Fair, Efficient and Envy-Free Bargaining : An Experimental Test of the Brams-Taylor Adjusted Winner Mechanism

The Adjusted Winner mechanism for two-player bargaining has been theoretically shown to produce fair (efficient and envy-free) outcomes (Brams and Taylor 1996). We study this claim experimentally in a bilateral bargaining game of incomplete information for two divisible goods using three information conditions. Results indicate that the Adjusted Winner mechanism can be quite effective in achieving both efficient, equitable and envy-free outcomes in an experimental setting but results vary greatly depending on the knowledge that each party has of the preferences of their bargaining partner. Under conditions of common knowledge, when players have competing preferences, their bilateral decisions converge toward efficient outcomes, yet only one-third can be classified as envy-free. In an environment where there is more uncertainty about preferences, pronounced strategic bidding emerges yet the result is twice as many envy-fee outcomes with increased levels of efficiency. Despite the fact that players dramatically misrepresent their true valuation for objects when ordinal preferences are commonly known to be complementary, both efficiency and envy-freeness approach maximum levels