When it comes to educating students, do private schools have an edge? This summer, dueling reports weighed in on this question, provoking yet another skirmish in the long-running school choice wars.
In July, the U.S. Department of Education’s National Center for Education Statistics released a study comparing the reading and math performance of fourth- and eighth-grade students in public and private schools. Using test results from the 2003 National Assessment of Educational Progress (NAEP) — given to a sample of students in public and private schools — researchers found that private school students performed significantly better than did students in public schools. But the story doesn’t end there: Researchers then chose to “put test scores into context,” and adjust for “race, ethnicity, income and parents’ educational backgrounds” in order to “make the comparisons more meaningful.” The new data told a completely different story.
“Adjusted” results revealed that private and public school students performed similarly in fourth-grade reading and eighth-grade math. However, public school fourth-grade math scores were significantly higher than those in private schools, while eighth-grade private-school reading scores were better than those in public schools. While the report’s executive summary urged caution in interpreting results, suggesting that “an overall comparison of the two types of schools” would be of “modest utility,” school-choice opponents speed-dialed reporters, barely suppressing their jubilation.
In response to the study, Reg Weaver, president of the National Education Association, the nation’s largest teachers’ union, trilled a paean of praise for public schools, claiming they were “doing an outstanding job.” Weaver heralded NCES results as evidence that poor families did not need school choice. Howard Nelson, a researcher for the American Federation of Teachers (another educators’ union) suggested NCES data provided proof that “private schools are not the silver bullet that voucher advocates say they are.” Even the mainstream media fixated on “adjusted” data, with the headlines, “Public Schools Perform Near Private Ones in Study,” (New York Times) and “Long-Delayed Education Study Casts Doubt on Value of Vouchers” (Wall Street Journal).
Fortunately, other researchers began to take a closer look at NCES’ “adjustments,” questioning the study’s methodology and conclusions. In a second study released this past August, Harvard University’s Paul Peterson found that the benefits shown by NCES for public schools were due entirely to flawed “adjusting” by researchers. Employing a different (and more accurate) method of measuring student characteristics, Dr. Peterson and his colleague Elena Llaudet found a “consistent, statistically significant private school advantage.”
According to the Harvard researchers, NCES’ income “adjustments” relied completely on student participation in federal programs for disadvantaged pupils (like Title I). This is inherently problematic since public schools participate in these programs at much greater rates than their private counterparts.
According to Peterson and Llaudet, during 2003-04, just 19 percent of private schools took part in Title I compared to 54 percent of public schools. As a result, NCES clearly underrepresented the number of disadvantaged students in private schools, skewing results in favor of public schools.
Second, it seems clear that private schools have a leg up on public schools academically. But that doesn’t negate the fact that lots of public schools are doing a commendable job. Parents know this and seldom demonize one whole “system” as intrinsically bad. After all, no two schools are the same, and this is true for both public and private school groupings — a fact affirmed by the authors of the NCES study.
Where does that leave us? When it comes to choosing a school, parents have the requisite smarts and savvy to make the right decision. That won’t ever change, no matter how you run the numbers.
Lindalyn Kakadelis is director of the North Carolina Education Alliance.