Accountability pressures apply disproportionately to tested grades and subjects, but the impacts of this disproportionate pressure have not been sufficiently investigated. Using longitudinal administrative data and teacher survey data from a large urban school district, we examine schools’ responses to those pressures in assigning teachers to high-stakes classrooms. We find that high-performing teachers, as measured by their students’ achievement gains in math and reading, are more likely to return to a tested grade-subject combination in the following year. These patterns hold not only when teachers remain in the same school but when they move to another school in the district. The relationship between prior teacher performance and assignment to a tested classroom is stronger in high-performing schools. Consistent with the idea that these patterns arise from school leadership behaving strategically to improve accountability standing, our survey data show that this relationship is also stronger in schools where principals—but not parents or students—have more influence over assignments.

Yet we also find evidence of some potential unintended negative consequences of this strategic assignment behavior. In elementary schools, teachers with low value-added systematically are reassigned to earlier grades; 58% of teachers reassigned from grades 3–5 teach second grade the next year, and 25% teach first grade. Using data from a low-stakes assessment given in the early grades, we find that achievement growth among second graders taught by a teacher reassigned from a higher grade is significantly lower than growth in other classrooms in both reading and math. We also find suggestive evidence of persistence of these effects into tested grades; students in second grade classrooms with reassigned teachers have lower test scores at the end of third grade than other students with similar characteristics. These results are troubling in the context of research on the long-term payoff of investments in students’ experiences in early grades.