Configurations Driving High Performance in Hydrogen Fuel Cell Vehicle Enterprises

Abstract

The hydrogen fuel cell vehicle (HFCV) market is growing rapidly, but technological limitations, high costs, and market constraints are hindering enterprise performance. Existing studies often analyze isolated factors, overlooking their configurational interactions. This study applies the Technology–Organization–Environment (TOE) framework and fuzzy-set Qualitative Comparative Analysis (fsQCA) to examine how R&D capability, human capital level, scale of enterprise, attention allocation, and government support shape high performance in 40 Chinese HFCV enterprises. The consistency of all antecedents does not exceed 0.9, indicating that high performance does not depend on any single factor. The sufficiency analysis identifies three effective configurations: technology-driven, internal–external synergy, and organization–policy-driven, with an overall solution consistency of 0.9206 and a coverage of 0.4167. Without adequate government support and human capital, achieving high performance in HFCV enterprises appears improbable. These findings reveal multiple pathways toward high performance and highlight the importance of condition combinations over isolated effects, offering theoretical and practical insights into sustainable development strategies for emerging green industries.

1. Introduction

Given the growing severity of climate issues caused by global fossil energy emissions, countries must urgently focus on developing clean energy [1]. Unlike conventional vehicles, hydrogen fuel cell vehicles (HFCVs) offer high energy efficiency and zero tailpipe emissions, making them ideal candidates for sustainable transportation systems [2]. In recent years, the global HFCV market has experienced steady development, particularly in countries such as China, Korea, Japan, the United States, and several European nations. Figure 1 illustrates the annual HFCV sales volume by country from 2022 to 2024, highlighting China’s rapid growth in this domain. In China, HFCV enterprises have entered a phase of rapid growth due to the joint efforts of the government and various enterprises, as shown in Figure 2. In addition to publicly listed companies like Beiqi Foton and the FAW Group, many private enterprises are also actively entering the green hydrogen and fuel cell vehicle (FCV) market. However, China’s HFCV enterprises face several challenges, including regional resource disparities, inadequate infrastructure, and a limited market scale [3]. As a result, even leading companies in the industry exhibit varying levels of development. As the largest consumer of hydrogen energy, the HFCV sector can accelerate the development of the entire hydrogen energy industry chain by establishing a large-scale industrial ecosystem. Consequently, it is crucial to identify the factors influencing HFCV enterprises and investigate avenues for enhancing their performance.

As representative actors in the emerging new energy sector, the performance of HFCV enterprises is shaped by the combined effects of internal mechanisms and external environmental factors. Internally, performance is affected by technological resources (e.g., R&D capacity) [4] and organizational attributes (e.g., firm size) [5], while externally, government policies, market dynamics, and societal pressures play critical roles [6]. Previous studies on the performance of new energy enterprises mainly focused on the impact of a single factor on outcome variables, neglecting multiple concurrent factors and paths that affect enterprise performance against the background of technological innovation. Existing studies often use econometric models or case studies focusing on net effects of single variables. However, these approaches tend to overlook the configurational nature of enterprise performance, where multiple conditions interact to produce outcomes. The Complex System Theory posits that industrial development is a system of interdependent elements, not merely a sum of parts [7,8]. Building on this concept, Qualitative Comparative Analysis (QCA) was introduced by the New Configurational School, viewing cases as configurations of features and focusing on synergies resulting from different factor combinations [9]. The fsQCA method is widely used in analyzing synergies in the renewable energy industry, as it is suitable for medium sample sizes. For example, Tran et al. [10] assessed key factors in electric vehicle (EV) deployment from a configurational perspective, emphasizing the interplay between technological scales and behaviors. Additionally, Yong and Park [11] applied fsQCA to analyze multiple factors influencing EV deployment, highlighting the impact of economic, social, and policy differences on EV penetration across countries. Building on this, Maqbool and Sudong [12] argue that the development of a new energy industry such as HFCVs involves systemic complexity, where the interaction and synergistic effects of various factors are often overlooked when analyzed in isolation. Therefore, similar to other renewable energy industries, we applied the fsQCA method developed by Ragin [9] to analyze synergies in the performance development of HFCV enterprises, overcoming the limitations of traditional regression analysis [8] and providing new insights into the combined effects of technological, organizational, and environmental factors.

To address the limitations of the resource-based view and institutional theory in analyzing the synergistic effects of internal and external factors, and in recognition of the critical role of technology in the development of strategic emerging industries, Tornatzky and Fleischer [13] proposed the Technology–Organization–Environment (TOE) framework based on innovation diffusion theory and the technology acceptance model. This framework has evolved into a multidimensional theoretical tool that integrates both internal and external environmental factors, offering an ideal analytical framework for the study of strategic emerging industries from a systematic and holistic perspective [14,15]. The TOE framework analyzes the impact of enterprise development at three levels: technology, organization, and environment. The technological level (T) focuses on the characteristics of the technology itself and related factors, such as technological complexity, innovation capacity, and human capital. The organizational level (O) focuses on the organization’s resources and characteristics, including factors like structure, capital input, and organizational size. The environmental level (E) examines the macro-environment of the organization, including the policy, market, and competitive environments.

This study aims to answer the following research question: What combinations of technological, organizational, and environmental factors drive high performance in Chinese HFCV enterprises? To address this, we applied the TOE framework and identified multiple complex factors influencing the performance improvement of HFCV enterprises, categorized into three dimensions: technology-driven, organization-driven, and environment-driven. We used fsQCA to analyze the configurations of 40 listed HFCV enterprises in China, examining how different factor combinations influence the enterprises’ high-performance development. The main contributions are threefold. First, this study shifts the focus from isolated factors to a configurational perspective, addressing the lack of research on performance pathways and resource allocation strategies at the enterprise level. Second, it extends the application of the TOE framework to green and complex industrial systems, offering new insights into how firm-level and contextual conditions interact to shape enterprise outcomes. Third, it enriches empirical understanding of China’ s hydrogen economy by highlighting the heterogeneity among HFCV enterprises. This study identifies the conditional configurations and mechanisms driving the high-performance development of HFCV enterprises, facilitating industry growth and contributing to the alleviation of the energy crisis and climate change.

2. Antecedent Conditions of HFCV Enterprise Performance Under the TOE Framework

2.1. Technical Dimension

According to Schumpeter’s innovation theory, continuous technological innovation is a key driver of economic growth [16,17] and significantly impacts enterprise performance development [18]. In the HFCV industry, technological innovation is both a source of competitive advantage and a gating constraint for commercialization due to the sector’s characteristic long R&D cycles and high capital intensity [19]. In the context of low carbon, the technical dimensions of the TOE framework for analyzing HFCV enterprise performance include R&D capability and human capital. R&D investment is broadly viewed as a direct measure of R&D capability. Empirical evidence from China’s new energy listed companies shows that R&D investment intensity has a significant positive correlation with enterprise performance and exhibits a lag effect, indicating that the benefits of technological inputs may take time to materialize [4]. This lag effect is expected to be even more pronounced for HFCV enterprises, given their typically longer R&D and commercialization cycles. However, R&D investment is costly and poses challenges to organizational stability and legitimacy [20,21]. Some scholars suggest that it can have a detrimental effect on enterprise development. Mazzelli and Nystrom argue that excessive R&D investment could notably detract from shareholder interests, potentially resulting in stock price declines and impacting the company’s overall value [21]. Additionally, R&D capability is closely tied to the human capital of the enterprise in the process of technological innovation [22]. In strategic emerging industries, the technology-intensive nature of enterprises amplifies their dependence on human capital [23]. According to human capital theory, the ongoing concentration of human capital facilitates resource sharing and reduces geographic barriers, consequently decreasing knowledge acquisition costs and boosting enterprise innovation capabilities [24].

In summary, an enterprise’s human capital and R&D capabilities are interdependent, jointly enhancing its competitiveness and productivity. Relevant findings from NEV enterprises—sharing similar high technology intensity, policy dependence, and market uncertainty—provide valuable analogies for understanding how R&D capabilities and human capital, as technological factors, impact HFCV enterprise performance. Given the complexity of enhancing HFCV enterprise performance, which necessitates significant technological breakthroughs, this study delves into how R&D capabilities and the level of human capital, as technological factors, impact HFCV enterprise performance.

2.2. Organizational Dimension

An enterprise’s competitiveness is influenced by the abundance, scarcity, and inherent value of its resources [25]. High performance of enterprises relies on the availability of irreplaceable resources within the organization, including human, material, financial, technological, and other support resources [26]. The organizational dimensions of the TOE framework for analyzing HFCV enterprise performance include firm size and attention allocation. Existing research suggests that the scale of enterprises is positively correlated with R&D investment and innovation performance [27,28]. However, an inverted U-shaped relationship has been observed, suggesting that while R&D investment and innovation performance increase with enterprise scale, they begin to decline after surpassing a specific threshold [29]. Empirical evidence from the new energy vehicle literature shows that firm size moderates how green/innovation investments translate into enterprise performance. Lin et al. (2019) found heterogeneous returns to green innovation across firm sizes in the automotive sector, implying that firm scale affects the efficiency and payoff of innovation expenditures [5]. As an intangible and cognitive resource, it distinguishes itself from a firm’s physical resources [30]. Bounded rationality theory says that managers have limited attention, and their distribution determines the strategic direction of the enterprise, thereby impacting economic and social performance [31]. He et al. (2021) show that managers reallocate attention toward R&D when facing performance shortfalls—effects that are especially relevant in high-tech, capital-intensive sectors [32].

Within the HFCV industry’s extended investment and R&D cycles, a larger enterprise scale promotes R&D and innovation, enhancing the enterprises’ fault tolerance. Moreover, managers need to direct their attention to crucial and potentially valuable areas for optimal decision-making, while short-sighted management risks lead companies astray. Hence, this study concentrates on examining the influence of the scale of an enterprise and attention allocation as organizational factors in HFCV enterprise performance.

2.3. Environmental Dimension

Organizational strategies and actions are often significantly shaped by their operating environment [33]. The environmental dimensions of the TOE framework for analyzing HFCV enterprise performance include the level of government support. Government support and subsidies within the policy environment foster innovation and boost enterprise performance [34,35]. A study on the green innovation performance of Chinese new energy vehicle enterprises reveals that government subsidies have a significant positive effect on enterprise innovation, but this impact varies depending on ownership type and regional location, with non-state-owned firms and those in economically developed regions being more sensitive to subsidy incentives [36]. However, the impact of government subsidies is not uniformly beneficial. When the cost of innovation or rent-seeking activities is substantially lower than government subsidies, enterprises might concentrate investments in subsidized areas to meet policy criteria, risking private capital displacement. This leads to a crowding-out effect, potentially hindering enterprise performance [37]. Given that the HFCV industry’s market is in its nascent stages, with a brief history of industrialization, it frequently encounters resource and capability limitations, complicating competition with other manufacturing sectors based solely on these factors. Thus, the policy environment’s guidance and support are vital for modulating market and economic conditions. Therefore, this study aims to examine the impact of government support as an environmental factor on HFCV enterprise performance.

Based on the TOE framework, five antecedent conditions were selected to capture the technological (T), organizational (O), and environmental (E) dimensions that influence HFCV enterprise performance. The selection reflects both theoretical underpinnings and the specific characteristics of the HFCV sector. Technologically, R&D capability and human capital level are essential given the industry’s long development cycles, high capital intensity, and dependence on skilled talent for breakthrough innovation. Organizationally, attention allocation and enterprise scale shape how firms mobilize and deploy scarce resources under uncertainty, directly affecting innovation efficiency and market responsiveness. Environmentally, government support plays a decisive role in shaping market formation and reducing entry barriers in strategic emerging industries like HFCVs. Together, these antecedents align with the multidimensional drivers of organizational performance under the TOE framework and have been repeatedly validated in analogous high-tech industries such as NEVs [10,11], ensuring their explanatory relevance in this context (see Figure 3).

3. Methods and Data

3.1. Research Methods

Qualitative Comparative Analysis (QCA), first introduced by Ragin in 1987, aims to identify causal relationships among multiple concurrent factors [9], offering deeper insights into the mechanisms that drive the performance differentiation of HFCV enterprises. The QCA method uses Boolean logic for formal operations [9]. In simple terms, for N antecedent conditions, each of which can have two states (e.g., yes or no), there are 2^N possible combinations of conditions. QCA assigns cases to these combinations based on their scores for each factor, identifying which combinations promote or inhibit a particular outcome. If a condition is required in all combinations to achieve the desired outcome, it is considered a necessary condition. A sufficient condition means that when condition A is present, the result must occur; however, A is not the sole condition—other combinations can also lead to the same result. Due to the limited sample size for HFCV enterprises, this study uses fsQCA, which is well-suited for medium sample sizes. This method retains truth value analysis while simplifying the processing of membership degrees and complex configurations, making it an ideal tool for exploring organizational strategy and innovation performance [38]. The detailed design process is provided in Section 4.

The reasons for selecting the fsQCA method are as follows: (1) The performance conditions of HFCV enterprises are synergistic and interdependent, and the fsQCA method offers unique advantages in addressing these complexities. (2) The sample enterprises are 40 HFCV enterprises, which represent a medium-sized sample, and the data scale aligns with the requirements of fsQCA. Furthermore, the ideal number of conditions for a medium-sized study (10−100 cases) typically ranges from 4 to 7 [39]. Five conditions were selected to meet the study’s requirements. (3) The fsQCA method facilitates the comparison and analysis of asymmetries influencing enterprise performance. In summary, this study employs the fsQCA method to analyze the configuration of antecedent conditions for achieving high performance in HFCV enterprises.

The first step of fsQCA involves calibrating raw data into fuzzy sets, transforming continuous or ordinal variables into values between 0 and 1, indicating degrees of membership [39]. This study employs the direct method of calibration, using three qualitative anchors: full membership (1), full non-membership (0), and the crossover point (0.5). The logistic function used to calculate fuzzy-set scores is defined as follows:

$$\mathrm{f}(\mathrm{x}) = {\displaystyle \frac{1}{1+e^{\left(- \frac{x-c}{s}\right)}}}$$

(1)

where x is the raw value, c is the crossover point, and s is the rate of change.

After calibration, a truth table is constructed listing all logically possible configurations of causal conditions. Next, we conduct necessity analysis, which examines whether certain conditions must be present for the outcome to occur. A condition is considered necessary when its consistency exceeds 0.9. The consistency of necessity is calculated as follows:

$$\mathit{C}\mathit{o}\mathit{n}\mathit{s}\mathit{i}\mathit{s}\mathit{t}\mathit{e}\mathit{n}\mathit{c}\mathit{y} ({\mathrm{X}}_{\mathrm{i}}\ge {\mathrm{Y}}_{\mathrm{i}})={\displaystyle \frac{\sum \left[\min ({\mathrm{X}}_{\mathrm{i}},{\mathrm{Y}}_{\mathrm{i}})\right]}{\sum \left({\mathrm{Y}}_{\mathrm{i}}\right)}}$$

(2)

$$\mathit{C}\mathit{o}\mathit{v}\mathit{e}\mathit{r}\mathit{a}\mathit{g}\mathit{e} ({\mathrm{X}}_{\mathrm{i}}\ge {\mathrm{Y}}_{\mathrm{i}})={\displaystyle \frac{\sum \left[\min ({\mathrm{X}}_{\mathrm{i}},{\mathrm{Y}}_{\mathrm{i}})\right]}{\sum \left({\mathit{X}}_{\mathrm{i}}\right)}}$$

(3)

where X_i represents the calibrated antecedent condition and Y_i represents the calibrated outcome condition for the unit.

The fourth step is sufficiency analysis, which identifies combinations of conditions that are sufficient to produce the outcome. A configuration is regarded as sufficient when it consistently leads to the outcome across cases. The formulas for sufficiency, consistency, and coverage are the following:

$$\mathit{C}\mathit{o}\mathit{n}\mathit{s}\mathit{i}\mathit{s}\mathit{t}\mathit{e}\mathit{n}\mathit{c}\mathit{y} ({\mathrm{X}}_{\mathrm{i}}\le {\mathrm{Y}}_{\mathrm{i}})={\displaystyle \frac{\sum \left[\min ({\mathrm{X}}_{\mathrm{i}},{\mathrm{Y}}_{\mathrm{i}})\right]}{\sum \left({\mathit{X}}_{\mathrm{i}}\right)}}$$

(4)

$$\mathit{C}\mathit{o}\mathit{v}\mathit{e}\mathit{r}\mathit{a}\mathit{g}\mathit{e} ({\mathrm{X}}_{\mathrm{i}}\le {\mathrm{Y}}_{\mathrm{i}})={\displaystyle \frac{\sum \left[\min ({\mathrm{X}}_{\mathrm{i}},{\mathrm{Y}}_{\mathrm{i}})\right]}{\sum \left({\mathit{Y}}_{\mathrm{i}}\right)}}$$

(5)

3.2. Data Construction

3.2.1. Sample Selection and Data Source

This study selects samples from the list of 100 leading enterprises in the hydrogen industry chain disclosed by Qixinbao in 2022, excluding ST and ST* companies, and ultimately selects 40 listed HFCV enterprises in China as case samples. The sample list is shown in Appendix A. The samples are processed as follows:

• Must have the 2021 company annual report with HFCV operating data.
• Exclude samples other than HFCV companies. The HFCV industry has the most obvious and extensive technological innovation in the hydrogen chain. The theme of this paper is the high performance of enterprises, so, for example, enterprises that produce hydrogen do not meet the research requirements.
• Exclude enterprises that do not use fuel cells, power systems, etc., as their main products. Some enterprises are group companies, and the related products are only a small branch of the enterprise layout, and their viability is unmatched by other enterprises in the sample.
• Exclude enterprises lacking sufficient information to support the study’s variables.

These 40 enterprises represent the core actors in China’s HFCV sector. Their diversity in ownership structure, regional distribution, and innovation strategies enhances the representativeness of the sample and supports cross-case configurational analysis. This ensures a meaningful exploration of systemic performance patterns across heterogeneous enterprise types.

Given the nascent stage of China’s hydrogen industry, enterprises are prioritizing technological advancement and market expansion. Numerous enterprises began publishing annual reports on the HFCV industry in 2021. Moreover, 40 samples adequately demonstrate the issue. To guarantee data novelty, this study designates 2021 as the research period to analyze how technology, organization, and environment collectively influence HFCV enterprise development. The case data are derived from the China Stock Market and Accounting Research Database (CSMAR) and company annual reports, while patent data are sourced from the China Research Data Service Platform (CNRDS).

In recent years, content analysis of annual reports has become an increasingly accepted method to evaluate firm performance through narrative disclosures and strategic representations [40,41,42]. Compared with single financial indicators, this approach captures both tangible and intangible aspects of firm performance, including innovation efforts, resource inputs, and market orientation, making it especially suitable for configurational analysis of strategic emerging industries like HFCVs.

3.2.2. Measurement of Variables

HFCV enterprise performance (EP). Past studies predominantly measured a single index, namely, total factor productivity. However, the concept of enterprise performance encompasses multifaceted dimensions [43,44] and should consider factors such as regional differences, periods, and the unique industry context and ownership structure of the enterprise. This necessitates time and geographical adaptation and even the consideration of conditions like industry characteristics and property rights. Therefore, to comprehensively capture the nature of HFCV enterprises’ performance, this study integrates relevant research by Haber and Reichel [45] with the specific attributes of HFCV enterprises and designs a scale comprising 19 items, encompassing three dimensions: economic value, social value, and enterprise capacity. Utilizing this scale seeks to reduce potential measurement biases and offer a more holistic evaluation of HFCV enterprises’ performance, as shown in Appendix B.

This study gathers the 2021 annual reports of 40 listed enterprises and incorporates them with relevant previous scholarly works [46]. To assess the high performance of the HFCV enterprise, a content analysis method was employed. Specifically, each measurement criterion was scored based on whether the enterprise’s annual report contained textual descriptions. A value of 1 was assigned if a text description was present, while a value of 0 was assigned otherwise. Subsequently, the scores for each measurement criterion were then summed to calculate the overall high-performance score. Finally, the average score was used as the indicator for the high-performance level of each enterprise.

R&D capability (RD). Previous research has determined that inventions and utility models exhibit the highest level of innovation when targeting core product functionalities [47]. Consequently, “the total number of invention patents and utility models by the end of 2021” is selected as an indicator of an enterprise’s R&D capability. To circumvent complications associated with zero values, the dataset undergoes a natural logarithm transformation after an increment of 1.

Human capital level (HC). Human capital levels are generally positively correlated with employees’ educational achievements [48]. Consequently, this study uses “the percentage of employees with at least a bachelor’s degree by the end of 2021” as an indicator of human capital levels in the sampled enterprises.

Attention allocation (AA). The timing of a company’s focus on hydrogen energy critically affects its success in developing a hydrogen energy division. Enterprises prioritizing hydrogen energy often demonstrate enhanced construction and development performance. According to Tan et al. [49], attention allocation was gauged by “the time interval between each enterprise’s announcement on expanding into the hydrogen energy sector and China’s 2014 official endorsement of ‘hydrogen energy and fuel cell’ as a strategic innovation direction in energy science and technology”. A longer interval suggests reduced attention by the enterprise toward the hydrogen energy sector.

Scale of enterprise (ES). Total assets are a key indicator of enterprise scale, effectively reflecting owned and controlled resources [50]. In measuring enterprise size, “total assets” by the end of 2021 has become the dominant metric. Hence, this study utilizes total assets (in hundreds of millions of Chinese Yuan) to gauge enterprise size.

Government support (GS). Government support levels are measured following [51]. A content analysis method is applied to quantify the “number of hydrogen energy policy clauses” issued by the provinces where the enterprises are based. The main information sources are the respective provinces’ government websites. Starting in 2011, China initiated the release of hydrogen energy support policies. To compile the dataset, provincial “policy documents” were searched on provincial government websites from 2011 to 2021, particularly under “government affairs disclosure.” We used keywords such as “hydrogen”, “strategic emerging industries”, “new energy”, “manufacturing” and “development” to identify relevant policies. As a result, 119 provincial policies from 18 provinces were selected. To facilitate content analysis of hydrogen energy policy texts based on policy instruments, we coded these texts in the format of “province—policy number—chapter—specific clause number” to form a coding unit table.

Table 1. Indicator descriptions.

Variable Type	Variable	Description of Indicators	Indicator Sources
Condition variable	RD	Total number of “invention patents and utility models granted as of the end of the reporting period plus 1 to take a logarithmic number”	CNRDS
	HC	Percentage of “employees with education level of bachelor’s degree or above”	CSMAR
	AA	“The time interval” between when companies published information on expanding hydrogen to 2014	Annual report of enterprises
	ES	Total “enterprise assets”	CSMAR
	GS	Number of “Hydrogen Energy Policy Provisions Issued in the Province”	Provincial government portals
Result variable	EP	“Frequency counts and averaging of words” related to corporate annual reports	Annual report of enterprises

shows the indicator descriptions.

3.3. Variable Calibration

Before conducting fsQCA, variables must be transformed into set concepts, specifically into fuzzy-set variables ranging from 0 to 1. Adhering to the calibration standards proposed by Du et al. [8], this study establishes calibration criteria for the high performance of HFCV enterprises, R&D capability, human capital level, scale of enterprise, and government support. These are assigned sub-points of 0.75, 0.5, and 0.25 as anchor points and are set as full out, crossover, and full in, respectively. The crossover point for attention allocation is set at 0.5, while full out and full in standards are set at 0.25 and 0.75, respectively. If the fuzzy-set membership score of an antecedent condition is exactly 0.5 after correction, replace it with 0.501 [9]. The calibration anchor points and descriptive statistics for each variable set are presented in

Table 2. Variable calibration and descriptive statistics.

Variable	Fuzzy-Set Calibration			Descriptive Statistics
	Full Out	Crossover	Full In	Minimum	Maximum	Mean	Standard Deviation
RD	5.6569	6.6307	8.0064	3.6636	9.7003	6.6923	1.4433
HC	0.1374	0.2572	0.3526	0.0176	0.6010	0.2572	0.1515
AA	31.0000	23.0000	10.0000	1.0000	40.0000	19.7500	11.8619
ES	61.0509	115.3969	488.3054	13.8436	9169.2270	540.8000	1505.2269
GS	16.0000	41.0000	45.0000	5.0000	66.0000	33.55500	19.0478
EP	7.7500	12.0000	14.2500	2.0000	19.0000	11.3000	4.2919

.

4. Results

4.1. Necessary Condition Analysis

In this study, fsQCA 3.0 software was used for data analysis. Before the configuration analysis, this study performed a necessity test on each antecedent, with the results detailed in

Table 3. Results of necessary condition analysis.

Variable	High Performance		Not-High Performance
	Consistency	Coverage	Consistency	Coverage
RD	0.5515	0.5591	0.4960	0.5131
~RD	0.5197	0.5027	0.5738	0.5662
HC	0.6707	0.6472	0.4431	0.4362
~HC	0.4157	0.4225	0.6416	0.6653
AA	0.6192	0.6135	0.4952	0.5005
~AA	0.4959	0.4905	0.6176	0.6233
ES	0.5783	0.5797	0.4823	0.4932
~ES	0.4944	0.4835	0.5891	0.5877
GS	0.6658	0.6642	0.4377	0.4456
~GS	0.4442	0.4365	0.6701	0.6716

Note: “~” = Negation (NOT). The following table is the same.

. The consistency of all antecedents does not exceed 0.9 [9], suggesting that no single condition fully explains the HFCV enterprises’ performance and thus that none of them is essential for either high or not-high performance. Consequently, HFCV enterprises should not solely focus on a specific dimension or antecedent but instead pursue diverse configurations that encompass advantages and disadvantages. Furthermore, substitution or complementary effects among different dimensions or antecedents could lead to “different pathways to the same destination” for achieving high performance.

4.2. Configuration Analysis

Drawing on prior studies, this study sets the original consistency threshold to 0.8 and the PRI consistency threshold to 0.75 [9,39]. The intermediate solution identifies core and peripheral conditions, complemented by the parsimonious solution. A core condition is an antecedent that is present in both the parsimonious and intermediate solutions, while a peripheral condition only appears in the intermediate solution [7]. This study identifies three-quarters of the driving pathways that elucidate the high/not-high performance of HFCV enterprises. The consistency and overall consistency levels for each configuration exceed 0.75, meeting the adequacy criteria.

4.2.1. Configuration Analysis of High Performance of Sample Enterprises

Table 4. Configuration analysis of high performance.

Condition Variables	Configuration of High Performance
	HLP1	HLP2	HLP3
RD	⬤	⭘	•
HC	⬤	⬤	⭘
AA	⭘	⬤	⬤
ES	•		⬤
GS		⬤	⬤
Case	FJF CSICP	BSH DPC	ATM SAIC
Consistency	0.8937	0.9482	0.8664
Raw coverage	0.1656	0.2405	0.0949
Unique coverage	0.1303	0.1819	0.0399
Overall solution consistency	0.9228
Overall solution coverage	0.4162

Note: “⬤” means that the core condition exists, “⭘” means that the core condition does not exist, “•” means that the auxiliary condition exists, “space” means that the condition can or cannot exist.

shows an overall consistency of 0.9206 for high performance, with a coverage of 0.4167, meaning the three configurations explain 92.06% of the high-performance variance and cover 41.67% of the high-performance instances in HFCV enterprises. Using the conditional configurations, distinct adaptation relationships that promote high performance in HFCV enterprises can be further identified. For a clearer presentation of the analysis results, the three pathways driving high performance in HFCV enterprises are visualized in a graph, depicted in Figure 4.

(1) Configuration 1 (HLP1) is defined by its technology-driven approach, characterized by high R&D capability, substantial human capital, and limited attention allocation. It shows that HFCV enterprises with limited attention allocation can still achieve high-performance industry growth through significant R&D capabilities, abundant human capital, and larger enterprise scale. This configuration represents 16.87% of high-performance cases in HFCV enterprises, with 13.08% of cases uniquely attributed to this configuration. In the technology-intensive HFCV industry, newcomers can gain a competitive edge by quickly achieving technological breakthroughs, as suggested by the dynamic capability-based enterprise capability model. As emphasized in “human capital theory”, human capital is a key driver for spurring technological innovation and improving R&D efficiency [52]. R&D capabilities supported by higher human capital facilitate structural optimization in high-tech manufacturing as human capital levels increase. However, reliance solely on technological R&D and talent is not enough to guarantee the growth of the capital-intensive HFCV industry. Additionally, substantial financial support and resources are crucial to enable further experimentation and innovation. Therefore, to offset the disadvantages of being a latecomer in the HFCV sector and support investment and expansion, enterprises need to fully leverage their R&D, human capital, and scale to persist in technological innovation and product advancement, aligning with market demands for high performance. Noteworthy cases of this configuration include FAW Jiefang (FJF), China Shipbuilding Industry Group Power (CSICP), and the Weifu High-Technology Group (WHTG).

Considering FJF as an example, this high-tech enterprise boasts total assets of 69.7 billion of Chinese Yuan, securing the second-highest rank among the sampled companies. Despite the “Qiyi Dengfeng Plan”, a new energy technology strategy establishing HFCV development as a strategic direction, not being released until July 2019, FJF had already invested more than 20 billion in R&D during the “14th Five-Year Plan” period. This investment was directed towards creating a global new energy innovation base in Changchun and an exclusive hub for FCVs and hydrogen energy industrial clusters, cumulatively exceeding 10 billion. Supporting technological R&D and innovation, FJF focuses on talent development, assembling an efficient and collaborative R&D team of nearly 2500 members. On 17 June 2022, FJF achieved a significant milestone by deploying 300 HFCVs from Changchun to Beijing, Shanghai, and Shanxi, demonstrating their advancement in hydrogen fuel cell technology and application. These developments underscore the crucial role of synergy among R&D capability, human capital, and enterprise scale in propelling FJF’s high performance.

(2) Configuration 2 (HLP2) is internal–external synergy, with core conditions including high human capital levels, elevated attention allocation, strong government support, and low R&D capability. This configuration shows that enterprises with an early strategic focus on the HFCV sector but limited R&D capabilities depend on specific conditions for high performance. First, significant government support is needed to adequately nurture local enterprises. Secondly, increasing investment in human capital is crucial for quickly overcoming technical shortcomings and achieving high performance in HFCV enterprises. This configuration represents 24.04% of high-performance cases in HFCV enterprises, with about 18.03% uniquely attributed to it. As the HFCV industry is still emerging, pioneering enterprises with limited R&D capacity must adopt specific strategies for high performance. Early-stage government subsidies are vital to stimulate market demand, encourage technology and process improvements, and create a virtuous cycle addressing challenges like high costs and technological bottlenecks. Enterprises should overcome their limitations by improving their human capital and utilizing high-quality employees to enhance R&D capabilities, thus strengthening market competitiveness. To further reinforce their first-mover advantage, enterprises must synergize human capital and government support, as illustrated by notable cases such as Beijing Sinohytec (BST), Beijing Dynamic Power (BDP), Zhongtong Bus Holding (ZBH), and others.

Take BST for example. Located in Beijing, a city chosen for FCV demonstration projects, BST is committed to delivering high-power fuel cell system products. The company has placed a strong emphasis on R&D talent acquisition, assembling a team of professionals with maturity and experience. This team’s independent development of advanced technologies in hydrogen fuel cell systems has afforded BST a considerable talent advantage. Following the national hydrogen energy support policies issued in 2011, BST initiated its hydrogen energy projects in 2012. They have developed the HD series hydrogen fuel cell engine and prioritized hydrogen fuel cell technology as their main business focus. Since August 2021, the government endorsement of five city clusters, including Beijing, Shanghai, Guangdong, Zhengzhou, and Zhangjiakou, for HFCV demonstrations has furthered BST’s advantage. The company’s client base primarily resides within these clusters, enjoying significant government financial support. BST has positioned itself as a leader in China’s hydrogen fuel cell system industry, being among the few companies capable of mass-producing fuel cell systems and demonstrating commendable high performance.

(3) Configuration 3 (H3) is characterized as organization–policy-driven. This pathway necessitates significant attention allocation, substantial scale of enterprise, strong government support, and limited human capital. It indicates that without enough human capital, high performance in HFCV enterprises depends on focused attention, large enterprise scale, extensive government support, and robust R&D capabilities. This configuration accounts for 9.64% of the cases of high performance in HFCV enterprises, with approximately 3.89% uniquely attributable to it. The HFCV industry presents high technological thresholds and market uncertainties, which often lead to a shortage of human capital. To enhance productivity effectively, enterprises must gradually expand to bridge the talent gap. Scaling up enterprises enables access to more resources and market share while providing a conducive developmental environment for employees. Yet, emerging HFCV enterprises face challenges like technical hurdles and limited funding during expansion efforts. Consequently, the government’s provision of financial support and infrastructure protection becomes crucial. With strong internal R&D capabilities, enterprises can attract and cultivate top-quality talent. The synergy among the three types of factors drives the industrialization and marketization of HFCVs, fostering the high performance of enterprises. Noteworthy examples of this configuration include Advanced Technology and Materials (ATM) and SAIC Motor Corporation Limited (SAIC).

Take SAIC as an example. SAIC is located in Shanghai, a city that is one of the cities within the FCV Demonstration Application Cluster. As the pioneering automobile enterprise in China, SAIC commenced research and development in fuel cell technology as early as 2001, securing the top rank in the sample. Over the years, SAIC has seen sustained growth, with its total assets amassing 916.9 billion, the highest in the study. The Shanghai FCV Industry’s Innovation and Development Implementation Plan expressly supports SAIC’s ambition to become both a globally and domestically leading manufacturer of FCVs, significantly propelling SAIC’s FCV development forward. Despite ranking 37th in terms of employees with at least a bachelor’s degree, SAIC holds 16,322 invention and utility patents, the highest in the study. Leveraging significant assets, early investments in fuel cell technology, considerable government support, and strong R&D, SAIC strategically develops “small giant enterprises” like Shanghai Hydrogen Propulsion Technology (HPT). As the leading entity in Shanghai’s FCV initiative, SAIC’s high performance is propelled by both technological progress and market demand.

4.2.2. Configuration Analysis of Non-Performance of Sample Enterprises

QCA analysis is characterized by causal asymmetry, which leads to a different configuration of antecedent variables for the high versus not-high performance of the research subject [53,54]. To obtain a comprehensive understanding of the driving pathways for performance in HFCV enterprises, this study further examines the configurations that result in not-high performance. The findings are presented in Table 4. The overall consistency among the sample enterprises experiencing not-high performance is 0.9091, indicating that four configurations of not-high performance of enterprises have an explanatory power of 90.91%. The coverage rate is 0.3267, signifying that the four configurations explain approximately 32.67% of the cases involving not-high performance in HFCV enterprises. Lower government support and lower human capital emerge are the core conditions in all not-high configurations. Thus, achieving high performance in HFCV enterprises necessitates robust government support and an elevated level of human capital. Furthermore, the five conditions of an enterprise’s R&D capability, human capital level, attention distribution, enterprise operation scale, and government support level are all very important for its high performance. Even if one or two of the conditions are high, it is difficult for an enterprise to achieve high performance when other conditions are low. The findings are presented in

Table 5. Configuration analysis of not-high performance.

Condition Variables	Configuration of Not-High Performance
	NLP1	NLP2	NLP3	NLP4
RD	•	⭘		⭘
HC	⭘	⭘	⭘	⭘
AA	⭘	•	⭘
ES			￮	￮
GS	⭘	⭘	⭘	⭘
Consistency	0.9664	0.9237	0.9021	0.8724
Raw coverage	0.1852	0.1619	0.2554	0.2133
Unique coverage	0.0297	0.0233	0.0178	0.0000
Overall solution consistency	0.9148
Overall solution coverage	0.3510

Note: “⭘” means that the core condition does not exist, “•” means that the auxiliary condition exists, “￮” means that the auxiliary condition does not exist, “space” means that the condition can or cannot exist.

.

4.3. Robustness Test

Based on Caves et al. [55], this study conducts robustness testing by adjusting the consistency threshold and calibration. Initially, the consistency threshold is elevated. More precisely, this investigation augments the original consistency threshold from 0.75 to 0.8, whilst keeping the other variables constant. The high performance of HFCV enterprises is re-evaluated under the revised configuration. Furthermore, the calibration is refined by adjusting the quartile points. Calibration parameters are adjusted from 0.75, 0.5, and 0.25 to 0.95, 0.5, and 0.05, respectively. Other variables are left intact. The results are shown in Appendix C and Appendix D. Through comparison and calculation, there is a clear subset relationship between the two robust test configurations [8]. Consequently, the findings of this study are robust.

5. Discussion and Insights

5.1. Discussion

This study examines the multiple pathways to high-performance development in 40 Chinese HFCV enterprises. It employs the TOE framework and the fsQCA method to analyze the synergistic effects of factors such as R&D capability, human capital, enterprise size, attention allocation, and government support on the development of HFCV enterprises. The results indicate that the high-performance development of HFCV enterprises is driven by the synergistic interaction of multiple factors, rather than any single factor. These findings provide theoretical support for the high-performance development of HFCV enterprises in China and offer valuable insights into the HFCV industry in other countries and regions.

In the context of the global energy transition, the HFCV industry is experiencing rapid growth in many countries. The experiences of various countries in promoting the HFCV industry provide valuable references. For instance, the South Korean government has accelerated the development of the HFCV market through active policy support, tax incentives, and the establishment of hydrogen refueling stations [56]. Additionally, India’s experience demonstrates that well-coordinated government support and innovative, sustainable product design will strengthen the HFCV ecosystem [6]. These practices align with the findings of this study, especially regarding the pivotal roles of government support and human capital development.

Specifically, in China, late entrants like FAW Jiefang have achieved significant success through technological breakthroughs and large-scale development, while early entrants such as the Shanghai Group have attained high-performance development by leveraging policy support and collaborative human capital efforts. This is in line with the “Pioneer Strategy” and “Late Follower Strategy” for fuel cell enterprises [57]. These strategies have not only proven successful in China but have also been tested in countries such as Germany. For example, the success of Toyota and BMW in Germany also relied on the dual drive of policy guidance and technological breakthroughs. This shows that the paths to promoting the high-performance development of the HFCV industry exhibit certain commonalities on a global scale. These cross-country experiences provide important implications for this study and further validate the proposition that “different paths lead to the same outcome.”

5.2. Theoretical Contributions

Based on the above, this study offers several significant theoretical contributions. First, it focuses on the high-performance development of HFCV enterprises, addressing a gap in existing research. While HFCVs have gradually developed in multinational markets and expanded into commercial vehicles, most existing research focuses on the hydrogen energy industry chain, technical challenges, and infrastructure at the macro level. Less attention is given to performance pathways and resource allocation strategies at the enterprise level. This paper advances the literature by analyzing key practices in HFCV enterprises, including technology R&D, government support, talent acquisition, and other key areas, through a configuration analysis. The three pathways identified provide solid evidence on how HFCV companies achieve high performance, as well as new insights into how domestic and international companies can adopt specific practices.

Second, this study shifts the focus of enterprise performance research from single factors to a more comprehensive perspective. Previous research on renewable energy performance often focused on a particular factor, lacking in-depth discussion of the comprehensive effects of multiple factors. This paper uses the fsQCA method to consider the synergistic effects of multidimensional factors such as technology, organization, and environment, supporting the proposition that “different paths lead to the same results” [8,53,56,58]. This perspective provides theoretical support for the diverse paths of HFCV corporate performance and responds to the call for a more comprehensive approach to studying corporate performance.

Third, this study introduces a novel methodological approach to enterprise performance research by adopting the “causal asymmetry” perspective. Traditional research on corporate performance often assumes symmetrical causality, where high- and not-high-performance conditions are opposites [59,60]. However, this study found that the conditions leading to high and not-high performance are not strictly opposed. By comparing the three high-performance paths with the four low-performance paths, this paper reveals that different path combinations can lead to either the same or different outcomes. As Fiss notes, the set of conditions leading to an outcome may differ from those that do not [7]. Thus, avoiding negative outcomes is as crucial as promoting positive ones.

Finally, this study reveals a paradox in enterprise development: the same factor may have a “double-edged sword” effect, promoting or hindering performance, depending on its configuration. This finding suggests that the net effect of each factor is not absolute but depends on its interaction with other factors. For instance, Jiang and Xu found that policy negatively affects the performance of new energy vehicle enterprises during the early implementation phase [61]. Conversely, Liu, Peng, and Cao proposed that combining government support with high-level technical input can effectively promote the high-quality development of new energy vehicle enterprises [62]. Therefore, the role of a single factor may vary depending on its combination with other conditions. Research must go beyond the causal inference of single variables and adopt the QCA method with a configuration perspective to comprehensively analyze the complex interactions among multiple factors.

5.3. Management Insights

According to the International Energy Agency (IEA) and related studies, each HFCV can reduce carbon dioxide emissions by 2.5 to 3.5 tons over its life cycle compared to conventional vehicles, equivalent to approximately 60 to 70 percent of annual carbon emissions. Transitioning from fossil fuel vehicles to HFCVs is crucial for achieving climate targets by 2050. This transformation requires policy support, continuous technological advancements, and effective strategic planning and resource allocation by enterprises. HFCV enterprises should prioritize “organizational coordination and overall improvement”, leverage their resource advantages, and promote specialization and refinement. The fsQCA results provide a foundation for identifying optimal strategies for HFCV enterprises.

For the HLP1 (RD*HC*~AA), large independent suppliers entering the HFCV sector late should focus on strengthening research and development and building core technologies to achieve high performance. Specifically, companies should recruit top technical talent and participate in the “Hydrogen and Fuel Cell Industry Talent Training Seminar” to meet market demands for HFCV technologies, including stacks, components, system control, and vehicle technology. Enterprises should also strengthen industry–university–research cooperation and enhance intellectual property protection to consolidate technological leadership.

For HLP2 (HC*AA*GS*~RD), enterprises with strong production capacity but limited independent R&D should leverage their forward-looking insight and human resource advantages to compensate for their R&D limitations. Specifically, these companies should adopt “product-forward development” as a core strategy and establish self-managed product development infrastructure using innovation platforms. They should strengthen international cooperation, expand into overseas markets, and support the development and application of “hydrogen products.”

For HLP3 (AA*ES*GS*~HC), early entrants facing talent shortages should assume the role of “experimenter”, filling the talent gap by applying for government support and leveraging enterprise scale to expand the market. Specifically, enterprises should leverage asset advantages to expand product lines and drive the development of vehicle integration, key components, intelligent manufacturing, and innovative scenario applications. Simultaneously, enterprises should enhance brand value and market recognition through policies like “gazelle enterprise” identification to further consolidate their market position.

Finally, the horizontal comparison of not-high performance development configurations (NLP1–4) reveals that “government support level” and “human capital level” are key factors that significantly hinder the high-performance development of HFCV enterprises, requiring significant attention. Many countries have implemented policies to promote the development of HFCVs. South Korea promotes the commercialization of Hyundai Motor’s Nexo series through tax incentives, financial support, and infrastructure development, while Japan has accelerated R&D and hydrogen refueling station construction through demonstration projects like Toyota Mirai. China can learn from these practices and use fiscal and tax policies to support technological development and promote HFCV applications in the commercial vehicle sector. At the enterprise level, technological breakthroughs and talent acquisition are crucial. Hyundai Motor and Toyota maintain their leadership through global talent acquisition and technological advancements. Chinese enterprises should implement a “talent reserve plan”, overcome bottlenecks in hydrogen fuel cell systems and key components, and seek policy support to accelerate demonstrations and applications to enhance competitiveness. These strategies apply not only to China but also provide valuable insights for enterprises in other countries.

6. Conclusions and Prospects

6.1. Conclusions

In the context of the accelerated global energy transition, the successful development of HFCV enterprises plays a pivotal role in reducing global greenhouse gas emissions and promoting the construction of sustainable transportation systems. Therefore, the high-performance development of HFCV enterprises is essential not only for enhancing corporate competitiveness but also for achieving environmental sustainability. This study conducts an fsQCA analysis of 40 Chinese HFCV enterprises and adopts the TOE framework and a configuration perspective to analyze the synergistic effects and multiple driving pathways of five antecedents, i.e., R&D capability, human capital level, attention allocation, scale of enterprise, and governmental support. The study reveals that, in general, no single antecedent condition alone is sufficient for the high performance of HFCV enterprises. Instead, multiple conditions need to synergistically contribute to improving performance. Additionally, the research innovatively reveals the complex path of high-performance development of HFCV enterprises, which can be split into the following three paths: (1) Technology-driven: For latecomers, gaining a competitive advantage entails rapid key technology breakthroughs through emphasizing technology research and development and human capital. This is complemented by a large-scale enterprise, which enables more trial-and-error opportunities and facilitates high performance. (2) Internal and external synergy: Enterprises that enter the HFCV industry early on but possess low R&D capability require not only high levels of government support to foster local enterprises but also adequate human capital to compensate for technological shortcomings. Achieving high performance necessitates leveraging internal and external synergies. (3) Organization–policy-driven: In cases where human capital is lacking, enterprises can achieve high performance by emphasizing strong corporate attention, a large scale of enterprise, and sufficient government support and complementing these with strong R&D capabilities. Furthermore, four configurations contribute to the not-high level of development in HFCV enterprises. Among these, “government support” and “human capital level” are core conditions in all not-high performance patterns. This finding offers a valuable reference for the policy and strategic adjustments within the global HFCV industry. In summary, the innovation of this study lies in its use of a combination of the TOE framework and the QCA method, creatively incorporating the synergistic effect of multidimensional factors into the study of high-performance development in HFCV enterprises. This approach offers a novel perspective on industry theory development and the diverse growth paths of HFCV enterprises. It also highlights the profound impact of innovative pathways on the high-performance development of enterprises within the context of the global energy transition and environmental sustainability goals.

6.2. Limitations and Prospects

This study adopts the TOE framework and the fsQCA method to investigate the intricate causality underlying the high performance of HFCV enterprises. However, some limitations in this study require further exploration. Firstly, there is room for increasing the sample size. The fsQCA method necessitates a substantial number of samples to uncover causality. However, as HFCV companies belong to emerging industries, the available sample size is relatively limited, which may introduce biases in analysis results. Secondly, while fsQCA is well-suited for exploring complex configurational relationships, it does not allow for testing mediating or moderating mechanisms among variables. Future studies could integrate fsQCA with structural equation modeling (SEM) to enhance causal inference and validate theoretical pathways. In addition, this study only discusses the configuration effects of five antecedents at the level of technology, organization, and environment on the high performance of HFCV enterprises. Future studies could integrate additional external environments (such as the market environment) and internal resources (such as senior management team experience) of enterprises to further analyze the coupling and synergy mechanism of conditions affecting the high performance of HFCV enterprises. Finally, given the global nature of the automotive industry, future research should extend to regional and cross-country comparisons to validate the universality and applicability of this study’s findings. Future research could explore how different countries adopt distinct strategic paths based on their local policy environments, market structures, and technological levels to promote HFCV industry development. Additionally, a comparative study of the energy consumption and environmental impact of HFCVs across different global markets will be a crucial aspect of future research.