So I write a column purporting to summarize the latest research on the effectiveness of charter schools, and then the next day one of the largest study of charter-school performance ever makes its appearance. Caroline Hoxby, a researcher at Harvard University, examined a data set much, much larger than the one the American Federation of Teachers cited, and came to a different conclusion.
Comparing charter-school students to those in nearby district-run public schools, Hoxby found that reading proficiency was about 4 percentage points higher for charters. Math proficiency was 2 points higher. Adjusting for race increased the charter-school effectiveness to 5 points and 3 points, respectively, and older charters have still better outcomes.
No big news media coverage of the study yet. On second thought, I snicker at the word ?yet.? You can download the actual report here (it’s the first one, the straightforward comparison).
How does all this square with the debate over charters in North Carolina? Here?s an interesting finding in the Hoxby study:
North Carolina is the only state in which charter students? proficiency is statistically significantly lower, by 4 percent, in both reading and math, compared to students in the nearest public school. In addition, Texas? charter students appear to be statistically significantly less proficient in math (not reading).