State legislators sometimes are skeptical about the claims of any program’s advocates. So, when early childhood program advocates claim great results for these programs, state legislators may be resistant. Why should they believe the claims of advocates for early childhood programs more than the claims of other programs’ advocates?
One reason is that the evidence for program effectiveness is stronger for early childhood programs than for almost any other publicly supported program. The evidence for early childhood programs is stronger because these programs have been studied with data that has good “comparison groups” for program participants. The underlying reason is that because access to early childhood programs is currently limited, we can either create or find good comparison groups.
So, for example, we know more about whether preschool works than whether 3rd grade works. For preschool, we have two very good random assignment experiments with long-term outcome data (Perry Preschool and the Abecedarian Program). Because not everyone gets access to free preschool, we have been willing to run experiments in which some individuals are lucky enough to get free access to preschool, and other randomly chosen individuals are denied access.
We haven’t been willing to do this random assignment experiment for 3rd grade, and we are unlikely ever to do so. It is easy to see that it would be difficult and morally problematic to flip a coin and assign some children to a “control group” that would be randomly assigned to not attend 3rd grade.
The restricted access to early childhood programs also helps us run good “natural experiments”. For example, because North Carolina’s Smart Start and More at Four programs were gradually implemented in various North Carolina counties, it was possible for Duke University researchers to do a rigorous study that linked the implementation of these programs to lagged effects on elementary school test scores in various counties. It is rare for such a large intervention as these programs to be only made available to some geographic areas. For example, because everyone in North Carolina has access to 3rd grade, we can’t use geographic variation in access to 3rd grade to evaluate its effects.
Similarly, for the Chicago Child Parent Center program, because the program was only made available in some low-income neighborhoods, and not all, researchers have been able to use these variations to neighborhood access to rigorously evaluate the long-term effects of CPC.
Early childhood programs have some unusual patterns of access. These programs have enough political support that these programs have been tried at a large scale in some places at some times. On the other hand, as of yet early childhood programs don’t have enough political support to be universally accessible. This unusual combination of large-scale but partial access is inequitable. But from a social science perspective, it has allowed for some powerful and rigorous evaluations of program effectiveness.