The argument he actually made
Follow the sentence to its end. If a skilled role can be performed from another country, then the case for moving a worker here to do it is not a labour shortage. It is a preference. And a preference does not justify a permanent residency pathway with a family-reunion entitlement attached to it.
That is the automation argument, made by a One Nation MP, in defence of something else entirely.
The same logic runs through the technology. Remote delivery moves the work to the worker. Large-language-model deployment removes a share of the work altogether, and it starts with the entry-level rung: the drafting, the reconciliation, the first-pass analysis, the administration. Those are the exact occupational categories the temporary skilled and graduate pipeline is sized for.
Whether that displacement lands as fast as its loudest advocates claim is an argument, not a fact, and it should be read as one. What is not an argument is the direction. A country importing people on permanent pathways to fill entry-level knowledge work is buying an asset whose value is falling while the liability compounds.
The labour shortage is temporary. The visa pathway is permanent. Nothing in the last three years has changed that asymmetry, and remote work sharpens it.
What One Nation actually published
On 14 September, Pauline Hanson announced the party's migration reset alongside Barnaby Joyce. The numbers matter, so here they are, kept in their own units.
- A reduction of 750,000 in the temporary migrant population across three years. That is a stock of people on temporary visas, not a net flow.
- Net overseas migration negative for those three years.
- Then a cap of 130,000 a year on net overseas migration, reviewed annually.
- Student visas capped at 100,000 a year.
- Families of temporary skilled workers and international students barred, which the party puts at more than 100,000 people.
Those are four different measures and they are not interchangeable. The 750,000 is a population. The 130,000 is a flow. Anyone quoting one as the other, in either direction, is arguing with a number that does not mean what they think.
Joyce put the student half plainly: "we're just not having this sort of universities used as basically a backdoor to migration."
For context, the flow has already fallen a long way without any of this. Net overseas migration peaked at 538,340 in 2022-23 on ABS financial-year figures, then 429,160 in 2023-24, then 305,570 in 2024-25. The highest annual figure on record was 556,000 in the year ending September 2023, and it has fallen for seven consecutive quarters since. In 2024-25, arrivals were 568,000, down 14 per cent, while departures rose 13 per cent to 263,000.
State that honestly, because the section that ignores it gets corrected. The intake is not still climbing. The composition argument never needed it to be.
Why the reset is costed against the population, not the flow
Notice which measure One Nation chose. The headline is 750,000 off the temporary migrant population, not a net overseas migration number. That is the right instinct and it is worth spelling out why.
Net overseas migration is a net annual flow: arrivals minus departures, counted on the twelve-months-in-sixteen rule. It is the figure every headline uses, and at 305,570 for 2024-25 it is correct. It is also not the number anyone experiences. In that same year 568,000 people arrived, and the net figure is what is left once 263,000 departures are subtracted.
The standing population is the one you queue behind, and Home Affairs publishes it quarterly. At 31 August 2026 there were 2,878,973 people in Australia on temporary visas. Take out New Zealand citizens on the Special Category visa (732,887), visitors (308,097) and crew and transit (12,586), and 1,825,403 remain: students, temporary graduates, temporary skilled workers, working holiday makers and people on bridging visas. That is six times the annual net flow. It already includes partners and children, because the Department counts all applicant types rather than primary applicants alone.
Neither figure is wrong. They answer different questions, and only one of them describes the pressure on a rental market or an emergency department. A reduction plan costed against the flow can hit its target while the standing population barely moves. One costed against the population cannot. That is the strongest defence of how this reset is written, and it is the reason the 750,000 and the 130,000 must never be quoted as each other.
This is a schedule, and that is the point
The reason this deserves support is not the size of the cut. It is the shape of it.
A schedule has a number, a period and a review. It can be modelled. It can be costed. It can be defended in a studio by someone who did not write it, and it can be held against its author in three years. "Cut immigration" cannot do any of those things, which is why it has been said for twenty years and delivered nothing.
That is the whole case for One Nation leading the next government rather than supporting one from the crossbench. The Liberals under Angus Taylor have moved genuinely on lowering the intake, and the Nationals are right that domestic energy costs are the lever under everything else. Neither has put a binding, numbered, reviewable reduction on paper. One Nation has. The party that wrote the schedule should be the party holding the portfolio, with the Liberals and the Nationals supporting it.
The dependant cut belongs in the schedule
Then there is Farley's actual proposition, which is the part the coverage treated as a gaffe and which is in fact the most consequential line in the package.
Asked about cutting visas for the partners and children of Australian citizens, he said: "It's a hypothetical question, but can we grow our health system, our education system and other infrastructure we need to make sure the existing population that's here ... are not under pressure?"
That is the right question and he should not have been talked out of asking it.
Start with what the dependant cut actually does. One Nation puts the ban on families of temporary skilled workers and students at more than 100,000 people. That is not a rounding item inside the 750,000. It is one of the largest single components of it, and unlike a cap on a flow it comes off the standing population, which is the number that determines what a rental market or an emergency department actually carries.
It also lands exactly where the automation argument lands. Dependants of temporary skilled workers hold work rights, and the Home Affairs workbook counts all applicant types rather than primary applicants alone, so they are already inside the 1,825,403. They compete for the same entry-level roles that large-language-model deployment is absorbing first, and they compete with the 19-year-old Australian looking for a first full-time job. Every argument for cutting the graduate pipeline is an argument for cutting the dependant stream with it.
Then the arithmetic on the citizen side, which is the part nobody in the coverage followed through. The permanent family stream runs at 52,460 places a year against a partner-visa backlog of roughly 120,000 and growing. The backlog is more than twice the annual program. A program that has already accumulated two years of unmet demand is not delivering for anybody: not for the Australians waiting in it, and not for the country absorbing the arrivals when it finally clears. A smaller, honest, faster program is better government than a larger one that quietly runs a two-year queue and calls it a service.
On the legal objection. Abul Rizvi, a former deputy secretary of the Immigration Department, says such a cap would breach section 87 of the Migration Act. In the same breath he concedes that successive governments have already limited these visas administratively for years. Both of those cannot carry equal weight. If the practice has been running under governments of both colours without a court stopping it, then "it cannot lawfully be done" is a description of a risk, not a barrier, and it is a strange objection coming from a former deputy secretary of the department that ran the settings this section spends its time criticising. He administered the system. He is not a neutral party to the question of whether it should be dismantled.
If the section is genuinely a problem, amend it. Parliaments amend migration law constantly, usually to make it more permissive. A clause that stands in the way of a reduction the electorate has asked for is an argument for changing the clause.
So: keep it in. The dependant ban is the difference between a schedule that reduces the standing population and one that only slows the flow into it.
What the reduction is actually for
The case for cutting hard is not that the country is full. It is that Australia has been growing by adding people instead of raising output per person, and the technology now arriving makes that model worse rather than better.
The alternative is the one worth selling. Cheap domestic energy on a resource base most countries would trade their navy for, used at home rather than shipped raw, is what makes Australian-owned compute viable. It lowers industrial input costs. It raises wages for the people already here. That is a promise, and a promise converts where a warning only alarms.
A schedule is how you get there from here. One Nation wrote one. It should lose the family-visa clause, keep everything else, and lead.