The Roles of Teacher and Peer Quality on Human Capital Accumulation in Primary and Secondary School

Policymakers and researchers are interested in factors that drive the production of student learning for determining policies to increase academic achievement and life chances. In this thesis I focus on two such factors, teacher and peer quality, and explore how distinct measures of these inputs map to student achievement both contemporaneously and in future periods. All studies use student-teacher matched administrative longitudinal education data from the North Carolina Education Research Data Center. Chapter 2 tests whether teacher quality measured by the National Board for Professional Teacher Standards certification (NBPTS) improves student achievement. Using panel methods we run models that estimate the effect of having a NBPTS teacher on academic achievement that account for both permanent and time-varying differences between families, and for selection to classrooms and schools on both fixed and time-varying unobservables. Our results show that students taught by NBPTS teachers have higher math and reading achievement. Stratification by grade suggests that the effect only exists for middle school teachers. Chapter 3 tests whether the effect of NBPTS teachers persists in future periods. To test this, I run models that estimate the effect of having a NBPTS teacher currently on student achievement in a lead year accounting for unobserved fixed and time-varying differences between schools and school cohorts. Results show the certification effect persists for reading and math. Relative to contemporaneous effects, the persistence estimates are slightly smaller for reading and drop by more than half for math. The findings suggest that the mathematics curriculum has less cumulative complementarity than reading. Chapter 4 explores how classroom disruption influences student learning. Little is known regarding how disruptive students influence classroom learning and teacher evaluation. In this study, we show that students with serious behavioral difficulties substantially reduce the academic performance of their peers in 4th and 5th grades.  Since standard value-added models fail to account for these peer effects, we find that some teachers’ value-added is penalized because of the students she is assigned.  Importantly, we show that the assignment of disruptive students to teachers is non-random, so these peer effects do not impact the evaluation of all teachers equally.