Comparing the 2021 Census and the Labour Force Survey
Explaining the different ways labour force data is used and collected
Introduction
The Census and the Labour Force Survey (LFS) both collect information about the labour market activity of people aged 15 years and over. While both collections seek to measure concepts related to employment, unemployment and being outside of the labour force, there are a number of differences between them. This paper outlines the strengths and key uses of each collection; how the collections differ; and explains why the statistics produced in each of these collections are not directly comparable. It also provides advice on how Census and LFS data can be used together in the analysis of change over time. Information is also provided on how the impacts of COVID-19 on the Australian labour market may be reflected in Census and LFS data.
Strengths and key uses of the Census and LFS
The LFS produces the most authoritative estimates of labour market information, including employment and unemployment. Labour force statistics are published monthly by the ABS in Labour Force, Australia. The LFS is designed specifically to measure changes over time in the Australian labour force, and to provide a high quality measure for use in international comparisons. It provides a highly accurate estimate of key labour force statistics of the Australian economy, including employment, unemployment and underemployment, as well as a range of more detailed labour market-specific data. The LFS is the leading source of data for monitoring Australia’s labour market conditions.
The Census provides a rich snapshot of all people living in Australia on Census Night. It is the leading source of information for small population groups and areas and allows for the analysis of labour market activities and industry and occupation data at a more detailed level. The Census also collects information about a range of characteristics of people, including, but not limited to, their labour force status, enabling analyses across a broader range of socioeconomic dimensions.
| Census | Labour Force Survey | |
|---|---|---|
| Purpose | Counts the number of people living in Australia on Census Night and the dwellings in which they live. Also collects information about a range of characteristics of people, including their labour force status. | Monthly household survey providing Australia's official estimates of employment and unemployment. It is the leading source of data for monitoring Australia’s labour market conditions. |
| Frequency | Conducted every five years. Provides a single snapshot of Australia on Census Night. | Monthly collection with respondents in the sample for 8 consecutive months. Provides a continuous time series of data on the Australian labour force for historical and international comparisons. |
| Reference period | The week prior to the Census Night. | The week prior to when the monthly survey is conducted. |
| Collection methodology | A self-completed online or paper form. | Telephone interview, face-to-face interview by highly trained interviewers, and self-completed online enumeration. |
| Survey questions | Questions cover a broad range of topics. Labour force status determined from four simple questions. Questions are used to ascertain if a person is working, actively looking for work and available to start work. | Questions mainly focused on measuring labour force status. Labour force status determined from an extensive range of questions, producing a highly accurate classification of people as employed, unemployed or not in the labour force, at a given point in time. |
| Sampling error | Not subject to sampling error as it aims to collect the information from everyone in Australia on Census Night. However, the Census is subject to some level of undercount. | Results are subject to sampling error as estimates based on information obtained from a sample of dwellings. The sample is designed to ensure sampling error is reduced to a minimum at the national and state/territory levels. However, it can be higher for labour market regions (SA4 data) or for detailed breakdowns. The estimates are therefore accompanied by information on the quality of the estimates, including relative standard errors. |
| Treatment of non-response | To account for unreturned Census forms, demographic characteristics of people in non-responding households are either imputed or included in the 'not stated' category. Labour force status is not imputed and data is not adjusted for non-responding households. Issues with response or coverage are identified through the Post Census Review which is conducted a few weeks after the Census to estimate the number and characteristics of people either not counted or counted multiple times on Census Night. | Only fully responding households contribute to the estimates. Non-responding households are treated as 'not stated' and excluded and adjusted for through the weighting process. As a sample survey, it is weighted to an independent population benchmark based on the Estimated Resident Population (ERP), which ensures estimates add up to an independently estimated distribution of the usual resident civilian population aged 15 years and over, regardless of any sample lost due to non-response. |
COVID-19 impacts on labour market data
The COVID-19 pandemic resulted in a period of unprecedented disruption to the labour market. Impacts can be seen across both Census and LFS data. COVID-19 related lockdowns were in place on Census Night, as well as in the weeks before and after Census Night. This has impacted data on a range of labour market indicators, such as the number of people working from home, the number of hours people worked (i.e. a person may have been stood down but still had a job, and therefore worked reduced or no hours that week) or whether someone considered themselves to be employed or unemployed during lockdown conditions.
As the Census is a self-completed form, a person’s response may differ between the Census and the LFS, where highly trained interviewers assist the person to complete the LFS form and ask a range of questions to determine if someone was employed. For example, if someone was away from their job for any reason (such as being stood down while in lockdown), they would still be counted as employed in the LFS if they were paid for some or part of the previous four weeks. This is similar to when someone may be on paid annual leave. In this scenario, they would be employed but worked zero hours for that week.
There were instructions on the Census website and online form on how to respond to questions on employment while in lockdown. The instructions were that if the person was employed in the last four weeks prior to lockdown, but was unable to work in the last week due to lockdown, select 'Yes, but absent on holidays, on paid leave, on strike or temporarily stood down'. Regardless, given the complexities in interpretation of employment during lockdown, and the level of uncertainty and pace of change at the time, there is likely an element of difference in response between the Census and LFS. These differences may be reflected in broader labour market measures such as hours worked, and industry and occupation information.