JobKeeper is a cornerstone of the Australian government's economic strategy for dealing with the impact of COVID-19.
Lockdown measures imposed to contain the spread of the virus have a devastating impact on the economy. Preventing deaths from COVID-19 was the top priority, but the government has also been aware of the consequences of that strategy for the economy and for employment in particular.
Unemployment is terrible. It causes massive pain and suffering. People who lose their jobs are more likely to commit suicide; modelling at the University of Sydney's Brain and Mind Centre suggests there could be an additional 1500 suicide-related deaths over the next five years due to the economic impact of COVID-19.
The more we can do to reduce the extent and the duration of unemployment the better off Australia will be.
The aim of the JobKeeper scheme is to keep employees attached to work. That's important. The longer a person remains unemployed the harder it is for them ever to regain employment. Remaining attached to the workforce is key to obtaining future work. If someone can't find a job after two years their chances of ever finding work again drop dramatically. So keeping employees connected to their employer in hard times, when the employer has cut back or closed, is good policy.
JobKeeper pays a subsidy to businesses that have had a drop in their income of 30 per cent (businesses with a turnover less than $1 billion) or 50 per cent (businesses with a turnover more than $1 billion).
It also applies to charities, other than schools and universities, where turnover has dropped by 15 per cent or more. There appears no real science to the income drop point at which JobKeeper applies.
The JobKeeper scheme has flaws - hardly surprising given the speed with which it was developed.
In normal times, a multi-billion dollar business subsidy program would take months, maybe years, to develop.
It could come from a major inquiry or commission, or from an idea worked up by a department, considered by the minister through numerous iterations, and then turned into a new policy proposal put to cabinet with coordination comments from all interested departments.
It would be subject to detailed scrutiny by the expenditure review committee and vigorous debate in backbench policy committees, the government party room and often other parliamentary committees.
The final step, passage through the Senate, is the backstop: a final opportunity for potential problems to be identified and sorted out via amendments to enabling legislation.
In hindsight, if the Australian government had been given months rather than days to devise JobKeeper, it might have been done better. It also might have been worse.
None of that happened with JobKeeper. It was done quickly for obvious reasons.
There are problems. Employers have to find money to make the payment up front (the subsidy is paid retrospectively), which some have found difficult.
The subsidy is due to end on September 27, regardless of economic conditions, leading to talk of employment numbers falling off a cliff.
JobKeeper does not cover some categories of employees including short term casuals and most temporary visa holders.
Businesses have to pass JobKeeper payments on to their employees. However, there's no test of whether they were planning to keep the staff on anyway.
For some businesses the payment thus represents a windfall gain: funds freed up from the wages bill as a result of the JobKeeper subsidy can be spent elsewhere.
It's also been observed that low paid staff can in some cases actually receive more under JobKeeper than their employer had previously been paying them. It is paid as a flat amount, regardless of hours worked.
Steven Hamilton at the Australian National University's Crawford School has identified these and other failings, but still argues the scheme should be left unchanged.
This is sound advice. In hindsight, if the Australian government had been given months rather than days to devise JobKeeper, it might have been done better. It also might have been worse - extensive consultation with interest groups can sometimes make policy convoluted and unworkable.
We will never know. It is the policy we now have. Although it may have problems it is also undeniably delivering economic stimulus and keeping thousands more people employed than without the scheme.
More recently Treasury and the Australian Taxation Office have been catching flak for errors in the publicly announced estimates of both the number of employees JobKeeper would cover and its likely cost.
The headlines ranged from the relatively mild "accounting error" to mistake, snafu, bungle and blunder.
A media release from Treasury provided the background.
The reason for the ATO overestimate of the scheme's coverage was said to be errors made by businesses in their application forms, errors not picked up by the ATO.
A more fundamental reason is likely to be confirmation bias. The incorrect applicant numbers were roughly in line with the original Treasury spending estimate (even though that was developed on a totally different basis) so was what the ATO expected to see.
The original Treasury estimate for spending was wrong for different reasons. The level and impact of health restrictions was not as great as estimated.
That's a good news story - the lockdown, and its negative effects, did not last as long as originally feared. Actual spending on government programs coming in higher or lower than estimated is not unusual, even with long established spending programs.
Having said that, there were much earlier signs the numbers first announced by the government were wrong.
More than a month ago, in an April 29 blog post, Newcastle University's Professor Bill Mitchell questioned the cost of JobKeeper, saying "the $A133 billion figure does not make sense given other statements the Treasury have been making about the scheme". He has been proven correct.
Now the dots have been connected and instead of the original $130 billion the JobKeeper scheme is estimated by Treasury to cost $70 billion.
Despite all this, JobKeeper is meeting its purpose. On an effectiveness scale, it rates reasonably well. Some of the other COVID-19 economic response policies may not.
For example, allowing people to withdraw superannuation early is good policy if economic recovery is rapid and then sustained. If once the crisis is over the people who drew on their superannuation to tide them over get better and higher paid jobs that last the rest of their lives, they will be better off in net terms.
A more worrying prospect is that recovery will be slow and patchy, leaving many of those people much worse off in the future. We then face the prospect of the Australian government having to devise new policy to patch over the problem created by the previous policy.
One encouraging aspect though is that in doing the further policy work it appears the government is turning to its public service for advice. It is already apparent that recent policy announcements are better thought out than those at the start of the crisis.
This is not because we have better information - there are so many unknowns about COVID-19 that policy is a textbook exercise in managing uncertainty. It is because the more often public servants work on policy the better they get at it: and this only happens when governments ask for policy and take action based on the advice.
If the problems with estimating the cost of JobKeeper lead the government to turn away from seeking policy advice from the public service it will be a bad outcome. Encouragingly, both the Treasurer and the Prime Minister have said they continue to have confidence in the public service.
This will be required for some time to come. While we appear to have contained the spread of the virus (keeping fingers crossed there is no second wave) the economic impact will continue for the foreseeable future.
If the economy does not recover quickly, additional stimulus measures to keep people in jobs will be needed.
Policy advice from the public service will be vital for the success of these measures.
Sign up for our newsletter to stay up to date.