Digital finance policy and the challenge of air pollution in China

Air pollution is a major global and Chinese public health and environmental challenge, driven by emissions from industry, transport, and agriculture, with significant impacts on health, ecosystems, and climate. In China, ambient particulate matter (PM2.5, PM10) causes substantial mortality and economic losses, with industrial sources dominating in major cities. China has implemented ambitious policies, including the 2013 National Air Quality Action Plan and a 2018 three-year plan to reduce major pollutants, alongside NDC commitments to peak CO2 before 2030 and achieve carbon neutrality before 2060. Digital finance policy is posited as a tool to enhance transparency, efficiency, and accountability in environmental management, mobilizing funds for renewable energy and pollution control via instruments like green bonds and carbon trading, and enabling behavior change through data-driven platforms. Advantages include robust IT infrastructure and rising digital literacy; limitations include regional disparities in infrastructure, digital/sustainable literacy, and HDI. The research question/hypothesis is that integrating digital finance mechanisms into environmental policy frameworks can improve the efficiency, transparency, and effectiveness of air pollution mitigation. The study contributes by focusing specifically on environmental outcomes of digital finance, identifying factors shaping its air quality impacts, and employing advanced empirical techniques to quantify these effects using annual data (1990–2022).

The literature review synthesizes three strands: (1) Air pollution and sustainability: Studies document severe health effects of air pollution on respiratory and cardiovascular systems and links to climate change. Conceptual work on sustainable development emphasizes balancing growth, equity, and environmental protection; tragedy of the commons illustrates drivers of pollution. (2) Digital finance policy: Evidence shows digital finance expands financial inclusion, entrepreneurship, and growth; in environmental contexts, tools such as green bonds and carbon trading mobilize investment in renewables and efficiency, aiding low-carbon transitions. (3) China’s environmental quality and sustainability: Rapid industrialization and urbanization have led to air, water, and soil pollution and biodiversity loss; policy effectiveness depends on institutional capacity, enforcement, and stakeholder engagement. Green technologies, circular economy, and sustainable urban planning are highlighted as solutions. Identified gaps include the need for integrated, interdisciplinary analyses linking environment, economy, and technology; better understanding of impacts on marginalized groups (rural, ethnic minorities, low-income households); and more empirical evidence on the efficacy and scalability of digital finance for achieving environmental improvements.

The study uses annual data for China spanning 1990–2022. The dependent variable is the Air Quality Index (AQI), where higher values indicate worse air quality. Key explanatory variables represent digital finance: internet banking and ATM banking. Control variables include coal consumption, urban population, inflation rate, and (as discussed in text) automobile production. Variable descriptions and sources include IQAir for AQI; China Statistical Yearbook for internet banking, ATM banking, coal consumption, and urban population; World Bank for inflation. Expected signs: internet and ATM banking are expected to reduce AQI by reducing travel to bank branches and improving operational efficiency; coal consumption, urbanization, and (potentially) inflation are expected to increase AQI through higher emissions and energy use. Econometric procedure: (1) Test order of integration using Augmented Dickey-Fuller (ADF) tests. (2) Conduct bounds cointegration testing to assess long-run relationships among variables. (3) Estimate an Autoregressive Distributed Lag (ARDL) model capturing both short- and long-run dynamics, with an associated Error Correction Term (ECT) for adjustment to equilibrium. Diagnostic tests include Breusch-Godfrey for serial correlation, White for heteroskedasticity, Jarque-Bera for normality, Ramsey RESET for functional form, and stability via CUSUM and CUSUMSQ. Robustness checks re-estimate models using PM2.5 concentration as the dependent variable.

- Stationarity and cointegration: ADF tests indicate variables are stationary at first differences. Bounds testing confirms cointegration with F-statistic = 13.085 exceeding upper critical bounds. - ARDL results (AQI as dependent variable): Internet banking shows significant negative associations with air pollution: approximately −0.43% (short-run) and −0.38% (long-run) change in AQI for a 1% increase in internet banking usage. ATM banking also negatively relates to AQI: roughly −0.13% (short-run) and −0.21% (long-run). Coal consumption has a positive association: +0.23% (short-run) and +0.54% (long-run) changes in AQI per 1% increase in coal use. Urban population and inflation rate are positively associated with AQI (worse air quality). The ECT is negative and significant (e.g., −0.756), indicating convergence to long-run equilibrium. - Diagnostics: Model meets normality, no heteroskedasticity, no serial correlation, correct functional form, and passes CUSUM/CUSUMSQ stability tests (reported statistics include BG 1.600, p=0.015; White 1.231, p=0.642; Jarque-Bera 0.600, p=0.592; Ramsey RESET 1.261, p=0.325). - Robustness (PM2.5 as dependent variable): Results remain qualitatively consistent, with internet banking and ATM banking maintaining negative coefficients, and coal consumption, urban population, and inflation remaining positive; statistical significance varies across short- and long-run estimates.

Findings support the hypothesis that digital finance mitigates air pollution by reducing travel-related emissions and improving financial sector efficiency. Internet banking exerts a stronger pollution-reducing effect than ATM banking, likely due to broader substitution of in-person services and more extensive digitalization of transactions and processes. Positive effects of coal consumption, urbanization, and inflation on AQI confirm the importance of energy transition, sustainable urban planning, and macroeconomic conditions in shaping air quality. The demonstrated cointegration and significant ECT indicate stable long-run relationships, while diagnostics and robustness checks enhance confidence in the empirical results. Policy relevance is high: scaling digital finance can complement environmental regulations by reducing mobility needs and enabling green financing mechanisms, while structural policies are needed to curb coal use and manage urban growth.

The study shows digital finance—particularly internet banking—significantly reduces air pollution in China, while coal consumption, urban population growth, and inflation increase it. Policy recommendations include leveraging digital finance platforms to reduce physical transactions and travel, expanding digitalization across sectors to improve efficiency, applying AI and big data for targeted interventions, and growing green financial instruments to finance renewable energy and efficiency projects. Future research directions include evaluating smart city initiatives, advancing sustainability education to build environmental literacy, and assessing circular economy implementation across provinces to understand environmental benefits, economic opportunities, and governance mechanisms for broader adoption.

The paper notes implementation constraints for digital finance solutions due to disparities in IT infrastructure, digital and sustainability literacy, and varying HDI levels across regions, especially between urban and rural areas. These disparities may limit access to and the effectiveness of digital finance in supporting environmental goals. Additionally, while robust econometric techniques (ADF, bounds testing, ARDL/ECT) are used, the analysis relies on aggregate annual time-series data, which may not capture localized heterogeneity in pollution sources and policy impacts.