A little more than a decade ago we embarked on what is arguably the most significant change in educational policy of the past half century – the introduction of No Child Left Behind. While many states had already introduced some form of test-based accountability by 2001, NCLB both made this mandatory for all states and introduced a very specific structure to accountability that importantly included consequences for schools that did not perform well.
I always viewed this as an experiment representing the best guesses of President George Bush and the U.S. Congress about how to improve national educational performance. While there was strong bipartisan support, NCLB was tempered as always by the conflicting political forces of interested parties, including the one anachronism about teacher quality that was based on inputs rather than outputs.
As with any one thousand page guess, I also thought the idea of revisiting the law in 2007, the date designated for its re-authorization, was an important part of the underlying wisdom of the act. Without researching it, I suspect that other Congressional acts have missed their re-authorization date by wider margins. But given the importance of this act to the hopes, aspirations, and operations of our schools, I am willing to assert that this ranks among the most consequential dropped balls of Congress.
Faced with this, Secretary Duncan did more than just rank historic Congressional missed deadlines. He established a waiver process that effectively allowed two-thirds of the states to deviate from various requirements of the law – most prominently the requirement that all students be proficient in math and reading by the end of this year.
While waiting to see where this new phase of accountability takes us, I think it is useful to return to the task of 2007 and to judge what might or might not usefully change in NCLB.
NCLB has a fairly simple structure: states were required to develop learning standards along with consistent assessments of student accomplishment of these standards; schools were required to be on a glide path to get all of their students up to a state-defined proficiency level by 2013; and the federal government established a series of corrective actions – including provision of supplemental services, broad student choice, and reconstitution – that were required of individual schools after continual deviation from this glide path.
NCLB has been a polarizing policy – in large part pitting school personnel against a coalition of civil rights groups, reformers, and, to a large extent, parents. While it is difficult to assess definitively the impact of NCLB, the best evidence suggests that it has had positive impacts on student performance as measured by the National Assessment of Educational Progress; it has narrowed some of the wide achievement gaps by race and ethnicity; and it has generally led to much more attention to the importance of student performance. But, on the other side, it has not brought all students anywhere close to being proficient; it may have narrowed instruction and the curriculum in general; it may have led to triaging of students and schools close to the cutoff while neglecting the rest of the distribution.
In my view, test-based accountability is both unlikely to go away and shouldn’t go away, regardless of the objections currently expressed by school personnel. Yet, as one who has studied many aspects of NCLB, I also believe that it has serious structural flaws (making its overall beneficial effect on achievement even more remarkable). Thus, I want to return to the task of 2007.
I can succinctly state what I think needs to change. First, the structure is backwards. NCLB has individual states determine “what” is to be accomplished and has the federal government determine “how” that should be done if schools fail to meet these goals. This allocation of responsibilities should in my opinion be reversed. Our achievement goals should be a national decision, not an individual state decision. The U.S. is really a single labor market that has common demands for skills of the population. On the other hand, the states (and districts) should be in charge of deciding how we achieve those goals – instead of trying to determine that from Washington. (The one caveat is having the federal government set goals is likely to be highly politicized, and it is important to find a way of insulating this from pure politics).
Second, the tests need to be improved so that they do not stop at the most basic levels. Third, accountability should include measures of learning growth across the entire spectrum and not be restricted to the bottom rung of performance. Fourth, we should retain the objective outcome focus for student performance. Fifth, subgroup disaggregation should be central, because this has led to some significant equity gains.
We have actually moved reasonably close to these changes in some dimensions, partly because of the standards and testing associated with the common core and partly because of the waiver process. If the tests being developed can support both enhanced accountability and measures of learning growth, we have the infrastructure in place for change. The waivers have also brought states into deciding how best to meet goals, albeit with a still excessive involvement of the federal government in process issues and in the “how” of education.
Congress should, in my opinion, move on the full agenda – rationalizing and solidifying the patchwork waiver process and reinforcing the need to improve our schools. Congress has shied away from making politically difficult educational decisions – but continuing on this course threatens long term damage to our economy and our nation.
Eric Hanushek is the Paul and Jean Hanna Senior Fellow at the Hoover Institution of Stanford University.
An earlier version of this discussion appeared on the “Bridging Differences” blog of Education Week, along with a response by Deborah Meier.