Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
I analyze a sequential bargaining model in which players are optimistic about their bargaining power (measured as the probability of making offers), but learn as they play the game. I show that there exists a uniquely predetermined settlement date, such that in equilibrium the players always reach an agreement at that date, but never reach one before it. Given any discount rate, if the learning is sufficiently slow, the players agree immediately. I show that, for any speed of learning, the agreement is delayed arbitrarily long, provided that the players are sufficiently patient. Therefore, although excessive optimism alone cannot cause delay, it can cause long delays if the players are expected to learn.