Do Charter Schools Improve Behavior?

By Jay Mathews
Washington Post Staff Writer
Tuesday, October 2, 2007 9:40 AM

More than 1.2 million students are attending more than 4,000 charter schools in 40 states and the District, up from 200,000 students in just 600 charter schools a decade ago. Charters are hot commodities, the public school equivalent of hybrid cars or left-handed relief pitchers. But many people are puzzled why that is so.

Charters are independent public schools that don't have to follow many school district rules. They can usually choose their own curriculums and hire and fire staff without dealing with the teachers union. Those freedoms are enough to win the support of some parents, but most, I think, also want to know what such schools would do for their children.

That is where the allure of charters becomes harder to figure out. Several studies have shown that, on average, they don't raise student achievement more than regular public schools with students of similar backgrounds. Yet many charters, even some with mediocre academic records, get lots of applications. What is going on?

Scott A. Imberman, an economist at the University of Houston, has come up with some interesting answers, based on his review of testing, attendance and disciplinary data from a large, unidentified school district that has experienced great growth in its charter schools since 1997. Imberman says he thinks parents who choose charters in this district, which is apparently not in the Washington area, may be "more concerned with discipline, safety and student satisfaction than academic performance."

His data indicate that the anonymous district is finding kids behave better in charters. It's a preliminary finding, and Imberman is cautious about drawing conclusions from one district. But in his study, he says, charter schools "provide improvements in student discipline and attendance with mixed effects on test scores."

The paper, available at, offers a national summary of the status of charter schools and looks at differences between charters that start from scratch and those that are created by converting a regular public school. He examines, for instance, the implications of his finding that students whose behavior improves when they enroll in charter schools tend to lose those good habits if they return to non-charters.

"Even if charter schools generate only temporary performance improvements," he says, "they also tend to spend less money than non-charter schools. In 2002, median per-student expenditures for charter districts were 13 percent lower than in non-charter districts. Thus, if charters provide the same level of long-term performance and cost less money, they still may enhance the efficiency of the education system."

That is an economist's take on school choice. It may not dovetail with what we parents consider important. But Imberman goes further, showing that almost all of the discipline improvements can be explained by the smaller school size and higher teacher-student ratios found in charters.

Charters are usually smaller. That is, in many ways, their best feature. What data we have show that the most likely solution for our massive high-school dropout rate, particularly in big cities, is to make the schools small enough that every student can develop a healthy personal relationship with at least some of the educators at the school and have an incentive to go to class each day.

Imberman's study buttresses the notion that charters in big cities are appealing to parents who don't have much money but do have jobs and ambitions for their children. The charter students in the study were on average from families less wealthy than non-charter students, but they also were less likely to be among the poorest students in the district and less likely to need help learning English. The charter students also had, on average, better behavior, better attendance and better test scores than non-charter students.

Pardon me if this is confusing. Remember, the study looks at what happens when students move from non-charters to charter schools. In that case, their behavior improves but their test-score improvement is mixed. That is different from saying that on average, students in charter schools have better test scores than students in non-charters. My conversations with parents who make these choices indicate that they often do not know the charter school's average test scores, but they do have some sense of what kind of students attend the charter, based on its location and what kind of people have recommended it. They make choices based on where they think their children will be comfortable and safe.

Charters seem to be evolving into a system of free public schools that mimic private schools, but are for families that cannot afford private school. Parents can be fairly sure that the charters will be smaller, and thus more likely to have congenial student-faculty relations. That is fine for those parents, but how hard would it be to create these charter-like conditions in all public schools?

The movement to break up big public high schools into smaller learning communities has had some success but has not reached many schools. It might be time to give it more attention and more money. With schools, smaller is often better, and it should not be just charter and private school families who benefit.

View all comments that have been posted about this article.

© 2007 Washingtonpost.Newsweek Interactive