Thursday, April 2, 2015
Effective school choice design not only helps school choice work efficiently, it also yields a treasure-trove of usable data about schools. Here's a paper that illustrates both of those features:
Centralized and coordinated school assignment systems are a growing part of recent education reforms. This paper estimates school demand using rank order lists submitted in New York City's high school assignment system launched in Fall 2003 to study the effects of coordinating admissions in a single-offer mechanism based on the deferred acceptance algorithm. In the previous mechanism, students were allowed to rank five choices and admissions offers were not coordinated across schools. While 18% of students obtained multiple first round offers, the mechanism's uncoordinated offers led more than a third of students to be unassigned after the main round and ultimately administratively assigned. Under the new mechanism, there is a 7.2 percentage point increase in matriculation at assigned school and students are assigned to schools that are on average 0.69 miles further from home. Even though students prefer nearby schools, our estimates suggest substantial heterogeneity in willingness to travel for school. The new mechanism generates higher utility on average and across numerous subgroups of students compared to either a neighborhood school alternative or the previous uncoordinated mechanism. Across a range of estimates, we find that the average student's welfare gain from coordinating assignment is substantially more than the disutility from increased travel. These gains are significantly larger than those from relaxing mechanism design constraints within the coordinated system. Preference heterogeneity implies that choice is far from zero-sum and coordinating admissions offers across schools increases allocative efficiency.