Home > Introduction > Limitations of the data

During 2003-04, extensive audits and error checking of APSED were undertaken and errors in the historical data were corrected. The following points, therefore, should be taken into consideration when making comparisons between data in this publication and previous publications.

• Previously published data for both ongoing and non-ongoing staff may have been revised and therefore may not necessarily be comparable with that in previous years' publications.
• In previous years some people who separated on 30 June were excluded from the count of the ongoing and non-ongoing employee population. The major impact was on the non-ongoing population. Data for earlier years has been corrected for this publication.
• A headcount approach underpins the Bulletin-that is, people working part-time are aggregated with people working full-time without weighting.
• Following the introduction of broadbanding of classifications in 1999, data in this publication reports the maximum classification level and is, therefore, not directly comparable with data for earlier years. This will result in some employees being reported at a higher than actual level (ie at the top of their broadband). Significant improvements in the quality of data provided by agencies this year have resulted in fewer agencies reporting broadbanded classifications. For example, at June 2003 over 31,000 employees had a different maximum and minimum classification. By June 2004 this figure had dropped to just over 2,600. Therefore, data on classifications is not directly comparable with that from earlier years. See Appendix 2.
• APSED data reports the actual classification of staff and includes temporary assignments to a higher level for periods of greater than 90 days. Previously, CRP data referred only to substantive classification.
• Data for the Senior Executive Service (SES) includes employees grouped with the equivalent SES bands under the Public Service Classification Rules 2000. Data also includes inoperative SES.
• As part of the process for loading data from the CRP onto APSED, an audit was undertaken of employees who were current at June 1999. A number of records were found where non-ongoing employees had been recorded as ongoing on the CRP. This led to a recalculation of ongoing employee numbers in the years preceding 1999. Employees who were employed at June 1999 were included in this audit. However, those who had been employed in earlier years but who had since left were not included. It can be assumed that there was an overestimate, which cannot be quantified, of ongoing employees in those years.
• Time series data reflect the Public Service as it was at that point in time. When comparing trends over time, the data may vary due to coverage changes within the Public Service Act. For example, in 1999 the Parliamentary agencies ceased to be covered by the PS Act. Administrative changes affecting APSED are detailed in Appendices 3 and 4.
• Due to differences in data sources and data definitions, there may be variations between data in this publication and that published by individual agencies.
• EEO data published by the APS Commission may differ from that published by some agencies. An independent review of the APS Commission's then current practice of retrospectively changing a person's EEO status if it changed during their employment was undertaken in early 2001. The review supported the continuation of the practice for reporting for non-English speaking background (NESB) and Indigenous status as these characteristics are likely to be constant throughout an employee's career. Disability, however, may change with time and APSED has been modified to record the status change for an individual and the date at which that change was notified by the agency.
• The provision of some EEO data by APS employees to their agency is voluntary. Therefore, as with any large voluntary data collection, APSED tends to under-represent the number of Indigenous Australians, people with a disability and people from a non-English speaking background (NESB).
• The proportion of employees in the NESB 2 category is lower than previously reported. This is because of an error in the algorithm used to calculate NESB status in earlier years when employees with missing data were incorrectly assigned a NESB status.

Note: Some classification data in this publication differs very slightly from that in this year's State of the Service report. This resulted from an error in data supplied by one provider which was found after the report was finalised. The correction affects a small number of employees, mainly at the APS 1-6 classification levels.