Empowering Sustainable Transformation: How Digital Finance Drives Productivity Growth in Resource-Based Enterprises

Abstract

Digital finance, representing the deep integration of finance and technology, has become a critical enabler of sustainable industrial transformation. Focusing on resource-based enterprises (RBEs)—key actors in transitioning towards sustainable practices—this study investigates how digital finance development fosters new quality productive forces (NQPFs), a core driver of high-quality, sustainable development. Utilizing panel data from Chinese A-share listed RBEs (2008–2022), we measure NQPF using the entropy method and gauge regional digital finance development with the Peking University Digital Financial Inclusion Index (DFII). Empirical analysis employing two-way fixed effects and panel threshold regression models provides robust evidence that digital finance significantly enhances NQPFs within RBEs. Crucially, mechanism analysis identifies three fundamental pathways underpinning sustainability: (1) mitigating financial constraints; (2) facilitating technological innovation and transformation; (3) strengthening green transition awareness. Furthermore, the impact of digital finance exhibits synergistic enhancement alongside increasing environmental regulation intensity and improved financial resource allocation efficiency. Heterogeneity analysis reveals that the effect is more pronounced in regions with lower marketization, within state-owned enterprises, and among RBEs in recession stages. Collectively, these findings offer significant implications for policymakers and industry practitioners aiming to strategically leverage digital finance to accelerate the sustainable transformation of resource-intensive industries, thereby contributing directly to environmentally sustainable and resilient economic development.

1. Introduction

The global imperative for sustainable industrial transformation necessitates reconciling a critical trilemma: balancing economic vitality, environmental stewardship, and resource efficiency. Within this context, New Quality Productive Forces (NQPFs) emerge as a pivotal paradigm shift. Representing an advanced form of productivity, NQPFs are intrinsically characterized by heightened efficiency, sustainability, and competitiveness. It signifies a qualitative leap beyond traditional models, driven by the deep integration of novel technologies, optimized production factors, and innovative structures enabled by scientific and technological progress. Propelled by the modern revolutions in digitalization, intelligence, and green transition, NQPF marks a new trajectory for global productivity development oriented toward long-term sustainability and resilience. It serves as the core engine for achieving genuinely high-quality, sustainable economic growth while overcoming persistent bottlenecks like inefficiency, resource scarcity, and environmental degradation.

Resource-based enterprises (RBEs), foundational yet often environmentally intensive industries reliant on mineral extraction and primary processing, hold strategic importance in national economies worldwide. However, they face intensifying pressures that underscore their central role in the sustainability challenge. Escalating factor costs and tightening environmental constraints have starkly exposed the limitations of their conventional high-input, high-consumption, high-pollution growth model, leading to significant challenges including overcapacity and depressed productivity [1,2]. Critically, RBEs generate substantial negative environmental externalities, such as land degradation, water system disruption, and air pollution [3]. Consequently, accelerating the formation of NQPFs within RBEs represents an essential and urgent pathway for achieving genuinely sustainable industrial development—balancing economic vitality with environmental integrity—a challenge faced by resource-dependent economies globally.

The rapid global advancement of Digital Finance (DF), significantly accelerated by supportive policies and technological progress, offers a promising enabler for the sustainable transformation of RBEs. DF leverages advanced technologies—including big data analytics, artificial intelligence, cloud computing, blockchain, and the Internet of Things—to provide novel, accessible financial services [4,5]. Distinct from earlier concepts, DF emphasizes the transformative potential of underlying digital technologies to reshape financial inclusion and efficiency. By deeply integrating digital innovation with financial services, DF optimizes processes and products, mitigating traditional issues like information asymmetry and high operational costs [6,7]. For RBEs seeking to cultivate NQPFs, reliance solely on market mechanisms, internal governance, or government mandates often proves insufficient. DF presents a compelling hybrid governance mechanism, potentially bridging the gap between market forces and regulatory enforcement by integrating financial instruments with environmental objectives [8]. This integration offers the potential to reduce the rigidity of administrative intervention while strengthening market efficiency, positioning DF as a novel and potent instrument for sustainable industrial policy.

This critical juncture raises fundamental research questions central to advancing sustainable industrial transformation globally. This study therefore addresses three core questions: First, can digital finance effectively foster the development of New Quality Productive Forces within resource-based enterprises? Second, through what mechanisms might DF exert this influence? Third, does the impact exhibit significant heterogeneity based on enterprise characteristics or external contexts? Addressing these questions is vital for designing effective policies to leverage DF for sustainable outcomes in resource-intensive sectors worldwide. Despite the growing importance of both NQPF and DF, significant knowledge gaps persist regarding their interplay, particularly at the micro-level within the critical RBE sector. Crucially, empirical evidence on DF’s specific pathways and contextual effectiveness in driving sustainable productivity gains in these industries remains limited.

To bridge these gaps, this study empirically examines the impact and underlying mechanisms of DF on the NQPF of Chinese resource-based listed companies (2008–2022), providing insights with potential relevance to resource-dependent economies globally. This research aims to make several key contributions to the literature on sustainable transitions: First, by developing a conceptual framework and novel evaluation system tailored to measure NQPFs within RBEs, it addresses the challenge of quantifying this complex construct at the firm level, addressing a critical micro-level knowledge gap in this sector [9,10,11]. Second, it seeks to elucidate the synergistic mechanisms through which DF influences RBE NQPF, particularly its role in integrating environmental and financial governance to overcome key barriers to sustainable transformation. Third, the study investigates contextual heterogeneity in DF’s effectiveness, providing nuanced insights for policymakers seeking to harness digital finance as a targeted instrument for accelerating the sustainable transformation of resource-intensive industries.

2. Literature Review and Theoretical Hypotheses

2.1. Literature Review

2.1.1. New Quality Productive Forces in Resource-Based Enterprises: Drivers and Knowledge Gaps

New Quality Productive Forces (NQPFs) signify a transformative productivity paradigm arising from technological revolutions, innovative resource allocation, and systemic industrial restructuring [12]. Systems theory conceptualizes NQPFs through evolved labor, tools, and infrastructure [13], while broader perspectives position digitalization, green technologies, and sustainable energy systems as its core pillars [14]. Despite NQPF’s centrality to sustainable development, its measurement remains methodologically challenging [9,10,11]. While total factor productivity (TFP) is a common proxy, emerging multi-dimensional frameworks are largely confined to regional analyses, with critically underdeveloped enterprise-level assessments. Research identifies data-driven innovation, digital transformation, and policy governance as key enablers of NQPFs in manufacturing and services [15,16,17], yet consistently neglects the unique sustainability challenges confronting resource-intensive sectors (RBEs).

Resource-based enterprises (RBEs)—strategically vital yet environmentally high-impact industries—exemplify this research gap. Economic analyses indicate capital deepening can marginally improve RBE efficiency [1], while institutional barriers, such as fragmented environmental regulations and weak sustainability governance, constrain their green transitions [18]. Although digital transformation helps RBEs overcome operational constraints [2,19], and fiscal tools (e.g., environmental taxes, resource pricing) incentivize incremental innovation [20,21,22], existing studies lack systematic frameworks linking these drivers to the sustainable core of NQPFs. Proposed industry models, such as integrated digital-green coal mining [23], remain theoretically promising but empirically unvalidated. This leaves a critical void in understanding how NQPF-driven sustainability transitions can be operationalized within RBEs.

Crucially, while adaptive financial systems are essential for catalyzing enterprise-level NQPFs—particularly in sectors grappling with significant environmental externalities—the specific role of digital finance (DF) remains unexplored. This oversight impedes the design of evidence-based policies for sustainable RBE transformation [24].

2.1.2. Digital Finance: An Emerging Catalyst for Sustainable Transitions

Digital finance (DF), characterized by the fusion of traditional finance with frontier technologies (e.g., AI, blockchain, IoT), reconstructs financial ecosystems through enhanced efficiency, inclusivity, and innovation. Its core capabilities—including real-time responsiveness, reduced transaction costs, data-driven risk management, and expanded market access—position DF as a potent enabler of sustainable enterprise development.

Empirical research reveals DF’s multi-scalar impacts. At the macro level, it stimulates inclusive growth and entrepreneurship [25,26,27]; at the micro level, it alleviates financing constraints, lowers capital costs, and fosters innovation within SMEs and real-economy firms [28,29,30,31,32]. Notably, DF’s positive effects are often amplified in underdeveloped regions [33], highlighting its capacity to bridge sustainability disparities [34]. However, current scholarship focuses predominantly on generic productivity outcomes (e.g., TFP), overlooking DF’s potential to drive intrinsically sustainable productivity—embodied by NQPFs—within environmentally critical sectors like RBEs.

2.1.3. Bridging the Gap: This Study’s Contribution to Sustainable Transition Research

Despite these advances, a pivotal knowledge gap persists: no study has systematically examined how DF activates NQPFs in resource-based enterprises—a sector where reconciling productivity gains with environmental stewardship is imperative. Understanding DF’s micro-mechanisms in this context is essential to decode its sustainability potential and design targeted transition policies [35]. This research addresses three critical limitations in the extant literature:

(1)

Sector-Specific NQPF Conceptualization: Prior work neglects RBEs’ dual identity as economic anchors and environmental burden-bearers, failing to conceptualize NQPFs within the unique sustainability challenges specific to this sector.

(2)

DF’s Synergistic Pathways: While DF’s generic benefits (e.g., financing alleviation) are established, its synergistic pathways to fostering NQPFs—particularly through enhancing green transition awareness, enabling technological transformation, and integrating with environmental regulation—remain unstudied.

(3)

Institutional Heterogeneity: The influence of contextual factors (e.g., marketization levels, ownership structures, firm life cycles) on DF’s sustainability impact within RBEs is significantly underexplored [36].

By empirically analyzing DF’s role in fostering NQPFs within Chinese RBEs (2008–2022), this study deciphers its core mechanisms—alleviating financing constraints, empowering technological transformation, and enhancing green transition awareness—alongside key contextual contingencies. Our findings provide empirically grounded policy levers for leveraging DF to accelerate sustainable transitions in resource-dependent economies globally.

2.2. Theoretical Framework and Hypotheses Development

Building on the identified gaps, we conceptualize DF as a catalytic enabler for NQPFs in RBEs. DF’s technological capabilities (e.g., blockchain transparency, AI-driven analytics) reconfigure financial and operational systems, aligning with Porter’s hypothesis that environmental and economic efficiency can be synergistic. We theorize three pathways through which DF overcomes RBE-specific sustainability bottlenecks: (1) financial constraint alleviation, (2) eco-technological empowerment, and (3) green institutionalization. These pathways operate within threshold conditions shaped by institutional maturity, justifying the following hypotheses:

2.2.1. Digital Finance as an Enabler of Sustainable Productivity

Digital finance (DF) transcends traditional financial models by integrating frontier technologies (e.g., blockchain, AI, IoT) to optimize capital allocation and service delivery. For resource-based enterprises (RBEs)—historically constrained by capital intensity and environmental externalities—DF bridges critical gaps between financial supply and sustainable development needs. By leveraging big data analytics and digital platforms, DF enhances credit risk assessment accuracy, expands financing channels, and reduces transaction costs [37,38]. This enables RBEs to secure essential funding for eco-innovation, circular production systems, and clean technology adoption—core drivers of New Quality Productive Forces (NQPFs) that harmonize efficiency with environmental stewardship.

Operationally, DF empowers real-time monitoring of resource flows and emissions, enabling sustainability-oriented optimization. Intelligent analytics streamline supply chains, reduce energy waste, and enhance responsiveness to green demand signals [39,40], positioning DF as a catalyst for decoupling productivity growth from resource depletion—a foundational objective of NQPFs. We therefore propose:

H1. 

Digital finance development significantly enhances new quality productive forces (NQPFs) in resource-based enterprises by enabling sustainable capital allocation and operational transformation.

2.2.2. Synergistic Pathways to Sustainable Transformation

The relationship between DF and NQPFs operates through three deeply interconnected and mutually reinforcing mechanisms that collectively advance sustainability transitions, underpinned by institutional theory and dynamic capability perspectives. First, DF alleviates green financing constraints by mitigating RBEs’ chronic underfunding of sustainability initiatives through innovative instruments such as green bonds and ESG-linked loans. Crucially, by reducing information asymmetry and monitoring costs via blockchain-enabled transparency, DF lowers capital costs for eco-innovation and broadens access for a wider range of firms, not merely those with pre-existing high NQPFs, thereby redirecting funds toward circular economy projects and pollution-control technologies [38]. This foundational financial enabler facilitates the second mechanism: DF accelerates eco-technological transformation by embedding digital tools—including AI diagnostics and IoT sensors—into RBE operations. This integration fosters dynamic capabilities for precision resource management, predictive maintenance to minimize waste, and automation of high-emission processes [41,42], yielding data-driven optimization that reduces ecological footprints while boosting productivity. Third, building upon this technological infrastructure, DF institutionalizes green transition capacity by cultivating organizational sustainability culture. Digital platforms act as carriers of institutional norms, disseminating environmental regulations and stakeholder expectations, thereby creating sustained normative and mimetic pressures for eco-innovation [8]. Concurrently, performance analytics quantify sustainability ROI, reinforcing institutionalization by incentivizing RBEs to structurally embed green objectives into strategic planning and routines [43]. Critically, these pathways operate synergistically in a self-reinforcing cycle: improved financing access (Mechanism a) enables the adoption and scaling of transformative eco-technologies (Mechanism b), which in turn generates the data, capabilities, and operational evidence necessary to build and legitimize enduring institutional commitment to sustainability (Mechanism c). This institutionalized green capacity then signals credibility, potentially further easing future financing constraints and driving continuous technological refinement. Thus, we hypothesize:

H2. 

Digital finance elevates new quality productive forces (NQPFs) through three mediating mechanisms: (a) alleviating green financing constraints, (b) driving eco-technological transformation, and (c) institutionalizing green transition capacity.

2.2.3. Threshold Dynamics in Sustainability Transitions

DF’s impact exhibits nonlinear scalability due to network externalities and institutional learning. While early-stage DF deployment may yield marginal NQPF gains, once adoption crosses critical thresholds in technological integration, stakeholder engagement, and regulatory maturity, DF triggers accelerating sustainability returns [2]. Post-threshold, network effects (e.g., industry-wide blockchain consortia) enable clean-tech knowledge spillovers, while unified ESG standards reduce compliance costs. This sustainability flywheel amplifies NQPFs through cross-enterprise resource sharing, collective green R&D, and market rewards for eco-efficiency. We hypothesize:

H3. 

The positive impact of digital finance on new quality productive forces (NQPFs) exhibits increasing marginal returns after surpassing critical thresholds in (a) technological integration, (b) stakeholder engagement, and (c) regulatory maturity.

3. Research Design

3.1. Model Building

3.1.1. Static Panel Model

Building upon the established methodological frameworks pioneered by Zheng Minggui et al. [1] and You Biying et al. [2], this study constructs and applies an empirical model to rigorously test the impact of digital finance (DF) development on new quality productivity (NQPF) within resource-based enterprises (RBEs). This approach provides a robust analytical foundation for examining the DF-NQPF relationship in the context of these strategically vital, yet environmentally intensive, sectors.

$${NQPF}_{i,t}={\alpha }_0+{\alpha }_1{DF}_{i,t}+\sum Controls+\sum Firm+\sum Industry+\sum Province+\sum Year+{\mu }_{i,t}$$

(1)

In Equation (1), NQPF represents the explanatory variable—the new qualitative productivity of resource-based enterprises, DF represents the explanatory variable—digital financial development, and i and t represent enterprises and years, respectively. Controls stand for Control Variable; Firm is to control the individual fixed effect; Province is the fixed effect of the control province; Year is the fixed effect of control time; ${\alpha }_0$ is the intercept term, ${\alpha }_1$ is the coefficient to be estimated for the development of digital finance, and ${\mu }_{i,t}$ is the random error term.

3.1.2. Mechanism Test Model

To empirically test the hypothesized relationships and elucidate whether digital finance (DF) development influences new quality productivity (NQPF) in resource-based enterprises (RBEs) through the mediating pathways of mitigating financing constraints, driving technological change, and enhancing green transition awareness, this study employs a stepwise regression analysis to assess mediation effects [44]. We construct the following regression model to formally test these mediating mechanisms:

$$M_{i,t}{ =\beta }_0+{\beta }_1{DF}_{i,t}+\sum Controls+\sum Firm+\sum Industry+\sum Province+\sum Year+{\mu }_{i,t}$$

(2)

$${{{NPF}_{i,t}=\theta }_0+\theta }_1{DF}_{i,t}+{\theta }_2{M}_{i,t}+\sum Controls+\sum Firm+\sum Industry+\sum Province+\sum Year+{\mu }_{i,t}$$

(3)

In Equations (2) and (3), M is the mediating variable, including the mitigation of financing constraints, technological change, and awareness of green transition, and the definitions of other variables are consistent with model (1).

Then, the actual test is carried out step by step: regression is carried out for model (1), and if ${\alpha }_1$ is not significant, the test is terminated; if ${\alpha }_1$ are significant, models (2) and (3) are tested. If both ${\beta }_1$ and ${\theta }_2$ are significant, the M mediating effect exists, and ${\theta }_1$ is also significant, indicating that there is a partial mediating effect, and ${\theta }_1$ is not significant, indicating that there is a complete mediating effect. If one or both of ${\beta }_1$ and ${\theta }_2$ are not significant, a Bootstrap test is required. If the product of ${\beta }_1$ and ${\theta }_2$ is the same as the ${\theta }_1$ sign, then M has a mediating effect; If the sign is reversed, M has a masking effect.

3.1.3. Panel Threshold Model

Zhao et al. [45] tested the idea and constructed a panel regression model:

$$\begin{array}{l}{NPF}_{i,t}={\gamma }_0+{\gamma }_1{DF}_{i,t}\times I\left(q_{i,t}\le \beta \right)+{\gamma }_2{DF}_{i,t}\times I\left(q_{i,t}>\beta \right)+\sum Controls+\sum Firm+\sum Industry\\ +\sum Province+\sum Year+{\mu }_{i,t}\end{array}$$

(4)

Equation (4), where I (·) represents the conditional function, if the condition in parentheses is true, I is assigned as 1, otherwise it is 0; ${\gamma }_0$ is the intercept term, and ${\gamma }_1$ and ${\gamma }_2$ are the coefficients to be evaluated for the development of digital finance. $q_{i,t}$ Are threshold variables such as digital finance development, financing constraints, technological change, and awareness of green transition? Equation (4) takes the form of a single threshold model, which can be extended to multiple thresholds when the presence of multiple thresholds needs to be detected.

3.1.4. Sample Selection and Data Sources

Following the methodological approach established by Ma Jie et al. [19], resource-based enterprises (RBEs) were defined according to China’s Industrial Classification of the National Economy standards, encompassing firms within 15 key industries including petroleum, coal, and non-ferrous metals. The research sample comprises China’s A-share listed RBEs spanning the period 2008 to 2022. To ensure data quality and relevance, we excluded enterprises designated as ST, *ST, or PT, along with those exhibiting significant deficiencies in financial data or other essential research variables. This rigorous screening process yielded a final balanced panel dataset of 14,369 firm-year observations, representing 958 distinct enterprises. To mitigate the potential distortionary effects of outliers, all continuous variables were winsorized at the 1st and 99th percentiles. Financial data were sourced from the authoritative CSMAR Database and Wind Database. Patent-related information was obtained from official publications of the State Intellectual Property Office and relevant WIPO listings. Indicators pertaining to enterprise green transformation were meticulously derived from environmental disclosures within the annual reports of the sampled listed companies. Finally, regional-level data were compiled from the China Statistical Yearbook.

3.2. Variable Measurement and Description

3.2.1. Dependent Variable: New Quality Productivity of Resource-Based Enterprises (NQPF)

Marx conceptualized productive forces as the material capacity enabling humans to transform nature through labor to fulfill societal needs, fundamentally comprising laborers, means of labor, and objects of labor [46]. The concept of New Quality Productive Forces (NQPFs) represents an innovative evolution of this foundational theory. For resource-based enterprises (RBEs), NQPF manifests at the micro-level as the optimization and upgrading of production factors during resource development and transformation. This conceptualization adheres to NQPF’s core principles while incorporating distinct RBE characteristics, most prominently through the advancement of Marx’s three elements:

(1)

New-type Laborers: Encompassing the knowledge enhancement, skill upgrading, and innovation capacity development of RBE employees and management. This evolution fosters improved resource development capabilities and collaborative efficiency, positioning laborers as the most dynamic element [10].

(2)

New-type Means of Labor: Including both the technological modernization of traditional resource extraction equipment and the digitalization of production tools, serving as key enablers of high technology and operational efficiency [13].

(3)

New-type Objects of Labor: Involving the development of resource-based products that meet market demand for green innovation while ensuring environmental compatibility with mining ecosystems, thereby embodying high-quality, sustainable characteristics [47,48].

These element-specific upgrades provide the critical theoretical foundation for investigating the relationship between factors like digital finance and NQPF in enterprises. Consequently, building upon this definition of NQPF’s core connotation and RBE-specific features, and informed by existing literature [11,49], we construct an enterprise-level NQPF evaluation index system.

Given that Marx’s three fundamental elements determine the character and developmental stage of productive forces [46], we posit that the advancement levels of new-type laborers, means of labor, and objects of labor within RBEs serve as primary indicators for assessing NQPF development.

Table 1. Evaluation index system for new quality productivity of resource-based enterprises.

Target	Level 1 Indicators	Secondary Indicators	Level 3 Indicators	Calculation Method:
The level of development of new quality productive forces	New quality workers	Quality of staff	Proportion of highly educated personnel	Number of master’s degree or above/total number of employees
			Proportion of R&D personnel	Number of R&D personnel/total number of employees
		Quality of management	Management’s green awareness [50]	ln (frequency of green development keywords in the annual report + 1)
			Richness of CEO functional experience [51]	CEO Functional Experience Count—Mean of CEO Functional Experience in the Resources Industry
	New quality labor materials	Mining science and technology labor data	The level of mining technology innovation	ln (number of patents granted by the enterprise + 1)
		Mine digital labor data	Proportion of digital assets in mines [52]	Total digitally related assets/intangible assets
		Intelligent labor data for mines	The level of investment in mine intelligence [53]	The total investment in intangible assets and fixed assets related to intelligence/the total annual investment of the enterprise
			Mining robot penetration [54]	Broken down by weights of the data from the International Federation of Robotics
	New quality labor objects	Green new products	Proportion of green patented products	The number of products applying for green patents/the number of products applying for patents
		Mine ecological environment	Mine environmental performance [55,56,57,58,59]	Adopt the environmental score in the ESG scoring system of China Securities

defines and operationalizes the specific metrics for these indicators. We subsequently apply the entropy weight method to determine indicator weights and compute the comprehensive NQPF evaluation metric.

3.2.2. Explanatory Variable: Digital Finance Development (DF)

This study employs the Peking University Digital Inclusive Finance Index to measure the dynamic development of digital finance (DF). This comprehensive index system evaluates DF across three core dimensions:

(1)

Breadth of Coverage (BCO): Reflecting the penetration of digital financial services across regions, as evidenced by metrics such as electronic account adoption, demonstrating the transcendence of geographical barriers by emerging internet finance models.

(2)

Usage Depth (UDE): Capturing the practical application intensity of digital financial services by end-users, encompassing critical domains including payments, money market funds, credit, insurance, and investments.

(3)

Digitization Level (DIG): Assessing the maturity and sophistication of digital financial services through factors such as the level of credit services enabled, convenience, and cost efficiency.

To ensure comparability and mitigate the influence of scale differences across the index components, all original index values were uniformly scaled by dividing by 100 prior to analysis.

3.2.3. Mediation Variables

To investigate the potential pathways through which digital finance development influences the new quality productivity (NQPF) of resource-based enterprises—specifically, financing constraints, technological change, and green transition awareness—we draw upon established research practices [44]. We operationalize these pathways using three mediating variables:

(1)

Financing Constraint Index (SA): Measured using the absolute value of the SA index.

(2)

R&D Intensity (RD): Calculated as the natural logarithm of the total number of utility model, design, and invention patent applications plus one, divided by the natural logarithm of R&D expenditure plus one.

(3)

Green Transition Focus (GF): A composite score derived from the environmental disclosures of listed companies. This score reflects the establishment of environmental management systems, implementation of environmental training, execution of environmental actions, and attainment of both ISO9001 and ISO14001 certifications. Each successfully disclosed item contributes one point, resulting in a total score ranging from 0 to 5.

3.2.4. Control Variables

To account for other factors impacting the dependent variable, these control variables are incorporated: Firm Size, using the natural log of year-end total assets; Firm Age, based on the natural log of the founding year; Leverage Ratio, calculated as total liabilities divided by total assets at year-end; Return on Assets (ROA), defined as net profit over average total assets; Total Asset Growth Rate, derived from current-year asset growth divided by beginning-of-year total assets; Largest Shareholder Ownership, indicating the top shareholder’s stake; Separation of Control and Ownership Rights (Dual), reflecting the excess of voting over cash flow rights; Growth Opportunity, measured by Tobin’s Q (market value of equity, preferred stock, and net debt divided by total assets’ book value); and Board Size, using the natural log of director count.

4. Empirical Analysis

4.1. Descriptive Statistics

Descriptive statistics for all study variables are presented in

Table 2. Descriptive statistical result of variables.

Variable	Variable Symbol	N	Mean	Std	Min	Max
New quality productivity of resource-based enterprises	NQPF	14,369	6.915	0.960	−1.352	11.418
Digital finance development	DF	14,369	2.298	0.817	0.188	3.766
Breadth of coverage	BCO	14,369	2.103	0.806	0.021	3.541
Use depth	UDE	14,369	2.339	0.899	0.088	4.011
Degree of digitalization	DIG	14,369	2.877	1.076	0.075	4.566
Technological change	RD	14,369	0.457	0.980	0.000	7.548
Green transformation of enterprises	GF	14,369	0.918	0.912	0.000	5.384
Financing constraints	SA	14,369	3.821	0.285	2.120	4.670
The size of the enterprise	Size	14,369	22.647	1.416	19.023	28.636
The age of the business	Age	14,369	2.77	0.36	0.00	4.79
Debt-to-asset ratio	Lev	14,369	0.482	0.189	0.014	1.056
Net profit margin on total assets	ROA	14,369	0.043	0.058	−0.644	0.517
The shareholding ratio of the largest shareholder	Top1	14,369	35.498	15.258	3.390	89.986
Separation of powers	Dual	14,369	0.160	0.367	0.000	1.000
Enterprise growth	TQ	14,369	1.889	1.377	0.641	31.400
Board size	Board	14,369	2.178	0.198	0.693	2.890

. Key observations emerge from this analysis. First, the new quality productivity (NQPF) variable exhibits a substantial difference between its maximum and minimum values. This significant dispersion indicates considerable variation in the development level of new quality productivity across the sampled resource-based enterprises. Second, an examination of the control variables reveals that, with the exception of the shareholding ratio of the largest shareholder (TOP1), which displays a relatively large standard deviation, the fluctuation ranges for all other control variables fall within expected norms. The observed distributions of these variables provide a suitable foundation for conducting deeper empirical analysis.

4.2. Baseline Regression Results

Table 3. Benchmark regression results.

	Overall Effect	Structural Effects
	(1)	(2)	(3)	(4)	(5)
Digital finance development	1.036 *** (2.66)
Breadth of coverage		0.633 *** (2.61)
Use depth			0.509 *** (2.02)
Degree of digitalization				−0.324 ** (1.71)	−0.533 *** (2.05)
Degree of digitalization quadratic terms					0.366 *** (2.06)
Control variables	YES	YES	YES	YES	YES
Fixed year	YES	YES	YES	YES	YES
Immobilization of individuals	YES	YES	YES	YES	YES
Stationary industries	YES	YES	YES	YES	YES
R2	0.887	0.886	0.654	0.883	0.866
N	14,369	14,369	14,369	14,369	14,369

Note: The standard deviation of firm cluster robustness is shown in parentheses; **, and *** indicate significance levels of 5%, and 1%, respectively. The same applies hereinafter.

presents the baseline regression results examining the relationship between digital finance development and the new quality productivity (NQPF) of resource-based enterprises. Column (1) reports the overall impact, revealing a statistically significant positive coefficient (at the 1% level) for digital finance development (DF). This indicates that the advancement of digital finance significantly enhances the NQPF of resource-based enterprises, thereby providing empirical support for Hypothesis H1.

The underlying mechanism driving this positive effect stems from the synergistic function of digital finance, which integrates environmental regulation with financial resource allocation. Firstly, DF incentivizes enhanced environmental information disclosure among resource-based enterprises. This reduces the opportunity costs associated with non-disclosure or misrepresentation (“fake information disclosure”), corrects short-termist behaviors, and signals potential for technological innovation. Consequently, enterprises are motivated to invest in green technologies, processes, product R&D, and management optimization. These actions not only offset the “compliance costs” imposed by environmental regulations but also generate a “transformation compensation” effect, ultimately fostering the development of new quality productivity. Secondly, DF internalizes the environmental external costs of resource-based enterprises. It achieves this by lowering transaction costs for investments in cleaner projects while simultaneously increasing the complexity and cost of financing polluting activities. Furthermore, DF facilitates the dynamic adjustment of opportunity costs related to environmental damage throughout the production lifecycle, establishing an effective pathway for financial resource allocation to guide NQPF formation.

Columns (2) through (5) explore the structural heterogeneity of this impact. Results in Columns (2) and (3) demonstrate that both the breadth of coverage (BCO) and usage depth (UDE) of digital finance exert significant positive influences on enterprise NQPF. Notably, the impact of coverage breadth is particularly pronounced. Digital finance possesses a distinct advantage over traditional finance in terms of its extensive geographical reach and population penetration, achieving wide coverage across both service groups and categories. This significantly enhances financial service accessibility and expands enterprise funding channels. The depth of usage, reflecting the intensity and outcomes of user engagement with digital financial services, is also crucial. The positive effect materializes when enterprises genuinely experience the tangible convenience and benefits offered by these services.

Column (4) reveals a negative coefficient for the digitization level (DIG). However, the inclusion of a quadratic term for DIG in Column (5) yields a positive coefficient for the linear term and a negative coefficient for the quadratic term (or vice versa, depending on the results), suggesting a non-linear, U-shaped relationship between digitization level and enterprise NQPF. Empirically, this nonlinearity arises from three interdependent mechanisms: First, adaptation costs dominate initial phases—regions with <3 years of DF development show significant efficiency losses, whereas mature implementations yield net gains, confirming temporal learning effects. Second, infrastructure gaps amplify early-stage inefficiencies—the U-curve steepens in low-broadband regions due to exacerbated resource misallocation from digital exclusion. Third, regulatory lag induces transient financialization—mediation analysis shows rising financial asset ratios during low-DIG phases divert resources from productive green innovation. Collectively, these mechanisms explain the inflection point: As digitization matures, infrastructure upgrades and regulatory catch-up progressively neutralize initial friction, enabling positive net returns.

4.3. Robustness Test

To further validate the reliability of the baseline regression model and its findings, we implemented several robustness checks using diverse methodological approaches:

4.3.1. Alternative Measurement of New Quality Productivity (NQPF)

Considering the historical characteristics of resource-based enterprises, particularly their association with “Three High” industries (high pollution, high energy consumption, high emissions), we substituted the core explanatory variable. Total Factor Productivity (TFP), a well-established indicator reflecting shifts towards new quality productivity, was employed as an alternative measure. TFP captures the average output per unit of combined inputs, representing overall input-output efficiency. Crucially, it correlates with technological progress, factor combination efficiency, and reflects underlying technical and managerial capabilities, making it suitable for observing efficiency changes linked to NQPF formation. Using the Levinsohn-Petrin (LP) method to calculate TFP for resource-based enterprises and re-estimating Model (1), the results (

Table 4. Robustness test results.

	TFP-LP I	NQPF (GLS) II	Replace the Sample III	Exclusion of the Environmental Protection Act IV	Fixed Provinces—Year V
DF	0.896 *** (2.79)	1.101 *** (2.63)	0.922 ** (2.46)	1.099 *** (4.23)	1.216 *** (4.26)
Control variables	YES	YES	YES	YES	YES
Fixed effect	YES	YES	YES	YES	YES
R2	0.927	0.899	0.928	0.887	0.888
N	14,325	14,369	6704	14,369	14,369

The standard deviation of firm cluster robustness is shown in parentheses; **, and *** indicate significance levels of 5%, and 1%, respectively.

, Column I) align substantially with the baseline regression, confirming robustness.

4.3.2. Alternative Estimation Method

We replaced the Ordinary Least Squares (OLS) estimation used in the baseline model with Generalized Least Squares (GLS). As shown in Table 4, Column II, the coefficient for digital finance development (DF) remains positive and statistically significant at the 1% level, consistent with the initial findings, thereby supporting result robustness.

4.3.3. Adjusted Sample Period

Given the extended sample timeframe (2008–2022), which encompasses significant exogenous shocks like the 2008 Global Financial Crisis and the COVID-19 pandemic starting in late 2019, we tested the sensitivity of our results to these events. These shocks could potentially impact the operations of resource-based enterprises and their green strategy implementation. Re-running the regression on the adjusted sample period yielded a DF coefficient of 0.039, significant at the 5% level (Table 4, Column III). This consistency with the baseline results underscores the robustness of our core conclusions.

4.3.4. Controlling for the 2015 Environmental Protection Law

The implementation of the new Environmental Protection Law in 2015 introduced stringent emission reduction and pollution control requirements for enterprises, potentially affecting TFP. To isolate the effect of digital finance development (DF) from this policy intervention, we excluded observations potentially confounded by this law. The results (Table 4, Column IV) show that the coefficient and significance level for DF remain largely consistent with the baseline regression, effectively mitigating concerns about policy confounding and reinforcing result robustness.

4.3.5. Inclusion of Province-Year Fixed Effects

Acknowledging that enterprise NQPFs might be influenced by other environmental protection policies, predominantly implemented at the provincial level in China, we augmented the regression model by adding province–year fixed effects. This controls for unobserved time-varying heterogeneity across provinces, including the potential influence of localized environmental policies. The findings, presented in Table 4, Column V, demonstrate that the positive and significant relationship between DF and NQPFs persists, further affirming the robustness of our primary results.

4.4. Addressing Endogeneity Through Instrumental Variables

Despite the robustness tests conducted, potential endogeneity arising from omitted variables remains a concern. To mitigate this issue, we employ the instrumental variable (IV) approach. We select the number of fixed telephones per 10,000 people and the per capita volume of postal and telecommunications services as instrumental variables. This selection is justified on two grounds. First, the development of broadband and Internet technology in China originated from the foundation laid by fixed telephone networks; consequently, regions with historically higher fixed telephone penetration likely exhibit more advanced levels of digital financial development. Second, these historical variables—fixed telephones per 10,000 people and per capita postal/telecommunications service volume—are plausibly exogenous to the current new quality productivity (NQPF) of enterprises, exerting minimal direct influence on it. Thus, these instruments satisfy the criteria of relevance and exogeneity.

To ensure data integrity, we utilize historical data from 1999: specifically, the number of fixed telephones per 10,000 people (Instrumental Variable I) and the per capita volume of postal and telecommunications services (Instrumental Variable II).

The first-stage regression results confirm the relevance of the instruments. The estimated coefficients for both Instrumental Variable I and II are significantly positive at the 1% level. This indicates that regions with higher historical fixed telephone penetration and more developed postal/telecommunication services indeed exhibit stronger contemporary digital financial development, consistent with our rationale. After controlling for potential endogeneity via the IV approach, the estimated coefficients for digital financial development remain significantly positive at the 1% level (1.033 and 1.127, respectively). This confirms that digital financial development significantly promotes the enhancement of enterprise NQPFs, and our core findings hold. The first-stage F-statistic of 247.588 substantially exceeds the critical value of 16.38, robustly rejecting the presence of weak instruments. Furthermore, the over-identification tests yield p-values of 0.581 and 0.717 (both > 0.1), failing to reject the null hypothesis of valid over-identifying restrictions.

4.5. Examination of Nonlinear Effects

Following Hansen’s method [60], we investigate potential threshold effects within the panel data framework. Utilizing 500 bootstrap replications, the statistical analysis (summarized in

Table 5. Panel threshold test results.

	Threshold Variable DF	Threshold Variable SA	Threshold Variable RD	Threshold Variable GF
Threshold value $q$	1.562	0.513	0.722	0.866
$\mathrm{D}\mathrm{F}\times \mathrm{I}\left(\mathrm{T}\mathrm{h}\le \mathsf{\beta }\right)$	0.866 *** (1.099)	0.266 *** (1.338)	0.401 *** (2.009)	0.557 *** (2.186)
$\mathrm{D}\mathrm{F}\times \mathrm{I}\left(\mathrm{T}\mathrm{h}>\mathsf{\beta }\right)$	1.011 *** (2.003)	0.899 *** (2.066)	1.006 *** (1.998)	0.865 *** (1.811)
Control variables	YES	YES	YES	YES
Fixed effect	YES	YES	YES	YES
N	14,369	14,369	14,369	14,369
Constant terms	7.669 *** (3.01)	4.193 *** (1.26)	9.166 *** (2.36)	9.003 *** (1.88)
R2	0.433	0.507	0.399	0.551

The standard deviation of firm cluster robustness is shown in parentheses; *** indicate significance levels of 1%.

) reveals significant single-threshold effects for several key variables: digital financial development (DF), financing constraints (SA), technological innovation (RD), and green transformation awareness (GF).

When digital financial development (DF) serves as the threshold variable, its impact on enhancing the new productive capabilities of resource-based enterprises exhibits a significant positive effect. Crucially, this positive effect intensifies progressively as the level of DF advances, with the coefficient increasing from 0.866 below the threshold to 1.011 above it. Similarly, when financing constraints (SA), technological innovation (RD), or green transformation awareness (GF) are employed as threshold variables, the positive impact of digital financial development on the new productive capabilities of resource-based enterprises consistently strengthens beyond the identified thresholds. This pattern consistently reveals a nonlinear characteristic characterized by increasing marginal returns.

4.6. Mediation Effect Analysis

Table 6. Mechanism analysis results.

	Financing Constraints		Technological Change		Green Transition Awareness
	SA	NQPF	RD	NQPF	GF	NQPF
	I	II	III	IV	V	VI
DF	0. 266 *** (6.66)	0.439 *** (2.61)	0.337 ** (4.00)	0.501 *** (2.77)	0.422 ** (3.50)	0.609 ** (1.96)
M		0.047 (0.46)		0.013 * (1.77)		0.036 * (1.65)
Control variables	YES	YES	YES	YES	YES	YES
Fixed effect	YES	YES	YES	YES	YES	YES
N	14,369	14,369	14,040	14,040	14,310	14,310
R2	0.886	0.803	0.758	0.796	0.553	0.809

The standard deviation of firm cluster robustness is shown in parentheses; *, **, and *** indicate significance levels of 10%, 5%, and 1%, respectively.

presents the mediation analysis results: Columns (I), (III), and (V) correspond to the regression results for Model (2), while Columns (II), (IV), and (VI) present the results for Model (3).

4.6.1. Mediating Role of Financing Constraints

Table 6, Column (I) shows a significantly positive coefficient (0.266, p < 0.01) for digital financial development (DF), indicating its effectiveness in easing financing constraints for resource-based enterprises. In Column (II), the DF coefficient remains significantly positive (0.439, p < 0.01), while the coefficient for the mediating variable (financing constraints, M) is insignificant (0.047). Bootstrap testing confirmed that the product of coefficients β₁θ₂ is significantly non-zero at the 99% confidence level. Given that θ₁, β₁, and θ₂ all share a positive sign, financing constraints act as a significant partial mediator between DF development and the new quality productivity (NQPF) of these firms.

This mediation arises because DF development enhances environmental information disclosure by resource-based enterprises. This reduces information asymmetry and corporate fraud, lowering perceived risk for external investors. This effect is particularly strong for state-owned resource enterprises benefiting from implicit government guarantees. These firms find it advantageous to exploit resources while cultivating a ‘green’ image. The resulting favorable perception facilitates access to financing through commercial credit and equity markets. Improved financing sustains operations and attracts greater external investment and collaboration. These conditions collectively enhance talent acquisition, accelerate digital transformation, boost total factor productivity, and ultimately foster NQPF development.

4.6.2. Mediating Role of Technological Change

Table 6, Column (III) shows a significantly positive DF coefficient (0.337, p < 0.05), confirming digital finance development spurs technological transformation in resource-based enterprises. In Column (IV), the DF coefficient (0.501, p < 0.01) remains significant, while the technological change mediator (M) coefficient is also significant (0.013, p < 0.10). This demonstrates that technological transformation partially mediates the relationship. Thus, digital finance development promotes NQPF formation partly by enhancing technological innovation in these firms.

4.6.3. Mediating Role of Green Transition Awareness

Column (V) reveals a significantly positive DF coefficient (0.422, p < 0.05), indicating digital finance development stimulates internal reflection on green, low-carbon, and environmental imperatives within resource-based enterprises. Column (VI) shows both the DF coefficient (0.609, p < 0.05) and the green transition awareness mediator (M) coefficient (0.036, p < 0.10) are significant. This confirms green transition awareness acts as a significant partial mediator with a positive effect on the DF-NQPF relationship.

Mechanism: Digital finance development mechanisms—such as pre-loan risk assessments, post-loan risk management, and market signaling—amplify external creditors’ demands for corporate environmental, social, and green transition performance. Consequently, managers are compelled to overcome short-term “green evasion” behavior driven by sunk costs and compliance concerns. These fosters heightened organizational awareness of environmental, social, and green transition responsibilities, creating an endogenous driver for NQPF development in resource-based enterprises.

4.7. Heterogeneity Analysis

Heterogeneity analysis of Model (1) was conducted across three dimensions.

(1)

Regional Marketization Levels: Using the China Provincial Marketization Index, enterprises were grouped into high, medium, and low marketization regions. Results (

Table 7. Heterogeneity analysis results.

	The Degree of Marketization		Nature of Property Rights		Life Cycle
	NQPF	NQPF	NQPF	NQPF	NQPF	NQPF	NQPF
	Lower	Higher	SOE	N-SOE	Growth Period	Maturity	Recession Period
DF	0.084 *** (2.66)	0.043 *** (2.58)	0.936 *** (−0.077)	0.558 *** (−0.095)	0.022 (1.07)	0.060 *** (2.63)	0.225 ** (2.10)
N	2562	11,868	5110	3629	6581	5720	2024
R2	0.891	0.875	0.14	0.051	0.911	0.924	0.903
Fixed effect	YES	YES	YES	YES	YES	YES	YES

The standard deviation of firm cluster robustness is shown in parentheses; **, and *** indicate significance levels of 5%, and 1%, respectively.

) show the impact of digital finance on enterprise total factor productivity (TFP) varies significantly: strongest and most significant in low-marketization regions, insignificant in medium-marketization regions, and significant but weaker in high-marketization regions. This pattern likely arises because advanced regions offer robust environmental regulations and diverse funding sources, diminishing digital finance’s relative impact. Conversely, enterprises in less marketized regions face greater environmental, logistical, and financial constraints, making them more responsive to digital finance development.

(2)

Ownership Structure: Grouping enterprises by ownership (state-owned vs. non-state-owned) reveals that digital finance has a stronger impact on state-owned enterprises (SOEs) (Table 7). This suggests SOEs possess greater responsiveness and execution capability regarding national digital finance initiatives. Key factors include their heightened national mission, social responsibility, established policy response mechanisms enabling swift adaptation, and greater external scrutiny (government, public, stakeholders), motivating them to demonstrate positive outcomes and foster productivity gains.

(3)

Enterprise Lifecycle Stage: Aligning with mineral resource exploitability, enterprises were classified into growth, maturity, and decline stages [61]. The impact of digital finance on TFP is strongest for enterprises in decline, moderate in maturity, and weakest in the growth stage (Table 7). Growth-stage enterprises prioritize economic performance over environmental concerns and have weaker profitability, reducing their responsiveness. Mature enterprises enjoy stronger profitability and stability but exhibit innovation inertia, balancing social reputation and environmental policy, leading to a moderate impact. Decline-stage enterprises face acute resource depletion, environmental remediation demands, operational decline, and intense pressure for transformation, making them most susceptible to digital finance development.

5. Discussion

This study fundamentally repositions digital finance as a multi-dimensional catalyst for New Quality Productivity (NQPF) in resource-based enterprises (RBEs), extending beyond its conventional conceptualization as a transactional efficiency tool. Our findings reveal a paradigm shift: digital finance functions as a systemic enabler that orchestrates synergistic alignment between environmental regulation and financial resource allocation—a mechanism previously underexplored in sustainable industrial transition literature. This resolves a critical gap in understanding how technologically integrated finance can structurally reconfigure resource-intensive sectors toward high-quality development. The resilience of this core relationship across robustness tests provides compelling firm-level validation for FinTech’s transformative potential within green finance frameworks, challenging the prevailing view of digital finance as merely optimizing existing processes.

The decomposition of digital finance yields nuanced theoretical implications. The primacy of coverage breadth underscores inclusive access as a foundational equity imperative, countering historical financial exclusion of smaller or remote RBEs—a finding that significantly advances regional convergence theory. The U-shaped relationship observed in digitization depth offers a pivotal qualification to linear digitization narratives: initial efficiency losses from advanced infrastructure adoption (e.g., blockchain, AI) reflect organizational absorption costs, yet surpassing this threshold unleashes exponential gains through data leverage and innovation recombination. This empirically validates Teece’s dynamic capability theory in digital transition contexts, suggesting that RBEs must develop absorptive capacity before capturing digital dividends.

Mechanistically, our analysis elucidates why digital finance uniquely addresses sustainable productivity bottlenecks. Its capacity to alleviate “double friction” (costly and difficult financing) transcends traditional credit markets by reducing information asymmetry—directly resolving Knightian uncertainty problems that deter green investments. More profoundly, by embedding ESG criteria into financial architecture (e.g., via green fintech platforms), digital finance institutionalizes environmental consciousness within corporate governance. This transforms compliance from a cost center into a strategic vector for innovation, echoing Porter’s hypothesis at the financial system level. The observed technological transformation pathway further demonstrates how data-driven capital allocation enables anticipatory adaptation—allowing RBEs to proactively upgrade resource efficiency rather than reactively mitigate environmental liabilities.

A theoretically significant contribution is the identification of nonlinear acceleration dynamics. The incremental marginal effect beyond critical thresholds in mediating factors (financing access, technological capability, environmental awareness) signifies the emergence of synergistic tipping points. This phenomenon aligns with Arthur’s complexity economics: reduced financial friction lowers innovation risk, advanced technologies unlock new solution spaces, and heightened environmental cognition fosters organizational resilience—collectively forming a self-reinforcing cycle that exponentially amplifies NQPFs. Such threshold effects necessitate reconceptualizing digital finance impacts through phase transition models rather than linear scaling assumptions.

Heterogeneity patterns provide critical theoretical refinements. The amplified effect in low-marketization regions demonstrates digital finance’s institutional substitution function—compensating for weak formal governance through technological intermediation, thus offering empirical support for Acemoglu’s institutional arbitrage framework. SOEs’ heightened responsiveness reveals how policy-aligned incentives can overcome inherent inefficiencies when coupled with market discipline mechanisms, refining theories of state-owned enterprise reform. Most notably, recession-stage RBEs’ exceptional transformative capacity exemplifies constraint-induced innovation, where existential pressures catalyze radical resource recombination through digital finance—a finding that extends the boundary conditions of organizational resilience theory.

Collectively, these insights reconfigure sustainable transition frameworks in three dimensions:

(1)

Theoretical: Digital finance emerges as a meta-governance mechanism coordinating environmental and financial systems, transcending its instrumental role.

(2)

Geographical: Its institutional substitution capability offers a transferable model for resource-dependent economies seeking leapfrog development.

(3)

Temporal: Threshold-dependent acceleration redefines transition timelines, emphasizing strategic patience during capability-building phases.

Policy implications must prioritize institutional innovations: Regulators should develop integrated digital platforms that embed environmental risk pricing into financial transactions, while incentivizing green fintech through sandbox frameworks. Concurrently, capacity-building programs must help RBEs navigate the U-curve of digitization, particularly SMEs in institutionally weak regions.

Limitations related to NQPF measurement and timeframe invite future research. Advancing real-time sustainability metrics and exploring nonlinear interactions with policy shocks (e.g., carbon pricing) would deepen dynamic analysis. Nevertheless, this study establishes digital finance as the linchpin of a new sustainability paradigm—one where financial technology, environmental governance, and industrial capability converge to transform resource constraints into innovation catalysts.

6. Conclusions and Policy Implications

6.1. Conclusions

This study establishes that digital finance development significantly elevates the New Quality Productivity (NQPF) of resource-based enterprises (RBEs) through a synergistic mechanism integrating environmental regulation with financial resource allocation. The expansion of digital financial coverage breadth emerges as the fundamental driver of this enhancement. By incentivizing authentic environmental information disclosure and mitigating corporate short-termism, digital finance concurrently signals viable pathways for technological transformation. This dual impetus encourages firms to develop green technologies and optimize management processes, effectively counterbalancing environmental compliance costs. Simultaneously, digital finance reduces transaction costs for clean investments while accentuating the complexity and risks of polluting investments, thereby dynamically adjusting the implicit costs of environmental damage. Such adjustments promote the internalization of negative externalities and optimize financial resource allocation efficiency.

The synergistic effect materializes through three interconnected pathways. First, through alleviating financing constraints via reduced information asymmetry and enhanced investor confidence, digital finance facilitates capital access through instruments like trade credit. This enables investments in digital transformation and specialized talent acquisition. Second, by compelling firms to respond to policy mandates and societal expectations, it drives technological transformation that breaks organizational inertia and activates a “change compensation” effect through green innovations. Third, institutional mechanisms such as preload environmental risk assessments and post-loan dynamic monitoring elevate green transition awareness. These mechanisms shift managerial focus from short-term compliance evasion toward strengthened environmental responsibility, cultivating endogenous drivers for NQPF advancement.

Critically, the positive impact of digital finance on RBE NQPF demonstrates nonlinear acceleration. When the alleviation of financing constraints, momentum for technological transformation, and green awareness surpass critical thresholds, digital finance’s marginal effect intensifies markedly, exhibiting self-reinforcing dynamics aligned with network effects. Significant heterogeneity further characterizes these effects. The enhancement proves stronger in regions with lower marketization levels, where underdeveloped traditional finance and environmental governance amplify digital finance’s substitutive role. State-owned enterprises experience heightened benefits due to their comparative advantage in policy responsiveness. Most notably, recession-stage RBEs—burdened by resource depletion and environmental pressures—demonstrate greater capacity to convert digital finance into an innovation accelerator than firms in maturity or growth stages, consistent with innovation-under-constraint theory. These findings collectively underscore digital finance’s core catalytic function in enabling sustainable transformation under acute operational and environmental duress.

6.2. Policy Implications

Based on the empirical findings of this study within the Chinese context, policymakers in similar resource-dependent economies, particularly China, should prioritize the strategic expansion of digital financial coverage breadth as a core lever to enhance the New Quality Productivity (NQPF) of resource-based enterprises (RBEs). First, regulatory frameworks must incentivize authentic environmental information disclosure by integrating digital finance platforms with real-time environmental data monitoring systems. This synergy can mitigate corporate short-termism and signal viable pathways for green technological transformation, enabling firms to offset compliance costs through optimized management and innovation. Second, financial authorities should design targeted mechanisms—such as reduced transaction fees for clean investments and enhanced risk-pricing models reflecting environmental performance for polluting assets—to internalize environmental externalities and redirect capital toward sustainable projects. Future research could specifically investigate the impact of digital finance on credit terms and interest rates to inform the development of more granular risk-pricing tools.

To amplify the synergistic effect between environmental regulation and financial resource allocation, three intervention pathways require institutional reinforcement:

(1)

Alleviating financing constraints through digital credit infrastructure (e.g., blockchain-enabled trade credit networks) to support RBE investments in digital tools and specialized talent.

(2)

Accelerating mandatory technological transformation by linking digital finance access to phased environmental benchmarks, leveraging policy pressure to break organizational inertia and activate “change compensation” via green innovation.

(3)

Embedding environmental accountability across corporate governance by mandating pre-loan ecological risk assessments and robust post-loan monitoring mechanisms (potentially leveraging AI). These measures, supported by the study’s findings on information disclosure and incentive alignment, can shift managerial incentives from compliance evasion to endogenous green transition. Based on the observed impact of digital finance on mitigating information asymmetry and facilitating targeted green investment, financial institutions are encouraged to explore the development of dynamic credit monitoring frameworks that incorporate environmental performance metrics.

Crucially, policymakers must recognize the nonlinear, threshold-dependent nature of digital finance’s impact. Scaling initiatives should focus on regions with low marketization levels within China where digital finance can substitute for weak traditional institutions, utilizing state-owned enterprises (SOEs) as policy-responsive “first movers” to demonstrate viability. Additionally, recession-stage RBEs—facing acute resource depletion—should receive prioritized access to transition-focused digital finance instruments, as their innovation elasticity under duress aligns with constrained innovation theory. Such stratified interventions can harness network effects to trigger self-reinforcing NQPF growth, ultimately catalyzing sustainable industrial restructuring under environmental and operational pressures in the Chinese resource sector.