Deciphering different sets of employment data

Greg Connolly and Carmel O’Regan are two of the authors of a Discussion Paper that adddresses the challenges of analysing employment data.

This story was first published on Thursday 12 April 2018. If you wish to use this content, please contact to confirm that the information is still current.

Inconsistencies between different sets of ABS employment data can be a trap for economic analysts and journalists trying to interpret and understand the figures.

This inconsistency prompted a Department of Jobs and Small Business staff discussion paper on Implications of differences in employment estimates by industry between ABS Labour Force Survey and ABS Business-survey data.

Co-author Dr Greg Connolly says an understanding of the reasons behind the differences between the two surveys and which statistics are collected will help ensure better, more informed analysis.

‘There are variations in collection methods and timing and different ways of including or excluding multiple jobholders, short-term temporary entrants working in Australia and child workers (persons aged 14 years and younger),’ Dr Connolly said.

‘These factors combine to make interpretation of labour market data by industry challenging.’

Dr Connolly says that since about the year 2000, there have been large differences in ABS employment data between the two sources for many industries.

For some industries, such as Mining and Information Media and Telecommunications, the Labour Force Survey figures are higher, while for others such as Administrative and Support Services, the business-survey figures are higher.

‘There is the better known household-based Labour Force Survey, but it is also straightforward to calculate a business survey estimate which is based on a combination of the Economic Activity Survey and the Employees and Earnings in the Public Sector survey.

‘Through our paper, we seek to shine a light onto the differences between the surveys and also to offer some insight into the consequences of ignoring these differences when analysing the data. Taking account of these differences is especially important when analysing labour productivity and average compensation of employees by industry.’

‘For some analyses, adjustments can be made for these differences by using the new ABS Labour Account data, first released in the ABS publication, Labour Account Australia, Experimental Estimates, July 2017 (ABS Cat. No. 6150.0.55.001).’

Dr Connolly says that even the definition of employment varies considerably across surveys.

‘The Economic Activity Survey definition is not as comprehensive as the Labour Force Survey,’ Dr Connolly said.

The paper was written by Dr Connolly, Jorge Medina and Carmel O’Regan and is the first in a series to be published regularly on topics related to the Jobs and Small Business portfolio.

Read Staff Discussion Paper Issue Number 1 — Implications of differences in employment estimates by industry between ABS Labour Force Survey and ABS Business-survey data

Feature caption: Carmel O’Regan and Dr Greg Connolly are two of the authors of a Discussion Paper that addresses the challenges of analysing employment data.

Correct at time of publication.