Green Innovation Policies in Complex Landscapes : An Agent-Based Approach

This study develops simulation scenarios to understand the impact of green innovation policies on carbon emissions and output in an oligopolistic manufacturing industry. We introduce an agent-based model of endogenous innovation in technologies with heterogeneous production costs and carbon intensities. The model combines evolutionary competition dynamics characterized by boundedly rational endogenous preferences over technologies with the heuristic search for innovations in a complex landscape with finite innovations. We consider multiple distributions of potential technologies and degrees of complexity involved in the learning process. The baseline scenario is capable of reproducing qualitative stylized facts that characterize manufacturing industries engaged in oligopolistic competition, such as persistent market concentration or sluggish diffusion of innovations. Our study tests various policies, including market mechanisms, regulatory frameworks, and innovation systems, finding a non-linear trade-off between output and emission reductions that can be optimized by combining more than two instruments. However, finding the intensity of each intervention in the policy package is a complex task that requires iterative experimentation. The results highlight the complementarity between different policy instruments in bringing cost-effective carbon transitions. Regulation plays a definitive role in lowering carbon emissions, while price incentives can offset the negative consequences from switching to costlier technologies. Innovation-oriented policies allow for a deeper extent of technological diffusion, reducing market concentration and improving the performance of regulation and market incentives. Our results relate to mission-oriented policies as they underscore the importance of combining instruments of different approaches and acknowledge the complementary roles of public and private institutions