FACT SHEET: EMPLOYMENT OR JOBS - WHAT DOES THE LABOUR FORCE SURVEY MEASURE?
The monthly Labour Force Survey (LFS) is the official source of Australian employment and unemployment statistics. Current estimates of the number of people who are employed, unemployed and not in the labour force, classified by sex, full-time / part-time status, and state and territory are released in the Labour Force, Australia (cat. no. 6202.0) publication each and every month. However, commentators often refer to the increase (decrease) in employment from month to month as the number of jobs created (lost). This is an incorrect inference as estimates of employment from the LFS (an ABS household survey) refer to counts of people rather than jobs.
The LFS is designed to produce estimates of the number of people engaged in economic activity and the definition used aligns closely with international standards and guidelines. The concept of employment used in the LFS (and other ABS household surveys) differs to the concept used in ABS business surveys, where estimates are based on the number of jobs involving paid employment. For example, a person holding multiple jobs with different employers would be counted in ABS household surveys as employed once, but in ABS business surveys would be counted once for each job that they held.
Estimates of the number of employee jobs from ABS business surveys are most commonly compared to estimates of the number of persons in paid employment (referred to as employees) from ABS household surveys. However, estimates of employees from household surveys are not equivalent to estimates of employee jobs from business surveys for the same reason as described above. An example of an ABS business survey which provides estimates of the number of employee jobs (limited to the public sector) is the Survey of Employment and Earnings (SEE). Estimates of employee jobs from Employment and Earnings, Public Sector, Australia (cat. no. 6248.0.55.002) can only be compared to estimates of the number of employees in the LFS, if the differences outlined above are considered and ideally quantified. For the purposes of this comparison, the estimates from the LFS would provide counts of the number of people in employee jobs whereas estimates from the SEE would provide counts of the number of jobs that are occupied by employees. People who appear on more than one payroll are only counted once in the LFS, whereas in the SEE they are counted once for each payroll on which they appear.
The situation of equating people with jobs arises most frequently as a result of commentators using the terms "jobs" and "employment" (or employed people / persons) interchangeably. This loose use of terminology often leads to a misinterpretation and misrepresentation of data. This misuse of data is exacerbated when the discussion includes a distinction between full-time and part-time work.1 Whether commentators refer to the number of or the change in the level of the number of employed persons, it is important to clarify the distinction between full-time employment and a full-time job. A person in full-time employment can hold more than one job (for example, two part-time jobs for which the combined number of hours worked totals 35 hours or more per week), whereas a full-time job represents one person employed full-time.
A number of examples illustrate this:
- if an unemployed person became employed full-time (by starting one full-time job), then the full-time employment estimate from the LFS would increase by one (in a business survey, or a 'jobs' count, this would lead to an increase in the jobs estimate by one);
- if an unemployed person became employed full-time (by starting two part-time jobs with a total of 35 hours of work or more per week), then the full-time employment estimate from the LFS would increase by one (however, in a business survey, or a 'jobs' count, this would lead to an increase in the jobs estimate by two);
- if a person who was already employed in one part-time job took on another part-time job, this would have differing impacts on the employment estimates from the LFS depending on the total number of hours worked: if the sum of hours worked in the two part-time jobs was fewer than 35 hours per week, the employment estimates from the LFS would remain unchanged, but if the sum of hours worked was 35 hours or more, the employment estimates from the LFS would show a decrease of one in part-time employment and an increase of one in full-time employment (however, in both cases this would lead to an increase of one in the jobs estimate from a business survey);
- if a person who was employed in three part-time jobs (working a total of more than 35 hours per week) resigned from these and assumed one full-time job, this would have no impact on the employment estimates from the LFS (however, this would lead to a decrease of two in the jobs estimate - the number of part-time jobs would decrease by three while the number of full-time jobs would increase by one); and
- if a person employed in two part-time jobs became unemployed, the employment estimate from the LFS would decrease by one (however, this would lead to a decrease of two in the jobs estimate from a business survey).
To correctly cite the employment estimates from the LFS, commentators should refer to employment or the number of people employed, not the number of jobs. Multiple job holding is the main reason why estimates of employment from the LFS can not be equated to estimates of jobs. One employed person does not necessarily equate to one job - one person can hold more than one job. For example, the employment estimates from the July 2011 LFS reported the number of people employed as 11,446,600. This represented 10,815,000 people who held only one job during the reference period and 11,400 people who changed jobs during the reference period. A further 620,200 people (or 5.4% of all employed persons) held more than one job concurrently during the reference period. If it were assumed that each one of those people who held more than one job concurrently in the reference period held only two jobs, then the number of jobs in July 2011 would be approximately 12,066,800 (which, as expected, is a number greater than the estimate of people employed). If some of those with more than one job had three jobs, then the estimate of the number of jobs would be slightly higher.
1. Full-time work is differentiated from part-time work based on an hours worked cut-off. Persons employed full-time are those who worked 35 or more hours in the reference week, or those who usually work 35 hours or more. Part-time equates to working fewer than 35 hours per week.