Many Happy Returns

Early Childhood Programs Entail Costs, but the Paybacks Could Be Substantial

1 | 2 | 3

By Lynn A. Karoly, M. Rebecca Kilburn, Jill S. Cannon, James H. Bigelow, and Rachel Christina

Lynn Karoly is a senior economist at RAND. Rebecca Kilburn and Jill Cannon are director and associate director, respectively, of RAND Child Policy. James Bigelow is an operations research analyst at RAND, and Rachel Christina is a RAND consultant.

For four decades, scientific research has sought to determine if early childhood intervention programs such as home visiting, parent education, and early childhood education — alone or in combination — could yield long-term benefits for participating children and families. Lifelong benefits could include increases in academic achievement (test scores), educational progress (timely promotion), educational attainment (years of schooling completed), behavioral and emotional competencies, health, employment, and earnings; and decreases in child abuse and neglect, delinquency and crime, and use of social welfare programs.

Almost all the early childhood programs studied so far have focused on children “at risk.” Common risks involve living in poverty, having a single parent, having a mother with less than a high school education, and living in a family whose primary language is not English. Nearly half of entering U.S. kindergartners face at least one of these risk factors. Nearly one in six is subject to more than one.

Well-designed early childhood interventions have been found to generate short- and long-term benefits to participating children and families in multiple domains of well-being.

For the substantial percentage of children who are disadvantaged in these ways, there is now a critical mass of completed scientific research to answer the underlying question of whether early childhood intervention programs can yield long-term benefits for participating children and families. We have compiled solid evidence that investments in high-quality early childhood programs do promote healthy physical, mental, social, and emotional development in the numerous ways described above among children at risk, improving their subsequent life outcomes during school age and beyond.

Two additional questions facing policymakers today are whether such programs would still be good investments if they were provided for all children, not just those in greatest need, and whether doing so is feasible. Recent RAND research has attempted to help answer these two additional questions as well.

First, we estimated the costs and benefits of one strategy for investing in an early childhood program for all: high-quality universal preschool, specifically in the state of California. We found that such an investment would indeed generate large positive net returns to society.

Second, we examined a range of state efforts to expand preschool, exploring the implementation choices made and some of the lessons that can be learned. While we found that there are significant challenges to the implementation of high-quality universal preschool programs, we believe that the challenges can be overcome.

We summarize our findings as follows:

  • Well-designed early childhood interventions have been found to generate short- and long-term benefits to participating children and families in multiple domains of well-being. When these benefits are valued in dollar terms, the return to society ranges from $1.26 to $17.07 for each dollar spent. The benefits that translate into the largest dollar values — educational attainment, earnings, and crime reduction — often accrue later in life. The best outcomes seem to be associated with programs that offer children better-trained caregivers, smaller child-to-staff ratios, and greater intensity of services.
  • Investing public money to make high-quality preschool available to every 4-year-old in California would generate an estimated $2 to $4 in benefits to society for every dollar spent. The cost of such a program would be more than offset by (1) a decline in the amount of special education services provided, (2) decreased grade repetition, (3) reduced youth and adult crime, and (4) increased lifetime earnings as a result of higher educational attainment, along with the associated tax revenues. The benefits would require several years to pay off, but they would be big enough to be worth waiting for (at anything less than a 10-percent annual discount rate).
  • Solid evidence does not yet exist about the best strategies for “scaling up” early childhood programs at a statewide level. Uneven or confused systems of state governance, program provision and support, and accountability can present unique challenges, particularly for states in transition from targeted to universal preschool programs. We suggest some ways to overcome these challenges, from reorganizing state bureaucracies to raising public awareness that investment in early childhood programs should not be justified entirely, or even primarily, on their academic benefits alone.

Numerous Investment Options

We found rigorous evaluations of 20 early childhood programs serving disadvantaged children in the United States (see the table). Four evaluations followed many or all participants only until ages 2 or 3, giving no indication of the ultimate effects of the programs on school performance or later success. The remaining 16 evaluations measured outcomes at the time of kindergarten entry or beyond, sometimes well into adulthood.


Rigorous Evaluations Exist of 20 Early Childhood Intervention Programs
Program Type Reach Initial year of evaluation
Carolina Abecedarian Project Combination (home visiting/ parent education combined with early childhood education) One site in North Carolina 1972
Chicago Child-Parent Centers Combination Chicago, Ill. 1983
Comprehensive Child Development Program Combination Nationwide demonstration projects 1990
DARE to be You Parent education Western states 1991
Developmentally Supportive Care: Newborn Individualized Developmental Care and Assessment Programt Home visiting National 1979
Early Head Start Combination National 1995
Early Training Project Combination Murfreesboro, Tenn. 1962
Head Start Combination National 1967
High/Scope Perry Preschool Project Combination Ypsilanti, Mich. 1962
HIPPY (Home Instruction Programs for Preschool Youngsters) USA Parent education/home visiting Multiple states 1990
Houston Parent-Child Development Center Combination Houston, Tex. 1970
Incredible Years Parent education Multiple states Mid-1990s
Infant Health and Development Program Combination Eight sites 1985
Nurse-Family Partnership Program Home visiting Multiple states 1978
Oklahoma Pre-K Early childhood education Oklahoma 2001
Parents as Teachers Parent education/home visiting National 1991
Project CARE (Carolina Approach to Responsive Education) Combination One site in North Carolina 1978
Project CARE Home visiting One site in North Carolina 1978
Reach Out and Read Parent education National 1996
Syracuse Family Development Research Program Combination Syracuse, N.Y. 1969
SOURCE: Early Childhood Interventions, 2005.

Significant and often sizable benefits were demonstrated for 19 of the 20 programs. In some cases, the benefits were demonstrated soon after the program had ended. In other cases, the benefits were observed throughout adolescence and in the transition to adulthood. In one case, lasting benefits were measured 35 years after the intervention had ceased.

Based on the results from these programs and other evidence in the research literature, there is some basis for identifying program features that are associated with more effective interventions. First, better-trained caregivers appear to generate stronger program outcomes. In the context of center-based programs, more favorable effects have been found when lead teachers have a college degree as opposed to no degree. Home visiting programs also appear to fare better with a trained nurse as opposed to a paraprofessional home visitor.

Second, for center-based programs, evidence suggests that programs with smaller child-to-staff ratios are more successful. Third, programs that provide a more intensive set of services are associated with better outcomes, but the evidence does not allow us to determine the optimal number of program hours or how that number might vary for children from various subgroups.

Of the 20 rigorously evaluated programs, one of them, the Comprehensive Child Development Program, was not shown to be effective. A second program, the Infant Health and Development Program, showed favorable outcomes as of the final follow-up at age 8, but the outcomes could not be translated into dollar savings. For the other programs, the estimates of net benefits to society ranged from about $1,400 per child to nearly $240,000 per child. In terms of benefit-cost ratios, the returns to society for each dollar spent ranged from $1.26 to $17.07 (see the centerpiece).

There were positive net benefits for programs that required a large investment (over $40,000 per child), for those that cost considerably less (under $2,000 per child), and for those in between. Likewise, there were favorable economic returns for programs that focused on parent education or home visiting, as well as for those that combined such services with early childhood education.

The largest benefit-cost ratios were associated with the programs that included longer-term follow-up data. Although the programs with evaluations that followed children only until kindergarten or a few years beyond typically did not measure the outcomes that are likely to be associated with the largest dollar benefits — educational attainment, delinquency and crime, and earnings — these programs might eventually generate these benefits as well.

Not all benefits could be translated into dollars. Therefore, the benefit estimates are likely to be conservative. The estimates generally did not account for any improved labor market performance among the parents of participating children or for economy-wide advantages that might result from the improved educational attainment of the workforce.

The returns were likely to be higher in those programs that targeted disadvantaged children than in those that served lower-risk children. In the Nurse-Family Partnership home visiting program, for example, the return for each dollar invested was $5.70 from a higher-risk population but only $1.26 from a lower-risk population.

It is unreasonable to expect the returns for a program serving a disadvantaged population to be matched when the same program serves a broader population. We certainly would not expect to see the same returns in a universal program, such as a state-run preschool program available to all. Nevertheless, the net benefits for universal programs could still be positive.

Next section: Universal Preschool
1 | 2 | 3