Executive Summary

Many consider Oklahoma a national leader in pre-K education, but how beneficial are pre-K programs? The empirical evidence is very uneven in scientific quality, especially as compared with the evidence on other education policy issues like school choice. A careful review of the research reveals that the better the studies are in scientific quality, the less likely they are to find benefits. The potential of expanded pre-K to disrupt the parent/child bond must also be considered, especially since the increasing fragility of the household is a leading factor in the perpetuation of poverty. Any large-scale expansion of pre-K would involve large financial costs, doubtful benefits, and the potential for unintended social harm.

I. Oklahoma Pre-K: “A National Leader”

Champions of expanding pre-K often point to Oklahoma. It was one of the first states to spend large amounts of money widening access to free pre-K education, and for a while it had comparatively high enrollment rates.¹ (Pre-K is voluntary, unlike compulsory K-12 education, so enrollments vary considerably.)

Today, Oklahoma isn’t quite the standout it used to be, but that’s largely because some other states have emulated Oklahoma’s embrace of expansive pre-K programs. According to the U.S. Department of Education, 48.5% of Oklahoma’s 3- and 4-year-old children are enrolled in school, placing Oklahoma 18th in the nation; the national average is 47.6%.² The U.S. Department of Education does not report pre-K enrollment separately by age, so including both 3- and 4-year-olds is the only way to make a reliable cross-state comparison using publicly available data. Regarding 4-year-olds specifically, the Oklahoma State Department of Education website claims 74% of Oklahoma 4-year-olds are enrolled in school.³

Pre-K programming costs Oklahoma about $140 million per year.⁴ Expanding pre-K toward “universal” enrollment would increase costs in unpredictable ways. If 3- and 4-year-olds were in view, as would likely be the case in any sustained push for expanded pre-K, enrollment would have to double (at least approximately) to become “universal.” Moreover, regardless of the age targeted, the families now enrolled are the families that choose to be enrolled. Rounding up families that haven’t chosen to be enrolled would have to cost substantially more per student than the existing system does. Innovations in the system designed to save money could, of course, mitigate the added expenses.

New Yorkers don’t typically look to Oklahoma as a model. But when New York City Mayor Bill de Blasio came into office in 2014 pushing a plan to expand pre-K in city schools, he asked Steven Dow of CAP Tulsa County to testify to his city council about what Gotham could learn from Oklahoma. Dow highlighted four aspects of Oklahoma’s 1998 decision to spend large amounts of its limited education funds on pre-K:

• Funding for pre-K is automatic as part of the state education funding system.
• Pre-K teachers are required to have both a bachelor’s degree and a certificate in early childhood education.
• Class sizes are limited; and
• school districts can work with local community organizations to provide services.⁵

The first of these is more important than it may appear to be. Where pre-K funding must be separately approved each year, program growth will be limited. Automatic funding ensures that spending for pre-K comes through on the same conveyor belt as all other education spending. In most years, historically, that has meant funding growth.

II. Empirical Evidence: “Raises Serious Doubts”

Is pre-K educationally effective? When asking this question, it is essential to look at long-term effects. It wouldn’t be surprising if 4-year-olds who are in pre-K have made more progress toward reading than 4-year-olds who aren’t. But what happens a year or two later, as children enter compulsory education? If the children who weren’t in pre-K catch up to their peers while in kindergarten and first grade, such that the differences disappear, pre-K won’t have been worth the investment.

Education research is difficult to do well. It’s hard to sort out which differences in student outcomes are due to a particular educational policy (like pre-K) and which are due to other policies, parental influence, social environment, and other factors. Since we can’t experiment on kids in laboratories to find out what works and what doesn’t, we have to do our best with the limited information available.

Researchers have developed a variety of methods for drawing accurate comparisons. The gold standard is random assignment, the same method used in medical trials. In cases where students apply to participate in a program with limited seats and access is determined by a random lottery, we can compare lottery winners (who got into the program) with lottery losers (who wanted to get in but didn’t) in the same way medical researchers compare treatment (medicine) and control (placebo) groups. When this kind of method isn’t possible, as is the case in Oklahoma’s programs, other methods can still be used, although they’re generally not as scientifically reliable.

Russ Whitehurst, former head of the U.S. Department of Education’s Institute of Education Sciences and now a senior fellow at the left-leaning Brookings Institution, conducted a comprehensive review of decades of research on pre-K programs. He found that the higher the scientific quality of a study, the less likely it was to find sustained benefits from pre-K. Of the 10 studies he identified that included data on later outcomes, six found meaningful benefits and four did not, but the four that found no long-term benefits were better-quality studies.

Specifically, Whitehurst noted that “not one of the studies that has suggested long-term positive impacts of center-based early childhood programs has been based on a well-implemented and appropriately analyzed randomized trial, and nearly all have serious limitations in external validity,” while “the only two studies in the list with both high internal and external validity (Head Start Impact and Tennessee) find null or negative impacts, and all of the studies that point to very small, null, or negative effects have high external validity.”⁶ Internal validity refers to the likelihood that the observed outcomes of participants were really caused by the program. External validity refers to the likelihood that their outcomes are representative of what others would experience, as opposed to being an idiosyncratic effect that we wouldn’t expect to see duplicated for others.

Since Whitehurst conducted that research review in 2014, one additional study has received special attention in Oklahoma. In a 2017 study, William Gormley, Jr., Deborah Phillips, and Sara Anderson found positive results from pre-K in Tulsa visible as late as seventh grade. However, to use the researchers’ own words, the size of the positive effects they found was “rather modest.”⁷ Also, because of data limitations, they had to abandon the high-quality scientific model used in earlier Tulsa pre-K research (regression discontinuity, a method comparable in reliability to random assignment) in favor of a propensity score weighting method that uses student characteristics to predict whether or not each seventh-grade student attended pre-K. This is a much less reliable method—its value is not zero, but it must be considered less informative than the studies we have that are of top quality. Hence this study continues the pattern identified by Whitehurst.

The question we want our research to answer is not simply, “does pre-K have a positive effect?” It is, “does pre-K have a large enough positive effect to make it worth the cost, and how sure are we that effect is real?” It seems plausible, on balance, that pre-K may have a modest positive effect. But we cannot be very sure it does. And any major expansion of pre-K would have to show not only benefits we can have confidence exist, but benefits of large size. Even advocates of pre-K generally avoid claiming that its benefits are dramatic, and the evidence seems to support this tendency.

III. What is Pre-K? “In a State with High Poverty…”

What drives interest in expanding pre-K as opposed to other policies, such as school choice, that have a much more solid track record in the empirical evidence—and that don’t increase taxpayer costs?⁸ One obvious explanation is the opportunity to grow the school system, which enriches education special interests such as teacher and staff unions; the desire of such interests to increase education spending and employment has been the dominant force in making education policy for generations.

However, another plausible explanation is paternalism toward households of low socioeconomic standing. Public advocacy for expanded pre-K often displays an implicit desire to have the government school system replace the role of parents in impoverished households. Oklahoma State Superintendent of Public Instruction Joy Hofmeister has justified expanded pre-K by saying: “A parent is a child’s first, and most important, teacher. While we recognize that a nurturing home is every child’s first classroom, in a state with high poverty, access to early childhood education is crucial to shaping the future trajectory of all learners.”⁹ If she is really concerned with “all learners,” why does she mention poverty? Clearly Hofmeister thinks pre-K exists particularly to serve impoverished households—which suggests its real function is not educational (in which case it would be of value to everyone) but as a substitute for the good parenting that she apparently thinks children of low socioeconomic standing don’t get at home. An even more stark example is Tulsa reporter Megan Allison, who has boldly asserted that Oklahoma should be interested in expanded pre-K because child care costs are a burden on poor families—which makes it very clear that for her the purpose of pre-K is not educational, but paternalistic government care for (i.e., control of) poor households.¹⁰

The fragility of the household is increasingly recognized as one of the primary factors in intergenerational poverty. While households of high socioeconomic standing have largely stabilized and are even recovering some of their earlier robustness after the disruption of the sexual revolution, households of low socioeconomic standing, which lack the social capital possessed by others, continue to face increasing rates of divorce, out-of-wedlock childbearing, and other difficulties.¹¹

One possible response to this crisis is to grow the bureaucratic state and have government employees incrementally replace the role of parents in the lives of children. Hofmeister’s comment on the role of parents, quoted above, suggests she is aware that incremental replacement of parents in the lives of children is what she is advocating. Why else would she bring up the subject? Another approach is to strengthen the social capital of impoverished households in ways that strengthen parents rather than replacing them.

IV. Policy Recommendations

The considerations above suggest the following policy recommendations:

• Avoid major expansions of pre-K given the high costs, uncertain benefits, and potential for disruption of family bonds among households of low socioeconomic standing.
• Take pre-K off the automatic-funding conveyor belt; since pre-K is not mandatory and its benefits are uncertain, it should have to make a case for itself like every other discretionary expense.
• Commission research using top-quality methods such as random assignment or regression discontinuity, to increase clarity about the long-term effectiveness of pre-K. (For the record, the author of this brief would not benefit financially from such research.)
• Consider whether pre-K is well served by regulations on class size and teacher qualifications that directly serve education special interests, but whose educational value is unclear.
• Consider introducing school choice design in existing pre-K programs, to strengthen the freedom and power of parents, and in light of the strong research consensus showing positive effects from school choice. Oklahoma’s existing program permits schools to partner with community organizations; why not allow community organizations to serve parents directly?

Oklahoma was among the first to spend large sums expanding Pre-K; it now has the opportunity to be among the first to ask fresh questions about what approach would best serve children who are approaching school age, especially the most vulnerable.