Includes updates and/or revisions.
A first-of-its-kind analysis suggests that in Chicago, at least, charter school students are more likely to graduate from high school and enroll in college than similar students in regular public high schools.
Released May 7 by researchers from the RAND Corp., Mathematica, and Florida State University in Tallahassee, is the first in the nation to track high school outcomes for students enrolled in charter schools. It finds that attending a charter high school in that city boosts a student鈥檚 chance of graduating from high school by 7 percentage points and increases the likelihood that a student will enroll in college by 11 percentage points.
But it also finds that, at the elementary and middle school level, students in Chicago鈥檚 charter schools don鈥檛 appear to make any greater learning gains than their peers in regular public schools.
鈥淭he bottom line is that the attainment effects that we鈥檙e seeing are very promising and quite substantial, and they鈥檙e going to be important for not only policy but for future research on charter schools,鈥� said Brian P. Gill, a study co-author and a senior social scientist at the Princeton, N.J.-based Mathematica Policy Research Inc.
Numbering more than 4,000 across the nation, charter schools are public schools that are allowed to operate with fewer of the bureaucratic constraints that apply to most regular public schools.
In the 408,000-student Chicago school system, charters serve an estimated 20,000 students in 28 schools, with two more such schools scheduled to open in the fall. The rapid growth of charters in that city is due in part to its 4-year-old Renaissance 2010 initiative, which seeks to create 100 new schools by 2010.
Part of Larger Project
The Chicago findings are among the first to emerge from a larger project looking at the growth of charter schools in four cities鈥擟hicago, Milwaukee, Philadelphia, and San Diego鈥攁nd in the states of Florida, Ohio, and Texas. It鈥檚 being underwritten by the Seattle-based Bill & Melinda Gates Foundation, the Joyce Foundation in Chicago, the Stranahan Foundation in Toledo, Ohio, and the William Penn Foundation in Philadelphia.
One of the stickiest problems in research comparing charter schools with regular public schools is how to account for the fact that charter school students might be more motivated to learn because they鈥攐r their families鈥攃hoose the schools they attend.
To get around that problem, the researchers in the Chicago study compared the achievement trajectories for the same students before and after they switched to charter schools. In keeping with other studies on charter schools, the researchers found that students make similar learning gains in reading and mathematics in both types of schools in grades 3 through 8.
The study also found that, both in terms of student achievement and in terms of their racial and ethnic mix, the charter schools to which students transferred looked no different than the traditional schools they left behind.
鈥淭hat鈥檚 important because there鈥檚 been a debate about whether charter schools are cream-skimming kids from traditional public schools or contributing to racial stratification,鈥� said Ron Zimmer, another study co-author and a senior policy researcher at the Santa Monica, Calif.-based RAND Corp.
The study鈥檚 main innovation, though, is its high school analysis. For that part of the study, researchers control for potential biases by focusing on a subset of about 1,000 students who attended charter schools in 8th grade and then went on to either a traditional public high school or remained in the same charter school. (The grade configurations available to Chicago charter school students include schools serving students in grades 7-12, 6-12, or K-12.)
Unusual Analysis
In 8th grade, Mr. Gill said, both groups were similar in terms of demographics and achievement levels. But the students who stayed in charter schools went on to earn higher scores on college-entrance exams and graduate from high school and enroll in college at higher rates than their counterparts in traditional high schools.
On the ACT college-entrance tests, the charter school students scored, on average, half a point higher than their traditional-school counterparts. (That鈥檚 in an analysis for which the median score was 16, out of a possible 36, according to Mr. Gill.)
鈥淲hat we have to acknowledge here, though,鈥� added Mr. Gill, 鈥渋s that it鈥檚 possible that the positive effects we鈥檙e seeing could be a result of the grade configuration itself. It could be that kids benefit from eliminating the transition that kids make from middle school to high school.鈥�
District school officials also acknowledged that the city鈥檚 charter high schools are typically smaller than its traditional high schools, many of which enroll 1,000 or more students. On the other hand, the study notes, those schools may have unconventional grade configurations or smaller enrollments because they are charters.
Independent experts praised the study for its rigor. Whether the findings will apply more generally to charter high schools in other states, though, remains an open question, they said. 鈥淓very charter school reform gets a different character,鈥� said Gary W. Miron, an education professor at Western Michigan University in Kalamazoo.
Charter high schools in Connecticut, for example, provide a more remedial function, he noted. In Delaware, they focus on college preparation. 鈥淏ut we need to get beyond student achievement alone in these kinds of evaluations,鈥� Mr. Miron said, 鈥渁nd this is one of the few studies that do that.鈥�
Preliminary findings from the larger multistate study do suggest that a similar pattern of high school attainment for charter schools is emerging in Florida. (鈥淗igh School Studies Eye Role of Charter Status, Teachers,鈥� April 9, 2008.)
Chicago school officials said they weren鈥檛 surprised by the positive findings. District data, however, shows charter school pupils at the elementary level outscoring the district average. The results differ because the district report, also released last week, gives a snapshot of student performance on 2006-07 tests rather than tracking achievement growth over time.
