Downstream ESG and Upstream New Quality Productive Forces: The Moderating Role of China-US Trade Friction
Description
We construct a downstream firm–upstream firm–year panel dataset using Chinese A-share listed companies from 2009 to 2023. Financial and corporate governance data are sourced from the China Stock Market & Accounting Research (CSMAR) database, ESG performance data are obtained from the Huazheng ESG rating system, and Sino-US trade friction data are collected from the China Trade Remedies Information Network. Following Tao et al. (2023), we restrict the sample to observations where both upstream and downstream entities are listed companies. To mitigate reverse causality, we employ a one-year lag structure: downstream ESG data cover 2009–2022, while upstream new quality productive forces data span 2010–2023. We remove observations with only a single cooperation episode during the sample period, observations where upstream and downstream firms belong to the same listed company, and observations with missing key variables. The final sample comprises 348 downstream enterprises, 462 upstream enterprises, and 2,002 downstream–upstream–year observations. The dependent variable, upstream firms’ new quality productive forces (NP), is an entropy-weighted composite index covering three dimensions—new quality laborers, new quality objects of labor, and new quality means of labor—with eleven indicators including R&D personnel ratio, green patent applications, digitalization degree, and AI investment level. The independent variable, downstream ESG performance (ESG_down), is proxied by the annual average of quarterly Huazheng ESG ratings (scored 1 to 9). We examine three mediating channels: stable cooperative relationships (cooperation duration Rel1 and capital occupation Rel2), pressure-induced technological upgrading (a dummy Qua for “Little Giant” enterprise designation), and improved governance mechanisms (social reputation Gov, measured as the natural logarithm of positive media reports plus one). Control variables include upstream firm age (Age), size (Size), revenue growth (Growth), leverage (Lev), board size (Board), ownership concentration (Sc), and state ownership (SOE), as well as downstream firm age (B_Age), growth (B_Growth), concentration (B_Sc), and state ownership (B_SOE). The moderating variable, Sino-US trade friction intensity (Impact), is the one-year lagged natural logarithm of one plus the number of US anti-dumping and countervailing investigations against the downstream firm’s industry; we also control for upstream trade friction intensity (Impact_up) and a COVID-19 dummy (Dis) for 2020–2022.