Flooding in New South Wales and Queensland in late February 2022 resulted in a major disruption to the operation of the Labour Force Survey. As a result, there was lower than usual numbers of responding households in some of the affected regions in March 2022.
Given the severity of these disruptions and to ensure that this loss of sample did not affect data for Australia, New South Wales and Queensland, the ABS imputed some sample within the following 15 statistical area level 4 (SA4s) regions for March 2022:
- Central Coast
- Coffs Harbour - Grafton
- Richmond - Tweed
- Sydney - Baulkham Hills and Hawkesbury
- Sydney - Blacktown
- Sydney - Inner South West
- Sydney - Outer West and Blue Mountains
- Sydney - South West
- Brisbane - South
- Brisbane - West
- Brisbane Inner City
- Darling Downs - Maranoa
- Logan - Beaudesert
This imputation approach drew upon previous information that had recently been collected from people in the affected areas, following the approach used for February 2019 for the Townsville Flood. For March 2022, the ABS also used some hours-related information for populations in affected areas and adjacent areas to account for the impact on hours worked (which is where labour market impacts from floods, bushfires and cyclones are generally observed, more than changes in employment or unemployment).
For March 2022, 2.9% of the sample was imputed using this approach (5.6% of the sample in New South Wales and 9.2% of the sample in Queensland).
The imputation approach will be re-assessed once April data has been collected from the affected areas. For the Townsville Flood, the ABS re-assessed the imputation approach used for February 2019 in the following month, and confirmed that revisions were not required.
Data for the above SA4s for March 2022 will not be published in Labour Force, Australia, Detailed on the 21st April. The ABS expects to resume publishing these estimates with the release of April data in May 2022.
Users of the matched sample analysis should also exercise some caution when looking at New South Wales and Queensland data between February and March, given the effect of the imputation for these two states.