QUALITY DECLARATION - SUMMARY

INSTITUTIONAL ENVIRONMENT

The experimental estimates use data sourced from a variety of institutional environments. Much of this data is collected by other organisations for administrative purposes.

The experimental estimates are predominantly derived using workplace location, industry and remuneration information from the WorkSafe dataset. This dataset is collected by WorkSafe Victoria through their registration of workers compensation insurance.

Commonwealth employers and Commonwealth government trading enterprises operating in Victoria, as well as businesses that tendered for work in competition with government departments or government business enterprises, are insured through Comcare rather than WorkSafe Victoria. Comcare data is included in the production of the experimental estimates for the 2010/11 financial year. The workplace location, employee and industry data for these business was obtained by linking Comcare data with data from the ABS Business Register.

The experimental estimates have been produced by the State and Territories Statistical Service (STSS) within the ABS. For information on the institutional environment of the Australian Bureau of Statistics (ABS), including the legislative obligations of the ABS, financing and governance arrangements, and mechanisms for scrutiny of ABS operations, please see ABS Institutional Environment.

RELEVANCE

The purpose of the experimental estimates is to produce Victorian employee location data for different industry classifications at a sub-state level. That data is currently only produced every five years as part of the Census of Population and Housing and so these experimental estimates address a data gap.

The scope of the estimates are based on the businesses with locations in Victoria that are registered with WorkSafe Victoria for the reference periods of 2005-06 to 2010-11, and for the 2010-11 reference period also included those registered in Victoria with Comcare.

These businesses include:

• businesses that employ workers and have an annual employer remuneration expense greater than $7,500, or those which employ trainees or apprentices, registered with WorkSafe Victoria,
• employers approved by WorkSafe Victoria to manage their own worker's compensation (Self-Insurers), and
• for the 2010-11 reference period also included Victorian located Commonwealth employers and Commonwealth government trading enterprises operating in Victoria registered with Comcare.

The experimental estimates have been produced at the Local Government Area (LGA) level for metropolitan Melbourne. LGAs outside of metropolitan Melbourne have been aggregated to Regional Growth Plan areas to enable the release of estimates at the ANZSIC subdivision level. LGAs are part of the Australian Statistical Geography Standard (ASGS) as a non-ABS structure. More information of Regional Growth Plan areas can be found of the Victorian Department of Planning and Community Development website. The Unincorporated component of the Victorian LGA classification has been excluded from the experimental estimates.

The businesses that contribute to the statistics in this publication are classified by industry in accordance with the Australian and New Zealand Standard Industrial Classification (ANZSIC), 2006 edition (cat. no. 1292.0). The following subdivisions have been removed from the scope of the project during production to support confidentiality and are not included in the estimates:

• 43 Non-store retailing and retail commission-based buying and selling - excluded from Retail trade
• 76 Defence - excluded from Public administration and safety

TIMELINESS

The WorkCover dataset is collected for each financial year and delivered to the ABS on an annual basis. The Comcare dataset for the 2010-11 financial year was requested by the ABS in the course of the production of this publication.

The estimates produced herein are for a six periods from the 2005/06 financial year to the 2010/11 financial year.

ACCURACY

This publication outlines a number of accuracy considerations that were taken into account by the ABS in the production of these estimates. These considerations relate to: the remediation work undertaken as part of the geocoding process, the method undertaken to ensure all workplaces were coded to industry classifications, the changes made to the Employee Earnings, Benefits and Trade Union Membership (EEBTUM, cat. no. 6310.0) survey from 2007, and the method used to produce the employee estimates themselves.

Estimates from the Employee Earnings, Benefits and Trade Union Membership Survey (EEBTUM, cat. no. 6310.0) are subject to sampling and non-sampling errors. The highest sampling errors are present for state and capital city/balance of state, so this project has utilised Mean Annual Earnings at the Australia level in order to minimise sampling error. Using estimates from this level, however, may result in a bias as the experimental estimates are derived at the Victorian level.

Due to the methodology used in the production of the experimental estimates, it is not possible to precisely calculate the error associated with each individual estimate. The estimates do however have an error at least as large as the error associated with the EEBTUM survey. Users should interpret the results with caution. Please refer to the EEBTUM Quality Declaration for more information.

COHERENCE

The ABS has a number of publications which produce estimates for the numbers of employees in a given area. The Labour Force Survey produces employee estimates at the state level, and the Census produces employee estimates at very small geographies.

While the experimental estimates for the 2006/07 and 2010/11 financial years produce a similar distribution of employees across industries and geographies as the 2006 and 2011 Censuses of Population and Housing respectively, the experimental estimates should be compared with other ABS data sets with caution, due to substantial differences in methodology.

Due to a change in the EEBTUM survey methodology in 2007, and the inclusion of Comcare data in the 2010/11 financial year estimates, the estimates contained in the publication do not represent an unbroken time series. There are three time series in this publication:

• 2005/06
• 2006/07 to 2009/10
• 2010/2011

INTERPRETABILITY

In using a statistical method to estimate the number of employees, the estimates produced in this publication are considered to be an 'employed person equivalent' rather than the number of employees. For example, an estimate of '1' in a given cell may represent one person, or it may represent two or more employees working a proportionately less amount of time. In this way the 'employed person equivalent' is effectively a proxy measure for the number of employees.

See the Estimation Methods chapter for more information about the method used to produce these estimates.

ACCESSIBILITY

These estimates are available on the ABS website as a data cube, downloadable in Microsoft Excel spreadsheet format.

Further detailed estimates may be available upon request. Contact National Information Referral Service on 1300 135 070.