Theses Doctoral

Leading from the Classroom: Teacher Leadership in New York City Schools

Sands, Sara

Schools are remarkably resilient institutions, with the core of instruction and the dynamics between teachers, students, and school leaders remaining largely unchanged despite a churn of reforms efforts. To move away from “tinkering” around the edges of the system, reach deeper into classrooms, and mitigate the effects of policy churn, teacher-driven approaches to organizing school improvement have been promoted.

One such approach is teacher leadership. Teacher leadership seeks to position teachers as decision makers in their school communities, so that they might influence their fellow teachers, school leaders, and other stakeholders to improve teaching and learning practices, subsequently improving school quality and, ultimately, student achievement. While teacher leadership initiatives have been popular with school leaders, policymakers, funders, and advocacy organizations since the 1980s, researchers and evaluators have struggled to quantify the impact of teacher leadership, thus leading to questions about the viability of programs to move the proverbial needle on student achievement and other measures. This dissertation explores how the New York City Department of Education (NYCDOE) attempted to impact instruction and school quality, and institute distributed leadership practices through Teacher Career Pathways (TCP), a teacher leadership program.

My investigation examined four questions: 1) Do stakeholders in schools with teacher leaders experience changes to organizational processes and relational dynamics over time?; 2) Do schools participating in TCP show greater improvement in school quality compared to schools that do not participate?; 3) Do TCP-participating schools show greater gains in student achievement scores than schools that do not participate?; and 4) Do schools implementing TCP under centralized oversight from the district observe different outcomes than schools that are entirely self-directed and implemented under decentralized control? To answer these questions, I use quantitative data collected by the NYCDOE, including TCP annual survey results, School Quality Review (SQR) ratings, and student achievement outcomes for ELA and math in grades 3 through 8. I apply descriptive and inferential statistics to examine changes in survey responses and SQR ratings, and difference-in-difference and event study approaches to analyze student achievement.

I find that where school leaders staff teacher leaders into formal roles with defined responsibilities, positional authority, and commensurate salary increases, schools show improvement in the experiences of stakeholders, school quality, and student achievement. In the case of student achievement, this improvement compounds over time, with schools exhibiting increasing gains in each year following the initial introduction of teacher leaders. Schools that have qualified teacher leaders who are never staffed see no changes in any areas examined. Finally, schools that staffed teacher leaders with funding and implementation support from the district experience even more improvement than schools staffing teacher leaders but with no district oversight. These findings offer crucial insight for districts and schools, making the case for increased teacher empowerment and teacher contributions to school-level policymaking, and highlighting the influence centralized governance at the district level can have on the success of initiatives aimed at decentralizing authority within schools.

Geographic Areas


  • thumnail for Sands_columbia_0054D_17659.pdf Sands_columbia_0054D_17659.pdf application/pdf 2.9 MB Download File

More About This Work

Academic Units
Politics and Education
Thesis Advisors
Henig, Jeffrey R.
Ph.D., Columbia University
Published Here
February 8, 2023