C&SD announces 2004/05-based Consumer Price Indices
***************************************************

    The Census and Statistics Department (C&SD) today (April 27) introduced the newly based Consumer Price Indices (CPIs).

     Speaking at the press conference today to announce the details, the Acting Commissioner for Census and Statistics, Mr Dominic Leung, said that a Household Expenditure Survey (HES) was conducted once every five years to provide up-to-date information for rebasing the CPIs. Information on expenditure patterns of households collected in the HES is used together with data on price movements collected on a monthly basis by the C&SD to compile the monthly CPIs.

     The existing series of the CPIs are based on the spending patterns obtained from the HES conducted in 1999/2000. Based on the results of a new round of HES conducted in 2004/05, the CPIs have been rebased. The expenditure ranges and weighting patterns of the CPIs have also been updated.

     The practice to rebase the CPIs once every five years in Hong Kong has been in place for a long time. Mr Leung explained that this practice, which conforms to international standards and practices, ensures that up-to-date expenditure patterns of households are accurately reflected in the compilation of the CPIs.

     "To maintain comparability of the various CPI series, the rebased series of the CPI(A) continues to relate to about 50% of households, in the relatively low expenditure range. Their average monthly expenditure during the new base period of October 2004 to September 2005 was between $4,000 and $15,499," Mr Leung said.

     "The rebased series of CPI(B) relates to about 30% of households, in the medium expenditure range. These households spent between $15,500 and $27,499 a month in the same base period.

     As for the rebased CPI(C), it relates to about 10% of households, in the relatively high expenditure range. Their monthly expenditure in the base period was between $27,500 and $59,999," Mr Leung added.

     The C&SD compiles different CPI series for reflecting the impact of consumer price changes on households in different expenditure groups. The CPI(A), CPI(B) and CPI(C) are compiled based on the consumption patterns of households in the relatively low, medium and relatively high expenditure ranges respectively. The Composite CPI is compiled based on the aggregate expenditure pattern of all of the above households taken together for reflecting the impact of consumer price changes on the overall household sector.

     The expenditure weights of the CPIs, which represent the shares of different commodities/services in total household expenditure, have been updated as shown in Table 1 at annex.

     The expenditure weights of 'Meals bought away from home', which account for a significant share of the 'Food' section, increased by 0.5 to 0.7 percentage point in each of the four CPIs.

     As regards 'Food, excluding meals bought away from home', the shares decreased by 0.5 and 1.4 percentage points in the CPI(A) and CPI(C) respectively but increased by 0.9 percentage point in CPI(B).

     For 'Housing' section, the weights for households in CPI(B) and CPI(C) dropped by 2.0 and 1.6 percentage points respectively. "This was mainly due to the general fall in rental for private housing compared to five years ago," explained Mr Leung. As for households in CPI(A), a large proportion of whom resided in public housing, the expenditure share of housing increased by 1.4 percentage points.  

     The expenditure weights of 'Electricity, gas and water' increased by 0.4 to 0.9 percentage point in all the CPIs.  Meanwhile, the weights of 'Alcoholic drinks and tobacco' decreased slightly in CPI(A) and CPI(B) but remained virtually unchanged in CPI(C).  

     As for 'Clothing and footwear', the expenditure shares dropped slightly in the CPI(A) and CPI(B) but rose marginally for households in CPI(C).  

     The weights of 'Durable goods' showed a decline of about 1 percentage point in the CPI(A) and CPI(B). This was mainly attributable to declines in the expenditure shares on video and sound equipment; and watches, clocks, camera and optical goods. Households in CPI(C), however, spent a slightly larger share of their expenditure on durable goods than five years ago.

     For 'Miscellaneous goods', the expenditure weights for households in all four types of CPI have dropped, by 0.6 to 1.5 percentage points.

     Regarding the 'Transport' section, the expenditure weights for households in different expenditure groups did not change in the same direction. The weight declined slightly in CPI(A) mainly due to decreases in the shares of taxi fares and expenses related to car ownership. The weight, however, rose in CPI(C) mainly due to increased expenditure shares on motor fuel and parking fees. As for the CPI(B), the weight remained unchanged.

     While a small decline in the expenditure share of 'Miscellaneous services' was recorded for households in CPI(A), notable increases of 2 to 3 percentage points are observed for the expenditure shares of 'Miscellaneous services' in CPI(B) and CPI(C). The latter was mainly due to the increases in the expenditure on education; medical services; and beauty treatment and fitness services.

     The latest HES was conducted from October 2004 to September 2005. About 6,100 households participated in the survey, representing a response rate of 81%.  

     "The 2004/05 HES has been conducted smoothly with the co-operation of the sampled households. I am very grateful to them and to the retail outlets which provide price data to us continuously. With their support, the quality of the CPIs can be guaranteed," Mr Leung said.

     The rebased CPI series and their year-on-year rates of changes for the reference months from October 2005 to March 2006 are presented in Table 2 at annex. The newly based CPIs for subsequent months will be published in future press releases and CPI monthly reports.

     Compared with the old (1999/2000-based) CPI series, the new (2004/05-based) CPI series generally shows smaller year-on-year rates of increase. This is because households tend to buy more of the goods and services with relatively smaller price increases (or relatively larger price decreases) and less of those with larger price increases (or smaller price decreases). Due to such substitution effect, CPIs with expenditure patterns fixed to a base year tend to over-estimate price increases or under-estimate price decreases over time.  

     For instance, compared with five years ago, households spent a greater share of their expenditure on products such as mobile phone, digital MP3 player/recorder, digital camera and electrical massage equipment, the prices of which have been falling. The effect of this change in spending pattern was to lower the rate of change in the overall CPI.

     "While the magnitudes of the year-on-year changes in the 2004/05-based and 1999/2000-based CPIs are somewhat different, the general trend of a gradual pick-up in the inflation rate from the 4th quarter of 2005 to the 1st quarter of 2006 is observed in all the four new CPI series, as is in the old CPI series," Mr Leung said.

     Concurrent with the 2004/05-based CPI series, the 1999/2000-based CPI series will continue to be compiled and published monthly until the reference month of December 2006.

     A detailed report on the results of the 2004/05 HES and the rebasing of the CPIs will be available in June this year.


Ends/Thursday, April 27, 2006
Issued at HKT 11:27

NNNN

Under the same press release series:

Related statistical subject(s):

Attachments
Table 1
Table 2