Working Paper: CEPR ID: DP10907
Authors: Gabrielle Fack; Julien Grenet; Yinghua He
Abstract: We propose novel approaches and tests for estimating student preferences with data from centralized matching mechanisms, like the Gale-Shapley Deferred Acceptance, when students are strictly ranked by, e.g., test scores. Without requiring truth-telling to be the unique equilibrium, we show that the matching is (asymptotically) stable, or justi?ed-envy-free, implying that every student is matched with her favorite school/college among those she is quali?ed for ex post. Having illustrated the approaches in simulations, we apply them to school choice data from Paris and demonstrate evidence supporting stability but not truth-telling. We discuss when each approach is more appropriate in real-life settings
Keywords: admission criteria; Gale-Shapley deferred acceptance mechanism; school choice; stable matching; student preferences
JEL Codes: C78; D47; D50; D61; I21
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| matching process under stability assumption (C62) | accurate estimation of student preferences (C92) |
| truthtelling assumption (D83) | underestimation of students' valuations of popular schools (D29) |
| stability-based estimators (C51) | better prediction of matching outcomes (C52) |
| statistical tests reject truthtelling assumption (C12) | support for stability (E63) |
| estimation methods based on stability (C51) | account for partial preferences (D11) |