ÇÑ×ÓÊÓƵ

To conservatives, it’s a classic example of big government gone awry, continuously expanding without regard for evidence of failure. The Heritage Foundation flatly: “Head Start Doesn’t Work.”

Democratic presidential nominee Hillary Clinton, by contrast, has to double investment in the program, saying it gives children “a strong foundation to learn.”

At first glance, Head Start skeptics seem to have a point: Federal using high-quality, randomized research techniques found small benefits that disappeared by the time a student finished kindergarten or first grade.

But a by Claire Montialoux of Berkeley’s Institute for Research on Labor and Employment summarizes a number of other studies showing that Head Start may actually be cost-effective and that the federal research understates the benefits of the program.

Indeed, shows that Head Start produced fairly large positive effects on test scores for children who would have otherwise been cared for at home. The researchers find that the effects of the initiative appear to gradually fade out over time, but some effects persist at least through first grade. (There were no effects of Head Start compared with students who attended other forms of center-based care.)

There is also that Head Start especially benefits students with the lowest baseline skills and has a on parental involvement.

The key question, though, is whether Head Start has lasting impact into adulthood. Famously, in the research world, the and preschool programs in the 1960s and ’70s were shown to have a number of long-run effects, including increased earnings and IQs and lower arrest rates. Critics, however, the applicability of research about these old, small-scale, intensive programs to the question of whether to expand existing programs like Head Start.

Research suggests that Head Start can in fact produce lasting benefits for students who otherwise would not attend preschool of any kind. A by David Deming of Harvard compared children who received Head Start with their siblings who didn’t. Beneficiaries of the program had better adult outcomes across a variety measures: high school graduation, college attendance, health, crime and teen parenthood. The gains were about 80 percent the size of those shown in Perry preschool.

Crucially, these positive long-term effects came about even though the test score benefits of the program faded, just as was observed in the federal study. “A projection of future benefits for these children based solely on test score gains would greatly understate the impact of the program,” Deming wrote. In other words, studies showing fade-out of Head Start gains are not strong evidence that the program is ineffective. This is in line with that short-run test score gains can dissipate over time but may “reappear” in the form of long-run benefits in adulthood.

using this sibling-comparison approach also find long-term positive effects of Head Start.

Stevens, who has of existing studies of pre-K, is skeptical of significantly expanding government-funded preschool. “There’s currently a $9 billion federal program with very uneven performance,” she said of Head Start, arguing that instead of spending large sums of money growing it, policymakers ought to experiment with and study a variety of “small pilot projects” to determine the best way to spend scarce resources.

No one argues against the need for more study, but there exists carefully done suggesting that expanding Head Start is likely a cost-effective policy lever. The research estimates almost a two-to-one return on investment for each dollar spent on Head Start, considering its impact on children’s income as adults. Although growing the program would (of course) cost money, it would also save some public dollars since it would likely replace state-financed preschool.

New York City, under Mayor Bill de Blasio, recently expanded its preschool program, with some of success.

Doubling down on Head Start, as Hillary Clinton proposes, should certainly be paired with a focus on quality. that Head Start centers vary significantly in their effectiveness and that those offering full-day classes and home visits are particularly good. There are to improve the program, though research will have to examine whether such changes work as intended.