Sage Journals: Discover world-class research

Abstract

Environmental regulation has emerged as an institutional response to the global ecological crisis, leading to discussions on its impact on green technology innovation. However, due to variations in indicator selection and data analysis, the academic community has yet to reach a consensus on this matter. To shed further light on the relationship between environmental regulation and green technology innovation, this study focuses on innovation investment in Chinese provincial regions from 2008 to 2019. By incorporating spatial econometrics, mediation effects, and threshold models, the study investigates the role of innovation investment in the influence of environmental regulation on the process of green technology innovation. The findings reveal the following: (1) The impact of environmental regulation on green technology innovation follows a non-linear “U-shaped” relationship. (2) Innovation investment acts as a mediator in the promotion of green technology innovation by environmental regulation, but this mediating effect exhibits stage-specific characteristics. (3) Environmental regulation has a threshold effect on the promotion of green technology innovation, with a more significant impact observed beyond the threshold value. Overall, these conclusions highlight the crucial role of innovation investment in driving the development of green technology innovation under environmental regulation.

Keywords

environmental regulation green technology innovation innovation investment

Introduction

In the post-pandemic era, the world faces unprecedented challenges in both climate change and economic recovery. The global climate change, exacerbated by the dominance of industrial civilization and its brown economy, has resulted in increasingly frequent extreme weather conditions, such as droughts, floods, and glacial melting (Azarpajouh, 2022; Sempertegui, 2022). The urgency to address these issues has become paramount. Green technology innovation (GT) emerges as a promising solution within this context. GT, as a novel technological system, offers avenues for energy conservation, emission reduction, and mitigating the impact of environmental pollution on the global climate. Moreover, it serves as a catalyst for the development of ecological civilization (Barbieri et al., 2020). Recognizing the significance of GT, the “14th Five-year Plan” explicitly emphasizes the necessity of constructing a market-oriented GT system and strengthening the waste pollution control system (Sun et al., 2017). However, despite its potential benefits, GT faces challenges, particularly regarding its substitutability. Without external intervention, high-polluting industries often prioritize traditional methods due to market scale effects and the perceived profit advantage of conventional innovation. This reluctance to adopt GT exacerbates environmental pollution and hampers progress toward sustainability. Hence, governmental intervention becomes imperative. Environmental regulations (ER) serve as effective tools to incentivize and guide heavy-polluting industry enterprises toward GT adoption (Y. J. Zhang et al., 2017). In essence, the synergy between GT and environmental regulations offers a pathway to address the intertwined challenges of climate change and economic recovery in the post-pandemic era. Through targeted policies and collaborative efforts, governments and industries can pave the way for a greener and more sustainable future.

Green technology innovation encompasses various techniques aimed at reducing energy consumption and environmental pollution across different sectors such as production, energy utilization, engineering construction, and logistics (Chen et al., 2016) . has remarkable positive spillover effects on the environment and stimulates innovation by sharing knowledge. (W. Cai & Li, 2018); environmental positive externalities may lead to market failure, and without external constraints and incentives, companies lack the enthusiasm to GT (Ghisetti & Pontoni, 2015; Marino et al., 2019); undoubtedly distorts the competition between GT and traditional innovation; means that GT is a strongly policy-driven innovation model in the absence of market forces, in which ER is an important driving force for GT (Guo et al., 2017; Johnstone et al., 2017), and scientifically reasonable ER can effectively resolve the positive externalities of GT and solve market failure issues (Hille & Möbius, 2019; F. Z. Wang et al., 2018); there is currently no consensus on whether ER can promote GT (D. Zhang & Vigne, 2021), mainly forming three levels:

(1) The promotion theory suggests that strengthening ER can stimulate companies’ willingness to innovate green technologies (Cainelli et al., 2019; Feichtinger et al., 2005); Feichtinger et al. (2005) and Cainelli et al. (2019) found that mandatory enforcement of policies during the implementation of controlled ER can internalize the external effects of environmental pollution (Xie et al., 2017). By imposing higher costs on pollution control, a “crowding-out effect” can be induced, which can bring energy-saving and emission-reducing pressure to companies and stimulate their willingness to innovate green technologies (X. Yang et al., 2020). On the flip side, incentive-based ER, With their inherent flexibility, companies are granted complete autonomy, enabling them to fully unleash their initiative, enhance investments in green technologies through diverse channels, convert environmental externalities into endogenous ones, and elevate their expectations regarding income generated from research and development (R&D; Xepapadeas & Zeeuw, 1998).

(2) The inhibition theory suggests that the introduction of ER has a “crowding-out effect” on corporate investment in innovation (Guo et al., 2017). The presence of ER escalates the expenses associated with corporate pollution control, consequently diverting funds away from investments in GT. Simultaneously, stringent ER shift significant environmental responsibilities onto enterprises, resulting in augmented investment costs. (Y. J. Zhang et al., 2017). Based on the cost-effectiveness principle, companies will reduce innovation investment. In line with the “pollution haven hypothesis,” companies often relocate polluting industries to regions with less stringent ER, consequently diminishing their contribution to local green technology investments (Wu et al., 2020). Furthermore, Schumpeter’s threshold theory implies that ER can erode the scale and capital advantages held by large enterprises, consequently impeding GT. This is primarily due to escalated production costs and the potential for government rent-seeking behavior, which further exacerbates the negative impact on innovation (Tang et al., 2020).

(3) Nonlinear relationship: mainly refers to the non-synchronous positive and negative effects of ER on GT—the negative effects start to play a role In the present period, we observe the emergence of negative effects, while the positive effects are delayed due to the lengthy cycle of green technological innovation activities and the associated high risks (Jiang et al., 2013). In the short term, companies have limited options and can only choose to adjust the types and quantities of products, which would result in lower corporate profits and directly weaken the company’s R&D capabilities. This is especially true for small-scale enterprises with low profits. However, in the long run, as ER improve, the relative profitability of GT increases. Enterprises can then adapt their technological innovation strategies to align with the level of ER, thereby promoting GT. This dynamic relationship ultimately leads to a U-shaped correlation between GT and the ER system (W. G. Cai & Li, 2019; J. Zhang et al., 2019); However, neighboring provinces exhibit an inverted U-shaped relationship with GT (Dong & Wang, 2019); In addition, Yang et al. (2012) argues that ER will force high-pollution, energy-consuming enterprises to treat waste, squeeze innovation investment, and that there doesn’t appear to be a clear correlation between increased emission reduction expenditures and GT (C. H. Yang et al., 2012). However, incentive ER such as fiscal, taxation and finance can effectively promote regional ER, which also verifies the differences in ER in different regions and causes differences in GT effects (Horváthová, 2012).

It is evident that the academic community has extensively researched the field of ER and GT, leading to a substantial accumulation of research findings. However, a consensus has not yet been reached, possibly due to variances in research subjects, sample types, sample regions, and sample time intervals. These factors have contributed to the divergence of opinions, offering valuable research clues for further in-depth exploration of the topic.

From the micro level, whether to achieve a win-win situation of ER and GT mainly depends on the compensation effect of GT. In the process of ER to resolve the problem of “market failure” caused by GT, innovation investment plays an indispensable key role. The reason is that GT requires more innovation capital investment than traditional innovation, and relying only on limited government innovation investment and financial subsidies cannot effectively make up for the gap of GT capital (Pei et al., 2019); In addition, GT has higher market and R & D risks than traditional technology innovation, and has a long return cycle. Enterprise innovation investment is still the main source of funds and guarantee for current GT; Strict ER can force enterprises to make aggressive venture capital decisions, increase innovation investment, and Facilitate the advancement of GT (Meng & Han, 2017); However, innovation investment of enterprises is still difficult, which is mainly reflected in the fact that ER force enterprises to dispose of pollution ends, increase the environmental treatment costs of enterprises, and form a “squeeze effect” and a “restraint effect” on innovation investment funds. Moreover, under the harsh financing conditions, innovation investment of enterprises is difficult (S. S. Wang et al., 2020). In short, innovation investment is among the key factors influencing the implementation of the “Porter Hypothesis,” but it is also constrained by ER.

Research insufficiency and textual research: Scholars have extensively analyzed the impact of ER on GT from various perspectives, providing a theoretical foundation for our research. However, there is little attention paid in the academic community to how innovation input affects the relationship between ER and GT. Moreover, there is a lack of in-depth analysis of the theoretical relationship and pathways among these three factors, providing a new research direction for this topic. Based on this, our paper constructs a theoretical model to deeply explore the theoretical relationship among ER, GI, and GT. We introduce spatial econometric models, mediation effect models, and threshold models to focus on how GI affects the relationship between ER and GT.

The article makes the following marginal contributions: (1) Innovative Research Perspective: This study adopts a novel approach by integrating GI, ER, and GT into a single research framework, with a focus on the role of innovation investment in the process of environmental regulation affecting green technology innovation. (2) Theoretical Model Construction: The study has established a theoretical model to systematically analyze the impact mechanism of ER on GT from the perspective of GI. It aims to uncover the relationships among these three factors. (2) Pathway of Influence: By considering both internal driving forces and external constraints, it expands the applicability boundaries of the “Porter hypothesis.” The study examines the mediating and threshold effects of ER as drivers for GI, thereby enhancing the mechanism of corporate GT.

The subsequent content arrangement of this article is as follows: The second part presents the construction of the theoretical model, providing a systematic explanation of the theoretical mechanism of the impact of ER on GT from the perspective of innovation investment. The third part focuses on the design of the model and the selection of indicators, laying the foundation for empirical research in the subsequent sections. The fourth part conducts empirical research to validate the relationships among variables. The concluding section summarizes the primary findings and offers policy recommendations.

Theoretical Model Construction

According to the different subjects of Innovation investment Innovation investment can be categorized into traditional industry Innovation investment and green industry Innovation investment. Green industry Innovation investment is mainly aimed at balancing the relationship between economy and environment, and pursuing maximum ecological effects, while traditional industry Innovation investment is aimed at maximizing market profits. Under the strong ER, Traditional industry innovation investment and green industry innovation investment have varying effects on GT. To scientifically unlock the mechanism of the role of enterprise Innovation investment in the process of ER affecting GT, this paper draws on the research of scholars such as Ji et al. (2021) and constructs a theoretical model in this paper.

Environmental Regulation and Innovation Investment

Assuming that the production functions corresponding to the Innovation investment of traditional industry and green industry are $y_{1} (x_{1}), y_{2} (x_{2})$ respectively, in terms of energy-saving and emission reduction effects, traditional industry is more likely to lean toward traditional technology innovation without ER constraints, while green industry Innovation investment pays more attention to the balance between economy and environment, leaning toward GT. Additionally, the function of energy-saving and emission reduction caused by Innovation investment can be designated as $R (y_{1}, y_{2})$ . Therefore, under the constraint of ER, the optimal production decision can be obtained as follows:

\begin{matrix} Max Π = P [(1 - G_{1} + S_{1}) y_{1} (x_{1}) + (1 - G_{2} + S_{2}) y_{2} (x_{2})] - C \\ st . R (r_{1} (x_{1}), r_{2} (x_{2})) \leq ER \end{matrix}

(1)

Among them, $G_{1}, S_{1}$ are the proportion of GT investment and government subsidy in traditional industry innovation investment, $G_{2}, S_{2}$ are the proportion of green R & D investment and government subsidy in green industry innovation investment, and Environmental Regulation is the ER.

From this, according to the Lagrange multiplier method, under the objective constraint function, the optimal decision equation is:

\begin{matrix} E = P {(1 - C_{1} + S_{1}) y_{1} (x_{1}) + (1 - C_{2} + S_{2}) y_{2} (x_{2})} \\ - C + λ R [y_{1} (x_{1}), y_{2} (x_{2})] \end{matrix}

(2)

\frac{\partial E}{\partial x_{1}} = (1 - G_{1} + S_{1}) {y^{'}}_{1} (x_{1}) + λ {r^{'}}_{1} (x_{1}) = 0

(3)

\frac{\partial E}{\partial x_{2}} = (1 - G_{2} + S_{2}) {y^{'}}_{2} (x_{2}) + λ {r^{'}}_{2} (x_{2}) = 0

(4)

r'_{1} (x_{1}) / y'_{1} (x_{1}) = - (1 - G_{1} + S_{1}) / λ

(5)

r'_{2} (x_{2}) / y'_{2} (x_{2}) = - (1 - G_{2} + S_{2}) / λ

(6)

\begin{matrix} 令 M_{1} = \frac{{r^{'}}_{1} (x_{1})}{{y^{'}}_{1} (x_{1})}, M_{2} = \frac{{r^{'}}_{2} (x_{2})}{{y^{'}}_{2} (x_{2})} \Rightarrow \\ \frac{M_{1}}{M_{2}} = \frac{{r'}_{1} (x_{1}) / {y'}_{1} (x_{1})}{{r'}_{2} (x_{2}) / {y'}_{2} (x_{2})} = \frac{1 - G_{1} + S_{1}}{1 - G_{2} + S_{2}} \end{matrix}

(7)

$M_{1}, M_{2}$ represent the output of traditional and green industry GT under ER. Based on the strength of ER, enterprise innovation investment is divided into two stages:

(1) Under the scenario of low ER, driven by profit, the traditional innovation model dominates, and rational enterprises may choose to “free ride” and reduce investment in GT. At this time, government subsidies tend to focus on GT R&D, resulting in $S_{2} > S_{1}$ . At this time, since the green industry focuses on green innovation, in the absence of external ER factors, the green innovation investment of the green industry is significantly higher than that of the traditional industry, resulting in $G_{2} > G_{1}$ . Therefore, it can be inferred that:

\frac{1 - G_{1} + S_{1}}{1 - G_{2} + S_{2}} < 1, \Rightarrow \frac{r' (x_{1}) / y' (x_{1})}{r' (x_{2}) / y' (x_{2})} < 1 \Rightarrow M_{1} < M_{2}

(8)

When ER is low, under the dominance of profit-driven traditional innovation, the scale of GT R&D investment within traditional industries is continuously reduced, indirectly weakening GT.

(2) Under the scenario of high ER, the government increases the punishment for waste emissions. Under regulatory pressure, enterprise innovation investment gradually shifts from traditional innovation to GT. In the market competition and scale effect, traditional industries will have to pay more investment costs on the road of green innovation than the green industry, that is, $G_{1} > G_{2}$ . At this time, traditional industries have more breakthrough potential in GT. To encourage traditional industries to renew and transform and promote green development, government subsidies tend to focus more on the green innovation investment of traditional enterprises, resulting in $S_{1} > S_{2}$ . Therefore, it can be inferred that:

\frac{1 - G_{1} + S_{1}}{1 - G_{2} + S_{2}} > 1, \Rightarrow \frac{R' (x_{1}) / y' (x_{1})}{R' (x_{2}) / y' (x_{2})} > 1 \Rightarrow M_{1} > M_{2}

(9)

That is, when ER are at a high level, the high cost of waste disposal forces companies to invest in GT, and the “innovation compensation” effect gradually emerges.

Innovation Investment and GT

$M_{1}, M_{2}$ respectively represent the output of GT in traditional industries and green industries under ER. Therefore, the output of social GT that can be introduced is:

M = \frac{R^{'} x}{Y^{'} (x)} = ε_{1} M_{1} + ε_{2} M_{2}

(10)

Among them, $R, Y$ are social energy conservation and emission reduction and social production functions, and $ε_{1}, ε_{2}$ are the contributions of GT under the innovation investment of traditional industries and green industries, respectively, meeting $, ε_{1} + ε_{2} = 1,$ ; When the proportion of GT in traditional industries increases to increase the contribution of GT to $Δ ε_{1}$ , the contribution rate of GT in green industries decreases by $Δ ε_{1}$ . From this, it can be seen that the increase in GT in society is:

Δ M = Δ ε_{1} M_{1} + Δ ε_{2} M_{2} = Δ ε_{1} M_{1} - Δ ε_{1} M_{2} = Δ ε_{1} (M_{1} - M_{2})

(11)

When ER is at a low level, the output of GT in the green industry is relatively high, which leads to $M_{1} - M_{2} < 0$ . With the increase in innovation investment by enterprises, their contribution to GT also increases, leading to $Δ ε_{1} > 0$ . Therefore, we can conclude that $Δ M < 0$ . This means that with low ER, increasing innovation investment by enterprises can further improve the output of GT in the green industry.

That is, when the ER is at a low level, the investment in green innovation in traditional industries is insufficient, which is not conducive to the output of GT in society as a whole.

(2) When ER are at a high level, traditional industries will increase their investment in green innovation and promote the output of GT under high-pressure policies, at which point $M_{1} - M_{2} > 0$ can be introduced. With the increase of innovation investment in traditional industries, their contribution to GT continues to increase, that is, $Δ ε_{1} > 0$ , and it can be concluded that $Δ M > 0$ that is, when ERs are at a high level, enterprises increase their investment in green innovation and improve the overall output of GT in society.

It can be seen that there is a certain nonlinear relationship between innovation investment and GT, and the promotion or inhibition depends on the intensity of ER. From the above theoretical models, it can be seen that the ER has different effects on GT at different stages.

Model Design and Indicator Selection

Model Design

Direct effect model: To elucidate the non-linear correlation between environmental regulations (ER) and green technology innovation (GT), the model will integrate both primary and secondary dimensions of ER. This approach aims to delineate how varying stages of ER distinctly influence GT, thereby providing a comprehensive understanding of the ER-GT relationship.

The SDM model combines the advantages of the SAR and SEM models and can effectively analyze the impact of ER on GT. Therefore, this article will be based on the SDM model to construct relevant models, as follows:

\begin{matrix} G T_{it} = β_{0} + ρ WG T_{it} + β_{1} E R_{it} + β_{2} ER R_{it} + β_{3} X_{it} \\ + θ_{1} WG T_{it} + θ_{2} E R_{it} + θ_{3} ER R_{it} + θ_{3} W X_{it} + ε_{it} \end{matrix}

(12)

Where $G T_{it}, E R_{it}, ER R_{it}$ represent green technology innovation, environmental regulation, and the quadratic term of environmental regulation in region i in year t, $X_{it}$ is the control variable set, W is the spatial weight matrix, and $ε_{it}$ is the random disturbance term. When the value in formula 10 is equal to $θ_{i}$ =0, the SDM model will transform into the SAR model, and when $θ_{i} = - β_{i}$ , the SDM model will transform into the SEM model.

Spatial weight matrix: Geographical Weight Matrix, Economic Weight Matrix, and Neighborhood Weight Matrix are widely utilized as weight matrices in spatial measurement models, and economic development is the key to regional GT. However, the knowledge stickiness between provinces and governments and “top-by-top competition” make the policies of neighboring provinces overflow and imitate GT, which indirectly contributes to local GT. Based on this, the spatial metrology model will be analyzed by introducing the geographical weight matrix.

Mediation model: To thoroughly examine the implications of ER on the GT, further research is needed, this paper introduces Innovation investment (GI) and construct a mediation model that specifically explores the role of innovation investment as a mediator.

\begin{matrix} G T_{it} = β_{0} + ρ WG T_{it} + β_{1} E R_{it} + β_{2} ER R_{it} + β_{3} X_{it} + θ_{1} WG T_{it} \\ + θ_{2} E R_{it} + θ_{3} ER R_{it} + θ_{3} W X_{it} + ε_{it} \end{matrix}

(13)

\begin{matrix} G I_{it} = β_{0} + ρ WG I_{it} + β_{1} E R_{it} + β_{2} ER R_{it} + β_{3} X_{it} + θ_{1} WG I_{it} \\ + θ_{2} E R_{it} + θ_{3} ER R_{it} + θ_{3} W X_{it} + ε_{it} \end{matrix}

(14)

\begin{matrix} G T_{it} = β_{0} + ρ WG T_{it} + β_{1} E R_{it} + β_{2} ER R_{it} + β_{3} GI + β_{4} X_{it} \\ + θ_{1} WG T_{it} + θ_{2} E R_{it} + θ_{3} ER R_{it} + θ_{3} G I_{it} + θ_{4} W X_{it} + ε_{it} \end{matrix}

(15)

Threshold Model: To promote the role mechanism of ER in GT more scientifically, the research introduced GI as a threshold variable, focusing on the nonlinear relationship between ER and GT. The threshold model is as follows:

G T_{it} = a_{0} + a_{1} E R_{it} {GI \leq ϕ} + a_{2} E R_{it} {GI \geq ϕ} + a_{3} X_{it} + ε_{it}

(16)

Indicator Selection

The variable data utilized in this study is derived from provincial-level panel data spanning from 2008 to 2019. However, it should be noted that due to significant data gaps in Tibet, it was not included in the analysis. The data primarily originate from 30 provinces and cities in China, with data sources including the “China Statistical Yearbook,” the “China Environmental Statistical Yearbook,” the “China Science and Technology Statistical Yearbook,” as well as the website of the State Intellectual Property Office of China (SIPO) and the National Intellectual Property Administration (CNIPA). These yearbooks and websites serve as the primary sources of raw data.

Dependent Variable

Green technological innovation (GT): Green technology innovation (GT) makes pivotal contributions in reduction of environmental burden. The measurement of GT has not reached consensus according to existing literature, among which, indicators regarding efficiency and patent are often applied. However, GT efficiency includes indicators that are not highly related to environmental pollution, which is less representative. Therefore, this paper selects the number of green technology invention patent grants to represent the progress of green technology. By utilizing the International Patent Classification (IPC) codes for green patents listed in the Green Patent Index provided by the World Intellectual Property Organization (WIPO; http://www.wipo.int/classifications/ipc/en/est), and setting parameters such as patent type, IPC classification codes, and addresses of the inventing entities (individuals), patent data at the provincial level is obtained from the State Intellectual Property Office of China’s patent publication and announcement website (Dong & Wang, 2019) .

Core Explanatory Variable

Environmental Regulation (ER): Scholars have measured environmental regulation from various perspectives, such as environmental policy, investment in environmental governance, pollution emissions, and environmental management performance. However, these indicators only reflect the level of environmental regulation from one aspect and do not capture its comprehensive level. Following the approach of Kang and others, this paper uses the ratio of regional gross domestic product (GDP) to total regional energy consumption to reflect the intensity of environmental regulation. This ratio can indicate the overall effect of regional environmental regulations. The higher the ratio, the more evident the energy-saving and emission-reduction effects are under a given level of GDP, which also means a greater intensity of environmental regulation (Kang & Ru, 2020) .

Mediation Variables (Threshold Variables)

Innovation Investment (GI): To scientifically represent innovation investment, this paper adopts Gu’s research findings and calculates the capital stock of innovation investment using the perpetual inventory method (Gu & Zhao, 2021) . The specific calculation formula is as follows:

K_{t} = (1 - δ) K_{t - 1} + RD I_{t}

(17)

K_{0} = RD I_{1} / (g + δ)

(18)

$K_{t}, K_{t - 1}$ respectively indicate the capital stock of innovation investment in T period and T-1period. $RD I_{t}$ is the innovation investment in T period, $δ$ is the depreciation rate of innovation investment, $δ$ =15%, and g is the growth rate of innovation investment, and the actual growth rate of innovation investment corresponding to the sample period.

Control Variables Cover FDI, ISU, URB, GF

Foreign Direct Investment (FDI):Foreign investment introduces capital and advanced technology, stimulating the vitality of enterprises’ green innovation, and positively impacts the research and development of green technologies. Based on this, the paper draws on the research findings of Sun et al., measuring it by the logarithm of the total actual foreign direct investment (Sun et al., 2017).

Urbanization level (URB) influences green technology innovation through agglomeration effects and improved infrastructure, which accelerate technology exchange and application. Based on this, the paper characterizes urbanization level by the proportion of urban population to total population at year-end (Guo & Tan, 2024)

Industrial structure upgradin (ISU): Industrial structure upgrading specifically denotes the transition of industries from low-end to high-end sectors, facilitating green technology innovation through optimized resource allocation. Regarding the measurement of industrial structure upgrading, there is no unified standard within the domestic academic community yet. To scientifically represent the level of domestic industrial structure upgrading, this paper constructs an index based on Wang’s method for industrial structure upgrading, with the specific formula as follows (W. Wang et al., 2015):

ISU = W_{1} + 2 W_{2} + 3 W_{3}

(19)

Where $W_{1}, W_{2}, W_{3}$ represent the proportions of the three major industries, and 1, 2, 3 are the weights assigned to these industries respectively. A higher composite index indicates a more significant degree of industrial structure upgrading.

Government support (GS): Government support boosts green technology innovation through financial incentives, regulations, and fostering industry-research collaboration, accelerating sustainable development. Based on this, the paper characterizes government support for green technology innovation by the proportion of innovation investment in government expenditures.

To scientifically reflect the accuracy of variable data, the paper conducts a descriptive statistical analysis of the variables, as detailed in Table 1.

Table 1.

Descriptive Statistics.

Variables		Obs	Mean	SD	Min	Max
Dependent variable	GT	360	6.759	1.496	2.079	10.097
Core explanatory variable	ER	360	1.451	0.704	0.373	4.844
Mediation variables	GI	360	15.641	1.402	11.436	18.357
Control variables	FDI	360	12.719	1.649	7.31	15.09
	ISU	360	2.348	0.131	2.102	2.832
	URB	360	55.521	13.186	29.11	89.6
	GF	360	0.169	0.102	0.057	0.793

Empirical Research

Direct Effect

Spatial correlation analysis: Due to the stickiness of knowledge and competition between governments, there is spillover and duplication of policies among neighboring provinces. Hence, it is crucial to examine the spatial correlation between enterprise GI, ER, and GT. In this regard, a weight distance matrix was constructed using geographical distance in the article, and a global Moran’s I index was calculated. The results show that the Moran’s I index of enterprise GI, ER, and GT is significantly positive, indicating a certain spatial correlation among Chinese provinces and cities(as detailed in Table 2). This provides a scientific basis for using spatial econometrics as an analytical tool.

Table 2.

Moran’s I Index of Core Variables From 2008 to 2019.

Year	ER	GI	GT	Year	ER	GI	GT
2008	0.595*** (3.838)	0.339*** (2.301)	0.360*** (2.436)	2014	0.546*** (3.602)	0.379*** (2.539)	0.224* (1.638)
2009	0.594*** (3.839)	0.345*** (2.335)	0.237** (1.752)	2015	0.527*** (3.494)	0.387*** (2.588)	0.231** (1.769)
2010	0.582*** (3.760)	0.346*** (2.343)	0.253** (1.794)	2016	0.482*** (3.193)	0.395*** (2.636)	0.269** (1.908)
2011	0.570*** (3.706)	0.353*** (2.381)	0.255** (1.848)	2017	0.513*** (3.401)	0.399*** (2.658)	0.250** (1.754)
2012	0.558*** (3.635)	0.362*** (2.435)	0.254** (1.820)	2018	0.524*** (3.481)	0.405*** (2.696)	0.241** (1.713)
2013	0.537*** (3.544)	0.372*** (2.492)	0.237** (1.728)	2019	0.526*** (3.472)	0.409*** (2.721)	0.274** (1.903)

***

p < 0.01. **p < 0.05. *p < 0.1.

Spatial Model Selection: After conducting a comprehensive comparison of multiple spatial econometric models, the fixed effect model was determined to be more appropriate for current study than the random effect model, as indicated by the Hausman test (p-value = 0; Anselin, 1988). Moreover, the LM and R-LM tests provided evidence in favor of the SAR model as opposed to the SEM model(as detailed in Table 3). Additionally, the WALD and LR tests confirmed the significance of the SDM model and its inability to be simplified into SAR or SEM. Hence, following analysis will primarily focus on the SDM model.

Table 3.

Model Test.

Test method	Statistical value	p	Test method	Statistical value	p
LM-Spatial_Lag	29.767	.000	Wald-Spatial_Lag	33.89	.000
Robust-LM-Spatial_Lag	43.636	.000	LR-Spatial_Lag	82.93	.000
LM-Spatial_Erro	14.001	.000	Wald-Spatial_Erro	96.82	.000
Robust-LM-Spatial_Erro	27.870	.000	LR-Spatial_Erro	89.69	.000
Hausman	67.653	.000

ER and GT: The specific results and implications of Model 1 SDM are likely presented in Table 4, The respective impacts of ER and its quadratic term on GT are −0.595 and 0.086, both statistically significant at a 1% level. This indicates a significant “U-shaped” relationship between ER and GT. ER can effectively promote GT, but it needs to reach a certain level of intensity. The reason for this is that, as profit-driven entities, companies tend to “free-ride” on low levels of government ER and are more inclined toward traditional technological innovation, which can lead to a “crowding-out” effect on GT. However, when ER is at a higher level, companies are incentivized to enhance their investment in GT through the imposition of environmental protection taxes and pollution permit trading fees, which compensates for the costs of innovation and leads to a “compensation effect” that promotes the development of GT.

Table 4.

Estimated Results of ER and GT.

GT	SDM: Model1		SRA: Model2		SEM: Model3
GT	$β$	T	$β$	T	$β$	T
ER	–0.595***	–3.480	–0.113	–0.790	–0.101	–0.650
ERR	0.086***	2.550	0.024	0.760	0.018	0.610
FDI	–0.068***	–2.730	–0.078***	–2.880	–0.060**	–2.160
ISU	–0.812**	–2.080	–0.903**	–2.150	–1.012**	–2.510
GF	–0.207	–0.280	–0.568	–0.780	–1.313**	–1.810
URB	0.022***	3.380	0.033***	4.890	0.025***	3.330
W*ER	0.569***	2.460
W*ERR	–0.066*	–1.820
W*FDI	–0.117***	–3.070
W*ISU	0.844	1.350
W*GF	3.427***	3.530
W*URB	0.073***	7.100
$ρ$	0.125**	2.220	0.292	5.870
$λ$					0.326***	5.600
Individual fixed	Control		Control		Control
Time fixed	Control		Control		Control
R-sq	.9275		.4118		.4037
Log-L	118.0275		76.5630		73.1845

***

p < .01. **p < .05. *p < .1.

Endogeneity Test

Due to omitted variables or reciprocal relationships between variables, endogeneity issues may arise, affecting the scientific validity of the conclusions. Based on this, the study employs the SYS-GMM model to conduct an endogeneity test, aiming to explore the scientific validity of the research findings. The research conclusions are presented in Table 5.

Table 5.

Endogeneity Test.

GT	$β$	St.Err.	t-Value	p-Value	CI
L.GT	0.655***	0.073	8.95	.000	[0.512, 0.799]
ER	–0.048***	0.012	–3.88	.000	[–0.072, –0.024]
ERR	0.113***	0.030	3.72	.000	[0.054, 0.173]
FDI	0.013	0.016	0.80	.426	[–0.018, 0.044]
ISU	0.594	0.493	1.21	.228	[–0.372, 1.559]
GF	0.054	0.037	1.44	.149	[–0.019, 0.127]
URB	–0.306*	0.159	–1.93	.054	[–0.616, 0.005]
Constant	0.912**	0.372	2.45	.014	[0.183, 1.641]
AR(1)	–2.542***	AR(2)	0.654	SARGAN	0.875

Note. ERR = ER*ER.

***

p < .01. **p < .05. *p < .1.

Table 5 reveals a significant positive coefficient for the lagged variable of GT (L.GT), indicating the presence of the “ratchet effect.” This analysis warrants the use of SYS-GMM. Furthermore, the absence of second-order serial correlation in the error terms, as confirmed by the AR(2) test, along with the validity of the results established by the Sargan test, reinforces the reliability of the SYS-GMM regression results.

From Table 5, it is evident that the first-order term of ER significantly inhibits green technology innovation, while the second-order term promotes it. Thus, considering endogeneity issues, there still exists a U-shaped non-linear relationship between ER and GT, confirming the accuracy of the conclusions.

Mediation Effect

According to Model 4 in Table 6, there is a “U”-shaped nonlinear relationship between ER and GT. Only when ER reaches a certain level can it significantly promote GT, confirming the existence of Porter’s hypothesis. Model 5 reflects the impact of ER on GI. The first-order term of ER has a driving effect on GI, however, it lacks statistical significance. The quadratic term significantly influences enterprise innovation investment. Hence, it is only when ER reaches a high level and the associated regulatory costs become burdensome for enterprises that they are compelled to increase their investment in technology innovation. This leads to an innovation compensation effect, which helps counterbalance the negative impact of compliance costs on enterprises.

Table 6.

Test of the Mediating Effect of ER on GT: GI as the Mediating Variable.

Mediating Effect	GT: Model4		GI: Model5		GT: Model6
Mediating Effect	$β$	T	$β$	T	$β$	T
ER	–0.595***	–3.480	0.019	0.250	–0.642***	–3.870
ERR	0.086***	2.550	0.041**	2.700	0.074**	2.240
GI					0.550***	4.850
FDI	–0.068***	–2.730	0.008	0.700	–0.070***	–2.910
ISU	–0.812**	–2.080	0.028	0.160	–0.846**	–2.250
GF	–0.207	–0.280	–1.394***	–4.190	0.731	0.980
URB	0.022***	3.380	0.018***	5.970	0.015**	2.270
W*ER	0.569***	2.460	0.030	0.280	0.684**	2.920
W*ERR	–0.066*	–1.820	0.029	1.370	–0.090**	–2.000
W*GI					–0.403**	–2.320
W*FDI	–0.117***	–3.070	–0.006	–0.360	–0.105**	–2.820
W*ISU	0.844	1.350	–0.431	–1.530	1.035*	1.810
W*GF	3.427***	3.530	0.275	0.630	3.103***	3.280
W*URB	0.073***	7.100	0.030***	6.260	0.070***	5.640
$ρ$	0.125**	2.220	0.156**	2.780	0.134**	2.400
R-sq	.9275		0.9430		0.9380
Log-L	118.0275		114.0292		114.0943

***

p < .01. **p < .05. *p < .1.

Based on Model 4, a mediating variable was added—innovation investment, forming Model 5. From Model 6, it is found that after incorporating innovation investment, the quadratic term of ER continues to have a significant positive impact on the promotion of GT, but the first-order term inhibits it. Compared with Model 4. The indirect effect of ER on GT through GI declined from 0.086 to 0.074. This suggests that the causal path of “ER → GI → GT” is established, and GI acts as a mediating role. However, this mediating effect is contingent upon the specific stage of development and the strength of ER, and the role of enterprise innovation investment cannot be ignored.

Threshold Model

The direct and mediating effects of ER on GT analysis demonstrate a non-linear relationship, characterized by a “U-shaped” pattern, between ER and GT, the mediating effect of innovation investment has stage-specific features. To scientifically analyze the variations in the influence of ER on GT at different stages of innovation investment, this study, in collaboration with Hansen’s threshold model. The Bootstrap sampling method was utilized to explore the threshold value of the mediating effect of innovation investment, as shown in Table 7.

Table 7.

Threshold Test.

Innovation investment as a threshold variable
N	F	p	10%	5%	1%	Threshold value	CI
1	29.770	.096	29.215	35.240	44.073	16.400	[16.327, 16.431]
2	28.250	.160	49.711	73.141	117.151	17.711	[17.600, 17.777]

Note. trim(0.01 0.03) grid(100) bs(500 500).

From Table 7, it can be seen that enterprise Innovation investment passed the single threshold effect test but failed the double threshold test. Therefore, it can be concluded that there is a single threshold effect of enterprise Innovation investment in the process of ER promoting GT. Furthermore, Figure 1 presents the relationship between threshold estimation values and LR. The corresponding confidence interval for the threshold value of 16.400 is [16.327, 16.431]. Confidence interval corresponding to the threshold value of 16.400 is [16.327, 16.431]. The critical value of the likelihood ratio test at the 5% significance level is 7.35, based on which it can be considered that the threshold effect estimation value of Innovation investment is reliable and valid.

Figure 1.

Estimate of innovation investment threshold and confidence interval.

Following the threshold test, an innovation investment threshold variable was utilized to conduct a panel regression analysis. The regression results presented in Table 8 indicate that there is a positive but statistically insignificant impact of ER on GT when the level of innovation investment is relatively low (GI < 16.400). This finding is in line with the theoretical analysis presented in this paper, which suggests that when enterprise innovation investment is at a relatively low level, traditional innovation profit-driven factors hinder breakthroughs in GT, and it is difficult to offset the “crowding out effect” of ER on environmental pollution end-of-pipe disposal resources. Furthermore, as the primary source of investment in GT, the government may lack the same level of sensitivity to market information as enterprises. This can make it challenging to capture and respond to the social demand for GT in a timely manner, hindering the formation of a mutually beneficial relationship between ER and GT.

Table 8.

Shows the Threshold Estimation Results.

GT		$β$	SE	T	p > t	CI
FDI		–0.056	0.040	–1.410	0.159	[–0.134, 0.022]
ISU		1.633***	0.468	3.490	0.001	[0.711, 2.554]
GF		7.004***	0.946	7.410	0.000	[5.144, 8.865]
URB		0.111***	0.007	16.660	0.000	[0.098, 0.124]
ER	GI < 16.400	0.137	0.100	1.360	0.174	[–0.061, 0.334]
ER	GI > 16.400	0.330***	0.108	3.050	0.002	[0.118, 0.543]

***

p < 0.01. **p < 0.05. *p < 0.1.

Further analysis reveals that when the level of innovation investment in enterprises exceeds the threshold value of GI > 16.400, the effect of ER on GT becomes statistically significant at the 1% level, with a path coefficient of 0.330(as detailed in Table 8). This indicates that ER can significantly promote GT when enterprise innovation investment surpasses the threshold. As the level of innovation investment increases, a higher proportion of funds is allocated to GT, driving breakthroughs in green technology and partially compensating for the “crowding-out effect” of ER on enterprise financial resources. This creates an “innovation compensation effect” and achieves a dual impact of ER on GT. The research conclusion indirectly reflects the threshold effect of innovation investment in promoting GT through ER.

Conclusion and Policy Recommendations

Conclusion and Discussion

Based on the above research conclusion, the study integrates GI, ER, and GT into a unified research framework, constructing a theoretical model to systematically delineate the relationships among them, thus revealing the theoretical “black box” among the three. Building upon this, utilizing spatial econometric models, mediation models, and threshold models with support from provincial panel data, the study elucidates the underlying mechanisms of how GI mediates the promotion of GT by ER. The research findings are as follows:

(1) Direct effect: ER has a significant nonlinear relationship with GT, meaning that while it has a certain positive influence on GT, it needs to reach a certain level of stringency to be evident. This conclusion integrates the findings of Tan and Xy’s research (Tang et al., 2020; X. Yang et al., 2020), and is in line with the conclusions of scholars such as Cai, Zhang, and Dong (W. G. Cai & Li, 2019; Dong & Wang, 2019; J. Zhang et al., 2019). The differences may stem from variations in the measurement of environmental regulation, selection of research subjects, and choice of control variables.

(2) Mediation effect: In the process of driving green technology innovation, innovation investment acts as a certain intermediary under environmental regulation. However, this intermediary effect demonstrates a certain stage-specificity: only when environmental regulation reaches a certain level does the intermediary effect of innovation investment become more pronounced. This aspect has received relatively less attention in past research. Traditional studies typically treat innovation investment and environmental regulation as independent factors (Dong & Wang, 2019; Yu et al., 2017; Liu et al., 2023), overlooking the potential complex interactions between them. Therefore, by incorporating innovation investment, this study offers a more comprehensive understanding of the mechanism through which environmental regulation influences green technology innovation, thereby enriching existing research (W. G. Cai & Li, 2019; Dong & Wang, 2019; J. Zhang et al., 2019; Tang et al. (2020; X. Yang et al., 2020) .

(3) Thresholdeffect: In the process of driving green technology innovation, innovation investment exhibits a certain threshold effect under environmental regulation. That is, when innovation investment reaches a certain level, the impact of environmental regulation on green technology innovation becomes more significant and robust. This research conclusion further broadens the understanding of the role of environmental regulation in green technology innovation (Horbach et al., 2012 ), enriching the theoretical framework of environmental regulation for green technology innovation (Porter & Linde, 1995 ).

In summary, this study not only expands the understanding of the relationship between ER, GI, and GT in theory but also provides a unified research framework in methodology. It helps to deeply understand the complex interaction among the three, offering important theoretical support for promoting green technology innovation.

Research Limitations

(1) Lack of detailed research on classification indicators: This study examines ER as a whole, without categorizing them, it is worth noting that such regulations can be classified as command-and-control, market-based, and participatory. By exploring the impact of ER on GT from multiple dimensions, a more objective understanding of their effectiveness can be achieved. Unfortunately, due to data limitations, this paper did not conduct a detailed study on the classification of ER. As such, Exploring the influence of various types of employee regulations on the effectiveness of GT remains an essential field for future research.

(2) Lack of historical comparative studies: From the perspective of investment in innovation, this study presents compelling empirical evidence supporting the link between ER and the advancement of GT. It is important to note that the statistical data used in this paper is limited to after 2008. Obtaining accurate data for the period before 2008, especially during the implementation of the national industrial development promotion strategy, is difficult due to existing conditions. However, a comparison of historical and existing statistical data would enhance the explanatory power of ER for GT from the perspective of innovation investment at a higher level.

(3) Theoretical model construction lacks consideration of variables: The constructed theory only focuses on costs, prices, and specifically examines the relationships between GI, ER, and GT. It fails to explore key elements such as Return on Assets (ROA). In future model construction processes, it would be beneficial to include the influence of multidimensional parameter variables to achieve a more scientifically rigorous analysis of the relationships between variables.

Policy Recommendations

(1) Develop ER tailored to local conditions. Based on regional characteristics, clarify the goals of ER under the context of green development, and scientifically and reasonably develop ER policies that are in line with regional development to ensure the credibility and effectiveness of policies. Establish regional environmental regulatory monitoring mechanisms in sections to adjust government ER efforts in a timely manner, Facilitate the advancement of regional GT systems and strengthen the depth of GT systems.

(2) Use ER as a guide for enterprise innovation investment. Enterprises are profit-driven, and their goal is to maximize profits through innovation investment. However, under traditional profit-oriented incentives, companies tend to invest funds in industries that are highly polluting and energy-consuming, which is not conducive to GT. To address this issue, the government should develop scientifically reasonable ER based on regional industrial characteristics and environmental governance pressure. By using external regulatory forces, the government can steer enterprise innovation investment toward green industries, promoting the progress of GT. This approach aims to strike a balance between economic development and environmental governance, ultimately achieving a win-win situation.

(3) Promote enterprise innovation investment. Chinese provincial governments should formulate scientifically reasonable policies to encourage innovation investment, taking into account environmental governance objectives and the stage of economic structural transformation. These policies should guide investment funds from enterprises toward green industries, Consequently, it will invigorate the growth of GT in Chinese provinces. The ultimate goal is to foster green development and achieve high-quality economic growth in these provinces.

Footnotes

ORCID iD

Chengliang Du

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The National Social Science Fund of China (24BGL207).

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data Availability Statement

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

References

Anselin

(1988). The Scope of Spatial Econometrics[M] Spatial Econometrics: Methods and Models. Springer.

Azarpajouh

(2022). Impact of climate change on broiler housing systems. Poultry World, 38(2), 119-124.

Barbieri

Marzucchi

Rizzo

(2020). Knowledge sources and impacts on subsequent inventions: Do green technologies differ from non-green ones? Research Policy, 49, 103901.

Cainelli

D’Amato

Mazzanti

(2019). Resource efficient eco-innovations for a circular economy: Evidence from EU firms. Research Policy, 49(1), 103827.

Cai

(2018). The drivers of eco-innovation and its impact on performance: Evidence from China. Journal of Cleaner Production, 176, 110–118.

Cai

W. G.

Q. Q.

(2019). The dual effect of environmental regulations on eco-technology innovation of enterprises. Science Research Management, 40(10), 87–95.

Chen

Han

Liu

(2016). Green technology innovation and energy intensity in China. Natural Hazards, 84(1), 317–416.

Dong

Z. Q.

Wang

(2019). Local-neighborhood effect of green technology of environmental regulation. China Industrial Economics, (01), 100–118.

Feichtinger

Hartl

R. F.

Kort

P. M.

Veliov

V. M.

(2005). Environmental policy, the porter hypothesis and the composition of capital: Effects of learning and technological progress. Journal of Environmental Economics and Management, 50(2), 434–446.

10.

Ghisetti

Pontoni

(2015). Investigating policy and R&D effects on environmental innovation: A meta-analysis. Ecological Economics, 118, 57–66.

11.

Zhao

(2021). How can China break through the dilemma of scientific and technological innovation? A new perspective based on the coordinated development of scientific innovation and human capital. Studies in Science of Science, 39(01), 129–138.

12.

Guo

Tan

(2024). Analyzing the synergistic influence of green credit and green technology innovation in driving the low-carbon transition of the energy consumption structure. Sustainable Energy Technologies and Assessments, 63, 103633.1-103633.11.

13.

Guo

L. L.

Tseng

M. L.

(2017). The interaction effects of environmental regulation and technological innovation on regional green growth performance. Journal of Cleaner Production, 162, 894–902.

14.

Hille

Möbius

(2019). Environmental policy, innovation, and productivity growth: Controlling the effects of regulation and endogeneity. Environmental and Resource Economics, 73(4), 1315–1355.

15.

Horváthová

(2012). The impact of environmental performance on firm performance: Short-term costs and long-term benefits? Ecological Economics, 84, 91–97.

16.

Horbach

Rammer

Rennings

(2012). Determinants of eco-innovations by type of environmental impact: the role of regulatory push/pull, technology push and market pull. Zew Discussion Papers, 78(32), 112–122.

17.

Jiang

F. X.

Wang

Z. J.

Bai

J. H.

(2013). The dual effect of environmental regulations’ impact on Innovation-an empirical study based on dynamic panel data of Jiangsu manufacturing. China Industrial Economics.

18.

Zhang

(2021). The impact of environmental regulation on green total factor productivity from the perspective of private investment. China Management Science, 11, 1–13.

19.

Johnstone

Managi

Rodríguez

M. C.

Hašcic

Fujii

Souchier

(2017). Environmental policy design, innovation and efficiency gains in electricity generation. Energy Economics, 63, 106–115.

20.

Kang

(2020). Bilateral effect decomposition of green technology innovation under environmental regulation. China Population Resources and Environment, 2020(10), 93–104.

21.

Liu

Sun

Chen

Wang

(2023). Optimization of carbon performance evaluation and its application to strategy decision for investment of green technology innovation. Journal of Environmental Management, 325, 116593.

22.

Marino

Parrotta

Valletta

(2019). Electricity regulation and innovation. Research Policy, 48(3), 748–758.

23.

Meng

F. S.

Han

(2017). Research on the mechanism of government environmental regulations on low-carbon technological innovation behavior of enterprises. Forecasting, 36(01), 74–80.

24.

Pei

Jiang

A. X.

(2019). Private investment, environmental regulations, and green technological innovation: Spatial Dubin model analysis of 11 provinces and cities in the Yangtze River economic belt. Science and Technology Progress and Policy, 36(8), 44–51.

25.

Porter

M. E.

Linde

C. V. D.

(1995). Toward a new conception of the environment-competitiveness relationship. Journal of Economic Perspectives, 9(4), 97-118.

26.

Sempertegui

(2022). The blind spot of the policies of the energy transition: The development imperative as the path to solve climate change. Journal of World Energy Law & Business, 15(2), 88–96.

27.

Sun

Wei

Wang

(2017). Analysis of the impact of environmental regulationson the green technology innovation of China’s OFDI: Based on the heterogeneous host country perspective. Research and Development Management, 29(06), 81–90.

28.

Tang

Qiu

Zhou

(2020). Does command-and-control regulation promote green innovation performance? Evidence from China’s industrial enterprises. Science of the Total Environment, 712, 136362.

29.

Wang

F. Z.

Jiang

Guo

X. C.

(2018). Government quality, environmental regulation, and green technology innovation of enterprises. Science Research Management, 39(1), 26–33.

30.

Wang

S. S.

Zhang

L. H.

Fan

T. J.

(2020). Research on investment balance strategy of competitive supply chain based on carbon emission reduction technology. Chinese Journal of Management Science, 28(6), 73–82.

31.

Wang

Liu

Peng

(2015). Research on the upgrading effect of industrial structure under population aging. China Industrial Economics, 2015(11), 47–61.

32.

Liu

C. H.

Tsai

S. B.

(2020). An empirical study on government direct environmental regulation and heterogeneous innovation investment. Journal of Cleaner Production, 254, 120079.

33.

Xepapadeas

Zeeuw

(1998). Environmental policy and competitiveness: The Porter hypothesis and the composition of capital. Discussion Paper.

34.

Xie

R. H.

Yuan

Y. J.

Huang

J. J.

(2017). Different types of environmental regulations and heterogeneous influence on “green” productivity: Evidence from China. Ecological Economics, 132, 104–112.

35.

Yang

C. H.

Tseng

Y. H.

Chen

C. P.

(2012). Environmental regulations, induced R&D, and productivity: Evidence from Taiwan’s manufacturing industries. Resource and Energy Economics, 34(4), 514–532.

36.

Yang

Jiang

Pan

(2020). Does China’s carbon emission trading policy have an employment double dividend and a Porter effect? Energy Policy, 142, 142.

37.

Hong

Ya-Jing

(2017). Relationship between R&D investment and business performance from the perspective of environmental regulation: Taking the heavy-polluting industries from 2011 to 2016 as an example. Ecological Economics, 14(3), 15–28.

38.

Zhang

Vigne

S. A.

(2021). How does innovation efficiency contribute to green productivity? A financial constraint perspective. Journal of Cleaner Production, 280, 124000.

39.

Zhang

Geng

G. W.

(2019). The impact of environmental regulations on green technology innovation. China Population Resources and Environment, 29(1), 168–176.

40.

Zhang

Y. J.

Peng

Y. L.

C. Q.

Shen

(2017). Can environmental innovation facilitate carbon emissions reduction? Evidence from China. Energy Policy, 100, 18–28.

Research on the Impact of Environmental Regulation on Green Technology Innovation From the Perspective of Innovation Investment

Abstract

Keywords

Introduction

Theoretical Model Construction

Environmental Regulation and Innovation Investment

Innovation Investment and GT

Model Design and Indicator Selection

Model Design

Indicator Selection

Dependent Variable

Core Explanatory Variable

Mediation Variables (Threshold Variables)

Control Variables Cover FDI, ISU, URB, GF

Empirical Research

Direct Effect

Endogeneity Test

Mediation Effect

Threshold Model

Conclusion and Policy Recommendations

Conclusion and Discussion

Research Limitations

Policy Recommendations

Footnotes

ORCID iD

Funding

Declaration of Conflicting Interests

Data Availability Statement

References