I did a “You Guest It” article on the Oklahoma Council of Public Affairs blog a few months ago. Here’s the text:
My daughter will be attending Tulsa’s excellent universal public prekindergarten program next year, at an already excellent elementary school. I am thrilled to be able to take advantage of this free program. But how excellent is this program, really?
In President Obama’s 2013 State of the Union address, he called for universal prekindergarten, and cited Oklahoma’s program. This week, the Washington Post ran an opinion piece lamenting the lack of momentum from that speech, citing—as many do—research indicating increased cognitive results and lifelong positive effects from a high-quality preschool program. Tulsa’s program has been studied by Georgetown researchers and by all indicators, according to them, meets the benchmarks of this type of quality program, not only by what the program provides but by the results.
I challenged this assertion in a piece written for The Federalist in 2013 (“The False Promise of Universal Pre-kindergarten”). The criteria for defining a quality program depend on a series of benchmarks related to class ratio and teacher degrees, and looking at the results of these programs, the output is suspect. Most commentators who cite the remarkable long-term effects of preschools are speaking of very intensive and small historical programs such as the Perry Preschool Project, the Chicago Parent-Child Center, and the Abecedarian experiment. Those programs are not the same in almost any way as larger universal programs such as the ones found in Oklahoma, Tennessee, Georgia, and the city of Boston.
In October 2014, the Cato Institute published a study by David Armor, a professor emeritus of public policy at George Mason University and the author of Maximizing Intelligence (Transaction Publishers, 2003). Armor took an even more in-depth look at the inputs and outputs related to the evaluation of preschool programs and found definitively that contemporary programs extolled for their high quality are not being evaluated in a way that accounts for discrepancies in control groups and treatment groups. In Tulsa, for example, the researchers used Regression Discontinuity Design (RDD), which is considered flawed and non-experimental and “biases outcomes upwards” by not considering students in the treatment groups that drop out of the preschool programs. The control group in Tulsa is from the next year’s cohort and has just started school, and therefore the control group has all the students with no dropouts.
The Tulsa results, as Armor highlights, are “truly staggering”; the verbal skills for the treatment group were almost a full year of schooling ahead of the control group. This result, he says, is larger than the result even from the intensive historical programs such as Perry and Abecedarian—and is likely inflated.
When evaluations of contemporary preschool programs are undertaken instead with randomized experimental designs—considered the “gold standard” for education evaluation—results are far less extraordinary. The Head Start Impact Study originally had discrepancy problems but a special analysis by Peter M. Bernardy, which accounted for these discrepancies and used correct methodology, found that any positive results began fading by kindergarten and first grade. Similarly, an evaluation of the Tennessee pre-K program found this fade-out. One reason for waning results at this grade level are the intense cognitive development that occurs during these years of children’s lives—which indicates the quality of kindergarten and first-grade programs may be much more influential than the quality of prekindergarten.
What does this mean for Oklahoma? As I mentioned in my Federalist piece, studies do show more benefit from preschool for disadvantaged children than for children who have educated parents. As I stated, “If you are an educated parent who spends time talking and learning with your children, your child probably will not gain any extra educational benefit from preschool. Oklahoma makes the classic mistake of assuming the government can do a better job of providing for our children than parents.”
Preschool for disadvantaged children may have its benefits, but that does not augur for a state-run program that catches all children. If Oklahoma is truly concerned about how the state is failing young children, offer Education Savings Accounts (ESAs) to the better off, and focus instead on reforming educational options for the most vulnerable. That may include bolstering Head Start statewide with private and non-profit partnerships, as it does in Tulsa. It may even include some version of the state-run prekindergarten, but instead a targeted intensive program for those who fall below the poverty line