The resource-based view (RBV) and the knowledge-based view (KBV) explain why innovation affects corporate FS from the perspective of competitive advantage. RBV points out that companies with scarce resources that are difficult to imitate will gain a stronger competitive advantage and show better and sustainable financial performance (Wernerfelt, 1984; Barney, 1991; Barney, 2001; Kuncoro and Suriani, 2017; Hameed et al., 2020). Inimitable resources come from technology innovation (Bakar and Ahmad, 2010), because technology innovation helps companies develop new products and services, construct the barrier from their competitors, and make companies expand their business scale, then enhance their competitiveness (Ireland et al., 2001; Su and Zhou, 2019). Among all the resources pointed out by RBV, knowledge is the most relevant to the competitiveness of companies (Villasalero, 2017). KBV shows that transforming tacit knowledge into explicit knowledge is the source of sustainable competitive advantage (Yang, 2008, 2012; Jawad and Mustafa, 2019). Technology Innovation contributes to knowledge management (Kamara et al., 2002), which leads to the explicit expression of accumulated knowledge. Therefore, innovation is a key factor for companies to seize the competition and promote corporate financial potential (Patterson, 1998; Veland and Shqipe, 2011; VanderPal, 2015; Zhang et al., 2018). Based on the analysis above, this paper proposes hypothesis 1:

Due to the reduction of natural resources and the intensification of global warming issues, social and corporate stakeholders are paying more and more attention to corporate environmental and social sustainability (Albort-Morant et al., 2018; Davenport et al., 2019), which is different from only focusing on financial growth. In the face of this revolution, green technology innovation characterized by minimizing impacts on the environment by conserving energy and resources (Lee and Kim, 2011), providing companies with the opportunity to use win-win logic to increase innovation to improve competitiveness (Porter and Van der Linde, 1995; Kong et al., 2016; Ishak et al., 2017; Li et al., 2019). Because producing green products, low resource consumption, high cleanliness, etc. are the characteristics of the renewable energy company, the technology innovation of the renewable company is consistent with the meaning of green technology innovation, which makes positive contributions to the realization of environmental and social sustainable development goals (Xie et al., 2019). So, to sum up, this paper proposes hypothesis 2:

This paper analyzes corporate sustainability from two dimensions: FS and SES. Firstly, for the FS, McKelvie and Wiklund (2010) indicates that companies with different types and companies in different periods show different growth characteristics. Higgins (1981) points out that FS reflects a company’s ability to use resources to obtain income. Tibor et al. (2015) and Peng and Liu (2018) define income as entrepreneurial companies’ growth capacities. However, revenue only reflects companies’ financial performance, not show the dynamic growth of the companies. Pandey (1994) and Zhao and Wijewardanab (2012) explained that in the long run, the company’s growth is an increase in company size and activities. Growth means that the company’s expansion activities involve sales, profits, and assets. Xu et al. (2020) point out that the FS is the maximum growth rate of operating revenue that can be achieved, and the average sales growth is a recurring indicator of performance because it discloses market acceptance and technical quality (Marino and De Noble, 1997; Utsch and Rauch, 2000), which is also regarded as an important parameter of corporate competitiveness (Bobillo et al., 2006; Ciro et al., 2020). Therefore, this paper selects the growth rate of operating revenue as an indicator of FS.

Sustainability needs to meet multiple aspects simultaneously (Elkington, 1998); FS alone cannot represent corporate sustainability (Gladwin et al., 1995). According to the “Triple bottom line” concept, society and environment are also two important parts of sustainability. So, this paper chooses ESG rating to measure SES. ESG is an important standard for measuring environmental protection and social responsibility of companies. The environmental rating takes into account waste of resources, green products, environmental violations, etc.; social rating is the concern for social contribution and external certification of the new energy company (Nirino et al., 2020). Nollet et al. (2016) believes that ESG rating is one of the best parameters for measuring the environmental, social, and business impact of the companies.

This paper selects Hua Zheng ESG Rating Index in the Wind database to measure the SES. Referring to the study of Zhang and Zhao (2019), we quantify the companies’ ESG performance as 0, if its ESG rating is BBB and below, and quantify the remaining companies’ ESG performance as 1.

Previous studies have used the intensity of R&D investment (Cumming et al., 2016), all intangible assets and capability (Sung, 2019), the proportion of technical staff, design or research, the proportion of sales or profits of new products (Chouaibi, 2020), and the number of patent applications (He et al., 2018; Plank and Doblinger, 2018) to measure technology innovation. It is a kind of strategic behavior that companies increase innovation inputs to raise profit opportunities (Clausen, 2009), and R&D investment can reflect companies’ efforts in promoting technology innovation and companies’ innovation ability (Hagedoorn and Cloodt, 2003). So, referring to the study of Xu et al. (2020), this paper uses the proportion of R&D expenditure to operating income to measure technology innovation.

Government subsidies are often measured by the natural logarithm of the total amount of government subsidies (Wu and Hu, 2020) or public expenditure for encouraging R&D and protection activities for industrial intellectual property (Sung, 2019). Hu et al. (2019) use the government subsidy amount reported under the non-operating income on the Statement of Financial Performance and divide it by the total assets as the indicator. Zhang and Guan (2021) choose the government subsidy amount in the annual report data of listed companies as the indicator. Combining the latest China Accounting Standards for government subsidies in 2017 and the study of Song et al. (2020), this paper uses government subsidies amount in deferred income and current income to measure government subsidies.

This paper selects five control variables based on previous studies: company size, equity concentration, return on total assets (ROA), company age, and capital structure.

The natural logarithm of revenue is used to measure the size of the company. The scale of a company is directly related to finances, and it is usually reflected in total assets, fixed assets, income, and the number of employees (Ahmedova, 2015). Considering that most of the renewable energy companies are in the start-up stage, revenue can more fully reflect the size of the company, so this paper chooses natural logarithm of revenue.

The capital structure is reflected by the debt-to-asset ratio. An increase of debt will aggravate financial risk and worsen capital constraints (Yu et al., 2020), and innovation is a risky activity with a long-lasting cycle (Li et al., 2020), so this paper chooses debt-to-asset ratio to measure the company’s risk situation.

ROA refers to the profitability of a company on its total assets. Return on total assets is usually defined as an accounting measure of performance (Waddock and Graves, 1997), which can be used to measure the company’s profitability. The most basic requirement of corporate sustainability is continuous profitability, so this paper chooses ROA as one of the control variables.

The H index is used to measure the concentration of equity. The share ratio of major shareholders may affect the way managers use government subsidies. Equity diversification can strengthen management supervision and restrain rent-seeking behavior (Yu et al., 2020). Equity concentration may lead to insufficient technology innovation activities because of “moral hazard” (Lin and Luan, 2020).

The company age. Young companies have additional resource constraints than old companies (Sung, 2019), and older companies have more time to accumulate knowledge and experience to support innovation (Li et al., 2020), so the company age is also an important control variable.

According to the study of Hu et al. (2019) and Yang et al. (2019), this paper constructs the regression model (1) as below to test the impact of technology innovation on FS. Referring to Ben-Akiva and Lerman (1985), this paper constructs the regression model (2) to test the effect of technology innovation on SES:

where i is the firm and t is the year; μ_i is the unobservable firm-specific effect and η_t refers to the unobservable time-specific effect; and β is the coefficients of the effect of technology innovation on FS and SES. Size, Top1, ROA, Lev, and Age are included in control variables, which represent the heterogeneity of companies. ε_i,t refers to the random error term.

Referring to Qing et al. (2021), existing studies typically add a quadratic term to the model or consider Hansen’s static threshold model. Hansen first proposed the threshold model (Hansen, 1999), which was used to describe the characteristics of skip or structural break as for the correlation between different variables. Seo and Shin (2016) proposed the first-difference GMM estimation of the dynamic panel threshold model. On this basis, this paper constructs the regression model (3) to test the threshold effect of government subsidies.

Hansen believes that the explanatory variable or the independent threshold variable can be set as the threshold variable. Therefore, when one of the threshold variables is selected, the other threshold variables could still be used as control variables. So, this paper selects government subsidies as the threshold.

Referring to Hong et al. (2019), Liu and Tsaur (2020), and Yu et al. (2020), this paper adds the interaction between government subsidies and technology innovation to construct the regression model (4) to test the moderating effect of government subsidies:

c₁ denotes threshold values of the government subsidies. β₁, β₂ are the coefficients of the effect of technology innovation on FS under the single threshold effect of government subsidies. β₃ indicates the impact of government subsidies on the effects of technology innovation on SES.

The result of the Hausman test (P < 0.0000) shows that the fixed effects regression model is suitable. The regression analysis results are as follows:

From the regression results of model (1) in Table 5, it can be seen that there is a significant positive relationship between Inn_L and FS (β = 6.578, p < 0.01), which means that technology innovation will promote the company’s FS. Therefore, hypothesis 1 is accepted. The reason may be that, on the one hand, the pioneer advantage gained by companies through technology innovation can help the company respond to rapid changes and the heterogeneity of markets (Chouaibi, 2020; Ma et al., 2021); on the other hand, technology innovation can provide organizations with continuous learning ability (Cheng et al., 2014).

Since ESG indicators are dummy variables, this paper uses the logit model to perform regression analysis. According to the regression results of model (2) in Table 5, Inn_L can promote corporate SES, so hypothesis 2 is accepted. It proves that renewable energy companies’ innovation has the characteristics of green technology innovation, which minimizes impacts on the environment by conserving energy and resources (Lee and Kim, 2011), is beneficial to reduce environmental impacts during a product’s life cycle (Christensen, 2011), and improves processes to reduce adverse environmental impacts. Renewable energy companies’ innovation brings them differentiation advantages (Cheng et al., 2014) and is beneficial to their SES.

According to the regression results of models (1) and (2) in Table 8, the coefficients of Inn_L in the two groups are 6.205 and 9.936, which are both significant at the 1% level; that is, Inn_L has a positive effect on FS. However, empirical p-value is 0.1730, suggesting that there is no significant difference in the coefficient of Inn_L between the two groups, which means that there is no difference in the promotion of technology innovation on FS between the east and the C&W.

The conclusion that the technology innovation of companies in the east and the C&W promotes the FS is consistent with the conclusion of Pandit et al. (2011), and their result shows that activities related to innovation have a positive impact on profits, while controlling the company’s leverage and its scale. A possible reason is that the technology innovation of companies can bring about a decrease in production and management costs, thereby increasing corporate profits and improving corporate financial performance. More importantly, Zhu et al. (2014) found that the accumulated technological progress will enhance the core competitiveness of the company, and external investors will be full of confidence in the future development of the company, so that the value of the company will continue to increase.

From the regression results of models (3) and (4) in Table 8, it can be seen that there is a significant positive relationship between Inn_L and SES (β = 10.547, p < 0.05) in the east. However, Inn_L has no significant effect in the C&W. The empirical p-value is 0.004, suggesting that there is a significant difference in the coefficient of Inn_L between the two groups, which means that there is difference in the relationship of technology innovation and SES between the east and C&W.

A possible reason is that there are differences in the regional environmental and social development between the east and the C&W. Yang and Wen (2017) found that the green development efficiency of the east and the C&W presents a serious polarization pattern. Chen and Wang (2018) believed that due to population migration, the C&W are aging faster than the east. The study of Hu et al. (2018) found that the average urban land use efficiency of China is higher in the east than in the C&W under the concept of green development. The quality of industrial development between the east and the west has the biggest difference (Deng and Liu, 2021). The east is no longer in the stage of relying on pollutant emissions to promote economic growth, while the increase in population and resource consumption in the central region leads to more pollutant emissions, and the west region is weak in environmental protection (He and Ran, 2009). With the same technology innovation resources, the contribution value of the economic indicators such as “GDP,” “GDP per capita,” and “industrial added value” in the northwest is lower than that in the east (Zhu and Zhang, 2005).

Due to the imbalance of regional social and environmental development, there are corresponding differences between eastern and C&W companies in the performance of social responsibility and innovation capabilities. Firstly, there are different degrees of emphasis placed on environmental and social responsibility. The eastern companies have a deeper understanding of the concept of social responsibility and related standards (Deng and He, 2013). Second, there are differences in the practice of company social responsibility, which means that eastern companies are at the highest level in terms of social responsibility for shareholders, employees, and environmental resources and the construction of a social responsibility management system, while company social responsibility in the northwest is the lowest level in China (Dong and Liu, 2017). Third, there are differences in the company’s innovation capabilities. There are obvious regional differences in the distribution of corporate innovation capabilities, and it shows the same spatial pattern as China’s regional economic development (Wang and Gao, 2017). Thus, there is obvious regional heterogeneity in the impact of company technology innovation on SES. Compared with the C&W, the technology innovation of eastern companies has played a significant role in promoting SES.

According to the regression results of model (2) and model (4) in Table 9, GSB inhabit the positive relationship between technology innovation and corporate sustainability.

The reason may be that the renewable energy industry is a technology-intensive industry; if companies want to gain a competitive advantage, they need to continue to carry out technology innovation. However, technology innovation requires a large amount of capital investment and has a high degree of uncertainty and spillover effects, which reduces the motivation for corporate innovation to a certain extent. The government needs to give certain subsidies and support before the start of the innovation project, that is, GSB. Because GSB are received before the investment in the project, they created a larger space for rent-seeking (Lee et al., 2014; Du and Mickiewicz, 2016; Wu and Hu, 2020). Therefore, when the threshold for subsidies is low, there are many companies attracted to squeeze into the renewable energy industry, especially large traditional energy companies (Peng and Liu, 2018). These large companies are more likely to receive subsidies due to their political connections, making start-ups that need subsidies for technology innovation unable to receive financial support. On the one hand, the intervention of several traditional energy companies has increased the intensity of competition in the renewable energy industry, and on the other hand, they have distorted the allocation efficiency of subsidies. Under the dual pressures of finance and excessive competition, the innovative motivation of renewable energy companies has weakened, which inhibits corporate sustainability.

According to the regression results of the threshold effect test in Table 10, only a single threshold model is significant at the 1% level for GSA, while GSB has no threshold effect. The estimated threshold for GSA is 0.4907, and the 95% confidence interval for threshold values is [0.4561, 0.4974]. In Table 8, When GSA exceed the threshold, for every 1% increase in technology innovation, the coefficient of financial growth decreases from 7.156 to 2.94%. The regression coefficient of the cross term between GSA and technology innovation is negative in model (3). These indicate that GSA are the main reason for the threshold effect of government subsidies and have a restraining effect on the relationship between technology innovation and corporate sustainability.

GSA aim to make up for the cost of completed projects, but they alleviate the promotion of technology innovation on corporate sustainability. Our results are different from the opinions of some scholars. Some scholars believe that GSA will be more effective than GSB (Peng and Liu, 2018), because GSA have a smaller rent-seeking space. This paper considers that although GSA are intended to help companies to expand market influence, if the actual production and sales capacity of renewable energy companies is not taken into account when formulating subsidy policies, only making the number of driving permits, registration certificates, and license plates reported as a basis for subsidies gives companies room to cheat for subsidies. Taking new energy vehicles as an example, 2013–2015 is a golden period for China to vigorously develop the new energy vehicle industry. The scope and amount of subsidies continue to expand, which provides opportunities for companies to cheat. For example, on September 8, 2016, the Ministry of Finance revealed that five new energy vehicle manufacturers, including Suzhou Jimxi Bus Manufacturing Co., Ltd., intended to defraud the government’s subsidies for more than one billion yuan. These companies used licenses without cars, cars without electricity, inconsistent signs, related parties and dealers idle, end-users idle, and other means to defraud government subsidies by inflating production quantity. On the one hand, cheating for subsidies reduces the efficiency of subsidies. Innovative companies that need subsidies fail to obtain funds, which undermines the companies’ enthusiasm for innovation. On the other hand, the actual sales situation is not taken into account in making the subsidy standards, falsely reporting their market influence through idle production after production or through expanding production blandly. GSA only play a role in mobilizing the companies’ enthusiasm in production, not promoting the enthusiasm for innovation, but have a negative impact on corporate sustainability.