DIRECTIONS FOR ABS LABOUR HOUSEHOLD SURVEYS
A number of changes are planned for the core monthly and quarterly LFS, the labour supplementary surveys, and the labour MPHS topics. These proposed changes are discussed below. As noted, the exact details may change subject to testing and an assessment of statistical impact.
THE LABOUR FORCE SURVEY
A number of developments are planned in relation to the content of the LFS that will improve conceptual robustness and relevance of the survey, improve the frequency and timeliness of key data, or enable critical labour market linkages that are not currently possible.
Measures of underemployment
Underemployment is increasingly being recognised as a key measure of spare capacity in the labour market, and a significant economic and social issue. In the late 1970s and 1980s, the underemployment rate was considerably lower than the unemployment rate, and there were approximately half as many underemployed people as unemployed people. In more recent times, the number underemployed workers has been significantly higher than the number of unemployed.
Currently, underemployment estimates are released quarterly. They are also not available to the same level of detail as unemployment estimates.
From July 2014, the ABS will increase the reporting frequency of the underemployment rate and labour force underutilisation rate from quarterly to monthly. The underemployment and underutilisation information will be integrated into LFS output so it accompanies unemployment information.
In addition to the changes identified above, the ABS also plans to increase the scope of the underemployment measures to include all employed people. The current approach to underemployment statistics only asks people who are employed part-time if they are wanting, and available, to work more hours. The change will ask all employed people about whether they want, and are available, to work more hours.
Volume measures of labour underutilisation
In addition to headcount underemployment and underutilisation measures, the ABS produces annual volume measures of underemployment and underutilisation, and releases them in Australian Labour Market Statistics (cat. no. 6105.0). Volume measures relate to the quantum of unused potential hours of labour, and are compiled using information collected in the Job Search Experience survey in July, the Underemployed Workers survey in September and the LFS in August. They are often more relevant for analysing the spare capacity of the labour force than headcount measures, as they take into account the number of hours sought and additional hours preferred by individuals.
From July 2014, the ABS will increase the frequency of the volume measures of labour underutilisation from annual to quarterly, and change the source from the Job Search Experience and Underemployed Workers supplementary surveys to the quarterly module of the LFS.
There are strong linkages between educational attainment and labour market outcomes, particularly earnings. Educational attainment is also important in the context of understanding productivity and the quality of labour inputs.
At present, educational attainment information is not collected in the LFS, and is only available from the MPHS and selected labour supplementary surveys. While the Survey of Education and Work (SEW) provides a comprehensive source of educational attainment information cross-classified with key labour force indicators, it is only conducted annually and does not include other labour market and employment-related characteristics.
From July 2014, the ABS will include educational attainment in the LFS on a monthly basis. This information will be collected from respondents who are in the 'incoming rotation group' of the survey, and will be 'rolled forward' for subsequent months (that is, respondents will only be asked about their educational attainment in their first month in the survey, with that information carried over for the respondent's subsequent seven months in the survey). For people who are currently studying (whose educational attainment may change over the course of their participation in the LFS), information will be sought throughout the LFS period on changes to their educational attainment. Educational attainment information will therefore be available in the monthly LFS product set, as well as in all labour supplementary surveys and MPHS topics.
The nature of employment in Australia has become more diverse over the past two decades, with growth in forms of employment other than the 'traditional' arrangement of full-time, ongoing wage or salary job, with regular hours and paid leave. One such form of employment is casual work. Casual employment is of particular interest because of concerns about the working conditions, career opportunities and job security of casual workers compared with other employees.
There is no single measure of casual employment, given the myriad of contexts in which casual work can be considered. However the most objective (and commonly used) measure of casual employment relates to the entitlement to paid holiday and sick leave, i.e. employees who do not have an entitlement to paid holiday leave or paid sick leave are regarded as casual. Other measures, such as whether someone identifies as a casual, or whether they have a guaranteed minimum number of hours, are also useful, albeit more subjective.
At present, entitlements to paid holiday leave and paid sick leave are collected in a number of supplementary surveys. However, it is not possible to easily produce a coherent time series as the data are contained in different datasets and in some cases based on only subsets of the employed population.
To improve the coherence of the time series of data available, and to increase the prominence and visibility of the data, entitlements to paid holiday and paid sick leave will be collected in the LFS on a quarterly basis, and incorporated into the LFS product set.
Public and private sector information
Following the discontinuation of the private sector component of the Survey of Employment and Earnings (SEE) in 2002, the ability for the ABS to provide a simple measure of 'sector of employment' has been limited. While some information can be obtained from the EEBTUM supplementary survey or other economic surveys, the review found significant unmet demand for robust estimates of sector of employment from a range of users.
From 2014, the ABS will include public and private sector employment estimates each quarter in the LFS product set. The information required to produce this information is already collected on a quarterly basis as part of the collection and coding of industry data, and will be added to the processing and compilation of LFS estimates.
There are currently two main employment classifications used in labour statistics: Status in Employment and Employment Type. The LFS uses the Status in Employment classification for its standard output, while the labour supplementary surveys (and other social surveys) predominantly use the Employment Type classification.
Status in Employment is necessary in the context of national accounting and the measurement of income, as 'compensation of employees' (the largest component of gross domestic product) is based on the System of National Accounts definition of 'employee'. However, it does not provide the most useful representation for analysis of the labour market.
The Employment Type classification is considered preferable for most labour market analyses. Unlike Status in Employment, Employment Type aims to capture the fundamental nature of employment, that is, whether a person works for an employer or operates their own business, regardless of the legal status of that business. This means that in the Status in Employment classification, people who operate their own incorporated enterprise (Owner Managers of Incorporate Enterprises - OMIEs) are included in the Employees category, whereas in the Employment type category, this group is identified separately within the Employee category.
To overcome the complexities, and potential misinterpretation or comparison of data, the ABS will be revising the Status in Employment classification so that it provides a single labour market relevant classification that can meet all uses. This will be achieved by removing OMIEs from the Employees category and having them separately identified. As a result, the Employment Type classification will no longer be required.
While the Status in Employment classification will be different, there will effectively be no break in LFS series as the categories will be conceptually consistent and able to be aggregated or disaggregated to match the previous version of the Status in Employment classification and the Employment Type classification.
Information on retrenchments is critical to understanding the performance of the labour market, particularly during economic downturns. Currently, retrenchment information is available from the two-yearly Labour Mobility supplementary survey. However the data are limited due to the frequency of the survey (every 2 years) and scope of retrenchments (during the 12 month period up to the survey).
Retrenchment data for the unemployed population can also be obtained on a quarterly basis from the LFS, although it is not available in any standard product. The article Retrenched Unemployed People in the April 2010 issue of Australian Labour Market Statistics (cat. no. 6105.0) provided an analysis of retrenchment using LFS data, and was well-received by key users. However, it could not include all retrenchments because the LFS does not currently identify people who had been retrenched and either subsequently found another job or are now not in the labour force.
To address the need for a comprehensive and frequent measure of retrenchment, information on retrenchment will be collected every quarter in the LFS and incorporated into the LFS product set. The quarterly retrenchment information will inform on all people aged 15 years and over who had been retrenched or made redundant in the past three months, irrespective of their current labour force status.
Duration with current employer/business
Information on the continuous duration with a current employer/business is currently collected from employed people in the LFS on a quarterly basis. While a considerable amount of detail is captured on the durations of less than one year, no detail is collected on durations of more than one year. Only FOES collects more detailed breakdowns of periods over one year (e.g. single years between 1 and 20 years, and 'More than 20 years').
From July 2014, the ABS will collect more detailed information on duration with current employer/business in the LFS on a quarterly basis. These data will be released as part of the quarterly LFS product set as well as the labour supplementary surveys.
Full-time and part-time job search
The LFS collects information from unemployed people on whether they are looking for full-time or part-time work. However, the current sequencing of the survey questionnaire results in people being asked if they are looking for part-time work only if they indicate that they have not been looking for full-time work. As a result, people who have looked for both full-time work and part-time work are not identified.
From July 2014, the ABS will change the sequencing of the LFS questions on full-time or part-time job search so that all unemployed people are asked whether they are looking for either full-time work, part-time work, or both.
Number of jobs in the economy
While estimates of employment from the LFS refer to counts of employed people rather than the number of jobs, there is continued interest from users and commentators in the number of jobs in the labour force, particularly in relation to full-time and part-time jobs. Multiple job holding is the main reason why estimates of employment from the LFS cannot be equated to estimates of jobs. The LFS does capture the existence of multiple job-holding on a monthly basis, through identifying people with more than one job, but this information is not included in any standard product.
Using the information that is already collected, the ABS intends to release more information from the LFS on the number of jobs in the labour market, to complement the measures of total employment.
Improving 'gross flows' data
'Gross flows' shows movements of people between the different labour force status from one month to the next, rather than just the status at a point in time. By considering the sample that responded to the LFS in consecutive months, the labour force status for both months can be compared. The rotating sample design of the LFS means that only about 80% of the sample for any given month can be matched to the corresponding record in the previous month, with the other 20% being people who are new to the sample, leaving the sample, or sample loss. Currently, 'gross flows' data are published using the weights from the 80% of the sample, which means the estimates only cover approximately 80% of the population (13.7 million out of 17.7 million). This limits their usefulness.
The ABS intends to improve the 'gross flow' estimates by making them more representative of the full sample and publish more information on the changing nature of the Labour Force sample.
Longitudinal Labour Force Survey Confidentialised Unit Record File
The review found that there is increasing demand from users for access to microdata to analyse people's transitions into and out of employment and unemployment, rather than solely using the standard LFS products or customised requests relating to point-in-time estimates. The ABS released a longitudinal Labour Force and Supplementary Surveys Confidentialised Unit Record File (CURF) in December 2012. This CURF provides a rich source of detailed labour market dynamics data, and will enable users and researchers to analyse the longitudinal microdata to gain a better understanding of people's movements over time in the labour market.
The CURF is available through the ABS's Remote Access Data Laboratory, with the first release of the CURF containing data for the months January 2008 to December 2010. Updates to the CURF are expected be released on an annual basis.
Aggregate monthly hours worked
The development of the 'aggregate monthly hours worked' series has been one of the most important developments in the LFS in recent years. Users have found the information invaluable in better understanding labour productivity and the performance of the labour market and economy, particularly during economic downturns. Currently the data are available for full-time/part-time status by sex; state; and industry sector.
The review found a strong demand for further disaggregations of aggregate monthly hours worked data, particularly full-time/part-time status by state by sex; and by industry division. As a result, aggregate monthly hours worked will be expanded to include a number of new series to meet this demand.
Monthly industry estimates
Data relating to industry of employment is currently available on a quarterly basis from the LFS. The review found strong user demand for industry information to be available on a monthly, rather than quarterly, basis. The current approach to collecting industry information on a quarterly basis could not be applied to the monthly collection, as it would result in even more volatile series, present an unacceptable burden on respondents, and is likely to have prohibitive costs.
Alternative approaches to producing industry data on a monthly basis from the LFS are being investigated, and the ABS will develop a strategy for monthly industry estimates in consultation with key users.
LABOUR SUPPLEMENTARY SURVEYS
A key outcome of the review is the consolidation of the labour supplementary surveys program into two surveys per year - run in February and August. One of the main considerations during the review was to design a program that was more coherent, and maximised the value of the information collected by ensuring that, as far as practicable, all data relevant to an area of the labour market was collected at the same time, and included within the same dataset. The proposed content for the two supplementary surveys has been developed based on an assessment of the subject matter and an analysis of user requirements, and will result in a more streamlined, less fragmented, program of labour supplementary surveys.
The changes are not expected to add significantly to respondent burden or to the average interview times in the respective months. This is due to reduced duplication in monthly content, the mutually exclusive nature of respondent groups within each of the surveys, and the dropping of some lower priority content. The overall outcome is expected to result in an overall reduction in respondent burden over the survey program.
The content of the two supplementary surveys are outlined in more detail below. The final content will be subject to testing and an assessment of statistical impact, to ensure that the changes to the surveys will not have an adverse impact on respondent burden or on LFS estimates.
The August survey: Characteristics of Employment
The August supplementary survey (Characteristics of Employment) will describe the key features of people's employment. It will integrate the Employee Earnings, Benefits and Trade Union Membership (EEBTUM), Forms of Employment (FOES), Working Time Arrangements (WTA) and Labour Hire (LH) surveys into one.
The Locations of Work (LOW) survey, which is a six-yearly module also collected in conjunction with FOES, will be ceased as some information on locations of work is collected on the same frequency in the Survey of Employment Arrangements, Retirement and Superannuation (SEARS).
The survey will inform users on the following broad labour market issues:
- earnings - in main job and all jobs
- employment arrangements
- trade union membership
- non-standard employment
- independent contracting
- work scheduling
- job security
- job flexibility
Including this range of topics into a single survey will result in a dataset that will enable analysis of the full range of key elements of people's employment and related outcomes. While EEBTUM, FOES and WTA have all collected information describing people's employment, each only provided part of the picture for understanding people's employment and related outcomes.
While the content of the August survey will largely be based on existing content, there are some changes that will be made to better inform on important issues. In particular, the survey will incorporate information on overwork, which was regarded as a high priority by many users during the consultation.
Due to the large amount of planned content, and to reflect that some of the data items are not required at an annual frequency, the August supplementary survey will comprise a core annual component, comprising around 2/3 of the overall content, with the remaining content included every two years on a rotating basis. The core component will contain question modules on characteristics of employment such as earnings, employment arrangements, job flexibility and trade union membership. The first of the two biennial modules will contain more detailed questions on earnings, trade union membership, independent contracting, and labour hire. The second of the two biennial modules will contain more detailed questions on job flexibility, stability and scheduling.
The February survey: Participation, Job Search and Mobility
The February supplementary survey (Participation, Job Search and Mobility) will inform on unemployment, underemployment, labour force participation, job change and job search. It will integrate the key elements of the Labour Mobility, Job Search Experience (JSE), Underemployed Workers (UEW) and Persons Not in the Labour Force (PNILF) surveys into one. Labour Force Experience (LFE), which is currently run every two years alternating with Labour Mobility, will be ceased as the release of the longitudinal LFS CURF will enable better analysis of labour force transitions.
The survey will inform users on the following broad labour market issues:
- job change
- job mobility
- job search
- participation and increasing participation
- marginal attachment.
Including this range of topics into a single survey will result in a dataset that will enable analysis of people's experiences relating to job search, job change and increasing participation. While JSE, UEW, PNILF and Labour Mobility have all collected information describing people's transitions within and out of the labour force, each only provided part of the picture for understanding labour force participation and related outcomes.
While the content of the February survey will largely be based on existing content, there are some changes that will be made to better inform on important issues. In particular, the survey will expand information collected on job search (and job churn), by adding to information already collected on job search by the unemployed, by asking whether employed people looked for a job (and if so, the steps taken) and asking those not in the labour force who looked for work the steps they took. It will also provide more comprehensive information on geographic mobility, by broadening the scope of the data items on whether people would be prepared to move interstate or intrastate for a job to the unemployed, employed and some people not in the labour force.
Impact of moving supplementary survey collection months
The move to create two consolidated supplementary surveys will involve a change in collection month for a number of the existing supplementary survey topics, namely:
- JSE, UEW and PNILF will move from their current July or September months, to February.
- FOES, WTA and LH will move from their current November month to August
There is a risk that changes to the collection month and consolidation of the JSE, PNILF and UEW surveys will impact on comparisons of the data over a number of years, due to the potential for seasonal variations in the experiences of some of these groups.
In order to better understand the impacts that the change in timing will have, all three surveys (JSE, PNILF and UEW) will be conducted in February 2014 in a format that is similar to their current format. This means that PNILF and UEW will be conducted in September 2013 and then again in February 2014. In February 2015, the new Participation, Job Search and Mobility survey will be begin.
The ABS also intends to provide information from available data sources - particularly the LFS - to assist users in understanding the impact of changing months, to support the transition from July and September to the February reference period. This will assist users to more easily determine the differences that may be attributed to changes in collection month.
LABOUR MULTIPURPOSE HOUSEHOLD SURVEY TOPICS
While the review was primarily focussed on the content of the LFS and labour supplementary surveys, it also considered the MPHS topics, including the inter-relationships between the topics, their timing and their frequency. Apart from some small improvements to the Barriers and Incentives to Labour Force Participation topic, the content and frequency of the labour MPHS topics will remain largely unchanged.
Barriers and Incentives to Labour Force Participation
The scope of Barriers and Incentives to Labour Force Participation is currently restricted to people who are working 15 hours or fewer per week (in addition to those who are not currently working), as they have been considered to be the group who could provide the largest additional contribution (in hours) to labour supply. Because people working more than 15 hours are also likely to want to work additional hours - and account for the majority of underemployed people - the scope of the topic will be expanded to provide a more comprehensive picture to inform on increasing participation amongst those who are already working.
From the next survey in 2012-13, the ABS has increased the scope to people working fewer than 35 hours per week (in addition to those who are not currently working).
Currently, information on incentives for increased participation is only collected every second time the Barriers and Incentives to Labour Force Participation survey is conducted (i.e. every four years). However, it is considered a fundamental part of the survey and worth being collected on a more frequent basis. The ABS will therefore include this information as a core component of the Barriers and Incentives to Labour Force Participation MPHS topic, collecting the data every two years from 2012-13.
The increase in scope and frequency may require a slight reduction in overall survey content in 2014-15. Investigations on the impact and resulting content will be progressed in the development of the 2014-15 survey.
OTHER POTENTIAL DEVELOPMENTS THAT ARE NOT BEING PROGRESSED
A range of other issues and data gaps were raised during the user consultation and were considered in the review, but are not being progressed by the ABS at this stage. These are discussed briefly below.
Residency and visa status
Temporary migrants may make a significant contribution to parts of the labour market, and there is demand to obtain more robust and frequent estimates of the labour force characteristics of this group, particularly in the context of labour shortages and the government's skilled migration programs. Information on the type of visa possessed by recent migrants is collected in the Characteristics of Recent Migrants supplementary survey, which is conducted on a three-yearly basis. However, there is no more frequent measure of this group.
Consideration was given to collecting information on residency and visa status in the LFS. While this is an important issue, concerns about the effect that collecting this information would have on the LFS, as well as the suitability of the Any Responsible Adult methodology, mean that this issue will not be pursued any further in considering the content of the survey program. Other strategies for supporting this information need will be investigated in due course.
The review considered the need for increasing the frequency of the annual families estimates released from the LFS. While it is acknowledged there is user demand for more frequent families estimates, this would require the production of the families benchmarks on a more frequent basis, for which funding and resources are currently unavailable. As a result of this, and the relatively stable nature of family dynamics, the families estimates will continue to be released on an annual basis.
Other areas that came up in the review, but will not be progressed relate to the collection of information on:
- methods of setting pay;
- dollar value of casual loadings;
- more frequent earnings data;
- journey to work and fly-in fly-out work;
- skill-based underemployment; and
- job and skills mismatch.
While it is recognised that these are important issues and would add value to the survey program, concerns about accurate data collection, as well as relative priorities amongst the other requirements, mean that these were not able to be incorporated into the proposed program at this time.