A neural network-based measurement of corporate environmental attention and its impact on green open innovation: Evidence from heavily polluting listed companies in China

Increasingly prominent global climate and environmental issues have raised concerns about the living environment and the scrutiny of production and innovation in heavily polluting industries. Environmental concerns can evolve into environmental regulation in the legal sense, as well as environmental attention in informal rules. Although there has been a lot of research on environmental regulation and green open innovation (proportion of collaborative filings in green patents), there is still a lack of identification of the impact of the role of environmental attention, especially in heavily polluting industries that are critical to the economy. Based on this background, this paper uses the relevant data of listed companies in China's heavily polluting industries as a research sample. We use neural networks (Word2Vec) to extract indicators of environmental attention from the text of listed companies' annual reports and measure their impact on green open innovation using multidimensional fixed effects regression. Our research shows that the environmental attention of listed companies in highly polluting industries will significantly reduce their ability to engage in green open innovation. Ultimately, our study will help managers in heavily polluting industries rebalance the economic benefits and environmental transformation of their companies.