Environmental regulation, green innovation, and productivity: Crowding-out or reallocation?

Does environmental regulation enhance firm productivity through reallocation toward green innovation, conditional on firms' pollution intensity and productivity levels? We analyze Chinese listed firms (2010–2018) using a Crépon–Duguet–Mairesse recursive framework, a three-equation causal chain that links regulation, innovation, and productivity. We find that a variety of environmental policies raise compliance costs, crowding-out average R&D investment. Non-green innovation yields higher productivity returns in low-pollution firms and exhibits a U-shaped return pattern across the productivity distribution, whereas green innovation's returns follow an inverted-U-shaped profile, peaking at medium-high productivity—where, for high-pollution firms, they exceed those of non-green innovation, boosting TFP. These results provide mechanism-based evidence for the strong Porter hypothesis as a context-dependent reconfiguration of technological and productive efficiency under regulatory stimuli. Policy support for green innovation should be targeted at high-pollution, mid-to-high-productivity firms based on verifiable outcomes, turning the crowding-out of R&D into productive reallocation.