A national report suggesting that charter school students lag behind their counterparts in regular public schools touched off a late-summer tempest among proponents and critics of charter schools.
The study, produced by the American Federation of Teachers and the subject of the lead story in The New York Times on Aug. 18, was made more controversial by its implicit suggestion that the U.S. Department of Education had tried to bury federal test data that reflect poorly on such schools.
See the accompanying table,
Enrolling 800,000-plus children in 38 states, charter schools are public schools that operate outside the bounds of school district authority鈥攐ften in exchange for guarantees of improved academic achievement. They are a favored education improvement strategy of the Bush administration, which supports them with grants and a Web site. The federal No Child Left Behind Act also points to charter schools鈥攏ow numbering about 3,000鈥攁s an important option for children in schools that fail to improve test scores.
Federal education officials, for their part, denied they had held back any data on charter schools, and they joined in a barrage of criticism of the AFT report.
In addition, an ad hoc group of 31 academics last week signed on to a full-page advertisement in The New York Times disputing the union鈥檚 report. The Aug. 25 ad was paid for by the Center on Education Reform, a Washington-based group that supports the charter movement.
Read the full report from the .
See also the North Carolina study, from Professor Helen F. Ladd at Duke University. (Both reports require .)
But there was no denying that the study, the first to assess children鈥檚 progress in charter schools nationally on a common set of tests, had made a splash in the continuing national policy debate over alternatives to traditional public education.
鈥淭he yellow light is up for charter schools now,鈥� said Priscilla Wohlstetter, a researcher from the University of Southern California, in Los Angeles, who has studied the startup schools. The study, she said, is 鈥渃ertainly a reason to make sure authorizers [of charter schools] use caution in doing their job.鈥�
Other longtime observers of charter schools saw in the findings a harsher indictment of the rhetoric that has helped fuel the nationwide movement.
鈥淭his was all based on how market forces would improve education, and now we know we really have no evidence of that,鈥� said Amy Stuart Wells, a sociology professor at Teachers College, Columbia University.
Delays Prompt Study
Once welcoming of charter schools, the AFT has grown critical of them in recent years. Its new analysis draws on data from the National Assessment of Educational Progress, a barometer of student achievement that is mandated by Congress. In 2003, for the first time, federal officials collected data on a nationally representative sample of 167 such schools as part of that assessment. They put the scores online in November, along with the regular state-by-state results.
But the Education Department never advertised the figures鈥� availability. In reports to the national board that sets policy for NAEP, officials said they planned instead to do a more finely grained analysis of the data and publish the findings in January 2004鈥攁 date that has since been moved to the end of this year.
The delay prompted union analysts to mine the data themselves. They found that 4th graders attending charter schools performed about half a year behind students in other public schools in reading and mathematics. In 8th grade, charter school students trailed in math, but for reading, the differences were not statistically significant.
Those patterns remained when researchers adjusted the numbers to account for the higher proportions of poor students who attend charter schools and for the fact that the schools are clumped in inner cities, where achievement is generally lower.
Students in charters and regular public schools scored about the same, however, after researchers controlled for differences in the racial makeups of the schools. Likewise, achievement gaps between poor students and their better-off peers were wide in both charter and traditional public schools, the report says.
鈥淥nce everyone quiets down and takes a good, hard look at this, I hope they鈥檒l say, 鈥楬ey, we鈥檝e got to do better here for the sake of the kids,鈥� 鈥� said Bella Rosenberg, a co-author of the study and a special assistant to the president of the AFT.
Critics Fire Back
Criticism of the report came swiftly, though, from many quarters.
U.S. Secretary of Education Rod Paige, three national groups that support charter schools, various state charter school associations, and the Republicans who chair both of the congressional committees that oversee education called the findings misleading.
In addition, an ad hoc group of 31 academics last week signed on to a full-page advertisement in The New York Times disputing the union鈥檚 report. The Aug. 25 ad was paid for by the Center on Education Reform, a Washington-based group that supports the charter movement.
But another, more sophisticated study released amid the furor over the AFT report reinforced some of the union鈥檚 conclusions. That analysis, based on six years of test-score data gathered on 6,000 North Carolina schoolchildren who had been enrolled in both regular public schools and charter schools, concludes that the academic gains that those students made in charter schools were smaller than the gains they made in regular schools.
The study is being reviewed for publication in a peer-reviewed economics journal, but the authors, Robert F. Bifulco of the University of Connecticut in West Hartford and Helen F. Ladd of Duke University in Durham, N.C., released it last week to weigh in on the national debate.
The problem with the union study, its critics said, meanwhile, is that charter schools may start out with lower achievement because they attract students who are unhappy in their regular schools.
They also said the study drew on a too-small sample of schools鈥攅nrolling only about 1 percent of charter school students nationwide鈥攁nd failed to distinguish long-standing charter schools from startups or to account for variations among charter schools.
鈥淚t鈥檚 wrong to think of charter schools as a monolith,鈥� Secretary Paige said in a statement to the press. 鈥淭here are schools for dropouts, schools for students who鈥檝e been expelled, schools serving the most economically disadvantaged families.鈥�
鈥淔or this to have taken on a life of its own,鈥� Jeanne Allen, the Center for Education Reform鈥檚 president, said of the AFT study, 鈥渢ells me there are politics at play, and some people are looking for bad news.鈥�
A better way to measure charter schools鈥� impact may be to look at the gains students make, said Tom Loveless, a scholar at the Brookings Institution in Washington who has also studied such schools. His own 2003 study, looking at 569 schools in 10 states, suggests that, even though charter school students start out behind their peers in other public schools, their test scores rise faster.
鈥淭hat鈥檚 a terrific argument to make,鈥� countered Ms. Rosenberg of the AFT, 鈥渂ut I don鈥檛 hear anyone making it when the results come out for regular public schools.鈥�
She also noted that the NAEP state-by-state assessments do not include special analyses to account for demographic and socioeconomic differences among states.
Still, Darvin M. Winick, the chairman of the National Assessment Governing Board, which sets policy for the national assessment, said he questions whether the AFT researchers should have based their conclusions on raw data from a limited pilot study.
In addition, he said he had no reason to suspect that the Education Department had held back findings.
Even so, he said, 鈥測ou could argue鈥攁nd I would not quarrel with you鈥攖hat it would be better if all these reports [from the department] were more prompt.鈥�
