Engaging Teachers: Measuring the Impact of Teachers on Student Attendance in Secondary School

administrative panel linking middle and high school teachers to class-period attendance of their students

US secondary school students and teachers

Value-added style decomposition, linking teachers to the class-period attendance of students in their classrooms and estimating teacher contributions to unexcused class absence. Observational - teacher-student assignment is not random - but the design does allow the two value-added dimensions to be compared within the same teachers, which is the point of interest here. Baseline fact reported - US secondary students are absent about three weeks per year.

There is systematic, stable variation across teachers in effectiveness at reducing unexcused class absence, and these differences are as stable as teacher effects on achievement - but teacher effectiveness on attendance correlates only WEAKLY with teacher effectiveness on achievement. The two are largely separate dimensions of teacher quality. High attendance-value-added teachers have a stronger effect on high school completion than high achievement-value-added teachers do. Effects are most salient for low-achieving and low-attendance students.

Medium. Teacher assignment is non-random and student sorting is a live threat, so the value-added magnitudes should not be taken as clean causal estimates. What survives the objection is the correlational structure between the two value-added dimensions, which is the finding used here - if attendance were the pathway through which teachers produce learning, the two should be tightly coupled, and they are not.