Using student-choice behaviour to estimate tuition elasticity in higher education

Prior research on student response to changes in university prices [tuition] finds that demand is inelastic. We present results, based on separate models for 11 colleges [n = 5606] at a major US university, that run counter to published findings. We discuss fundamental differences between our methods and those used in [most] previous research to explain our findings. Rather than use market level data, we model individual student-choice behaviour and derive market level implications via upward aggregation. Our modelling uses discrete-choice experiments in which choice sets are customised in real-time to reflect each respondent's true consideration set of schools as well as to capture elements of inter-university competition. Published research using market-level data, though appropriate for national policy debates, is not necessarily useful for governance decisions at the university level. We illustrate using tuition elasticities estimated by college and further show that elastic demand can have deleterious effects on the quality of an incoming class even when demand for seats far outstrips supply.