Secretary of Education Arne Duncan has called for a speedy re-authorization of the Elementary and Secondary Education Act (ESEA), branded the “No Child Left Behind” law (NCLB) by the Bush administration. After expiring in 2007, NCLB has hung on with annual temporary extensions; its provisions are now so controversial that no Congressional majority has been able to coalesce around a proposal for modification. But “re-authorization can’t wait,” Duncan said recently, and he has pledged to get Congress to enact a new education law in 2010.Many critics of NCLB have been dismayed by Duncan’s apparent resolve. But they may be reacting too soon. The principles for re-authorization that Duncan has announced represent a radical departure from NCLB and open uncharted territory in education policy. Fleshing out these principles will require extraordinary creativity and complexity as well as require the Department of Education and Congress to usefully explore accountability tools for which there is little precedent, either at the state or national levels. If Duncan sticks to these principles, the worst of NCLB will be behind us, although designing a new federal education policy will take us well beyond 2010.

That’s not a bad thing. After rushing into an ill-considered NCLB whose perverse consequences far outweighed its benefits, it would be wise now to move more carefully in the design of new federal education policy, rather than rushing in again.

NCLB requires all schools, by 2014, to get all students, in all racial, ethnic, and income subgroups, to demonstrate proficiency by passing “challenging” tests in reading and math. Schools that fail to make progress toward this goal at first are required to permit students to transfer elsewhere. If the goal remains out of reach after five years, the school can be closed, or its staff replaced. NCLB, its promoters insisted, permit “no excuses” for children not meeting the goal of proficiency. The underlying concept is that “all children can learn” if schools are held sufficiently accountable for successful instruction: children from impoverished backgrounds may have a tougher time, but if teachers have high expectations, children will respond by overcoming their obstacles.

These sentiments are understandable, but there are perverse consequences of wrongly believing that every student, without exception, will reach proficiency, at a minimal standard, let alone a “challenging” one. In a September speech outlining his principles for ESEA re-authorization, Duncan called this idea “utopian” and observed that in a fruitless effort to get increasing numbers of students to pass, states have made tests easier by lowering the passing scores and eliminating all but the least-challenging subject matter. Schools have thrown out a balanced curriculum—eliminating the arts, science, history, social studies, and foreign language—to spend ever more time in boring drills of basic math and reading, along with incessant practice in test-taking. Physical education and even recess have been cut out to free more seat-time for basic skills “remediation.” Inordinate attention has been devoted to students whose skills are just below the passing point, while students with either very poor or adequate skills have been ignored, because moving just a few students from failing to passing qualifies as progress toward the goal. The transfer provisions have been used mostly by the better students in “failing” schools, doing nothing to help students most in need, and making it all the more impossible for their schools to progress toward getting all students to proficiency.

Duncan has denounced each of these corruptions of American education and has vowed that a re-authorized ESEA won’t replicate them. He has criticized NCLB’s requirement that, rather than improve schools where test scores are inadequate, students are instead told to transfer out. In his September speech calling for re-authorization, Duncan charged that the NCLB system is “not education” but “game-playing tied to bad tests with the wrong goals.”

Duncan has vowed that a new federal law would not continue to create incentives for states to adopt easy tests of basic skills: “The biggest problem with NCLB,” he said, “is that it doesn’t encourage high learning standards. In fact, it inadvertently encourages states to lower them.” To remedy this, Duncan announced that he will use funds within his discretion under the stimulus bill to subsidize states’ development of tougher tests.

In his speech, Duncan repudiated the insistence that all students achieve the same pre-determined passing point. Instead, he called for

“…a law that encourages educators to work with children at every level, the gifted and the struggling—and not just the tiny percent near the middle who can be lifted over [a] mediocre bar of proficiency with minimal effort.

The Secretary vowed that a new law would not result in narrowing the curriculum by holding schools accountable only for math and reading. “Let us build a law,” he said

“…that discourages a narrowing of curriculum and promotes a well-rounded education that draws children into sciences and history, languages and the arts in order to build a society distinguished by both intellectual and economic prowess.“

And in an article he authored in the October issue of the American School Board Journal, Duncan also rejected the notion, implicit in NCLB, that schools alone can be held responsible for student success. He argued that mayors, not school boards, should take control of urban education because mayors can mobilize and coordinate the many public services in addition to schools that also affect academic achievement. “Mayors can facilitate the cradle-to-career health and social service networks that support student learning,” Duncan writes, and adds:

“It takes more than a school to educate a student. It takes a city that can provide support from the parks department, health services, law enforcement, social services, after-school programs, nonprofits, businesses, and churches.”

How can such principles—a radical departure from the old NCLB approach—be incorporated into a new ESEA? If, for example, Congress comes to believe, as Duncan does, that “it takes more than a school to educate a student,” then federal law cannot hold schools that are fortunate enough to benefit from health and social services, or after-school programs, to the same outcome goals as schools that are not so fortunate. The impact on student achievement of these services, or lack of them, will become ever more serious as the long-term effects of our current 9.8% unemployment rate work their way through the next generation of disadvantaged youth. As John Irons notes, “Unemployment and income losses can reduce educational achievement by threatening early childhood nutrition; reducing families’ abilities to provide a supportive learning environment (including adequate health care, summer activities, and stable housing); and by forcing a delay or abandonment of college plans.”

If, as Arne Duncan wrote, it really does take more than a school to educate a student, then a scho
ol struggling on its own, without the support of surrounding services, may have poorer outcomes than a school that benefits from these services. Yet the school with the poorer outcomes might be more successful and be a better school than one that faces lesser social and economic challenges, or that has more community support for its mission. In other words, a school with worse outcomes but more disadvantaged students might deliver more “value-added” than a school with better outcomes and more affluent students. Many schools now glibly categorized as “failing” might not be so inadequate after all. How can a re-authorized ESEA develop transparent federal rules and regulations that fairly make such difficult distinctions?

Duncan wants a new ESEA to avoid creating incentives for schools to narrow their curricula, but such narrowing necessarily results from presently available accountability methods. NCLB demonstrated that if schools are held accountable only for some skills, they will respond rationally and ignore those for which they are not held to account. Incentives work. But while it is relatively easy to design standardized tests, even higher quality tests, of math and reading, American education has yet developed no reliable ways to measure whether students are developing “intellectual prowess” in “sciences and history, languages and the arts.” So long as an accountability system is built around math and reading tests, narrowing of the curriculum will be an inevitable result.

Only a few days after Secretary Duncan’s speech outlining his principles for re-authorization, the Government Accountability Office (GAO) reported on how states’ standardized test development has responded to NCLB. It confirmed that some states lowered the skill levels of state tests in math and reading in order meet NCLB requirements, and suggested that this may be an inevitable consequence of NCLB’s demand that test results be released before the next school year begins. Such time pressure tends to preclude the use of test items that ask for student writing, because only multiple-choice questions can be machine-scored quickly. Some states, therefore, have developed two sets of standards—one for instruction and another for NCLB tests—making the test results invalid as a measure of school quality. This will be hard to avoid in any new ESEA that also insists on quick scoring of standardized tests for purposes of accountability.

The task of re-authorization will be even tougher because some other policies Duncan has adopted actually undermine the principles he’s enunciated for ESEA. For example, while the Secretary has denounced the curriculum narrowing that results from holding schools accountable only for math and reading scores, his commitment to subsidize the development of better tests is tied to efforts to develop national standards in math and reading alone. Likewise, Duncan has promised to subsidize states to develop data systems for the evaluation of teachers by the gains their students make on standardized tests. Yet because the only standardized tests that can be used for this purpose are in math and reading, teachers being evaluated in this way will have even more incentives to stress only basic skills, and downplay other aspects of a well-rounded curriculum. (That the statistical technology does not exist to accurately link a teacher’s quality to her students’ test scores is yet another problem.)

Beyond the unresolved substantive issues hindering a serious proposal for re-authorizing NCLB-like provisions in ESEA, there are daunting political problems. NCLB was virtually the only major bi-partisan domestic legislation of George W. Bush’s tenure. Indeed, it had more support from Democrats, led by the late Senator Edward Kennedy and Congressman George Miller, than from Republicans. But the conversion of America’s schools into testing factories has now made NCLB so unpopular that the law’s name, again in Duncan’s words, has become “toxic.” NCLB-type accountability seems to be fairly popular in only one Congressional district—Washington, D.C., where there is no voting representative—but unpopular elsewhere. A recent Gallup poll found that 48% of Americans had an unfavorable view of NCLB, compared to 28% favorable. Among Democrats, it was even more uneven: 55% unfavorable to 22% favorable.

Many Republican legislators who supported NCLB when it was first adopted in 2001 can no longer be counted upon for support. Now that Democrats are in control of the legislative and executive branches, these Republicans have rediscovered their faith in the local control of education and want nothing more to do with federal testing requirements. Democrats first elected to the House or Senate in 2006 or 2008 mostly campaigned against NCLB. Barack Obama himself denounced NCLB during the 2008 campaign, charging that it had caused schools to “teach to the test” and to abandon a well-rounded curriculum that includes physical education, art, and music. The law, Obama charged, was designed to punish schools, not to improve them.

This all leaves a very small base of senior Democrats who are committed to making only minor modifications in the law they sponsored eight years ago.

It is not, however, impossible to imagine an education accountability system that creates incentives for a well-rounded education, that encourages raising achievement of students of various abilities, that requires higher standards and not just basic skills, and that avoids blaming schools for failures that stem from social and economic conditions beyond their control. Over the last two years, a project convened by the Economic Policy Institute has assembled a diverse and distinguished group of national experts from a variety of policy fields and political perspectives, to outline such a system. Called the Broader, Bolder Approach to Education (BBA) campaign, it has issued a set of principles that meet the standards for a new ESEA set forth in Duncan’s speech. It is a very different model from that of NCLB.

BBA’s proposal for ESEA re-authorization recognizes that no techniques presently exist for evaluating the quality of schools in fair manner. So it calls for the federal government to back away from trying to micro-manage school quality, and instead require states to develop new and experimental systems of qualitative evaluation to complement standardized testing. Such systems would rely on the judgment of experts who visit schools to make on-site determinations of whether a balanced curriculum aligned with high standards is being delivered, whether quality instruction is being undermined by test preparation, and whether efforts are being made to mobilize and coordinate surrounding health, social service, after-school, and early childhood programs that make it more likely that children come to school ready to learn. The purpose of such evaluation would not be to punish schools, but to identify areas where improvement is required, and to direct additional resources and professional support where they are needed.

BBA’s proposal for a qualitative evaluation system would be much more expensive than the machine-scored standardized tests we currently use in a corrupt accountability system. But because it will take many years of trial and reformulation before states are ready to implement a qualitative evaluation system, the initial costs will be modest. Ultimately, we estimat
e a fully-scaled up system of qualitative evaluation will cost only about $16 per pupil, representing less than one-sixth of one percent (0.15%) of total United States school spending. This is not a lot to pay for an accountability system.

Arne Duncan recognizes that it will take time to develop better ways to evaluate student success. In his September speech calling for ESEA re-authorization, he said: “Until states develop better assessments—which we will support and fund through [stimulus fund subsidies]—we must rely on standardized tests to monitor progress—but this is an important area for reform and an important conversation to have.” This, however, is the wrong approach. If standardized test-based accountability is doing the damage that Duncan has identified, it makes no sense to exacerbate that damage by continuing to rely on these tests to monitor progress. Such reliance will make it even harder to rescue American education in the future.

With the political prospects dismal both in Congress and the country for a new test-based ESEA, trying to maintain the existing flawed system while figuring out a new one is likely to be a frustrating endeavor. Rather than rush into a hastily designed policy that risks unforeseen and dangerous consequences, as was done in 2001, now is the opportunity to make a fresh start. The Obama administration’s ESEA re-authorization proposal should reject the continued punishment of schools based on flawed standardized tests, and instead focus on the careful and cautious design of new forms of qualitative evaluation.