metricas
covid
Buscar en
Journal of Innovation & Knowledge
Toda la web
Inicio Journal of Innovation & Knowledge Can Industrial Digitalization Promote Regional Green Technology Innovation?
Información de la revista
Vol. 9. Núm. 1.
(enero - marzo 2024)
Compartir
Compartir
Descargar PDF
Más opciones de artículo
Visitas
735
Vol. 9. Núm. 1.
(enero - marzo 2024)
Acceso a texto completo
Can Industrial Digitalization Promote Regional Green Technology Innovation?
Visitas
735
Xiaoli Haoa,b, Yi Lianga,c, Cunyi Yangi, Haitao Wud,e,
, Yu Haod,e,f,g,h,**
a School of Economics and Management, Xinjiang University, Urumqi 830047, China
b Center for Innovation Management Research, Xinjiang University, Urumqi 830047, China
c School of Economics, Minzu University of China, Beijing Beijing 100081, China
d School of Management and Economics, Beijing Institute of Technology, Beijing 100081, China
e Center for Energy and Environmental Policy Research, Beijing Institute of Technology, Beijing 100081, China
f Collaborative Innovation Center of Electric Vehicles in Beijing, Beijing 100081, China
g Sustainable Development Research Institute for Economy and Society of Beijing, Beijing 100081, China
h Beijing Key Lab of Energy Economics and Environmental Management, Beijing 100081, China
i Lingnan College, Sun Yat-Sen University, Guangzhou, China
Ver más
Este artículo ha recibido
Información del artículo
Resumen
Texto completo
Bibliografía
Descargar PDF
Estadísticas
Figuras (3)
Mostrar másMostrar menos
Tablas (13)
Table 1. Evaluation index system of industrial digitalization level.
Table 2. Description of control variables.
Table 3. Statistic description of variables.
Table 4. Benchmark regression results.
Table 5. Test for the threshold effect of industry digitization level.
Table 6. Threshold regression results.
Table 7. Results of the mediation mechanism test.
Table 8. Regional heterogeneity.
Table A1. Results of lag period test.
Table A2. Results of replacing dependent variable.
Table A3. The results of 2SLS regression with instrumental variable.
Table A4. The results of GMM regression with instrumental variable.
Table B. Panel threshold robustness tests.
Mostrar másMostrar menos
Abstract

Regional green technology innovation (RGTI) carries ecological and technological externalities, serving as a crucial force in aligning economic development with the imperative of environmental preservation. Industrial digitalization (indig) exhibits innovative, quality-enhancing, and efficiency-enhancing traits. Yet, the exact effects of indig on RGTI have not been extensively studied. In this study, based on provincial data from 2013-2019 in China, we investigated the linear and nonlinear relationships, spatial characteristics and underlying mechanisms between indig and RGTI through a Hansen threshold model and a mediating effect model. The results showed that: i. indig fluctuates while RGTI increases steadily, both displaying evident spatial agglomeration traits and a unipolar characteristic of regional heterogeneity; ii. indig has effectively boosted RGTI at both national and regional levels, with the exception of the northeastern region; iii. There was a single threshold for industrial digitalization and the positive driving trend exhibited a “marginally increasing” characteristic; iv. Two critical elements, namely industrial structure advancement and marketization, have partial intermediary roles. Basing on these results, this study presents several policy recommendations.

Keywords:
Industrial digitization
green technology innovation
spatial agglomeration
intermediary model
threshold effect
regional heterogeneity
JEL:
O13
O15
P28
Texto completo
Introduction

The new round of the information technology revolution, exemplified by big data and the Internet, has promoted the rapid development of new technologies, intelligent manufacturing upgrades, and other fields. Moreover, countries around the world have embarked on the digital revolution process (Shi, 2022; Xiao & Liu,2022). The digital economy is a new type of industry that utilizes data as a production factor, leverages the Internet as a platform, and integrates digital technology with other domains (Lailag & Chen, 2022; Sui & Yao, 2023). From an economic structure perspective, it consists of industrial digitalization and digital industrialization. Industrial digitalization (indig) means that digital technology is applied to traditional industries to realize the digitalization-characterized process. Digital industrialization refers to various economic activities that completely rely on digital technologies and data elements, which are collectively referred to as the information industry (Dai & Ma, 2023). Traditional industries are faced with the impact of external shocks, contradictory supply and demand structures, overcapacity, unreasonable factor structures and other difficulties. Industrial digitalization has innovation-driven, quality-enhancing and efficiency-enhancing characteristics. It is an effective way for traditional enterprises to transform, upgrade and improve resource allocation efficiency. Different types of economies, such as developed markets, emerging markets, and frontier markets, exhibit unique contexts and distinct features (Piñeiro-Chousa et al., 2018). According to the China Academy of Information and Communication Technology, in 2021, the proportion of industrial digitalization reached 37.18 trillion yuan, accounting for 81.7% of the digital economy, and 32.5% of GDP. Under the strategy of "Digital China" and "Smart Society", it is increasingly important to promote deep integration of digital technology and real economy. The Chinese economy has undergone an extensive phase of rapid growth, emphasizing the magnitude and pace of economic development, resulting in environmental degradation, elevated carbon emissions, and pollution. In response, the Chinese government introduced a new development paradigm in 2015, encompassing five key elements: "innovation, greenness, coordination, openness, and sharing." This signifies a shift for China from high-speed development to a focus on high-quality development (Hao et al., 2023c; Hao & Deng, 2019), with innovation as the first driving force and a key way to realize green development (Hao et al., 2023d; Mondalet al., 2024; Yanget al., 2024). China's clean energy continues to rapidly develop, and the energy consumption structure has continued to be optimized (Wu et al.,2022; Mao et al.,2023). However, China is still dominated by fossil energy, which accounted for about 84.3% of energy consumption in 2021(Feng et al., 2022; Hao et al.,2022). Among them, coal is the most dominant, accounting for 56% of consumed energy sources, followed by oil (18.5%). In 2020, China established the goals of "carbon peaking" and "carbon neutrality", reflecting China's determination and desire to achieve green economic transformation and high-quality development (Liu et al.,2023; Wu et al.,2024). Regional green technology innovation (RGTI) improves the economic quality (Wang et al., 2018). Particularly in the aftermath of the COVID-19 pandemic, there has been a growing endorsement of the green economy (Piñeiro-Chousa et al., 2022). The RGTI is also of significance in achieving "carbon neutrality" and "carbon peaking". It can balance the dual objectives of economic development and energy transformation, reduce the premium price of green products, and also promote sustainable development (Romero-Castro et al., 2022). Digital technology drives the transformation of production's social relations, the reengineering of those relations, and a comprehensive change in economic and social structures (Knickrehm et al., 2016). It carries significant, disruptive changes to the innovation process (Nambisan et al., 2017). Industrial digitalization is the main driving factor for digital economy development, and is an important starting point for the digital economy to empower high-quality economic development (Qiu et al., 2022). Digital transformation of industries has the ability to empower green and low-carbon development of industries. Industrial digitalization realizes deep integration of digital technology and real economy to provide technological support for RGTI.

Existing studies have primarily investigated the relationship between digital economy and technological innovation (Miller & Wilsdon, 2001; Ning et al., 2022). The impact of indig on RGTI has not been extensively investigated, raising the question: can industrial digitalization effectively promote regional green technology innovation? Complex and changing external dilemmas and internal contradictions have increased the complexity of the impact of digital transformation of traditional industries on RGTI. Industrial digitalization is innovation-driven, however, at the beginning of industrial digitalization, a large amount of capital investment is required to build digital facilities and platforms, which will delay the effects on RGTI. This paper aimed to explore strategies for optimizing the impact of industrial digitization in fostering regional green technology innovation. Concurrently, it aims to investigate its interrelation with the marketization process and industrial structure. We still need to find out the development levels, characteristics and spatial distribution trends of indig and RGTI. Industrial digitization displays distinct stage characteristics and exhibits regional variations. This paper aims to further investigate into its non-linear association with regional green technology innovation and analyze its heterogeneous effects across different regions. This paper will provide a deeper understanding of the impact of indig of traditional industries on green innovation, and promote green innovation to achieve high-quality economic development by transforming the traditional industrial development mode.

While studies have investigated the environmental impact of the digital economy and explored the implications and influencing factors of RGTI, there remains a gap in research specifically examining the nuanced sub-dimensional relationship between industrial digitalization (indig) and RGTI. This may undermine the efficacy of targeted innovation policies, management models, and green development practices. The marginal contributions are as follows: i. The overall level of indig, as well as its variations in different regions, is quantified by constructing a comprehensive evaluation system and utilizing the projection pursuit function; ii. The spatial agglomeration characteristics and regional heterogeneities of indig and RGTI are explored; iii. The linear and non-linear relationships between indig and RGTI are investigated via OLS and Hansen threshold model, as well as the special driving trend and the regional heterogeneous effects and iv. The potential intermediary mechanisms of two critical factors in the process are tested — industrial structure advancement and marketization.

The rest of the paper is presented as: Part 2 is the literature review and Hypotheses; Part 3 is the research design; Part 4 is the results and analysis; Part 5 is discussion; the last part is the conclusion and recommendation.

Literature review and hypotheses

The digital economy, a concept that was first introduced in the 1990s (Tapscott, 1996), has received a lot of attention. Scholars have studied the definition of digital economy (Mesenbourg, 2001), digital divide (Martin, 2003; Torrent-Sellens et al., 2022), and monopoly effect (Armstrong, 2006). The digital economy is closely associated with rapid progress of the digital technology, which has efficiency-enhancing and quality-enhancing characteristics. Wang et al. (2022) examined the three paths by which the Internet economy impacts eastern Central Europe: human capital, clean technology innovation, and energy structure. Cette et al. (2022) argue that the digital technology will increase labor productivity and total factor productivity of enterprises. Digital technologies can be applied to the financial sector to enable credit risk management of data (Bisht et al., 2022). The digital economy is composed of two main parts: digital industrialization and industrial digitalization. The latter focuses on digital shift or digital technology applications in existing traditional industries, which is all-round transformation of traditional industries. The main advantages of industrial digitization include accelerating economic development (Bjorkdahl, 2020), optimizing industrial environment (Verhoef & Broekhuizen, 2021), improving environmental innovation performance (Hung & Nham, 2023) and corporate performance (Heredia et al., 2022; Peng & Tao, 2022) among others. These advantages are conducive to enterprise development because data is extremely easy to gather and the marginal cost of information exchange is extremely low (Gölzer & Fritzsche, 2017; Reddy, 2023). The green effect of digitalization has been studied in terms of its influencing factors. Researchers argue that this green effect is influenced by the size of the economy (Li & Liao, 2022; Wang & Du, 2022) and the industrial chain (Zhang et al., 2022). Wang et al. (2022) argue that digital transformation significantly affects electric consumption efficiency. Yang & Yee (2022) investigated the three factors that affect process digitization (PDI): differentiation strategy, absorptive capacity, and lean manufacturing.

Innovation means constantly destroying old structures and creating new structures (Schumpeter, 1982). Romer (1986) reported that knowledge has spillover effects.

In addition to digitalization, a nation's innovation can be nurtured by various factors, including public expenditure on education, foreign direct investment (FDI), and entrepreneurial activity (López-Cabarcos et al., 2020). The digital economy enables the change from old to new dynamics, reshaping economic structures such as factor changes, and being knowledge-driven. This has laid the theoretical foundation for studies on digital economy externalities. To influence regional green technology innovation, first, digital technology creates complementarity with links in organizational chains, reconfigures factor resources, triggers production paradigm improvement as well as industry linkage effects, and promotes structural optimization of production sectors (Heo & Lee, 2019; Cao et al., 2024). Digital technology, for example big data, can promote green innovation willingness and green process innovation (Tian et al., 2022). Second, as digital infrastructure continues to improve, it promotes the optimization of the production chain (Kohli & Melville, 2019), changes the traditional information dissemination and enhances the optimization of information dissemination structures (Henfridsson & Bygstad, 2013). Studies have investigated its green effects from the perspectives of its externalities, specifically, the rationality of energy consumption (Ren et al., 2021; Hao et al., 2022; Wang et al., 2022), energy efficiency (Wu et al., 2021; Hao et al., 2023b), carbon reduction feasibility (Hao et al., 2023a; Yi et al., 2022; Zhou et al., 2022; Hao et al., 2023e; Wu et al., 2024), and increasing the export value of green products (Thanh, 2022).

In value chain promotion, the digital shift and upgrading of existing industries, driven by integration with digital technology, are conducive to RGTI in the following ways: i. Green innovation requires more involved information, multi-dimensional and multi-level knowledge synergy, and integration of technology fields. Therefore, it is difficult to carry out green innovation by relying on enterprises' original technology experience and knowledge accumulation in a single technological field (Mubarak et al., 2021; Park,2022). The indig can promote industrial chain informatization and transparency, increase labor productivity, thereby improve resource management (Lange et al., 2020), accelerate knowledge integration and reconstruction, promote cross-border industrial integration, expand cognitive domain of enterprises, and thereby promote RGTI; ii. Influenced by the green development concept, people began to practice environmental protection behaviors, which promotes the demand side to choose environment-friendly products while indig can promote the exchange between the supply and demand sides, reducing information asymmetry; iii. Digital technology applications can effectively change the input structure of production factors of enterprises, and the digital technology itself has green and innovation-driven attributes, which can empower traditional technology of enterprises, such as resource development, and business model among others. These will further reduce technical energy consumption, improve industrial access threshold, and promote RGTI. Based on the above analysis, Hypothesis 1 is derived:

Hypothesis 1. Industrial digitalization can promote the improvement of regional green technology innovation level.

Due to high growth and high marginal rewards of the digital shift, it can effectively improve production efficiency, and development of enterprise information networks, which enhances low-cost dissemination of information and information sharing of enterprises (Goldfarb & Tucker, 2019). These can eliminate the spatial limitations of enterprise technology spillover effects, which benefits technological innovations. However, due to high permeability of the digital technology, construction of a large amount of infrastructure is required at the beginning of industrial digitization. Therefore, a lag-effect in play of ecological technology benefits exists, which restricts RGTI to a certain extent. With improvement of indig and the maturation of information network, the asymmetry of outside information will be further reduced and enrichment of factor inputs will be promoted, which can assist enterprises in enhancing resource utilization cognition, broadening the green innovation of enterprises boundary. Based on the above analysis, Hypothesis 2 is derived:

Hypothesis 2. Industrial digitization has different development stages, and the marginal effect of industrial digitization is increasing.

From the supply side, indig offers high technology and high marginal rewards. Upgrading of traditional industries, such as intelligent manufacturing, makes the products more competitive, and factors of production such as technology and data flow into these industries with higher profitability, which has a "crowding out effect" on industries with a low degree of integration with digital technology, such as agriculture and light industry, and "strengthening effect" on some industries with a higher degree of integration. Moreover, traditional industrial interconnection accelerates extension, improves industrial production efficiency and complementary resources level. In turn, the overall competition level of the industry will be improved, and industrial structure optimization will be driven by labor division, model innovation, factor endowment and value transfer. In addition, indig diversifies the consumption mode, enhances intelligence and services (Zhao et al., 2022). Simultaneously, industrial structure optimization must break constraints from energy consumption, environmental and other factors. Sticking to the high-end and green direction is the internal requirement of advanced industrial structures. It will meet production side demands as well as internal and external demands of the consumption side, then continuously strengthen RGTI, and improve core competitiveness. Based on the above analysis, Hypothesis 3 can be obtained:

Hypothesis 3. Industrial digitalization can improve regional green technology innovation by upgrading the industrial structure.

With its high technology, high payoff, high integration and high synergy, indig promotes structural innovation of production factors, empowers the marketization of factors and promotes the marketization process. First, the intelligence brought about by the high technology of indig drives the demand for complex labor that cannot realize automated jobs, increases the demand for high-end labor (Staab, 2017), and facilitates production efficiency improvement. Second, high integration and synergy enhances the synergistic effects for each factor, increases the matching speed of various factors. Third, marketization enhances the accu racy of price signals and market information, guides effective allocation of innovation resources, forms an effective market incentive and constraint mechanisms, and promotes enterprises to obtain innovation resources by virtue of their own strength. Gradually, marketization eliminates market barriers, and backward regions improve RGTI by learning clean technology. Regions with higher marketization levels tend to have more active economies, which can easily attract the entry of enterprises with clean technologies (Li, 2014), expanding green technology spillover effects and promoting RGTI. Based on the above analysis, Hypothesis 4 can be obtained:

Hypothesis 4. Industrial digitalization can improve regional green technology innovation by promoting the process of marketization.

The diagram of theoretical analysis is shown in Fig. 1.

Fig. 1.

The diagram of theoretical analysis.

(0.41MB).
Research design

This section provides an explanation of the research design, incorporating definitions and descriptive statistics for variables, along with the presentation of methods and data sources.

Variable descriptionsCore variables

(1) Core dependent variable: regional green technology innovation (RGTI)

Regional Green Technology Innovation (RGTI) is a paradigm of technological innovation aimed at protecting the ecological environment. It considers sustainable economic development, advances in science and technology, and improvements in current processes to enhance resource utilization efficiency, reduce pollutant emissions, and achieve energy conservation synergy. A patent is an important criterion that reflects the original innovation level of a region or country. This indicator is widely recognized by scholars worldwide (Wagner, 2007; Wurlod & Noailly, 2018). Based on the study by Wang & Zhang (2020), we collected green technology patents granted in 30 provinces to represent the RGTI level.

Changes in RGTI and its percentage across the 4 major regions (east region, central region, west region, and northeast region) of China are presented in Fig. 2. (Specifically, east region refers to Beijing, Tianjing, Hebei, Shanghai, Jiangsu, Zhejiang, Fujian, Shandong, Guangdong, Hainan; Middle region refers to Shanxi, Anhui, Jiangxi, Henan, Hubei, Hunan; West region refers to Sichuan, Guizhou, Yunnan, Shannxi, Gansu, Qinghai, Ningxia, Xinjiang, Chongqing, Inner Mongolia, Guangxi; and northeast region refers to Jilin, Heilongjiang, Liaoning).

Fig. 2.

The number of granted patents of green innovation and the corresponding shares in the four major regions.

(0.14MB).

According to Fig. 2, for RGTI, the eastern region displays a pronounced unipolar characteristic in regional heterogeneity: the gap in RGTI compared to the other three regions is widening rapidly. The number of green patents in the eastern region accounts for more than 50% of the national level, and has an increasing tendency. This indicates that a certain spillover effect of technology exists. The technology's “inertia acceleration effect” will foster the continuous advancement of the entire region, which may cause polarization effects with time and unbalance regional development in the country. The unbalanced RGTI level and unipolar characteristic during the study period are reflected in the total number of green patents in each region: east (67011) > central (25284) > northeast (12691) > west (11697).

(2) Core independent and threshold variable: industry digitalization (indig) level

Industrial digitalization is defined as the digitalization transformation of industries, which can potentially accelerate economic development, optimize industrial environment, and improve environmental innovation performance. To accurately represent the indig level, with reference to the study by Zhao et al. (2020), 13 indicators are selected, considering the following aspects: level of digital finance, industry digital application level, and industry digital convergence level (Table 1). Given the high growth and high marginal rewards of the digital shift, industrial digitization can effectively enhance the production and communication efficiency of enterprises (Goldfarb & Tucker, 2019). Nonetheless, owing to the pervasive nature of digital technology, substantial infrastructure construction should be developed at the onset of industrial digitization. Industrial digitization exhibits distinct stage characteristics, with the initial phase demanding significant funds for digital infrastructure development. This infrastructure construction profoundly influences the widespread adoption of digital technology, consequently impacting the efficacy of industrial digitization in promoting green technology innovation. Consequently, a lag-effect is observed, leading to the selection of industrial digitization as both an independent and threshold variable. The industrial digitization level in each province, city, and autonomous region is assessed using the projection pursuit function. The specific model and data processing steps are outlined below:

Table 1.

Evaluation index system of industrial digitalization level.

Target level  Criterion level  Index level  Data source 
Industrial digitization levelDigital finance level  Provincial Digital Financial Inclusion Index  Peking University Digital Inclusive Finance Index 
Digital application level of industriesProportion of the number of enterprises engaged in e-commerce trading activities  China Statistical Yearbook 
Number of websites per 100 companies  China Statistical Yearbook 
Provincial E-commerce Development Index  China E-Commerce Development Index Report 
Number of enterprises implementing the standard of integration of informatization and industrialization  Ministry of Industry and Information Technology 
Number of Taobao villages (towns)  Taobao Village Research Report of China 
Digital convergence level of industriesOnline retail sales accounted for the proportion of total retail sales of consumer goods  China Statistical Yearbook 
Enterprise e-commerce sales  China Statistical Yearbook 
Enterprise e-commerce procurement volume  China Statistical Yearbook 
Courier volume  China Statistical Yearbook 
Express business revenue  China Statistical Yearbook 
Total exports of hi-tech enterprises  China Torch Statistical Yearbook 
Share of technology revenue in operating income in high-tech enterprises  China Torch Statistical Yearbook 

① Employ the extremum method for dimensionless processing of the 13 indicators.

② Construct the projection indicator function:

SZ denotes the standard deviation of P(u)t while Dp is the local density of P(u)t;

③ Calculate the optimal projection value in the ideal projection direction:

Considering the limitations of traditional methods in solving complex nonlinear optimization problems, an accelerated genetic algorithm (RAGA) is utilized. This algorithm optimizes the projection indicator function using MATLAB programming to achieve global optimization for high-dimensional data:

Then, the best projection value (z) is calculated, which is the indig level value.

indig levels and its changes in the four major regions are shown in Fig. 3.

Fig. 3.

Industrial digitalization level in the four major regions and on the whole.

(0.18MB).

Based on data shown in Fig. 3, the indig level is highly dynamic in the east and progresses significantly. In the other three regions, particularly the western and northeastern regions, its development is slower with a risk of stagnation. There's an imbalance in the characteristics of indig. This may be due to slow infrastructure construction and insufficient talent reserves, leading to a slowdown in the momentum of indig development, while the high industrial digitization level in the east attracts talents for clustering due to technological spillover effects and the more complete infrastructure. The unbalanced indig level during the study period is as: east (2.977) > central (2.565) > west (2.477) > northeast (2.449).

Mediating variables

(1) Advanced industrial structure (ts). First, indig has a "crowding out effect" on industries with a low degree of integration with digital technology, such as agriculture and light industry, and a "strengthening effect" on industries with a high integration degree. Second, it also facilitates structural optimization in areas such as cooperative labor division, innovation modes, factor endowments, and value transfer. Third, it can promote consumption pattern renewal. The advanced industrial structure is measured by the ratio of tertiary industry output value to secondary industry output value (Gan et al., 2011). The specific formula is as follows:

Whereby tsab denotes the advanced industrial structure level in province b in year a, tioab is the gross value of tertiary industry in province b in year a, while wioab denotes the gross value of secondary industry in province b in year a.

(2) Marketization level (mi). It is the extent to which the market influences resource allocation. At the beginning of marketization, the pursuit of economic efficiency and restricted efficiency of resource allocation affects RGTI. As the marketization level continues to increase, it will guide enterprises to efficiently allocate innovation resources, and attract enterprises with clean technology to enter (Li, 2014). This paper employs the total marketization index, considering factors like marketization extent, speed, and depth. This index was proposed by Fan Gang and has been widely adopted (Fan et al., 2011).

Control variables

The relevant control variables are described in Table 2

Table 2.

Description of control variables.

Control Variables  Definition  Reference 
Government intervention (goverThe share of fiscal spending in GDP represents the level of the government's intervention in the economy. More fiscal spending on science and technology is conducive for technological innovation. Therefore, the ratio of local government general budget fiscal expenditure to GDP represents the level of the government intervention degree (gover).  Dong & Wang(2021
Foreign direct investment (fdiFDI has a technology spillover and demonstration effect, and has a positive impact on improving innovation capability to a certain extent. The ratio of foreign direct investment to GDP represents the level of foreign direct investment (fdi) level in each region.  Zhong & Shao (2022) 
Environmental regulation(erEnvironmental regulation can improve enterprise environmental awareness and promote RGTI. Considering industrial wastewater, industrial SO2, and industrial smoke, standardize the three pollutants, find the weight for each pollutant, and the comprehensive index of environmental regulation(er)can be obtained.  Li & Du (2014) 
Financial scale (finsizeSolomon Tadesse (2002) argues that a good financial system is able to provide the technological innovation system with large-scale input financing needed for technological innovation. It promotes long-term, stable, and sustainable behaviors of technological innovation. The financial scale (finsize)is represented by the ratio of added value in the financial industry to regional GDP.  Li & Yang (2018) 
Degree of openness (open)(used in the robustness test)  Opening up to the outside world will introduce advanced technologies, play a demonstration effect role and promote technological innovation. The proportion of total import and export to regional GDP represents the level of openness (open).   
Science and technology investment (rd)(used in the robustness test)  The more investment in science and technology, the greater the possibility of RGTI. The proportion of R & D expenditure in GDP in each province represents the level of rd 
Model designPanel regression model

Based on the analyses above, we established a panel regression model to empirically verify whether indig can promote RGTI as follows.

Whereby RGTI it is the level of green technology innovation in province i in year t, indig it is the industrial digitalization level in province i in year t. Control variables affecting RGTI include goverit, fdiit, erit, and finsizeit. Where, goverit denotes government intervention, fdiit is the foreign investment level, erit denotes environmental regulation, finsizeit is the scale of financial development.

Dynamic Panel threshold model

Equation (6) explores the direct effects of indig on RGTI. Given the development stage of indig, there may be nonlinear relationships between its development and RGTI. Given the dynamic nature of industrial digitization and to mitigate estimation bias arising from endogeneity, we adopt the approach proposed by Kremer et al. (2013) by employing a dynamic panel threshold model.

Whereby γ1, γ2···γn denotes the threshold value at different levels, I is the indicative function, and indig is the threshold variable. Control variables are the same while δit are the random disturbance terms.

Mediation effect model

To investigate how indig influences RGTI, we explored whether the advanced industrial structure and marketization may act as mediating variables. According to the study by Baron & Kenny (1986), the mediation effect model is constructed based on the panel regression model as follows.

M is the mediating variable, including tsit, miit. Tsit denotes the advanced industrial structure while miit denotes marketization.

Data sources

The study is based on data for 30 provinces from 2013 to 2019 in China (Taiwan, Hongkong, Macao and Tibet are not included due to data availability). Unless specified, the data sources for this research include the China Finance Yearbook, Wind Database, China Statistical Yearbook, China Environmental Statistical Yearbook, China Urban Statistical Yearbook, China Tertiary Industry Statistical Yearbook, China Research Data Services Platform (CNRDS), as well as provincial statistical yearbooks, provincial statistical bulletins, and official websites of provincial statistical bureaus. A descriptive statistic of the variables can be found in Table 3. The results show that there are large gaps in government intervention (gover) and financial size (finsize) among others, indicating an unbalanced regional development in China.

Table 3.

Statistic description of variables.

Variables  Obs  Mean  Std.Dev.  Min  Max 
lnRGTI  210  7.772  1.273  3.689  10.44 
indig  210  2.658  0.339  2.334  4.173 
gover  210  0.267  0.111  0.120  0.658 
er  210  0.525  0.541  2.585 
finsize  210  0.0726  0.0297  0.0306  0.185 
fdi  210  0.0189  0.0174  0.000107  0.121 
ts  210  1.279  0.699  0.633  5.169 
mi  210  6.950  1.974  2.530  11.40 
lnRGTI 1  210  8.376  1.310  4.564  11.12 
open  210  0.257  0.277  0.0128  1.342 
rd  210  0.0169  0.0112  0.00458  0.0631 
Analysis and results

Estimation results of different models are presented sequentially.

Baseline regression resultsAnalysis of benchmark regression results

Based on the Hausman test to determine whether industrial digitization promotes RGTI, we employed the fixed-effects model. The results are presented in Table 4.

Table 4.

Benchmark regression results.

ln RGTI  (1)  (2)  (3)  (4)  (5) 
indig  2.0215***  1.9924***  1.9348***  1.4004***  1.4009*** 
  (5.08)  (4.97)  (4.74)  (4.04)  (4.02) 
gover    1.1455  1.3150  0.3782  0.3971 
    (0.85)  (1.04)  (0.25)  (0.26) 
er      0.2871  0.3539**  0.3497** 
      (1.66)  (2.22)  (2.14) 
finsize        15.6430***  15.3705*** 
        (3.96)  (3.32) 
fdi          -0.5954 
          (-0.21) 
_cons  2.3979**  2.1697*  2.1270*  2.6270***  2.6537*** 
  (2.26)  (2.01)  (1.94)  (3.03)  (3.09) 
N  210  210  210  210  210 
R2  0.480  0.485  0.497  0.587  0.587 

Note: t-statistic values in parentheses; *** p < 0.01, ** p < 0.05, * p < 0.1. The following tables are the same (unless otherwise noted).

In Table 4, from columns (1) to (5), with the sequential addition of control variables, the coefficients of indig on RGTI were all significant. The influence of the indig coefficient on RGTI is (1.4009), significant at level 1% with adding all control variables, indicating that industrial digitalization in China can effectively enhance RGTI, consistent with findings by Nin et al. (2023).

For the control variables, the coefficient of gover on RGTI is (0.3971), which is insignificant, indicating that even though administrative interventions can effectively guide enterprises to achieve technological direction changes, its promotion effects on RGTI are not conclusive. The coefficient of er is (0.3497), significant at the 5% level, indicating that enterprises promote green technology R&D investment to reduce environmental costs that are due to environmental regulation, which promotes RGTI, consistent with the conclusion by Tao et al. (2021). Environmental regulation will force the government to implement environmental governance by putting pressure on government officials. Besides, enterprises will also aim at energy conservation and emission reduction, and carry out technological transformation as well as innovation. The coefficient of the financial scale is significant at the 1% level, indicating that finance can serve the real economy and a good financial environment is conducive for RGTI. The coefficient of fdi is (-0.5954), which is not significant. It maybe because the introduction of foreign capital results in technological dependence and is harmful for progression of RGTI.

Threshold test and analysis of industrial digitalization

(1) Threshold test

Considering the phased characteristics of indig development, there might be a nonlinear relationship with RGTI, as seen in Table 5.

Table 5.

Test for the threshold effect of industry digitization level.

Number of thresholdsF-statistic valueP-valueThreshold valueNumber of Sampling
10%  5%  1% 
Industrial digitization (indigSingle Threshold  35.62⁎⁎⁎  0.0000  11.3615  13.2229  19.0610  600 
  Double threshold  9.17  0.1317  10.0495  12.0353  16.3915  600 
  Triple threshold  4.49  0.8250  14.1259  15.8463  19.6503  600 

Single threshold is significant, which was selected to report the regression results.

(2) Analysis of threshold results

Single threshold regression results are presented in Table 6.

Table 6.

Threshold regression results.

  indig 
  indig<2.35  indig>2.35  l.green  gover  er  finsize  fdi 
Coefficient  0.2543*  0.4357⁎⁎⁎  0.7276⁎⁎⁎  1.5213⁎⁎⁎  0.0943  3.7041*  -0.6183 
T-value  1.70  2.98  15.05  3.00  1.25  1.93  -0.42 

In Table 6, there is a single threshold effect between indig and RGTI. The threshold value for indig is 2.35. The indig coefficients show that the driving trend has a “marginally increasing positive” characteristic. (1) When indig < 2.35, the indig coefficient is significantly positive (0.2543) at 1% level. Due to infrastructure construction of digital technology and lag effects of ecological technology benefits, promotion of indig on RGTI cannot be fully realized. (2) When indig > 2.35, impact coefficient of indig on RGTI is significantly positive (0.4357) at 1% level. Upon reaching maturity, industrial digitalization, accompanied by fully established infrastructure, significantly diversifies the types of production factors and optimizes their structures. Importantly, the promotional impact of industrial digitization on RGTI becomes more pronounced.

Further analysisAnalysis of mediating effects

The transmission mechanism in the process is analyzed in part 3. The development of indig can promote RGTI and broaden cognitive boundaries of enterprises by industrial structure advancement (ts) and marketization (mi). The mediation effect model is used to examine the potential mechanism (Table 7).

Table 7.

Results of the mediation mechanism test.

  modelmodelmodelmodel4  model5 
  ln RGTI  ts  ln RGTI  mi  ln RGTI 
indig  1.401⁎⁎⁎(8.26)  0.4424⁎⁎⁎(5.03)  1.011⁎⁎⁎(6.25)  1.749⁎⁎⁎(9.08)  0.7995⁎⁎⁎(4.21) 
ts      0.8818⁎⁎⁎(6.79)     
mi          0.3439⁎⁎⁎ 
          (5.60) 
Control variable  YES  YES  YES  YES  YES 
Province FE  YES  YES  YES  YES  YES 
R-squared  0.9545  0.9593  0.9640  0.9756  0.9614 
Obs.  210  210  210  210  210 
Sobel test  Z=4.041, P=0.0000Z=4.765, P=0.0000
Proportion of total effect that is mediated 0.27850.4293
Bootstrap Test  Z=2.82, P=0.005Z=4.53  P=0.000 

Model (1) of Table 7 shows the direct positive effects of indig on RGTI, and in model (2), the coefficient of indig on ts is (0.4424), which is significantly positive at the 1% level, proving that industry digitalization may enhance industrial structure advancement. Model (3) adds ts into model (1), and the indig coefficient is still significant (1.011), with some decrease compared with model (1), but positive at 1% level. Accordingly, the mediating effect size of ts is 0.390=1.401-1.011, and accounts for 27.85% of total effect, indicating that indig improves RGTI by enhancing advanced industrial structure.

Model (4) confirms that the indig coefficient on mi is (1.5054), significantly positive at 1% level, proving that industry digitization can promote marketization. Model (5) adds the mediating variable of marketization level into model (1), and the indig coefficient is still (0.7995), which is significantly positive at 1% level, with obvious decreases compared with model (1). Accordingly, the mediating effect size of mi is 0.6015=1.401-0.7995, accounting for 42.93%, indicating that indig can enhance RGTI by promoting marketization.

Regional Heterogeneity Analysis

Due to factors such as resource endowments and historical influences, the levels of industrial digitization (indig) exhibit variation among different regions in China. To investigate potential spatial heterogeneity in the impact of indig on RGTI, four major regions are taken into account: the east region, central region, west region, and northeast region. The results are in Table 8.

Table 8.

Regional heterogeneity.

  East  Central  West  Northeast 
  lnRGTI  lnRGTI  lnRGTI  lnRGTI 
indig  1.0730***  3.4832***  3.2047***  1.7822* 
  (5.39)  (9.91)  (3.24)  (4.27) 
gover  0.2601  2.2996*  0.5648  -1.2721 
  (0.16)  (2.22)  (0.33)  (-0.75) 
er  0.0887  0.1937  0.7549***  0.5914* 
  (0.39)  (0.80)  (5.07)  (3.50) 
finsize  9.5965*  9.6535  14.5214**  16.3257* 
  (1.98)  (1.26)  (2.82)  (3.70) 
fdi  -4.9465*  -10.3881  -18.5149  5.9922* 
  (-1.95)  (-0.76)  (-0.98)  (2.99) 
_cons  4.6272***  -1.8295  -2.3029  1.9274 
  (8.02)  (-1.24)  (-0.90)  (2.43) 
N  70  42  77  21 
R2  0.731  0.837  0.698  0.887 

In Table 8, it is shown that indig in the eastern region promotes RGTI at 1% level, but with the lowest coefficient. This may be because the eastern region has a higher level of industrial digitalization and is in the exploration period, which has a certain blindness for digital inputs (Griliches et al., 1991). Indig development will also intensify the financial burden and cause resource mismatch, increasing the price of green technology or green products and limiting green development (Li et al., 2016). Therefore, its marginal effect on RGTI is diminishing.

Compared to the other two regions, indig in the central and western regions significantly promotes RGTI. This outcome could be because these regions have more potential promotion space for indig: with completed digital infrastructure, the use of digital technology has expanded adaptability of regional resources and enhanced the possibility as well as efficiency of carrying out green innovation activities; in the overall industrial chain, deep integration of digital technology and traditional industries promotes synergistic clustering of industries, and improves the cooperative innovation levels, as well as RGTI.

It's noteworthy that the indig coefficient for RGTI in the northeast region stands at 1.7882, which is positive but only significant at the 10% level. This indicates potential benefits from green dividends due to digitization. However, RGTI is limited due to reasons such as beginning of industrial digitalization, underdeveloped infrastructure, and contradictions in supply and demand of talents.

Discussion

Based on Model (1) shown in Table 4, we infer that indig enhances RGTI, thereby confirming Hypothesis 1. As the industrial digitization progresses, deep integration of digital technology and traditional industries has been enhanced. The extensive penetration and use of digital technology is expected to increase the demand for highly skilled and educated personnel. This will also continuously optimize the human capital structure, and lay the foundation for the proliferation of innovative factors and RGTI (Sun & Hou., 2019). Furthermore, deep integration will elevate employee innovation (Wang, 2022) and enhance open innovation by easing the identification, access, matching, and utilization of digital resources (Wu, 2022).

The regression results obtained from the panel threshold model presented in Table 6 reveal a development characterized by distinct phases in industrial digitization. These phases demonstrate increasing marginal effects, thus confirming Hypothesis 2. (1) Specifically, when indig is less than 2.35, the indig coefficient is positive (0.2543) at the 1% significance level. RGTI promotes information sharing effects, including multi-dimensional and multi-level knowledge, as well as the integration of technological domains. Undertaking RGTI solely based on enterprises' past technological experience and knowledge accumulation becomes challenging in such circumstances (Mubarak et al., 2021). Indig promotes industrial chain informatization and transparency, accelerates knowledge integration and reconstruction, promotes industrial cross-border integration, and expands the cognitive domain of enterprises, effectively improving the production efficiency. Besides, there are regional imbalances in industrial digitalization. The western region, represented by Xinjiang, Inner Mongolia, and the northeast region are lagging behind in industrial digitalization, and the indig degree is low. Due to infrastructure construction of digital technology and lag effects of ecological technology benefits, promotion of indig on RGTI cannot be fully realized. (2) When indig > 2.35, impact coefficient of indig on RGTI is significantly positive (0.4357) at 1% level. Once industrial digitalization reaches maturity, as indicated by a mature infrastructure, it unlocks a wider range of production tools, arranged in the most efficient manner. Moreover, sophisticated digital technologies optimize traditional business models and resource utilization, leading to improved productivity and innovation. These will further reduce technical energy consumption, improve the industrial access threshold, promote the development of green technology of enterprises, and enhance RGTI. Besides, it decreases information asymmetry between enterprises and promotes resource allocation, enabling enterprises to improve their knowledge of resource utilization and efficiently integrate resources. The eastern region, represented by Guangdong and Zhejiang, has been rapidly developing in terms of industrial digitalization. With the national "Beautiful China" development strategy, the environmental protection concept is deeply rooted, promoting the demand side to choose environmentally friendly products. Increased indig can further promote supply-side and demand-side communication, which reduces information asymmetry, improves the efficiency of enterprises to change the production mode, and facilitate enterprises to conduct RGTI.

Table 7 's mediation effect test results highlight that the advanced industrial structure and marketization processes lays the foundation for industrial digitalization to foster green technology innovation, corroborating Hypothesis 3 and Hypothesis 4. Industrial digitalization can upgrade traditional industries, strengthen the competitiveness of highly integrated industries, and ultimately lead to industrial transformation and upgrading. Of note, indig promotes structural innovation of production factors, empowers the marketization of factors and enhances the marketization process owing to its high technology, high payoff, high integration and high synergy. Marketization strengthens inter-regional technology exchanges, which favors regional green technology innovation.

Conclusions, Recommendations and Future research directionConclusions and recommendations

Given the dual constraints of energy and environment, promoting regional green technology innovation through industrial digitalization becomes a pivotal strategy to achieve the "dual carbon" goal and foster high-quality economic development. We have empirically analyzed the mechanism and effects (linear & non-linear) of indig on RGTI. The main conclusions are: i. Indig and RGTI have positive spatial correlations, with obvious spatial agglomeration characteristics. The indig level fluctuates, while the RGTI level continuously rises, showing regional heterogeneity and a unipolar characteristic; ii. Indig can effectively promote RGTI, both at national and regional levels (apart from the northeastern region), and during the research period, the RGTI level improved by about 0.4446, which was attributed to indig; iii. There is a single threshold for indig, with a threshold value of 2.35, and the positive driving trend has a “marginally increasing” characteristic; iv. There are two critical factors with partly intermediary roles —industrial structure advancement (0.2390) and marketization (0.6015). Accordingly, the following policy recommendations were made:

  • (1)

    Enhance industrial chain digitalization and intelligent transformation. Smoothen the digital trade circulation system and develop digital service trade. Focus on industrial digitalization and improve service efficiency as well as quality. As an important driver for regional green technology innovation, industrial digitization can optimize the structure of talents, and improve the development vitality of the region as a whole. However, giving full play of its effect requires strengthening digital infrastructure in the whole existing industrial chain. Steadily promote digital industry cooperation, promote digital infrastructure construction and scientific research and innovation.

  • (2)

    Due to regional heterogeneity, each region should formulate own corresponding policies to development characteristics. By formulating differentiated policies, the huge potential of industrial digitalization service economy will be released and imbalance of development between regions will be alleviated. Moreover, increasing the allocation of innovation resources and capital investment is essential to stimulate enterprise-led green technology innovation. The low level of industrial digitization in the northeast region cannot promote RGTI while in the west and northeast, it is slow and at risk of stagnation. Therefore, it is necessary for relatively backward regions to undertake industry transfer from the eastern and central regions to synergistically build an industrial chain system, ensuring steady introduction of advanced technologies. Due to the high demand of digital technology for talents, it is helpful for backward regions to establish supporting policies to attract short-term talents in key fields and accelerate talent flow.

  • (3)

    Promote industrial structure advancement and marketization. Upgrade traditional industries by digital technology, digitalize the traditional operation model, and promote information exchange on supply and demand sides. Besides, optimize the configuration of heterogeneous resources and improve resource utilization efficiency via deep integration in the industrial structure advancement process, promoting the cognitive level of overall industry and improving the innovation boundary of enterprises. Continuous development of marketization is an effective motive force for enhancing green innovation. Enterprises should employ feedback signals from the market in time to alleviate resource mismatches and further promote technological spillovers.

  • (4)

    Enterprises should actively develop digital technology and realize digital transformation. The digital economy can promote the progress of green technology, thereby improving the green production efficiency of enterprises. If enterprises build digital platforms, while promoting information sharing and improving technical levels, they can also prevent potential risks and achieve smooth development of enterprises. In addition, the creation of a digital platform enables collaborative development among enterprises, facilitating more precise professional division of labor through the utilization of digital technology.

Limitations and future research direction

The purpose of this empirical study is to investigate the linear and non-linear impact of industrial digitalization on regional green technology innovation. These results indicate that the former can significantly promote the latter. However, this paper investigates the influence of both internal and external factors on green technology innovation, utilizing macro-level data. While a micro-level perspective holds greater sway and significance for companies and enterprises in policymaking, it is crucial to consider the path of industrial digitization for attaining high-quality economic development from a micro standpoint (Ning et al., 2022; Ning et al., 2023). In addition, since both industrial digitalization and regional green technology innovation show certain spatial correlation, whether industrial digitalization as the main body of digital economy has a spatial spillover effect on regional green technology innovation need to be further discussed.

Acknowledgments

This study was funded by Major Project of the National Social Science Fund of China 21&ZD133, Social Science Fund of Xinjiang [grant number 21BJY049], Postgraduate Education Teaching Reform Project of the Xinjiang Uygur Autonomous Region [grant number XJ2023GY08]; Xinjiang University's Graduate Education and Teaching Reform Project [grant number XJDX2023YJG07]; General Project of the Institute of Central Asian Studies of Xinjiang University [grant number 24ZYYB018]; the 2024 University Scientific Research Project of Xinjiang Autonomous Region [Research on the impact of digital government construction on the green innovation of enterprises in resource-based areas].

Appendix A

Robustness tests of basic regression

(1) Lag period test

  • indig with one-period lag (indig1) and two-period lag (indig2) are introduced into the benchmark regression model to mitigate the possible estimation bias caused by reverse causality. In Table A1, coefficients of indig1 and indig2 are significantly positive. Thus, the results are robust.

Table A1.

Results of lag period test.

  (1)  (2)  (3)  (4) 
indig1  2.0423***  1.5485***     
  (5.00)  (4.05)     
indig2      1.9616***  1.7662*** 
      (4.93)  (4.40) 
gover    0.9505    1.1437 
    (0.71)    (1.21) 
er    0.2995*    0.1984 
    (1.96)    (1.69) 
finsize    13.0246***    4.9100 
    (2.78)    (1.19) 
fdi    -2.4013    -4.8354* 
    (-0.96)    (-1.87) 
_cons  2.4869**  2.4445**  2.8618***  2.6635** 
  (2.31)  (2.63)  (2.76)  (2.51) 
N  180  180  150  150 
R2  0.446  0.543  0.494  0.553 

(2) Replacing the dependent variable

While some studies measure regional green technology innovation using the quantity of granted green patents, others utilize the quantity of green technology applications (Xu & Cui., 2020). Therefore, we adopted the number of green patent applications for the robustness test. In models (1) and (2) of Table A2, the coefficient of indig is positive and significant, indicating that the results are robust.

Table A2.

Results of replacing dependent variable.

  (1)  (2)  Non-municipalities1  Non-municipalities2 
  ln RGTI 1  ln RGTI 1  RGTI  RGTI 
indig  2.5588***  1.8466***  2.0617***  1.4273*** 
  (5.34)  (4.65)  (4.40)  (3.60) 
gover    0.5073    0.4338 
    (0.28)    (0.24) 
er    0.2602    0.3562** 
    (1.60)    (2.37) 
finsize    19.4098***    18.9843*** 
    (3.81)    (3.88) 
fdi    1.6920    0.8291 
    (0.58)    (0.18) 
_cons  1.5733  1.7538*  2.2266*  2.3409** 
  (1.24)  (1.76)  (1.81)  (2.38) 
N  210  210  182  182 
R2  0.530  0.627  0.471  0.599 

(3) Reducing samples

This paper directly removes municipality samples for robustness test, and results are in columns (4) and (5) of Table A2. The coefficients of indig on RGTI are significantly positive, proving that the results are robust.

(4) Endogenous test

① (Two-stage least squares) regression with instrumental variable

Empirical analyses might encounter endogenous problems due to reverse causality. Regions with high levels of green technology innovation will pursue high-quality economic development, which is matched with characteristics of industrial digitalization with quality and efficiency, green benefits. Regional human capital level with high level of regional green technology innovation is generally higher. Based on employee organization matching theory, digital transformation and upgrading is inseparable from matching high-quality employees (Xiao et al., 2022). Thus, interactions and a reverse causal relationship may exist between industrial digitalization and regional green technology innovation.

Referring to Arellano & Bond (1991), we adopt the two-stage least squares method (2SLS). The first and second-order lag terms of the independent variables serve as instrumental variables for the endogeneity test. Regression results are shown in Table A3. After considering the possible endogenous problems, industrial digitalization still shows a significant positive impact on regional green technology innovation, indicating robust benchmark regression results.

Table A3.

The results of 2SLS regression with instrumental variable.

  (1)  (2)  (3) 
indig  2.6097***  1.6898***  2.1055*** 
  (13.06)  (8.36)  (6.54) 
Control variable  No  YES  YES 
Province FE  No  No  YES 
Kleibergen-Paap rk LM  23.753  16.858  17.474 
Kleibergen-Paap rk Wald F  2270.253  1133.876  274.908 
150  150  150 

② GMM with instrumental variable

Given the endogeneity problem, the systematic GMM model is used for regression analysis, and the Sargan statistic is selected to test instrumental variable selection reliability of the GMM model (Table A4). The Sargan test result is 0.986, implying that instrumental variable selection is valid. In addition, the Arellano-Bond statistic AR(2)=0.215, thus, estimation results of system GMM model are consistent and valid. After considering the influence of endogeneity, industrial digitalization can promote regional green technology innovation, implying robust results.

Table A4.

The results of GMM regression with instrumental variable.

  (1) 
  lngreen 
L.lngreen  0.9268*** 
  (0.0409) 
   
Indig  0.2318* 
  (0.1375) 
   
Controls variables  YES 
   
Province FE  YES 
   
AR(1)  0.002 
AR(2)  0.275 
Sargan test  0.740 
180 

Note: Standard errors in parentheses.

Appendix B

Robustness test of panel threshold

Conduct robustness test by adding control variables to the threshold model. Model (1) adds open, model (2) adds rd, and model (3) adds both to the original threshold model (See Table B). Models (1)-(3) show that indig plays an increasing marginal effect on RGTI. Therefore, the result is robust.

Table B.

Panel threshold robustness tests.

  (1)  (2)  (3) 
  lngreen  lngreen  lngreen 
  b/z  b/z  b/z 
indig·I(indig<2.35)  0.2105  0.2326  0.1811 
  (1.36)  (1.49)  (1.12) 
indig·I(indig≥2.35)  0.3905**  0.4174***  0.3652** 
  (2.58)  (2.77)  (2.33) 
L.lngreen  0.7260***  0.7185***  0.7148*** 
  (15.03)  (14.01)  (13.93) 
gover  1.5990***  1.5491***  1.6378*** 
  (3.13)  (3.03)  (3.18) 
fdi  -0.9087  -0.5477  -0.8390 
  (-0.61)  (-0.37)  (-0.56) 
er  0.0950  0.0933  0.0937 
  (1.27)  (1.24)  (1.24) 
finsize  3.4088*  3.5471*  3.1979 
  (1.76)  (1.82)  (1.63) 
open  -0.3447    -0.3653 
  (-1.15)    (-1.21) 
rd    6.1447  7.5670 
    (0.54)  (0.66) 
_cons  0.6070  0.3993  0.6422 
  (1.58)  (1.20)  (1.65) 
180  180  180 
R2  0.853  0.851  0.853 

References
[Arellano and Bond, 1991]
M. Arellano, S. Bond.
Some Tests of Specification for Panel Data: Monte Carlo Evidence and an Application to Employment Equations.
The Review of Economic Studies, 58 (1991), pp. 277-297
[Armstrong, 2006]
M. Armstrong.
Competition in two-sided markets.
The RAND journal of economics, 37 (2006), pp. 668-691
[Baro and Kenny, 1986]
R.M. Baro, D.A Kenny.
The moderator–mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations.
Journal of personality and social psychology, 51 (1986), pp. 1173
[Bisht et al., 2022]
D. Bisht, R. Singh, A. Gehlot, S.V. Akram, A. Singh, E.C. Montero, B. Twala.
Imperative role of integrating digitalization in the firms finance: A technological perspective.
Electronics, 11 (2022), pp. 3252
[Cao, Feng, Yang, & Yang, 2024]
Q. Cao, Z. Feng, R. Yang, C. Yang.
Conflict and natural resource condition: An examination based on national power heterogeneity.
Resources Policy, 89 (2024), pp. 104549
[Cette et al., 2022]
G. Cette, S. Nevoux, L. Py.
The impact of ICTs and digitalization on productivity and labor share: evidence from French firms.
Economics of innovation and new technology, 31 (2022), pp. 669-692
[Dai and Ma, 2023]
X. Dai, H.W. Ma.
Digital transformation, export growth and low markup trap.
China Industrial Economics, (2023), pp. 61-79
[Dong and Wang, 2021]
Z. Dong, H. Wang.
Urban wealth and green technology choice.
Econ Res J, 56 (2021), pp. 143-159
[Fan et al., 2011]
G. Fan, X. Wang, G. Ma.
Contribution of marketization to China's economic growth.
Economic Research Journal, 9 (2011), pp. 1997-2011
[Feng and Zheng, 2022]
G.F. Feng, M. Zheng.
Economic policy uncertainty and renewable energy innovation: international evidence.
Innovation and Green Development, 1 (2022),
[Gan et al., 2011]
C.H. Gan, R.G. Zheng, D.F. Yu.
An empirical study on the effects of industrial structure on economic growth and fluctuations in China.
Economic Research Journal, 5 (2011), pp. 4-16
[Goldfarb and Tucker, 2019]
A. Goldfarb, C. Tucker.
Digital economics.
Journal of Economic Literature, 57 (2019), pp. 3-43
[Gölzer and Fritzsche, 2017]
P. Gölzer, A. Fritzsche.
Data-driven operations management: organisational implications of the digital transformation in industrial practice.
Production Planning & Control, 28 (2017), pp. 1332-1343
[Griliches et al., 1991]
Z. Griliches, B.H. Hall, A. Pakes.
R&D, patents, and market value revisited: is there a second (technological opportunity) factor?.
Economics of Innovation and new technology, 1 (1991), pp. 183-201
[Hao and Deng, 2019]
X. Hao, F. Deng.
The marginal and double threshold effects of regional innovation on energy consumption structure: evidence from resource-based regions in China.
Energy Policy, 131 (2019), pp. 144-154
[Hao et al., 2023a]
X. Hao, Y. Li, S. Ren, H. Wu, Y. Hao.
The role of digitalization on green economic growth: Does industrial structure optimization and green innovation matter?.
Journal of Environmental Management, 325 (2023),
[Hao et al., 2023b]
X. Hao, Y. Li, H. Wu.
Ways to improve the efficiency of clean energy utilization: Does digitalization matter?.
Energy Strategy Reviews, 50 (2023),
[Hao et al., 2023c]
X. Hao, S. Wen, K. Li, J. Wu, H. Wu, Y. Hao.
Environmental governance, executive incentive, and enterprise performance: Evidence from Chinese mineral enterprises.
Resources Policy, 85 (2023),
[Hao et al., 2023d]
X. Hao, S. Wen, Q. Sun, M. Irfan, H. Wu, Y. Hao.
How does the target of green innovation for cleaner production change in management process? Quality targeting and link targeting.
Journal of Environmental Management, 345 (2023),
[Hao et al., 2023e]
X. Hao, S. Wen, Y. Xue, H. Wu, Y. Hao.
How to improve environment, resources and economic efficiency in the digital era?.
Resources Policy, 80 (2023),
[Hao et al., 2022]
Y. Hao, Y. Guo, S. Li, S. Luo, X. Jiang, Z. Shen, H. Wu.
Towards achieving the sustainable development goal of industry: How does industrial agglomeration affect air pollution?.
Innovation and Green Development, 1 (2022),
[Henfridsson and Bygstad, 2013]
O. Henfridsson, B. Bygstad.
The generative mechanisms of digital infrastructure evolution.
MIS quarterly, (2013), pp. 907-931
[Heo and Lee, 2019]
P.S. Heo, D.H. Lee.
Evolution of the linkage structure of ICT industry and its role in the economic system: The case of Korea.
Information technology for development, 25 (2019), pp. 424-454
[Heredia et al., 2022]
J. Heredia, M. Castillo-Vergara, C. Geldes, F.M.C. Gamarra, A. Flores, W. Heredia.
How do digital capabilities affect firm performance? The mediating role of technological capabilities in the “new normal”.
Journal of Innovation & Knowledge, 7 (2022),
[Hung and Nham, 2023]
B.Q. Hung, N.T.H. Nham.
The importance of digitalization in powering environmental innovation performance of European countries.
Journal of Innovation & Knowledge, 8 (2023),
[Knickrehm et al., 2016]
M. Knickrehm, B. Berthon, P. Daugherty.
Digital disruption: The growth multiplier.
Accenture Strategy, (2016),
[Kohli and Melville, 2019]
R. Kohli, N.P. Melville.
Digital innovation: A review and synthesis.
Information Systems Journal, 29 (2019), pp. 200-223
[Kremer et al., 2013]
S. Kremer, A. Bick, D. Nautz.
Inflation and growth: new evidence from a dynamic panel threshold analysis.
Empirical Economics, 44 (2013), pp. 861-878
[Lailag & Chen, 2022]
U. Lailag, W. Chen.
How Does Digitization Affect Sports Industry Development and Public Health?.
Economic Analysis Letters, 1 (2022), pp. 8-14
[Lange et al., 2020]
S. Lange, J. Pohl, T. Santarius.
Digitalization and energy consumption. Does ICT reduce energy demand?.
Ecological economics, 176 (2020),
[Li et al., 2016]
F. Li, A. Nucciarelli, S. Roden, G. Graham.
How smart cities transform operations models: a new research agenda for operations management in the digital economy.
Production Planning & Control, 27 (2016), pp. 514-528
[Li and Liao, 2022]
G. Li, F. Liao.
Input digitalization and green total factor productivity under the constraint of carbon emissions.
Journal of Cleaner Production, 377 (2022),
[Li and Du, 2014]
M.J. Li, W.J. Du.
Does the double dividend of environmental regulation and employment apply to China at this stage.
Empirical analysis based on provincial panel data Economic Science, 4 (2014), pp. 14-25
[Li, 2014]
X.L. Li.
An empirical analysis based on marketization, industrial agglomeration and environmental pollution.
Stat Res, 8 (2014), pp. 39-45
[Li and Yang, 2018]
Z. Li, S.Y. Yang.
Fiscal decentralization, government innovation preferences and regional innovation efficiency.
Management World, 34 (2018), pp. 29-194
[Liu et al., 2023]
J. Liu, Y. Xue, Z. Mao, M. Irfan, H. Wu.
How to improve total factor energy efficiency under climate change: does export sophistication matter?.
Environmental Science and Pollution Research, 30 (2023), pp. 28162-28172
[Mao et al., 2023]
J. Mao, J. Xie, Z. Hu, L. Deng, H. Wu, Y. Hao.
Sustainable development through green innovation and resource allocation in cities: Evidence from machine learning.
Sustainable Development, (2023), http://dx.doi.org/10.1002/sd.2516
[Martin, 2003]
S.P. Martin.
Is the digital divide really closing? A critique of inequality measurement in a nation online.
IT & society, 1 (2003), pp. 1-13
[Mesenbourg, 2001]
T.L. Mesenbourg.
Measuring the digital economy.
US Bureau of the Census, 1 (2001), pp. 1-19
[Miller and Wilsdon, 2001]
P. Miller, J. Wilsdon.
Digital futures—an agenda for a sustainable digital economy.
Corporate Environmental Strategy, 8 (2001), pp. 275-280
[Mondal, Al-Kfairy, & Mellor, 2024]
C. Mondal, M. Al-Kfairy, R.B. Mellor.
Entrepreneurial universities: Modelling the link between innovation producers and innovation users shows that team structures in the tech transfer function improves performance.
Economic Analysis Letters, 3 (2024), pp. 54
[Mubarak et al., 2021]
M.F. Mubarak, S. Tiwari, M. Petraite, M. Mubarik, R.Z.R.M. Rasi.
How Industry 4.0 technologies and open innovation can improve green innovation performance?.
Management of Environmental Quality: An International Journal, 32 (2021), pp. 1007-1022
[Nambisan, 2017]
S. Nambisan.
Digital entrepreneurship: Toward a digital technology perspective of entrepreneurship.
Entrepreneurship theory and practice, 41 (2017), pp. 1029-1055
[Ning et al., 2023]
J. Ning, X. Jiang, J. Luo.
Relationship between enterprise digitalization and green innovation: A mediated moderation model.
Journal of Innovation & Knowledge, 8 (2023),
[Ning et al., 2022]
J. Ning, Q. Yin, A. Yan.
How does the digital economy promote green technology innovation by manufacturing enterprises? Evidence from China.
Frontiers in Environmental Science, 10 (2022),
[Park, 2022]
S. Park.
Do Bank Capital Requirements Make Resource Allocation Suboptimal?.
Journal of Economic Analysis, 1 (2022), pp. 35-49
[Peng and Tao, 2022]
Y. Peng, C. Tao.
Can digital transformation promote enterprise performance?—From the perspective of public policy and innovation.
Journal of Innovation & Knowledge, 7 (2022),
[Piñeiro-Chousa et al., 2022]
Juan Piñeiro-Chousa, M.A. Lopez-Cabarcos, L. Quinoa-Pinerio, A.M. Perez-Pico.
Us biopharmaceutical companies' stock market reaction to the covid-19 pandemic. understanding the concept of the 'paradoxical spiral' from a sustainability perspective.
Technological forecasting and social change, (2022),
[Piñeiro-Chousa et al., 2018]
Juan Piñeiro-Chousa, Marcos Vizcaíno-González, Jérôme. Caby.
Financial development and standardized reporting: a comparison among developed, emerging, and frontier markets.
Journal of Business Research, (2018), pp. 797-802
[Qiu et al., 2022]
Y. Qiu, S. Jia, J. Liao, X. Yang.
Evaluation of Urban High quality Development Level Based on Entropy Weight-TOPSIS Two step Method.
Journal of Economic Analysis, 1 (2022), pp. 50-65
[Reddy, 2023]
K. Reddy.
Are CEOs Paid for Performance? A Study of CEO's Compensation in the Public Sector Corporations.
Journal of Economic Analysis, 2 (2023), pp. 16-36
[Ren et al., 2021]
S. Ren, Y. Hao, L. Xu, H. Wu, N. Ba.
Digitalization and energy: how does internet development affect China's energy consumption?.
Energy Economics, Elsevier, 98 (2021),
[Romero-Castro et al., 2022]
N. Romero-Castro, V. Miramontes-Viña, M.Á. López-Cabarcos.
Understanding the antecedents of entrepreneurship and renewable energies to promote the development of community renewable energy in rural areas.
Sustainability, 14 (2022), pp. 1234
[Romer, 1986]
P.M. Romer.
Increasing returns and long-run growth.
Journal of political economy, 94 (1986), pp. 1002-1037
[Schumpeter, 1982]
J.A. Schumpeter.
The theory of economic development: An inquiry into profits, capital, credit, interest, and the business cycle (1912/1934).
Transaction Publishers, (1982),
[Shi, 2022]
D. Shi.
Evolution of Industrial Development Trend under Digital Economy.
China Industrial Economics, (2022), pp. 26-42
[Staab, 2017]
P. Staab.
The consumption dilemma of digital capitalism.
Transfer: European Review of Labour and Research, 23 (2017), pp. 281-294
[Sui and Yao, 2023]
B. Sui, L. Yao.
The impact of digital transformation on corporate financialization: The mediating effect of green technology innovation.
Innovation and Green Development, 2 (2023),
[Sun and Hou, 2019]
Z. Sun, Y.L. Hou.
How Does Industrial Intelligence Reshape the Employment Structure of Chinese Labor Force.
China Industrial Economies, 5 (2019), pp. 61-79
[Tadesse, 2002]
S. Tadesse.
Financial architecture and economic performance: international evidence.
Journal of financial intermediation, 11 (2002), pp. 429-454
[Tao et al., 2021]
F. Tao, J. Zhao, H. Zhou.
Does Environmental Regulation Improve the Quantity and Quality of Green Innovation——Evidence from the Target Responsibility System of Environmental Protection.
China Industrial Economics, (2021), pp. 136-154
[Tapscott, 1996]
D. Tapscott.
The Digital Economy: Promise and Peril in the Age of Networked Intelligence.
McGrawHill, (1996),
[Thanh, 2022]
T.T. Thanh.
Effects of digital public services on trades in green goods: does institutional quality matter?.
Journal of Innovation & Knowledge, 7 (2022),
[Tian et al., 2022]
H. Tian, Y. Li, Y. Zhang.
Digital and intelligent empowerment: Can big data capability drive green process innovation of manufacturing enterprises?.
Journal of Cleaner Production, 377 (2022),
[Torrent-Sellens et al., 2022]
J. Torrent-Sellens, A. Diaz-Chao, A.P. Miro-Perez, J. Sainz.
Towards the Tyrell corporation? Digitisation, firm-size and productivity divergence in Spain.
Journal of Innovation & Knowledge, 7 (2022),
[Verhoef et al., 2021]
P.C. Verhoef, T. Broekhuizen, Y. Bart, A. Bhattacharya, J.Q. Dong, N. Fabian, M. Haenlein.
Digital transformation: A multidisciplinary reflection and research agenda.
Journal of Business Research, 122 (2021), pp. 889-901
[Wagner, 2007]
M. Wagner.
On the relationship between environmental management, environmental innovation and patenting: Evidence from German manufacturing firms.
Research policy, 36 (2007), pp. 1587-1602
[Wang and Zhang, 2020]
H. Wang, Y. Zhang.
Trade Structure Upgrading, Environmental Regulation and Green Technology Innovation in Different Regions of China.
China Soft Science, (2020), pp. 174-181
[Wang and Du, 2022]
Q. Wang, Z.Y. Du.
Changing the impact of banking concentration on corporate innovation: The moderating effect of digital transformation.
Technology in Society, 71 (2022),
[Wang et al., 2022]
Q. Wang, A. Hu, Z. Tian.
Digital transformation and electricity consumption: Evidence from the Broadband China pilot policy.
Energy Economics, 115 (2022),
[Wang, 2022]
Y. Wang.
Analyzing the mechanism of strategic orientation towards digitization and organizational performance settings enduring employee resistance to innovation and performance capabilities.
Frontiers in Psychology, (2022), pp. 13
[Wu et al., 2021]
H. Wu, Y. Hao, S. Ren, X. Yang, G. Xie.
Does Internet development improve green total factor energy efficiency?.
evidence from China. Energy Policy, (2021),
[Wu et al., 2024]
H. Wu, R. Zhong, Z. Wang, Y. Qu, X. Yang, Y. Hao.
How Does Industrial Intellectualization Affect Energy Intensity? Evidence from China.
The Energy Journal, 45 (2024),
[Wu, Yang, & Chen, 2024]
J. Wu, C. Yang, L. Chen.
Examining the non-linear effects of monetary policy on carbon emissions.
Energy Economics, 107206 (2024),
[Wu et al., 2022]
L. Wu, L. Sun, Q. Chang, D. Zhang, P. Qi.
How do digitalization capabilities enable open innovation in manufacturing enterprises? A multiple case study based on resource integration perspective.
Technological Forecasting and Social Change, 184 (2022),
[Wurlod and Noailly, 2018]
J.D. Wurlod, J. Noailly.
The impact of green innovation on energy intensity: An empirical analysis for 14 industrial sectors in OECD countries.
Energy Economics, 71 (2018), pp. 47-61
[Xiao and Liu, 2022]
H. Xiao, J. Liu.
The impact of digital economy development on local fiscal revenue efficiency.
Economic Analysis Letters, 1 (2022), pp. 1-7
[Xiao et al., 2022]
T. Xiao, R. Sun, C. Yuan, J. Sun.
Digital Transformation, Human Capital Structure Adjustment and Labor Income Share.
Journal of Management World, 38 (2022), pp. 220-237
[Xu and Cui, 2020]
J. Xu, J.B. Cui.
Low-carbon cities and firms’ green technological innovation.
China Ind. Econ, 12 (2020), pp. 178-196
[Yang, Zhu, & Albitar, 2024]
C. Yang, C. Zhu, K. Albitar.
ESG ratings and green innovation: AU‐shaped journey towards sustainable development.
Business Strategy and the Environment, (2024),
[Yang and Yee, 2022]
Y. Yang, R.W. Yee.
The effect of process digitalization initiative on firm performance: A dynamic capability development perspective.
International Journal of Production Economics, 254 (2022),
[Yi et al., 2022]
M. Yi, Y. Liu, M.S. Sheng, L. Wen.
Effects of digital economy on carbon emission reduction: New evidence from China.
Energy Policy, 171 (2022),
[Zhang et al., 2022]
W. Zhang, T. Zhang, H. Li, H. Zhang.
Dynamic spillover capacity of R&D and digital investments in China's manufacturing industry under long-term technological progress based on the industry chain perspective.
Technology in Society, 71 (2022),
[Zhao et al., 2022]
S. Zhao, D. Peng, H. Wen, Y. Wu.
Nonlinear and spatial spillover effects of the digital economy on green total factor energy efficiency: Evidence from 281 cities in China.
Environmental Science and Pollution Research, (2022), pp. 1-21
[Zhao et al., 2020]
T. Zhao, Z. Zhang, S.K. Liang.
Digital economy, entrepreneurship, and high-quality economic development: empirical evidence from urban China.
Management World, 36 (2020), pp. 65-76
[Zhong and Shao, 2022]
S. Zhong, J. Shao.
Distribution Dynamics, Spatial Difference and Convergence of Innovative Development in the Yellow River Basin.
The Journal of Quantitative & Technical Economics, 39 (2022), pp. 25-46
[Zhou et al., 2022]
B. Zhou, H. Zhao, J. Yu, T. He, J. Liu.
Does the growth of the digital economy boost the efficiency of synergistic carbon-haze governance? Evidence from China.
Frontiers in Environmental Science, (2022), pp. 1621

Google Scholar: https://scholar.google.com/citations?user=6ZN5hFcAAAAJ&hl=zh-CN

Assoc. Prof. Dr. Xiaoli Hao is currently a full associate professor at School of Economics and Management of Xinjiang University. Her main research interests include innovation management, energy economics and environmental economics. E-mail: haoxiaoli@126.com

Mr. Yi Liang is currently a master student of School of Economics, Minzu University. E-mail: ly_xjdx_edu@163.com

Mr. Cunyi Yang is currently a Ph.D. student on economics at Lingnan College, Sun Yat-Sen University, Guangzhou, China. E-mail: yangcy9@mail2.sysu.edu.cn.

Present address: School of Management and Economics, Beijing Institute of Technology. 5 Zhongguancun South Street, Haidian District, Beijing 100081, China https://ideas.repec.org/d/cebitcn.html, http://www.ceep.net.cn/english/index.htm

Copyright © 2024. The Authors
Descargar PDF
Opciones de artículo