Child Care Cost and Quality

Child Care Cost and QualitySuzanne W. HelburnCarollee Howes

Abstract

This article summarizes what is known about the cost and quality of full-time child carein centers and family child care homes, and about parents’ attention to quality in mak-ing child care choices. It relies primarily upon two recent studies which are among thefirst to collect detailed information about child care operating costs: the Cost, Quality,and Child Outcomes in Child Care Centers study1 and the Economics of Family ChildCare study.2,3 Results indicate that mediocre quality is the rule and that parents oftendo not choose quality settings for their children. At the present time, child care qual-ity is only modestly related to the cost of providing services. In part, the modesty of thisrelationship reflects the low wages of child care staff, the availability of in-kind dona-tions in the nonprofit sector, and the altruistic motivations of many providers thatdepress direct costs and the fees charged for child care. The article concludes with rec-ommendations for future action: (1) launch consumer education efforts; (2) imple-ment higher standards for child care at the state level; (3) avoid public policies thatencourage people to become child care providers if they have no interest in such acareer; (4) increase public and private investments in child care; and (5) develop themeans to compensate child care workers as is appropriate for their levels of training,experience, and responsibility.

While there is an extensive literature on child care quality devel-oped from the perspectives of psychology and education, thestudy of the cost of child care by economists, and especially the

relation between cost and quality of services, is fairly limited. This article, acollaboration between an economist and a psychologist, examines what isknown about child care cost and quality, and the linkages between the two.

The quality of child care is important because it is closely linked withchildren’s social, cognitive, and language development.4 Children in high-quality early childhood programs are more likely to be emotionally secureand self-confident, proficient in language use, able to regulate impulsiveand aggressive inclinations, and advanced in cognitive development.5 Overtime, these children may experience enhanced school achievement, higherearnings, and decreased involvement with the criminal justice system.6 Incontrast, children who experience poor-quality child care are at risk for

6262

The Future of Children FINANCING CHILD CARE Vol. 6 • No. 2 – Summer/Fall 1996

Suzanne W. Helburn,Ph.D., is professor emer-ita of economics at theUniversity of Coloradoat Denver.

Carollee Howes, Ph.D.,is a professor of educa-tion at the GraduateSchool of Education atthe University of Cali-fornia, Los Angeles.

63

poor long-term developmental outcomes, including apathy, poor schoolskills, and heightened aggression.5

Knowledge of the cost-quality relationship in child care matters to bothpolicymakers and the participants in the child care market. It is importantin designing policy concerning the financing of child care to know howmuch good care costs, which factors determine costs, and how likely it is thatthese factors will persist into the future. It is of equal importance to con-sumers who want to make sure that they are receiving the quality of carethey believe they are paying for.

Generally, policymakers, consumers, and economists expect that higher-quality services should cost more than lower-quality services. This is par-ticularly so in a labor-intensive industry such as child care, where the qual-ity of staff is critical to good services. Wages for better-qualified staffmembers are expected to be higher than wages for less-qualified staff mem-bers, and higher wages will drive upward the costs of providing child careand the fees charged for services. On the other hand, competitive pressureswithin the child care market should keep costs, fees, and profits low, whichmight be expected to depress quality. Consumer demand also plays a role:if parents and third-party payers demand quality child care, providers arelikely to supply it, and costs and fees will increase. In other words, economictheory and consumer intuition suggest that the cost and quality of childcare should rise or fall together and that competition among providers andconsumer demand will provide important checks and balances on the lev-els of each.

If these expectations about staff wages, competition, or consumerdemand in child care do not hold, however, then traditional economicassumptions about the relationship between cost and quality may fail. Forexample, if child care staff members are so committed to caring for childrenthat they are willing to work for less-than-market wages, then child care costsmay be low, no matter the quality of care. In-kind donations can also keepcosts low, no matter the quality. If parents and purchasers of care do not rec-ognize and demand good-quality child care, then no incentive exists formany providers to offer it, no matter what it costs. Thus, if expectations fail,then even a highly competitive child care market will not assure provision ofgood-quality services at the lowest possible prices. These complexities makeit imperative to examine how the child care market operates in order toinform public policy concerning child care financing.

This article relies primarily upon two recent studies, the Cost, Quality,and Child Outcomes in Child Care Centers study1 and the Economics of

64 THE FUTURE OF CHILDREN – SUMMER/FALL 1996

Child Care QualityFor 20 years, developmental psychologistshave investigated the impact of child careattendance on children’s development.Initially, researchers sought to determine ifcare by someone other than a parent harmschildren, and they found that children’slong- and short-term development does notappear to be harmed by child care.5However, this research also suggests that thequality of child care varies widely and thatchildren’s outcomes differ depending uponthe quality of their child care experiences.1,7

There is today considerable agreementamong developmental researchers on theaspects of child care quality that are directlylinked to positive outcomes for children.Quality child care is primarily defined bytwo highly interrelated components: processquality, which captures the experiences chil-dren have in child care, and structural quali-ty, which captures the aspects of the child

care environment that are often regulatedby government.1,2,8,9 In addition, researchindicates that the adult work environmentindirectly influences the children in carebecause it is closely tied to both caregiverbehavior and to caregiver longevity in childcare.9,10 Table 1 describes these aspects ofchild care quality and how they are mea-sured in most studies, including the studiesreported in this article.

Process QualityProcess quality refers primarily to how chil-dren experience child care—their interac-tions with the adults who care for them and

their exposure to materials and activities thatenhance learning. Process quality is consid-ered basic to child care quality because it ismost directly related to children’s behaviorin the child care environment.2,11 Qualitychild care means that caregivers respond tochildren’s social behaviors in a sensitive andpositive fashion, are involved in their playand learning activities, and are not harsh intheir management of children’s behavior.Children in high-quality care spend theirchild care hours in socially appropriate playwith adults and peers,5 and they explorematerials in ways that fit their age and devel-opmental stage.12 Children in such settingshave higher scores on cognitive, social, andlanguage measures of development.1

Structural QualityStructural quality refers to objective aspectsof the child care environment that are oftenregulated by government, such as the num-ber of children cared for in a group and byeach caregiver (adult-child ratio), caregiverformal education and specialized training inchild development or related fields, caregiv-er experience, and aspects of the facilitythat houses the child care programs, suchas the amount of floor space per child.Optimal structural conditions create oppor-tunities for more favorable process condi-tions to occur in child care settings, andthese, in turn, lead to more positive childoutcomes.11 For example, when an adult isresponsible for providing care to a smallergroup of children, she is more likely to tailorcare to the particular needs of eachchild.2,5,13 Adults who have more formal edu-cation and specialized training tend to usethis knowledge when they interact with chil-dren, and this enables them to be more sen-sitive and responsive in their caregiv-ing.1,2,5,9,13 Children cared for under theseconditions have more positive developmen-tal outcomes, that is, they are more likelythan children cared for under less optimal

The quality of child care varies widely andchildren’s outcomes differ depending uponthe quality of their child care experiences.

Family Child Care study,2,3 to describe the relationship between cost andquality in center-based and family child care programs. The first section ofthe article describes how quality child care is defined and measured, sum-marizes the quality of child care in the United States today, and examinesparents’ understanding of child care quality. The next sections of the articledescribe child care operating costs, revenues, and fees, and their relation-ship to quality. The article concludes with recommendations for policy andpractical action.

65Child Care Cost and Quality

Table 1

Definition and Measurement of Quality in Child Care Programs

Source: Harms, T., and Clifford, R.M. Early childhood environment rating scale. New York: Teachers College Press, 1980; Palacios, J., and Lera, N.Observation in preschool. Unpublished rating scale, 1991; Harms, T., Cryer, D., and Clifford, R.M. Infant/toddler environment rating scale. New York:Teachers College Press, 1990; Howes, C., and Stewart, P. Child’s play with adults, toys and peers: An examination of family child care influences.Developmental Psychology (1987) 23:423–30; Stipek, D., Daniels, D., Galuzzo, D., and Milburn, S. Characterizing early childhood education programsfor poor and middle-class children. Early Childhood Research Quarterly (1992) 7:1–19; Harms, T., and Clifford, R.M. Family day care rating scale. NewYork: Teachers College Press, 1989.

Areas of ComparisonWhat Was Measured

and How Adult WorkProcess Quality Structural Quality Environment Quality

Aspects of Child Care Interactions between Group size Child care director’sProgram child and caregiver, and caregiver’s salary

including caregiver Adult-child ratio and benefitssensitivity, harshness,detachment, and Caregiver’s previous Annual turnover ratesinvolvement with experience in caring of caregivers and children for children directors

Caregiver’s attitudes Caregiver’s formal Caregiver’s work toward children education satisfaction

Presence of learning Caregiver’s special- Caregiver’s work activities ized training in child commitment

careHealth and safety Caregiver’s perceptionaspects of the child of job stresscare environment

Presence of appropri-ate furnishings, equip-ment, and curricularmaterials

Techniques Used to Observations of Observation Interviews with orMeasure Quality children, caregivers, questionnaires to child

and child care Questionnaires care program directorsenvironment or teachers

Interviews

Review of personnelrecords

Instruments Used to Early Childhood Observations of Staff questionnaireMeasure Quality Environment Rating activities in preschool

Scale (ECERS) (for group sizes and ratios)

Infant/Toddler Environment RatingScale (ITERS)

Caregiver InteractionScale

UCLA Early ChildhoodObservation Form

Family Day Care Rating Scale


structural quality conditions to be develop-mentally on target. In addition, adults withmore formal education and training(rather than just more experience work-ing with children) are more likely to pro-vide age-appropriate learning experiencesfor children.2,3,9,13

While, in theory, child care programscould score well on a measure of one aspectof structural quality and score poorly onanother, in practice in the United States,elements of structural quality are highlyassociated with one another. For example,in theory, a very well-educated and well-trained staff person might be able to man-age a large group of children more effec-tively than a less-qualified staff person, butthis pattern is seldom found in the UnitedStates. In most instances, child care pro-grams that hire well-trained staff membersalso tend to have small groups and highadult-child ratios.

Because children of different ages needdifferent levels of care and degrees of super-vision, recommended group sizes and adult-child ratios differ by age group. For instance,quality programs for infants and toddlers(up to age two) generally have smaller groupsizes and a higher adult-child ratio than pro-

grams serving preschool-age children,meaning that more staff members areneeded (see Box 1). As will be discussedbelow, personnel costs are the singlelargest expense in a child care program’sbudget. Therefore, altering quality featuressuch as staffing ratios can have an enormouseffect on a child care program’s cost andprofitability.

The Adult Work EnvironmentSalary, benefits, and working conditionsaffect caregiver behavior and longevity inchild care, and so indirectly they influencechildren’s experience in the care setting.5,9

For example, providers who are committedto their jobs, satisfied, and compensated ade-quately are more sensitive to the children,

more responsively involved, and more nur-turant.2 Centers that offer higher wages andbetter benefits can attract and retain a better-trained staff. Caregivers who earn more areless likely to leave their center jobs9,14 or toclose their family child care homes.15 Staffturnover is an important indicator of poor-quality care because children who lose theirregular caregivers tend to experience nega-tive outcomes such as poor language andsocial development9 and, in at least onestudy, increased aggression.16

The Quality of U.S. Child CareTodayA consistent and dismaying profile of thequality of child care in America emergesfrom two recent, in-depth, multisite observa-tional studies on child care centers and fam-ily child care homes,1,2 as well as from earlierstudies.11,15 (The article by Hofferth in thisjournal issue provides additional data ontrends in child care quality based on largenational surveys.)

The Cost, Quality and Child Outcomesin Child Care Centers (CQO) study, con-ducted in 1993–94, examined 401 child carecenters representing 749 classrooms inCalifornia, Colorado, Connecticut, andNorth Carolina.1 Classroom observationsand interviews with center directors and staffmembers provided data on center charac-teristics, program quality, and staff qualifica-tions, commitment, and compensationissues. About 826 children were also individ-ually assessed to learn the effects of programcharacteristics on children’s development.

In 1991–92, the Study of Children inFamily Child Care and Relative Care exam-ined care provided by 226 family child careproviders in communities in California,North Carolina, and Texas. The study alsoinvolved 820 mothers and 225 of their chil-dren.2 Observations and interviews withmothers and providers yielded data on fam-ily child care home characteristics, programquality, and other issues.

The Quality of Center-Based CareThe CQO study indicated that about 86% ofthe centers in the study provided mediocreor poor-quality services. Only 14% of thecenters surveyed met levels of process quali-ty that were high enough to support chil-dren’s development. Twelve percent were

Child care programs that hire well-trainedstaff members also tend to have small groupsand high adult-child ratios.


Accreditation of Child Care Programs

Several voluntary systems exist nationally to establish higher-quality standards than arerequired by law for both child care centers and family child care homes.

CHILD CARE CENTERSOne of the most widespread accreditation systems for centers is that created by the NationalAssociation for the Education of Young Children (NAEYC). In this system, center-based pro-grams conduct self-evaluations involving staff and parents. Professional validators fromNAEYC conduct visits to determine whether or not standards have been met, and if theyhave, programs are accredited for three years. Standards are designed for programs thatserve children from infancy through age eight in centers caring for 10 or more children;school-age programs are eligible if a majority of children are eight years old or younger. Asof January 31, 1996, there were 4,523 accredited centers in the United States and 16 foreigncountries, and another 8,815 centers were in the process of becoming accredited.

Initially developed in 1984, the NAEYC criteria for accreditation were based on researchand professional consensus. The criteria include guidelines for staff-child interactions; cur-riculum content; parental involvement; staff qualifications and training; administration;staffing patterns (group size and adult-child ratios); physical environment; health and safe-ty; and nutrition and food service. As an example, NAEYC standards for group size andadult-child ratio are as follows:

Age of Children Number of Children per Group Adult-Child Ratio0 to 12 months 6 to 8 1:3 to 1:412 to 24 months 6 to 12 1:3 to 1:42 years 8 to 12 1:4 to 1:63 years 14 to 20 1:7 to 1:104 to 5 years 16 to 20 1:8 to 1:106 to 8 years 20 to 24 1:10 to 1:129 to 12 years 24 to 28 1:12 to 1:14

FAMILY CHILD CARE HOMESIn 1988, the National Association for Family Day Care (now the National Association forFamily Child Care, or NAFCC) began a program of voluntary accreditation for family childcare homes. The process includes self-evaluation as well as external validation of aspects ofprogram operation, including health and safety, nutrition, indoor and outdoor play envi-ronments, interactions, and professional responsibility. Accreditation of family child carehomes is less common than accreditation of child care centers. As of early 1996, only 1,083providers were accredited, and another 272 were in the process of becoming accredited. Allbut six states had accredited providers. These six states were Louisiana, Mississippi, Nevada,New Mexico, South Carolina, and South Dakota.

The National Association for Family Child Care is working with the Family Child CareProject at Wheelock College, Boston, MA, and a national network of providers, parents, andothers to develop a new accreditation process and instrument, scheduled to replace the cur-rent system in 1998.

Selected References: Hayes, C.D., Palmer, J.L., and Zaslow, M.J., eds. Who cares for America’s children? Child care policy for the 1990s.Washington, DC: National Academy Press, 1987; National Association for the Education of Young Children. Accreditation criteria and pro-cedures. Rev. ed. Position statement of the National Academy of Early Childhood Programs, a division of the National Association for theEducation of Young Children. Washington, DC: NAEYC, 1991.

Box 1


judged to be of such poor quality that chil-dren’s basic health and safety needs wereonly partly met and few learning experi-ences were provided. Quality in the rest(74%) of the centers was judged mediocre.Care for infants and toddlers was particular-ly poor. Only 8% of classrooms servinginfants were rated high quality; fully 40%were judged low quality. At least half theinfant and toddler classrooms observed hadpoor general health practices.

Results for structural quality in theCQO study revealed that 28% of the teach-ing staff had college degrees, 46% hadsome college course work, and 26% hadonly a high school education or less. Not allof this training was in the child care field:of those staff members with a high schooleducation or less, for example, 25% hadnever had any training in early childhoodeducation.

The average group size was 8 children forinfants and toddlers, and 14 for preschool-ers, within acceptable limits recommendedby the National Association for the

Education of Young Children (NAEYC), thepremier professional organization for staffin early childhood programs. The averageratio was one adult for every six preschool-ers, and one adult for every three or fourinfants—again within acceptable limits rec-ommended by NAEYC.

Levels of staff education, group size, andratios all varied across the four states in thestudy, perhaps reflecting the variation inchild care regulations across the states. Forexample, the state that had no requirementfor staff experience in its regulations (NorthCarolina) had the lowest average level ofteachers’ prior experience.

Some centers met additional standardsbeyond those required for state licensing,and these tended to provide higher-qualityservices. For example, centers that wereaccredited by the NAEYC were of higher

quality than centers that were not accredited(see Box 1).

The CQO study also assessed the adultwork environment. Data on staff wages arediscussed in the section about the costs ofchild care programs, but staff turnover(which is related to wages) is discussed here.Annual staff turnover rates were 37%, con-sistent with rates in multisite child care stud-ies conducted in 19889 and 1992.14 This isalmost four times the annual turnover rateof 9.6% reported by all U.S. companies in1992, and it is more than double the 12%turnover rate reported by employers such asgovernment, school, and other nonprofitorganizations.14

Finally, the CQO study indicated no over-all differences in quality between nonprofitand for-profit centers except in NorthCarolina, where state licensing standardspermitted low staff-child ratios and low stafftraining requirements. In earlier studies,care in nonprofit centers has sometimesbeen rated as being of higher quality thancare in for-profit centers,9,13 but the CQOstudy found instead that quality varied morewithin nonprofit and for-profit sectors thanbetween them. Church-affiliated nonprofitcenters were of lower quality than other non-profits. In addition, for-profit centers thatwere part of national systems seemed to haverelatively higher quality than other for-profitcenters. These differences in quality amongcenters were associated with differences insources of revenues. Centers dependentupon parent fees as their primary source ofrevenues tended to provide lower qualitycare. Centers with additional revenuesources that were designed to enhance qual-ity did indeed offer higher-quality services.These differences suggest that financingmechanisms that rely solely on parent feesare unlikely to lead to improvements in thequality of child care.

The Quality of Family Child Care Like care in centers, care in family child carehomes varies greatly in quality, but overall, ismediocre on measures of process quality.The 1991–92 Study of Children in FamilyChild Care and Relative Care rated only 9%of homes surveyed as good quality; 56% pro-vided care considered adequate or custodial;and a full 35% were judged to be inade-quate.2 The average provider was rated as

Financing mechanisms that rely solely on parent fees are unlikely to lead toimprovements in the quality of child care.


nonresponsive or inappropriate in interac-tions with the children close to half the time.

With respect to structural quality mea-sures, family child care providers had rela-tively low levels of formal education andexperience in child care. Only 17% of thosesurveyed had college degrees, and 36% hada high school education or less. Mostproviders (98%) did have some specializedtraining in child care, often through work-shops or child care conferences outside theformal education system. As was true incenter-based care, providers with more for-mal education were rated as more sensitiveand less detached, and were observed to bemore responsive in their interactions withchildren in their care.

The groups in family child care averagedfour children but were larger for toddlersand preschoolers than for infants. Fully 36%of homes had at least one infant present.The typical family child care home had alone caregiver and a staff-child ratio of aboutone to three.

Results suggest that family child careproviders who had chosen child care as acareer tended to provide better-quality carethan did those who viewed it as interim workwhile their own children were young or whofelt obliged to provide care for a relative.The professionally oriented providers tend-ed to care for larger groups of children (forexample, three to six children rather than

one or two), to have slightly lower staff-childratios, and to be regulated—all factors asso-ciated in the study with better processquality scores.

In addition, the study gathered data onthe family child care work environment.Providers who were better compensated andreported higher levels of job commitmentwere more sensitive toward the children intheir care. Overall, however, family childcare providers remain in the field for a rela-tively short time. The typical provider hadonly one year of experience as a provider;half had experience of three years or less.

Do Parents Choose ChildCare on the Basis ofQuality?Parents want the best for their children, andmost undoubtedly try to choose a high-quality child care arrangement over amediocre or poor one. Several research stud-ies suggest, however, that parents andresearchers sometimes have conflictingviews about what aspects of a child care pro-gram are associated with quality. Some stud-ies suggest that parents place cost and con-venience above warm interactions withcaregivers,15,17 while others find that parentsendorse the importance of the nature of theinteractions between the child and the care-giver (process quality).18–21 Parents seem topay little attention to the structural indica-tors of quality such as regulatory status and

caregiver training. For example, an Illinoisstudy indicated that more than 40% of thelow-income parents surveyed did not consid-er it important if their child’s provider werelicensed,22 and the Study of Family ChildCare and Relative Care found that, whenasked about their reasons for selecting theircurrent provider, mothers almost nevermentioned formal credentials such as edu-cation and licensing.

In the CQO study,1 parents andresearchers rated the quality of child careprovided for the children. While parentssaid they valued the process characteristicsmeasured by the researchers, they consis-tently overestimated the quality of care theirown children received. That is, parents andresearchers agreed about what was impor-tant in a child care setting, but the twogroups saw the same settings differently:parents perceived care as high in qualitywhile, on average, researchers rated thequality as mediocre. These differencesbetween parent and observer qualityscores were greater for aspects of care thatparents were unable to observe, such asnap time, or for aspects that they valuedmore highly.

These findings—general agreementbetween researchers and parents on theaspects of care that are important but incon-sistency in ratings of care—suggest that par-ents are not well-informed consumers anddo not accurately judge child care quality.This has implications for child care financ-ing because parents are usually the only per-

sons responsible for selecting a child carearrangement. These study findings suggestthat it cannot be assumed that parents willpurchase high-quality child care simplybecause they want the best for their chil-dren. As is discussed below, high-quality carecosts only a little more than mediocre care,so parents cannot use the cost of care as anindicator of its quality. Even parents who arewilling to pay more for child care have no

assurance that their extra dollars will pur-chase higher-quality care.

Child Care OperatingCosts, Revenues, and FeesIt is important to understand costs, rev-enues, and fees in order to understand com-petition and profitability in the child caremarket. How family child care homes andfor-profit and nonprofit centers derive andspend their revenues has important implica-tions for whether or not particular sectors inthe market can or do enjoy any competitiveadvantages. And, of course, an understand-ing of these issues is necessary before therelationship between costs and quality canbe understood.

Only a few researchers have examinedthe operating costs of child care centers23–26

or of family child care homes despite theimportance of these costs.3 Data from theCQO study and the Economics of FamilyChild Care (EFCC)27 study summarizedbelow indicate that (1) parent fees comprisethe largest single source of revenues forchild care programs; (2) personnel costs arethe largest single cost in child care pro-grams; (3) the cost structures in family childcare homes and centers differ, but the feesand costs to parents are similar across thetwo settings for preschoolers, while infantfees are lower in family child care homes;(4) differences exist in revenue sources andcost allocation across for-profit and non-profit centers; (5) parents benefit from hid-den subsidies that hold down the costs ofchild care.

Defining and Measuring theCost of Child CareThe cost of child care can be viewed fromthe perspective of the provider, the parent,and society. From a societal perspective, it isimportant to estimate the value of allresources used in providing child care. Childcare is provided in a mixed market of for-profit, nonprofit, and publicly operated cen-ters, and in licensed and unlicensed childcare homes. Tax benefits, philanthropic con-tributions, and government subsidies reducecosts and fees paid by some parents so thatfees may not reflect the actual cost of pro-viding the care. (See the article by Stoneyand Greenberg in this journal issue for adescription of current subsidy programs.)


Even parents who are willing to pay morefor child care have no assurance that theirextra dollars will purchase higher-qualitycare.


Subsidies can be classified as either sup-ply subsidies or demand subsidies. Supplysubsidies—donations, forgone earnings ofstaff, the U.S. Department of Agriculture’sChild and Adult Care Food Program—arecontributions that reduce the expendedcosts incurred by centers and family childcare providers and, thereby, indirectly affectfees charged by lowering expenses. Becauseconsumers do not know about theirproviders’ finances, they may be unaware ofthe supply subsidies from which they proba-bly derive considerable benefits throughlower fees.

Demand subsidies directly affect what par-ents pay for services. Demand subsidiesinclude direct payments for specific chil-dren, such as state or federal reimbursementto centers, or for vouchers, United Way lumpsum payments for low-income children, andthe Child and Dependent Care Tax Credit.By reducing the fees for services to parents,demand subsidies can affect the level ofquality or the amount of service they are will-ing to purchase.

Table 2 gives estimates from the CQOstudy of the costs of child care, breakingcosts into three categories: (1) Expendedcenter costs, which are the cash costsincurred by providers to produce services.These equal labor, occupancy, food, otheroperating costs, and overhead; (2) cost within-kind donations, which is an estimate ofsociety’s cost, including donations fromindividuals and agencies for facilities, volun-teer services, equipment, materials, food,and so on; and (3) the full cost of care,which is a more inclusive measure of thevalue of all resources used to provide childcare services, made up of expended centercosts plus donations plus the forgone earn-ings of child care staff.

Some of these subsidies (for example,donations and forgone earnings of childcare providers) are hard to estimate. Forinstance, donated goods and services maynot have a market price. Where possible, theCQO researchers attempted to evaluate themarket value of donations to centers, such aslabor, space, and utilities.

The CQO researchers also contend thatthe low wages of providers constitute supplysubsidies—in effect, they comprise a labor

donation, as providers are forgoing thehigher wages they could presumably earn inanother occupation based on their educa-tion, gender, age, and other characteristics.A considerable amount of economic litera-ture argues that, because nonprofit workersvalue the social benefits provided by theirwork, they are willing to donate labor tononprofits,28–33 and one author has suggest-ed that child care workers do this whetherthey work in nonprofit or for-profit childcare programs. The CQO researchers con-tend that, like any donation, estimates ofthe forgone earnings of providers should,therefore, be included in calculating thefull cost of providing child care. They calcu-lated forgone wages by estimating the wagea provider could earn in another occupa-tion, based on demographic characteristics(the provider’s forecast wage), and subtract-ing what the provider actually earned inchild care.34

Measuring the magnitude of forgonewages is subject to error. Unmeasuredcharacteristics of the worker (such as nega-tive personality traits) might explain theworker’s lower wages and mean that anyparticular child care worker actually could

not earn more in a different job. Or, ifchild care employees accept low wagesbecause they get nonmonetary rewards(such as enjoyment from working withchildren) from their jobs, then their lowwages are a function of worker prefer-ences, and one could argue that no for-gone wages exist.

Undoubtedly, enjoyment of the workdoes motivate many of the women whoprovide child care for relatively low wages.For instance, in the CQO study, one of themain reasons center staff members citedfor working in the child care field was theopportunity to work with children. How-ever, the high staff turnover rates in childcare described above suggest that an initialwillingness to work for low wages may beeventually outweighed by negative aspects

As current child care workers leave theirjobs, they are being replaced with a less well-educated work force.


Sou

rce

: He

lbu

rn, S

., e

d. C

ost

, qu

alit

y, a

nd

ch

ild o

utc

om

es

in c

hild

ca

re c

en

ters

: Te

ch

nic

al r

ep

ort

. De

nve

r, C

O: D

ep

art

me

nt

of

Eco

no

mic

s, C

en

ter

for

Re

sea

rch

on

Ec

on

om

ics

an

d S

oc

ial P

olic

y, U

niv

ers

ity o

f C

olo

rad

o, 1

995.

Tab

le 2

Mo

nthl

y C

ost

s fo

r Ful

l-Ti

me

Chi

ld in

No

npro

fit a

nd F

or-

Pro

fit C

ent

ers

, by

Sta

te

Mea

ns b

y St

ate

and

Sec

tor

Ca

lifor

nia

Col

ora

do

Con

nect

icut

Nor

th C

aro

lina

Perc

enta

ge

Non

pro

fitFo

r-Pr

ofit

Non

pro

fitFo

r-Pr

ofit

Non

pro

fitFo

r-Pr

ofit

Non

pro

fitFo

r-Pr

ofit

Dol

lar T

ota

lof

Ful

l Cos

t

Cos

t Est

ima

tes

Full

co

st o

f ca

re$

548

$ 50

9$

551

$ 53

2$

788

$ 66

8$

420

$ 32

3$

548

100

%Le

ss s

taff

forg

on

e e

arn

ing

s10

699

135

143

8511

889

7010

619

Co

st w

ith in

-kin

d d

on

atio

ns

449

408

433

401

708

551

331

253

442

82Le

ss in

-kin

d d

on

atio

ns

5810

458

9350

454

397

Exp

en

de

d c

en

ter c

ost

sa39

140

138

939

361

550

128

625

240

374

Tota

l re

ven

ue$

404

$ 42

6$

402

$ 40

1$

627

$ 51

7$

296

$ 27

7$

420

77 %

Less

ca

sh c

on

trib

utio

ns

163

1417

864

283

6212

854

118

22

Ce

nte

r re

ven

ue fr

om

242

412

224

337

344

455

168

223

302

55p

are

nt

fee

pa

yme

nts

b

Less

inc

om

e t

ax

cre

dit

3346

2939

3352

2731

377

Ne

t p

are

nt

co

st20

936

619

529

831

240

214

119

226

548

Ma

xim

um M

onth

ly F

ees

F-te

stsc

Infa

nts

$ 51

7$

562

$ 46

4$

474

$ 63

0$

632

$ 28

9$

310

$ 45

4C

T>C

A>C

O>N

CPr

esc

ho

ol

368

404

338

357

487

492

248

279

372

CT>

CA

>CO

>NC

P >

NP

aEl

em

en

ts m

ay

no

t a

dd

up

to

to

tals

an

d p

erc

en

tag

es

ma

y d

iffe

r sli

gh

tly f

rom

ac

tua

l be

ca

use

of

rou

nd

ing

.

bTh

is e

stim

ate

is g

rea

ter t

ha

n t

he

ma

xim

um

cre

dit

for a

n in

div

idu

al f

am

ily o

f 20

% x

$2,

400

or $

40 b

ec

au

se f

ull-

time

eq

uiv

ale

nt

en

rollm

en

t in

clu

de

s p

art

-tim

e c

hild

ren

an

d t

wo

ch

ildre

n re

pre

sen

ting

on

e

full-

time

eq

uiv

ale

nt

ca

n p

ay

mo

re t

ha

n $

40 p

er

mo

nth

.

cC

T=C

on

ne

ctic

ut;

CA

=C

alif

orn

ia; C

O=

Co

lora

do

; NC

=N

ort

h C

aro

lina

.P=

For-

pro

fit c

en

ters

; NP=

No

np

rofit

ce

nte

rs.

Ou

tco

me

s o

f F-t

est

s lis

ted

we

re s

tatis

tica

lly s

ign

ific

an

t a

t th

e p

<.0

5 le

vel,

at

lea

st (

tha

tis,

like

ly t

o h

ave

oc

cu

rred

by

ch

an

ce

no

mo

re t

ha

n f

ive

tim

es

ou

t o

f 10

0).

of the job. In addition, as more womenenter the general work force, their person-al preferences concerning the type of workthey do may shift away from child care, andthey may also learn they can earn more inother jobs. The consequence for the fieldappears to be that, as current child careworkers leave their jobs, they are beingreplaced with a less well-educated workforce.9,14

For these reasons, estimates of the for-gone earnings of providers are included incalculating the full cost of providing childcare in the CQO study, the EFCC study, andin this article, although a more conservativeapproach would be to consider the full cost-of-care estimates reported here as upperlimits.

Center FinancesTables 2 and 3 summarize key results of theCQO study concerning the costs of provid-ing child care in centers. The tables compareresults by region, as well as by profit sector,that is, private for-profit and private non-profit (including publicly operated centers).

Expended and Full CostsTable 3 reports monthly budgets showingaverage expended costs and revenuesources for all centers, and for-profit andnonprofit centers. Total expended costs forall centers averaged $403 per month. Laborcosts represented 70% of costs, occupancy14%, food 5%, other operating expenses9%, and overhead 2%. These percentagesare roughly comparable to findings fromearlier studies.4,35 When donations and


Cost

Item Nonprofit Centers For-Profit Centers All Centers

Percentage Percentage PercentageDollars of Total Dollars of Total Dollars of Total

Labora $ 331 78.9 % $ 239 61.8 % $ 285 70.3 %Occupancya 31 7.4 78 20.2 55 13.8Fooda 21 4.9 16 4.3 19 4.6Other operatinga 28 6.7 40 10.4 34 8.6Overhead 9 2.0 11 2.9 10 2.5

Total costb $ 420 100.0 % $ 386 100.0 % $ 403 100.0 %

In-kind donationsa $ 60 14.0 % $ 17 4.0 % $ 39 10.0 %Forgone earnings 104 25.0 108 28.0 106 26.0

Full costc $ 584 139.0 % $ 511 132.0 % $ 548 136.0 %

Revenue

Parent feesa $ 244 56.5 % $ 357 88.2 % $ 302 72.3 %Public feesa 88 20.4 38 9.4 63 14.9USDA CACFP (food)a 13 3.0 3 0.6 8 1.8Other public fundsa 46 10.7 0 0.0 23 5.3Other private fundsa 41 9.4 7 1.8 24 5.6

Total Revenueb $ 433 100.0 % $ 405 100.0 % $ 420 100.0 %

Surplus or Profit $ 13 3.0 % $ 18 4.5 % $ 16 3.7 %

a Significant difference between for-profit and nonprofit sector: p < 0.001.b Elements may not add up to totals and percentages may differ slightly from actual because of rounding.c Significant difference between for-profit and nonprofit sector: p < 0.01.

Source: Helburn, S., ed. Cost, quality, and child outcomes in child care centers: Technical report. Denver, CO: Department of Economics, Centerfor Research on Economics and Social Policy, University of Colorado, 1995.

Table 3

Budget for a Typical Child Care Center, Means per Child per Month


forgone earnings were considered, the fullcost of care rose to $548 per month or 136%of total expended cost, with forgone earn-ings representing 26% of expended costs,volunteer time about 2%, occupancy subsi-dies 7%, and other donations 1% ofexpended costs.

Although for-profit and nonprofit cen-ters did not differ in the total cost per child,they did differ in how they allocated costs.Nonprofit centers spent significantly morethan for-profit centers on labor and food;for-profit centers spent significantly morethan nonprofit centers on facilities andother operating expenses. For instance,labor costs represented 79% of nonprofitcenter costs, but only 62% of for-profit cen-ter costs; occupancy costs represented 7% ofnonprofit center expended costs, but 20%of for-profit center costs.

These differences in the composition ofcosts by profit sector reflect significant differ-ences in receipt of in-kind donations betweensectors. Most of the donations to nonprofitcenters (70%) were in the form of subsidizedfacilities (for example, the use of space in a

church building), whereas for-profit centersrarely received occupancy donations. Despitesubsidization of facilities, for-profit centersstill had higher occupancy costs, indicatingthat nonprofit centers apparently used lower-quality space than for-profit centers. (SeeAppendix B for additional information aboutthe costs of child care facilities.)

■ Regional Differences Expended costs variedby state (see Table 2). Connecticut had sig-nificantly higher costs and North Carolinasignificantly lower costs than the other states.Costs in California and Colorado were closeto the average. These tendencies held truewhen cost-of-living adjustments were madefor differences in consumer prices in thefour states, although the range of differencediminished.

Regional differences in the full cost ofcare reflect differences in expended costs,donations, and forgone earnings. The fullcost of care was also higher in Connecticutthan in the other three states and lower inNorth Carolina, reflecting regional differ-ences in both expended costs and dona-tions. Larger subsidies were found inConnecticut than in other states, except forforgone wages, which were highest inColorado and lowest in Connecticut.

Revenues, Profits, and FeesTable 2 also describes center revenues,breaking out revenue gained from parent fee pay-ments, which equals center total revenueminus other revenue sources (for example,state reimbursement, public and privategrants, cash donations, other income), andnet parent cost, which equals the averageparent fee after deducting a 20% incometax credit.

Although for-profit and nonprofit cen-ters did not differ in total revenue per childper month, they did differ in the sources ofthose revenues. As indicated in Table 3, non-profit centers had much more diversifiedrevenue streams than for-profit centers.Compared with nonprofit centers, for-profitcenters were highly dependent on parentfees, which comprised 88% of total revenuecompared with 57% for nonprofit centers.For-profit centers, which historically havereceived fewer public dollars, received littlerevenue from state fee reimbursement,USDA food grants, or other public grants.These sector differences are even more dra-matic when church-affiliated centers are sep-arated from other nonprofit centers.Although these results are not listed in Table3, parent fees represented 83% of the totalrevenue for church-affiliated centers, whichis very similar to the for-profit profile. Inother words, within the nonprofit sector,the independent nonprofit and publiclyoperated centers that serve subsidizedchildren rely least on parent fees.

For the whole sample, the rate of profitor surplus on revenue was low, averaging3.7%. By comparison, corporate profit onrevenue for the nation as a whole averagedabout 9% in recent years.36

Table 2 indicates that the mean maxi-mum monthly fee was $454 for full-time

Although for-profit and nonprofit centersdid not differ in total revenue per child permonth, nonprofit centers had much morediversified revenue streams than for-profitcenters.


infants and $372 for full-time preschoolchildren. Fees were highest in Connecticutand lowest in North Carolina. When feeswere adjusted to account for regional differ-ences in the cost of living, these differencesin fees across states diminished somewhat.Fees charged by for-profit centers remainedhigher than those charged by nonprofitcenters.

Costs to Society, Centers, and ParentsTable 2 illustrates how the burden forchild care costs is shared by society,providers, and parents. On average, asmentioned above, the full cost of care was$548 per child per month; center expend-ed costs were $403 per child per month or74% of full cost; and net parent costs were$265 per child per month or 48% of the fullcost of care.

Table 3 shows that parent fee paymentsmade up 72% of total revenues, or an aver-age of $302 of total revenues of $420 permonth. Net parent cost is the cost to the par-ent after deducting a 20% federal incometax credit.37 In for-profit centers, overall, par-ent fees represented almost 90% of centertotal revenue and 70% of the full cost ofcare. After deducting the child care taxcredit, parent fees amounted to 62% of for-profit full cost, but in nonprofit centers,because a smaller percentage of parentspaid full tuition, parents paid on averageabout 40% of the full cost of care (notdepicted in tables).

In other words, when only the revenuescoming into centers each month are con-sidered, parents are shouldering most ofthe burden, especially parents whosechildren are in for-profit child care pro-grams. When all subsidies (including for-gone wages) are considered, however, theparental share declines to about half of thefull cost of care.

Importance of Center SizeThe CQO study corroborated earlier find-ings25 that economies of scale exist in childcare. Centers that had longer hours of oper-ation, operated closer to capacity, or servedlarger numbers of children had lowerexpended costs per child per hour with noapparent ill effects on quality of care. Whenthe study controlled statistically for theeffects of operating in different states, results

indicated that labor cost, total cost, and totalrevenue per child were significantly higherin centers serving fewer than 40 children ona full-time basis than in centers serving morethan 40.

Wages and Forgone EarningsIn 1993, the mean wages for teaching staffwere as follows: $7.22 per hour for teachers,$5.70 per hour for assistants, and $11.33 foradministrative directors.1 Significant region-al differences reflect differences in cost ofliving and in labor market conditions: Wagesin California and Connecticut were compa-rable and were $2.50 per hour higher thanin Colorado and $3.00 per hour higher

than in North Carolina. Adjustments for dif-ferences in labor market conditions in thefour states suggested that Connecticut childcare jobs pay higher wages relative to otherjobs in Connecticut hiring people withsimilar characteristics, and this may helpexplain the higher quality observed inConnecticut centers. Based on a compari-son of the CQO study and the earlierNational Child Care Staffing Study(NCCSS),9 real wages for child care staffmembers do not appear to have changedsignificantly since 1988, when the NCCSScollected wage data in different but compa-rable regions of the country.

The CQO study estimated forgone earn-ings. On average, all categories of staff in allstates earned less than their forecast wage.Fully 93% of teachers and assistants earnedless in child care than their counterpartsearned in other occupations and industries.Teachers, on average, earned $5,238 peryear less in child care than they could earnin other professions, given their education,racial and ethnic status, gender, and age.Forgone earnings for assistants, while lowerin absolute terms ($3,582), were propor-tionately higher than for teachers. There waswide variation in forgone wages for adminis-trators, but many administrators (32% ofthose in the study) were earning their mar-ket wage.

When all subsidies (including forgonewages) are considered, the parental sharedeclines to about half of the full cost of care.


20% of Providers withAll Providers Highest Child Care Income

Percentagea Annual per Annual perAnnual per of Gross Full-Time Annual per Full-Time

Provider Income Equivalent Child Provider Equivalent Child

Number of Observations 133 27

Gross Income from $ 19,242 100.0 % $ 4,120 $ 38,062 $ 4,806Child Care

Less CostAssistants and substitutes 1,874 9.7 401 4,420 558Food for children 2,376 12.3 509 3,122 394Toys, materials, equip- 655 3.4 140 965 122

ment, equipment repairMileage 255 1.3 55 528 67Household supplies 691 3.6 148 846 107Advertising, office 294 1.5 63 517 65

expenses, license feesDues, tuition, entertain- 262 1.4 56 573 72

ment, gifts, professionaltravel

Accountant, lawyer, 218 1.1 47 417 53insurance

Repairs, remodel, furni- 1,259 6.5 270 2,134 269ture, loan interest, otherexpenses

Utilitiesb 1,012 5.3 217 1,500 189Total Cost $ 8,895 46.2 % $ 1,905 $ 15,023 $ 1,896

Net Provider Income $ 10,347 53.8 % $ 2,215 $ 23,039 $ 2,910

Other Total Benefits from $ 2,517 13.0 % $ 539 $ 2,700 $ 341Child Care Provisionc

Provider Income and $ 12,864 66.9 % $ 2,755 $ 25,883 $ 3,262Benefits from Child Care

Other Provider InformationMean full-time equiva- 4.67 7.92

lent childrend

Staff ratioe 3.87 4.46Quality indexe 0.26 0.72Percentage with own 55 41

child in programFamily income $ 39,420 $ 47,950Age of provider 40 41Weekly full-time $ 74 $ 88

preschool feef

a Percentages may differ slightly from actual because of rounding.b Utilities expenditures based on the time/space formula allowed by IRS are included as a proxy for facilities expenses related

to child care provision. Assuming that the provider does not use specialized facilities for child care, housing costs equal the extra costs for maintenance, replacement, and utilities created by child care provision.

c Includes tax benefit minus Social Security taxes plus value of own child care.d Significant difference between all providers and 20% of providers with the highest child care income: p < 0.001.e Significant difference between all providers and 20% of providers with the highest child care income: p < 0.05.f Significant difference between all providers and 20% of providers with the highest child care income: p < 0.01.

Source: Modigliani, K., Helburn, S., Morris, J., and Culkin, M. The economics of family child care project. Unpublished manuscript. WheelockCollege, Boston, 1996.

Table 4

Income and Costs per Year, and Other Characteristics for Family ChildCare Providers, in 1993 Dollars

Family Child Care FinancesTables 4 and 5 report results from theEconomics of Family Child Care (EFCC)study,3 which interviewed 133 family childcare providers in California, North Carolina,and Texas who had also participated in thestudy on Quality in Family Child Care andRelative Care. The methods used to estimatecosts and income in family child care werecomparable to those used for estimatingcenter-based costs and revenues, with modi-fications where necessary to take intoaccount that, because family child careproviders work at home, work and familytime (and expenses) may sometimes overlap.

Provider Costs and IncomeTable 4 compares mean income, costs andother provider characteristics for the wholesample and for the 20% of family child careproviders with the highest child careincome.38 The table reveals that family childcare providers had low average earnings,and that serving a larger number of childrenis generally in the interest of both providersand their clientele. On average, all providersserved about five children; the higher-income group served about eight. Averagenet provider income and benefits (includingsavings generated by not working outsidethe home, such as not needing to pay forchild care for their own children) was$12,864 per year for all providers, comparedwith $25,833 for the higher-income group.For the high-income group, annual childcare income approaches a living wage, andquality of care is significantly higher than forthe sample as a whole. These higher-incomechild care providers spend more per childthan the average provider on assistants andsome administrative expenses, but otherwisespend less per child than the averageprovider.

Total costs of providing care represent46% of the provider’s gross income fromchild care. The largest categories of spend-ing are for assistants (9.7% overall, but assis-tants are used by relatively few providers),food (12%), and utilities (5%). As in centercare, labor cost (which includes theprovider’s net income from child care) is byfar the largest component of total cost (64%of gross income). This is a smaller percent-age than nonprofit centers allocate for per-sonnel but is roughly comparable to that offor-profit centers.

Given the low income generated byfamily child care, it is more appropriate tothink of family child care providers as self-employed individuals who work at homefor a variety of reasons rather than asentrepreneurs who are motivated by prof-its. The EFCC study, for instance, estimatesmean forgone earnings for family childcare providers at $3,841 per year (seeTable 5); these providers could hardly bethought of as earning a profit. In fact,most respondents reported that theychose family child care for altruistic orfamily reasons (usually, to care for theirown children).

Comparison of Center andFamily Child Care FinancesTable 5 compares the financial picture forcenter and family child care provision usingdata from the CQO and EFCC studies. Themost interesting finding shown in Table 5 isthe apparent similarity in total costsbetween the two child care sectors whenthe provider’s income is included as themajor component in labor costs. Occu-pancy costs were lower, and food andother costs were higher in family childcare than in center care. The full costof care represented 135% of family childcare expended cost, but 156% of centerexpended cost.

Fees were somewhat lower for pre-schoolers, but dramatically lower for infantsin family child care homes than in centers.Lower fees may be part of the reasonmany parents of infants choose family childcare over center-based care, as reported by

Hofferth in this journal issue. Centerteacher wages, while low, were neverthelesshigher than net family child care providerincome and benefits. The forgone wages ofteachers in centers were higher than theforgone earnings of family child careproviders, partly because of the reducedexpenses of working incurred by familychild care providers (for example, they didnot need to pay for child care).


Fees were somewhat lower for preschoolers,but dramatically lower for infants in familychild care homes than in centers.

Relationships BetweenChild Care Quality andCosts While a few studies have considered rela-tionships between child care costs and qual-ity using only structural measures of quali-ty,39–41 the CQO and EFCC studies examinedrelationships between cost and quality usingmeasures of process quality. In both centerand family child care programs, the costs ofproviding child care were moderately andpositively related to the quality of care pro-vided. In the family child care study, themain comparison was between mediocreand poor-quality child care homes because

there were too few good-quality homes inthe sample (only 11) to be able to makemeaningful comparisons. Results indicatedthat total expended cost per child per hourwas higher in mediocre homes than in poor-quality homes. In the CQO study, total laborcosts were higher in better-quality centerswhen other center characteristics42 wereconsidered. This was expected becausestaffing ratios and staff education were asso-ciated with quality, as described earlier inthis article.

To determine how much it costs toincrease center quality, a cost function was esti-mated in the CQO study. Based on econom-


Centers Family Care

Mean Monthly Percentageb of Mean Monthly Percentageb ofValues per Child Total Cost Values per Child Total Cost

Number of Observations 300 133

CostsLabor $ 236 70 % $ 218 64 %Occupancyc 53 14 18 5Food 19 5 42 12Other 44 11 65 19

Total Cost 352 100 343 100

Full Cost $ 548 156 % $ 462 135 %

Fees and EarningsMaximum monthly fees

Preschool $ 372 $ 318Infants 454 341

Teachers wage or net 15,018 12,864provider income andbenefits

Teacher or provider 5,238 3,841annual forgone earningsd

a The family child care providers sample comes from Los Angeles, California; Charlotte, North Carolina; and Dallas–FortWorth, Texas. The center data came from Los Angeles, California; Greensboro–High Point and Winston-Salem, in NorthCarolina; the Front Range in Colorado; and Connecticut. To permit a reasonable comparison, Connecticut centers wereeliminated from this analysis because of the high cost of living in that state. Colorado and Dallas–Fort Worth are compa-rable.

b Percentages may differ slightly from actual because of rounding.c Occupancy costs were estimated at the utilities costs included as a child care expense based on the time/space formula.d Forgone earnings are not strictly comparable in that family provider income and benefits include the value of the provider’s

Table 5

Comparison of Center and Family Child Care Costs, Fees, andIncomea, Dollars per Child per Month

Sources: Helburn, S., ed. Cost, quality, and child outcomes in child care centers: Technical report. Denver, CO: Department of Economics,Center for Research on Economics and Social Policy, University of Colorado, 1995; Modigliani, K., Helburn, S., Morris, J., and Culkin, M. Theeconomics of family child care project. Unpublished manuscript. Wheelock College, Boston, 1996.

ic theory, a cost function is a mathematicalexpression of the relationship between thecosts of a service or product and factors suchas the price of labor and the quality andquantity of services provided which arethought to influence the cost. In the CQOstudy, the cost function related center totalvariable costs to wages of staff of differenteducation levels (a high school diploma,some college, or bachelor’s degree or more),hours of child care provided for each agegroup, child care process quality, physicalsize of the center, volunteer hours, region ofthe country, and whether the center was for-profit or nonprofit. Total variable costincludes expended costs (except for facilitiescost), in-kind donations, and the extrapolat-ed salary of owner-operators.

Results of the analyses indicated that pos-itive, significant, but modest relationshipsexist between cost and quality. Specifically,the results suggest that raising quality by25% from mediocre to good would increasetotal variable costs about 10%-—that is, by 13cents per child hour or about $300 per childper year. For a child care center of averagesize, this would mean an increase of $18,000per year.43 This assumes that wages, hours ofservice provided, space, and volunteer hoursremain the same. The analysis also indicatedthat it costs more to increase quality when acenter’s quality is already relatively high (forexample, to increase it from good to excel-lent). It costs less to increase quality frompoor to mediocre.

This finding that improving quality frommediocre to good would require cost increas-es of only about 10% seems to contradict thefinding reported earlier, based on severalstudies, that process quality is related tostaffing ratios. The cost of increasing staffingratios enough to raise quality from mediocreto good is substantial. This suggests that atleast some of the qualities or characteristicsthat enable centers to provide good-qualitycare relatively inexpensively are not includedin the model. For instance, the CQO studyfound that a child care director’s administra-tive experience and effectiveness positivelyaffected quality, but this factor was not (bydesign) included in the cost function. Otherfactors, such as personality traits of staff, staffcommitment to good quality, and effectiveteamwork, undoubtedly also contribute tooverall program quality.

The CQO study also looked for differ-ences in the cost-quality relationship by sec-tor (for-profit versus nonprofit) and cor-roborated findings of earlier studies.39–41 Nosignificant differences were seen in variablecost per child hour between the two sectors,for the same size of center, wages paid, qual-ity, and amount of physical space. Lookingwithin the for-profit and nonprofit sectorsreveals overall similarities in cost and quali-ty between church-affiliated and for-profitcenters. Both groups of centers seem to

occupy the same market niche, dependenton attracting mainly middle-income fami-lies. Average values of structural andprocess quality in both sectors are signifi-cantly lower than for independent non-profits and publicly operated centers. Itappears that these centers face stiff compe-tition with each other and that parents’inadequate knowledge about quality meansthat affordability, not quality of services,drives competition.

Overall, the results of this analysis indi-cate that cost and quality are related, butthat it takes only a small amount of moneyto raise quality. As mentioned earlier, par-ents do not appear to differentiate amongvarying levels of quality of child care pro-grams, and these data indicate that parentscannot use fees as an indicator of quality.Furthermore, there does not seem to bemuch consumer pressure to increase quali-ty. In analyses not otherwise reported here,the CQO study found that the differentialin fees between mediocre and good centerswas smaller than the cost differential, imply-ing that there is an economic disincentivefor centers that are highly dependent onparent fees to improve quality.

ConclusionsA review of the literature on the cost andquality of child care for children of workingparents indicates that (1) the quality of ser-vices is mediocre, on average; (2) the cost offull-time care is high; (3) at the present time,


Parents do not appear to differentiateamong varying qualities of child care pro-grams, and these data indicate that parentscannot use fees as an indicator of quality.

the cost of increasing quality from mediocreto good is not great, about 10%; (4) goodchild care is dependent on professionallyapproved staffing ratios, well-educated staff,low staff turnover, a good adult work envi-ronment, and effective leadership by experi-enced directors; (5) these characteristics areinterrelated and, taken together, are costly;and (6) therefore, the present modest linkbetween cost and quality reflects the impactof other factors that have not been capturedin studies thus far, such as the staff charac-teristics and teamwork of committed staffproviding a labor donation.

Other conclusions are also notable.Licensed family child care fees are not muchlower than center care fees, except for infantcare, although on average the quality of thatcare is somewhat lower. The quality of ser-vices and earnings of family child careproviders are higher for those who care for alarger number of children with a commit-ment to their profession and who operate asa small business.

Fees are set through competitive forcesthat maintain a continual downward pres-sure. Donations of facilities, equipment andmaterials, volunteers, and labor mean thatfees do not reflect the cost of production.There is not much variation, on average,between fees for mediocre and good care, sothat parents, who have little knowledgeabout the quality of care their children arereceiving cannot use the fees charged tothem as a means of differentiating betweengood and not-so-good care. Furthermore,the fee differential is smaller than the differ-

ence between the cost of mediocre and ofgood care, so that centers have an economicdisincentive to improve quality.

These findings are depressing, to say theleast. They reflect the low priority given tochildren’s care and women’s work inAmerican society. This must change if chil-dren’s needs are to be met now and in thefuture. Further research is needed to bet-ter understand such aspects of child careprovision as the nature of parents’ choicesfor child care and the underlying featuresof care that affect cost and quality, asthese will help to illuminate relevant policydirections.

In the meantime, the following steps arerecommended: (1) launch consumer educa-tion efforts to help parents identify good-quality child care programs and to informthe American public of the consequences ofpoor-quality programs for children; (2)implement higher standards for child careat the state level as a step toward eliminatingpoor-quality child care; (3) avoid public poli-cies that encourage people to become childcare providers if they have no interest insuch a career; (4) increase public and pri-vate investments in child care, using strate-gies such as tying funding to standards thatproduce high-quality care, providing finan-cial incentives to enable the hiring andretention of trained professionals, and tai-loring employee benefits to provide childcare support; and (5) develop the means tocompensate child care workers as is appro-priate for their levels of training, experience,and responsibility.


1. Helburn, S., ed. Cost, quality, and child outcomes in child care centers: Technical report. Denver, CO:Department of Economics, Center for Research in Economic and Social Policy, University ofColorado, 1995.

2. Kontos, S., Howes, C., Shinn, M., and Galinsky, E. Quality in family child care and relative care.New York: Teachers College Press, 1994.

3. Modigliani, K., Helburn, S., Morris, J., and Culkin, M. The economics of family child careproject. Unpublished manuscript. Wheelock College, Boston, 1996.

4. Hayes, C.D., Palmer, J.L., and Zaslow, M.J., eds. Who cares for America’s children? Child carepolicy for the 1990s. Washington, DC: National Academy Press, 1990.

5. Howes, C., and Hamilton, C.E. Child care for young children. In Handbook of research on theeducation of young children, B. Spodek, ed. New York: Macmillan, 1993.

6. See, for example, Gomby, D.S., and Larner, M.B., eds. Long-term outcomes of early child-hood programs. The Future of Children (Winter 1995) 5,3:1–224.

7. Phillips, D. Predictors of quality child care. Washington, DC: National Association for theEducation of Young Children, 1987.

8. Clifford, R.M., Harms, T., Pepper, S., and Stuart, B. Assessing quality in family child care.In Family day care: Current research for informed policy. D. Peters and A. Pence, eds. New York:Teachers College Press, 1992.

9. Whitebook, M., Howes, C., and Phillips, D. Who cares? Child care teachers and the quality ofcare in America. Oakland, CA: Child Care Employee Project, 1990.

10. Phillips, D., Mekos, D., Scarr, S., et al. Paths to quality in child care: Structural and contex-tual influences in children’s classroom environments. Unpublished manuscript, 1995.

11. Howes, C., Phillips, D., and Whitebook, M. Thresholds of quality: Implications for thesocial development of children in center based care. Child Development (1992) 63: 449–60.

12. Howes, C., and Smith, E.W. Relations among child-care quality, teacher-behavior, chil-dren’s play activities, emotional security, and cognitive activity in child care. Early ChildhoodResearch Quarterly (1995) 10,4:381–404.

13. Phillips, D.A., Howes, C., and Whitebook, M. The social policy context of child care:Effects on quality. American Journal of Community Psychology (1992) 20,1:25–51.

14. Whitebook, M., Phillips, D., and Howes, C. National Child Care Staffing Study revisited: Fouryears in the life of center-based child care. Oakland, CA: Child Care Employee Project, 1993.

15. Hofferth, S.L. National Child Care Survey, 1990 . Washington, DC: The Urban Institute,1991.

16. Howes, C., and Hamilton, C. E. The changing experience of child care: Changes in teach-ers and in teacher-child relationships and children’s social competence with peers. EarlyChildhood Research Quarterly (1993) 8:15–32.

17. Kisker, E., and Maynard, R. Quality, cost and parental choice of child care. In The economicsof child care. D.M. Blau, ed. New York: Russell Sage Foundation, 1991.

18. Shinn, M., Phillips, D.A., Howes, C., et al. Correspondence between mothers’ perceptionsand observer ratings of quality in child care centers. Unpublished manuscript, New YorkUniversity, 1990.

19. Browne-Miller, A. The day care dilemma: Critical concerns for American families. New York:Plenum, 1990.

20. Mitchell, A., Cooperstein, E., and Larner, M. Child care choices, consumer education and low-income families. New York: National Center for Children in Poverty, 1992.

21. Cryer, D. Infant child care: Parent perceptions and child experiences. Unpublished manu-script, University of North Carolina, 1989.

22. Siegel, G., and Loman, L. Child care and AFDC recipients in Illinois: Patterns, problems, andneeds. St. Louis, MO: Institute of Applied Research, 1991.

23. Brush, L.R. Projecting the costs of full-day child care from the costs of Head Start.Unpublished manuscript for the National Academy of Sciences Panel on Child CarePolicies, 1988.

24. Clifford, R., and Russell, S.D. Financing programs for pre-school children. Theory IntoPractice (1989) 28,1:19–27.

25. U.S. General Accounting Office. Early childhood programs: What are the costs of high-quality pro-grams? GAO/HRD-90-43Br. Washington, DC: U.S. GAO, January 24, 1990.

26. Ruopp, R., Travers, J., Glantz, F.M., and Coelen, C. Children at the center: Summary findingsand their implications. Final report of the National Day Care Study. Cambridge, MA: AbtAssociates, 1979.

27. See note no. 3, Modigliani, Helburn, Morris, and Culkin. This study surveyed a subsampleof 133 family child care providers who also participated in the Study of Children in FamilyChild Care and Relative Care. Providers from California, North Carolina, and Texas wereinterviewed by phone in 1993 to obtain extensive cost, revenue, subsidies, and tax informa-tion, as well as information about enrollment, program characteristics, and provider atti-tudes. These data were combined with information gathered through the companionstudy on process and structural quality.

28. Ferguson, W.D., and Cosgrove, J. Nonprofit status and wages: An investigation of the childcare industry with firm-specific effects. Unpublished paper delivered at the WesternEconomics Association meetings, June 1994.

29. Preston, A.E. The effects of property rights on labor costs of nonprofit firms: An applica-tion to the day care industry. Journal of Industrial Economics (1988) 36,3:337–50.


30. Preston, A.E. The nonprofit worker in a for-profit world. Journal of Labor Economics (1989)7,4:438–63.

31. Weisbrod, B.A. Nonprofit proprietary sector behavior: Wage differentials among lawyers.Journal of Labor Economics (1983) 1,3:246–63.

32. Mirvis, P.H., and Hackett, E.J. Work and work force characteristics in the nonprofit sector.Monthly Labor Review (April 1983) 106,2:3–12.

33. Rose-Ackerman, S. Altruistic nonprofit firms in competitive markets: The case of day-carecenters in the United States. Journal of Consumer Policy (1986) 9:291–310.

34. In the two studies, forgone earnings were measured for each staff member and for eachfamily provider as the difference between the person’s actual child care earnings and thewages the person could earn in the same labor market (state, in this case), given the per-son’s age, gender, racial and ethnic background, and number of years of education. Tomake these calculations, wage equations were estimated for the regions from the 1992Census of Population Statistics (CPS) tapes. These equations were then used to estimatethe forgone earnings of each center teaching staff person and each family child careprovider. The CPS tape does not include data on prior experience of workers, and thestudy data set did not include marital status of staff or their number of children, so thesemeasures could not be included in the wage equations. These exclusions could affect theestimates to some extent.

35. In the GAO study (see note no. 25), labor costs represented 65%, occupancy (includingdonated space) 11%, food 7%, and other operating costs 19%. As cited in Powell, R.,Eisenberg, D. R., Moy, L., and Vogel, J. Cost and characteristics of high-quality early child-hood education programs. Child and Youth Care Forum (April 1994) 23,2:103–18.

36. The economic report of the President, 1993. Washington, DC: U.S. Government Printing Office,1993, p. 361.

37. Most parents are entitled to credit on their federal income taxes ranging from 30% ofchild care expenses for low-income parents who pay taxes down to 20% for parents withan adjusted gross income greater than $28,000. Given the average incomes of parents inthe sample centers, most parents would qualify only for the 20% credit. This estimate over-states the effect of the tax credit because not all parents who are eligible actually take it.(See also the article by Stoney and Greenberg in this journal issue.)

38. Estimates were also calculated for providers with family incomes below 185% of povertyand for providers serving at least three full-time equivalent children. For these two cate-gories, differences from the average were not, on the whole, significant, and so are notreported here.

39. Mukerjee, S., and Witte, A.D. Provision of child care: Cost functions for profit-making andnot-for-profit day care centers. Journal of Productivity Analysis (1993) 4: 145–63.

40. Preston, A. Efficiency, quality and social externalities in the provision of day care:Comparison of nonprofit and for-profit firms. Journal of Productivity Analysis (1993)4:165–82.

41. Powell, I., and Cosgrove, J. Quality and cost in early childhood education. Journal ofHuman Resources (1992) 27:472–84.

42. Factors included the state in which the center was located, whether centers were for profitor nonprofit, and the mix of age groups of children served.

43. This analysis may understate the cost-quality relationship somewhat because it holds wagesconstant; it is based on the assumption that wages for particular quality levels of staff mem-bers (as indicated by their levels of education) are set in the labor market and that centerspay the going wages. Although this assumption is undoubtedly true for the majority of cen-ters, it is also possible that good-quality centers offer higher-than-market wages, either toattract the best people or to increase productivity of existing staff members by improvingtheir morale and loyalty. In the former instance, wages based on education levels do notcapture other worker qualities for which centers have to pay. In the latter instance, it prob-ably costs somewhat more than an average of 13 cents per child per hour to raise qualityfrom mediocre to good levels.


Documents

Child Care Cost and Quality