CHARTER SCHOOLS -- A RESPONSE TO ANDREW BLAIS

Andrew Blais [an old friend and former student, whose fine doctoral dissertation I directed], nudges me to post something about charter schools. My familiarity with the subject, which is shallow and quite recent, comes from a session of my doctoral seminar on "Normative Dimensions of Public Policy," which I am teaching this semester in the graduate program of the Public Policy Department at UNC Chapel Hill. The first of what will be six two-hour seminar presentations by pairs of students in the seminar was devoted to the subject of charter schools, and in preparation for the session all of us in the seminar read several essays and chapters from books set us by the two students presenting.

The charter school movement, which now enrolls more than a million students in almost five thousand schools nation-wide, grew out of critiques mounted both from the left and from the right of the traditional public schools in which America's children have been educated for several centuries now. From the left came a demand for community control of what were until then centrally run city school systems, as well as a call from Albert Shanker, the progressive head of the New York City teachers' union, for small-scale experimentation. From the right came calls for privatization, the introduction of market mechanisms via a voucher system, and the breaking of the teachers' unions, as well as for an assortment of religiously based revisions to the curriculum [such as the excision of any mention of evolution, the blanchissement of American history, the enshrining of the Christian God in a position of prominence, and so forth.]

A variety of reforms or experiments have been proposed and tried, among them making classes smaller, making schools smaller, lengthening the school day, and lengthening the school year. The principal impetus for all this churning about is the comparative standing of American school children in international measures of competency in math, science, and language arts. One response to all of this by the federal government has been to mandate testing of school children repeatedly, roughly on the theory that if a patient has a fever, the best medicine is to take the temperature repeatedly. The subject of testing is too large to be discussed here, and is obviously worth a full-scale post of its own.

A large amount of foundation money has been poured into the charter school experiment, including some very large donations by the Bill and Melinda Gates Foundation and several others. I think it is fair to say that driving the movement is a deep-seated hostility to traditional schools and teachers, coupled with the conviction that smart, innovative businessmen and women ought to be able to run schools better than moss-backed career educationists comfortably nestled in tenured sinecures. [Am I too blatantly revealing my sympathies here?]

Charter Schools are privately run schools, "chartered" by the state and supported by funds that would otherwise be spent on educating the same children in public schools. The charter schools are liberated from the supposed tyranny of the teachers' unions, and are free to adopt innovative curricula and teaching methods and any other ideas that the traditional schools are thought to be too hidebound and encrusted with bureaucratic barnacles to try.

The charter schools have certain obvious advantages, the most important of which is that they are not required by law to accept every child in a neighborhood. Nor are charter school entrepreneurs required to maintain schools in every neighborhood in a city. In addition, the children enrolled in charter schools are there because their parents chose to put them there, and that implies a level of concern for education that would, one might suppose, constitute a sort of positive self-selection process resulting in better outcomes for the children.

One of the readings set by Casey Megan and Lisa Spees, the two students responsible for the presentation, was a report by the Center for Research on Education Outcomes [CREDO] located at Stanford University of a 2009 study called "Multiple Choice: Charter School Performance in 16 States." The structure of the study strikes me as really quite brilliant, and I want to take a little time to explain it as well as I can.

There are two points that we must begin with. First, what really matters is what is called the "value added" effect of charter schools, which is to say how much improvement in student learning they accomplish, as measured by standard state-wide tests, compared to what students would have learned in the traditional public schools, had they stayed there rather than transferring to the charter schools. Now, this is asking a counterfactual question: What would the students have learned if, contrary to fact, they had stayed in the traditional schools. Since it is impossible to rewind history so as to run the students through both kinds of schools, the challenge is to come up with some way of answering the question.

What the CREDO study did was to create a huge number of "virtual twins," by pairing each charter school student whose performance was being monitored with a traditional public school student who matched the charter school student in a number of ways thought to be significant. The match factors included grade-level, gender, race/ethnicity, free or reduced price lunch status [an indicator of family income level], English language learner status, special education status, and prior test score on state achievement tests.

Each pair of students thus consisted of a charter school student and a public school student who with respect to all of these factors were identical. The report says that 1.7 million records were analyzed, but I do not think that means 1.7 million students [although it may have.] The study then took a look at the performance of these pairs on the next year's state achievement tests [can you imagine the computer program required to do this?] Some standard statistical number crunching then ascertained both what the average differences were, state by state, between the charter school students and their paired traditional public school matches, and what statistical significance, if any, the data exhibited. I may just be new to this sort of experimental investigation, but I think this is an extraordinarily imaginative way to try to get at the effect of charter schools on the education of their pupils.

I confess that as I began to read this report, I assumed it would reveal important relative gains by the charter school students. How could it be otherwise? The effect of the self-selection process alone, I thought, was sure to produce some important relative gains, and I was all prepared [given my preliminary bias] to conjure up arguments about equity and the hidden injuries of class [to steal Richard Sennett's fine phrase]. But the results astonished me.

To put it as simply as I can, the study showed that by and large transfer to charter schools made little or no difference at all in the gains of educational gains of students, and where there was a difference, it was a small but statistically significant edge for the traditional public schools! I was astonished by these results.

In the class discussion, which was very interesting and informed by a good deal of hands-on experience and knowledge among the students [not me!], one point was made that merits repeating. Because the study amalgamated such large swaths of data, it was unable to disambiguate the performances of especially successful charter schools. Several students spoke about the Knowledge is Power Program [KIPP] charter schools, which apparently have quite dramatic successes to their credit. The KIPP schools combine a longer school day and school year with heavy homework assignments and a strong emphasis on team performance, all of which seem to be very effective in improving student performance. That sounds right to me, intuitively.

I would be very interested in anecdotal or other reactions from the readers of this blog.