How Digital Data Are Turning Factories into Green Heroes?

Chinese companies are increasingly using digital tools – sensors, artificial intelligence and cloud platforms – to make production cleaner. A study of thousands of firms found that these "digital inputs" push companies to cut emissions and become more efficient. The effect is strongest in large plants, publicly owned firms, and those that pollute the most.

A surprising side effect is a reshaping of the market: digital technologies favor the entry of new, high‑efficiency, low‑emission firms while forcing the most polluting ones to exit. This cleans up the sector from within.

The impact also spreads beyond the same industry. When firms adopt digital tools, the green push ripples upstream to suppliers because of a more sustainable demand. In contrast, downstream industries gain little from a "supply‑led" approach. In practice, the demand for greener products drives suppliers to renew capital and invest in technological innovation, creating an upward chain reaction.

These findings show that developing countries can cut industrial pollution and speed up the shift to a greener economy by betting on data and technology.

The study employs a high‑dimensional fixed‑effects model that simultaneously controls for unobserved variables linked to time, sector and region. This approach isolates the pure effect of digital inputs – such as ERP systems, IoT devices and analytics platforms – on CO₂ and other pollutant reductions.

The analysis reveals significant heterogeneity: large firms reap digital economies of scale, while SMEs show a modest improvement, likely due to limited financial resources for tech investments. State‑owned enterprises also display a higher propensity to integrate digital solutions thanks to government incentive policies.

The entry/exit mechanism is quantified using a hazard model that estimates firm survival probability as a function of digital intensity. Results indicate that a 10% rise in digital investment reduces the closure probability of high‑pollution firms by 3%, while increasing the entry probability of high‑efficiency start‑ups by 5%.

Regarding inter‑industry spillovers, a two‑stage regression distinguishes upstream and downstream effects. The upstream effect is mediated by two channels: physical capital renewal – via smart machinery investments – and technological progress, measured through patents and scientific publications. These channels boost the energy productivity of suppliers, lowering overall chain emissions.

Conversely, the downstream effect is weaker because downstream firms depend more on external market factors, such as global demand and environmental regulations, which are not directly driven by suppliers' digital inputs.

Policy implications suggest focusing tax incentives and subsidies on digital projects that enhance energy transparency and environmental data management, especially for high‑pollution firms. Moreover, promoting data‑sharing platforms between suppliers and customers can amplify upstream spillovers, accelerating the green transition of the entire industrial ecosystem.