Cognition and framing in sequential bargaining for gains and losses

摘要: Noncooperative game-theoretic models of sequential bargaining give an underpinning to cooperative solution concepts derived from axioms, and have proved useful in applications (see Osborne and Rubinstein 1990). But experimental studies with discounting have generally found systematic deviations between the offers people make perfect equilibrium backward induction (e.g., Ochs and Roth 1989). We extended this experimental literature two ways. First, we used a novel software system record information subjects looked at while they bargained. Measuring patterns search helped us draw inferences about how think, testing as directly as possible whether use compute offers. Second, we compared over gains that shrink time (because of discounting) equivalent losses expand over time. In games studied, players bargain by making finite number of alternating offers. A unique subgame-perfect equilibrium can be computed by induction. The begins last period and works forward. Our experiments three-round game pie of $5.00 50-percent discount factor (so shrinks $2.50 and $1.25 second third rounds). In perfect first player player $1.25 keeps $3.75.