The Bluegrass Institute for Public Policy Solutions

View Original

Comments on the Charter School Debate: What's really in those CREDO charter school reports?

If you have not taken time to read our major charter school debate with UK professors Wayne Lewis and Martin Solomon, you owe it to yourself and Kentucky’s children to take some time to do so. The professors provide a good introduction into the issues of establishing charter schools in Kentucky from the viewpoint of both a strong proponent of charters and a sharp critic of these school choice options for parents.

Now that the professors have weighed in, I’m adding more to the discussion. Today I deal with an apparent incongruity created by our two honored debaters – both somehow manage to pull points from the same report to support their pro- and con- positions on charter schools.

For example, in his opening statement, Professor Solomon made a point that sounds pretty bad for charter schools. He said:

“A 2013 Stanford follow-up study of 27 states, which included 95 percent of the students in U.S. charter schools, found a slight improvement. But in that study, 71 percent of charter schools were still either inferior to or no better than public schools.”

Professor Solomon is referring to the “National Charter School Study, 2013” report from a research group at Stanford University called the Center for Research on Educational Outcomes, widely referred to by its initials as CREDO.

On the other side, in his Response #2, Professor Lewis also mentions the same CREDO study to support his pro-charter position. Writes Lewis:

“In CREDO’s 2013 study, which looked at the performance of students attending charter schools in 26 states and New York City, students attending charter schools had greater learning gains than their traditional public school peers in reading and equivalent learning gains in mathematics.

Further, economically disadvantaged students, African-American students and students who were English Language Learners (ELLs) gained significantly more days of learning each year in reading and mathematics compared to their traditional public school-attending peers. And the performance differences between students attending charter schools and their traditional public school peers were strongest among economically disadvantaged African-American and Hispanic students and Hispanic students who were ELLs.”

So, what’s going on here? How can both a proponent and a critic of charter schools draw support from the same study? I have some answers, and they point to an important reason why many studies of charter schools may not be providing an accurate picture of the real performance of these alternative schools of choice.

For reasons that elude me, reports from CREDO consistently fail to provide much prominence to what is perhaps the most important finding in all of their studies. Very simply, that finding, which is even mentioned on Pages 32 and 33 in the first CREDO study from 2009, is that, across the country, once students spend about two to three years in a charter school, they do start to outperform their traditional public school counterparts.

This CREDO finding is actually unsurprising. It makes sense that charter schools need a couple of years to help students establish good study skills and work habits. During that time, students are still adjusting to the more rigorous demands of the charter school world. While they are still getting their feet on the ground, quite often new charter school student performance is no better, and sometimes worse, than their counterparts in the traditional school system.

So, CREDO shows any charter school study that simply looks at the performance of all charter school students lumped together, including those still-adjusting new students, will probably under-report the real potential of charter school education.

By the way, this time-is-required situation actually impacts many of the other findings in CREDO’s studies because much of the material they discuss does include all charter school students – first year up – in the numbers.

So, while, as Professor Solomon points out, CREDO does show that “71 percent of charter schools were still either inferior to or no better than public schools,” that statistic is loaded against showing charter schools in a fully revealing light.

But, here is what the 2013 CREDO report shows for how charters compare to traditional public schools over time.

CREDO 2013 Report Graph of Charter Impact by Years of Enrollment Extract

On this graph, the traditional public school student performance is shown by the zero axis about half way up from the bottom of the chart. Relative performance of charter school students by year of enrollment in a charter school is shown by the pink and light green lines (please ignore the numbers and the asterisks, which relate to technical measurements in standard deviation units that will probably confuse most of our readers).For example, first year charter school students do notably worse than their traditional public school counterparts, performing at an equivalent of having received around 60 days less instruction in mathematics (pink line) and around 50 days worse in reading (light green line).However, by the second year, charter school students are moving slightly ahead, performing at a level that would indicate about 15 days more math instruction and 25 days or so more in reading.

Now take a look at the third year charter school student’s performance compared to his typical traditional public school counterpart. He is about 25 days ahead in math and around 50 instructional days ahead in reading. This gets even better after the charter school student has been in their school of choice for four years.

The nationwide results shown above get even stronger if we look at states with better charter school programs. For example, consider the latest CREDO 2013 update study in Louisiana.

Here is the relevant graph (again, ignore the standard deviation numbers and the scale on the left side of the graph unless you are into that level of detail).

CREDO LA Report - Charter Impact by Years of Enrollment

By the time a student spends three years in a Louisiana charter school, he or she gains the equivalent of more than 100 days of extra instruction in both reading and math compared to traditional public school counterparts! By the time the student has spent another year or two in Louisiana charters, they are about a full school year ahead (Keep in mind, the average school year length runs about 180 days, depending upon the state).

In fact, in Louisiana charter students actually start to outperform their traditional public school counterparts in math from the very first year of charter enrollment. That is a tough act for Louisiana charters to master, but they are doing it.

So, here is the important take-away: reports on charter schools that just average all student performance together to compare to traditional public schools will tend to under-report the real benefit of charter schools. This failing is found in many charter school studies, even in some of the data from CREDO. This is why the 2013 CREDO statistic Dr. Solomon cites puts charters at a very unfair disadvantage. However, the very same study also shows how dramatically charter schools really perform once they have enough time with students to truly make a difference. It’s a difference Kentucky’s under-privileged students badly need. It’s sad that kids in Kentucky are denied that advantage while kids in Louisiana are flourishing with it.