CHANGES THIS MONTH: HISTORICAL REVISIONS


OVERVIEW

From the February 2011 issue of Labour Force, Australia (cat. no. 6202.0), historical labour force estimates have been revised due to the Australian Bureau of Statistics' commitment to maintaining accuracy and coherence between our published labour force products. These revisions will also be applied to data in the February 2011 issues of Labour Force, Australia, Detailed - Electronic Delivery (cat. no. 6291.0.55.001) and Labour Force, Australia, Detailed, Quarterly (cat. no. 6291.0.55.003), which are both due to be released on 17 March 2011. These revisions apply broadly to data relating to the period November 1984 to August 1997, and there are three components:

• Revisions to employment by industry estimates for the period November 1984 to May 1994, and also November 2010;
• Revisions to time series spreadsheets for the period April 1986 to August 1997; and
• Introduction of a break in series for duration of unemployment estimates.

INDUSTRY REVISIONS

For the period November 1984 to May 1994, employment by industry estimates have been revised to reflect improvements made to the concordance between the formerly used Australian Standard Industrial Classification (ASIC 1983) and the current Australian and New Zealand Industry Classification (ANZSIC 2006). This new concordance remains largely the same as what was previously used, except for amendments to the estimated employment ratio between seven ASIC 1983 Industry Groups and their 26 attributable ANZSIC 2006 Groups. The most notable impact of these revisions is on the estimates of Division I: Transport, Postal and Warehousing and Division J: Information Media and Telecommunications.

The revised historical estimates of employment by ANZSIC 2006 Division and Subdivision will be published as of the February 2011 issue of Labour Force, Detailed, Quarterly (cat. no. 6291.0.55.003) in the following tables:

• Table 04. Employed persons by Industry - Trend, Seasonally adjusted, Original
• Table 05. Employed persons by State and Industry
• Table 06. Employed persons by Industry Subdivision and Sex
• Table 11. Employed persons by Actual hours worked, Industry and Sex

Further details of these amendments are available in Information Paper: Revisions to Historical ANZSIC Industry Data (cat. no. 6259.0), released on 25 February 2011.

Revisions to employment by industry estimates for the month November 2010 have also been applied this month due to correction of a coding error. Correction of this error has resulted in minor revisions to the division level estimates, but improvements have been made to the underlying subdivision and group level estimates. These changes are mostly isolated to the estimates within Division P: Education and Training.


TIME SERIES REVISIONS

For the period April 1986 to August 1997, revisions have been applied to the time series data contained in the Excel spreadsheet products to incorporate more accurate estimates from the source data and improve coherence with other labour force products. These revisions have been applied to the time series spreadsheets released in Labour Force, Australia (cat. no. 6202.0), Labour Force, Australia, Detailed - Electronic Delivery (cat. no. 6291.0.55.001) and Labour Force, Australia, Detailed, Quarterly (cat. no. 6291.0.55.003).

The time series data contained in the SuperTABLE data cubes remain unchanged. The revisions to the time series spreadsheets brings them in line with the SuperTABLE data cubes, and both types of products now accurately reflect the source data.

The impact of these revisions to estimates are small, and in most cases there is little difference to those previously published. Seasonally adjusted and trend estimates have also been revised to incorporate the revisions to the original estimates.

DURATION OF UNEMPLOYMENT

Prior to April 2001, estimates for duration of unemployment were based on the responses to two questions in the Labour Force Survey:

• When did you begin looking for work?
• When did you last work for at least 2 weeks in a job of 35 hours or more a week?

Duration of unemployment was then based on the shortest of these two time periods; time spent looking for work, or the time since last employed in a full-time job. For example:

• If someone last worked in a full-time job 9 months ago, but only started actively looking for work 3 months ago, then their duration of unemployment was 3 months
• If someone employed full-time was told they were going to lose their job and began looking for work 9 months ago, but didn't actually stop working until 6 months ago, then their duration of unemployment was 6 months.

In April 2001, the definition of duration of unemployment was changed to reflect the international standard; the shortest time period out of time spent looking for work, or the time since last employed for 2 weeks or more in any job, regardless of whether it was full-time or part-time. The redesigned Labour Force Survey questionnaire introduced at this time incorporated an additional question to reflect this change in definition:

• When did you last work for 2 weeks or more?

In most cases, this new definition of duration of unemployment only differs from the old definition in particular circumstances. For example:

• If someone last worked full-time 2 years ago and began looking for work 9 months ago, then under the old definition their duration of unemployment would be 9 months. However, during their job search, if they worked part-time for 2 weeks in a temporary job 3 months ago, then their duration of unemployment under the new definition would be 3 months.

The duration of unemployment data collected between February 1978 and March 2001 under the old definition was reclassified as "duration of unemployment since last full-time job" and the ABS has continued to collect data based on the original questionnaire design and published the estimates in Tables 15a, 15b and UM1 of Labour Force, Australia, Detailed - Electronic Delivery (cat. no. 6291.0.55.001).

The estimates for duration of unemployment under the new definition are published in Tables 14a, 14b, UM2 and UM3 of Labour Force, Australia, Detailed - Electronic Delivery (cat. no. 6291.0.55.001). Although collection of data under this new duration of unemployment definition only began with the questionnaire redesign in April 2001, historical estimates have been provided in tables 14a, 14b and UM3 for months prior to April 2001 based on an analysis of the now reclassified "duration of unemployment since last full-time job" data. However, these historical estimates are not obtainable from the ABS source data and so cannot be updated to match the newly revised unemployment estimates provided in other labour force products. These outdated historical estimates will be removed from their respective tables and provided in the following:

• Table 14a_Sep97: Sep 1997 to Mar 2001
• Table 14b_Apr86: Apr 1986 to Mar 2001
• UM3_Sep97: Sep 1997 to Mar 2001

These historical tables will be provided in the February 2011 issue of Labour Force, Australia, Detailed - Electronic Delivery (cat. no. 6291.0.55.001) for one month only, before being moved to the labour force archive of superseded and outdated data stored in the publication Labour Force Historical Timeseries, Australia (cat. no. 6204.0.55.001).

Seasonally adjusted and trend estimates provided in the historical Table 14b_Apr86 are based on a seasonal analysis for the time period April 1986 to January 2011. From February 2011 onwards, the new Table 14b with the historical data removed will have seasonal adjustment and trend estimation based on an analysis starting from April 2001. This will result in minor revisions to the trend, which are noticeable near April 2001, but there is no noticeable difference to the seasonally adjusted and trend estimates for recent years.

FURTHER INFORMATION

For any queries regarding the implementation of any of these changes to the LFS, contact Labour Force Estimates on Canberra 02 6252 6525, or via email at <labourforce@abs.gov.au>.