6602.0 - Microdata: Longitudinal Labour Force, Australia , 1982-2019 Quality Declaration 
ARCHIVED ISSUE Released at 11:30 AM (CANBERRA TIME) 06/12/2019   
   Page tools: Print Print Page Print all pages in this productPrint All

Using DataLab

DataLab allows real time access to detailed microdata files through a portal to a secure ABS environment. Using detailed microdata in DataLab allows users to run advanced statistical analyses using recent analytical software.

For information about the data items available on the detailed microdata files, see the Data Item List in the Downloads tab.

About DataLab

Detailed microdata files in DataLab can be accessed on-site at ABS offices or in a secure virtual environment from your own computer. All unit record data remains in DataLab, and any analysis results or tables are checked by the ABS before being provided to users.

For more information, including prerequisites for DataLab access, please see the About DataLab page.

Detailed microdata test file

A test file is available as a free download in the Downloads tab.

The test file mimics the structure of the detailed microdata in that it has the same data items and allowed values. This allows users to become familiar with the data structure and prepare code/programs before applying for or beginning a DataLab session.

All data on the test file is false, created through a randomisation process and therefore cannot be used for analysis.

Counts and weights

The Longitudinal LFS (LLFS) includes 434 monthly cross sections covering the period from October 1982 to August 2019 (Table 1). On average each cross section has close to 62,000 observations but the size of the cross sections has been closer to 50,000 since late 2009. Between July 2008 and September 2009, the sample size was temporarily reduced to be around 40,000. The size of cross sections generally declines over time reflecting improvements in sampling methodology.

Each cross section is used to produce the headline LFS statistics. The weights provided with the LLFS ensure that each cross-section reproduces the statistics found in the LFS.

Table 1: Summary of monthly cross sections(a)



(a) Excluding the missing cross sections for May 1983, June 1983, November 1983, January 1984, September 1984, December 1984, January 1985, July 1985 and July 1986.

For more information about the file structure, see the Data and file structure page.

For more information about the weights, see the Weighting and Benchmarks section of the Data Item List, available in the Downloads tab.

Unit identifiers

Every record of the file is uniquely identified by the item ABSRID. This identifier is a combination of the household identifier ABSHID and the person identifier ABSPID. All of these identifiers are used consistently across months to create longitudinal links.

Non-private dwellings and dwellings selected in indigenous communities are excluded from longitudinal linking. These include hotels, motels, hostels, hospitals, religious institutions providing accommodation, educational institutions providing accommodation, prisons, boarding houses, and short-stay caravan parks. These are given a new household identifier each month in this dataset. People in non-private dwellings are more likely to be older and not in the labour force than those in private dwellings.

Family units are identified by the item FAMNUM. This identifier is not used consistently across months due to the dynamic nature of family relationships. Its purpose is to identify family units within multiple family households for the particular circumstances of each month. The identification of family units in one month may not necessarily correspond to how family units are identified in subsequent or preceding months.

For more information about the identifiers, see the Record Identifiers section of the Data Item List, available in the Downloads tab.

Longitudinal analysis

The Labour Force Survey (LFS) is designed to survey the same household for eight consecutive months. However, this means that individuals can move in and out of the LLFS for a number of reasons:
  • they do not complete the survey in a month,
  • they are visiting the household,
  • they move house permanently, or
  • they were in a non-private dwelling.

While over 5 million individuals are observed in total, only 1.8 million are observed for the full eight months (Table 2).

Table 2: Counts of individuals

Number of responses

Non-private dwellings

Some individuals are more likely to leave the LFS than others.
  • Males are more likely to leave relative to females
  • Younger individuals are more likely to leave relative to older individuals
  • Unemployed individuals are more likely to leave relative to employed individuals or those not in the labour force
  • Individuals in regional areas are more likely to leave relative to those in capital cities
  • Single individuals are more likely to leave relative to married individuals
  • People who were born overseas are more likely to leave than those born in Australia
  • Those visiting a household are more likely to leave than those that live in the household

When linking individuals, this variability in the types of people who leave and stay in the LFS results in attrition bias overtime. It is important when looking at aggregate statistics that use linked observations to be aware of this bias and to attempt to control for it if possible. This can be done by adjusting the weights appropriately - increasing the weights for those who are more likely to leave the LFS.

It is also important to consider how the collection of the LFS has changed over time. The method of data collection has changed to allow a greater choice in how people response, from face-to-face interviews to phone interviews to online self-completion.