EXPLANATORY NOTES
INTRODUCTION
1 This publication presents monthly estimates of the value of turnover of retail businesses classified by industry, and by state and territory. The principal objective of the series is to show month to month movement of turnover.
2 Estimates of turnover contained in this publication are compiled from the Retail Business survey. Following a new sample design introduced in the July 2004 issue, the survey includes about 4,350 retail and selected service businesses. All 'large' businesses are included in the survey, while a sample of about 3,500 'smaller' businesses is selected. The 'large' business' contribution of approximately 55% of the total estimate ensures a highly reliable Australian total turnover estimate.
SCOPE AND COVERAGE
3 The scope of the Retail Business survey is all employing businesses with at least one retail outlet. Like most Australian Bureau of Statistics (ABS) economic surveys, the frame used for the Retail Business survey is taken from the ABS Business Register which includes registrations to the Australian Taxation Office's (ATO) pay-as-you-go withholding (PAYGW) scheme. Each statistical unit (as defined below) included on the ABS Business Register is classified to the Australian and New Zealand Standard Industrial Classification (ANZSIC) industry in which it mainly operates. The frame is supplemented with information about businesses which are classified as non-retail but which have significant retail activity.
4 The frame is updated quarterly to take account of new businesses, businesses which have ceased employing, changes in industry and other general business changes. The estimates include an allowance for the time it takes a newly registered business to get on to the survey frame. Businesses which have ceased employing are identified when the ATO cancels their Australian Business Number (ABN) and/or PAYGW registration. In addition, businesses with less than 50 employees, and which do not remit under the PAYGW scheme in each of the previous five quarters are removed from the frame.
5 The following industries included in the survey are as defined in ANZSIC:
- Food retailing
Supermarkets and grocery stores (5110) and non-petrol sales of convenience stores of selected petrol stations
Takeaway food retailing (5125)
Other food retailing
Fresh meat, fish and poultry retailing (5121)
Fruit and vegetable retailing (5122)
Liquor retailing (5123)
Bread and cake retailing (5124)
Specialised food retailing n.e.c. (5129)
- Department stores (5210)
- Clothing and soft good retailing
Clothing retailing (5221)
Footwear, fabric and other soft good retailing
Footwear retailing (5222)
Fabric and other soft good retailing (5223)
- Household good retailing
Furniture and floor covering retailing
Furniture retailing (5231)
Floor covering retailing (5232)
Domestic hardware and houseware retailing (5233)
Domestic appliance and recorded music retailing
Domestic appliance retailing (5234)
Recorded music retailing (5235)
- Recreational good retailing
Newspaper, book and stationery retailing (5243)
Other recreational goods retailing
Sport and camping equipment retailing (5241)
Toy and game retailing (5242)
Photographic equipment retailing (5244)
- Other retailing
Pharmaceutical, cosmetic and toiletry retailing (5251)
Other retailing n.e.c.
Antique and used good retailing (5252)
Garden supplies retailing (5253)
Flower retailing (5254)
Watch and jewellery retailing (5255)
Retailing n.e.c. (5259)
- Hospitality and services
Hotels and licensed clubs
Pubs, taverns and bars (5720)
Clubs (Hospitality) (5740)
Cafes and restaurants (5730)
Selected services
Video hire outlets (9511)
Hairdressing and beauty salons (9526).
STATISTICAL UNITS DEFINED ON THE ABS BUSINESS REGISTER
6 The ABS uses an economic statistics units model on the ABS Business Register to describe the characteristics of businesses, and the structural relationships between related businesses. The units model is also used to break groups of related businesses into relatively homogeneous components that can provide data to the ABS.
7 In mid-2002, to better use the information available as a result of The New Tax System, the ABS changed its economic statistics units model. The new units model allocates businesses to one of two sub-populations. The vast majority of businesses are in what is called the ATO Maintained Population, while the remaining businesses are in the ABS Maintained Population. Together, these two sub-populations make up the ABS Business Register population.
ATO Maintained Population
8 Most businesses and organisations in Australia need to obtain an ABN, and are then included on the ATO Australian Business Register. Most of these businesses have simple structures; therefore the unit registered for an ABN will satisfy ABS statistical requirements. For these businesses, the ABS has aligned its statistical units structure with the ABN unit. The businesses with simple structures constitute the ATO Maintained Population, and the ABN unit is used as the statistical unit for all economic collections.
ABS Maintained Population
9 For the population of businesses where the ABN unit is not suitable for ABS statistical requirements, the ABS maintains its own units structure through direct contact with each business. These businesses constitute the ABS Maintained Population. This population consists typically of large, complex and diverse businesses. The new statistical units model described below has been introduced to cover such businesses.
Enterprise Group: This is a unit covering all the operations in Australia of one or more legal entities under common ownership and/or control. It covers all the operations in Australia of legal entities which are related in terms of the current Corporations Law (as amended by the Corporations Legislation Amendment Act 1991), including legal entities such as companies, trusts, and partnerships. Majority ownership is not required for control to be exercised.
Enterprise: The enterprise is an institutional unit comprising (i) a single legal entity or business entity, or (ii) more than one legal entity or business entity within the same Enterprise Group and in the same institutional subsector (i.e. they are all classified to a single Standard Institutional Sector Classification of Australia subsector).
Type of Activity Unit (TAU): The TAU is comprised of one or more business entities, sub-entities or branches of a business entity within an Enterprise Group that can report production and employment data for similar economic activities. When a minimum set of data items are available, a TAU is created which covers all the operations within an industry subdivision (and the TAU is classified to the relevant subdivision of the ANZSIC). Where a business cannot supply adequate data for each industry, a TAU is formed which contains activity in more than one industry subdivision.
10 For more information on the impacts of the introduction of the new economic statistics units model, refer to Information Paper: Improvements in ABS Economic Statistics [Arising from the New Tax System] (cat. no. 1372.0).
11 Prior to the July 2002 reference month, the Retail Business survey used the management unit as the statistical unit. From the July 2002 reference month onwards, the statistical unit is the ABN unit for businesses with simple structures, and the TAU for businesses with complex structures. In most cases, ABN/TAU units concord with the management units previously used.
SURVEY METHODOLOGY
12 The survey is conducted monthly by both telephone interview and a questionnaire mailed to businesses. The businesses included in the survey are selected by random sample from a frame stratified by state, industry and business size. Following the new sample design introduced in the July 2004 issue, the survey uses annualised turnover as the measure of business size. For the ATO Maintained Population, the annualised turnover is based on the ATO's Business Activity Statement item Total sales and for the ABS Maintained Population a modelled annualised turnover is used. For stratification purposes the annualised turnover allocated to each business is not updated each quarter as to do so would result in increased volatility in the estimates.
13 The July 2004 issue also saw the introduction of the generalised regression estimation methodology which replaced ratio estimation. For estimation purposes the annualised turnover allocated to each business is updated each quarter. The introduction of the new sample design and new estimation methodology resulted in changes to the level of the Retail Trade series. However, to facilitate comparisons over time, the historical series were revised to make the time series of estimates as continuous as possible. For more information about the changes introduced in the July 2004 issue refer to Information Paper: Changes to the Retail Trade Series (cat. no. 8501.0.55.002).
14 In the first month of each quarter, some businesses in the sample are replaced, at random, by other businesses so that the reporting load can be spread across smaller retailers.
15 Most businesses can provide turnover on a calendar month basis and this is how the data are presented. When businesses cannot provide turnover on a calendar month basis, the reported data and the period they relate to are used to estimate turnover for the calendar month.
16 Most retailers operate in a single state/territory. For this reason, estimates of turnover by state/territory are only collected from the larger retailers which are included in the survey each month. These retailers are asked to provide turnover for sales from each state/territory in which the business operates. Turnover for the smaller businesses is allocated to the state of their head office or main outlet.
DEFINITION OF TURNOVER
17 Turnover includes retail sales; wholesale sales; takings from repairs, meals and hiring of goods (except for rent, leasing and hiring of land and buildings); commissions from agency activity (e.g. commissions received from collecting dry cleaning, selling lottery tickets, etc.); and net takings from gaming machines etc. From July 2000, turnover includes the goods and services tax.
18 Turnover presented in the Retail Trade series includes net proceeds from licensed gambling activities undertaken in the Hotels and licensed clubs industry. The impact of net proceeds from gambling on movements in the Retail Trade series was discussed in Feature article: Contribution of gambling to retail estimates included in the December 2002 issue of this publication. The article concluded that net proceeds from gambling had not had a significant impact on quarterly movements for the series but net proceeds from gambling had increased over time and users should be aware of this when interpreting the series. An electronic release, Contribution of Gambling to Retail Estimates (cat. no. 8501.0.55.003), provides updated quarterly information and is available free of charge from the ABS web site. It is released approximately a week after the release of the March, June, September and December issues of this publication.
SEASONAL ADJUSTMENT
19 Seasonally adjusted estimates are derived by estimating and removing systematic calendar related effects from the original series. In the Retail trade series, these calendar related effects are known as seasonal (e.g. increased spending in December as a result of Christmas) and trading day influences (arising from the varying length of each month and the varying number of Sundays, Mondays, Tuesdays, etc. in each month). Each influence is estimated by separate seasonal and trading day factors which, when combined, are referred to as the combined adjustment factors.
20 The seasonally adjusted estimates also have an allowance for an Easter proximity effect, which is caused when Easter falls late in March or early in April. This effect, when present, is combined with the seasonal and trading day factors to form the combined adjustment factors. There is also a similar allowance for the variable timing of Father's Day. See the Appendix of the July 2001 and August 2002 issues respectively of this publication for more information.
21 The Retail series uses a concurrent seasonal adjustment methodology to derive the combined adjustment factors. This means that data from the current month are used in estimating seasonal and trading day factors for the current and previous months. For more information see Information Paper: Introduction of Concurrent Seasonal Adjustment into the Retail Trade Series (cat. no. 8514.0).
22 Concurrent adjustment can result in revisions each month to estimates for earlier periods. However, in most instances, the only noticeable revisions will be to the combined adjustment factors for the current month, the previous month and the same month a year ago. The following table shows how the combined adjustment factor for these months, at the total Australian Retail and Hospitality/Services level, evolved under the concurrent seasonal adjustment methodology. The table presents two different estimates of the combined adjustment factors. The first row gives the combined adjustment factors estimated following the last annual reanalysis in August 2006 using data up to and including the June 2006 reference month. The second row gives the most recent combined adjustment factors estimated and used in this month's calculation of the concurrent seasonally adjusted series.
23 The revision properties of the seasonally adjusted and trend estimates can be improved by the use of autoregressive integrated moving average (ARIMA) modelling. ARIMA modelling relies on the characteristics of the series being analysed to project future period data. The projected values are temporary, intermediate values, that are only used internally to improve the estimation of the seasonal factors. The projected data do not affect the original estimates and are discarded at the end of the seasonal adjustment process. The retail collection uses ARIMA modelling where appropriate for individual time series. The ARIMA model is assessed as part of the annual reanalysis and following the 2006 annual reanalysis 94% of Retail series use an ARIMA model. For more information on the details of ARIMA modelling see Feature article: Use of ARIMA modelling to reduce revisions in the October 2004 issue of Australian Economic Indicators (cat. no. 1350.0).
24 The seasonal adjustment methodology is able to produce combined adjustment factors for future months. The latest factors for some future months are shown in the following table. While these factors represent the best current estimate, the actual factors used for estimating the seasonally adjusted estimates in these months will differ because they will incorporate subsequent months' data as they become available.
25 The seasonal and trading day factors are reviewed annually at a more detailed level than possible in the monthly processing cycle. The annual reanalysis will not normally result in significant changes. For Retail Trade, the results of the latest review are shown in the July 2006 issue.
26 In the seasonal adjustment process, both the seasonal and trading day factors evolve over time to reflect changes in spending and trading patterns. Examples of this evolution include the slow move in spending from December to January; and, increased trading activity on weekends and public holidays. The seasonally adjusted estimates still reflect the sampling and non-sampling errors to which the original estimates are subject.
27 A "two-dimensional reconciliation" methodology has been used on the seasonally adjusted time series in this publication to force additivity - that is, to force the sum of fine-level (state by industry) estimates to be equal to the relevant state and industry totals, and Australian total. This methodology was first implemented in the November 2006 publication, and in publications prior to November the total for a state or industry did not necessarily equal the sum of the component series. The new methodology has been implemented for all timepoints back to the beginning of all series. For further details on the methodology employed, contact <time.series.analysis@abs.gov.au>.
TREND ESTIMATES
28 The monthly trend estimates are derived by applying a 13-term Henderson moving average to the seasonally adjusted estimates (7-term for quarterly series). The Henderson moving average is symmetric, but as the end of a time series is approached, asymmetric forms of the moving average have to be applied. The asymmetric moving averages have been tailored to suit the particular characteristics of individual series and enable trend estimates for recent periods to be produced. Estimates of the trend will be improved at the current end of the time series as additional observations become available. This improvement is due to the combined effect of the concurrent seasonal adjustment methodology and the application of different asymmetric moving averages for the most recent six months (or three quarters). As a result of the improvement, most revisions to the trend estimates will be observed for the most recent six months (or three quarters).
29 Trend estimates are used to analyse the underlying behaviour of the series over time. As a result of the introduction of The New Tax System, a break in the monthly trend series has been inserted between June and July 2000. Care should therefore be taken if comparisons span this period. For more details refer to the Appendix in the December 2000 issue of this publication.
30 For further information on trend estimates, see Information Paper: A Guide to Interpreting Time Series - Monitoring Trends, 2003 (cat. no. 1349.0) or contact the Assistant Director, Time Series Analysis on Canberra (02) 6252 6345 or by email at <time.series.analysis@abs.gov.au>.
CHAIN VOLUME MEASURES
31 The chain volume measures of retail turnover appearing in the quarterly issue of this publication are annually reweighted chain Laspeyres indexes referenced to current price values in a chosen reference year. The reference year is advanced in each June issue and is currently 2004-05. Each year’s data in the Retail chain volume series are based on the prices of the previous year, except for the quarters of the latest incomplete year. Data for the 2006-07 financial year will initially be based upon price data for the 2004-05 financial year. Comparability with previous years is achieved by linking (or chaining) the series together to form a continuous time series. While current price estimates reflect both price and volume changes, chain volume estimates measure changes in value after the direct effects of price changes have been eliminated and hence only reflect volume changes. Further information on the nature and concepts of chain volume measures is contained in the ABS publication Information Paper: Introduction of Chain Volume Measures in the Australian National Accounts (cat. no. 5248.0).
RELIABILITY OF ESTIMATES
32 There are two types of error possible in estimates of retail turnover:
Sampling error which occurs because a sample, rather than the entire population, is surveyed. One measure of the likely difference resulting from not including all establishments in the survey is given by the standard error, see below. Sampling error may be larger for the first month of each quarter, when some of the businesses in the sample are replaced by other businesses so that the reporting load can be spread across retailers.
Non sampling error which arises from inaccuracies in collecting, recording and processing the data. The most significant of these errors are: misreporting of data items; deficiencies in coverage; non-response; and processing errors. Every effort is made to minimise reporting error by the careful design of questionnaires, intensive training and supervision of interviewers, and efficient data processing procedures.
STANDARD ERRORS
33 Seasonally adjusted and trend estimates and chain volume measures are also subject to sampling variability. For seasonally adjusted estimates, the standard errors are approximately the same as for the original estimates. For trend estimates, the standard errors are likely to be smaller. For chain volume measures, the standard errors may be up to 10% higher than those for the corresponding current price estimates because of the sampling variability contained in the prices data used to deflate the current price estimates.
34 Estimates, in original terms, that have an estimated relative standard error (RSE) between 10% and 25% are annotated with the symbol '^'. These estimates should be used with caution as they are subject to sampling variability too high for some purposes. Estimates with an RSE between 25% and 50% are annotated with the symbol '*', indicating that the estimates should be used with caution as they are subject to sampling variability too high for most practical purposes. Estimates with an RSE greater than 50% are annotated with the symbol '**' indicating that the sampling variability causes the estimates to be considered too unreliable for general use.
35 To further assist users in assessing the reliability of estimates, key data series has been given a grading of A to E. Where:
- A represents a relative standard error on level of less than 2%. The published estimates are highly reliable for movement analysis.
- B represents a relative standard error on level between 2% and 5%, meaning the estimate is reliable for movement analysis purposes.
- C represents a relative standard error on level between 5% and 10%, meaning users are advised to exercise some caution in interpreting movements for such series.
- D represents a relative standard error on level between 10% and 15% meaning users are advised to exercise caution in interpreting movements for such series.
- E represents a relative standard error on level greater than 15% (mainly affects unpublished state by industry series).
36 The table below provides an indicator of reliability for key retail turnover estimates.
![](/icons/ecblank.gif) | Food
retailing | Department
Stores | Clothing
and soft
good
retailing | Household
good
retailing | Recreational
good
retailing | Other
retailing | Hospitality
and
services | Total | ![](/icons/ecblank.gif) |
| ![](/icons/ecblank.gif) |
NSW | B | A | C | B | C | C | B | A | ![](/icons/ecblank.gif) |
Vic. | B | A | C | C | C | C | C | A | ![](/icons/ecblank.gif) |
Qld | B | A | C | B | D | C | C | B | ![](/icons/ecblank.gif) |
SA | B | A | C | C | D | C | C | A | ![](/icons/ecblank.gif) |
WA | B | A | B | C | D | C | C | A | ![](/icons/ecblank.gif) |
Tas. | B | np | C | C | D | np | C | A | ![](/icons/ecblank.gif) |
NT | B | np | C | B | E | np | C | A | ![](/icons/ecblank.gif) |
ACT | B | A | C | C | D | D | C | B | ![](/icons/ecblank.gif) |
Aust. | A | A | B | B | B | B | B | A | ![](/icons/ecblank.gif) |
| ![](/icons/ecblank.gif) |
np not available for publication but included in totals where applicable, unless otherwise indicated |
ABS DATA AVAILABLE ON REQUEST
37 Retail Survey Special Data Service provides additional retail trade statistics which include further state industry dissections through to ‘top ten’ industry reports. For more information, contact the Retail Trade Special Data Services manager on Canberra (02) 6252 5220.
RELATED PUBLICATIONS
38 Current publications and other products released by the ABS are available from the Statistics View. The ABS also issues a daily Release Advice on the web site which details products to be released in the week ahead.