This appendix discusses the existing literature on approaches to determining what share of total early care and education (ECE) costs is reasonable for families at different income levels to pay. It describes the advantages and disadvantages of the four main approaches that the committee found in this literature: no-fee payments, share of income determined by families’ current average ECE expenditures, share of family income after protecting for other necessities, and share of income that minimizes impact on family utilization decisions.
CRITERIA FOR DETERMINING A REASONABLE SHARE FOR FAMILIES TO PAY
The committee identified several important criteria to use to assess different approaches to determining a reasonable share of total ECE costs for families, or in other words, in defining what is affordable for families. These criteria reflect the committee’s view that children’s access to high-quality early care and education should not be constrained by a family’s income, and the committee therefore agreed that an affordability standard should simultaneously promote access and equity. First, for an approach to be considered affordable, it must enable families at all income levels to access high-quality ECE services for their children of all ages from birth to 5 years old. Second, to be equitable, if an approach requires family payment of fees, then the share of income expected to be paid out of pocket for ECE services must increase progressively across income levels, reflecting the
fact that as family income increases, the share of income needed for other necessities decreases.1
The committee noted two additional desirable attributes for a method of determining a reasonable share for families to pay: clarity or transparency and ease or cost of implementation. The basis for determining family payments and assistance levels should be clear and understandable to policy makers and families. Minimizing the complexity of appropriate payment shares can improve the transparency and acceptability of the system and promote uniform application of rules across families and jurisdictions. To the extent practicable, minimizing the cost of implementing a payment share system is another desirable goal, as it uses public and private resources more efficiently. Moreover, an efficient payment share system may also allow administrators to focus on service provision rather than payment management.
As described in Chapter 6, within a given geographic market, the cost of high-quality early care and education on a per-child basis varies by type of care (home based or center based) and age of child as well as the particular service needs of the child (e.g., special needs). However, every child, regardless of family income, should have access to services of equally high quality. Therefore, regardless of the financing mechanism, the process for gaining access to high-quality ECE services must be evaluated based on the aforementioned criteria of affordability (to ensure access for all), equity, transparency, and efficiency.
APPROACHES TO SPECIFYING AN AFFORDABLE SHARE OF COSTS FOR FAMILIES
There is no universally accepted definition of affordability for ECE services or agreement on how it should be measured. Definitions for affordability of housing, health care, and higher education face similar challenges (see, e.g., Harkness and Newman, 2005).2 The committee reviewed four
1 A progressive tax is one that imposes a heavier tax burden, as a percentage of income, on higher-income households than on lower-income households. In the tax literature, progressivity is often justified in terms of promoting equal sacrifice, on the ground that a dollar given up by a higher-income individual requires a smaller sacrifice than a dollar given up by a lower-income individual. This assumption of diminishing satisfaction (which economists call “utility”) as income rises is plausible and clearly implies that higher-income families should pay more taxes than lower-income families. However, unless one knows the specific form of the relationship between income and satisfaction, it is impossible to be specific about the appropriate degree of progressivity or indeed even if tax burdens should rise as a share of income, as income rises. Thus, legal and economic tax experts are careful to note that the degree of progressivity must ultimately be based on value judgements about what is fair (see, e.g., Slemrod, 1996).
2 In presentations to the committee, representatives from the health care, housing, and higher-education fields discussed definitions of affordability in their sectors.
different approaches to determining a reasonable share for families to pay, or in other words, defining an affordability standard for families. These approaches include (1) no-fee approaches, (2) share of income based on equitable cost burden, (3) share of income after protecting for necessities (basic-needs budget approach), and (4) affordability as minimizing impact on utilization decisions (economic modeling approach).
Many complexities arise in defining an affordable share for families to pay, in terms of defining both family income and payments. The federal standard for family payments in the Child Care and Development Block Grant (CCDBG) program is based on gross income, not accounting for tax benefits currently available to middle-income families. Prices of ECE services and availability of tax preferences differ across states; thus a standard based on national averages is likely to be too high (with respect to affordability) in some states and too low in others. Accounting for multiple children of different ages in families also complicates the discussion; in other words, should the affordability standard refer only to payments for children age zero to 5 years (the focus of this report) or to all children in the family? As noted elsewhere in the report, expenditures for care of school-age children are substantial for many families who also have children younger than 5 years old. Dealing effectively with these complexities adds to the challenge of designing a system that is affordable, equitable, transparent, and not costly to administer.
In a few states in the United States, courts have included early education for children of certain ages as part of the right to education protected by the state, as it is for older children in the birth to age 8 years range. Oklahoma and Georgia have established universal prekindergarten programs, some of which are offered at no out-of-pocket costs to parents. Other localities, such as Washington, D.C., and New York City, have also implemented universal prekindergarten programs that do not require parental payments. In some countries, for instance Portugal and the Nordic countries, access to ECE services is defined as a legal right; therefore, demand at both national and local levels must be met and relevant resources provided (Penn, 2017).3
As limited U.S. experiences demonstrate, no-fee approaches eliminate financial barriers to accessing certain ECE programs and ensure access
3 However, most countries in the Organisation for Economic Cooperation and Development (OECD) charge fees for early care and education. According to the OECD Family Database, “On average across OECD countries, the net cost of childcare (for two children aged 2 and 3 in full-time centre-based care) for a two-earner couple family works out to just under 17.5 percent of average earnings, but there is substantial variation across countries (Chart PF3.4.B).” See www.oecd.org/els/family/database.htm [June 2018].
to early care and education, regardless of family circumstances. No-fee approaches can also help to reduce economic insecurity and boost discretionary income of families with young children in areas or groups where poverty is highly concentrated. A no-fee approach may also promote economic integration of children if programs are designed and located to serve diverse groups of children without regard to family income. Such integration has been shown to benefit all children, but if the greater public cost of no-fee programs causes them to be limited to low-income children, the effect is to promote harmful economic segregation.
No-fee systems may also be more transparent and simpler to administer, as they avoid the need for complex fee and copayment schedules, for administrative structures to determine family income and eligibility, and for the ability to complete complex tax return documents.
No-fee approaches also have disadvantages. If a no-fee approach is structured so that no fess are charged to families only up to a certain income level, with a significant fee imposed above that level, a classic “cliff” with work disincentives will emerge (see Chapter 4). In addition, if a no-fee approach is structured so that families at all income levels do not pay for services, higher-income families will receive the same subsidization as lower-income families, yielding a regressive financing structure (unless the revenue sources supporting the spending are sufficiently progressive to offset the subsidies given to upper-income families).
Family Payment Based on Current Average ECE Expenditure as Share of Income
A common approach across the housing, higher education, and health care sectors is to define affordability based on a share of family income. For example, a widely used criterion for affordability of housing costs is that housing should cost not more than 30 percent of income. With respect to health insurance, the Commonwealth Fund Affordability Index identifies “high” premium costs to be 10 percent or more of income (7% for low income), “high” deductibles to be 5 percent of income or more, and “high” out-of-pocket costs to be 10 percent or more of income (5% for low-income families, defined as household income below 200 percent of the federal poverty level) (Collins et al., 2015). Current federal childcare subsidy policy also uses this cost burden approach, indicating that family payments for CCDBG recipients should not exceed 7 percent of income.4 Until recently, federal policy had specified 10 percent of income as a measure of affordable copays, and states may choose to exempt families below the federal poverty
4 The data underlying this standard, and the policies to which it applies, include payments for school-age children as well as children from birth to 5 years.
level from copays.5 Across sectors, however, there is no generally accepted rationale for determining what share of income is appropriate. Because of this uncertainty, the committee considered alternative ways to assess what share of income should be considered as affordable for a family to pay for ECE services.
Given differences in needs and preferences, two families with the same income level may choose to spend their resources differently. Thus, a share-of-income approach is not intended to determine (or assume) that every family will spend the designated percentage of their income on ECE services. Some will choose to spend more, others will want to spend less. To ascertain what an average family would consider affordable, one approach is to examine current levels of ECE expenditures as a basis for what is affordable. This is a market-concept approach, assuming that if families currently pay this amount, it is affordable to them.
One advantage of using current ECE expenditures as the basis for a share-of-family-income affordability standard is its grounding in the economic theory of “revealed preferences.” Asking families in a survey what is affordable is not likely to result in reliable numbers, whereas using data on actual expenditures reveals what families spend when taking into account their preferences for different goods and services. Current federal guidelines for ECE subsidy copays are based on national survey data indicating the share of income that is paid out of pocket by families (about 7%, on average).6 The income share could be proportional (set at the same level for all families) or progressive (where the share of income increases as income level increases). A proportional share of income that is affordable for very-low-income groups will not generate substantial resources for the system and likely would benefit some affluent families who would pay less than they currently pay. In contrast, a progressive approach that increases the required family share for higher income families could promote greater equity because as family income increases, the share of income needed for other necessities decreases.
Setting an affordable share of income based on current expenditures by families provides one approach to defining what is affordable. However, families’ current expenditures on ECE are driven by a number of factors including cost of programs (see Chapter 2), and some families may currently be spending large shares of their income on ECE at the expense of other necessities. Determining how to set a benchmark or affordability standard for a typical family presents a number of challenges. Families differ in both
5Federal Register, vo. 81, no. 190, p. 67440. CCDBG Final Rule. Published September 30, 2016.
6 The 7 percent average is based on all families’ current payments, which includes families who are currently paying zero and includes payments for school-age children.
their needs and their resources, and so even families with the same income level will not necessarily find the same income share to be “affordable.” Taking into account differences in family needs as well as resources could be done but could result in a complicated formula or determination process. A related approach that determines affordability by protecting a share of income for other necessities is described in the next section.
Protecting a Share of Income for Other Necessities (basic-needs budget approach)
The basic notion of affordability of a good or service is measured by its cost relative to what a family can pay, or whether the cost is within the family’s financial means. But the criterion of being “within the family’s financial means” is not sufficiently specific (e.g., the income share could be any number below 100%). A related approach is to establish an affordability standard that accounts for the share of income needed for other basic necessities. Like the cost burden approach, this approach is based on affordability as a share of family income, but in this case, the family contribution is based on income above the amount for necessities. By setting aside a certain amount of income for necessities, this standard ensures ECE access for low- and moderate-income families. A share of the remaining “discretionary” income is charged as family payment for early care and education, with the remainder of ECE cost covered by public subsidy. For example, Helburn and Bergmann (2002) proposed setting aside income equivalent to twice the federal poverty level, which has been shown by a number of analyses to be the amount required for a basic standard of living with assistance. However, determination of a basic-needs budget typically includes ECE costs, so in developing this approach one would want to adjust to avoid double counting.
As noted above, one critique of basing an affordability standard on a share of family income is that families differ in their need for other necessities. By setting aside a basic income level for necessities, affordability is implicitly defined as a level of expenditures that does not impinge on the family’s ability to purchase other necessities. Family contribution to higher-education costs are based on a similar concept, accounting for both family income and family needs (such as having more than one student attending higher education at the same time).
One key advantage of this method is the clear conceptual basis for the set-aside and the potential for accounting for some differences across families in terms of needs. For instance, the federal poverty level varies by number of children in the household, although neither their ages nor disability status are factors.
One critique of this approach is that setting aside the full cost of meeting other basic needs effectively exempts families from having to make
tradeoffs among different goods and services. Within their budgets, families may spend more on other goods or services that may benefit children, such as health or housing, or activities that families value and can be enriching for children and contribute to family well-being, such as travel. The first method of setting an affordability standard, based on current income share spent on ECE, reflects the tradeoffs families make among different goods and services and the value they place on them. But it also results in a cliff at the set-aside amount, as discussed above.
The basic-needs set-aside approach does not necessarily provide equitable access, if the share of income above basic needs is fixed at the same level for all income groups. In addition, the federal poverty level is set nationally, yet costs of living vary considerably across locations. The National Center for Children in Poverty estimates a basic-needs budget (including ECE costs) ranging from 175 percent to 327 percent of the federal poverty level, depending on location and family structure.7 However, this approach to an affordability standard could be modified to allow for differences in costs of living in different locations.
Implementation of this approach to an affordability requirement would require first, determining what is the level of basic income to set aside (which might vary by location and family structure), and then determining the share of the remaining income to designate as affordable for families (which likely would need to vary by income level to ensure access for moderate-income families). While there are estimates of basic-needs budgets, the determination of these numbers is fundamentally a policy decision. Keeping in mind the criteria identified by the committee in the introduction to this appendix, this approach adds complexity although it may improve access by taking into account variation in families’ needs.
Minimizing Impact on Utilization Decisions (economic demand modeling approach)
A fourth approach to defining an affordability standard is to use economic analysis and data on families’ use of ECE services to extract information about what families would be willing to pay. The economists’ concept of “willingness to pay” refers to the maximum amount someone is willing to pay for a good or service, given their income level. However, to most people, an “affordable” amount is less than the maximum amount one would pay. In economics, a demand curve indicates how much of a good or service consumers will purchase at different prices, given their incomes, prices of other goods and services, and other factors. An economic model of demand for high-quality ECE services could be used to estimate the degree
to which the net current price charged to families at different income levels affects their decisions regarding the type and hours of ECE service utilized. Estimates of the degree of price responsiveness (called “elasticiticies” by economists) could be used to determine how changes in the amount families pay would affect their utilization decisions.
The economic demand modeling approach to setting an affordability standard would directly address the objective of making high-quality early care and education “affordable” by determining what families would pay for high-quality ECE services while continuing to use the same or greater level of those services. This approach could differentiate among different income groups (promoting the goal of equity) and adjust for costs of care for children of different ages. By estimating families’ responsiveness to prices, this approach reflects families’ preferences and tradeoffs, including their spending on early care and education and other goods and services.
If higher costs lead to higher prices for families, economic demand modeling can provide information about how families are likely to respond. If price is increased, families likely will reduce the number of hours of ECE services they use, or they may switch from higher- to lower-price (and potentially lower-quality) providers, such as from center-based to home-based ECE providers. The current pattern of lower use of center-based care by middle-income families than either higher-income families or lower-income families (who have more access to subsidies and free public programs) provides evidence of how families respond to prices (see Chapter 2 for details). Blau (2001) estimated that a 10 percent increase in the price of center-based care (holding other prices constant) would decrease the use of centers by 2.4 percent. If prices of all types of care increased 10 percent, he estimated a drop of about 3 percent in use of paid child care. These estimates indicate that families will use more paid care and more center-based care than currently used if the amount they would pay out-of-pocket decreases.
Additional evidence of the response of parents to more affordable early care and education comes from studies that have demonstrated that families with ECE subsidies use more center-based care, and higher-quality care, than those without subsidies (Berger and Black, 1992; Davis, Krafft, and Forry, 2017; Johnson, Ryan, and Brooks-Gunn, 2012; Marshall et al., 2013; Ryan et al., 2011). Parents’ rate and hours of employment also respond to the price of ECE services; if price increases, some may remain outside the labor market entirely. In a review of the literature, Morrissey (2017) concluded that mothers’ employment would decline 0.5 to 2.5 percent if ECE costs (to families) increased by 10 percent.
Current estimates of families’ out-of-pocket expenditures on early care and education and their price responsiveness reflect current market conditions, in which the quality of early care and education is predominantly mediocre. It is possible that families would have higher rates of utilization
or would pay more for early care and education that was of higher quality than they currently find available. There is limited research measuring families’ willingness (and ability) to pay for high-quality early care and education. Studies by Blau (2001) and Blau and Hagy (1998) concluded that families are not willing to pay (much) higher prices for higher-quality care. Blau (2001, p. 113) notes that “consumer willingness to pay for higher quality is . . . weak on average . . . and highly variable across markets.” Both studies noted that the measures of quality used in the analysis are limited and may not be closely tied to quality of care valued by parents. In addition, a weak relationship between price and quality could be due to a lack of information about quality; that is, whether parents can determine the relative quality of ECE options (Cryer and Burchinal, 1997; Mocan, 2007). However, the relationships among price, quality, and ECE utilization estimated in these studies may be less relevant today and in the future, given the changes in the ECE landscape over the past two decades. In particular, the introduction of quality rating systems may give parents more information about quality. If higher-quality ECE services are available and identifiable, some families may be willing to spend more than they currently do, and they may use more ECE services.
The economic modeling approach would also face challenges because the required analysis is complex and could be difficult for policy makers and stakeholders to understand fully. There are multiple factors in addition to the price families pay that affect their utilization decisions (e.g., availability of family caregivers, work schedules, cultural preferences, urban/rural location, and number of nearby facilities). For the economic demand modeling approach to take these factors into account, the data requirements would be substantial. Whether family payments would vary by these factors would need to be determined, although federal and state policies do reflect some differences across families, such as differences in copay by family size. Similarly, if price responsiveness varies by state, that would open the question of whether family payments or the affordability standard ought to vary across states. As with any of the methods that use data to set an affordability standard, there would be a need to update over time, and the added complexity might increase the cost of updating.
To summarize, economic modeling of family demand for high-quality ECE would provide important information about families’ preferences and responsiveness to prices and quality. This information could be helpful in determining a level (or levels) of family contribution that does not reduce utilization of high-quality early care and education. Further exploration of the economic demand modeling approach is beyond the scope of this study but may be warranted for its potential to inform the phases needed in the transition to the envisioned new ECE system.
The financial burden on parents affects their decisions about using ECE services, including the amount, type, and quality of ECE service they use. While parents may contribute some portion to the costs of early care and education, relying solely on parents to shoulder the burden of higher costs of higher-quality early care and education would likely lead to reductions in the use of higher-quality ECE options and less support for children’s early learning and development. Yet determining what level of expenditures is affordable to families is challenging for a number of reasons. First, there is no universally accepted measure or standard of ECE affordability. In addition, the share of income families spend on early care and education varies with their resources, needs, and preferences.
Declaring that a specific share of income is “affordable” does not imply or assume that every family will be willing to spend that percentage of their income on early care and education. Some will want to spend more; others will want to spend less. Changes in out-of-pocket costs to families will alter the size of the contributions from families and from the public sector, but such changes will also affect families’ decisions with regards to how much and what kind of early care and education to use.