|Illustration by Timothy Cook.
Do Charter Schools Measure Up? The Charter School Experiment After 10 Years
American Federation of Teachers, 2002.
Teacher unions are pulled in different directions. On the one hand, many of their staffers have devoted their lives to education and are genuinely committed to improving schools. Indeed, under the late Al Shanker, the American Federation of Teachers (AFT) was known for tough-minded talk on issues like standards, testing, and failing schools.
On the other hand, as unions their primary mission is to protect the welfare of their members. Sometimes the interests of teachers and students are aligned, but not always. For instance, unions have steadfastly defended policies that undermine education reform, such as: No junior teacher, however competent and in demand, should ever be paid more than someone who has been on the job longer. No teacher, however talented and knowledgeable, may enter a classroom without state certification. In a 1997 speech, Bob Chase of the National Education Association (NEA) declared, “If there is a bad teacher in one of our schools, then we must do something about it.” The union acted as if this were a bold, meaningful concession. To the rest of us it was a no-brainer.
The national unions deliver the occasional encomiums to accountability. However, such rhetoric is rarely matched by action among their relatively autonomous local satellites, like the United Federation of Teachers in New York. Charged with negotiating actual contracts, these tend to behave more like traditional urban unions, defending their members at every turn, no matter what the consequences for schools. After all, serious academic standards might highlight the poor performance of some teachers, who might face sanctions (like–shock!–being fired). Likewise, school choice would empower parents, who might choose nonunionized private and charter schools.
Nowhere have we seen a better example of the unions’ clumsy attempts to straddle representation and reform than in the AFT’s handling of charter schools. Consider the AFT’s recently released “study” of charter schools, Do Charter Schools Measure Up? The Charter School Experiment After 10 Years. The report touts the fact that Shanker supported charter schools, but then claims that “these schools are a diversion from reformers’ and policymakers’ efforts to improve education in America.” It alleges that a review of the research on charter schools leads to the conclusions that, overall, charter schools: 1) fail to raise student achievement more than traditional district schools do; 2) aren’t innovative and don’t pass innovations along to district schools; 3) exacerbate the racial and ethnic isolation of students; 4) provide a worse environment for teachers than district schools; and 5) spend more on administration and less on instruction than public schools. With all this evidence apparently stacked against charter schools, it seems downright responsible of the AFT to call for a moratorium on further charter school expansion “until more convincing evidence of their effectiveness and viability is presented.”
If only that were all true. The AFT’s conclusions, you’ll see, are based on a selective reading of the research, shameless spinning of research findings, and a failure to place findings in context. The report ignores the judgments of parents and students, uses bizarre definitions of such terms as innovation and accountability, compares charter schools with the ideal school rather than with traditional district schools, and presents confusing and out-of-context discussions of such admittedly complex matters as school finance and student achievement. In short, the report poses as serious research, but is more about lobbying than a search for clarity.
There is no doubt that the AFT cites some good research, usually accurately. For instance, the AFT acknowledges, after some hemming and hawing, that most charter schools spend less public money than most district schools. The AFT admits that charter schools are generally not “creaming” the more capable students. The AFT is right that charters are no panacea, as a few charter boosters had predicted. The AFT expresses legitimate though perhaps overstated concerns about the viability and effectiveness of for-profit education.
In addition, the AFT rightly points out that not enough has been done to measure how much learning occurs in charter schools. But the analytical flaw here repeats itself throughout the report. The AFT persistently compares the situation in charter schools with that of some mythical school district where only innovative, research-tested curricula are used; students of all races, ethnicities, and income levels mix happily; and schools and school employees are held accountable for student performance. Yes, the performance of charter schools has been inadequately measured. Yet the same is true of district schools, and the unions are partly responsible. They have fought off efforts to yardstick student achievement for decades, fearing that good measures will make some teachers and schools look bad. The AFT’s sudden cheerleading for accountability in the case of charter schools reminds me of the old Yiddish definition of “chutzpah”: a young man kills his parents and then asks the judge for leniency because he is an orphan.
Moreover, the AFT uses a remarkably static definition of accountability. It claims that charter schools are no more accountable than district schools since most states now have accountability plans. Unlike district schools, however, charter schools are actually held accountable. When a charter school fails to satisfy its parents and/or sponsors, it loses customers and closes down. While only 4 percent of charters have shut their doors, every operator of a charter school lives in fear of demise. By contrast, when a district school fails, school employees face the truly awful consequence, in most cases, of drafting so-called “school improvement” plans that promise long-term reform. They may even get more money and a raft of consultants to help them shape up. Once in a while more drastic measures will be taken, such as reconstituting the school with new staff, but the old staff is usually just shuffled around the district (the venerable tradition known as “passing the lemons”). Charter schools face market accountability. They either provide an education that attracts parents or they lose their enrollment and funding.
The AFT uses an equally peculiar definition of innovation. To the AFT, innovation seems to mean inventing something never before seen on Earth. This is akin to saying that Apple Computer in 1980 was not an innovative company since it did not invent computers; it merely made and marketed computers small enough for home use. By contrast, economists like Douglas Greer describe three types of innovation: inventing a new product or service; modifying it for public use; and disseminating it to the public. Charter schools have done a little of the first and a whole lot of the second and third types of innovation. For example, Montessori curricula were invented nearly a century ago, but remain too “innovative” for the vast majority of district schools. In the past, parents had to turn to expensive private schools if they liked the Montessori approach, but now many charter schools offer Montessori curricula in response to parental demand. (Notably, some district schools have established Montessori programs to fend off competition from charters.) Similarly, the AFT claims that for-profit management companies “do not contribute to innovation because they offer a single, ‘cookie-cutter’ school design, curriculum, and technology package to all the schools they operate.” But in the environment of public education, where successful programs are rarely studied and replicated, any company that manages to disseminate effective school designs and curricula to a large number of schools should be considered very innovative.
In my research I have identified 34 different examples of charter school innovation, including small size; untenured teachers; contracts with parents; real parent and teacher involvement in school governance; outcome- (rather than input-) based accreditation; service learning fully integrated into the curricula; unusual grade configurations; split sessions and extended school days and years to accommodate working students; and computer-assisted instruction for at-risk and other frequently absent students. It appears that most charter schools were founded to pursue one or more of these 34 innovations. Of course, the biggest innovation of all is inherent to charters: allowing parents to choose their children’s schools or even start new schools.
The AFT is correct to point out that not all charter schools are innovative. Yet on the whole they seem much more innovative than district schools, which is after all the point. A few charter schools, such as Edupreneurship in Arizona, which uses a token economy in an open classroom setting to motivate students, seem to have invented new curricula. A few charters have invented new modes of school governance, such as the Charter School of Sedona, where master teachers control their classroom budgets, including their own salaries and those of their aides (whom they hire and fire). More typically, charter schools have refined and disseminated existing practices that district schools were reluctant to use–a nice service for parents who may not care whether a program is “innovative” as long as it works for their children.
Sins of Omission
When the AFT isn’t using definitions convenient to its conclusions, it simply ignores information altogether. For instance, numerous surveys have found that students and parents who transferred from district schools to charter schools thought the charters were safer, friendlier, and more effective, often by margins of more than 50 percent (see Figure 1). For example, in a 1997 survey of charter school students who used to attend traditional public schools, 65 percent said their charter school teachers were better than their previous teachers. Only 6 percent rated charter teachers worse. This may explain why roughly seven out of ten charter schools have waiting lists. My own fieldwork suggests that older charter schools are particularly likely to be oversubscribed, based on their track records. The AFT apparently thinks that the opinions of parents and students just don’t matter in evaluating charter schools. In fact, they cite with contempt the fact that in some instances “teachers, students, and parents successfully lobby to keep their charter school open” when authorizers attempt to shut them down, often for political rather than academic reasons.
The AFT’s strategy of selective reporting also colors its approach to the question of whether competition from charter schools has forced changes in district schools. At the state level, teacher unions and school districts, when not opposing charter schools altogether, have lobbied intensively to place strict caps on the number of charter schools and to limit the number of institutions that can grant charters. (Now this stance, once the province of just local unions bent on protecting their monopolies, appears to have become national policy for the AFT.) In the 20 states where the AFT has succeeded in restricting the authority to grant charters to school districts, an average of 26 charter schools are open. The 18 states where other institutions, such as universities and local governments, can grant charters have an average of 96 schools. The AFT fails to note this in criticizing charter schools for not providing enough competitive pressures for district schools. In other words, the AFT chastises charter schools for a policy environment it played a major role in creating.
Where strong charter laws exist, as in Arizona and a few cities, considerable peer-reviewed research (not cited by the AFT) finds that districts do in fact respond to competition by working to improve. As my team reported in Small Districts in Big Trouble, Arizona districts that lost a tenth to a third of their students to charter competition reacted with changes in leadership and curricula. In School Choice in the Real World, we reported that competition prompts districts to empower their teachers and increase outreach to parents.
Similarly, in Revolution at the Margins, Frederick Hess reports that limited competition had little impact, but the threat of serious competition from charter schools and vouchers in 1995-’96 led Milwaukee Public Schools to reform with Montessori options, decentralization, tougher graduation requirements, more transparent school report cards, advertising, and empowerment of their more innovative principals, who had previously been treated with contempt. A school board member supportive of the city’s teacher union recalled, “It was choice and vouchers that encouraged the [school] board to pull together with the union to create innovative schools. Suddenly you had a union that said, ‘Yeah, we like this idea, let’s do it.’ ” After a court challenge blocked charter school expansion, the union and its allies “went back to their old ways.” Harvard economist Caroline Hoxby’s quantitative analyses suggest that competition from vouchers in Milwaukee and from charters in Michigan and Arizona have improved the test scores of all students, even those “left behind” in district schools.
When the AFT can’t ignore the numbers, it just changes them. For instance, the AFT claims that Humboldt State professor Eric Rofes, in How Are School Districts Responding to Charter Laws and Charter Schools? found that only a quarter of district schools responded to competition from nearby charters. Actually, Rofes found that 48 percent of the districts exhibited moderate or high levels of response to charter school competition. He also found that districts that had lost funding to charters were particularly likely to respond, just as market theory predicts. This would seem to support arguments for more charter schools, not fewer.
Similarly, the AFT misstates the findings of a report by economists Michael Podgursky and Dale Ballou. The AFT claims that the authors found that charter schools determine pay “in a similar manner to most school districts,” but Podgursky and Ballou in fact found charters far more likely to use merit pay and far less likely to use traditional salary schedules. Notably, while most teachers in traditional public schools are tenured and have multiyear contracts, 96 percent of charter teachers in their study were either at-will employees or had annual contracts; thus charters can and do separate ineffective teachers.
Likewise, the AFT says that about 8 percent of charter schools have closed, but this figure includes charter schools that never actually opened or that were consolidated into local school districts. In fact, as noted above, the Center for Education Reform reports that only about 4 percent of charter schools have closed, not a bad failure rate for a new program.
Finally, in an extremely muddled discussion, the AFT reports that charter school student-to-teacher ratios “generally match or exceed” those of their host districts. I find more credible the statistics from the U.S. Department of Education-sponsored report The State of Charter Schools 2000 showing that charter schools have a median student-teacher ratio of 16 to 1, 7 percent lower than that of district schools.
The AFT’s misreporting of charter school finances may be more understandable, since school finance is less transparent than Enron’s balance sheet. For example, the AFT first states, “Charter school salaries tend to be competitive with other public schools at the beginning-teacher salary level and less competitive for more experienced teachers.” Yet later the AFT declares, “Charter school teachers are paid less than other public school teachers, particularly when their teaching experience and education are considered.” Further on, the AFT reports that charter schools spend more on administration and less on instruction than traditional schools.
I suspect that the AFT’s confused “findings” reflect three interrelated factors. First, the vast majority of charter schools are new schools. As such, the vast majority of charter school teachers have less than six years of teaching experience, typically less than half the average for nearby public schools. Since teacher salaries are the primary instructional expenses for schools, it should come as no surprise that charters spend less on classroom instruction than traditional public schools, whose teachers are older and thus further along on the salary schedule. My fieldwork and other surveys suggest that teachers who choose charter schools tend to be attracted by a school’s curricula and mission. The more senior teachers who come to charter schools (often as a second career) are frequently willing to forgo higher salaries out of dedication to their school’s mission. The AFT may find this difficult to believe, but many teachers are idealistic: they are willing to trade lower salaries for the in-kind benefit of working with a curriculum they believe in.
Second, as new and underfunded schools, charters must spend a higher than normal percentage of their resources on their buildings, leaving less money for salaries. Nearly all the operators of the 29 charter schools I’ve visited lamented facilities problems. Indeed, there is a long line of research on charter school facilities challenges. State legislatures have been reluctant to fund building programs, and charter schools, unlike school districts, can’t float bonds to pay for capital spending.
Third, the finding that charter schools have more “administrators” than traditional schools ignores the fact that most charters are very small schools. The U.S. Department of Education reports that as of 1998-’99, the median charter school had 137 students, compared with 475 in all public schools. Many charter principals still teach, while principals at traditional public schools are a decade or more out of the classroom. Which sort of administrator is more likely to inspire the teaching staff?
A similar lack of context plagues the AFT’s discussion of the makeup of student bodies in charter schools. The AFT claims that charter schools are more racially homogeneous than district schools, citing research that makes much of very small differences (normally less than 10 percent) between charter and nearby district school student bodies. What the AFT fails to acknowledge is that charter schools are more likely than district schools to promote integration, since in most charter schools white and minority kids take the same courses, while in many district schools minority kids are placed into nonacademic tracks.
The AFT also fails to present a sophisticated discussion of school market locations. Due to their problems with obtaining facilities, charter schools tend to locate in low-rent areas, while drawing students from miles around. This makes comparisons with the nearest district schools highly misleading. Scott Milliman and I found, after correcting for this and other errors, that one of the key studies cited by the AFT as alleging racial concentration in charter schools in fact found charter schools no more segregated than district schools, with the notable exception of those charters that had converted from private schools. Furthermore, findings that charter schools for at-risk students tend to have proportionately more minorities than district schools are no surprise, since proportionately more minority students are at risk. Nationally, charter schools are 52 percent nonwhite while district schools are 41 percent nonwhite, suggesting that on the whole charters are serving traditionally underserved populations.
Context is a major issue in the AFT’s reporting on student achievement as well. The AFT correctly reports that most kids in charter schools seem to do about as well as in district schools, controlling for demographic factors. This is in fact less success than charter boosters predicted. The AFT fails to note, however, that most charters are very new schools; 14 percent of the nearly 2,400 charter schools operating in the 2001-’02 school year were in their first year, and another 23 percent were in their second year. As a result, the studies cited by the AFT compare many charter schools in their first or second year with district schools with decades of experience and deep pockets behind them. A wealth of scholarship suggests that first-year charters face serious start-up problems, particularly regarding curricula and personnel. As the RAND study of charter schools and vouchers, Rhetoric Versus Reality, argued, “Judging the long-term effectiveness of the charter school movement based on outcomes of infant schools in their first two years of operation may be unfair, or at least premature.” A more apt comparison would analyze charter schools in their third year or older. As it is, for the charters to be doing as well as traditional schools is nothing short of remarkable. Furthermore, many parents chose charters because their children were failing in district schools, meaning that charters have very challenging kids to teach.
The schizophrenic personalities of the teacher unions are on full display in Do Charter Schools Measure Up? A decade ago, the AFT is fond of telling us, the AFT claimed the mantle of reform by advocating charter schools as a way of promoting innovation and sidestepping administrative bureaucracy. But now that the charter school movement has grown to a point where it actually threatens the monopoly of unionized school districts and the salaries and perks of teachers, the AFT is changing its tune. This is unfortunate. As scholar Bruce Fuller points out, charter school proponents need “a devil’s advocate, a loyal opposition,” a role played by the RAND Corporation and by academics like Fuller himself. But whereas RAND calls for more experimentation so that more evidence can be gathered, the AFT, revealing its real purpose here, wants charter schools to be choked off in their infancy. The scary thing is how powerful their lobby can be. But so far, parents appear to be more powerful.
-Robert Maranto is an assistant professor of political science at Villanova University and associate scholar at the Goldwater Institute.