How do we feel about a national daycare program?

Those who use simple heuristics to make up their minds about policy questions (e.g. “government good, markets bad”) will undoubtedly already know how they feel about the federal NDP’s recent announcement of a bold new plan for a national daycare system. The headline feature of the system is that it will be subsidized, so that the cost to parents should come in at about $15 per day.

There has of course been a sense of welcome relief that the NDP has stopped doing small-bore, pseudo-populist politicking, and is actually coming forward with a genuine proposal to expand the Canadian welfare state, moving it away from the American toward the Scandinavian model. But if you look at the issue from first principles, or from a social justice perspective, public involvement in the daycare sector is not an easy issue to assess. In this respect, it is not like the other two “big fish” out there in the policy space – carbon pricing and national pharmacare – which are no-brainers by comparison. The efficiency case for carbon pricing is both clear-cut and overwhelming, the only reason we don’t have it is that the federal government has essentially been captured by a special interest group. Pharmacare as well would generate significant efficiency gains, the failure to include it in the original plans for medicare was simply a consequence of the fact that drugs made up such a low percentage of health care spending at the time. Again, the only reason we don’t have it is lack of political leadership – specifically, a willingness to face down the anti-tax hysteria that a program of this sort would inevitably confront.

With daycare, however, there is no obvious market failure, and therefore the most uncontroversial rationale for state involvement is absent. To the extent that a national system of subsidized childcare centres is justified, the argument would have to appeal to some conception of equality or distributive justice. And while there are some programs that have such a rationale (Old Age Security comes to mind), the arguments are inherently trickier, because egalitarian arguments involve comparing the status or treatment of individuals against one another, in a way that efficiency arguments do not. Furthermore, these arguments run into an obvious challenge, which is that daycare is only one of the different modalities of childcare that are available to parents (the other two major ones being stay-at-home parenting and nannies/babysitting). Thus there is a potential violation of equality in subsidizing only one of these three modalities – the state may thereby fail to show “equal concern and respect” for the different choices made by people with different values. This is the argument that informs existing Conservative government policy (i.e. the “Child Benefit” payments), and it is not one that should be casually dismissed.

(I should pause perhaps just to note that bringing children into the world does not result in any signficant positive externalities, and so there is no “market failure” in that regard. Sometimes when you gaze lovingly at your own child, it’s difficult to imagine that you haven’t done the world a huge favour by bringing this special creature into existence. This is, however, simply bourgeois sentimentality, which a moment’s reflection is sufficient to disabuse anyone of. The world is not underpopulated. It is actually the people who don’t have children who are, on balance, generating the positive externalities. It is irresponsible for the state to encourage people to have children – and to the extent that any state does, it is usually part of a dubious ethnonationalist project.)

So the first question, when trying to formulate an egalitarian argument for public childcare, is to ask what sort of equality it is seeking to promote. Most obviously, such a program gives money to the parents of young children (indirectly, by subsidizing their consumption of a particular good). And as Stephen Gordon and others were quick to point out, “universal” social programs of the sort that the NDP is proposing are generally regressive, because they subsidize the consumption of both the rich and poor alike. When my children were in daycare, not that long ago, I was paying around $70 per day, a bit less as they got older. (Friends in Quebec find this number somewhat eye-popping – but this is at the University of Toronto daycare, which is actually subsidized by the university. Prices have even gone up since my kids were there. Incidentally, there are still huge queues for positions, even at these prices, which says something about the availability of spaces in downtown Toronto.) If government subsidies had reduced my fees to only $15 per day, then with two children that would have put over $25,000 in my pocket per year. Needless to say, that’s a very large chunk of money, which frankly myself and a lot of other Toronto fatcats don’t need. A means-tested program (which is what we have right now), by contrast, gives the money only to parents below a certain threshold of income, who find daycare unaffordable.

Gordon though is a bit quick to assume that there is only one worthwhile form of redistribution (viz. transferring money from the rich to the poor). One might instead choose to look at subsidized childcare as an instance of what Gosta Esping-Andersen and John Myles refer to as “horizontal redistribution,” or more prosaically, “income smoothing.” Basically, it involves people using the state to redistribute their own income to themselves at different stages of life. If you look at health care and pensions, for instance, and take only a snap-shot, it looks like wealth being redistributed from the young to the old. But if you take the long view (along with seeing the overlapping-generations structure of the system), you can think of it as people redistributing wealth from themselves-now to themselves-later. They might have a variety of reasons for doing this (even setting aside the insurance functions of the way the state pools these savings). The most obvious is self-control problems. The other one is failure or limitations of capital markets, so that people have inadequate access to credit, making it difficult for them to match their income to their consumption needs on an annual basis. So they wind up with not enough money when they most need it (e.g. when they are young) and too much money when they no longer need it (e.g. when they are old). Presumably this is starting to sound familiar…

Before getting into the details, it is worth observing that subsidizing childrearing is actually something that the welfare state does a lot of, most obviously with maternity and parental leave benefits, generous tax deductions for childcare expenses (not to mention sports!) as well as free primary and secondary education. It is worth pausing to ask why any of this should be happening. If people decide to have children – and in this day and age, it is clearly a decision – why shouldn’t they just pay for them themselves? Many people describe raising children as the most fulfilling part of their lives, so there is clearly a motive of self-interest at work. Most people love their children more than their houses, so why shouldn’t they pay for their children, just like they pay for their houses? Why should others be expected to contribute – especially those who do not have children? (I have certainly heard more than one gay man questioning why he should be forced to subsidize the expensive lifestyle choice of “breeders.”)

The comparison between houses and children is not meant to be a joke. Indeed, to pose the question in those terms is already to suggest the answer. People can’t afford houses either, the only reason they are able to buy them is that they have access to credit. And the reason they have access to credit is that they are able to pledge the house as collateral. With children, on the other hand, people have no access to credit – because they are unable to pledge the child as collateral (or the child’s future labour). This is actually one of many examples of how the “prohibition of slavery contracts” leads to significant imcompleteness in credit markets. As a result, many parents find themselves in a serious financial pinch when they first have children, since they generally cannot borrow against their own future earnings in order to cover the expenses they are experiencing. (In my own case, with two young children my daycare payments were higher than my mortgage payment – and not many people have a household budget that can easily absorb a doubling of the monthly mortgage payment.)

So one can definitely find an argument in all of this for offering some financial assistance to parents with young children, then getting it back in the form of higher taxes later on. One might seek to justify the tax deductions for childcare expenses in this way. I think however it would be difficult to explain why large-scale public involvement in the childcare sector is required. In other words, if the argument is going to appeal to the financial “pinch” that parents find themselves in – with reference perhaps to horizontal redistribution and incompleteness in credit markets – then there are a lot of other programs that might deliver the sort of relief required in a far more efficient manner (e.g. subsidized loans). I’m not going to get into it, however, since I don’t think that this argument is doing much of the work in motivating the NDP position, or the views of those who support it.

So then let me get to the major argument. The purpose of the plan is to promote gender equality, by encouraging greater female labour-force participation. And there seems to be little doubt that the plan would encourage such participation. Apart from the data showing very high rates of such participation in countries such as Sweden, that have a highly subsidized childcare system, there is the obvious economic calculus involved when women (not always, but largely women) are deciding whether to return to work or stay home with their child. Basically, the market income on offer has to be enough to cover the cost of childcare, as well as the cost of transportation to and from work, and offer enough of a surplus to offset the quality of life reduction most people experience when going to work rather than being at home with their child. At market prices for childcare, there are a lot of people for whom the math just doesn’t work out in favour of working. Subsidizing childcare obviously would change that.

I guess the big question about this argument is whether encouraging female labour force participation really does promote gender equality, or whether it reflects a perfectionist commitment on the part of some people to a particular conception of how a “modern” woman should conduct her life. There is no question that women who drop out of the labour force for a long time disadvantage themselves in several ways, at least as far as future income is concerned (difficulties that are compounded if they subsequently divorce). And yet there are also great benefits to be had from not working and spending time with one’s child. (As the primary caregiver in my own family, I spend a lot of time – ridiculous amounts of time – as the only man in a room full of moms who aren’t working. They all seem pretty happy with the arrangement. I certainly haven’t heard a lot of “I really want to get back to work, but I just can’t afford to.” Of course, I don’t exactly interact with a representative sample of the population.)

Which brings me back to the conservative argument for the Child Benefit payment. The Conservatives, it will be recalled, upon election cancelled the plans that Paul Martin had been making for a national childcare system, on the grounds that it arbitrarily privileged one particular life choice, and was therefore discriminatory against women (parents, but most of the time women) who choose to stay home with their children. (It also discriminates against people with nannies, although apparently no one cares about this, due to the perception that people with nannies are “rich,” and therefore do not need subsidies. This is not entirely true – the rule of thumb in Toronto is that once you have two young children a nanny is about the same price as daycare, and with three a nanny is cheaper. Furthermore, people with twins often have nannies, both because of the scale economies and the simple mechanics of looking after two babies simultaneously. Finally, there are entire occupational classes – lawyers and doctors, for instance – where daycare centres are simply not open long enough, and so every woman who wants to work full-time has to have either a nanny or stay-at-home husband – and it’s a lot easier to find a nanny.)

This was the rationale for just sending cheques to people, to spend as they please. Money, at least, is neutral between the three different modalities of childcare, and so does not discriminate against stay-at-home parents the way that subsidized daycare does. Of course, it also put the Conservatives in the position of doing the thing that conservatives hate most, which is using the tax system to take money away from some people and give to someone else, the perfect example of an “unproductive transfer” (images of Okun’s leaky bucket spring to mind). My sense is that what they would really like to have been doing is subsidizing stay-at-home moms (based on hazy recollections of their own idealized childhoods), but that they lacked the courage to just come right out and do so, and so they were willing to accept the inefficience of the Child Benefit payments as a way of at least staving off the demand to subsidize daycare.

Since then, however, they have come up with a new, somewhat indirect way of encouraging women to stay at home, which is income-splitting. This has not been introduced yet, but with Flaherty gone it seems likely to be in the works. Allowing a working husband to split his income with his stay-at-home wife creates a huge deterrent to female labour-force participation, because it means that the woman’s income, if she returns to work after having a child, is taxed at a much higher marginal rate than it is right now – because the earned income crowds out the income being transferred from her spouse. So the women is now facing not just the cost of childcare, the cost of transportation, the loss of quality of life, but also a much larger tax bill.

The Conservatives have a lot of rhetoric about “fairness” surrounding this plan. I don’t think any of these arguments, hold water, but that’s a bit too complicated to get into now. Underneath it all, I think it’s fairly easy to see that the plan reflects a commitment to a set of perfectionist values. Namely, they think that its better for mothers to stay at home with their children, and they want to use the various policy levers at their disposal to encourage women to make what they think of as the “correct” choice. But from this perspective, I think it’s also not that difficult to see the NDP plan for subsidized daycare as very much the mirror image, but simply reflecting a different set of values. They think that the “correct” choice for women is to be working, and they want to use the various policy levers at their disposal to encourage women to make this choice.

So in the end, I’m inclined to see the NDP proposal as very much the mirror image of the Conservative plan, both representing expensive attempts to push women in one direction or another when it comes to their childrearing choices. Although in this context, it is perhaps worth pointing out that the NDP plan, at merely $1.8 billion, is an awful lot cheaper than the Conservative plan, which will cost between $4 to $5 billion annually. But that having been said, and being an inveterate centrist, I am left longing for a solution that exhibits greater neutrality toward the decisions that women themselves make about how to organize their lives. I do think there is an argument for state involvement in the childcare sector (again, too long to get into here, having to do with the fact that childcare is best delivered through non-profits, and that as a result markets tend to undersupply it), but I think that this argument justifies a lot less than massive subsidization.

Just to round things out, I do want to mention that, in their attempts to avoid coming across as simply pushing women to work, supporters of the NDP policy have come up with a variety of bogus efficiency arguments, purporting to show that subsidized daycare produces more general “benefits” for society. Typically these benefits are either illusory or overstated:

1. Appeals to early childhood development. This is a really thin reed to hang an argument on. The simple fact is, daycares are not that great, and affordable daycares even less so. Anyone who has ever spent more than 15 minutes in a daycare can sense that. When it comes to child development, nannies perform surprisingly well. Why not have state-subsidized nannies? Or mother’s helpers to come by the home 2-3 hours per day? Or whatever… point is, if you started with an open mind and a commitment to enhancing child development, it is not obvious that you would end up wanting to pour billions of dollars into daycares.

2. It will pay for itself through increased tax revenue. This argument is like a left-wing version of the Laffer curve. The idea is that with subsidized daycare, more women will work, leading to more tax revenue for the government, paying for the program. But of course, if these women are able to pay so much more in taxes that it can fund the system, then obviously they could easily have afforded to pay for the childcare privately out of their wages, so the subsidy would provide no incentive to labour-force participation. Duh.

3. Increases GDP by promoting labour-force participation. True, but it’s important to remember that hurricanes and floods also increase GDP, by forcing people to rebuild their houses. And labour is a curse, not a blessing. GDP is in some ways a strange measure, so when a woman goes to work, you can’t just add her salary to the GDP and call it a gain. When I repair my own fence, while my neighbour looks after her children, all this work contributes nothing to the GDP. But if I pay my neighbour to repair my fence, while she pays me to babysit her kids (and for some strange reason we both report the income) then this does contribute to the GDP. But that contribution overstates the real contribution to national wealth. So at very least, when a woman returns to the workforce, the increment in national wealth is the difference between what she is earning and the implicit wage she was earning staying at home (e.g. the foregone childcare expense).

In my view, there are only two good efficiency arguments. The first is not very glamourous, it refers simply to the economies of scale that can be achieved through collective childrearing. Instead of having 5 people looking after one child each, you can have one person looking after 5 children, leaving 4 free to do something else. The second is slightly more subtle, and it refers to the “mommy track” that gets created in many occupations in countries where women are less likely to return to work after childbirth. This in turn discourages firms from making human capital investments in women (e.g. training), thereby reducing female labour productivity across the board (even among those who never have children, but are merely suspected of harbouring a desire to do so). But just to be clear, these are arguments for wanting to increase female labour-force participation, they are not necessarily arguments for state involvement in the child care sector.


How do we feel about a national daycare program? — 5 Comments

  1. Whew! Lots to digest here – and with no slight intended,I can’t help feeling that the philosopher has travelled all the way around the world in order to go from the front door of his house to the back door.

    At the risk of relying too much on “simple heuristics”, this non-breeder really doesn’t mind paying some more taxes to make it easier for his breeder friends and neighbours to negotiate their child-rearing years. I concede that there is some selfishness in this. I might have been able to spend more time with some of my friends over the past 20 years or so if it had been easier for them to emerge from the Babyzone a bit more often.

    Persons unknown helped to pay for my education and medical care when I was a kid, and they did so to an extent that was much greater than for my parents or grandparents when they were youngsters. I’m grateful to live in such a generous and civilized society. Call me irrational.

  2. Bah–reposting comments. My inequality signs were interpreted as HTML markups….feel free to delete first version.
    Hi Joseph,

    Thanks for a thoughtful post. A few points:

    1) I agree with you that externalities are hard to find with this program. I wrote on this in a bit more detail here:

    2) There is good evidence that extra quality is not priced in the childcare market; people don’t seem willing to pay for quality. (E.g. David Blau’s work: Interesting to think about why–is it that parents correctly can’t assess things that will be good for their kids in the long run? Or is it that parents are correct that the things that experts think make for good quality are actually good for their kids.

    3) on the benefits of childcare for kids, there is pretty good evidence for kids from distressed households that a quality ECE environment can improve their outcomes. (E.g. the Perry Preschool evaluation from the 1960s that took kids with 70 .LE. IQ .LE. 85 and gave them an intense quality immersion.)

    4)On ‘encouraging’ labour force participation. I think you’re right to question whether we really ought to be encouraging it. But childcare is a cost of doing work. We should be taxed on the net return to work, after expenses. This is a basic tax principle to ensure *neutrality* between home work and out-of-home work. Because the current Child Care Expense Deduction is so low (at $7K), the tax system currently favours home work for people facing childcare costs >$7K per year. So, that’s why my public commentary has focused on restoring the CCED as a first step–to *restore* neutrality.

    5) On the ‘pays for itself’ tax argument. You’re right to compare it to the Laffer Curve. In my own work, I estimated that about 40% of the cost of the Quebec subsidy comes back as income tax. I think that’s not a bad estimate. To get to the ‘pays for itself’ level of tax boost, Pierre Fortin and coauthors map the labour supply increase into GDP, and then count sales tax and corporate tax revenue. I have some issues with some of the mechanics of how they do that, but the mechanism is not crazy–in the end I’m not as sure as they are about the ‘pays for itself’ argument.

    6) on the ‘Laffer curve’ aspect, I commented on twitter the other day that the NDP’s embrace of this argument is a pretty large strategic weakness in policy wonk arguments. They’ve now embraced what we call ‘dynamic scoring’, allowing feedback effects of tax policy to enter the calculation. I have a strong suspicion that if I was allowed to use dynamic scoring for a corporate tax cut, I could make that CIT cut look very attractive. To heap on the irony, the NDP plans to pay for the childcare plan with a CIT increase, which they price by assuming absolutely no feedback effects on the economy. I think dynamic scoring should be used for both childcare and CIT. I also think if you use it for only one of the two you are exposing your bias.

    7) the gender equity argument is real. In some of my own work and in work by Phil Merrigan and his UQaM team, we’ve found that the boost to mother labour supply is more or less *permanent*. That is, after the kid reaches school age, mother exposed to the Quebec program continue to have elevated labour supply relative to other mothers. That really shifts the nature of work/gender in the household for the long-run. As you’ve noted, the gender aspects of this program should not be understated.

    Best wishes.

  3. The credit argument is a good one. Why not loans for daycare? I have to say most people look at me like an idiot when. I suggest this. Something about the packaging isn’t good, people don;t like the idea of a direct loan for putting their kid in daycare (I suspect many do so indirectly by allowing credit card balances to rise while their kids are young, but it’s not seen the same way). For some reason, people react better to the idea of “extended payments for daycare” (put the kid in daycare for 3 years, pay over 6). I;m working from a very small “n” here, but I think there’s something in this.

  4. Eugénie Bouchard’s father actually tried to sell her future income but the CRA wouldn’t let him

  5. Here’s a proposal I haven’t heard discussed: what about simply expanding and extending maternity leaves by another 6-12 months? There’s an enormous difference in costs between daycare for infants (0-18 months) and for pre-schoolers (30 months+). At my kids’ campus daycare, for example, infants are around $1450/month, and pre-schoolers just over $800/month. The roughly $600/month difference is pretty big, and doesn’t even take into account the cross-subsidization built into the fees: the preschoolers subsidize the infants, so the actual difference in cost to deliver the care is even larger. Since maternity leaves only last 12 months (less, if the other parent also takes some leave), mothers returning to work are facing very high upfront costs: at least 6 months of over $1400/month childcare. That is, understandably, a price lots of families choose not to pay. But they may choose to pay the lower costs that come along with older children– the problem is the min 6 months before costs start falling. The proposal wouldn’t be obviously regressive (I could be wrong about that, though), and it would go some way to accomplishing the gender-equality aims of the national daycare program. It might also be a lot cheaper, since it subsidizes another year of childcare, rather than 3-4.