Working Paper: NBER ID: w26394
Authors: Bnaya Dreyfuss; Ori Heffetz; Matthew Rabin
Abstract: Deferred Acceptance (DA), a widely implemented algorithm, is meant to improve allocations: under classical preferences, it induces preference-concordant rankings. However, recent evidence shows that—in both real, large-stakes applications and experiments—participants frequently play seemingly dominated, significantly costly, strategies that avoid small chances of good outcomes. We show theoretically why, with expectations-based loss aversion, this behavior may be partly intentional. Reanalyzing existing experimental data on random serial dictatorship (a restriction of DA), we show that such reference-dependent preferences, with a degree and distribution of loss aversion that explain common levels of risk aversion elsewhere, fit the data better than no-loss-aversion preferences.
Keywords: loss aversion; strategy-proof mechanisms; deferred acceptance algorithm; preference misrepresentation; expectations-based reference-dependent preferences
JEL Codes: B49; D47; D82; D84; D91
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| loss aversion (G41) | intentional misrepresentation of preferences (D91) |
| expectations-based loss aversion (G41) | ranking lower-value options higher (D91) |
| low probability of receiving higher-value options (D81) | misrepresentation of preferences (D91) |
| EBRD model (C51) | deviation from expected value rankings (D81) |
| misrepresentation of preferences (D91) | violation of first-order stochastic dominance (FOSD) (C69) |
| coefficient of loss aversion (1.20) (D11) | better fit of EBRD model (C51) |
| likelihood-ratio test (p<0.0001) (C52) | strong statistical significance of EBRD model (C51) |