Futile Accountability Systems Should Be Abandoned
Forum: Is Test-Based Accountability Dead?
Is test-based accountability “on the wane”? The question is based on a fallacy. For something to be on the wane, it has to exist, and test-based accountability has never truly existed in the United States. Holding people accountable requires that they face significant consequences as a result of their actions. Despite years of “high stakes” student testing, very few of the nation’s 3.14 million public-school teachers have ever lost a job, had their pay reduced, or otherwise faced meaningful consequences because of these test results.
It’s true that under the No Child Left Behind (NCLB) accountability regime, schools have been given labels, such as “in need of improvement.” Some have even been threatened with reorganization or closure. But these threats have only rarely been carried out, and the educators in these schools have typically just been reshuffled to other locations or new management. This enterprise does not constitute true accountability. It’s more akin to “double secret probation,” the toothless threat imposed on the party-mad frat boys of the 1978 film Animal House.
The real question we should address here is whether the hollow threats of this double-secret probation are on the wane. I think they are, and I say good riddance. While testing has failed to produce meaningful accountability, it has distorted the operation of schools to the detriment of educational quality—and it has proven politically unviable.
Test-based accountability is essentially a central-planning exercise similar to that used by officials in the Soviet Union in attempting to manage the country’s economy. In both cases, a distant official selected a particular goal for production, focused on a limited set of metrics to assess whether goals were met, and then threatened to impose rewards or sanctions based on whether those metrics showed desired results. Central planning failed in the Soviet Union, and it is failing here in public education—and for similar reasons.
First, education goals established by distant officials cannot possibly capture the diverse spectrum of local priorities in our nation. Officials have focused on improving math and reading ability, but emphasizing those subjects has come at the expense of other goals. Several studies, including a recent paper by the University of Virginia’s Daphna Bassok and colleagues, as well as widespread reports from educators, show that schools have shortchanged history, science, physical education, art, music, and civics. They’ve also cut back on culturally enriching field trips. Even within math and reading, schools tend to focus narrowly on tested items, which often exclude poetry, literature, and more abstract math.
Providing students with math and reading skills that are useful in the workplace is a worthy goal of education, but so is helping students become good citizens—cultured, tolerant, self-disciplined, and creative. With test-based accountability, distant officials have imposed their preferences on the rest of us. In addition, studies such as the ongoing research of David Grissmer and colleagues indicate that long-term achievement in math and reading depends on a broader education that includes the type of general knowledge conveyed by history, science, art, and music. Paradoxically, a narrow focus on math and reading may undermine later success in math and reading.
Second, the limited metrics used to assess math and reading achievement are easily gamed and further distort the educational process. If success is defined by the percentage of students who exceed a threshold for proficiency, officials will be tempted to lower the bar for what constitutes “proficient.” Schools will also be tempted to focus on students whose performance is below but close to the proficiency threshold, neglecting both high achievers and students who are unlikely to pass even with a reasonable amount of extra attention. School administrators and teachers will be tempted to cheat, as in the recent scandal in Atlanta, or to narrow their instruction, as mentioned earlier. And given that test-based accountability systems are almost entirely built around proficiency levels rather than growth, schools can appear more successful if they can avoid serving too many students who are difficult to educate.
Third, schools are gradually figuring out that few real consequences will befall them if they fail to meet the imposed metrics. School leaders’ bluffs about mass firings could only be sustained for so long. As a result, scores on the National Assessment of Educational Progress rose in the early years of test-based accountability, but more recently, those gains have stalled.
Furthermore, test-based accountability is built on the assumption that test results are reliable proxies for success later in life, but research has called that assumption into question. It’s true that test scores are correlated with some measures of later life success, but for test-based accountability to work we would need to see that changes in test scores caused by schools are associated with changes in later life success for students. Test-based accountability proponents can point to research by Raj Chetty and colleagues that shows a connection between improvements in test scores and improved outcomes in adulthood, but their work examines testing from the 1980s, prior to the high-stakes era, and therefore does not capture how the threat of consequences might distort the relationship between test-score changes and later life outcomes.
Furthermore, findings such as theirs are more the exception than the rule. A growing number of studies show a disconnect between short-term progress on test scores and long-term success. That is, even when we measure the extent to which schools contribute to student test-score growth—something that test-based accountability systems rarely do—we cannot consistently predict which programs or schools will help students be more successful later. We cannot centrally plan success if we cannot reliably predict success.
The educational failures of test-based accountability, as detrimental as they are, will not spell its demise. Rather, accountability that centers on testing is doomed because it has many political adversaries but no enduring political constituency. Parents have never rallied to demand that their children be tested more, that tests be used to retain students or prevent them from graduating, or that tests be used to determine teacher pay or employment. Educators revile test-based accountability even more. Test-based accountability was initiated by policy elites frustrated over rising education costs and subpar results. But elites cannot sustain such a policy in the face of opposition from educators and families. American politics is shaped by the activity of organized interests, not poll results. Other countries may be able to impose meaningful systems of test-based accountability, but the decentralized nature of American education and politics gives far more power to organized groups of upper-middle-class families and educators than to the technocratic elite.
The political weakness of test-based accountability helps explain why there are no meaningful consequences attached to it. Opponents have not been able to repeal testing, since there is broad support for information on student achievement—even partial and distorted information—but these adversaries have effectively neutered the consequences of accountability. So, the Every Student Succeeds Act continues to require testing, but the accountability piece is even more anemic than it was under NCLB.
The collapse of the Common Core State Standards illustrates the political folly of test-based accountability. Common Core attempted to transform largely symbolic accountability systems into something tougher, which is precisely why it failed. The standards were an effort to better articulate the proper goals of education. The federally subsidized tests aligned to Common Core and developed by the SBAC and PARCC consortia were intended as the rigorous metrics for this stronger accountability regime (see “The Politics of the Common Core Assessments,” features, Fall 2016). And centralized teacher-evaluation systems being pioneered by the Gates Foundation in their Measures of Effective Teaching effort were supposed to impose meaningful consequences for failure to perform well on those metrics.
Even these baby steps toward a real accountability system produced a fierce political backlash, led largely by suburban middle- and upper-middle-class families. Such families are accustomed to having significant autonomy with respect to what and how their children are taught, either by choosing the public or private schools their children attend or by influencing locally elected and appointed school officials. By its nature, test-based accountability shifts control away from these parents. Suburban families see Common Core as an infringement on their autonomy, and they have the savvy to fight back. As they do, we are seeing fewer than half of the states sticking to one of the Common Core testing consortia. Soon Common Core will become the same type of nonentity it was meant to replace.
What might constitute real accountability in K–12 education? The power of middle- and upper-middle-class families to exercise control over how and what their children are taught is one example. Suburban schools that stray from parental preferences may lose students and revenue or have to answer to angry parents. Test-based systems are politically doomed because middle- and upper-middle-class families tend to prevail in education politics. This phony accountability harms education and undermines schools’ direct accountability to parents. Rather than doubling down on such futile efforts, education reformers should seek to expand true accountability by increasing school choice for more families. The solution to rising costs and subpar results is not central planning but greater control over education on the part of all families, rich and poor.
This is part of a forum on test-based accountability. For alternate takes, see “Why Accountability Matters, and Why It Must Evolve,” by Morgan S. Polikoff, or “If Parents Push for It, Accountability Can Work,” by Kevin Huffman.