MAJOR REBENCHMARKING OF LABOUR FORCE SERIES
Overview
“Benchmarks” refer to the population estimates that Labour Force Survey (LFS) data are aligned to, so that they effectively represent the population. That is, how the large monthly sample of 50,000 people can effectively represent the 20 million usually resident civilian population aged 15 years and over.
Every five years, population estimates are updated to include the latest information from the Census, and this updated information is used to "rebenchmark" the Labour Force estimates.
This article provides a simple explanation of what “rebenchmarking” is, the reasons for rebenchmarking and the impacts on Labour Force estimates.
From benchmarks to estimates
Labour Force statistics measure key aspects of the Australian labour market, and provide some of the most important economic and social indicators. For this reason, it is important to ensure that Labour Force statistics are of the highest quality, and most effectively represent the labour market status of the population.
There are two key inputs to the Labour Force estimates:
· A measure of the size and profile of the population potentially available to the labour force; and
· The labour force status of people in the population.
The first of these are population benchmarks, which are based on population estimates, while the second is based on the data that are collected in the LFS.
To produce Labour Force estimates, data from the LFS are "weighted up" to the total population benchmarks. Since only a sample of the population can be surveyed every month, this weighting process transforms the survey data into representative estimates, which provide a picture of the entire population.
Rebenchmarking to 2016-based benchmarks
Rebenchmarking is the process of updating the total population figures that the Labour Force estimates are weighted up to, as new Estimated Resident Population (ERP) figures become available in Australian Demographic Statistics (cat. no. 3101.0).
Prior to this release, Labour Force estimates were based on 2011 Census-based benchmarks. Following consultation with key users, the ABS has rebenchmarked the Labour Force estimates to 2016-based ERP population benchmarks, resulting in revisions to Labour Force estimates for the period from July 2011 onwards.
Rebenchmarked estimates will be progressively released from 7 December to 21 December 2017, per the following schedule:
· Rebenchmarked person estimates in Labour Force, Australia - Rebenchmarked Estimates, Oct 2017 (cat. no. 6202.0.55.003) on 7 December 2017;
· Rebenchmarked hours worked estimates in Labour Force, Australia, Nov 2017 (cat. no. 6202.0) – Tables 19 and 20 – on 14 December 2017;
· Rebenchmarked detailed monthly data in Labour Force, Australia, Detailed - Electronic Delivery, Nov 2017 (cat. no. 6291.0.55.001) – on 21 December 2017; and
· Rebenchmarked detailed quarterly data in Labour Force, Australia, Detailed, Quarterly, Nov 2017 (cat. no. 6291.0.55.003) – on 21 December 2017.
Labour Force statistics will undergo a second major rebenchmarking to the final 2016-based ERP population benchmarks, in December 2018.
The ABS will also continue with the regular quarterly rebenchmarking, as per the process outlined in the article ‘Rebenchmarking of Labour Force Series’, published in the February 2015 issue of
Labour Force, Australia (cat. no. 6202.0).
What is the impact of Rebenchmarking on the Labour Force estimates?
While rebenchmarking results in revisions to published data, it does not generally result in any material change to headline indicators (e.g. unemployment rates, participation rates or employment to population ratios) at the Australia or state and territory levels. Changes to the population benchmarks are more likely to impact on estimates that are directly related to the size of the population (e.g. the number of employed people).
Changes in population composition such as age, sex or region of residence may result in a different rate of change, particularly at finer levels of geography. A summary of the impact of the major rebenchmarking is outlined below.
- The overall labour market narrative is unchanged as a result of rebenchmarking, with the headline unemployment rate, participation rate and employment to population ratio showing very little change. For example, as at October 2017, the change to the trend unemployment rate was less than 0.1 percentage points at the Australia level.
- As at October 2017, trend employment has been revised upwards by around 40,000 persons, while the number of unemployed persons was revised up by around 500 persons.
- For the rebenchmarking period (July 2011 to October 2017), the median change to the trend estimate of employed persons was a 27,500 person increase. The magnitude of change increases through the series with the July 2011 difference being 1,100 (0.009%) and the October 2017 difference being 41,700 (0.34 per cent). The impact on state trend employment estimates are also relatively small, with the largest changes in Victoria (up around 55,000) and Western Australia (down around 35,000).
- There have been no changes to survey content or questionnaire design as part of the rebenchmarking process.
Table 1: Trend Estimates, October 2017
Trend - Australia | Rebenchmarked (2016 ERP Base) | Pre- Rebenchmarked (2011 ERP Base) | Difference | Difference (%) |
| Employed ('000) | 12342.9 | 12301.2 | 41.7 | 0.34 |
| Unemployed ('000) | 709.9 | 709.4 | 0.5 | 0.08 |
| Unemployment Rate (%) | 5.4 | 5.5 | -0.01 | -0.25 |
| Employment to Population Ratio (%) | 61.7 | 61.7 | 0.08 | 0.13 |
| Participation Rate (%) | 65.3 | 65.2 | 0.08 | 0.12 |
Graph 1: Trend employed persons, Australia, July 2011 to October 2017
Table 2: Seasonally Adjusted Estimates, October 2017
Seasonally Adjusted - Australia | Rebenchmarked (2016 ERP Base) | Pre- Rebenchmarked (2011 ERP Base) | Difference | Difference (%) |
| Employed ('000) | 12337.8 | 12297.1 | 40.7 | 0.33 |
| Unemployed ('000) | 702.4 | 701.5 | 0.9 | 0.13 |
| Unemployment Rate (%) | 5.4 | 5.4 | 0.00 | 0.00 |
| Employment to Population Ratio (%) | 61.7 | 61.6 | 0.10 | 0.16 |
| Participation Rate (%) | 65.2 | 65.1 | 0.10 | 0.15 |
Graph 2: Seasonally Adjusted Employed Persons, Australia, July 2011 to October 2017
Print Page
Print All