TECHNICAL NOTE 1 METHODOLOGY
2 The counts of businesses in the diagram refer to the numbers of businesses on the survey 'frame', or list of business units, from which the sample to produce these estimates is selected. During processing of the survey, units no longer in operation can be detected, and the counts of the numbers of businesses adjusted accordingly. Hence the total number of businesses shown in the diagram is higher than the count of operating businesses presented in the tables.
PRODUCING EAS ESTIMATES
1 The estimates in this publication, for each of 2001-02 and 2002-03, are produced using a combination of ABS directly collected data and business income tax (BIT) data sourced from the Australian Taxation Office (ATO). The following diagram (using data from the 2002-03 collection) illustrates the ways in which Australian businesses contribute to these estimates.
3 Estimates for 2000-01 are shown on two bases in table 1.1. For details of the derivation of the 'old basis' management unit estimates shown, see Technical Note 1 in the previous (2000-01) issue of this publication. For details of the 'new basis' bridged ABN unit / TAU estimates, see Appendix 1 of this publication.
TYPES OF BUSINESS, AND DATA STREAMING
4 For the purpose of compiling the estimates in this publication, data for businesses as recorded on the ABS Business Register (ABSBR) contribute via one of three categories (or 'streams') in accordance with significance and collection-related characteristics.
Stream D: direct collection of data
5 STREAM D, consisting of directly collected survey data, comprises data for:
6 Information for these businesses was sourced from the ABS EAS collection. The main reasons for this approach were:
- those businesses from the Australian Taxation Office maintained population (ATOMP) of the ABSBR that are either:
- tax-exempt, or
- completely enumerated in the Economic Activity Survey (EAS) - a business may be completely enumerated when it has significantly high employment, significantly high assets, or a significantly high level of turnover
- businesses from the ABS maintained population (ABSMP) of the ABSBR.
Stream B: Both direct collection and BIT data
7 STREAM B, comprises data for ATOMP businesses not selected in STREAM D and which are above the cut-off for inclusion in STREAM T. This cut-off is usually set so that businesses in STREAM T contribute 15% of industry turnover, as determined from ATO Business Activity Statement (BAS) data.
- ABSMP businesses may have more than one Australian Business Number (ABN), making it difficult to identify all ABNs for any such business on the BIT files.
- BIT data do not include all of the detailed information that the ABS requires from completely enumerated businesses.
- Tax exempt businesses, such as charities, are not required to complete business income tax returns.
8 Data for businesses in STREAM B are sourced from the BIT files. However, a sample of these businesses were also included in the direct EAS collection, for the purpose of calculating 'proration' factors. These factors were applied to the STREAM B BIT data in order to produce estimates of items not available from the BIT files.
Stream T: BIT data only
9 STREAM T comprises data for ATOMP businesses not selected in STREAM D and which are below the STREAM T cut-off. Data for businesses in STREAM T are sourced from the BIT files. Directly collected survey data were used to prorate the STREAM T BIT data, in order to produce estimates of items not available from the BIT files.
10 For businesses in STREAMS B and T, the more detailed information presented in this publication is derived by using a two phase estimation technique. Firstly, a set of key data items are sourced from the BIT files. Then these data items are 'prorated' to generate a set of detailed financial information for each business. The proration factors are sourced from businesses in the EAS in the appropriate industry.
11 The sampling and estimation methods used by the ABS take into account those selected businesses that are not able to be identified on the BIT files. The ABS was unable to match all of the selected ABSBR businesses to records on the BIT files, as:
- not all businesses have completed their business income tax return by the time the ABS receives the snapshot of the BIT file, and
- a proportion of businesses that are included on the ABSBR would not have traded for the reference year.
AUSTRALIAN INDUSTRY ESTIMATES
12 Estimates for each of the selected industries were produced by aggregating the contributing data streams.
STATE AND TERRITORY EXPERIMENTAL ESTIMATES
13 The methodology for producing the state and territory experimental estimates separates businesses entities into two groups:
14 For businesses in STREAMS B and D, the EAS collection uses information collected by other ABS surveys that share similar concepts but ask a variety of state/territory based questions. Where the EAS collection finds data from one of these 'donor' surveys for a business, the state/territory proportions collected from the donor survey are used to allocate the business's data across the states/territories.
- those businesses identified as operating in only one state or territory ('single state' businesses) and
- those businesses identified as operating across more than one state and/or territory ('multi-state' businesses).
15 It is from these donor surveys that the majority of state/territory proportions are produced. In descending order of relevance, the ABS collections used to obtain state/territory proportions for sales and wages and salaries were the following:
16 The EAS attempts to match businesses to these ABS donor collections. If matched to more than one collection, the most relevant collection as per the list above takes precedence. Sales proportions obtained for each business were used to allocate EAS total income, total expenses and profit data across the states/territories for that business. Similarly, wages and salaries proportions were used to allocate EAS wages and salaries data across states/territories for that business. It is worth noting that some bias may arise from obtaining state/territory dissections from various ABS collections with different reporting periods, definitions and scope. To understand these differences more clearly, please refer to the relevant publications' Explanatory Notes.
Construction industry survey
Quarterly business indicator survey
17 The ABSBR includes a 'States of operation' field to identify the states/territories in which each business operates. For businesses in STREAMS B and D for which no donor information is available, this field is used to apportion the data. Where a unit is classified as multi-state, the state/territory proportions are assigned based on an industry average.
18 Businesses in STREAM T are assumed to operate in a single state/territory, and are treated accordingly. For other businesses that are classified as single state, all activity is assigned to that state/territory. Single state/territory units from STREAM T comprise the majority of units and represent almost 15% of total income. STREAM D has the fewest units and makes the highest contribution to total income.