The Future of Test-based Accountability
In the U.S., the principal lever for K-12 public education reform for the last 40 years has been test-based accountability. Prior to the 1970s, individual school districts bore nearly all of the responsibility for determining what the students within their purview needed to know and be able to do to advance from grade to grade and graduate from high school. Districts, in turn, deeded this responsibility to teachers in the form of the end-of-year or end-of-course grades they assigned to their students.
State-mandated minimum competency tests (MCTs), established in the 1970s, were the first wave of accountability systems designed and overseen at the state or federal level. An MCT is a standardized exam of basic skills on which a passing score is intended to signify that a student has acquired at least the minimum necessary skills and knowledge for promotion to the next grade or for graduation from high school. In 1973, only two states had test-based minimum competency requirements. By 1978, 30 states had passed legislation requiring such exams.
Today’s approach to standards and accountability is exemplified by the goals of the No Child Left Behind Act of 2001 (NCLB), the Obama administration’s revisions of NCLB through Race to the Top and discretionary state waivers, and the Common Core State Standards movement. The standards and accountability systems under which every public school in the nation operates in this century differ in many respects from earlier MCT systems. The most obvious of the differences lies in the conceptualization of what states and the federal government should hold educators and students accountable for: minimum competency vs. college and career ready skills.
Think of the MCT movement as based on a philosophy akin to that which has historically motivated central government’s role in regulating the food supply. The goal is not to assure that the public eats well, but that it avoids diseases that are transmitted through contaminated foods. Under present day standards and accountability systems, states, pushed and prodded by the federal government, have moved from trying to force districts to educate students to a minimum level of basic skills and to do something about schools that are obviously failing, to holding districts, schools and teachers accountable for (in the words of the Common Core State Standards Initiative) “preparing all students for success in college, career, and life.” It is as if the FDA switched from its role of keeping wiggly things out of our food stock, at which it is pretty good, to assuring that we eat a healthy diet. 
Despite its lofty goals for every student and the support of some of the most powerful national players in education reform, the modern standards and accountability movement has hit a rough patch, if you haven’t noticed. It seems that not a week passes without a setback. Notable recently were the Gates Foundation’s call for a two-year moratorium on tying results from assessments aligned to the Common Core to consequences for teachers or students; Florida’s legislation to eliminate consequences for schools that receive low grades on the state’s pioneering A-F school grading system; the teetering of the multi-state Partnership for Assessment of Readiness for College and Careers (PARCC) assessment consortium (down from 24 to 15 members, and with its contract with Pearson to deliver the assessments in limbo because of a lawsuit that alleges bid-rigging); and the groundswell of opposition from parents, teachers, and political groups to the content of the Common Core.
In this context, there are three different future scenarios. The first is that the teacher unions win, and we head back to a future in which we just let teachers teach without meddlesome interference from above. The second is that the Common Core movement muddles through, meaning that we end up, eventually, with a nearly national set of standards for what students need to know and do at each grade, high quality assessments aligned to those standards, cut scores for proficiency on those assessments that are challenging and equal across the nation, and a set of meaningful carrots and sticks for holding educators accountable for preparing all their students for success.
The third possibility – the one I’m betting on – is that we’re in a transformative period fueled by a kind of restlessness that nobody is getting accountability right, the achievement problem remains, and ideas are not manifold about what to do next. At some point the prevailing standards and accountability approach to education reform will be replaced with new designs that are more productive, or at least different. If I’m right about this, it would be fruitful to identify the variables that could be in play if and when the nation or individual states are ready to try something different.
In that spirit I’ll offer a taxonomy that has helped me to think systematically about the ways that standards and accountability systems differ. Note that my focus is test-based accountability. Other things I’ll not cover here, such as students’ aspirations and soft skills, are important too.
|Dimensions of Test-Based Accountability Systems*|
|Who||For What||By Whom||How|
|Student||Standards-based Basic Skills||Federal Government||Tangible Consequences|
|Teacher||Standards-based Proficiency||State Government|
|Principal||Comparative Performance||District||Market Mechanisms|
|Superintendent||Grades in class||School|
*This taxonomy comprises four separate dimensions, with all combinations of entries across dimensions theoretically possible.
Who is held accountable? The accountability focus of NCLB was school principals and district superintendents. This is evident in the federal law’s requirement that each state’s accountability system generate a report card for each school and district indicating the proportion of students meeting proficiency standards on state tests of math and reading. It is also evident in that the real teeth of the law were reserved for schools, e.g., mandatory school restructuring for schools persistently failing to make progress towards targets of student proficiency. Teachers and students, while certainly affected by the accountability consequences applied to schools and districts, were not directly accountable under NCLB.
The Obama administration’s revisions of NCLB have shifted test-based accountability towards teachers, whose tenure and retention is to be dependent on their classroom ratings and students’ test score gains, and away from schools. Under NCLB, all schools were accountable for the proficiency of all subgroups of students, whereas under the Obama administration’s waivers, only the bottom tranche of schools face consequences.
Note that students, by and large, face no accountability for their performance on state tests under either NCLB or the Obama revisions. This is very different from the MCT-based accountability systems of the 1970s, under which students were held accountable, for example, for passing a high school exit exam if they were to receive a regular high school diploma.
For what are students, teachers, principals, or superintendents held accountable? Prior to the 1970s, students were accountable for the grades they received from their teachers. There was no test-based accountability outside the classroom. During the reign of MCTs, the focus was on basic skills, frequently indexed by a cut score on norm-based assessments such as the Iowa Test of Basic Skills. Under NCLB, the responsible parties were accountable for what, in concept, was to be student proficiency on rigorous standards. In fact, some states, e.g., Massachusetts, adopted rigorous standards and assessments against those standards, but most states defined proficiency such that it was closer to minimum competencies or basic skills. The Common Core is supposed to address the “dumbing down of standards” under NCLB by requiring all students to demonstrate proficiency on challenging material.
By whom are students, teachers, principals, and superintendents held accountable? Under NCLB and the Obama revisions, central government controls the accountability system. It need not be that way. Up until the 1970s, individual school districts bore that responsibility. In most of the private school sector today, accountability is at the level of the individual school, i.e., with the principal and the school’s board.
How is accountability realized? Most of the accountability consequences in NCLB and the Obama revisions have been tangible top-down actions. Many have been punitive, e.g., the restructuring of a school means the principal is replaced. Salary and bonuses have frequently been part of the contracts of school principals, contingent on students meeting test-based performance targets. Superintendents have understood that they are expected to increase test scores, and their jobs are on the line.
An alternative approach to accountability, prevalent in the charter school sector, and increasingly finding its way into traditional public school districts, is based on choice and competition. The concept is straightforward and similar to the way the higher education sector operates. Let students and their parents choose the school the student will attend. Arm those shopping for schools with information on how the individual schools available to them are performing. Let the money in the form of taxpayer dollars follow students to their school of choice. In this model, schools close or are restructured because they are unpopular, i.e., they don’t attract enough students to pay their bills, not because they fail to meet a regulatory requirement that their students perform well on an internationally benchmarked assessment, or their teachers have particular credentials, or they utilize a certain curriculum, etc.
I was part of the administration that brought you NCLB. That law, passed in 2001, required that every state have a federally approved plan under which every child would be proficient in math and reading by 2014 – in other words, by right about now.
How did that work out? The graph below presents the most recent national NAEP results in reading for 4th graders, plotting the percentage of students who achieved below the basic level for each year from 1998 through 2013. There’s been some progress in this outcome since the onset of NCLB in 2002, but as of last year 32 percent of the nation’s fourth graders were still at the lowest level on NAEP. These students could not locate relevant information, make simple inferences, and use their understanding of the text to identify details that support a given interpretation or conclusion. Nor could they reliably interpret the meaning of age-appropriate words as used in the text.
What happens when high stakes are attached to a test score criterion of success that cannot be reliably reached by the vast majority of schools and students? The system falls apart.
This is exactly what happened to NCLB. For the 2009-2010 school year, 38 percent of schools in the nation were failing to make their achievement targets under NLCB. Secretary of Education Arne Duncan testified before Congress in 2011 that the percentage of schools failing to make their targets would soon become 82 percent. This marked the end of NCLB as an accountability system. Duncan soon implemented NCLB waivers under which states were relieved of the responsibility of intervening except in the lowest performing schools.
We’re now moving into Alice-In-Wonderland territory: States in which the majority of schools and students were failing under No Child Left Behind are required to adopt even more rigorous standards and assessments under the Common Core. The present rallying cry of college and career ready is distinctly more challenging than no child left behind. So we have an accountability system that can’t be sustained politically because it labels too many schools as failing, and the solution is to replace it with a new system that will surely produce even more bad news? In New York State, the percentage of students scoring proficient in math under the new Common Core Assessments in 2013 plummeted to 31 percent from 65 percent the prior year under the NCLB assessments. The consequence – immense pressure from teachers and the public to place a moratorium on accountability – was entirely predictable.
One way forward
The dimensions of accountability I’ve described represent the building blocks of systems that could look quite a bit different from what we have now and where we seem to be heading.
For example, what if top-down accountability as defined and enforced by states and the federal government were limited to basic skills, something like the MCTs of the 1970s but based on clear, empirically validated standards about what students need to know and be able to do to hold down the types of jobs that do not require a college degree? In other words, what if we had NCLB with standards for student achievement that would actually be reachable? Accountability for not teaching students to read and do math at this basic level would be severe: a school not doing so would be shut down by government as surely as a restaurant serving food that sends its customers to the hospital.
With the responsibility of central government trimmed back, school districts and individual schools would have the authority and freedom to innovate around curriculum, staffing, and other matters central to what students are expected to learn and how they are taught. Some would focus on a traditional college preparatory curriculum for which Common Core standards and assessments might be the best vehicle, whereas others might focus on career and technical education, science, or the arts. Or they might focus on college preparation with a different vision of how that can best be realized than is found in the Common Core. Or they might do all of these things, but for different sets of students.
In such a split system, in which federal and state government holds all schools accountable for assuring that students acquire basic skills, whereas local school districts or individual schools are free to define their goals for student learning beyond basic skills, there would be the opportunity for variety and innovation of the sort that has served us well in other sectors of the economy.
The accountability for schools under such a varied system of education providers and services would probably need to be market-based. How else to mate the student who wants what school A is providing to that school if not by the student’s choice? Such a system would require clear publicly-available information on school performance, but the performance metrics would have to be relevant to the school’s mission and comparable among schools with similar missions rather than one-size-fits-all.
The dearth of student accountability for test score results other than tests given by teachers in the classroom is a remarkable aspect of current test-based accountability systems in the U.S. There are exceptions, of course, in the form of high school exit exams, end-of-course exams, and grade promotion requirements. But the exceptions are under siege, as for example in the recent roll back of high school exams by the Texas legislature. High stakes exams for students play a prominent role in the education systems of many other countries, with suggestive evidence that they enhance student achievement. Why shouldn’t students in the U.S. be held accountable for their own learning? With the inevitable growth of web-based course materials and the corresponding displacement of traditional forms of classroom instruction, standardized high-stakes end-of-course exams are the future, even if they are under challenge now.
One thing I’m sure about in education is how little we know about what works best for which students under what circumstances to produce the best long-term outcomes. This includes how to design an optimal standards and accountability system. Accountability 3.0 doesn’t have to look like Accountability 2.0. That’s certain. Perhaps it is time for central government to yield some of the control over standards and accountability that it has amassed in the last 30 years in exchange for the opportunity for districts and schools to innovate around what students are taught and how the actors in the system are held accountable.
This post originally appeared on the Brown Center Chalkboard.
 With appreciation to the late Tom Greene by way of Tom Loveless for the “wiggly things” analogy.