Summary
Improving energy-environmental efficiency (EEE) is vital for meeting energy conservation and emission reduction objectives. This research investigates the dynamics by which green technology innovation (GTI) influences EEE and explores the role of environmental regulation (ER) in shaping this relationship. Employing a Dynamic Spatial Durbin Model and analyzing panel data spanning 2003 to 2017 across 30 Chinese provinces, the study examines how GTI impacts EEE in the context of ER. The empirical findings indicate the following: (1) GTI exhibits a U-shaped impact on EEE, driven mainly by SubGI. (2) GTI’s influence on EEE is prominently manifested in PTE, also stemming from SubGI. (3) The interaction term between ER and GTI is 0.0022, with the GTI coefficient at -0.0741 and the GTI quadratic term coefficient at 0.0007—all statistically significant. This suggests that ER mitigates the negative impact of GTI on EEE while reinforcing its positive effect. These results offer empirical evidence and policy insights for optimizing the use of GTI and ER to enhance EEE and achieve energy conservation and emissions reduction objectives.
Opening Statement
Since the initiation of economic reforms and opening-up policies, China has experienced rapid economic development, giving rise to a growing contradiction among economic growth, energy consumption, and environmental pollution. According to the BP Statistical Review of World Energy in 2019, China’s primary energy consumption surged to 141.7 EJ, marking a 4.4% increase from the previous year and representing the fastest-growing energy consumer for 19 consecutive years. This escalating primary energy consumption not only signals a faster depletion of limited natural energy resources but also raises concerns about energy security.
As a crucial indicator of a nation or region’s sustainable economic development, China’s carbon intensity in 2019 was 48.1% lower than in 2005. However, it still exceeded the global average during the period from 1990 to 2019. This underscores the imperative of controlling carbon emission intensity within the context of energy conservation and emissions reduction. Consequently, exploring ways to enhance energy-environmental efficiency (EEE) becomes a critical measure for reducing energy consumption and improving carbon intensity, serving as a key driver for achieving sustainable economic development in China.
Green technology innovation (GTI) has emerged as a significant approach to ecological and environmental protection in various regions. It has demonstrated promising outcomes in controlling industrial energy consumption, reducing energy intensity, encouraging the development of green technologies, curbing the use of non-renewable energy sources, and harmonizing ecological conservation with economic growth. However, studies, such as that by Wang and Chen, have revealed that the relationship between resource dependency and haze pollution in Chinese prefecture-level cities is intricate. While GTI can reduce haze pollution when resource dependency is low or moderate, it may lead to the disappearance of optimization effects when resource dependency is high. Similarly, research by Mongo et al. indicates that environmental innovation may reduce carbon emissions in the long term but could have opposite effects in the short term. This suggests potential rebound effects associated with GTI, leading to increased resource consumption and carbon emissions. Consequently, it is crucial to assess the impact of GTI on EEE, considering the potential complexities of its effects.
Moreover, studies by Zhang et al. and Xing and Dong highlight the importance of distinguishing between genuine green technology innovation and strategic green technology innovation, with the latter being influenced more by environmental regulations. This distinction reveals that the impact of GTI on EEE may vary depending on the nature of innovation, but current assessments often focus solely on GTI itself, neglecting the influence of its components on EEE.
Furthermore, China has implemented a series of environmental policies since 1970 aimed at improving the environment through legal and market mechanisms, such as pollution fees and carbon trading markets. The interaction between environmental regulation (ER) and value-added tax has motivated enterprises to reduce the intensity of SO2 emissions and promote innovation in pollution reduction. Thus, when researching how to enhance EEE, considering ER is essential. However, existing research predominantly focuses on the impact of ER on EEE, overlooking the role of ER in the mechanism through which GTI affects EEE.
To address these gaps, this study integrates the definition of GTI, decomposes GTI into Substantial green innovation (SubGI) and Symbolic green innovation (SymGI), and elucidates the underlying mechanisms through which GTI affects EEE. It aims to clarify whether the impact of GTI on EEE is driven by the development and promotion of new clean energy technologies or by improvements in ecological environmental protection and resource recycling technologies. Additionally, the study introduces the interaction effects of ER with GTI, SubGI, and SymGI, elucidating the role of ER in the mechanisms through which these components influence EEE. Lastly, spatial factors are incorporated through the use of spatial econometric models to explore the impact of GTI on EEE under the influence of ER, aiming for more realistic and reliable outcomes.
This research contributes to the existing literature by providing a deeper understanding of the mechanisms of GTI through its decomposition. Simultaneously considering the moderating role of ER and spatial spillover effects aligns the study with the real-world context, offering a theoretical basis and reference for China and other developing countries to more accurately utilize GTI and ER for energy conservation and emissions reduction.
Review of Existing Studies
The trajectory of energy is molded by technology, and the innovation of technology paves the way for the future of energy. Within the confines of China’s dual carbon targets, the imperative of attaining energy conservation, emission reduction, and enhancing energy-environmental efficiency (EEE) has taken precedence. In the scrutiny of energy and the environment, EEE stands out as the most promising instrument for forging a balanced relationship between economic growth and resource consumption. Consequently, the mechanisms propelling EEE have become a focal point of extensive inquiry. The structure of the literature review is detailed in Table 1.
Indicators | Dimension | Summary | References |
---|---|---|---|
GTI | Positive impact | Enhanced environmental performance, profitability, core competitiveness, total factor carbon productivity in high-income economies, and carbon performance | 20, 21, 22, 23, 24 |
Non-linear variation | Critical point reached before a boost in green productivity, with an inverted U-shaped effect on regional carbon emissions | 28, 29, 30 | |
ER | Positive effect | Promotes green transformation of enterprises, fosters long-term economic growth, increases green innovation outputs, improves energy efficiency, and enhances green technological efficiency | 11, 14, 33, 34, 35, 36 |
Moderating effect | Enhances the impact of green knowledge innovation on CO2 emissions | 11, 26 | |
EEE | Positive effect | Reduces pollution emissions, enhances sustainability of resource utilization, stimulates technological innovation, effectively counters environmental problems, and promotes sustainable development | 38, 39, 40, 41 |
Affected by multiple factors | Influenced by technological innovation, policies and regulations, FDI absorptive capacity, level of economic development, industrial and energy structure, etc | 42, 43, 44, 45, 46 | |
Consists of SE, PTE | Multiplication of SE and PTE | 47, 48, 49 |
GTI, positioned as a pivotal driver of sustainable development, has become a focal point for scholars across diverse fields. Initially, green innovation contributes not only to a company’s environmental performance, profitability, and core competitiveness but also bolsters total factor carbon productivity in high-income economies and carbon performance. This is attributed to green innovation steering companies toward more environmentally friendly production and operational methods, thus curbing resource waste and pollution emissions. Despite these varied benefits, the specific impact of GTI may differ depending on the industry or context. For instance, environmental regulations (ER) can amplify the effect of green knowledge innovation on CO2 reduction, prompting heavily-polluting enterprises to prioritize substantial green technology innovation during their green transformation. However, the fundamental reasons behind these impacts of GTI remain underexplored. Moreover, some studies neglect intrinsic driving factors of GTI.
On another note, ER, functioning as a political tool, drives corporate green transformation, promotes long-term economic growth, enhances green innovation output, improves energy-environmental efficiency (EEE), and augments green technological efficiency. The rationale lies in moderate ER partially or fully offsetting enterprise innovation costs through the compensating effects of green innovation. However, ER’s implementation can also influence other factors, such as significantly increasing the impact of green knowledge innovation on CO2 emissions while having a less pronounced impact on green process innovation. Therefore, ER’s role is crucial in understanding the overall impact mechanism. Furthermore, environmental policies and carbon emissions exhibit spatial spillover effects, making the inclusion of spatial factors more realistic.
EEE, a pivotal concept in the realms of the green economy and sustainable development, holds the potential to reduce pollution emissions, enhance sustainable resource utilization, stimulate technological innovation, and effectively address environmental issues, thereby promoting sustainable development. However, EEE is intricately influenced by various factors such as technological innovation, policy regulations, absorptive capacity of foreign direct investment, economic development level, industrial structure, and energy structure. It is a complex indicator requiring in-depth research and comprehensive consideration. Moreover, EEE is comprised of scale efficiency (SE) and pure technical efficiency (PTE), representing the product of these two efficiencies. SE is influenced by scale factors, while PTE is affected by management and technological factors. Hence, understanding which part of EEE is affected by various influencing factors is essential.
In summary, despite extensive research on EEE, GTI, and ER in the field of energy-environment management, a research gap persists regarding the underlying mechanisms of GTI and the moderating role of ER. Given China’s energy structure, there is a critical need to enhance energy utilization efficiency through GTI, a cornerstone in achieving dual carbon objectives. To address the limitations in evaluating GTI’s impact on EEE, this study decomposes GTI into SubGI and SymGI, introduces interaction terms of ER with GTI, SubGI, and SymGI to explore the sources of GTI’s impact on EEE and the moderating effects of ER on this impact and its sources. Empirical analysis is conducted using spatial Durbin models, incorporating spatial spillover effects to better align with real-world development scenarios.
Data Collection and Research Methodology
Taking the year 2003 as the starting point due to the emphasis on development during the 16th National Congress of the Communist Party of China in November 2002, this study calculates the EEE of each province in China. The study investigates the impact of GTI on EEE under the influence of ER, with the calculation of EEE relying on the input indicator of capital stock measured by gross fixed capital formation. Unfortunately, this data is unavailable after 2017 in the China Statistical Yearbook. To ensure data availability, the study selects data from 30 provinces and regions in China (excluding Tibet, Hong Kong, Macao, and Taiwan) from 2003 to 2017.
A two-step approach is employed to investigate the impact of GTI on EEE under the influence of ER. Firstly, EEE in 30 provinces is measured. Secondly, the impact of GTI on EEE under the influence of ER is analyzed, considering EEE, SE, and PTE as explanatory variables, and GTI, SubGI, SymGI as explanatory variables, with ER as a moderating variable. Control variables include gross domestic product per capita, industrial structure, foreign direct investment, energy consumption structure, urbanization level, R&D investment intensity, energy intensity, fixed asset investment, and R&D personnel input.

Description of Data
Green Technology Innovation (GTI): GTI encompasses the advancement and application of novel products and technologies dedicated to environmental protection, pollution reduction, energy and resource conservation, and the promotion of sustainable development. In this context, green patents include inventions, utility models, and design patents specifically related to resource conservation, energy efficiency, and pollution prevention. To align with the definition of GTI, considering design patents only cover product shapes and patterns, this study employs the total number of green invention patents and green utility model patents to measure GTI. Given the time-consuming nature of the patent granting process in China, utilizing the total number of patent applications serves as a prompt and accurate reflection of enterprises’ willingness and motivation to engage in GTI. Moreover, in line with the comprehensive essence of GTI, encompassing both the development and promotion of new clean energy technologies and improvements in ecological environment protection and resource recycling technologies, this research attempts to decompose GTI into Substantive Green Innovation (SubGI) and Symbolic Green Innovation (SymGI).
Substantive Green Innovation (SubGI): Substantive Green Technology Innovation focuses on the development and adoption of environmentally beneficial technologies that result in substantial reductions in resource consumption and carbon emissions. Green invention patents, exhibiting creativity, novelty, and energy-efficient features, align with the fundamental principles of Substantive Green Innovation. Consequently, this study uses the quantity of green invention patent applications as a metric to assess SubGI.
Symbolic Green Innovation (SymGI): Symbolic Green Innovation aims to respond to government environmental policies, emphasizing improvements to existing technologies but typically lacks significant positive environmental impacts. Green utility model patents, representing product or process improvements with energy-saving and emission reduction characteristics, align with the core principles of Symbolic Green Technology Innovation. Thus, this study employs the quantity of green utility model patent applications as a metric to assess SymGI.
Environmental Regulation (ER): In adherence to the polluter-pays principle, China initiated the imposition of emission charges in 1982. By levying charges on enterprises and individuals emitting pollutants, this policy incentivizes the adoption of environmentally friendly measures, reduction of pollutant emissions, and consequent mitigation of environmental pollution and resource wastage. In this study, the emission charges to GDP ratio is used as a measure of ER.
This research incorporates control variables, following the definitions provided by several scholars: gross domestic product per capita (PGDP), industrial structure (IS), foreign direct investment (FDI), energy consumption structure (ECStruc), urbanization level (Urban), research and development investment intensity (RDI), energy intensity (EI), fixed asset investment (Fix), and research and development personnel input (RDP). The definitions of each variable are detailed in Table 10.
Distance Function in Non-Radial Directions
Calculation of Energy-Environmental Efficiency (EEE)
This section of the study focuses on determining the dependent variable, EEE. Leveraging the flexibility of Data Envelopment Analysis (DEA), which does not necessitate a specific functional form, proves effective in measuring the EEE of Decision Making Units (DMU). In this context, DMU pertains to the 30 provinces from 2003 to 2017.
The Non-radial Directional Distance Function (NDDF) emerges as a variant of the DEA method, providing distinct advantages in the assessment of EEE. Firstly, it facilitates the simultaneous consideration of multiple input and output indicators, allowing for a more comprehensive evaluation of regional efficiency. Secondly, NDDF introduces directionality, specifying the optimization direction and yielding more precise assessment results. Moreover, it permits the assignment of varying weights to different input and output indicators during the distance function calculation, reflecting their significance in the evaluation process. In essence, NDDF proves to be more versatile in assessing EEE by comprehensively incorporating multiple indicators and their respective weights, thus contributing to a more accurate evaluation of the efficiency levels of businesses or regions.
The NDDF function is defined based on the principle of output expansion while minimizing pollutant emissions, as follows.

In the formula, where K, L, and E represent input variables, Y stands for the desired output, and C denotes the undesired output. The specific input and output variables are detailed below:
Input Indicators:
- Capital (K): The capital stock is estimated using the perpetual inventory method34.
- Labor (L): Labor is quantified based on the number of people employed at the conclusion of each year in each region.
- Energy (E): Energy consumption is gauged using the consumption of tons of standard coal in each region.
Expected Output:
- Total Output Value of Each Province (Y): This is derived by converting the nominal GDP to 2003 constant price GDP through a price index.
Unintended Output:
- CO2 Emissions (C): CO2 emissions are computed based on the calorific value of consumption of nine energy sources: raw coal, coking coal, crude oil, gasoline, kerosene, diesel fuel, fuel oil, natural gas, and electricity.

Given the panel nature of the data encompassing both time series and cross-sectional dimensions, employing Ordinary Least Squares (OLS) for estimation would yield biased and inconsistent results19. While Generalized Least Squares (GLS) can be utilized for panel data, it fails to consider spatial effects, leading to biased estimates. Additionally, spatial autoregressive model (SAR), spatial error model (SEM), and spatial Durbin’s model (SDM) all address spatial factors. However, SAR overlooks cross-variable autocorrelation, and SEM neglects spatial dependency among variables. Therefore, this study has chosen SDM as the foundational model, extending it to include lag effects and thereby creating a DSDM.
Findings and Discourse
Energy-Environmental Efficiency (EEE)
Year | PTE | SE | ER |
---|---|---|---|
2003 | 0.341*** | 0.107*** | 0.270*** |
2004 | 0.376*** | 0.118*** | 0.288*** |
2005 | 0.392*** | 0.121*** | 0.237*** |
2006 | 0.422*** | 0.127*** | 0.225*** |
2007 | 0.412*** | 0.120*** | 0.203*** |
2008 | 0.408*** | 0.117*** | 0.198*** |
2009 | 0.391*** | 0.109*** | 0.198*** |
2010 | 0.383*** | 0.102*** | 0.197*** |
2011 | 0.377*** | 0.097*** | 0.209*** |
2012 | 0.398*** | 0.101*** | 0.219*** |
2013 | 0.399*** | 0.099*** | 0.240*** |
2014 | 0.400*** | 0.102*** | 0.240*** |
2015 | 0.373*** | 0.097*** | 0.244*** |
2016 | 0.395*** | 0.104*** | 0.247*** |
2017 | 0.412*** | 0.110*** | 0.259*** |
Environmental Regulation (ER)
Year | GTI | SubGI | SymGI |
---|---|---|---|
2003 | 0.263*** | 0.150* | 0.036** |
2004 | 0 |
Model
(5) | (6) | |
---|---|---|
Weight | ||
0.7690*** | 0.7740*** | |
(40.53) | (46.03) | |
0.7608*** | 0.7741*** | |
(43.60) | (41.95) | |
0.7734*** | 0.7689*** | |
(46.25) | (43.51) | |
− 0.0787 | (− 1.31) | |
0.3383* | (1.81) | |
0.0941 | (1.16) | |
− 0.0761 | (− 1.28) | |
0.2852 | (1.52) | |
0.1110 | (1.32) | |
0.0077 | (1.35) | |
0.0048 | (0.93) | |
0.0027 | (0.50) | |
0.0084 | (1.48) | |
0.0021 | (0.39) | |
− 0.0020 | (− 0.36) | |
0.0017* | (3.90) | |
0.0014* | (3.50) | |
0.0013* | (2.78) | |
0.0016* | (3.60) | |
0.0010* | (2.60) | |
0.0008 | (1.60) | |
− 0.0162* | (− 3.07) | |
− 0.1272* | (− 4.99) | |
− 0.0135 | (− 1.16) | |
− 0.0223* | (− 4.24) | |
− 0.1083* | (− 4.28) | |
− 0.0122 | (− 1.05) | |
− 0.0105 | (− 2.67) | Yes |
− 0.0075 | (− 2.06) | Yes |
− 0.0055 | (− 1.46) | Yes |
Yes | ||
Yes | ||
0.0231* | (3.18) | |
0.0609* | (3.08) | |
− 0.0086 | (− 0.69) | Yes |
− 0.0000 | (− 0.22) | |
0.0001 | (0.63) | |
0.0003* | (1.80) | Yes |
Yes | Yes | |
0.0014* | (6.09) | |
0.0031* | (3.13) | |
0.0006 | (1.19) | |
Yes | Yes | Yes |
Spatial rho | 0.0482 | (0.69) |
0.4532 | (2.33) | |
0.1268 | (1.50) | |
0.0080 | (0.11) | |
0.4498 | (2.29) | |
0.1548* | (1.78) | |
Sigma2_e | 0.0004* | (15.50) |
0.0003* | (15.74) | |
0.0004* | (15.62) | |
0.0003* | (15.51) | |
0.0003* | (15.73) | |
0.0004* | (15.64) | |
R-squared | 0.980 | |
0.854 | ||
0.971 | ||
0.974 | ||
0.933 | ||
0.961 | ||
Observations | 420 | |
420 | ||
420 | ||
420 | ||
420 | ||
420 |
Table 4: Effects of Green Technology Innovation (GTI) and Environmental Regulation (ER) on Energy-Environment Efficiency (EEE)
The results in Table 4 unveil the direct, spatial, and total impacts of GTI on EEE while considering the influence of ER. The coefficients representing the total effect of GTI are notably negative under the weighting matrices ( W_1 ) and ( W_2 )65. Moreover, the total effect mirrors the spatial effect, indicating that the restraining impact of GTI on neighboring regions surpasses its stimulating effect on the local region66. A plausible explanation for this phenomenon lies in the resource competition triggered by the promotion of GTI in neighboring areas. This study stands out by incorporating the quadratic term of GTI26,47. Remarkably, the analysis reveals that the coefficient for the quadratic term of GTI is significantly positive30, aligning with the trend of direct effects. This implies the existence of a critical point in the impact of GTI on EEE. Prior to reaching this critical point, the negative spatial spillover effect of GTI on EEE outweighs the positive direct effect, leading to a decline in EEE. However, post the critical point, the positive direct effect of GTI on EEE surpasses the negative spatial spillover effect, ultimately enhancing EEE. The underlying reason for this shift is that as GTI advances, the prerequisites for its success become more defined, fostering cooperative relationships among neighboring regions and expediting the adoption of green technology across various regions.
Similarly, under matrices ( W_1 ) and ( W_2 ), the total effect coefficient of ER on EEE is significantly positive, indicating that the enforcement of ER contributes to EEE19. This result may be attributed to the use of the emission fees to GDP ratio as the ER measurement indicator in this study. Enterprises with excessive emissions face higher emission costs, elevating operational expenses and prompting businesses to update production equipment35. Consequently, this encourages companies to reduce pollutant emissions, leading to an improvement in EEE. Furthermore, the coefficient for the interaction term between ER and GTI is significantly positive, even though the impact of GTI on EEE follows a U-shaped curve. This implies that before GTI reaches its critical point, the implementation of ER mitigates the negative impact of GTI on EEE. The rationale behind this observation could be that in the initial phases of GTI implementation, businesses encounter high sunk costs. The implementation of environmental policies encourages businesses and individuals to invest in the research and application of green technology, thereby increasing financial support for green technology development. Once GTI reaches its critical point, the implementation of ER reinforces the stimulating effect of GTI on EEE. The primary reason for this is that the conditions required for green technology innovation are already mature, and businesses have developed awareness of green technology innovation. At this juncture, the implementation of ER can stimulate long-term investment and collaboration by businesses, expediting the development and application of green technology.

Unveiling the Source of GTI’s Impact on EEE: A Deeper Exploration
While our understanding of the influence of Green Technology Innovation (GTI) on Energy-Environment Efficiency (EEE) and the role of Environmental Regulation (ER) in shaping this impact has advanced, the specific origins of GTI’s influence on EEE still elude us. GTI encompasses both the development of new energy technologies and the enhancement of environmental protection and resource recycling technologies. To unravel this intricacy, the study delved into the decomposition of GTI into Substantive Green Innovation (SubGI) and Symbolic Green Innovation (SymGI), leading to a fresh empirical analysis detailed in Table 5. Under the weight matrix ( W_1 ), the coefficient of SubGI exhibits a significantly negative impact in the total effect, aligning with the trend observed in the GTI coefficient in Table 4. However, the coefficient of SymGI does not emerge as significant. This suggests that the effect of GTI on EEE predominantly emanates from SubGI31, implying that GTI chiefly influences EEE through the development and expansion of new clean energy technologies11. The potential reason behind this lies in the fact that new clean energy technologies not only curtail resource wastage but also diminish reliance on traditional fossil fuels, thereby further reducing greenhouse gas emissions66. Moreover, the role of ER and its impact on SubGI’s influence on EEE aligns consistently with the results obtained before the decomposition of GTI.sni
Leave a Reply