School reform was a politically popular issue. In 1988, President George H. W. Bush said he wanted to be the “Education President,” and many governors in the 1980s and 1990s claimed the mantle of “Education Governor.” Among them were James Hunt in North Carolina, Bill Clinton in Arkansas, Lamar Alexander in Tennessee, Richard Riley in South Carolina, George W. Bush in Texas, Booth Gardner in Washington State, and Roy Romer in Colorado. Some governors made their mark as education reformers by expanding funding for pre-kindergarten or raising teachers’ salaries (or both), but most of the time their reforms consisted of new requirements for testing and accountability.
In Washington, the calls for testing and accountability grew louder during the 1990s. The first President Bush released his America 2000 program in 1991, recommending voluntary national standards and voluntary national testing, but it was never authorized by the Democratic majority in Congress. President Clinton came to office promising to create a system of national standards and national tests, but he was stymied by a Republican majority in Congress. Clinton’s Goals 2000 program, enacted before Republicans took control of Congress in the fall of 1994, encouraged states to develop their own standards and tests. It became a ritual for Republicans and Democrats alike to bemoan the lack of accountability in American public education and to grouse that no teacher, principal, or student was held accountable for poor test scores.
In that light, the large test score gains in Texas over the previous half-dozen years must have impressed Congress. Not only were increasing numbers of students passing the state tests, according to the Texas state education department, but the achievement gap between white students and minority students was steadily shrinking, as was the number of students dropping out before high school graduation. So when President George W. Bush arrived in Washington with a plan based on what appeared to be a successful model of accountability in Texas, members of both parties were willing and ready to sign on, so long as they could add one or two or three of their own priorities to the bill. Almost everyone wanted an accountability plan, and almost everyone embraced the main elements of the one President Bush set forth.
A few scholars had warned in 2000 that the gains in Texas were a mirage; they said the testing system actually caused rising numbers of dropouts, especially among African American and Hispanic students, many of whom were held back repeatedly and quit school in discouragement. These scholars insisted that the state’s rising test scores and graduation rates were a direct result of the soaring dropout rate: As low-performing students gave up on education, the statistics got better and better. In separate studies, Walt Haney of Boston College and Stephen Klein of RAND maintained that the dramatic gains in Texas on its state tests were not reflected on other measures of academic performance, such as the SAT and NAEP, or even the state’s own test for college readiness. Haney argued that the Texas high-stakes testing system had other negative effects. As teachers spent more time preparing students to take standardized tests, the curriculum was narrowed: Such subjects as science, social studies, and the arts were pushed aside to make time for test preparation. Consequently, students in Texas were actually getting a worse education tied solely to taking the state tests.3
In its eagerness to endorse education reform, Congress paid no attention to these red flags and passed a program that was closely aligned with the Texas model.
NCLB was complex and contained many programs. Its accountability plan included these features:1. All states were expected to choose their own tests, adopt three performance levels (such as basic, proficient, and advanced), and decide for themselves how to define “proficiency.”
2. All public schools receiving federal funding were required to test all students in grades three through eight annually and once in high school in reading and mathematics and to disaggregate (i.e., separate) their scores by race, ethnicity, low-income status, disability status, and limited English proficiency. Disaggregation of scores would ensure that every group’s progress was monitored, not hidden in an overall average.
3. All states were required to establish timelines showing how 100 percent of their students would reach proficiency in reading and mathematics by 2013-2014.
4. All schools and school districts were expected to make “adequate yearly progress” (AYP) for every subgroup toward the goal of 100 percent proficiency by 2013-2014.
5. Any school that did not make adequate progress for every subgroup toward the goal of 100 percent proficiency would be labeled a school in need of improvement (SINI). It would face a series of increasingly onerous sanctions. In the first year of failing to make AYP, the school would be put on notice. In the second year, it would be required to offer all its students the right to transfer to a successful school, with transportation paid from the district’s allotment of federal funds. In the third year, the school would be required to offer free tutoring to low-income students, paid from the district’s federal funds. In the fourth year, the school would be required to undertake “corrective action,” which might mean curriculum changes, staff changes, or a longer school day or year. If a school missed its targets for any subgroup for five consecutive years, it would be required to “restructure.”
6. Schools that were required to restructure had five options: convert to a charter school; replace the principal and staff; relinquish control to private management; turn over control of the school to the state; or “any other major restructuring of the school’s governance.” (Most states and districts ended up choosing the last, most ambiguous alternative, hoping to avoid the other prospects.)
7. NCLB required all states to participate in the federal National Assessment of Educational Progress (NAEP), which would henceforth test reading and mathematics in grades four and eight in every state every other year (before NCLB, state participation in NAEP was voluntary, and some states did not participate; also, the NAEP reading and math tests were not administered every other year). The NAEP scores, which had no consequences for any student or school or district, served as an external audit to monitor the progress of the states in meeting their goals.4
Although the law contained a host of other programs and priorities (most notably, a requirement that all children be taught by a “highly qualified teacher”), the central focus of NCLB was accountability. It was the issue that brought together Republicans and Democrats. Had there not been bipartisan agreement on accountability, NCLB would never have become law. Both parties believed that accountability was the lever that would raise achievement.
Over the next few years, most complaints about NCLB focused on funding. Some states complained that the federal government did not give them enough extra money to do what the law required. Federal funding for elementary and secondary programs was increased by nearly 60 percent in the early years of NCLB, but Democrats complained that it was way below what was needed and what Congress had authorized. I remember thinking at the time that the law only required the schools to teach reading and math effectively, so why should that be a huge additional cost? Isn’t that what they should have been doing anyway? Could there be a more fundamental responsibility of schools than to teach everyone basic skills? Then there were periodic outbursts by parents and activists against excessive testing and even some organized protests against mandatory state tests. I was not sympathetic to the anti-testing movement. I didn’t see why anyone would object to an annual test of reading and mathematics.
My support for NCLB remained strong until November 30, 2006. I can pinpoint the date exactly because that was the day I realized that NCLB was a failure. I went to a conference at the American Enterprise Institute in Washington, D.C.—a well-respected conservative think tank—to hear a dozen or so scholars present their analyses of NCLB’s remedies. Organized by Frederick M. Hess and Chester E. Finn Jr., the conference examined whether the major remedies prescribed by NCLB—especially choice and after-school tutoring—were effective. Was the “NCLB toolkit” working? Were the various sanctions prescribed by the law improving achieve
ment? The various presentations that day demonstrated that state education departments were drowning in new bureaucratic requirements, procedures, and routines, and that none of the prescribed remedies was making a difference.
Choice was not working, they all agreed. The scholars presented persuasive evidence that only a tiny percentage of eligible students asked to transfer to better schools. In California, less than 1 percent of eligible students in “failing” schools sought to transfer to another school; in Colorado, less than 2 percent did; in Michigan, the number of transfers under NCLB was negligible; in Miami, where public school choice was already commonplace, less than ½ of 1 percent asked to move because of NCLB; in New Jersey, almost no eligible students transferred, because most districts had only one school at each grade level, and the state’s urban districts did not have enough seats available in successful schools to accommodate students from “failing” schools. Julian Betts of the University of California at San Diego questioned whether choice was even a successful strategy, because his own studies found that choice had little or no effect on student achievement.5
The scholars suggested many reasons why students were not transferring out of allegedly failing schools. In the first year or two, the letters informing parents of their right to switch their children to a better school were unclear or arrived too late, after the school year had already started. Even when the letters were clear and arrived on time, some parents did not want to send their children on a bus to a faraway school. In some districts, there were already so many public school choice programs that NCLB added nothing new. In others, there were far more eligible students than seats.
But what was especially striking was that many parents and students did not want to leave their neighborhood school, even if the federal government offered them free transportation and the promise of a better school. The parents of English-language learners tended to prefer their neighborhood school, which was familiar to them, even if the federal government said it was failing. A school superintendent told Betts that choice was not popular in his county, because “most people want their local school to be successful, and because they don’t find it convenient to get their children across town.”6 Some excellent schools failed to meet AYP because only one subgroup—usually children with disabilities—did not make adequate progress. In such schools, the children in every other subgroup did make progress, were very happy with the school, did not consider it a failing school, and saw no reason to leave.
Thus, while advocates of choice were certain that most families wanted only the chance to escape their neighborhood school, the first four years of NCLB demonstrated the opposite. When offered a chance to leave their failing school and to attend a supposedly better school in another part of town, less than 5 percent—and in some cases, less than 1 percent—of students actually sought to transfer.
Free after-school tutoring (called Supplementary Educational Services, or SES) fared only a bit better than choice, according to the papers presented that day. In California, 7 percent of eligible students received tutoring; in New Jersey, 20 percent did; in Colorado, 10 percent; and in Kentucky, 9 percent. The law implicitly created a “voucher” program for tutoring companies, a marketplace where tutoring companies and school districts could compete for students. Any organization could step forward to register with state departments of education to provide tutoring, whether they were a public school, a school district, a community group, a mom-and-pop operation, a faith-based agency, a for-profit corporation, a college, or a social services organization. Across the nation, nearly 2,000 providers registered to offer tutoring to needy students. But no more than 20 percent of eligible students in any state actually received it, even though it was free and readily available.
Why so little interest in free tutoring? The tutoring agencies blamed the districts for not giving them space in the public schools, and the public schools blamed the tutoring agencies for demanding space that was needed for extracurricular activities. The tutors complained about the cost of liability insurance, and the districts complained that some of the tutoring companies were ineffective or were offering students gifts or money if they signed up for their classes. It also seemed likely that large numbers of low-performing students did not want a longer school day, even though they needed the extra help.
As I listened to the day’s discussion, it became clear to me that NCLB’s remedies were not working. Students were offered the choice to go to another school, and they weren’t accepting the offer. They were offered free tutoring, and 80 percent or more turned it down. Enough students signed up to create handsome profits for tutoring companies, but the quality of their services was seldom monitored. I recalled a scandal in New York City when investigators discovered that a tutoring company, created specifically to take advantage of NCLB largesse, was recruiting students by giving money to their principals and gifts to the children; several of the firm’s employees had criminal records.7
Adult interests were well served by NCLB. The law generated huge revenues for tutoring and testing services, which became a sizable industry. Companies that offered tutoring, tests, and test-prep materials were raking in billions of dollars annually from federal, state, and local governments, but the advantages to the nation’s students were not obvious.8
At the conference, I was on a panel charged with summing up the lessons of the day. I proposed that the states and the federal government were trying to assume tasks for which they were ill suited. I suggested that they should flip their roles, so that the federal government was gathering and disseminating reliable information on progress, and the states were designing and implementing improvements. Under NCLB, the federal government was dictating ineffectual remedies, which had no track record of success. Neither Congress nor the U.S. Department of Education knows how to fix low-performing schools. Meanwhile, the law required the states to set their own standards and grade their own progress; this led to vastly inflated claims of progress and confusion about standards, with fifty standards for fifty states. Every state was able to define proficiency as it saw fit, which allowed states to claim gains even when there were none. The proper role of the federal government is to supply valid information and leave the remedies and sanctions to those who are closest to the unique problems of individual schools.
What I learned that day fundamentally changed my view of No Child Left Behind. When I realized that the remedies were not working, I started to doubt the entire approach to school reform that NCLB represented. I realized that incentives and sanctions were not the right levers to improve education; incentives and sanctions may be right for business organizations, where the bottom line—profit—is the highest priority, but they are not right for schools. I started to see the danger of the culture of testing that was spreading through every school in every community, town, city, and state. I began to question ideas that I once embraced, such as choice and accountability, that were central to NCLB. As time went by, my doubts multiplied. I came to realize that the sanctions embedded in NCLB were, in fact, not only ineffective but certain to contribute to the privatization of large chunks of public education. I wonder whether the members of Congress intended this outcome. I doubt that they did. In a bill whose length exceeds 1,000 pages, it is unlikely that many members of Congress read it thoroughly and fully understood all the eventual consequences.
The most toxic flaw in NCLB was its legislative command that all students in every school must be proficient in reading and mathematics by 2014. By that magical date, every single student must achieve proficiency, including students with special needs, students whose native language is not English, students who are homeless and lacking in any societal advantage, and students who have every societal advantage but are not interested in their schoolwork. All will be proficient by 2014, or so the law mandates. And if they are not, then their schools and teachers will suffer the consequences.
The term “proficiency”—which is the goal of the law—is not the same as “minimal literacy
.” The term “proficiency” has been used since the early 1990s by the federal testing program, the National Assessment of Educational Progress, where it connotes a very high level of academic achievement. The federal assessment refers to four levels of achievement. The lowest is “below basic,” which means a student who is unable to meet the standards for his or her grade. The next level is “basic,” which means that a student has partially mastered the expectations for the grade. Then comes “proficient,” indicating that a student has fully mastered the standards for the grade. And at the very top of the performance levels is “advanced,” which represents truly superior achievement. On the 2007 NAEP for fourth-grade reading, 33 percent of the nation’s students were below basic; 34 percent were basic; 25 percent scored proficient; and 8 percent were advanced. In that same year, 28 percent of students in eighth grade were reading at the proficient level, and an additional 3 percent were advanced.9 Now, in a nation where only one-third of students meet the federal standard for proficiency, we are expected to believe that fully 100 percent will meet that standard by 2014. It will not happen. Unless, that is, the term “proficiency” is redefined to mean functional literacy, minimal literacy, or something akin to a low passing mark (say, a 60 on a test with a 100-point scale, a score that once would have merited a D, at best).
The Death and Life of the Great American School System Page 13