5. GETTING SCHOOL REFORM WRONG The Arrogance of Top-Down Coercion

Do It My Way ... or Else

THERE ARE basically two ways to change what happens in schools. The first might be called the support model, and the second, the demand model.¹ The support model begins with the premise that the role of teachers, administrators, parents, public officials, and the community at large is to help students act on their desire to make sense of the world. The school should guide and stimulate their interest in exploring what is unfamiliar, constructing meaning, and developing a competence at (and a passion for) playing with words and numbers and ideas. Improvement is seen as something that tends to follow when students are provided with stimulating, worthwhile tasks. Students are not just expected to take responsibility for their own learning but are actively assisted in doing so.

In the demand model, by contrast, those outside—and, figuratively speaking, above—the classroom decide what the people in it are required to do. Lists of specific achievement goals are imposed on teachers and students. The methods and metaphors of this process are often borrowed from the corporate world, with much talk of results, performance, accountability, and incentives. Children are even described as "workers" who have an obligation to do a better job. Schools represent an "investment" and must become more "competitive," the idea being that test scores in the United States ought to surpass those in other countries. Education is described as though it were a hybrid of an assembly line and a sports match.

The latter approach, of course, describes the Tougher Standards movement to a T. In fact, that movement is true to the historical roots of the word "standard," which originally referred to the place on the battlefield where the king issued orders to his army.² A cartoon published in the early 1990s nicely captured the current direction of school reform: the president is shown atop a tank, its weapon trained on a schoolchild sitting in a desk down below. The president bellows a single command: "LEARN!" This may be the sort of image that Elliot Eisner of Stanford University had in mind when he warned that

the language of standards is by and large a limiting rather than a liberating language.... It distracts us from paying attention to the importance of building a culture of schooling that is genuinely intellectual in character, that values questions and ideas at least as much as getting right answers.... The challenge in teaching is to provide the conditions that will foster the growth of those personal characteristics that are socially important and, at the same time, personally satisfying to the student. The aim of education is not to train an army that marches to the same drummer, at the same pace, toward the same destination. Such an aim may be appropriate for totalitarian societies, but it is incompatible with democratic ideals.³

Not content with requiring that all students learn the same things, some people want them to learn the same things at the same time. Some even want this to happen across the whole country. Such standardization probably wouldn't enhance learning,⁴ but it is the whole idea of imposed standards, not just the prospect of nationalizing them, that should give us pause. Let's face it: many of the people doing the imposing, even at the state or local level, know precious little about teaching and learning.⁵ Of course we wouldn't expect most state legislators to be aware of the latest research on pedagogy, just as we wouldn't expect them to understand the latest research on kidney disease. The difference is that while few lawmakers would presume to tell physicians when to recommend dialysis, many think nothing of dictating a particular method of teaching children to read.^*

The demand model doesn't make much sense for other reasons as well. "Nothing has brought pedagogical theory into greater disrepute than the belief that it is identified with handing out to teachers recipes and models to be followed in teaching." So said John Dewey when the twentieth century was new, and his point is still well taken.⁶ Yet these pedagogical theorists share with legislators, curriculum publishers, and others a persistent belief in the feasibility of a "teacher-proof curriculum." This belief is not just insulting to teachers but corrosive to the very enterprise of schooling, which becomes less like educating and more like training. In the words of two school reform experts, "The more the curriculum is specified and defined externally, the more the role of the teacher becomes that of the technician, expected to put into play decisions made by others outside the school. This is true whether the external source is the state department of education, a textbook company, or a standardized test.... How can we expect students to be problem solvers, thinkers, and decision makers when we do not expect the same of their teachers?"⁷

In fact, we can't expect it and we don't get it. Here is a study you may want to track down, photocopy, and mail to every school board member, education reporter, and state legislator in your area. Researchers at the University of Colorado asked a group of fourth-grade teachers to teach a specific task. About half of them were told that when they were finished, their students must "perform up to standards" and do well on a test. The other teachers were simply invited to "facilitate the children's learning." The result: students in the "standards" classrooms did not learn the task as well as those whose teachers felt less pressured.⁸ A carefully controlled scientific experiment, then, confirms real-life experience: even on a traditional test, the demand model that practically defines contemporary education reform turns out to be counterproductive.

Telling teachers exactly what to do and then holding them "accountable" for the results⁹ does not reflect a commitment to excellence. It reflects a commitment to an outmoded, top-down model of control that is reminiscent of Frederick Taylor's "scientific management" method for speeding up factory production.¹⁰ Is it really surprising that this approach tends to backfire? The factory offers one of many examples outside education where people often rely on power to try to compel certain results. Those higher up in a hierarchy attempt to maximize efficiency by tightly controlling the actions of those down below. Unfortunately, we humans just don't respond very well when people do things to us rather than working with us.

That doesn't necessarily doom the whole concept of accountability. The idea can be valid and valuable if we define it as a sense of responsibility to oneself, to one another, and to the community—and if it's nested in a support model.¹¹ But these days teachers are held accountable for the wrong things (specifically, for producing higher standardized test scores) and in the wrong way. The word has come to be a euphemism for more control over what goes on in classrooms by people who aren't in classrooms, and it has approximately the same effect on learning that a noose has on breathing.

When teachers are told exactly what and how to teach, when they feel pressured to produce results, they in turn tend to pressure their students. That is exactly what another study found: teachers who felt controlled became more controlling, removing virtually any opportunities for students to direct their own learning.¹² (The same thing is on display every day in corporations: the middle managers who are most rigidly controlled by top executives tend to do the same to their subordinates.)¹³ Since people don't do their best when they feel controlled, the results of the Colorado study make perfect sense: the more teachers are pressured to make their students perform, the less well they actually perform.

Burnt at the High Stakes

Suppose you were a superintendent or a school board member or a state legislator. Suppose all these arguments and data just rolled off your back, and you remained determined to impose your version of standards on the schools, complete with an exaggerated emphasis on achievement and a traditional model of instruction. The chances are, you would use standardized tests to impose this agenda, and the chances are that this strategy would be pretty darned effective—providing that the people in the schools came to worry a lot about the test results. The question then becomes: How can you make sure they do so?

To begin with, you can order that such tests be given frequently, thus raising their visibility among teachers and students. Then you can see to it that the scores are published in the newspapers, preferably in a chart that encourages readers to see it as a way of comparing the quality of schools or districts. The prospect of looking bad may serve as a kind of "public shaming" and thereby pressure educators to do anything necessary to crank up their scores.

Best of all, though, you can create a structure to "motivate" educators to become more concerned about the tests. Here you would say, in effect, "Oh, so you don't care about these tests, huh? You think they're shallow and based on a simplistic approach to instruction? You'd rather pursue other kinds of learning? Well, we'll force you to care about them!" Students can be punished for low scores by being held back a year—or, in the case of "exit exams," prevented from graduating on the basis of a single test regardless of their academic records.¹⁴ Teachers can be evaluated on the basis of their students' scores and then find their paychecks swollen or shrunk accordingly. Principals can be suspended or fired; their schools may receive bundles of money for high scores or, in extreme cases, be shut down or taken over by the state for low scores.

This collection of tactics, known as "high-stakes" testing, has spread like the plague it is. Unhappily, most of the discussion about it has tended to focus on the particulars of implementation, with people asking whether rewards are better than punishments, which specific strategies work best, or whether threats should be concentrated on teachers or students.¹⁵ Rarely do we hear more substantive criticisms concerning the validity and effects of the tests themselves. (If the tests are norm-referenced, then we really don't have any legitimate basis for rewarding or punishing.) And rarely do we examine high-stakes testing in light of the misconceived view of human psychology that underlies it. Apart from the inherent problems of control, there are specific drawbacks to dangling something desirable (such as money) in front of people or threatening them with something unpleasant.

The use of punishments, even if referred to euphemistically as negative incentives, sanctions, or consequences, creates a climate of fear, and fear generates anger and resentment. It also leads people to switch into damage-control mode and act more cautiously. Human beings simply do not think creatively and reach for excellence when they perceive themselves to be threatened. The philosophy behind high-stakes testing has been deftly satirized by a sign in some offices and classrooms: THE BEATINGS WILL CONTINUE UNTIL MORALE IMPROVES. When teachers are deprived of job security or pay raises in an effort to make them perform better, they usually become demoralized rather than motivated.

What is true of sticks is ultimately true of carrots as well.¹⁶ The simplistic call to reward excellence (or what passes for excellence in the minds of many observers) overlooks an enormous body of research and experience demonstrating that rewards are just a different way of controlling people. Moreover, rewards are based on a fundamental psychological misconception—namely, that there is a single entity called "motivation," something of which one can have a lot or a little. If this premise were true, then rewards and punishments might well make sense because they would presumably make someone's level of motivation go up. Unfortunately, the premise is false: there are different kinds of motivation, and the kind matters more than the amount.

Psychologists typically distinguish between "intrinsic motivation," a fancy term for finding something worth doing in its own right, and "extrinsic motivation," which means that you do one thing so that something else, outside the task, will happen. (Specifically, you'll receive a reward or escape a punishment.) It is quite clear—to everyone except a small group of orthodox behaviorists, at least—that these two kinds of motivation are qualitatively different. It is equally clear to most of us that intrinsic motivation is more desirable than extrinsic, and that no amount of the latter can make up for the absence of the former. Adults who consistently do excellent work, and students whose learning is most impressive, are usually those who love what they do, not those who see what they do as a tedious prerequisite for getting dollars or A's.

But the real revelation is not that extrinsic motivation is different or inferior; it's that it is corrosive: it tends to undermine intrinsic motivation. This is one of the most thoroughly replicated findings in the field of social psychology: the more you reward people for doing something, the more they tend to lose interest in whatever they had to do to get the reward. Thus, the intrinsic motivation that is so vital to quality—to say nothing of quality of life—often evaporates in the face of extrinsic incentives. More specifically, researchers have found that people's interest in a task ordinarily plummets when they are acutely aware of being evaluated on their performance—even if the evaluation is positive. "When subjects expect rewards for attaining a specified level of competence, the anticipation of performance evaluation interferes with intrinsic interest," one group of researchers concluded.¹⁷ And when interest drops, so too does excellence.

Beyond these general hazards of carrot-and-stick psychology, high-stakes testing to hold schools accountable is profoundly unfair for several reasons. First, even assuming the tests themselves were valid measures of meaningful learning, low scores (in absolute and especially in relative terms) are to a large extent reflections of how educated and affluent the student's family is. To this extent, punishing people for those scores is cruel and pointless. Second, to the extent the scores do reflect school experience, that experience is hardly limited to the current year. Thus, it seems hard to justify holding a fourth-grade teacher accountable for her students' test scores when they reflect all that has happened to the children before they even got to her class.

Third, the fairest way of allocating money for education—considered common sense in some other countries—is to provide higher levels of support for the schools and districts that are in the most trouble. The high-stakes approach, to the extent it involves money, usually amounts to precisely the opposite basis for funding: give more to those already successful and less to those who really need it. Finally, because high-stakes strategies lead to teaching to the test, and because "teaching to the test can raise scores more dramatically than can instruction designed to improve achievement, the incentive system could reward the worst practices," as the noted researcher Lorrie Shepard points out.¹⁸ It could also be based on distorted information, since test scores don't reflect achievement per se. (The more teachers end up teaching to the test, the less valid the tests become as measures of what students can do.)

The practical effects of high-stakes testing are no less appalling for being completely predictable. Teaching to the test—and all the ways it damages education—is obviously far more likely to take place in the crudest and most extreme way when the stakes are high. Indeed, when the carrot is appealing enough or the stick is aversive enough, people may do more than adjust the curriculum to assure the desired results. In a desperate effort to raise scores, educators are sometimes reduced to playing games. Some focus all their attention on the students who, they believe, are just shy of getting high scores, thereby slighting everyone else. Some flunk low achievers on the theory that they'll do better on the exam after repeating a grade and in the meantime won't bring down the average of those in the next grade.¹⁹ Some try to assign such students to special education classes so they won't have to take the tests at all. And some cheat in more conventional ways: principals in numerous states have been accused of "excessive coaching" and even of altering students' answer sheets.

Far from being a constructive force, high-stakes testing makes teachers defensive as they try to show that low scores are not their fault. It sets them against each other and skews their priorities. David Berliner and Bruce Biddle recall how one state

based high school mathematics teachers' annual raises on gains in their students' achievement scores. The next year some of the top teachers in the state resigned in disgust. Those who remained entered into intense competition with one another, which disrupted school programs and caused morale to drop throughout the state. (Among other things, some math teachers demanded that their schools restrict extracurricular activities, cancel school assemblies, and abolish out-of-school trips that might interfere with their instructional efforts.) The following year the incentive program was dropped.²⁰

That story, at least, had a moderately happy ending, but often the program remains and the educators don't. High-stakes testing routinely drives good people out of the profession, and it is particularly hard to find qualified educators in those areas who will agree to take a position as principal. (The paradoxical result, once again, is that the Tougher Standards movement leads to lower standards.) One middle school principal in Kentucky says he has watched his colleagues "disappear from the ranks. No one wants to blame it on [high-stakes testing programs], but from my perspective as a practicing principal, many of them made it clear they weren't going to put up with unreasonable demands."²¹ Or perhaps they refused to relinquish control over the curriculum to the faceless individuals who design the tests.

By the time they have a choice, students may react to more tests and threats in exactly the same way. According to a study jointly prepared by U.S. and Irish researchers, "Evidence from a number of countries indicates that the use of competency tests for graduation or retention results in increasing the student drop-out rate."²² Some observers have begun to suspect that this is precisely the intent of these mandatory tests for graduation, which at this writing are used by almost half the states. Ironically, students who drop out are not necessarily the lowest achievers;²³ they may just be the kind of people who don't want any part of a school system that views them as dispensable or uses coercion to heighten their concern about test scores. Or maybe they sense that standardized testing doesn't exist for their benefit.

We come back to the contrast between a demand model and a support model. High-stakes testing must surely be the ultimate example of the former: impose tougher standards and accountability, use standardized tests, and try to manipulate everyone in sight by bribing and threatening them into raising scores. Apart from the disadvantages of this strategy, it has, as Linda Darling-Hammond summarizes the evidence, basically "failed wherever it has been tried."²⁴ Even when you judge by a standardized test (the NAEP), "States without high-stakes tests are showing more improvement than those with them."²⁵ And the evidence from other countries is similarly discouraging.²⁶

But the response to all these cautionary data on the part of the Tougher Standards crowd is true to form: do it some more! Thus, when a new superintendent was installed in a Maryland school district several years ago—a man who said he didn't see how education is "any different from any other major business"²⁷—his pledge was not to help students learn more effectively. It was to raise test scores. As a reporter describes it, "He carried out this pledge by removing student artwork that had graced the walls of the conference room adjacent to his office and replacing it with charts that showed the test performance of every school in the district, along with the principal's name."²⁸ Here is as fitting, and as poignant, an image as any for the prevailing version of school reform.