Cory Koedel
Department of Economics - 0508
University of California, San Diego
9500 Gilman Drive
La Jolla, CA 92093-0508
Office: Economics 126
Phone: 858-220-3449
Email: ckoedel@ucsd.edu
Curriculum Vitae
Dissertation Abstract

Job Market Paper
Teacher Quality and Educational Production in Secondary School
Using administrative data linking students and teachers at the classroom level, this study evaluates teacher quality and team production in the secondary school educational production function. Although empirical research has struggled to link observable teacher qualifications to student achievement, I show that teacher quality measured by student performance varies significantly and has important effects on educational outcomes. In my test-score analysis, I allow multiple teacher inputs to affect student performance and estimate teacher effects from a within-student, value-added specification to control for selective matching between students and teachers. I identify which teacher inputs affect output in both math and reading and find strong evidence of joint production. I also consider the extent to which teacher quality affects whether students graduate from high school. I use an exogenous set of instrumental variables based on school-level staffing changes from year to year to identify teacher effects and show that students’ graduation decisions are indeed influenced by teacher quality. Furthermore, teacher quality measured by high school completion rates is positively correlated with that measured by test-score performance. The results from my analysis are applicable to incentive design and teacher accountability at the secondary level.

Papers in Submission
Cory Koedel and Julian R. Betts, "Re-Examining the Role of Teacher Quality in the Educational Production Function," 2006.
This study uses data from San Diego elementary schools linking students and teachers at the classroom level to estimate teacher value-added to student test scores. Our findings indicate that variation in teacher quality is an important contributor to student achievement in both math and reading – much more important than has been implied by previous work. We use our estimates of teacher fixed effects to examine whether it is feasible to incorporate performance-based measures of teacher quality into evaluation or merit pay programs. The estimation error inherent in the statistical modeling of teacher fixed effects is quite large, implying that value-added estimates may be ill-suited as stand-alone determinants of accountability. However, there is also a sizeable quality signal in the estimated teacher coefficients, particularly when compared to the current standards by which most teachers are evaluated. Indeed, the observable characteristics and qualifications that generally determine teacher recruitment, retention and salaries are almost entirely unable to predict teacher quality measured by student outcomes.

Other Research
Teacher Quality, the Achievement Gap and Efficiency in Educational Production
The outcome-based teacher quality literature has unambiguously shown that high-quality teachers have important effects on skill acquisition. Thus, it is straightforward to conclude that a shift in the resource of teacher quality toward disadvantaged students will reduce the achievement gap, possibly substantially. However, the efficiency costs of such a shift are less clear. On the one hand, if disadvantaged students are more responsive to differences in teacher quality than advantaged students, such a shift would increase total educational output. On the other, if they are less responsive, the equity benefits associated with this policy prescription would be coupled with an efficiency loss. To evaluate this question, I estimate an alternative distribution of teacher quality that is generally faced by disadvantaged students using various proxies for socioeconomic status. This alternative distribution is estimated based on teachers that teach both advantaged and disadvantaged students. Preliminary results indicate that disadvantaged students are not differentially affected by differences in teacher quality. This implies that the equity gains associated with a shift in teacher quality toward disadvantaged students could be obtained without lowering efficiency.

Evidence of a Positive Effect of Race-Based Matching on Labor Productivity
This study considers the effects of race-based matching on labor market outcomes in a bargaining environment. Using a unique dataset based on survey data that I collected, I observe economic outcomes when agents of matched and mismatched race interact in the market for new and used automobiles. An important attribute of the data is that interacting agents are randomly matched. My findings indicate that racial matching enhances productivity and that its effects are non-trivial. This may help to explain why differential labor market outcomes by race continue to persist, particularly in labor markets in which productivity is determined by success in a bargaining environment.

School Choice and Integration (with Julian R. Betts and Lorien Rice)
We examine the supply and demand for school choice in San Diego and evaluate how each of these components of the “choice market” affects integration. Because of generally high participation rates among non-white and disadvantaged students in San Diego, demand for school choice ultimately serves to promote integration by race, achievement status, and parental education levels. However, this is only the case when free busing is provided. The open enrollment program in San Diego, for which busing is not provided, generally serves to further segregate students across all measured dimensions. On the supply side, participation in all of the choice programs in San Diego is constrained. Supply constraints limit both the integrating and segregating effects of the school choice programs.

The Effects of Reading Interventions on Student Performance (with Julian R. Betts and Andrew C. Zau)
This study uses administrative data from San Diego to evaluate an intervention program targeted at students who have fallen behind in reading. Specifically, students who are identified as being one or more grade levels behind in reading are assigned to additional reading classes in summer school and/or double-length English classes during the school year. Our preliminary results are mixed. For elementary students, these interventions improve student performance on standardized tests. However, for middle and high school students they are less successful and in some cases actually stall student-achievement growth. We do not find any effects on high school graduation rates or on the probability of completing the course requirements necessary for entry into the University of California