Landmark studies of preschool programs in the 1960s and 70s showed that they could make a big difference for children from low-income families. But more recent experimental studies of preschool don’t show as strongly positive results for students’ academic and social outcomes. Why?
That’s the question that Anamarie Whitaker, an assistant professor in human development and family sciences at the University of Delaware, set out to investigate in a new working paper. She and seven colleagues outlined the findings from randomized studies of modern preschool programs—which show smaller effects than the well-known Perry Preschool program that began in 1962, or the Abecedarian project of the 1970s.
Children that participated in these programs saw a range of positive outcomes that lasted well into adulthood: They were healthier, more likely to graduate from high school, and made more money than their peers. Many elements of the Perry Preschool program inspired early childhood education design for years afterward.
Education Week spoke with Whitaker about what might be driving these lower effects, and how the preschool landscape is changing. This interview has been edited for length and clarity.
What made the Perry Preschool and Abecedarian programs distinctive—and effective?
Perry Preschool was a preschool program designed to demonstrate that early life experiences influenced children’s early development. It was a small demonstration program that took place in the 1960s, in Ypsilanti, Mich. It was a small-scale study with a random assignment design: 123 African American children were part of the study, ages 3 and 4. Fifty-eight children were randomly assigned to receive the treatment condition, and the others were randomly assigned not to receive the treatment, being in the control.
The treatment condition received high-quality preschool services, that was a school year program. It provided half-day programming five days a week and a weekly home visit by the child’s teacher. The program was really focused on close teacher-child interactions, children selecting their own activities, teachers supporting children’s interest and extending the activity that the child had selected, if appropriate. There was a really strong emphasis on involving children in designing their activities and planning their own time.
Abecedarian was also an early-care and -education program. This program took place in the 1970s in Chapel Hill, N.C. The majority of participants in the program were African American as well.
Abecedarian offered full day, center-based intensive early-care and -education services, including some medical care, starting when children were infants and continuing until children entered kindergarten. There’s a school-age component to Abecedarian, but I won’t talk about that since we’re just focusing on early care.
Abecedarian also had a random assignment design, where 111 children were part of the study—57 of those during infancy were assigned to the treatment condition. And much like Perry, the program was designed to be child-centric, focus really on close teacher-child relationships, frequent conversations between teachers and children, and children’s selecting their own activities.
And it’s a really important thing to know that both of these programs were small programs, run by researchers. And both programs, importantly, were evaluated using an experimental design and followed these children into adulthood. And both of these programs offer lifelong benefits of participation, including in domains like educational attainment, economic outcomes, and overall health and well being.
You mentioned both of them being designed and also run by researchers. Is that different from the more recent program evaluations?
Yes, and it’s really important to think about. The programs that are operating now, these are at-scale statewide, or citywide programs serving many more children, operating across many more sites, and locations.
What are some of the other ways that they might differ from the programs with really strong results in the past?
This is kind of the central question of our paper: why we’re seeing different results than these demonstration programs or earlier preschool programs in general. We really try to discuss these common theories, review the evidence behind them. We think scale is one reason.
One of the potential reasons mentioned for decline in program effects in the paper is improved counterfactual conditions. Can you explain what that means?
Since the 1960s, we’ve seen a dramatic increase in the safety net for low-income families. We’ve seen increases in funding for social support, such as income support, like TANF [Temporary Assistance for Needy Families], nutritional support through SNAP [the Supplemental Nutrition Assistance Program], increases in education spending on a per-child basis. We’ve also seen declines in infant and child mortality, suggesting these improved conditions medically and/or environmentally for children and families.
While these social supports may not be universal, children that are not attending preschool now likely have more access to, and potentially higher quality services, than they likely did 50 or 60 years ago. So this is a really good development, and it’s also likely that this is one of the reasons why we’re not seeing the same type of effects from our current preschool programs.
Does this mean that preschool isn’t worth the investment—or that the form of preschool should change?
I think preschool still plays a really essential role for most parents of young children. Working parents really rely on preschool; over half of 3-5-year-olds are attending a preschool center. Our question is really, how do we ensure that these experiences are best supporting their short- and long-term development?
The other kind of theory that we put forth as a really good area for future research is change in instructional models. We really need to understand what’s happening in these classrooms: What those instructional modes, practices, content, structure, are that can best promote children’s development. And this is an area that’s really ripe for future research, specifically experimentation, and just understanding more about what’s happening in those classrooms and what is really effective for children’s development.
You noted that now there is more academic focus in preschool. How do we understand that difference?
This is an area where we need really good, future high-quality research. We have evidence that children are receiving increased academics in the early grades, maybe receiving similar instruction in kindergarten, maybe being part of a classroom where there’s lots of large-group time that could be developmentally inappropriate—[that could] result in disengagement or boredom. But again, this is all an area that we need to explore more and learn specifically about how the mode of instruction and content being taught is related to children’s outcomes, particularly medium- and longer-term outcomes.
You and your co-authors note that funding levels for preschool are lower than in the past. How might that affect outcomes?
Perry and Abecedarian were smaller demonstration programs, very intensive research design. At scale, you know, programs have to serve a lot more children across a lot of different sites. It may not be really surprising that the funding levels differ. So it maybe is not as surprising that we see differences in impacts from our current programs, compared with the demonstration programs.
And it may not be reasonable to expect our programs to produce those same results, especially given changes in the safety net, generally. This is an area where future research can really provide important insights on how and what specifically to invest in: for example, funding for increased teacher salaries, improving teacher-child ratios, purchasing additional classroom materials, etc.
Is the answer more funding across the board, or do we not know necessarily which levers might be most important to pull?
It’s a question of where to invest, and that’s where that future research can show us: What specifically about classroom experiences are effective? And what is effective in both the short, medium, and long term?
The pandemic saw a dip in the number of children in preschool programs. How might this change in enrollment affect the research landscape?
You’re completely right; we saw this decrease in preschool attendance, and it’s beginning to rebound, but not quite yet back to the pre-pandemic levels. This is also coupled with the teacher shortage. I think it’s reasonable to believe that the preschool landscape is at least slightly different than prior to the pandemic.
Given these changes, I think we have to be super thoughtful about what we evaluate, especially if our programs are not over-enrolled, which doesn’t give us an opportunity to use lotteries for admission as a way to evaluate programs.
But it really could give us a chance to examine what practices are being implemented, conduct more descriptive studies on classroom practices to better understand what’s occurring in programs, and focus on evaluating practices that do have a strong research base.
A version of this article appeared in the January 31, 2024 edition of Education Week as Preschool Studies Show Lagging Results. Why?