The Future of Head Start

The Nation’s Dual System of Child Care and Preschool Education

There have been dramatic changes in both child care and preschool education in the nearly forty years since Head Start’s birth. With the increase in labor force participation by mothers over this period, the need for child care increased apace. By 1970, about 40 percent of mothers were in the labor force; by 1985, over 60 percent; by 2000, over 70 percent. Although there have been numerous legislative campaigns since the early 1970s to create a universal federal child care program or to enact federal regulations that apply to most or all child care facilities, the federal government has not accepted general responsibility for either the financing or quality of child care. However, in 1990 the federal government did establish a major child care program for poor and low-income children called the Child Care and Development Block Grant (CCDBG). The block grant provides nearly $5 billion yearly to states to help poor and low-income families (below 85 percent of state median income) pay for child care while parents are at work or school. In the face of congressional reluctance to create national standards, every state has enacted its own child care standards—standards that most child advocates feel are inadequate.

A central characteristic of state child care programs is that parents choose their own care. It follows that a diverse array of care—including centers, family child care in neighborhoods, and relative care—receives subsidies. Although many see market diversity as a strength of the system, others see it as chaos.

Alongside this bustling but arguably flawed child care market, a moderate but growing number of preschool programs aimed at stimulating school preparation have been created over the years. When Head Start began nearly forty years ago, very few children were in facilities that aimed specifically to prepare them for school. Forty states and the District of Columbia have either established their own preschool programs or have used state funds to expand Head Start. Some of the state preschool programs are elaborate, like those in Georgia, Oklahoma, and New York, while others are quite modest.

This brief overview shows that programs for preschool children have been driven primarily by two forces. First, the child care market exists largely to provide care for children while parents work or go to school. This market has facilities of diverse size and quality that are only lightly regulated. According to the Census Bureau, in 1999 there were as many as 13 million preschool children in market child care facilities. Research suggests that much of this care is of mediocre or poor quality, although a small fraction of the centers are of high quality and are probably the equivalent of preschool programs. By contrast, there is a second set of facilities designed specifically to prepare children for school. This sector includes Head Start and the preschool programs established in recent years by states. Perhaps as many as 1.5 million preschoolers are in these facilities, 900,000 in Head Start and the remainder in state-supported facilities. In addition, Head Start began a program in 1995 that provides child and family services to poor pregnant women and their children through age three. This Early Head Start program, however, enrolls only about 45,000 children.

Both child care and preschool education are important. Child care is essential to enable both single and married mothers and fathers to work. Employment in turn is central, not only to economic opportunity for women and to the health of the national economy, but also to the economic viability of families, especially mother-headed families. The 1996 welfare reform legislation that has been associated with remarkable increases in employment and earnings by low-income single mothers underlines the importance of this work support function of child care.

Preschool education programs are important because they help prepare low-income children for school. This issue bears emphasis. Since at least the 1960s, when President Lyndon B. Johnson initiated the War on Poverty, a major goal of federal policy has been to improve the educational achievement of poor children. Now almost forty years later, we seem to have learned some hard lessons. Despite the expenditure of billions of dollars on programs for poor preschool children, as a recent study by Valerie Lee and David Burkam of the University of Michigan shows, the school readiness gap between poor and more advantaged children persists. Not only do poor children enter school with serious educational deficits, but the achievement gap between poor and more advantaged children actually increases during the school years. President Johnson’s goal of using preschool programs to bring all children to the same “starting line” as a strategy for equalizing educational opportunity goes largely unrealized.

Evidence on Effectiveness of Preschool Education

Even so, high quality preschool programs such as the Abecedarian program in North Carolina and the Perry Preschool project in Michigan have shown what is possible. High quality preschool can reduce grade retention, reduce placement in special education, increase high school graduation rates, increase college attendance, and produce a host of related effects. A preschool program in inner-city Chicago involving more than 1500 preschoolers conducted by Arthur Reynolds and others at the University of Wisconsin suggests that similar effects are possible in larger-scale programs.

But the effects of Head Start are not as clear. After nearly forty years of operation, there is not a national random-assignment evaluation of the long-term impacts of Head Start. A study of this type, ordered by Congress in 1997, is now underway, but even initial impacts will not be known until next year. Because high-quality studies that meet scientific criteria are not available, the Head Start literature is somewhat weak. A comprehensive review in 1985 found that most studies of Head Start were of very poor quality and would not permit reliable conclusions. But based on the best studies that existed at that time, the review concluded that Head Start produced immediate impacts on IQ, school readiness, and three measures of socioemotional development, but that the effects faded within a year or two of the time children entered the schools. Recent studies of Head Start are somewhat more encouraging. Janet Currie and her colleagues at UCLA have conducted two studies based on national data that show some effects of Head Start on the school achievement of white children, but no long-term effects on the school achievement of black students. The Currie studies did, however, show some effects on improving the health of black children as well as on reducing criminal behavior.

Results from the FACES surveys of Head Start youngsters and families in 1997 and 2000 are at best modestly encouraging. Perhaps the worst news coming from these surveys is that in both years, four- year-olds attending Head Start scored on average only slightly above the twentieth percentile on tests of vocabulary, letter recognition, early writing, and early mathematics. Better news is that these children scored slightly better at the end of the year on some skills and slightly improved their performance on letter recognition between 1997 and 2000. Even so, all the skill scores for both years were below the thirty-second percentile.

Two conclusions about preschool education seem justified. First, high quality preschool education can very substantially improve the school readiness and school performance of poor and minority children. Second, Head Start produces results that are more modest than the results produced by high-quality preschool education programs such as Abecedarian, Perry Preschool, and the Reynolds Child-Parent Center program in Chicago. Most observers, including researchers, think Head Start is a good program that needs improvement. More specifically, researchers and others point to the uneven quality among Head Start centers, the need to improve the quality of teachers, and the need for increased accountability for results. Against this background, the Bush administration has proposed a potentially radical reform of Head Start.

The Bush Proposal and Its Critics

President Bush proposes to give states the option of assuming control of Head Start and its funding. A major justification for the president’s proposal is that increased coordination of Head Start and state preschool programs could lead to more efficient use of resources and greater accountability. Further, in order to gain control of Head Start funds, states would have to create a strong plan to focus all preschool programs on achieving the academic and social skills needed to succeed in school.

Thus, in order to take control of Head Start funds, a given state would have to agree to several conditions in a written proposal submitted to a Board composed of the Department of Health and Human Services (HHS) and the Department of Education. To be approved, the proposal must, among other requirements:

Explain how the state will work with the public schools to define the academic and social skills that five-year-olds must have to succeed in kindergarten

Provide a plan for developing preschool activities and materials that will help poor children acquire the academic and social skills needed to succeed in kindergarten

Outline a state accountability program for determining whether four-year-olds are acquiring these academic and social skills

Demonstrate that the state will continue to serve as many poor children as are currently being served by Head Start

Report the level of state funding of state preschool programs or Head Start supplements, and provide assurances that the state will continue spending at least an equivalent number of state dollars on preschool programs

Assure that the state will continue to provide comprehensive services that include social, family, and health services

Provide information on how the state will assure professional development among preschool teachers and administrators

Outline how the state will ensure coordination among Head Start, state preschool programs, Title I preschool programs, and other preschool programs in the state.

States that decide to accept the challenge of meeting all these conditions and assume control of Head Start would be provided with modest additional funds, paid out of the current technical assistance fund of $165 million controlled by HHS, to implement their state plan.

The Administration proposal requires states to find ways to do what Head Start has not done sufficiently—improve the school readiness of poor children. The Administration believes that states already have lots of incentive to improve preschool because of the rigors of the No Child Left Behind Act of 2001. This Act requires states to test children in grades 3 through 8 and to publish the results in a form that reveals the performance of individual schools over time. Students attending schools that consistently fail are allowed to transfer to different schools. Many states, after reviewing years of experience with Title I and other programs designed to improve the school performance of poor and minority children, are concluding that they cannot be successful unless these children have a better preschool foundation. Hence the states’ growing interest in improving and expanding preschool programs.

Critics of the Administration plan grant that Head Start has flaws, and that there are many individual programs that are of poor quality. But they argue that a large majority of Head Start programs are reasonably successful, and that it is unclear whether states could improve Head Start. They also contend that the process of turning over the program to states could harm Head Start and that the replacement programs developed by states might be worse.

Critics also contend that although coordination and efficiency may help states cover more children with the same amount of money, the size of this effect would probably be small. There is little evidence to determine whether the president’s proposal will actually increase efficiency, and even less evidence to support any estimate of the number of additional children that could be served with the savings from increased efficiency. Certainly, expansion through efficiency could not be expected to allow states to cover all the 190,000 or so poor four-year-olds who are eligible but not enrolled in Head Start. Proponents of the Bush proposal respond that many of those 190,000 children are now in state preschool programs and many others are in child care paid for by the Child Care and Development Block Grant. Thus, nontrivial sums are already being spent on many of the poor four-year-olds who are not in Head Start, thereby reducing the cost of ensuring that all of them receive a high-quality program before entering the schools.

On balance, it seems very likely that additional funds will be needed if all children from poor families are to receive at least one year of preschool education. Nor is it clear that only children below the poverty level (about $15,000 for a family of three in 2003) need preschool education. If eligibility were moved to 125 percent of the poverty level, an additional 200,000 or so four-year-olds would become eligible and costs would increase by around $1.4 billion per year.

Many critics of the Bush proposal also believe that strengthening the impact of Head Start on school readiness will cost more money per child. The per-child cost of Head Start is now around $7,000 per year. States could reduce this cost by increasing classroom size, hiring teachers for lower salaries than Head Start teachers are paid (about $25,000 per year, including benefits), or through other means. Each of these approaches has drawbacks, however, and none has been demonstrated to be compatible with increased school readiness. Moreover, one of the major criticisms of Head Start is that its teachers are underqualified. A straightforward approach to improving teacher quality is to hire teachers with better qualifications. But this reform would increase the per-child cost of Head Start.

Finally, critics of the Administration proposal hold that preparing children for school is not the only purpose of good preschool programs. The current Head Start program provides comprehensive services that include health screening, dental checkups, and social services. In addition, most of the programs have parents who are active on their boards of directors and in providing assistance to classroom teachers, many of whom are themselves current or former Head Start parents.

Administration officials respond that their proposal requires states to maintain these comprehensive services and to explain in their application how they plan to do so. Further, states are responsible for administering most social service programs, thereby creating an opportunity for states to coordinate this array of services with Head Start at moderate cost. According to this view, there is little reason to worry that states will reduce the comprehensive services of the current Head Start program.


Perhaps the best arguments for the Bush proposal are that Head Start does not now achieve the goal of adequately preparing poor and minority children for school, and that states would have the authority to coordinate all the major preschool programs in order to increase efficiency and improve the school readiness of poor children. The evidence indicates that the average child in Head Start is probably somewhat better prepared for school than she would be without Head Start. Even so, national data show unequivocally that poor children as a group are substantially behind their more fortunate peers when they enter the schools, and that they fall further behind during the elementary school years. Given the vital importance of education to achieving equality of opportunity, the nation must find ways to improve both preschool education and the K through 12 school system. The president’s plan requires states to take all the reasonable actions that would be expected to improve the school readiness of poor children, including improved curriculum, better coordination with the public schools, and increased accountability.

Though reasonable, the president’s plan is untested. Moreover, the Head Start program is better than many preschool programs and much better than the average child care program. Head Start has also achieved high levels of parental involvement and even produced some evidence of lasting effects on health and school performance. In short, turning Head Start over to the states carries risks.

Thus, we recommend that Congress enact legislation this year that would allow up to five states to implement the president’s proposal. In addition to the conditions states must meet under the president’s plan, a further condition for state participation should be an agreement to cooperate with a five-year, third-party evaluation of state reforms. States would have a year of planning before they begin implementation, during which time the third-party evaluator would be selected and the evaluation plan established. Where possible, the plan should call for random assignment. The secretaries of HHS and Education would be responsible for working with the evaluators and the states to develop a common set of performance measures that would be used to test children across all the demonstration programs—as well as children from states still operating under the regular Head Start program. The federal government would pay for the evaluations.

Although the president’s plan does not require states to coordinate with child care programs to participate in the demonstration, the secretaries should be encouraged to select at least one state that would attempt to coordinate its Head Start, state preschool, Title I, CCDBG child care, and perhaps even its preschool programs for disabled youngsters. States are now spending almost $9 billion on child care of uncertain quality through the CCDBG. Perhaps states could begin demonstrating how to coordinate their preschool and child care programs in such a way as to create or strengthen the school readiness component of child care programs. Recent experience in North Carolina and other states seems to show that operators of child care facilities, including informal family child care facilities, are quite willing to receive advice on how to improve their programs.

This demonstration plan represents a reasonable compromise between those who are concerned that the quality and even existence of Head Start would be jeopardized by turning responsibility for the program over to states, and those who believe that states can improve preparation for school through increased coordination and accountability. Given the immensity of the task and the modest success achieved thus far, new ideas are worth trying.


Scholars have long noted that families and neighborhoods play a major part in child development and school readiness. Whether a preschool program can completely overcome the deficits poor children acquire from their home environments is an open question. Expecting a one-year preschool program to overcome the huge gap in school readiness between poor and more fortunate children may be unreasonable. Indeed, some researchers and educators have concluded that more than one year of high quality preschool education will be required to reduce the school readiness gap, and that even such interventions will need to begin well before the age of three or four. One consequence of this conclusion would be the need for substantial additional funding. Finally, no matter what the outcome of these demonstrations, and even without the need for two or more years of preschool, additional funds will almost certainly be necessary to provide all poor children with at least a year of high quality preparation for schooling.