본 연구에서는 다자 협상이 양자 협상과 달리 타결이 어려운 점을 살펴보고 가능한 해결전략을 시도해보았다. 양자협상의 경우는 상호 이해구조가 합치하면 통나무굴리기식 방법으로 통합적 해결이 가능하지만, 다자 협상에서는 협상인원의 증가와 함께 이해구조의 복합성이 증가하고 따라서 양자협상에서 쓰이는 방법으로는 합의를 보기 어렵고 오히려 중요한 것을 포기하는 역통나무굴리기식 방법을 사용함으로써 좋은 결과를 얻는 경우가 있다. 하지만 이 방법은 이익을 추구하는 개인에게는 반직관적이며 따라서 발견되고 사용되기 어려운 점이 있다. 그렇다면 어떤 조건하에서 사람들은 자신의 이익을 양보하고 전체 협상결과를 최적으로 이끌 수 있을까. 이를 알아보기 위해 본 연구에서는 세 차례의 연구조사에서 게임 이론을 이용한 다자간 협상의 과제를 통해 해결 여부와 이에 영향을 주는 요소를 분석하였다.

In this study, the method was sought that all the parties of the multilateral negotiation in interest conflict can win together without anyone to take advantage or handicap. The method that used very well in the negotiations is logrolling method that every party in negotiation sticks their most important issues and concedes the less important one to the another party. This method works best, because it gives better outcomes than general bargaining, and is known as a very superior method, is widely used in the actual negotiation also. However, logrolling was being invested and used quite in bilateral relations, but whether it can be applied to multilateral negotiations has not been studied well yet, except for some works (e.g. Binder, 2005, Binder & Diehl, 2008). In this study was complemented Binders research and examined the issues through experimental research on the multilateral negotiations and invested under what conditions it can be conduce to optimal resolution. Results of three rounds of the studies of the experiment are as follows: 1. Difficulties of multilateral negotiation grows with the increase the number of parties to negotiate and the increasing complexity of the conflict structure, the negotiation results are also suboptimal. 2. The more important than the number of negotiators matters is the complexity of the interest structure. While the main point of logrolling, to yield less important interest to win the more important interest, is relatively easy to understand and also actually well used, the reverse logrolling methods, to yield the most important interest for the benefits of all parties is difficult to understand, not acceptable, and difficult to practice. Information change and accountability for the negotiation`s success enhance the negotiation behavior, but not optimal. 3. It has been tried to reach the optimal results by the ingroup-leader and the mediation by third person. Mediator ist more effective than ingroup-leader of negotiators. Therefore, for the enhancing of the use of reverse logrolling methods it is necessary a variety of ways to try.