Journal Issue: Children and Poverty Volume 7 Number 2 Summer/Fall 1997
Head Start was created in 1965 to "strike at the basic cause of poverty,"52 which was seen as a lack of education among the poor. Over the years the program has matured into the nation's primary federally sponsored child development preschool program. Ninety percent of the children served are three and four years old; two-thirds are four-year-olds. About 13% are children with disabilities. The program is implemented with considerable variation around the nation to meet diverse community needs. More than 90% of the children are served in centers, but among offered programs, the number of days per week and hours per day is variable. In 1994, the annual federal cost per child was $4,345. Like WIC, Head Start is not an entitlement. In addition to the federal income eligibility criterion (see Table 1), local programs add other criteria relating to family characteristics designed to help Head Start better serve community needs.
The overall goal of Head Start is to bring about "a greater degree of social competence in preschool children from low-income families."53 Social competence includes cognitive, intellectual, and social development; physical and mental health; and adequate nutrition. Key principles of Head Start include providing comprehensive services (including education, health, nutrition, social services, and parent involvement), fostering the parent's role as the principal influence on the child's development, encouraging parents to be involved in policy and program decisions, and establishing partnerships with community agencies to improve the delivery of services to children and families.Target Population Coverage
In terms of the incomes of Head Start families, the program achieves its goal of targeting families below the poverty line; in 1994–95, 64% of participating families had annual incomes under $9,000, and 83% had yearly incomes of less than $12,000.54 In terms of coverage, however, the limited budget of Head Start means that only 38% of eligible three- and four-year-olds are served.55 Additionally, in an era of increasing need for child care by parents who are either working or in job training programs, the largely part-day structure of Head Start programs makes it difficult for them to serve this share of the eligible population. Head Start's part-day structure is likely to become increasingly problematic with the work requirements imposed by the recent welfare reform legislation. Head Start's traditional emphasis on parent involvement—both in the educational program and in policymaking roles—may also become more difficult to achieve as eligible parents cope with the increased demands of education and/or work requirements.Achievement of Program Goals
Head Start research has examined program outcomes across the full range of children's well-being, although the bulk of the research has focused on cognitive development. This review focuses on evidence of Head Start's short-term outcomes as they provide the most robust evidence of program effectiveness. The findings show positive effects of Head Start in both the cognitive and socioemotional domains of development at the end of the Head Start year. This evidence of beneficial effects of programs is critical, given the recent evidence that long-term poverty depresses the cognitive and socioemotional development of young children.56
No national evaluations have been conducted of the Head Start program in the past 20 years, although many specific demonstration programs within Head Start have been rigorously studied.57 Head Start's parent agency—the Administration for Children, Youth and Families (ACYF) in the Department of Health and Human Services (DHHS)—has placed greater emphasis on conducting smaller-scale studies that are able to focus on particular policy questions. Still, Head Start evaluations have examined all of the primary outcomes listed above, with an emphasis on cognitive development. Because Head Start serves predominantly four-year-olds, this review focuses on findings relating to their development.58 In 1985, DHHS published a synthesis of existing studies based on meta-analytic methods.59 Although this synthesis is now more than 10 years old, it still provides the most comprehensive assessment of Head Start's benefits. The review that follows begins with this synthesis, supplementing its findings with more recent studies where appropriate.
Across 72 studies, regardless of the particular measures used, the synthesis study found Head Start to have sizable effects on children's cognitive development when measured at the end of the Head Start year. Furthermore, the 21 most rigorous studies (those that employed an experimental treatment-control group design) showed an average effect size of 0.52 across all cognitive measures. An effect size of 0.52 indicates that the mean of the cognitive measures for the Head Start group was more than half a standard deviation above the mean for the control group. Differences of this magnitude are generally considered to be educationally meaningful. As Barnett recently noted, differences of this magnitude constitute "meaningful improvement in cognitive ability and can have important implications for children in terms of academic performance and placement in special education classes."60
A recent study based on assessments conducted as part of the National Longitudinal Survey of Youth (NLSY) bolsters these cognitive development findings.61 The study involved a comparison of Head Start children with siblings who had another type of preschool experience or no preschool. The study is based on a national sample, an advantage over many of the smaller local studies included in the synthesis report. However, there are difficulties with the sibling-comparison design, and the sample may not be representative of all Head Start programs and participants. Study findings associate Head Start with large and significant gains in test scores among both whites and African Americans. With careful controls for family background factors, not only did Head Start children perform better than siblings who attended no preschool, but also, for white children in the sample, the effects of Head Start were much greater than the effects of attending other preschool programs. The cognitive outcome measure in this case was the Peabody Picture Vocabulary Test, which, although flawed in a number of respects, remains a strong predictor of later school achievement.
Head Start has also shown beneficial effects on children's socioemotional development at the end of the Head Start year. Across 17 studies, the Head Start synthesis report found benefits of Head Start participation in educationally meaningful effect sizes on both achievement motivation and social behavior.59 These findings cannot be attributed to a particular outcome measure: the 17 studies included 18 different measures of achievement motivation and 11 measures of social behavior. Even though the national Head Start program has, from the beginning, emphasized the importance of health services for children and their families, research focused on children's health or physical development outcomes has been sparse. However, a random assignment evaluation in the mid-1980s in four counties with highly underserved populations found that, at the end of the year, Head Start children were much more likely than the controls to have received basic health services, enjoyed better access to health care services, experienced improved health status, eaten meals significantly higher in nutrient quality, and exhibited better motor coordination and development.62 Further, Head Start children who displayed pediatric problems upon program entry were less likely to have the same problems remaining one year later. An analysis of NLSY's limited health data supported these findings of improved delivery of health services, associating Head Start with an 8% to 9% higher probability of being immunized.61Indirect Program Outcomes
Most reviews of Head Start's effectiveness include a discussion of the long-term benefits. However, policy considerations and methodological weaknesses are reasons for de-emphasizing the research of long-term gains. Although Head Start program planners had hoped that their program efforts would sustain benefits well into the elementary school years, a one-year program such as Head Start may not be sufficient to protect children from future risk. Additionally, while a number of studies claim to show a fade-out effect (such that gains present immediately after the end of the Head Start year are no longer found in second or third grade), their methodologies have not been sufficiently rigorous to support the conclusion that Head Start fails to have positive benefits in the long term.63
Longitudinal studies of a limited number of other early childhood programs (many of which have features in common with Head Start) have demonstrated long-term effectiveness. Effects include positive outcomes on school achievement beyond third grade and reduced rates of grade retention, enrollment in special education, and delinquency.64 As the authors of the non-Head Start studies point out, however, the non-Head Start programs studied are smaller in scale than Head Start and thus able to be implemented with greater consistency. Model programs generally have been found to have larger effects than large-scale programs. Also, the "average" Head Start program might not be expected to produce the same results as high-quality Head Start programs.64
Studies of Head Start's indirect effects for families, staff, and communities have been few, but the synthesis report found several areas in which the benefits of Head Start programs can be seen to extend beyond enhancing children's development.59 Head Start parents participate in various paid and volunteer program positions that range from clerical to classroom to policymaking. In 1992, some 94% of enrolled families received needed support services through the program.65 Some studies have shown that parental child-rearing practices have been positively affected by Head Start participation, but others have not.59 Career development is only secondary to Head Start's aims, yet some 35% of Head Start staff in 1992 were former children or parents of the enrolled Head Start children.65 Such employment opportunities become an increasingly important program benefit as low-income parents seek alternatives to public assistance.