TECHNICAL NOTE 1 ESTIMATION METHODOLOGY
INTRODUCTION
1 The estimates in this publication have been sourced from a combination of data directly collected from businesses as part of the Construction Industry Survey conducted by the ABS, and business income tax data provided to the Australian Tax Office (ATO).
COLLECTION DESIGN
2 In order to decrease the statistical reporting load placed on providers while maintaining the range of information available to users of statistical data, the strategy for this survey was to adopt the use of both directly collected data from a smaller sample of providers and data from the ATO. Under taxation law, data may be passed by the Commissioner for Taxation to the ABS for specified statistical purposes.
3 Businesses were only selected to participate in the Construction Industry Survey (i.e. the direct collect sample) if they met a threshold level of activity. Very small businesses with annual income less than the threshold were not eligible for selection. Thresholds were set to account for 85% of total turnover in each ANZSIC subdivision.
4 In order to minimise the statistical reporting load even further for very small businesses, the direct collect population was divided into two streams. Businesses with employment of less than four persons completed a 'short form' (i.e. not all detailed income and expense items were collected from these businesses) while all other businesses completed a form with finer level income and expense items.
ESTIMATION METHODOLOGY
5 Large and/or complex businesses were completely enumerated by the Construction Industry Survey. Estimates for these businesses were formed by adding together the values reported to the survey. ATO business income tax data (BIT) was not used for these businesses due to difficulty in identifying all of the ABNs that were used by large and complex businesses.
6 For all other businesses that could be readily matched to BIT data, the strategy was to use this data source for broad economic data and apply proration techniques for data items not available from ATO.
7 This twofold approach was required to overcome some of the limitations in using ATO data. These limitations included: the non-availability, on the BIT file, of detailed income and expense items required for the compilation of the national accounts; the inability to identify all legal entities associated with large businesses or some small and medium employing businesses; and the absence of tax exempt businesses on the income tax file.
Two Stage Proration
8 Detailed income and expense items directly collected from the Construction Industry Survey were used to produce a series of 'factors' to prorate broad economic data items, such as sales, into finer components, such as contracting and subcontracting income. Two stages of proration were used.
First stage
9 One set of proration factors was estimated within the direct collect population using the full detailed income and expense items from businesses with employment between 5-19 persons. These factors were then applied to produce the detailed data items not directly collected from businesses with employment of less than four persons that were sent the short form. At the end of this stage each business that responded to the Construction Industry Survey had been assigned values for the full range of Construction Industry Survey data items.
Second stage
10 The survey sample was used to produce a second set of proration factors. These factors were used to produce any data items that were not included on the BIT file. Only broad income and expense items were available from the BIT file.
11 At the end of this second stage each business in the Construction Industry Survey population (i.e. direct collect sample and ATO populations) held data for the full range of Construction Industry Survey data items. These data were then aggregated and adjusted to account for survey non-response, ABNs that did not match to business income tax data, and other sources of error such as newly established businesses that were not fully represented on the ABS Business Register.
12 For more information, contact Geoff Frost on (03) 9615 7787
Print Page
Print All