“A Straightforward Comparison of Charter Schools and Regular Public Schools in the United States” by Caroline M. Hoxby, at http://post.economics.harvard.edu/faculty/hoxby/papers/hoxbyallcharters.pdf (Sept. 2004).

The emotional debate over charter schools has raged for years without much solid evidence on either side. Now, on the heels of a widely publicized American Federation of Teachers (AFT) study last summer that found charter students lagging behind their peers in regular public schools, comes an unusually comprehensive research paper by Hoxby, a Harvard University economist. Her conclusion: Charter schools do a better job of producing proficient students.

Thirty-six states and the District of Columbia now have charter schools, which are publicly funded but free of many of the strictures that bind the conventional public school system. Founded by community members, entrepreneurs, and others, charter schools tend to stress innovative teaching practices and parent involvement. Nationwide, they enroll 1.5 percent of all students.

Hoxby looked at how some 50,000 charter-school fourth graders did on state proficiency exams in reading and math, relative to their peers in comparable regular public schools. The results: The percentage of charter students who were proficient was four points higher in reading and two points higher in math.

The charter schools’ superiority was greater in states where they had been in existence longer and enrolled more students. In the District of Columbia, which has a larger proportion of students in charter schools (11 percent) than any state, the advantage was 35 percentage points in reading and 40 points in math. Only in North Carolina, where less than 2 percent of  students are enrolled in charter schools, were there charter disadvantages in both reading and math.

Because so few American students attend charter schools, Hoxby says, the tiny sample (3 percent) used in the AFT study (www.aft.org) was statistically meaningless. And the AFT compared the performance of charter schools, which often serve low-income neighborhoods, with statewide school averages. Hoxby’s research, by contrast, encompassed 99 percent of all charter-school fourth graders in the 2002–03 school year, and it compared charter schools with nearby conventional public schools, where the student body is more likely to be similar in composition.

But the debate is far from over. After Hoxby’s study appeared, the U.S. Department of Education released a report (available at http: //www.ed.gov/rschstat/eval/choice/pcsp-final/finalreport.pdf) comparing schools rather than students. The results: In all five states studied, charter schools were less likely than conventional public schools to meet state proficiency standards. Even after adjusting for differences in the composition of the student body and other factors, charter schools in two states came up short.