Reducing spending on consultants, contractors and labour hire is a signature commitment and multi-billion-dollar savings have been claimed against it. Two substantial datasets bear on whether the claim holds. AusTender publishes contract-level detail, and the Department of Finance publishes the Audit of Employment, which reports the external labour workforce and total expenditure on consultants, contractors and labour hire -- $6.0 billion and 21,970 full-time equivalent staff in 2023-24.

What is missing is the join. Neither series is presented in a way that lets a reader reconstruct the savings figure: which procurement categories were counted, what the counterfactual baseline was, how roles converted to public service employment are valued against the contracts they replaced, and how movement between the consultancy, contractor and labour hire categories is treated. The method that produced the headline number has been described in documents released under freedom of information rather than published alongside the claim.

The result is that several different expenditure figures for the same years circulate legitimately, Senate committees have repeatedly probed which of them the savings are measured against, and the question stays open. A government confident in the claim has the strongest interest in closing it, because the data needed to check the arithmetic is already public and only the arithmetic is missing.

The Audit of Employment reports are the best available starting point and give a consistent external labour series across years. AusTender can be aggregated by category to produce a spend series independently.

Either will produce a number. Neither will produce the government's number, and without the definitions there is no way to tell whether the difference is disagreement or double counting.

Checked on 2026-09-01 and again on 2026-09-02. This entry is weaker than it was first drafted, and the record of how it was narrowed is kept below rather than tidied away.

The earlier version implied there was no published measurement framework at all. There is one: the Audit of Employment is a recurring published report with a consistent series for external labour expenditure and headcount, and it is a better dataset than the entry originally allowed for. The verdict has moved from no to derived and not_published has been dropped, because the material needed is largely public and the obstruction is that it does not join up.

The one thing that would settle the entry is whether Finance has published, anywhere, the reconciliation between the Audit of Employment series and the claimed savings. That was the open question after the first check.

inconsistent leads because the obstruction is definitional rather than one of secrecy: the same underlying transactions support several different totals depending on categorisation, and no published rule says which categorisation the headline figure used.

Useful boundary case alongside How often is a government contract extended or varied without going back to market?. In both, everything needed is public and the published shape answers a different question from the one being asked. Here the missing piece is a page of definitions.

Related: Which companies receive the most federal government contract money?, Commonwealth entity annual reports.

Re-checked on 2026-09-02, narrowing the open question above as far as search could take it.

Two of the three things needed are published. The 2023-24 Audit of Employment sets an explicit baseline -- 21,970 FTE and $6.0 billion of expenditure on consultants, contractors and labour hire -- and the APS Strategic Commissioning Framework publishes definitions distinguishing labour hire, contracting and consulting, including the direction-and-control test that separates a consultancy from a contractor.

The third is not. The calculation connecting that baseline to the headline savings figure was located only in documents released under freedom of information, not published alongside the claim. That is the whole of the residual gap, and it is a page of arithmetic rather than a dataset.

The Audit of Employment was downloaded and searched on 2026-09-03, which upgrades this entry to verified. It confirms the figures relied on here: total reported external labour of 21,970 FTE in 2023-24 and $6.0 billion of expenditure on consultants, contractors and labour hire, broken down by agency.

The residual gap is unchanged and now rests on a document that has been read rather than described. The baseline is published; the arithmetic joining it to the savings claim is not.