Here’s an approach to charter schools that should seem obvious—to those on both sides of the acrimonious debate on the future of charters in public education.
In places where charter schools are not achieving results, they should be suspended or at least curtailed until whatever isn’t working can be fixed.
And in cities where charters are making striking gains compared to traditional public schools, enrollment opportunities should be expanded, so that more kids can take advantage of them.
A case-specific approach reflects the truism that schools are not a monolith; some traditional public schools achieve terrific results and some—particularly in inner cities—do not. Charter schools are similarly heterogeneous. School districts should take a page from pasta chefs: Throw the results against the wall and see what sticks.
The Center for Research on Education Outcomes (CREDO), at Stanford University, has done that in a new study, and it turns out that charters, in general, are strongest exactly where the need is greatest—in urban areas. In some cities, such as Boston, students are achieving six times the growth in math knowledge as are their traditional school counterparts; in reading, four times as much.
The CREDO study also fingers cities where charters are plainly failing, although on average in the 41 urban areas it studied, charter students are clearly outpacing traditional-school peers. Notably, the methodology employed by CREDO seems to rule out the persistent accusation that charter schools get better results merely by “cherry-picking” abler or more motivated students.
For anyone who has spent the last couple of decades on Mars: Charter schools are publicly funded schools that are run outside traditional public systems. They are not bound by many of the rules that govern conventional schools, nor, typically, must they hire unionized teachers. Enrollment is open, with lotteries when there is a surfeit of applicants.
Although forests have been leveled for all the studies on charter schools, CREDO’s new study took an unusual tack. It studied students in multiple areas of the country—and exclusively studied urban areas. Three points emerged. When suburban charters were excluded, the smaller average gains registered in previous studies were suddenly magnified. In other words, charters seem to be remedying a particular defect of schools in the most challenged areas. Second, within those schools, gains were greatest among students—those in poverty, African-Americans, Hispanics, English-as-a-second-language (ESL) students—whose performance typically lags. Disadvantaged students gain the equivalent of months (or more) of extra learning for every year in a charter school.
And the third point was the great divergence among charter organizations (each of which has its own board and often a distinct approach, with varying levels of community engagement). Some are offering a superior alternative; some are not.
Tracking demographic ‘twins’
CREDO’s methodology was to pair each urban charter student with up to seven students at nearby traditional schools of a similar achievement level and demographic. Thus, a black male third-grader with strong reading and weak math scores enrolled in a charter school in Detroit was paired with a group of similar students at traditional public schools in Detroit. A girl in special education in Memphis would be similarly paired with a peer group in that city.
Over a six-year period, CREDO compared each charter student with the average of their demographic “twins.” This matching was done for more than one million students—a large enough sample so that the results would not be driven by chance.
It is true that charter students are drawn from a motivated group—those that care enough to apply. However, if a charter student’s home environment produced similar test scores to those of the matching group at the beginning of the sample period, one would expect equivalent results at the end as well.
But they weren’t equivalent. The charter students, according to the study, achieved “significantly higher levels of annual growth in both math and reading.” On average, students achieved the equivalent of 40 days of extra learning in math per year, and 28 days in reading. Moreover, with each year that a student stayed in charter school, relative results improved.
As noted, African-American and other disadvantaged groups performed particularly well, relative to their traditional school peers. Gains were strongest among students with “multiple disadvantages.” For instance, black students who were also poor achieved the equivalent of 59 days of additional learning in math compared to their traditional-school peers. Hispanic students who were also ESL advanced in math at the same rate as white peers in traditional schools—erasing a seemingly intractable learning gap.
More leaders than laggards
The CREDO study disaggregates the data by individual cities, revealing what it called truly “dramatic” gains in some, and disappointing results in others. Those in Boston, Newark, N.J., and Memphis were the most impressive, whereas charter students in Fort Worth and Las Vegas badly lagged their traditional school peers. Why have an alternative if it produces inferior results?
Overall, there were 26 cities in which charter students did demonstrably better than traditional school peers in math and 11 in which they lagged; in reading, those numbers were 23 and 10. Margaret Raymond, CREDO’s director, says “the idea here is that the charter formula is achieving” superior results across diverse areas in a two-to-one ratio of cities. But the results are not a blanket endorsement of charters.
I was especially interested in what CREDO had to say about my hometown, Boston. I have visited charter schools in the area and been impressed with the students’ engagement, the discipline and decorum maintained in class, and the schools’ emphasis on learning.
Boston charter schools have tested well in the past; however, charters here, and in all of Massachusetts, are hugely controversial. The state limits spending on charters to 9% of funding per district; this “cap” rises to 18% in underperforming districts. Boston has hit the cap; consequently, 18,000 students in the city are on a waiting list to get into charter schools (nearly 42,000 are on such lists state-wide).
While the waiting list grows, Boston public schools remain the classic hole in the donut—an island of underperformance within a high-achieving state. Students in Boston public schools perform at dramatically lower levels than the average in Massachusetts. Despite this geographic handicap, students in Boston charter schools are on pace to catch students elsewhere in the Bay State. “Boston is closing the achievement gap,” Raymond told me.
Confronted with such evidence, legislators introduced a bill last year to lift the cap, but the Massachusetts Teachers Association vehemently lobbied against it. The measure died in the state senate.
Richard Stutman, president of the Boston Teachers Union, said he wasn’t surprised by the CREDO results. “They’re cherry-picking” the students, he said. He pointed out that Boston charters had dramatically fewer ESL students and slightly fewer special-ed students. Neither factor, though, would affect the study’s conclusions.
I asked Stutman if he thought traditional schools could learn anything from charters. “They don’t do anything different,” he said. “All I ever hear is they have a longer school day.” In response, Jon Clark, co-director of Brooke Charter Schools (which I visited), enumerated a list: better discipline, professional development for teachers, principals focused on instruction rather than non-academic chores, and more tutoring and feedback for students.
We have enough experience with charters now so that, in places where they are clearly succeeding, rote opposition doesn’t make sense. And that doesn’t just go for Boston. If certain schools are getting better results with disadvantaged kids, you don’t put a cap on them.
Roger Lowenstein is the author, most recently, of "The End of Wall Street." He is writing a book on the origins of the U.S. Federal Reserve.