Labour Force, Australia methodology

1 This publication contains estimates of the civilian labour force derived from the Labour Force Survey component of the Monthly Population Survey. The full time series for estimates from this publication are also available electronically. More detailed estimates are released one week after this publication in various electronic formats - see Labour Force, Australia, Detailed - Electronic Delivery (cat. no. 6291.0.55.001) and Labour Force, Australia, Detailed, Quarterly (cat. no. 6291.0.55.003).

2 The conceptual framework used in Australia’s Labour Force Survey aligns closely with the standards and guidelines set out in Resolutions of International Conferences of Labour Statisticians. Descriptions of the underlying concepts and structure of Australia's labour force statistics, and the sources and methods used in compiling the estimates, are presented in Labour Statistics: Concepts, Sources and Methods (cat. no. 6102.0.55.001) which is available on the ABS website http://www.abs.gov.au.

3 The Labour Force Survey is based on a multi-stage area sample of private dwellings (currently approximately 26,000 houses, flats, etc.), a list sample of non-private dwellings (hotels, motels, etc.), and covers approximately 0.32% of the civilian population of Australia aged 15 years and over.

4 Information is obtained from the occupants of selected dwellings by specially trained interviewers using computer-assisted interviewing, or self-completion online.

5 Households selected for the Labour Force Survey are interviewed each month for eight months, with one-eighth of the sample being replaced each month. The first interview is generally conducted face-to-face. Subsequent interviews are conducted by telephone (if acceptable to the respondent).

6 From December 2012 to April 2013, the ABS conducted a trial of online electronic data collection. Respondents in a single rotation group (i.e. one-eighth of the survey sample) were offered the option of self completing their labour force survey questionnaire online instead of via a face-to-face or telephone interview. From May 2013, the ABS expanded the offer of online electronic collection to 50% of each new incoming rotation group. For more information see the article in the April 2013 issue of this publication. From September 2013, online electronic collection has been offered to 100% of private dwellings in each incoming rotation group. From April 2014, 100% of private dwellings are being offered online electronic collection.

7 The interviews are generally conducted during the two weeks beginning on the Sunday between the 5th and 11th of each month. The information obtained relates to the week before the interview (i.e. the reference week). Each year, to deal with the operational difficulties involved with collecting and processing the Labour Force Survey around the Christmas and New Year holiday period, interviews for December start four weeks after November interviews start (i.e. between the 3rd and 9th December), and January interviews start five weeks after December interviews start. As a result, January interviewing may commence as early as the 7th or as late as the 13th, depending on the year. Occasionally, circumstances that present significant operational difficulties for survey collection can result in a change to the normal pattern for the start of interviewing.

8 Estimates from the Labour Force Survey are usually published first in this publication 39 days after the commencement of interviews for that month, with the exception of estimates for each December which are usually published 46 days after the commencement of interviews.

9 The Labour Force Survey includes all people aged 15 years and over except members of the permanent defence forces, certain diplomatic personnel of overseas governments customarily excluded from census and estimated population counts, overseas residents in Australia, and members of non-Australian defence forces (and their dependants) stationed in Australia.

10 In the Labour Force Survey, coverage rules are applied which aim to ensure that each person is associated with only one dwelling, and hence has only one chance of selection. The coverage rules are necessarily a balance between theoretical and operational considerations. Nevertheless, the chance of a person being enumerated at two separate dwellings in the survey is considered to be negligible.

11 The Labour Force Survey estimates are calculated in such a way as to add to independent estimates of the civilian population aged 15 years and over (population benchmarks). These population benchmarks are based on the most recently released estimates of Final, Revised and Preliminary quarterly Estimated Resident Population (ERP). For information on the methodology used to produce the ERP see Australian Demographic Statistics (cat. no. 3101.0). Since the most recently released ERP estimates lag the current time period for Labour Force estimates by nine months, the Labour Force population benchmarks are created by projecting forward three quarters past the most recently released quarterly ERP estimates. The projection is based on the historical pattern of each population component - births, deaths, interstate migration and net overseas migration (NOM). Estimates of NOM are supplemented with other data sources to better reflect short-term population changes.

12 The revision status of quarterly ERP data changes over time from, preliminary, to revised, to final, as natural increase, overseas migration, and interstate migration component data is revised to incorporate more up to date data. These revisions flow through to the population benchmarks used to rebenchmark the Labour Force estimates on a quarterly basis.

13 Every five years, the ERP series are revised to incorporate additional information available from the latest Census of Population and Housing (Census). Labour Force Survey population benchmarks, and the estimates, are revised following this five-yearly revision in the ERP. The process of incorporating the revised population benchmarks is referred to as 'rebasing'. The rebasing process is subject to a revision going from, preliminary rebasing approximately a year after a census, to final rebasing approximately 2 years after a census. Preliminary rebasing occurred in the November 2017 issue of Labour Force, Australia (cat. no. 6202.0), at the same time revisions were made to historical Labour Force estimates from July 2011 to October 2017. The final rebasing occurred in November 2018 Labour Force, Australia - Rebenchmarked Estimates, Oct 2018 (cat. no. 6202.0.55.003) with the series from July 2011 to November 2018 rebenchmarked to align with final ERP benchmarks from the 2016 Census. For more information on revised ERP estimates, refer to the December 2016 issue of Australian Demographic Statistics (cat. no. 3101.0) released in June 2017.

14 In between Censuses, the ABS revises the Labour Force population benchmarks using the latest ERP according to the paragraphs above. These were introduced in the July 2010, November 2012 and April 2013 issues. The revisions planned for the October 2013, April 2014 and November 2014 issues were not implemented (see What's New in the Labour Force in the September 2013 issue and Changes in this Issue in the October 2014 issue of this publication). From the February 2015 issue, rebenchmarking were undertaken quarterly in the February, May, August and November issues, apart from May 2015. For more information, refer to the article Rebenchmarking of Labour Force Series in the February 2015 issue of this publication.

15 The estimation method used in the Labour Force Survey is Composite Estimation, which was introduced in May 2007. In January 2014 composite estimation was applied to all estimates from July 1991 as part of the 2011 Census rebenchmarking. Composite Estimation combines data collected in the previous six months with current month's data to produce the current month's estimates, thereby exploiting the high correlation between overlapping samples across months in the Labour Force Survey. The Composite Estimator combines the previous and current months' data by applying different factors according to length of time in the survey. After these factors are applied, the seven months of data are weighted to align with current month population benchmarks. For details see Information Paper: Forthcoming Changes to Labour Force Statistics, 2007 (cat. no. 6292.0).

16 From April 1986, the definition of employed people was changed to include people who worked without pay between 1 and 14 hours per week in a family business or on a farm (i.e. contributing family workers). For further information, see paragraphs 22 and 23 of the Explanatory notes section in the February 2003 issue of Labour Force, Australia (cat. no. 6203.0).

17 The ABS introduced telephone interviewing into the Labour Force Survey in August 1996. Implementation was phased in for each new sample group from August 1996 to February 1997. During the period of implementation, the new method produced different estimates than would have been obtained under the old methodology. The effect dissipated over the final months of implementation and was no longer discernible from February 1997. The estimates for February 1997 and onwards are directly comparable to estimates for periods prior to August 1996. For further details, see the feature article in the June 1997 issue of Labour Force, Australia (cat. no. 6203.0).

18 From April 2001 the Labour Force Survey was conducted using a redesigned questionnaire containing additional data items and some minor definitional changes. The definition of unemployed people was changed to include all people who were waiting to start work and were available to start in the reference week. This change was introduced in February 2004, when historical unit record data were revised from April 2001 to January 2004. This revision created a small trend break at April 2001 in unemployed people and unemployment rate series. For further details, see Information Paper: Forthcoming Changes to Labour Force Statistics, 2003 (cat. no. 6292.0), released in December 2003. From July 2014 the Labour Force Survey questionnaire was further redesigned and definitional changes made to active job search steps and duration of job search. For further details, see the Glossary and Information Paper: Forthcoming Changes to Labour Force Statistics, (cat. no. 6292.0), released in October 2014.

19 Core labour force series were revised in April 2001 for the period April 1986 to March 2001 for the remaining definitional changes introduced with the redesigned questionnaire, to reduce the impact of the changes on labour force series. For further details, see Information Paper: Implementing the Redesigned Labour Force Survey Questionnaire (cat. no. 6295.0) and the 2004 issue of Information Paper: Questionnaires Used in the Labour Force Survey (cat. no. 6232.0).

20 In May 2007, an improved method of estimation, known as composite estimation, was introduced into the Labour Force Survey. In introducing this change, the ABS revised unit record data from April 2001 to April 2007 based on the new estimation method. No change was identified in the trend breaks in the unemployed people and unemployment rate series which arose with the introduction of a redesigned survey form in April 2001 (as noted above in paragraph 18). In January 2014 composite estimation was applied to all estimates from July 1991 as part of the 2011 Census rebenchmarking. For further details, see Information Paper: Forthcoming Changes to Labour Force Statistics, 2007 (cat. no. 6292.0).

21 As one of a range of ABS savings initiatives for the 2008-09 financial year, there was a 24% reduction in the LFS sample size for the period July 2008 to August 2009, relative to the June 2008 sample size. The sample reduction was reversed from September 2009 to December 2009, with December 2009 estimates being the first produced under the fully reinstated sample.

22 The current Labour Force Survey sample has been selected using information from the 2016 Census of Population and Housing and the ABS Address Register.

23 Between July 2018 and February 2019, a new sample was phased in over eight months with the incoming rotation group being selected from the ABS Address Register.

24 Updated sampling fractions were also phased in with the introduction of the ABS Address Register selected sample. The state sampling fraction method applies a predefined skip interval for each state which defines the probability of an address within each respective state being selected for the survey. Each incoming rotation group that was phased in had this new state skip applied.

25 Detailed information on the new sample, including the use of the ABS Address Register, can be found in Information Paper: Labour Force Survey Sample Design, July 2018 (cat. no. 6269.0), which was released on 30 July 2018.

26 Two types of error are possible in an estimate based on a sample survey: sampling error and non-sampling error.

27 Sampling error occurs because a sample, rather than the entire population, is surveyed. One measure of the likely difference resulting from not including all dwellings in the survey is given by the standard error. There are about two chances in three that a sample estimate will differ by less than one standard error from the figure that would have been obtained if all dwellings had been included in the survey, and about nineteen chances in twenty that the difference will be less than two standard errors. Standard errors of key estimates for the latest month and of movements since the previous month of these estimates are shown in the standard errors section of this publication. Standard errors for other estimates and other movements may be calculated by using the spreadsheet contained in Labour Force Survey Standard Errors, Data Cube (cat. no. 6298.0.55.001) which is available free of charge on the ABS website http://www.abs.gov.au.

28 Non-sampling error arises from inaccuracies in collecting, recording and processing the data. Every effort is made to minimise reporting error by the careful design of questionnaires, intensive training and supervision of interviewers, and efficient data processing procedures. Non-sampling error also arises because information cannot be obtained from all people selected in the survey. The Australian Labour Force Survey receives a higher level of co-operation from individuals in selected dwellings compared to other countries, with the average response rate over the past 3 years being 92.0 per cent, and the average rate over the past year being 91.75 per cent (to the nearest quarter of a per cent, in rounded terms). See Glossary for definition of response rate.

29 Any original time series can be thought of as a combination of three broad and distinctly different types of behaviour, each representing the impact of certain types of real world events on the information being collected: systematic calendar related events, short-term irregular fluctuations and long-term cyclical behaviour. A multiplicative decomposition model is applied in the seasonal adjustment of Labour Force Time Series, where the original time series (O) is considered as the product of the underlying trend (T), a systematic calendar related or seasonal component (S) and an irregular component (I). This can be expressed as O = T*S*I. The contributions of each of these behaviours varies from series to series, as well as throughout time for a given series, depending on the nature of the interactions of real world events and the data of interest.

30 Seasonal adjustment is a statistical technique that attempts to measure and remove the effects of systematic calendar related patterns including seasonal variation to reveal how a series changes from period to period. Seasonal adjustment does not aim to remove the irregular or non-seasonal influences which may be present in any particular month. This means that month-to-month movements of the seasonally adjusted estimates may not be reliable indicators of trend behaviour.

31 The Labour Force Survey uses the concurrent seasonal adjustment method to derive seasonal factors. Concurrent seasonal adjustment uses data up to the current month to estimate seasonal factors for the current and all previous months. This process can result in revisions each month to estimates for earlier periods. However, in most instances, the only noticeable revisions will be to the seasonally adjusted estimates for the previous month and one year prior to the current month. From the March 2015 issue of this publication, the effects of supplementary surveys are removed prior to the estimation of seasonal factors for key Labour Force series from February 1978 onwards. While this methodology has addressed short term volatility in the seasonally adjusted series arising from changes to the timing and content of the supplementary survey program, in general prior corrections and resulting changed seasonal patterns can be identified and measured to a more reliable degree of certainty after three successive observations (in this case after three years). For further details refer to the October and December 2014 issues of this publication.

32 The revision properties of the seasonally adjusted and trend estimates can be improved by the use of Autoregressive Integrated Moving Average (ARIMA) modelling. ARIMA modelling relies on the characteristics of the series being analysed to project future period data. The projected values are temporary, intermediate values, that are only used internally to improve the estimation of the seasonal factors. The projected data do not affect the original estimates and are discarded at the end of the seasonal adjustment process. The Labour Force Survey uses an ARIMA model for the majority of the individual time series. The ARIMA model is assessed as part of the annual reanalysis. For further details, see the feature article in the October 2004 issue of Australian Economic Indicators (cat. no. 1350.0).

33 Seasonal adjustment is able to remove the effect of events which occur at the same time in the survey every year. However, there are some events, like holidays, which are not always at the same time in the survey cycle or which are not at the same time across Australia. The effects of these types of events on Labour Force Survey estimates cannot in all cases be removed, because the pattern of their effects cannot be determined. However, two events for which adjustment is made in the seasonally adjusted series are the January interview start date and the timing of Easter. For further details, see Information Paper: Forthcoming Changes to Labour Force Statistics (cat. no. 6292.0) released in December 2003.

34 While seasonal factors for the complete time series are estimated each month, they will continue to be reviewed annually at a more detailed level to take into account each additional year's original data. This annual review will not normally result in significant changes to published estimates. The review is usually conducted early each year with the results released in this publication shortly thereafter.

35 The smoothing of seasonally adjusted series to produce 'trend' series reduces the impact of the irregular component of the seasonally adjusted series. These trend estimates are derived by applying a 13-term Henderson-weighted moving average to all months except the last six. The last six monthly trend estimates are obtained by applying surrogates of the Henderson average to the seasonally adjusted series. Trend estimates are used to analyse the underlying behaviour of a series over time.

36 While this smoothing technique enables estimates to be produced for the latest month, it does result in revisions in addition to those caused by the revision of seasonally adjusted estimates. Generally, revisions due to the use of surrogates of the Henderson average become smaller, and after three months have a negligible impact on the series

37 Trend estimates are published for the Northern Territory in table 10 and for the Australian Capital Territory in table 11. Unadjusted series for the two Territories have shown, historically, a high degree of variability, which can lead to considerable revisions to the seasonally adjusted estimates each month when seasonal factors are estimated. For this reason, seasonally adjusted estimates are not currently published for the two Territories. In addition, caution should be exercised in the interpretation of trend estimates for the two Territories, particularly for the three most recent months, where revisions may be relatively large.

38 For further information, see A Guide to Interpreting Time Series - Monitoring Trends (cat. no. 1349.0).

39 Users may also wish to refer to the following publications and other data products that are available free of charge from the ABS website:

• Labour Force Survey Standard Products and Data Item Guide (cat. no. 6103.0)
• Labour Force, Australia, Detailed (cat. no. 6291.0.55.001)
• Labour Force, Australia, Detailed Quarterly (cat. no. 6291.0.55.003)
• Labour Force Survey Standard Errors, Data Cube (cat. no. 6298.0.55.001)
• Labour force, Australia - Rebenchmarked Estimates (cat. no. 6202.0.55.003)
• Labour Statistics: Concepts, Sources and Methods (cat. no. 6102.0.55.001)
• Australian Demographic Statistics (cat. no. 3101.0)
• Australian Labour Market Statistics (cat. no. 6105.0)
• Labour Force, Australia (cat. no. 6203.0)
• Information Paper: Outcomes of the Labour Household Surveys Content Review (cat. no. 6107.0)
• Information Paper: Questionnaires Used in the Labour Force Survey (cat. no. 6232.0)
• Information Paper: Regional Labour Force Statistics (cat. no. 6262.0)
• Information Paper: Labour Force Sample Design (cat. no. 6269.0)
• Information Paper: Forthcoming Changes to Labour Force Statistics (cat. no. 6292.0)
• Information Paper: Expansion of Hours Worked Estimates from the Labour Force Survey (cat. no. 6290.0.55.001)
   

40 Current publications and other products released by the ABS are listed on the ABS website http://www.abs.gov.au. The ABS also issues a daily Release Advice on the website which details products to be released in the week ahead.

41 As well as the statistics included in this and related publications, the ABS may have other relevant data available. Inquiries should be made via email client.services@abs.gov.au or contact the National Information and Referral Service on 1300 135 070.

42 Estimates have been rounded and discrepancies may occur between sums of the component items and totals. Estimates of movement shown in this publication are obtained by taking the difference of unrounded estimates. The movement estimate is then rounded. Where a discrepancy occurs between the reported movement and the difference of the rounded estimates, the reported movement will be more accurate.

The estimates in this publication are based on information gained from the occupants of a sample survey of dwellings. Because the entire population is not surveyed, the published original, seasonally adjusted and trend estimates are subject to sampling error. The most common way of quantifying such sampling error is to calculate the standard error for the published estimate or statistic. For more information, see paragraphs 25 to 27 of the Explanatory notes section.

To illustrate, let us say the published level estimate for employed people aged 15-19 years is 700,000 and the associated standard error is 9,000. The standard error is then used to interpret the level estimate of 700,000. For instance, the standard error of 9,000 indicates that:

• There are approximately two chances in three that the real value falls within the range 691,000 to 709,000 (700,000 + or - 9,000)
• There are approximately nineteen chances in twenty that the real value falls within the range 682,000 to 718,000 (700,000 + or - 18,000).
   

The real value in this case is the result we would obtain if we could enumerate the total population.

The following table shows the standard errors for this month's level estimates.

The following example illustrates how to use the standard error to interpret a movement estimate. Let us say that one month the published level estimate for females employed part-time in Australia is 1,890,000; the next month the published level estimate is 1,900,000 and the associated standard error for the movement estimate is 11,900. The standard error is then used to interpret the published movement estimate of 10,000. For instance, the standard error of 11,900 indicates that:

• There are approximately two chances in three that the real movement between the two months falls within the range - 1,900 to 21,900 (10,000 + or - 11,900)
• There are approximately nineteen chances in twenty that the real movement falls within the range - 13,800 to 33,800 (10,000 + or - 23,800).
   

The following table shows the standard errors for this month's movement estimates.