// fiscal_flow_analysis

🏛️

AU Federal Budget 2024-25 — Trace the Money

$738.5B decomposed into a 4-level Sankey — from revenue bucket to sub-program. Follow a dollar from your tax return to NDIS, defence, or debt servicing.

Gov DataSankeySVG VizFiscal AnalysisATONDIS

$738.5B

total budget 2024-25

all Commonwealth outlays

47.1%

from income tax

$347.6B individual

$58.6B

structural deficit

net borrowing to close gap

4 levels

decomposition depth

bucket → source → portfolio → program

loading visualisation...

// key_findings

  1. 01.

    Individual income tax is 47.1% of all revenue — bracket creep is quietly the government's most reliable revenue lever, growing automatically as wages rise without a budget vote.

  2. 02.

    NDIS at $42B is the single fastest-growing line item. Demand is structural (disability prevalence + scheme access), not policy discretionary — making it functionally unbudgetable.

  3. 03.

    Net borrowing funds 8% of spending. The deficit is structural, not cyclical — even at full employment (3.8% in 2024), the Commonwealth spends $58.6B more than it collects.

  4. 04.

    Defence grew 23% in two years. AUKUS commitments and the Indo-Pacific strategic reassessment have rebaselined defence as a non-discretionary expenditure bloc.

  5. 05.

    GST ($89.6B) is distributed to states unaltered — the Commonwealth collects it but the federal government doesn't control how it's spent, creating accountability diffusion.

  6. 06.

    Superannuation tax concessions (~$50B, off-balance-sheet) rival NDIS in fiscal cost but appear nowhere in the main budget papers — the invisible transfer to high earners.

// data_lineage

  • Budget Paper No.1 2024-25 (Treasury)
  • MYEFO 2024 — Mid-Year Economic & Fiscal Outlook
  • ATO Taxation Statistics 2022-23
  • NDIS Quarterly Report Q4 FY2024
  • AIHW — Government Health Expenditure

// built_with

React 19TypeScriptCustom SVG Sankey (no D3)Force-directed node layout (hand-built)Next.js

// methodology_note

All Sankey node positions computed analytically from column assignments and vertical flow balancing — no library layout engine. Values cross-validated against MYEFO final estimates.