Does Green Finance Promote Export Sophistication? An Analysis of the Moderating Effect Based on Green Taxes

: This study examines the impact of green ﬁnance on export technological complexity by using panel data from 30 provincial-level administrative units in China from 2011 to 2019. The study ﬁnds that green ﬁnance signiﬁcantly promotes export sophistication; with the promotion effect varying by the geographical location and institutional environment, the mechanism test shows that upgrading industrial structure and enhancing technological innovation are the two transmission paths for green ﬁnance to enhance export sophistication. Additionally, the study ﬁnds that green tax moderates the impact of green ﬁnance on export sophistication. The threshold effect test reveals that industrial structures, as well as their upgrades and technological innovation, have a single threshold. However, they need to reach a certain threshold value before they can play their role to the fullest, while green tax has a marginal increasing effect. The study provides a new perspective on the relationship between green ﬁnance and export sophistication, and the empirical evidence for current green ﬁnance policies promotes the development of the real economy.


Introduction
Since reforming and opening up, China has accelerated its integration into the international division of labor system and gained a foothold in the global value chain, based on the comparative advantage of production factors such as natural resources and labor. However, while obtaining huge economic benefits, the extensive external trade model has kept China at the middle and low ends of the global value chain for a long time, which has hindered the high-quality development of China's foreign trade [1]. With the rising cost of factors, as well as the impact of domestic and foreign epidemics, China's foreign trade growth rate has declined. According to statistics from the Ministry of Commerce, the foreign trade growth rate in US dollars, from January to April 2022, was 10.1%, which was nearly three-quarters lower than the 38.7% growth rate in the same period in the previous year. Notably, China's high economic growth has been accompanied by a critical waste of resources and environmental pollution. Revisiting environmental issues and promoting the economy in a green and low-carbon direction have become crucial to achieving the nation's aspirations for a better life. In this context, promoting stable and quality improvement in foreign trade has become a critical priority for China's economic development. The 20th National Congress of the Communist Party of China emphasized the need to optimize and upgrade the trade of goods, innovate the development mechanism of service trade, and accelerate the construction of a robust trading system to establish China as a strong trading nation. Therefore, China must achieve high-quality development. Enhancing the competitiveness of export products is crucial to improving the international competitiveness of foreign trade and achieving a transition from low to middle and high-end global value chains [2][3][4]. Export product competitiveness is determined, mainly, by export sophistication [5][6][7]. Therefore, under the new development pattern of domestic and international dual circulation, vigorously improving export sophistication is of great significance for changing traditional trade modes, optimizing the allocation of production factors, and reshaping the global economic structure.
Studies have found that financial development is important for improving and affecting export sophistication [8,9]. With the smooth progress of the financial reform, financial development has gradually become a "lubricant" for China's economic growth. After experiencing transitional reforms of a non-market-oriented financial system, exploratory reforms of framework construction, scale reforms of market-oriented guidance, and endogenous reforms of diversified optimization, China entered a stage of open-style reform with comprehensive innovation in the financial system, among which the most prominent intuitive manifestation was the reform and implementation of green finance policies. The "Green Credit Guidelines" document, issued in 2012 by the CBRC, guided financial institutions to implement green credit policies and set higher requirements for bank credit behavior, leading to the gradual establishment of green finance as a powerful constraint on market activities. The 'Guiding Opinions on Building a Green Finance' document, issued in 2016 by People's Bank of China and the Ministry of Finance, clarified the scope of the green finance system and its gradual penetration into all aspects of the real economy to achieve coordinated and sustainable development of the environment and the economy.
Does green finance improve export sophistication? There is one view that holds that financial institutions' credit decisions generally affect enterprises' export behavior choices [10,11], and bank institutions' green credit content directly relates to the regional green credit funds supply scale. To some extent, the implementation of green finance policy has improved green technological innovation in exports [12]. However, this view mostly examines export behavior from an enterprise perspective. At a stage where the financial system's open-style reform depends on government guidance, room exists for further exploration of the relationship between green finance and regional export sophistication. If green finance development can improve regional export sophistication, government departments should pay attention to local green finance development levels, continue to increase the effective supply of local green finance, improve the green finance system, and remove financing obstacles for the rapid promotion of export sophistication improvement. Otherwise, public policies formulated by the government should find a better balance between the development path and the orderly coordination between financial institutions and the real economy.
Based on the above analysis, this study attempts to introduce green finance into the research framework of regional trade development from the perspective of financial supply, and it comprehensively examines the impact of green finance on export sophistication. The literature related to this study concerns the factors that influence export sophistication. Gorodnichenko and Schnitzer (2013) previously found the impact of financing constraints on export sophistication, and they argued that financing constraints have a negative effect on firms' innovation decisions, thus affecting the improvement speed and catch-up level of national export sophistication [13]. Liu et al. (2014) also conclude that financing constraints inhibit improvements in export sophistication. However, no consensus exists on how to measure financing constraints in existing studies [14]. In addition, some scholars examined the factors influencing export sophistication from multiple perspectives, such as trade liberalization [15,16], digital trade [17][18][19], factor endowment [20,21], and institutional quality [22,23]. Although some scholars have explored the impact of financial openness [8,9] and digital finance [24] on export sophistication, from the perspective of financial openness and digital finance, there exists little discussion on the green factor of the financial supply side and export sophistication. There is even less research on the impact of the regional green finance development level on export sophistication.
Another stream of literature related to the theme of this study is the economic effects of green financing. Existing studies on the economic effects of green finance have primarily focused on the real economy, such as high-quality economic development [25,26] and manufacturing upgrades [27,28]. The literature most closely related to the main body of this study concerns green finance and export behavior. Jin Xiangyi et al. (2022) based their research on a quasi-natural experiment of Green Credit Guidelines to analyze the relationship between green finance and China's export trade. Jin Xiangyi et al. [29] found that green finance plays a significant role in promoting export trade. Yu Maomao and Ma Yanyan (2022) started with provincial panel data and examined the impact of green finance policy on export quality by using a synthetic control method. They concluded that the establishment of a green finance pilot zone significantly improved regional export quality, and provinces with a high degree of pollution had a more obvious effect on improving export quality [30]. In contrast, this study attempts to redraw the regional green finance development level from a more scientific perspective, focusing on the impact of green finance on export sophistication, rather than export trade scale and quality, to provide empirical evidence for accelerating trade power construction, promoting trade optimization, and upgrading in the current stage of high-quality economic development.
This study has three innovations. First, it explores the impact of green finance on export sophistication, which is unique compared to previous studies that have mainly focused on how financial development affects export trade. This study reveals that green finance mainly affects the improvement of export sophistication through two paths: accelerating green technology transformation and promoting industrial structure upgrading. Second, it constructs a scientific system to measure regional green finance development levels. It considers both the basic elements of regional objective green credit, green investment, green securities, carbon finance and green insurance, as well as subjective factors, such as government attention to green finance. It calculates the level of regional green finance development based on both objective and subjective perspectives. Third, it examines the moderating and threshold effects of regional green tax revenue on the relationship between green finance and export sophistication, providing empirical evidence of the conditions affecting green finance and export sophistication.
The rest of this paper is structured as follows: Section 2 presents the theoretical analysis and research hypotheses for the impact of green finance on export sophistication; Section 3 presents the model design and variable explanation, including model setting variable measurement and data sources; Section 4 presents the empirical test results and the analysis; Section 5 provides further analysis on the impact mechanism and institutional environment threshold effect of green finance and export sophistication, as well as the moderating effect of green tax revenue; Section 6 presents the conclusion and policy recommendations.

Green Finance and Export Technology Complexity
Green finance provides several ways of promoting export sophistication. First, it can provide financial support to the environmental protection industry, which requires significant R&D investments for developing new technologies, products, and services. Second, green finance can aid export enterprises in developing and producing high-tech green products with high added value and market competitiveness, thereby driving technological upgrading and industrial structure adjustment. Third, the development of green finance can drive the upgrading and expansion of related industrial chains, such as solar, wind, and biomass energy industries, by providing financial support for their expansion. By doing so, it can help reduce the complexity of export technology. Hence, we propose the following hypothesis: Hypothesis 1. Green finance is conducive to the optimization and upgrading of regional industrial chains and promotes the improvement of export technology complexity.

Mechanism of Green Finance Affecting Export Technology Complexity
In the modern financial environment, different financial governance methods and directions affect the inter-regional credit allocation ratio and efficiency [31]. The main goal of "dual carbon" is to build a green low-carbon circular development economic system, so green transformation development also emerges as a crucial response, and the role of green finance in this process cannot be overlooked [32]. The development of green finance affects the availability of green credit funds in regions [33], which consequently affects the regional trade structure. By combining relevant research, this study identifies three mechanisms by which green finance affects export technology complexity: first, green finance affects regional industrial structure upgrading; second, it affects regional green technology transformation efficiency; third, it can regulate the impact of green finance on export technology complexity.

Green Finance Promotes Export Technology Complexity Improvement by Promoting Industrial Structure Upgrading
Industrial structure upgrading has long been a key link for China to accelerate its modernization process. Owing to the obvious information asymmetry problem in the financial market, financing constraints have also become an important factor affecting local industrial structure upgrades. Green finance is related to the financing constraint problem of the green industry, which provides financing support for the green export industry and promotes industrial structure upgrading. At present, energy conservation and environmental protection, clean energy, sustainable agriculture, and other new green industries are in the emerging development stage, which still requires funds to purchase new equipment and achieve technological upgrades and transformation. Additionally, green finance for green development can reduce financing costs [34], provide targeted, long-term low-cost financial support for the green industry, control capital flow direction, help industrial structure upgrade to green low-carbon, and promote the sustainable development of the industry [35]. However, green finance also improves the investment attractiveness of the green industry, expands the scale effect of the export industry, and promotes industrial structure upgrading. First, green finance can provide risk management services for local enterprises, from the perspective of risk investors' preference, provide the timely launch of green financial tools and financial products that meet their needs, use risk diversification mechanisms to reduce financing risks, use capital efficiently, and create new opportunities for industrial structure upgrading [36]. Second, green finance provides technical support services for the green industry by cooperating with scientific research institutes and industrial organizations, building and promoting green standards, and using a certification system, providing environmental protection technology specifications and consulting services for local investors and enterprises. Finally, green finance provides marketing services, publicizes the concept of green finance through multiple channels, guides the public and enterprises to pay attention to environmental protection issues, raises environmental awareness, and emphasizes the necessity of upgrading the industrial structure. It can be seen that green finance can provide risk management, technical support, marketing, and other services to help green industries reduce risk, increase their return on investment, and promote industrial structure upgrades. Therefore, green finance can reduce information asymmetry in the financial market, promote the upgrading of the local industrial structure by alleviating financing constraints, improve the scale effect of green industries, and ultimately increase the added value of export products, driving the increase in the complexity of export technology. In summary, the following hypotheses are proposed: Promoting the upgrading of industrial structures is an effective path for green finance to increase the complexity of export technologies.

Green Finance Promotes the Improvement of Export Technology Complexity by Increasing the Conversion Rate of Technology
Green finance can increase the proportion of green funds in the region, and it can promote the improvement of export technology complexity by providing loans, venture capital services, and other ways to innovate and enhance green financial products and services. Green funds mainly come from funds generated by local governments and financial institutions because of the national green development policy orientation, which are the "lubricants" for achieving economically and environmentally sustainable development. Regions with a high level of green finance development have strong liquidity of green funds, which can encourage financial entities to actively participate in the green finance market and perform innovation activities for green financial products and services [37]. Undoubtedly, green development funds inject fresh blood into the innovation of export products and services, meet R&D funding needs for green innovation, and provide an important financing environment. However, through the funding of green innovation and improving the conversion rate of green technology, green finance can accelerate the transformation of R&D achievements and promote the development of advanced production equipment, technology, and management. The successful transformation of technological achievements is crucial for improving technological innovation in green products and services, as well as maximizing their market value [38]. The resulting improvement in the conversion rate of green technology can reduce the resource consumption, drive the development of green imports and exports, and ultimately promote the improvement of export technology complexity. Based on this, we propose the following hypothesis: Hypothesis 3. Improving the conversion rate of technology is an effective path for green financing to reduce export technology complexity.

Green Tax Can Regulate the Impact of Green Finance on Export Technology Complexity
Green tax, also known as environmental taxation, is an important means of government macro-control levied on enterprises with high pollution emissions and energy consumption, and it encourages enterprises to adopt more environmentally friendly and low-carbon production methods. Specifically, green taxation is a "green" tax system that aims to protect the environment, rationalize the exploitation of natural resources, promote green production and consumption, and establish an ecological tax system that protects the environment, thereby maintaining sustainable human development. Green taxes can regulate the impact of regional green finance on export technology complexity to a certain extent.
First, green taxes can motivate regional enterprises to innovate in environmental protection technologies. Levying a green tax increases the production cost of enterprises, prompting them to reduce energy consumption and environmental pollution in the production process, thus encouraging the adoption of more complex and environmentally friendly technologies in exports, ultimately increasing the added value of export product technology innovation [39]. Second, green taxes can increase the support for regional green financing, as well as government revenue and investment in environmental protection industries [40], thereby promoting the development of environmental protection industries, improving regional green technology level, and increasing export product competitiveness. Additionally, green taxes can optimize the structure of the environmental protection industry and encourage enterprises' environmental protection standards. Levying a green tax can guide enterprises to transform from traditional high-pollution and high-energy consumption industries to green low-carbon industries, thereby encouraging export enterprises to pay more attention to technology complexity and environmental performance in terms of environmental protection, optimizing the regional industry structure [41], and promoting regional export technology complexity improvement. Based on this, we propose the following hypothesis: Hypothesis 4. Green tax plays a moderating role in the relationship between green finance and export technology complexity, that is, in regions with a high degree of green tax, the promotion effect of green finance on export technology complexity is stronger.

Baseline Model Setting
Regarding econometric model design, the econometric regression model constructed to test hypothesis Hypothesis 1-that is, to verify the impact of green finance on the technological complexity of exports-is as follows: In Equation (1), Expy it denotes the export technology complexity of province i in year t; GF it denotes the green finance of province i in year t, which is used to measure the green finance development of each province, with higher values indicating a higher level of green finance development in that province; Control it denotes a series of control variables in the regression of this paper, α 0 is a constant, δ i is a province-fixed effect, and ε it is a random disturbance term.
To test Hypothesis 2 and verify whether industrial structural upgrading (Ind) is a mediating variable in the technical complexity of green finance for exports, the following model was constructed: To test Hypothesis 3 and verify whether the green technology conversion rate (Tech) is a mediating variable in the technological sophistication of green finance for exports, the following model was constructed:

Variable Measurement and Explanation
Export sophistication (Expy). Following the idea of Hausmann et al., (2007) [6], the measurement steps are specified as follows: (1) Calculate the technical complexity (Prody) of each category of exported products in each province, based on information on commodity exports provided by the Chinese customs database, as shown in Equation (6).
In Equation (6), j denotes the commodity export category, i represents the province, x ij is the export trade value of commodity category j in province I; X i denotes the total export trade value of all commodities in province i, and Y i is the GDP per capita of province i, using the resident price index and real value in 2000 as the base year.
(2) Based on the technical complexity (Prody) of each type of export product for each province calculated in Equation (6), the technical complexity (Expy) of exports for each province is calculated using a weighting method, as shown in Equation (7).
Considering the effect of data heteroscedasticity on the regression results, regression analysis was conducted after taking the logarithmic values of Expy obtained above.
Green Finance (GF). The discussion on green finance can be traced back to the 1970s, and for China, the construction of a green finance system can be traced back to 2005 when the State Council issued the "Decision on Implementing the Scientific Development Concept and Strengthening Environmental Protection," emphasizing that environmental and economic development must work simultaneously. By 2022, the "Guidelines for Green Finance in Banking and Insurance Industry" (CBIRC [2022] No. 15), issued by the China Banking and Insurance Regulatory Commission, required banks and insurance institutions to actively support the construction of a clean and low-carbon energy system, support energy conservation, pollution reduction, carbon reduction, greening, and disaster prevention in key industries and fields. They were also required to promote the application of green and low-carbon technologies. With the expansion of green finance research, scholars have depicted the level of green finance development in China from multiple perspectives [42][43][44].
Green finance refers to the financial services provided to support economic activities for environmental improvement, climate change, and the efficient use of resources, such as project investment and financing, project operation, and risk management in the fields of environmental protection, energy conservation, clean energy, green transportation, and green buildings [45][46][47]. Based on the connotation of green finance, considering that the role of green finance policy guidance in China is relatively strong, this study introduces the government's attention to the indicator system of green finance depiction by many scholars; starting from the annual government work reports of each province, we selected 81 keywords related to the connotation of green finance, including green credit, green finance, carbon finance, and green development (see Table 1). Subsequently, we conducted a word frequency search, and we added up the obtained word frequency numbers. These numbers are defined as the attention of provincial government departments to green finance.
This study constructs a more comprehensive indicator system of green finance development, in China, from two perspectives: subjective (i.e., government attention to green finance) and objective (i.e., green credit, green investment, green securities, carbon finance, and green insurance). It includes six first-level indicators (green credit, green investment, green securities, carbon finance, green insurance, and green finance attention) and eight second-level indicators, calculates their corresponding weights using the principal component analysis method, and finally obtains a comprehensive indicator reflecting the level of green finance development in each province. Table 2 presents the indicator system used for green finance construction.
Control variables. (1) Foreign direct investment (Fdi) is measured as the logarithm of actual foreign capital utilization in each province. (2) The degree of openness (Open) is measured as the ratio of import and export trade volumes to the regional GDP in each province. (3) Human capital (Hr) is measured as the logarithm of college students in each province. (4) The economic development level (GDP) is measured as the logarithm of the per capita GDP for each province. (5) Government intervention (Gov) is measured as the ratio of R&D expenditure input to the GDP in each province. Mediating variables. According to the theoretical mechanism analysis in the previous section, the green technology conversion rate and industrial structure upgrading are two channels for green finance that affect export technology complexity. Therefore, this study's mediating variables include the green technology conversion rate (Tech) and industrial structure upgrades (Ind). Databases such as the National Bureau of Statistics have not yet published relevant information on provincial green technology conversion indicators, so the green technology conversion rate (Tech) is measured by the logarithm of authorized patent applications. To measure industrial structure upgrading (Ind), we calculated the industrial structure upgrading coefficient for each province.
Moderating variable. Green Tax (GT). Regarding the connotation of green tax, scholars mainly define it as the total revenue of various types of green taxes, which can be divided into two categories: narrow-sense green tax and broad-sense green tax. The narrowsense green tax mainly refers to the tax policy specially formulated by the government for environmental protection, energy conservation, and emissions reduction, which is manifested as environmental protection tax collection. The broad-sense green tax mainly refers to a series of tax policies related to environmental protection formulated by the government, which are manifested as a resource tax, vehicle and vessel tax, consumption tax, environmental protection tax, urban land use tax, urban maintenance and construction tax, vehicle purchase tax, and cultivated land occupation tax. Considering China's basic green tax situation, this study selected the ratio of the sum of the eight major taxes to the total tax revenue as a measure of each province's degree of green taxation. Table 3 presents the descriptive statistics of these variables.
A prerequisite for regression analysis is that the dependent variable must meet normal distribution characteristics, so this study also conducted D'Agostino and Shapiro-Wilk tests, for the dependent variable Expy, to determine the normal distribution characteristics of Expy, respectively. The results showed that the skewness of Expy was 0.9039, kurtosis was 0.1074, the p-value was 0.2691 in the D'Agostino test, and the p-value of Expy was 0.4977 in the Shapiro-Wilk test. Both tests rejected the original hypothesis, indicating that the dependent variable Expy obeyed the characteristics of a normal distribution. Number of words related to green finance in government work reports + Note: Green credit is often referred to as sustainable finance or environmental finance, and it is an important economic tool for the environmental protection sector and banking industry to work together to counteract corporate environmental violations, promote energy conservation and emission reduction, and avoid financial risks. Green investment is a new type of investment that aims to achieve ecological environmental protection, comprehensive pollution control, rational use of resources, economic recycling, and harmony between people and society. Green securities refer to modern securities models in which the securities industry adheres to the concepts of green, circular, and low-carbon development; they incorporate environmental protection verification, environmental performance assessment, and environmental information disclosure into the securities market index system; they promote the comprehensive, coordinated, and sustainable development of material, spiritual, and ecological civilization through the innovation and internal development of the securities foundation and regulatory systems. Carbon finance refers to the financial system and various services provided by financial trading activities to reduce greenhouse gas emissions. Green insurance refers to the insurance industry's economic actions in providing risk protection and financial support for environmental resource protection and social governance, green industry operations, and green life consumption. Green financial concern refers to the extent to which governments focus on developing local green finance.

Data Sources
Owing to serious missing data for most variables in Tibet, this study selected panel data from provinces other than Tibet to explore the impact of green finance on export technology complexity. Export-related data come from the China Customs Import and Export Database; the rest of the variable-related data come from the China Statistical Yearbook, China Science and Technology Statistical Yearbook, China Taxation Yearbook, Provincial Statistical Yearbook, and National Bureau of Statistics. Finally, panel data for 30 provincial administrative units, from 2011 to 2019, were obtained.

Baseline Regression Results
Based on the theoretical analysis in the previous section, this study examined the impact of green finance on export sophistication through regression analysis. Table 4 presents the regression results for the full sample, while controlling for provincial fixed effects. Column (1) shows the regression result for green finance on export sophistication without adding control variables. The result shows that the regression coefficient of GF passes the test at the 1% significance level and is significantly positive, indicating that green financial development can improve export sophistication. Columns (2)(3)(4)(5)(6) show the regression results after adding control variables, and the GF coefficient remains significantly positive at the 1% level, indicating the core explanatory variable. Therefore, Hypothesis 1 is verified.

Heterogeneity Analysis
Owing to variations in green finance development and export sophistication across different regions, the impact of green finance on export sophistication in different regions may also be different. Therefore, this study conducted a heterogeneity analysis from the following two aspects: First, differences in geographical location. The total sample was divided into three groups-eastern, central, and western regions-and heterogeneity analysis was conducted at the regional level. Second, institutional environmental differences existed. Specifically, the marketization index compiled by Wang et al. (2019) [48] was used. Based on the median of the marketization index, the total sample was divided into regions with perfect institutional environments and regions with imperfect institutional environments.
Columns (1-3) of Table 5 give the regression results after distinguishing the geographical locations. After dividing into eastern, central, and western regions, the regression coefficient of green finance (GF) is not significant in the central and western regions, while it is significantly positive in the eastern regions, indicating that green finance development has the most significant effect on improving export sophistication in the eastern regions. The reasons for this are as follows: First, different industrial structures exist. Compared with central and western regions, the eastern region has more developed manufacturing and high-tech industries that have higher requirements for technological complexity and environmental protection. Therefore, regarding green finance development, the eastern regions can promote technological innovation and upgrade faster, thereby promoting sophisticated export improvements. Second, different technological innovation capabilities exist. The eastern regions have stronger technological innovation capabilities, which are particularly evident in green finance development. Green finance is currently an emerging field that requires significant technological innovation, so eastern regions with more advantages in technological innovation can promote green finance development faster. Third, different intensities of policy support exist. Policy support is an important means of promoting green financial development. Eastern regions have more policy support, including tax incentives and subsidies, which help enterprises in the eastern regions conduct green finance-related businesses more successfully, thereby promoting technological innovation and export sophistication improvement. Eastern regions have more advantages than the central and western regions regarding industrial structure, technological innovation capability, and policy support intensity; hence, eastern regions can promote technological innovation and upgrade faster regarding green finance development, thereby promoting export sophistication improvement. Columns (4) and (5) present the regression results after accounting for institutional environments. Regarding the subsample with a perfect institutional environment, the regression coefficient of green finance, GF, is significantly positive at the 1% level, while for the subsample with an imperfect institutional environment, the coefficient is not significant. Generally, financial development positively correlated with the degree of institutional perfection. Regions with more perfect institutional environments tend to have higher marketization degrees, sounder financial supervision and legal systems, and better-behaved financial market participants, which enhance financial market transparency, fair-ness, and ultimately promote financial market development. Moreover, a perfect institutional environment simplifies financial business operations, reduces transaction costs, encourages financial innovation, promotes financial product diversification, improves financial market diversity and liquidity, and further advances financial market development. Based on these factors, green finance has a stronger effect on improving export sophistication in the regions with more perfect institutional environments.

Impact Mechanism of Green Finance on the Improvement of Export Technology Complexity
According to the theoretical mechanism analysis in the previous section, technological innovation and industrial structure upgrading are the two paths through which green finance affects the complexity of export technology. This study first estimates the impact of green finance on technological innovation and industrial structure upgrading, and then, it examines the impact of technological innovation and industrial structure upgrading on export technology complexity, thus obtaining discovering how green finance affects export technology complexity by affecting technological innovation and industrial structure upgrading. Table 6 shows the regression results of the impact-mechanism tests described in the previous section. Observing Table 6, columns (1) and (2) test the impact of green finance on export technology complexity through the channel of industrial structure upgrading, where column (1) estimates the impact of green finance on industrial structure upgrading, and the coefficient of GF passes the test at the 1% significance level with a positive sign. This means that an increase in the level of green finance can significantly promote the improvement of export technology complexity. Column (2) estimates the impact of industrial structure upgrading on export technology complexity under the control of green finance, and the coefficient of Ind is significantly positive and passes the test at the 1% significance level, indicating that increasing the industrial structure upgrading coefficient does promote the improvement of export technology complexity, and hypothesis Hypothesis 2 is tested. Similarly, columns (3)(4) test the impact of green finance on export technology complexity through the channel of technological innovation. Specifically, column (3) tests the impact of green finance on regional technological innovation, and the regression coefficient of GF passes the test at the 1% significance level with a positive sign, meaning that an increase in the level of green finance can increase the level of regional technological innovation. Column (4) estimates the impact of technological innovation on export technology complexity under the control of green finance, and the regression coefficient of Tech is significantly positive and passes the test at the 1% significance level, meaning that increasing regional technological innovation does promote regional export technology complexity improvement, and hypothesis Hypothesis 3 is tested. Therefore, green finance promotes the improvement in regional export technology complexity through two channels: industrial structure upgrades and technological innovation.

Moderating Effect of Green Taxation
Considering that the regional green taxation degree affects the impact of green finance on export technology complexity, this study constructs an interaction term between green finance and green taxation (GF*GT) and adds it to the baseline regression model to verify the moderating effect of green taxation, as shown in Column (5) of Table 6. It can be found that the estimated coefficient of the interaction term between green finance and green taxation (GF*GT) is 3.0936, which passes the test at a 1% significance level. Green taxation has a significant moderating effect on green finance's impact on export technology complexity: the higher the degree of green taxation, the stronger the promotional effect of green finance on export technology complexity, thus verifying Hypothesis 4.

Threshold Effect Test
Based on the preliminary verification of industrial structure upgrading, technological innovation, and green taxation's impact on green finance's influence on export technology complexity, to further examine whether different stages of industrial structure upgrading, technological innovation, and green taxation will lead to non-linear characteristics of green finance's impact on export technology complexity, this study considered using a threshold effect model for testing. This study used a fixed-effects model to test the assumption that there is no threshold, single threshold, double threshold, or triple threshold. The constructed threshold regression model was as follows: where Y it represents threshold variables for industrial structure upgrading, technological innovation, and green taxation, respectively; I(·) represents the indicator function; τ 1 represents the threshold value to be estimated; other variable definitions remain unchanged. Table 7 shows threshold affect test results. It is evidenced that industrial structure upgrading, technological innovation, and green taxation all have single thresholds. Among them, industrial structure upgrading's threshold value is 2.3405, which passes the 5% significance test; technological innovation's threshold value is 8.3018, which passes the 10% significance test; green taxation's threshold value is 24.4623, which also passes the 10% significance test. Table 8 shows panel threshold regression results. When industrial structure upgrading is below threshold value 2.3405, the estimated coefficient of green finance is 0.1034, which does not pass the significance level test. When industrial structure upgrading is above threshold value 2.3405, the regression coefficient rises to 0.4919, indicating that, when industrial structure reaches a certain height, green finance has a positive impact on export technology complexity. When technological innovation is below threshold value 8.3018, the estimated coefficient of green finance is 0.1410, which does not pass the significance level test. When technological innovation is above threshold value 8.3018, the regression coefficient rises to 0.8552, indicating that technological innovation also needs to reach a certain height for green finance to have a positive promotion effect on export technology complexity. When green taxation is below the threshold value of 24.4623, the estimated coefficient of green finance is 0.502, which passes the 1% significance level test; when green taxation is above the threshold value of 24.4623, the estimated coefficient of green finance is 0.954, and it passes 1% significance level, indicating that green taxation crosses a certain threshold value and can effectively improve the impact on export technology complexity. Therefore, it can be seen that industrial structure upgrading and technological innovation need to cross a certain threshold to maximize the impact on export technology complexity, and green taxation has a marginal increasing effect on the impact of green finance on export technology complexity. The previous regression analysis measured the export sophistication of each province using the per capita GDP. However, per-capita GDP cannot accurately measure the development level of the manufacturing sector in each province. Moreover, Expy measures the similarity between the export structure of the manufacturing sector in each province and that of provinces with a high per capita GDP. Using per capita GDP to measure export sophistication may reduce the accuracy of measurement results. Therefore, following Li et al. (2022) [49], this study replaces per capita GDP with the labor productivity of the manufacturing sector in each province to re-measure export sophistication. The regression results are shown in Column (1) of Table 9. After replacing the measurement method for export sophistication, the positive effect of green finance on export sophistication remains robust.

Endogeneity Test
Regions with higher export sophistication may have more developed financial markets and more advanced technological innovation and, thus, may have higher levels of green finance development, so an endogeneity problems may exist between the dependent and independent variables. Therefore, to better solve the endogeneity problem in the regression equation, this study adopts the system GMM estimation for the robustness test. In addition, we used the lagged value of green finance and the average value of the green finance development level in neighboring provinces as indirect variables for green finance. The test results are represented in Columns (2-4) of Table 9. It is evidenced that, whether it is system GMM or instrumental variable method, green finance still significantly promotes regional export sophistication, which is consistent with the baseline regression results, indicating the robustness of the basic regression results.

Conclusions and Policy Recommendations
Given the gradual emergence of drawbacks in traditional trade modes and new competitive trade advantages, improving export sophistication has become an effective path for China's high-quality foreign trade development.
Improving export sophistication cannot be separated from effective support of the regional financial system. However, few studies have explored the impact of regional green finance on export sophistication from the financial supply perspective. This study finds that green finance significantly promotes export sophistication, and this promotion effect has heterogeneity in the geographical location and the institutional environment, indicating that promoting industrial structure upgrading and technological innovation are two transmission channels for green finance to promote export sophistication improvement, and green tax regulates the impact of green finance on export sophistication. The threshold effect test shows that industrial structure upgrading, technological innovation and green tax all have single thresholds; however, industrial structure upgrading and technological innovation need to reach certain threshold values before they can play their role to the fullest, while the green tax has a marginally increasing effect on the impact of green finance on export sophistication.
Based on these conclusions, this study proposes suggestions on using green finance development to promote export sophistication.
First, it should build a good overall plan for green finance development, improve the positive role of green finance in export sophistication, and highlight China's new competitive advantage in exports. With the international community's attention on "carbon neutrality" and "carbon peaks", green transformation development has become crucial to China's green high-quality sustainable development. We should accelerate the construction of green finance systems, fully consider the coverage and penetration rate of financial institutions in various regions in the spatial layout, improve the layout of small and medium-sized banks, promote the steady development of green finance, increase green credit fund allocation, improve regional enterprises' convenience in obtaining green finance services, and thus improve regional export sophistication.
Second, considering regional differences allows for the leveraging of financial policy advantages to improve financing for export enterprises, as well as enhance the exportupgrading effect of green finance. Given the heterogeneity in the geographical location and institutional environment of the impact of green finance on export sophistication, local governments should make use of macro-control measures to establish a robust market institutional environment, guide financial institutions to provide credit support to green development enterprises, enhance credit supplies for regional enterprises, encourage and support financial institution personnel to engage in deep communication with other regional financial institutions, and fully utilize the positive role of financial institutions.
Third, increasing financial services for the real economy, improving the regional industrial structure and technological innovation levels, and improving the green tax supervision system are important channels through which green finance promotes export sophistication. If the regional green finance development level improves, the regional financial services for the real economy will be stronger. Optimizing regional industrial structures also provides a good development environment for improving technological innovation levels. In the digital economy era, relevant government departments can establish a green tax supervision and management committee, as well as green finance development innovation bases. They can clarify market access standards, improve green tax collection and management systems, thereby promoting enterprises to perform the intelligent transformation and improving export processes, allowing for the leveraging of financial policy advantages in improving financing for export enterprises and enhancing the export-upgrading effect of green finance.