Back in May 1997, 14 economists, statisticians and representatives of groups regarded as users of inflation statistics were appointed to advise the Minister of Statistics, then Maurice Williamson, on revision of the Consumers Price Index, or CPI. Collectively, they were known as the Consumers Price Index Revision Advisory Committee.
The committee included John Albertson, from the Retail and Wholesale Merchants Association; Peter Crawford, from the Manufacturers Federation; Jeremy Harding, from Federated Farmers; Grant Hannis, from the Consumers Institute; Peter Harris, from the Council of Trade Unions; Paul Stewart, from the Bankers Association. Oh, and Scott Roger, from the Reserve Bank, whose use of the CPI has significant influence on the rates we pay for the money we borrow, which in turn influences the exchange rate and the prices we fetch for our exports.
The terms of reference required the committee to:
* Investigate, review and make recommendations about the general nature, form and objective of the CPI, (looking at the range of uses to which the CPI is applied, for example, and at alternative measures that might be appropriate for CPI users);
* Investigate and make recommendations about the construction of the CPI (such as the range of goods and services, the household population and the urban areas covered by the index).
The committee reported back at the end of November that year.
The CPI is subjected to an exhaustive examination of this sort every six years or so, and Statistics New Zealand is limbering up for similar exercise next year.
A hefty bundle of papers was prepared by CPI experts for the 1997 review, addressing questions such as what should the CPI measure, and how should its measurement limitations be measured?
This might all sound somewhat arcane, but if your next pay rise or interest rate is geared to the CPI, you should have strong vested interest in hoping the index is properly constructed and gives reasonably accurate reading, and you may be bothered to learn there are measurement limitations that need to be managed.
But hey, the CPI is not precision instrument. It has margin of error and the statisticians who compile it know it has margin error. They don’t know the size of the margin, but one frailty is that the published figures don’t account for qualitative improvements in the goods we buy.
American business writer Joseph Spiers, some seven years ago, cited problems with the CPI as an example of why several sets of statistics needed upgrading to account for technological changes and other advances. The CPI’s major flaw, he pointed out, is its inability to adjust for quality changes in services like health care and banking.
The US Bureau of Labor Statistics at that time was spending considerable energy distinguishing quality changes from pure inflation in goods such as autos and apparel. car with two airbags this year, for example, is clearly superior to the same model with only one airbag last year, so the full price increase shouldn’t go into the American CPI.
But the specialists did not make quality adjustments for health care services, which are large and growing share of GDP. If course of treatment requires shorter hospital stay than previously while producing the same outcome, it should be recorded as price reduction. But such adjustments weren’t being made because they’re conceptually difficult to do and because research funds didn’t stretch that far.
According to Spiers, most analysts reckoned the American CPI overstated inflation by at least half percentage point per year and US Federal Reserve chairman Alan Greenspan suspected the CPI could be awry by full percentage point or more.
A week or so before the recent election campaign, Finance Minister Michael Cullen expressed the Government’s desire to have the Reserve Bank do its bidding, when it shoves interest rates up or down to keep annual inflation somewhere between 0-3 percent as measured by the CPI. This led to great deal of huffing and puffing about the propriety of Dr Cullen publicly chiding the bank, whose “independence” was of more importance to the minister’s critics than the economic effects of its monetary policy management.
The monetary management issue raised by Cullen was whether the bank should be targeting 1.5 percent, the mid-way point in the target band, or whether it should be content to keep the CPI anywhere between zero and three percent. Either way is neither here nor there, however, if the CPI is out by as much as one percent or more. Anyway, the CPI is not gauge of an overall economic price level because it ignores the prices of assets like jewellery, antiques and vintage cars along with the prices of houses, shares, farmland and commercial real estate. M
Bob Edlin is regular contributor to Management magazine.
