Friday, December 31, 2004

In defence of charter schools

Below are some excerpts from a big defence of charter schools on A Constrained Vision

"Amy Stuart Wells, professor at Teachers College of Columbia University and long-time critic of school choice, tries to dismiss those pesky charter schools once and for all. She blames those evil free-market conservatives and their "well-funded think tanks" for this loser of a reform movement.... Even ignoring her gratuitous swipes at think tanks, conservatives, and free markets, there are a lot of problems with Wells's analysis. Three major points.

One, if charter schools are so terrible, why are they expanding so rapidly? Like Wells says, the Center for Education Reform reports almost 700,000 students in 2,993 charter schools in 40 states and the District of Columbia; this year alone, 405 new charter schools opened. Either all those families have been brainwashed by the evil conservatives or they're stupid enough to send their kids to schools that are bad for them or they have found something about charter schools that they like better than public schools. Since we generally don't assume that people are brainwashed or stupid, this point is difficult for charter school opponents to address and Wells hasn't done it adequately.

Two, Wells misrepresents the recent research on charter schools. Contrary to Wells's quick dismissal of disagreement, the methods used in these studies are hotly contested--see, for example, Joanne Jacobs's summary, Eduwonk's round-up of reaction to the latest NAEP study and William G. Howell and Martin R. West's thorough trashing of the AFT study (just a taste: "[O]n a methodological level, the AFT analyses are sufficiently pedestrian to be laughable.")--but even if we all agreed that the methods were fine, Wells gets the punchline wrong! From the Department of Education's NAEP study:

[T]he mathematics performance of White, Black, and Hispanic fourth-graders in charter schools was not measurably different from the performance of fourth-graders with similar racial/ethnic backgrounds in other public schools.

In reading, there was no measurable difference in performance between charter school students in the fourth grade and their public school counterparts as a whole. This was true even though, on average, charter schools have higher proportions of students from groups that typically perform lower on NAEP than other public schools have. In reading, as in mathematics, the performance of fourth-grade students with similar racial/ethnic backgrounds in charter schools and other public schools was not measurably different.

Because charter school demographics are very different from those of public schools, it is not a fair comparison to look just at overall test scores, as Wells seems to do when she declares unequivocally that "charter schools performed more poorly than public schools on the same tests." When you look at similar types of students, charter schools and public schools perform about the same. Wells and other charter school opponents would probably point out that this is not the magic improvement that advocates promised, but why does that matter? Charter schools are cheaper than regular public schools, parents are happier with charter schools, and bad charter schools can go out of business, unlike regular public schools--if the charter schools aren't doing any harm, why not allow them to exist just for those advantages? Charter school opponents have the burden of proof to show that charter schools are actually harming students, and again, Wells has not met that burden.

Three, evaluating charter school performance is exceptionally difficult. From issues like the one above--what is the proper comparison group?--to nitty-gritty econometric debates, there is very little agreement about what are the proper methods. Given the demographic differences, it does seem wrong to compare overall charter school performance to overall public school performance, as Wells does, and it does seem wrong to look only at one year of data, as the studies she relies on do. And there are many more difficult issues that Wells does not seem to consider. Here's a sampling:

- Charter schools have very different purposes. Some promote studying the arts, some promote studying cultural heritage; some focus on at-risk students, some focus on academically gifted students. Is it right to lump all of these schools into one big study? One study that attempts to address this issue, the Manhattan Institute's "Apples to Apples: An Evaluation of Charter Schools Serving General Student Populations," puts it well: "[C]omparing targeted charter schools to regular public schools is like comparing apples and zebras." When the authors compare "apples to apples," they find that charter schools outperform similar public schools.

- Charter schools with different purposes have different goals. Charter school opponents are some of the same people who vehemently criticize standardized testing for being one-dimensional, yet they now rely on those supposedly one-dimensional tests to malign charter schools. What about evaluating charter schools on other dimensions, such as graduation rates, student discipline problems, parent satisfaction, teacher turnover? Furthermore, should a charter school whose mission is to help students succeed in the arts, for example, be judged on its test scores? Or should charter schools whose students are special education students or high-school dropouts or kids in the juvenile justice system be judged on their test scores?

- Charter school students are different from public school students, if for no other reason than that they chose to leave the public schools. What is that difference and how do we control for it? Without some reliable way to adjust for that difference, we are again comparing apples to zebras.

Evaluating charter schools is difficult, but the schools are clearly popular. They must be doing something right and we owe it to all students to find out what that is. We should study the issue dispassionately and not resort to political cheap shots. We should use the most careful analysis that we can and not rely on simplistic analysis just because it is easy."

Judicial obstructionism again: "In a ruling the dissent characterized as driving 'a semi-truck' through a 'small window' in the U.S. Constitution, the full 1st District Court of Appeals in Florida on November 12 struck down the state's five-year-old Opportunity Scholarship program, ruling it violated a provision in Florida's Constitution barring public funds being used to aid any religious institution. ... Although they will be able to continue in their choice schools while the ruling is appealed to the Florida Supreme Court, the breadth of the appeals court decision places a cloud not only over the future of the Opportunity Scholarship program, but also over similarly funded post-secondary scholarship programs ..."


For greatest efficiency, lowest cost and maximum choice, ALL schools should be privately owned and run -- with government-paid vouchers for the poor and minimal regulation.

Comments? Email me here. For times when is playing up, there is a mirror of this site (viewable even in China!) here


No comments: