Has Preschool Been "Oversold"? A Reponse to Russ Whitehurst
Russ Whitehurst says that the results of the recently released Head Start Impact study--whose results were disappointing, if not surprising to those who've been following the study--call for "hard-headed" thinking about the performance of early childhood programs generally, and Head Start in particular. There's a lot I disagree with in Russ's piece: I think the Head Start Impact study results--which found that children in Head Start learned more than a control group while in Head Start, but performed no better by 3rd grade--are really discouraging and suggest a need for major improvements. But I don't think the findings--which showed the kids did make gains while in Head Start--justify calling the program a total failure. I'd also dispute Russ's contention that preschool has been "oversold," is also off-base. Yes, early learning advocates sometime come across like Billy Mays, and our discourse and children would be better served by a much more honest discourse about the massive execution and delivery challenges to delivering quality pre-k at scale. But the reality is that the evidence base about the beneficial impacts of quality pre-k--particularly for low-income children--is one of the most robust anywhere in education. We now have not only the High/Scope Perry Preschool and Chicago Child Parent Center studies that used rigorous randomized-controlled methodologies and found long-term results for children. We also have studies that find positive results from an increasing number of state pre-k programs operating at scale with modest resources--including evidence from states as disparate as New Jersey and Texas of pre-k educational benefits that last into the elementary grades. And given that the U.S. still doesn't have anywhere near the widespread access to publicly funded early childhood we see in other developed countries, the evidence for pre-k seems more "unpurchased" than "oversold."
The problem we have in early childhood isn't an overselling problem--it's an execution problem. At that should inform how we respond to it.
In my next post, I'll explain why the approach Russ proposes is not just hard-headed, but wrong-headed.