Exchange in Negotiations. A Dynamic Model of Exchange Processes, applied to the Amsterdam Intergovernmental Conference 1996

The later changed initial name of the project was "Intergovernmental Negotiations. Comparative Dynamic Simulations". The central question of this project is how individual actors (i. e. governments) with conflicting preferences reach consensus under the unanimity decision rule (i. e. in intergovernmental negotiations) in spite of the possibility to veto. In literature the unanimity decision rule is often seen as costly and ineffective, leading rather to deadlock than to substantial reforms. In this project a theoretical model of the convergence process taking place in negotiations by step-wise concessions is provided. Game theoretic approaches are bound to equilibrium analysis, but it has to be questioned if this leads to a realistic representation of negotiations – especially if there are many issues and actors involved. Accordingly, the model developed here explicitly accounts for negotiation dynamics. Further, the importance of political exchange for compromising and the negotiation outcome is explored. The theoretical model is implemented using agent based computer simulation. This method allows for modelling of individual level decisions of potentially bounded rational actors in a changing environment. Furthermore, with the possibility of varying model assumptions, it allows for systematic testing of hypotheses derived from theory. The simulation model extends the exchange model of Stokman and van Oosten, specifying the negotiation as a truly dynamic concession process. Additionally, the risk taking behaviour of the negotiators is taken into account when modelling their concession decisions. The model is empirically tested using a data set on Endgame of the Amsterdam Intergovernmental Conference 1996-97. Very good predictions of the actual negotiation outcome can be provided. Comparison of model predictions using competing explanation models shows that dynamic modelling contributes significantly to prediction accuracy. Moreover, the results give evidence that exchange processes play an important role in the Amsterdam Conference. Political exchange appears as means negotiators employ to improve their individual utility, not to reach consensus. Anyhow, exchange leads to efficient negotiation outcomes. The compromising character of exchange prevents deadlock and results in far-reaching institutional reforms, in spite of the unanimity decision rule.