The Tragedy of the Commons : Pricing Social Welfare in Multiagent Systems

Markets o er an e cient and decentralized method of coordinating the actions of autonomous, self-interested agents. However, they fail in the presence of incentive con icts between individual goals and system-wide goals. The actions of one agent often a ect other agents. The tragedy of the commons occurs when selfinterested agents, acting without concern for other agents, have a negative e ect on the overall e ciency of the outcome. The classic economic solution is to impose taxes that penalize such actions. However the center needs complete information on the preferences of all the agents in the system to impose the correct level of tax, and in general agents can gain by revealing incorrect information. We present a simple decentralized mechanism that produces an e cient outcome through payo s between agents, and requires no information in the center. We show an application of the mechanism to a multi-commodity ow

[1]  Jeffrey S. Rosenschein,et al.  Time and the Prisoner's Dilemma , 2007, ICMAS.

[2]  A RAPOPORT,et al.  Three-person non-zero-sum nonnegotiable games. , 2007, Behavioral science.

[3]  M. Morris,et al.  The Design , 1998 .

[4]  Victor R. Lesser,et al.  Issues in Automated Negotiation and Electronic Commerce: Extending the Contract Net Framework , 1997, ICMAS.

[5]  Michael P. Wellman,et al.  Market-oriented programming: some early lessons , 1996 .

[6]  Tad Hogg,et al.  Social Dilemmas in Computational Ecosystems , 1995, IJCAI.

[7]  Jeffrey S. Rosenschein,et al.  Designing Conventions for Automated Negotiation , 1994, AI Mag..

[8]  Edmund H. Durfee,et al.  On the design of an adaptive meeting scheduler , 1994, Proceedings of the Tenth Conference on Artificial Intelligence for Applications.

[9]  H. Varian.  A Solution to the Problem of Externalities When Agents Are Well-Informed , 1994 .

[10]  Michael P. Wellman.  A Market-Oriented Programming Environment and its Application to Distributed Multicommodity Flow Problems , 1993, J. Artif. Intell. Res..

[11]  Tad Hogg,et al.  Spawn: A Distributed Computational Economy , 1992, IEEE Trans. Software Eng..

[12]  J. Rosenschein,et al.  The Clarke Tax as a Consensus Mechanism Among Automated Agents , 1991, AAAI.

[13]  Norman M. Sadeh,et al.  Resource allocation in distributed factory scheduling , 1991, IEEE Expert.

[14]  John. Moore,et al.  Subgame Perfect Implementation , 1988 .

[15]  Robert M. May,et al.  The evolution of cooperation , 1981, Science.

[16]  E. H. Clarke.  Multipart pricing of public goods , 1971 .

[17]  Jacques H. Dreze,et al.  A Tâtonnement Process for Public Goods , 1971 .

[18]  E. Malinvaud,et al.  A Planning Approach to the Public Good Problem , 1971 .

[19]  J. R. RINGROSE,et al.  Solution of a Problem , 1874 .

[20]  Makoto Yokoo,et al.  Cooperative Behavior in an Iterated Game with a Change of the Payoff Value , 1996 .

[21]  E. Durfee,et al.  The Impact of Nested Agent Models in an Information Economy , 1996 .

[22]  Moshe Tennenholtz,et al.  On Social Laws for Arti cial AgentSocieties : OLine Design , 1995 .

[23]  K. Eric Drexler,et al.  Markets and computation: agoric open systems , 1988 .

[24]  L. Hurwicz.  On informationally decentralized systems , 1977 .

[25]  John Roberts,et al.  On the Incentive Problem with Public Goods , 1975 .

[26]  G. Hardin,et al.  The Tragedy of the Commons , 1968, Green Planet Blues.