At its peak, more than 3.8 million Australians were on JobKeeper - three in every 10 Australian workers.
Adding in those workers already employed by government, it meant four in every 10 received a paycheck that originated from government, more than in Russia.
Yet when JobKeeper ended at the end of March, it looked like a mere blip in employment. The unemployment rate actually fell, for the sixth consecutive month.
The Bureau of Statistics said the cutoff had no "discernible impact".
Treasurer Josh Frydenberg went further. The economy had "strengthened, even after the end of JobKeeper".
Since the end of JobKeeper 132,000 people had come off income support.
The Treasurer is right. Since March the number of Australians on JobSeeker and related support payments has fallen 9 per cent.
Changed rules pushed people off benefits
But it's possible for people to come off benefits at the same time as people are losing jobs, especially if something else is driving them off, as it was at the end of March.
At the end of March the coronavirus supplement used to top up unemployment benefits stopped. The payment dropped from $715.70 to $620.80 per fortnight.
And jobseekers were once again required to search for a minimum of 15 jobs a month, climbing to 20 from July.
While burdensome for employers (if all of Australia's jobseekers actually applied for those jobs, employers would be lumbered with 17 million applications per month, climbing to 23 million) it's also unhelpful for jobseekers.
There's evidence to suggest jobseekers get real jobs sooner if they don't have to go through charades.
The "dob in a jobseeker" hotline will have further dissuaded people from applying for benefits.
These changes make the drop in the number of claimants understandable, much more so than the suggestion they got jobs, which in net terms they did not. Employment fell after the end of March, by 30,600 according to Bureau of Statistics figures which will be updated on Thursday.
As many as 97,000 fewer workers?
How is a drop in employment consistent with a drop in the unemployment rate?
The unemployment rate fell to 5.5 per cent in April not because employment grew but because 33,600 people who had previously identified themselves as unemployed dropped out of contention, changing their status to "not in the labour force".
Had they continued to not work but continued to describe themselves as "unemployed" the unemployment rate would have been 5.7 per cent.
And it would have been higher still if the number of people shifted to zero or reduced hours with the end of JobKeeper had been taken into account. The number employed fell 0.2 per cent, but the number of hours worked fell 0.7 per cent.
Rough maths suggests this means the number of people actually working might have fallen 94,100.
An analysis prepared by Melbourne University employment specialist Jeff Borland for the Fair Work Commission puts the number between 45,000 and 97,000.
He gets 45,000 by comparing the number of people who left employment between March and April this year with the number who left between March and April in previous years.
He gets 97,000 by comparing the average (rapid) growth over the previous four months as we emerged from recession with the growth between March and April.
One in 11 JobKeeper jobs
A touch over one million Australians remained on JobKeeper to the end, suggesting that as many as 10 in every 11 of them kept their jobs when JobKeeper ended. One in every 11 might have lost their jobs.
Borland says his calculations are best interpreted as showing that the end of JobKeeper constituted a "temporary stalling" rather than a major setback to recovery.
The Australians most knocked around were the youngest. Since the end of JobKeeper women under the age of 30 have on balance lost jobs while women over that age have continued to gain jobs. Men under the age of 40 have lost jobs while men over that age have gained them.
The treasury's deepest concern about jobs since the start of Covid (it mentions it right at the top of that section of the budget) has been scarring.
The unlucky young people who happen not to secure jobs during downturns can fail to secure them for years, being passed over by newer, fresher young people who are barely affected.
Even where these people get jobs, if they enter the market when the youth unemployment is 5 percentage points higher than normal they can expect to earn roughly 8 per cent less in their first year, and 3.5 per cent less after five years. It takes around a decade for the effect to fully disappear, and it's worse for women than men.
Vacancies, plus mismatch
Australia's record-high vacancy rate (2 per cent of all jobs were vacant at the end of March) makes it look as if scarring needn't be much of a concern. But jobs are vacant for reasons.
It might be that the mix of jobs we will need is changing, or that employers can for the moment no longer rely on migration to give them the mix of skills they need, or it might simply be that the general bounce back in jobs has been so fast that the right employers and the right workers are still working out how to find each other.
Many businesses will die as a result of the end of JobKeeper. Businesses are forever dying. Some have been kept alive for longer than would have been the case, and some have exited JobKeeper into a changed environment.
We've managed to end JobKeeper without a catastrophe, but that doesn't mean there hasn't been damage, and it doesn't mean young lives won't be scarred.
After a textbook exit from a recession with the sharpest V-shaped recovery ever, it would be awful if we left a slice of young Australia behind.
- Peter Martin is a former economics editor of The Canberra Times. This article was first published in The Conversation.