1. Introduction
With the aggravation of global environmental pollution, oil crisis and other problems, the NEV industry with low dependence on oil and low pollution is highly valued by the governments of all countries [
1]. Currently, more than 20 percent of global carbon emissions come from the world’s transport industry. In China, where oil consumption in the transportation sector accounts for 70 percent of total domestic consumption, the issue of energy security is becoming more prominent. Different from the traditional automobile industry, NEVs mainly use batteries as power to replace the traditional internal-combustion engine, so as to reduce emissions and greatly relieve the dependence on oil [
2]. The NEV industry has not only become the core of the energy transformation and upgrading of the global automobile industry, but also has a close relationship with energy reform and the promotion of sustainable development [
3]. At the 75th Session of the United Nations General Assembly, Xi Jinping declared China’s goal of “peak carbon dioxide emission and carbon neutrality” to the international community. As the world’s largest automobile production and consumer, China is in the process of moving from “Made in China” to “Intelligent Manufacturing in China”. Therefore, nurturing and developing the NEV industry is a strategic option to alleviate the dual pressure of energy and the environment, and achieve sustainable development. This is of great significance to promoting China’s energy transformation and realizing the goal of “carbon peaking and carbon neutrality”.
As a material part of the global energy transformation, it is both an opportunity and challenge for China to realize the sustainable development of NEVs [
4]. To conform to the development trend of the automobile industry, the Chinese government actively promotes the application, infrastructure construction and research of NEVs [
1]. As Build Your Dream (BYD), Shanghai Automotive Industry Corporation (SAIC) and other traditional automobile enterprises have begun to transform to the production and research of NEVs, this industry has entered a new stage of rapid development. The annual production and sales scale of automobiles and number of power lithium battery assemblies ranks first in the world [
5]. However, there is still a gap between the “three power system” and leading international level. It is difficult to break through the core technical barriers [
6], low return rate of R&D investment, and narrowing of the profit space of enterprises, which are still the critical factors restricting industrial development. Facing an increasingly volatile market environment, TI has become a significant means for enterprises to participate in the market competition [
7,
8,
9].
As a typical technology-intensive industry [
10], NEVs are characterized by technical complexity and market uncertainty [
11,
12], and it is difficult for enterprises to complete technical upgrading and progress alone [
6,
13]. At this time, the innovative network of resource sharing and risk dispersion has become efficient alongside the development of innovative organization forms for the industry of NEV to reduce the cost of technology development, overcome the core technical difficulties, and narrow the gap with developed countries [
11,
14,
15]. It is also a crucial means to realize the complementary advantages of enterprises, enhance industrial competitiveness, and achieve extraordinary development. With the gradual maturity of industry technologies, the market type of industry has gradually shifted from “policy-oriented” to “market-oriented”, and the challenges facing enterprises are more severe. How to make better use of the innovative network, transform the external knowledge into core advantages [
16,
17], and realize leapfrog development of the industry are significant issues faced by NEV enterprises.
At present, research using patent data for TI has become a mature analysis method [
14,
18]. The research in the field of NEV, especially the evolution of the innovative network and influence on the TI of the enterprise is still inadequate [
19]. First, there are the limitations of the data studied. Most of the existing studies focus on the joint patent application between a certain enterprise or single type of institutions, with limited data evidence and research objects [
20]. In this way, the characteristics of innovative networks cannot be explored from the perspective of enterprises and universities. Second, the inadequacy of the systematic research perspective. Existing studies only analyze a particular technique or type of vehicle, and less analysis is done on the overall patent situation of the industry. The conclusions are not highly applicable to the industry as a whole. Third, the deficiency of the methodology of the study. Most of the existing studies focus on the macro comparison of patent data between countries and statistical analysis of industry development [
21]. These research methods are limited to descriptive statistics based on patents, and there are few systematic mechanistic studies through social networks and empirical analysis methods. For technology-intensive industries, however, macro-level analysis is far from sufficient. It is necessary to focus on the enterprises and study the evolution of innovative networks and the influence of NEVs.
Based on the above theoretical and realistic background, this study follows the logical framework of “Patent analysis—Network evolution—Empirical research”, and attempts to make beneficial explorations in the following three aspects: (1) Expansion of research data: This study obtains technical high-frequency words in this field based on the LDA model; and determines a patent search formula combined with IPC to obtain the patent data of China’s NEVs from 2001 to 2022. Additionally, by systematically cleaning the data, the joint patent application data of this field is obtained; and the patent situation of this industry is visually displayed. (2) Innovation of research perspective: Different from the previous single macro comparison and statistical analysis of patent data, this study applied the theory of industrial life-cycle to divide the development of China’s NEV industry into three stages. Moreover; the innovative networks of multi-type participants of NEVs are constructed separately. We systematically analyze the characteristics and evolution of innovative networks from the three aspects of network structure, key participants, and network content [
22]. (3) Innovation of research methods: This study breaks the fixed paradigm of “structure—effect” in the field of network and integrates network structure and network content into an analytical framework. The impact of the enterprise innovative network on TI is analyzed through social networks and empirical analysis. Additionally, we also conduct heterogeneity checks based on the differences in ownership, type, and region of the enterprise to improve the research of the relationship between the innovative networks and TI of NEV enterprises. These conclusions provide theoretical and practical guidance for the NEV enterprises to improve their TI and achieve technological catch-up on the curve, overtaking competition by leveraging innovative networks.
The composition of this study is as follows.
Section 2 is devoted to the analysis of the network evolution.
Section 3 and
Section 4 introduce the assumptions and study design.
Section 5 contains the empirical research, robustness tests, and heterogeneity checks.
Section 6 discusses the research conclusions, reveals theoretical and practical implications, and presents limitations and future research.
5. Empirical Research
5.1. Descriptive Evidence
Table 3 shows the basic description of the variables and the correlation coefficients for which the variables are reasonable. Meanwhile, the variables were tested for the variance inflation factor and the maximum VIF was 4.81, indicating that there was no multivariate linearity among the variables. The empirical tests in this study were analyzed using Sata17.0.
5.2. Testing Hypotheses
The regression results are shown in
Table 4. Model 1 is the reference model with control variables added. Model 2 is added to the cooperation breadth and its square term on the basis of Model 1. According to Model 2, the coefficient of the cooperation breadth is significantly positive and the coefficient of the quadratic term is significantly negative. Therefore, there is an inverted U-shaped relationship between the cooperation breadth of enterprises and TI, which has been verified in H
1. Model 3 is added to the structural hole on the basis of Model 1. As can be seen from model 3 (
p < 0.01,
β = 3.852), the structural holes occupied by enterprises have a significant positive impact on TI, and H
2 is verified. Meanwhile, Model 4 and Model 5 are added to the knowledge diversity and technical value on the basis of Model 1, and H
3 (
p < 0.01,
β = 0.044) and H
4 (
p < 0.01,
β = 0.304) are verified. The results show that the knowledge diversity and technical value of enterprises play a positive role in promoting TI. Model 6 is a complete model, that is, all variables are included.
5.3. Robustness Tests
To further validate the regression results, the key variables are replaced. The TI of NEV enterprises in China is measured from multiple perspectives.
First, to avoid endogenous problems due to measurement errors, the number of licensed patents is replaced by the number of cited patents. The regression results are shown in
Table 5, and the conclusions further support the robustness of the article. Second, because the patents from the application to the authorization also need to go through acceptance, preliminary examination, public, actual trial, and other processes. Generally speaking, it is announced 18 months from acceptance, and about 3 years to obtain authorization (or even longer) [
24].To avoid the influence of the network hysteresis in this study, the number of patent applications is delayed for two cycles. Subsequently, the lagged variables are put into the model for re-estimation, effectively avoiding the endogeneity problem. The results show that the regression results are still significant. These conclusions further verify that the study does not suffer from the network problem caused by reverse causality.
5.4. Heterogeneity Checks
With the empirical analysis presented above, we have fully tested the null hypothesis. A more material question, however, is which companies are better promoted by innovative networks. Heterogeneity analyses are required to identify the sources of heterogeneity in our findings.
5.4.1. Heterogeneity in Ownership
At present, there are two developmental patterns for the innovative cooperation of NEV in China: private enterprises as the core and state-owned enterprises as the core [
56]. Although these automobile enterprises entered the industry relatively late, they have developed at a fast pace, relying on abundant capital and policy support. On the contrary, private enterprises, which are sensitive to the market and flexible in their strategic adjustments, also have certain advantages in terms of TI. Therefore, we distinguished two sub-samples of state (21.86%) and non-state enterprises (78.14%) to examine the impacts of innovative networks on different ownership enterprises.
Table 6 shows that the results are in general agreement with the conclusions of the original model, but the coefficient of regression for non-state-owned enterprises is larger and more significant than that for state-owned enterprises. This suggests that the influence of innovative networks on the technical innovation of non-state-owned enterprises is more significant. The reason is that: compared with state-owned enterprises, non-state-owned enterprises face greater market competition. They are in urgent need of sharing the risks of technical innovation with their partners and improving technical innovation through the innovative network.
5.4.2. Heterogeneity in Type
The type of enterprise is further subdivided. Some are traditional automobile manufacturers, such as: BYD, Geely, and others. These enterprises have long-term technical experience, occupy certain automobile markets, and can update their product lines in a timely manner according to the market demand. There is also a new breed of emerging automotive companies, represented by NIO, Ideal, and others. Businesses in this category have updated their offerings with an Internet-minded focus on smart control experiences. Its products and services are popular with younger customers. Therefore, according to the division of the development stage of NEV, this study takes 2015 as the time point to distinguish the two types of NEV enterprises, and divides the sample into traditional (53.87%) and emerging automobile enterprises (46.13%), to investigate the impact of the innovative network on different types of enterprises.
As can be seen from
Table 7, the function of innovative network characteristics is more prominent for emerging automobile companies. The reason is: traditional automobile enterprises have strong industrial foundations and technical reserves, and mastered relatively mature R&D resources. They have competitive advantages such as market drive and technology leadership in the transformation from traditional (fuel vehicles) to NEV. Otherwise, emerging enterprises, as new entrants to the industry, need to continuously enhance technical cooperation and academic drive. They need to rely on innovative networks to realize the transfer and sharing of critical technologies and core resources within the industry, improve the efficiency of R&D and the quality of innovative products, and seize market opportunities. Finally, they can find novel growth points and secure a place in the fast-growing industry competition.
5.4.3. Heterogeneity in Region
According to the data disclosed by the China Association of Automobile Manufacturers, there are more NEV enterprises in Guangdong, Zhejiang, and Jiangsu provinces. The reasons are as follows: (1) The development level and policy implementation effect of enterprises in different geographical locations are different. For example: Guangdong province, on China’s southeast coast, is adjacent to Hong Kong and Macao. As its representative city, Shenzhen is the central city of the “Guangdong-Hong Kong-Macao Greater Bay Area” and also the gateway city of China’s opening to the outside world. BYD, Waltmal, and other well-known enterprises in the field of NEV are located here. Therefore, the region has location, policy, and technical advantages that can contribute to the R&D and product renewal. (2) Due to technical limitations, the performance of power batteries in a low temperature environment will decrease significantly. This reduces the capacity of the battery, which in turn reduces the range of the car. If air conditioners are frequently used in low-temperature environments, the loss rate will be higher. To a large extent, it leads to the situation of “hot in the south and cold in the north” in the use of NEVs in China.
Therefore, according to the degree of economic development in each region, this study divides the enterprises into three sub-samples: the eastern region (77.91%), central region (13.77%) and western region (8.32%). An analysis of heterogeneity based on differences in enterprise location shows that the breadth of cooperation plays the most significant role in TI in the eastern and western regions (see
Table 8).
The reason is: (1) the eastern region has a relatively developed economy. The Pearl River Delta region and the Yangtze River Delta region have gathered more than 100 industrial parks with an annual industrial output value of more than 10 billion yuan, with good industrial foundations and technical reserves. At the same time, thousands of large enterprises such as SAIC and Geely are located here, making it an industrial hub for NEV. Therefore, the eastern region has a strong clustering effect, which can give full play to the dominance position of enterprises in TI. Compared with the status of the central and western regions, the industrial cooperation in the eastern region is relatively close, and the role of the breadth of cooperation is more significant. (2) the number of structural holes, knowledge diversity, and technical value play a more significant role in promoting the western region. In terms of financial support and technical assistance, the policies of Chongqing, Sichuan, Guangxi, and Shaanxi provinces support the development of the western region. Meanwhile, the regions also have famous domestic universities such as Xi’an Jiao Tong University and Chongqing University. To a certain extent, each participant can innovate and disseminate knowledge through the innovative network to realize complementary resources and transform scientific and technical achievements. In general, the eastern region has unique technical advantages, while the western region has strong policy support. In contrast to the central region, the innovative networks play a more pronounced role in TI in the eastern and western regions.
5.5. Discussion
According to the different joint applicants of patents, this study divides the cooperation types into three categories: “external cooperation, internal cooperation and industry-university-research cooperation” (see
Figure 10). Among them, the cooperation between enterprises and research institutions (include: universities and research institutions) accounts for a large proportion (43.73%), and there is no competitive relationship between the research institutions and enterprises. For enterprises, in the development process of frontier technology, research institutions are the most reliable cooperative partner [
57,
58]. Therefore, this study introduces the virtual variable “cooperation type”. If the enterprise has cooperation with a research institution, it is recorded as 1; otherwise, it is recorded as 0. According to the results, the industry-university-research cooperation plays a major role in promoting the TI of enterprises.
The reason is as follows: As a major location for knowledge innovation, research institutions have an advanced base of knowledge and innovative talent. In particular, the structural holes of Tsinghua University, Zhejiang University, Sun Yat-sen University, and other famous universities are at the forefront of the key participants in TI. It means that universities, as the “broker” of the network, can connect other different participants, and provide a path for the technology and knowledge exchange between participants with a large gap in innovative ability. They play a significant role in promoting innovative cooperation. Therefore, enterprises should strengthen cooperation with research institutions. With the support of the innovative network, each participant should realize complementary resources to rapidly improve TI.
At the same time, compared with external cooperation, internal cooperation is more common in the industry of NEV. The reason for this phenomenon is that: compared with the industry of traditional automobile manufacturing, the NEV industry, as an emerging technology, has the characteristics of “winner-takes-all”. A few enterprises occupy a core position in the network, and they have mastered all kinds of resources and knowledge needed for technical development to guide the direction of technical development. To maintain the core competitiveness of their own technologies, the behavior of the technical monopoly is relatively common. Most of the leading technologies are developed independently by these companies, and they do not collaborate with other companies.
6. Conclusions
6.1. Theoretical Contribution
According to the logical framework of “Patent Analysis—Network Evolution—Empirical Inspection”, this study extracts the keywords of technical hot-spot in the field of NEV through the LDA model, obtains the patent data of China’s NEV from 2001 to 2022, and visually shows the indicators of joint innovation within the industry and overall patent situation in this field. Furthermore, based on the perspective of multiple types of participants such as enterprises and research institutions, combined with the product life cycle theory, the development stage of the NEV enterprise is divided and identified, and the innovative network of the three stages under the integration of multiple participants of NEV is completely depicted. Subsequently, the evolution trends and characteristics of the innovative network of the NEV industry are systematically analyzed from three aspects: network structure, the key participants, and network content. The conclusion is that there are large differences in the network structure at different evolutionary stages. The position of core organizations is becoming increasingly prominent; the network presents diversified cooperation and development; and the joint patent applications among different types of organizations are becoming more and more significant. By virtue of their central position in the network, enterprises facilitate the flow and distribution of innovative elements such as knowledge, information and technology among nodes, and play a leading and organizational role in the agglomeration of industrial innovation. These conclusions could deepen the understanding of innovative networks in the management of NEV domains.
Most of the previous studies on NEV remained at the level of statistical analysis of patent data. They lack systematic and quantitative research of the NEV innovative network from a network perspective, which does not accurately characterize the impact of the innovative network. In this study, the characteristics of innovative networks are introduced into the discussion of TI. Based on the data of 1706 NEV enterprises, the influence of network characteristics on the TI of enterprises is deeply analyzed from the two dimensions of network structure and network content. The research found that: (1) Moderate cooperation breadth can significantly promote TI. However, when the cooperation breadth is too large, enterprises easily to fall into the relationship trap, which is not conducive to the benign development of TI. (2) The structural holes occupied by enterprises have a significant positive effect on TI. Enterprises in the structural hole position benefit from the differences in knowledge and technology held by different partners. By controlling the allocation of information and resources in the network, they can obtain innovative information earlier than other colleagues, and enhance the ability to acquire resources in the innovative network. (3) The knowledge diversity of NEV enterprises has a significant positive effect on TI, which can avoid enterprises from falling into the dilemma of rigid technical development, promote the knowledge flow and transfer within the network, and improve the efficiency of innovation. (4) The technical value of NEV enterprises has a significant positive effect on TI. The greater the patent value, the more likely the enterprise has the double insurance of “leading technology + expected profit”, which can reduce the complexity and uncertainty in the innovative process, attract more participants to join in, and promote its own TI. (5) Finally, this study deeply analyzes the results of the potential heterogeneity of different regions and enterprise types. The results show that, first, compared with state-owned enterprises, the innovative network plays a prominent role in the TI of non-state-owned enterprises. Second, the characteristics of innovative networks have a more prominent role for emerging automobile enterprises. Third, innovation networks have a more significant impact on the eastern and western regions due to the regional difference. To some extent, our research has enriched and improved the theory of technical innovation and social networks, broadened the research perspective of the industry of NEV, and provided theoretical inspiration for exploring the influence of innovative networks on enterprise TI.
6.2. Practical Implications
Based on the conclusions of the study, this study presents proposals from both government and enterprises.
From the perspective of government policy formulation and industrial layout: The global automobile industry is accelerating its integration with energy, transportation, information, and communications. The NEV industry will have greater development opportunities and wider scope for development. The government should encourage both horizontal and vertical cooperation. On the one hand, the horizontal cooperation refers to the cooperation between NEV enterprises. For industrial development, enhancing the size and strength of the innovation networks and strengthening the division of labor and cooperation among different participants are of significance to reduce TI and promote industrial cluster development. The government should provide innovative resources to enterprises, promote cross-linkage and cross-field cooperation among enterprises, and strengthen the construction of the innovative consortium and innovative consortiums for the NEV industry. On the other hand, vertical cooperation means cooperation between universities, research institutions, and NEV enterprises. As the main position of knowledge output and scientific and technical personnel training, universities have an advantage in absorbing and integrating the knowledge of different types of participants. The government should give full play to the advantages of scientific resources in universities and research institutes. Through the establishment of an innovative ecosystem with deep integration of industry, university, and research, the industry will carry out basic R&D, the transformation of achievements and scientific research personnel training, to help the high-quality development of the industry of NEV. Finally, the global NEV industry will be increasingly competitive as subsidies in China decline and the global industrial landscape changes. The Chinese government should give full play to the leading role of enterprises in the integrated innovation of industrial chains, make breakthroughs in generic industrial technologies and key technologies, and promote the upgrading of traditional industries. At the same time, we will explore cutting-edge technologies and develop future industries to seize a leading role in the commanding heights of global future industries.
The conclusions have significant theoretical and practical implications for promoting the growth and development of enterprises. The transformation of the traditional automobile industry into NEVs has become an unstoppable development trend due to non-renewable oil, the continuous deterioration of the environment and growing environmental awareness. Enterprises should strengthen cooperation with external organizations and maintain the openness of innovative networks. The choice of partners should not only be limited to within the group or in competition, but should also be moderate, with partnerships with research institutions, universities, and upstream and downstream enterprises in the industrial chain to lead and drive the industry’s integrated innovation and rapid development. At the same time, the enterprise should not only give full play to the “relationship advantage”, absorb the knowledge and resources needed for innovation, enhance its own information and control advantage, improve the transformation of the innovative achievements; but also avoid the “trap of relationships”, moderately expand the patent innovative network, avoid the homogenization of knowledge in the enterprise, and maintain its heterogeneous advantages.
6.3. Limitations and Future Research
Our study has the following limitations: (1) This study analyzes the influence of network characteristics on TI based on the network perspective. However, what factors affect the internal driving force of enterprise innovative cooperation, and how the internal and external factors affect the embedded behavior of enterprises, and then affect the role of network characteristics are all issues worthy of further discussion. (2) Patent data is used in our article. Although the TI of an enterprise can be reliably measured based on patent data to some extent, not all technical knowledge can be translated into patents. This study does not have access to external patent and innovation data other than the patent database. Thus, there are some limitations to the conclusions. (3) The TI of enterprises is closely related to the external environment. Under different economic environments and political systems, innovative networks will evolve in different ways and have differential impacts. However, the research perspective of this study is limited to the NEV industry in China, and a comparative analysis of samples from different countries is lacking.
Combined with the limitations of this study, key future work is presented as follows: On the one hand, enrich the research perspective. In the future work, we will further extend the research perspective to the global NEV industry, including special samples such as multinational companies and joint ventures in the research framework of the article. At the same time, based on the perspective of knowledge spillover and knowledge flow, this study will deepen the research and enlightenment on the development of the global NEV industry [
26]. On the other hand, the data sources are expanded. To better describe the sharing and flow of technology, knowledge, and other resources among organizations, in the subsequent research, we can obtain the cooperative relationship of participants through the combination of patent data, literature, annual reports, and other information of enterprises, and then confirm each other and enrich the research conclusions of the article.