Next Article in Journal
The Frequency Spectrum Analysis of Wideband Oscillations of Grid-Connected Voltage Source Converter
Next Article in Special Issue
Digital Literacy, Farmers’ Income Increase and Rural Internal Income Gap
Previous Article in Journal
Effect of Psychological Meaningfulness on Job Involvement, Proactive Behavior, and Performance: Focusing on the Mediating Effect of Self-Efficacy
Previous Article in Special Issue
Building a Super Smart Nation: Scenario Analysis and Framework of Essential Stakeholders, Characteristics, Pillars, and Challenges
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Financial Inclusion through Digitalization: Improving Emerging Drivers of Industrial Pollution—Evidence from China

1
Post Graduate Centre, Management and Science University, Shah Alam 40100, Malaysia
2
School of Economics and Trade, Henan University of Technology, Zhengzhou 450001, China
*
Authors to whom correspondence should be addressed.
Sustainability 2023, 15(13), 10203; https://doi.org/10.3390/su151310203
Submission received: 24 May 2023 / Revised: 19 June 2023 / Accepted: 23 June 2023 / Published: 27 June 2023
(This article belongs to the Special Issue Digital Society/Society 5.0 and Sustainable Development)

Abstract

:
As an emerging product of the coupling of digital technique and traditional finance, digital inclusive finance (DIF) may play a vital role in alleviating the contradiction between economic growth and environmental contamination. This paper utilises the panel data from various provinces in China as a sample to empirically test the effect of DIF on industrial pollution. The study found that (1) DIF and its sub-dimension coverage (DIF_B) and depth of use (DIF_D) have significant governance effects on industrial pollution, and the conclusion remains valid even when endogeneity is considered; (2) the mediation effect test found that the upgrading of the industrial structure and the degree of technological innovation are important transmission paths for DIF to reduce industrial pollution; (3) the heterogeneity test found that the effect of DIF on industrial pollution control successively showed a pattern of weakening in the centre, eastern, and western regions, while the treatment effect of DIF on industrial wastewater is better than that of industrial waste gas, and the effect on industrial solid pollutant emissions has a U-shaped non-linear relation that is first suppressed and then promoted; (4) the threshold effect test found that DIF, DIF_B, and DIF_D all have a double threshold effect on industrial pollution. Based on the empirical outcomes, this paper proposes measures to improve the development mechanism of DIF, formulate differentiated monetary support and oversight policies under local conditions, and build and enhance the supervision mechanism of the digital financial industry and prevent systemic risks.

1. Introduction

From the end of the 20th century to the present, tremendous advances in global economic and social development have been made. However, the extensive development pattern has also had a serious effect on the natural environment on which human beings depend. Human health problems and economic losses caused by environmental pollution have become the focus of attention worldwide. Therefore, ecological environmental governance has also attracted profound attention from governments and people around the world. China, as the second largest economy in the world, has rapidly advanced its industrialisation process since its reform and opening up and has swiftly transformed from an undeveloped farming country to an industrial country [1]. At the same time, China’s economy grows steadily every year and maintains an average annual growth rate of 9.5%. Such a rapid and sustained increase is rare in the history of human economic development. The remarkable achievements in the process of industrialisation and economic development are inevitably causing serious environmental pollution problems, and industrial pollution is the main source of current environmental pollution [2,3]. According to the “China Environmental Statistical Yearbook” and the data released by the National Bureau of Statistics of China, Figure 1 shows the changes in GDP and industrial solid waste generation in China from 2000 to 2019. It can be seen from Figure 1 that while China has made great achievements in economic development, it is inevitably accompanied by the problem of aggravated industrial pollution. Understanding how to achieve the coordination and unity of economic development and ecological environment is an emerging demand for China’s pollution-reduction work in the new era and the new stage and is also an unavoidable choice for China’s economic transformation and development within the ‘14th Five-Year Plan’ period.
Existing research shows that factors affecting industrial pollution emissions include openness, economic scale, industrial structure, environmental regulation, technological progress, industrial agglomeration and financial development [4,5]. Finance is not only the core element of economic operation and social production but also an important regulatory tool to save energy and reduce emissions from government departments. Consequently, scholars have been studying the association between finance and the environment for a long time. Some scholars believe that the financial system realises the coordination and unity of economic and social development and environmental conservation through the optimisation of allocating resources and technological progress [6]. However, the essential characteristics of traditional finance ‘dislike the poor and love the rich’, and the information dissymmetry between the offer and need sides makes it difficult for startups and other poor groups to obtain financial resources [7], which in turn makes it difficult for the energy-saving and emission-reduction effects of finance to fully play their role. Nonetheless, with the profound coupling of digital technology and traditional finance, DIF has emerged, which has largely improved the limitations of finance on environmental governance. The characteristics of the low cost, wide coverage, and high efficiency of DIF have lowered the financing threshold for startups and overcome the disadvantages of traditional financial transactions such as high transaction costs and insufficient supply of financial resources. At the same time, DIF covers the ‘service blind spots’ of traditional finance, which is conducive to reshaping the economic production pattern; this scenario promotes green technology innovation and offers new possibilities for encouraging the coordinated development of the economy and the environment [8,9,10]. Few studies are related to DIF and industrial pollution. Consequently, studying the relationship between DIF and industrial pollution is significant for promoting environmental governance, building an ecologically civilised society, and promoting economic and social transformation in an environmentally friendly direction. To clarify the internal relationship between DIF and industrial pollution, this paper further tests the relationship between DIF and industrial pollution through empirical research based on theoretical analysis. The study found that DIF has a significant governance effect on industrial pollution, and there are intermediary transmission mechanisms and non-linear effects.
The novelties of the article include the following: (1) this paper uses China’s provincial panel data as a sample to empirically test the impact of DIF on industrial pollution; (2) taking the upgrade of industrial structure and technological innovation as mediating variables, this paper analyses the role of DIF in decreasing industry pollution discharges and conducts empirical tests; and (3) based on traditional linear regression, the non-linear relationship between DIF and industrial pollution is tested through the threshold effect.
The rest of this paper consists of four parts. The second part includes a literature review and research hypothesis. The third part includes data sources, variable measurement, and model specification. Section 4 presents the empirical results. The fifth part focuses on the results and countermeasures.

2. Literature Review and Research Hypothesis

Theoretical research on environmental problems and economic growth began in the 1960s. Owing to the sharp decrease in the stock of earth resources at that time, humanity’s living environment was severely damaged, which triggered two environmental revolutions. In the first environmental revolution, people debated the relationship between ambient quality and economic development. Economists led by Meadows published a report called ‘Limits to Growth’, which found that exponential increases in population growth, food supply, capital investment, environmental pollution, and resource depletion can stall economic growth and that technological progress can only ease the time to reach the limits. The second environmental revolution raised the issue of sustainable development and proposed many reports related to sustainable development, which prompted economists to study and discuss the circumstance for the coordinated development of the economy and environment. The two environmental revolutions encouraged economists to discuss and research the relationship between the economy and the environment on a theoretical level in general. Empirical research comes later than theoretical research. In the early 1990s, some scholars conducted an empirical study on the relationship between the economy and the environment and discovered that the relationship between many environment pollution indicators and income per person showed a reversed ‘U-shaped’ curve; that is, the ambient quality worsened first and then enhanced with economic development [11,12,13]. Panayotou (1993) found that this curve is similar to the Kuznets curve hypothesis proposed by Kuznets (1955); thus, the curve was termed the ‘environment Kuznets curve (EKC) hypothesis’ for the first time. The latter is also known as the EKC curve [14].
Finance was included in environmental-related research, which was also initially developed around the EKC hypothesis [15,16]. With the deepening research, its content gradually reduces the influence of the EKC hypothesis and pays extra attention to the relationship between finance and the environment. Currently, three main viewpoints exist on the correlation between finance and the environment. The first viewpoint is that financial development contributes to enhancing environmental quality. Some scholars believe that finance can enhance environmental quality by promoting technological progress [17], adjusting the industry structure, alleviating the shortage of environmental protection funds for enterprises, optimising the allocation of financial resources [18], improving the conversion rate of deposits and loans [19], and promoting green project investment and financing [20]. The second view is that financial development will aggravate environmental pollution. Some scholars believe that financial development increases energy consumption and pollution emissions by easing corporate financing constraints to expand production capacity [21,22] and easing consumer budget constraints to stimulate commodity consumption [23,24]. The third point of view is that financial growth has a reversed U-shaped non-linear relation with environmental pollution that first promotes and then inhibits it [25,26]. In recent years, with the emergence of inclusive finance, researchers have studied the relationship between finance and the environment from the viewpoint of financial resource equality and found that inclusive finance can affect environmental pollution emissions through economic growth and technological innovation mechanisms [27]. Meanwhile, the research results show that the emission-reduction effect of inclusive finance shows threshold characteristics [28]. As a new financial product formed by the combination of financial technology and inclusive finance, DIF has been around for a short time, and relatively few studies exist on DIF and pollution emissions. However, existing research generally believes that DIF can affect pollution emissions in direct and indirect ways. In terms of direct effects, DIF is an organic integration of advanced digital technology and inclusive finance. Its essence is to transform the traditional financial system with the help of new digital technologies. DIF has realised the innovation of traditional inclusive financial products and services, giving DIF wider coverage, lower service costs, higher service efficiency, and more accurate service targets. According to the transaction cost theory, the renewal of and improvement in transaction methods caused by technological progress have greatly improved transaction efficiency. At the same time, it effectively reduces search costs, negotiation costs, information transmission costs, supervision costs, and breach of contract costs in the transaction process. As a technological revolution in the financial field, DIF has provided many conveniences for enterprises and consumers in terms of reducing transaction costs and breaking through geographical distance restrictions and reducing various inefficient market transactions. Based on the above theoretical logic that DIF can reduce transaction costs, it can be found that DIF itself has green attributes and is environmentally friendly. Some scholars believe that DIF is a significant path to enhance financial inclusion, which has an optimistic effect on economic development and people’s lives [29,30]. DIF has improved green economic benefits by advocating the concept of low-carbon life, raising public consciousness of environmental protection, and building an environmental protection service platform [31]. At the same time, DIF uses digital techniques to effectively alleviate the problem of information asymmetry between the offer and need sides of financial institutions and enterprises, alleviate the misallocation of financial resources, and restrain environmental pollution emissions [32]. Based on the above analysis, this paper proposes the following hypothesis:
Hypothesis 1 (H1). 
The development of DIF has a significant governance effect on industrial pollution discharges.
In terms of indirect effects, some scholars believe that DIF can reduce pollution emissions through industrial structure upgrades [33], innovation effects [34], and entrepreneurial effects [35]. Specifically, the industrial structure upgrading path of DIF includes three aspects [36]. On the one hand, DIF improves the scope and efficiency of financial services, effectively reduces the problem of financial exclusion, optimises the allocation of resources, and guides the flow of factors to high-productivity industries, thereby encouraging the upgrade of industry structure; on the other hand, DIF promotes the growth of the digital economy, development of logistics, transportation, and other service industries through online consumption and industrial transformation and upgrading. Simultaneously, DIF promotes the upgrade of the consuming structure by easing the consumption budget constraints of low- and middle-income groups, thereby encouraging the upgrade of industry structure. Compared with the secondary industry, the service industries have the peculiarities of low investment and low pollution. Thus, industrial transformation and upgrading can help reduce pollution emissions. Some scholars believe that the innovation effect, the emission reduction path, of DIF occurs by alleviating the external financing constraints of technological innovation enterprises [37] and reducing the financial resource misallocation problem caused by financial exclusion and information asymmetry of startups [38] to promote technological innovation of enterprises and reduce pollution emissions. On the basis of the above analysis, this article puts forward the following hypotheses:
Hypothesis 2 (H2). 
The development of DIF will help upgrade the industry structure, thereby reducing industrial pollution emissions.
Hypothesis 3 (H3). 
DIF contributes to enhancing the level of technological innovation, thereby reducing industrial pollution emissions.

3. Data and Empirical Model

3.1. Data Sources

According to China’s current standards for dividing provincial administrative regions, there are 34 provincial administrative regions in China. However, due to the serious lack of data in Taiwan Province, Tibet Autonomous Region, Hong Kong Special Administrative Region, and Macao Special Administrative Region, the sample data of four administrative regions were excluded. This paper uses the panel data of 30 provinces (autonomous regions) in China from 2011 to 2019. The original data come from China Education Statistical Yearbook, China Statistical Yearbook, China Financial Yearbook, China Environmental Statistical Yearbook, Provincial Statistical Yearbook, and wind database. For very few missing data, the interpolation method was used for completion. DIF adopts the ‘Peking University Digital Financial Inclusion Index (2011–2020)’ issued by the Numeric Finance Research Centre of Peking University. To avoid the influence of heteroscedasticity, the DIF index and its three sub-indicators are all processed by logarithm.

3.2. Variable Measurement

Dependent variable: Industrial Pollution Index (IP). The commonly used measures of industrial pollutant discharge level are the sum of industry pollution emissions and the industrial pollution emissions strength [39]. This study cites the research methods of some scholars [40]; the total industrial pollutant discharge is utilised to evaluate the extent of regional industrial pollutant emissions. From the three dimensions of the industry’s ‘three wastes’ (exhaust gas, wastewater and solid waste), we construct the industrial pollutant emission level indicator system as shown in Table 1. The industrial pollutant discharge index is calculated using the entropy weight method.
Independent variable: digital inclusive finance index (DIF). DIF includes three dimensions: DIF_B, DIF_D, and the degree of digitisation (DIG), and each dimension is measured using multiple indicators. Consequently, it systematically and comprehensively reflects the development level of China’s DIF. DIF not only reflects the inclusiveness of digital finance but can also effectively reflect the impact and trend changes in digital technology on financial development, which is representative and authoritative.
Mediating variables: (1) industrial structure upgrading (IS). The upgrade of industry structure refers to the change and improvement in industry development from low end to high end. This paper draws on existing research methods, and the upgrading of the industry structure is characterised by the proportion between the tertiary and secondary sectors of industry. The higher the ratio, the better the degree of regional industry structure transformation and upgrading [41]. (2) Level of technological innovation (RD). Two methods exist for measuring technological innovation capability in the academic circles: one uses technical innovation input as the input variable [42], and the other takes the number of applications for patents as the output variable [43]. Referring to Sánchez-Sellero et al., this study uses the R&D investment strength as a substitute variable for the degree of scientific and technological innovation. The level of regional scientific and technological innovation is positively correlated with the intensity of R&D investment; the higher the intensity of R&D investment, the higher the level of innovation in the region [44].
Control variables: (1) Human capital level (EDU): Not only can the level of human capital enable enterprises to adopt more advanced manufacturing technologies and environmentally friendly technology to decrease pollution discharges, but a high level of human capital and strong environmental awareness can also promote energy conservation and emission reduction in enterprises. This study uses the mean years of education of the people in each area to measure the level of human capital. The mean number of years of education is calculated by multiplying the proportion of the population at each educational stage by the number of years of education and then adding up the average years of each educational stage. (2) Government support (GS): GS is expressed by the percentage of local fiscal environmental preservation spending to general budget spending. (3) Foreign direct investment (FDI): On the one hand, FDI may enhance production efficiency and decrease pollution discharges by introducing advanced production and clean technologies. On the other hand, it may also increase the industrial pollutant discharge of the host country through the international transfer of the production process of polluting products. The calculation of the FDI indicator is expressed by the percentage of regional overseas direct investment in GDP. (4) Financial development level (finance): Industrial pollution control requires a substantial amount of financial support, and regional financial development can have an impact on the local industrial pollution emission level through corresponding financial support. This study chooses the proportion of the surplus of deposits and loans of banking institution to the local GDP to measure the level of financial development in the region. (5) Marketisation degree (Market): The level of marketability affects the effect of industrial agglomeration. The better the level of industrial agglomeration, the higher the possibility of industrial pollution emissions. This paper uses the proportion of staff in non-nationalised business to measure the degree of marketisation in different regions. Table 2 shows the variable descriptive statistics.

3.3. Model Specification

According to theoretical analyses and hypotheses, this study sets the following benchmark model to investigate the direct control impact of DIF development on industrial pollution:
I P i j t = β 0 + β 1 D I F j t + β j X i t +   year   + city + ε i t
I P i j t = β 0 + β 1 D I F _ B j t + β j X i t +   year   + city + ε i t
I P i j t = β 0 + β 1 D I F _ D j t + β j X i t +   year   + city + ε i t
I P i j t = β 0 + β 1 D I G j t + β j X i t +   year   + city + ε i t
In Equations (1)–(4), IP represents the industrial pollution degree of the explained variable, and the specific calculation method is described below. DIF represents the development index of digital financial inclusion, DIF_B represents the coverage of DIF, DIF_D represents the depth of use of DIF, and DIG represents the extent of digitisation. X i t represents the series of control variables. ε i t is the disturbance term.
To effectively explore the possible indirect mechanism of DIF in industrial pollution control, based on the above analysis, it is verified whether industrial structure upgrading (IS) is an intermediary variable between the two. The test method is as follows: Based on the fact that the estimated coefficient β 1 of the benchmark regression model (1) of DIF on industrial pollution emissions is significant, we establish a linear regression Equation (5) of DIF on the intermediary variable industrial structure upgrade (IS). Simultaneously, the regression Equation (6) of DIF and the upgrade of the industry structure (IS) on industrial pollution emissions are constructed. We determine whether a mediating effect exists on the basis of the significance of the regression estimates of α 1 , γ 1 , and γ 2 . The mediation effect test of the level of technological innovation is the same as the above-mentioned industrial structure upgrade. To this end, this paper establishes the following mediation effect model:
I S i j t = α 0 + α 1 D I F j t + α j X i t +   year   + city + ε i t
I P i j t = γ 0 + γ 1 D I F j t + γ 2 I S i j t + γ j X i t +   year   + city + ε i t
This study considers the possible non-linear relation between DIF and industry pollution and the possible interval heterogeneity of the effect of DIF development on industrial pollution discharges. To avoid the subjective prejudice of the model estimation outcomes due to the artificial division of the region, this study refers to the research method of Hansen [45] in carrying out the threshold effect regression to determine the accurate threshold value. The panel threshold model is constructed as follows.
I P i t = + 1 D I F i t I ( Q < γ 1 ) i , t + 2 D I F i t I ( γ 1 Q < γ 2 ) i , t + + n D I F i t I ( Q γ n ) i , t + n + 1 c o n t r y i , t + ε i , t
where Q is the threshold variable and I() is the indicator function. If the expression in the brackets is true, then, I() takes 1; otherwise, 0; , …, is the threshold value of Q to be estimated. The rest of the variables are identical to the previous one.

4. Empirical Results and Analysis

The correlation test results of the main variables are shown in Table 3. The dependent variable industrial pollution index and DIF were significantly negative at the 1% level, which preliminarily verified the correlation between DIF and industrial pollution. At the same time, there is a significant correlation between the industrial pollution index and all control variables, indicating that the selection of control variables is scientific, and the correlation coefficients among other variables are all less than 0.5, indicating that the model does not have the problem of multicollinearity.

4.1. Effect of DIF on Industrial Pollution Discharges

Based on the econometric model constructed above, this study analyses the governance impact of DIF on industrial pollution emissions. Commonly used panel fitting models in research are mixed least squares (0LS), random effects (REs), and fixed effects (FEs). To improve the robustness of the regression results, this study uses three models. Table 4 indicates the test outcomes, indicating that the regression estimation coefficients of DIF on industry pollution emissions in the three equations are all significantly negative, and the conclusions of the three models are consistent regardless of whether control variables are included or not. This shows that the development of DIF has a salient repressive effect on industrial pollution discharges. According to the regression results of the fixed effects model, column (6) in Table 4 shows that for each unit enhancement in the development level of DIF, the level of industry pollution discharges decreases by 0.062 units, which strongly demonstrates the governance effect of DIF on industrial pollution emissions, thus proving Hypothesis 1.
With respect to the control variables, the human capital level and financial development level have a significant negative correlation with industrial pollution at a 1% level. The regression coefficient of overseas direct investing is significantly positive, and the regression coefficient of marketisation degree is significantly positive at 10%. The outcomes indicate that both human capital and financial development can significantly inhibit industrial pollution emissions, whilst the level of overseas direct investing and marketisation can promote industrial pollution emissions. The estimated coefficient of government support for industrial pollution emissions is negative but insignificant, probably because the structure of environmental protection expenditures is unreasonable and the effect on industrial pollution control is not obvious.

4.2. Impact of DIF Sub-Indicators on Industrial Pollution Emissions

DIF includes three sub-indicators: DIF_B, DIF_D, and DIG. Referring to the research methods of some scholars, this study utilises the three sub-indicators of DIF to further study the governance effect of DIF on industry pollution discharges. The fixed-effects model was chosen as the research model, and Table 5 presents the regression outcomes. Columns (1)–(3) in Table 4 are the regression outcomes of DIF_B, DIF_D, and DIG on industry pollution discharges, respectively. The outcomes show that the DIF_B and DIF_D have a notable governance impact on industry pollution discharges, and the DIF_D is the major motive power for industrial pollution governance. The estimated coefficient of DIG on industrial pollution emissions is a negative but unremarkable value. The cause may be that the DIF_B reflects the coverage of financial services, which allows underdeveloped regions to access financial services. For the individualised financial needs of enterprises, the DIF_B cannot be better satisfied. DIF_D reflects the extent to which financial functions are exerted, which can serve the needs of enterprises more deeply, to better exert the environmental governance effect. Therefore, the emission reduction effect of DIF_D is better than that of DIF_B. DIG reflects the low threshold and low-cost advantages of DIF, mainly for the convenience of financial services. Enterprises are less sensitive to the DIG. Thus, the effect on industrial emission reduction is insignificant.

4.3. Mediation Effect Analysis

We now verify the mediating effect of industry structure upgrade and technological innovation in the suppression of industry pollution discharges by DIF. In this paper, DIF is utilised as an explanatory variable, and the upgrade of industry structure and the degree of technological innovation are used as intermediary variables to conduct empirical tests. Table 6 presents the test outcomes. Columns (1)–(3) of Table 6 are the test outcomes of industrial structure upgrading, and columns (4)–(6) of Table 6 are the examination results of the degree of scientific and technological innovation. Columns (1) and (4) demonstrate the inhibitory effect of DIF on industrial pollution discharges. The outcomes of columns (2) and (5) indicate that the effect of DIF on the upgrade of industry structure and scientific and technological innovation are both significantly positively correlated at the 1% level, showing that DIF has an important role in encouraging the upgrade of industry structure and scientific and technological innovation. Column (3) in Table 6 examines the effect of DIF and the upgrade of industry structure on industrial pollution discharges. The outcomes indicate that when considering the upgrade of industry structure, the estimated coefficient of DIF on industry pollution emissions is a remarkably negative value, and the absolute figure of the coefficient is reduced compared with column (1) in Table 6. The outcomes indicate that the upgrade of industry structure has a mediating effect on the suppression of industry pollution emissions by DIF. Column (6) in Table 6 examines the effect of DIF and technology innovation on industrial pollution emissions. The outcomes indicate that when technology innovation is considered, the estimated coefficient of DIF on industrial pollution emissions is a remarkably negative value, and the absolute figure of the coefficient is reduced compared with column (4). The outcomes present that the degree of the technological invention has a mediating effect on the suppression of industrialised pollution emissions by DIF. To sum up, DIF can efficiently decrease industrialised pollution discharges by promoting the upgrade of industry structure and technological innovation. The empirical results verify Hypotheses 2 and 3.

4.4. Robustness Test

To ensure the robustness of the empirical outcomes, this paper uses four main methods for robustness testing. Firstly, we exclude the sample of municipalities directly under the Central Government. Considering that the growth rates of DIF in the four municipalities of Beijing, Shanghai, Chongqing, and Tianjin are quite different from those of other provinces in China, such differences may affect the empirical results. Therefore, utilising a sample that excludes four cities, the fixed effect model is utilised for testing. The regression outcomes are presented in column (1) of Table 7. Excluding the sample of municipalities directly under the central government, the results remain stable. Secondly, we replace the explained variable. This study cites the study of Wen et al. using PM2.5 as a proxy variable for industrial pollution emissions [33]. The regression outcomes are presented in column (2) of Table 7. After replacing the explained variables, the estimated coefficient of DIF on industrial pollution emissions remains remarkably negative, which once again confirms the robustness of the outcomes. Thirdly, we determine the independent variable with a lag of one period. Nonetheless, this paper selects more control variables to ease the endogeneity problem caused by the neglect of significant variables when analysing the relationship between DIF and industrial pollution emissions. However, the model setting still needs to face the simultaneous endogeneity problem that the two are mutually causal. To eradicate the endogeneity problem due to two-way causality, this study selects the independent variables with a lag of one period for regression. The regression outcomes from column (3) of Table 7 indicate that the independent variables with a lag period also have a remarkable negative effect on industrial pollution emissions. Therefore, the robustness of results is further demonstrated. Fourthly, we utilise the instrumental variable method. To further settle the endogeneity problem, this study cites the existing research methods and uses the proportion of the number of Internet users to the people at the end of the year from 2002 to 2010 to obtain the regional Internet penetration [46]. The growth of the Internet is a prerequisite for the progress of DIF. However, the growth of the Internet in previous years has had no impact on industrial pollution at the current stage, thereby satisfying the conditions of instrumental variables. As presented by the outcomes of columns (4) and (5) in Table 7, the estimated value of DIF is remarkably negative after the introduction of instrumental variables. This finding shows that DIF has a significant inhibitory impact on industrial pollution discharges. The robustness of the results is verified.

4.5. Heterogeneity Analysis

4.5.1. Regional Heterogeneity Test

The above conclusions confirm that DIF has a salient governance role in industrial pollution. However, owing to China’s vast geographical area, the level of economic development in different regions is not balanced, and the level of environmental governance and DIF development is different. Therefore, errors may occur when only the regression results of the national sample are considered to make a conclusion. To this end, this paper divides each province in the country into three areas, namely eastern, central, and western, based on China’s criteria for regional division, and conducts empirical tests afterwards. The outcomes are shown in columns (1)–(3) in Table 8. The outcomes indicate that the impact of DIF on industry pollution is remarkably negative in eastern, central, and western China, once again proving that DIF has a salient governance effect on industry pollution. By comparing the outcomes, DIF was found to have the strongest impact on industrial pollution control in the middle region, and it is weakened in the eastern and western regions. The cause may be that the central region is dominated by the energy industry and resource-based industries, and the ecological environment is relatively fragile. By optimising quality of service, DIF can significantly suppress industrial pollution emissions. As far as the eastern region is concerned, it is mostly the knowledge-intensive and capital-intensive industries, such as high-tech industries, and the environment is relatively good. Thus, the effect of DIF on industrial pollution control is slightly weaker than that in the midland. The industries in the West are relatively backward, and most of them are energy and resource industries. Polluting industries in developed regions are also considered. However, owing to the relatively backward development level of DIF in the West, the consequence of DIF on industrial pollution control is relatively weak.

4.5.2. Heterogeneity of Pollutants

To investigate whether differences exist in the governance effects of DIF on different industrial pollutants, this paper divides industrial pollutants into three categories: industry exhaust pollutants, industrial effluent pollutants, and industrial solid pollutants. According to the category of industrial pollutants, the sample data were subsequently tested empirically, and columns (4)–(6) of Table 8 present the empirical outcomes. Columns (4) and (5) of Table 8 show that the impact of DIF on industrial waste gas pollutants and industrial wastewater pollutants is remarkably negative, which indicates that the development of DIF can significantly inhibit the discharge of industrial waste gas and wastewater, and the treatment effect of industrial wastewater is better than that of industrial waste gas. Column (6) in Table 8 shows that the impact of DIF on industrial solid pollutants is remarkably positive, showing that DIF has a facilitating impact on industrial solid pollutant emissions. To further study the effect of DIF on industrial solid pollutant emissions, the square term of DIF was added to the original model, and the regression outcomes are presented in column (7) in Table 8. Column (7) in Table 8 shows that the effect of DIF on industrial solid pollutants has a positive U-shaped non-linear relationship that first inhibits and then promotes industrial solid pollutants. The reason may be that the effect of the growth of DIF on the emission of solid pollutants of enterprises includes two aspects. On the one hand, DIF can expand the production scale of enterprises and increase the emission of solid pollutants by easing financing constraints. On the other hand, DIF can offer financial support for enterprises to control solid pollutants. When the governance role is better than the scale effect, DIF has an inhibitory effect on the discharge of solid pollutants. When the scale effect is greater than the governance role, DIF has a promoting effect on solid pollutant emissions.

4.6. Threshold Effect Test

The above panel data model regression results confirm that DIF and its sub-dimensions DIF_B and DIF_D have a linear inhibitory effect on industrial pollution emissions. We examine whether a non-linear association exists between DIF and its sub-dimensions on industry pollution discharges due to the different development levels of DIF. In this paper, DIF is utilised as the threshold variable to examine the panel threshold model. Table 9 presents the examined outcomes. The regression outcomes in Table 9 indicate that the DIF index has passed the double threshold effect test. DIF_B and DIF_D passed the single threshold effect test at the 1% level and the double threshold effect test at the 5% level. Therefore, DIF, DIF_B, and DIF_D have a double threshold effect. Table 10 reports the threshold values with DIF, DIF_B, and DIF_D as threshold variables, respectively.
Table 9 and Table 10 show that DIF and its sub-dimensions DIF_B and DIF_D have a threshold effect on industry pollution discharges, and Table 11 reports the threshold regression outcomes. The regression outcomes of column (1) in Table 11 present that the governance effect of DIF on industrial pollution emissions divided into three stages. When LnDIF ≤ 5.3735, DIF has a certain repressive effect on industrial pollution emissions, but the impact is not obvious. When 5.3735 < LnDIF ≤ 5.5273, the inhibitory role of DIF on industry pollution discharges is significantly improved. When LnDIF > 5.5273, the inhibitory effect of DIF on industry pollution discharges is further enhanced. Therefore, the impact of DIF on industrial pollution reduction is enhanced with the development of DIF. It can be seen from column (2) in Table 11 that the DIF_B regression outcomes are the same as the above conclusions. When DIF_B is less than the low threshold value, the impact of DIF_B on reducing industrial pollution is insignificant. As the DIF_B index increases, the effect of reducing industrial pollution is gradually enhanced. In terms of the DIF sub-dimension DIF_D, the governance effect of DIF_D on industrial pollution emissions can be divided into three stages. When the DIF_D index is less than 2.5463, DIF_D significantly inhibits industrial pollution emissions; when the DIF_D index is between 2.5463 and 5.2894, the effect of DIF_D on industrial pollution control is significantly weakened; when the DIF_D index is greater than 5.2894, the effect of DIF_D on industrial pollution control is again improved.

5. Research Conclusions and Recommendations

As a new form of finance formed by the coupling of digital technique and inclusive finance, DIF provides new ideas and opportunities for using financial tools to control industrial pollution emissions. This study utilises the panel data of various provinces in China from 2011 to 2019 to empirically test the governance effect, heterogeneity characteristics, and transmission mechanism of DIF on industrial pollution discharges through multiple linear regression, the mediation effect model, and the panel threshold model. The empirical outcomes indicate that (1) DIF and its sub-indicators-DIF_B and DIF_D have significant governance effects on industrial pollution, and the conclusion remains valid even when endogeneity is considered; (2) the mediation effect test found that the upgrade of industry structure and the degree of technological innovation are important transmission paths for DIF to reduce industrial pollution; (3) the heterogeneity test found that the effect of DIF on industrial pollution control showed a pattern of weakening in the central, eastern, and western areas in turn. Moreover, the treatment effect of DIF on industrial wastewater is better than that of industrial waste gas, whilst the effect on industrial solid pollutant emissions has a positive U-shaped non-linear relationship that is suppressed first and then promoted; (4) threshold effect test found that DIF, DIF_B, and DIF_D all have a double threshold effect on industrial pollution. The above research conclusions not only provide ideas for China to achieve industrial pollution reduction through DIF but also have reference significance for other countries. As a result, this paper puts forward the following policy recommendations.
Firstly, all countries should continue to improve the DIF development mechanism and improve the service quality and efficiency of DIF. Although the current DIF is developing rapidly, the problem of the imbalance in regional development persists, and all countries and regions should strengthen the construction of DIF infrastructure in remote and backward areas. In addition, an information communication platform needs to be built between the financial market and startups and the environmental protection industry. Financial institutions must be encouraged to formulate diversified financial security policies, the efficiency of financial resource utilisation needs enhancement, and the emission reduction effect of DIF needs further improvement through direct and indirect means. Secondly, differentiated policies and measures must be formulated based on local conditions. Owing to the large differences in the level of industrial pollution in different countries and regions, a diversified policy plan should be in place to guide the growth of local digital finance in the direction of diversification and efficiency. For areas with relatively fragile ecological environments, support for DIF should be increased to give full play to the emission reduction effect of DIF. For regions with a better ecological environment, a sound evaluation mechanism for the environmental performance of DIF should be established, and the reduction effect of DIF should be strengthened through policy means. Finally, the supervision mechanism of the digital financial industry must be built and enhanced to prevent systemic risks. DIF has obvious advantages. However, owing to the short development time and rapid development speed, the current regulatory system cannot comprehensively prevent systemic risks. Therefore, each country and region should build a more scientific and effective regulatory system through cutting-edge technologies, and an all-weather and all-around digital financial risk prevention and control work should be carried out effectively.

Current Limitations and Future Directions of Research

The limitation of this study is that due to the availability of data, the entropy weight index of industrial pollution is relatively low, and the current comprehensive index of industrial pollution has certain limitations. In addition, the empirical data uses data from 2011 to 2019, the time interval spans only 9 years, the sample size is small, only short-term analysis can be carried out, and the research has certain limitations. In the future, researchers can consider increasing the period to explore the relationship between the two under long-term effects, or they can take other countries as the research object to further expand the research results.

Author Contributions

Conceptualisation, M.X. and W.L.; formal analysis, calculating and writing—original draft preparation, M.X.; methodology and data curation M.X.; writing—review and editing, W.L., P.W. and C.J.; supervision, W.L. All authors have read and agreed to the published version of the manuscript.

Funding

This research received no external funding.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

Most of the data generated or analysed during this study are included in this published article. The rest of the datasets used for analysis can be found in the CSMAR Database (https://www.gtarsc.com/, accessed on 10 October 2021) and the WIND Database (https://www.wind.com.cn/, accessed on 10 October 2021).

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Islam, N.; Yokota, K. Lewis growth model and China’s industrialization. Asian Econ. J. 2008, 22, 359–396. [Google Scholar] [CrossRef]
  2. Siddique, H.M.A. Industrialization, energy consumption, and environmental pollution: Evidence from South Asia. Environ. Sci. Pollut. Res. 2022, 30, 4103. [Google Scholar] [CrossRef]
  3. Zhou, Y.; Liu, Z.; Liu, S.; Chen, M.; Zhang, X.; Wang, Y. Analysis of industrial eco-efficiency and its influencing factors in China. Clean Technol. Environ. Policy 2020, 22, 2023–2038. [Google Scholar] [CrossRef]
  4. Nasrollahi, Z.; Hashemi, M.S.; Bameri, S.; Mohamad Taghvaee, V. Environmental pollution, economic growth, population, industrialization, and technology in weak and strong sustainability: Using STIRPAT model. Environ. Dev. Sustain. 2020, 22, 1105–1122. [Google Scholar] [CrossRef]
  5. Han, C.; Hua, D.; Li, J. A View of Industrial Agglomeration, Air Pollution and Economic Sustainability from Spatial Econometric Analysis of 273 Cities in China. Sustainability 2023, 15, 7091. [Google Scholar] [CrossRef]
  6. Triki, R.; Kahouli, B.; Tissaoui, K.; Tlili, H. Assessing the Link between Environmental Quality, Green Finance, Health Expenditure, Renewable Energy, and Technology Innovation. Sustainability 2023, 15, 4286. [Google Scholar] [CrossRef]
  7. Caplan, M.A.; Birkenmaier, J.; Bae, J. Financial exclusion in OECD countries: A scoping review. Int. J. Soc. Welf. 2021, 30, 58–71. [Google Scholar] [CrossRef]
  8. Hasan, M.M.; Yajuan, L.; Khan, S. Promoting China’s inclusive finance through digital financial services. Glob. Bus. Rev. 2022, 23, 984–1006. [Google Scholar] [CrossRef]
  9. Sun, Y.; Tang, X. The impact of digital inclusive finance on sustainable economic growth in China. Financ. Res. Lett. 2022, 50, 103234. [Google Scholar] [CrossRef]
  10. Chopra, S.; Dwivedi, R.; Sherry, A.M. Leveraging technology options for financial inclusion in India. Int. J. Asian Bus. Inf. Manag. 2013, 4, 10–20. [Google Scholar] [CrossRef]
  11. Grossman, G.M.; Krueger, A.B. Environmental Impacts of a North American Free Trade Agreement; NBER: Cambridge, UK, 1991. [Google Scholar]
  12. Shafik, N.; Bandyopadhyay, S. Economic Growth and Environmental Quality: Time Series and Cross Country Evidence; Background Paper for the World Development Report; The World Bank: Washington DC, USA, 1992. [Google Scholar]
  13. Yang, R.; Wong, C.W.; Miao, X. Evaluation of the coordinated development of economic, urbanization and environmental systems: A case study of China. Clean Technol. Environ. Policy 2021, 23, 685–708. [Google Scholar] [CrossRef]
  14. Panayotou, T. The economics of environments in transition. Environ. Dev. Econ. 1999, 4, 401–412. [Google Scholar] [CrossRef]
  15. Dinda, S. Environmental Kuznets curve hypothesis: A survey. Ecol. Econ. 2004, 49, 431–455. [Google Scholar] [CrossRef] [Green Version]
  16. Jalil, A.; Feridun, M. The impact of growth, energy and financial development on the environment in China: A cointegration analysis. Energy Econ. 2011, 33, 284–291. [Google Scholar] [CrossRef]
  17. Hu, Y.; Dai, X.; Zhao, L. Digital finance, environmental regulation, and green technology innovation: An Empirical Study of 278 Cities in China. Sustainability 2022, 14, 8652. [Google Scholar] [CrossRef]
  18. Gu, B.; Chen, F.; Zhang, K. The policy effect of green finance in promoting industrial transformation and upgrading efficiency in China: Analysis from the perspective of government regulation and public environmental demands. Environ. Sci. Pollut. Res. 2021, 28, 47474–47491. [Google Scholar] [CrossRef]
  19. Tamazian, A.; Chousa, J.P.; Vadlamannati, K.C. Does higher economic and financial development lead to environmental degradation: Evidence from BRIC countries. Energy Policy 2009, 37, 246–253. [Google Scholar] [CrossRef]
  20. Shahbaz, M.; Hye, Q.M.A.; Tiwari, A.K.; Leitão, N.C. Economic growth, energy consumption, financial development, international trade and CO2 emissions in Indonesia. Renew. Sustain. Energy Rev. 2013, 25, 109–121. [Google Scholar] [CrossRef] [Green Version]
  21. Cole, M.A.; Elliott, R.J. FDI and the capital intensity of “dirty” sectors: A missing piece of the pollution haven puzzle. Rev. Dev. Econ. 2005, 9, 530–548. [Google Scholar] [CrossRef]
  22. Shahbaz, M.; Solarin, S.A.; Mahmood, H.; Arouri, M. Does financial development rEDUce CO2 emissions in Malaysian economy? A time series analysis. Econ. Model. 2013, 35, 145–152. [Google Scholar] [CrossRef] [Green Version]
  23. Zhang, Y.J. The impact of financial development on carbon emissions: An empirical analysis in China. Energy Policy 2011, 39, 2197–2203. [Google Scholar] [CrossRef]
  24. Sadorsky, P. The impact of financial development on energy consumption in emerging economies. Energy Policy 2010, 38, 2528–2535. [Google Scholar] [CrossRef]
  25. Charfeddine, L.; Khediri, K.B. Financial development and environmental quality in UAE: Cointegration with structural breaks. Renew. Sustain. Energy Rev. 2016, 55, 1322–1335. [Google Scholar] [CrossRef]
  26. Abbasi, F.; Riaz, K. CO2 emissions and financial development in an emerging economy: An augmented VAR approach. Energy Policy 2016, 90, 102–114. [Google Scholar] [CrossRef]
  27. Lv, J.; Wang, N.; Ju, H.; Cui, X. Influence of green technology, tourism, and inclusive financial development on ecological sustainability: Exploring the path toward green revolution. Econ. Res. 2022, 36, 1–23. [Google Scholar] [CrossRef]
  28. Jiang, S.; Qiu, S.; Zhou, J. Re-Examination of the Relationship between Agricultural Economic Growth and Non-Point Source Pollution in China: Evidence from the Threshold Model of Financial Development. Water 2020, 12, 2609. [Google Scholar] [CrossRef]
  29. Risman, A.; Mulyana, B.; Silvatika, B.; Sulaeman, A. The effect of digital Finance on financial stability. Manag. Sci. Lett. 2021, 11, 1979–1984. [Google Scholar] [CrossRef]
  30. Dwivedi, R.; Alrasheedi, M.; Dwivedi, P.; Starešinić, B. Leveraging financial inclusion through technology-enabled services innovation: A case of economic development in India. Int. J. E-Serv. Mob. Appl. 2022, 14, 1–13. [Google Scholar] [CrossRef]
  31. Li, G.; Fang, X.; Liu, M. Will Digital Inclusive Finance Make Economic Development Greener? Evidence From China. Front. Environ. Sci. 2021, 9, 452. [Google Scholar] [CrossRef]
  32. Shi, F.; Ding, R.; Li, H.; Hao, S. Environmental Regulation, Digital Financial Inclusion, and Environmental Pollution: An Empirical Study Based on the Spatial Spillover Effect and Panel Threshold Effect. Sustainability 2022, 14, 6869. [Google Scholar] [CrossRef]
  33. Wen, H.; Yue, J.; Li, J.; Xiu, X.; Zhong, S. Can digital Finance rEDUce industrial pollution? New evidence from 260 cities in China. PLoS ONE 2022, 17, e0266564. [Google Scholar] [CrossRef] [PubMed]
  34. Xue, L.; Zhang, X. Can Digital Financial Inclusion Promote Green Innovation in Heavily Polluting Companies? Int. J. Environ. Res. Public Health 2022, 19, 7323. [Google Scholar] [CrossRef] [PubMed]
  35. Chang, J. The role of digital finance in reducing agricultural carbon emissions: Evidence from China’s provincial panel data. Environ. Sci. Pollut. Res. 2022, 29, 87730–87745. [Google Scholar] [CrossRef] [PubMed]
  36. Li, F.; Wu, Y.; Liu, J.; Zhong, S. Does digital inclusive finance promote industrial transformation? New evidence from 115 resource-based cities in China. PLoS ONE 2022, 17, e0273680. [Google Scholar] [CrossRef] [PubMed]
  37. Lin, B.; Ma, R. How does digital finance influence green technology innovation in China? Evidence from the financing constraints perspective. J. Environ. Manag. 2022, 320, 115833. [Google Scholar] [CrossRef]
  38. Wang, X.; Zhu, Y.; Ren, X.; Gozgor, G. The impact of digital inclusive finance on the spatial convergence of the green total factor productivity in the Chinese cities. Appl. Econ. 2022, 35, 1–19. [Google Scholar] [CrossRef]
  39. Xiaoyu, L.; Jun, Y.; Pengfei, S. Structure and application of a new comprehensive environmental pollution index. Energy Procedia 2011, 5, 1049–1054. [Google Scholar] [CrossRef] [Green Version]
  40. Dong, F.; Pan, Y.; Li, Y.; Zhang, S. How public and government matter in industrial pollution mitigation performance: Evidence from China. J. Clean. Prod. 2021, 306, 127099. [Google Scholar] [CrossRef]
  41. Xue, L.; Li, H.; Xu, C.; Zhao, X.; Zheng, Z.; Li, Y.; Liu, W. Impacts of industrial structure adjustment, upgrade and coordination on energy efficiency: Empirical research based on the extended STIRPAT model. Energy Strategy Rev. 2022, 43, 100911. [Google Scholar] [CrossRef]
  42. Wu, N.; Liu, Z. Higher education development, technological innovation and industrial structure upgrade. Technol. Forecast. Soc. Change 2021, 162, 120400. [Google Scholar] [CrossRef]
  43. Hu, G.G. Is knowledge spillover from human capital investment a catalyst for technological innovation? The curious case of fourth industrial revolution in BRICS economies. Technol. Forecast. Soc. Change 2021, 162, 120327. [Google Scholar] [CrossRef]
  44. Sánchez-Sellero, P.; Bataineh, M.J. How R&D cooperation, R&D expenditures, public funds and R&D intensity affect green innovation? Technol. Anal. Strateg. Manag. 2022, 34, 1095–1108. [Google Scholar] [CrossRef]
  45. Hansen, B.E. Threshold effects in non-dynamic panels: Estimation, testing, and inference. J. Econom. 1999, 93, 345–368. [Google Scholar] [CrossRef] [Green Version]
  46. Zhong, S.; Xin, C.; Shen, H.; Chen, X. Effects of land urbanization and internet penetration on environmental sustainability: A cross-regional study of China. Environ. Sci. Pollut. Res. 2021, 28, 66751–66771. [Google Scholar] [CrossRef]
Figure 1. Changes in GDP and industrial solid waste generation in China from 2000 to 2019.
Figure 1. Changes in GDP and industrial solid waste generation in China from 2000 to 2019.
Sustainability 15 10203 g001
Table 1. Indicators of Industrial Pollutant Emission Levels.
Table 1. Indicators of Industrial Pollutant Emission Levels.
First-Level IndexesSecond-Level IndexesIndicator DefinitionAttribute
Industry exhaustSO2Industrial sulphur dioxide emissions (tons)+
S&DIndustrial smoke and dust emissions (tons)+
Industrial wastewaterCODIndustrial Chemical Oxygen Demand Emissions (tons)+
NH3-NIndustrial ammonia nitrogen emissions (tons)+
Industrial solid wasteSolid WasteIndustrial general solid waste discharge (10,000 tons)+
Table 2. Descriptive statistics of variables.
Table 2. Descriptive statistics of variables.
VariableMeanStandard DeviationMinMaxSample Size
Dependent variableIP0.2390.1610.001190.770270
Explanatory variablesDIF5.1510.6702.9096.017270
DIF_B4.9950.8270.6735.952270
DIF_D5.1330.6461.9116.087270
DIG5.4580.7162.0266.136270
Mediating variablesIS1.2920.7120.5275.234270
RD1.5921.1300.1906.315270
Control variablesEDU9.1940.8887.51412.80270
GS0.03000.009630.01180.0681270
FDI2.1221.913012.10270
Finance3.2510.9931.6787.035270
Market0.5740.1510.2440.866270
Table 3. The results of the correlation test of the variables.
Table 3. The results of the correlation test of the variables.
IPLnDUFEDUGSFDIFinanceMarket
IP1
LnDUF−0.359 ***1
EDU−0.299 ***0.07901
GS−0.125 **0.05900.03601
FDI0.417 ***−0.0360−0.108 *−0.223 ***1
Finance−0.26 2***0.314 ***−0.02300.269 ***0.155 **1
Market−0.180 ***0.257 ***0.499 ***−0.0330−0.09900.04201
*, **, *** indicate significance at the 10%, 5%, and 1% levels, respectively, whereas the brackets represent the t values.
Table 4. Baseline estimation results of the impact of DIF on industrial pollution emissions.
Table 4. Baseline estimation results of the impact of DIF on industrial pollution emissions.
VariablesOLSREFE
(1)(2)(3)(4)(5)(6)
LnDIF−0.086 ***−0.062 ***−0.100 ***−0.065 ***−0.101 ***−0.062 ***
(−6.30)(−5.03)(−12.79)(−5.61)(−12.94)(−4.05)
EDU −0.05 1*** −0.073 *** −0.085 ***
(−5.17) (−3.77) (−2.83)
GS 1.307 −0.839 −1.165
(1.52) (−1.13) (−1.50)
FDI 0.038 *** 0.029 *** 0.027 ***
(8.88) (7.06) (6.26)
Finance −0.046 *** −0.055 *** −0.069 ***
(−5.27) (−3.52) (−3.16)
Market 0.093 0.200 * 0.266 *
(1.55) (1.88) (1.86)
Constant0.683 ***1.009 ***0.754 ***1.269 ***0.758 ***1.390 ***
(9.61)(9.90)(16.29)(8.02)(18.74)(5.71)
Observations270270270270270270
R-squared0.1290.413 0.4120.561
F39.7530.84 167.449.90
* and *** indicate significance at the 10%, and 1% levels, respectively, whereas the brackets represent the t values; the same is true below.
Table 5. Regression results of DIF sub-dimensions on industrial pollution emissions.
Table 5. Regression results of DIF sub-dimensions on industrial pollution emissions.
Variables(1)(2)(3)
IPIPIP
LnDIF_B−0.026 **
(−2.19)
LnDIF_D −0.069 ***
(−4.87)
LnDIG −0.018
(−1.59)
EDU−0.103 ***−0.068 **−0.119 ***
(−3.35)(−2.24)(−3.99)
GS−1.081−1.119−0.885
(−1.36)(−1.47)(−1.11)
FDI0.028 ***0.026 ***0.028 ***
(6.37)(6.22)(6.23)
Finance−0.089 ***−0.068 ***−0.099 ***
(−4.10)(−3.27)(−4.73)
Market0.0730.2010.021
(0.52)(1.62)(0.15)
Constant1.538 ***1.304 ***1.706 ***
(6.11)(5.40)(7.27)
Observations270270270
R-squared0.5400.5740.536
F45.7752.5144.97
**, *** indicate significance at the 5%, and 1% levels, respectively, whereas the brackets represent the t values.
Table 6. Regression results of mediation effect.
Table 6. Regression results of mediation effect.
VariablesIndustrial Structure UpgradingLevel of Technological Innovation
IP(1)IS(2)IP(3)IP(4)RD(5)IP(6)
LnDIF−0.062 ***0.086 ***−0.047 ***−0.062 ***0.147 ***−0.047 ***
(−4.05)(3.09)(-3.17)(−4.05)(4.91)(−2.98)
IS −0.171 ***
(−5.03)
RD −0.101 ***
(−3.11)
EDU−0.085 ***0.275 ***−0.038−0.085 ***0.023−0.083 ***
(−2.83)(4.98)(−1.27)(−2.83)(0.38)(−2.80)
GS−1.1658.226 ***0.242−1.1656.708 ***−0.486
(−1.50)(5.80)(0.31)(−1.50)(4.39)(−0.61)
FDI0.027 ***−0.021 ***0.023 ***0.027 ***0.0060.028 ***
(6.26)(−2.65)(5.63)(6.26)(0.74)(6.52)
Finance−0.069 ***0.276 ***−0.021−0.069 ***−0.015−0.070 ***
(−3.16)(6.97)(−0.94)(−3.16)(−0.34)(−3.29)
Market0.266 *−0.3060.2140.266 *0.1840.285 **
(1.86)(−1.17)(1.57)(1.86)(0.65)(2.03)
Constant1.390 ***−2.609 ***0.944 ***1.390 ***0.3841.426 ***
(5.71)(−5.86)(3.80)(5.71)(0.77)(5.95)
Observations270270270270270270
R-squared0.5610.7050.6040.5610.3950.577
F49.9093.3150.8249.9025.4245.45
Sobel test−0.0147 *** (z = −2.635)−0.015 *** (z = −2.625)
Intermediary effects23.9%24.17%
*, **, *** indicate significance at the 10%, 5%, and 1% levels, respectively, whereas the brackets represent the t values.
Table 7. Results of robustness tests.
Table 7. Results of robustness tests.
Variables(1)(2)(3)(4)(5)
IPPM2.5IPLnDIFIP
L.LnDIF −0.088 ***
(−5.73)
rate_inter 0.054 ***
(15.64)
LnDIF−0.065 ***−4.520 *** −0.083 ***
(−3.93)(−4.23) (−4.67)
EDU−0.071 **−5.179 **−0.085 ***0.021−0.052 ***
(−2.15)(−2.44)(−2.81)(0.58)(−5.29)
GS−1.625−281.551 ***−0.872−3.4431.259
(−1.58)(−5.17)(−1.07)(−1.12)(1.48)
FDI0.036 ***0.729 **0.026 ***−0.0070.037 ***
(6.23)(2.41)(5.72)(−0.45)(8.80)
Finance−0.072 ***−1.638−0.045 *0.078 **−0.041 ***
(−2.90)(−1.08)(−1.84)(2.55)(−4.56)
Market0.298 *18.119 *0.575 ***0.709 ***0.117 *
(1.93)(1.81)(3.34)(3.35)(1.89)
Constant1.251 ***112.646 ***1.241 ***3.373 ***1.095 ***
(4.69)(6.60)(4.73)(10.65)(9.54)
Observations234270240270270
R-squared0.5610.4560.5790.5670.407
F43.0632.6846.6857.46
*, **, *** indicate significance at the 10%, 5%, and 1% levels, respectively, whereas the brackets represent the t values.
Table 8. Results of heterogeneity test.
Table 8. Results of heterogeneity test.
Variables(1)(2)(3)(4)(5)(6)(7)
Eastern PartMiddle PartWestern PartWastegasWastewaterSolid WasteSolid Waste
LnDIF−0.074 ***−0.082 ***−0.067 ***−0.065 ***−0.092 ***0.052 ***−0.336 ***
(−3.34)(−3.59)(−3.01)(−3.35)(−3.66)(3.32)(−3.99)
LnDIF2 0.045 ***
(4.68)
EDU−0.126 ***−0.139 **−0.020−0.075 *−0.141 ***0.071 **0.027
(−3.26)(−2.22)(−0.42)(−1.94)(−2.82)(2.29)(0.87)
GS−0.0140.328−3.133 *−3.366 ***−1.468−0.093−1.223
(−0.02)(0.17)(−1.88)(−3.42)(−1.15)(−0.12)(−1.53)
FDI0.026 ***0.063 ***0.0270.005−0.000−0.0000.004
(6.88)(2.80)(1.43)(0.90)(−0.03)(−0.08)(0.87)
Finance−0.083 ***−0.165 ***0.021−0.058 **−0.126 ***−0.037 *−0.055 **
(−2.83)(−3.38)(0.68)(−2.12)(−3.52)(−1.68)(−2.55)
Market0.506 ***0.650 *0.1700.2400.119−0.108−0.228
(2.73)(1.83)(0.84)(1.32)(0.51)(−0.74)(−1.60)
Constant1.716 ***1.912 ***0.5711.414 ***2.387 ***3.168 ***4.513 ***
(5.83)(3.70)(1.42)(4.57)(5.95)(12.74)(12.09)
Observations1088181270270270270
R-squared0.7520.6570.3870.3840.5090.1800.251
F45.5221.046.95324.2840.458.56911.13
*, **, *** indicate significance at the 10%, 5%, and 1% levels, respectively, whereas the brackets represent the t values.
Table 9. Threshold effect test results.
Table 9. Threshold effect test results.
Threshold VariablesModelsF-Valuep-ValueNumber of BS1% Threshold5% Threshold10% Threshold
DIFSingle threshold106.78 ***0.00050019.79815.37913.819
double threshold29.08 ***0.00050017.22211.5659.867
three threshold14.220.32650038.08728.76424.032
DIF_BSingle threshold127.48 ***0.00050017.37113.52611.834
double threshold12.11 **0.03250014.76911.1469.505
three threshold11.240.38250026.89921.57218.915
DIF_DSingle threshold127.52 ***0.00050034.79129.22127.075
double threshold18.79 **0.02450021.04716.96814.526
three threshold19.110.67450056.15344.05538.646
The number of BS is the number of the threshold self-sampling, **, *** indicate significance at the 5%, and 1% levels, respectively.
Table 10. Estimation results of threshold values.
Table 10. Estimation results of threshold values.
Threshold VariablesModelsThreshold Estimates95% Confidence Interval
DIFSingle threshold5.37355.36855.3802
double threshold5.52735.50095.5403
DIF_BSingle threshold5.29655.28325.3192
double threshold5.46815.43695.4801
DIF_DSingle threshold2.54632.54634.1307
double threshold5.28945.28255.3038
Table 11. Threshold regression results.
Table 11. Threshold regression results.
Variables(1)(2)(3)
LnDIF (LnDIF ≤ 5.3735)−0.022 *
(−1.73)
LnDIF (5.3735 < LnDIF ≤ 5.5273)−0.036 **
(−2.68)
LnDIF (5.5273 < LnDIF)−0.048 ***
(−3.64)
LnDIF_B (LnDIF_B ≤ 5.2965) −0.007
(−0.59)
LnDIF_B (5.2965 < LnDIF_B ≤ 5.4681) −0.024 *
(−1.84)
LnDIF_B (LnDIF_B > 5.4681) −0.032 **
(−2.61)
LnDIF_D (LnDIF_D ≤ 2.5463) −0.137 ***
(−5.00)
LnDIF_D (2.5463 < LnDIF_D ≤ 5.2894) −0.045 ***
(−2.49)
LnDIF_D (5.2894 < LnDIF_D) −0.065 ***
(−4.39)
EDU0.0190.0070.003
(0.44)(0.17)(0.07)
GS1.0250.8550.470
(1.03)(0.85)(0.53)
FDI0.020 **0.020 **0.019 **
(2.62)(2.60)(2.33)
Finance−0.045−0.064 *−0.060 **
(−1.43)(−1.86)(−2.21)
Market0.1330.0310.209
(0.96)(0.21)(1.58)
Constant0.2380.3860.519
(0.62)(1.01)(1.52)
Observations270270270
R-squared0.7290.7140.742
F17.9016.1418.28
*, **, *** indicate significance at the 10%, 5%, and 1% levels, respectively, whereas the brackets represent the t values.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Xiong, M.; Li, W.; Jenny, C.; Wang, P. Financial Inclusion through Digitalization: Improving Emerging Drivers of Industrial Pollution—Evidence from China. Sustainability 2023, 15, 10203. https://doi.org/10.3390/su151310203

AMA Style

Xiong M, Li W, Jenny C, Wang P. Financial Inclusion through Digitalization: Improving Emerging Drivers of Industrial Pollution—Evidence from China. Sustainability. 2023; 15(13):10203. https://doi.org/10.3390/su151310203

Chicago/Turabian Style

Xiong, Mingzhao, Wenqi Li, Chenjie Jenny, and Peixu Wang. 2023. "Financial Inclusion through Digitalization: Improving Emerging Drivers of Industrial Pollution—Evidence from China" Sustainability 15, no. 13: 10203. https://doi.org/10.3390/su151310203

APA Style

Xiong, M., Li, W., Jenny, C., & Wang, P. (2023). Financial Inclusion through Digitalization: Improving Emerging Drivers of Industrial Pollution—Evidence from China. Sustainability, 15(13), 10203. https://doi.org/10.3390/su151310203

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop