The Allure of Order: High Hopes, Dashed Expectations, and the Troubled Quest to Remake American Schooling, by Jal Mehta, Oxford University Press, 416 pages, $29.95.

Last year there was a major change at the private Waldorf elementary school my son attends. His second-grade class had only about 10 kids, and the third-grade class above him had a similarly small number. The teachers were concerned that so few kids in the class would hurt learning; in a tiny class, they felt, kids would have fewer opportunities to form friendships and fewer opportunities to learn from and teach each other. So the teachers consulted with the board, and they eventually decided to combine the two classes into a 15- to 20-student third/fourth-grade class. The change seems to have worked out well.

In some sense, the way this decision was made seems natural. Teachers know what's going on in the classroom; teachers, therefore, should be the main people making major decisions about their classes. And yet in most public schools, this local knowledge is ignored. Teachers are rarely consulted about administrative changes. And this is not an accident. Subjecting teachers to outside control has been the major goal of school reform policies for the past century, as Jal Mehta, an assistant professor of education at Harvard, explains in The Allure of Order.

Most people think of the current wave of school reform—the push for centrally defined standards, testing, and "accountability"—as something that began in the 1990s. Mehta suggests that the blueprint was set in the Progressive Era. It was tried again in the 1960s and '70s before finally culminating in George W. Bush's No Child Left Behind Act and our current rage for standardized tests.

All of these movements, Mehta explains, were based on the idea of rationalizing schools. The Progressives worked to systematize and control (mostly female) teachers under the control of (mostly male) superintendents and administrators. The reformers of the '60s and '70s worked, with sporadic success, to establish standards and increase state control over district schooling. And Bill Clinton, George W. Bush, and Barack Obama have overseen a nationalization of education policy, demanding testing and "accountability" nationwide. The effort to control schools from afar through a stern bureaucracy, and to impose excellence by fiat—this is not a new solution. It's the same old problem.

For Mehta, the issue is not so much teachers themselves as the teaching profession. That profession—in part because it was early on dominated by women, and so was considered low-status—never fully professionalized. Unlike medicine or law or higher education, teaching never staked out a rigorous body of knowledge which practitioners had to master; it never established self-imposed criteria for proficiency; it never created self-administered barriers to entrance.

Because teachers were never systematically able to separate themselves from laborers, Mehta argues, they have been unable to defend themselves from outside administrators who seek to impose a factory discipline upon them. They are not viewed as professionals striving for excellence, but as recalcitrant, incompetent workers who need managing if they are going to work at all. In reaction, teachers, understandably, did what workers have often done when they feel alienated from management, and unionized. But while this gave them a better bargaining position in terms of benefits and wages, it compounded the problem; in acceding to the definition of themselves as workers, teachers ceded the professional claim to moral authority and expertise. Rather than participating in shaping the educational direction of their schools (as they do at my son's Waldorf school), they instead hunkered down into the traditional antipathy between management and labor.

One result, Mehta argues, is that teaching has little cultural ground on which to defend itself. Teachers unions can protects instructors' jobs, but they have had little ability to combat the testing regime. Thus educators have gotten swept along (with mostly ineffective protest) by the vacillating administrative dictats of the moment. The Chicago school system decides to create small schools; five years later they change their mind and get rid of the small schools. They decide they need to build new schools; then they decide that 50+ schools need to be closed. Or they decide that Persepolis should be banned, and while teachers can criticize, they can't really launch a counterattack. After all, reading Persepolis isn't going to help you on tests. And what else are schools there for, if not for the tests? The idea that education has a value in itself, or that schools are first and foremost communities of learning, has little cachet and little power. Instead, first the administrative bureaucracy and then teachers themselves have locked themselves into a vision in which educators are little more than balky automatons hammering out widgets in accord with an endless succession of government-imposed Five-Year Plans.

Children aren't in fact widgets, so it's not much surprise that treating them as such has not worked especially well. But when you fail to turn children into widgets, the answer offered is more reforms aimed at making better widgets. And then that fails, and there are more reforms, and on and on, scantron without end.

Mehta suggests suggests several alternatives. Pay teachers a lot more. (Mehta points out that in the high-achieving South Korean school system, teachers are paid more than lawyers or engineers.) Shift education research away from theory (traditionally considered high-status) and towards practical questions of pedagogy (traditionally disdained). Use tests as guides to improve teacher practice rather than as cudgels to punish. Give local schools control over hiring. Make tenure (which Mehta sees as a potentially useful tool of professionalization) more dependent on performance—and have teachers themselves be the ones who assess their colleagues for this purpose. Stop micromanaging teachers in the classroom, and instead work to hire good people in the first place. In sum, Mehta believes that if you take steps to start treating teachers like professionals, more and higher skilled people will want to enter the profession, so that people start treating teachers like professionals, producing a spiral of increasing quality and improvement.

In addition to finding models in Finland, South Korea, and other foreign countries, Mehta thinks we can look to the more successful stateside charter schools—those which treat teachers as knowledgeable contributors to a culture of learning. You can look, for that matter, to Waldorf schools, where the faculty body is in large part responsible for administration and school policy. The key to good schools is not to punish untrustworthy workers until they somehow, despite themselves, do what some distant bureaucrat decides is the right thing. The key is to hire good people who care about learning and students, then give them the support and resources they need to do their jobs. Support and resources here means more money, especially for schools in poor or segregated neighborhoods. (For Mehta, it doesn't matter whether that money comes from direct subsidies, from vouchers, or from some other means.)

Mehta is pessimistic that anything will change soon. Government bureaucrats don't let go of money, or control, willingly. Neither do unions. Nor, for that matter, will university researchers happily start turning their attention to lower-status practical research. Teachers are likely to embrace the idea of higher pay, but not so much the idea of clearing out the deadwood and getting better people in the profession. Charter school reforms can be a means to decentralize the system and create communities of learning—but they can also be an excuse for even more obsessive micromanagement, and even lower pay. In practical terms, teachers being treated contemptuously by local corporate school administrators isn't much different from teachers being treated contemptuously by distant federal bureaucrats.

The libertarian argument, of course, is that teachers should in effect be managed by parents. If school funding is distributed through vouchers, parents can choose the schools that serve them best, those schools will thrive, and the education system will improve. Mehta doesn't necessarily share libertarians' desire to empower parents and distribute funds through market mechanisms rather than political grants. But he doesn't have a doctrinaire objection to that scenario either, and he seems to believe that, in many ways, charter schools can push back against the century-long wave of centralization.

Yet Mehta makes a good case that vouchers and charters can't work without an effort to improve teachers' status. As long as teachers do not have control over their work, someone else is going to have control over it. And if control comes from outside, teachers are going to feel like manipulated widgets. It is difficult to hire adventurous, engaged, talented widgets. Perhaps some charters will manage to create learning communities that respect teachers and their local knowledge. But as long as teacher status in general is low, another wave of reform will come along and another wave of accountability and centralization, and those learning communities will be swept away in the interest of making it look like some politician is keeping busy.

Not that professionalization is always or everywhere a magic solution to everything. The doctors' guild, for example, has marginalized midwives, with disastrous results. Mehta doesn't discuss these potential pitfalls of legally enforced monopoly power, but he acknowledges that there has been increasing public accountability for doctors and lawyers in recent years, and that this is a good thing. Complete professional autonomy is dangerous—but so is obsessive micromanagement by distant politicians or nearby bureaucrats. If we don't want our kids taught by slavish, debased drones, then we need to stop treating teachers like slavish, debased drones.