Characteristics of Employment, Australia methodology

1 The statistics in this publication were compiled from information collected in the Characteristics of Employment (COE) survey conducted throughout Australia in August 2019 as a supplement to the Australian Bureau of Statistics' (ABS) monthly Labour Force Survey (LFS). Respondents to the LFS who fell within the scope of the supplementary survey were asked further questions.

2 Information about survey design, scope, coverage and population benchmarks relevant to the monthly LFS, which also applies to supplementary surveys, can be found in the publication Labour Force, Australia (cat. no. 6202.0).

3 The conceptual frameworks used in the monthly LFS align closely with the standards and guidelines set out in Resolutions of the International Conference of Labour Statisticians. Descriptions of the underlying concepts and structure of Australia's labour force statistics, and the sources and methods used in compiling these estimates, are presented in Labour Statistics: Concepts, Sources and Methods (cat. no. 6102.0.55.001).

4 The conceptual framework for measures of mean and median earnings aligns closely with the standards and guidelines set out in the System of National Accounts 2008, and Resolutions of the International Conference of Labour Statisticians.

5 The scope of the LFS is restricted to people aged 15 years and over and excludes the following people:

• members of the permanent defence forces;
• certain diplomatic personnel of overseas governments, customarily excluded from the Census and estimated populations;
• overseas residents in Australia; and
• members of non-Australian defence forces (and their dependants).
   

6 Students at boarding schools, patients in hospitals, residents of homes (e.g. retirement homes, homes for people with disabilities), and inmates of prisons are excluded from all supplementary surveys.

7 This supplementary survey was conducted in both urban and rural areas in all states and territories, but excluded persons living in Aboriginal and Torres Strait Islander communities.

8 In addition to those already excluded from the LFS, contributing family workers, persons not in the labour force and unemployed persons were also excluded.

9 The estimates in this publication relate to persons included in the survey in August 2019. In the LFS, 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 in the survey. See Labour Force, Australia (cat. no. 6202.0) for more details.

10 Supplementary surveys are not always conducted on the full LFS sample. Since August 1994 the sample for supplementary surveys has been restricted to no more than seven-eighths of the LFS sample.

11 This survey is based on the new sample introduced into LFS in July 2018. The new sample design has adopted the use of the Address Register as the sampling frame for unit selection, and the sampling fractions for selection probabilities within each state have been updated to reflect the most recent population distribution based on results from the 2016 Census of Population and Housing. As with each regular sample design, the impacts on the data are expected to be minimal. For more information, see the Information Paper: Labour Force Survey Sample Design, Jul 2018 (cat. no. 6269.0).

12 Estimates in this publication are subject to sampling and non-sampling errors:

• Sampling error is the difference between the published estimate and the value that would have been produced if all dwellings had been included in the survey. For more information, see the Technical Note.
• Non-sampling errors are inaccuracies that occur because of imperfections in reporting by respondents and interviewers, and errors made in coding and processing data. These inaccuracies may occur in any enumeration, whether it be a full count or a sample. Every effort is made to reduce the non-sampling error to a minimum by careful design of questionnaires, intensive training and effective processing procedures.
   

13 The estimates are based on information collected in the survey month (August) and, due to seasonality, may not be representative of other months of the year. For example, the numbers of employees working on weekdays and weekends will be representative for an August month but not necessarily representative of all months in the year.

14 To reduce the impact of seasonality on total employment, the estimates have been adjusted by factors based on trend LFS estimates. These factors were applied at the State and Territory, Sex, Full-time and Total employment levels, based on the trend LFS series as published in the September 2019 issue of Labour Force, Australia (cat. no. 6202.0), published 17 October 2019. This adjustment accounts for August seasonality and irregular effects, resulting in an increase to the typically lower original employed estimates for August.

15 In the August 2017 issue of Characteristics of Employment, Australia (cat. no. 6333.0), historical estimates re-published from surveys conducted in different survey months (May and November) will be subject to different seasonal impacts, which may result in an observable break in series between the historical data and data collected in COE. Trend factors have also been applied to these historical estimates to reduce the impact of seasonality on total employment estimates.

16 Country of birth data are classified according to the Standard Australian Classification of Countries (SACC), 2011 (cat. no. 1269.0).

17 Occupation data, including skill level of main job, are classified according to ANZSCO - Australian and New Zealand Standard Classification of Occupations, 2013, Version 1.2 (cat. no. 1220.0).

18 Industry data are classified according to the Australian and New Zealand Standard Industrial Classification (ANZSIC), 2006 (Revision 2.0) (cat. no. 1292.0).

19 Education data are classified according to the Australian Standard Classification of Education (ASCED), 2001 (cat. no. 1272.0).

20 Geography data are classified according to the Australian Statistical Geography Standard (ASGS), 2011 (cat. no. 1270.0.55.001).

21 Where information relating to earnings in both main job and/or second job was not provided by the respondent, values have been imputed. In August 2019, there were 4,440 cases where information relating to earnings in main job was not provided by the respondent and 191 cases where information relating to earnings in second job was not provided by the respondent. Where this was the only information missing from the respondent record, the value was imputed based on answers provided from another respondent with similar characteristics (referred to as the "donor"). Donor records were selected for imputation of earnings in main job by matching information on sex, age, state or territory of usual residence and selected labour force characteristics (full-time or part-time in main job, industry, occupation, hours worked in main job, owner manager status) of the person with missing information.

22 Donor records were selected for imputation of earnings in second job by matching information on age, state or territory of usual residence, area of usual residence, owner manager status, hours worked in second job and frequency of pay in second job. Depending on which values were imputed, donors were chosen from the pool of individual records with complete information for the block of questions where the information was missing.

23 Estimates relating to mean and median weekly earnings generally exclude owner managers of incorporated enterprises (OMIEs) unless otherwise stated. Employees who only received payment in kind were also excluded.

24 Estimates relating to mean and median hourly rate generally exclude owner managers of incorporated enterprises (OMIEs) unless otherwise stated. Employees who only received payment in kind or worked zero hours while on workers compensation were also excluded.

25 Employees have been classified as 'With paid leave entitlements' if they were entitled to paid sick leave and/or paid holiday leave. In all other cases, employees have been classified as 'Without paid leave entitlements' and are also referred to as “casual employees”.

26 The LFS estimates and estimates from the supplementary surveys, (e.g. COE) are calculated in such a way as to sum to the independent estimates of the civilian population aged 15 years and over (population benchmarks). These population benchmarks are updated quarterly based on Estimated Resident Population (ERP) data.

27 From August 2015, Labour Force Estimates have been compiled using population benchmarks based on the most recently available release of ERP data, continually revised on a quarterly basis. At the time of publication, this issue's estimates are comparable with the published labour force estimates for August 2019.

28 From August 2017, the estimates in this publication have moved to regular rebenchmarking to reflect the latest revisions to ERP data and updated trend LFS estimates.

29 In August 2017, historical estimates re-published from previous supplementary surveys have been revised to reflect the latest benchmarks and trend LFS estimates for employment (as at November 2017). These include estimates from the previous surveys:

• Employee Earnings, Benefits and Trade Union Membership, Australia, August 2004-13 (cat. no. 6310.0) (EEBTUM)
• Forms of Employment, Australia, November 2008-13 (cat. no. 6359.0) (FOES)
• Working Time Arrangements, Australia, November 2006, 2009 and 2012 (cat. no. 6342.0) (WTA)
• Education and Work, Australia, May 2004-05 and 2007-12 (cat. no. 6227.0). (SEW)
   

30 For the historical estimates re-published from surveys run in months other than August, two kinds of estimates have been produced.

• For estimates relating to the number of employees or number of employed persons, estimates have been revised based on the full set of respondents who completed the survey for that month.
• For estimates relating to weekly earnings and hourly rates, the data are based on the respondents who remained employed, remained in sample, and provided information in both the non-August survey and the nearest next or previous August survey. This allowed for the earnings information collected in August to be merged with the complementary data collected in the other months (for example, merging August EEBTUM earnings with the highest non-school qualification data collected in the May SEW). Relative Standard Errors (RSEs) for these estimates are higher than usual, with less than half of the full sample common between both surveys.
   

31 When comparing results from the 2018 and 2019 issues of COE to previous surveys, it is recommended to use the revised and re-published estimates provided within the current issue. In previous publications, caution should be exercised when comparing results, as the definition of employees is not always directly comparable to the current definition. Changes to the employee definition involved excluding Owner Managers of Incorporated Enterprises (OMIEs) and including persons who worked for a commission only without a retainer. In this publication, time series of employee estimates are presented on a consistent basis.

32 From August 2014 collection of earnings in second job was changed to match the collection of earnings in main job. Previously, earnings in second job was collected from respondents who were employees in their second job who actually worked some hours in their second job in the reference week. Earnings were reported for those hours actually worked in that job. From 2014, earnings in second job were collected from employees in their second job regardless of whether they worked in that job in the reference week. Earnings data and frequency of pay in that second job were subsequently collected. This change will result in a break in series of earnings in all jobs and earnings in second job. Caution should be exercised when comparing second and all job earnings data from COE with previous EEBTUM data.

33 Prior to 2014, information about trade union membership was collected only of employees and owner managers or incorporated enterprises. From 2014 onwards, information on trade union membership is collected from all employed people.

34 For information on the history of changes to EEBTUM, see the Explanatory notes section (cat. no. 6310.0).

35 For information on the history of changes to FOE, see the Explanatory notes section (cat. no. 6359.0).

36 The estimates of earnings in this publication are produced in accordance with the conceptual framework for measures of employee remuneration, as outlined in Information paper: Changes to ABS Measure of Employee Remuneration, Australia 2006 (cat. no. 6313.0).

37 From 2007, as a result of a change in the concept of earnings being measured, employees and OMIEs were asked to include salary sacrifice when estimating their earnings. In previous years, there was no explicit reference to the treatment of salary sacrifice. It is probable that some employees were already including amounts of salary sacrifice in their estimates of earnings, depending upon how their pay was reported. This change has resulted in a break in series. Users need to exercise care when comparing the earnings of employees and OMIEs in this release with those prior to 2007.

38 From 2017, additional information relating to the hourly rate and the skill level of main job were added to the imputation process for main job earnings.

39 From 2014, additional information relating to the number of hours usually worked and the frequency of pay in a respondent's second job were added to the imputation process for second job earnings.

40 From 2009, additional information relating to the number of hours that a respondent's last pay period covered in their main job was added to the imputation process for main job earnings.

41 Aside from the changes listed above, the current imputation method has been used since the 2005 survey. A similar method of imputation was used for the 2004 survey. The differences between the 2004 and the current imputation method are that donors are matched, where possible at a finer level of detail; and second job earnings are imputed whereas in 2004 they were not.

42 Prior to 2004, imputation was not used. Employees whose weekly earnings could not be determined were excluded from estimates of mean or median weekly earnings. Care should be taken when comparing earnings data from 2004 onwards with earnings data prior to 2004. To compare the change in methodology from 2003 to 2004 see paragraph 28 of the Explanatory notes section  in the August 2004 Employee Earnings, Benefits and Trade Union Membership (cat. no. 6310.0).

43 Due to differences in the scope and sample size of this supplementary survey and that of the monthly LFS, the estimation procedure may lead to some small variations between labour force estimates from this survey and those from the LFS.

44 Caution should be exercised when comparing estimates of earnings in this release with estimates of earnings included in the biannual Average Weekly Earnings, Australia (cat. no. 6302.0) and two-yearly Employee Earnings and Hours, Australia (cat. no. 6306.0) publications. The data in both these publications are compiled from employer based surveys. There are important differences in the scope, coverage and methodology of these surveys which can result in different estimates of earnings from each survey.

45 The survey of Average Weekly Earnings (AWE) collects information from employers who provide details of their employees' total gross earnings and their total number of employees. The survey of Employee Earnings and Hours (EEH) collects information about weekly earnings and hours paid for, and the individual characteristics of a sample of employees within each selected employer unit. Both AWE and EEH are completed by employers with information from their payroll. However, for COE and EEBTUM, respondents are either the employed person or another adult member of their household who responds on their behalf. Where earnings are not known exactly an estimate is reported. There are also scoping differences between both household and employer surveys. For example, AWE and EEH exclude employees in the Agriculture, forestry and fishing industry, and also employees of Private households, whereas these employees are included in the COE and EEBTUM surveys.

46 The earnings series from AWE historically excluded amounts salary sacrificed. However, since the May 2011 AWE publication, the Average Weekly Cash Earnings (AWCE) series have also been released. These series are inclusive of salary sacrificed amounts. The key earnings series from AWE have continued to be published on the old conceptual basis (i.e. exclusive of amounts salary sacrificed) to maintain long term comparability of the key series. In EEH, the salary sacrificed amounts have been included in the estimates of mean and median weekly earnings from 2006 onwards. From 2007, COE and EEBTUM have included amounts salary sacrificed in the estimates of mean and median weekly earnings.

47 For further information on a number of earning series available from ABS sources, please refer to the feature article Understanding earnings in Australia using ABS statistics published in Australian Labour Market Statistics, July 2014 (cat. no. 6105.0).

48 Similar surveys on weekly earnings have been conducted annually in August since 1975, except in 1991 when the survey was conducted in July, and in 1996 when the survey was not conducted. Prior to COE, weekly earnings were most recently published in Employee Earnings, Benefits and Trade Union Membership, Australia (cat. no. 6310.0) (EEBTUM).

49 Prior to 1999, the EEBTUM publication was titled Weekly Earnings of Employees (Distribution), Australia (cat. no. 6310.0). The change in title reflects the inclusion of employment benefits and trade union membership data previously released in other publications.

50 Results of previous surveys on employment benefits have been published in Weekly Earnings of Employees (Distribution), Australia, August 1997 (cat. no. 6310.0), The Labour Force, Australia, Jan 1995 (cat. no. 6203.0), and Employment Benefits, Australia (6334.0).

51 Information on trade union membership was first collected in a supplementary survey in 1976, again in 1982, then biennially in its current format from 1986 to 1990. Between 1992 and 2013, it was conducted annually (with only limited data available every second year). Prior to COE, results of previous surveys were published in EEBTUM and in Trade Union Members, Australia (cat. no. 6325.0).

52 Limited data on trade union membership have also been published in:

• Employment Arrangements, Retirement and Superannuation, Australia, April to July 2007 (Re-issue) (cat. no. 6361.0);
• Weekly Earnings of Employees (Distribution), Australia, August 1997 (cat. no. 6310.0);
• Working Arrangements, Australia, November 2000 (cat. no. 6342.0); and
• Working Arrangements, Australia, November 2003 (cat. no. 6342.0).
   

53 Information on trade union membership provided from an annual census of trade unions is available in the following reports between 1891 and 1996:

• Labour and Industrial Branch Report, 1891-1912 (cat. no. 6101.0);
• Labour and Industrial Branch Report, 1911-1921 (cat. no. 6101.0);
• Labour Report, 1922-1973 (cat. no. 6101.0);
• Labour Statistics, 1975-1978 (cat. no. 6101.0);
• Labour Statistics, Australia 1979-1997 (cat. no. 6101.0);
• Trade Union Statistics, Australia, 1969-1996 (cat. no. 6323.0);
   

54 Information on Forms of employment was originally collected every 3 years between 1998 and 2004, followed by surveys in 2006 and 2007. In 2008, the survey was redeveloped to better capture information of independent contractors, other business operators and employees, and was collected annually on this basis until 2013. Results of previous surveys were published in the final issue of Forms of Employment, Australia, November 2013 (cat. no. 6359.0).

55 Information on Working Arrangements has been collected in a variety of surveys since 1976, as follows:

• Work Patterns of Employees, November 1976 (cat. no. 6328.0 / Ref. No. 6.70)
• Evening and Night Work, November 1976 (cat. no. 6329.0)
• Working Hours Arrangements, Australia, February to May 1981 (cat. no. 6338.0)
• Alternative Working Arrangements, Australia, 1982 and 1986 (cat. no. 6341.0)
• Working Arrangements, Australia, August 1993 and 1997 (cat. no. 6342.0)
• Working Arrangements, Australia, November 2000 and 2003 (cat. no. 6342.0)
• Working Time Arrangements, Australia, November 2006, 2009 and 2012 (cat. no. 6342.0)
   

56 A number of Datacubes (spreadsheets) containing all tables produced for this publication are available from the Data downloads section of the publication. The Datacubes present tables of estimates and their corresponding Relative Standard Errors (RSEs).

57 For users who wish to undertake a more detailed analysis of the data, the survey microdata will be released through the TableBuilder product. For more details, refer to the TableBuilder information, Microdata: Characteristics of Employment, Australia (cat. no. 6333.0.00.001). For more information see About TableBuilder.

58 Special tabulations are available on request. Subject to confidentiality and sampling variability constraints, tabulations can be produced from the survey incorporating data items, populations and geographic area selections to meet individual requirements. These can be provided in printed or electronic form. All enquiries should be made to the National Information and Referral Service on 1300 135 070.

59 The next survey will be conducted in August 2020 and will contain information on trade union membership, independent contractors and employment found through an employment agency or labour hire firm. Data on overwork, job flexibility, working patterns and locations of work will not be collected in August 2020.

60 ABS surveys draw extensively on information provided freely by individuals, businesses, governments and other organisations. Their continued cooperation is very much appreciated: without it, the wide range of statistics published by the ABS would not be available. Information received by the ABS is treated in strict confidence as required by the Census and Statistics Act, 1905.

61 Current publications and other products released by the ABS are available from the Statistics Page on the ABS website. The ABS also issues a daily Release Advice on the website which details products to be released in the week ahead.

62 As estimates have been rounded, discrepancies may occur between sums of the component items and totals.

1 Since the estimates in this publication are based on information obtained from occupants of a sample of dwellings, they are subject to sampling variability. That is, they may differ from those estimates that would have been produced if all dwellings had been included in the survey. One measure of the likely difference is given by the standard error (SE), which indicates the extent to which an estimate might have varied by chance because only a sample of dwellings was included. There are about two chances in three (67%) that a sample estimate will differ by less than one SE from the number that would have been obtained if all dwellings had been included, and about 19 chances in 20 (95%) that the difference will be less than two SEs.

2 Another measure of the likely difference is the relative standard error (RSE), which is obtained by expressing the SE as a percentage of the estimate.

RSE% = (SE/estimate ) x 100

3 RSEs for Characteristics of Employment estimates have been calculated using the Jackknife method of variance estimation. This process involves the calculation of 30 'replicate' estimates based on 30 different sub-samples of the original sample. The variability of estimates obtained from these sub-samples is used to estimate the sample variability surrounding the main estimate.

4 The Excel spreadsheets in the Data downloads section contain all the tables produced for this release and the calculated RSEs for each of the estimates. The RSEs for estimates other than medians have been calculated using the Jackknife method, and RSEs for the medians have been calculated using the Woodruff method.

5 In the tables in this publication, only estimates (numbers, percentages, means and medians) with RSEs less than 25% are considered sufficiently reliable for most purposes. However, estimates with larger RSEs have been included. Estimates with an RSE in the range 25% to 50% should be used with caution while estimates with RSEs greater than 50% are considered too unreliable for general use. All cells in the Excel spreadsheets with RSEs greater than 25% contain a comment indicating the size of the RSE. These cells can be identified by a red indicator in the corner of the cell. The comment appears when the mouse pointer hovers over the cell.

6 RSEs are routinely presented as the measure of sampling error in this publication and related products. SEs can be calculated using the estimates (counts or means) and the corresponding RSEs.

7 An example of the calculation of the SE from an RSE follows. An estimate of males aged 55–59 years who were employed part-time was 81,000, which has an RSE of 7.5%. The SE is:

SE of estimate
= (RSE / 100) x estimate
= 0.075 x 81,000
= 6,100 (rounded to the nearest 100)

8 Therefore, there are about two chances in three that the value that would have been produced if all dwellings had been included in the survey would fall within the range 74,900 to 87,100 and about 19 chances in 20 that the value would fall within the range 68,800 to 93,200. This example is illustrated in the following diagram.

9 Proportions and percentages formed from the ratio of two estimates are also subject to sampling errors. The size of the error depends on the accuracy of both the numerator and the denominator. A formula to approximate the RSEs of proportions not provided in the spreadsheets is given below. This formula is only valid when x is a subset of y.

10 Considering an estimate of 1,532,300 males aged 25-34 years who were employed, 1,334,500 or 87.1% were full-time workers. The RSE for 1,334,500 is 1.2% and the RSE for 1,532,300 is 1.0%. Applying the above formula, the RSE for the proportion who were full-time workers:

\(R S E=\sqrt{(1.2)^{2}-(1.0)^{2}}=0.7 \%\)

11 Therefore, the SE for the proportion who were full-time workers was 0.6 percentage points (= (87.1/100) x 0.7). Therefore, there are about two chances in three that the proportion of full-time workers is between 86.5% and 87.7%, and 19 chances in 20 that the proportion was within the range 85.9% to 88.3%.

12 Published estimates may also be used to calculate the sum of two or more estimates, or the difference between two survey estimates (of numbers, means or percentages) where these are not provided in the spreadsheets. Such estimates are also subject to sampling error.

13 The sampling error of the difference between two estimates depends on their SEs and the relationship (correlation) between them. An approximate SE of the difference between two estimates (x–y) may be calculated by the following formula:

\(R S E\left(\frac{x}{y}\right)=\sqrt{[R S E(x)]^{2}-[R S E(y)]^{2}}\)

14 The sampling error of the sum of two estimates is calculated in a similar way. An approximate SE of the sum of two estimates (x+y) may be calculated by the following formula:

\(S E(x+y)=\sqrt{[R S E(x)]^{2}+[R S E(y)]^{2}}\)

15 For example, an estimate of males aged 55–59 years who were employed part-time was 81,000, and the SE for this estimate was 6,100. For males aged 60-64 years who were employed part-time was 96,700 and the SE was 5,500. The estimate of the combined age group i.e. males aged 55–64 years who were employed part-time is:

81,000 + 96,700 = 177,700

16 The SE of the estimate of males aged 55-64 years who were employed part-time is:

\(S E=\sqrt{(6,100)^{2}+(5,500)^{2}}=8,200\)

17 Therefore, there are about two chances in three that the value that would have been produced if all dwellings had been included in the survey would fall within the range 169,500 to 185,900 and about 19 chances in 20 that the value would fall within the range 161,300 to 194,100.

18 While these formulae will only be exact for sums of, or differences between, separate and uncorrelated characteristics or subpopulations, it is expected to provide a good approximation for all sums or differences likely to be of interest in this publication.and efficient operating procedures.

19 The estimates of means and sums of continuous variables are subject to sampling variability and random adjustment. As for population estimates, the variability due to sampling and random adjustment is combined into the calculated Standard Error, and the Relative Standard Error is reported. The component of variability arising from sampling is calculated using the Jackknife method.

20 The estimates of quantiles such as medians, quartiles, quintiles and deciles are subject to sampling variability and random adjustment. As for population estimates, the variability due to sampling and random adjustment is combined into the calculated Standard Error, and the Relative Standard Error is reported. The component of variability arising from sampling is calculated using the Woodruff method. This is also true for Equal Distribution Quantiles. 

21 A statistical test for any comparisons between estimates can be performed to determine whether it is likely that there is a significant difference between two corresponding population characteristics. The standard error of the difference between two corresponding estimates (x and y) can be calculated using the formula in paragraph 9. This standard error is then used to calculate the following test statistic:

\(\large\left(\frac{x-y}{S E(x-y)}\right)\)

22 If the value of this test statistic is greater than 1.96 then there is evidence, with a 95% level of confidence, of a statistically significant difference in the two populations with respect to that characteristic. Otherwise, it cannot be stated with confidence that there is a difference between the populations with respect to that characteristic.

23 The imprecision due to sampling variability, which is measured by the SE, should not be confused with inaccuracies that may occur because of imperfections in reporting by respondents and recording by interviewers, and errors made in coding and processing data. Inaccuracies of this kind are referred to as non-sampling error, and they occur in any enumeration, whether it be a full count or sample. Every effort is made to reduce non-sampling error to a minimum by careful design of questionnaires, intensive training and supervision of interviewers, and efficient operating procedures.