Measuring Human Capital and Teacher Value-Added with Course Grades (JMP)
Abstract: This paper develops theory and methods to recover human capital from course grades while allowing for selection and grading heterogeneity and then estimates the technology of human capital accumulation emphasizing the role of high school teachers. The reliance on standardized test scores implicitly shapes the scope, methods, and study populations of research in ways that may bias empirical results and create blind spots for policy. By identifying and estimating the measurement system of course grades, this paper facilitates education and human capital research without resorting to over-testing or raw grades which can cause bias. I model course grades using a factor model and characterize sufficiency conditions for identification in terms of the connectivity of the network using graph theory. Using this measurement innovation, I then recover teacher effects in non-tested subjects and grades in high school. This advances two frontier topics in the value-added space and improves the representation of teachers and vulnerable populations of students, like high school dropouts, who are central to policy but frequently have missing test scores.