Association for Public Policy Analysis & Management

The role of principals in the evaluation and support of teachers is receiving renewed attention due to the recent spread of multiple measure teacher evaluation systems and the growing awareness of how complex it is to implement such programs. In these systems, principals create observation-based measures of teacher performance and use the information derived from these observations to improve teacher practice. Recent literature suggests that teacher evaluation systems incorporating classroom observations, value added measures of teacher performance (VAMs), and other performance measures improve teacher effectiveness and student achievement. However, there is little evidence about the role of principals in these evaluation systems and their impacts on teacher quality.

In this paper we examine relationships between principal and teacher quality in the context of the Los Angeles Unified School District’s Educator Development and Support System for Teachers (EDST). Designed to improve teachers’ instructional practice, the EDST involves bi-annual principal observations of classroom instruction in which teachers are rated on their instructional practice using Charlotte Danielson’s Teaching and Learning Framework. These classroom observations, along with other assessments of teachers’ contributions to professional practice (e.g., student surveys), provide data on teacher performance which, in turn, should inform how principals manage their teacher workforce (e.g., recommending professional development). Since teacher participation in the EDST should, in theory, increase the quality of interactions between principals and teachers, we use this setting to examine the following research questions: 1) What is the relationship between principal and teacher quality?; 2) To what extent is this relationship enhanced through teacher participation in the EDST?; and 3) To what extent is this relationship moderated by school, principal, and teacher characteristics?

We answer these questions using administrative data from 2013-14 to 2015-16 with information on teacher, principal, and school characteristics, principals’ certification scores for observing teacher practice, and measures of teacher quality such as VAMs and indicators of teachers’ EDST performance. We combine this panel with principal survey data (62% and 59% response rates in 2013-14 and 2014-15) with measures for principal instructional leadership, perceptions of EDST, and their use of EDST data to inform teacher staffing and training. We run linear regressions with teacher fixed effects to examine whether principal quality (measured using survey data, certification scores, and prior fixed effects estimates of principals’ impact on student learning) is related to teacher performance in terms of VAMs among all teachers and specifically among EDST participants. We also examine relationships between principal quality and EDST teacher participants’ growth in observation ratings between observation cycles within each year and across years. Finally, we examine whether the above relationships are affected by a variety of school, principal and teacher characteristics such as school size and demographics, and teacher and principal years of experience.

Preliminary evidence shows that higher quality principals improve the performance of all teachers in the district, and that these relationships are even more robust among EDST-participating teachers. This suggests that the EDST process allows high quality principals to increase their effectiveness through accessing rich data on teacher performance.