In the era of globalization, financial and natural resources are important indicators that contribute significantly to mitigating environmental degradation and promoting economic growth (Usman et al., 2022a; Usman et al., 2022b; Zhang C et al., 2022). Green finance integrates financial and natural resources into consideration and directs financial resources to energy-saving and emission-reducing production activities through environmental regulation to achieve sustainable development while ensuring economic growth (Criscuolo & Menon, 2015). On the one hand, green finance can promote environmental protection and governance (Al Mamun et al., 2022), allocating resources from highly polluting and energy-consuming industries to industries with advanced production concepts and environmental technologies (Falcone, 2020). On the other hand, environmental protection regulations can also promote green finance (Cojoianu et al., 2020), as confirmed by several studies, and the regulations are spatially heterogeneous and firm-heterogeneous in promoting green finance in different regions of China (Xu et al., 2022).

In recent years, research on green finance and green finance policies has attracted the attention of an increasing number of scholars (Pu et al., 2018; Zhang C et al., 2019; Zhang D et al., 2019; Azhgaliyeva & Liddle, 2020; Akomea-Frimpong et al., 2021; Sun et al., 2021; Jahanger et al., 2022; Jiang et al., 2022; Pu & Yang, 2022). Several scholars have developed a framework for green finance and defined it (Wang & Zhi, 2016; Taghizadeh-Hesary & Yoshino, 2019; Hafner et al., 2020; Zhang et al., 2020). Lindenberg explains the concept of green finance as policies from financial institutions to sustain a green economy (Blinder, 2012; Lindenberg, 2014). The ‘financial’ aspect of the concept demonstrates the allocation and investment of capital through the financial system (Berensmann et al., 2017; Weber & ElAlfy, 2019). The ‘green’ attribute requires that the allocation of financial resources should extend to environmental protection, clean energy, green buildings, climate change, and corporate governance in all economic sectors (Yuan & Gallagher, 2018; Urban & Wójcik, 2019).

According to the official definition from the Chinese government ³ , green finance includes the following aspects. First, green finance aims to provide financial support for projects related to environmental protection to address the problems caused by climate change and improve energy efficiency. Secondly, green finance is divided into green bonds (Sinha et al., 2021), green credits, and green stock indices. Furthermore, green securities, green investments, climate finance, carbon finance, green insurance, green credit, and green infrastructural bonds are banks’ major green finance products. Finally, it clearly states that green finance involves green project financing and risk management, and carbon finance issues. A clear definition of green finance helps label green financial products and attract green companies and investors to make more green investments. Since then, green finance policies have been introduced in various regions of China (Lee, 2020; Muganyi et al., 2021), and many scholars have conducted studies on the impact of green finance policies (Xu & Li, 2020; D.; Zhang D et al., 2021).

In terms of the relationship between green finance and regional economic development, the development of green finance in China has shown a general upward trend, with high-value regions located mainly in the east, followed by the central and western regions, and the coupled coordination between green finance and environmental performance has increased from uncoordinated to primary coordination over time (H. Zhang H et al., 2022). It was found that green finance can improve total factor productivity in Chinese regions, and this effect is more pronounced in provinces with high economic and social conditions and higher pollution levels (Lee & Lee, 2022). Meanwhile, the green finance development index shows a general upward trend. However, the overall development level in China is not high, which restricts the coordinated development of green finance and the economy in different aspects and leads to the insignificant support of green finance for economic growth (Yin & Xu, 2022). The gap in green finance development between different regions is narrowing, and the development level is distributed according to a ladder of East China, Central China, West China, and Northeast China (Lv et al., 2021).

In terms of promoting industrial development, green finance has the highest correlation with the output value of tertiary industries, followed by primary and secondary industries. The highest impact on industrial structure upgrading is in the east, followed by central and western China (Wang & Wang, 2021). In addition, green finance can impact high-quality economic development through three aspects: ecological environment, economic efficiency, and economic structure, and the development of financial technology can help strengthen this impact (Yang et al., 2021). In terms of energy market impact, the impact of green finance on energy consumption shows spatial heterogeneity, with the level of green finance development in the eastern region having a more pronounced positive impact on energy consumption structure (Sun & Chen, 2022). Green finance has also played a significant and positive role in improving environmental ecology and investment in renewable resources (Li et al., 2022), especially after the Covid-19 outbreak in 2020 (Arif et al., 2022). It has been shown that green finance has a U-shaped effect on China’s regional economic and ecological development, which is more pronounced in the central region (Sun & Chen, 2022). This fact has also been verified in other studies such as the study that found a negative correlation between green finance and eco-efficiency, showing a non-linear U-shaped trend, also more pronounced in the central and western regions (Wang et al., 2022).

In the policy research on green finance, some researchers used data-driven scientometric methods from the literature and found four policies that are considered to have great future potential, including carbon taxes, government subsidies, green bonds, and some other policies. Green bonds have received the most attention (Wang et al., 2021). Green finance policies effectively mitigate the effects of financing constraints, and green innovation suffers when firms face higher financing constraints, but green finance policies do not seem to benefit private firms (Yu et al., 2021). Firms in economically developed regions are more strongly affected by green credit than firms in less economically developed regions, and both green credit policies and green credit development increase the cost of debt financing for energy-intensive and polluting firms but reduce the cost of debt financing for green firms (Xu & Li, 2020). The debt financing of heavily polluting Chinese firms declined significantly after introducing green credit policies, suggesting that the green credit policy system plays a guiding role in credit resource allocation (Liu et al., 2019). In addition, industrial gas emissions decreased significantly during the green finance policy review period (Muganyi et al., 2021; Meo & Abd Karim, 2022), while there are also studies showing that green credit policies help reduce sulfur dioxide and wastewater emissions (S. Zhang S et al., 2021).

Regarding the impact of pilot zones for green finance reform and innovation, due to the short period after the establishment, most studies on pilot zones for green finance reform and innovation have focused on theoretical aspects, emphasizing the inspiration of relevant policy experiences on regional economic development (Jingzhi et al., 2017; Wencong & Weijun, 2020; Yunfan et al., 2021), or discuss the impact of pilot zones on firms’ green innovation (Lu & Xianchun, 2020; Ying & Yao, 2021). A synthesis of existing research on green finance and green financial policies reveals that the existing literature is rich in research on the policy effects of green finance. However, it mainly focuses on the implementation effects of green credit policies in green finance, and the research on pilot zones for green finance reform and innovation is still very limited. Second, based on the availability of data, most of the existing studies focus on the effects of green finance on regional economic development, environmental governance, and corporate investment and financing behaviors, and there is a lack of literature that examines the implementation effects of pilot zones from the perspective of innovation. Third, the research on green financial policies on green technology innovation is focused on the enterprise level, and there is a lack of spatial perspective to test the effect of policy implementation. Because of this, this paper takes Chinese cities as the research object and adopts a DID model to conduct a multidimensional empirical test on the policy effects of pilot zones for green finance reform and innovation.

In this paper, we select the data of Chinese cities and green patents from 2010 to 2019. In the end, 2,890 observations were finally obtained from 297 cities, and the samples were standardized. The city economic data were obtained from China City Statistical Yearbook from 2011 to 2020, and the regional green patent data were obtained from the Chinese Research Data Services (CNRDS) Platform.

This paper aims to test whether pilot zones for green finance reform and innovation can promote regional green technology innovation. In the existing literature, DID is a more effective method used to test policy effect. In this paper, the five provinces of Zhejiang, Jiangxi, Guangdong, Guizhou, and Xinjiang are used as the treatment group, and the remaining provinces are used as the control group. To verify whether the pilot zones for green finance reform and innovation help promote regional green technological innovation, we construct a model 1) with the following model settings. T g r e e n i t = β 0 + β 1 T r e a t r ∗ T i m e t + β 2 T r e a t r + β 3 T i m e t + ρ X i t + δ r + γ t + ε i r t (1) T g r e e n i t denotes the number of green patents of city i in the year t . T r e a t denotes the dummy variable for pilot zones for green finance reform and innovation and takes one if the city is located in a pilot zone, otherwise takes 0. T i m e is the dummy variable before and after the policy, this paper assigns the year 2017 to 1, the years after 2017 to 1, and the years before 2017 to 0. X i t is a matrix of control variables, including nine indicators of the city, as detailed in Table 2.

In the benchmark analysis, the coefficient β 1 of the DID term Treat   r * Time   t is the focus of this paper. The coefficient reflects the impact of the pilot zones on regional green technology innovation before and after the announcement between pilot and non-pilot zones. If β 1 is significantly positive, it indicates that pilot zone helps promote green technology innovation. In addition, the model controls for province fixed effects and year fixed effects over time, denoted as δ r and γ t , respectively, and ε i r t is a random disturbance term.

1) Explained variables.

The research perspective of this paper is to examine the effectiveness of pilot zones on regional green technology innovation activities. Therefore, the number of regional green patents is adopted as the explained variable. Patents are an effective indicator of technological innovation, and green patents can most intuitively reflect the output of regional green technological innovation activities. The patent data can be further classified according to different technological properties to reflect innovation activities’ different value connotations and contributions. In the specific study, this paper explores the influence of pilot zones on different kinds of green innovation behaviors in the region from two aspects of patent application and patent acquisition, respectively. There are four indicators for each aspect, and eight indicators in total, as detailed in Table 3. The indicators for each aspect have been subdivided into two subdivisions, i.e., the number of patents and the percentage of the number of patents, representing the absolute achievement and relative advantage of the regional green innovation level, respectively. Table 3shows the descriptive statistical information of the explained variables after standardization. From the descriptive statistics of the explained variables, the level of green innovation in technology varies unevenly and widely among regions according to the mean, maximum, and standard deviation.

2) Control variables.

This paper selects a series of city-level influencing factors as control variables. First, the larger the total population in a city jurisdiction, the more people are likely to engage in green technology innovation, and the more patents are obtained. Second, if students’ level of education in the area is higher, the greater the probability that students will engage in innovative activities and achieve results in school and their graduate work. Finally, as the size of firms in a city increases, firms’ profitability increases due to scale expansion. At this time, the enterprises are more likely to provide their employees with platform support and financial assistance to conduct green innovation activities. Therefore, the above considerations led to selecting nine control variables, as shown in Table 4.

According to the benchmark regression model, this section examines the quantitative impact of pilot zones on urban green technology innovation, and the estimated results are shown in Table 5.

The regression results in Table 5 indicate that pilot zones have promoted green technology innovation. According to Table 5, the regression coefficients of Treat*Time in columns (3), (5), (7), 8) are 0.303, 0.370, 0.335, and -0.146, respectively, and all of them are significant at different confidence levels. This result indicates that pilot zones have promoted the number of green utility patents obtained, the number of green invention patents applied, and the number of green utility patents applied. However, pilot zones are negative for the percentage of the number of green utility patents to the total number of patents applied. Among the control variables, population size exhibits a significant adverse effect in several models, indicating that regional innovation does not necessarily increase with regional population size. The increase in employees and people engaged in scientific and technical activities leads to regional green innovation, which is also reflected in research expenditures and education levels.

The reasons for the above empirical results may be as follows: on the one hand, since the pilot zone was established in 2017, the data covers from 2010 to 2019, and the sample period is short, which may cause policy effects of the pilot zone cannot be fully demonstrated; on the other hand, according to the Patent Law of the People’s Republic of China, an invention patent refers to a new technical solution for a product, method or its improvement; a utility patent refers to a new technical solution for the shape, structure or combination thereof of a product that is suitable for practical use. From the above explanation, the level of inventiveness of a green utility patent is lower than that of a green invention patent. Therefore, the incentive effect of the policy is firstly reflected in the application and acquisition of green utility patents and then, over time, in the number of green inventions patents. In addition, since the technical field of invention patents is broader, their examination period is generally more than one and a half years. In contrast, the examination period of a utility patent is generally half a year, and the difference in time may also be one of the reasons for the insignificant effect on the number of green inventions patents obtained.