Uncertainty about admission chances and constraints on the number of application choices are two integral aspects of most coordinated school assignment systems. And yet, there is sparse empirical evidence on the behavioral and welfare consequences of these features. This paper builds and estimates a model of optimal portfolio choice to quantify the effects of constrained choice and uncertainty. Using the Marginal Improvement Algorithm proposed by Chade and Smith (2006), our strategy allows us to handle cases with large numbers of students and schools and to recover underlying structural parameters. An application based on administrative data from Ghana finds that total welfare is a concave function of the number of ranked choices – expanding the number of choices from 1 to 2 increases total welfare by 26% and reduces the number of unassigned students by approximately 18%. Allowing students to submit an unrestricted number of choices generates three times as much total welfare as allowing for a single choice.
Ajayi, K.; Sidibe, M. An Empirical Analysis of School Choice Under Uncertainty. (2015) 23 pp.