With the decision in August by Connecticut’s attorney general to file a lawsuit against the U.S. Department of Education, and with the prospect of similar suits from other states, the No Child Left Behind Act faces its greatest challenge since adoption by Congress in 2001. (“Connecticut Files Court Challenge to NCLB,” Aug. 31, 2005) The federal court system has played an important role in education during the last century, protecting the rights of minority students, students with disabilities, and the First Amendment rights of all students. The question now is whether federal mandates for annual testing in reading and math are also matters to be resolved by the courts.
There is a temptation to cast the players in this legal drama in the extremes of courtroom rhetoric. Proponents of federal mandates are caricatured as engaged in a thinly disguised effort to use children as political pawns to humiliate teachers and school administrators and, as an alternative, to make vouchers and charter schools appear more appealing. Those objecting to the testing provisions of the federal law are lambasted as education bureaucrats hiding from responsibility, preferring job security to the welfare of children. From such mutual vilification is litigation born.
There is a better way. We can respect the legitimate concerns of both sides in the debate and mediate, rather than litigate, a solution.
To be fair, the U.S. Department of Education has already exhibited some flexibility, with modifications of rules regarding special education, English-language learners, and participation rates in mandatory testing. Most recently, Secretary of Education Margaret Spellings has announced measures that could give schools affected by the Gulf Coast hurricanes some relief from the accountability provisions of the law. Moreover, the language of the No Child Left Behind statute itself, too frequently overlooked in this debate, explicitly provides for broad state discretion in matters of standards, assessment, and accountability.
States have also exhibited creativity, providing a wide range of content and formats in tests, as well as some exceptionally innovative ways to involve teachers, parents, and other stakeholders in the processes of standards, assessment, and accountability. Nebraska, Massachusetts, Virginia, and Wyoming, states remarkably divergent in politics and philosophy, have all provided models of ingenuity, rigor, and balance in meeting the requirements of the federal law and, more important, improving the achievement of their most disadvantaged students.
There is a balance between anarchy and micromanagement, and people of good will on both sides of this debate must be able to reason together and invest our resources in education rather than litigation. Four keys provide the way to a mediated solution to the current No Child Left Behind debate.
First, let the states acknowledge essential “hygiene standards” in education standards and testing. If a hundred guests came to dinner—50 governors and 50 state education leaders—we would have at least a hundred different ideas about menus and the presentation of food. But surely we would all agree that the cutlery should be clean and the kitchen should be sterile. What are the comparable hygiene standards for education standards and testing?
While the states can provide significant variation with regard to the content of standards and the format of tests, we should at least agree that the tests are to be related to the curriculum and standards the students received, and that they must be administered in a manner that meets minimally acceptable requirements for fairness, validity, and reliability. We should also agree that real data from annual tests are superior to estimates of student performance. When in doubt, let the data speak, as the actual performance of students sometimes confounds the predictions of experts.
Second, let the federal government acknowledge that “fairness, validity, and reliability” can be achieved in multiple ways. Specifically, the Education Department should allow value-added assessment, in which “adequate yearly progress” is calculated based on a same-student-to-same-student formula, rather than a comparison of this year’s 4th grade students and last year’s 4th grade students. In addition, while statistical reliability is a function of consistent testing conditions and, in practice, a high number of test items, the federal government should subordinate traditional reliability measurements to the higher goal of gaining meaningful assessment data.
There is a balance between anarchy and micromanagement, and people of good will on both sides of this debate must be able to reason together and invest our resources wisely.
Such a trade-off means that public policy might be well served by a 20-item test with identical items for all students, followed by another 20 items that are differentiated to meet individual student needs. The statistical reliability of a differentiated assessment will be lower than that of a 40-item test that is “one size fits all,” but the gains to students, teachers, and policymakers of a more nuanced test that shows performance over a greater range of students far outweigh the value of an extra one-tenth of a point of statistical reliability.
It is important to note that not all “value added” models are equal. Only those models that show genuine growth through same-student comparison are completely transparent and, subject to independent validation, will justify public confidence. Statistical complexity is not a substitute for public confidence.
Third, let the federal government acknowledge that locally created tests have the potential to be valid and appropriate measures of student learning. During recent years of rampant growth in large-scale testing, it is fair to say that there have been embarrassing errors by corporate and local test providers. When locally created tests are found wanting, public exposure is a better cure than federal mandates for external test vendors. At the heart of potential litigation is not the issue of “testing vs. no testing.” In Connecticut, the first state to file suit against the federal government, it is highly unlikely that a single classroom fails to conduct a test of reading and math at least once a year. The central issue is whether those tests should be standardized and publicly reported. States are right to defend the potential validity of local assessment; but states are wrong to assert that the locally created tests are immune from reasonable requirements of validity and reliability.
Some might argue that local control of testing and the lack of statewide standardization will result in chaos. I would argue that in the absence of statewide standardization, the compulsion of public reporting will require every district, school, and classroom to publish its tests along with student work it regards as “proficient.” Such transparency will validate the “Goldilocks syndrome,” in which we will acknowledge that “just right” does not exist outside of fairy tales. Local standards will be either “too hard” or “too soft,” and public exposure of those standards will validate the former and force improvements upon the latter. But with sufficient state latitude, we could create precisely the right incentives for local districts. Rather than advertise that they have more students who are “proficient,” they can compete for the honor of contending that their neighbor’s “proficient” would need improvement in a local school. Now we appear to reward states for creatively avoiding rigor. Why not endorse states that encourage districts to creatively avoid mediocrity?
Fourth, let states acknowledge that complaints about “unfunded mandates” are a double-edged sword. Education funding has risen faster than any other part of the federal budget outside of defense, and, lest memory fail us, the pre-9/11 budget of 2001 included a higher increase for education than for defense. There are unfunded mandates in transportation, public safety, housing, civil rights, and scores of other areas. It is a short step from the complaint that educational accountability is an unfunded mandate that should be ignored, to the complaint that requirements for curb cuts for wheelchairs are an inappropriate imposition by the federal government on the will of the local populace. One person’s unfunded mandate is another person’s federally protected civil right.
There are no villains here. State and federal officials are decent people of good will who love children and strive for educational equity. The parties to this litigation are my friends, and it breaks my heart to see them go hammer and tong at one another. Litigation demonizes opposing parties, while education promotes mutual respect and a search for common ground.
Our society needs lawyers, professionals who protect our rights and help enforce the rule of law. But in this case, we are better served by investing our limited resources in education, not litigation.
A version of this article appeared in the October 12, 2005 edition of Education Week as Education, Not Litigation