Appendix 2 : Expenditure-Based Forward Planning Assessment Methodology

Introduction

1.          This appendix presents a worked example of an expenditure-based methodology which has been developed for the assessment of retail floorspace requirements at the sub-regional planning level. The same methodology may also be used to assess the floorspace requirement of individual developments.

2.          The assessment model is applicable for any size of centre. The example presented here is to assess the shopping needs of Tung Chung New Town - defined as a regional level centre upon full development. The approach can be adapted for estimating the requirements for smaller local centres as described in Paragraph 21 below.

3.          The methodology is carried out in 6 steps:

1. Define the Retail System

2. Collect Data for the Base Year

3. Calibrate a Base Year Model

4. Project the Variables to the Design Year

5. Assess the Supportable Floorspace in the Design Year

6. Assess the Need for Additional Floorspace.

             Step 1 : Define the Retail System

4.          Identify the probable extent of consumer expenditure sources - the "Catchment Area". This is a matter of judgement as to the total extent of the area in which residents are likely to regularly shop at Tung Chung. Regard should be given to the distance and location of alternative shopping venues.

5.          The catchment area is then divided into catchment zones. In the case of Tung Chung 5 zones were identified (see Figure A2.1) representing expected differences in consumer shopping habits by virtue of distance from Tung Chung Centre and competing centres, the presence of other shopping facilities in the zone (district and local centres), social/income characteristics and other factors which might be expected to affect the level of expenditure "captured" from the zone (see Step 2 below). In this case 6 zones are identified:

TC1 : Tung Chung New Town : planned as the regional centre and the subject of the assessment

TC2 : Tai Ho : planned as a district centre

TC3 : CLK Airport : an employment zone which may have some retail facilities

TC4 : Discovery Bay : existing area served by small local centre

TC5 : North Lantau : rural area with limited population and port facilities

TC6 : Rest of Lantau : remaining rural communities with limited traditional retail facilities.

             Step 2 : Collect Data for the Base Year (1995)

6.          For the above retail system of zones data is required on households, workers and visitors. A territorial total is also required. In the example, data for the base year is presented in Table A2.1 below.

Table A2.1 Numbers of Households, Workers and Visitors in each Zone

[ Table Summary ]

	TC1	TC2	TC3	TC4	TC5	TC6	Territorial Total
Household (1)	349	0	95	4,791	1,446	3,343	1,777,000
Worker (2)	98	0	13,800	765	1,783	16,446	2,905,100
Visitor (3)	0	0	0	0	0	39,810	10,199,994

  Note:

(1)    Figures for catchment zones are calculated by dividing the population figure with person-per-occupied-flat (PPOF) ratio. Data source for this is the Central Data Unit, Planning Department. The territorial figure is extracted from "Report of Population and Household Statistics", the C&SD.

(2)    Territorial figure is extracted from the "General Household Survey", the C&SD. Figures for catchment zones are estimated from their proportionate share of commercial floorspace in the Islands District. Data source is the "Employment and Vacancies Statistics (Detailed Tables)", the C&SD and the "Property Review", the R&VD.

(3)    Territorial figure is extracted from the "A Statistical Review of Tourism", the Hong Kong Tourist Association (HKTA). Figures for the catchment zones are estimated from their proportionate share of hotel rooms. Data source for this is the "Hotel Supply Situation", the HKTA.

7.          Collect data on average annual (territorial) retail expenditure for the main categories of retail expenditure by households, workers from their place of work, and visitors. For the year 1995, the expenditure in each category is presented in Table A2.2 below.

Table A2.2 Household, Worker and Visitor Expenditure

[ Table Summary ]

	Convenience Goods (1)	Comparison Goods (2)	Restaurant (3)
Household (per year less worker spending) (4)	32,803	28,662	32,539
Worker (per year) (5)	2,583	2,304	9,555
Visitor (per visit) (6)	0	4,023	819

  Note:

(1)    Include food (except meals bought away from home), alcoholic drinks, medicines, newspapers, and other personal effects.

(2)    Include clothing and footwear, durables, soft furnishings and other household goods.

(3)    Include meals bought away from home and expenses on dinner parties in restaurants.

(4)    Data source is the 1994/95 Household Expenditure Survey (HES), the C&SD.

(5)    Obtained by multiplying household expenditure by the percentage of household expenditure spent near workplace. Data source for the percentage is the "Shopping Habits Survey". Data source for household expenditure is the 1994/95 HES.

(6)    Data source is "A Statistical Review of Tourism", the HKTA. Non-retail expenditure such as hotel bills and tour expenses are excluded.

8.          A correction factor estimated as 1.314 in 1995 is required to be applied to the consumer expenditure data reported in the Household Expenditure Survey to bring it in line with the territorial turnover sales data reported by retailers.

9.          Estimate market share percentages for each category of goods. This is a professional judgement made about the percentage of expenditure by residents, workers and visitors in each catchment zone likely to be retained within the zone. This judgement is based on knowledge of the facilities already available or proposed in the zone, the distance and accessibility of other retail facilities in the zones (this varies with category of goods - most convenience shopping is carried out within the zone where facilities are available), the quality and mix of existing and proposed facilities and the habits of shoppers in the area. Some information on shopping habits is available in the study on Shopping Habits and Revision of HKPSG Chapter 6 (1998) based on a household survey. Current, local data on shopping habits may also be obtained by carrying out a local survey in conjunction with the assessment. A typical market share distribution for Convenience Goods for Tung Chung is shown in Table A2.3 below.

Table A2.3 Market Share for Convenience Goods

[ Simple Table Format ]

Retail Facilities in Catchment Zone	Flow of Retail Expenditure from Each Catchment Zone to Retail Facilities in Each Catchment Zone (%)
	TC1	TC2	TC3	TC4	TC5	TC6	Rest of HK
TC1	91.5	0	0	0	0	0	0
TC2	0	0	0	0	0	0	0
TC3	0	0	0	0	0	0	0
TC4	0	0	0	81.5	7	0	0
TC5	0	0	0	0	0	0	0
TC6	8.5	0	0	18.5	93	100	0
Rest of HK	0	0	100	0	0	0	100
Total	100	100	100	100	100	100	100

10.        Estimate territorial floorspace efficiency. This is the ratio between retail turnover and retail floor area as specified in the Annual Survey of Wholesale, Retail, Import and Export Trades, Restaurants and Hotels (C&SD). For the Tung Chung example the 1994 data was projected to 1995 giving a figure of $26,983 for Lantau Island and $51,414 for the rest of Hong Kong.

11.        Existing retail floorspace data for each catchment zone is obtained from the R&VD. A factor of 0.7778 in 1995 - is applied to discount the floorspace occupied by retail services.

            Step 3: Calibrate the Model for the Base Year

12.        The collected data is then used to calculate the supportable level of retail floorspace and the additional need for retail provision in 1995.

13.        A simple spreadsheet model has been developed which:

• estimates and aggregates the expenditure available from/to each zone from each category of shopper to each category of retail expenditure;

• converts the available expenditure to required floorspace in each zone by applying the floorspace efficiency ratio and correction factor; and

• compares the required floorspace with the existing floorspace to estimate the requirement in 1995.

14.        The spreadsheet model is set out in Appendix 7 to the Study on Shopping Habits and Revision of HKPSG Chapter 6 (1997).

15.        In the Tung Chung example the requirement for Zone TC1 (Tung Chung Centre) is 981 sqm which was not believed to represent a material difference. Supply and demand were therefore estimated to be in balance in 1995.

            Step 4: Project the Variables to the Design Year

16.        The data collected in Step 2 are projected forward to the "design year" - usually the first year of full operation - in this case 2011. Thus:

• projections of households, workers and visitors are obtained for 2011;

• annual retail expenditure is projected on past trends based on the HES and HKTA surveys;

• the same correction factor is applied;

• existing committed and proposed retail floorspace by 2011 in each Zone are added to the 1995 floorspace total; and

• market share percentages in each catchment zone are re-estimated based on the redistribution of floorspace proposed.

17.        An example of the changed pattern estimate for Tung Chung in 2011 for convenience goods is shown in Table A2.4 below.

Table A2.4 Market Share for Convenience Goods in 2011

[ Simple Table Format ]

Retail Facilities in Catchment Zone	Flow of Retail Expenditure from Each Catchment Zone to Retail Facilities in Each Catchment Zone (%)
	TC1	TC2	TC3	TC4	TC5	TC6	Rest of HK
TC1	100	20	10	14	70	5	0
TC2	0	80	0	14	25	5	0
TC3	0	0	30	0	0	0	0
TC4	0	0	0	66	0	0	0
TC5	0	0	0	0	0	0	0
TC6	0	0	0	6	5	90	0
Rest of HK	0	0	60	0	0	0	100
Total	100	100	100	100	100	100	100

18.        Floorspace efficiency change has been estimated to be negligible in the past 10 years and this is assumed not to change to 2011.

            Step 5: Assess Supportable Floorspace in the Design Year

19.        The data collected in Step 4 are again input to the spreadsheet model in the same way as Step 3. Total floorspace supportable for all zones (TC1 to TC6) amounts to 510,000 sq.m.IFA.

            Step 6: Assess the Need for Additional Floorspace Provision

20.        The final step is to compare this 2011 requirement with the existing and future provision. In Tung Chung Centre (TC1) the additional retail provision by 2011 is estimated to be about 264,000 sq.m. IFA.

            Variations in Modelling Procedures for Small Centre Developments

21.        The model can also be applied to assess the retail floorspace requirement for smaller local or district centres using the same six steps. However some variations in judgement are sometimes necessary:

In Step 1 the retail system defined is likely to be smaller in area and the number of zones fewer. There are likely to be more and closer competing centres in the retail hierarchy restricting the scale of the retail system.

In Step 2 it may be appropriate, where there is a higher proportion of convenience shopping in smaller centres, to use a different correction factor - estimated as 1:1.27 in 1995.

In Step 3 more care is required in assessing market share percentages where larger centres outside the system may attract expenditure from the zone and competition between centres is more localised. Local current survey data is even more valuable.

In Step 4 careful consideration should be given to proposals for major shopping centres outside the system which may attract expenditure in future.