Internet Vacancy Index – Methodology


Overview

The Internet Vacancy Index (IVI) is a monthly data series measuring online job advertisements, compiled by the National Skills Commission (NSC).

IVI data count job advertisements newly lodged on the SEEK, CareerOne and Australian JobSearch online job boards during the reference month.

These job advertisements are coded to a fine level of detail by the NSC, with data available by occupational groups (down to the 4-digit ANZSCO level), skill level groups and by state or territory, as well as a best-fit regional geographical structure.

The IVI is the only publicly available source of detailed information of this kind.

To view the latest data return to the detailed Internet Vacancy Index page


Scope

All job advertisements newly lodged on the contributing job boards for vacancies within Australia during the reference month fall within scope of the IVI.

IVI monthly results have been compiled into a time series that dates back to January 2006.

  • The regional series was added to the IVI in 2010, hence regional IVI results are not available prior to that time.

The IVI is based on administrative data provided by contributing job boards and, as such, is not subject to the sampling error present in survey-based estimates.

The IVI can be used as a proxy for the level of Australian recruitment activity and an indicator for labour demand, however there are some important conceptual limitations described following that can help you to best understand the data.


Conceptual limitations

The IVI does not reflect the total number of job advertisements in the Australian labour market.

    • The IVI does not account for jobs advertised through other online job boards, employer websites, social media, newspapers, or other informal methods such as word of mouth.
    • The IVI does not take account of multiple positions being advertised in a single job advertisement.

Job vacancies and job advertisements are different. Some employment opportunities are not advertised by employers, who may instead fill their vacancies via internal promotion or alternative recruitment methods.

Online job advertisements can be slightly biased towards higher skilled positions. Employers with lower skilled vacancies tend to use informal recruitment methods like social media or word of mouth more regularly (see NSC Recruitment Experience and Outlook Survey findings).


 


Coding process

As part of the NSC’s IVI data preparation process, job advertisements are coded to:

  • Occupation at the 4-digit ANZSCO (Australian and New Zealand Standard Classification of Occupations) level of detail.

For more information on the ANZSCO framework, visit the ABS website.

  • Skill Level group.
  • State/Territory.

Job advertisements that cannot be coded to at least this level of detail are dropped from the IVI count due to insufficient information.

Additionally, job advertisements are coded to the IVI Regional structure, comprising 37 best-fit labour market regions (see Regional Structure section below). If a job advertisement cannot be coded at the regional level of detail, it is assigned to an ‘other regions’ group for the state or territory to which it is coded to.

Duplicate job advertisements are also removed as part of the IVI coding process. The IVI coding accounts for both ‘within site’ and ‘across site’ duplicates in its de-duplication process:

  • Within site duplicates: refer to duplicate job advertisements posted to the same job board during the reference month.
  • Across site duplicates: refer to duplicate job advertisements posted to multiple contributing IVI jobs boards during the reference month.

IVI coded results are compiled into a monthly time series and seasonally adjusted and trended.

  • Each individual job advertisement series is seasonally adjusted and trended separately.
  • The separate seasonal adjustment and trending processes mean IVI seasonally adjusted and trended results are non-additive.

The NSC applies seasonal adjustment and trending to the IVI results to account for the pronounced seasonal patterns within the Australian labour market and to reduce data volatility. This helps with the identification of underlying data trends. 


Correspondence files

Excel document
LGA (2016) To IVI Region

(46 Kb)

Excel document
SA4 (2016 ASGS) To IVI Region

(22 Kb)

The correspondence files above can be used to understand the relationship between IVI Regions and other geographic units.

The files list the geographic units which make up each IVI Region. Geographic units that do not fall entirely within a single IVI Region are given a percentage according to how much of the population falls within the IVI Region. For these IVI Regions, the correspondence method may lead to small errors.

The correspondence files contain ABS quality ratings which indicate how closely the geographic units match each IVI Region. Care should be exercised in using correspondence files to convert data from one geographic basis to another, particularly where the quality rating is ‘Poor’, as the resulting data may not reflect the actual characteristics of the geographic areas involved. Further information about the ABS quality ratings is available from the ABS website.

Please note that the correspondence files are based on the total population counts at the time of the 2016 Census. 


Regional structure

The IVI uses a custom geographical structure for its regional coding based on the differing area groupings used by the job boards contributing to the data. The IVI regional structure is outlined in the following documentation.

PDF document
IVI Regions Structure

(312 Kb)

IVI Shape Files

(9.18 Mb)

IVI Geojson Files

(11.60 Mb)


Additional notes

The Vacancy Report usually focuses analysis on the trend IVI data series. Trend analysis provides a more stable series for comparisons over time, however, such analysis is initially resistant to large movements and does not fully capture monthly fluctuations in job advertisements following an economic shock. Accordingly, the National Skills Commission has temporarily shifted the focus of the analysis in the Vacancy Report to the seasonally adjusted IVI series during the COVID-19 period.

Currently, the Vacancy Report replaces much of the standard year-on-year analysis with comparisons to pre-COVID-levels. Comparisons to pre-COVID-19 levels provide a better indication of the current state of recruitment activity relative to previous levels.

Pre-COVID-19 job advertisement levels are defined as the 12-month average in the seasonally adjusted IVI series to February 2020.

A 12-month average is used to reduce the impacts of:

  • End-point revisions of seasonally adjusted and trend estimates.
  • The varying times at which particular data series were first impacted by the effects of COVID-19.

Glossary

Reference month: refers to the entire month from which the IVI data is compiled. The reference month has a one-month lag to the month of data release.   

Newly lodged job advertisement: the IVI only counts job advertisements newly posted to contributing jobs boards during the reference month.

Contributing jobs boards: The IVI has the defined scope of covering job advertisements posted to SEEK, CareerOne and Australian JobSearch.

Level: the amount of detail a data series corresponds to within the hierarchy of the given downloadable IVI file.

ANZSCO code: the Australian and New Zealand Standard Classification of Occupations code. At the 2-digit level the NSC has aggregated customised 2-digit ANZSCO groups that better fit the structure of job advertisement postings.

ANZSCO title: the Australian and New Zealand Standard Classification of Occupations title.

  • At the 2-digit level the NSC has aggregated customised 2-digit ANZSCO groups that better fit the structure of job advertisement postings.

Acknowledgements:

The National Skills Commission thanks the following job boards for their ongoing contribution to the IVI:

Back to top