| In Chapter 1 I consider three separate explanations for how price affects the usage rate of a purchased product: Screening, signaling, and sunk-cost bias. I propose an experimental design that disentangles the three effects. Furthermore, in order to quantify and compare these effects I introduce a simple structural model and show that the parameters are separately identified by the experimental data. Finally, I implement the proposed experiment and estimate the model by selling vouchers for drinks at a local cafe in Evanston. In Chapter 2 I again use this experimental design to ask whether students who pay more for their education put forth more effort. In general I find that price does not screen for, nor does it cause, greater effort among students. In both chapters, when I am able to detect a causal link between price and utilization, it is based on the informational aspect of price, and not on a behavioral sunk-cost bias.;In chapter 3 I describe the optimal allocation of students to teachers when teachers have heterogeneous ability. Under this allocation, high-ability teachers have more students but not necessarily better outcomes as measured by average student achievement. I argue that precisely because the principal wants to maximize average student gains, it is problematic to evaluate teachers based on this same metric. I document empirically that some teachers persistently have more students than other teachers, which suggests that this optimizing behavior may already take place. For principals that want to achieve the optimal allocation, I propose a pair of similar bonus payments that induce efficient effort without creating incentives for teachers to manipulate enrollment. One mechanism requires the principal to know teacher abilities while the other elicits this information. Finally, I use a simulation to quantify how the optimal allocation (relative to equal class sizes) affects student gains and measures of teacher effectiveness. |
