Charter schools do not work as promised, writes Wayne Au, a University of Washington, Bothell education professor. One study indicates charters are two times as likely to widen achievement gaps as close them, he says.

Share story

As the Washington state Legislature considers a bill proposing the introduction of charter schools, it is important that voters and representatives understand a critical point: While advocates claim charter schools will raise the achievement of African American, Latino and low-income students, there is little evidence to support this claim.

For instance, the study completed by the pro-charter Center for Research on Education Outcomes (CREDO), a research group at Stanford’s politically conservative Hoover Institute, found that charter schools performed worse than public schools 36 percent of the time, performed better 17 percent of the time, and performed no differently the rest of the time. This study raises a question: Are the people of Washington state willing to gamble on an education reform that is twice as likely to widen achievement gaps than it is to close them?

Citing the CREDO study, advocates suggest that we can just replicate “good” charters on a large scale. This has never been done successfully anywhere. To attempt to do so would amount to crafting education policy purely on speculation.

Or take the more recent study of charter schools by Mathematica, done with assistance from the pro-charter Center for Reinventing Public Education, which also has ties to the Hoover Institute. This study has been referenced by charter advocates on statewide Listservs and discussion boards to make a bold claim: Over the course of two to three years, high-performing charter schools can make up an extra year or more of learning.

In actuality, the Mathematica study says this: “The differences between high-performing and low- performing [charters] after two years of enrollment are large enough to be equivalent to a year or more of learning.” It is a simple comparison of high-performing charters to low-performing charters, and it says nothing about charters outperforming regular public schools.

In comparing charters to regular public schools, the Mathematica study does say: “Although overall average two- and three-year test score impacts are positive in all four subjects, they are not statistically significant.” “Not statistically significant” is research-speak for, “there is no difference.” The Mathematica study simply found that charters do not outperform regular public schools, despite the fact that the charters in their study had higher-performing African-American and Latino students, fewer English-language learners, fewer students with disabilities, and smaller class sizes than regular public schools.

Even the success of Knowledge Is Power Program (KIPP) charters, often offered as a model of success, is misleading: According to one major study, KIPP test scores and college-entrance rates have been artificially inflated because they have either kicked out or lost African-American and Latino students at up to three and four times district averages.

Research has revealed other points to consider about charter schools: Charters underserve English-language learners and students with disabilities; they do not keep accurate track of student data, such as who is on free and reduced lunch; their governing boards regularly lack public accountability; they have also reached levels of racial segregation not seen since before the 1954 Brown v. Board of Education ruling that legally ended “separate but equal” schooling — prompting the NAACP to issue a statement in 2010 opposing charter schools.

Yes, the achievement of African-American, Latino and low-income students absolutely demands a fix. Based on the evidence, charter schools are not that fix.

Wayne Au is an assistant professor in the Education Program at the University of Washington, Bothell, and an editor for Rethinking Schools magazine.