Policy Portfolios for Green Technology Innovation: An Evolutionary Game Analysis of Heterogeneous Firms

Abstract Enterprise-led green technology innovation (GTI) is pivotal for achieving China’s “dual carbon” goals. This study develops a heterogeneous-agent evolutionary game model to evaluate how a portfolio of market-based policies which contains environmental taxes, carbon emission trading, and subsidies influences the innovation strategies of large and medium-sized enterprises (LMEs) versus small and micro enterprises (SMEs). By incorporating “innovation success probability” as a key parameter, our model analyzes the individual and synergistic effects of these instruments. Our findings are threefold. First, policy effectiveness is highly contingent on firm scale; SMEs require significantly stronger policy support than LMEs to adopt GTI as an evolutionarily stable strategy. Second, a combined policy portfolio is markedly superior to any single instrument due to both synergistic and substitution effects, which allow for flexible regulatory design. Third, external factors such as the “technological dividends” stemming from intelligent supervision, the “institutional dividends” from optimized tax structures, and the “dynamic efficiency dividend” arising from firms’ “learning effects” can amplify policy impact, thereby creating a pathway to Pareto-improving outcomes. Our heterogeneous-agent evolutionary game framework captures firms' differentiated responses to climate policy instruments, revealing when market-based policies succeed or fail in inducing genuine technological transitions, thereby offering microfoundational insights into macro climate outcomes and guiding targeted, differentiated policy design for China’s green transformation.