For Survey 19 the methodology was changed to address the Survey 18 ‘non response’ bias and the use of ABR contact details only bias. The ABR worked closely with the Australian Bureau of Statistics (ABS) to develop a methodology to address these issues and to optimise the accuracy of the survey.
Non response bias: In Survey 18 we had a non response rate of around 80%. The approach taken in Survey 18 was to just keep adding new ABN holders to the survey until we managed to contact the number required to give a statistically valid sample. No conclusions on the accuracy of the details for the 80% we were unable to contact were included in the survey results. In Survey 19 if we could not contact the ABN holder using ABR or ATO details and we saw no signs of business activity (tax returns or activity statements lodged) we treated these as ‘No longer in business’.
ABR Contact details: In Survey 18 we used the contact details on the ABR only. If we could not reach the ABN holder using these details it was treated as a ‘no contact’ and another ABN holder was selected to survey. In Survey 19, we used the ABR contact details and if these were unsuccessful we used alternative contact details from ATO systems.
From Survey 18 to 19, there has also been a shift from the quota based sampling approach used in Survey 18, to the fixed sampling approach in Survey 19 where the fixed sample group of 2,520 was mined extensively and exhaustively to achieve a final outcome for each ABN in the sample provided.
In summary, the introduction of the revised survey methodology in this year’s survey has consistently lowered the estimates on the accuracy of each key data field in comparison to Survey 18. However, these estimates from Survey 19 represent a more accurate estimate of the integrity of the ABR register.
The main survey was undertaken by an external contractor from 7 May to 11 June. In this survey a random stratified sample of ABN holders was invited to undertake a voluntary 10 minute phone survey about their ABN and the information captured on the Register. A total of 1,121 interviews were completed from an overall sample of 2,520 ABN entities, representing an overall response rate of 44%.
Whilst Survey 19 results vary from previous years, the revised methodology has allowed us to present a more accurate assessment of the integrity of the Register. The findings from Survey 19 provide a benchmark for the accuracy of the Register for the future.
The ABN holders we interviewed were across all entity types in the Register and we asked them about their ABN listing, including:
At a glance
Of the 44% of ABN holders who completed this survey:
The survey sample size selected was 2,520. From this sample:
In Survey 19, where survey participants could not be contacted by the external contractor, extensive follow up work was undertaken by the ABR from June to July, utilising ATO systems to contact them.
In the event the ABN holder could not be contacted and signs of business activity were not evident (no tax returns or activity statements lodged) these entities were treated as ‘no longer in business’. A survey outcome was therefore achieved for each ABN in the survey sample.
28% of ABNs are no longer in use – the main reason reported was ‘used ABN for a while but enterprise no longer operating’.