Third HKCSI Position Paper on Service Industries Statistics
September 1996

1. In October 1994, the HKCSI submitted a second position paper on services statistics to the government. In that paper the HKCSI sought to review the progress of development in services statistics since the first HKCSI position paper published in October 1992. Government's response to the second paper was received in April 1995. Considerable progress has been made since the two HKCSI position papers, both for economic statistics in general and for services statistics in particular. In the former case, for example, significant advances have been achieved in the production of gross national product (GNP) statistics and the development of balance of payment statistics. Likewise, there has been considerable improvement in the quality and coverage of services statistics.
2. This third position paper by the HKCSI is aimed at reviewing progress since the second position paper, as well as bringing into focus further issues in the development of services statistics. In the following sections, where reference is made to the second position paper, the paragraph numbers of the previous paper will be given in square brackets.

Price statistics

3. Consumer prices [4]

   3.1

   There have been a number of positive developments including the following:

   1. Since January 1994 detailed composite CPI figures were available from the official CPI reports.
   2. Since July 1995 the Hang Seng CPI index figures were no longer published separately by the Hang Seng Bank, but incorporated into the official CPI report. With the Hang Seng CPI and the Composite CPI, the official CPI report now offers much more comprehensive coverage and well as clearer presentation.
   3. The CPI figures have now been rebased to 1994/95 prices according to the new round of the Household Expenditures Survey. In the newly-based figures, some of the groupings hitherto consisting of both goods and services items have now been disaggregated. This has enabled the CSPI indices to be more representative of services prices than previously calculated under the 1989/90-based figures.
   4. Since July 1996 both government's CPI and HKCSI's Consumer Services Price Index and their year-on-year rates of changes are based on the new 1994/95 series of prices.

   3.2

   Hitherto the CSPI figures have included indices corresponding to households in the CPI(A), CPI(B) and the Hang Seng CPI, but not that of the composite CPI. In order not to make too many changes to the CSPI report (having just changed the base-year to 1994/95), the HKCSI will maintain the present format of the CSPI report for the time being but will consider adding the composite CSPI index in due course.

4. Producer price index [12].

   4.1

   In their reply to the HKCSI paper, the Census & Statistics Department has undertaken to first develop PPI for the manufacturing sector, before working on PPI on the service industries. This is to be welcomed.

   4.2

   In view of the potential usefulness of the PPI as a "leading" indicator of the economy, we suggest that the Census and Statistics Department should

   1. step up its work on PPI for manufacturing;
   2. in spite of the methodological and practical difficulties which we fully acknowledge, start work on developing PPI for services;
   3. draw up a timetable on the development of PPI's for both manufacturing and services, including target dates for publishing these indices.

Production-based GDP [14]

5. In our last position paper we proposed that production-based GDP figures be broken down into meaningful subsectors as follows:

   Sectors	Subsectors
   Electricity, gas & water	Electricity Gas & water
   Construction	Architectural & civil engineering works Other construction
   Wholesale, retail, import/export trades, restaurants & hotels	Distributive trades	Wholesale
   		Retail
   		Import/export trades
   	Catering trades	Restaurants
   		Hotels
   Transport, storage and communication	Transport and related services Storage Communication
   Financing, insurance, real estate and business services	Financing services (including banks, DTCs) Insurance services Real estate services Business services
   Community, social and personal services	Government services Commercial services Private non-profit bodies

6. Many of the subsectoral breakdowns are now available upon enquiry from the Census and Statistics Department, although in the official "Estimates of Gross Domestic Products", only the broad sectoral figures are shown. A number of subsectoral figures are, however, not available from the Census and Statistics Department, notably, those in italics in the above table. It is believed figures for these groups are withheld for lack of precision.
7. To increase the usefulness of the production-based figures, we propose Census and Statistics Department consider the following:
   1. Census and Statistics Department should endeavour to publish detailed figures for all subsectors as shown above. In addition, for the "transport and related services" sector, it would be very useful and informative if this could be further broken down into passenger transport and cargo transport. To this end, Census and Statistics Department may have to develop more precise methodology for estimation, for example, through increasing the sample size of its surveys.
   2. Figures for the more detailed subsector breakdowns should be published in its annual "Estimates of Gross Domestic Products".
   3. Currently, all production-GDP figures are given at current prices. Census and Statistics Department should consider looking into developing GDP deflators for production-based GDP.
   4. Because of the need to conduct surveys to estimate value-added, there is currently a significant gap in the timing of the publication of expenditure-based and production-based GDP. This gap should, as far as possible, be shortened.
   5. At present, expenditure-based GDP is published at quarterly intervals but production-based GDP is available annually. We fully appreciate that it would be very difficult to produce quarterly figures for production-based GDP. However, we wonder if some compromise could not be made with regard to precision and level of detail so as to enable quarterly figures to be estimated to cover, for example, major sectors only.

Trade statistics [15]

8. Hong Kong's exports and imports of services are classified into six major groupings, namely, transportation, travel, insurance, financial, trade-related and other business. According to the notes in the "Estimates of Gross Domestic Product":

   Transportation	includes sea, air and transport
   Travel	includes business and personal travel
   Insurance	includes direct insurance and re-insurance
   Financial	includes banking, financial assets dealing and brokerage services
   Trade-related	includes offshore trading (merchanting), purchasing services and other trade-related services
   Other business	includes communication, legal, construction, industrial, advertising, marketing research, management consultancy, computer, information, accounting, real estate, architectural, news transmission, production and distribuion of film, hotel management services etc.

9. Such grouping is too general and in our second position paper we have proposed that there should be more detailed sectoral as well as by-country breakdowns. In their reply to the CSI position paper, Census and Statistics Department undertook to expand coverage of trade in services as well as breakdown by importing/exporting countries. We strongly support Census and Statistics Department's efforts to move in that direction.
10. To facilitate the development of trade in services statistics, we propose that Census and Statistics Department consider the following:
    1. Increasing the sample size of its survey on trade in services to the extent that meaningful disaggregation of subsectors hitherto lumped together as "other services" can be achieved.
    2. Participate actively in international forums on development in services statistics. Hong Kong's trade authorities have taken a positive and sometimes leading role in multilateral trade negotiations. Our statistics authority should, likewise, be actively involved in the policymaking of international statistical development, for example, taking part in resolving difficult statistical issues such as that of the four modes of delivery in trade in services.
    3. Enhance the quality of services trade data dissemination. At present there is no separate report on trade in services. As the volume of services exports already exceeds that of domestic goods, the time has come for services trade data to be more fully reported and analysed in an independent volume, rather than just as a table in the "Estimates of Gross Domestic Product". Although this may have to wait until more details on sectoral and by-country breakdowns are available, a specific and realistic timetable should be drawn up.

External direct investment [6]

11. With the publication of the first results of the Census & Statistics Department's new annual survey of external investment in non-manufacturing sectors (of 1993/94 data) in May 1996, an important gap in our knowledge of external direct investment is now being filled.
12. Notwithstanding the very useful and often interesting findings of the survey, some areas of the survey may merit a closer look, in particular, the classification of some of the surveyed companies as "holding companies". Although the number of such companies is small, together they occupy a significant proportion in terms of net assets, thus affecting considerably the sectoral picture. It would appear that the nature of these companies would merit closer study.

Productivity measures [16]

13. In our previous paper we suggested that there should be more constructive discussion and research into productivity of services. We were encouraged to learn that the Hong Kong Productivity Council is of the same view and in a consultation paper of its future role, the HKPC has proposed that it be involved in the development of services productivity measures.
14. We appreciate that productivity measures for services will be much more difficult to develop than that for goods, with formidable practical and conceptual problems to overcome. However, we suggest Census and Statistics Department should take advantage of the current enthusiasm over services productivity and react positively to the HKPC's proposal. The HKCSI will be happy to be suitably involved in the discussion over the development of services productivity measures.

River trade and transshipment

15. River trade statistics [13]. In Census & Statistics Department's reply to the CSI position paper, government undertook to upgrade river trade statistics with more detailed breakdown by port, commodity and cargo handling location. The first of a more detailed report on river trade statistics was published in the September 1995 issue of the Monthly Digest of Statistics. This is warmly welcomed.
16. Transshipment. In line with the continual restructuring of Hong Kong towards a service economy, there is a gradual change in the nature of re-exports towards more transshipment. There is a greater need, therefore, to understand more about the nature of transshipment. At present, figures on transshipment tonnage are available but not on value and cargo nature. We suggest Census and Statistics Department develop estimates for a more detailed picture of transshipment.

Towards better quality services statistics

17. Many of our proposals necessitate more work on the part of both government and the private sector. Census and Statistics Department will have to conduct more surveys and there will be more work on analysis and dissemination, whereas the private sector will have to accept an increased reporting burden.
18. Notwithstanding our reservation towards bigger government, we would support more funding and resources for the purpose of improving our statistical system for services. If more funding were not possible, then existing expenditure should be progressively reallocated to match the relative weight of the sectors.
19. For our part the HKCSI will be happy to persuade service sector firms to accept the greater reporting burden which can be anticipated as services statistics are being upgraded. At the end of the day, however, the greater efforts by both government and the private can only be justified by better-quality statistics. To ensure that collection of data and its analysis and presentation are done in the most cost-effective manner, we suggest that Census and Statistics Department consider forming technical consultative groups comprising industry representatives to advise on the form and content of the industry surveys.
20. As a means of improving cost-effectiveness of statistical collection and analysis, there may be room for some of the statistical surveys to be conducted by private enterprises. We suggest that the feasibility of privatising some of the statistical collection and compilation should be explored, with an open mind, by Census and Statistics Department and other interested parties.
21. In the longer term, the quality of statistics will benefit from more interaction between the Census authorities and an informed public. In this connection, we commend Census and Statistics Department for the bilingualism which it recently adopted in its reports, and for setting up its homepage on the Internet. To improve effective dissemination of statistics, we suggest Census and Statistics Department increase the coverage and make more liberal use of its homepage.
22. Another long-term development which would help improve quality of statistics is a strengthening of statistician as a profession. Although the statisticians of Hong Kong are already of the highest international standards, the local statistician profession as such is not yet a fully accreditted profession, separate accreditation having to be sought from international professional authorities. The HKCSI fully supports a strong local statistician profession and will be happy to support the Hong Kong Statistical Society in upgrading the international professional status of local statisticians.

(Ends)

¡@