The important policy memo below, written by Kevin Welner and William J. Mathis, explores how the debates in over the reauthorization of the Elementary and Secondary Education Act (which in its current version is NCLB), ignore the harm caused by NCLB’s test-based reforms. They write:
As a result, the proposals now on the table simply gild a demonstrably ineffective strategy, while crowding out policies with proven effectiveness. Deep-rooted trends of ever-increasing social and educational needs, as well as fewer or stagnant resources, will inevitably lead to larger opportunity gaps and achievement gaps. Testing will document this, but it will do nothing to change it. Instead, the gaps will only close with sustained investment and improvement based on proven strategies that directly increase children’s opportunities to learn.
Welner is the director of the National Education Policy Center, an attorney and a professor education policy at the University of Colorado Boulder. Mathis is the managing director of the center and a former Vermont superintendent. You can find the fully annotated memo here, at the National Center of Education Policy website.
By Kevin Welner and William Mathis
Today’s 21-year-olds were in third grade in 2002, when the No Child Left Behind Act became law. For them and their younger siblings and neighbors, test-driven accountability policies are all they’ve known. The federal government entrusted their educations to an unproven but ambitious belief that if we test children and hold educators responsible for improving test scores, we would have almost everyone scoring as “proficient” by 2014. Thus, we would achieve “equality.” This approach has not worked.
Yet over the past 13 years, Presidents Bush and Obama remained steadfastly committed to test-based policies. These two administrations have offered federal grants through Race to the Top, so-called Flexibility Waivers under NCLB, School Improvement Grants, and various other programs to push states, districts, and schools to line up behind policies that use these same test scores in high-stakes evaluations of teachers and principals, in addition to the NCLB focus on schools. The proposed new Teacher Preparation Regulations under Title II of the Higher Education Act now attempt to expand the testing regime to teacher education programs. These expansions of test-driven accountability policies require testing even beyond that mandated by NCLB.
Not surprisingly, current debates over the reauthorization of the Elementary and Secondary Education Act (ESEA), of which NCLB is the most recent iteration, now center around specific assessment issues such as how many tests should be given and which grades should be tested, as well as the respective roles of state and federal governments. Largely lost in these debates is whether test-based accountability policies will produce equitable educational opportunities through substantially improved schooling. This NEPC Policy Memo explains why they will not. Instead, we argue that as a nation we must engage in a serious, responsible conversation about evidence-based approaches that have the potential to meaningfully improve student opportunities and school outcomes.
Where We Started
NCLB was an ineffective solution to some very real problems. Opportunities for children to grow, learn, and thrive were inequitably distributed in 2001, just as they are now. For very good reasons, many civil rights groups lined up behind NCLB (just as some now continue to support test-based reforms). As articulated by President Bush: we need to end the “soft bigotry of low expectations.” We must maintain universally high expectations for our students, teachers, principals, and schools. Nobody should advocate that we return to an idealized version of pre-NCLB schooling.
Current policy fails to provide the schooling supports necessary for student success and largely ignores the many opportunity gaps children face outside of school that powerfully affect school performance. Standards-based accountability reform as originally conceived in the late 1980s acknowledged the importance of these factors. As originally adopted, NCLB included additional, albeit insufficient, federal funding for public schools, as did the stimulus funding that included Race to the Top. But the Great Recession wreaked havoc on state budgets, and the additional federal funding also (twice) dried up. Moreover, the federal incentives created through Race to the Top and the ESEA flexibility waivers failed to pressure states to tackle resource inequities. As a result, adequate school funding remains a key, unaddressed equity issue.
Students from low-income families are now a majority in U.S. public schools. Many of these children live in conditions of concentrated poverty, with few educational resources in their homes or communities. Their parents face unemployment, underemployment, food insecurity, a lack of stable housing, and many other obstacles that seriously undermine children’s opportunities to learn. Well-supported schools with substantial resources can make an important difference in these children’s lives, but it is not realistic to expect schools to be the nation’s primary anti-poverty program. Doing so is as unfair to children as it is to educators. It is a false promise.
What’s Included in, and What’s Missing from, the Current Testing Debate
There is now a parent-led backlash against “over-testing,” and politicians in both major parties are paying attention. These parents point to the time spent administering the tests themselves as well as to the diversionary effects of high-stakes testing on curriculum and instruction—which include narrowed curriculum, teaching to the test, and time spent preparing for the high-stakes assessments.
Nevertheless, the debate in Washington, D.C., largely ignores the fundamental criticism leveled by parents and others: testing should not be driving reform. Often missing this point, many politicians have begun to call merely for reducing or shortening the tests. Some also want to eliminate the federal push to use the tests for teacher evaluation while at the same time leaving untouched the test-driven accountability policies at the center of education reform. Other politicians are less interested in whether testing mandates continue than whether those mandates come from the states or from the federal government.
This kind of tinkering at the margins is just more of the same; the past decades have seen a great deal of attention paid to technical refinement of assessments—their content, details, administration, and consequences. In the words of long-time accountability hawk Chester Finn, “NCLB Accountability is Dead; Long Live ESEA Testing.” But the problem is not how to do testing correctly. In fact, today’s standardized assessments are probably the best they’ve ever been. The problem is a system that favors a largely automated accounting of a narrow slice of students’ capacity and then attaches huge consequences to that limited information.
Testing used as a diagnostic or summary instrument for children’s learning can be a helpful tool. It is harmful, however, to use students’ test scores as a lever to drive educational improvement. This use of testing is ill-advised because, as described below, it has demonstrably failed to achieve its intended goal, and it has potent negative, unintended consequences.
Since NCLB became law in 2002, students may have shown slight increases in test scores, relative to pre-NCLB students. Looking at the results of the National Assessment of Educational Progress (NAEP), however, any test score increases over the pre-NCLB trend are very small, and they are miniscule compared to what early advocates of NCLB promised. We as a nation have devoted enormous amounts of time and money to the focused goal of increasing test scores, and we have almost nothing to show for it. Just as importantly, there is no evidence that any test score increases represent the broader learning increases that were the true goals of the policy—goals such as critical thinking; the creation of lifelong learners; and more students graduating high school ready for college, career, and civic participation. While testing advocates proclaim that testing drives student learning, they resist evidence-based explanations for why, after two decades of test-driven accountability, these reforms have yielded such unimpressive results.
The National Academy of Sciences is among many that have critiqued test –driven incentives. The specific ways that tests are used within accountability policies is problematic. Again and again, experts have pointed to the violation of basic rules for test use, to the weaknesses in the scope and capacity of tests, and to the limited ability of growth models to make valid inferences about whether a given input (such as an individual teacher) actually caused a given student’s measured changes in test scores.
As important as those concerns are, the unintended consequences of testing are even more alarming. Researchers have long documented how top-down mandates can result in goal displacement. That is, the measured goal of increased test scores displaces the larger goal of increased learning. Thus, even while test scores may have inched up, other facets of students’ education have suffered. All the standardized tests we’ve given, past and present, capture only a small part of what we care about in education and in our children. Whether our goals are for citizenship or a well-prepared workforce, the narrowing of curriculum and constraining of instruction is harmful to the nation’s democratic and economic goals and those of the citizenry. To be specific, our singular focus on those test scores had negative consequences. For example:
- making schooling less engaging and creative;
- deprofessionalizing teachers and teaching;
- abandoning our past pursuit of learning that fully encompasses arts, music, social studies, and science; and
- marginalizing values and skills that help students develop the ability to cooperate, solve problems, reason, make sound judgments,and function effectively as democratic citizens.
We stress here that tests are useful when applied to their intended purposes and when there is legitimate evidence to support those purposes. Although measuring outcomes does not directly enrich learning, our schools do need disaggregated and useful information about how schools are serving students. This is an important part of a healthy evaluative feedback loop. The problem is not in the measurements; it is in the fetishizing of those measurements. It is in the belief that measurements will magically drive improvements in teaching and learning. It is in the use of test scores to issue facile admonishments: try harder! teach smarter! retain the child in third grade! reconstitute the school staff! It is in the singular focus on achievement outcomes to the exclusion of focusing on children’s opportunities to learn or on the system’s needs.
Policymakers know how to do these things and often do good work when they listen to and educate their constituents. Some state and federal initiatives are aimed at evidence-based reforms, such as expanding high-quality early childhood education and community schools. These remain small exceptions, however, within a system that still has test-based accountability at its core.
Considering the Equity Argument for Test-Based Reform
For those focused on educational equity, the appeal of NCLB and other test -based accountability policies arises from the nation’s history of ignoring fundamental educational needs of entire subgroups of children: those with special needs, those whose first language is not English, and those who are neither wealthy nor white. The requirement in NCLB that test scores be disaggregated served to call greater attention to achievement gaps. The hope was that this greater attention would be followed by greater resources and greater opportunities. In fact, reading recent statements from some civil rights groups today, calling for Congress to continue NCLB’s testing requirements, this still seems to be the hope. But we do not see any reason to believe that a test-focused ESEA in 2015 would yield any greater focus on opportunities to learn than did a test – focused ESEA in 2002.
It is important to note that achievement gaps were well known prior to NCLB. The disaggregation of NAEP test results has provided clear documentation of achievement gaps for many decades. What NCLB and related policies added was a set of punitive interventions, not a guiding knowledge of the gaps and not a set of strategies and resources to close the gaps. Test-focused policies, whether we start with 2002’s NCLB or 1994’s “Goals 2000” and “Improving America’s Schools Act” (when the federal push for test – based accountability began), have not closed the nation’s achievement gaps. “While the states’ progress was uneven among different grades, subjects, and subgroups, NCLB did not yet evidence sustainable and generalizable high-stakes accountability policy effects.”Disaggregating data and “shining a light” does not teach children. Teaching children requires good teachers and resources.
The apparent embrace of test-focused policies by some civil rights groups is all the more puzzling given the recent history of achievement gaps. As Prof. Jaekyung Lee explained in 2006:
Racial and socioeconomic achievement gaps narrowed substantially in the 1970s and early 1980s. During the 1970s, education and social policies worked to narrow the achievement gap by guaranteeing a minimally adequate level of achievement for minorities through compensatory education, minimum competency testing, school desegregation, equalization of school funding, the war on poverty, and affirmative action. As the focus of education policy . . . shifted from equity to excellence during the last two decades, there [arose] a potential tension between academic excellence and equity. In the 1990s, racial achievement gaps stopped narrowing or began to widen, signaling setbacks in the progress the nation made toward educational equity.
Achievement gaps are important. More specifically, gaps in scores on high-quality tests are important, and it is important to continue measuring those gaps. But the NAEP serves that purpose; policymakers do not need additional test results to identify the core needs and problems.
Closing Opportunity Gaps in Communities of Concentrated Poverty
In government, the process of enacting laws is generally separate from the appropriations process. This often leads to disjointed, disconnected, and underfunded laws. If we as a nation are to continue asking our schools to somehow counteract the effects of poverty and other societal ills, we will need to provide children in resource-starved communities with extraordinarily enriching opportunities within those schools. Looking to the adequacy studies across the nation, each economically deprived child should receive between 40 percent and 100 percent greater funding than the average student, while they actually receive about 19 percent greater funding. In fact, by one measure, urban districts serving our most needy children have only 89 percent of the national average in revenues.
The original language of ESEA’s Title I program provided that each child living in poverty would receive an additional 40 percent of the state’s average spending. Neither the federal government nor the states have ever appropriated sums of this magnitude. The current discussion in Congress similarly ignores this promise and this need. In fact, one proposal is to make Title I funds “portable,” which would have the effect of moving even more funds away from schools with the greatest needs. Moreover, next year’s sequester, with the automatic reductions in ESEA and other education and social services funding, remains a real threat. This dire picture is compounded by the United States now having one-third of all our children living in poverty, half our public school students living at or near poverty, a 2 percent annual increase in children in poverty, and one of the worst income disparities among developed nations.
The above-described pattern of ever-increasing social needs and educational needs, as well as fewer or stagnant resources, will inevitably lead to larger—not smaller—opportunity gaps and achievement gaps. Testing will document this, but it will do nothing to change it. Instead, the gaps will only close with sustained investment and improvement based on proven strategies that directly increase children’s opportunities to learn.
NCLB and similar policies have done a disservice to the word “accountability.” Our nation and our nation’s education system need accountability, but it must be fair and it must be universal. Holding teachers accountable but excusing the policymakers who fail to provide necessary supports is as harmful and illogical as holding students accountable but excusing poor teaching. Today’s demoralized teaching force has been given too much responsibility for outcomes and too little control over those outcomes. A system of universal accountability would continue to make strong demands on teachers and principals, but equally strong demands must be made on the leaders and policymakers in district offices, state and federal legislatures, and state and federal departments of education. This, too, was the basic idea behind standards-based accountability reform as originally conceived in the late 1980s. But it has largely disappeared from the system. When one part of the educational accountability system disappears, the system cannot provide the necessary opportunities to learn.
The purpose of ESEA has long been to ensure equal opportunities for all children, particularly those living in poverty, and the reauthorization debates offer us an opportunity to remember that purpose and to re-think our national directions in education. This must begin with a sober and honest look at the effects of the No Child Left Behind Act. The broad consensus among researchers is that this system is at best ineffective and at worst counterproductive. The issues now being debated in Washington largely ignore these facts about the failure of test-based educational reform, and the proposals now on the table simply gild, rather than transform, a strategy with little or no promise.
The ultimate question we should be asking isn’t whether test scores are good measures of learning, whether growth modeling captures what we want it to, or even whether test scores are increasing; it is whether the overall impact of the reform approach can improve or is improving education. Boosting test scores can, as we have all learned, be accomplished in lots of different ways, some of which focus on real learning but many of which do not. An incremental increase in reading or math scores means almost nothing, particularly if children’s engagement is decreased; if test-prep comes at a substantial cost to science, civics, and the arts; and if the focus of schooling as a whole shifts from learning to testing.
The way forward is not to tinker further with failed test-based accountability mechanisms; it is to learn from the best of our knowledge. We should not give up on reaching the Promised Land of equitable educational opportunities through substantially improved schooling, but we must study our maps and plan a wise path. This calls for a fundamental rebalancing—which requires a sustained, fair, adequate and equitable investment in all our children sufficient to provide them their educational birthright, and an evaluation system that focuses on the quality of the educational opportunities we provide to all of our children. As a nation, we made our greatest progress when we invested in all our children and in our society.
You can find the fully annotated memo here, at the National Center of Education Policy website.
You may also be interested in: