With their passage of charter laws during the 1990s, the states made a dramatic system change in public education. This was an unexpected development in state policy. It was radical. It was, and remains, controversial. It continues to puzzle people. What is it? Why are we doing it? Is it working? How do we know?
At the moment, the country is not thinking clearly about these questions. The confusion is visible as people struggle for some term to distinguish the system change from the schools created ... often failing, and using “charter schools” to refer to both these different things.
Sorting out the confusion begins with getting the terms right. This is easy. Let’s bury “charter schools.” Let’s say chartered schools when we refer to the schools created—and chartering when we mean the new state strategy.
Any sensible effort to evaluate the strategy starts by understanding this: Chartering is an institutional innovation. It is the state moving beyond the effort to transform existing schools, moving to create quality schools new. To do this, the state lets someone other than the superintendent start and run a public school; lets someone other than the local school board offer public education.
Most evaluations of this strategy have not examined chartering as an institutional innovation. They have simply looked at the schools created, asked about their characteristics and about the students they enroll and their test scores. This is too simplistic. The central question is whether chartering is helping state policy leadership out of its probably hopeless effort simply to change existing schools.
Changing schools is hard to do within the traditional, regulated-public-utility arrangement. With students required by law to attend, and holding an exclusive franchise on public education within their boundaries, school districts could—as Albert Shanker said in Minnesota in 1988—"take their customers for granted.” This circumstance has made K-12 education an inert institution. And it has made “school improvement” essentially an effort to push in changes from the outside. We have been treating public education like a patient in intensive care that is hooked up to an external life-support system which, through an arrangement of ropes, pulleys, tubes, and wires, supplies the needed nourishment and stimulus.
When policymakers grew impatient with the schools, they pushed in the standards, measurements, and consequences that the institution could not introduce itself. In 2001, with the impetus of the federal No Child Left Behind Act, they commanded that the K-12 education institution Do Better! But it does not work to order people to do what in fact cannot be done.
Clearly, it is ridiculous to be pushing in improvement while leaving the institution structured to resist change. Good sense suggests changing arrangements, so that the inert institution becomes a self-improving institution: one in which organizations do the improvement themselves—on their own initiative, in their own interest, from their own resources—and in which “consequences” are imposed without political action.
Is chartering replicating good models and good practice?
In the 1990s, the charter idea showed the states that they need not be captive to the public-utility arrangement of K-12 education; they are not limited to simply changing the schools we have. The arrangement exists in state law; the state can change the existing law.
Governors and state legislators probably would not explain their actions this way. But it is impossible to account for the chartering laws—enacted without strong outside support and against the opposition of powerful interests—except in terms of the state policy leadership’s frustration with the inability or unwillingness of districts to educate students and to control their operating costs. Also playing a role, no doubt, was an instinctive sense that the state’s moving to “get somebody else who will” would introduce a useful dynamic into school reform.
How should we assess this new state strategy? A commissioner of education preparing to evaluate chartering might ask the following questions:
1. Is chartering replicating good models and good practice?
Districts and schools must change traditional practices radically for all students to learn. The cultures and values of organizations work powerfully against radical change, as the Harvard University business professor Clayton M. Christensen showed in The Innovator’s Dilemma. Paul Houston, the executive director of the American Association of School Administrators, says: “If you are a superintendent, you are in the ‘incremental’ business.”
More-than- incremental changes, needed quickly, may be introduced faster into K-12 education if they are built into schools started new. Evaluation should ask how far proven pedagogical, personnel, financial, and other practices are appearing; in what kinds of schools, operating under what provisions of law. Chartering is partly about replication; not every new school needs to be innovative.
2. Is chartering generating new models and new practices?
Some of these new schools may be innovative. The people organizing them get to try new things. So chartering is partly a research-and-development program, producing new models for learning, governance, and management—teacher professional partnerships, for example.
Evaluation must factor in this character of chartering as a trial-and-error program. Some experiments will not work well in their early years. Some will not work at all. But a failed experiment is not a failed program. And we learn from failure. No sensible business would evaluate its R&D process by the no-defects standard it sets for the proven models in its production processes. So chartering can be succeeding even though not every chartered school is succeeding.
3. Are the chartered schools succeeding with student learning?
Learning is one measure of chartering’s success. But evaluation must be cautious about generalizing, about talking as if chartered schools were all the same, as people do when they ask: “Do ‘they’ work? What students do ‘they’ enroll? How are ‘their’ students scoring?”
Studies that generalize about "charter schools" are almost always inconclusive. They're an embarrassment to the research community.
The institutional innovation does mean that the schools are alike in one sense: The chartered sector is organized on principles different from those in the traditional, district-led sector. Its schools have boards removed from electoral politics. The schools have authority to decide how to teach, select their own personnel, and handle their finances. They are to be accountable for performance, not for “compliance.” This autonomy is important.
The differences that result are enormous, among schools within a state and between states, especially in terms of what directly affects learning. The learning model adopted is a school-level decision. Some schools will use a proven model; some will try a new model. Some have teachers talking to kids in groups; some use project-based learning. Some are high-technology; some low- technology. Some are small; some are not small. In some, kids dress as they choose; in others, they wear uniforms. Some staff up on the “bureau” model used in districts; some contract for services. Some have principals; some are run by teacher-partnerships. Some are free-standing organizations; some belong to a managed group. Some are nonprofit; some commercial. And on and on.
Because of this, studies that generalize about “charter schools” are almost always inconclusive. They usually report that “the evidence is mixed,” as of course it would be. Studies of this sort are an embarrassment to the research community.
The tendency to generalize is driven also by the policy, and by political controversy about opening public education to schools other than those owned and run by the local board of education. Some people like this idea. Others hate it—because it offends their ideology, or because they feel their interests threatened by it. The intensity of the debate puts advocates on both sides under intense pressure. Too often, the advocates for one side or another argue simply in terms of “public (district) schools” and “private schools,” and “voucher schools” and, unfortunately, “charter schools.”
Evaluation should stay clear of this impulse to generalize. It should, instead, look beneath the label, look inside the sector to identify, then compare, the different kinds of schools. The schools chartered differ, and the differences make a difference. We want evaluation to show which differences make what difference. In schools of what sort do which students do better? In schools of what sort do they do less well?
Jeffrey R. Henig, a professor of political science at Teachers College, Columbia University, has been stressing the need to disaggregate data in order to draw meaningful conclusions. In looking at chartered schools in the District of Columbia, his research team identified essentially “mission oriented” nonprofits and schools run by “market oriented” groups, then examined how the two differed in terms of school size and autonomy. This proved useful. Evaluation should respect the different dimensions—and measures—of success. The academic realm is important, but the public also wants schools to be safe, orderly, and responsive, and to develop life skills, physical fitness, good character, and the motivation for lifelong learning. Doing well on tests is not the only measure of student learning, and the level of test scores may be a poorer measure of school success than the rate of gain. Schools pressed to show high scores may respond like ceos pressed to show high quarterly earnings, training their employees in the tricks of scoring high, reducing what they truly know and are able to do.
4. How do the two strategies compare?
The prevailing notion is that existing schools will be transformed, like caterpillars into butterflies, as people who mean well and get resources try hard and act decisively after experts train them in how to “do better.” The suggestion implicit in chartering is that the schools we want will be developed faster by creating them new.
Evaluation should simply tell us whether chartering does some things that cannot be done as well in "regular" schools, or does them more quickly.
Neither theory need work perfectly: Evaluation should simply compare the two; tell us whether chartering does some things that cannot be done as well in “regular” schools, or does them more quickly. We are getting some of this evaluation now, as people compare the two approaches to creating quality high schools—by breaking down large existing schools, and by creating small schools new, often through chartering.
Policy is very curious whether districts change faster when chartered schools appear, whether districts pick up improvements and innovations that appear in the chartered sector. But this is tricky: The “ripple” analogy is false. Chartered schools do not, cannot, change district schools. Their presence has an impact on the district enrollment, perhaps on the district’s pride. But evaluation must understand that only the district can change district schools. Evaluation should look for responses on this question; where they appear, it should ask which boards or superintendents decided to respond, which did not, and why.
Chartering will need to be evaluated over time. Nobody gets everything right on the first try. Most things improve, evolve. This applies both to the schools and to the laws, the state policies supporting chartering. Perhaps as the chartered sector grows, the districts will change more rapidly.
Evaluation should draw conclusions; make recommendations. It should be diagnostic, as testing is diagnostic, telling us what is going well and what is not, and how to adjust. With chartering, we want to know which pedagogical, governance, and management practices succeed—and what provisions of law are responsible—so that policy can do more of what works better.
An appropriately sophisticated and dispassionate evaluation of chartering will be immensely important in answering what is now the critical question in the country’s effort at improvement.
Up to now, the notion has been that we would—and could—get the schools we need just by changing the schools we have. Probably this will not work, has not worked. It is an astonishing risk—imprudent—to be leaving all the chips bet on districts’ doing what districts have never been able to do. We are not compelled to take that risk, since we can now generate quality schools new. So the risk is really not an acceptable one to be taking, especially with other people’s children.
We can reduce this risk by adding an effort now to create schools new. In almost every field, we balance building-new and rebuilding. In K-12 education, too, change and improvement probably will come more through building-new than through reform; more by replacing schools than by transforming them.