A new study of pre-K came out of Tennessee this week that reinforces the need for a federal initiative to support state pre-K programs along the lines proposed by the Obama Administration. No study informs policy on its own, so my purpose here is to look at what the Tennessee study adds to the larger body of research. The research as whole finds that observed preschool program impacts tend to decline after school entry. This is not entirely “fade-out.” It also results from “catch-up,” in part because of compensatory efforts by the schools for children who do not attend preschool. Some of these are obvious, as when children who do not attend quality pre-K have higher rates of grade repetition. Other efforts such as increased individual attention from teachers or more classroom time devoted to review of the basics are less easily measured. Such later compensatory efforts are expensive, however, which is one of the reasons that high quality preschool has been found to be a sound economic investment.
Researchers at Vanderbilt University are conducting a rigorous evaluation of Tennessee’s state-funded pre-K program in a randomized trial, considered by many to be the “gold standard” approach to measuring program impact. The ultimate goal is to evaluate the effects of Tennessee’s voluntary pre-K program when children take the state mandated third grade tests. Of course, it takes time for children to get from preschool to third grade, and in the meantime the researchers have been administering tests to a subsample from the larger study to track progress. This subsample is in two waves, and they now have results from the end of pre-K into kindergarten and first grade. What did they find?
Initial results in Tennessee at the end of pre-K were modest. Compared to results from Head Start’s national impact study, effects were somewhat larger for literacy but not mathematics on the commonly used “Applied Problems” test. When the Tennessee results are compared to estimates from some other recent studies of public pre-K, they appear to be on the low side. For example, studies of pre-K in Boston, Tulsa, and New Jersey find larger impacts on literacy and much larger effects on math (again comparing the same Applied Problems test). Methodological differences might explain some of the difference in findings, but it seems fair to say that Tennessee’s state pre-K did not produce large gains right out of the gate, especially in the deeper, more difficult domains of language and mathematics.
It is no surprise then that Tennessee pre-K’s estimated effects at kindergarten and first grade also resemble those of Head Start. No persistent effects on achievement were apparent. This may be disappointing, but it is not unexpected. Programs that have modest initial impacts have even smaller lasting impacts. Nevertheless, the Tennessee program was found to reduce grade repetition at the end of kindergarten. In the sub-study the effect was a reduction from 6 to 4 percent. Importantly, the researchers were able to evaluate impacts on grade repetition for the full randomized trial sample and found a reduction from 8 to 4 percent. This effect is 50 percent larger than the estimated effect on grade repetition in the substudy which produces the estimated effects on achievement. As the substudy sample was compromised by differential participation rates between treatment and control groups, this could reflect a bias that has been introduced into the subsample analyses. If so, the effects on achievement may be similarly underestimated.
Yet, taking the Tennessee results at face value produces a picture that is very much aligned with the research as a whole. Modest programs produce modest achievement advantages that are smaller in subsequent years, in part because schools compensate children who lag behind. For this reason, other benefits are apparent in later years even while achievement gains decline. The grade repetition results are just one manifestation of costly efforts schools make later for the control group children to help them catch up. And, programs that produce larger initial gains do not see their achievement impacts erode completely.
Program’s like the Chicago Child Parent Centers and New Jersey’s Abbott program that produce larger initial gains across broad domains of learning maintain substantial gains through the school years. However, these more effective programs have accountability and continuous improvement systems that ensure quality in the classroom. The quality of Tennessee’s pre-K is unknown as no system is in place to measure it, much less improve and guarantee quality. Like many other states, Tennessee has emphasized expanded enrollment over adequate funding. NIEER has estimated that pre-K funding in Tennessee is more than $2,000 per pupil short of what is required to provide high quality preschool education in that state.
Finally, it seems likely that universal pre-K programs have an important advantage in producing long-term impacts. When the vast majority of children attend pre-K, the entire kindergarten class gets a boost—children learn more from each other and teachers can raise the average level of instruction and have fewer children who need special attention because they are far behind. In New Jersey’s Abbott districts, preschool was the leading edge of education reform that pushed richer content up through all the subsequent grades.
There are important lessons for Tennessee and the nation here. If Tennessee is to get value for its pre-K dollar, it must ensure that programs are of sufficient quality. An accountability and continuous improvement system is a prerequisite for quality, as is adequate funding for those being held accountable. Expanding access to pre-K for all children is likely to improve program performance and the extent to which program effects “stick.” Federal policy can help assure that states like Tennessee take these steps by offering matching funds for quality enhancement and expansion of access to high quality programs. Whatever policy emerges from Congress in response to the President’s proposal must retain this dual focus on raising quality while expanding access. The real high cost policy is failing to invest in programs capable of producing large long-term gains in achievement and other domains of child development.
– Steve Barnett, Director, NIEER