Yesterday, I explained that I don’t think it makes sense to think and talk about the evidence on pre-k in the same way that we think and talk about the evidence on specific instructional and pedagogical innovations.
Instead, I’d suggest that the body of evidence for pre-k is more comparable to the body of evidence on charter schools. In both areas, we have a strong body of evidence that specific models--such as the High/Scope Perry Preschool and the KIPP network of charter schools--"work.” But we also have evidence showing that the broader range of early childhood or charter school operators are something of a mixed bag. When it comes to charter schools, a recent study by the Center for Research on Education Outcomes (CREDO) at Stanford found that only 17 percent of charter schools nationally outperformed their district school peers, 37 percent did worse, and the remainder were about on par with their district peers. When we look at the larger body of evidence on publicly funded preschool, we see similarly mixed results. While there is evidence that large scale state pre-k programs in New Jersey, Tulsa, and New Mexico are producing benefits, these programs are not identical with one another or with pre-k programs in other states. Further, in at least New Jersey, which I know the most about, there is variation in quality among different pre-k providers even as the program overall produces positive results. National studies of state pre-k programs have found low levels of average instructional quality across these programs. We also need to take into account the disappointing results from the Head Start impact study, which found that, although Head Start does produce learning gains for participating youngsters, these gains disappear a few years after children enter school.
Now, does admitting that there is some mixed evidence here discredit the value of charter schooling or pre-k as a strategy to improve educational outcomes? NO! Why? Because charter schooling and publicly funded pre-kindergarten are both ultimately structural reforms that expand the boundaries of public education and create new spaces and opportunities for educators to serve children. Charter schools do this by allowing organizations other than school districts to operate public schools. Publicly funded pre-k does this by allowing districts and early childhood providers--both existing and new--to receive public funds to serve 4-year-olds. The organizations that fill these spaces and take advantage of those opportunities will vary in quality, with some providing high-quality services and others not.
The question, then, becomes not whether pre-k or charter schools “work.” Rather, once you decide--based on a variety of types of evidence, including empirical research and theory--that allowing innovation and choice in public education, or providing education to 4-year-olds is a desirable thing to do: How do you design structural arrangements so as to maximize the number and reach of the highest-quality providers, and minimize the number of low-quality providers? I’ll close this series out tomorrow by talking a little bit more about that, and why the issue of whether pre-k is a specific intervention or a structural approach has policy implications that extend beyond how we talk about research.
First, though, two caveats to the above discussion: (1) The discussion about pre-k is different from the discussion about charter schools in that charter schools in most places offer an alternative to an existing public education system that, while often producing suboptimal outcomes does in fact offer universal access to educational services to all children. That’s not true in pre-k, where significant numbers of children, particularly low- and moderate-income children, are not getting any early childhood education services at all. There is an argument, which Jens Ludwig and Deborah Phillips make compellingly here, that simply extending early childhood services to poor children who do not currently receive them can carry benefits for those children even if those services are not particularly good. This does not mean, though, that we should not try to ensure that we do not design pre-k arrangements to be as efficient and effective as possible in serving children given existing resource constraints. (2) There is evidence that the level of resources and quality of inputs in many settings where young children are currently served is simply too low to meet even very basic conditions for quality. To the extent this is the case, simply increasing resources or raising standards for inputs may produce improvements in quality and child outcomes--and is probably necessary to do so. Individuals who approach pre-k debates from a K-12 reform perspective and fail to understand this reality risk looking foolish. Again, though, we should seek to ensure that efforts to increase resources and/or raise input standards in pre-k settings are done in the most efficient and effective ways possible.