Until the mid 1800s, education in the United States was local. Virtually all decisions were made at the school level, whether the schools operated independently or were run by local government.
A chief argument for the creation of our modern state-level school systems was that the centralization of decision-making in the hands of experts would usher in a new era of more effective teaching methods and materials, and thus lead to significant improvement in student achievement. These same arguments are heard today, though it is now widely suggested that standards and testing should be directed by the federal government. In a May 2001 survey, 78 percent of Americans supported nationally standardized tests.
Is our support for nationalizing curriculum and testing justified? Or is it another example of something that seems like it ought to work but really doesn't? There are a couple of ways to answer that question. Most obviously, we can look at how well the move to state-level curricula and testing have played out.
One of the most decisive examples was California's adoption of a new statewide reading instruction curriculum in 1987. The adoption process was very much what one might expect, with curriculum designers from all over the country being solicited to submit their proposals, and a state committee of education experts selecting the winner from among them. It was the same sort of process by which many other curriculum- and textbook-selection decisions are made around the country.
Over the next several years, reading scores on the California Assessment Program declined steadily. Seven years into the new reading program, California's fourth-grade reading scores on the National Assessment of Educational Progress (NAEP) had fallen to dead last in the nation.
Though some suggested that this abysmal showing might be due to the state's high immigrant population or other demographic factors, the evidence indicated otherwise. In fourth grade reading, California's white students – when considered separately from their minority counterparts – scored worst in the nation in reading. The state's Hispanic students also scored worst, and its black students were second worst in the nation.
What had gone wrong? As it happened, the education experts on the state curriculum committee had selected a program that eliminated structured, synthetic phonics, the practice of systematically teaching children to read words by sounding out their constituent parts and then blending those sounds together. Evidence showing the great importance of phonics in early reading instruction was already well established before the committee members made their decision.
So why did the committee's experts reject all the instructional programs that adhered to the consensus of reading research? Wishful thinking. Structured, teacher-directed methods of instruction run counter to the philosophy that has held sway in teacher training programs for more than 70 years. Education philosophers and theorists, those who have guided the curriculum in colleges of education for a century, espouse a naturalistic teaching philosophy that states all learning is natural, and that structured, teacher-directed lessons are stifling and harmful. So, when it came time for a group of educators to choose a method of reading instruction for California's millions of children, they eschewed structured, empirically proven phonics lessons, opting instead for the unproven but philosophically appealing "whole language" approach.
To blame California's rejection of effective reading instruction methods on this committee would be a terrible mistake. These were well-meaning individuals doing what they thought, and what they had been taught, was right. There is no reason to think that an entirely different group of people would not have made the same choice, in spite of the evidence in favor of phonics.
The problem was not with the people involved, but with the system in which they operated. If there had been a mechanism in place that would have encouraged committee members to heed the most reliable research, and that would have strongly tied their own professional futures to student outcomes, then and only then could this committee, and the next, and the next, be expected to consistently make sound decisions.
The case of California is not unique. It is just one episode in a long history of arbitrary curriculum and textbook selection decisions made by well-meaning expert educators. Occasionally their decisions have been good ones; more often they have been dubious or even disastrous.
Dismayed by the lack of evidence mustered by supporters of centralized curriculum guidelines, Columbia University researcher Richard M. Wolf decided to compare the results of the nations participating the Third International Mathematics and Science Study to determine whether or not having national achievement standards correlated with higher achievement. They didn't. Though most of the participating countries did "have a national curriculum or syllabus," Wolf wrote, there was "virtually no relationship between student performance and having a national curriculum or syllabus."
The 150-year-old prediction that centralizing power over the curriculum in the hands of experts would ensure sound pedagogical decisions has been proved false by hard, sad experience. Centralization of authority doesn't provide an incentive structure that consistently forces educators to heed research evidence and to focus on outcomes rather than on idealized notions of how children should learn.
The record of state testing programs is equally troubled. A great body of evidence has been amassed over the past decade pointing to widespread corruption in these programs. High-stakes test results are referred to in the scholarly literature as "polluted" and "contaminated" by fraud and are considered to be virtually useless as measures of actual student achievement. The problem ranges from occasional outright cheating by teachers and principals to inflate student scores, to the more common practices of "teaching to the test" and of preventing potentially low-scoring students from taking the tests.
These problems occur all over the country, from Michigan to California, from suburban Connecticut to central Chicago, and everywhere in between. By far, the most telling evidence of cheating comes from public-school educators themselves. A 1992 survey asked 2,256 teachers, principals, testing coordinators and superintendents from around the country whether their colleagues engaged in blatant cheating. Forty-four percent said yes. Fifty-five percent of the teachers surveyed were aware of flagrantly unethical testing practices such as changing student's answers, teaching specific test items in advance of the test, and giving hints during the test. The higher the stakes associated with a particular test, the greater the incentive to massage its results.
Apart from the fact that high-stakes tests have become unreliable measures of student achievement, there are other reasons to be apprehensive about them. Unbeknownst to most education reformers, high-stakes national testing programs are not a recent idea. More than a century ago, England put in place a system whereby schools were paid based on the number of students passing a set of government tests. Scientist T. H. Huxley observed at the time that this "Payment by Results" program "did not compel any schoolmaster to leave off teaching anything; but, by the very simple process of refusing to pay for many kinds of teaching, it has practically put an end to them."
Even when money is not directly at stake, evaluating schools solely on the basis of high-profile tests can have the same deleterious effects. As already noted, U.S. public schools already alter their curricula to fit the material they know to be on state- or district-level tests. To the extent that a subject is not part of a mandatory state curriculum or testing program, it is likely to be marginalized. This trend is also evident in England under its current National Curriculum. According to the National Association of Head Teachers, the "obsession with passing tests in English, mathematics and science [means] other subjects [are] being overlooked."
Most of us agree on the importance of a thorough, well-designed curriculum. And the public is solidly in favor of academic testing as one way to find out how much children are learning and how well schools are performing. But the benefits to be derived from mandatory imposition of a particular curriculum or set of tests by state and federal governments are not supported by the evidence.