A case for universal free childcare

The idea of increasing the amount of public money going to childcare keeps popping up. For instance, Eleanor Robertson and Anne Summers recently singled it out as a desirable step forward for Australia. I think it sounds great, but I’ve been wondering how feasible it would be. I decided to take a closer look at the numbers.


Increasing public support for childcare would mean transferring money to households with kids (author interest disclosed) and taking it from those without. Is it reasonable to expect workers without children to pay for raising other people’s kids?

I believe the answer is a clear yes.

Public support for childcare pushes against one of the main reasons for women’s economic disadvantage. As Andrea Fox puts it,

the economy is built upon the toil of unpaid care, largely undertaken by women. That the provision of this essential care work leaves women financially depleted is evidenced by their eventual over-representation in numbers on the age pension.

Public daycare would help to correct this imbalance, and also counteract the tendency for policy to act as if men were the natural inhabitants of public life, with a woman’s situation implicitly treated as abnormal or deviant.

But there are self-interested reasons for citizens without children to support this policy, even if they don’t care about gender imbalances. An expansion of daycare services would generate a net increase in GDP, due to the economies of scale in childcare—one person can look after five other workers’ kids. And its impact would be long lasting, since it’d help to keep a cohort of women in the labour force for decades. (People are much less likely to re-enter the labour force, having left, than they are to stay in it when employed or looking for work.) A sustained increase in economic output will benefit everyone, including through higher government revenue that can fund other public services. For that reason, the Grattan Institute has argued that increasing women’s participation in the labour force should be one of the government’s main priorities.

What exactly are we talking about?

The policy I suggest is upgrading the public subsidy for childcare to 100 per cent, capped at a certain dollar amount per hour. In other words, Medicare applied to preschool.

A simple Econ101 model would predict two things: as the sticker price approaches zero, consumers will demand a huge number of childcare hours per week; and, if consumers never see a bill, firms will jack up their prices until most of the public subsidy is going into monopoly profit. But I think in this particular case a 100 per cent subsidy would work. Firms could only earn more than the designated subsidy cap by explicitly charging consumers, as in the Medicare system. And there are natural limits on the amount of childcare each household can use, such as transport costs and the demands of family life, so it’s not essential to use prices as a rationing device.

The subsidy should not be means tested. Consequently, a chunk of public money will flow to middle-class and well-off households, as usually happens with publicly provided services, but that’s outweighed by some major benefits. A 100 per cent subsidy is simple to understand and straightforward to implement. The current system of rebates is complex and confusing, and requires a lot of interaction with Centrelink. (As a matter of explicit policy, talking to Centrelink is made to be as annoying as possible, a strategy designed to discourage people from seeking welfare.) As well, a 100 per cent childcare subsidy eliminates the disincentive to work, by lowering and smoothing out the effective marginal tax rate faced by a working woman with kids.

The voluntary take-up rate of the new subsidy will probably be different across the income distribution anyway: households in the upper half of the income distribution are less likely to be budget-constrained when deciding whether to use childcare.

How much would it cost?

The next step is to figure out roughly how much money we’re talking about. I don’t have access to the complex models Treasury uses to estimate policy impacts, but it’s enough at this stage to write something out on the back of an envelope. If the annual cost turns out to be, say, $50 million, then the government could implement it without breaking a sweat, and if it’s $500 billion then we should forget it.

Right now there are just under a million children using childcare services. That includes long daycare, family daycare, and before/after school care.

Another quarter of a million kids need childcare, in the sense that their parents have told the ABS that they would increase their working hours if childcare was available. You end up with a pretty similar number if you start by looking at the participation gap between men and women aged 20 to 44. Closing that gap means putting 550,000 women into the labour force. Each of those women has, on average, 0.6 children under 11, giving an upper estimate of about 300,000 extra children needing childcare.

In 2014, the average household that used childcare was spending $70 a week on it, after subsidies. Increasing that to $80 to make a generous allowance for inflation, and making the conservative assumption that the $80 fee applies to each single child, gives the result that we need to find an extra $4.6 billion. (The current budget for childcare subsidies is roughly $7 billion.)

Alternatively, we could assume that each child uses 2 days of childcare per week, at a cost of $10 per hour. Those are conservative assumptions, since the current averages are more like 1.5 days per week at $8.50 per hour, so it allows for a fair number of current childcare users to increase their hours if they want. Multiplying that hourly cost by the number of children needing care then produces an annual bill of $11.7 billion, or a $4.7 billion increase over what’s now funded.

(I’ve printed the first decimal place on those figures, but I don’t mean to suggest any great precision. If you adjust the assumptions this way or that, you can make the net impact move up or down by about a billion dollars.)

Based on this analysis, I think it’s clear that we could fully fund universal childcare if we chose to. A $5 billion cost would require a single major tax change. For example, reintroducing an estate tax could raise about $3 billion, and broadening the GST with a generous compensation package would raise a net $11 billion. Given that the childcare policy would lead to more employment, and hence more income tax revenue, the net cost would probably be a fair bit less than $5 billion, but we’d need a proper econometric model to figure that out.

The cock-up in Quebec

Okay, so let’s imagine that we increase the demand for childcare by $5 billion or so. What might happen on the supply side?

Quebec introduced a near-complete public subsidy for childcare, charging only a nominal daily fee, in 1997. The policy was taken up enthusiastically, as new childcare providers rushed into the market, and female labour force participation soared. However, the quality of the childcare itself has been terrible, with wide-ranging negative impacts on the children that have used it.

My hunch is that we won’t have the same problem. The national standards used in Australia seem to be working pretty well, at least as far as I can tell, based on the half-dozen daycare places my kids have used over the years. Even at a mediocre daycare centre, where meeting the standards is just a box-ticking exercise, the kids end up having a stimulating and well-structured day.

If our childcare sector is operating under stringent regulations, then one might imagine that we’ll face the opposite problem to Quebec. Perhaps the amount of childcare provided will stay more or less fixed, with prices rising until the industry soaks up the extra money.

But that’s not the picture that emerges from the Productivity Commission’s mammoth report on childcare. They find that childcare suppliers have responded pretty flexibly to the steady growth in demand over the last decade. There’s been a big increase in home-based care, which can respond more quickly to ebbs and flows in demand than childcare centers. And there’s a precedent for what a demand-side stimulus might do: during the GFC, the government increased the amount of Child Care Rebate. In the months following, childcare prices grew at an annual rate that was 5 percentage points higher than usual, then returned to their average rate of growth as more supply came online. In other words, exactly what’s supposed to happen.

Conclusion: Pay up

There’s a clear need for more public childcare funding. Making childcare universal and free to use, like Medicare, is the simplest way to get that done. Since it will cost somewhere around $5 billion, we’ll probably need to raise taxes a little. It seems reasonable to expect that the extra childcare will be supplied efficiently, though there may be a bit of an adjustment period. The benefits will include higher workforce participation, an improvement in one of the structural factors that impinge unfairly on women, and a chunky increase in GDP. Let’s make it happen.