How we measure competition

Purpose and scope of this section

This section summarises the datasets and shared definitions used across the tracker measures, including the unit of analysis, time basis, coverage rules, and industry classification. The next sections describe how these inputs are used to construct each indicator and how results are aggregated for reporting.

Data sources at a glance

Use the in-depth cards linked above for dataset specific details such as variable definitions, access notes, and any measure specific caveats.

Core concepts used across measures

These definitions apply across the competition measures unless stated otherwise in the indicator specific documentation.

• Time period: Financial year (for example 2023-24).
• Business population: Businesses classified as active in the relevant financial year.
• Industry classification: ANZSIC, with core calculations produced at the four digit industry class level.
• Market share activity measure (for share based measures): Turnover reported via BAS.
• Sector coverage (for BLADE based competition measures): Restricted to the non-financial market sector, excluding ANZSIC subdivisions 62-64 and non-market divisions covering public administration, education, and health (subdivisions 75-77, 80-82, and 84-87).

Where turnover is missing for an otherwise active unit in a given year, that unit is not used in the turnover share calculations for that specific industry and year.

For BLADE based measures, we treat Type of Activity Units (TAUs) as the firm identifier. This separates large enterprises into core operating functions so groups active in multiple industries can be represented separately (for example Wesfarmers).

BLADE microdata used in the tracker

We estimate concentration, displacement, and incumbent turnover using de-identified business microdata from the ABS Business Longitudinal Analysis Data Environment (BLADE). BLADE integrates firm records from ATO administrative sources (Business Income Tax and Business Activity Statements) with ABS Business Surveys.

For this work we rely on BAS turnover and firm demographics from the Australian Business Register. For more detail, see the ABS BLADE page covering microdata and TableBuilder projects: https://www.abs.gov.au/statistics/microdata-tablebuilder/available-microdata-tablebuilder/business-longitudinal-analysis-data-environment-blade

Industry definitions and base calculation level

The core concentration calculations are produced for each four digit ANZSIC industry class in each financial year. This is the most detailed industry level used in the published framework.

The diagram below illustrates how broader ANZSIC levels are formed by grouping together more detailed levels.

Reporting levels and aggregation approach

We also report concentration at broader industry groupings:

• Three digit level: Industry group
• Two digit level: Industry subdivision
• One digit level: Industry division
• National summary: All industries combined

At these broader levels, concentration is not recalculated using pooled turnover across all included industries. Instead, it is created by averaging the detailed four digit results, as described in the calculation steps.

Dataset notes by measure

The table above links each tracker measure to its primary dataset. The notes below clarify how each dataset is used at a high level. For detailed definitions and any measure specific exceptions, refer to the corresponding in-depth card.

• Market concentration: Estimated from BLADE using BAS turnover to form industry market shares at the TAU level.
• Firm displacement and incumbent turnover: Estimated from BLADE using TAU based firm identifiers and longitudinal firm demographics.
• Entry and exit: Estimated from ABS CABEE (see in-depth card for unit definitions and scope).
• Labour mobility: Estimated from the ABS PJSM survey (see in-depth card for mobility definitions and scope).

Data quality, confidentiality, and limitations

Australian Bureau of Statistics, Business Longitudinal Analysis Data Environment (BLADE), financial years 2003-04 to 2023-24, accessed via ABS DataLab. Findings are based on use of BLADE data. Outputs were confidentialised and cleared by the ABS prior to release, consistent with DataLab confidentiality requirements.

Purpose

These measures describe how concentrated business activity is within Australian industries. In practical terms, they show whether revenue in a detailed industry is spread across many businesses or dominated by a small number of firms.

Step 1: Build a TAU by year dataset

For each financial year:

• Identify TAUs that are active during the year.
• Assign each TAU to its four digit ANZSIC class for that year.
• Attach BAS turnover for the same year.

This produces one record per TAU per year with:

• Industry class
• Turnover
• Active status

Step 2: Calculate turnover shares within each four digit industry class

For each four digit ANZSIC class and financial year:

• Add up turnover across all included TAUs to get total turnover for that class and year.
• Compute each TAU's share of turnover within that class.

TAU turnover share = (TAU turnover) divided by (total turnover in the class).

These shares sum to approximately 1 for each class and year (allowing for rounding).

Step 3: Calculate the concentration measure for each four digit industry class

Using turnover shares, a concentration statistic is calculated for each four digit class and year.

Common concentration statistics include:

• Top 4 share (CR4): the combined turnover share of the four largest TAUs in the class.

If your published results use one of these, the interpretation is straightforward:

• A higher value indicates a more concentrated industry.
• A lower value indicates turnover is spread more evenly across many businesses.

For example: a value of 0.60 means the four largest units account for 60 percent of turnover in that detailed industry class for that year.

Step 4: Create broader industry level results by averaging four digit outcomes

Broader level figures are produced by taking a simple average of the four digit class concentration results that sit underneath the broader category.

This is an equal weight approach, meaning each four digit class contributes the same weight to the broader level outcome, regardless of the size of that class.

Specifically:

• Three digit level result: average of the four digit class results belonging to that three digit group.
• Two digit level result: average of the four digit class results belonging to that two digit subdivision.
• One digit level result: average of the four digit class results belonging to that one digit division.
• National summary: average of all four digit class results across the economy.

This approach produces a measure of the "typical" concentration across detailed industries within a broader category. For example, a CR4 of 0.55 in the beverage and tobbacco manufacturing industry means that the average four digit sub-industry within beverage and tobacco manufacturing has a CR4 of 55 percent.

Step 5: Output rules and publication

All outputs are produced and released in line with ABS DataLab confidentiality and disclosure requirements. Results are published only where they meet the relevant minimum contributor and confidentiality thresholds. Where requirements are not met, results are suppressed or reported at a higher aggregation level.

Data quality and interpretation notes

• Industry classifications are not the same as competition "markets." ANZSIC classes are designed for statistical reporting and may not match how firms compete in practice (for example across regions or product segments).
• Equal weighting at broader levels changes the meaning of the metric. The broader level and national summary figures are not dominated by large industries, but they also do not represent an economy-wide, turnover-weighted concentration measure.
• Administrative definitions differ from accounting concepts. BAS turnover is a consistent administrative measure, but it may not align perfectly with financial statement revenue concepts for every business.

Data source and unit of observation

Business entry and exit measures are sourced from the ABS Counts of Australian Businesses, including Entries and Exits (CABEE) release available on the ABS website. CABEE is produced from the Australian Bureau of Statistics Business Register (ABSBR), which is populated using Australian Business Register (ABR) registrations and Australian Taxation Office (ATO) administrative data (including BAS related information), with additional ABS profiling for large or complex businesses.

The CABEE statistical unit defines a "business" by the ABS Type of Activity Unit (TAU), which represents an operationally separate unit within an enterprise group. For small businesses without complex structures, the TAU often corresponds to the ABN level. For large or multi-divisional enterprises, the TAU separates different business functions that may operate in different industries. This unit of observation supports more accurate industry classification and better reflects business dynamics compared to using ABN level data alone.

Reference period and scope

CABEE is an annual financial year series, with counts benchmarked to 30 June each year. The ABS compiles the data using an extract taken in mid-May to produce 30 June snapshots for the most recent five financial years in the release.

Counts are limited to actively trading businesses in the Australian market sector. In CABEE, actively trading businesses are TAUs (profiled population) and ABNs (non-profiled population) that are actively remitting GST.

A business can still be counted as actively trading at 30 June even if it has ceased operating, provided the ABN has not been cancelled and it has remitted GST within the last five quarters (or three years for annual remitters). Businesses that have not lodged BAS (and/or have reported zero amounts) over five consecutive quarters (or three consecutive years for annual remitters) are treated as long term non-remitters (LTNRs) and are excluded from actively trading counts.

Industry and grouping

Each business is classified to a single ANZSIC 2006 industry class (4 digit). Within a CABEE release, the business’s final year classification is applied across all years in that release (a back-cast classification), which supports consistent industry groupings across the five-year window.

For reporting, ANZSIC classes can be aggregated to broader groupings (division 1 digit, subdivision 2 digit, group 3 digit) using standard ANZSIC concordances.

Defining entries and exits

CABEE identifies business events by comparing consecutive 30 June snapshots.

Business entry (count): a business is an entry in financial year $t$ when it is actively trading on the ABSBR at 30 June of the reference year, but was not actively trading at 30 June of the previous year. CABEE separates entries into business births (new registrations or new profiled units) and other entries (for example ABN reactivations, LTNRs that resume BAS activity, or businesses moving into scope due to reclassification or first GST role registration).

Business exit (count): a business is an exit in financial year $t$ when it was actively trading at 30 June of the previous year but is no longer actively trading at 30 June of the reference year. CABEE separates exits into business cancellations (for example ABN cancellation or cessation of a profiled unit) and other exits (for example becoming an LTNR or moving out of CABEE scope).

Movements between the profiled and non-profiled populations are recorded as an exit and corresponding entry (usually a small number).

Intra-period events: a business that enters after 30 June in one year and exits before 30 June in the next year will not appear in either snapshot and is not included in CABEE entry or exit counts.

Calculating entry and exit rates

For a given industry grouping $g$ and financial year $t$ (for example 2024-25), define:

• $S_{g,t}$: businesses operating at the start of the financial year (count at 30 June of the prior year).
• $E_{g,t}$: business entries during the year (present at 30 June $t$, not present at 30 June $t-1$).
• $X_{g,t}$: business exits during the year (present at 30 June $t-1$, not present at 30 June $t$).

CABEE entry and exit rates are calculated using the start-of-year business count as the denominator:

This denominator convention is consistent with how CABEE tables present Operating at start of financial year, Entries, Exits, and the associated Entry rate and Exit rate.

If required, a derived net measure can be reported as:

and the net change in business counts over the year as:

Some CABEE stratifications (for example turnover or employment size ranges) can also include net movement of surviving businesses between size strata, which affects reconciliation within those specific tables.

Quality and comparability considerations

CABEE is a complete enumeration of economically active businesses on the ABSBR (not a sample survey), so it is not subject to sampling error, but it can be affected by non-sampling error and administrative updates to the underlying register and classifications.

To protect confidentiality, the ABS applies perturbation to many detailed cells in CABEE data cubes, which can mean detailed components may not sum exactly to totals; high-level totals are generally unperturbed.

Data source and population

Labour mobility is measured using the ABS Job mobility release from the Participation, Job Search and Mobility (PJSM) survey. PJSM is collected in February as a supplement to the monthly Labour Force Survey (LFS), where in-scope respondents are asked additional questions relating to job mobility.

The population covered is the civilian population aged 15 years and over (consistent with the LFS scope) with standard exclusions such as permanent defence force members and certain overseas residents, and supplementary-survey exclusions such as people in institutions (for example hospitals and prisons).

Reference period and timing

Estimates relate to the year ending February reference period (for example, year ending February 2025). Data are collected during February, and survey responses relate to the week prior to interview (the reference week).

If results need to be aligned to financial years for comparison with other indicators, a practical approach is to map year ending February YYYY to the financial year that contains February (for example, February 2025 maps to FY2024-25). This mapping should be documented as a convention because the ABS release itself is not a financial-year series.

Constructing the labour mobility measure

The primary indicator is the ABS job mobility rate, defined as:

Changed jobs is defined as changing employer or business in the previous 12 months.

Industry allocation and grouping (ANZSIC)

Industry breakdowns are taken from the ABS Job mobility release tables for:

• Industry of job that was left (industry at the start of the year), which is useful for describing mobility out of an industry.
• Industry of job that was entered (industry at the end of the year), which is useful for describing mobility into an industry.

These industry categories correspond to ANZSIC industry divisions (1-digit).

Quality notes

PJSM estimates are survey-based and therefore subject to sampling and non-sampling error; ABS provides relative standard errors (RSEs) in the data downloads for assessing reliability of disaggregated estimates.

Data and unit of observation

Firm displacement ratios are calculated from ABS' BLADE dataset. Top four firms in a given industry year combination are calculated using total annual turnover sourced from Business Activity Statement (BAS) items available within BLADE.

Because BAS turnover is recorded at the Australian Business Number (ABN) level, turnover is allocated from ABN to Activity Unit using the allocation procedure described in the Data and definitions section. The resulting analysis file contains, for each financial year, an Activity Unit level turnover measure and an associated ANZSIC industry classification.

Identifying industry turnover leaders

For each financial year $t$ and each ANZSIC class at the 4 digit level $c$:

• Restrict the data to Activity Units classified to $c$ in year $t$.
• Rank Activity Units by allocated turnover in descending order.
• Define the leader set $\text{Top4}_{c,t}$ as the four Activity Units with the highest turnover in that year and class.

Three year persistence and displacement

For each Activity Unit $i$ that belongs to $\text{Top4}_{c,t}$, construct an indicator of whether it remains a turnover leader three financial years later within the same 4 digit class:

Activity Units that are not observed in $t+3$, have missing turnover at $t+3$, or are no longer present in the class leader set at $t+3$ are coded as 0, since they do not persist as class level leaders.

Aggregation and displacement ratios

Each 4 digit ANZSIC class is mapped to broader ANZSIC groupings (1 digit division, 2 digit subdivision, and 3 digit group).

For a given financial year $t$ and grouping $g$, compute the persistence rate as the share of baseline leaders that remain leaders after three years:

The sum is over all Activity Units $i$ in $\text{Top4}_{c,t}$ whose class $c$ belongs to grouping $g$; $N_{g,t}$ is the number of baseline leader observations.

The firm displacement ratio is defined as the complement of persistence:

This produces, for each year and industry aggregation level, the proportion of turnover leaders that are replaced within a three year window.

Sample window

Because persistence is evaluated at $t+3$, results are calculated only for baseline years $t$ where data are available in year $t+3$.