1504.0 - Methodological News, Mar 2017  
ARCHIVED ISSUE Released at 11:30 AM (CANBERRA TIME) 22/03/2017   
   Page tools: Print Print Page Print all pages in this productPrint All

SAMPLE REDESIGNS FOR QBIS AND CAPEX

New sample designs have recently been implemented for the Quarterly Business Indicators Survey (QBIS) and the Survey of Private New Capital Expenditure and Expected Expenditure (Capex). These surveys were last redesigned in 2009, hence there was significant scope to improve the efficiency of the sample designs by reflecting changes to the population of in-scope businesses that have occurred since then.

Business Statistics Methodology (BSM) conducted a range of investigations in the first half of 2016 to create the new sample designs. Both surveys remain stratified by State, Industry, Size, and Employment status (whether or not a business employs staff), but within this framework there have been several improvements. Some of these improvements are as follows:

  • The size dimension of the stratification remains principally based on a business’s number of employees, but there have been improvements to the treatment of units with small employment size but very large turnover. These units are now identified and moved to a bigger size group using a method consistent with that used for the Economic Activity Survey, which collects similar data on an annual basis. This replaces a less effective method that was used under the previous designs.
  • ABS business surveys typically have a ‘Completely Enumerated’ (CE'd) sector, where all businesses above a certain size cut-off are guaranteed selection in the survey. For these new designs an industry-dependent CE'd cutoff was implemented in place of the previous single cutoff. This meant the CE'd cutoffs could be increased for some industries. This reduces the provider burden for businesses that no longer fall above the cut-offs while maintaining the quality of the survey outputs.
  • The industry dimension of the stratification was rationalised to achieve much closer alignment in industry stratification between the two surveys. This will assist with survey consolidation initiatives which may occur in conjunction with the ABS’ Statistical Business Transformation Program.
  • Micro non-employing units remain out of scope of the surveys. These are non-employing units with turnover below certain industry dependent cutoffs. These cutoffs have been updated to account for inflation and changes to the economy since the last redesign.
  • Public sector businesses remain out of scope of both surveys but the derivation of the variable used to identify public and private sector businesses has been improved.
  • Due to population changes since the last redesign, the weights of selected businesses in some strata had increased undesirably. Sample was optimally allocated to the new stratification to meet specified relative standard error constraints for the key survey variables, while ensuring no individual unit received a weight of greater than 400.

BSM supported the implementation and quality assurance of the new designs, which took effect for the September 2016 reference period.


Further Information
For more information, please contact Carrie Samuels (methodology@abs.gov.au).

The ABS Privacy Policy outlines how the ABS will handle any personal information that you provide to us.