For a mechanism to be highly impactful, it must have both good game-theoretic properties and computational efficiency. A poster child in this respect is the Gale-Shapley work (1962) on stable matching. This talk concerns cardinal-utility matching markets, for which the most prominent mechanism was due to Hylland and Zeckhauser (1979); this pricing-based mechanism has all the game-theoretic properties one could ask for. However, recent work has established it to be computationally intractable in theory and practice.
This talk will summarize several papers which:
a. Established intractability.
b. Proposed a replacement mechanism based on Nash bargaining.
c. Established its game-theoretic and computational properties and gave evidence that there may not be a better alternative.