Introduction

In many OECD countries, governments have invested in early childhood education and care (ECEC) policies and programmes for reasons that include gender equality, poverty alleviation, labour market activation, supporting life-long learning, encouraging social cohesion and inclusion, facilitating Indigenous reconciliation, integrating children with special support needs, and more. While researchers have focused on the positive child development outcomes that can occur when children have access to high-quality early years programmes,Footnote ¹ labour market concerns have also spurred governments to develop policies to support maternal employment and to counteract what researchers have called the motherhood penalty (Reference Budig, Misra and BoeckmannBudig, Misra, & Boeckmann, 2016).

Social investment patterns in Canada and the United States generally echo those of other liberal welfare states, yet they diverge when it comes to ECEC. In Australia, New Zealand and the U.K., government spending on ECEC programmes has increased dramatically as a percentage of national GDP in recent decades. Yet, in Canada and the U.S.A, services and spending on ECEC lag behind these comparator countries (Reference WhiteWhite, 2017). While federal and provincial governments have invested in childcare and early childhood education, programmes remain under-financed and weakly developed relative to other OECD countries. In Canada, with the exception of child care investments in Québec and the expansion of publicly provided full- or part-day kindergarten programmes for five-year-olds in some provinces (and four-year-olds in the case of Ontario), families confront a spotty childcare market of not-for-profit and commercial programmes, characterised by high fees and low access. As a result, many families have to rely on unregulated care or to do without any care arrangement, even when both parents work or study full-time. In this sense, childcare is markedly unlike primary and secondary education, which has historically been both publicly financed and publicly delivered and which is heavily regulated.

In countries such as Canada and the United States, childcare services originated as ‘split’ or bifurcated systems, with one set of policies and programmes targeted to the poor as part of social welfare services, and another set of policies directed to education. We argue that the current patchwork of programmes and services in Canada reflects the legacy of this split system and the associated decisions that led governments to rely on demand-side funding and tax relief instead of public provision of childcare services. Legacies of this policy history include regional inequality in access – what recent literature has labelled childcare ‘deserts’ – as well as inequities in access based on family income and other factors.

We thus begin by tracking some of the inequities in access that can be observed in Canada. We also review the range of cognate family policies that support parental employment to track the inequalities in access to those programmes and services. We argue that a liberal welfare state tradition that historically has encouraged private and market based and gendered system of care, in the context of a relatively decentralised federal institutional system, drives distributional inequalities and makes coordinated policy-making challenging. Recent initiatives by provincial and federal governments, including the 2017 Multilateral Framework on Early Learning and Child Care, are not sufficient to induce what Hall (Reference Hall1993) labels third-order paradigmatic change to meaningfully improve cost, quality, and childcare access across regions and income groups in Canada.

Distributional disadvantages in existing Canadian childcare markets

American researchers have recently introduced the concept of ‘childcare deserts’ as a metaphor for variation in access to childcare services (Reference Malik and HammMalik & Hamm, 2017). Malik and Hamm (Reference Malik and Hamm2017) have operationalised a childcare desert as a census track where there is a ratio of under 3:1 childcare capacity to children (where childcare is defined as ‘all childcare centres; family childcare providers; Head Start providers; and public and private preschools’) and where there are at least 50 children under age 5. In a study of the most populous 22 states, these researchers found 51 per cent of Americans live in a childcare desert. In childcare deserts, childcare waiting lists, unlicensed childcare arrangements, and/or negative effects on parents’ employment decisions are common. Deserts are more common in rural areas (at 58 per cent), than in suburban neighbourhoods (44 per cent). Low-income urban areas have ‘high rates’ of childcare deserts, with 55 per cent of urban census tracts showing childcare supply gaps. Affluent rural, suburban and urban neighbourhoods fare better. Hispanic/Latino, as well as American Indian and Alaska Native (AIAN) communities are over-represented in deserts, with about 60 per cent of their combined child population living in an area of low supply (which worsens to 75 per cent for the rural AIAN population). In American childcare deserts, maternal employment is a full three percentage points lower than in communities with better childcare supply; and maternal employment falls even further in deserts where median family incomes are below the national average (Reference Malik and HammMalik & Hamm, 2017, pp. 2–4).

Canadian researchers have not yet systematically mapped out childcare deserts. Yet, variation in families’ access to high-quality childcare services by income, municipality, and province and territory creates a number of distributional disadvantages which we trace below. First, some geographically focused research has sought to quantify the spatialised nature of Canadian childcare access. In a study of the distribution of childcare centres in Winnipeg, for example, Prentice (Reference Prentice2007) found that spatial inequity characterised all neighbourhoods. Access rates ranged from a low of 10.8 per cent to a high of 24.3 per cent – a difference of more than two-fold. Poorer neighbourhoods, where Indigenous populations are larger, were particularly disadvantaged, having less access and fewer childcare spaces than more affluent suburban areas. The worst access was found where the poorest children lived.

Similar results have been found in other parts of Canada. A Vancouver study found a 10-fold differential in neighbourhood childcare access between the best- and least-served parts of the city (Reference HertzmanHertzman, 2004). Close analysis shows that areas with the lowest access to licensed child care are the areas with the lowest levels of education among the adult population (Hertzman, McLean, Kohen, Dunn, & Evans, Reference Hertzman, McLean, Kohen, Dunn and Evans2002). Hertzman (Reference Hertzman2004, p. 3) found that the least-served areas are in the working class areas of the east side, where quality childcare would likely have the greatest developmental benefit: barriers to access therefore ‘are clearly more significant in lower socioeconomic neighbourhoods’.

Research in Québec has also found a spatialised socio-economic gradient in access and quality, notwithstanding the province's comparatively generous childcare policy relative to other Canadian provinces and territories. Japel, Tremblay, and Côté (Reference Japel, Tremblay and Côté2005), for example, found a few years after the roll out of the innovative provincial system that with significant growth in childcare spaces, ‘it is the more privileged rather than the less privileged Québec families who are reaping the greatest benefit’ (p. 30) including in urban centres such as Montréal.

Spatialised inequity is a logical consequence of the delivery model long-established in Canada (Reference Friendly and PrenticeFriendly & Prentice, 2009). Public policy in virtually all provinces and territories supports a mixed market of childcare, with virtually no publicly delivered childcare (Reference PrenticePrentice, 2006). In relying on non-state actors (primarily not-for-profit organisations) to implement public policy, governments therefore have only indirect mechanisms to encourage service expansion in underserved areas, to promote quality, or to regulate costs. For these reasons, Canada's existing childcare policy architecture is profoundly inadequate for meeting child, family, women's and social needs. Below we document in greater detail the distributional disadvantages in access by income.

Despite federal and provincial investments, licensed childcare spaces are scarce in Canada. Regulated childcare services include full-time, full-day services for younger children (0–6 years) and part-day and summer programmes for school-age children (ages 6–12 years). For the country's approximately 4.9 million children aged 0–12 years, 1.35 million licensed childcare spaces were available as of 2016 (Friendly et al., Reference Friendly, Larsen, Feltham, Grady, Forer and Jones2018). There was thus a licensed childcare space for 27 per cent of all children ages 0–12 and 29 per cent of children ages 0–5. The most recent data also reveal vast differences in availability across provinces and territories, with coverage ranging from a low of 8 per cent of children aged 0–12 in Saskatchewan, to a high of 55 per cent of children aged 0–12 in Québec (Friendly et al., Reference Friendly, Larsen, Feltham, Grady, Forer and Jones2018, p. 146). Studies show at least 70 per cent of childcare facilities report a waiting list (that figure is higher in some cities, such as Toronto, where 95 per cent of centres report a waiting list (Reference Macdonald and FriendlyMacdonald & Friendly, 2017, p. 20)). In many places, the wait is long. A 2016 Manitoba survey found that waiting periods of up to twenty or more months are common, with 15 months being the average (Manitoba Child Care Association, 2016).

When parents manage to secure a scarce childcare space, they are responsible for the cost (unless they qualify for a fee subsidy, discussed below.) Despite provincial policies that may regulate childcare fees,Footnote ² childcare expenses can be staggering. A 2017 report found that median infant fees in Toronto and Vancouver are $21,096/year per child and that Calgary preschool fees are $14,544/year per child (Reference Macdonald and FriendlyMacdonald & Friendly, 2017). A Toronto or Vancouver family with an infant and a three-year-old pays close to $36,000 annually, if they are able to find two regulated spaces. A Toronto economic modelling study found that licensed childcare is unaffordable for 75 per cent of families in Toronto (Cleveland, Krashinsky, Colley, & Avery-Nunez, Reference Cleveland, Krashinsky, Colley and Avery-Nunez2016, p. 4).Footnote ³ Parents who have paid childcare fees may partially offset the cost through the federal Child Care Expense Deduction (CCED) at tax time, but this tax deduction is regressive in nature, with higher income families (who can afford to pay more for childcare) receiving greater tax deductions. And still, net childcare costs in Canada are higher than the OECD average for all family types, even after taking into account tax-based benefits related to childcare (Cleveland, Forer, Hyatt, Japel, & Krashinsky, Reference Cleveland, Forer, Hyatt, Japel and Krashinsky2008, p. 16).

Outside Québec (which provides high levels of low-cost care as well as tax credits to about 55 per cent of children),Footnote ⁴ low-income parents may qualify for fee relief through the provincial fee subsidy system. Fee subsidies are demand-side financing directed to users to help them purchase childcare (unlike supply-side funding directed to the operation of facilities). Such subsidies paid on behalf of low-income families represent a significant portion of provincial and territorial childcare spending (Reference Beach and FriendlyBeach & Friendly, 2005) but demand for subsidies is much higher than supply. The hurdles that parents must pass to qualify for eligibility are onerous.

In order to qualify for a fee subsidy, parents must have demonstrably very low income and ordinarily must be working, attending school, or in training. Parents must continuously prove ongoing eligibility. One common requirement is that parents be employed: when parents lose a job, they usually find they are ineligible for subsidy after a very short period, and so must relinquish a precious space. Sometimes, social need or special condition of risk may qualify a child for a subsidy. In most cases, families must be well below the poverty line to qualify for a maximum subsidy. Even where families qualify for a maximum subsidy, a co-payment is often required. In Manitoba and Saskatchewan, a mandatory minimum user fee is required of even fully subsidised parents (Friendly et al., Reference Friendly, Larsen, Feltham, Grady, Forer and Jones2018, pp. 78 and 89–90). In other provinces, such as Nova Scotia, childcare programmes may charge parents more than the maximum subsidy rate. In Ontario, waiting lists are common for eligible families who cannot secure a fee subsidy even if they have been able to secure a licensed space – and vice versa, especially for parents seeking scarce infant spaces.

The Canadian Centre for Policy Alternatives (Macdonald and Klinger, Reference Macdonald and Klinger2015) has recently estimated the impact of each province's subsidy system on the fees for a preschooler in a hypothetical two-parent, two-child family living below the after-tax low-income measure poverty line of $30,480 in 2015. The preschooler's childcare fees would be $4–7/day in Newfoundland, Ontario, PEI, and New Brunswick; would rise to $12.19/day in Manitoba; and would be as high as $20.74/day in Saskatoon (Reference Macdonald and KlingerMacdonald & Klinger, 2015, Appendix C). Between co-pay requirements and low subsidy rates, ‘fully subsidised’ families can be charged over $400/month per child; moreover, they may be unable to even access a fee subsidy, despite meeting restrictive eligibility criteria, as subsidies are rationed in some jurisdictions. Bluntly put, childcare subsidy programmes rarely produce full affordability for low-income families.

Provincial and territorial funds are directed to the demand-side (fee subsidies) as well as to recurring funding (operational supply-side). Childcare is primarily delivered by not-for-profit organisations (70 per cent), with commercial owners supplying 30 per cent of all spaces – although the share by auspice enormously within and between provinces. Commercial operators provide 72 per cent of all spaces in Newfoundland and Labrador but not a single space in Nunavut and the Northwest Territories. In the largest provinces – Ontario and Québec – for-profit providers comprise about one-fifth of the market (Friendly et al., Reference Friendly, Larsen, Feltham, Grady, Forer and Jones2018, p. 171). Research evidence has found that highest quality care is generally provided by not-for-profit services (Reference Sosinsky, Lord and ZiglerSosinsky, Lord, & Zigler, 2007). British Columbia, Alberta, Ontario,Footnote ⁵ PEI, Nova Scotia, New Brunswick, and Newfoundland and Labrador permit a parental fee subsidy to be used at either a commercial or a not-for-profit centre; only Saskatchewan and Manitoba restrict fee subsidies to not-for-profit facilities. Provincial and territorial policies have long supported a mixed economy of childcare providers (Reference Friendly and PrenticeFriendly & Prentice, 2009).

Even in Québec, despite the many real gains in access, quality levels remain troublingly low (as in the rest of Canada), with more lower income children in lower, rather than higher, quality settings (Doherty, Lero, Goelman, Tougas, & LaGrange, Reference Doherty, Lero, Goelman, Tougas and LaGrange2000; Reference Japel and WelpJapel & Welp, 2009) – a concerning socio-economic gradient in access. Data from 2009 and reported by the l'Institut de la Statistique du Québec (Institut de la Statistique du Québec, 2011, p. 177) confirms that, with the exception of the very highest income earners, utilisation of subsidised child care spaces increases with parental income (see also Fortin, Reference Fortin2018). Québec is further challenged by growthFootnote ⁶ in the commercial market and concomitant political pressure from for-profit operators to establish policy conditions and regulations that promote their profitability. Gaps in access reveals some troubling patterns of what recent literature has identified as ‘the Matthew effect’ in access to social provision, whereby increasing public provision ends up advantaging higher income rather than lower income groups (Pavolini & Van Lancker, Reference Pavolini and Van Lancker2018; Van Lancker & Ghysels, Reference Van Lancker and Ghysels2012, Reference Van Lancker and Ghysels2016).

As a result of shortage of spaces in licensed childcare all across Canada, a disproportionate number of low-income parents rely on the unlicensed childcare sector (Reference Varmuza, Perlman and WhiteVarmuza, Perlman, & White, 2018). Data from the 2011 cycle of the General Social Survey reveal that 23 per cent of working parents report having no regular form of non-parental childcare (Varmuza et al., Reference Varmuza, Perlman and White2018). Some families do not require non-parental care for work, education or training; in other families, kin provide care. Some two-parent couples ‘off-shift’, organising paid work to ensure one parent is available to care for children.

While some unlicensed childcare is of good quality, research demonstrates that unlicensed care is generally of lower quality than licensed services (Reference Bassok, Fitzpatrick, Greenberg and LoebBassok, Fitzpatrick, Greenberg, & Loeb, 2016), for reasons that include lack of requirements around health and safety, as well as lack of caregiver credentials. Data from the 2011 cycle of the General Social Survey, based on parents’ self-reports, reveal that about 17 per cent of children are cared for in unlicensed settings (Varmuza et al., Reference Varmuza, Perlman and White2018). But it appears that some portion of parents using home-based childcare often do not know if their childcare facility is even licensed. In all provinces, regulators only investigate unlicensed childcare on a complaint basis; there is no proactive monitoring. Public polling has shed light on public perceptions about unregulated care: one recent study found that three of four respondents believed that all daycares have to meet regulations (Reference Ferns and FriendlyFerns & Friendly, 2015, p. 6). Parents have difficulty evaluating the quality of the childcare they purchase (Reference MocanMocan, 2007). And, especially in the absence of strong market competition, researchers have found that quality is highly variable (Bassok et al., Reference Bassok, Fitzpatrick, Greenberg and Loeb2016; Sosinsky et al., Reference Sosinsky, Lord and Zigler2007).

Distributional disadvantages in Canadian work and family policies

As with childcare, other elements of Canadian family policy and work-family reconciliation are also weakly developed compared to other OECD countries. Eligible new parents in all provinces and territories (except Québec, which has its own paid leave planFootnote ⁷) have access to either standard or extended parental leave benefits under federal Employment Insurance. Under both plans, 15 weeks of maternity leave are reserved for birth mothers. In the standard leave option, parents have 35 weeks at 55 per cent of previous salary to be taken by either parent or shared between them; under extended benefits, parents have up to 61 individual or shared weeks at 33 per cent of previous salary (Government of Canada, 2018). Beginning in June 2019, a ‘use it or lose’ and additional five-week ‘Parental Sharing Benefit’ will begin, providing extra time for a second parent (paradigmatically, a father). Benefit rates under both standard and extended scenarios, as well as the shared benefit, are low and average weekly insurable earnings are further capped at annual salaries of $51,700 (as of January 2018).

Eligibility for maternity and parental leave and benefits maps onto Employment Insurance, a programme that has been steadily eroded (Reference PorterPorter, 2015). As a result, many new parents discover they are ineligible for the leave. A recent study has found nearly four in ten mothers outside Québec do not qualify (McKay et al., Reference McKay, Mathieu and Doucet2016). In one in three families with a new child, neither parent claims EI family benefits (Reference RobsonRobson, 2017, p. 20). Canadian parents thus do not have equal access to maternity and parental leave benefits, as some are ‘parent-leave rich’ while others are ‘parent-leave poor’, and this gap maps onto class differences.

Class gaps are exacerbated where supplemental unemployment benefit plans permit employer top-ups. Only 17 per cent of new mothers outside Québec receive a top-up. In Québec, where EI eligibility is more generous, employees are 2.7 times more likely to receive a top-up. Generally, employer top-ups are found in large, public sector, and unionised workplaces, where wages and benefits are already better than average: half of all public sector employees are in receipt of top-ups, usually for the first fifteen weeks of their leave (Reference MarshallMarshall, 2010). The top-up rate in 2010 among those earning less than $20/hour was 9 per cent, compared with 30 per cent for those with a $20–24.99 hourly wage, and 36 per cent for those who earned $25 per hour or more (Reference MarshallMarshall, 2010, p. 9). Most Canadian families rely therefore on the limited parental leave programme and experience income disruption upon the birth or adoption of a child – and precarious and poorly paid workers experience proportionally the most income loss, compounding inequalities between women.

Even where parents do qualify for paid parental leave, the design of current policy carries strongly gendered effects. As McKay et al. point out, ‘leaves of absence straddle the intersection of employment relations and family life’ (Reference McKay, Mathieu and Doucet2016, p. 7). Parental leave benefits are modelled on employment insurance and are a ratio of previous earnings, factors that explain how households attempt to minimise income disruption. Where there are two income earners, income loss can be reduced by ensuring the lower income earner is the one who takes the leave: in most cases, this means the mother. Research by Statistics Canada has demonstrated that mothers’ earnings are ‘clearly the overriding factor’ in determining the length of leave mothers take from employment (Reference MarshallMarshall, 2003). Few fathers in two-parent families take a leave: just four per cent of couples outside Québec share benefits (Reference RobsonRobson, 2017, p. 20). This may change after 2019, when Canadians will begin to have access to a five-week dedicated second parent leave, with ‘use it or lose it’ provisions.

Under other recent changes to parental leave, families now have the new option taking a longer parental leave at a reduced weekly payout. Parents can take up to an 18 month leave, with the same total benefit, in a policy that was launched under the banner of promoting ‘more choice and flexibility’ (Government of Canada, 2017a). Yet, whether parents come off parental leave when their babies are 12 or 18 months, they confront a landscape of few infant spaces, and exorbitant fees where a space can be found. For this reason, one comprehensive study of parental leave in Canada dryly notes that ‘Canada's parental leave systems are generally not well synchronized with childcare policies’ (Reference RobsonRobson, 2017, p. 12). Comparative research points to the ways that long leaves as well as low remuneration can compound the maternal penalty (Budig et al., Reference Budig, Misra and Boeckmann2016).

Other federal programmes provide support to families with children. Families are now eligible for the monthly Canada Child Benefit (CCB), launched July 20, 2016. The CCB consolidated and replaced previous programmes, offering up $6400/year for children under age 6 and $5400 for children aged 6–17, scaled to family income. Initial calculations are that a total of 4,056,060 households, with just over 7.1 million Canadian children, are eligible for the CCB (Office of the Parliamentary Budget Officer, 2016). The actual pool of recipients in the first year, however, was 3.6 million households (or 88 per cent of the anticipated total). The gap between potential and actual recipients is troubling, since indications strongly suggest that it is the most marginalised households who are missed. Families must have filed taxes to be eligible, and there is a socio-economic gradient to tax filing. Indigenous families are particularly unlikely to enrol for the CCB (Reference McKieMcKie, 2017). Cindy Blackstock, Executive Director of the First Nations Child and Family Caring Society of Canada, points out that many Indigenous families have not yet heard of the new benefit and have little trust in government programmes. The Assembly of First Nations concurs, arguing ‘relying on tax returns is not effective’ (Reference McKieMcKie, 2017). In total close to 500,000 households, where the cash transfers would be most needed, are not benefitting from the new programme.

Financial forecasts show that there is a further problem: the CCB is not indexed. The independent Parliamentary Budget Office has calculated that in 2016–2017, payments under the new system cost the government $3.4 billion more than under the previous regime, yet by 2024–2025, total net spending on the CCB is projected to fall to levels that would have equalled the old system of children's benefits. As the report summarises, ‘the real value of the payments, the nominal level of payments for which families are eligible, and the number of eligible families will decline over time’ (Office of the Parliamentary Budget Officer, 2016, p. 4). Since the CCB is sensitive to family income and income thresholds are not indexed, the PBO observed that over time ‘more families will see their benefits decline and fewer families will receive benefits’. The time horizon of the new benefit, therefore, is less than ten years of more generous support to families. The formal agenda thus contradicts important elements of the public political agenda (in particular, for support to the middle class and families).

Low uptake of the new benefit is profoundly concerning. It means that the most vulnerable children, those with the highest needs, are drawing the least benefit. A socio-economic gradient in access to the new child benefit has strong parallels with inequitable access quality childcare and childcare use patterns, discussed earlier.

In summary, Canada has a system of benefits and services for children and families that are unevenly distributed and are in many cases not benefiting families who are most in need. Provincial spending on childcare is primarily oriented to demand-side subsidies and each province and territory has its own policies, programmes, and regulations, with significant variation in supply and regulation. A small number of lucky families are able to find and pay for (or are eligible to receive a fee subsidy for) a regulated childcare space. But middle-income families are generally priced out of licensed childcare. Even in Québec, lower income children are less likely than affluent children to use regulated childcare. The practical effect of public policy is to compound social stratification.

For the moment, most Canadian families with children are benefitting from a relatively generous new Child Care Benefit; however, the most marginalised and low-income families find the programme's bureaucratic requirements too onerous to manage. Low-income, Indigenous, immigrant and newcomer families are markedly disserved by the socio-economic gradient of childcare and family policy. Higher income parents, especially those with employer top-ups, can better manage the income disruption associated with birth or adoption while parents in occupations not covered by federal EI rules or who have not accumulated the requisite number of hours of work prior to pregnancy are at a loss. Canadian parents thus grapple with starkly divergent realities of childcare service, maternity and parental leave, and family support.

These recognised market failures and policy gaps in other countries have typically triggered public investment in the supply of childcare and other policies to help reconcile work and family life and support parental labour market participation. The OECD and UNICEF recommend a minimum national expenditure of one per cent of GDP on ECEC services to support families (UNICEF, 2008). Yet Canada fails to come near to meeting this benchmark.

How did we get here?

As political commentators have noted when asked about why electoral losses occur, ‘the party/candidate didn't get enough votes’, a history of inadequate public funding is the primary explanation for why licensed childcare services are rare and/or expensive in Canada and family supports are unevenly distributed. The more important question is what explains this systemic underinvestment. Drawing on historical institutional approaches and varieties of capitalism literature (Reference Esping-AndersenEsping-Andersen, 1990; Reference Hall and SoskiceHall & Soskice, 2001), part of the answer lies in Canada's particular political economy. Empirical research has demonstrated that while active labour market policies and work-family reconciliation policies have characterised, in different ways, both the social democratic Nordic states and the ‘conservative’ world of continental European countries, such policies have traditionally not been priorities in the liberal welfare states and liberal market economies. Morgan (Reference Morgan2005) argues that there are fewer incentives on the part of governments and employers in countries with an abundance of low-wage, low-skill employment – i.e. liberal market economies – to provide labour market enhancing policies; and, indeed, low-wage female workers are drawn upon to provide childcare in markets. In coordinated market economies characterised by higher unionisation rates, governments have incentives to subsidise the cost of services because of the high costs of delivery of such services due to quality regulations. Some conservative welfare states have also begun to subsidise the cost of childcare in order to encourage maternal participation in the labour market (Oliver & Mätzke, Reference Oliver and Mätzke2014).

A second explanation for Canada's relative underinvestment stems from traditional societal norms regarding the extent to which the state supports families and encourage women's – and particularly mothers’ – labour market participation. Liberal welfare norms have traditionally reinforced the view that the state is the provider of last resort. Historically, governments in liberal welfare states have provided funding and services for poor and highest-risk families but encouraged ‘the bulk of the population to rely as much as possible on private sources of economic security, including occupational benefits and personal savings’ (Reference Banting, Myles, Banting and MylesBanting & Myles, 2013, p. 4). Middle and upper income families are expected to manage the tasks of social reproduction by themselves, either through mothers’ own caregiving or through the purchase of private services.

Institutional factors reinforce those liberal norms that discourage state presence in the financing and delivery of childcare and family policies. Comparative research of other countries’ social spending demonstrates the relationship between high levels of social spending and factors such as high union density, proportional electoral systems, and left parties in government or governing coalitions (Reference Iversen and SoskiceIversen & Soskice, 2009; Reference Iversen and StephensIversen & Stephens, 2008). Canada does not have a strong left-party tradition nationally, although left parties have seen some electoral success at the provincial level. While the level of unionisation has been higher in Canada than in the U.S.A (Reference Card, Lemieux and RiddellCard, Lemieux, & Riddell, 2004), union density is not as high in Europe and Canadian labour unions are both more ideologically divided and decentralised (Reference Banting, Myles, Banting and MylesBanting & Myles, 2013). The strongly decentralised nature of Canadian federalism has meant, with a few exceptions in the post-World War II era of cooperative federalism, a smaller and less generous range of social programmes, particularly in the area of children and families, than in many other OECD countries (Reference Banting, Myles, Banting and MylesBanting & Myles, 2013). Federal funding of childcare services has been sporadic, and highly politically contested (Reference White and FriendlyWhite & Friendly, 2012).

Access to social services has also been subject to gendered norms in Canada. Early post-war policies assumed a ‘housewife’ at home, and so questions of social reproduction and care were largely assigned to the private sphere (Reference ChristieChristie, 2000; Reference Langford, Prentice and AlbaneseLangford, Prentice, & Albanese, 2017). Policies such as unemployment assistance and pensions are tied to employment, and thus for decades primarily benefitted mainly male breadwinners, while family policy to support women's employment was not a national priority. As Mahon, Bergqvist, and Brennan (Reference Mahon, Bergqvist and Brennan2016) argue, Canada traditionally did not offer much support to female caregivers. Social norms and political values thus justified the institutional arrangements that have created the social reproduction care crisis and care deficit in Canada (Langford et al., Reference Langford, Prentice and Albanese2017; Reference WhiteWhite, 2017). The defunding and delegitimising of the women's movement in recent years has muted many gender critiques of existing policy (Reference BrodieBrodie, 2010).

This institutional distribution of political power, along with political economic shifts such as global economic pressures, compounded by the relative waning influence of civil society organisations that advocate on behalf of disadvantaged groups in Canada, have meant many policies have remained ‘frozen’. As the Mowat Centre (2015) has observed, ‘Canada's social policies and programs have barely changed since the 1960s, despite major transformations in our economy and society.’ Many aspects of Canadian social policy are affected by the policy architecture established in the country's foundational constitutional documents, wherein provincial governments maintain jurisdictional authority over many aspects of social policy but the federal government maintains the greater fiscal levers. The long history of costly intergovernmental battles has left many governments reluctant to engage in new efforts at coordination in many policy sectors. Thus, with some important exceptions – such as pensions federally, and health care and primary and secondary education (two of the largest areas of provincial expenditure) – many aspects of Canada's social policy have atrophied; the historical institutionalist literature describes such atrophy as policy ‘drift’ (Reference Streeck and ThelenStreeck & Thelen, 2005, p. 24). Policy drift occurs when the context within which policies are made changes yet institutions and policies do not change such as to alter their effects. Policy drift is more likely when existing rules prevent change (such as electoral institutions that ensure one of two major parties are elected nationally) or where policy makers encounter numerous institutional veto points such as in federal–provincial bargaining.

Banting and Myles (Reference Banting, Myles, Banting and Myles2013, p. 3) additionally argue that governments in Canada have failed to modernise social policy in light of new social risks. Very little by way of policy has been added to the policy architecture established in the post-World War II period to support families with children, with the exception of some tax programmes such as the geared-to-income Canada Child Tax Benefit in 1993 and the National Child Benefit Supplement in 1997 that targeted low-income families. Both were rolled into the CCB in 2016, along with the previous Conservative government's Universal Child Care Benefit (UCCB) programme. Kindergarten programmes have recently been extended to full day in a majority of provinces and territories; that was accomplished via institutionally ‘layering’ onto kindergarten programmes already in existence rather than creating whole new programmes from scratch.

In summary, policies such as childcare and parental leave have remained comparatively less generous than in many of the European welfare states. On the whole, childcare is still largely treated as a private good garnering private utility to parent consumers. In the case of low-income parents, childcare has been transformed from a private utility to being instrumental in assisting parents become and remain employed. Governments in all provinces and territories thus subsidise some of the cost of purchasing care for low-income families but are less attentive to concerns about quality that would yield positive child development outcomes. Until children reach kindergarten age – making the jump from ‘care’ to ‘education’ – childcare is treated as a private good to be purchased in markets with a varying range of quality and price. With the exception of Québec, where there is a growing policy awareness of the connection between childcare and expansion of mothers’ labour market participation, childcare is not treated in policy terms as part of human capital investment – let alone as a social investment that can improve the lives of children and increase gender equity, as is the case in the Nordic countries.

Weak prospects for third-order policy change

Hall (Reference Hall1993) characterises policy paradigm change as not only changes in policy instruments and settings but also change in policy makers’ goals. Given that so much of child care policy support is directed to families rather than to services, current policy initiatives are unlikely to bring about ‘third-order’ policy paradigm change.

At the national level, the last federal effort to significantly expand childcare service funding through bilateral agreements with the provinces was launched under a minority Liberal government in 2005. These agreements were cancelled by the incoming federal Conservative Party in 2007. In 2015, the federal Liberal government campaigned again on funding for childcare. During the 2015 campaign, the Liberal Party made a priority delivering large infrastructure investments, strengthening the middle class, and expanding Canada's economy. The party identified childcare as a key element of social infrastructure that included affordable housing, seniors’ services, and culture/ recreation. It vowed to ‘fund the creation of thousands of new childcare spaces, enhance their quality, and ensure that affordable childcare spaces are available to more families who need them’ (Liberal Party of Canada, 2016). The 2016 and 2017 federal budgets committed $7.5 billion to childcare over 11 years, starting in 2017–2018.

In June 2017, the federal government released a Multilateral Framework on Early Learning and Child Care to begin the work to deliver ‘affordable, high-quality, flexible, and fully inclusive’ childcare for Canadian families (Government of Canada, 2017b).

The Multilateral Framework layers onto existing structures and approaches, in particular, the provincial reliance on mixed markets and predominately voluntary-sector delivery. The funding mechanism transfers funds to provinces, whose provincial governments will largely use the new dollars to maintain or expand existing financing models, premised on parent fee subsidies and generally modest recurring support. The unique arrangements of federalism will permit Québec to continue to organise its own childcare system.

Thus, the Multilateral Agreement cannot be considered a transformational reform, as it simply layers new funds onto the existing architecture – a classic first-order change, using Hall's typology (Reference HallHall, 1993). The total funding committed is significant, but not in comparison to the cancelled 2005 bilateral agreements which committed $5 billion over five years. In those 2005 agreements, the federal government required funds to be spent on services that met the QUAD principles: quality, universality, accessibility, and developmentally appropriate. The Child Care Advocacy Association of Canada notes the new agreement does not even require funds to be spent on regulated childcare services (CCAAC, 2017).

Moreover, the existing federal tax system in the form of the CCED maintains the policy of reimbursing parents for multiple types of expenses. Few restrictions are placed on the types of care that qualify other than the provider's willingness to provide receipts. Childcare, day camps, and boarding schools are among the long list of eligible expenses. Taxpayers can claim up to $8000 per child for children under the age of 7 and $5000 per child for children aged 7–16; generally, the higher the child care expenses accrued, the greater the deduction. Under the terms of the CCED, there are no criteria related to quality or education. Any child minding, across a range of settings, is permissible. The claimed expenses cannot exceed two-thirds of a parent's earned income for the year and must be deducted from the income of the lower earning parent in a two-parent family (generally the mother).

Thus the federal childcare architecture is tightly constrained. The tax and provincial transfer system reinforces distributional disadvantages and reinforces wide provincial variations in spending and regulatory regimes and attendant provincial and regional gaps in access and affordability.

At the provincial level, some experimentation is going on, particularly around efforts to regulate parent fees. Some provincial initiatives have promising potential. For example, the Ontario government announced, as part of the 2018 budget plan, and based on the recommendation of the province's expert review, a plan to implement free ECEC services for children ages 2.5 to kindergarten, while continuing to fund the expansion of childcare spaces (Ontario Office of the Premier, 2018). This marked the first time a government in Canada committed to free at the point of service childcare for an entire age group but, at the time of this writing, the plan is unlikely to survive a June 2018 provincial election Other scholars have explored municipal initiatives, such as those in Montréal, Toronto and Vancouver, where local organising, more open political opportunity structures, and greater resources (including in the Vancouver case, innovative use of zoning regulations) have enabled innovative local childcare development (Reference MahonMahon, 2006; Mahon et al., Reference Mahon, Bergqvist and Brennan2016; Reference McGraneMcGrane, 2010; Reference PrenticePrentice, 2004).

Conclusion: childcare deserts and distributional disadvantages

Childcare policy and services in Canada exhibit contradictions and marked limitations. Families confront a disorganised market of not-for-profit and commercial programmes, characterised by high fees and low access. Beyond the barrier of costs, many families live in childcare deserts, where services are entirely absent. Affluent families, by contrast, find it easier to purchase services sold through a growing, pricey, and lucrative care market. This array of services and policy is proving to be increasingly unsustainable, as women's persistently strong presence in the labour market makes childcare services more and more necessary. Across Canada (outside Québec), provincial childcare policy essentially relies on non-state actors (primarily the voluntary sector, although with a growing role for commercial operators) to deliver services. Federally, as well as provincially, governments have sought to intervene, attempting to broaden services through financing but not provision.

Funding levels remain too low, demand-side funding approaches are inadequate, and market-based programme delivery perpetuates and entrenches social and spatial inequity. Parental leave policies do not coordinate with childcare services. As welcome as they are, the new CCB will not build the services that parents needs and will miss many of the most vulnerable and marginalised families.

The institutional factors that built Canada's welfare state – marked by a strong role for the third sector in the delivery of social services; the treatment of childcare as a private maternal responsibility where the state need only intervene for the poorest families; and the strongly decentralised federal system of government that makes intergovernmental cooperation and coordination on social policy matters challenging – have led to childcare underinvestment relative to nearly all other OECD countries. The changing conditions of mothers, the majority of whom participate in paid labour, have not yet generated the kind of value change that is a precondition for a new policy agenda. The gendered and classed care deficit has not been addressed through the path-shifting third-order change that is required. Absent such change, the current national initiative cannot ensure all children experience quality early learning and childcare and will perpetuate childcare deserts and distributional disadvantages.