Researchers fail, once again, to find educational gains from public spending on preschool.
Yet another study has found that preschool yields no significant educational benefits. After spending £2 billion annually for almost a decade to fund an expansion of preschool to cover all three-year-olds, the United Kingdom had no lasting improvements in educational achievement to show for it, researchers have found.
In a study published in the current edition of The Economic Journal, the researchers found that “the policy led to small improvements in attainment at age 5,” yet “the effects are short-lived for all groups, becoming essentially zero by age 7.” This finding echoes those of many U.S. studies, which consistently find that any educational gains from preschool fade out over time.
Nevertheless, U.S. preschool advocates from President Obama and presumptive Democratic presidential nominee Hillary Clinton to New York City mayor Bill de Blasio continue to press for the adoption of universal pre-K, or “UPK” in the lingo, as a way to elevate student achievement.
Well, the only thing “universal” about preschool is its failure to produce educational gains. Preschool programs of all shapes and sizes, both in the U.S. and in the U.K., have failed in this goal. Preschool programs do, however, serve an essential function in providing day care so parents can work.
When the U.K. program was introduced in 2000, about 82 percent of British parents already were placing their three-year-olds in preschool or day care, and about half of them were in private programs for which parents paid in full. Self-funders in the private sector took advantage of the program’s offer of government funding (£1,130 per child in the initial year) with the result that government-funded places increased from 37 percent to 88 percent of total enrollment by 2007, the end year of the study period, when national preschool attendance by three-year-olds hit 96 percent. There was no expansion of publicly run preschools, only of publicly subsidized private programs.
Obviously, a mere change of the source of tuition funding wouldn’t generate educational benefits. A modest and gradual 14-percentage-point increase in participation would be unlikely to change the average achievement of the group, especially since it didn’t involve much change in the schools themselves — none in the public sector and only incremental increases in enrollment in the private sector.
Accordingly, while a central aim of the program was to elevate educational achievement, it is not surprising that it did not. Indeed, the lead researcher, Dr. Birgitta Rabe, told me that she was not surprised that it did not.
Unfortunately, advocates of preschool generally refuse to accept evidence contrary to their beliefs.
Unfortunately, advocates of preschool generally refuse to accept evidence contrary to their beliefs. They usually complain that programs fail because of insufficient “quality.” They argue that, if “quality” programs were provided, it would enable preschool to achieve its inherent, but heretofore unrealized, potential to deliver real educational impact.
Even Rabe and her co-authors fall into this trap, blaming the failure of the studied program on its exclusive reliance upon an “increase in private provision, where quality is lower on average than in the public sector.” They posit that preschool’s efficacy should logically be assumed: “A priori, one might think that early education has benefits for children’s development.”
Acting on this supposition, they undertook an additional study of children who attended preschool in the academic years 2008–10, as distinct from the 2000–07 cohort that was the subject of the main study. This additional study, using a key data set unavailable before 2008, is covered in the Economic Journal article as well.
In the additional study, the authors find that “quality,” defined as the “presence of staff qualified to degree level,” raised children’s performance at age five on the U.K.’s national academic performance test, the Foundation Stage Profile, “by around 2-3% of standard deviation.” The authors characterize this as “statistically significant.”
The authors’ finding is hardly a slam dunk. First, not all statisticians would consider 2 to 3 percent of standard deviation to be meaningful. Second, there’s no later-age data yet available for children who were three-year-olds in the years 2008–2010. There’s no telling whether the gain would survive the famous “fade-out” effect found in most studies of preschoolers, including the authors’ main study. The authors acknowledge elsewhere in the article that “fade-out is a common empirical finding in many studies analyzing early educational interventions.”
The authors reference other research allegedly supporting their belief that preschool can improve school achievement, but they do not mention the $9 billion-per-year U.S. Head Start preschool program for low-income kids. During the Bush and Obama administrations, the Department of Health and Human Services conducted a multiyear study of Head Start, concluding in December 2012 that the program had “no significant impacts” on educational achievement.
Instead, the authors cite the so-called Perry preschool project as “the strongest evidence in support of early years interventions.” Indeed, Perry is mythic in the pre-K advocacy community. It is cited frequently — and wrongly — as evidence of pre-K efficacy. It is so legendary and so foundational in the UPK movement that no discussion of pre-K is complete without comment on Perry.
Perry was an extremely small (too small to be statistically valid) 1960s experiment (too old to be relevant today) involving 123 “at risk” low-IQ (70–85) children from one poor minority neighborhood in Ypsilanti, Michigan. The kids were divided into a study group that received two years of “high-quality” preschool and a control group that did not.
The Perry study group had the same 2.5 hours of preschool instruction each weekday as was funded by the U.K.’s UPK program. However, it employed four certified public-school teachers for 20–25 kids — the highest “quality” imaginable and something patently unaffordable on a larger scale. Also, there were weekly 1.5-hour teacher visits to the home and monthly teacher meetings with the parent group.
So what did Perry achieve? A large claimed gain in IQ of 15 points, but the gain disappeared by the end of third grade.
So what did Perry achieve? A large claimed gain in IQ of 15 points, but the gain disappeared by the end of third grade. The famous “fade out” again.
There was a vast difference between the Perry study group and its control group. Perry pre-dated Head Start. The control group experienced no intervention, just life in a poor minority neighborhood. One would have thought that the Perry study group’s intensive enrichment should have achieved some lasting educational advantage compared with a control group that was left so destitute Rabe et al acknowledge as much, writing, “One would expect the effects of the policy to be positive if the quality of the counterfactual is low and/or if the quality of the additional childcare is high.”Given this standard, one wonders why they cited Perry as evidence of the educational efficacy of preschool. The only educational gains found in the Perry study group were less time spent in special education at unspecified grade levels and a lower high-school drop-out rate. These are imprecise measures at best, occurring randomly and/or many years afterward. The only credible evidence of preschool efficacy would be better educational test scores consistently thereafter: better scores in elementary, middle, and high school.
The U.K. researchers are self-contradictory also with respect to the effects of “quality,” which they define as having three basic elements: more hours per day, a lower child-staff ratio, and more “qualified” staff. They claim “quality” is important, yet they also refer to “the school effectiveness literature where interventions such as class size reductions and improvements in teacher quality are found to diminish over time.”
Most troubling is the researchers’ comment about the Conservative Party’s announced plan to expand the U.K.’s decade-and-a-half-old universal preschool program to fund 30 hours per week for working parents The researchers say the plan “is unlikely to generate substantial positive impacts for affected children’s outcomes.” This is absolutely correct. Expanding the program might, however, enhance parents’ employment prospects, since most jobs require that parents work at least 30 hours per week.
In fact, Dr. Rabe and her colleagues are now engaged in a study of the impact of longer school hours for young children on parental labor-force participation. Unsurprisingly, the initial findings are that longer hours do lead to increased participation. Given the lack of evidence that preschool generates any measureable and lasting improvement in educational outcomes no matter the setting, the format, or the “quality” level, perhaps emphasis should be placed on its role in assisting parents to work, rather than on its ever-elusive educational benefits to children.