Why Early Childhood Education Matters and Why We Should Pay for It
by elizabeth u. cascio
liz cascio teaches economics at Dartmouth College.
Illustrations by Shaw Nielson
Published July 31, 2018
It’s tough to be a parent. But to be a working parent — particularly when the other parent also works or is absent — is even harder. Who will care for my child when I work? Should I stay home instead?
These choices are especially fraught in the United States, which invests only about as many public dollars per child in early childhood education and care (ECEC) as the typical country that is half as rich. It’s not because ECEC in the United States is cheaper or easier to finance. The price tag for full-time, full-year, center-based ECEC can easily run upward of $20,000 per year, and relatively few Americans with young children can afford to cover the cost without private loans that are simply unavailable.
The paucity of government support for ECEC can’t be attributed to public preferences. Vast majorities of voters, regardless of party affiliation, claim interest in making high-quality ECEC more affordable and accessible. Yet, the U.S. remains locked in place, with working parents facing the difficult choice of paying through the nose for center-based care, putting their children in more compromised care situations (for instance, babysitting by a retired neighbor) or leaving the labor force altogether.
Why, then, does America spend so little on child care compared with comparably affluent countries? The short answer: terminology, institutions and history all matter, and they are inextricably linked. Here, I offer some perspective and propose a politically palatable (if only partial) fix.
Terminology: The Failure to See Child Care as an Investment
Say the words “child care,” and you’ll almost certainly conjure a vision of a service that meets families’ short-term needs. Like other forms of consumption — purchases of food, TV programming or vacations — the value of child care would seem to be realized right away. Parents benefit from having their children supervised while they work, just as we might get nourishment from a barbequed chicken, pleasure from bingeing on Stranger Things or relaxation from a week in Kauai in January. Terminology matters: as with food, TV and recreation, there appears to be little in the way of benefits from one person’s “child care” spending to society at large.
The word “education,” on the other hand, implies investment. Outlays today yield returns over the long term. Many investments don’t yield public returns, but we typically think of education as being one that does. Sure, people with more education earn more in the job market, but their added education also benefits others. A more educated populace is better prepared to participate in democracy and less likely to commit crime or rely on public assistance. Better educated people make their co-workers more productive – and perhaps better paid.
And yet in practice, K-12 education is clearly a service delivering short-term benefits to households as well as an investment. Public school teachers are the unsung child care providers, offering essentially free child care during the school day. In fact, public education does more to subsidize child care for Americans than any other government intervention: spending on elementary and secondary education amounts to about $630 billion a year, whereas the next largest ECEC program costs less than $10 billion.
By the same token, child care is an investment as well as a household service. Parents are their children’s first teachers, and, for many children, child care providers are their second. How children spend their earliest years matters a lot over the long term; it is an investment in their futures.
Institutions: The Limited Federal Role in Education
If we accept that child care is an investment — and, accordingly, call it ECEC — who should pay and how? We clearly have a system in place for funding elementary and secondary education; as one of the biggest line items in public budgets, it rivals spending on Medicare and accounts for over 3 percent of the GDP. A big distinction between public health insurance and public education, however, is which government’s budget is affected — which level of government raises the revenue and does the spending. Public schooling has historically been a state and local responsibility; the federal government has played — and continues to play — a very limited role.
Local financing of education through property taxes arose organically in the United States. It has the appeal of solving a fundamental financing problem: it costs more to educate children than parents typically could pay as they go. By funding local schools through the property tax, young households effectively “borrow” from their older neighbors because they pay less in tax than it costs to educate their children. Later in life, they pay back this “loan” by financing schools that their children no longer attend.
Local finance through the property tax remains an important source of cash for schools. But since the 1970s, state financing and control have come to dominate as states have gotten in the business of redistributing income and educational opportunity across communities within their boundaries. The federal role remains roughly what it was 40 years ago, focused on providing schools with additional resources for difficult-to-educate populations, such as the poor and the disabled.
It should therefore come as no surprise that, with notions of educational investment in ever-younger children in hand, grassroots movements at the state and local levels have been wellsprings of ECEC finance in the U.S. The downward extension of public school systems to include kindergarten provides a case in point: in the 1960s and 1970s, states that lacked kindergartens introduced their first subsidies for school districts offering them. Since then, state and local initiatives have been critical to swelling the ranks of 3- and 4-year-olds in pre-K programs.
In the early 1980s, only four states funded pre-K. By contrast, in 2015-16, 43 states and the District of Columbia did, and statefunded pre-Ks enrolled an estimated 32 percent of 4-year-olds across the country. Some large cities — Boston, New York, Seattle — have gone further, financing universal programs through local initiatives.
By funding local schools through the property tax, young households effectively “borrow” from their older neighbors because they pay less in tax than it costs to educate their children.
Yet there is great diversity in what these programs look like. Many are mandated to serve only disadvantaged children or simply aren’t given the resources to serve all kids who are categorically eligible. Many also set minimum standards below those set for K-12 education, particularly in terms of teacher training. And with such piecemeal adoption, it could be a while before pre-K becomes the new kindergarten.
It took more than a century from their conception for kindergartens to appear in the public schools of the last state to subsidize them (Mississippi, in the 1980s). Circumstances are, of course, different now — parents are hungrier for ECEC than they were 40 or 50 years ago. But relying on state and local funding could still mean it will be decades before pre-K is available to all 4-year-olds. Universal pre-K for 3-year-olds is an even more distant dream: in the 2015-16 school year, only 5 percent of 3-year-olds attended statefunded pre-K programs.
It’s therefore very likely that a larger federal role would be needed to speed up this process and to meet the ECEC needs of infants and toddlers. Yet, beyond getting the job done faster, is there a good case for a larger federal role in ECEC than in elementary and secondary education? Would it give Washington the leverage to impose ideas on communities that don’t appeal to their residents?
I think a bigger federal presence in ECEC can be justified on both theoretical and empirical grounds. For one thing, local finance of public schools through the property tax makes rough and ready sense because much of the non-private return to the investment is realized within the community’s borders. Keeping idle teenagers off the streets is an example of a localized social return to public education.
Other societal returns — those that stem from better preparing children to participate in democracy, in work, in life — don’t respect boundaries. Economists generally agree, however, that these kinds of spillovers make up a relatively small share of the overall return on elementary and secondary educational investment. So local finance is capable of getting the amount of educational spending about right, especially if states chip in more resources in order to offset the impact of inequalities in income across communities.
For ECEC, on the other hand, a larger share of the overall return appears to come in the form of border-crossing societal externalities, where much of the benefit goes to other communities. Re-examination of several small field experiments from the ’60s and ’70s by Nobel Laureate James Heckman revealed that reductions in violent crime in adulthood generate much of the startlingly high return on investment in the preschool education of disadvantaged populations. And, of course, these benefits extend beyond the locality providing the education. Hence, there are good reasons to believe that a community or even a state will not invest as much in ECEC as would be best for society as a whole. And an increased federal role need not mean mandates for provision or enrollment — it could just be financial incentives to get programs off the ground.
History: Echoes of the Cold War
The federal government actually does play the dominant role in financing ECEC. On a per-child basis, it certainly does more today for ECEC than it does for elementary and secondary education. Yet, those federal programs remain small relative to the cost and the numbers of young children. This is where history becomes important.
Washington got in the business of funding ECEC in the 1960s with the establishment of Head Start as part of President Johnson’s War on Poverty. Child care is not (nor ever was) Head Start’s intent — healthy child development is. But like public education, it has the effect of providing free child care during the program day.
Unlike public education, though, Head Start narrowly focuses on children from very low-income families. While it is the largest single ongoing federal investment in ECEC (with an annual appropriation nearing $9 billion and rolls totaling about one million children under age 5) it serves only 12 percent of 4-year-olds and 10 percent of 3-year-olds nationally.
This is not to say that there weren’t later efforts to create a larger federal ECEC program, explicitly providing child care as well as education to larger numbers of children. With the continuing rise in labor force participation among women with young children, pressure mounted in the early 1970s for a comprehensive approach to child care. Washington’s response, the Comprehensive Child Development Act (CCDA), would have established a network of universally accessible, subsidized child care centers nationwide, with a federal funding commitment well eclipsing that for Head Start. Yet, though the act passed both houses of Congress with widespread bipartisan support in 1971, President Nixon vetoed it. His rationale: it would promote “communal approaches to childrearing” over a “family-centered approach.”
An increased federal role need not mean mandates for provision or enrollment — it could just be financial incentives to get programs off the ground..
The purpose of the legislation alone might have doomed it. But some have argued that, with Nixon’s veto, public funding for child care became synonymous with communism and therefore a threat to the American identity, if not the nation’s security — and that the veto had a chilling effect on efforts to fund child care with federal dollars that continues to this day. It has even affected conversations about ECEC at the state and local levels. Concerns that it would be “babysitting” or “socialistic,” for example, derailed then-governor Jimmy Carter’s early attempts to introduce kindergarten programs in Georgia.
That said, one can’t deny that efforts to expand the federal role in ECEC since Nixon’s veto of the CCDA have been limited and largely unsuccessful. What federal spending on ECEC there is comes in the form of tax subsidies and modest direct subsidy programs.
In 1976, Congress morphed a small itemized deduction for child care dating to the 1950s into the Child and Dependent Care Tax Credit (CDCTC). The Revenue Act of 1978 established a second tax program — Dependent Care Flexible Spending Accounts (DCFSAs) — which allows families to set aside pretax dollars for child care expenses. The direct subsidy program came more than a decade later. The Child Care and Development Fund (CCDF), established by the Child Care and Development Block Grant Act of 1990, provides cash to states to offer means-tested vouchers for child care to working parents with children ages 12 and under.
Only about 4 percent of children under age 5 have benefited from CCDF subsidies.
Until the promise to expand CCDF subsidies under a budget deal earlier this year, none of these programs has been close to Head Start in size. In recent years, CCDF expenditure has been about $5.4 billion annually, whereas the tax programs together reduce federal revenues by about $4.6 billion. It’s also true in terms of the number of young children served. Only about 4 percent of children under age 5 have benefited from CCDF subsidies.
Reimagining the Federal Role
All that said, is there a path forward? In a recent report commissioned by the Brookings Institution’s Hamilton Project, I argue that even some small changes to how existing federal dollars are allocated could yield big returns. These changes would recognize both where the need is greatest and where the federal government is uniquely positioned to act. Briefly:
Treat children under the age 5 more generously. A curious feature of the federal subsidy programs outside of Head Start is that they don’t really distinguish between the care of school-aged children and children under age 5 despite the fact that caring for infants, toddlers and preschoolers is considerably more expensive. As a result, substantial program resources go toward funding the care of children ages 5 to 12, who already receive highly subsidized care and education through public schools.
It may be more efficient in administrative terms to not make eligibility distinctions based on age. But it’s hard to believe that doing so would generate more costs than benefits — the latter realized in terms of improved-quality care of children at earlier ages and moving more women with young children into the labor force.
Even budget-neutral changes to the age-eligibility criteria for the CCDF and the tax subsidies would make a difference. By my rough calculation, at typical subsidy rates in recent years an additional 435,000 infants, toddlers and preschool-aged children could be served simply by shifting CCDF dollars currently spent on older children to those under age 5.
This figure would be smaller if subsidy rates per child were increased sufficiently to meet the health, safety and quality standards laid forth in the program’s 2014 congressional reauthorization. But the number would still be sizable — and, indeed, perhaps comparable in magnitude to the estimated 230,000 additional children who will be served annually by the new dollars set aside for the Child Care and Development Fund. Similarly, even if the tax programs’ terms remained otherwise unchanged, shifting at least some of the implicit tax expenditure to favor child care for families with younger children would generate more bang for the federal buck.
Make Tax-Based Subsidies More Progressive.
The subsidy shuffle outlined above would help. But it would be even better to fundamentally change the structure of tax programs for child care. The existing programs are highly regressive. Higher-income people not only receive a larger share of the benefits, but also receive larger benefits per family
One reason is that the child and dependent care tax credit is “non-refundable” — that is, it can’t be converted into a cash refund for a household otherwise owing no income tax. So it offers no net benefit for most families in the income range to be eligible for the maximum credit (about $2,100). DCFSAs are also available only when an employer chooses to offer them. The ones that do offer them tend to be larger enterprises with relatively skilled, well-paid workers. All of this means that much of the benefit of these tax programs is going to families who would have the resources to buy child care and participate in the labor force even if the program didn’t exist.
In my Hamilton Project essay, I suggested eliminating the DCFSAs altogether and replacing the CDCTC with a new, refundable tax credit that is more generous for families both with lower incomes and with younger children. The benefit structure would feature a credit of $4,000 per child for each of the first two children under the age of 5 and $2,000 for the third for families with adjusted gross incomes up to $25,000. The first- and second-child credits would diminish with income, phasing out as income rose at the rate of $400 for every $5,000 above $25,000 (and a phase-out rate of $200 per $5,000 for the third child).
It would thus disappear entirely for families with adjusted gross incomes above $70,000.
In my proposal, older children would not be entirely left out in the cold. Care for children ages 5 through 12 would qualify for tax subsidies half as large at any given level of income.
The refundability of the credit is key here; the only way to redistribute income effectively through the tax system is to allow families with no tax liability to receive refunds. Indeed, simply making the current credit refundable would make the benefit progressive, both in terms of the likelihood of claiming the credit and dollars awarded per child. Under the more generous benefit structure I outlined above, the overall cost would exceed current expenditure by only $1 billion annually.
This figure, though, only reflects the impact on the implausible assumption that it would not have the secondary effect of coaxing more children into better care. How much more it would cost would depend on just how responsive child care demand is to the subsidy.
Estimates vary widely. We do know, however, that free child care in the form of public education induces a significant shift away from private care. What is different here is that the tax subsidy would reduce, but generally not eliminate, the out-of-pocket costs of center-based care. It would also be a much blunter policy tool if available only as a lump-sum payment once a year rather than on an ongoing basis that gives families living near the edge enough liquidity to pay child care expenses weekly or monthly. But the uncertainty hardly means the program change wouldn’t be worth exploring. And my proposal is just a starting point; child care expenses are still a considerable burden for families with incomes over $70,000, suggesting that a more expansive phase-out region could be justifiable.
The marginal dollar invested in pre-K would go much further if spent on expanding access rather than on improving standards that have not been scientifically assessed.
Create incentives for more state and local support of ECEC. Nothing about the policy proposals described thus far acknowledges the “education” part of ECEC. My focus has been on the “care” part and on redirecting federal resources to the households most in need. While there have been recent efforts to professionalize the Head Start workforce and to improve the quality of care subsidized by the CCDF, the federal government could do much beyond making sure that those particular efforts are adequately funded. In fact, there is a lot of low-hanging fruit here ready to be picked.
The federal government could, for example, help to create a more transparent market for child care in which parents have easy access to information on the quality of individual centers. Armed with this information, parents could make better choices and in the process stimulate competition on quality in local markets for child care.
This is the idea behind quality rating and improvement systems (QRIS). Though now in place in many states, it takes money and effort to implement QRIS with the metrics that best represent program quality. But a growing body of research hints that the return could be high.
Now consider policymakers. The officials who administer or regulate child care at the state and local levels sometimes seem about as uninformed about pre-K programs as parents are when choosing a center. There is great diversity in the form these programs take — not only in terms of size, but also in the populations they serve and the quality standards they enforce. States might have different preferences, but it’s hard to believe that the current wild variation in pre-K programs reflects anyone’s informed choice.
My own recent research, for example, suggests that the marginal dollar invested in pre-K would go much further if spent on expanding access rather than on improving standards that have not been scientifically assessed. As we learn more about the value of pre-K dollars spent in different areas — say, requiring more teacher education or lowering class sizes or expanding access — it is important that the knowledge is shared with policymakers. Here, then, is another place where the federal government could play a valuable role.
Is there hope?
We can’t expect sensible tinkering at the margins to transform an inadequate ECEC system into a good one. Changing incentives, rearranging subsidies and maybe throwing a few billion dollars more into the hopper would get us only a few steps down the road. Such tinkering is still worth doing, of course. But to properly address the big challenges posed by raising little kids would realistically require a much bolder federal effort. And that would require the political will to recognize ECEC for what it is — both child care and an investment in the future — and to jettison the traditions and assumptions that weigh down the effort.