Abstract This study examines whether digital financial inclusion (DFI) is associated with the growth of artificial intelligence (AI) enterprises and through which potential channels this relationship may operate. Using panel data from 271 Chinese cities from 2013 to 2023, we construct a city-level measure of AI enterprise net growth based on firm registration records and AI-related keyword identification. The results show that DFI is positively associated with AI enterprise growth, and this relationship remains robust after addressing endogeneity concerns and conducting multiple robustness checks. The channel analysis provides suggestive evidence consistent with both supply-side and demand-side interpretations, including credit availability, professional labor allocation, and consumption demand. Dynamics decomposition analysis further shows that DFI is associated with more active enterprise entry and exit as well as a lower exit rate, suggesting dynamic adjustment within the AI enterprise ecosystem. Heterogeneity analysis indicates that this positive association is stronger in eastern regions and in cities with higher human capital, stronger innovation capacity, and larger population sizes. This paper provides an integrated “factor supply–market demand” framework for understanding how digital financial inclusion supports emerging technology enterprises.
Abstract Improving investment efficiency is important for sustainable development, yet the mechanisms through which China’s Green Finance Reform and Innovation (GFRI) policy affects corporate investment efficiency remain underexplored. Using data from Chinese listed companies, this paper finds that the GFRI policy significantly improves corporate investment efficiency. The external channel operates through increased analyst attention, while the internal channel works through improved ESG performance. The effect is more pronounced among firms with over-investment, non-state ownership, greater financing constraints, lower accounting transparency, lower environmental information disclosure, and non-heavy-polluting industries. The findings suggest that green finance policy can improve capital allocation when market-based external monitoring is combined with stronger internal governance incentives.
Abstract Understanding how green finance (GF) affects agricultural carbon emissions (ACE) is essential for advancing agricultural green transformation. This study utilizes panel data from 333 prefecture-level cities in China, covering the years 2006–2022, to systematically investigate the nonlinear impacts, transmission pathways, and conditional constraints of GF on ACE. The key findings are: (1) a significant inverted U-shaped relationship between GF and ACE, where GF initially increases emissions but suppresses them after exceeding a turning point (approximately 0.276). This result remains robust across a series of tests including U-shape verification, system GMM, and instrumental variable approaches. (2) Farmer income plays a nonlinear mediating role in this relationship. Specifically, GF affects farmer income in a U-shaped manner, while farmer income is negatively correlated with ACE, thereby transmitting the inverted U-shaped relationship between GF and ACE. (3) The emission reduction effect of GF is limited by threshold effects related to rural human capital (single threshold) and digitalization level (double threshold). Exceeding these thresholds results in a stepwise enhancement of GF’s ability to suppress ACE. (4) A heterogeneity analysis reveals that the emission reduction effect is more pronounced in major grain producing areas once the turning point is surpassed, while it is weaker in national green finance reform and innovation pilot zones, indicating a “policy ineffectiveness” phenomenon. Based on these findings, several policy recommendations are proposed, including phased guidance for GF, region-specific strategies, the development of a synergistic “finance-income-emission reduction” promotion mechanism, and coordinated enhancements of human capital and digital infrastructure.
Abstract In the gig economy context, algorithmic control has become an important tool for platform governance, but its impact on the negative emotions of frontline workers still lacks systematic empirical testing. Grounded in stress appraisal theory, this study investigates a moderated mediation model, examining how algorithmic control influences riders’ negative emotions via challenge-hindrance appraisals and the role of work gamification in this process. Data from a two-wave survey of 401 food delivery riders were analyzed using structural equation modeling and the Bootstrap method. The results indicate that algorithmic control not only directly exacerbates riders’ negative emotions but also exerts an indirect effect through heightened hindrance stress; conversely, the mediating role of challenge stress is non-significant. Notably, rather than mitigating the negative impact of algorithmic control, work gamification showed a positive but non-significant moderating trend ( p = 0.122), descriptively strengthening the relationship between algorithmic control and hindrance stress. This study reveals the dual-path mechanism by which algorithmic control affects riders’ emotions through empirical data, pointing out that work gamification may have the double-edged sword characteristic of “both incentives and constraints,” and providing data support and practical insights for optimizing algorithmic governance and promoting human-centered system design on gig platforms.
Abstract In the platform economy, Chinese intangible cultural heritage (ICH) inheritors are transitioning from cultural custodians to digital entrepreneurs, yet the mechanisms linking digital skills to economic success remain unclear. Integrating human capital and identity theories, this study proposes a chain mediation model examining how digital human capital is associated with economic performance through identity and communicative behavior. Using an explanatory sequential mixed-method design, we analyze survey data from 317 ICH inheritors in Zhejiang Province (collected during October–November 2025), 15 in-depth interviews, and 20 public cases. Quantitative results suggest that digital human capital is positively associated with economic performance ( β = 0.78, p < 0.001), partially mediated by digital entrepreneurial identity and livestreaming e-commerce adoption depth. Age moderates these paths: for older inheritors, digital skills more strongly influence identity formation; for younger inheritors, identity more strongly drives adoption depth. Qualitative analysis identifies four identity transition narratives and triple barriers – subjective, technical, and institutional. This study extends human capital theory by suggesting that identity-mediated behavioral pathways may serve as important mechanisms through which digital skills are associated with economic returns, offering policy implications for empowering cultural producers in the digital economy.
Abstract Exploring the impact of the Carbon Emission Trading Pilot Policy (CETPP) on firm-level digital innovation, this study analyzes a dataset encompassing 25,371 Chinese A-share listed firms from 2007 to 2022 using a difference-in-differences approach. This study reveals that CETPP significantly enhances digital innovation capabilities within firms, with effects particularly strong in capital-intensive, technology-intensive, high-tech, manufacturing, and non-SOE. This research highlights two main mechanisms through which CETPP fosters digital innovation: first, by alleviating financing constraints, and second, by enhancing the visibility of firms within media outlets, which in turn attracts further investment and public support. These dynamics are shown to be robust across multiple tests addressing potential endogeneity concerns. The results not only underscore CETPP’s role in advancing environmental management but also illuminate its broader economic implications by facilitating a synergy between environmental sustainability and technological advancement. This study contributes novel insights into the policy’s effectiveness, offering a robust empirical foundation for policymakers and firms focusing on sustainable development strategies.
Abstract China’s real estate market demonstrates a systemic sensitivity to macroeconomic regulation, necessitating a robust legal and empirical evaluation of policy efficacy. This research fills a critical gap by proposing a novel framework that integrates an economic law perspective with a Pressure-State-Response (PSR) model, weighted via the Analytic Hierarchy Process (AHP). Analyzing national market data from 2004 to 2024, the study reveals that housing prices are predominantly driven by state-led land, monetary, and tax interventions rather than fundamental socio-economic variables like per capita income or marriage rates. A key empirical finding is the persistent misalignment between the pressure index and the response index, alongside a significant temporal lag where regulatory “state” indices consistently follow market “response” trends. These results confirm a policy-driven market mechanism and highlight the need for proactive institutional optimizations within the economic law framework to ensure long-term market stability and social welfare.
This study investigates the impact of generative artificial intelligence (AI) acceptance on strategic decision-making speed and innovation performance within small and medium-sized enterprises (SMEs). The research utilizes data collected through a cross-sectional survey of 392 SME employees. The collected data were analyzed using Partial Least Squares Structural Equation Modeling (PLS-SEM) with SmartPLS 4. The findings indicate significant relationships between generative AI acceptance and both innovation performance and strategic decision-making speed in SMEs. In particular, mediation analyses reveal that strategic decision-making speed acts as a partial mediator in the relationship between generative AI acceptance and innovation performance, suggesting that higher performance enhances employees’ work pace, thereby strengthening their propensity to adopt AI technologies. These results contribute to a deeper understanding of the dynamics of AI integration in organizational contexts and offer valuable insights for SME managers and policymakers. The study provides meaningful implications for how AI acceptance can influence business processes, strategic decision-making, and innovation outcomes, highlighting the complex and nuanced effects of AI adoption on organizational performance and innovation.
Amid the rapid evolution of the digital economy, English for Specific Purposes (ESP) teachers – who serve as crucial links between higher education and industry – face both the pressures of transformation and the opportunities of strategic growth. Drawing on a systematic review of 50 articles from the Science Citation Index Expanded (SCIE), Social Sciences Citation Index (SSCI), and Arts & Humanities Citation Index (A&HCI) databases, this study identifies key pathways, barriers, and policy implications for ESP teachers’ career development. The findings reveal that data-driven approaches and artificial intelligence technologies are reshaping teacher training. Collaborative university–industry programs, online autonomous learning platforms, and intelligent feedback systems have become mainstream models. However, three major governance challenges persist: (1) The integration of industry knowledge and language pedagogy remains incomplete, limiting instructional effectiveness; (2) Career pathways for teachers lack institutionalization and sustainability; (3) Structural inequalities in technology access and application undermine organizational performance and educational equity. To foster the joint advancement of ESP teachers and institutional capacity, it is essential to develop an institutionalized digital empowerment platform, enhance performance feedback and decision-support systems, implement ESG-oriented teacher development policies to ensure equitable technology access, and strengthen cross-departmental collaboration. These measures can support a strategic and inclusive response to digital transformation within the education system.
Digital economy has fundamentally reconfigured the relationship between corporations and investors. With their substantial data reserves and multifaceted content, Social media platforms are becoming important channels for investors to acquire information, express perspectives, and shape investment decisions. This study utilizes data from Zhihu, a prominent Chinese social media platform, to examine listed A-share companies on the Shanghai and Shenzhen stock exchanges. It further explores how topic-based interactions on corporate social media influence investor behavior and stock market dynamics, with particular emphasis on the functional and emotional implications of earnings-related and non-earnings-related discussions. Findings show that: first, there is a positive correlation between the level of engagement with corporate accounting topics and stock returns; Second, interactions on earnings-related topics provide investors with functional investment benefits, while discussions on non-earnings topics offer emotional benefits; Third, compared with interactions on non-earnings topics, engagement with earnings-related discussions exerts a stronger influence on investor sentiment and, consequently, on stock returns.The findings reveal the fundamental patterns of investor online engagement, thus providing support for enterprises to develop and implement market intervention strategies.
This study examines how oil market uncertainty influences stock market integration among four major energy-dependent European economies (France, Germany, Italy, and Spain). Oil uncertainty is proxied by the CBOE Crude Oil Volatility Index (OVX), a forward-looking implied volatility measure. Using weekly data from May 2007 to December 2025, the analysis employs univariate GARCH-X models to assess the direct effects of uncertainty on individual markets, a multivariate DCC-GARCH framework to capture time-varying correlations, and ARDL models to investigate short-run transmission mechanisms. The results indicate that oil uncertainty exerts a statistically significant negative effect on all markets and strengthens cross-market integration during periods of heightened uncertainty. However, this effect is not persistent. Evidence suggests a two-stage dynamic in which initial uncertainty shocks are associated with herding-like behavior and financial contagion, followed by portfolio rebalancing and partial market decoupling. The findings highlight the non-linear and time-varying nature of financial integration under energy-related uncertainty and offer important implications for portfolio diversification, risk management, and energy-dependent economies. Furthermore, rolling-window estimates and structural break tests reveal that this integration dynamic is highly regime-dependent, intensifying notably during major global crises. The robustness of these behavioral mechanisms is confirmed using Brent crude oil realized volatility.
As climate risks intensify, climate investment and financing policies have become an important tool for mitigating supply chain disruption and enhancing resilience. This study examines whether China's Climate Investment and Finance Pilot Policies (CIFPPs) reduce disruption risks among upstream and downstream firms linked to focal firms and explores the underlying mechanisms. Using panel data on Chinese A-share listed firms from 2017 to 2023, we employ a Double Machine Learning model based on the Lasso algorithm to identify causal effects. The results show that CIFPPs generate significant supply chain spillover effects through focal firms, reducing disruption risks for associated partners. The effects are mainly transmitted downstream, significantly benefiting customer firms rather than suppliers, which suggests asymmetric policy absorption across supply chain positions. Mechanism analysis indicates that CIFPPs operate by promoting collaborative green talent innovation and improving the quality of green technological innovation. These effects are stronger among geographically proximate and highly digitalized firms. In addition, CIFPPs indirectly improve firms' carbon performance by reducing disruption risks. Overall, the findings highlight the spillover logic of climate finance policies and their implications for economic resilience and environmental sustainability.
Traditional sectors such as construction face persistent challenges in technological upgrading, digital adoption, and productivity transformation. However, the role of managerial knowledge attributes in facilitating such transitions remains largely unexplored. This study investigates whether the academic background of senior executives enhances new-quality productivity (NQP) in the construction industry. Using data from Chinese A-share listed construction contracting firms from 2013 to 2023, estimate the conditional causal effects using generalized random forests (GRF) with orthogonalized scores, complemented by entry-exit difference-in-differences (DID) identification. The results show that executives with academic backgrounds significantly improve firms' NQP outcomes. Mechanism analysis reveals that the effect operates through accelerated digital transformation and the adoption of AI applications. Furthermore, the impact is conditioned by firm size, equity incentives, and market valuation. These findings highlight academic-oriented managerial attributes as an overlooked driver of productivity transformation in traditional sectors and provide evidence relevant not only for construction enterprises but also for other traditional and low-digitalization sectors undergoing structural upgrading.
Valuing ecological products is a crucial strategy for achieving carbon neutrality goals through nature-based solutions. Focusing on the Northeast China black soil region, this study provides a quantitative simulation of an Ecological Product Tax by integrating the InVEST model with a Computable General Equilibrium (CGE) model. The integrated framework first quantifies the region's carbon sink value at 636.468 billion RMB, establishing a numerical benchmark. Subsequently, the CGE model simulations demonstrate that internalizing this ecological value through taxation effectively transforms ecological capital into economic momentum. This process is shown to notably boost regional employment, capital returns, social consumption, and investment, thereby facilitating a transition toward a domestic demand-driven economy. A heterogeneity analysis further reveals significant disparities in these effects across industries and regions, Liaoning benefits most from capital deepening, while Jilin shows a stronger consumption-driven response. The study provides robust simulation evidence that the "Ecological Product Tax" mechanism optimizes industrial structure and enhances ecological efficiency. Consequently, we recommend the implementation of a tiered ecological product tax system synchronized with existing environmental tax frameworks to achieve sustainable industrial revitalization.
The persistent decline in the relative price of investment goods has emerged as a global stylized fact. This paper develops a three-period Overlapping Generations model to delineate the theoretical mechanism through which population age structure affects the relative price of investment goods via household savings and corporate investment decisions. Based on a comprehensive panel dataset of 155 countries from 1970 to 2024, the empirical results demonstrate that population ageing systematically depresses the relative price of investment goods by elevating household savings rates while simultaneously suppressing corporate investment demand, thereby altering the capital supply-demand dynamics. Heterogeneity analysis reveals that this effect is particularly pronounced in economies with more advanced financial systems and better-developed pension systems. By incorporating demographic structure into the analytical framework of long-term asset pricing, this study provides a unified demographic explanation for the global downward trend in relative price of investment goods, while offering crucial empirical evidence to help countries undergoing demographic transition mitigate potential asset price volatility and financial risks.
In the context of the rapid growth of China's new energy vehicle industry and the rise of short-video platforms, the internet celebrityization of CEO personal brands has emerged as an important channel for brand communication. However, existing research has paid limited attention to the components of CEO personal brands and their influence mechanisms in such dynamic digital environments. Drawing on social identity theory and industry characteristics, this study identifies four core components of CEO personal brand: morality, ability, personalization, and attractiveness. Using a mixed-method approach, content analysis was first conducted to capture key traits, followed by a survey yielding 529 valid responses. The results show that all four dimensions have significant positive effects on corporate brand image, with social identity playing a mediating role. Furthermore, the moderating effect of internet celebrityization follows an inverted U-shape, with the strongest support found for morality, while evidence for ability is comparatively weaker. The findings remain robust after controlling for demographic factors. This study extends research on CEO branding and offers practical implications for optimizing brand communication strategies in the new energy vehicle industry.
The Beijing-Tianjin-Hebei (BTH) region of China faces challenges in achieving sustainable development, necessitating a delicate balance between rapid economic growth, environmental protection, and efficient energy utilization. Against this background, this paper develops a System Dynamics (SD) model that simulates the Economy-Energy-Environment-Policy (3E-P) system from 2008 to 2035, calibrated with data from 2008 to 2020. The model comprises four interconnected subsystems: (i) economic subsystem, which includes primary, secondary and tertiary industries, as well as labor, capital, energy, and technological factors; (ii) energy subsystem, encompassing the supply and demand dynamics of both fossil and renewable energy sources; (iii) environmental subsystem, accounting for carbon dioxide emissions, water usage, air quality, and solid waste; (iv) policy-subsystem, addressing fiscal, industrial, energy, low-carbon, and pollution control policies. Five development scenarios are simulated: the current state, economy-focused, environment-focused, technology-driven, and coordinated development scenarios. Results indicate that the coordinated development scenario can reduce annual CO2 emissions by 20 % while maintaining an average GDP growth rate of 6.0 %, demonstrating the effectiveness of integrated policy measures. The study highlights the significance of synergistic strategies across economic, energy, and environmental domains. It also offers a robust decision-making tool to support sustainable and high-quality development in the BTH region.
With the rapid development of the digital economy, digital finance has become an important factor shaping household consumption dynamics. Using balanced panel data from 31 Chinese provinces during 2011-2023, this study examines the relationship between digital finance and household consumption from the perspectives of average effects, distributional heterogeneity, regional variation, spatial dependence diagnostics, and nonlinear predictive validation. Methodologically, two-way fixed-effects models provide baseline estimation, panel quantile regression captures distributional heterogeneity, Moran's I statistics evaluate spatial dependence, and machine-learning models serve as external predictive validation tools. The results show that the mean effect of digital finance on consumption is statistically weak, but significant distributional heterogeneity exists across the consumption distribution. Digital finance exhibits stronger positive effects at lower consumption quantiles and weaker effects at higher quantiles, indicating a pro-consumption and distribution-sensitive pattern. Mediation analysis provides limited short-run evidence for indirect transmission channels, while regional differences are modest and spatial dependence reflects clustering rather than structural spillovers. Machine-learning validation confirms the stable importance of digital finance under nonlinear prediction. This study highlights distributional heterogeneity as the central empirical feature of digital finance's consumption impact in China's mature digital-finance stage, offering policy implications for inclusive consumption growth and regionally balanced development.
This study explores the way digital transformation relates to business model innovation by using a bibliometric analysis. The goal is to trace the development of the field and clarify its underlying theoretical foundations and emerging research themes. It is based on 164 publications from the Web of Science core collection spanning 2017 to 2024. Utilizing VOSviewer and SCImago Graphica for data visualization, the analysis identifies publication trends, influential authors, leading journals and institutions, and the geographic distribution of research outputs. The findings reveal a significant increase in publications after 2020, with 64.02 % of the studies related to these two key concepts indexed in the SSCI. Despite the rapid growth of digital transformation and business model innovation research, existing reviews lack a coherent synthesis of their intersection, a methodologically unified understanding of enterprise adaptation across contexts, and an up-to-date bibliometric review based on consistently curated datasets. Addressing these gaps, this study identifies 7 stable thematic clusters using a transparent, PRISMA-guided bibliometric mapping approach with VOSviewer and SCImago Graphica, and reveals emerging themes that offer actionable insights for academics, practitioners, consultants, and policy makers.
Links between wages, prices, and inflation expectations are examined for the United States (US) and United Kingdom (UK) using time-varying Granger causality tests applied to monthly (US 1983-2025; UK 2001-2025) and quarterly (US 1967-2025; UK 1985-2025) data. Pass-through is episodic and sector-specific rather than a stable feature of the macroeconomy. In the US, wage inflation Granger-causes price inflation during the dot-com period, the mid-2000s, and again post-COVID-19 (2020-2025), with two-way wage-price feedback emerging in the private sector from 2021 onward. In the UK, wage-to-price links reappear in 2021-2023, while recent manufacturing evidence indicates price-led wage adjustments (price inflation Granger-causing wage inflation). Inflation expectations are predominantly adaptive - price inflation Granger-causes inflation expectations - but feedback from expectations to prices strengthens during turbulent periods, including the Global Financial Crisis (GFC) and the post-pandemic inflation surge. Aggregation to quarterly frequency attenuates but does not eliminate these regime-dependent dynamics. By dating the onsets and terminations of predictive links, the analysis reconciles muted full-sample pass-through with short-lived but policy-relevant bursts, and highlights the value of real-time sectoral monitoring for assessing second-round inflation risks.
