The Paris Agreement relies on nationally determined contributions to reach its targets and asks countries to increase ambitions over time, leaving open the details of this process. Although overcoming countries’ myopic ‘free-riding’ incentives requires cooperation, the global public good character of mitigation makes forming coalitions difficult. To cooperate, countries may link their carbon markets¹, but is this option beneficial²? Some countries might not participate, not agree to lower caps, or not comply to agreements. While non-compliance might be deterred³, countries can hope that if they don’t participate, others might still form a coalition. When considering only one coalition whose members can leave freely, the literature following the publication of refs ^4,5 finds meagre prospects for effective collaboration⁶. Countries also face incentives to increase emissions when linking their markets without a cap agreement^7,8. Here, we analyse the dynamics of market linkage using a game-theoretic model of far-sighted coalition formation. In contrast to non-dynamic models and dynamic models without far-sightedness^9,10, in our model an efficient global coalition always forms eventually if players are sufficiently far-sighted or caps are coordinated immediately when markets are linked.

《巴黎协定》依靠国家确定的贡献来实现其目标，并要求各国随着时间的推移增加野心，并公开这一过程的细节。尽管要克服各国的近视“搭便车”激励措施需要合作，但减轻全球公共利益的性质却使组建联盟变得困难。为了进行合作，国家可以将其碳市场联系起来¹，但是这种选择是否有益²？一些国家可能不参与，不同意降低上限或不遵守协议。虽然可能会阻止违规³，各国可以希望，如果他们不参加，其他国家可能仍会组成联盟。当只考虑一个成员可以自由离开的联盟时，参考文献^4,5的出版后的文献发现有效合作^6的前景不佳。各国在没有上限协议的情况下联系其市场时，也面临着增加排放的诱因^7,8。在这里，我们使用具有远见的联盟形成的博弈论模型来分析市场联系的动力。与非动态模型和没有远见的动态模型相比[ ^9,10]，在我们的模型中，如果参与者具有足够的远见或在链接市场时立即调整上限，则最终会形成有效的全球联盟。