Pennsylvania’s charter school debate attracts a lot of heated rhetoric.
But this week, the conversation got some cold, hard numbers.
The Center for Research on Education Outcomes, a group based at Stanford University, released a deep dive into Pennsylvania’s charter schools, which now serve roughly 140,000 students.
Debates about the quality of the growing sector can be especially fraught because comparing schools is rarely an apples-to-apples exercise.
A charter school serving many low-income students might not post top results on state tests, but may actually do a better job serving disadvantaged students than a nearby traditional public school.
On the flip side, some studies show charters sidestep the toughest-to-serve students, like those with extreme special needs or those who are learning English. These skeptics worry that traditional public schools end up with these cast-aside students, and thus, lower test scores.
CREDO’s analysis is an attempt to control for these variables.
The researchers — who have done similar analyses in cities and states across the country — look at charter school students and try to find students in nearby traditional public schools who are their “virtual twins.” They create these “twins” by looking at a student’s age, gender, race, socioeconomic status, English language ability, and special education status.
Then the CREDO researchers compare the test-score growth of charter school students with their “twins” to see if the charter school is making a difference.
This latest Pennsylvania study covered the years 2013-17.
So what did CREDO find?
Lesson 1: Overall, Pa. charter schools aren’t moving the needle
CREDO found that students in charter schools did about as well as their public school peers in reading. And on math tests, charter students actually fared worse.
A student at a charter school in Pennsylvania received the equivalent of “30 fewer days of learning in math” each school year, according to CREDO.
Those middling results come at a cost. Taxpayers spend over $1.5 billion each year on charter schools, and studies suggest increased charter enrollment creates a financial burden for traditional students.
Lesson 2: Cyber schools are dragging the entire charter sector down
Cyber charter schools are barely a blip in some states.
But in Pennsylvania, they educate about a quarter of all students who attend charter schools. CREDO’s analysis suggests they don’t educate these students well.
The average student at a cyber charter in Pennsylvania lost 106 days of learning in reading and 118 days in math compared to their “twins” in traditional public schools.
Cyber school skeptics say these results prove that the virtual classroom isn’t working for Pennsylvania students, and that the roughly half a billion dollars cyber charters collect each year from taxpayers is a bad investment.
“In cyber schools we get abysmal results,” said Susan Spicka, executive director of Education Voters PA.
Some aren’t convinced, though.
They say cyber charters serve students on the fringes of the public education system, students who’ve floundered in traditional school settings. They don’t think CREDO’s analysis can properly gauge the difficulty of serving these students. And they point to parent demand for cyber charters as evidence Pennsylvania needs virtual options.
“In many cases, if students did not have the cyber charter school option, many would most likely drop out of the traditional public school system and would be excluded from the education process altogether,” said the Pennsylvania Coalition of Public Charter Schools
Lesson 3: Some charters are working very well
The lackluster showing from Pennsylvania’s cyber charters masks another critical data point: Brick-and-mortar charter schools show some promise.
Overall, Pennsylvania students at brick-and-mortar charters performed better in reading than their “twins” in traditional public schools, and about the same in math. CREDO estimates that it’s as if charter students have an additional 24 days of reading instruction each year.
In reading, charters seem especially effective for poor students, black students, and hispanic students enrolled at brick-and-mortar charter schools. Overall, charters in urban areas were particularly effective at boosting reading scores — while suburban and rural charters struggled.
“[Charters] are far from perfect,” said Mark Gleason, a school-choice backer who heads the financially influential Philadelphia School Partnership. “But for a significant number of families they create an opportunity that would not exist.”
Lesson 4: This may be a Pennsylvania problem
CREDO published a similar analysis of Pennsylvania charter schools in 2011 and found largely similar results.
Does this lack of progress suggest a problem with the charter school concept or a problem with how Pennsylvania specifically implements charter policy?
CREDO has looked at charter schools elsewhere and found stronger overall gains in student achievement. The latest report highlights Indiana, Illinois, and North Carolina as states where the sector seems to show stronger results.
“When we look at charters overall, Pennsylvania is one of the less impressive states,” said Mark Gleason with the Philadelphia School Partnership.
Pennsylvania’s charter law, drafted in the late 1990s, receives criticism from just about every corner of the political world.
Susan Spicka thinks it’s too hard to close bad charters in Pennsylvania.
The state, which authorizes cyber charter schools, doesn’t have the bandwidth to hold cyber schools accountable, she said. And when districts try to shut down brick-and-mortar charters, Spicka argued, they get caught up in years of appeals and legal muck.
“It just encourages this wild west of charter schools where if you can just get yourself approved one way or another you can just keep operating,” Spicka said.
Gleason acknowledges there are shortcomings in the charter sector. Some of it he puts on individual charter school boards. Some of it he blames on a slow and “cumbersome” process for closing bad charters and expanding good ones.
The CREDO analysis found that 45 percent of charters did “significantly better” in reading than traditional public schools and that 33 percent did “significantly better” in math.
But there are enough middling and poor charters that the overall numbers look uninspiring.
Even with signs of success in urban areas, the “evidence isn’t strong enough yet” that the charter sector writ large has lived up to its promise, Gleason said.
“But from my vantage point there’s enough evidence that that is happening slowly,” he argued. “It’s too early to say this is a failed experiment.”
This story has been updated to clarify Mark Gleason’s perspective.