Report Free TAFE by starting cohort, so the argument about it can be settled
More than 800,000 enrolments and about 258,000 completions. One side calls that a failure; the other says the data is immature. Both are reasoning from a statistic that cannot answer the question, and the program was made permanent at $1.6 billion before anyone fixed that.
Steps this would change
Rationale
Free TAFE was made permanent by the Free TAFE Act 2025, with more than $1.6 billion committed to 2034-35 for at least 100,000 places a year from 2027, on top of around $30 billion over five years through the National Skills Agreement.
The published headline is more than 800,000 enrolments and about 258,000 completions since January 2023.
That ratio is doing work it cannot do, and both sides of the public argument are misusing it.
The critics' version — roughly a third completing, therefore a failure — is comparing a stock of enrolments accumulated over three years against a flow of completions that has not finished arriving. Vocational qualifications take years. A student who enrolled in 2025 has not failed to complete; they have not completed yet.
The defenders' version — the data is immature, completions are conventionally measured over four years, rates are rising — is true and is also unfalsifiable in its current form. It will remain available as an answer indefinitely, because there is always a recent cohort still in progress.
Neither can be resolved with the statistic being published, and the honest position is that nobody currently knows whether Free TAFE works, including the people confidently asserting that it does and the people confidently asserting that it does not.
The fix is a reporting change, not a policy change. Publish outcomes by starting cohort: of the students who commenced in a given year, what share had completed, were still enrolled, or had withdrawn, at 1, 2, 3 and 4 years — broken down by qualification and by the priority cohorts the program targets, since places are deliberately directed to people facing barriers to education and a lower completion rate among them may be the program working rather than failing.
The data exists. It is collected at enrolment and at completion, against a Unique Student Identifier designed for exactly this. What is missing is the published presentation.
Days saved: zero. This changes no elapsed time. It changes whether a $1.6 billion permanent commitment can be evaluated, which on this site's own standards — every number sourced, and the working shown — is the more important property.
Who acts: the Commonwealth, in partnership with the states that hold the enrolment data.
What is being traded away
Cohort reporting will produce a number that looks bad, and it will be misused. Completion rates for students facing genuine barriers to education are lower than for the general population, by design and for good reasons. Publishing that clearly hands a weapon to anyone who wants to argue that disadvantaged students are a poor investment. The mitigation — reporting alongside the cohort characteristics and against a comparable baseline — is exactly the sort of nuance that does not survive a headline.
Completion is not the only outcome, and elevating it distorts. A student who completes four units, gets the job, and never finishes the qualification is counted as a non-completion and may be a policy success. Employment outcomes matter more than certificates for much of this cohort, and a reporting regime built around completion will pull provider behaviour toward completion, including by enrolling students likely to complete rather than students who need the place.
And this site should own its own exposure here. This project publishes a figure — 800,000 against 258,000 — that it then says cannot support a conclusion. Publishing a number while explaining why it does not mean what people will take it to mean is a genuine risk of being quoted for the number and not the explanation. It is done because the alternative is not mentioning the largest public argument about the program, and a database that omits inconvenient statistics is doing the thing it criticises.
Precedent
Cohort tracking is standard in higher education reporting and in NCVER's own longitudinal work. The data is collected; it is the published presentation that aggregates it into a ratio that cannot bear the weight put on it.
Sources
- Free TAFE Act 2025 (Cth); more than $1.6 billion to 2034-35 for at least 100,000 places per year from 2027
- Prime Minister of Australia: more than 800,000 Free TAFE enrolments and about 258,000 completions since January 2023
- DEWR, Fee-Free TAFE Program Snapshot, Quarter 1 2026
- Deakin Centre for Research for Educational Impact / The Conversation, analysis of Free TAFE completion measurement, August 2026