An analysis released in today's Education Gadfly finds that new charter schools in disadvantaged communities are almost four times as likely to reach above-average rates of student achievement as the closest district school. This raises serious questions about the wisdom of the federal government pumping $3 billion into school turnaround efforts instead of using some of the money to replicate and scale up successful charter models.
However, the finding comes with several big caveats. First, because of the small sample size, the results cannot be deemed statistically significant. And second, it's impossible to know whether "selection effects" played a role--whether the new charter schools performed better because they attracted better students.
The analysis was by David Stuit, a Vanderbilt PhD. who authored a previous Fordham study on school turnarounds last December.
The idea for this analysis came from Public Impact's Bryan Hassel. After the release of our December study--which found that just one percent of district and charter school turnarounds were successful, as defined as reaching at least the 50th percentile in state proficiency in reading and math--Bryan wondered whether charter start-ups in similar neighborhoods would fare any better against such rigorous criteria.
So we asked Stuit to give it a look. As he explains in his essay:
Across ten states (Arizona, California, Florida, Michigan, Minnesota, North Carolina, Ohio, Pennsylvania, Texas, and Wisconsin), I located all incidents (between 2002-03 and 2006-07) of a charter school opening in close proximity to a district school that had reading and math proficiency rates in the bottom 10 percent of its state at the time the charter appeared in its neighborhood. To qualify as a fair match-up, the charter and district schools had to be the nearest neighboring public schools of the same type (elementary or middle) and be located less than three miles apart as the crow flies. The schools also had to be demographically similar, with no more than a 10 percentage point difference in their subsidized lunch and minority enrollments. After identifying these matches, I examined the reading and math proficiency rates of the schools in 2008-09 to determine how many schools had become ?successes? by that year (success defined here as performing above the state average).
And sure enough, the charters did better--a lot better. He found 81 pairs of schools; 15 of the charters (19%) made the cut versus four of the district schools (5%). To be sure, that means that the vast majority of the charter start-ups remained in the lower half of student proficiency--and 40 percent remained stuck in the lowest quartile. But if you're playing the odds, they look better for new charters than for turnarounds.
But another finding of Stuit's is just as important: He could find very few charters located near the lowest-performing district schools:
Even if start-up charters are more likely to succeed than turnarounds, there currently are not enough of them available to kids stuck in failing schools.
We simply don't have the luxury of choosing charter start-ups over district turnarounds, because charters are nowhere near the scale that we would need them to be, especially in the neediest areas. If we want to "do something" about the very worst public schools in the country, we have to pursue a "both, and" strategy, not an "either, or" one.
Still, one thing's clear to me: it's screwy for federal taxpayers to spend 12 times as much on school turnarounds ($3 billion) as charter start-ups ($250 million) when the latter appear to be four times more likely to succeed than the former. Team Obama: Wanna fix that?
-Mike Petrilli
