Events
REPORTS ON LATEST RESEARCH
Report on the 5th Coalition Theory Network Workshop
Lydia Kernevez*
*Fondazione Eni Enrico Mattei
The Formation of coalitions is a central question in game theory. It
examines the possibilities of co-operation between agents or regions
and can be applied to various concrete economic fields such as the co-ordination
of trade, taxation, technological, or environmental policies. The Fondazione
Eni Enrico Mattei, Milan and Venice, together with the initial membership
of CORE in Louvain-la-Neuve, GREQAM at the Université la Méditérannée
and the Johnston Center at SMU in Dallas, is part of the Coalition Theory
Network whose aim is to gather researchers working on coalitions theory.
This year the 5th Workshop on Coalition Formation was hosted by the
CODE (Center for the Study of Organisations and Decisions in Economics)
located at the University Autonoma of Barcelona. The aim of this workshop
is to give the latest insights on coalition formation theory with applications
covering a large range of interests such as trade, environmental, fiscal,
research and development policies. It intends to provide analytical
tools and results to be used in the current theoretical framework as
well as support for an interdisciplinary exchange. Compared to the 4th
workshop held in Aix en Provence, this last workshop was characterised
by an impressive number of papers (18 presentations in two days) and
participants. In fact it gathered more than 30 scholars and researchers
coming from research centers located in different countries: Greqam
(France), CODE (Spain), CentER (Netherlands), CORE (Belgium), University
of Warwick (England), CNRS-EHESS-CIRED (France), University of Vrije
(Netherlands), University of Copenhagen (Denmark), University of Paris
I (France), The Pennsylvania State University (USA), Southern Methodist
University (USA).
In the first session, María Paz Espinosa, in collaboration with
Inés Macho-Stadler, studied the equilibrium coalition structures
in a model characterised by moral hazard within coalitions. The authors
concluded that when moral hazard between coalitions is very severe,
no partnership will form. However, when moral hazard is not too severe,
the coalition structure will be either similar or more concentrated
than without moral hazard. Moreover they stressed the fact that the
presence of moral hazard can lead to an insufficient number of coalitions
in the equilibrium. Then, Maria Montero examined coalition formation
in games with random proposers in which coalitions impose externalities
on other players. She showed that an agreement will be reached without
delay provided that any set of coalitions profit from merging. Even
under this strong assumption however, the formation of the grand coalition
is not guaranteed. Thirdly, Francis Bloch developed a formal model of
agenda control to analyse the power of the head of state in influencing
the political coalitions formed. In particular, he related the power
of the agenda setter to the underlying decisive structure among players
using a co-operative sequential game in which coalitions are divided
between losing and winning coalitions. The main result of his analysis
is that the only structure in which the agenda setter can fully manipulate
the outcome of coalitional bargaining is the apex structure, with one
large and small players, where the large player only needs one of the
small players to form a winning coalition. In all other structures,
the power of the agenda setter is limited. During the second session
Paola Conconi's model of international policy co-ordination focused
on the effects of "tie-in negotiation" (a kind of issue linkage)
for the stability of multilateral co-operation over trade and environmental
policies. She concluded that this mechanism can either help or hinder
international co-operation depending on the payoff structure of the
underlying non co-operative game. More precisely tie-in negotiation
appears to be relevant since the value of the inverse elasticity of
marginal damage valuation is large. Richard Tol, during his presentation,
investigated the impact of technological diffusion and learning-by-doing
on the stability of climate coalitions. In this respect, he demonstrated
that a free diffusion of technology leads to an increase in global welfare
but also in the incentives to free ride on a Climate agreement. Imposing
a restriction on technological diffusion can therefore in certain cases
help build larger environmental coalitions. In the end, Jean-Christophe
Péreau presented a model of environmental negotiations in which
he introduced three strategies : co-operation, free-riding and precautionous
unilateral commitment. This last one characterises countries that decide
to commit unilaterally to reduce their emissions in the event of failure
of international bargaining. They demonstrated that when countries can
choose between the three kinds of strategies, the emergence of a small
coalition creates an incentive for all the non signatories to commit
unilaterally and lead to a non- co-ordinated global co-operation.
In the first paper of the 4th session Birgit Grodal introduced a theory
of production in which a specified set of technologies is accessible
to each firm. Labour skills are differentiated and acquired endogenously
by workers. She showed what price systems are required so that competitive
equilibrium exists and core outcomes are equivalent to competitive outcomes.
Hubert Kempf, analysed then the formation and size of a club formed
by wealthy people consuming conspicuous goods. It aims at investigating
the effects of an overtaxation on luxurious goods and established that
such a tax can't help in maximising global welfare. Michel Le Breton
and Shlomo Weber studied next a co-operative game in which they defined
an optimal rule for the distribution of the utilisation costs of a public
good that allows avoiding secession of a part of the Community. They
proved that, in this game, the core exists under certain conditions.
In conclusion Salvador Barberà and Carmen Bevia discussed the
merits of different social choice procedures by proposing allocation
rules that are defined "self selection consistent" (that's
to say allocation consistent and envy free) and that respect the Condorcet
criterion.
Jordi Masso and Flip Klijn initiated the 5th session by introducing
a concept of weak stability in a matching model applied to marriage.
They demonstrated that the set of weakly stable and weakly efficient
matchings coincides with the bargaining set of Zhou (1994). David Perez-Castrillo's
paper present a mechanism, called "bidding mechanism" in which
the players bid for the right to propose a sharing of the surplus generated
by co-operation. They established that the payoff outcome of the subgame
perfect equilibria of this game coincides with the Shapley value of
the Game.
The 6th and last session directly adressed the Climate Change problem
and was based on simulations from the RICE model (Nordhaus and Yang,
1996), a large-scale dynamic model of climate and economies. Firstly,
Johan Eyckmans and Henry Tulkens tested the core property, which is
a necessary condition for full voluntary co-operation, of transfers
scheme designed to sustain full co-operation. They showed that these
transfers yield an allocation in the core of the RICE carbon emission
abatement game. In others words, under the transfers scheme presented
here all coalitions are better off compared to the Nash equilibrium
situation. Secondly, Johan Eyckmans investigated the stability of the
Kyoto Protocol on Emissions of greenhouse gases when dealing with farsighted
players (in the sense that they take into account possible subsequent
deviations by other players). This approach, which considers the players
as no longer myopic, leads to the conclusion that the Kyoto Protocol,
including the possibility of emissions trading among Annex B countries,
is not stable in that farsighted sense. However, simulations show that
a coalition consisting of Annex B countries minus Japan is stable in
a farsighted sense. Finally, Zili Yang's paper dealt with the determination
of feasible and incentive-compatible coalitions to cope with potential
climate change. He examined strategic interactions between regions that
join a coalition and those that do not. He applied the concepts of "Incentives-compatible
coalitions", meaning that a coalition in which the members' payoff
is higher compared to the situation in which they do not belong to the
coalition) and feasible coalitions (that's to say a coalition in which
members' payoff are higher compared to the Nash situation). Numerical
simulations showed that full co-operation is always a definitive Pareto
improvement over the Nash Equilibrium and that it can be a feasible
coalition after a redistribution of transfers. Moreover, they stressed
the importance of partial co-operation since some sub-coalitions prove
to be both feasible and incentive compatible coalitions.
To conclude, we can state that all papers presented have enlightened
new important findings and promising results in the field of coalition
formation and others issues that deserve further research efforts.
