Réseau universitaire international de Genève
Geneva International Academic Network

Français | English
Homepage > Research > Outputs > Description

Learning in Multilateral Trade Negotiations: Some Results from Simulation for Developing Countries


Mr Cosimo Beverelli , Pompeu Fabra University, Spain .

Research Project

Developing Countries and the Trade Negotiation Process

> see the project description


Domestic Feasibility Sets - TNSE - Trade Negotiation Simulation Exercise - Two-level Game Metaphor - Win-sets - WTO - WTO Training Institute


Whereas there are still many gaps in our understanding of negotiation processes, one well-established pattern is that the way negotiators communicate and interpret information does matter, and co-determines both the process and outcome of bargaining. In the study of international negotiations, authors have particularly focused on how actors behave when they have, at the beginning of the process, limited information regarding the resources and resolve of negotiating partners as well as on their domestic context. A large body of literature has explored such situations with the use of formal game-theoretic analytical tools, making use of so-called incomplete information models to derive consequences of limited information on actors’ strategies. Incomplete information about domestic politics has been a particular focus of study, especially since the advent of the two-level game metaphor (Putnam 1988), which gave a new start to an older research tradition on domestic politics and international negotiations (Schelling 1960, Iklé 1964). Authors have paid particular attention to the influence of domestic feasibility sets, win-sets in Putnam’s words, on international negotiations, starting from the assumption that negotiators can never, or at least very rarely, know exactly the domestic constraints of their counterparts, and may also have difficulty with their own constraints. While that recent formal work has helped to better determine the conditions of applicability of Schelling’s famous strength in (domestic) weakness, it has remained problematic, to say the least, for many analysts, observers and practitioners of international negotiations. We see two main reasons for this. A first one is linked to what Allan and Dupont (1999) call the lack of robustness of most game-theoretic models that use advanced solution concepts (relying on refined criteria for unexpected behavior) for reducing indeterminacy in complex settings. Particularly problematic are the rationality demands placed upon negotiators who are assumed to be very sophisticated maximizers, carefully updating any piece and bit of information that they get. There is a kind of sophistication to the square. Actors are assumed to form sophisticated beliefs about the state of the world and to change them in sophisticated fashion. A second reason is that game-theoretic models have almost uniquely looked at bilateral bargaining processes, mostly due to the technical limits, or deficit, of game theory for situations of multiple actors.


Learning in Multilateral Trade Negotiations: Some Results from Simulation for Developing Countries
English | [166 ko] > download