School Turnarounds: Evidence From the 2009 Stimulus
In 2009, Fordham fellow Andy Smarick wrote: “School turnaround efforts have consistently fallen short of hopes and expectations.” And we’ve generally agreed. This research from Thomas Dee scuffs up that pristine position, however, at least a little. It examines first-year impacts of the federal School Improvement Grants (SIG) program (background here and here) in California. Dee analyzed data from roughly 2,800 schools situated just above and below the eligibility cut-off for SIG funds (eighty-two of which received SIG awards, averaging roughly $1,500 per pupil)—looking specifically at schools that opted for either of the two most popular models: transformation or turnaround (more on those here). Dee found that SIG reforms raised the scores on California’s Academic Performance Index by an impressive thirty-four scale points over the course of one year (2010-11). Before the interventions, the average SIG-eligible school scored roughly 150 points below the state’s performance target of 800, which implies that SIG closed this gap by 23 percent. (Still, API is a complex metric, and it is not clear what this means for average student-level growth on the California Standards Tests.) Dee found the most improvement in the turnaround schools (where the principal and most of the staff is replaced) and in schools that had been the “lowest-achieving.” That said, CA’s “lack of progress” schools (Title I schools that posted very minimal improvement in the five years before undergoing SIG intervention) were not significantly impacted by the federal program. Dee conducted a number of analyses attempting to refute his findings—including attention to non-random student sorting (might the SIG program attract or repel certain students?). But results held. (Sidebar: Dee also compared SIG’s efficacy to class-size reduction [CSR] initiatives and found that SIG generates about half the achievement gains of CSR at a third of the cost. Of course raising class size is notoriously expensive.) Granted, these are short-term impacts. It’s unclear whether they’ll carry over through years two, three, and beyond, especially as SIG funding runs out. And with an average SIG grant of $1.5 million per school, the well will run dry.
Thomas Dee, School Turnarounds: Evidence from the 2009 Stimulus (Cambridge, MA: National Bureau of Economic Research, April 2012).