The Tragedy of the Commons: Can It Be Overcome?
Florian Diekert’s paper “The Tragedy of the Commons from a Game-Theoretic Perspective” begins by examining Hardin’s paper on “The Tragedy of the Commons” and the criticism that it has received over the years. Hardin claimed that freedom in a common space leads to private benefits and public burden, therefore “bring[ing] ruin to all.” However, Diekert concludes that this is not necessarily true and that cooperation, combined with punishment for non-cooperative behavior, allows for the overcoming of the tragedy. The punishment is such that the defectors, who break the cooperation in order to take surplus resources for themselves, eventually loses just as much of the stock that they would have if they had cooperated in the beginning. In such a scenario, any potential defector will have no incentive for doing so, and the others will have no incentive for not enacting the punishment on anyone who does decide to defect.
The author then proceeds to reference three other papers’ punishment schemes through the lens of anthropogenic climate change. This situation is treated as an “inverse” of the tragedy of the commons, as the benefits of avoiding resultant catastrophe is public, but the costs of avoidance are private to countries or businesses. This sets up the basis for the mentioned papers, whose models assume that an individual would want to defect because they can then receive the benefits of avoidance without having to invest in the effort themselves.
The model in the first paper is very similar to the one mentioned above, with one player’s actions having no overall consequence for the total payoff after the punishment has been enacted. The second paper models the situation of pollution: the payoffs from reducing emissions depends on the total pollution that has been accumulated till time of payoff-calculation. The third mentioned paper proposes two models: one in which the players are aware of the threshold for catastrophe and one in which they are unaware of the exact location. In the first case, players have full incentive to coordinate to avoid catastrophe, but in the latter situation, each player has incentive to keep moving as far forward as safely possible and coordination is much more difficult to enforce.
Relevance to SOC 2090:
Diekert’s paper is built on the assumption that the tragedy of the commons is conquerable, and this claim has game-theoretic foundations. First, a payoff matrix was constructed with players who could each choose to cooperate (remain with original stock) or defect (get additional stock, which would then consume resources). When Diekert examined this matrix, he found that everyone defecting was actually not a Nash equilibrium and therefore not a long-term inevitability, as Hardin had assumed.
Another aspect that I found interesting in its connections to course material was the different assumptions made in the three referenced papers about climate change. So far in our discussion of game theory, we have been assuming a model with common knowledge, where players are aware of the risks (payoffs) of their strategies, as well as others’ strategies. In the latter model of the third paper, however, this is not the case, as the location of the threshold (and therefore the risk of any move forward) is unknown to all players, and the paper mentioned that this one fundamental change results in a largely different model. Another of our assumption is slightly altered in the second paper, which models a dynamic situation where individual payoff depends on overall efforts. While this might still fall into our assumption of a “rational model” where each player works to maximize their own payoffs, the paper’s model requires the optimization of overall results in order to increase one’s own payoff, which has not been present in our models. These different papers show not only the applicability of game theory in climate change but also how a situation can be represented in several distinct ways depending on the assumptions made in making the model.
Link: https://doi.org/10.3390/su4081776