Abstract
We review what is known about Head Start's impacts on children and argue that the program is likely to generate benefits to participants and society as a whole that are larger than program costs. Our conclusions differ from those in some previous reviews because we use a more appropriate standard to judge program effectiveness (benefit-cost analysis), draw on a body of new evidence for Head Start's long-term effects on early cohorts of participating children, and discuss why common interpretations of a recent randomized experimental evaluation of Head Start's short-term impacts may be overly pessimistic. Estimating the long-term benefits of Head Start for recent participants necessarily requires a number of assumptions. But we believe there is a plausible case that short-term effects on achievement scores of .1 to .2 standard deviations might be large enough for Head Start to pass a benefit-cost test. Data from the experiment imply that Head Start enrollment &em; as distinct from assignment to the experimental treatment group &em; usually generates impacts of at least this magnitude. While, in principle, there could be more beneficial ways of deploying Head Start resources, the benefits of such changes remain uncertain and there is some downside risk. There is a growing scientific consensus that a variety of early childhood interventions generate benefits in excess of costs at current levels of spending, which suggests the value of increased spending in this area. However there remains considerable uncertainty about what form any additional investment should take. Additional government funding to support rigorous research to identify the relative strengths of Head Start and its alternatives, as well as the critical "active ingredients" in these programs that most effectively produce short- and long-term developmental benefits, would be a particularly high value-added activity.