|Page tools: Print Page|
28.4 Data can also be cross-classified by: State/Territory; sector (public/private); level of government; industry (4 digit ANZSIC); employer size; sex; full-time/part-time; adult/junior; managerial/non-managerial; permanent/temporary/casual; status of employee (working proprietor, managerial/executive, supervisor, apprentice/trainee, other); and occupation (4 digit ASCO).
28.5 Data on how pay is set were first collected in the May 2000 survey and include: proportion of employees whose pay is set by each pay setting mechanism; and average weekly total earnings (by pay setting mechanism).
28.6 Data are compiled according to the concepts and definitions outlined in Chapters 4 (employees) and 11 (earnings).
28.7 For the first stage sample of employing businesses (MUSs), the standard scope exclusions for ABS labour-related business surveys (outlined in Chapter 23) apply to this survey.
28.8 The scope of the second stage sample is restricted to civilian employees based in Australia who received payments during the survey reference period. Self-employment jobs (such as proprietors/partners of unincorporated businesses), jobs for which payments were not made during the reference period (such as jobs held by persons on unpaid leave), jobs based outside Australia, and members of the Australian permanent defence forces are not in scope of the survey.
28.9 Detailed information is obtained about a sample of employees from each selected business using a mail-out/mail-back collection methodology.
28.10 The survey reference period is the last pay period ending on or before the third Friday in May of the survey year. Businesses are asked to include only one week's proportion of hours paid and related earnings if their payroll is not weekly.
28.11 Businesses which do not mail back their completed questionnaire within a reasonable period of time after the survey reference date are followed up by mail and then phone if necessary. The response rate for the 1998 collection was 98.5% at the MUS level and 97.4% at the employee level.
28.12 A probability sample design is used. The sample of employees is obtained using a stratified two stage selection approach which involves: first, the ABS selecting a sample of businesses from the ABS Business Register; and second, each selected business selecting a sample of employees from their payroll(s).
28.13 The selection unit for the first stage selection is the Management Unit/State (MUS). The collection and reporting units used in the survey usually correspond to the selection unit. However, where the MUS unit is unable to provide information required for the survey, it may be split into a number of 'reporting units'.
28.14 Second stage sampling units (employees) are selected using systematic sampling within the selected first-stage units. For further information on statistical units used in ABS business surveys refer to Chapter 23.
STAGE ONE SELECTION
28.15 A probability sample of employing businesses (MUS units) is drawn from the ABS Business Register using the process outlined in Chapter 23. Variables used to stratify the survey frame at stage one of the sample selection are:
28.16 Strata on the survey frame that are completely enumerated include those containing MUSs with benchmark employment greater than a set cutoff (this cutoff will vary for different States/Territories) and strata with a very small number of MUSs. Strata which are completely enumerated because they contain a low number of MUSs may become sampled strata if the number of MUSs in those strata increases sufficiently.
28.17 In addition to constraints outlined in Chapter 23, sample selection is constrained by the need to ensure that there is maximum overlap with the quarterly Survey of Average Weekly Earnings and minimum overlap with other surveys.
STAGE TWO SELECTION
28.18 Second stage sampling units (employees) are selected using systematic sampling within selected first-stage units. A random start and a skip are provided to each selected business for use in selecting the second-stage sample from its payroll(s). Units with an unknown employment size are given a skip of one. Businesses are not required to order their payroll in any particular way when selecting the sample.
SAMPLE SIZE AND ALLOCATION
28.19 For the first-stage sample, approximately 8,000 MUSs are selected to yield a live sample of approximately 6,800 MUSs.
28.20 For the second-stage sample, approximately 59,000 employees are selected. The maximum number of employees for any reporting unit is 40.
28.21 The ABS reselects the sample for the Survey of Employee Earnings and Hours each time it is conducted. At the same time the overall design of the sample is examined to ensure that it remains efficient and cost-effective.
28.22 The estimation process occurs in two stages. In the first stage, number raised estimation is used to estimate the number of employees in each business. In the second stage, number raised estimation is again used to estimate the total number of businesses, and therefore employees, in the target population.
28.23 In the sampled strata, the Live Respondent Mean method is used to impute for missing data items. In the completely enumerated strata a ratio imputation model is used.
28.24 Survey outliers are handled using the 'surprise outlier' technique.
28.25 New Business Provisions were introduced in the 2000 survey. Adjustments are made to survey estimates to allow for births and deaths of businesses that have occurred up to the end of the survey reference period but which are not reflected on the survey frame.
28.26 For further information on estimation methods used in ABS Business Surveys, refer to Chapter 23.
TIME SERIES ESTIMATES
28.27 Seasonally adjusted and trend estimates are not produced for this survey.
RELIABILITY OF THE ESTIMATES
28.28 Estimates from the survey are subject to both sampling and non-sampling error. The relative standard errors of survey estimates are published in both Employee Earnings and Hours, Australia, Preliminary (Cat. no. 6305.0) and Employee Earnings and Hours, Australia (Cat. no. 6306.0).
28.29 The ultimate cluster variance estimation technique is used to calculate estimates of variance for this survey. For further information on the ultimate cluster variance estimation technique, or on sampling and non-sampling error, refer to Chapter 16.
DATA COMPARABILITY OVER TIME
28.30 Although many estimates from previous years or collection cycles are on a consistent basis, care should be taken in using data from this survey on a time series basis. The survey is designed to give an accurate 'snapshot' of data rather than an ongoing series of observations over time. Hence the survey methodology and sample design are not specifically set up to provide time series data e.g. the sample is not set up to have the same employee jobs in it for more than one year. Nevertheless, average weekly earnings and hours data at aggregate levels are compiled in a consistent manner over time, although successive estimates would not be strongly correlated because of low sample overlap. The following changes have been made to survey methods, survey concepts, data item definitions, and frequency of collection.
28.31 For further details contact the Assistant Director, Labour Employer Surveys Section, on Perth (08) 9360 5240.