Understanding the impact of trade policy effect uncertainty on firm‐level innovation investment

Abstract Drawing on real options and resource dependence theories, this study examines how firms adjust their innovation investments to address trade policy effect uncertainty (TPEU), a type of firm‐specific, perceived environmental uncertainty capturing managers' difficulty in predicting the impacts of potential policy changes on business operations. To develop a context‐dependent, time‐varying measure of TPEU, we apply bidirectional encoder representations from transformers, an advanced deep learning technique. We analyze the texts of mandatory management discussion and analysis sections of annual reports from 3181 publicly listed Chinese firms. Our sample comprises 22,669 firm‐year observations spanning the years 2007 to 2019. The econometric analyses show that firms experiencing higher TPEU will reduce innovation investments. This effect is stronger for firms facing lower competition, involving more foreign sales, and not owned by the state. These findings provide clarity on previously inconclusive results by showcasing the significant influence of policy effect uncertainty, as opposed to policy state uncertainty, on firms' decisions regarding innovation investments. Additionally, these findings underscore the importance of resource dependence factors as crucial contextual factors in this decision‐making process.