AUSTRALIAN BUSINESS GROWTH 2026: DATA AND CHART NOTES
Peter Ngo Digital, prepared 3 November 2026. Local V4 draft edition.

2026 SOURCE
https://www.abs.gov.au/statistics/economy/business-indicators/counts-australian-businesses-including-entries-and-exits/jul2022-jun2026
Workbook: the source URL followed by /8165DC01.xlsx. Released 18 August 2026.
Tables 1/10: start B, end I, change J, growth K. Tables 4/13: end J, change K, growth L.
Both annual comparison rates come from the same 2026 release, including revised baselines.

CSV DICTIONARY
Current period: 2025-26 unless a period column specifies another year.
start_count/end_count: businesses at the beginning/end of the financial year.
net_change: end minus start; growth_pct: ABS published annual percentage change.
previous_net_change/previous_growth_pct: 2024-25 observations in the SAME 2026 release.
entries/exits: ABS entry/exit totals, not startup/insolvency counts.
entry_rate_pct/exit_rate_pct: percentages against the starting business count.
net_movement_surviving: surviving businesses moving between employment groups/states.
source/previous_source: original workbook, sheet and row range for independent checks.
Survival: original_businesses is the original cohort; survived_june_2026 and survival_rate_pct
refer to that cohort through June 2026. Unknown-category rates are blank, never zero.

HISTORICAL COUNCIL SOURCE AND METHOD
https://www.abs.gov.au/statistics/economy/business-indicators/counts-australian-businesses-including-entries-and-exits/jul2021-jun2025
Workbook: this older URL followed by /8165DC10.xlsx. Released 16 December 2025.
Tables 1/2/3 are June 2025/2024/2023 respectively. Total column K, non-employing column F.
536 areas match by council/industry code with numeric observations across all three years,
excluding placeholders/unincorporated areas. Missing observations are never zero.
135 meet the minimum of 5,000 businesses in June 2024. Rank unrounded annual percentage
growth descending; council code breaks ties. Rate = (2025 count / 2024 count - 1) x 100.
Two-year growth uses 2025/2023 and is cumulative, not annualised. Empty rates can mean a zero
denominator. Empty non-employing fields mean unavailable data; empty ranks mean ineligible.
Leading industry means largest division net increase, not largest existing industry.
Transport is sole leader in 8/10 top areas and 14/125 others (80.0% versus 11.2%). Including
joint leaders Greater Shepparton/Melville gives 16/125 (12.8%); the top ten stays 8/10.
Threshold sensitivity, sole leaders: minimum 1,000 gives 9/10 versus 25/291;
minimum 5,000 gives 8/10 versus 14/125; minimum 10,000 gives 7/10 versus 8/71.
Transport is included in council total increases: never add the paired chart bars.

LIMITS AND REUSE
Counts cover actively trading businesses in ABS market-sector scope, not all ABNs,
premises, jobs, revenue, demand or investment merit. Multi-location businesses are assigned
to one geography. Local cells are perturbed and need not add to totals. Employment can lag.
Broad transport counts cannot isolate couriers, rideshare or warehouses. No causal claim.
Detailed 2026 local tables were scheduled for 18 December 2026 and are not in this edition.
For editorial reuse credit Peter Ngo Digital and ABS, retain dates, definitions and footnotes,
and identify changes. Independent analysis, no ABS endorsement. Copyright/licence exceptions:
https://www.abs.gov.au/website-privacy-copyright-and-disclaimer
Queries/corrections: info@peterngodigital.com.au

V3: 2026 INTEGRATION
2026 now supplies the study lead, employment breakdown, all 19 industry divisions, all eight
states/territories, national entries/exits, legal organisation and survival observations.
2025 council results are retained as historical detail, never extrapolated to 2026.
New CSVs include all seven employment rows (subtotals overlap; do not sum every row),
2026 industry/state values with prior-year change/rate from the SAME latest release,
four financial years of entry/exit totals, five legal organisation categories and two
survival cohorts by industry. Net growth is not entries, entries are not simply startups,
exits are not insolvencies, and non-employing is not a legal structure.
Table 10 covers legal organisation. Tables 2 and 3 cover distinct cohorts' survival to June 2026.
The survival percentage is relative to the original cohort, not an annual failure rate.
Full national totals include Currently Unknown and Other Territories where applicable;
industry/state comparison tables omit those placeholders and must not reconstruct national totals.
Chart 06 shows the top eight national industry rates; chart 07 shows all eight state/territory rates.
Chart 08 replaces the earlier national bar with a pie: 83,105 / 85,130 = 97.6%; 2,025 / 85,130 = 2.4%. These are shares of the net increase, not all businesses. Charts 01/02/04 remain June 2025 local data.
Source credits: Peter Ngo Digital analysis of ABS Counts of Australian Businesses,
July 2022 to June 2026, data cube 1; historical council analysis uses July 2021 to June 2025,
data cube 10. ABS material copyright Commonwealth of Australia, CC BY 4.0 subject to exceptions.

V4 DERIVED FINDINGS AND REPRODUCTION
Chart 09: difference between annual net gains: 85,130 - 66,650 = 18,480.
Construction: 15,603 - 10,018 = 5,585. Professional/scientific/technical: 12,804 - 8,663 = 4,141.
(5,585 + 4,141) / 18,480 = 52.6%. All19divisions plus unknown reconcile to18,480.
Chart10: non-employing entries354,098 minus exits299,120 plus net movement28,127 =83,105.
28,127 /83,105 =33.8%. Employing:106,363 -76,211 -28,127 =2,025.
Reclassification cancels nationally; it is not a share of national growth caused by movement.
industry-step-up-2026.csv includes both years' counts, net changes, exact growth-rate changes and sourcecells.
Classification of exact rate changes is17accelerations/2slowdowns; printed one-decimal rates conceal tiny changes.
All CSV rows now carry stable source URLs, release dates, reference period, licence and credit.
Full council-industry input rows allow independent recomputation of division maxima/ties and threshold comparisons.
Run the text-supplied reproducibility script with Python's standard library from the extracted pack.
The council finder searches/filters/sorts the 536 council summaries, exports visible rows and individual source-complete rows.
Its displayed main rank always means the national qualifying135, never a reranked state subset.
No network request, data collection or third-party dependency is used by the finder.
