A new study shows that charter school students are now outpacing their peers in traditional public schools in math and reading achievement, cementing a long-term trend of positive charter school outcomes.
The study is the third in a series , which has researched charter school performance since 2000. The third study, released June 6, is notable because it shows superior outcomes among charter school students while the center鈥檚 earlier studies showed charter school students performed either worse than or about the same as their peers in traditional public schools.
The researchers used standardized testing data from over 1.8 million students at 6,200 charter schools to determine how student learning at the schools compares to traditional public schools.
The most recent study covers student learning from 2014 to 2019. Its previous studies covered 2000-01 through 2007-08 and 2006-07 through 2010-11. CREDO鈥檚 study of that first period, released in 2009, found that charters failed to meet traditional public school outcomes.
The 2023 study reversed that narrative and showed that charters have drastically improved, producing better reading and math scores than traditional public schools. The results are 鈥渞emarkable,鈥� said Margaret 鈥淢acke鈥� Raymond, founder and director of CREDO.
鈥淭he bigger lesson now in the post-COVID world is, hey guys if you鈥檙e looking for a way to improve outcomes for kids, here is an absolutely demonstrated framework that you can look at and maybe apply it in other contexts,鈥� Raymond said.
Raymond said the improvement has come largely from existing schools tweaking their practices over time and getting better as a result鈥�"evolution鈥� rather than 鈥渞evolution,鈥� she said.
Over the same time period, Raymond said, charter school authorizers鈥攖he school districts and state bodies that allow charter schools to start and oversee their performance鈥攈ave become more focused on ensuring quality.
However, the study isn鈥檛 a completely positive assessment of charter schools. Outcomes vary depending on location, school type, student circumstances, and leadership, according to the study. And charters have a long way to go in serving students with disabilities.
Here are six major takeaways from the new report.
1. Charters have improved demonstrably over time
When CREDO released its first study on charter school performance in 2009, Raymond described it as a 鈥渃old shower鈥� for the charter community. The study showed results that indicated the schools didn鈥檛 match the achievement of traditional public schools.
In the decade that followed, charters managed to turn that story around. The study, which measures student achievement in days of learning gained or lost compared with peers in traditional public schools, showed that charter students were able to go from losing six days of learning in reading and 17 days in math in 2007 to gaining 16 days of learning in reading and six days in math in 2019.
2. Students of color performed better at charter schools
In both math and reading, students of color at charter schools performed better than their traditional public school peers, but they didn鈥檛 fully close achievement gaps.
For example, Black charter school students had 35 more days in reading growth and 29 more days in math growth than Black students in traditional public schools. But neither group managed to meet the 180-day benchmark for standard learning.
Hispanic charter school students also outpaced their peers, gaining 30 days in reading and 19 days in math over Hispanic students in traditional public schools.
Black and Hispanic students in poverty had even stronger results. Black charter students in poverty gained 37 days of learning in reading and 36 days in math over their counterparts in traditional public schools, and Hispanic students in poverty gained 36 days of reading and 30 days of math over their traditional public school peers.
English learners in charter schools also saw gains over English learners in traditional public schools, with 11 additional days of learning in reading and eight additional days of learning in math.
3. Charter schools do a poor job educating students with disabilities
While students of color seem to perform better in charter schools, it鈥檚 a different story for another subgroup鈥攕tudents with disabilities.
The study showed that students with disabilities lost around 13 days of learning in reading and 14 days in math compared with their traditional public school peers.
The result is 鈥渁bysmal,鈥� Raymond said, and indicates that charter schools need to do more to support students with disabilities.
The study doesn鈥檛 have data on the types of disabilities for which charter school students in special education received services, but charters tend to serve fewer students with developmental delays, intellectual disabilities, and multiple disabilities than traditional public schools, according to an organization focused on improving outcomes for charter school students with disabilities.
4. Results vary by location
While the charter schools studied in the report had a superior performance on average, students didn鈥檛 improve across the board compared with their peers in traditional public schools.
Some states鈥攍ike Maryland, New York, Rhode Island, and Tennessee鈥攕aw drastic improvement in the latest study. Rhode Island saw the largest growth, with students gaining 90 days of learning in reading and 88 days of math over their peers in traditional public schools.
Other states鈥擮hio, Oregon, South Carolina, and Utah鈥攕aw charter school students lose days of reading, with Oregon losing the most, 18.5 days.
And in nine states鈥擜rkansas, Florida, Indiana, Ohio, Oregon, Pennsylvania, South Carolina, Texas, and Utah鈥攃harter school students lost days of learning in math compared with their peers in traditional public schools. Ohio charter schools saw the largest drop in learning days in math, losing 37 days compared with traditional public schools.
Charter school students in urban and suburban settings also performed better than charter students living in town or rural settings. Rural charter students gained five days of learning in reading compared with their traditional public school peers and lost 10 days in math, while charter students in urban settings gained 30 days in reading and 28 days in math.
5. Online charter schools have 鈥榟orrific鈥� performance
Students in fully online charter schools lost 58 days of learning in reading and 124 days in math compared with their traditional public school peers.
Meanwhile, students in brick-and-mortar charter schools gained 22 days of learning in reading and 15 days in math.
The study results show that fully online charters can鈥檛 keep up with traditional public schools, and the students enrolled in them suffer because of it, Raymond said.
鈥淭he learning there is just completely dreadful,鈥� Raymond said. 鈥淭he kids in online schools on average lose a third of the year in reading and two-thirds of a year of growth in math.鈥�
Data from the online schools drag down the average performance for students in all charter schools, Raymond said.
6. Schools run by charter management organizations do better overall
Schools affiliated with charter management organizations, which generally oversee networks of connected charter schools, showed greater growth than stand-alone charter schools and the charter school sector as a whole.
According to the study, 43 percent of students in charter schools run by a management organization performed better than their traditional public school peers, 42 percent performed the same as their traditional public school peers, and 15 percent performed worse.
At stand-alone charter schools, 32 percent of students performed better than their traditional public school peers, 50 percent performed the same, and 18 percent performed worse.
Many of the online schools fall into the stand-alone category, so they are likely dragging the average down, Raymond said. But the data do indicate that charter management organizations have become skilled at taking what works and replicating it elsewhere.
鈥淪omebody has already figured out how to do a good school so the challenge there is just replicating it鈥攊t鈥檚 not figuring out the model all over again,鈥� Raymond said.
