Horace Meister, a regular contributor, has discovered a shocking instance of contradictory research, posted a year apart by the same “independent” governmental agency. The first report, published a year ago, criticized New York City’s charter schools for enrolling small proportions of high-need students; the second report, published a month ago, claimed that the city’s charter schools had a lower attrition rate of high-needs students than public schools. Meister read the two reports carefully and with growing disgust. He concluded that the Independent Budget Office had massaged the data to reach a conclusion favoring the powerful charter lobby. Eva Moskowitz read the second report and wrote an opinion piece for the Wall Street Journal called “The Myth of Charter School ‘Cherry Picking.’” Horace Meister says it is not myth: it is reality.



Meister writes:



In January 2014 the Independent Budget Office in New York City released a report on student attrition rates in the charter school sector in New York City.[i] A year later, in January 2015, the very same office released an “update” of the earlier report.[ii] The story behind this “updated”[iii] report reveals all that is wrong with education policy in the United States.


The original report had a number of fascinating revelations. It turns out that, as a sector, charter schools in New York City are using student demographics and attrition to boost their “performance” in ways that public schools cannot. This, of course, is not an indictment of any particular charter school or the dedicated staff in a specific charter school. It is an indictment of the overall corporate reform policy that favors charter schools over public schools, allows the charter sector to operate under a different set of rules than public schools so that charter schools can employ these sorts of gimmicks, and then dares to claim that charter schools are somehow better overall for students than public schools.


What did the original report reveal?


  1. Charter schools in New York City serve a much more advantaged student population than public schools. The very first table in the report showed that charter schools served 80% fewer English Language Learner students than nearby public schools. Charter schools serve 1/9th the proportion of the highest needs special education students as nearby public schools. Charter schools served a much more economically advantaged student body than nearby public schools—with three times as many students paying full-price for lunch than nearby public schools.
  2. By only accepting students in certain grades charter schools are able to artificially boost their outcomes as compared to nearby public schools. “The increased incidence of transfer to a traditional public school, instead of a charter school, might be due to the fact that many charters limit admissions to traditional starting points (such as kindergarten for elementary schools).” Of course it is the students with higher needs and higher absentee rates who are most likely to transfer at points other than the traditional ones. And, of course, it is illegal for public schools to bar students from admission at points other than the traditional ones, though this tactic is widely employed by charter schools. The disruptions caused by in-migration are also eliminated.
  3. Students with low test scores are more likely to leave charter schools. “The results are revealing. Among students in charter schools, those who remained in their kindergarten schools through third grade had higher average scale scores in both reading (English Language Arts) and mathematics in third grade compared with those who had left for another New York City public school (Figure 3)… One important difference between the two types of schools, particularly manifest when the percentage of students meeting or exceeding proficiency standard is used as the metric, is that the gap between the stayers and movers was significantly larger in charters compared with those in traditional public school.”
  4. Special education students with the highest needs are significantly more likely to leave charter schools than public schools. After leaving the charter schools these students go to public schools. “The attrition rates are higher for special education students who start kindergarten in charter schools than for special education students who start in neighboring traditional public schools. Only 20 percent of students classified as requiring special education who started kindergarten in charter schools remained in the same school after three years, with the vast majority transferring to another New York City public school (see Table 5). The corresponding persistence rate for students in nearby traditional public schools is 50 percent…Of those continuing in the same charter school, 10 percent were identified as special education students by the third year, and of those transferring out to another charter school, 16 percent were special education students (see Figure 2). But of those transferring out to another traditional public school, fully 27 percent were classified as special education students.” Of course the highest need special education students are also, as a rule, the students who perform the poorest on standardized tests.


Reading the original report a couple of unanswered questions suggest themselves. How do individual charter schools or charter school chains differ in the extent to which they employ the four tactics described above? Why did the report only look at the data on students in kindergarten through 3rd grade? In middle schools, where every grade takes high-stakes standardized tests, does the charter sector employ the four tactics to an even greater effect? How does the fact that charter schools only accept students who actively apply to their school impact the overall attrition patterns? As the original report asks, do “other factors such as unobserved differences in student characteristics contribute to some of the gaps in mobility patterns?”


An objective observer would expect that any updated report, such as the one the Independent Budget Office just released, would address at least some of the above questions. But it did not. Instead the “Independent” Budget Office folded under the pressure brought to bear by charter school advocates and their paid researchers. Immediately after the original report was released, a “researcher,” paid for by the Walton Foundation, complained that the report only looked at the highest need special education students and not all special education students.[iv] While true, this has no bearing on the four tactics that the report conclusively showed the charter sector employs to game their results.


A couple of weeks ago, the “Independent” Budget Office, caving to the pressure, “updated” the report to include a cohort of students through 4th grade. Their “finding:” across all special education students, such students are slightly more likely to remain in charter schools than public schools. The media parroted these claims. This “finding” is, however, entirely bogus. Instead of using the categories that generally correspond to the level of need (namely whether the student requires a self-contained class or can be supported in mixed classes or even classes that are entirely general education), the report uses the named disability category (namely speech impaired, learning disabled, other health impaired, all other disabilities) of each student. This, of course, tells us nothing about the severity of each student’s need within the category. Instead it covers up the fact, which we already know from the original report, that charter schools are much more likely than public schools to selectively attrite the students with the highest level of special education needs, the very same students who are most likely to bring down their test scores. But now special interests and the media can trumpet the fact that charter schools in New York City keep their special education students at higher rates than public schools.


Ignored in the new analysis is the fact that the charters serve an entirely different mix of special education students, i.e. students much, much less likely to require the highest level of accommodations and supports. Importantly, the “Independent” Budget Office did not just add these broader-brushed approaches to the analyses of the original report; it declined to repeat, with the updated dataset, the analysis of the sky-high attrition rates of highest need special education students at charter schools. It declined to repeat, with the updated dataset, the analysis of higher attrition among students with low test scores at charter schools. It declined to repeat, with the updated dataset, the analysis of student creaming by charter schools. It declined to break down the updated dataset so that we could learn more about the tactics employed by individual charter schools and charter school chains. It declined to even look at the unanswered questions about charter middle schools.



Fortunately, for those interested in the truth, at about the same time, other data were released showing just how much the charter sector in New York City relies on tricks, rather than true educational innovations, to produce their “results.”[v] The data break down the comparisons between charter schools and public schools, school by school and district by district.[vi]


It turns out that the charter sector suspends students at rates up to twenty-six times higher than public schools in the same geographic region, despite the fact that the charter schools serve only a self-selected student body.[vii] These data may explain how charters are able to selectively attrite the most troublesome students who bring down their test scores. They harass the students until they leave for the public schools, which are of course morally and legally obligated to accept every student.


It turns out that public schools serve up to five times as many students living in temporary housing as charter schools in the same geographic region.[viii] This little fact may be one of those “other factors,” mentioned in passing by the “Independent” Budget Office, that explain why public schools have a slightly higher overall student transition rate than charter schools. Obviously kids with no permanent home are more likely to move around and switch schools.


It turns out that all of the highest-flying charter schools serve a much, much more advantaged student body than the local public schools.[ix] It is almost shocking to see not a single charter school represented among the schools in the top quarter of student need, in any of New York City’s 32 geographic regions. The gaps in student need are even higher when looking at charter schools co-located in the same buildings as public schools.[x] In co-locations the public school serves up to six times as many students living in temporary housing, up to twenty times as many English Language Learners, and many multiples the number of special education students as the charter school in the very same building!


What we have here is a failure to tell the truth. The “Independent” Budget Office, aided by a compliant press, has whitewashed the story of inequity that it itself had helped flesh out just a year earlier.


The data could not be any clearer. Charter schools have no secret sauce. In fact, they are creating more segregation and greater inequity in our school system. The time has come to end the charade. Charter schools must be folded under the umbrella of the public school system. We must then have the difficult conversations that have been avoided due to all the tumult and distractions caused by the charter school corporate reform agenda.[xi] How do we serve all students in a nation with significant, perhaps increasing, opportunity gaps? What can schools do to help mitigate the overwhelming disadvantages that students growing up in poverty face? Since it is obvious that schools can’t do much in isolation what can we, as a nation, do to support schools in their work of providing opportunity to all students?




[ii] http://www.ibo.nyc.ny.us/iboreports/2015charter_schools_public_schools_attrition.html


[iii] Apologies in advance for the generous use of scare quotes. But it’s almost impossible to tell this sordid tale without them.


[iv] This researcher had, by this time, already made many claims about special education students and charter schools in New York City that had been debunked. See for example http://nepc.colorado.edu/thinktank/review-why-the-gap


[v] http://www.uft.org/files/attachments/secure/charter-school-suspensions.pdf


[vi] The fact that the teacher’s union had to collate this data and not a single “independent” journalist could be bothered to do so, despite the regular appearance of newspaper stories and editorials praising charter schools, tells us just how biased the media is when it comes to education policy. It suggests that media outlets would benefit by being more skeptical of charter school claims when deciding upon and reporting upon their stories.


[vii] http://www.uft.org/files/attachments/secure/charter-school-suspensions.pdf


[viii] http://www.uft.org/files/attachments/secure/demographics-charters-v-traditional.pdf


[ix] http://www.uft.org/files/attachments/secure/peer-index-explainer.pdf


[x] http://www.uft.org/files/attachments/secure/colocated-schools-sharing-buildings.pdf


[xi] It may not take a conspiracy theorist to assume that this conversation is exactly what the special interests groups that back charter schools want to prevent from happening.