Publications

 

CTN Newsletter n.12 - September 2009

Index

 

Editorial

by Hubert Kempf, Centre d'Économie de la Sorbonne

Dear All,

The programme committee of the 2009 Public Economic Theory Meeting, hosted last 18-19 June by the National University of Galway, in Ireland, have offered to include a special session on coalition theory, under the responsibility of CTN. This generous offer increases the visibility both of CTN and highlights the vitality of academic research on coalitions and its usefulness for economic theory and applications. So we are all very grateful to the members of the Programme Committee.

This session was chaired by Vincent Vannetelbosch, from CORE. Three papers have been presented by Jose Semperer-Monerris (U. of Valencia), Jean-François Caulier (Facultés Universitaires Saint-Louis) and Vincent himself. It attracted a large audience.

In addition there were a session on "Clubs and coalitions" and five sessions on networks ("Networks", "Networks dynamics and incomplete information", "Networks and markets", "Networks and financial stability", "Networks and the diffusion of information and behaviour"). Several papers on coalitions were presented in various topical sessions.

Finally, Myrna Wooders made a keynote speech on clubs (entitled "Small groups in large economies: Past, present and future") at the closing of the conference.

This flurry of research confirms the interest of academics for coalitions and networks, and more generally the groupings of individuals. This gives hope that the next CTN meeting, in June 2010 and hosted by GREQAM, in Marseille, will attract as many good and challenging submissions as in the past.

Hubert Kempf

 

CTN Announcements

International Workshop on Fairness and the Commons: Socio-Economic Strategies and Resource Dynamics
Venice, Italy, October 19-20, 2009
http://www.iccgov.org/events-1_CONF_2009-03.htm

Sustainably managing the local and the global commons requires not only an understanding of the environmental factors that affect them, but also a knowledge of the interactions and feedback cycles that operate between such resource dynamics and the socio-economic dynamics attributable to human intervention. This, in turn, calls for an investigation of the behavioural drivers behind human action.
The workshop aims to bring a multidisciplinary approach to the environmental challenges inherent in the provision and utilization of the services originating from common-pool resources. By establishing bridges between the socio-economic, the ecological and the behavioural traditions, the goal is to find new insights into the mechanisms that can promote and sustain cooperation among the end-users of the commons.

The workshop will bring together a broad audience of selected international researchers from fields ranging from theoretical biology and economics, to behavioural and computational social science. Such diversity of backgrounds will promote the exchange of the latest research and policy proposals among the participants, as well as provide an opportunity to embark in subsequent common efforts.
The workshop is structured around three main sessions.

On the first morning, the emphasis will be placed on establishing the role of fairness motives and spatial patterns in affecting the conservation and utilization of the commons.
The afternoon session will provide an opportunity to explore the resilience of coupled systems, that is of systems characterised by interactions between the environmental and the socio-economic dynamics, to endogenous (i.e. behavioural) as well as exogenous (i.e. climate change) pressures exacerbating the resource scarcity.

The second day's session aims at drawing some conclusions and identifying research directions. Particular attention will be given to the integration of behavioural experiments and environmental investigations, leading to a general discussion aimed at setting a common future research agenda.

Organisers: ICCG, FEEM, CMCC and PEI
Workshop agenda: http://www.iccgov.org/events-1_CONF_2009-03_program.htm
Secretariat : ughetta.molin@feem.it

15th Coalition Theory Network Workshop - Announcement
Marseilles, France, June 17-18, 2010
http://www.feem-web.it/ctn/events/10_Marseilles/ctn15i.htm

The GREQAM CTN group (University of Mediterranée, University of Provence, CNRS and EHESS) will be organizing the 15th Coalition Theory Network Workshop in Marseilles, France, on June 17-18, 2010. This workshop will present the state of the art of theoretical and empirical aspects in coalition/network formation and matching.

Keynote speakers: T.B.A.

Local organizing committee (GREQAM): Sebastian Bervoets and Frédéric Deroïan.

Papers presented in the Workshop are eligible for publication in the FEEM CTN Working Paper series.

Call for papers and further information about the workshop will soon be available on this website.

Note that the Journées Louis-André Gerard Varet will be held in the same week in Marseilles.


JPET Special Issue on CTN Topics
http://www.feem-web.it/ctn/events/09_Maastricht/ctn14j.htm

The special issue of Journal of Public Economic Theory, centered on the 2009 Maastricht meeting of the CTN, is in progress. The official deadline is passed, but if you are interested in submitting a paper that you believe would be a good fit, please contact the editors, Francis Bloch (Francis.Bloch@polytechnique.edu), Jean-Jacques Herings (p.herings@maastrichtuniversity.nl), or Bettina Klaus (b.klaus@algec.unimaas.nl), who will decide whether the paper might still find a place in the issue.


 

Report: Workshop on "Coalitions for Climate Cooperation. A Game-Theoretic Analysis of Post 2012 Climate Policy"

Giovanni Ursino, Ca' Foscari University of Venice

In June 15 and 16 it has been held on the Island of San Giorgio, Venice, a workshop on Climate Change sponsored by FEEM, ICCG, CMCC and ETH Zürich and organized by FEEM and ETH Zürich. The event, "Coalitions for Climate Cooperation. A Game-Theoretic Analysis of Post 2012 Climate Policy",was attended by several among the major researchers on game theoretic modeling of climate change policy agreements. Several papers were thoroughly discussed by an active audience. The contributions were mainly theoretical while simulation studies were also presented.
The following report mentions a subset of papers presented as it focuses on the issues raised at the workshop which more closely relate to the interests of CTN. It also summarizes the current challenges and hot topics of the climate change research agenda with a special focus on its coalition theory implications. The report is organized by themes with a concluding summary.

Tariffs & Transfers
An major focus of the workshop has been on the implementation of transfer schemes and tariffs to sustain International Environmental Agreements (IEAs). As stressed by de Zeeuw research in this direction should be encouraged because of its relevant policy implications.

Fuentes-Albero and Rubio show that, once transfer schemes are allowed in a standard model with Partial Agreement Nash Equilibrium and transfers cannot be reneged on, then cooperation is possible and the Shapley value is the outcome. More importantly, Rubio and Fuentes-Albero show that the ability of buying cooperation, while it is not affected by heterogeneity in abatement costs, it is actually enhanced by heterogeneity in environmental damages. Their result establishes a positive relationship between the scope of cooperation and the degree of asymmetry independently of the effect that a change in the heterogeneity has on the gains to full cooperation.

Edenhofer and Lessmann investigate whether tariffs -or trade sanctions- are a good instrument for fostering IEAs. They present a trade model á-la Armington -i.e. with national product differentiation- where coalition members are allowed to impose tariffs on imports from non-coalition members. The model is then solved numerically using a version of Negishi's algorithm. It is found that indeed trade sanctions rise participation whenever the welfare increase of coalition members outweighs the cost of constraining trade. This is true in the model for a wide range of parameters and also for nearly substitute goods as long as the elasticity of substitution is conveniently adjusted.


Simulations
An important spin was given to the debate by several contributions highlighting the relevance of simulations to draw reliable policy implications out of theoretical models. Both Henry Tulkens and Joahan Eyckmans stressed the importance of simulations in their discussions and called for more cooperation in sharing payoff matrices (Eyckmans) and programming codes (Tulkens).

Bosetti et al. numerically simulate scenarios based on the WITCH model. They are interested in checking profitability and stability of Potentially Effective Coalitions -a subset of the possible coalitions identified through assumptions on their political feasibility. They analyze the role of different types of externalities and leakage effects. Their major findings are: sizable mitigation is economically rationale provided one assumes high damages and low discounting; for large coalitions, namely the grand coalition, the surplus is not large enough to compensate all free riders from abstaining; the largest stable coalition is the one composed by OECD countries.

Blanford and Rutherford use the MERGE model to assess the viability of self enforcing abatement plans. They first calibrate damages -included non market ones- in the model, then develop a strategic non cooperative toy model with sovereign states deciding abatement levels, and finally numerically simulate scenarios in MERGE. They find that, without transfers, the abatement is limited to large nations while with transfers largest nations can bring about abatement up to its marginal willingness to pay in all countries.

Finally, Dellink investigates what are the main drivers of stability. He first develops a dynamic cartel formation game with open membership, transfers -i.e. emission permits- and grandfathering. He then simulate it in the SATCO model under different transfer schemes. He finds that: as expected, free-rider incentives are the major problem for stability; transfer schemes incorporating expectations on growth and emissions, -dynamic grandfathering- are essential to get effective coalitions; regional damage shares are a main driver for membership as compared to mitigation costs and global damage levels.


R&D & Dynamics
Two papers focused on R&D expenditures, which, contrarily to previous findings, seem to be important means to improve coalition stability.

Hoel and de Zeeuw develop a very stylized one shot game with no uncertainty where the cost of adopting a green technology is downward convex in the global level of R&D expenditures and investment cost is linear. They find that non-cooperative behavior may lead to full adoption with a sufficiently higher than first-best level of the public good R&D, while the cooperative solution improves on the non-cooperative payoffs by limiting overinvestment.

Harstad analyzes dynamic private provision of public goods when countries can invest in cost-reducing technologies and sign incomplete contracts. The unique (Markov Perfect) equilibrium leads to a dynamic common pool problem where short-term agreements on contribution levels make everyone worse off since all agents invest less when they anticipate future negotiations. Long-term agreements raise investments. If renegotiation is possible, long-term agreements implement the first best. The results have implications for the optimal design of climate treaties and they hold whether permits are tradable or non-tradable and with or without side payments.


Minimum Participation Rules
Weikard, Wangler and Freytag analyze a cartel game with open membership and heterogeneous countries to study the endogenous choice of MPRs and their role for the success of international agreements. In particular they extend previous studies to the case of heterogeneous countries. In their model, where the threat is full break-up and not one-by-one defections, the outcome is almost full cooperation.


Uncertainty
Finus and Pintassilgo addressed the role of uncertainty and learning for the formation of self-enforcing IEAs. They develop a coalition formation model with zero, partial or full learning where the uncertainty is about the level of gains from cooperation, the distribution across countries or both. Compared to previous literature learning is surely bad only if there is pure uncertainty about the distribution of the gains from cooperation. They show that this problem can be mitigated, fixed or even turned into an asset through an appropriate transfer scheme.


Summary
The main lessons from the two days workshop can be summarized as follows:
1. from a theoretical perspective the concepts of internal and external stability are shortsighted compared to the more compelling concept of core stability which should be preferred (Tulkens)
2. the notion of self-enforcing agreement is not clear as it is sometimes refers to either Internal/External stability or core stability or both; it should be either dropped or used more carefully (Tulkens)
3. simulation studies are useful to concretely apply theoretical models to stringent policy questions and should be encouraged while codes and results should be disseminated more widely (Bedford and Rutherford, Bosetti et al., Dellink, Eyckmans, Tulkens)
4. the idea that transfers are capable of stabilizing otherwise unstable coalitions is an important one and should be extensively studied (Bedford and Rutherford, Dellink, Edenhofer and Lessmann, Eyckmans, Finus and Pintassilgo, Fuentes-Albero and Rubio, Harstad, Tulkens, de Zeeuw)
5. linking IEAs to R&D expenditures seems an important way to improve the chances of forming stable coalitions; more research in this direction is needed to draw neat policy conclusions (Harstad, Hoel and de Zeeuw)
6. more attention should be devoted to dynamic aspects of coalition formation like, for instance, the possibility of renegotiation (Harstad, de Zeeuw)

The full program of the workshop as well as the list of papers and presentations can be found at: http://www.iccgov.org/events-1_CONF_2009-02.htm.

 

Article: Conflict Networks

Jörg Franke, Technische Unversität Dortmund
Tahir Ozturk, Universitat Autónoma de Barcelona

Game theoretical models of violent conflicts and wars are widespread with a recent upswing in the literature (some recent examples are: Esteban and Ray (2008), Beviá and Corchón (2008), and Jackson and Morelli (2007)). Those types of studies are focus on distinctive issues but they are usually based on a similar setup based on two or more conflict parties that are involved into singular or isolated conflicts.

The objective of our paper is the extension of this framework with respect to the last mentioned assumption, i.e., in our setup conflict parties might be engaged into several conflicts with different opponents at the same time. Hence, we consider a structure of interrelated conflicts which also induces local externalities between conflict parties. The relation between conflict structure and conflict intensity and its consequences for peaceful conflict resolution are naturally arising questions that we try to address in our framework.

The interrelated conflict structure is set up in the following way: We assume that if there is a conflictive relation between two agents then they are direct conflict parties, i.e., there is a direct link between them. Hence, conflicts are always bilateral which allows to represent the conflict structure as a fixed and undirected network of conflictive relations among agents. An agent can affect its probability to win a specific bilateral conflict against a direct rival by investing into conflict specific and costly technology. Winning a conflict implies that a fixed amount of resources is transferred from the loosing to the winning agent. Hence, the aggregated value of contested resources is zero and conflict investment is social loss because resources are purely redistributed.

The above specification allows to describe the overall conflict structure as a non-cooperative game where several bilateral conflict games are simultaneously played on a fixed and given network. Note, that in this setup (and contrary to most of the network literature, see Goyal et al. (2008) for a recent exception) the action of a player is link specific because conflict investment is specific for each bilateral conflict in which an agent is involved. As in Ballester et al. (2006), the main objective is then to determine the consequences of network structure on aggregated equilibrium strategy. Hence, for the conflict network game we are especially interested in analyzing the connection between conflict structure and total conflict investment in equilibrium, interpreted here as conflict intensity.

Results
For a general specification of the above mentioned conflict network game it is possible to show that a unique and interior equilibrium exists based on the notion of diagonally strictly concave games introduced by Rosen (1965) and Goodman (1980). Interiority of equilibrium has two important consequences. First, it implies that all conflict parties will invest positive amounts into all the bilateral conflicts in which they are involved. Hence, the conflict investment equilibrium has similarities to a prisoner's dilemma because the socially efficient strategy would be to investment nothing at all. However, the socially efficient strategy can be exploited by the respective rival. As agents do not have commitment power, the strategy of investing nothing into conflict specific technology is not part of an equilibrium strategy. Second, interiority implies that the equilibrium is characterized by the system of first order conditions. This system is in general non-linear, i.e., there do not exist closed form solutions which are necessary to determine conflict intensity. Therefore, more structure is needed with respect to the conflict network. The following types of networks provide a more structured environment that guarantees closed form solutions:

o Regular conflict networks,
o Star-shaped conflict networks, and
o Complete bipartite conflict networks.

Additionally, the considered classes of conflict networks have some importance as a representation of potential conflict structures from an interdisciplinary perspective: We discuss concepts stemming from social anthropology that seem to be closely related to the network classes mentioned above.

Analyzing the three classes separately leads to the intuitive result that conflict intensity in increasing in the number of bilateral conflicts and the density of the conflict network. Based on the concept of eigenvector centrality it is also possible to compare equilibrium results across the three different classes. This comparison can be interpreted from the perspective of peaceful conflict resolution which means in this context ad-hoc deletion of specific links. In fact it can be shown that ad-hoc deletion of links implies a reduction in conflict intensity if this process starts and ends with a conflict network belonging to one of the considered three classes. Hence, for the considered classes of conflict networks peaceful conflict resolution turns out to be beneficial because conflict intensity is reduced.

Additionally, the reformulation based on eigenvector centrality can be used to show the existence of local externalities which tends to reduce conflict intensity if the network is rather asymmetric. The reason is that agents are negatively affected by their centrality (more rivals means more total individual equilibrium investment without resulting in increased winning probability for a specific bilateral conflict) which is beneficial for their direct rivals. This effect can be so important that it even dominates the beneficial effects of peaceful conflict resolution in specific networks: We present an example outside the considered classes of conflict networks where peaceful conflict resolution does in fact imply an increase in conflict intensity. Hence, peaceful conflict resolution might have adverse consequences for conflict intensity if the fact that conflict parties are embedded into a structure of multiple conflicts is neglected.

Although deriving general results for irregular networks (outside the mentioned classes) is a complex issue, the following indirect statement is possible: Conflict parties with high total conflict investment in equilibrium loose in expectation against rivals with lower total conflict investment. The intuition fits to the above mentioned result that being central is detrimental for an agent because in the considered classes of conflict networks total conflict investment of an agent is increasing in its centrality.

Several extensions of the model might be possible and interesting. For instance, using a different probability function might allow to differentiate between conflict and war while allowing at the same time comparative static analysis. Introducing additional sources of heterogeneity (beside the heterogeneity in locations) would make the framework more realistic such that theoretical predictions for an empirical analysis of specific conflict structures might be possible. Lastly, the simple and intuitive specification of the probability function in combination with uniqueness of equilibrium suggests the verification of the theoretical predictions of the model in the controlled environment of a laboratory experiments. Some of the mentioned extensions are already work in progress; comments and suggestions are highly welcomed.

References
Ballester, C., A. Calvó-Armengol, and Y. Zenou (2006). "Who´s Who in Networks. Wanted: The Key Player," Econometrica 74(5): 1403-1417.
Basu, K. (2005). "Racial Conflict and the Malignancy of Identity," Journal of Economic Inequality, 3: 221-241.
Beviá, C., and L.C. Corchón (2008). "Peace Agreements without Commitment," Barcelona Economics Working Paper Series No. 340.
Esteban, J., and D. Ray (2008). "On the Salience of Ethnic Con?ict," American Economic Review 98(5): 2185-2202.
Goodman, J.C. (1980). "A Note on Existence and Uniqueness of Equilibrium Points for Concave N-Person Games," Econometrica 48(1): 251.
Goyal, S., A. Konovalov, and J.L. Moraga-González (2008). "Hybrid R&D," Journal of the European Economic Association 6(6): 1309-1338.
Jackson, M.O., and M. Morelli (2007). "Political Bias and War," American Economic Review 97(4): 1353-1373.
Rosen, J. (1965). "Existence and Uniqueness of Equilibrium Points for Concave N-Person Games", Econometrica 33(3): 520-534.




Visit the CTN Members' seminars web pages

CODE: http://idea.uab.es/code/workshops.htm
CORE: http://www.uclouvain.be/en-43617.html
CES: http://ces.univ-paris1.fr/seminaires_ces.htm
FEEM: http://www.feem.it/Feem/Pub/Seminars/default.htm
Greqam: http://greqam.univ-mrs.fr/seminars.php
Maastricht: http://www.fdewb.unimaas.nl/meteor-seminar-et
Vanderbilt: http://www.vanderbilt.edu/econ/seminars-research.htm
Warwick: http://www2.warwick.ac.uk/fac/soc/economics/forums/

 

About the CTN Newsletter

The CTN Newsletter is prepared with the contribution of all the CTN Partner Institutions. Please send comments and questions to: silvia.bertolin@feem.it.
The next issue will be published in March 2010

Subscribe to the CTN Newsletter