PUBLISHED : Sunday, 27 August, 1995, 12:00am
UPDATED : Sunday, 27 August, 1995, 12:00am

I REFER to the article headlined, 'Inflation data hiding disaster', published in the Sunday Morning Post on July 16.

Mr Michael Green, of the Nomura Research Institute, made several remarks on the rent component of the Consumer Price Index (CPI) in the article. However, there are certain misconceptions in these remarks and I would like to make some clarifications below.

The rent index in the CPI measures the average impact of rent increases on consumers. In compiling the index, new lettings as well as continued lettings and renewed lettings are taken into account. This resembles the market situation in which one would find a mix of new tenants, sitting tenants and tenants who are about to renew their tenancies.

For renewed lettings, rent increases normally range from 20 per cent to 30 per cent (for a two-year tenancy agreement) for residential premises falling within the purview of the CPI (A) and CPI (B).

Also relatively more rapid increases in the rentals of private flats would occur among new lettings.

The inclusion of continued lettings in the rent index of the CPI hence has a 'diluting effect', resulting in a comparatively smaller overall increase in the rent index.

It appears that the author focused only on new lettings, hence overstating the actual overall increase in residential rentals as well as the expenditure weight of rentals.

Past experience shows that average household expenditure patterns change only gradually over time. So as at present the expenditure weights are not reckoned to be significantly out-dated for the purpose of CPI compilation.

Under the current practice, the expenditure patterns of households are updated once every five years in a household expenditure survey (HES) to provide a reliable set of expenditure weights for the compilation of the CPI. The current round of HES, which is the 1994/95 HES, has already been conducted for some months and is now proceeding to an advanced stage. It covers the expenditure patterns of households over the period from October 1994 to September 1995.

When the summary results of this survey become available, the expenditure weights of the CPIs will be updated accordingly.

The new 1994/95-based CPIs and the updated set of expenditure weights will be released by the Government in April, 1996.

I hope the above explanations has helped readers to gain a proper understanding of the compilation basis of the CPI.

LILY OU-YANG Commissioner for Census and Statistics