Charters' Path to Success or Failure Set Early, Study Finds
First year is key predictor of achievement in later years, study concludes
Charter schools' academic success or failure during their first year is a strong predictor of whether they will excel or struggle in later years, a far-reaching study finds.
The study, released Jan. 30 by Stanford University's Center for Research on Education Outcomes, or CREDO, which has conducted extensive analysis of charters around the nation, also concludes that significant improvement in charter school performance over time is rare among middle and high schools, though it occurs more often in elementary schools.
The analysis seeks to test a number of the most pressing questions about charter schools and their "life cycle of quality," said Margaret E. "Macke" Raymond, the study's lead author. Those questions include the extent to which charters can improve over time; whether the academic strategies and other policies they put in place out of the gate determine their success as they reach maturity; and whether charters can expand beyond their original, flagship schools to form networks of successful schools.
There is now an assumption within the charter sector that even if "the first few years are rocky" at a school, charters can eventually rise to higher performance over time, the authors say. But the study casts doubt on that notion.
A new, nationwide study by Stanford University examines whether charters improve over time, and whether successful schools’ models are easily replicated. Among the findings:
Performance: The vast majority of charters that start out in the lowest- or highest categories of performance tend to remain at those levels of performance five years later.
Improvement: Student-achievement gains over time are “largely absent” from middle and high school charters, but elementary charter schools have a record of making academic improvements over time.
Poor and Minority Student Performance: The academic performance of charter-management organizations—consisting of three or more charter schools—is not better than non-CMO charters. However, CMO-affiliated charters outperform both traditional public schools and independent charters in showing gains among poor and minority students.
Student Gains: New CMO-affiliated charters produce larger student gains than new independent charters, but both types of charters lag behind average learning gains in traditional public schools.
CMOs: Roughly one-third of CMOs get better over time; one-quarter of them show declines, and performance remains stable in the rest.
Learning Gains: The average student in education-management organizations, which provide services to charters under contract but do not hold the charters, posted larger learning gains than CMOs, independent charters, or traditional public schools.
On the one hand, the Stanford researchers found that many new charters, regardless of the grade levels they serve, succeed from their inception, and that there is no structural "new school" phenomenon that would cause them to founder at the outset.
At the same time, charter schools' early academic performance is a strong predictor of how well they perform later, the researchers found. The vast majority of schools in the lowest quintiles of performance, 80 percent, remained low performers through their fifth year in operation.
Top-tier charters followed a similar pattern. Ninety-four percent of schools that started in the top quintile of performance remained at that level over time. When the researchers waited until the third year of charters' operation to make predictions, the patterns for predicting success or failure were even stronger, the authors write.
The authors put forward a number of lessons for charter school authorizers, the entities that typically approve and oversee charters. Authorizers need to pay close attention to charters' performance early in those schools' lives, and be willing to close stragglers if they show signs of not improving.
"For the majority of schools, poor first-year performance will give way to poor second-year performance," the study says. "Once this has happened, the future is predictable and extremely bleak. For the students enrolled in these schools, this is a tragedy that cannot be dismissed."
The CREDO study also sought to examine the performance of charters overseen by various types of operators, large and small. It found that charter-management organizations, defined as operators of three or more charters, score at about the same level as non-CMOs, with a broad range of strong and weak performers within those schools. Charter-management groups posted superior results in serving minority and impoverished students, however.
There are roughly 6,000 charter schools in the United States, serving more than 2 million students. The study analyzed 167 charter-management organizations, with 1,372 schools, in their analysis, which drew data from 25 agencies for states or jurisdictions. The Stanford researchers followed student-level performance in schools from the time they opened through their fifth year, mapping charters' academic showing against a "static set of performance thresholds" so that academic trends over time could be evaluated.
The analysis of charter-management organizations is based on a "virtual control record" method, in which students in those schools are compared with "virtual twins" who attend the regular public schools the charter students would otherwise have attended.
The study also examined the performance of students in charter school networks whose member schools have been replicated in local communities—organizations that the authors dub "super-networks."
Of four super-networks reviewed in the study, a pair of them, the Knowledge Is Power Program, or KIPP, and Uncommon Schools, generally had a strong academic showings, the authors found, while the other two, Responsive Education Solutions and White Hat Management, did not fare as well.
For example, both kipp and Uncommon Schools had a large, significant positive effect on the academic growth of students in both reading and math, the study found.
But Responsive Education Solutions had a significant negative impact on student reading and nonsignificant effects in math. Meanwhile, White Hat had a small but significant positive impact on reading progress, but a significant negative effect in math.
Alan Wimberley, the chief learning officer for Responsive Education Solutions, headquartered in Lewisville, Texas, said the study did not present an accurate picture of the challenges the charter operator faces across its 60-school, 10,000-student population, or its broad academic mission.
Responsive Education "serves students across the entire spectrum of charter school missions," Mr. Wimberly said in an email. "We have schools focused on children with autism, college-preparatory schools, residential treatment schools and dropout recovery schools."
White Hat Management officials, in a statement, said they had not reviewed the study's methods, but said their schools are providing "consistent value to students and their families."
The study also examined the academic showing of students in charters operated by education-management organizations, defined as organizations that work at a school under contract but do not hold the charter. The average student in an EMO-managed school outperformed peers in CMOs, independent charters, and regular public schools, the study found.
Vol. 32, Issue 20, Pages 12-13
Get more stories and free e-newsletters!