The Effect of Corporate Resource Abundance on the Transformation and Upgrading of Manufacturing Enterprises from the Perspective of Whole Process Innovation

Abstract

Exploiting a comparative advantage through resource endowment is a way to promote the transformation and upgrading of enterprises and high-quality economic development. Based on resource-based theory, this paper takes the listed companies of China’s manufacturing industry as a sample and classifies them into leading, potential, catching-up, and exiting enterprises according to the abundance of financial and human resources. In terms of the research perspective, this paper adopts resource abundance as the starting point to explore its impact on enterprise transformation and upgrading and incorporates technological innovation into the framework to investigate its transmission mechanism. The results indicate that the resource abundance of enterprises directly promotes the transformation and upgrading of enterprises. Notably, the mechanism identification test suggests the following. (1) From the perspective of innovation investment, innovation investment intensity is an important channel for manufacturing companies to transform and upgrade. (2) From the perspective of innovation direction, higher levels of resources allow leading enterprises to accelerate transformation and upgrading through product and process innovation. In contrast, catching-up enterprises tend to foster product innovation rather than process innovation to transform and update, while potential enterprises are likely to foster process innovation to transform and update rather than product innovation. On the other hand, the path of transformation and upgrading through product or process innovation is obstructed for the exiting enterprises. (3) From the perspective of innovation output, all but the exiting enterprises can be transformed and upgraded by increasing their innovation output. Overall, considering the impact of corporate innovation behavior, this research offers new insights into the relationship between resource abundance and transformation and upgrading, and it provides inspiration for promoting transformation and upgrading in Chinese manufacturing enterprises.

1. Introduction

Currently, with the fierce competition of globalization, China’s manufacturing industry has long been plagued by inelasticity and weak growth, lack of core technology and innovation, and serious product homogeneity [1], while facing the huge pressure of rapid industrial transformation and upgrading. In response to these challenges, China has proposed the ‘Made in China 2025’ and ‘Building an Innovative Nation’ manufacturing strategies to accelerate the pace of transformation and upgrading [2]. In this context, enterprises, as the basic unit of industrial transformation and upgrading, should spare no effort to play the primary role in implementing industrial transformation and upgrading actions.

The research on transformation and upgrading in the existing literature range widely. Most of the existing studies have primarily focused on the macro- and mesolevel [3,4,5], with a relatively limited number of studies using the case study method to qualitatively analyze the transformation and upgrading process at the microlevel [6,7]. While case studies offer a visual representation of enterprise transformation and upgrading, their scope is restricted as findings from a single sample may not universally apply to all enterprises. Recently, researchers have started to adopt empirical studies to explore enterprise transformation and upgrading. However, the majority of these studies primarily concentrate on examining the role of external policy regimes (e.g., National Innovation Demonstration Zones, green credit, et al.) in promoting the transformation and upgrading of enterprises from a “macro–micro” perspective [8,9,10], making it challenging to gain insights into the internal workings of enterprises.

Based on the above literature review, it can be seen that some achievements have been made in the research of transformation and upgrading. However, there are several research gaps in this field.

➢

First, the promotion effect of enterprise resource abundance on the transformation and upgrading of manufacturing enterprises has not been examined from a micro perspective.

➢

Second, the analysis of the heterogeneous effect of different types of enterprise resource abundance on transformation and upgrading in the manufacturing industry needs to be further deepened.

➢

Third, most studies only stay at the basic level, and there are few attempts to explore transformation and upgrading from the perspective of mechanism research. In addition, while the significance of technological innovation for transformation and upgrading has been highlighted by a large body of literature [11,12], there has been a lack of exploration into the “bridge role” of technological innovation in the process of linking resource abundance with the transformation and upgrading of the manufacturing industry.

To address these gaps, the following questions are explored in this paper:

➢

Does resource abundance stimulate the transformation and upgrading of manufacturing enterprises, and why?

➢

Going further, how does the resource abundance in various types of enterprises differ in its contribution to their transformation and upgrading?

➢

What are the underlying mechanisms within the process? To what extent does technological innovation exert a mediating effect at different stages?

In response to these questions, this paper takes listed Chinese manufacturing enterprises as the research object, and it uses a mediation model to empirically examine the influence of corporate resources on enterprise transformation and upgrading and the mechanisms thereof (Figure 1). The specific analysis process is as follows. At the outset, we divide manufacturing enterprises into four groups based on their resource abundance, linking the four company types to enterprise transformation and upgrading. Afterward, we adopt a technological innovation perspective to test the impact mechanism of resource abundance on manufacturing transformation and upgrading and strive to open the “black box” in three aspects:

• Innovation investment intensity (innovation input).
• Innovation direction (innovation process).
• Innovation quantity (innovation output).

In so doing, the research makes the following contributions.

➢

First, differently from previous works [13,14], this paper pioneers the development of four enterprise type models to investigate the impact of corporate resource abundance on transformation and upgrading, based on the resource-based view. By adopting a unique perspective, this research sheds light on the transformation and upgrading of microenterprises, thus expanding the existing literature on the factors influencing corporate transformation and upgrading.

➢

Second, prior studies mainly identify the driving factors of enterprise transformation and upgrading [15,16]. Nevertheless, in-depth analysis of the mechanisms by which organizational resource factors affect transformation and upgrading is scarce. This paper addresses this research gap by employing a mediation model to investigate the relationship among corporate resources, innovation, and transformation and upgrading. It not only highlights the significance of corporate resources as catalysts for transformation and upgrading, but also introduces a new paradigm for understanding the intricate mechanisms by which strategic corporate resources impact corporate transformation and upgrading.

➢

Third, while previous literature has extensively examined the impact of strategic resource attributes on innovation models within enterprises [17], a comprehensive understanding of the entire innovation process remains unexplored. By integrating the literature on technological innovation with a resource-based view, this paper innovatively introduces the whole innovation process into the model to investigate the mechanisms and boundary conditions related to resource abundance that impact the transformation of manufacturing enterprises. Our findings enrich the theoretical research on enterprise transformation and upgrading and provide further enlightenment for managers to help them make reasonable but innovative strategic and high-quality transformation and upgrading decisions.

The remainder of this paper proceeds as follows. Section 2 reviews the related literature. Section 3 develops our hypotheses. Section 4 describes the sample selection, variable definitions, and empirical model. Section 5 presents our empirical results. Section 6 reports robustness tests. Section 7 presents the discussion and conclusions.

2. Literature Review

2.1. The Determinants of Corporate Transformation and Upgrading

Attention to exploring the factors that influence transformation and upgrading has been developing. In the previous literature, the transformation and upgrading behaviors of firms are influenced by both internal and external factors. With respect to external factors, scholars argue that government environmental regulations [18], government subsidies [19], and the “broadband China” policy [20] give full play to the ability of government pressure and incentives to promote the successful transformation and upgrading of the manufacturing industry. With respect to internal factors, scholars believe that technological innovation plays a pivotal role in supporting the transformation and upgrading of manufacturing industries throughout the various stages of enterprise innovation activities. For example, in the innovation input stage, firms can provide a foundation and guarantee for their transformation and upgrading by enhancing human capital and increasing their R&D investment [21]. In the innovation process stage, firms are able to carry out technological change [22] by exploratory and exploitative innovation [23], providing innovative direction for transformation and upgrading. In the innovation output stage, Lin et al. [21] show that innovation output has a positive association with the transformation and upgrading of manufacturing firms. The stronger the innovation capability is, the greater the probability of transformation and upgrading will be [24]. In addition, other scholars argue that enterprises can perform transformation and upgrading by enhancing marketing capabilities [25] and dynamic capabilities [26], which provide new avenues for future study.

2.2. The Impact of Corporate Resource Abundance

A number of studies have already demonstrated that resource endowments have diverse impacts on facilitating internal activities and expanding external relationships. In terms of internal activities, resource-based theory postulates that resources serve as the engine for firm creation and growth, enabling firms with richer resource endowments to effectively address environmental threats and challenges [27], swiftly pursue risky strategic changes [28], and engage in high-reward innovation activities [29] to bolster their competitiveness and performance [30]. In terms of external relationships, companies having resource advantages not only enhances their interorganizational cooperation [31], but also facilitates manufacturing companies being able to choose the matching value chain integration model in alignment with their resource endowment and improve the effectiveness of their enterprise value chain integration [32].

In summary, prior studies have commonly focused on the issue of resource endowment, technological innovation, and transformation and upgrading, but how different types of manufacturing enterprises utilize their resource endowments to transform and upgrade via technological innovation from the perspective of whole innovation process remains unknown. Therefore, this study addresses this gap by conducting a detailed empirical test of the above mechanism within the manufacturing industry.

3. Research Hypothesis

3.1. Four Strategic Resource Enterprise Types

In this study, we specifically focused on two crucial resources, namely, financial and human resources, since they are essential for firms’ growth and development [33]. We divided the manufacturing companies into four groups based on abundance. (1) The first type, the so-called ‘leading enterprises’, are rich in financial and human resources compared to other firms. (2) ‘Catching-up enterprises’ are rich in financial resources but low in human resources when compared to other firms in the marketplace. (3) ‘Potential enterprises’ are rich in human resources but low in financial resources in comparison to their competitors. (4) Finally, ‘exiting enterprises’ are firms with a paucity of financial and human resources.

From the discussion above, an enterprise type model is depicted in Figure 2.

3.2. Resource-Based Innovation Model

Innovation direction can be divided into product innovation and process innovation [34]. Each firm can be categorized as one of four types based on the innovation type they perform, as illustrated in Figure 3. The present study categorized innovation into product and process innovation, and innovation directions are divided into four categories: product and process-oriented, product-oriented, process-oriented, and non-oriented innovation. Combining Figure 1 and Figure 2 shows that leading companies that focus more on both product and process innovation than others would be located in II. Catching-up enterprises mainly focusing on product innovation but not process innovation would be categorized as I, while potential enterprises that focus primarily on process innovation but not product innovation would be located in IV. Type III refers to the exiting enterprises that have non-oriented innovation in terms of both product and process innovation.

3.3. Analysis of the Impact of Resource Abundance on Enterprise Transformation and Upgrading

Without resources, no business can expand sustainably [36]. The abundance of resource endowment, which serves as one of the essential elements of a firm’s transformation and upgrading model, influences the willingness of enterprises to transform and upgrade and the way it will happen [37]. Specifically, leveraging comparative advantages based on resource endowment, optimizing resource allocation, and adjusting incremental changes are key strategies to drive the transformation and upgrading of the manufacturing industry and promote high-quality economic development [38].

First, the “stock effect” of resources brought about by the accumulation of factors is helpful for enterprises to cope with the threats brought about by external uncertainties, implement technological upgrading, constantly pursue innovation changes, and achieve transformation and upgrading. In addition, the “incremental effect” of resources brought about by a dynamic change in factors can produce a large stock of resources through a small incremental input, boost the level of the enterprise’s resource stock, realize the optimal allocation of resources, and activate the transformation and upgrading of enterprises through a “leveraging mechanism”. Based on the above analysis, we propose the following:

H1. 

Resource abundance is positively related to enterprise transformation and upgrading.

3.4. Analysis of the Channels of Resource Abundance for Enterprise Transformation and Upgrading

Technology innovation activities, differing from ordinary investment activities, involve developing new technology, new products, or new services [39]. The theory of endogenous growth, which posits that technological innovation propels economic development, serves as a core motivating force in guiding enterprises with diverse resource endowments to choose transformation and upgrading paths, transitioning them from factor-driven to innovation-driven. Therefore, we take technological innovation behavior as an essential mediator and portray the technological innovation behavior of enterprises in three stages, namely, innovation intensity (innovation input), innovation direction (innovation process), and innovation quantity (innovation output), to investigate the implementation paths for the transformation and upgrading of Chinese manufacturing enterprises with different resource endowments.

3.4.1. Mediating Role of Innovation Investment Intensity

The transformation and upgrading of manufacturing enterprises evolve dynamically from a low-tech, low-value-added state of products to a high-tech, high-value-added state of products [20]. With long production chains and high innovation risks and costs, the dynamics of transformation and upgrading require financial resources to offer material security for continuous innovation [40] and human resources to make scientific decisions and grasp the direction of investment.

Investment in R&D represents a company’s most direct innovation behavior and signifies a strategic commitment made by the company towards innovation. Without ensuring a certain level of R&D investment, an enterprise cannot attain the necessary innovation capability, resulting in stagnant production technology and rendering transformation and upgrading unachievable. To be specific, financial resources as the initial resources of enterprises, with their strong leveraging ability [41], are a necessity for enterprises to develop various strategies. The abundant financial resources of enterprises alleviate resource constraints in the decision-making process, free the enterprises from the pressure of capital shortages, help them seize investment opportunities [42], and increase their investment in innovation [40], providing a guarantee for technological innovation and transformation activities and enhancing their ability to withstand the risk of investment failure [43]. In addition, human resources, which encompass knowledge and technology, play a pivotal role in the process of transformation and upgrading. Specifically, high-quality human resources combined with higher resource allocation capacity [44] are more likely to promote the efficiency of enterprise resource allocation and the speed as well as quality of decision-making. Therefore, they are conducive for enterprises being able to innovate [45] and utilize technological advantages to implement diversified transformation and upgrading strategies. In light of this, we propose the following:

H2. 

Innovation investment intensity plays a mediating role in the influence of resource abundance on enterprise transformation and upgrading.

3.4.2. The Mediating Role of Innovation Direction

Technological innovation can be classified by type and content into product innovation and process innovation [46]. Product innovation refers to “a new or improved good or service that meets market needs”, which could improve the properties and performance of the finished product, and process innovation refers to “a new or improved business process for one or more business functions that meets company requirements”, which aims to reduce cost and increase productivity [1,34,47].

Choosing appropriate innovation models that match their resource endowments is favorable for enterprises to gain competitive advantages and stimulate their innovation capabilities [18,29,48], thus contributing to the transformation and upgrading of enterprises. Drawing on resource-based logic, companies with resource reserves enrich their decision-making options for innovation models. By learning and synergizing different innovation approaches, companies can effectively achieve twice the result with half the effort in the transformation and upgrading process [49], avoiding the problem of relying solely on a single innovation approach which could weaken their competitive advantage. As a result, enterprises engage in product innovation and process innovation simultaneously, learning from their experiences and mistakes in each innovation process and thus improving them individually and together, which makes it easier to break through innovation bottlenecks and raises their success rates in transformation and upgrading. Our argument is supported by the existing literature [50], which suggests that successful manufacturing firms pursue both product and process innovation, obtaining the highest return on investment when they focus on both types of innovation [51]. In our paper, we argue that leading firms with abundant resources can capitalize on resource complementarity and redundancy to synergistically couple product and process innovation, reduce competition for scarce resources between the two types of innovation, and facilitate successful corporate transformation and upgrading.

Accordingly, firms with limited resources face formidable challenges when engaging concurrently in product and process innovation, necessitating effective resource allocation and strategic decision-making to solve this problem. Specifically, product innovation, which is required to integrate market demand into the process of developing product innovations [52], is relatively risky and requires original and complex thinking, given the low success rate in research and development, creating more uncertainty for companies. Furthermore, the entire lifecycle of a new product, from development to output, involves multiple intricate processes and incurs high adjustment costs [53]. Therefore, before embarking on product innovation, companies should thoroughly evaluate their innovation capacity and financial situation [54]. In contrast, process innovation, with low innovation risk [51] and a low probability of failure in transformation and upgrading, does not require long-term accumulation of new knowledge and continuous investment, and the adjustment costs are relatively small. As a result, catching-up enterprises with ample financial resources possess greater flexibility in utilizing product innovation to drive transformation and upgrades, as it allows for faster adaptation to changing market needs. In contrast, potential enterprises with limited financial resources prioritize process innovation over product innovation to increase product output rates, which is more suitable for achieving transformation and upgrading. Finally, the blind pursuit of product innovation or process innovation by existing enterprises under constraints of financial and human resources, relative to other types of firms, may also be counterproductive to transformation and upgrading. Grounded in the above, we propose the following:

H3a. 

Leading companies tend to focus on both process and product-oriented innovation for transformation and upgrading.

H3b. 

Catching-up enterprises tend to transform and upgrade through product-oriented innovation rather than process innovation.

H3c. 

Potential enterprises tend to transform and upgrade through process-oriented innovation rather than product innovation.

H3d. 

Exiting companies tend to transform and upgrade through non-oriented innovation.

3.4.3. The Mediating Role of Innovation Output

Sufficient resources and advanced innovative technology are the prerequisites for the manufacturing industry to transform and upgrade. Specifically, by harnessing abundant resources, optimizing resource utilization, and leveraging a wide range of technical knowledge [55], enterprises can implement differentiation strategies to carry out innovation activities, reduce development cycles, improve innovation performance [56], and guarantee quality and efficiency improvement [57,58], ultimately realizing the transformation and upgrading of the manufacturing industry. However, when firms face resource constraints, their ability to engage in innovation activities is greatly diminished, leading to a decrease in innovation output and impeding their path towards transformation and upgrading. Given this, exiting enterprises have fewer financial and human resources to bear the losses caused by innovation failures and thus cannot transform and upgrade through innovation output. On this basis, we propose:

H4. 

Apart from exiting enterprises, innovation output plays a mediating role in the influence of resource abundance on enterprise transformation and upgrading.

From the above discussions, a framework is depicted in Figure 4.

4. Methodology

4.1. Sample and Data Collection

This paper used listed manufacturing companies from 2015–2020 as the research object, which were obtained from CSMAR and the INCOPAT database. “In 2015, the Chinese government officially proposed the action plan ‘Made in China 2025’, which clearly defined the key directions for the future transformation and upgrading of manufacturing industry. Therefore, taking 2015 as the starting point for this research can assess the extent of enterprises’ strategic responsiveness to transformation and upgrading”. In 2020, a global outbreak of COVID-19 occurred, leading to a global economic slowdown. To avoid the COVID-19 pandemic having an impact on the data for the transformation and upgrading of the manufacturing industry, this study covered the period from 2015 to 2020. After excluding *ST companies and deleting companies with incomplete or abnormal financial data, final sample data for 9709 companies were eventually obtained. Winsorize tailing was applied to all continuous variables at the 1% and 99% levels to avoid the effects of extreme values. Finally, Stata was used to process the data.

4.2. Measures

4.2.1. Dependent Variable

Corporate transformation and upgrading served as a dependent variable in this study, and we selected total factor productivity as the proxy variable of transformation and upgrading (TFP). The reasons were as follows. Firstly, the essence of enterprise transformation and upgrading results from an overall improvement in resource allocation, the technology level, and organizational efficiency. No matter how the enterprise modifies its product or industrial value chain, its total factor productivity will ultimately reflect those changes [59]. Secondly, compared with other variables, total factor productivity, as a comprehensive index, contains richer information, is more comprehensive, and has higher recognition. For instance, Zhao et al. [60] use enterprise total factor productivity to measure the transformation and upgrading of manufacturing enterprises and study the influence of central environmental inspection on the transformation and upgrading of manufacturing enterprises. Thirdly, the core of the transformation of the economic growth mode is the improvement of efficiency [61], while the growth of total factor productivity can reflect the transformation and upgrading of enterprises. Therefore, this paper selected total factor productivity as a proxy for the level of transformation and upgrading of manufacturing enterprises, which was calculated by the GMM method to effectively solve the possible endogeneity problem of the model [62,63,64].

4.2.2. Independent Variable

Our independent variable was resource abundance (Fyd), composed of financial and human resources, since adequate financial resources and well-trained human resources are known to be beneficial to productivity [65]. Specifically, financial resources were measured using cash flow [66]. Human resources were measured by the number of employees [67]. Then, this paper used the entropy weight method to measure resource abundance.

4.2.3. Mediating Variables

Technological innovation is a dynamic process that involves innovation investment, innovation direction, and innovation output. Consequently, we used these as mediator variables. To be specific, innovative investment is the beginning of a firm’s innovation activities and implies the attention given to technological innovation, measured as the ratio of corporate innovation investment to the level of the same industry. To further analyze the role of different innovation models, innovation direction was divided into product and process innovation, measured as the enterprise’s number of process and product innovation patents [68]. Finally, innovation output is the result of a firm’s innovation activities and was measured by the number of innovation patents for product and process innovations. This paper used textual analysis to identify patent categories in two steps: (1) We referred to the definitions of product innovation and process innovation and the questionnaire in the literature to find the keywords for product innovation and process innovation. (2) On the basis of the patent descriptions in the INCOPAT database, we conducted a keyword text search to identify the number of patents for product and process innovation. The keywords for these two types of innovation patents are shown in

Table 1. Data extraction.

Dimension 1. Product Innovation

Typical Context Matching from the Original Analysis of the Text Keywords

product, new product, product quality, product structure, product design, product development, product formalization, product service, product distribution, product maintenance, product function, product appearance, product type, product performance, product commissioning, product added value, new services, product portfolio, new functions, product market, product development and design, product function upgrade, product life, product differentiation, product design, product aesthetics, product practicality, product stability, product comfort, economic efficiency

Dimension 2. Process Innovation

Typical Context Matching from the Original Analysis of the Text Keywords

process, manufacturing technique, production process, cost advantage, production performance, production cost, production efficiency, process technology, process innovation, technique innovation, production method, process flow, management process, labor productivity, management method, processing method, testing method, process technology innovation, process equipment innovation, process technology upgrade, intelligent transformation, cost reduction, cost savings, energy saving and emission reduction, product processing sequence, processing technology, processing procedures, production side, process upgrading and optimization, equipment upgrading and optimization, production equipment, process transformation and upgrading, resource utilization, cost cutting, processing methods, process equipment flow, production processing, equipment renewal, process flow optimization, assembly processes, processing efficiency, industrialized production, production lines, product conversion rate, new equipment, control devices, processing equipment, industrial preparation methods, mass production, yield, production processes, process times, manufacturing costs, process simplicity, failure rates, improved yields, scrap rates, yield rates, processing accuracy

.

4.2.4. Control Variables

Following previous studies [69,70,71,72,73,74,75], we controlled a number of variables to alleviate the problem of bias in the regression results due to omitted variables in the model. Firm size (Size) was equal to the natural logarithm of total assets at year’s end. Leverage (Lev) was estimated by total debts/total assets. Return on net assets (Roe) was the ratio of net profit to net assets. Operating profit margins (Pro) was the ratio of operating profit to operating income. The remuneration of the top three directors (Inc) was measured by the natural logarithm of remuneration of the top three directors. Net profit growth rate (Growth) was the ratio of the increase in net profit for the period to net profit for the previous period. Capital intensity (Zbmw) was derived as fixed assets/the number of employees/10,000. CEO duality (Dual) was a dummy variable indicating whether the Chairman and CEO were the same people. Ownership concentration (Shrcr) was the sum of the shareholding ratio of the top 3 largest shareholders of the company. Funding liquidity (Ldx) was the ratio of monetary funds to total assets. Management concentration (Mna) was expressed as the ratio of the number of senior managers to the number of employees. The total operating cost rate (Cbl) was represented as operating profit/operating costs. The fixed asset growth rate (Fgr) was the ratio of growth in fixed assets for the period to fixed assets at the beginning of the period. The sustainable growth rate (Sus) was calculated as Roe × Earnings retention ratio/(1 − Roe × earnings retention ratio). The establishment of committees (Com) was the number of committees of the company. The percentage of independent directors (Dratio) was represented as the number of independent directors/number of directors. In addition, this study also controlled for some factors that influence innovation and transformation and upgrading, for example, management concentration (Mna), the total operating cost rate (Cbl), the establishment of committees (Com), an annual dummy variable, and an industry dummy variable.

In summary, all of the variables used in this study are shown in

Table 2. Variable definition.

Variable Type	Variable Name	Variable Symbol	Variable Definition
Dependent variable	Transformation and upgrading	Tfp	Calculated by the GMM method
Mediating variables	Innovation investment intensity	Inv	Ratio of corporate innovation investment to the level of the same industry
	Product innovation	Product	The number of product patents
	Process innovation	Process	The number of process patents
	Innovation quantity	Count	Sum of product innovation and process innovation patents
Independent variable	Resource abundance	Fyd	Calculated by using entropy weight method
Control variables	Firm size	Size	Ln (total assets)
	Leverage	Lev	The ratio of total debts to total assets
	Operating profit margins	Pro	The ratio of operating profit to operating income
	Remuneration of the top three directors	Inc	The natural logarithm of remuneration of the top three directors
	Net profit growth rate	Growth	(Net profit for the period − net profit for the previous period)/net profit for the previous period
	Ownership concentration	Shrcr	The sum of the shareholding ratio of the top 3 largest shareholders of the company
	Capital intensity	Zbmw	Fixed assets/number of employees/10,000
	CEO duality	Dual	Take 1 if the Chairman and CEO are the same people, otherwise, take 0
	Funding liquidity	Ldx	Monetary funds/total assets
	Management concentration	Mna	Number of senior managers/number of employees
	Total operating cost rate	Cbl	Operating profit/operating costs
	Fixed asset growth	Fgr	(Closing fixed assets − beginning fixed assets)/beginning fixed assets
	Sustainable growth rate	Sus	Roe × earnings retention ratio/(1 − Roe × earnings retention ratio)
	Establishment of committees	Com	The number of the committees of the company
	Return on net assets	Roe	Net profit/net assets
	Percentage of independent directors	Dratio	Number of independent directors/number of directors
	Year	Year	Year dummy variables
	Industry	Ind	Industry dummy variables

.

4.3. Methods of Analysis

Methodologically, this paper adopted the method of Wen and Ye [76] to examine the mediating effect and we supplemented this analysis with Sobel’s [77] test to determine the type and significance of the mediation effect [78]. Specifically, the proposed research question aimed at discovering the underlying mechanism producing a relationship between corporate resource abundance and transformation and upgrading; thus, the paper’s research question concerned issues of mediation. Indeed, adopting a mediation model was a reasonable approach as it goes beyond simple cause and effect relationships to explore the underlying mechanisms that lead to the outcome variable (Figure 1). Within this framework, a mediating variable shed light on the mechanism through which an independent variable influenced an outcome variable [79]. In addition, among the various methods available to test for a mediation effect, such as the production of coefficients or the difference in coefficients, we selected the production of coefficients method due to its higher statistical power and lower risk of Type I error [80].

The specific steps were as follows:

(1)

After the regression between resource abundance (X) and enterprise transformation and upgrading (Y), we can check to see if c is significant. If it is significant, we proceed to the second-step regression. Otherwise, the analysis is not further conducted.

(2)

After the regression between resource abundance (X) and technological innovation (M), we can check to see if a is significant.

(3)

Resource abundance (X) and technological innovation (M) are regressed with enterprise transformation and upgrading (Y). If both a and b are significant, and c′ is not significant, it means that there is a full mediating effect. If c′ is still significant but decreases, it implies that there is a partial mediating effect. If at least one of a and b is not significant, then a Sobel test is needed. If it is significant, it implies that there is a mediation effect. Otherwise, there is no mediation effect.

$$\mathrm{Y}=\mathrm{c}\mathrm{X}+{\displaystyle \sum \mathrm{c}\mathrm{o}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{o}\mathrm{l}}+{\mathrm{e}}_1$$

(1)

$$\mathrm{M}=\mathrm{a}\mathrm{X}+{\displaystyle \sum \mathrm{c}\mathrm{o}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{o}\mathrm{l}}+{\mathrm{e}}_2$$

(2)

$$\mathrm{Y}={\mathrm{c}}^{\prime }\mathrm{X}+\mathrm{b}\mathrm{M}+{\displaystyle \sum \mathrm{c}\mathrm{o}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{o}\mathrm{l}}+{\mathrm{e}}_3$$

(3)

In the above model, Y represents enterprise transformation and upgrading (Tfp). M represents technological innovation, including innovation investment intensity (Inv), product innovation (Product), process innovation (Process), and innovation quantity (Count). X represents resource abundance (Fyd). Control variables include Size, Lev, Roe, Pro, Inc, Growth, Shrcr, Zbm, Dual, Ldx, Mna, Cbl, Fgr, Sus, Com, and Dratio, and e_n represents residual terms.

5. Empirical Results

Figure 5 represents the mediation model to determine the total effect of resource abundance as well as its direct and indirect effects (through mediation variables) on transformation and upgrading in different types of enterprises.

Table 3. Identification of the mechanisms of resource abundance affecting transformation and upgrading in leading enterprises: innovation input, innovation direction, and innovation output.

Variables	(1) Tfp	(2) Inv	(3) Tfp	(4) Product	(5) Process	(6) Tfp	(7) Tfp	(8) Count	(9) Tfp
Fyd	11.9558 ***	24.1892 ***	11.4484 ***	1.8848 **	4.3417 **	11.8897 ***	11.8826 ***	6.2265 **	11.8768 ***
	(18.7693)	(6.3651)	(19.9744)	(1.9735)	(2.0110)	(18.6998)	(18.6825)	(2.1368)	(18.6885)
Inv			0.0210 ***
			(3.5942)
Product						0.0351 ***
						(3.6134)
Process							0.0169 **
							(2.1022)
Count									0.0127 ***
									(2.6142)
Size	2.8305 ***	5.9512 ***	2.7056 ***	−0.4647 ***	−0.6903 **	2.8468 ***	2.8421 ***	−1.1550 ***	2.8451 ***
	(38.0252)	(12.5587)	(31.1863)	(−2.8687)	(−2.4407)	(38.3910)	(38.2470)	(−2.7054)	(38.3321)
Lev	−0.2055	−2.4394 ***	−0.1543	0.2958	0.0391	−0.2158	−0.2061	0.3349	−0.2097
	(−1.1498)	(−3.7361)	(−0.8655)	(1.4742)	(0.1066)	(−1.2099)	(−1.1572)	(0.6292)	(−1.1774)
Pro	−4.2276 ***	−7.9065 ***	−4.0617 ***	0.0325	−0.3946	−4.2287 ***	−4.2209 ***	−0.3621	−4.2230 ***
	(−6.4721)	(−4.0108)	(−6.2826)	(0.0844)	(−0.5984)	(−6.4703)	(−6.4783)	(−0.3751)	(−6.4768)
Growth	−0.0081	−0.0044	−0.0080	0.0026	0.0010	−0.0082	−0.0081	0.0036	−0.0082
	(−1.4912)	(−0.3045)	(−1.5053)	(0.6363)	(0.1441)	(−1.5111)	(−1.5003)	(0.3394)	(−1.5053)
Inc	0.0205	0.2798 ***	0.0146	−0.0606	−0.0959	0.0226	0.0221	−0.1565	0.0225
	(1.2323)	(2.8290)	(0.8785)	(−1.5358)	(−1.0831)	(1.3570)	(1.3257)	(−1.2950)	(1.3479)
Dual	−0.0674 **	0.2670	−0.0730 ***	0.0312	0.0464	−0.0685 **	−0.0682 **	0.0776	−0.0684 **
	(−2.4715)	(1.4303)	(−2.6556)	(0.5362)	(0.4655)	(−2.5195)	(−2.5096)	(0.5108)	(−2.5175)
Shrcr	0.0040 ***	−0.0028	0.0041 ***	−0.0005	−0.0023	0.0040 ***	0.0040 ***	−0.0028	0.0040 ***
	(3.5990)	(−0.4648)	(3.6770)	(−0.2663)	(−0.6898)	(3.6161)	(3.6339)	(−0.5686)	(3.6317)
Zbmw	0.0062 ***	−0.0093 ***	0.0064 ***	0.0017 **	0.0043 **	0.0061 ***	0.0061 ***	0.0060 **	0.0061 ***
	(12.8292)	(−3.1898)	(13.2579)	(1.9854)	(2.3799)	(12.7649)	(12.6704)	(2.3871)	(12.6927)
Ldx	−0.4612 ***	−1.4796 *	−0.4302 ***	0.6186	1.1834	−0.4829 ***	−0.4812 ***	1.8020	−0.4841 ***
	(−3.6858)	(−1.8170)	(−3.4875)	(1.5138)	(1.5594)	(−3.8601)	(−3.8363)	(1.5800)	(−3.8616)
Mna	−152.1335 ***	798.4957 ***	−168.8834 ***	34.7795	78.6731	−153.3527 ***	−153.4608 ***	113.4526	−153.5722 ***
	(−4.2021)	(4.9873)	(−4.7641)	(0.3928)	(0.5770)	(−4.2596)	(−4.2685)	(0.5353)	(−4.2738)
Dratio	0.1016	3.7660 ***	0.0226	−0.0601	−0.2760	0.1037	0.1063	−0.3361	0.1059
	(0.4743)	(2.8867)	(0.1077)	(−0.1460)	(−0.2851)	(0.4854)	(0.4963)	(−0.2549)	(0.4949)
Cbl	−1.4108 ***	−2.6506 *	−1.3552 ***	0.0374	−0.5276	−1.4121 ***	−1.4019 ***	−0.4903	−1.4045 ***
	(−3.7369)	(−1.7983)	(−3.6424)	(0.1057)	(−0.6837)	(−3.7491)	(−3.7512)	(−0.4622)	(−3.7522)
Fgr	0.0395	−0.2088	0.0439	−0.0254	0.1011	0.0404	0.0378	0.0756	0.0385
	(0.5900)	(−0.6910)	(0.6624)	(−0.2733)	(0.4283)	(0.6036)	(0.5638)	(0.2385)	(0.5753)
Com	−0.0322	−0.5205 ***	−0.0212	0.0235	0.0617	−0.0330	−0.0332	0.0852	−0.0332
	(−1.3103)	(−2.9440)	(−0.8772)	(0.4889)	(0.4607)	(−1.3350)	(−1.3391)	(0.4850)	(−1.3407)
Sus	−0.3366	3.4378 **	−0.4088	0.7104	0.8051	−0.3615	−0.3502	1.5155	−0.3559
	(−1.2458)	(1.9819)	(−1.5218)	(1.4114)	(0.8224)	(−1.3349)	(−1.2889)	(1.0990)	(−1.3103)
Roe	3.8724 ***	3.3480	3.8021 ***	−0.3116	−0.8421	3.8833 ***	3.8866 ***	−1.1537	3.8870 ***
	(7.9712)	(1.6236)	(7.9036)	(−0.5668)	(−0.7343)	(8.0172)	(8.0503)	(−0.7215)	(8.0476)
Constant	12.0549 ***	−61.1922 ***	13.3385 ***	5.3034 ***	8.5810 ***	11.8690 ***	11.9101 ***	13.8844 ***	11.8788 ***
	(16.5270)	(−13.3153)	(16.9716)	(3.6983)	(3.5101)	(16.3202)	(16.3457)	(3.7781)	(16.3232)
Year dummy	YES	YES	YES	YES	YES	YES	YES	YES	YES
Ind dummy	YES	YES	YES	YES	YES	YES	YES	YES	YES
Observations	1356	1356	1356	1356	1356	1356	1356	1356	1356
R−squared	0.9568	0.6964	0.9573	0.0785	0.0558	0.9570	0.9570	0.0616	0.9570

Note: ***, **, and * represent statistical significance at the levels of 1%, 5%, and 10%, respectively. The numbers in parentheses are T values.

reports the regression results for the effects of the channels of resource abundance on transformation and upgrading in leading enterprises. We focused on testing the channels of resource abundance stimulating transformation and upgrading in leading enterprises, that is, whether leading enterprises transform and upgrade by increasing innovation intensity and output or by adopting different innovation models. The outcomes indicated that leading enterprises with abundant resources reap substantial benefits throughout the entire innovation process, leading to transformative changes and upgrades.

Columns (1)–(3) in Table 3 show the results of testing H1 and H2, which were set to examine whether innovation investment intensity mediates the relationship between resource abundance and transformation and upgrading in leading enterprises. Specifically, Columns (1) and (2) showed that leading enterprises with higher levels of resource abundance were prone to transform and upgrade (β = 11.9558, p < 0.01) and invested more in innovation activities (β = 24.1892, p < 0.01). When the mediator variable Inv was added to the model in Column (3), the coefficients of Tfp decreased only slightly (from β = 11.9558–11.4484, which was significant). This meant that innovation investment intensity (Inv) partially mediated the relationship between resource abundance (Fyd) and transformation and upgrading (Tfp) in leading enterprises. H1 and H2 were supported. Accordingly, it implied that the intensification of financing and human capabilities can elevate the level of innovation intensity, thereby facilitating the transformation and upgrading of enterprises. The result was in agreement with [81,82], who highlighted the pivotal role of financial and talent support in facilitating R&D activities and promoting transformation and upgrading initiatives within enterprises.

Columns (1), (4)–(7) in Table 3 show the results of testing H3a, and were set to test the mediating effects of the two types of innovation directions between resource abundance and leading enterprise transformation and upgrading. Specifically, Columns (1), (4), and (6) were the regression results for resource abundance, product innovation, and enterprise transformation and upgrading. In addition, Columns (1), (5), and (7) presented the regression results for resource abundance, process innovation, and enterprise transformation and upgrading. It can be seen that all regression results were positive and significant. The findings demonstrated that resource abundance (Fyd) positively influenced the transformation and upgrading of leading enterprises (Tfp) through both process and product innovation, indicating a partial mediating role. H3a was validated. The results agreed with Ettlie [50].

Columns (1) and (8)–(9) in Table 3 show the results of testing H4, and were set to test the innovation output mediation effect. The results implied that innovation output (Count) was an effective mediating variable for resource abundance (Fyd) to improve transformation and upgrading (Tfp) in leading enterprises (β = 11.9558, p < 0.01; β = 6.2265, p < 0.05; from β = 11.9558–11.8768, p < 0.01). H4 was established. Specifically, on the one hand, technology innovation possesses traits such as long cycles, high risks, and significant capital requirements, necessitating the support of both R&D personnel and capital [83]. On the other hand, the transformation and upgrading of the manufacturing industry involve multifaceted aspects such as production and technology. Given that, enterprise resources and innovation capabilities are vital factors that enhance enterprise transformation and upgrading [84]. Consequently, the ample resources possessed by leading firms ultimately result in increased innovation output and the attainment of high-quality transformation and upgrading.

Table 4. Identification of the mechanisms of resource abundance affecting transformation and upgrading in catching-up enterprises: innovation input, innovation direction, and innovation output.

Variables	(1) Tfp	(2) Inv	(3) Tfp	(4) Product	(5) Process	(6) Tfp	(7) Tfp	(8) Count	(9) Tfp
Fyd	111.2636 ***	9.7127 ***	110.0454 ***	21.7236 **	16.3754	111.0839 ***	111.1208 ***	38.0990	111.0620 ***
	(30.2816)	(2.9333)	(30.5544)	(1.9845)	(0.9813)	(30.2254)	(30.2222)	(1.5088)	(30.2110)
Inv			0.1254 ***
			(4.8329)
Product						0.0083
						(1.5770)
Process							0.0087 ***
							(2.6481)
Count									0.0053 **
									(2.4033)
Size	2.3670 ***	0.9028 ***	2.2538 ***	−0.4045 ***	−0.2538	2.3703 ***	2.3692 ***	−0.6583 **	2.3705 ***
	(44.2971)	(12.1851)	(36.3672)	(−3.1910)	(−1.4528)	(44.3466)	(44.3416)	(−2.3930)	(44.3684)
Lev	0.1416	−0.1393 **	0.1591	−0.0390	−0.1265	0.1420	0.1427	−0.1655	0.1425
	(1.3966)	(−2.1827)	(1.5937)	(−0.2325)	(−0.5555)	(1.3991)	(1.4067)	(−0.4530)	(1.4043)
Pro	−3.0033 ***	−0.8566 ***	−2.8958 ***	0.2432	0.2332	−3.0053 ***	−3.0053 ***	0.4764	−3.0058 ***
	(−7.9407)	(−5.7829)	(−7.6888)	(0.9264)	(0.6239)	(−7.9530)	(−7.9568)	(0.8114)	(−7.9585)
Growth	0.0002	−0.0025 *	0.0005	−0.0035	−0.0067	0.0003	0.0003	−0.0102	0.0003
	(0.0933)	(−1.8659)	(0.2257)	(−1.1816)	(−1.4469)	(0.1055)	(0.1176)	(−1.4426)	(0.1159)
Inc	0.0579 ***	0.0341 **	0.0536 ***	−0.0014	−0.0500	0.0579 ***	0.0583 ***	−0.0514	0.0582 ***
	(3.6823)	(2.4557)	(3.4972)	(−0.0340)	(−0.8041)	(3.6809)	(3.7091)	(−0.5490)	(3.6980)
Dual	−0.0225	0.0222	−0.0253	−0.0768	−0.0665	−0.0219	−0.0220	−0.1433	−0.0218
	(−1.3160)	(1.5943)	(−1.4789)	(−1.5179)	(−0.8741)	(−1.2781)	(−1.2824)	(−1.2328)	(−1.2716)
Shrcr	−0.0008	−0.0005	−0.0008	−0.0003	−0.0020	−0.0008	−0.0008	−0.0023	−0.0008
	(−1.2600)	(−0.8970)	(−1.1697)	(−0.1848)	(−0.7922)	(−1.2557)	(−1.2336)	(−0.5785)	(−1.2413)
Zbmw	0.0083 ***	−0.0008 ***	0.0084 ***	0.0017 **	0.0012	0.0083 ***	0.0083 ***	0.0029 **	0.0083 ***
	(23.3860)	(−2.7585)	(23.5150)	(2.3956)	(1.3324)	(23.3706)	(23.4267)	(1.9728)	(23.4089)
Ldx	−0.6309 ***	0.1238 **	−0.6464 ***	−0.1826	−0.2438	−0.6293 ***	−0.6287 ***	−0.4264	−0.6286 ***
	(−8.7296)	(2.2559)	(−9.0175)	(−0.8815)	(−0.7891)	(−8.7233)	(−8.7139)	(−0.9094)	(−8.7158)
Mna	−57.5735 ***	8.1049 ***	−58.5900 ***	4.1321	4.5326	−57.6077 ***	−57.6130 ***	8.6647	−57.6193 ***
	(−18.2623)	(5.6793)	(−18.4703)	(0.4875)	(0.4262)	(−18.2504)	(−18.2851)	(0.4989)	(−18.2683)
Dratio	−0.3239 **	0.1527	−0.3430 **	−0.1434	−0.1164	−0.3227 **	−0.3228 **	−0.2598	−0.3225 **
	(−2.0420)	(1.3313)	(−2.1857)	(−0.3078)	(−0.1701)	(−2.0352)	(−2.0377)	(−0.2468)	(−2.0353)
Cbl	−1.0911 ***	−0.1371	−1.0739 ***	−0.3983 *	−0.8329 **	−1.0878 ***	−1.0839 ***	−1.2312 **	−1.0846 ***
	(−4.2143)	(−1.2984)	(−4.2419)	(−1.6492)	(−2.2044)	(−4.2023)	(−4.1871)	(−2.1530)	(−4.1903)
Fgr	0.0272	−0.0020	0.0274	0.0538	−0.0066	0.0267	0.0272	0.0471	0.0269
	(1.3836)	(−0.1746)	(1.4042)	(0.8795)	(−0.0719)	(1.3587)	(1.3768)	(0.3332)	(1.3637)
Com	0.0194	−0.0669 ***	0.0278	0.0414	−0.0319	0.0190	0.0197	0.0095	0.0193
	(0.6931)	(−2.7710)	(1.0114)	(0.6613)	(−0.4039)	(0.6817)	(0.7035)	(0.0720)	(0.6922)
Sus	−0.5901 ***	0.4180 **	−0.6426 ***	0.6871	0.7290	−0.5958 ***	−0.5965 ***	1.4161	−0.5976 ***
	(−2.8448)	(2.2529)	(−3.1209)	(1.4287)	(1.1593)	(−2.8701)	(−2.8757)	(1.4354)	(−2.8801)
Roe	3.8996 ***	0.9565 ***	3.7796 ***	−0.6623	−0.7711	3.9051 ***	3.9063 ***	−1.4334	3.9072 ***
	(9.0527)	(4.1967)	(8.8594)	(−1.3468)	(−1.2641)	(9.0546)	(9.0539)	(−1.4444)	(9.0550)
Constant	13.6436 ***	−8.5310 ***	14.7136 ***	4.1476 ***	4.6525 ***	13.6093 ***	13.6031 ***	8.8002 ***	13.5971 ***
	(24.1382)	(−13.2162)	(24.1723)	(3.6020)	(2.8191)	(24.0371)	(24.0302)	(3.4257)	(24.0127)
Year dummy	YES	YES	YES	YES	YES	YES	YES	YES	YES
Ind dummy	YES	YES	YES	YES	YES	YES	YES	YES	YES
Observations	3406	3406	3406	3406	3406	3406	3406	3406	3406
R−squared	0.9310	0.4667	0.9317	0.0575	0.0516	0.9310	0.9311	0.0600	0.9311
Sobel Test for Product Mediator Variables				Sobel Test for Process Mediator Variables			Sobel Test for Count Mediator Variables
z		mediating effect		z	mediating effect		z	mediating effect
1.2438		significant		0.9197	insignificant		1.2787	significant

Note: ***, **, and * represent statistical significance at the levels of 1%, 5%, and 10%, respectively. The numbers in parentheses are T values.

reports the regression results for the effects of the channels of resource abundance on transformation and upgrading in catching-up enterprises. We focused on examining the pathways through which resource abundance enhances the transformation and upgrading process in catching-up enterprises, that is, whether catching-up enterprises transform and upgrade by adopting product innovation activities or process innovation activities and increasing innovation output. The results indicated that catching-up firms with a high level of financial resources had greater flexibility in applying resources to invest in product innovation activities and increasing innovation output, thus accomplishing transformation and upgrading.

Specifically, Columns (1)–(3) in Table 4 show the results of testing H1 and H2. The findings suggested that resource abundance (Fyd) enhanced catching-up enterprises’ transformation and upgrading (Tfp) by allowing them to spend more on innovation investments (Inv) (β = 111.2636, p < 0.01; β = 9.7127, p < 0.01; from β = 111.2636–110.0454, which was significant). H2 was validated. Specifically, R&D activities and transformation and upgrading have a high degree of uncertainty. Financial constraints limit the resources available to innovation activities [85] or even interrupt R&D projects [86]. Therefore, catching-up enterprises with substantial financial resources can ensure investment persistence and promote successful transformation and upgrading.

Columns (1), (4)–(7) in Table 4 show the results of testing H3b. Specifically, Columns (1), (4), and (6) are the regression results for resource abundance, product innovation, and enterprise transformation and upgrading. As shown in Column (1), the regression coefficient of Fyd was positively and significantly associated with Tfp (β = 111.2636, p < 0.01). As shown in Column (4), Fyd had a significant positive effect on Product (β = 21.7236, p < 0.05). When the mediator variable Product was added to the model in Column (6), the coefficient of Tfp was 111.0839, which was smaller than that in Column (1) (β = 111.2636, p < 0.01). However, the coefficient between Product and Tfp was positive and failed to pass the significance test at the 10% level (β = 0.0083, p > 0.1). Following the method of Wen and Ye [52], if there was a regression coefficient failing the test in the second or third step, the Sobel test was carried out. This method is mainly used to test whether the coefficient of the a × b product term is significant, and the test statistic is calculated as $z=(\widehat{a}\times \widehat{b})/s_{ab}$. Unlike the standard normal distribution, the critical value of Sobel test statistics at the significance level of 5% was approximately 0.97 for the test of mediating effect [77]. After calculation, the Sobel test results demonstrated the significant presence of the Product intermediation effect (Z = 1.2438 > 0.97). Columns (1), (5), and (7) are the regression results for resource abundance, process innovation, and enterprise transformation and upgrading. Column (1) shows that the regression coefficient of Fyd was positively and significantly associated with Tfp (β = 111.2636, p < 0.01). Nevertheless, in Column (5), the regression coefficients of Fyd were not significantly related to Process (β = 16.3754, p > 0.1). When the mediator variable Process was added to the model in Column (7), the coefficients of Tfp was 111.1208, which was less than 111.2636 in the total effect regression (p < 0.01). The Sobel test needed to be carried out [76]. After calculation, Z = 0.9197 < 0.97, suggesting that the mediating effect of process innovation did not exist. This was consistent with H3b. Our results conflicted with those of Cho and Linderman [17], as we found that catching-up enterprises, which are rich in human resources, were prone to engage in product innovation activities rather than process innovation. Taking it further, our paper reveals that product innovation played a partial mediating role among catching-up enterprises as they adapt to transform and upgrade.

Columns (1) and (8)–(9) in Table 4 show the results of testing H4. Specifically, Column (1) shows that the regression coefficient of Fyd was positively and significantly associated with Tfp (β = 111.2636, p < 0.01). As seen from Column (8), the effect of Fyd on Count was not significant (β = 38.0990, p > 0.1). When the mediator variable Count was added to the model in Column (9), the coefficient of Tfp was lower than that of Column (1) (from β = 111.2636–111.0620, p < 0.01). The Sobel test needed to be carried out [76]. After calculation, Z = 1.2787 > 0.97, which meant that innovation output (Count) was the mediating variable for resource abundance (Fyd) to push transformation and upgrading (Tfp) in catching-up enterprises. H4 was validated. Compared with R&D input, enterprises were more concerned about innovation output. The advantage of enterprises having financial resources can motivate them to conduct R&D activities and increase innovation output. Meanwhile, the stronger the innovation capability of enterprises, the higher the probability of implementing transformation and upgrading.

Table 5. Identification of the mechanisms of resource abundance affecting transformation and upgrading in potential enterprises: innovation input, innovation direction, and innovation output.

Variables	(1) Tfp	(2) Inv	(3) Tfp	(4) Product	(5) Process	(6) Tfp	(7) Tfp	(8) Count	(9) Tfp
Fyd	14.3935 ***	30.8727 ***	13.5997 ***	3.9560	3.0301 ***	14.3589 ***	14.2899 ***	7.2981**	14.2867 ***
	(11.2535)	(5.0069)	(10.6030)	(1.4591)	(2.6060)	(11.2292)	(11.1546)	(2.4785)	(11.1622)
Inv			0.0257 **
			(2.3261)
Product						0.0087
						(0.8171)
Process							0.0342
							(1.2126)
Count									0.0146
									(1.3214)
Size	2.7288 ***	3.7240 ***	2.6331 ***	−0.6094 **	−0.3823 ***	2.7341 ***	2.7419 ***	−0.8595 ***	2.7414 ***
	(27.0765)	(7.6766)	(23.5904)	(−2.0450)	(−3.5809)	(27.1668)	(27.0692)	(−3.2655)	(27.0972)
Lev	−0.1527	−0.7874	−0.1324	0.6318 *	0.2054	−0.1582	−0.1597	0.4335	−0.1590
	(−0.9141)	(−1.4116)	(−0.8040)	(1.6743)	(1.3900)	(−0.9474)	(−0.9582)	(1.2620)	(−0.9539)
Pro	−3.5026 ***	−2.2705	−3.4442 ***	−0.3482	−0.3017	−3.4995 ***	−3.4923 ***	−0.8444	−3.4902 ***
	(−4.4460)	(−1.0374)	(−4.5772)	(−0.3932)	(−0.8124)	(−4.4357)	(−4.4318)	(−0.9665)	(−4.4289)
Growth	−0.0093 **	−0.0135	−0.0089 **	−0.0001	0.0030	−0.0093 **	−0.0094 **	0.0060	−0.0094 **
	(−2.3237)	(−0.8781)	(−2.3114)	(−0.0137)	(1.1570)	(−2.3240)	(−2.3506)	(1.0132)	(−2.3485)
Inc	0.0293	0.4002 ***	0.0190	−0.1195 *	−0.0213	0.0303	0.0300	−0.0619	0.0302
	(1.1057)	(3.6533)	(0.7015)	(−1.9029)	(−0.7312)	(1.1394)	(1.1320)	(−0.8985)	(1.1384)
Dual	−0.0902 ***	−0.0527	−0.0888 ***	−0.1327	−0.0487	−0.0890 ***	−0.0885 ***	−0.1178	−0.0885 ***
	(−2.7200)	(−0.3978)	(−2.6746)	(−1.2392)	(−1.1323)	(−2.6818)	(−2.6679)	(−1.1373)	(−2.6667)
Shrcr	0.0031 ***	−0.0040	0.0032 ***	−0.0042	−0.0013	0.0032 ***	0.0032 ***	−0.0026	0.0032 ***
	(2.7684)	(−0.7158)	(2.8571)	(−1.3004)	(−1.0050)	(2.8015)	(2.8122)	(−0.8766)	(2.8074)
Zbmw	0.0055 ***	−0.0008	0.0055 ***	0.0022	0.0012 **	0.0054 ***	0.0054 ***	0.0027 *	0.0054 ***
	(10.6940)	(−0.2758)	(10.6370)	(1.5914)	(1.9874)	(10.6831)	(10.6137)	(1.8092)	(10.6275)
Ldx	−1.1292 ***	−0.4069	−1.1188 ***	0.5814	0.2588	−1.1343 ***	−1.1381 ***	0.5573	−1.1374 ***
	(−5.3644)	(−0.4647)	(−5.4882)	(1.1252)	(1.1593)	(−5.3930)	(−5.4065)	(1.0644)	(−5.4065)
Mna	−112.0752 **	595.5835 ***	−127.3880 ***	71.3323	61.4725	−112.6985 **	−114.1779 **	175.1449	−114.6375 **
	(−2.4657)	(3.5763)	(−2.8225)	(0.6867)	(1.1745)	(−2.4788)	(−2.5156)	(1.3586)	(−2.5271)
Dratio	−0.1975	0.5725	−0.2122	1.7873 *	1.0547 **	−0.2131	−0.2336	2.5492 **	−0.2348
	(−0.7350)	(0.4595)	(−0.7695)	(1.7511)	(2.3301)	(−0.7858)	(−0.8519)	(2.3107)	(−0.8576)
Cbl	−0.1052	−0.3769	−0.0955	−0.8933	−0.0915	−0.0974	−0.1021	−0.3395	−0.1002
	(−0.1411)	(−0.1647)	(−0.1352)	(−0.7729)	(−0.2263)	(−0.1303)	(−0.1367)	(−0.3465)	(−0.1342)
Fgr	−0.0530	−0.0259	−0.0523	−0.0223	0.0055	−0.0528	−0.0532	0.0199	−0.0533
	(−1.4545)	(−0.1422)	(−1.4186)	(−0.2720)	(0.1758)	(−1.4475)	(−1.4620)	(0.2448)	(−1.4650)
Com	0.0612	−0.2058	0.0665	−0.0078	0.0003	0.0613	0.0612	0.0106	0.0611
	(1.4002)	(−1.0479)	(1.5356)	(−0.1137)	(0.0081)	(1.4019)	(1.4010)	(0.1288)	(1.3976)
Sus	1.1147	4.8776 *	0.9893	−0.3201	0.0397	1.1175	1.1133	0.0261	1.1143
	(1.4470)	(1.9062)	(1.3149)	(−0.2301)	(0.0670)	(1.4509)	(1.4443)	(0.0176)	(1.4465)
Roe	2.3282 **	−1.7937	2.3743 **	0.3967	0.0851	2.3247 **	2.3252 **	0.3017	2.3237 **
	(2.4122)	(−0.6823)	(2.5308)	(0.2918)	(0.1532)	(2.4065)	(2.4048)	(0.2285)	(2.4049)
Constant	10.9422 ***	−46.4026 ***	12.1353 ***	6.8846 **	3.2431 ***	10.8821 ***	10.8313 ***	7.4463 ***	10.8333 ***
	(8.7821)	(−8.9576)	(8.8353)	(2.2552)	(2.9895)	(8.7292)	(8.6635)	(2.8555)	(8.6695)
Year dummy	YES	YES	YES	YES	YES	YES	YES	YES	YES
Ind dummy	YES	YES	YES	YES	YES	YES	YES	YES	YES
Observations	848	848	848	848	848	848	848	848	848
R−squared	0.9400	0.5725	0.9406	0.1616	0.1970	0.9400	0.9401	0.1976	0.9401
Sobel Test for Product Mediator Variables				Sobel Test for Process Mediator Variables			Sobel Test for Count Mediator Variables
z		mediating effect		z	mediating effect		z	mediating effect
0.7103		insignificant		1.0995	significant		1.1619	insignificant

Note: ***, **, and * represent statistical significance at the levels of 1%, 5%, and 10%, respectively. The numbers in parentheses are T values.

reports the regression results for the effects of the channels of resource abundance on transformation and upgrading in potential enterprises. The results revealed that potential firms with more excellent human resources and fewer financial resources invested in process innovation activities and increased innovation output, ultimately driving transformation and upgrading.

The results of Columns (1)–(3) in Table 5 related to H1 and H2 and indicated that companies with high human capital levels, even in the face of financial constraints, can ensure R&D investment through the efficient use of human capital to promote the transformation and upgrading of enterprises. Specifically, Column (1) shows a significant positive effect of Fyd on Tfp (β = 14.3935, p < 0.01), indicating that resource abundance had a stimulating effect on transformation and upgrading in potential enterprises. H1 was verified. Column (2) shows that Fyd positively impacted Inv at the 1% level, signaling that a high abundance of potential enterprises increased innovation investment. Column (3) shows that Fyd had a positive effect on Tfp at the 1% level, and Inv positively affected Tfp at the 5% level. This meant that innovation investment intensity (Inv) served as a bridge between resource abundance (Fyd) and enterprise transformation and upgrading (Tfp). This outcome coincided with that of Kong et al. [87].

The results of Columns (1), (4)–(7) in Table 5 related to H3c and indicated that potential enterprises with ample financial resources but limited human resources were prone to carry out process innovation activities rather than product innovation. Specifically, Columns (1), (4), and (6) are the regression results for resource abundance, product innovation, and enterprise transformation and upgrading. Column (1) shows that the regression coefficient of Fyd was positively and significantly associated with Tfp (β = 14.3935, p < 0.01). As seen from Columns (4) and (6), after calculation, Z = 0.7103 < 0.97. This meant that the mediating effect of product innovation did not exist. Columns (1), (5), and (7) are the regression results for resource abundance, process innovation, and enterprise transformation and upgrading. Columns (1) and (5) show that Fyd promoted Tfp and Process, respectively (β = 14.3935, p < 0.01; β = 3.0301, p < 0.01). When the mediator variable Process was added to the model in Column (7), the coefficients of Fyd on Tfp dropped from 14.3935 to 14.2899 (p < 0.01). However, the coefficient between Process and Tfp was nonsignificant (β = 0.0342, p > 0.1). The Sobel test needed to be carried out [76]. After calculation, Z = 1.0995 > 0.97, indicating that the mediating effect of process innovation (Process) on resource abundance (Fyd) and enterprise transformation and upgrading (Tfp) was significant. This was consistent with H3c. The results of this study were similar to those in Cho and Linderman [17], who suggested that companies relying heavily on knowledge-based resources tend to centralize their efforts on process innovation rather than product innovation. We extended Cho and Linderman’s findings to show that process innovation plays a partial mediating role among potential enterprises as they adapt to transform and upgrade.

The results of Columns (1) and (8)–(9) in Table 5 related to H4 and indicated that innovation output played a partial mediating role in the relationship between resource abundance and transformation and upgrading in potential enterprises. Specifically, Column (1) shows that the regression coefficient of Fyd was positively and significantly associated with Tfp (β = 14.3935, p < 0.01). In Column (8), the coefficient between Fyd and Count was significant (β = 7.2981, p < 0.05). When the mediator variable Count was added to the model in Column (9), the coefficients of Tfp decreased only slightly (from β = 14.3589–14.2867, p < 0.01). The Sobel test needed to be carried out [76]. After calculation, Z = 1.1619 > 0.97, indicating a partially mediated relationship. H4 was proven. This finding was compatible with the work of Chemmanur et al. [88], which underscored the significance of high-quality human capital.

Table 6. Identification of the mechanisms of resource abundance affecting transformation and upgrading in exiting enterprises: innovation input, innovation direction, and innovation output.

Variables	(1) Tfp	(2) Inv	(3) Tfp	(4) Product	(5) Process	(6) Tfp	(7) Tfp	(8) Count	(9) Tfp
Fyd	106.7431 ***	15.9935 ***	104.7915 ***	−5.9139	−2.2215	106.7345 ***	106.7365 ***	−8.1354	106.7312 ***
	(32.1978)	(5.2207)	(32.5094)	(−0.7125)	(−0.2091)	(32.1870)	(32.1974)	(−0.4629)	(32.1945)
Inv			0.1220 ***
			(5.6624)
Product						−0.0015
						(−0.3252)
Process							−0.0030
							(−0.8542)
Count									−0.0015
									(−0.6897)
Size	2.3403 ***	0.8453 ***	2.2372 ***	−0.3359 ***	−0.4261 ***	2.3398 ***	2.3390 ***	−0.7620 ***	2.3392 ***
	(43.9387)	(12.6110)	(38.4008)	(−2.9770)	(−2.5840)	(43.9317)	(43.8145)	(−2.9773)	(43.8434)
Lev	−0.0681	−0.2359 ***	−0.0394	0.1288	0.5907 *	−0.0680	−0.0664	0.7195	−0.0671
	(−0.7110)	(−3.3103)	(−0.4149)	(0.6752)	(1.9096)	(−0.7090)	(−0.6916)	(1.5577)	(−0.6995)
Pro	−3.3157 ***	−0.6034 ***	−3.2420 ***	−0.7332 *	−0.7443	−3.3167 ***	−3.3179 ***	−1.4775	−3.3178 ***
	(−7.5265)	(−3.8898)	(−7.4625)	(−1.8577)	(−0.9779)	(−7.5248)	(−7.5449)	(−1.4112)	(−7.5372)
Growth	−0.0035	0.0005	−0.0035 *	0.0019	0.0049	−0.0034	−0.0034	0.0068	−0.0034
	(−1.6436)	(0.2861)	(−1.7046)	(0.4538)	(0.9644)	(−1.6424)	(−1.6363)	(0.7988)	(−1.6389)
Inc	0.0424 ***	0.0389 **	0.0377 ***	−0.0131	−0.0315	0.0424 ***	0.0423 ***	−0.0446	0.0424 ***
	(2.8857)	(2.0357)	(2.5905)	(−0.3149)	(−0.5266)	(2.8843)	(2.8806)	(−0.4815)	(2.8820)
Dual	−0.0501 ***	−0.0019	−0.0499 ***	0.0583	0.0772	−0.0501 ***	−0.0499 ***	0.1355	−0.0499 ***
	(−3.1737)	(−0.1472)	(−3.1687)	(1.1496)	(1.1371)	(−3.1670)	(−3.1599)	(1.2420)	(−3.1610)
Shrcr	0.0008	0.0001	0.0008	−0.0011	−0.0014	0.0008	0.0008	−0.0025	0.0008
	(1.2791)	(0.2292)	(1.2630)	(−0.5605)	(−0.5545)	(1.2758)	(1.2722)	(−0.6093)	(1.2727)
Zbmw	0.0077 ***	−0.0010 ***	0.0078 ***	0.0009	0.0020	0.0077 ***	0.0077 ***	0.0028	0.0077 ***
	(21.4204)	(−2.8967)	(21.8479)	(1.1363)	(1.3307)	(21.4246)	(21.4187)	(1.3502)	(21.4237)
Ldx	−0.7636 ***	0.2344 ***	−0.7922 ***	0.2089	0.6305 *	−0.7633 ***	−0.7617 ***	0.8394	−0.7624 ***
	(−8.8965)	(2.7955)	(−9.1905)	(0.7505)	(1.6974)	(−8.8880)	(−8.8668)	(1.4110)	(−8.8742)
Mna	−58.7177 ***	9.5855 ***	−59.8874 ***	3.5100	9.0352	−58.7126 ***	−58.6908 ***	12.5452	−58.6993 ***
	(−21.2506)	(6.7048)	(−21.5070)	(0.4464)	(0.9580)	(−21.2412)	(−21.2459)	(0.7721)	(−21.2412)
Dratio	−0.5301 ***	0.2010 *	−0.5546 ***	0.4929	1.3084 *	−0.5294 ***	−0.5262 ***	1.8013 *	−0.5275 ***
	(−3.6179)	(1.7037)	(−3.7984)	(1.0454)	(1.8493)	(−3.6130)	(−3.5866)	(1.6678)	(−3.5976)
Cbl	−1.0628 ***	0.3331 ***	−1.1034 ***	−0.6864 **	−1.4680 ***	−1.0638 ***	−1.0671 ***	−2.1544 ***	−1.0659 ***
	(−2.7858)	(2.6865)	(−2.9468)	(−2.0337)	(−2.7384)	(−2.7866)	(−2.7987)	(−2.7090)	(−2.7937)
Fgr	−0.0067	−0.0038	−0.0063	−0.0140	0.0010	−0.0068	−0.0067	−0.0130	−0.0068
	(−0.4805)	(−0.3835)	(−0.4531)	(−0.3555)	(0.0135)	(−0.4820)	(−0.4805)	(−0.1297)	(−0.4820)
Com	0.0198	−0.0539 **	0.0264	−0.0217	0.0209	0.0198	0.0199	−0.0008	0.0198
	(0.7236)	(−2.5068)	(0.9745)	(−0.3006)	(0.1656)	(0.7221)	(0.7254)	(−0.0042)	(0.7231)
Sus	−0.3696	0.5045 *	−0.4311	0.3594	−0.1281	−0.3691	−0.3700	0.2313	−0.3692
	(−1.0026)	(1.8912)	(−1.1689)	(0.3529)	(−0.1231)	(−1.0010)	(−1.0030)	(0.1204)	(−1.0011)
Roe	4.3784 ***	0.6706 **	4.2966 ***	0.4194	−0.0298	4.3790 ***	4.3783 ***	0.3896	4.3790 ***
	(8.4750)	(1.9885)	(8.3821)	(0.4974)	(−0.0284)	(8.4743)	(8.4768)	(0.2216)	(8.4768)
Constant	14.3816 ***	−8.5918 ***	15.4300 ***	4.1448 ***	5.4524 ***	14.3877 ***	14.3978 ***	9.5972 ***	14.3957 ***
	(23.3666)	(−15.5587)	(23.6080)	(3.3205)	(3.1797)	(23.3250)	(23.3408)	(3.4701)	(23.3277)
Year dummy	YES	YES	YES	YES	YES	YES	YES	YES	YES
Ind dummy	YES	YES	YES	YES	YES	YES	YES	YES	YES
Observations	4099	4099	4099	4099	4099	4099	4099	4099	4099
R−squared	0.9191	0.4003	0.9200	0.0577	0.0441	0.9191	0.9191	0.0531	0.9191

Note: ***, **, and * represent statistical significance at the levels of 1%, 5%, and 10%, respectively. The numbers in parentheses are T values.

reports the regression results for the effects of the channels of resource abundance on transformation and upgrading in exiting enterprises. The findings revealed that for existing firms, the most effective pathway for transformation and upgrading lay in increasing investment in innovation. However, the pathways of innovation direction and innovation output did not lead to significant transformation and upgrading.

The results of testing H1 and H2 were tested with Columns (1)–(3) in Table 6. The conclusion was that resource abundance had a “push effect” on innovation investment, significantly facilitating the transformation and upgrading of exiting enterprises. Specifically, in all regressions, the coefficients of the independent variable (Fyd) with the dependent variable (Tfp) and the mediator (Inv) were significant at a confidence level of 1% (β = 106.7431, p < 0.01; β = 15.9935, p < 0.01), indicating that when an exiting firm had adequate resources, it had more incentive to conduct transformation and upgrading activities and increase innovation investment intensity. Additionally, the effect of the independent variable (Fyd) on the dependent variable was smaller in Model (3) than in Model (1) (from β = 106.7431–104.7915, p <  0.01). This meant that innovation investment intensity (Inv) exerted a partial mediating effect. That is to say, resource abundance (Fyd) promoted transformation and upgrading (Tfp) through innovation investment intensity (Inv). Based on these data, we accepted H1 and H2.

The results of testing H3d were tested with Columns (1) and (4)–(7) in Table 6. The conclusion was that exiting enterprises facing constraints in terms of human and financial resources lacked the capability to engage in innovation activities and that the mediating channel of transformation and upgrading in exiting enterprises disappeared. Specifically, Columns (1), (4), and (6) were the regression results for resource abundance, product innovation, and enterprise transformation and upgrading. Column (1) showed that the regression coefficient of Fyd was positively and significantly associated with Tfp (β = 106.7431, p < 0.01). As seen from Columns (4) and (6), the partial mediating effect of product was insignificant. Columns (1), (5), and (7) were the regression results for resource abundance, process innovation, and enterprise transformation and upgrading. Column (1) shows that the regression coefficient of Fyd was positively and significantly associated with Tfp (β = 106.7431, p < 0.01). As seen from Columns (5) and (7), the mediating effect of process innovation did not exist. This was consistent with H3d. The results were similar to those in Cho and Linderman [17], who proved that companies lacking resources tend to minimize product and process innovation activities.

The results of testing H4 were tested with Columns (1) and (8)–(9) in Table 6. The outcome revealed that the mediating effect of innovation output was not significant. H4 was validated. This indicated that resource-constrained, less innovative enterprises were relatively technologically backward, lacked competitive advantages, and had a lower success rate in transformation and upgrading. To conclude, exiting enterprises needed more financial and human resources to transform and update through increased innovation intensity.

6. Robustness Tests

To check the robustness of the overall model, variable substitution was carried out for the independent variable. Following Chen and Chang [89], we used the net profit margin as an alternative proxy for financial resources. The financial and human resources were then calculated using the entropy method to obtain the enterprise resource abundance (Ffyd).

Table 7. Robustness test for effects of channels of resource abundance on transformation and upgrading in leading enterprises.

Variables	(1) Tfp	(2) Inv	(3) Tfp	(4) Product	(5) Process	(6) Tfp	(7) Tfp	(8) Count	(9) Tfp
Ffyd	11.9392 ***	24.1511 ***	11.4327 ***	1.8808 **	4.3332 **	11.8733 ***	11.8661 ***	6.2140 **	11.8604 ***
	(18.7633)	(6.3661)	(19.9661)	(1.9729)	(2.0105)	(18.6941)	(18.6767)	(2.1363)	(18.6827)
Inv			0.0210 ***
			(3.5947)
Product						0.0351 ***
						(3.6141)
Process							0.0169 **
							(2.1027)
Count									0.0127 ***
									(2.6148)
Size	2.8310 ***	5.9526 ***	2.7061 ***	−0.4645 ***	−0.6899 **	2.8473 ***	2.8426 ***	−1.1544 ***	2.8456 ***
	(38.0408)	(12.5653)	(31.1964)	(−2.8684)	(−2.4401)	(38.4068)	(38.2626)	(−2.7048)	(38.3478)
Lev	−0.2055	−2.4396 ***	−0.1543	0.2957	0.0390	−0.2159	−0.2061	0.3347	−0.2097
	(−1.1501)	(−3.7366)	(−0.8658)	(1.4739)	(0.1063)	(−1.2101)	(−1.1575)	(0.6289)	(−1.1777)
Pro	−4.2267 ***	−7.9052 ***	−4.0609 ***	0.0325	−0.3946	−4.2278 ***	−4.2200 ***	−0.3621	−4.2221 ***
	(−6.4726)	(−4.0108)	(−6.2830)	(0.0844)	(−0.5984)	(−6.4708)	(−6.4788)	(−0.3751)	(−6.4773)
Growth	−0.0081	−0.0044	−0.0080	0.0026	0.0010	−0.0082	−0.0081	0.0036	−0.0082
	(−1.4920)	(−0.3050)	(−1.5061)	(0.6361)	(0.1438)	(−1.5119)	(−1.5011)	(0.3392)	(−1.5061)
Inc	0.0205	0.2797 ***	0.0146	−0.0606	−0.0959	0.0226	0.0221	−0.1565	0.0225
	(1.2312)	(2.8288)	(0.8775)	(−1.5357)	(−1.0831)	(1.3560)	(1.3247)	(−1.2950)	(1.3469)
Dual	−0.0674 **	0.2672	−0.0730 ***	0.0312	0.0465	−0.0685 **	−0.0681 **	0.0777	−0.0683 **
	(−2.4695)	(1.4310)	(−2.6537)	(0.5364)	(0.4658)	(−2.5176)	(−2.5076)	(0.5110)	(−2.5156)
Shrcr	0.0040 ***	−0.0028	0.0041 ***	−0.0005	−0.0023	0.0040 ***	0.0040 ***	−0.0028	0.0040 ***
	(3.5989)	(−0.4651)	(3.6769)	(−0.2664)	(−0.6899)	(3.6160)	(3.6338)	(−0.5688)	(3.6316)
Zbmw	0.0062 ***	−0.0093 ***	0.0064 ***	0.0017 **	0.0043 **	0.0061 ***	0.0061 ***	0.0060 **	0.0061 ***
	(12.8358)	(−3.1893)	(13.2640)	(1.9851)	(2.3798)	(12.7717)	(12.6771)	(2.3870)	(12.6994)
Ldx	−0.4601 ***	−1.4774 *	−0.4291 ***	0.6187	1.1838	−0.4818 ***	−0.4801 ***	1.8025	−0.4829 ***
	(−3.6768)	(−1.8143)	(−3.4786)	(1.5139)	(1.5597)	(−3.8513)	(−3.8275)	(1.5802)	(−3.8528)
Mna	−152.1573 ***	798.3378 ***	−168.9017 ***	34.7417	78.6043	−153.3754 ***	−153.4836 ***	113.3460	−153.5949 ***
	(−4.2025)	(4.9864)	(−4.7644)	(0.3923)	(0.5765)	(−4.2600)	(−4.2689)	(0.5348)	(−4.2742)
Dratio	0.1017	3.7664 ***	0.0227	−0.0601	−0.2759	0.1038	0.1064	−0.3359	0.1060
	(0.4748)	(2.8868)	(0.1082)	(−0.1458)	(−0.2850)	(0.4859)	(0.4968)	(−0.2548)	(0.4954)
Cbl	−1.4110 ***	−2.6511 *	−1.3554 ***	0.0373	−0.5277	−1.4123 ***	−1.4021 ***	−0.4904	−1.4048 ***
	(−3.7382)	(−1.7988)	(−3.6436)	(0.1055)	(−0.6838)	(−3.7504)	(−3.7525)	(−0.4623)	(−3.7535)
Fgr	0.0393	−0.2090	0.0437	−0.0255	0.1010	0.0402	0.0376	0.0756	0.0384
	(0.5880)	(−0.6917)	(0.6605)	(−0.2735)	(0.4282)	(0.6017)	(0.5619)	(0.2383)	(0.5734)
Com	−0.0321	−0.5205 ***	−0.0212	0.0235	0.0617	−0.0330	−0.0332	0.0852	−0.0332
	(−1.3099)	(−2.9440)	(−0.8768)	(0.4888)	(0.4607)	(−1.3346)	(−1.3387)	(0.4850)	(−1.3403)
Sus	−0.3365	3.4378 **	−0.4086	0.7103	0.8051	−0.3614	−0.3501	1.5154	−0.3558
	(−1.2454)	(1.9819)	(−1.5214)	(1.4113)	(0.8224)	(−1.3345)	(−1.2886)	(1.0990)	(−1.3100)
Roe	3.8723 ***	3.3482	3.8020 ***	−0.3115	−0.8419	3.8832 ***	3.8865 ***	−1.1534	3.8869 ***
	(7.9713)	(1.6238)	(7.9037)	(−0.5666)	(−0.7342)	(8.0174)	(8.0504)	(−0.7213)	(8.0477)
Constant	12.0559 ***	−61.1938 ***	13.3394 ***	5.3024 ***	8.5793 ***	11.8700 ***	11.9112 ***	13.8818 ***	11.8799 ***
	(16.5303)	(−13.3171)	(16.9721)	(3.6980)	(3.5099)	(16.3236)	(16.3490)	(3.7779)	(16.3266)
Year dummy	YES	YES	YES	YES	YES	YES	YES	YES	YES
Ind dummy	YES	YES	YES	YES	YES	YES	YES	YES	YES
Observations	1356	1356	1356	1356	1356	1356	1356	1356	1356
R−squared	0.9568	0.6964	0.9574	0.0785	0.0558	0.9570	0.9570	0.0616	0.9570

This table reports the robustness regression results for the effects of the channels of resource abundance on transformation and upgrading in leading enterprises. Columns (1)–(3) were set to explore the mediating effect of innovation investment intensity. Columns (1), (4), and (6) were set to explore the mediating effect of product innovation. Columns (1), (5), and (7) were set to explore the mediating effect of process innovation. Columns (1) and (8)–(9) were set to explore the mediating effect of innovation output. T-statistics are reported in parenthesis, where the significance is defined as * p < 0.1, ** p < 0.05, *** p < 0.01.

reports the robustness results for the effects of the channels of resource abundance on transformation and upgrading in leading enterprises. It can be seen that all regression results were still positive and significant. This meant that the effects of resource abundance on transformation and upgrading were mediated through the whole innovation process, including innovation investment intensity (Inv), product innovation (Product), process innovation (Process), and innovation output (Count). These results illustrated that the regression results in this study were robust.

Table 8. Robustness test for the effects of channels of resource abundance on transformation and upgrading in catching-up enterprises.

Variables	(1) Tfp	(2) Inv	(3) Tfp	(4) Product	(5) Process	(6) Tfp	(7) Tfp	(8) Count	(9) Tfp
Ffyd	111.1115 ***	9.6652 ***	109.8977 ***	21.7277 **	16.3970	110.9321 ***	110.9687 ***	38.1247	110.9100 ***
	(30.2767)	(2.9197)	(30.5504)	(1.9876)	(0.9840)	(30.2204)	(30.2172)	(1.5120)	(30.2059)
Inv			0.1256 ***
			(4.8362)
Product						0.0083
						(1.5751)
Process							0.0087 ***
							(2.6459)
Count									0.0053 **
									(2.4010)
Size	2.3680 ***	0.9031 ***	2.2545 ***	−0.4046 ***	−0.2539	2.3713 ***	2.3702 ***	−0.6585 **	2.3714 ***
	(44.3241)	(12.1864)	(36.3830)	(−3.1929)	(−1.4545)	(44.3733)	(44.3685)	(−2.3951)	(44.3952)
Lev	0.1416	−0.1393 **	0.1591	−0.0390	−0.1266	0.1419	0.1427	−0.1656	0.1425
	(1.3962)	(−2.1824)	(1.5936)	(−0.2326)	(−0.5555)	(1.3986)	(1.4062)	(−0.4531)	(1.4039)
Pro	−3.0033 ***	−0.8568 ***	−2.8957 ***	0.2434	0.2334	−3.0053 ***	−3.0053 ***	0.4768	−3.0058 ***
	(−7.9412)	(−5.7830)	(−7.6891)	(0.9269)	(0.6244)	(−7.9535)	(−7.9572)	(0.8120)	(−7.9590)
Growth	0.0002	−0.0025 *	0.0005	−0.0035	−0.0067	0.0002	0.0003	−0.0102	0.0003
	(0.0905)	(−1.8662)	(0.2231)	(−1.1820)	(−1.4471)	(0.1027)	(0.1148)	(−1.4430)	(0.1131)
Inc	0.0579 ***	0.0341 **	0.0536 ***	−0.0014	−0.0500	0.0579 ***	0.0583 ***	−0.0514	0.0581 ***
	(3.6810)	(2.4560)	(3.4959)	(−0.0342)	(−0.8042)	(3.6796)	(3.7079)	(−0.5492)	(3.6967)
Dual	−0.0226	0.0222	−0.0253	−0.0768	−0.0665	−0.0219	−0.0220	−0.1433	−0.0218
	(−1.3172)	(1.5942)	(−1.4802)	(−1.5179)	(−0.8741)	(−1.2793)	(−1.2835)	(−1.2328)	(−1.2727)
Shrcr	−0.0008	−0.0005	−0.0008	−0.0003	−0.0020	−0.0008	−0.0008	−0.0023	−0.0008
	(−1.2569)	(−0.8961)	(−1.1666)	(−0.1847)	(−0.7923)	(−1.2526)	(−1.2305)	(−0.5785)	(−1.2382)
Zbmw	0.0083 ***	−0.0008 ***	0.0084 ***	0.0017 **	0.0012	0.0083 ***	0.0083 ***	0.0029 **	0.0083 ***
	(23.3862)	(−2.7602)	(23.5166)	(2.3968)	(1.3337)	(23.3707)	(23.4268)	(1.9742)	(23.4090)
Ldx	−0.6300 ***	0.1238 **	−0.6456 ***	−0.1824	−0.2435	−0.6285 ***	−0.6279 ***	−0.4259	−0.6278 ***
	(−8.7180)	(2.2562)	(−9.0063)	(−0.8803)	(−0.7883)	(−8.7117)	(−8.7024)	(−0.9083)	(−8.7042)
Mna	−57.5729 ***	8.0953 ***	−58.5896 ***	4.1416	4.5450	−57.6072 ***	−57.6125 ***	8.6865	−57.6189 ***
	(−18.2614)	(5.6733)	(−18.4701)	(0.4887)	(0.4273)	(−18.2495)	(−18.2842)	(0.5001)	(−18.2673)
Dratio	−0.3235 **	0.1527	−0.3427 **	−0.1433	−0.1163	−0.3223 **	−0.3225 **	−0.2596	−0.3221 **
	(−2.0398)	(1.3311)	(−2.1836)	(−0.3076)	(−0.1699)	(−2.0330)	(−2.0355)	(−0.2466)	(−2.0331)
Cbl	−1.0909 ***	−0.1371	−1.0737 ***	−0.3982 *	−0.8328 **	−1.0876 ***	−1.0836 ***	−1.2310 **	−1.0844 ***
	(−4.2135)	(−1.2984)	(−4.2412)	(−1.6489)	(−2.2041)	(−4.2016)	(−4.1863)	(−2.1527)	(−4.1895)
Fgr	0.0271	−0.0020	0.0274	0.0538	−0.0067	0.0267	0.0272	0.0471	0.0269
	(1.3807)	(−0.1746)	(1.4013)	(0.8793)	(−0.0721)	(1.3558)	(1.3739)	(0.3330)	(1.3608)
Com	0.0194	−0.0669 ***	0.0278	0.0414	−0.0319	0.0190	0.0197	0.0095	0.0193
	(0.6927)	(−2.7712)	(1.0114)	(0.6613)	(−0.4039)	(0.6814)	(0.7031)	(0.0720)	(0.6918)
Sus	−0.5899 ***	0.4180 **	−0.6424 ***	0.6872	0.7292	−0.5956 ***	−0.5962 ***	1.4163	−0.5974 ***
	(−2.8433)	(2.2528)	(−3.1198)	(1.4289)	(1.1595)	(−2.8686)	(−2.8742)	(1.4356)	(−2.8786)
Roe	3.9005 ***	0.9567 ***	3.7803 ***	−0.6623	−0.7712	3.9060 ***	3.9072 ***	−1.4334	3.9081 ***
	(9.0519)	(4.1975)	(8.8585)	(−1.3468)	(−1.2642)	(9.0537)	(9.0530)	(−1.4445)	(9.0541)
Constant	13.6886 ***	−8.5290 ***	14.7597 ***	4.1582 ***	4.6615 ***	13.6542 ***	13.6479 ***	8.8197 ***	13.6419 ***
	(24.1860)	(−13.1914)	(24.2182)	(3.6060)	(2.8199)	(24.0847)	(24.0778)	(3.4278)	(24.0603)
Year dummy	YES	YES	YES	YES	YES	YES	YES	YES	YES
Ind dummy	YES	YES	YES	YES	YES	YES	YES	YES	YES
Observations	3406	3406	3406	3406	3406	3406	3406	3406	3406
R−squared	0.9310	0.4667	0.9317	0.0575	0.0516	0.9310	0.9311	0.0600	0.9311
Sobel Test for Product Mediator Variables				Sobel Test for Process Mediator Variables			Sobel Test for Count Mediator Variables
z		mediating effect		z	mediating effect		z	mediating effect
1.2445		significant		0.9219	insignificant		1.2806	significant

This table reports the robustness regression results for the effects of the channels of resource abundance on transformation and upgrading in catching-up enterprises. Columns (1)–(3) were set to explore the mediating effect of innovation investment intensity. Columns (1), (4), and (6) were set to explore the mediating effect of product innovation. Columns (1), (5), and (7) were set to explore the mediating effect of process innovation. Columns (1) and (8)–(9) were set to explore the mediating effect of innovation output. T-statistics are reported in parenthesis, where the significance is defined as * p < 0.1, ** p < 0.05, *** p < 0.01.

reports the robustness regression results for the effects of the channels of resource abundance on transformation and upgrading in catching-up enterprises. Based on the three-path mediation model and the Sobel test, the findings indicated that the sign direction of the coefficients of core variables remained unchanged. The results revealed that the mediating effects of innovation investment intensity, product innovation, and innovation output were still significant, but process innovation failed (Z = 0.9219 < 0.97). This meant that resource abundance affected the transformation and upgrading of manufacturing in catching-up enterprises through increasing innovation investment intensity, driving product innovation activities, and increasing innovation output. The results indicated that the findings were very reliable.

Table 9. Robustness test for the effects of channels of resource abundance on transformation and upgrading in potential enterprises.

Variables	(1) Tfp	(2) Inv	(3) Tfp	(4) Product	(5) Process	(6) Tfp	(7) Tfp	(8) Count	(9) Tfp
Ffyd	14.3678 ***	30.8332 ***	13.5745 ***	3.9548	3.0252 ***	14.3332 ***	14.2642 ***	7.2871 **	14.2611 ***
	(11.2528)	(5.0055)	(10.6020)	(1.4597)	(2.6042)	(11.2286)	(11.1540)	(2.4771)	(11.1617)
Inv			0.0257 **
			(2.3277)
Product						0.0087
						(0.8167)
Process							0.0342
							(1.2137)
Count									0.0146
									(1.3224)
Size	2.7285 ***	3.7226 ***	2.6327 ***	−0.6098 **	−0.3824 ***	2.7338 ***	2.7416 ***	−0.8597 ***	2.7411 ***
	(27.0616)	(7.6716)	(23.5815)	(−2.0455)	(−3.5811)	(27.1517)	(27.0547)	(−3.2659)	(27.0826)
Lev	−0.1528	−0.7875	−0.1326	0.6318 *	0.2054	−0.1584	−0.1599	0.4335	−0.1592
	(−0.9148)	(−1.4118)	(−0.8047)	(1.6743)	(1.3898)	(−0.9481)	(−0.9589)	(1.2619)	(−0.9547)
Pro	−3.5030 ***	−2.2705	−3.4445 ***	−0.3480	−0.3017	−3.4999 ***	−3.4926 ***	−0.8445	−3.4906 ***
	(−4.4458)	(−1.0372)	(−4.5771)	(−0.3930)	(−0.8126)	(−4.4356)	(−4.4316)	(−0.9667)	(−4.4287)
Growth	−0.0093 **	−0.0135	−0.0089 **	−0.0001	0.0030	−0.0093 **	−0.0094 **	0.0060	−0.0094 **
	(−2.3229)	(−0.8777)	(−2.3106)	(−0.0134)	(1.1571)	(−2.3232)	(−2.3499)	(1.0133)	(−2.3477)
Inc	0.0293	0.4003 ***	0.0190	−0.1195 *	−0.0213	0.0303	0.0300	−0.0619	0.0302
	(1.1062)	(3.6536)	(0.7017)	(−1.9028)	(−0.7311)	(1.1398)	(1.1325)	(−0.8983)	(1.1389)
Dual	−0.0903 ***	−0.0530	−0.0889 ***	−0.1327	−0.0487	−0.0891 ***	−0.0886 ***	−0.1179	−0.0886 ***
	(−2.7224)	(−0.3999)	(−2.6769)	(−1.2395)	(−1.1328)	(−2.6842)	(−2.6703)	(−1.1379)	(−2.6692)
Shrcr	0.0031 ***	−0.0040	0.0032 ***	−0.0042	−0.0013	0.0032 ***	0.0032 ***	−0.0026	0.0032 ***
	(2.7642)	(−0.7174)	(2.8531)	(−1.3007)	(−1.0055)	(2.7973)	(2.8080)	(−0.8772)	(2.8032)
Zbmw	0.0055 ***	−0.0008	0.0055 ***	0.0022	0.0012 **	0.0054 ***	0.0054 ***	0.0027 *	0.0054 ***
	(10.6901)	(−0.2776)	(10.6330)	(1.5917)	(1.9870)	(10.6793)	(10.6099)	(1.8089)	(10.6237)
Ldx	−1.1292 ***	−0.4067	−1.1187 ***	0.5815	0.2589	−1.1343 ***	−1.1381 ***	0.5573	−1.1374 ***
	(−5.3635)	(−0.4644)	(−5.4874)	(1.1253)	(1.1593)	(−5.3921)	(−5.4056)	(1.0644)	(−5.4056)
Mna	−112.1046 **	595.7847 ***	−127.4327 ***	71.4228	61.4751	−112.7285 **	−114.2093 **	175.1650	−114.6691 **
	(−2.4660)	(3.5770)	(−2.8231)	(0.6873)	(1.1743)	(−2.4791)	(−2.5160)	(1.3584)	(−2.5275)
Dratio	−0.1973	0.5730	−0.2120	1.7874 *	1.0548 **	−0.2129	−0.2334	2.5493 **	−0.2346
	(−0.7341)	(0.4599)	(−0.7687)	(1.7511)	(2.3302)	(−0.7850)	(−0.8512)	(2.3108)	(−0.8568)
Cbl	−0.1054	−0.3771	−0.0957	−0.8933	−0.0915	−0.0976	−0.1023	−0.3396	−0.1005
	(−0.1414)	(−0.1648)	(−0.1356)	(−0.7729)	(−0.2264)	(−0.1306)	(−0.1370)	(−0.3466)	(−0.1345)
Fgr	−0.0533	−0.0266	−0.0526	−0.0224	0.0054	−0.0531	−0.0535	0.0198	−0.0536
	(−1.4632)	(−0.1459)	(−1.4267)	(−0.2731)	(0.1737)	(−1.4562)	(−1.4706)	(0.2428)	(−1.4736)
Com	0.0611	−0.2061	0.0664	−0.0078	0.0003	0.0612	0.0611	0.0105	0.0609
	(1.3971)	(−1.0495)	(1.5327)	(−0.1143)	(0.0072)	(1.3987)	(1.3979)	(0.1279)	(1.3945)
Sus	1.1156	4.8794 *	0.9900	−0.3199	0.0399	1.1184	1.1142	0.0265	1.1152
	(1.4481)	(1.9070)	(1.3159)	(−0.2299)	(0.0673)	(1.4521)	(1.4455)	(0.0179)	(1.4476)
Roe	2.3277 **	−1.7953	2.3738 **	0.3963	0.0850	2.3242 **	2.3247 **	0.3014	2.3232 **
	(2.4121)	(−0.6829)	(2.5308)	(0.2915)	(0.1530)	(2.4064)	(2.4047)	(0.2283)	(2.4048)
Constant	10.9542 ***	−46.3714 ***	12.1472 ***	6.8899 **	3.2458 ***	10.8940 ***	10.8431 ***	7.4531 ***	10.8451 ***
	(8.7884)	(−8.9492)	(8.8427)	(2.2554)	(2.9904)	(8.7354)	(8.6696)	(2.8564)	(8.6756)
Year dummy	YES	YES	YES	YES	YES	YES	YES	YES	YES
Ind dummy	YES	YES	YES	YES	YES	YES	YES	YES	YES
Observations	848	848	848	848	848	848	848	848	848
R−squared	0.9399	0.5725	0.9406	0.1616	0.1970	0.9400	0.9401	0.1976	0.9401
Sobel Test for Product Mediator Variables				Sobel Test for Process Mediator Variables			Sobel Test for Count Mediator Variables
z		mediating effect		z	mediating effect		z	mediating effect
0.7103		insignificant		1.0994	significant		1.1617	significant

This table reports the robustness regression results for the effects of the channels of resource abundance on transformation and upgrading in potential enterprises. Columns (1)–(3) were set to explore the mediating effect of innovation investment intensity. Columns (1), (4), and (6) were set to explore the mediating effect of product innovation. Columns (1), (5), and (7) were set to explore the mediating effect of process innovation. Columns (1) and (8)–(9) were set to explore the mediating effect of innovation output. T-statistics are reported in parenthesis, where the significance is defined as * p < 0.1, ** p < 0.05, *** p < 0.01.

reports the robustness regression results for the effects of the channels of resource abundance on transformation and upgrading in potential enterprises. In a similar vein, the results revealed that the mediating effects of innovation investment intensity, process innovation, and innovation output were still significant, but product innovation failed (Z = 0.7103 < 0.97). This meant that resource abundance affected the transformation and upgrading of manufacturing in potential enterprises through increasing innovation investment intensity, driving process innovation activities, and increasing innovation output. This indicated that the research conclusion of this paper was robust.

Table 10. Robustness test for the effects of channels of resource abundance on transformation and upgrading in exiting enterprises.

Variables	(1) Tfp	(2) Inv	(3) Tfp	(4) Product	(5) Process	(6) Tfp	(7) Tfp	(8) Count	(9) Tfp
Ffyd	106.5453 ***	15.9880 ***	104.5945 ***	−5.8923	−2.2310	106.5367 ***	106.5453 ***	−8.1233	106.5334 ***
	(32.1752)	(5.2225)	(32.4849)	(−0.7107)	(−0.2102)	(32.1644)	(32.1752)	(−0.4627)	(32.1719)
Inv			0.1220 ***
			(5.6575)
Product						−0.0015
						(−0.3266)
Process							−0.0030
							(−0.8535)
Count									−0.0015
									(−0.6899)
Size	2.3391 ***	0.8450 ***	2.2360 ***	−0.3359 ***	−0.4260 ***	2.3387 ***	2.3391 ***	−0.7619 ***	2.3380 ***
	(43.8754)	(12.6075)	(38.3456)	(−2.9755)	(−2.5815)	(43.8684)	(43.8754)	(−2.9750)	(43.7804)
Lev	−0.0682	−0.2359 ***	−0.0394	0.1288	0.5907 *	−0.0680	−0.0682	0.7195	−0.0671
	(−0.7107)	(−3.3106)	(−0.4148)	(0.6752)	(1.9096)	(−0.7087)	(−0.7107)	(1.5577)	(−0.6993)
Pro	−3.3173 ***	−0.6035 ***	−3.2437 ***	−0.7330 *	−0.7443	−3.3184 ***	−3.3173 ***	−1.4774	−3.3195 ***
	(−7.5259)	(−3.8908)	(−7.4619)	(−1.8574)	(−0.9780)	(−7.5242)	(−7.5259)	(−1.4111)	(−7.5367)
Growth	−0.0035	0.0005	−0.0035 *	0.0019	0.0049	−0.0034	−0.0035	0.0068	−0.0034
	(−1.6437)	(0.2859)	(−1.7047)	(0.4538)	(0.9645)	(−1.6425)	(−1.6437)	(0.7988)	(−1.6390)
Inc	0.0424 ***	0.0389 **	0.0376 ***	−0.0131	−0.0315	0.0424 ***	0.0424 ***	−0.0446	0.0423 ***
	(2.8821)	(2.0354)	(2.5869)	(−0.3149)	(−0.5266)	(2.8807)	(2.8821)	(−0.4815)	(2.8784)
Dual	−0.0502 ***	−0.0019	−0.0499 ***	0.0583	0.0772	−0.0501 ***	−0.0502 ***	0.1355	−0.0500 ***
	(−3.1751)	(−0.1480)	(−3.1700)	(1.1496)	(1.1372)	(−3.1683)	(−3.1751)	(1.2420)	(−3.1624)
Shrcr	0.0008	0.0001	0.0008	−0.0011	−0.0014	0.0008	0.0008	−0.0025	0.0008
	(1.2733)	(0.2279)	(1.2573)	(−0.5605)	(−0.5545)	(1.2700)	(1.2733)	(−0.6092)	(1.2669)
Zbmw	0.0077 ***	−0.0010 ***	0.0078 ***	0.0009	0.0020	0.0077 ***	0.0077 ***	0.0028	0.0077 ***
	(21.4121)	(−2.8952)	(21.8390)	(1.1367)	(1.3305)	(21.4163)	(21.4121)	(1.3503)	(21.4153)
Ldx	−0.7636 ***	0.2344 ***	−0.7922 ***	0.2090	0.6304 *	−0.7633 ***	−0.7636 ***	0.8394	−0.7624 ***
	(−8.8930)	(2.7961)	(−9.1864)	(0.7506)	(1.6973)	(−8.8845)	(−8.8930)	(1.4110)	(−8.8706)
Mna	−58.7568 ***	9.5860 ***	−59.9265 ***	3.5150	9.0324	−58.7517 ***	−58.7568 ***	12.5474	−58.7384 ***
	(−21.2607)	(6.7023)	(−21.5166)	(0.4470)	(0.9578)	(−21.2512)	(−21.2607)	(0.7723)	(−21.2513)
Dratio	−0.5308 ***	0.2009 *	−0.5553 ***	0.4929	1.3084 *	−0.5301 ***	−0.5308 ***	1.8013 *	−0.5282 ***
	(−3.6216)	(1.7032)	(−3.8020)	(1.0456)	(1.8493)	(−3.6168)	(−3.6216)	(1.6679)	(−3.6014)
Cbl	−1.0632 ***	0.3330 ***	−1.1039 ***	−0.6863 **	−1.4680 ***	−1.0642 ***	−1.0632 ***	−2.1543 ***	−1.0664 ***
	(−2.7853)	(2.6862)	(−2.9462)	(−2.0336)	(−2.7384)	(−2.7860)	(−2.7853)	(−2.7089)	(−2.7932)
Fgr	−0.0068	−0.0038	−0.0064	−0.0140	0.0010	−0.0069	−0.0068	−0.0130	−0.0068
	(−0.4863)	(−0.3848)	(−0.4588)	(−0.3554)	(0.0135)	(−0.4877)	(−0.4863)	(−0.1296)	(−0.4878)
Com	0.0199	−0.0539 **	0.0265	−0.0217	0.0209	0.0199	0.0199	−0.0008	0.0199
	(0.7260)	(−2.5067)	(0.9769)	(−0.3007)	(0.1656)	(0.7245)	(0.7260)	(−0.0042)	(0.7255)
Sus	−0.3718	0.5044 *	−0.4334	0.3596	−0.1282	−0.3713	−0.3718	0.2314	−0.3715
	(−1.0081)	(1.8905)	(−1.1742)	(0.3531)	(−0.1232)	(−1.0064)	(−1.0081)	(0.1205)	(−1.0066)
Roe	4.3819 ***	0.6709 **	4.3000 ***	0.4191	−0.0298	4.3825 ***	4.3819 ***	0.3894	4.3825 ***
	(8.4766)	(1.9894)	(8.3835)	(0.4971)	(−0.0284)	(8.4759)	(8.4766)	(0.2215)	(8.4784)
Constant	14.4461 ***	−8.5808 ***	15.4931 ***	4.1419 ***	5.4503 ***	14.4521 ***	14.4461 ***	9.5922 ***	14.4602 ***
	(23.4248)	(−15.5302)	(23.6601)	(3.3126)	(3.1720)	(23.3833)	(23.4248)	(3.4617)	(23.3859)
Year dummy	YES	YES	YES	YES	YES	YES	YES	YES	YES
Ind dummy	YES	YES	YES	YES	YES	YES	YES	YES	YES
Observations	4099	4099	4099	4099	4099	4099	4099	4099	4099
R−squared	0.9190	0.4004	0.9199	0.0577	0.0441	0.9190	0.9190	0.0531	0.9190

This table reports the robustness regression results for the effects of the channels of resource abundance on transformation and upgrading in exiting enterprises. Columns (1)–(3) were set to explore the mediating effect of innovation investment intensity. Columns (1), (4), and (6) were set to explore the mediating effect of product innovation. Columns (1), (5), and (7) were set to explore the mediating effect of process innovation. Columns (1) and (8)–(9) were set to explore the mediating effect of innovation output. T-statistics are reported in parenthesis, where the significance is defined as * p < 0.1, ** p < 0.05, *** p < 0.01.

reports the robustness regression results for the effects of the channels of resource abundance on transformation and upgrading in exiting enterprises. According to the three-path mediation model and after calculation, the results showed that the sign direction and significance level of the coefficients of the core variable had not changed. Exiting enterprise resource abundance still significantly increased innovation intensity to promote transformation and upgrading, but low levels of resources as appeared to be insufficient for successful transformation and upgrading through product innovation, process innovation, and innovation output. Above all, the conclusions of this paper were robust and reliable, supporting all hypotheses.

7. Conclusions and Discussion

7.1. Conclusions

Based on the theories of resources and innovation perspectives, this paper took listed manufacturing enterprises from 2015 to 2020 as research samples to analyze the direct effect of resource abundance on enterprise transformation and upgrading and the indirect effects on them through technological innovation.

The main conclusions are as follows:

(1)

Resource abundance has a significant positive impact on enterprise transformation and upgrading. H1 was verified. The result agreed with [81].

(2)

Regarding innovation input, innovation investment intensity serves as a crucial pathway for manufacturing companies to achieve transformation and upgrading. H2 was verified. The result was aligned with Peng et al. [82].

(3)

Regarding the innovation process, four types of enterprises with different resource statuses are key factors influencing the innovation model and the success of transformation and upgrading. More specifically, leading enterprises with adequate resources can undertake product and process innovation simultaneously to transform and upgrade. In contrast, catching-up enterprises drive transformation and upgrading through product innovation, while potential enterprises do so through process innovation. Meanwhile, exiting enterprises are unable to transform and upgrade through product or process innovation. H3a, H3b, H3c, and H3d were verified. The outcomes coincided with those of Ettlie [50] and Gagnon [90].

(4)

Regarding innovation output, with the exception of exiting enterprises, all remaining manufacturing enterprises can achieve transformation and upgrading by increasing their innovation output. H4 was verified. The result agreed with that of Chemmanur et al. [88].

7.2. Discussion

In the era of the digital economy, how to promote transformation and upgrading and gain competitive advantages is becoming an important issue for enterprises. Our research has several theoretical and practical implications.

7.2.1. Theoretical Implications

The theory implications of this paper are as follows:

(1) Existing research has highlighted the significant role of firm resources in maintaining competitive advantage [11] and innovation capabilities [91], but ignored the impact of different types of firms’ resources on transformation and upgrading. To fill this research gap, this paper categorized manufacturing enterprises into four types and incorporated them into a unified theoretical model, exploring the relationships between enterprise resources and transformation and upgrading unique to each type. This played a significant theoretical role in uncovering whether resource abundance stimulates enterprise transformation and upgrading, and why it does so, as well as the heterogeneity existing among different types of enterprises.

More specifically, resource-advantage theory posits that the heterogeneous market positions of firms result from heterogeneity in their resources and capabilities [92]. Firms with superior resources and the ability to effectively deploy them in value-creating strategies achieve better financial performance [93]. Additionally, Barney and Ketchen [94] think that resource-based theory can be used by poorly performing firms to achieve competitive parity by learning and imitating the resources and capabilities of successful companies. From this, by classifying enterprises into different types for comparison, the revealed strengths, weaknesses, and characteristics of each type can enhance the competitiveness of enterprises and improve the quality of decision making. Our research expands the existing literature examining the impact of firm resources on various organizational economic consequences [95,96], which provides a more comprehensive understanding of transformation and upgrading in the manufacturing industry and serves as a valuable reference for strategic planning in the manufacturing sector, ultimately fostering high-quality development.

(2) Existing studies lack a comprehensive explanation of the mechanisms through which resource abundance influences transformation and upgrading. To bridge this research gap, our study highlighted the important value of technological innovation activities in stimulating the transformation and upgrading of enterprises and it conducted theoretical deduction on the varying “bridge roles” of technological innovation at different stages. This answers how resource abundance within enterprises facilitates transformation and upgrading through technological innovation and explicitly identifies the mediating effects of technological innovation in each step of the whole innovation process.

In particular, most studies have taken a narrow perspective by focusing solely on a specific stage of enterprise technological innovation [17,97,98]. We adopted a holistic approach and examined the complete innovation process, aiming to analyze the mechanisms by which firms’ resource endowments contribute to their transformation and upgrading through three distinct channels of technological innovation: innovation input, innovation modes, and innovation output. This study not only enriches the theoretical research on the intrinsic influence mechanisms of technological innovation [99,100], but also complements the existing transformation and upgrading paths [101,102], which assist in establishing a “fast track” to transformation.

(3) The existing literature has demonstrated that sustainability is intricately connected to transformation [103] and emphasized the importance of adopting green technology innovation for the smooth transformation and upgrading of manufacturing and the long-term sustainability of the modern economy [104,105,106]. In order to verify the implementation process and result value of corporate innovation behavioral decisions, our paper presented a theoretical framework that examined the intrinsic logical mechanisms of enterprise resource endowment affecting transformation and upgrading from the perspectives of innovation, which offers a new perspective on the issue of sustainable transformations of manufacturing. The findings of our study, which revealed that robust resource support and innovative technological change serve as sources of vitality for enterprise transformation and upgrading, acting as crucial prerequisites for sustainable development [107], contribute to a deeper understanding of enterprise transformation and upgrading in the pursuit of sustainability.

7.2.2. Practical Implications

The practical implications of this paper are as follows:

(1) Nurturing the concept of transformation and fostering innovation awareness is essential for accelerating the pace of transformation and upgrading. Managers of manufacturing enterprises should prioritize the strategic idea of the sustainable development of manufacturing transformation, leverage the unique characteristics of these strategic resources, select the most optimal innovation solutions, address resource–innovation mismatches, and form a manufacturing model that saves resources and steadily innovates to ultimately achieve high-quality transformation. For example, adopting a product innovation model is recommended for catching-up companies and a process innovation model for potential enterprises. Nevertheless, leading companies equipped with abundant resources should utilize their advantages to strengthen the synergy of the two innovation types, smoothly accomplishing transformation and upgrading.

(2) Promoting collaboration and knowledge exchange is crucial for driving the transformation and upgrading of the manufacturing industry. For the whole manufacturing industry, transformation and upgrading are crucial for the advancement of modern manufacturing and serve as the direction for future development [108]. However, there exist disparities in the level of transformation and upgrading among different types of manufacturing companies, and while some large enterprises have taken the lead, the outcomes are still uncertain. Many smaller manufacturing enterprises find themselves at a critical juncture and require the experience and guidance of industry leaders. Therefore, it is essential to strengthen communication and cooperation between manufacturing enterprises, to allow representatives of some transforming enterprises in the industry to play a leading role in communication, and to guide other manufacturing companies in the industry to engage in dialogue. By doing so, the full potential of transformation and upgrading in the manufacturing industry can be realized, fostering a robust mechanism for collaboration and communication.

(3) Establishing a comprehensive policy system is indispensable for facilitating the transformation and upgrading of the manufacturing industry. The government should actively implement government subsidies and preferential innovation policies, build a platform to help enterprises gather capital, recruit talent, and introduce technology, and formulate a sound policy system that aligns with the specific needs of different types of manufacturing enterprises, thus facilitating their transformation and upgrading processes. For example, an advisable approach would be to adopt supervisory regulations for leading enterprises while utilizing incentive mechanisms for other enterprises.

7.2.3. Limitations and Prospects

This study has limitations that provide avenues for future research.

(1) Due to the complexity and uncertainty of the transformation environment of the manufacturing industry, the moderating effects need to be further expanded upon. According to Zhang et al. [109], the institutional environment is a crucial reference point for organizational behavior, decision making, and strategy formulation. Therefore, to advance our understanding of the relationship between resources and transformation, future research needs to examine moderating variables, such as government governance, market regulation, and legalization level.

(2) Given the data availability, this study focused on listed Chinese manufacturing enterprises as the primary source of data, which may limit the generalizability of the findings to non-listed companies and manufacturing firms in other countries. It is worth noting that the structure of the manufacturing industry varies across different national institutional contexts [110], consequently influencing organizational innovation strategies. Therefore, it is recommended to include samples from other countries to test the research questions and enable comparative analyses.

(3) In the current digital economy era, data elements, as new modern factors, release huge value potential far beyond traditional factors. It helps to improve the existing problems of traditional factors of production and promote the new development of traditional factors. Specifically, empowering capital and labor with data improves the efficiency of using traditional factors, optimizes the allocation of factor resources, and acts as a significant driving force for the transformation and upgrading of Chinese manufacturing enterprises. Although the importance of data resources has been discussed in detail [111,112], few empirical studies have explored the critical role of data resources in transformation and upgrading, which is one of the limitations of this study. Therefore, future research can use complementarity theory to explore the impact of the mechanism of combining data elements with traditional factors of production on the transformation and upgrading of enterprises.