Markovovské správanie a efektívne bilaterálne rokovanie

The paper analyzes a hypothetical negotiation between two players (e.g., two countries) about levels of their activities that may be harmful to one another. Negotiation takes place in noncontinuous time with no time constraints, with alternating offers from both players. Until the players reach an agreement, they can freely choose the levels of their activities. Once an agreement is reached, however, it is binding. Such a negotiation process is modeled on a non-cooperative extensive- form game. My analysis focuses on Markov strategies, which proscribe tactics that do not affect current or future payoffs. In each Markov-perfect equilibrium (a subgame-perfect equilibrium in Markov strategies), the players reach an agreement on Pareto-efficient levels of their activities in the first round of negotiation.