One intriguing figure in the new book, Childhood Programs and Practices in the First Decade of Life, examines the benefits per dollar of different early childhood and early school-age programs at different ages. (The figure is on p. 182, in a chapter by Arthur Reynolds, Judy Temple, and Suh-Ruu Ou. The chapter is mostly about the Chicago Child-Parent Center program. But it also includes some more general discussion of “what works”.)
The figure shows two things. First, there are a variety of programs at all these ages, from birth to age 8 (the maximum age the figure considers), that have dollar benefits greater than costs. Second, as Reynolds, Temple and Ou say, “preschool programs for 3- and 4-year olds generally show the highest returns” (p. 181).
I have similar findings in my book Investing in Kids. For each dollar invested in universal pre-K at age 4, state economic development benefits are $2.78. For each dollar invested in the Abecedarian program, which provided child care and preschool services from birth to age 5, the state economic development benefits are $2.25. For each dollar invested in the Nurse Family Partnership program, which helps disadvantaged first-time mothers from the pre-natal period to the child’s second birthday, the state economic development benefits are $1.85.
I also compare universal pre-K at age 4 with reducing class size in early elementary school. Universal pre-K has about twice the effects per dollar of lower early elementary class size. However, lowering early elementary class size does have state economic development benefits exceeding costs.
I was reminded of this when I heard an excellent talk the other night by Richard Rothstein. (Rothstein is a long-term commentator on the role of social and economic class in influencing educational outcomes. He has helped develop the so-called “broader bolder” approach to improving American education.)
I was reminded of this when Mr. Rothstein mentioned, as a side point, that although preschool is needed, for many kids it is too late, because their language development is already way behind by the time they are age 4.
Similar points are sometimes made when people talk about brain development and when it occurs. The compelling logic is that if so much development happens early on, before ages 3 and 4, shouldn’t the returns be highest for programs that intervene early on?
I would agree that potential returns are higher for programs that intervene earlier, when children’s brains and characters are more malleable. But realizing such potential returns may be a little harder to do in a cost-effective way when children are younger, at least based on what we currently know.
When children are ages 3 and 4, we can improve early childhood development in larger group settings than are effective for younger children. That lowers the costs per child of preschool relative to some earlier interventions. It also seems that although we need well-trained teachers and a good curriculum and reasonable class sizes for preschool, we don’t need to have some genius government administrator to run such programs. An average American state and local government or school district of reasonable competence can successfully run such preschool programs in a way that is effective for many children in improving their life course.
Earlier interventions such as child care need to have smaller group settings. Parenting programs seem most effective, based on current knowledge, when they are one-on-one. The parenting program with the most rigorous evidence of success, the Nurse Family Partnership program, requires that the home visitation with parents be delivered by nurses, which obviously raises costs compared to other possible delivery systems.
None of this argues that these programs don’t work. As I point out above, high quality parenting programs such as NFP do have benefits that significantly exceed costs, as do high-quality child care programs. However, as of right now, the most rigorous evidence for these programs suggests that their benefits per dollar are somewhat less than those of high-quality preschool programs.
More intensive programs such as full-time child care may also deliver higher total benefits than one year of preschool. For example, as I pointed out in a recent post, full-time child care and preschool from birth to age 5 can raise the future earnings of the lowest income children by about one-third. One year of preschool has effects more on the order of perhaps raising future earnings of low income children by 10% on average. However, these greater total returns for the earlier intervention come at more than proportionately greater costs, so benefit to cost ratios are somewhat lower for the earlier intervention than for preschool at age 4.
Does this pattern reflect some inevitable “law” of social science? No. I would not be surprised if over time we figure out some earlier interventions than preschool that would be significantly more cost-effective than preschool in providing benefits to children. For example, there potentially would be major increases in benefit/cost ratios for a parenting assistance program that could be delivered more in a group setting, with less one-on-one attention — but only if that cheaper program proved to be significantly effective in improving parenting. It would certainly be worthwhile if some national foundations funded some experiments in this area.