The adoption of clean technologies (CTs) plays an important role in reducing carbon dioxide (CO$_2$) emissions. We study CT adoption in a large population of consumers with heterogeneous behavioral tendencies. We model the interaction among the agents as a multi-type mean-field game in which the agents choose between clean and polluting technology based products and may either behave as rationals (trading off price and moral incentives), herding agents (just follow the majority), or lethargic agents exhibiting inertia toward adopting the new technologies. We characterize equilibrium CT adoption levels using the recently introduced notion of $\boldsymbolα$-Rational Nash Equilibrium ($\boldsymbolα$-RNE) and its multi-type extension. We then identify a stable subset using the limits of a stochastic turn-by-turn behavioral dynamics. Our results highlight the role of population composition in determining CT adoption. In particular, widespread adoption requires either a sufficiently small price disadvantage for CTs or the presence of a sufficiently large herding population that can be influenced through social awareness programs. Surprisingly, we could prove that environmental damages do not provide sufficient incentives to increase CT adoption.

翻译：清洁技术（CT）的推广在减少二氧化碳（CO$_2$）排放中起着关键作用。我们研究了一类具有异质性行为倾向的大规模消费者群体中清洁技术的采用问题。将主体间的互动建模为多类型平均场博弈，其中主体在清洁技术与污染技术产品间进行选择，并可能表现为理性者（权衡价格与道德激励）、从众者（仅追随多数）或对新技术表现出惯性的惰性者。我们利用新近提出的$\boldsymbolα$-理性纳什均衡（$\boldsymbolα$-RNE）概念及其多类型扩展，刻画了均衡状态下的清洁技术采用水平。随后，通过随机逐轮行为动力学的极限，识别出一个稳定子集。研究结果揭示了群体构成在决定清洁技术采用中的关键作用。特别是，广泛采用要求清洁技术价格劣势足够小，或存在可通过社会意识项目影响的足够大的从众群体。令人惊讶的是，我们证明了环境损害并不能提供充分的激励来促进清洁技术的采用。