Why High Stakes Accountability Sounds Good But Doesn‘t Work— And Why We Keep on Doing It Anyway
We have bet the future of federal education policy on a theory of accountability that does not work. It has been the dominant educational reform theory for decades and the No Child Left Behind Act (NCLB) is its extreme expression. It turns out, after studying research results from across the country, it does not make much sense either as a managerial or an educational strategy. It has very good intentions but often sanctions those institutions where progress is most difficult and most urgently needed rather than offer the kind of help that could really make a difference. This report, by researchers Heinrich Mintrop and Gail Sunderman, dissects the logic of high-stakes accountability policies, explores why they have failed, and concludes that the failure was not one of implementation (though that made things worse) but of the basic structure of the policy.
The Civil Rights Project has been studying the results of NCLB in six states since it was passed and has previously issued 12 reports, as well as two books and a number of articles, on its implementation and the results. Gail Sunderman has led this research. Professor Mintrop is a leading expert on the impacts of sanctions-based policies at the state as well as national level. Years ago, we showed that the standards were inconsistent and sometimes meaningless, the goals were incorrectly set and unfairly punished integrated schools and those serving English language learners and other minority groups. Our work showed that the law‘s assumptions about teachers and sanctions were wrong and that the sanction process was undermining the good goal of keeping experienced teachers where they were most needed. We showed early on that neither the transfer option nor the supplemental educational services provisions were working. Work we commissioned demonstrated that the dropout provisions had been gutted and that the requirements placed on the states went far beyond the limited capacities of state agencies to fulfill. We have recognized all along that the goals of more equal outcomes, good statistics on outcomes by subgroups, and a number of other provisions in the Act could be part of a good policy. The Civil Rights Project has joined many other researchers in recommending the replacement of the very narrow and arbitrary goals of test scores in two subjects with a much richer accountability scheme. When we originally raised a number of these issues we were attacked by the law‘s defenders, but, increasingly, the issues we raised have become part of much more broadly shared views of the NCLB experience. We believe that applying the lessons of the past eight years could produce a much more effective federal policy.
Now, as the country thinks about what to do next, it is important to focus on some fundamental design problems with the NCLB that undermine its very important goal of increasing the equity and success of American schools. The first is that it was not designed around real educational experience, nor does it utilize what research has shown about the sources of educational inequality or the possibilities and conditions necessary for reform to work. Instead, NCLB is based on the dual assumptions that children are falling behind very largely because educators don‘t care enough and that deadlines and strong sanctions imposed by the federal government can cure the problem so that all subgroups of children will become proficient by 2014. The second problem is that it often punishes schools that are making a positive difference for students, discouraging the staff and undermining future prospects for the school. The third is that it has a very narrow definition of education that not only diverts attention from other vital goals but also produces a strong focus on tactics that create a semblance rather than reality of success in those limited areas. The fourth is that all schools are being required to attain goals that are impossible to attain on any broad level given what we know about both the impact of schools relative to other forces in children‘s lives and the distribution of talent and achievement that appears in all human populations. Finally, while the law obviously hopes schools will experience deep reform, the deadlines and yearly goals do not connect with what is actually known about the time and capacity-building required to actually turn around a school. I believe that there are good ways to correct each of these problems.
This study, commissioned by the Civil Rights Project, finds that some fundamental assumptions of the law are in error and, if continuously pursued, are very likely to do more harm than good. Since state and local educational institutions have the primary responsibility for public education—paying nine-tenths of the bills and setting most of the rules—the first requirement for federal policy should be that it does no additional harm to the public school systems. This report shows that, in that respect, NCLB falls short—not only in operation but even in its design and basic assumptions. A reasonable standard would be that a policy not weaken key institutions, not undermine support for public schools, and not try to impose impossible requirements. Historically the role of the federal government has been to encourage new initiatives, to commission research, to disseminate information and statistics, and to provide resources to the schools. The sudden decision by NCLB to define the most important subject matter, to mandate the grades that are tested, to control teacher requirements, to set detailed requirements for yearly educational gains, and to order what can be very drastic sanctions represent truly radical interventions on state and local authority. A reasonable standard for such interventions is that: 1) they are based on a solid understanding of schools and school reform; 2) they make sense to and win cooperation from the teachers, local administrators and state officials who must try to make them work; and 3) they provide the resources needed to meet the additional demands. As it now stands, NCLB does not meet any of those standards.