How School Accountability Systems Create Incentives for Principals to Assign Weaker Teachers to Early Grades

Jason Grissom, Vanderbilt University

Test-based accountability systems – like the ones enshrined in No Child Left Behind and continued under the Every Student Succeeds Act – are intended to give schools strong incentives to improve student achievement. Numerous studies show that schools indeed respond to those incentives – both productively, as when they increase instructional time, and at times in unproductive ways, as when they “teach to the test” or engage in outright cheating.

Less apparent is another important feature of test-based pressure: it affects some classrooms more than others. For example, testing is not required for every grade under the Every Student Succeeds Act. State educational authorities only need to test students in grades three through eight and once during high school. For elementary schools, these requirements mean that student performance in the upper grades “counts” toward evaluation of the school as a whole, while student performance remains unmeasured in kindergarten, first grade, and second grade classrooms. In a recently published study, my colleagues and I show that school principals are responding to those uneven incentives in ways that may have negative long-term consequences.

Incentives to Move Lower-Performing Teachers to Early Grades

To improve student test scores, perhaps the most effective tool in a principal’s toolbox is to put strong teachers in classrooms. A principal who wants her school to get high marks from the state may well try to make sure that her best teachers are placed in grades where testing occurs.

At least two prior studies have found that schools tend to keep teachers whose students perform well on state tests in tested classrooms the next year. Using detailed data on teacher and students in Miami-Dade County Public Schools, the fourth-largest school district in the United States, we replicated those results, showing that teachers whose students did well on year-end tests were much more likely to teach in a tested classroom. Moreover, this tendency was stronger in schools receiving a grade of “F” from the state’s accountability system – schools that therefore faced the most pressure to show improvement. And the tendency was especially strong in schools where teachers told us on surveys that principals were closely involved in assigning students and teachers to particular classrooms. Evidence from our survey suggests that principals are making classroom assignments strategically, to maximize strong test results.

But what happened to teachers whose students didn’t do well? Those teachers were often moved to untested classrooms the next year. In elementary schools, such teachers were usually moved to kindergarten, first-grade, and (especially) second-grade classrooms. What is more, teachers moved at the end of the year from grades three to five to the lower grades were much less effective, on average, than the teachers who stayed in the upper-grade classrooms.

Pressure to put high performers in tested grades works the other way, too. Miami-Dade County Public Schools administers a test not used for official accountability to students in grades kindergarten through second grade. Using scores on that test, we found that kindergarten, first-grade, and second-grade teachers whose students did better were more likely to be moved into state-tested upper-grades classrooms the next year.

Negative Consequences for Crucial Early Learning

For school principals worried about how their schools fare on state accountability metrics, it makes sense to place their strongest teachers in the state- tested classrooms. But placing weaker teachers in early grades has clear downsides.

We examined the achievement gains of first- and second-graders taught by teachers moved from upper-elementary classrooms (teachers we had already determined to be much less effective). Students learned appreciably less under the reassigned teachers; in fact, their gains in math and reading looked similar to those of students taught by brand-new teachers. Teachers are known to be least effective in their first year. In other words, by doing teacher assignments in response to the accountability system, principals undermine learning in the early grades.

Educational research more generally shows that students’ early learning experiences set the foundation for their entire schooling careers. Because this is true, reassigning teachers to deploy those who are most effective to higher grades may have big negative impacts for students in the long run. In fact, the negative effects may even occur immediately, as we discovered when we probed learning gains of students in third and fourth grade (the first years tested by the accountability system). We discovered that students who had been taught in second grade by reassigned teachers did not perform as well on the third- and fourth-grade tests as they would have performed had they been taught by a high-quality second-grade teacher.

What Can Be Done? 

Principals should resist strong incentives to put their most effective teachers in tested classrooms, inadvertently weakening instruction in the all-important earliest school years. The overall goal, obviously, is to get effective teachers into every classroom. But if compromises are necessary, principals should be aware of research showing that weak instruction in the earliest grades may be harmful down the line for their schools and students.

Future reconsiderations of the accountability requirements in state and federal policies should recognize the consequences of uneven testing across grade levels. How secondary schools respond may not matter much; we find, for example, that teachers whose students do not test so well in ninth or tenth grade are simply moved to later grades, where negative impacts on students may be minimal. In elementary schools, however, the structure of the testing regime creates perverse incentives that deserve attention from educational policymakers. By failing to hold schools accountable for student outcomes in early grades, the current testing regime may undermine itself. Policymakers need to make sure that schools and teachers have strong incentives to maximize student learning in the earliest grades, not just when state tests are administered in third grade and beyond. 

Read more in Jason Grissom, Demetra Kalogrides, and Susanna Loeb "Strategic Staffing? How Performance Pressures Affect the Distribution of Teachers within Schools and Resulting Student Achievement,” Stanford Center for Education Policy Analysis (2017).

November 2017