Media draw big, misleading conclusions from a study with a small sample size.
You have to spend money to make money. That is the conclusion of a new article by Nobel Prize–winning economist Jim Heckman and his team of researchers who studied two 1970s-era preschool programs for disadvantaged children and found large positive results persisting into adulthood. Participating students saw benefits in health and quality of life, labor income, crime reduction, and education gains, and their mothers saw increases in their income because of the child care provided by the program.
Sure, the cost was enormous, as the authors readily admit. The program cost $18,514 per child per year (in 2014 dollars) and lasted from birth to age five. But it produced $6.30 in benefits for every dollar spent. That equates to a 13 percent rate of return, leading NPR to headline a story “How Investing in Preschool Beats the Stock Market, Hands Down.” The Atlantic went so far as to ask, “Why Doesn’t Public School Start at Birth?” Taking a more measured tone, the Christian Science Monitor led with “High ROI: Why Preschool Programs Are a Good Investment for Society.”
But we need to clarify something right off the bat: These are the findings from a study of 70 kids.
If you look at the table (page 8 of the study by Heckman et al.) that describes the cohort of students the authors studied, you see an initially recruited sample of 121 students. The actual “treatment” of center-based child care from ages zero to five had 53 participants in one of the two programs and 17 in the other, for a total sample of 70 students.
It is a huge leap to argue that such an intensive, hothouse study of such a small sample is proof that such an intervention would work at scale. Just look at the description (page 7) of the programs themselves:
The programs individualized treatment. Each child’s progress was recorded and learning activities were appropriately adjusted every 2 to 3 weeks. Environments were organized to promote pre-literacy and provide access to a rich set of learning tools. The curriculum emphasized active learning experiences, dramatic play, and basic concepts of order and category (“pre-academic skills”), as well as discipline and the ability to interact with and respect others. At later ages (3 through 5), the program focused on the development of “socio-linguistic and communicative competence.”
The program is substantially more robust than your average pre-K center. Extending that care all the way from birth makes it even more labor-intensive and more reliant on a highly trained, well-coordinated staff. If we’re going to take a program like this state- or nationwide, we have to ask: Is the average pre-K center capable of doing such intentional and intensive work?
According to the authors, for ages zero to one, these programs had a staff-to-child ratio of 1 to 3; for ages one to four, a ratio of 1.4 to 5; and ages four to five, a ratio of 1:5 to 6. In my state, Missouri, we have around 78,500 four-year-olds alone. Implementing a program like this would mean recruiting tens of thousands of teachers and professionals. It would also mean spending over $1.45 billion for just the four-year-olds for just one year. Should we expect them to do as well as those who worked in the highly controlled, research-intensive environment that these authors studied?
#related#To be clear: I’m not saying that the authors didn’t find what they found. I’m not surprised that disadvantaged children who received such intensive care at a young age did better than students who did not receive it. But I am intensely skeptical of drawing strong implications for policy from a tiny sample. What happens when the program is expanded to 700 kids, or 7,000 kids, or 700,000 kids? Should we expect the same magnitude of results? I wouldn’t.
Selling these programs as long-term cost savers or wise investments at the state level based on this data is misleading. While there is reason to think that pre-K can benefit low-income students, believing that it can create such fantastic returns requires an enormous leap of faith.