Speaking of charter schools, here's a brilliant op ed in the WSJ earlier this week by Harvard's Paul Peterson, who rebuts some of the myths and faulty studies about charters and shares some very interesting data I'd never seen on how well choice works in other countries:

According to a 2009 Education Next survey, the public approves of steady charter growth. Though a sizeable portion of Americans remain undecided, charter supporters outnumber opponents two to one. Among African Americans, those who favor charters outnumber opponents four to one. Even among public-school teachers, the percentage who favor charters is 37%, while the percentage who oppose them is 31%.

A school can have short-term popularity without being good, of course. Union leaders would have us believe that charter popularity is due to the "motivated" students who attend them, not the education they provide. But charters hold lotteries when applications exceed available seats. As a result—and also because they are usually located in urban areas—over half of all charter students are either African American or Hispanic. More than a third of charter school students are eligible for the federal free or reduced lunch program.

To identify the effects of a charter education, a wide variety of studies have been conducted. The best studies are randomized experiments, the gold standard in both medical and educational research. Stanford University's Caroline Hoxby and Harvard University's Thomas Kane have conducted randomized experiments that compare students who win a charter lottery with those who applied but were not given a seat. Winners and losers can be assumed to be equally motivated because they both tried to go to a charter school. Ms. Hoxby and Mr. Kane have found that lottery winners subsequently scored considerably higher on math and reading tests than did applicants who remained in district schools.

In another good study, the RAND Corp. found that charter high school graduation rates and college attendance rates were better than regular district school rates by 15 percentage points and eight percentage points respectively.

Instead of taking seriously these high quality studies, charter critics rely heavily on a report released in 2004 by the American Federation of Teachers (AFT). The AFT is hardly a disinterested investigator, and its report makes inappropriate comparisons and pays insufficient attention to the fact that charters are serving an educationally deprived segment of the population. Others base their criticism of charters on a report from an ongoing study by Stanford's Center for Research on Education Outcomes (Credo), which found that there are more weak charter schools than strong ones. Though this report is superior to AFT's study, its results are dominated by a large number of students who are in their first year at a charter school and a large number of charter schools that are in their first year of operation.

Credo's work will be more informative when it presents findings for students in charters that have been up and running for several years. You can't judge the long-term potential of schools that have not amassed a multi-year track record.

To identify the long-term benefits of school choice, Harvard's Martin West and German economist Ludger Woessmann examined the impact of school choice on the performance of 15-year-old students in 29 industrialized countries. They discovered that the greater the competition between the public and private sector, the better all students do in math, science and reading. Their findings imply that expanding charters to include 50% of all students would eventually raise American students' math scores to be competitive with the highest-scoring countries in the world.

-----------------------