Examining the Impact of a Teacher-Leader Program

• In Event: The Intersection of K-12 School Accountability and Human Capital

Abstract

Starting in 2017-18, North Carolina (NC) began a new teacher-leader program, called Advanced Teaching Roles or ATR. ATR was designed to give teachers leadership responsibilities in their school for increased pay, without having to leave the classroom. ATR teacher-leaders could work in one of two roles: (1) adult leadership teachers support a team of colleagues through activities such as coaching, co-teaching, or team planning and (2) classroom excellence teachers take on additional students in their classrooms. Overall, the goal of ATR is to develop a career ladder that gives effective teachers opportunities to expand their reach, support colleagues, and ultimately improve student achievement in their schools. 

To estimate the impact of ATR, we use longitudinal administrative data collected by the NC Department of Public Instruction. These longitudinal data capture student, educator, and school level data for all NC public schools in each year between 2009-10 and 2022-23. Our 14-year panel captures 8 years before most schools began implementing ATR (2009-10 through 2016-17), two years after ATR began statewide but before the COVID-19 pandemic began (2017-18 through 2018-19), and four years of data when schools were implementing ATR after the pandemic began (2019-20 through 2022-23).

Leveraging these data, we use difference-in-difference (DID) models that compare a pre-post difference in outcomes for ATR schools minus a corresponding pre-post difference for comparison schools that never implement ATR. This DID approach allows us to account for outcomes before and after schools begin implementing ATR. The core intuition behind this DID approach is that the pre-post difference in the ATR and comparison schools will not be different unless implementing ATR has an effect that changes the student outcomes in ATR schools. Our DID models address the staggered nature of when schools began implementing ATR (Callaway & Sant’Anna, 2020). This staggered DID approach accounts for any bias that could occur if the effect of ATR differs as different schools implement the model across time (Goodman-Bacon, 2018). We estimate separate DID models on two main categories types of outcomes: teacher turnover and student test scores.

We find that ATR had positive effects on student test scores in math (0.07 SD) and marginally significant effects in reading (0.03 SD, p < .10). However, we find no significant effects on teacher turnover in ATR schools, relative to a demographically similar group of comparison schools. These results shed light on a novel teacher-leader policy and extends existing scholarship on how teacher career ladders may affect the teacher workforce and student achievement. 

Authors

• Lam D Pham, North Carolina State University
  Presenting Author

• Sora Yoo, North Carolina State University
  Non-Presenting Co-Author