Kevin Milligan is a professor of economics at the University of British Columbia. He is the author of a number of papers pertaining to maternal benefits, employment and child care, including an assessment of Quebec’s 7-dollar-a-day daycare system. Universal childcare, maternal labor supply, and family wellbeing, co-authored with Michael Baker and Jonathan Gruber, concluded that while Quebec’s statefunded care did increase the number of working women, it also resulted in poorer behaviour outcomes for kids and more illness. To read about this study in the New York Times, you can click here.
More recently, as claims are being made in both Ontario and British Columbia that all-day kindergarten will bring untold economic benefits, Professor Milligan has been a voice of reason. His concerns stem from an examination of the research. The Perry Preschool Project, for example, is a study oft-cited by advocates of all-day kindergarten/universal daycare that shows benefits of interventions for at-risk kids.
But it was a small-scale targeted intervention that worked with parents too. The question remains: Why should it be extrapolated to show benefits for an entire province? Professor Milligan outlined his concerns in a post at a Canadian economics blog; we had the opportunity to ask Professor Milligan more about his work over the phone.
(Please note that because it is the stated intention of those who advocate for all-day kindergarten to also push for universal government-funded daycare, those two terms are used side by side or interchangeably throughout this interview.)
Institute of Marriage and Family Canada: In many places the case for all-day kindergarten/universal daycare is made on the premise that we are going to see a return on investment down the road. What are the main problems that you see with that?
Kevin Milligan: I’ve looked at the Pascal report [a blueprint for child care in Ontario submitted to Premier McGuinty in 2009, ed.] and many of the other reports that have been cited in defence of this big investment return hypothesis and the main problem I see is that the evidence that is drawn comes in large part from studies on at-risk populations. And these studies have, in some cases, shown big returns to such investment. Not to investments like full-day kindergarten but to some very intensive investments.
The criticism I have is that it’s not obvious that what we learn from those studies can be applied to the case of full-day kindergarten. You could make that leap, and say, yes, the return from these intensive interventions on at-risk kids would be just the same for all families. However, that is a very strong assumption and I don’t see that assumption being discussed very appropriately or openly in the Pascal report or other places.
IMFC: I wonder if you see evidence, perhaps from other jurisdictions, that this type of all-day kindergarten program/universal daycare might actually have a negative return on investment or no return on investment? Alternatively, do you have examples where there is a positive return on investment?
KM: One thing I should make clear to be fair to the Pascal report and to others is, we don’t have the exact experimental counterpart to the Ontario full-day kindergarten, or the B.C. or the P.E.I full-day kindergarten experiment. It would be wonderful to rerun history a hundred times and take the average and see what happens in Ontario over those hundred times. We can’t do that, so we are always going to have to make some assumptions and leaps to apply what we know from other situations to Ontario. The question is, how credible is that leap? My criticism before, about the benefits of intensive interventions on at-risk kids being applied to an entire province, is one example of a leap that is, perhaps, a fairly big one.
What you are essentially asking is “Can you find something that is a better comparator to what we are seeing in Ontario?” And in my view, one of the best things you can look at is what happened in Quebec.
Now there, we are talking about a full-day program for age zero to four-year-old kids. They also did a fullday kindergarten at the same time, but the evidence I am familiar with is mostly on the zero to four-yearold stuff. What is interesting and makes this comparison relevant, is the evidence on the zero to fouryear- olds from these older interventions on at-risk kids looks very positive. But in Quebec, and this is in my work with Michael Baker and Jonathan Gruber, we found evidence that there were negative behavioural consequences on many of these kids. By behavioural consequences I’m talking about things like their behavioural scores on negative behaviours [like aggression, ed.]. What was interesting about this Quebec program is there was already heavily subsidized child care for lower income families, for single moms who have low incomes trying to get back in the workforce. They already had heavily subsidized child care available to them.
What was different about the Quebec program was that it entitled a broad swath of middle income and higher income families to heavily subsidized child care. There were two things of note about the Quebec program versus what we saw elsewhere. One thing is, it was talking about a bunch of middle class families rather than at-risk families and second it’s a real, full scale province-wide program rather than a little experimental program. For those reasons we think that looking at the Quebec program can be informative to something like the Ontario full-day kindergarten.
It’s been in place since 1997 now and I don’t see evidence of a huge economic return. Thirteen years in the program do we see Quebec leaping ahead in terms of economic measurements when compared with other provinces? I’m not sure that we do.
IMFC: Any other long term concerns about all-day kindergarten/universal daycare you might have?
KM: The other long term concern is the funding for such a program can be expensive. Education is expensive. It’s expensive to have grade one, grade two… these things are all expensive. The question is what you get for what you are paying. You can see in Quebec that the expense of the daycare program went up. Some strikes were involved that led to increases in pay of the workers and other things. Maybe that was the right thing to do, but it certainly increased the cost relative to what they thought it would be. So the cost of this kind of universal program can be very high. Education is expensive at all levels and the question should be whether the return here is worth the cost.
IMFC: That leads nicely into another question I had for you, which is what are the opportunity costs of going forward with this type of program? (In other words, what else could be done with the money that will go toward all-day kindergarten?)
KM: The question is, if that money were kept within the education envelope, where else you might spend it? It seems to me the evidence is strong about returns to at-risk families, children in at-risk families. That’s one place you could think about putting it within the education envelope. If you didn’t keep it in the education envelope, the cost is the cost of raising money through taxes or through debt. The broad criteria that I would suggest would be to ask what’s the bang for the buck?
IMFC: Are there any other approaches, policy imperatives on this issue of daycare/all day kindergarten that you think people in government should be considering that they have not considered, be it vouchers, family taxation or something else entirely?
KM: Those are certainly some of the options that are out there. Economists typically love the idea of vouchers for things; they have many attractive features. For this kind of childcare program you can imagine a lot of advantages to a properly structured voucher program. But beyond that, I don’t have any deep thoughts on alternative programs.
IMFC: You have written about how the Nobel Prize winning economist James Heckman has been misrepresented by advocates for universal daycare. Why do you suppose this sort of misrepresentation is happening with such regularity when it seems quite clear that Dr. Heckman didn’t intend his work to be used in that way?
KM: As I mentioned in the blog post, it is really quite jaw-dropping to me how his work has been grossly misrepresented by the advocates of full-day kindergarten. I think this comes from the fact that he is a Nobel Prize winner and quite deservedly so. He is an absolute leader in how we view econometric evaluations. His work has fundamentally changed how we have done many things and he is a person of great authority. He’s not infallible, but he’s a person of great authority. And for that reason, they want to ride on those coattails.
But what they have done is to grossly misrepresent what he has written. He has written very clearly that the evidence is quite strong on at-risk interventions. And he is quite passionate in defending that and saying this is the closest thing to a free lunch you are going to find in economics; it’s both efficient and equitable and it has a very positive return. But he very, very, very clearly states that he is not in favour of universal programs. For the reasons similar to what I’ve laid out: It’s not obvious that you can make comparisons from at-risk families to non-at-risk families. And to use his evidence as intellectual cover for the defence of a universal program is, I think, a misuse of his evidence.
I can’t put myself in the place of the advocates for universal programs, but my suspicion might be, and it’s natural, if there’s a Nobel laureate whose research you can cite you want to cite that, but it’s really important to cite that properly.
IMFC: Dr. Heckman definitely has brought attention to the idea that the zero to three age range is really important for child development.
KM: Yes, what he says is extremely important; the environment in which these zero to three-year-olds grow up in matters. That’s really exciting and interesting. He has really strong evidence and he’s linked it to biology and to other fields on how children develop. But that environment does not necessarily need to be government provided universal care. There’s other ways you can provide an enriched environment and many families do a very good job of that.
IMFC: Is there any other comment on this issue that you would like to make?
KM: One thing I think is important to stress is a rebuttal that I’ve seen to some of my comments. Premier McGuinty was asked: Why are you using these at-risk families and applying it to everyone. And his response, and this comes from Pascal as well, was that everyone is at risk! Forty per cent of kids are at risk in some way, he says, they come from high income families, low income families and middle income familes and so that’s why we can apply the findings to the universal program.
Premier McGuinty is welcome to his own definition of at risk, if he likes, if defining us all as at risk is convenient, but not if you are going to use the Perry Preschool Project or these other interventions as proof. There, the definition of at risk was quite specific. To be in the Perry Preschool Project, which was one of these evaluations that is often cited, the child had to have an IQ in the range of 70 to 85. Now IQ is not the only measure of wellbeing, far from it, but the point is that kids who have a 70 to 85 IQ are really struggling. So what I learned from the interventions on those kids, it’s hard to understand how I might apply that to others. That’s just one example. The exact definition of at risk used in those studies is very different from the one that Pascal and McGuinty use.