It’s hard to find a federal program more popular than Head Start. Especially since the end of the Reagan administration, it has enjoyed bipartisan favor, with its budget quadrupling to $6.2 billion. So it is surprising to be reminded that there’s very little empirical evidence that the program actually does give a head start to the underprivileged preschoolers it serves. President George W. Bush has now proposed moving the program from the Department of Health and Human Services (HHS) to the Department of Education and increasing Head Start’s emphasis on teaching language skills. (He has also proposed a two percent budget increase.) That has touched off a debate about what Head Start should be asked to do. When President Lyndon B. Johnson launched Head Start in 1965 as part of his War on Poverty, the goal was to give economically disadvantaged children a leg up by providing a range of educational, medical, social, and psychological services so that they could enter kindergarten on a more equal footing with their better-off peers. Today, Head Start serves more than 800,000 preschoolers—about half the eligible population. "The jury is still out on Head Start," notes economist Janet Currie of the University of California, Los Angeles, in her survey of research on early childhood education programs in the Journal of Economic Perspectives (Spring 2001). There’s never been a large-scale, long-term study of Head Start children (though HHS is now planning one). One reason: There’s no single Head Start; the roughly 1,500 Head Start programs are locally administered. Also, such studies are costly and difficult. The children (including a non-Head Start control group) would have to be tracked over many years to determine whether Head Start had any measurable effects on their school performance or other aspects of their lives. Other influences, such as differences in family income and parents’ marital status, would have to be taken into account.

The research that does exist tends to point to one conclusion: Head Start’s academic effects fade out as kids grow older. A 1990 Educational Testing Service study, for example, found that involvement in the program "had positive effects on both verbal test scores and measures of social adjustment." But by the end of second grade, the Head Start kids were statistically indistinguishable from their peers.

That’s where today’s debate begins: What’s responsible for the apparent "fadeout," and what should be done about it?

Two sides of the argument are presented in Education Matters (Summer 2001, online at edmatters.org). David Elkind, a professor of child development at Tufts University, says that the fade-out should come as no surprise. "The giants of earlychildhood development," such as Maria Montessori and Jean Piaget, all agreed on at least one thing: Children’s minds develop in stages, and they’re not equipped "until the age of five or six" to reason their way through reading and math. It’s far more important for young children "to explore and conceptualize" by "seeing, touching, and handling new things and . . . experiencing new sensations." In Elkind’s view, it "makes little sense to introduce formal instruction in reading and math" to preschoolers, and it’s "simplistic" to think that early schooling will give disadvantaged youngsters "the skills and motivation to continue their education and break the cycle of poverty."

In the same issue of Education Matters, the arguments of Grover J. Whitehurst, chairman of the Department of Psychology and a professor of pediatrics at the State University of New York at Stony Brook, could not offer a greater contrast. He contends that the problem is the century-old "child-centered" style of education featured in Head Start (and many American primary and secondary schools). Yes, children can be harmed by schooling that’s beyond them, he says, but the average child attending Head Start now "exits that program in the summer before kindergarten being able to name only one—yes, one—letter of the alphabet." Whitehurst favors "content-centered" schooling "organized around the principle that there are skills and dispositions that children need to be taught if they are to be prepared for later schooling and life." He scoffs at Elkind’s "giants," who conducted no empirical research, but he allows that the evidence for "contentcentered" education is only "inferential" at this point. That evidence is strongest in the case of reading. For example, there are studies showing a strong link between the literacy skills children possess upon entering kindergarten and their subsequent school performance, while other studies reveal a link between student reading difficulties and other problems, such as dropping out or committing crimes.

One of Head Start’s founders, Yale University psychologist Edward Zigler, offers yet another perspective in Education Matters. Go ahead and strengthen the preschool education component of the program, he and a Yale associate say, but don’t forget Head Start’s other purposes, from identifying children who are malnourished or have vision problems to providing emotional support to troubled kids.

That’s similar to the tack Janet Currie takes. Her own research suggests that the Head Start fade-out afflicts only African American children. She thinks the problem may be what happens after Head Start, when black children go off to inferior schools. But Currie still thinks Head Start has a lot to offer. Her "back-of-the-envelope" calculations suggest that the short- and medium-term social benefits of Head Start cover 40 to 60 percent of its costs. Those benefits include everything from improving child nutrition to saving kids from costly special education programs later in their school careers. Above all, the benefits include the value of quality child care. The alternatives to Head Start are frequently dismal. Stir in hard-to-estimate longer-term benefits (e.g., better school attainment, reductions in crime), and Currie believes that the program could pay for itself. One study suggests that the Perry Preschool Project in Ypsilanti, Michigan—a much more expensive version of preschool than Head Start—has yielded a total package of benefits that far outweigh the costs. Currie believes that the evidence is "compelling enough" to warrant a recommendation. To her, it makes the most social sense to expand Head Start into a year-round, full-time program and open it up to more poor and children who are otherwise at risk.