Abstract
When the global community is actively combating climate change and advancing the sustainable development process, the crucial role of the science and technology finance ecosystem (STFE) in carbon reduction has not been fully explored. In this paper, an evaluation system for the STFE is constructed from two dimensions, namely the “macro-environment subsystem” and the “micro-community subsystem”. From a new perspective of the STFE, it is introduced into the endogenous economic growth model, and combined with the STIRPAT model, a theoretical model of the carbon emission reduction effect of the STFE is explored. Based on the panel data of 284 prefecture-level cities from 2011 to 2020, the dynamic spatial Durbin model is adopted to examine the spatial spillover effect of the STFE on carbon emission reduction and its influencing mechanism. The results are as follows: (1) STFE in China has a spatial imbalance with a pattern of “high in the east and low in the west”. Total factor carbon productivity (TFCP) shows a spatial distribution pattern where “block distribution prevails in mid and low-level cities, supplemented by point distribution in high-level cities”. (2) Local STFE can facilitate local city carbon emission reduction; however, the development of local STFE is harmful to carbon emission reduction in neighboring cities, resulting in the negative externality of the “beggar-thy-neighbor” spatial spillover effect. (3) Heterogeneity analysis indicates that only the STFE in the eastern region is harmful to the improvement of the carbon emission reduction level in neighboring regions, while the positive spatial spillover effect of the STFE carbon emission reduction effect is more prominent in non-resource cities and high-financial-development areas. (4) The negative spatial spillover effect of the development of local STFE on carbon emission reduction in neighboring areas operates through the mechanisms of green technology innovation, human capital siphoning, the digital divide in informatization development, and the competitive exclusion in financial resource allocation. These conclusions not only provide a new research perspective and complementary empirical evidence but also offer a practical guideline for strengthening STFE and facilitating the growth of a low-carbon economy.
Similar content being viewed by others
Introduction
According to the annual report released by the Global Monitoring Laboratory of the National Oceanic and Atmospheric Administration (NOAA), the global average atmospheric carbon dioxide concentration reached 419.3 parts per million (ppm) in 2023, representing a rise of over 50% from the pre-industrial level (~280 ppm) and setting an all-time high. The escalation of greenhouse gas concentrations has triggered systemic climate risks, prompting the international community to accelerate the formulation of emission reduction policies. As the world’s largest carbon emitter, China has established a systematic emission reduction framework through its “dual-carbon” goals (Dong et al., 2022). However, the deep adjustment of its energy structure remains constrained by the transformation pressures of the traditional industrial system. The global temperature control targets have imposed stringent time-window requirements on emission reduction actions across all sectors (IPCC, 2021). This indicates that China, as a country in the late stage of industrialization, needs to establish a new balance mechanism between economic growth and low-carbon transformation. Currently, while China’s emission reduction practices rely on technological innovation and industrial upgrading, strengthening the construction of a science and technology financial ecosystem (STFE) is more critical to inject new momentum into low-carbon transition and provide replicable Chinese experience for global climate governance.
Carbon productivity, defined as the ratio of gross domestic product to total carbon emissions, has become an important indicator for measuring the development efficiency of low-carbon economy in a country or region (Wu and Yao, 2022). Meanwhile, the rate of its improvement can also reflect the efforts made by countries or regions to address climate change (Zhang et al., 2018). However, since carbon productivity solely emphasizes the direct relationship between carbon emissions and economic growth while ignoring the inherent connections between carbon emissions and factors such as capital, labor, energy, and technology, it is difficult to capture the comprehensive effects of multi-input and multi-output systems. Related research then introduced capital, labor, and energy as inputs, and carbon emissions as unexpected outcomes, employing data envelopment analysis (DEA) and directional distance function (DDF) to evaluate TFCP. This methodology incorporates the full effects of input elements like capital, labor, resources, and technology, thus appraising the efficiency of a low-carbon economy in a comprehensive manner (Li et al., 2018; Mo, 2021). Furthermore, carbon emissions, akin to production elements such as capital and labor, possess strong space flow (Deng et al., 2022). Hence, the examination of TFCP ought to contemplate spatial correlation and spatial spillover effects whilst accounting for the complete effects of multiple input factors.
Science and technology finance are the allocation and services of a series of financial resources to promote science and technology innovation, and it is also an institutional arrangement to realize the close combination of science and technology and finance. Science and technology and finance provide financial support for low-carbon projects and related industries, fostering a mutually beneficial scenario where economic progress and environmental improvement can flourish simultaneously (Gao et al., 2023). In 2016, the “Thirteenth Five-Year Plan” for China’s scientific and technological innovation initiated the conceptualization of a STFE for synergistic integration of various financial instruments. It regards technology and finance policies and related systems as a living ecosystem. The synergistic integration of multiple financial instruments in the STFE provides a crucial funding source for science technology innovation enterprises and projects. Meanwhile, the STFE can direct capital flow to the most promising and valuable science, technology, and innovation areas. Through the financial market’s price mechanism and risk assessment, resources are allocated to technology projects with high returns and great social significance, effectively enhancing overall social innovation efficiency.
In addition to improving the efficiency of resource allocation, the STFE is capable of providing both financial and technical support for the digital and intelligent transformation of traditional industries. It also facilitates the cultivation and development of emerging industries (George and Prabhu, 2003). Through obtaining financial support from venture capital and government-guided funds within the STFE, a large number of talents and resources have been attracted to invest in the industry. Gradually, a new industry has emerged, covering multiple application scenarios such as digital currency and supply-chain finance, creating a new growth point for economic development and promoting the development of the economic structure towards diversification and high-end development. Moreover, the technology industry developed under the STFE usually has strong innovation and adaptability. When faced with economic fluctuations and external shocks, technology enterprises can explore new markets and generate new demand through technological innovation, thereby alleviating the pressure of an economic downturn (Liu et al., 2020). Meanwhile, STFE focuses on supporting sustainable development areas such as environmental protection and new energy, which is beneficial for realizing the coordination between economic development and environmental protection and ensuring the long-term sustainable development of the economy. Consequently, exploring the STFE is of great significance in enhancing the efficiency of social innovation and promoting the high-quality development of the economy.
The deep integration of science, technology, and finance can resolve the structural mismatch problem in the traditional financial model and effectively and healthily guide technological innovation for carbon emission reduction (Li et al., 2023), thereby promoting the carbon emission reduction effect of the STFE. This ecology can enhance the independent innovation ability, correct the resource mismatch between the financial industry and science technology enterprises, better solve the capital chain shortage problem for green technological innovation, and thus improve carbon productivity (Wei and Huang, 2022). Moreover, the STFE breaks the information and talent constraints of enterprises, promotes innovation in energy-saving and emission-reduction technologies, unleashes the double spillover effects of talent and technology (Wang et al., 2024), empowers the digitalization and green transformation of traditional enterprises, and drives the development of a low-carbon economy. However, existing research mainly focuses on the emission reduction impact of science technology finance development or relevant policies, and few studies have used carbon productivity to measure carbon emission reduction to explore the emission reduction effect of the STFE (Zhang and Liu, 2022). Even fewer studies have investigated the spatial spillover effect, intrinsic mechanism, and heterogeneity of the impact of the STFE on carbon emission reduction based on this.
Therefore, this paper innovatively takes the perspective of STFE. It considers science and technology financial development, policies, and related systems as a system with living characteristics. It uses TFCP to measure carbon emission reduction and green economic growth and explores the impact of STFE on TFCP. Additionally, based on this impact, the spatial spillover effect, internal mechanism, and various heterogeneous impacts are further explored. This broadens the multi-dimensional social impacts of the STFE, enriches its internal impact mechanism, effectively supplements the existing literature, and provides valuable reference and experience for China’s future formulation of the synergy between science technology finance policies and carbon emission reduction policies to continuously promote the achievement of the “dual carbon” goal.
The research presents three main contributions. First, the enrichment of research perspectives mainly lies in two aspects. On the one hand, most existing research on science technology finance focuses on science technology financial policy or development. Only a few scholars have conducted research based on the STFE, and even fewer have comprehensively measured it. Therefore, based on the economic environment and financial environment within the “macro-environmental subsystem” and the scientific and technological support system, financial inclusion system, and capital market investment within the “micro-community subsystem”, the index system of the STFE was constructed. On the other hand, TFCP is an important indicator for measuring the development of the economy with low carbon emissions. Only a few studies have explored the carbon emission reduction effect of the STFE from the perspective of TFCP. This paper conducts an in-depth exploration of this issue, which effectively enriches the literature related to the science technology financial development, thereby realizing a win-win situation for both the economy and the environment.
Second, this paper presents certain innovations in both theoretical and empirical approaches. Firstly, very few scholars have explored the theoretical foundation of the impact of STFE on carbon emission reduction through the derivation of mathematical models, and there is a significant scarcity of theoretical research in this regard. This paper incorporates STFE into the endogenous economic growth model and, in combination with the STIRPAT model, constructs a theoretical model for analyzing the relationship between STFE and carbon emissions. Through rigorous mathematical derivations, it demonstrates the carbon-emission-reducing effect of STFE at the theoretical level. This holds certain theoretical value for further expanding research in the related fields of science and technology finance and carbon emissions. Secondly, most existing research on the spatial spillover effect focuses on considering the spatial correlation characteristics of variables while overlooking their time-lagged nature. This paper employs the dynamic spatial Durbin model to explore the carbon-emission-reducing effect of STFE. The DSDM not only takes into account the impact of local STFE on carbon emission reduction but also incorporates the impact of STFE on carbon emission reduction in adjacent regions through the spatial spillover effect. Meanwhile, through the lagged terms of the explained variables, it reflects the cumulative effect and inertia of STFE over time, enabling a more comprehensive and accurate portrayal of the dynamic change process and spatial transmission mechanism of STFE. Therefore, it is reasonable and innovative to use the dynamic spatial Durbin model to study the spatial spillover effect of STFE on TFCP.
Third, the existing analysis of the impact mechanisms of science and technology finance on carbon emission reduction remains insufficiently comprehensive. Centering on the theme of “the spatial spillover effect of STFE on carbon emission reduction”, this paper systematically analyzes the internal mechanisms of the negative spatial spillover effect of STFE on carbon emission reduction from four aspects: the green technology innovation mechanism, the human capital siphon mechanism, the digital divide mechanism in informatization development, and the competitive exclusion mechanism of financial resource allocation. Moreover, this paper further explores the spatially heterogeneous impacts of STFE on carbon emission reduction, covering three aspects: geographical location, resource endowment, and the level of financial development. By deeply investigating the internal impact mechanisms of STFE on carbon emission reduction from a spatial perspective, this research fills the gaps in relevant literature.
The remainder of the study is structured as follows: Section 2 provides the literature review; Section 3 presents the research hypotheses; Section 4 presents the research design, including variables and data; Section 5 reports the empirical results; and Section 6 discusses the limitations of this paper, summarizes the research findings and provides suggestions for further studies.
Literature review
Research on the measurement of STFE
The field of the measurement of STFE is not well explored, with much of the research centering on two science and technology finance policies, and the development of science and technology finance. Primarily, when evaluating science and technology finance policies, scholars typically regard it as a regulated natural experiment, utilizing the difference-in-differences method (DID) to assess its economic and environmental impacts (Liu et al., 2022; Lu et al., 2022). A select few scholars have quantified this policy, constructing a thorough evaluation index system encapsulating policy strength, goals, and measures to assess the effectiveness of science and technology financial policies (Mao et al., 2021; Liu et al., 2020).
Secondly, academicians have yet to achieve consensus on the methodology to measure the development of science and technology finance. The majority of investigations contemplate establishing an index system for science and technology and finance, primarily drawing upon technology enterprises, financial institutions, government spending, intermediaries, and the economic and social milieu. Lu et al. (2023) employed entropy weighting to assess technological and financial development, utilizing variables such as governmental outlay on science and technology, bank loans in science and technology, venture capitalist capital allocation, and stock market funding of science and technology enterprises. Sheng et al. (2021) leveraged a three-stage DEA model to scrutinize the science and technology and financial efficiency of the marine industry, constructing indexes including internal R&D expenditures, scientific and technological projects, authorized patent applications, marine scientific research education value-added management service industry, local economic development level, local government support, and local openness.
However, limited research has delved into the definition and quantification of STFE concepts. Drawing from literature review and exploration, this study classifies STFE as a dynamic equilibrium system where science and technology enterprises, financial institutions, governments, and intermediary organizations interact with the external environment, thriving in harmony. Built upon the two facets of the “macro-environmental subsystem” and the “micro-community subsystem”, this paper constructs an indicator system and utilizes the coefficient of variation methodology to assess the STFE.
Research on carbon productivity
Carbon productivity, an important concept in the low-carbon economy field, can combine the goals of reducing carbon emissions and realizing economic growth (Wu and Yao, 2022). Its use in measuring low-carbon economic development is widely recognized by the academic community (Wang et al., 2023). From an economic perspective, carbon productivity takes carbon as a binding indicator of economic and social development. It measures economic output per unit of carbon dioxide, considering energy, material, and other input factors under a certain technological level (Gao et al., 2021). In this new historical development phase, green low-carbon development has obtained a new meaning. It emphasizes not only carbon emission reduction and continuous economic growth but also rests on China’s resource endowment, the profound progress of the energy revolution, and the contemplation of inputs from numerous factors. Hence, at present, single-factor carbon productivity, which only contemplates the economic benefits per unit of carbon dioxide, has some limitations in gauging the carbon emission reduction level. It is more fitting to measure carbon productivity by scientific means within a comprehensive evaluation system that considers a broad range of relevant factors.
Carbon productivity is a hot topic in current research, with existing studies mainly focusing on its measurement and related influencing factors. First, regarding carbon productivity measurement, there are two main evaluation methods: single-factor and total-factor. The single-factor evaluation method uses the ratio of total carbon emissions to a certain factor as the carbon emission performance measurement index but ignores other input factors (Gao and Zhu, 2016). The total-factor evaluation method incorporates carbon emissions and input factors into the total factor productivity growth measurement framework to obtain total factor carbon productivity (TFCP) (Wang et al., 2022). Total factor productivity measurement methods mainly include the Stochastic Frontier Approach (SFA) and Data Envelopment Analysis (DEA). The DEA method can avoid equation form setting errors, is easier to operate, and is more commonly applied (Sueyoshi and Goto, 2012). To overcome the shortcomings of the traditional DEA method such as inability to distinguish environmental factors, the non-radial and non-angle SBM model was proposed (Tone, 2002). For further dynamic efficiency analysis, the SBM model combined with the ML index is commonly used to measure dynamic changes in total factor productivity (Li and Chen, 2021).
Second, in the study of factors influencing carbon productivity, based on its measurement, scholars have identified that factors such as Internet development (Zhang et al., 2021), technological innovation (Liu and Zhang, 2021), financial development (Murshed et al., 2022), the digital economy (Han et al., 2022), environmental regulation (Zhang et al., 2024), foreign direct investment (Long et al., 2020), and manufacturing agglomeration (Xu et al., 2023) impact carbon productivity and commonly present a spatial spillover effect. Additionally, some scholars explore the impact of policy implementation on carbon productivity from a policy evaluation perspective (Kong et al., 2024).
A review of the literature reveals numerous factors affecting carbon productivity. While scholars usually study the impact of digital finance (Sun et al., 2024), green finance (Li et al., 2024), and other topics on carbon productivity, there is a lack of research focusing on the STFE. Based on the “macro-environmental subsystem” (encompassing the economic environment and financing environment) and the “micro-community subsystem” (including the science and technology support system, the financial inclusion system, and capital market inputs), the impact of STFE on carbon productivity is worthy of further exploration. STFE can directly promote the enhancement of carbon productivity by providing financial support for low-carbon technological innovation. Indirectly, it can facilitate the economic transformation towards low-carbon industries such as high-end manufacturing and services, thereby enhancing carbon productivity. Moreover, by influencing enterprise business decisions and behaviors, guiding them to strengthen environmental management, improve energy efficiency, and reduce carbon emissions, STFE can also improve carbon productivity. Consequently, an in-depth study of this issue has substantial theoretical and practical significance for addressing climate change, promoting sustainable development, and formulating effective science and technology finance policy measures.
Research on the effects of STFE on carbon emission reduction
At present, the results of academic research concerning the relationship between STFE and carbon emission reduction are relatively insufficient. Similar studies mainly focus on the relationship between science technology financial development or related policies and carbon emission reduction, and the relationship between the two can be summarized in the following two aspects.
First, most scholars maintain that the implementation of science technology financial policies can effectively reduce the scale and intensity of carbon emissions in pilot cities (Lu et al., 2022). Some scholars believe that digital green finance facilitates the development of renewable energy and aids in reducing carbon emissions (Yu et al., 2022). Other scholars also consider that low-carbon technology financing helps to enhance the R&D and absorption capacity of carbon emission reduction technologies in underdeveloped countries (Gu et al., 2022). Second, some scholars have found that when technological change only promotes economic development rather than being directed towards cleaner technologies, the expansion of the production scale is likely to cause an energy rebound effect, resulting in more carbon emissions (Liang et al., 2022). Similarly, some scholars have argued that the technological revolution brought about by the gradual spread of digital technologies and infrastructure has triggered an energy rebound effect (Peng and Qin, 2024). When energy use increases, the intensity of carbon emissions rises and adversely affects carbon mitigation. The inconsistent conclusions of the research may be due to differences in research time and samples or individual differences. With the continuous development of technological innovation, there are systematic differences in the impact of technological innovation on carbon emissions in different periods, regions, and even in different types of industries or enterprises.
By comprehensively reviewing the relevant literature, we discover that relatively few studies analyze the dynamic evolution of its spatial distribution and the impact on carbon emission reduction from the perspective of STFE. Even fewer studies use TFCP as an entry point. Therefore, this paper adopts the perspective of STFE and utilizes TFCP to comprehensively measure the carbon emission reduction effect and economic development. Firstly, STFE is incorporated into the endogenous economic growth model and combined with the STIRPAT model to deduce the theoretical relationship between STFE and carbon emissions at the theoretical level. Then, the DSDM is employed to explore the spatial spillover effect and the influence mechanism of carbon emissions, which provides a reference for China to strengthen the development of science technology finance and achieve the “dual-carbon” goal.
Research hypothesis
Emission reduction effects of the STFE
The carbon emission reduction effect of the STFE can be explained from two aspects: the efficiency of resource allocation and the reconstruction of the industrial structure. In terms of resource allocation efficiency, the STFE can reduce the social cost of carbon emissions and improve the efficiency of resource allocation by constructing a coupling mechanism between innovative elements and green development. In the STFE, financial institutions direct funds to low-carbon fields such as the development of new energy sources and the research and development of carbon capture technologies through tools such as green credit, securitization of carbon assets, and climate bonds (Kashif et al., 2024). This addresses the problem in traditional markets where it is difficult to price the “positive externality” of environmental benefits. Such market-oriented capital allocation not only reduces the innovation financing costs of science and technology enterprises, but also accelerates the commercial application of low-carbon technologies through the risk-sharing mechanism, achieving the scale economic effect of emission reduction across the whole society (Xu et al., 2023). In this process, government policies make up for market failures through means such as carbon pricing and research and development subsidies, forming a virtuous cycle from policy guidance to the development of green technology innovation (Xue et al., 2022), and prompting the production function of enterprises to shift towards a low-carbon factor-intensive pattern.
In terms of the reconstruction of the industrial structure, intermediary organizations within the STFE incorporate environmental factors into the enterprise value evaluation system by establishing the infrastructure of the carbon finance market, such as carbon exchanges and green credit rating systems (Eleftheriadis and Anagnostopoulou, 2015). This helps to address the issue of insufficient green investment caused by information asymmetry. Meanwhile, the application of blockchain technology in carbon footprint tracking and carbon account management has significantly reduced the marginal costs of carbon emission monitoring and trading. This has promoted the expansion of carbon trading from a quota-based system to a voluntary emission reduction market, and formed a more flexible price discovery mechanism (Zhang et al., 2023). This collaborative innovation based on the ecosystem not only enhances the efficiency of emission reduction, but also transforms carbon emission reduction from an external constraint into an internal driving force for enterprises’ cost control and value creation by establishing a closed loop from technological innovation to financial services and then to institutional guarantees. Ultimately, it achieves a Pareto improvement in both economic growth and environmental sustainability.
In addition, this paper incorporates the STFE into Romer’s (1990) endogenous growth model and combines it with the STIRPAT model. Subsequently, through mathematical derivation, it analyzes the impact of the STFE on carbon emission intensity. It is hypothesized that the national economy consists of a product production sector and an R&D sector. Moreover, it is assumed that there are only two economic factors, capital and labor, which are distributed between these two sectors. Based on this, the production function of the R&D sector is set as:
All variables are assumed to be functions of time \(t\). Here, \(K\left(t\right)\) represents total capital, \(L\left(t\right)\) represents labor force, and \(A\left(t\right)\) represents the technology level. \(\widetilde{A\left(t\right)}\) represents the amount of change in the technology level. It is given that \(\beta > 0\), \(\gamma > 0\), and \(\beta +\theta < 1\). \(\mu\) is a transfer parameter greater than 0, which represents the factors influencing R&D other than capital and labor force. \({\alpha }_{K}\) represents the ratio of capital to total capital in the R&D sector, and \({\alpha }_{L}\) represents the ratio of labor to total labor in the R&D sector. Equation (1) indicates that the amount of change in the technology level depends on the amount of capital and labor invested in the R&D sector and the technology level itself. Dividing both sides of the equal sign of Eq. (1) by \(A\left(t\right)\) gives:
where \({g\left(t\right)}_{A}=\frac{\widetilde{A(t)}}{A(t)}\) represents the rate of change in the level of technology. The average capital of effective labor can be expressed as follows:
where \(A\left(t\right)L\left(t\right)\) represents the effective labor force and \(k\left(t\right)\) represents the average capital of effective labor. By taking the derivative with respect to \(t\) on both sides of the equal sign of Eq. (3), we obtain:
where \(\widetilde{k\left(t\right)}\) represents the amount of change in the average capital of effective labor, \(\widetilde{K\left(t\right)}\) represents the amount of change in total capital, \(n\left(t\right)=\frac{\widetilde{L(t)}}{L(t)}\) represents the rate of change in labor force, and \(\widetilde{L(t)}\) represents the amount of change in labor force.
The traditional endogenous growth model assumes that the economic system includes a product production sector and a research and development sector (Shao and Yang, 2014). The capital accumulation equation mainly focuses on the change of the total capital. It is expressed by the formula \(\widetilde{K\left(t\right)}=I\left(t\right)-\delta K\left(t\right)={sY}\left(t\right)-\delta K\left(t\right)\), where \(I\left(t\right)\) represents investment, \(Y\left(t\right)\) represents total output, \(s\) represents the savings rate, and \(\delta\) represents the depreciation rate. The traditional endogenous growth model has been expanded by introducing the STFE variable. The expanded capital accumulation equation is:
The development of STFE spawns new financial services and forms, offering enterprises more financing options. \(\varphi \left(t\right)\) gauges the share of extra investment from STFE in total investment (Yuan et al., 2022). STFE channels funds into tech firms via venture capital and tech credit, with this new investment’s proportion in total investment being part of \(\varphi \left(t\right)\)’s meaning. A well-developed STFE in a region draws in capital, leading to a larger \(\varphi \left(t\right)\), signifying its high investment-growth contribution. A sound STFE system, with rich financial products, an active market, and good policies, can enhance tech-finance integration, raise \(\varphi \left(t\right)\), drive tech firms’ growth, and positively affect the regional economy and carbon emissions (Han et al., 2023).
Assuming constant returns to scale, there exists \(y\left(t\right)=f(k\left(t\right))\) at this point. \(y\left(t\right)\) represents the average output of effective labor and is a function of the average capital of effective labor. Substituting \(y\left(t\right)=f(k\left(t\right))\) and Eq. (5) into Eq. (4), we obtain:
When the economy is in long-run equilibrium, the average capital of effective labor remains stable for an extended period. The new investment equals capital depreciation, and \(k\left(t\right)\) reaches the equilibrium level \(\widetilde{k\left(t\right)}\). At this time, \(\widetilde{k\left(t\right)}=0\) in Eq. (6), which can be simplified to obtain:
where Eq. (7) represents the equilibrium average capital of effective labor. To introduce the STFE into the endogenous growth model, substituting Eq. (7) into Eq. (2) yields.
Taking the natural logarithm of both sides of Eq. (8) yields:
Differentiating Eq. (9) with respect to \(t\), we obtain:
At this point, to avoid the complex fluctuations in the labor force change rate from interfering with the relationship between the STFE and carbon emissions, this paper regards \(n\left(t\right)\) as a relatively stable variable, whose derivative is zero, that is, \(\acute{n\left(t\right)}=0\). By further simplifying Eq. (10), we can obtain:
After further simplification of Eq. (11), we obtain:
Among them, \(\widetilde{{g\left(t\right)}_{A}}\) represents the change in the growth rate of the technological level; \(\frac{\widetilde{\varphi \left(t\right)}}{\varphi \left(t\right)}\) is the rate of change of the development of the STFE, and assume that \(\varnothing \left(t\right)=\frac{\widetilde{\varphi \left(t\right)}}{\varphi \left(t\right)}\). When the growth rate of the technological level reaches equilibrium, the technological level no longer changes. At this time, \(\widetilde{{g\left(t\right)}_{A}}=0\) in Eq. (12). Since, in the general economic sense, the technological level is changing, \({g\left(t\right)}_{A}\,\ne\, 0\). Therefore, there exists:
Under the normal state of economic growth, it is highly unlikely that the sum of the labor force growth rate, the depreciation rate, and the rate of change of the technological level is zero. Therefore, \(n\left(t\right)+{g\left(t\right)}_{A}+\delta \,\ne\, 0\).
Equation (16) reveals that, after incorporating the STFE into the endogenous growth model, the rate of change of the development level of STFE in equilibrium significantly influences the rate of change of the technology level. Specifically, given that \(\beta > 0\) and \(\beta +\theta < 1\), it follows that \(\frac{\beta }{1-\beta -\theta } > 0\). This indicates that when the development level of STFE increases, the technology level also rises.
The STIRPAT model is developed based on the IPAT model, which aims to illustrate the impact of population, affluence, and technology on environmental pressures (Kilbourne and Thyroff, 2020). By referring to relevant literature (Dong and Li, 2022), this paper sets the carbon emission intensity function as follows:
where \({CEI}\left(t\right)\) represents the carbon emission intensity, \(A\left(t\right)\) represents the technology level, and \(\rho\) represents the coefficient of the technology level with respect to the carbon emission intensity, that is, the technology effect. Technological innovation can inhibit carbon dioxide emissions (Cai et al., 2023), then \(\rho < 0\). \(S\left(t\right)\) represents the optimization level of the industrial structure, and \(\omega\) represents the coefficient of the industrial structure optimization level with respect to the carbon emission intensity, that is, the structural effect. Since the optimization of the industrial structure has a negative impact on carbon emissions (Zhao et al., 2022), then \(\omega < 0\). \(\eta\) represents the influence of other factors, apart from the technology level and industrial structure optimization level, on the carbon emission intensity. Taking the natural logarithm of both sides of Eq. (17) and then differentiating, we can obtain:
where \({g\left(t\right)}_{{CEI}}=\frac{\widetilde{{CEI}(t)}}{{CEI}(t)}\) represents the rate of change of carbon emission intensity, \({g\left(t\right)}_{A}=\frac{\widetilde{A(t)}}{A(t)}\) represents the rate of change of technology level, and \({g\left(t\right)}_{s}=\frac{\widetilde{S(t)}}{S(t)}\) represents the rate of change of the optimization level of industrial structure. Since \(\rho < 0\), the rate of change of the technology level negatively affects the rate of change of carbon emission intensity, indicating that when the technology level rises, the carbon emission intensity will decrease. Because \(\omega < 0\), the rate of change of the industrial structure optimization level negatively affects the rate of change of carbon emission intensity, meaning that the rise in the industrial structure optimization level will lead to a decrease in carbon emission intensity.
When the economy converges to the balanced growth path, the growth rate of the technology level and that of the industrial structure optimization level are the same (Lu et al., 2017). Thus, \({g\left(t\right)}_{s}={g\left(t\right)}_{A}={{g\left(t\right)}_{A}}^{* }\). By substituting Eq. (16) into Eq. (19), we obtain:
The derivation of \(\varnothing \left(t\right)\) yields:
where \(\frac{\beta }{1-\beta -\theta } > 0\). Then, whether the STFE can exert a negative effect on carbon emission intensity depends on the sign of \((\rho +\omega )\). According to the model settings in the previous section, \(\rho < 0\) representing the effect of technological innovation on carbon dioxide, and \(\omega < 0\) representing the effect of industrial structure optimization on carbon dioxide. Based on this, it can be concluded that \(\left(\rho +\omega \right) < 0\) and \(\frac{\partial {g\left(t\right)}_{{CEI}}}{\partial \left[\varnothing \left(t\right)\right]} < 0\). Therefore, this paper proposes that:
Hypothesis 1: The development of the STFE promotes urban carbon emission reduction.
Spatial spillover effects and mechanism analysis of the impact of the STFE on TFCP
In the spatial dimension, the development of the STFE in a region not only affects the carbon emissions within this region but also influences the carbon emissions of neighboring regions through multiple channels, giving rise to a spatial spillover effect (Liu et al., 2022; Sheng et al., 2021). Science and technology finance resources tend to concentrate in regions with favorable development conditions, thus forming the “Matthew effect” (Shi, 2023). When a region actively develops its STFE, its high-quality infrastructure, preferential policies, and broad development prospects will attract the inflow of factors such as capital and talent. On one hand, sufficient capital supports local enterprises in conducting research and development of energy conservation and emission reduction technologies and optimizing production processes, enabling a reduction in local carbon emissions. On the other hand, due to the loss of factors, neighboring regions lack adequate resources for energy conservation and emission reduction transformation. Moreover, in order to maintain economic growth, they have to rely on high-carbon industries, which leads to an increase in carbon emissions in neighboring regions. Therefore, during the development process of the STFE, it may remarkably exhibit a negative spatial externality of “local emission reduction-neighboring region emission increase”. This negative spatial spillover effect may stem from four moderating effects: green technology innovation, human capital, informatization development, and financial resource allocation efficiency.
-
(1)
Green technology innovation
During the development process of the STFE, the local area will, by virtue of abundant science and technology finance resources, make substantial investments in the research, development, and innovation of green technologies. This encourages local enterprises to adopt more advanced green technologies in the production process (Lu et al., 2023; Gao et al., 2022). This not only reduces local carbon emissions but also alters the trade pattern among regions (Liu et al., 2023; Lin et al., 2022). As green technologies enhance the competitiveness of local products, the demand for local green products from neighboring regions increases, prompting local enterprises to expand their production scale. At the same time, due to the lack of sufficient science and technology finance support, some production activities in neighboring regions find it difficult to achieve green technology upgrades synchronously (Yang, 2022). Moreover, in order to meet local consumption demands, some high-carbon production links have increased, leading to an increase in carbon emissions in neighboring regions and generating negative spatial externalities.
-
(2)
Human capital
In regions where the STFE is well-developed, with abundant financial resources and numerous development opportunities, a large number of high-quality talents are attracted to flow in. This influx of talents promotes the deep integration of local science and technology finance and green industries, thereby achieving a reduction in local carbon emissions. However, due to the brain drain (Hu et al., 2023), neighboring regions suffer from a lack of vitality in scientific and technological research, development, and innovation, and the development of their green industries is hindered (Yuan et al., 2024). In order to sustain economic development, enterprises in neighboring regions, in the absence of guidance from professional talents, still rely on traditional high-carbon production models. Consequently, this leads to an increase in carbon emissions in neighboring regions. Eventually, the carbon emission reduction effect of the development of the STFE exhibits a negative spatial spillover effect.
-
(3)
Informatization development
The development of the local STFE significantly boosts the level of informatization. This enables local enterprises to promptly obtain and process various types of information, optimize production processes, and reduce energy consumption and carbon emissions (Zhang et al., 2024). In contrast, due to the lagging informatization construction in neighboring regions, enterprises there struggle to quickly access cutting-edge green technologies and management experience, placing them at a disadvantage in market competition. To maintain their market share, enterprises in neighboring regions may maintain or expand high-carbon production scales to cut costs. As a result, carbon emissions in neighboring regions increase, triggering a negative spatial spillover effect.
-
(4)
Financial resource allocation efficiency
In the local area with a sound STFE, financial institutions can allocate funds to low-carbon and environmentally friendly high-quality projects more precisely and efficiently. This facilitates the low-carbon transformation of local enterprises and leads to a reduction in carbon emissions (Arian and Sands, 2024). Conversely, in neighboring regions, due to the low efficiency of financial resource allocation, it is difficult for funds to flow effectively into the green industry. As a result, the development of the green industry in these regions is sluggish. Meanwhile, to sustain economic growth, traditional high-carbon industries in neighboring regions continue the high-carbon production model without sufficient financial support for transformation and upgrading (Wang and Gao, 2024). This causes an increase in carbon emissions in neighboring regions, generating a negative spatial spillover effect. Therefore, this paper proposes:
Hypothesis 2: There exists a negative spatial spillover effect in the impact of STFE on carbon emission reduction.
Hypothesis 2a: Green technology innovation, human capital, informatization development, and financial resource allocation efficiency exert moderating influences during the spatial spillover process through which the STFE impacts carbon emission reduction.
Research design
Data source
This paper covers 284 prefecture-level cities in China from 2011 to 2020. To ensure data validity and reliability, the main sources include the China Urban Statistical Yearbook, China Science and Technology Statistical Yearbook, China Financial Statistical Yearbook, China Technology and Finance Development Report, China Venture Capital Development Report, and the Wind database. Additionally, it should be noted that since the start of Beijing University’s Digital Pratt and Whitney Index in 2011, the study period of this paper is from 2011 to 2020. Figure 1 shows the mathematical analysis of this paper.
Flow chart of research methodology.
Variable selection
Explained variable
This study employs the super-efficiency SBM-DDF model and Luenberger analysis to measure the TFCP of prefecture-level cities (Long et al., 2024). Specifically, input indicators include capital, labor, and energy. Capital input is represented by annual fixed capital stock, calculated using the perpetual inventory method (Nehru et al., 1995); labor input is represented by the total number of employees in each year; and energy input is represented by annual total energy consumption. Output indicators include desired and undesired outputs, where the desired output indicator is GDP of each region, and the undesired output refers to carbon dioxide emissions. Detailed descriptions of the indicators are shown in Table A1 of Appendix A.
Explanatory variable
STFE is a systemic organization with ecological characteristics, incorporating the development level of science and technology in finance, related policies, and the surrounding external environment. It is complex and highly interrelated. In the context of modern economic and innovation development, science and technology enterprises, financial institutions, governments, and intermediary organizations play crucial roles.
First, science and technology enterprises are the principal innovation entities and the core driving force within the STFE (Chen et al., 2024). Their R&D investment and technological innovation outcomes directly reflect the development level of science and technology finance. These enterprises continuously increase R&D investment, drive technological innovation, and develop competitive new products and services in the market, thereby enhancing the innovation capacity and development level of the entire STFE. Simultaneously, the development and expansion of science and technology enterprises can also promote the development of related industries, further contributing to the STFE’s prosperity. Science and technology enterprises are the main beneficiaries and implementers of science and technology finance policies. The government has introduced support policies such as tax incentives, financial subsidies, and additional deductions for research and development expenses. These policies create a favorable environment, reduce innovation costs, and stimulate innovation vitality. Furthermore, the development of science and technology enterprises is deeply affected by the external environment (Song et al., 2018). A favorable external environment with perfect infrastructure, high-quality talent resources, and open market competition provides advantageous conditions for their development.
Second, financial institutions are capital suppliers and risk managers within the STFE. The financial products and services they provide offer financial support for the innovative development of science and technology enterprises. The professional competence and risk management level of financial institutions directly influence the STFE’s development level. Financial institutions are significant implementers and promoters of science and technology finance policies. The government-released science and technology finance policies have a crucial impact on financial institutions’ operating behavior and business development. Financial institutions need to respond actively to the policy, enhance support for science and technology enterprises, and drive the STFE. Moreover, external elements like the macroeconomic situation, market competition environment, and financial regulatory policies affect financial institutions’ operating performance and risk profile (Ofoeda et al., 2024). Financial institutions must closely monitor changes in the external environment and adjust business strategies in a timely manner to prevent financial risks.
Third, the government is the policy creator and guide in the STFE. It promotes the STFE’s development by creating science and technology finance policies and guiding financial resources to support science and technology enterprises. The government’s policy support and investment are crucial in elevating the STFE’s development level (Chen et al., 2024). External factors such as the international economic situation, domestic macroeconomic environment, and the trend of science and technology innovation all impact the government’s policy formulation and implementation. The government needs to adjust policies in a timely manner based on changes in the external environment to meet the STFE’s development requirements.
Fourth, intermediary organizations are service providers and connection bridges in the STFE. The professional services they provide offer important support for the cooperation between science and technology enterprises and financial institutions. The professional level and service quality of intermediary organizations directly influence the STFE’s development level (Clayton et al., 2018). Intermediary organizations can enhance science and technology enterprises’ and financial institutions’ understanding and implementation of policies through activities like policy publicity, training, and consultation.
Therefore, this study posits that STFE is a dynamic equilibrium system manifested through the constructive interaction and balanced growth of science and technology enterprises, financial institutions, government agencies, and intermediate bodies, along with the external environment. Referring to Liu (2022), the assessment of STFE considers five dimensions: the science and technology support system, the inclusive financial system, capital market investments, the economic environment, and the financial environment.
First, the technological support dimension focuses on the R&D capacity and achievement transformation efficiency in science and technology. It encompasses the number and quality of researchers and the expenditure on science and technology funds. Second, the financial inclusion degree reflects the coverage scope and depth of financial services among science and technology enterprises. It relates to financial institutions’ ability to serve science and technology enterprises of different sizes and at different stages. Third, the capital market is crucial in the STFE. The capital market input dimension considers banks’ new deposits to loan ratios and insurance depth. Fourth, the macroeconomic environment significantly impacts the STFE (Chen et al., 2021). A favorable economic growth trend provides more opportunities and resources for science and technology financial activities. Thus, the economic environment mainly includes economic development and the degree of openness. Fifth, the financing environment covers the ease of financing policies and the strength of market support (Zhou et al., 2024). The former is reflected in the government’s support for science and technology enterprises’ financing, while the latter is determined by the technology market turnover percentage.
By measuring the STFE through these five dimensions, its development status and problems can be more comprehensively and accurately assessed, providing a solid basis for further research and policy formulation. Table A2 in Appendix A presents the index composition of STFE and its specific meanings. In this paper, the coefficient of variation method is used to measure STFE, and the detailed introduction of the coefficient of variation method is also provided in Appendix A.
Control variables
By reviewing relevant literature, this study controls a series of variables potentially influencing urban TFCP in the empirical analysis, primarily including education level (Edu), industrial structure (Industry), informatization level (Infor), financial development level (Fin), environmental regulation (ENV), and government intervention (GOV). The specific explanations of control variables and referenced citations are presented in Table A3 of Appendix A.
Moderating variables
-
(1)
Green technology innovation (GI): This paper selects the number of green patent grants to represent the level of green technological innovation (Guo et al., 2022). Compared with innovation input, the number of patent grants can more effectively reflect innovation capabilities.
-
(2)
Informatization level (Infor): Referencing the study by Noh and Lee (2022), the informatization level of cities is measured by the ratio of total urban post and telecommunications business (including internet, telephone, mail, and newspaper distribution services) to regional gross domestic product (GDP).
-
(3)
Human capital level (HC): It is measured by the number of college students per 10000 people in each prefecture-level city.
-
(4)
Efficiency of Financial Resource Allocation (FIN): In this paper, the Data Envelopment Analysis (DEA) method based on input-output is used to measure the efficiency of financial resource allocation. Drawing on relevant literature (Liao et al., 2020; Qiao and Guo, 2020), the number of employees in the financial industry, the amount of government expenditure, and the balance of loans of financial institutions are selected as input indicators, and the added value of the financial industry is taken as the output indicator. Descriptive statistics of the variables are shown in Table A4 in the Appendix A.
Empirical analysis
Exploratory spatial data analysis (ESDA)
Exploratory spatial data analysis is a crucial research area in spatial econometrics, exploring spatial dependence and correlation associated with spatial location. This paper aims to evaluate the spatial distribution patterns of STFE and urban TFCP by using exploratory spatial data analysis, and we employed the global Moran’s I index for spatial correlation tests.
Temporal variation characteristics of TFCP and STFE
A line chart is used to demonstrate the changing trend in the average value of TFCP across 284 prefecture-level municipalities from 2011 to 2020. Figure 2 illustrates the trend in total factor carbon productivity (TFCHCH) and its decomposition items, technological progress (TECHCH) and technological efficiency (EFFCH). From 2011 to 2020, the TFCP demonstrated a steady fluctuation around the value of 1, exhibiting a “W” shaped trend with minimum and maximum points in 2013 and 2016, respectively. Moreover, the trend in technological progress has always been consistent with TFCP, and it is more volatile than TFCP. Conversely, technological efficiency and TFCP follow almost contrasting trends, suggesting that technological progress is the primary driver for promoting TFCP.
Temporal variation characteristics of TFCP and its decomposition terms from 2011 to 2020.
As illustrated in Fig. 3, the average level of STFE in both the country as a whole and its subregions showed a consistent upward trend between 2011 and 2020. In terms of regional variation, the trend is “east > central > west”. Notably, the eastern region’s STFE level surpasses the national average, due to its status as a significant hub for China’s economic participation in the global economy. The region boasts fully functional financial institutions, ample financial resources, unimpeded financing channels, and extensive data sharing, and plays an increasingly prominent role in driving scientific and technological innovation and development.
Temporal variation characteristics of STFE in China and different regions 2011 to 2020.
Spatial distribution patterns of TFCP and STFE
Figure 4 illustrates the spatial evolution patterns of TFCP across prefecture-level cities in 2011 and 2020. It can be observed that during 2011–2020, cities with low and very low TFCP accounted for more than 50% of all cities, those with medium TFCP constituted 40% of the sample, and the remaining were cities with relatively high and high TFCP levels. Generally, there is a spatial distribution pattern of TFCP characterized by the predominance of blocky distribution in middle and low-level cities, complemented by point distribution in high-level cities. This spatial distribution pattern persisted substantially unchanged for a span of 10 years, exemplifying the characteristics of a robust spatial structure.
Spatial evolution trend of TFCP in China.
This study also drawn a spatial visualization map of the STFE (Fig. 5). Comparing the maps of 2011 and 2020, it’s evident that the dark blue and red zones have increased significantly, indicating an overall increasing trend in the level of China’s STFE, and there is significant spatial heterogeneity.
Spatial evolution trend of STFE in China.
Regionally, the STFE in the eastern region is significantly higher than that in the central and western regions, presenting a characteristic of “high in the east and low in the west”, which is consistent with the trend of China’s economic development (Zhao et al., 2022a). Among them, the southeast coastal cities and urban agglomeration of the Yangtze River Delta have higher STFE levels than other regions. With the increasing richness of technology finance entities, the continuous optimization of the STFE environment, and the gathering of high-level scientific and financial talents, the STFE in these areas is at a higher level. Hunan, Hubei, and Henan provinces in the central region have developed STFE to a certain extent with the backing of the strategy of the rise of central China. Apart from Sichuan province and Chongqing city, STFE levels in other regions are low. This may be due to the fact that the western region has not yet developed a sound technology financial service system, which is consistent with Shen et al.’ (2023) perspective.
Spatial correlation analysis
As illustrated in Table 1, the global Moran’sIindex distinctly registered as positively significant at the 1% level under the three weight matrices of geographic distance (\({W}_{1}\)), economic distance (\({W}_{2}\)), and geographic and economic nested distance (\({W}_{3}\)). This suggested that China’s TFCP exhibits a positive spatial correlation from 2011 to 2020, revealing a tendency towards spatial aggregation. This implies regions surrounded by regions with high TFCP typically have relatively elevated TFCP.
Regression results analysis of DSDM
To determine whether to employ a fixed-effects model or a random-effects model, the Hausman test yielded a statistic of 60.31 with a p value of 0.0000, indicating that the fixed-effects model should be used. Therefore, the DSDM model with bidirectional fixed effect was identified as optimal.
In contrast to the spatial lag and error models, the SDM comprehensively considers the spatial correlation of both the explanatory and explained variables. To take into account the dynamics and lags of TFCP and avoid potential endogeneity issues that might bias estimation results, a DSDM was constructed to explore the impact of STFE on TFCP based on relevant literature (Zhao et al., 2022b).
where \({{TFCP}}_{{it}}\) represents total factor carbon productivity; \({{TFCP}}_{i,{t}-1}\) represents the lagged period of total factor carbon productivity; \({{STFE}}_{{it}}\) stands for science and technology financial ecology; \({X}_{{it}}\) denotes the control variable; \(W\) represents the spatial weight matrix; \(\rho\) represents the spatial effect coefficient; \({\mu }_{i}\) denotes city fixed effect; \({v}_{t}\) denotes the time fixed effect; \({\varepsilon }_{{it}}\) denotes a random error term.
Table 2 presents the influences of STFE on TFCP under different weight matrices. The findings of the static SDM are presented in columns 1, 3, and 5, while those of the DSDM are in columns 2, 4, and 6. It is observable that the coefficients of the lagged TFCP are significantly positive in the economic distance matrix and the economic-geographical nested matrix, denoting the persistence of TFCP within these two matrices. The coefficients of the spatially lagged TFCP are significantly positive at the 1% level only in the economic matrix, indicating that the TFCP exhibits significant spatial spillover effects and spatial-dimension synergistic correlation in the economic distance matrix. Compared with the geographical channel, the improvement of TFCP in this region has a certain “demonstration effect” on neighboring regions through the economic channel, facilitating the improvement of TFCP in those regions.
For the core explanatory variables, the coefficients of STFE display notable positivity under both static and dynamic SDM, suggesting that STFE’s influence on TFCP possesses a substantial spatial enhancement effect. By comparing the coefficient values for columns 2, 4, and 6, it is found that this impact primarily diffuses through economic channels. Furthermore, the coefficient of spatial lag term in STFE exhibits considerable negativity under the three weight matrices, implying that the enhancement of STFE may procure the resources of adjacent regions and potentially generate adverse externalities of “beggar-thy-neighbor”. The rapid development of local STFE can attract high-end labor and resources, such as financial and scientific personnel, from neighboring areas. This can lead to the loss of talent and resources from these areas and hinder the development of neighboring STFE.
The results of the direct effect show that, for both the static SDM model and the dynamic SDM model, the coefficient values of the STFE are significantly positive. This indicates that the local STFE can promote local urban carbon emission reduction, thus verifying Hypothesis 1. For both the static and dynamic SDM models, the coefficients of the indirect effects of the STFE are significantly negative at the 1% level. This indicates that local development of an STFE has a negative impact on carbon emission reduction in neighboring cities, namely producing a negative spatial spillover effect, which validates Hypothesis 2.
The possible reasons for this situation are as follows. Firstly, there is resource competition and siphoning effect. In terms of financial resources, local areas with well-developed science and technology finance tend to attract more financial institution investments and capital inflows, and local STFE development may lead to intense resource competition. Regarding talent resources, a high-quality local STFE offers more development opportunities and high salaries, attracting talent from neighboring regions. Neighboring cities, due to talent loss, have reduced scientific and technological R&D and innovation capabilities and are underdeveloped in new technologies and modes for promoting carbon emission reduction. This resource siphoning effect undermines the resource base of neighboring cities’ carbon emission reduction efforts and is unfavorable to their carbon emission reduction process. Secondly, it may be caused by regional barriers to technological innovation. There may be regional diffusion barriers to technological innovations driven by the local STFE. Low-carbon technologies developed by local enterprises and research institutions with the support of science and technology finance may be hard to be promoted and applied in neighboring cities because of excessive intellectual property rights protection and corporate competitive strategies.
Robustness and endogenous tests
Replaced the explained variable
This study recalculated the TFCP under the economic distance weight matrix, using the formula below that mainly expresses the ratio of each city’s GDP to its carbon dioxide emissions. As revealed within the column 1 of Table 3, the coefficients and significance were in alignment with the benchmark regression outcomes, signifying that the conclusions were reliable.
Replaced spatial weight matrix
The inverse geographic weight matrix was applied to replace the original spatial matrices. The distance between any two cities is calculated based on their longitude and latitude, and the reciprocal is taken. The farther the distance between the two places, the less the ability to influence each other. The results in column 2 of Table 3 show basically consistent coefficients and significance with the benchmark regression, signifying robust conclusions from this paper.
Replacement of the indicator weighting methodology
Calculating the weights of the indicators through the coefficient of variation method may rely too much on the objective data performance, thus resulting in dynamic changes in the weighting results as the sample varies. Therefore, this paper further adopts the equal weight method and principal component analysis method to calculate the STFE development level and conducts the robustness test. Among them, the equal weight method indicates that each first-level indicator has the same weight, and its subordinate second-level indicators are also assigned values with the same weight. As can be seen from the results in columns 3 and 4 of Table 3, changing the weight calculation method does not affect the empirical results, which remain consistent with the benchmark regression.
Endogenous test
Although the theoretical analysis indicates that the development of the STFE is one of the factors influencing carbon emission reduction, there still exists an endogeneity problem caused by reverse causality. To address the endogeneity issue resulting from reverse causality between the explanatory variables and the explained variable, this paper, drawing on the research of Jin (2019), re-estimates the parameters of the benchmark regression equation using the Generalized Spatial Two-Stage Least Squares (GS2SLS) method. This approach does not require the independent selection of external instrumental variables. Instead, it only needs to generate instrumental variables by utilizing the spatial lag terms and various explanatory variables for a robust estimation, which can effectively reduce the bias in the estimation results caused by the endogeneity problem. Table 4 reports the parameter estimation results of GS2SLS. Firstly, the F-statistics in the first stage under the settings of the three weight matrices are all >10. Therefore, there is no need to worry about the problem of weak instrumental variables. Secondly, as can be seen from the parameter regression results, after taking the endogeneity problem into account, the STFE still exhibits a significant carbon emission reduction effect, which verifies the robustness of the conclusions of this paper.
Heterogeneity analysis
Regional heterogeneity
Due to variations in resource endowments, innovation bases, and levels of financial development across different regions in China, there may be disparities in the effectiveness of promoting regional STFE to enhance TFCP. Therefore, this section aims to further analyze regional heterogeneity impacts under the economic weight matrix.
Columns 1–3 of Table 5 present the results of regional heterogeneity in the impact of STFE on the spatial spillover effects of TFCP. For the direct effect results, only the STFE coefficient in the eastern region is significantly positive at the 1% level, while the STFE coefficients in the central and western regions are not significant. This indicates that among the three major regions, only the local STFE development in the eastern region has a significant carbon emission reduction effect. For the indirect effect results, among the three major regions, only the STFE coefficient in the eastern region is significantly negative at the 1% level, while those in the central and western regions are negative but not significant. This suggests that the development of STFE in the eastern region has a negative impact on carbon emission reduction in surrounding areas, that is, it generates a negative spatial spillover effect.
The possible reason is that the eastern region undergoes faster industrial upgrading during the development of STFE (Chen et al., 2024). Due to the industrial gradient difference between the east and the central and western regions, some high-carbon industries in the east will be transferred to the central and western regions during the industrial structure optimization process. Such industrial transfer often lacks effective low-carbonization constraints and guidance mechanisms. When the central and western regions receive these industries, they may face difficulties in greening them because of their own limited technical level and capital, thereby increasing local carbon emissions and impeding the improvement of carbon emission reduction. Moreover, this industrial transfer may disrupt the original industrial ecological balance in the central and western regions and hinder the development of low-carbon industries that originally had development potential.
Resource endowment heterogeneity
Compared with non-resource-based cities that rely on emerging industries, resource-based cities generally depend more on the development and processing of local natural resources. Therefore, the exploitation and utilization of abundant natural resources in resource-based cities may lead to higher carbon emissions. In this study, sample cities were divided into two groups (resource-based cities and non-resource-based cities) for heterogeneous comparative analysis by referring to the list of resource-based cities specified in the “National Sustainable Development Plan for Resource-Based Cities (2013–2020)”. The specific results are presented in Columns 4–5 of Table 5.
The results of direct effects show that the coefficient of STFE is significantly positive at the 5% level only in non-resource-based cities, indicating that the carbon emission reduction effect of the STFE is more pronounced in non-resource-based cities compared to resource-based cities. The results of indirect effects reveal that the coefficient of the STFE is significantly positive at the 5% level only in non-resource-based cities. This suggests that the promotion effect of local STFE development on carbon emission reduction in neighboring areas is more significant in non-resource-based cities than in resource-based cities, meaning that the spatial spillover effect of the carbon emission reduction effect of the STFE is more evident in non-resource-based cities.
The possible reason is that resource cities usually have a unitary industrial structure with resource extraction and processing as the dominant industry. In contrast, non-resource cities possess a more diversified industrial structure (Li et al., 2024), incorporating manufacturing, service, and high-tech industries. These industries have a higher demand for scientific and technological innovation, and the STFE can offer them more development opportunities. Technical exchanges and cooperation among non-resource cities are also more frequent, which facilitates the rapid dissemination and spillover of low-carbon technologies.
With the adjustment and upgrading of the industrial structure of non-resource cities, some highly polluting and energy-consuming industries may be relocated to other regions. In this process, the STFE can guide the industry transfer in a low-carbon direction, preventing the disorderly transfer of polluting and high-carbon-emission industries among different regions. Such industry transfer can promote the synergistic development of industries between regions, enhance the efficiency of resource utilization, reduce carbon emissions, and increase TFCP.
Heterogeneity of financial development level
Regional differences in the level of financial development may affect the carbon reduction effects of STFE. The ratio of financial institutions’ loan balances to GDP at the end of the year was used to construct the index of financial development level in this study. Based on the mean value of the financial development level, all samples were classified as high financial development level group and low financial development level group, then regressed sequentially. Regression results are shown in columns 6 and 7 of Table 5.
By comparing columns 6 and 7, it can be seen that in areas with a high level of financial development, the coefficient values of STFE are significantly positive at the 5% level for both direct and indirect effects. This means that the development of STFE in these areas significantly contributes to carbon emission reduction both locally and in neighboring areas. However, in areas with a low level of financial development, these two coefficients are not significant, indicating that the development of STFE in these areas makes no significant contribution to carbon emission reduction either locally or in neighboring areas.
The level of financial development is crucial for promoting the carbon emission reduction effect of the STFE. When the financial development level is higher, more funds will be allocated to green enterprises, and the inhibitory effect on the carbon emission behavior of heavily polluting enterprises will be more obvious, which helps to enhance the TFCP. However, when the financial development level is low, the regional financial system is less complete, and the implementation efficiency of financial policies will decrease. The financial system may not be able to execute the provisions in the STFE, failing to reasonably allocate funds from heavily polluting enterprises to green enterprises, and thus being unable to effectively restrain the pollutant emission behaviors of heavily polluting enterprises. As a result, carbon emissions increase, and the enhancement of TFCP is inhibited.
Mechanisms analysis
The test results of the moderating effect are shown in Table 6. The results of columns 1–4 indicate that the development of STFE has significantly promoted the improvement of green technology innovation, informatization level, human capital level, and the efficiency level of financial resource allocation. Columns 5–8 report the test results of the interaction between STFE and the moderating variables. It can be seen that the interaction terms of STFE with green technology innovation, STFE with the informatization level, and STFE with the human capital level are all significantly positive at the 5% level. This indicates that the development of STFE has strengthened the emission reduction effects of green technology innovation, informatization level, and human capital level on local urban carbon emissions. However, the moderating effect of the efficiency of financial resource allocation on the impact of local carbon emission reduction is not significant.
Possible explanations are as follows. For green technology innovation, the STFE offers financial support, cutting innovation costs for enterprises. This allows them to invest more in green R&D, promoting emission-reduction tech breakthroughs and applications. In terms of informatization, a better-developed ecosystem boosts information flow. Enterprises can access low-carbon tech info faster, adjust production strategies, and optimize processes to cut waste and emissions. As for human capital, the ecosystem channels funds into education and training. More innovative talents are cultivated, who can drive green tech progress and implement low-carbon strategies in enterprises, thus reducing local urban emissions.
Columns 5–8 also report the spatial effects of the interaction terms between STFE and the moderating variables. The spatial terms of the interaction between STFE and the four moderating variables are all significantly negative. This indicates that the development of STFE, through green technology innovation, informatization level, human capital level, and the efficiency of financial resource allocation, promotes the increase in carbon emissions in neighboring regions, exacerbating the “beggar-thy-neighbor” negative externality spatial impact of STFE, which verifies Hypothesis 2a. In contrast, the moderating mechanisms of the informatization level and human capital level exacerbate the negative spatial externality of STFE’s impact on carbon emission reduction to a greater extent.
The possible explanations are as follows. In the aspect of green technology innovation, the development of STFE enables local enterprises to secure more funds for research and development. Since the R&D achievements are preferentially applied locally, neighboring regions are put at a competitive disadvantage. This may impel them to resort to high-carbon technologies to sustain production, thereby leading to an increase in carbon emissions. At the level of informatization, the development of STFE empowers local enterprises to access information more efficiently. While optimizing their own production processes, local enterprises leave neighboring regions lagging in information acquisition. As a result, the industrial adjustment in neighboring regions is slow, and the proportion of high-carbon industries relatively increases, causing a rise in carbon emissions. Regarding the human capital level, the development of STFE attracts talent to flow into the local area. This leads to a talent drain in neighboring regions, leaving them short of the intellectual support essential for low-carbon transformation. Consequently, it becomes difficult to reduce the carbon emissions of industries in these regions. In terms of the efficiency of financial resource allocation, the development of STFE enables a more rational allocation of local financial resources, which promotes the development of local industries. However, neighboring regions face difficulties in obtaining funds and may have to rely on traditional high-carbon industries, thus resulting in an increase in carbon emissions.
Limitation
Some limitations should be taken into account for a more comprehensive study. First, concerning the quantification of STFE metrics, additional metrics can be included in the STFE metric system in the future, for example, government institutions and science and technology companies, to make sure that the quantification of STFE metrics is more comprehensive. Second, future research can explore the environmental implications of STFE from various perspectives, such as a study on the impacts of STFE on pollution mitigation and carbon reduction. Third, with the accelerated development of science and technology finance, the sample scope can be expanded to a longer time dimension to further investigate the economic and environmental implications of STFE, ensuring that the study’s findings are more convincing.
Conclusion and policy recommendations
Based on panel data from 284 Chinese prefecture-level cities and above from 2011 to 2020, this study uses the DSDM model to investigate the spatial spillover effect of STFE on TFCP. The main conclusions are as follows:
-
(1)
There is a significant spatial imbalance in the STFE in China. The level of STFE in the eastern region, gathering rich financial resources, is significantly higher than that in the central and western regions, displaying a characteristic of “high in the east and low in the west”. TFCP exhibits a spatial distribution pattern of “block distribution mainly in middle and low-level cities, complemented by point distribution in high-level cities.” Furthermore, technological progress is the primary driver for the improvement of TFCP.
-
(2)
In the regression analysis concerning the spatial spillover effect of STFE’s influence on carbon emission reduction, the direct effect outcomes imply that, in both the static SDM model and the dynamic SDM model, local STFE can facilitate local urban carbon emission reduction. The indirect effect findings, however, reveal that in both these models, local STFE development is prejudicial to carbon emission reduction in neighboring cities, thereby generating the negative externality of the “beggar-thy-neighbor” spatial spillover effect. This conclusion retains its significance following a series of robustness and endogeneity tests, highlighting the complexity of the relationship between STFE and carbon emissions within different spatial contexts.
-
(3)
The regional heterogeneity analysis uncovers that only STFE development in the eastern region is inimical to the enhancement of the carbon emission reduction level in adjacent regions, while the spatial spillover of the STFE carbon emission reduction effect is negligible in the central and western regions. The analysis of resource endowment heterogeneity further indicates that the positive spatial spillover effect of the STFE carbon emission reduction effect is more pronounced in non-resource cities as opposed to resource-rich ones. The examination of financial development level heterogeneity also shows that, in contrast to areas with low financial development, STFE development in high-financial-development areas significantly contributes to carbon emission reduction both locally and in neighboring areas. These regional, resource-based, and financial-level differences demonstrate the diverse nature of how STFE affects carbon emissions across various settings.
-
(4)
The results of the mechanism analysis indicate that the development of STFE promotes the increase in carbon emissions in neighboring regions through the green technology innovation mechanism, the human capital siphon mechanism, the digital divide mechanism in informatization development, and the competitive exclusion mechanism of financial resource allocation, exacerbating the “beggar-thy-neighbor” negative externality spatial impact brought about by STFE. In contrast, the moderating mechanisms of the informatization level and the human capital level can, to a greater extent, intensify the negative external spatial effects of STFE on carbon emission reduction.
According to the above empirical conclusions, the following suggestions are put forward:
-
(1)
Region-specific differential policies for supporting science and technology finance should be formulated. For the central and western regions, means such as financial subsidies and tax incentives should be employed to guide the inflow of financial resources. Encourage financial institutions to establish specialized science and technology finance institutions, so as to enhance the local STFE level and narrow the gap with the eastern region. Meanwhile, establish special funds to support the research and application of low-carbon technologies in the central and western regions. Focus on supporting the block-distributed cities with medium-low total factor carbon productivity, assist them in advancing towards a high level, and optimize the spatial distribution pattern.
-
(2)
It is necessary to establish a collaborative development mechanism for the STFE among regions and break through local protectionism. Through regional cooperation agreements, clarify the responsibilities and obligations of each region to avoid the negative externality spatial spillover effect of “beggar-thy-neighbor”. For example, construct a regional carbon emission reduction compensation mechanism to provide economic compensation to the regions that are damaged by the development of science and technology finance in neighboring areas. In addition, strengthen cross-regional environmental supervision and policy coordination to ensure that all regions strictly implement carbon emission reduction standards during the development of science and technology finance, and prevent the transfer of carbon pollution.
-
(3)
The eastern region should continuously leverage the STFE to promote local carbon emission reduction, increase investment in the green technology industry, explore cooperation models with neighboring areas, and reduce negative spillovers. The central and western regions should introduce targeted policies to guide financial resources to flow towards science and technology and low-carbon industries, and enhance the role of the STFE in local carbon emission reduction. Resource-based cities need to formulate transformation policies, utilize the STFE to develop green transformation industries, and improve the carbon emission reduction effect. Non-resource-based cities should strengthen the positive spatial spillover of the STFE and promote the exchange and cooperation of low-carbon technologies among regions. In areas with low-level financial development, increase investment in financial infrastructure construction, improve the level of financial services, and create favorable conditions for the development of the STFE. In areas with high-level financial development, improve the STFE, strengthen cooperation with neighboring areas, and jointly promote regional carbon emission reduction.
-
(4)
The government should construct a cross-regional collaborative governance framework to break the spatial lock-in of negative externalities. Establish a cross-domain transfer compensation mechanism for green technologies, and weaken technological monopolies through patent pool sharing and innovation revenue sharing. Implement a gradient cultivation plan for human capital, and bridge the gap in environmental governance capabilities with targeted educational investment and flexible talent flow policies. Promote the equalized layout of digital infrastructure, and eliminate the space for regulatory arbitrage relying on the blockchain-based carbon footprint tracking system. Innovate collaborative green finance tools, and reconstruct the logic of resource allocation through cross-regional trading of carbon emission rights and joint credit mechanisms. In view of the strengthening of the moderating effects of informatization and human capital, it is necessary to focus on enhancing the digital regulatory capabilities of neighboring regions and the reserve of human capital adaptable to low-carbon technologies. At the same time, establish a horizontal transfer payment system for ecological compensation, and achieve the sharing of emission reduction costs and governance benefits through the reconstruction of the rights and responsibilities in spatial planning.
Data availability
The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.
References
Arian G, Sands J (2024) Do corporate carbon emissions affect risk and capital costs? Int Rev Econ Financ 93:1363–1377. https://doi.org/10.1016/j.iref.2024.04.018
Cai J, Zheng H, Vardanyan M, Shen Z(2023) Achieving carbon neutrality through green technological progress: evidence from China Energy Policy 173:113397. https://doi.org/10.1016/j.enpol.2022.113397
Chen B, Lin H, Shan B, Xiao Y (2024) Government investment in science and technology, digital transformation, and innovation in manufacturing enterprises. Financ Res Lett 69(B):106299. https://doi.org/10.1016/j.frl.2024.106299
Chen W, He K, Wang L (2021) Blockchain technology, macroeconomic uncertainty and investment efficiency. Int J Emerg Mark 18(7):1493–1514. https://doi.org/10.1108/IJOEM-10-2020-1250
Chen W, Li Q, Zhang S, Zhou X (2024) Can industrial structure upgrading restrain industrial land expansion? Evidence from China. Chin Geogr Sci 34(3):504–518. https://doi.org/10.1007/s11769-024-1431-z
Clayton P, Feldman M, Lowe N (2018) Behind the scenes: intermediary organizations that facilitate science commercialization through entrepreneurship. Acad Manag Perspect 32(1):104–124. https://doi.org/10.5465/amp.2016.0133
Deng Y, Guang F, Hong S, Wen L (2022) How does power technology innovation affect carbon productivity? A spatial perspective in China. Environ Sci Pollut Res 29(55):82888–82902. https://doi.org/10.1007/s11356-022-21488-0
Dong F, Zhu J, Li Y, Chen Y, Gao Y, Hu M, Qin C, Sun J (2022) How green technology innovation affects carbon emission efficiency: evidence from developed countries proposing carbon neutrality targets. Environ Sci Pollut Res 29(24):35780–35799. https://doi.org/10.1007/s11356-022-18581-9
Dong J, Li C (2022) Scenario prediction and decoupling analysis of carbon emission in Jiangsu Province, China. Technol Forecast Soc Change 185:122074. https://doi.org/10.1016/j.techfore.2022.122074
Eleftheriadis I, Anagnostopoulou E (2015) Relationship between corporate climate change disclosures and firm factors. Bus Strat Environ 24(8):780–789. https://doi.org/10.1002/bse.1845
Gao C, Song P, Wen Y, Yang D (2022) Effect of science and technology finance policy on urban green development in China. Front Environ Sci 10:918422. https://doi.org/10.3389/fenvs.2022.918422
Gao W, Zhu Z (2016) The technological progress route alternative of carbon productivity promotion in China’s industrial sector. Nat Hazards 82(3):1803–1815. https://doi.org/10.1007/s11069-016-2271-7
Gao Y, Zhang M, Zheng J (2021) Accounting and determinants analysis of China’s provincial total factor productivity considering carbon emissions. China Econ Rev 65:101576. https://doi.org/10.1016/j.chieco.2020.101576
George G, Prabhu G (2003) Developmental financial institutions as technology policy instruments: implications for innovation and entrepreneurship in emerging economies. Res Policy 32(1):89–108. https://doi.org/10.1016/S0048-7333(02)00002-1
Gu G, Zhang W, Cheng C (2022) Mitigation effects of global low carbon technology financing and its technological and economic impacts in the context of climate cooperation. J Clean Prod 381(1):135182. https://doi.org/10.1016/j.jclepro.2022.135182
Guo K, Zhang T, Liang Y, Zhao J, Zhang X (2022) Research on the promotion path of green technology innovation of an enterprise from the perspective of technology convergence: configuration analysis using new energy vehicles as an example. Environ Dev Sustain 25(6):4989–5008. https://doi.org/10.1007/s10668-022-02253-2
Han D, Ding Y, Shi Z, He Y (2022) The impact of digital economy on total factor carbon productivity: the threshold effect of technology accumulation. Environ Sci Pollut Res 29(37):55691–55706. https://doi.org/10.1007/s11356-022-19721-x
Han J, Zhang W, Isik C, Muhammad A, Yan J (2023) General equilibrium model-based green finance, decarbonization and high-quality economic development: a new perspective from knowledge networks. Environ Dev Sustain https://doi.org/10.1007/s10668-023-04072-5
Hu J, Hu M, Zhang H (2023) Has the construction of ecological civilization promoted green technology innovation? Environ Technol Innov 29:102960. https://doi.org/10.1016/j.eti.2022.102960
Intergovernmental Panel on Climate Change. Climate change 2021: the physical science basis. https://www.ipcc.ch/
Jin J (2019) The effects of labor market spatial structure and the built environment on commuting behavior: considering spatial effects and self-selection. Cities 95:102392. https://doi.org/10.1016/j.cities.2019.102392
Kashif M, Ullah A, Ullah S, Qian N (2024) Towards a greener future: the impact of financial technology (FinTech) and climate finance on ecological sustainability. J Environ Manag 370:122876. https://doi.org/10.1016/j.jenvman.2024.122876
Kilbourne W, Thyroff A (2020) STIRPAT for marketing: an introduction, expansion, and suggestions for future use. J Bus Res 108:351–361. https://doi.org/10.1016/j.jbusres.2019.10.033
Kong L, Jin X, Xu Y, Xu K (2024) The impact of China’s total carbon emission control policy on low carbon total factor productivity. Ind Innov https://doi.org/10.1080/13662716.2024.2388648
Li C, Wang Y, Zhou Z, Wang Z, Mardani A (2023) Digital finance and enterprise financing constraints: structural characteristics and mechanism identification. J Bus Res 165:114074. https://doi.org/10.1016/j.jbusres.2023.114074
Li L, Ma X, Ma S, Gao F (2024) Role of green finance in regional heterogeneous green innovation: evidence from China. Humanit Soc Sci Commun 11(1):1011. https://doi.org/10.1057/s41599-024-03517-0
Li W, Wang W, Wang Y, Ali M (2018) Historical growth in total factor carbon productivity of the Chinese industry-A comprehensive analysis. J Clean Prod 170:471–485. https://doi.org/10.1016/j.jclepro.2017.09.145
Li Y, Chen Y (2021) Development of an SBM-ML model for the measurement of green total factor productivity: the case of Pearl River Delta urban agglomeration. Renew Sustain Energy Rev 145:111131. https://doi.org/10.1016/j.rser.2021.111131
Liang H, Lin S, Wang J (2022) Impact of technological innovation on carbon emissions in China?s logistics industry: based on the rebound effect. J Clean Prod 377:134371. https://doi.org/10.1016/j.jclepro.2022.134371
Liao G, Yao D, Hu Z (2020) The spatial effect of the efficiency of regional financial resource allocation from the perspective of internet finance: evidence from Chinese provinces. Emerg Mark Financ Trade 56(6):1211–1223. https://doi.org/10.1080/1540496X.2018.1564658
Lin Z, Liao X, Yang Y (2022) China’s experience in developing green finance to reduce carbon emissions: from spatial econometric model evidence. Environ Sci Pollut Res 30(6):15531–15547. https://doi.org/10.1007/s11356-022-23246-8
Liu F, Tan D, Deng F, Wang Y (2023) How does green credit reduce carbon emissions? Dynamic spatial interactions and regional disparities. Environ Sci Pollut Res 30(26):68504–68523. https://doi.org/10.1007/s11356-023-27239-z
Liu L (2022) Quantitative impact analysis of financial support on regional science and technology innovation and productivity based on the multivariate statistical model. Math Probl Eng 2022:7175807. https://doi.org/10.1155/2022/7175807
Liu R, Zhou X, Xiang G, Yang J (2020) Effects of private enterprise participation in military industrial development on enterprise performances: evidences from China. Transform Bus Econ 19(2B):886–906
Liu W, Ye B, Liu Y (2020) Marine finance and marine science-tech innovation: an industrial panel data-based analysis. J Coastal Res 276–280. https://doi.org/10.2112/SI106-065.1
Liu X, Chai Z, Luo Y, Wang S, Liu B (2022) How to achieve sustainable development: from the perspective of science and technology financial policy in China. Environ Sci Pollut Res 30(10):26078–26093. https://doi.org/10.1007/s11356-022-23874-0
Liu X, Zhang X (2021) Industrial agglomeration, technological innovation and carbon productivity: Evidence from China. Resour Conserv Recycl 166:105330. https://doi.org/10.1016/j.resconrec.2020.105330
Long R, Gan X, Chen H, Wang J, Li Q (2020) Spatial econometric analysis of foreign direct investment and carbon productivity in China: two-tier moderating roles of industrialization development. Resour Conserv Recycl 155:104677. https://doi.org/10.1016/j.resconrec.2019.104677
Long Y, Liu L, Yang B (2024) The effects of enterprise digital transformation on low-carbon urban development: empirical evidence from China. Technol Forecast Soc Change 201:123259. https://doi.org/10.1016/j.techfore.2024.123259
Lu S, Wang J, Shang Y, Bao H, Chen H (2017) Potential assessment of optimizing energy structure in the city of carbon intensity target. Appl Energy 194:765–773. https://doi.org/10.1016/j.apenergy.2016.07.116
Lu Y, Ahmad M, Zhang H, Guo J (2023) Effects of science and technology finance on green total factor productivity in China: insights from an empirical spatial durbin model. J Know Econ https://doi.org/10.1007/s13132-023-01306-9
Lu Y, Guo J, Ahmad M, Zhang H (2022) Can sci-tech finance pilot policies reduce carbon emissions? Evidence from 252 cities in China. Front Environ Sci 10:933162. https://doi.org/10.3389/fenvs.2022.933162
Mao J, Tang S, Xiao Z, Zhi Q (2021) Industrial policy intensity, technological change, and productivity growth: evidence from China. Res Policy 50(7):104287. https://doi.org/10.1016/j.respol.2021.104287
Mo J (2021) Do environmental policy and innovation improve carbon productivity? Evidence from the Korean Emission Trading Scheme. Energy Environ 34(3):445–462. https://doi.org/10.1177/0958305X211064575
Murshed M, Apergis N, Alam M, Khan U, Mahmud S (2022) The impacts of renewable energy, financial inclusivity, globalization, economic growth, and urbanization on carbon productivity: evidence from net moderation and mediation effects of energy efficiency gains. Renew Energy 196:824–838. https://doi.org/10.1016/j.renene.2022.07.012
Nehru V, Swanson E, Dubey A (1995) A new database on human-capital stock in developing and industrial countries: sources, methodology, and results. J Dev Econ 46(2):379–401. https://doi.org/10.1016/0304-3878(94)00054-G
Noh H, Lee S (2022) Forecasting forward patent citations: comparison of citation-lag distribution, tobit regression, and deep learning approaches. IEEE Trans Eng Manag 69(4):1185–1196. https://doi.org/10.1109/TEM.2020.2978528
Ofoeda I, Amoah L, Anarfo E, Abor J (2024) Financial inclusion and economic growth: What roles do institutions and financial regulation play? Int J Financ Econ 29(1):832–848. https://doi.org/10.1002/ijfe.2709
Peng H, Qin X (2024) Digitalization as a trigger for a rebound effect of electricity use. Energy 300:131585. https://doi.org/10.1016/j.energy.2024.131585
Qiao Y, Guo C (2020) Evaluation model of financial resource allocation efficiency in coastal areas based on supply chain. J Coast Res 186-189:103. https://doi.org/10.2112/SI103-040.1
Romer P (1990) Endogenous technological change. J Political Econ 98(5):71–102
Shao S, Yang L (2014) Natural resource dependence, human capital accumulation, and economic growth: a combined explanation for the resource curse and the resource blessing. Energy Policy 74:632–642. https://doi.org/10.1016/j.enpol.2014.07.007
Shen Y, Fu Y, Song M (2023) Digital finance promotes sustainable total factor eco-efficiency: evidence from China. Appl Econ. https://doi.org/10.1080/00036846.2023.2211337
Sheng X, Lu B, Yue Q (2021) Impact of sci-tech finance on the innovation efficiency of China’s marine industry. Mar Policy 133:104708. https://doi.org/10.1016/j.marpol.2021.104708
Shi Y (2023) Big data simulation of science and technology finance supporting Shaanxi urban-rural integration development based on BP neural network algorithm. Soft Comput https://doi.org/10.1007/s00500-023-08371-x
Song M, Wang J, Wang S, Zhao D (2018) Knowledge accumulation, development potential and efficiency evaluation: an example using the Hainan free trade zone. J Knowl Manag 23(9):1673–1690. https://doi.org/10.1108/JKM-06-2018-0368
Sueyoshi T, Goto M (2012) DEA environmental assessment of coal fired power plants: methodological comparison between radial and non-radial models. Energy Econ 34(6):1854–1863. https://doi.org/10.1016/j.eneco.2012.07.008
Sun H, Chen T, Wang C (2024) Spatial impact of digital finance on carbon productivity. Geosci Front 15(3):101674. https://doi.org/10.1016/j.gsf.2023.101674
Tone K (2002) A Slack-based measure of super-efficiency in data envelopment analysis. Eur J Operat Res 143(1):32–41. https://doi.org/10.1016/S0377-2217(01)00324-1
Wang L, Wang H, Cao Z, He Y, Dong Z, Wang S (2022) Can industrial intellectualization reduce carbon emissions? - Empirical evidence from the perspective of carbon total factor productivity in China. Technol Forecast Soc Change 184:121969. https://doi.org/10.1016/j.techfore.2022.121969
Wang X, Gao C (2024) Does green finance policy help to improve carbon reduction welfare performance? Evidence from China. Energy Econ 132:107452. https://doi.org/10.1016/j.eneco.2024.107452
Wang X, Wang S, Zhang X (2023) Heterogeneity and influence factors of carbon productivity: evidence from Chinese manufacturing enterprises. J Clean Prod 404:136938. https://doi.org/10.1016/j.jclepro.2023.136938
Wang Z, Xue X, Ding W, Feng M (2024) The promotion effect and spillover effect of financial technology on regional innovation: evidence from China. Technol Anal Strat Manag https://doi.org/10.1080/09537325.2024.2328144
Wei Z, Huang L (2022) Invading the dynamics of economic growth and CO2 emission: panel data error correction model (ECM) approach. Environ Sci Pollut Res 29(48):73365–73381. https://doi.org/10.1007/s11356-022-20189-y
Wu Y, Yao L (2022) Carbon productivity and economic growth patterns in China. Econ Res Ekon Istraz https://doi.org/10.1080/1331677X.2022.2084437
Xu H, Liu W, Zhang D (2023) Exploring the role of co-agglomeration of manufacturing and producer services on carbon productivity: an empirical study of 282 cities in China. J Clean Prod 399:136674. https://doi.org/10.1016/j.jclepro.2023.136674
Xu J, Chen F, Zhang W, Liu Y, Li T (2023) Analysis of the carbon emission reduction effect of Fintech and the transmission channel of green finance. Financ Res Lett 56:104127. https://doi.org/10.1016/j.frl.2023.104127
Xue Q, Bai C, Xiao W (2022) Fintech and corporate green technology innovation: impacts and mechanisms. Manag Decis Econ 43(8):3898–3914. https://doi.org/10.1002/mde.3636
Yang Y (2022) Research on the impact of environmental regulation on China’s regional green technology innovation: insights from threshold effect model. Pol J Environ Stud 31(2):1427–1439. https://doi.org/10.15244/pjoes/141801
Yu H, Wei W, Li J, Li Y (2022) The impact of green digital finance on energy resources and climate change mitigation in carbon neutrality: case of 60 economies. Resour Policy 79:103116. https://doi.org/10.1016/j.resourpol.2022.103116
Yuan C, Shang M, Han Z, Wang J (2024) Research on the impact of the national ecological demonstration zone on green total factor productivity: evidence from China. J Environ Manag 356:120421. https://doi.org/10.1016/j.jenvman.2024.120421
Yuan H, Zhang T, Hu K, Feng Y, Feng C, Jia P (2022) Influences and transmission mechanisms of financial agglomeration on environmental pollution. J Environ Manag 303:114136. https://doi.org/10.1016/j.jenvman.2021.114136
Zhang L, Xiong L, Cheng B, Yu C (2018) How does foreign trade influence China’s carbon productivity? Based on panel spatial lag model analysis. Struct Change Econ Dyn 47:171–179. https://doi.org/10.1016/j.strueco.2018.08.008
Zhang M, Liu Y (2022) Influence of digital finance and green technology innovation on China’s carbon emission efficiency: empirical analysis based on spatial metrology. Sci Total Environ 838(3):156463. https://doi.org/10.1016/j.scitotenv.2022.156463
Zhang T, Feng T, Cui M (2023) Smart contract design and process optimization of carbon trading based on blockchain: the case of China’s electric power sector. J Clean Prod 397:136509. https://doi.org/10.1016/j.jclepro.2023.136509
Zhang W, Lu J, Tian W (2024) Environmental regulation and total factor carbon productivity. Chem Technol Fuels Oils 59(6):1239–1247. https://doi.org/10.1007/s10553-024-01640-x
Zhang W, You J, Lin W (2021) Internet plus and China industrial system’s low-carbon development. Renew Sustain Energy Rev 151:111499. https://doi.org/10.1016/j.rser.2021.111499
Zhang Z, Chen Y, Yan Q (2024) Effects of agricultural informatization on agricultural carbon emissions: a quasi natural experiment study in China. Appl Econ https://doi.org/10.1080/00036846.2024.2354118
Zhao J, Jiang Q, Dong X, Dong K, Jiang H(2022) How does industrial structure adjustment reduce CO2 emissions? Spatial and mediation effects analysis for China Energy Econ 105:105704. https://doi.org/10.1016/j.eneco.2021.105704
Zhao M, Lv L, Wu J, Wang S, Zhang N, Bai Z, Luo H (2022b) Total factor productivity of high coal-consuming industries and provincial coal consumption: based on the dynamic spatial Durbin model. Energy 251:123917. https://doi.org/10.1016/j.energy.2022.123917
Zhao Y, Goodell J, Dong Q, Wang Y, Abedin M (2022a) Overcoming spatial stratification of fintech inclusion: inferences from across Chinese provinces to guide policy makers. Int Rev Financ Anal 84:102411. https://doi.org/10.1016/j.irfa.2022.102411
Zhou X, Liu P, Liu Y, Luo X (2024) Understanding the effects of the tiered system in the National Equities Exchange and Quotations (Neeq) on the growth of technology-based SMEs from the perspective of financing constraints. Transform Bus Econ 22(3A):809–831
Author information
Authors and Affiliations
Contributions
Zhang Jinyue: writing—original draft, methodology, formal analysis, data curation. Sun Zhenglin: supervision, writing—review and editing, validation, and resources.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Ethical approval
Ethical approval was not required as the study did not involve human participants.
Informed consent
The participant has consented to the submission of the article research to the journal.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary information
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
About this article
Cite this article
Zhang, J., Sun, Z. Mechanisms and spatial spillover effects of science and technology financial ecosystem on carbon emission reduction from multiple perspectives: evidence from China. Humanit Soc Sci Commun 12, 1025 (2025). https://doi.org/10.1057/s41599-025-05423-5
Received:
Accepted:
Published:
Version of record:
DOI: https://doi.org/10.1057/s41599-025-05423-5
