistics adjusts shelf prices for changes in perceived quality, because it doesn't publish its detailed methodology.

"All that we know is that many years ago quality adjustments were limited to obviously-needed adjustments for things such as a change in the size of a tin of baked beans. Today the Bureau seeks information from many manufacturers and importers about quality improvements for all kinds of products and converts that data into pricing reductions based partly on obscure statistical techniques and partly on judgement," he says.

The Harmer Review itself found that rents had increased by 41 per cent at a time when the CPI said they had increased 23 per cent after adjusting for changes in the quality of the accommodation being rented. New car prices had climbed 20 top 50 per cent at a time when the CPI said they had increased 7.3 per cent.

"Aspects of this new index might well be more relevant to pensioners than the existing CPI," said Mr Hayes, "But they will be completely swamped by the use of pure pricing and not shelf pricing."

The ABS will put increase the weight it gives to food prices in the index to be released today in recognition of the fact that age pensioner households spend 21 per cent of their budgets on food, compared to 15 per cent for other households.

It has decided to include spending on alcohol and tobacco in the index on the basis that it should not make "social or moral judgments".

Costing $18.5 million over four years, the pensioner and beneficiary index will be updated every six months.

Published in today's SMH and Age

ABS 6466.0 - Information Paper: Introduction of the Pensioner and Beneficiary Living Cost Index, Australia, 2009

Comment on the ABS PBLCI Paper:


Bruce Bradbury said...

Just about all price indices (including the CPI) attempt to only measure 'pure' price increases. So there is nothing special about the pensioner price index in this regard. I think this is appropriate. People _are_ better-off if the quality of something has improved.

The washing machine example implies that people are forced to purchase a higher quality machine than they would prefer. However, you can get a poorer quality machine by purchasing second-hand, by paying to repair an old machine, or simply keeping your working old machine for a longer time. I can see that some people might find these other options impractical or difficult - but I'm not convinced this is empirically important.

These arguments apply even more so to cars, where there is a thriving second-hand market.

As for rent, the ABS says that the quality of the average rented house has gone up (which it then uses to adjust the rental price index down). This does not mean that one cannot still obtain poor quality housing. If increases in rent per unit of quality are constant across the board, then the ABS approach should also accurately reflect the increase in this poorer quality housing.

The ABS approach might be an imprecise method for measuring rent increases among poor quality dwellings, but the bias could equally well be in either direction.

Anonymous said...

ABS also deliberately omits land component of housing = CPI that bears no resemblence to reality of 300% land price increase over last decade.

Post a Comment