Working Paper: NBER ID: w29265
Authors: Margherita Comola; Marcel Fafchamps
Abstract: We conduct a laboratory experiment to study a decentralized market where goods are differentiated and evaluations are private. We implement different semi-structured bargaining protocols based on deferred acceptance, and we compare their performance to the benchmark scenario of a sealed-bid auction. We show that bargaining dramatically improves efficiency, mainly to the benefit of players rather than the silent auctioneer. A protocol of unconstrained simultaneous bargaining performs best, doubling the proportion of deals relative to the benchmark. This is because participants seek to reveal information through a gradual bidding-up strategy that favors bargaining environments. Aggregate efficiency nonetheless suffers from the fact that buyers bargain harder than sellers, and that some players over-bargain to appropriate a larger share of the unknown surplus.
Keywords: No keywords provided
JEL Codes: C78; C91; D47; D82
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| type of bargaining protocol (C78) | market efficiency (G14) |
| semistructured bargaining (J52) | market efficiency (G14) |
| unconstrained simultaneous bargaining protocol (C79) | market efficiency (G14) |
| buyers bargaining harder than sellers (D44) | aggregate efficiency (E10) |
| overbargaining (C78) | transaction success (L14) |