New Support for the Economic Benefits of Universal Child Care

I met Sebastien Montpetit at the Canadian Economics Association meetings in Winnipeg last year.  He is a Canadian and Quebecer who has been studying for his PhD in economics at the University of Toulouse.  And he, with co-authors, has come up with a really fascinating analysis of the impacts of Quebec’s universal child care program ushered in the late 1990s and the early 2000s. 

The paper is complex, has multiple parts, and the latest version of it is available here.  It has been selected as one of three finalists for the Canadian Labour Economics Forum prize at the upcoming Canadian Economics Association meetings in Toronto.  I’ll give you the main take-home points right away, and then delve into where the results come from.

Sebastien’s main conclusions?

  • The importance of the supply of child care services has been underrated.  Greater supply of child care – availability – is as important as improvements in affordability.  In Quebec, the regions that had the largest increases in child care supply had the biggest impacts on mother’s employment and increased child care use.  Lowering fees without increasing coverage has modest effects on the benefits to families.  The bottom line: increasing local child care supply is key to the effectiveness of child care reforms.

  • The economic benefits from improved maternal labour supply in Quebec have been well studied and Sebastien confirms them.  But, there are very substantial non-monetary benefits for mothers too.  Think of this as work-family balance, things like the reduced search time for child care, the shorter distances that have to travelled each day when child care is much more available and affordable. 

  • When all the benefits are summed, benefits total more than 3.5 dollars of benefit per dollar of net government spending – more than twice the benefit that comes from looking only at increased mothers’ earnings.
  • Earnings gains for mothers impacted by Quebec’s child care reforms are concentrated in the fifth through the eighth decile of income. In other words, many of the fiscal benefits to governments of a universal child care reform come from mothers who can earn moderate to reasonably high incomes.  These are mothers who will not be reached by a targeted approach to child care spending.  A universal approach may therefore be more fiscally responsible than targeted child care initiatives.
  • Michael Baker, Kevin Milligan and Johnathan Gruber became renowned for their paper concluding that there were a range of negative effects on children who lived in Quebec during the early years of Quebec’s child care reforms (and may have participated in child care).  Sebastien looks at data on those children many years later and assesses whether their educational development was negatively impacted.  He finds no evidence of this; educational attainment of students in Quebec and the rest of Canada is very much the same.
  • Michael Baker, Kevin Milligan and Johnathan Gruber gained some additional notoriety for a follow-on paper that found increased juvenile criminality amongst Quebec children who were exposed to Quebec’s child care reforms.  Sebastien Montpetit looks at the evidence on juvenile crimes and finds that most of the increased juvenile crime that may have occurred was very minor and that the societal cost is relatively small.

The main data source for all of his analyses is the National Longitudinal Study on Children and Youth.  He also uses data from the Canadian Censuses of 2016 and 2021. 

There are four types of analysis that compose this complex paper.  First, with new data on regional child care coverage rates, Sebastien uses a difference-in-differences approach to compare mothers in Quebec to those in the rest of Canada.  He finds that in regions where child care supply increased the most, employment and child care use increased much more when other factors are controlled.

In particular, in regions where child care supply expanded more, the child care reforms boosted mothers’ labour force participation by 40% more than in other regions

Further, Sebastien finds that mothers with low levels of education also respond more in these regions with high levels of expansion.

Results suggest that for high educated mothers with a post-secondary qualification, the main incentive to take up employment was the fee reduction.  For mothers without a post-secondary qualification, access to a space was key. 

Sebastien uses a non-linear difference-in-differences model to estimate earnings gains across mothers’ income distribution.  Mothers’ earnings gains from the child care reforms are found to amount to $1.42 per $1.00 of net government spending.

Baker, Gruber and Milligan found that eligible children in two-parent families experienced worse developmental outcomes and lower consistency in parenting.  Other researchers found substantial heterogeneity in these results.  Haeck et al (2015,2018, 2022) found that most negative impacts on children and parental behaviour fade away over time.

In order to look at children’s educational attainment later in life, Sebastien employs a triple-difference model which compares education levels of same age individuals born before or after the reforms in Quebec to similar individuals in the rest of Canada.

The paper concludes: “We find no evidence of negative effects on educational attainment of eligible children in the long-run. This pattern is true for each educational level, namely for university, high school, and college completion….

 As a result: “…the negative impacts on child behavior documented by Baker et al. (2008, 2019) do not translate into depressed economic outcomes later in life.” (p. 2)  “…this evidence thus suggests the absence of negative fiscal impacts stemming from eligible children’s economic outcomes in the long run.” (pp. 2-3).[1]

Triple-difference estimator compares same-age individuals who vary in eligibility status based on the census year and their province of birth.   He finds no evidence of negative effects on educational attainment of eligible children in the long run.  This pattern is true for every educational level. 

Sebastien Montpetit takes Baker and colleagues’ estimates of increases in youth criminal activity (2019) and estimates what the victimization costs and productivity losses would be.  Using recent estimates of the costs of crime, he finds that these social costs are small.

Difference-in-differences estimates seek to use good control groups to help judge the effectiveness of some policy change.  So, for instance, children 0-4 years of age in the rest of Canada where there was no major child care reform, might be considered to be a good control group to compare to what happened with children 0-4 or the mothers of those children in Quebec.  Why is it called difference-in-differences?  Because this statistical technique does not compare the level of a variable (like mothers’ labour force participation) in Quebec to the same level in Canada.  Instead, it compares the change in mothers’ labour force participation (called a difference) in Quebec to the change over a few years (another difference) in the mothers’ labour force participation in the rest of Canada.  This analysis is done in a regression framework including other variables, so that we can see the impact of those variables on the policy result.

Montpetit then estimates a structural model of maternal labour supply and child care choice in order to make inferences about the size of the non-monetary benefits that mothers receive from Quebec’s universal child care system.  The non-monetary benefits are found to be substantial.  Using the model to do additional simulations, Sebastien concludes that these non-monetary benefits are particularly closely related to the availability of child care services in the local area.  He concludes that universal child care policies for children 0-4 can generate substantial social returns.  And he concludes that increased availability of child care is particularly important to these returns.

Sebastien notes that the quality of Quebec child care in this period was very uneven with CPEs having higher quality and other child care centres having lower quality.   Sebastien is not able to include quality measures in his analyses. 

Altogether a very interesting, carefully crafted and timely paper.  Congratulations Sebastien and co-authors!

[1] Montpetit, S., Beauregard, P., & Carrer, L. (2024). A Welfare Analysis of Universal Childcare: Lessons From a Canadian Reform

Supply-Side or Demand-Side – A Contribution to the Australian Discussion

John Cherry, from Goodstart Early Learning, has written an evaluation of child care in Quebec and New South Wales.  Apparently his purpose is to determine whether supply-side funded systems (like Quebec’s) are better or worse than demand-side funded systems (like in Australia). 

To summarize briefly, John finds that Quebec does better on workforce participation and affordability, NSW does better on child care accessibility and quality.  So, John concludes that Australia’s system is pretty good.  His conclusion appears to be that Australia shouldn’t flirt with Quebec’s fixed-fee, supply-side-funded system. 

It’s a problematic paper for several reasons.  First, some of the details about Quebec are wrong.  Second and more fundamentally, only part of Quebec’s child care system is supply-side funded and charges parents a fixed fee of approximately $10 a day.  The other part (about 20% of the total) is demand-side funded like in Australia.  In the demand-side-funded part, child care providers can set whatever parent fees the market will bear and some of this later gets reimbursed to parents.  So, some of John’s comparisons, particularly on affordability and quality, are actually comparing a mixed system (Quebec) to a demand-side-funded system (New South Wales).  These comparisons don’t tell us much about how a supply-side funded system would perform in Australia.  Third, John does not explain how a demand-side funded child care system can deliver what we want from a universal child care system – dependably low fees, financial accountability for public funds, and planned expansion of capacity according to need.  Let me explain.

Much of John’s paper is captured in Table 1 – Summary of ECEC Indicators.  There’s a column for Quebec and one for New South Wales, comparing results on different indicators of ECEC health.  I reprint it below.

Workforce Participation
John agrees that Quebec does a better job than New South Wales in workforce participation.  Absolutely true.  85% labour force participation for Quebec mothers with young children vs 71% in Australia.  Add on top of that the fact that most Quebec mothers work full-time vs. Australian mothers mostly part-time and it does appear that a fixed low parent fee really does have a very substantial impact on mothers’ employment. 

John then presents comparisons of affordability, but his numbers are too generous to New South Wales and not generous enough to Quebec.  The differences in parent fees between supply-side funding and demand-side funding are much bigger than he admits.  On NSW, John calculates that for a family with average income, the parent fee for a first child is $29.50 per day and for a second child it is $10.05 per day.  In fact, the Productivity Commission draft report says that the average per child out-of-pocket parent fee across Australia (and therefore likely in NSW) is just shy of $45 per day. That includes the extra charges for centres open more than 10 hours per day, where parents have to pay the full fee for these extra hours even though they don’t use them.

And the Quebec numbers on parent fees are too high.  In the supply-side funded centres and family homes, the daily fee for every child in 2024 is CA $9.10 (or about AU $10).  The figure John quotes for Quebec of  CA $17.20 per day includes the children who pay $9.10 but it also includes the high parent fees paid by demand-side funded parents before the tax credit reimburses them.  In a fair comparison, Quebec’s child care is cheaper than in NSW by a considerable amount, not just by a little bit.  That helps us understand why mothers’ employment has been so responsive in Quebec.

Then there is accessibility.  According to John, NSW scores high on accessibility of child care.  But, he chooses a strange way of measuring it.  He chooses the growth in the number of centre-based child care spaces in the last 5 years.  NSW has added more child care spaces so therefore he concludes that accessibility is better in NSW.  

A much better measure of accessibility would have been the coverage rate – what percent of the child population could be accommodated in approved services (licenced services in Canada).  John provides these numbers on page 6 of his paper, but not in Table1 and not in his conclusions about accessibility.  In fact, as he records, about 75% of  children 0-5 in Quebec are in early childhood services.  This compares to about 60% in New South Wales.  John makes a big deal about services growing in New South Wales and not growing in Quebec.  Of course, that’s what you would expect if accessibility was already better in Quebec; it wouldn’t need to grow its services as fast.  The current rate of growth of services is not a good measure of current accessibility.

And if you compare the number of days of child care attended in Quebec and NSW, the accessibility in Quebec is even stronger.  Over 90% of the children in Quebec who attend ECEC do so on a  full-time basis, compared to about 30% in Australia (with another 25% in Australia attending 4 days a week).

Finally, we get to a part of the comparison between Quebec and NSW on which John and I agree.  The quality of child care in Quebec is lower than it should be, and probably is lower than it is in NSW.  The most obvious indicators of this are the child-staff ratios.  5 children to 1 staff member for very young infants in Quebec vs. 4 to 1 in NSW.  Personally, I think both of these ratios are too high for the very young, but I agree that a 5 to 1 ratio is shocking.  As is a ratio of 8 to 1 in Quebec after children turn 19 months of age. 

Quebec is an outlier here in Canada too.  In Ontario, the required ratios are 10 children to 3 staff members for children 0-17 months, 5:1 for children 18 months to 35 months, and 8:1 for children 3 years to 6 years (except for before-and-after school care for kindergarten children).  NSW’s ratios are comparable to Ontario’s. 

Similarly, the wage rates paid to educators in Quebec are worse than in New South Wales.  John is right on this.

John overstates the differences in percent of educators required to be qualified.  He says it is 50% in Quebec and 100% in New South Wales.  The regulated percent in Quebec is really 66.6% or 2/3rds.  It was temporarily lower due to staff shortages during the pandemic. And the requirement in NSW is for 100% of front-line staff to be certified.  But this is a bit misleading because only 50% of the front-line staff in NSW must have an ECE Diploma or above.  The other 50% can have a Certificate III which is a qualification well below what is needed to provide good quality care for children on one’s own.

However, the inadequate quality of Quebec’s child care system is not really evidence that supply-side funding does not work.  Instead it is evidence that Quebec services have not been adequately funded.  The history of Quebec’s system explains this.  Back in the 1990s, Quebec struck out on its own to build a universal child care system, without any funding from Canada’s federal government.  Relying only on its own funding, Quebec ended up cutting corners on quality.  If New South Wales were operating either a demand-side funded or a supply-side funded system with no Commonwealth funding – relying only on state funds – I am sure that quality would suffer too.  But Quebec’s history is not New South Wales’ inevitable destiny.  With strong Commonwealth commitment to spending on universal child care, New South Wales can have both supply-side funding and good quality care.  As you can see in John Cherry’s Table 1, public funding of child care in New South Wales is already 50% higher than in Quebec – AU $5.7 bn vs. AU $3.7 bn annually.

What Conclusions Should We Draw From This Comparison?
I understand John Cherry and Goodstart’s hesitation about a switch to supply-side funding.  It would be a big transformation of funding arrangements and would constrain the power of child care operators to set their own fee levels.  If it was done badly, it could have negative effects. 

However, I think John and Goodstart need to explain how they will build a publicly-accountable universal low-fee high-quality child care system on Australia’s existing demand-side funding base.  In my opinion, they need to answer (at least) three questions.  How would they guarantee that the system will have low child care fees in the future?  How can they build financial accountability for public funds into the existing system?  And, what mechanisms of public planning for location of new services can ensure an equitable and efficient growth of new services in Australia?

Australia has seen parent fees rise consistently as public funding has increased over the years.  The average parent fee per child is now about AU $135.00 per day.  Every time the Commonwealth government pours more money into the system, out-of-pocket child care fees fall temporarily.  After a short while, these out-of-pocket costs gradually rise back to previous levels.  Nothing has worked to keep fees down in the long term.  Supply-side funded systems guarantee low dependable out-of-pocket fees.  Until Australia’s demand-side-funded child care system can provide the same guarantee, it cannot be considered a good basis for a universal system.

In a universal child care system, the vast majority of operator revenues come from governments.  It is unacceptable to continue to have no public accountability for these substantial amounts of public funds.  Currently, child care operators do not have to justify the fees they charge or show that public moneys are spent on legitimate costs of service provision.  Goodstart should explain how this will be remedied in their plans for a universal child care system built on the existing demand-side foundations.

Finally, an equitable universal system of child care services needs to plan where new child care services will be located.  It cannot leave this to the whims of private investors who all want to crowd their new services into higher income areas.  How will this be accomplished within Australia’s demand-side funded system?  These are the tough questions that need to be answered by the champions of a continuation of demand-side funding for Australian child care. 

The Fraser Institute’s Evaluation of the $10 a Day Child Care Reforms

This is not his best work.  Phillip Cross has had a notable career at Statistics Canada.  He’s an expert in macroeconomic trends.  But, one thing that he knows very little about is child care.    Unfortunately, he has written a short paper for the Fraser Institute evaluating the success or failure of the Canada-Wide Early Learning and Child Care reforms so far. 

It’s bad. Almost everything in the paper is either wrong or misleading.

So what does Phillip Cross say?

  • He says that the Canada-Wide early learning and child care program had 3 goals:

(1) providing more jobs in the child care industry,

(2) enabling mothers to join the labour force, and

(3) providing better care for young children. 

His paper will look at the first two.

  • He looks at some evidence and concludes that there has been no change in the employment trends in child care staff.
  • Then, he looks at evidence about women’s labour force participation and concludes that it has hardly changed since 2015.
  • Having concluded that the Canada-Wide child care reforms are a failure, he goes on to take pot shots at Quebec’s child care system concluding that its universal child care system doesn’t really help low-income families, wasn’t really responsible for the boost in its labour force participation, has long waiting lists due to inadequate supply, and isn’t really universal.

Phillip Cross is wrong on all counts, contributing yet more false information to child care discussions in Canada.  There are many problems with the rollout of the Canada-Wide program across the provinces and territories – particularly slow rates of growth in child care capacity.  However, the Fraser Institute paper does not grapple with real issues and propose real solutions.

Phillip Cross, believe it or not, ignores improving the affordability of licensed child care in his list of goals of the Canada-Wide program.  This, of course, is the greatest success of the program so far.  Hundreds of thousands of children and families have benefited from less expensive child care.  Their very high child care costs have been cut by half or more.  These parents are very happy with the marvellous success of the program.

Employment in the Child Care Industry

There has been substantial growth in employment in the day care industry (NAICS Code 6244) since April 2021 when the Canada-Wide program was announced.  By my reckoning, the number of persons employed in Canada outside Quebec has risen by 36.9%, a total of 32,885 additional persons employed.  Phillip Cross hides this growth in two ways.  First, he looks at Canada including Quebec, which is inappropriate.  Quebec has a mature child care system and its employment of child care staff is not growing quickly.  The focus of growth in the Canada-Wide program is on the provinces and territories outside Quebec.

Second, Phillip Cross ignores the collapse of child care employment during the pandemic and assumes that child care employment should have grown as if the pandemic did not happen.  In fact, child care employment in Canada outside Quebec collapsed from over 100,000 at the beginning of 2020 to less than half of that a few months later.  Employment did not climb above 100,000 until March of 2022.  So, the Canada-Wide program has helped the revival of employment in the child care industry and gone well beyond.  We should celebrate this, rather than hiding it.  This evidence can be found in Statistics Canada CANSIM Table 14100201.

Mothers in the Labour Force

Phillip Cross concludes that the Canada-Wide program has also shown no progress in supporting mothers to enter the labour market.  According to him, labour force participation hit its peak in 2015 and even after all this money spent on child care, it has only just about reached the same level.  As he notes, the participation rate was 61.7% in 2015 and now it is just 61.5%.

But, Cross is not looking at the right statistics.  He is looking at the labour force participation of all women 25-54 years of age.  However, most women do not currently have a child 0-5 years of age.  Women without young children would not have their labour force participation affected by the Canada-Wide child care program.

The Fraser Institute report should instead be looking at labour force participation of mothers with children 0-5 who are the target of the program.  Here, participation rates are up by several percentage points from April 2021 to now (from 76.9% to 79.9%) even though expansion of child care has been slower than it should be.  And compared to 2015, which the Fraser Institute cites as the high water mark, the labour force participation of mothers with children 0-5 is over 6 percentage points higher now than it was then.  So this evidence of “failure” is false news and should not be left to become conventional wisdom.  This data can be found in Statistics Canada CANSIM Table 14100397.

Quebec’s Universal Child Care System

Phillip Cross would presumably be very surprised to hear that Quebec’s child care system is very popular with parents and with the Quebec government.  He believes that low-income families have been squeezed out of access to child care.  In fact, there is good evidence that a much higher percentage of low-income families in Quebec have been able to access child care than was true for low-income families in the rest of Canada in the period before the Canada-wide system[1]. The universal system of child care in Quebec encouraged many more low income mothers into the labour force and into using child care.  It is true, and a problem, that on average low-income families are more likely to have their children in the lower-quality for-profit child care services.  The Quebec government is expanding not-for-profit centres as a partial remedy.

Cross claims that Quebec’s child care system is not universal.  His evidence for this seems to be that there are 51,000 families on a waiting list for child care services.  Here his lack of child care knowledge is really showing.  This is a waiting list to get into one part of their child care system – the preferred part with a fixed fee and many better quality services. 

There is no overall shortage of child care spaces in Quebec; in fact there are many empty spaces in the for-profit child care services funded by a tax credit.  But parents don’t prefer these for-profit spaces where there is no guaranteed parent fee.  These services have been shown to be much poorer quality than the not-for-profit spaces in CPEs (early childhood centres).  So, yes, there are 51,000 children on a waiting list to get out of these tax-credit-funded spaces and into the fixed-fee services that they prefer.

Finally, Phillip Cross tries to deny that the universal child care system in Quebec has been responsible for dramatic increases in labour force participation of mothers.  He writes that “proponents attribute the increase in female participation in Quebec to its childcare program” and “Clearly, some determinants of female labour force participation are not understood by researchers, who nevertheless loudly endorse Quebec’s initiative.”  This is a bit strange, because if there is one thing that all economic studies of the Quebec child care program are agreed upon, it is that there was a substantial boost to mothers’ labour force participation and hours of work as a result of universal child care.

A summary of the results of one of the many studies goes like this:  “Lefebvre and Merrigan[2] (2008) use Statistics Canada’s Survey of Labour and Income Dynamics (SLID) annual data from 1993 to 2002. Using the sample of all Canadian mothers with at least one child aged 1 to 5, they find that the policy had substantial effects on a diversity of labour supply indicators (participation, labour earnings, annual weeks and hours worked). In 2002, the effects on participation, earnings, annual hours and weeks worked of the childcare policy are respectively between 8.1 and 12 percentage points, $5,000 to $6,000 (2001 dollars), 231 to 270 annual hours at work, and 5 to 6 annual weeks of work.“   

The Fraser Institute is not noted for the complete accuracy of its studies, but this is a bit ridiculous.  As an evaluation of the success or failure of the Canada-Wide Early Learning and Child Care program, the Fraser Institute study is worse than useless. It is, perhaps deliberately, misleading. 

Instead, we should conclude that:

  • Hundreds of thousands of children and families have benefited from more affordable licensed child care
  • There are now nearly 33,000 more persons working in the day care industry than there were when the program was announced in April 2021 – an increase of nearly 37%.  Many more qualified educators are needed, but this is a good start.
  • Even though the growth in the number of child care spaces has been too slow, there has still been a rise of 3 percentage points in the labour force participation rate of mothers with children 0-5 since April 2021.  Again, only a start, but definitely a start.
  • Quebec does have a universal child care program and many families access child care for less than $10 a day.  It is a very popular program with families.  There is no overall shortage of child care spaces in Quebec, but many families want to get into the fixed-fee part of the child care system, especially the better-quality not-for-profit CPEs.  Many of these families are on a waiting list.  A large number of low-income families have benefited from the universal child care program in Quebec, a much larger percentage than benefited from Canada’s targeted child care assistance.  There is still important work to do to ensure that low-income families also benefit equally from better quality in child care services.

[1] Cleveland, G. (2017) “What is the Role of Early Childhood Education and Care in an Equality Agenda?” pp. 75-98 in Robert J. Brym ed. Income Inequality and the Future of Canadian Society ISBN-13:978-1-77244-044-7 Oakville, ON: Rock’s Mills Press. Proceedings of the inaugural S.D.Clark memorial symposium.  That study found that:” In Quebec, 61.8 percent of children 1-5 years with an employed or studying mother with a high school education or less use licensed child care. Including children with a mother who is not employed, 43.1 percent of Quebec children whose mother has a high school education or less are using licensed child care — about 30 percentage points higher than the comparable figure in the rest of Canada.“

[2] Lefebvre, P., Merrigan, P. (2008). Childcare policy and the labor supply of mothers with young children: a natural experiment from Canada. Journal of Labor Economics 23, 519–548.

Some Thoughts About Australian Child Care Policy

The Labor federal (i.e., Commonwealth) government of Australia has declared its intention to move towards universal child care. There is a lot of interest in the Quebec model. The Commonwealth government asked the Productivity Commission to investigate and to provide a roadmap towards universal early childhood education and care throughout Australia.

The post below is my submission in response to the draft report of the Productivity Commission which you can find here. As you can see, my advice and comments are strongly informed by Canada and Quebec’s experiences.

Response to the Productivity Commission Draft Report

Main Messages

  • The final report of the Productivity Commission should lay out a 10-20 year vision of recommended steps to achieve universal affordable, accessible, high quality child care.  The recommendations in the draft report go only\ part way to universal child care.  The recommendations should include ways in which there can be guaranteed fee levels for parents, much greater financial accountability of operators, and substantial introduction of supply-side operational  funding. 
  • There should be a much stronger gender equity lens by which recommendations are judged and through which recommendations are presented.  This would affect recommendations that imply that 3 days a week is the norm for child care attendance and mothers’ participation in the labour force.
  • The commercialization of child care provision should be an issue of concern.  Child care growth has been very unbalanced; nearly all new centre-based child care for at least 10 years, and probably 20 years, has been commercial.  There are not adequate supports needed for expansion of not-for-profit services.
  • In the draft report, the description and lessons learned from the experience of child care reforms in Quebec is one-sided.

There are some good things about the lengthy and detailed Productivity Commission Draft Report. 

If there is not enough money to do everything right away, it is often sensible to prioritize providing child care services to children in lower-income families.  Moving to 100% subsidy and getting rid of the activity requirement for 3 days a week of child care services will address some important barriers to participation by children in lower-income families while directing over half of the additional assistance to families in the lowest 20% of the income distribution.  Even here, there are potential issues with the proposals[1].

Getting rid of the activity requirement for 3 days a week will also help some middle-income families where parents have irregular work activity and will tend to normalize regular child care attendance for children.

And the Productivity Commission dips its toe in the water of supply-side funding in remote communities where the profit motive clearly does not adequately encourage needed supply.  This is an important start, even if a minor one.

However, as a guide to the pathway to universal child care in Australia,  the Productivity Commission’s draft report is disappointing.

  1. The government asked for a plan to move towards universally accessible, affordable and high-quality child care.  This draft report does not deliver this.  Instead it chooses to primarily fill one hole in the current state of accessibility – access by lower-income families.  Unless the Productivity Commission believes that all other families already had affordable access to child care (which is unlikely since the average out-of-pocket amount that parents pay for centre-based child care is $44.42 a day per child), remedying this one (important) problem will certainly not deliver universal child care. As long as there is no legislative or regulatory limitation on parent fees and no limitation on centres charging full fees for unused hours above 50 in a week, child care in Australia will be unaffordable and inaccessible for some families, perhaps many families, who have middle and higher incomes, as well as families with lower incomes.   As long as there are either financial or supply barriers that prevent access, early childhood education and care is not universal.  Frankly, despite the Productivity Commission’s mandate to study how universal child care can be achieved, there is evidence in the draft report of some bias against universal child care, reflected in the cautious nature of the recommendations and in the one-sided evaluation of Quebec’s system of universal early childhood services.

2. The Productivity Commission’s draft report appears to reflect a view that child care markets work well in Australia, and that strong competitive pressures already compel commercial operators of early childhood services to keep costs low, expand to serve new needs and continually enhance quality.  In other words, the Productivity Commission believes that current funding and regulatory arrangements provide the appropriate incentives and controls to make child care providers serve the public interest.  Apparently, only a few tweaks are necessary to make these services more accessible.

This optimistic view is less true than the Productivity Commission believes; the problems are larger and the need for reform is greater.  First, we know that competition in child care markets is very localised, largely because few parents want to regularly transport their children more than a couple of kilometres to a child care service.  So, each centre only really competes – on price, services and quality – with other centres close by.  Generally, that means that competitive pressures are not that strong. 

Fees have not been kept down by competition; they have been continually rising for many years. The current average daily fee for centre-based child care is $133.96 per child. Over 20% of child care centres charge more than the hourly rate cap (currently $13.73 per hour for centre-based day care for children younger than school age), particularly for-profit centres.  There is little evidence that costs and fees are controlled by strong competitive pressures.

One of the hallmarks of competitive markets is that prices charged are forced down close to actual costs. If the price of one product or service is much higher than its per-unit costs, we would expect profit-seeking producers in a competitive market to offer this product or service at a lower price and take a large number of customers away from existing providers. In centre-based child care, given the required staff-child ratios, the labour costs for infant care must be close to 3 times the labour costs of child care for three- and four-year-old children.  And labour costs are the large majority of total costs.  Yet, competition does not drive centres to charge much lower fees for older children than they do for infants. There is a large variation in per-child costs and there is virtually no difference in fees.  And there are long waiting lists for child care for children less than two years of age, mostly because infant care is less profitable. These facts are a strong signal suggesting that child care markets in Australia are only weakly competitive. 

Figure 4 of the interim ACCC report suggests that average occupancy rates of large providers of centre-based day care are about 75%.  We know that occupancy rates are a key driver of per-unit costs.  In a competitive market, we would expect strong pressures on operators to cut fees in order to increase occupancy, lower per-child costs and maintain quality.  This does not appear to be widespread in child care markets.

In short, the main mechanism that makes the Productivity Commission so complacent – competitive pressure – cannot realistically be assumed to deliver publicly beneficial results on its own.  There is a need for more public management – active market stewardship – and financial accountability.

3. There is no realistic plan to keep child care fees from rising faster than the CPI (which they have been doing for many years)[2].  Draft recommendation 6.2 suggests a new hourly rate cap for Child Care Subsidy based on the “average efficient costs of providing early childhood education and care services”.   Unfortunately, there is no unique average efficient cost.  As mentioned above, just think of infant care with required child-staff ratios of 4 to 1 vs. care for children over 36 months of age with required child-staff ratios of 11 to 1 in many states and territories.  How could there possibly be a unique average efficient cost per unit across these different age groups?  And look at cost variations that are recognized in supply-side-funded jurisdictions.  In Quebec and New Zealand[3], for instance, child care operating payments vary across a number of important factors that drive key cost variations – staff experience levels and qualifications and pay rates, legitimate variations in arms-length occupancy costs, higher per-unit costs in thinner markets, etc.  Unless the Productivity Commission can propose a realistic set of rate caps tailored to different circumstances and a means of regularly updating them and enforcing them, this recommendation may not work.

4. The recommendations in the report would establish 3 days a week as a norm for the number of days a mother should work.  This is negative for gender equity, which is already dramatically impacted by the almost universal assumption that women are primarily or solely responsible for the day-time care of children before school.  The draft Productivity Commission report shows that the average size of the motherhood penalty in Australia – the amount of previous earnings that is lost when mothers bear children – is 55% (!), higher than in many other countries.  The motherhood penalty is explained by lower rates of employment, lower hours per week of employment, and lower hourly pay of mothers. The Productivity Commission is doing a good thing by reducing the impact of the activity test on access to child care.  That will lower barriers to employment for mothers. However, they should recommend its elimination entirely for 5 days a week.  To me, the recommendation as it stands suggests that children only need child care for 3 days a week, and that child care for more than 3 days a week may be negative for children and is done only for the mothers who insist on working too long weekly hours (to whom the activity test is applied).   There is increasingly strong evidence[4] that universal child care in Quebec and elsewhere has reduced motherhood penalties substantially.

5. The Productivity Commission appears to believe that the widespread use of only three days a week of child care is due to maternal preference rather than to the unaffordability of 5 day a week child care.  They show self-reported numbers that allegedly prove that very few mothers would work longer hours each week (and use 5 days of child care) under any circumstances.  In other words, the motherhood penalty in Australia is the result of mothers’ deliberate and free choices.  I doubt it.  In contrast, the ACCC believes that “the price of childcare significantly impacts how much childcare households use.” (p. 22).

 It is true that lower labour force participation and part-time work for mothers are strong norms in Australia, compared to many other countries.  However, there are reasons to believe that if child care was universally affordable and accessible in Australia, those norms would change.  As evidence, look at the very substantial changes in labour force participation of mothers with children 0-4 in Australia over the 12 years from 2009-2021.  In 2009, 48% of mothers with young children stayed outside the labour force.  By 2021, that number had fallen by one-third (16 percentage points!) to 32%.   That would seem to indicate that mothers’ employment decisions may be quite sensitive to changes in policy, rather than fixed by historical norms.  This matters for the motherhood penalty, but it also matters a lot for the funding of child care programs; in Quebec, a large portion of the fiscal costs of child care programs is funded by increased incomes and taxes due to changed employment.

6. There is no plan for requiring financial accountability of providers for the vast sums of government money they receive.  The legal fiction is that parents who receive subsidies for the purchase of child care are effective watchdogs of how the money is spent.  This is so obviously not true that it needs little argument to reject it.  But, there is no requirement for providers to show that they have spent money wisely to achieve publicly desirable purposes.  There are some serious red flags that the Productivity Commission does not really address.  They report that there are many hours of ECEC services that are paid for each day (by parents and the government) but are not used. This sounds like evidence of substantial inefficiency in current funding and attendance arrangements.  The Australian Competition and Consumer Commission (ACCC) report concludes that for-profit child care providers pay more for occupancy costs than not-for-profit providers (and that part of this may be due to the use of facilities for which ownership is not at arms-length)[5].    Further, for-profit services are found by the ACCC to be of worse average quality[6] than that provided by not-for-profit providers.  The Productivity Commission should be making recommendations about compulsory and regular financial accountability. I believe that, in Australia as in Canada, child care is fundamentally a public service (with about 80% of costs paid by the public purse) but one that is delivered by private operators. Detailed and regular reporting on how public moneys are spent should be an obvious requirement.

7. The final report of the Productivity Commission should lay out a 10- to 20-year vision for the establishment of universal child care services in Australia.  The recommended National Partnership Agreement would be a part of this plan.  Wrap-around child care for preschools would be a part of this plan.  The expansion of supply-side funding of services with fees controlled would be part of this plan.  The new independent ECEC Commission would monitor and report on progress towards universal access and make ongoing recommendations to move towards it.  The Productivity Commission hints at a long-term vision but is not explicit.  This allows the Productivity Commission to duck a lot of longer-term questions about affordability, commercialization of the system, financial accountability, and generally the evolution towards serving public interests better.

8. Australia has a long-established demand-side (voucher) funding system for child care.  It allows providers to set their own fees, decide on staff compensation conditional on meeting the award levels set by the Fair Work Commission, choose the children and parents they will serve from those who apply and choose the hours of service to provide.  This is not, in my opinion, the best system going forward; I believe that a system of supply-side-funded services with a guaranteed set fee level (plus fees reduced below the set-fee level or to zero for some families) would be better.  However, changing funding systems is not easy and there is often a lot of push-back from those in the system.  Why not think outside the box? Why not establish an alternative supply-side funding system that would exist in parallel with the existing demand-side funding system with incentives for centres to switch? 

Centres that were funded on the supply-side would have a fixed fee, and enhanced regulatory requirements.  In exchange, they would have guaranteed funding to cover costs above parent fees. Set-fees that are known and predictable are very popular with parents at all income levels and, in Quebec, have encouraged high child care participation by children in lower-income families.  There would be strong elements of financial accountability and reporting by centres, requirements to pay above-award wages, reduced ability to rely on part-time and casual staff and other requirements related to quality of care, but some centres and some parents would prefer this.  There would be obvious transition difficulties, but this kind of recommendation would boldly look towards transforming Australia’s system into a universal and affordable one.

9. The Productivity Commission does not address the increasing commercialization of child care services in Australia.  Virtually all of the expansion of centre-based child care services (not preschools) in the last decade – a 50% increase in the number of spaces available – has been in the for-profit sector.  As the ACCC interim report notes: “the child care sector is widely viewed as a safe and strong investment with guaranteed returns, backed by a government safety net.” The Productivity Commission report does not even raise the question of whether this extremely unbalanced growth pattern is desirable. The growth in services that has occurred is disproportionately located where returns are higher, rather than where need is greater, as shown in Figures 3 and 7 of the draft report.  1% of providers now provide 35% of all centre-based child care services. The Productivity Commission should be making recommendations about means of encouraging growth in not-for-profit and public provision of services.  These recommendations would call for planned development and dedicated loan guarantees or other capital funding targeted at not-for-profit providers.  I believe that Australian children and families are unlikely to prefer a universal child care system with unplanned expansion and complete domination of service provision by commercial incentives and ethics.

10. The Productivity Commission draft report provides a one-sided summary of the experience and effects of Quebec’s universal child care system.  Although it is true that economic researchers found short-run negative effects on some children (effects were found to vary substantially across different child groups[7]), the most recent and comprehensive work on Quebec, using a triple-difference estimator similar to other studies (Montpetit et al., 2024[8]) does not find any long-run negative effects on children’s completed education.  Rather, they find that the long-run education levels of Quebec children who had been eligible for $5 a day child care were no different than their peers in other parts of Canada.  In particular they write: “We find no evidence of negative effects on educational attainment of eligible children in the long-run. This pattern is true for each educational level, namely for university, high school, and college completion….

The results suggest a positive but statistically insignificant impact on completion of a university degree, the most comparable outcome across provinces, and no impact at lower levels.”(p. 21).  Further, Montpetit’s study calculates the social cost of increased “youth criminal activity” identified by Baker, Gruber and Milligan (2019[9]) and finds negligible social costs because the identified transgressions were minor. 

11. The Productivity Commission draft report gives little sense that this fixed-parent-fee child care program is an incredibly popular social program with Quebec parents.  The reader will struggle to understand why the Canadian federal government decided in 2021 to spend $30 billion over 5 years to spread the Quebec child care model of a fixed-fee, supply-side-funded program across the country.  The reader of the draft report will not be told that Quebec’s child care reforms had sufficient impacts on mothers’ employment and economic growth to more than pay for the costs of the program according to the influential opinion of prominent Canadian economists (Fortin, Godbout, St. Cerny, 2013[10]).  Lefebvre and Merrigan (2008[11]) find that Quebec’s policy reforms increased labour force participation of mothers with children 0-4 by 7.6 percentage points from 61.4%  before the policy.  They estimate the labour force elasticity to child care price to be 0.25.  In addition the child care reforms increased the annual hours worked, weeks worked and earnings; these elasticities were 0.26, 0.28, and 0.34, respectively.  With these elasticities, a 10% decrease in the fee would increase annual hours worked by 2.6%, increase weeks worked by 2.8% or increase earnings by 3.4% on average.

Lefebvre, Merrigan and Verstraete (2009[12]) found that the labour force impacts lasted beyond the preschool child care years when mothers no longer had any children 0-5 years of age, and that the positive labour force impacts were particularly strong amongst mothers with lower levels of education. Even if long run labour force effects are ignored, the recent study by Montpetit and colleagues (2024) finds that the overall benefits of universal child care in Quebec are three and a half times the costs.  This includes a careful evaluation of the value of the improvements in the well-being of Quebec mothers from universal child care services.

12. The Quebec model of funding and management of child care services is not a perfect one.  Two factors made its birth particularly difficult.  First, when they initiated the $5 a day program, Quebec only had enough child care supply to provide services to 15% of the child population 0-4 years.  For 20 years, they scrambled to increase supply and have now reached nearly 70%. However, this scramble to increase supply meant relying too heavily on both family child care and for-profit child care with weaker regulation.  These types of care have been the Achilles heel of quality[13] in the Quebec system, a problem that is now being addressed.  Second, this was a program funded exclusively by the provincial government; at that time, the federal government was unwilling to provide any financial support.  The provincial governments in Canada have modest taxing powers, so services were not as generously funded as they should have been.  With the federal government coming to the table in 2021 with billions of dollars of additional funding, child care services in Quebec will now be funded more appropriately.  I have described the problems of the Quebec model of child care here[14], warts and all.   However, these problems are not inherent in a universal program; Australia already has a large child care supply and substantial financial resources available to support good quality programming. It can gain the substantial benefits of Quebec’s universal program without the birth pangs that Quebec has faced.

Commentators have noted that low-income families in Quebec do not have as much access to good quality child care as do middle income families.  That is true and is a problem. As far as I can tell, that is true and is a problem in most countries whether child care systems are universal or not; it is certainly true in Australia[15]. However under Quebec’s universal program it is also true that a much higher percentage of low-income families were able to access licensed child care than was the case with the targeted funding that predominated in the rest of Canada[16].  Children from low-income families also were particularly likely to benefit from their access to early childhood programs[17].

13. The terms of reference of the Productivity Commission enquiry require that it study “the operation and adequacy of the market, including types of care and the roles of for-profit and not-for-profit providers, and the appropriate role for government.” Further, these terms of reference direct that “The Commission should have regard to any findings from the Australian Competition and Consumer Commission’s Price Inquiry into child care prices….”   However, the findings in the ACCC draft report about the child care industry scarcely get any mention, including differences in costs and priorities of for-profit and not-for-profit providers.  The ACCC report provides important insights about costs and performance not available elsewhere.

14. I hope that many of these issues will be addressed directly in the final report of the Productivity Commission.

Gordon Cleveland, Ph.D.,
Associate Professor of Economics Emeritus,
Department of Management,
University of Toronto Scarborough


[1] These policy changes -removing activity requirements for 3 day attendance and 100% subsidy up to $80,000 -should mean many more lower-income families wanting access to child care.  Some operators prefer to serve a more exclusive clientele; this is known as creaming.  Under current rules, centres that charge a fee that is above the maximum hourly-fee limit are likely to effectively exclude most of these children.  Perhaps the Productivity Commission should require that centres be compelled to serve these children at the maximum hourly fee if parents apply to attend.

[2] The cost of child care in Australia is pretty high.  Centre-based child care fees per hour (averaged across ages 0-5) were $11.72 in 2022.  The Productivity Commission reports that the average daily fee is $124 per day.   From 2018 to 2022, gross fees in Australia increased by 20.6% in comparison to the OECD average of 9.5%.  The OECD ranks Australia as 26th out of 32 countries on affordability of child care for a typical couple family with two children.  This is despite the Australian Government contribution to fees being significantly higher than most other OECD countries – 16% in Australia compared to the OECD average of 7%.

[3] See for a discussion of details of child care funding in Quebec and see for a discussion of details of child care funding in New Zealand.

[4] See Connolly, M., Mélanie-Fontaine, M. & Haeck, C. (2023). Child Penalties in Canada.   Canadian Public Policy doi:10.3138/cpp.2023-015.  See also Karademir, S., J.-W. Laliberté, and S. Staubli. (2023). “The Multigenerational Impact of Children and Childcare Policies.” IZA Discussion Papers No. 15894, Institute of Labor Economics (IZA), Bonn, Germany.  As Karademir et al indicate “The disproportionate impact of children on women’s earnings constitutes the primary factor contributing to persistent gender inequality in many countries.”

[5] Land and occupancy costs are about 18% of the total of all costs for large for-profit providers compared to about 10% for large not-for-profit providers. This is not due to what the Aussies call “peppercorn rents” (i.e., below-market rents provided on a goodwill basis).  The average profit margin for large centre based day care providers was about 9% for for-profit providers and about 6% for not-for- profit providers in 2022. 

[6] About 95% of the staff in not-for-profit centres are paid “above-award” compared to 64% in for-profit centres.  Not-for-profit providers are much more likely to hire their staff on a full-time basis, whereas for-profit providers primarily rely on part-time staff.  As the ACCC report suggests: “large not-for-profit centre-based day care providers invest savings from lower land costs into labour costs, to improve the quality of their services and their ability to compete in their relevant markets.”  The ACCC finds that centre-based day care services with a higher proportion of staff paid above award and with lower staff turnover have a higher quality rating under the National Quality Standard. 

[7] Kottelenberg and Lehrer provide evidence of substantial heterogeneity in the impacts of the Quebec child care reforms by the age of the child, the child’s gender and by initial abilities in a series of studies published in 2013, 2014, 2017 and 2018.  Kottelenberg, M. J. and Lehrer, S. F. (2013). New evidence on the impacts of access to and attending universal child-care in Canada. Canadian Public Policy, 39(2):263–286. Kottelenberg, M. J. and Lehrer, S. F. (2014). Do the perils of universal childcare depend on the child’s age? CESifo Economic Studies, 60(2):338–365. Kottelenberg, M. J. and Lehrer, S. F. (2017). Targeted or universal coverage? assessing heterogeneity in the effects of universal child care. Journal of Labor Economics, 35(3):609–653. Kottelenberg, M. J. and Lehrer, S. F. (2018). Does Quebec’s subsidized child care policy give boys and girls an equal start? Canadian Journal of Economics/Revue canadienne d’ ́economique, 51(2):627–659. Kottelenberg and Lehrer (2017) finds that levels and changes in home learning environments by some parents in response to the Quebec reforms were an important explanatory factor of differential effects on children.

[8] Montpetit, S., Beauregard, P., & Carrer, L. (2024). A Welfare Analysis of Universal Childcare: Lessons From a Canadian Reform

[9] Baker M., Gruber J., & Milligan K. (2019). The Long-Run Impacts of a Universal Child Care Program American Economic Journal. Economic Policy, Vol.11 (3), p.1-26; American Economic Association.

[10] Fortin, P., Godbout, L. and St.Cerny, S.. (2013). “Impacts of Quebec’s Universal Low-fee Childcare Program on Female Labour Force Participation, Domestic Income and Government Budgets. University of Toronto. Toronto, ON.  Translated from French  Original reference is Fortin, P., Godbout, L., and St-Cerny, S. (2013). L’impact des services de garde a contribution reduite du quebec sur le taux d’activite feminin, le revenu interieur et les budgets gouvernementaux. Revue Interventions economiques. Papers in Political Economy, 47.

[11] Lefebvre, P., Merrigan, P. (2008). Childcare policy and the labor supply of mothers with young children: a natural experiment from Canada. Journal of Labor Economics 23, 519–548.

[12] Lefebvre, P., Merrigan, P., Verstraete, M. (2009) Dynamic Labour Supply Effects of Childcare Subsidies: Evidence from a Canadian Natural Experiment on Low-fee Universal Child Care.  Labour Economics 16: 490-502.

[13] Couillard, K. (2018) Early Childhood: The Quality of Educational Childcare Services in Quebec. Observatoire des tout-petits. Montreal, Quebec, Fondation Lucie et André Chagnon.  Page 25 of this document charts the measured quality differences between CPEs (not-for-profit centres) and the for-profit non-subsidized daycares.  In the CPEs that are the heart of the supply-side funded system, in two age categories, 4% or fewer of centre rooms are of poor quality.  In the for-profit child care centres funded by demand-side tax credits to quickly boost supply, 36%-41% are of poor quality.

[14] Cleveland, G., Mathieu, S., and Japel, C. (2021) What is “the Quebec Model” of Early Learning and Child Care? Policy Options, Institute for Research on Public Policy, Montreal QC.,educational%20child%20care%20after%20that.

[15] See Cloney, D., Cleveland, G., Hattie, J., and Tayler, C. (2016) Variations in the Availability and Quality of Early Childhood Education and Care by Socioeconomic Status of Neighborhoods Early Education and Development Vol. 27(3 ):384 – 401, and also see : Australian Children’s Education and Care Quality Authority (ACECQA) (2020) Quality ratings by socio-economic status of areas, ACECQA, Sydney

[16] Cleveland, G. (2017) “What is the Role of Early Childhood Education and Care in an Equality Agenda?” pp. 75-98 in Robert J. Brym ed. Income Inequality and the Future of Canadian Society ISBN-13:978-1-77244-044-7 Oakville, ON: Rock’s Mills Press. Proceedings of the inaugural S.D.Clark memorial symposium.

[17] Kottelenberg and Lehrer (2017) op. cit.

The story coming from the CSELCC survey – I don’t think we’re going to make it…not even close!

We know that child care affordability is improving dramatically because of the $10-a-day program (otherwise known as CWELCC or the Canada-Wide Early Learning and Child Care Program).  But what about access and availability?  It’s difficult to know.  There is some activity, and lots of announcements, but are there actually more children using licensed child care?  A really important question, because most of the social and economic benefits of the $10-a-day program come from improving access to children and families that haven’t used child care before.

Finally we have some solid answers.  Statistics Canada just completed a massive survey of parents across the country that tells us how many children have access to centre-based child care (the overwhelming bulk of licensed child care in the CWELCC program is in centres).  We can compare this to the situation before the pandemic in 2019.  Unfortunately, the picture is not positive.

Looking only at the provinces and territories that are part of the CWELCC program (i.e., leaving out Quebec), there are 521,800 children 0-5 using centre-based child care in 2023.  There were 483,200 children 0-5 using centre-based child care in 2019.  That’s an increase of centre-based spaces in the provinces and territories participating in CWELCC of 38,600 spaces over the course of the last 4 years, an increase of about 8%

However, the agreements signed between the federal government and the provinces and territories promised that there will be 250,000 additional child care spaces available by March 31st, 2026.  That would be an increase of over 50% compared to the spaces that were available in 2019.  That’s just over two years away.  I don’t think we’re going to make it.  Not even close!

The CSELCC survey indicates that 49% of parents using child care reported difficulty finding it.  Up from 36% in 2019. 

In 2023, 26% of parents with children 0-5 who are not using child care reported that their child is on a waitlist for child care, up from 19% in 2019.  Almost half (47%) of infants younger than one year who are not using child care are on a waitlist!!!  That’s up from 38% in 2022.

Yes, the affordability problem has improved.  But availability or access is either worse or not much better depending on your point of view.  And accessibility is improving at a snail’s pace compared to the promised additional 250,000 spaces.  Hurray for Statistics Canada giving us a clear picture of this problem.  Now federal and provincial/territorial governments have to seriously address the problems of how to grow our wonderful child care system in the not-for-profit and public sectors that are the priority.