The Effect of Corporate Venture Capital on Labor Income Share: Evidence from China

Abstract

This study examines the impact of corporate venture capital (CVC) on the labor income share of science and innovation enterprises, focusing on data from China’s Science and Technology Innovation Board (STIB) and Growth Enterprise Market (GEM) between 2010 and 2022. Empirical results reveal a significant inverted U-shaped relationship between CVC shareholding and the labor income share of invested firms. CVC increases the labor income share by enhancing corporate governance, encouraging digital transformation, and improving human capital quality, but this effect diminishes when CVC shareholding exceeds a certain threshold. The moderating role of media attention and the heterogeneity of this relationship across regions and financial conditions are further explored. Additionally, the study identifies a positive U-shaped connection between CVC shareholding and the corporate pay gap, highlighting CVC’s complex role in influencing income inequality within firms. This research contributes to the literature by unveiling the nonlinear effects of CVC on income distribution, offering new insights into its dual role in promoting innovation and equity. Practically, it provides actionable recommendations for firms to optimize CVC ownership and for policymakers to design targeted interventions that address regional and financial disparities. By bridging the gap between CVC investment strategies and labor income fairness, this study lays the foundation for a balanced approach to sustainable economic development.

1. Introduction

One of the fundamental assumptions underlying the theory of economic growth is the well-known “Kaldor’s fact”, which postulates that the proportions of capital and labor remain constant over time (Kaldor, 1961). However, this assumption has been increasingly challenged in recent decades, as nations across the globe experience significant declines in labor income shares. For example, Chi et al. (2020) found that labor shares have decreased in 27 countries since the 1990s, with an average drop of approximately 4% in advanced economies. Similarly, in China, the labor share has significantly declined since the 1990s, despite some recovery since 2008 (Bai & Qian, 2010; Luo & Zhang, 2010). This persistent decrease distorts economic growth models, exacerbates income inequality, and increases the potential for social instability (Chen et al., 2013). Given that labor income constitutes a substantial portion of household earnings, raising the labor income share is critical for promoting equitable income distribution, boosting domestic demand, and fostering sustainable economic growth.

The Chinese government has recognized the importance of addressing this issue and has implemented a series of policies to increase the labor income share, particularly in the entrepreneurial sector. These initiatives include raising the minimum wage, providing financial support for entrepreneurship, and promoting technological innovation. Such efforts aim to optimize income distribution, enhance employment opportunities, and stimulate market vitality, thereby contributing to both social stability and economic development.

Amid these challenges, Corporate Venture Capital (CVC) has emerged as a significant force in modern financial systems, with its investment activities extending beyond mere financial support to foster innovation, expand markets, and create employment opportunities (Dushnitsky et al., 2021). CVC decision making is deeply influenced by the investor’s industrial background and strategic needs, and it tends to establish a long-term cooperative relationship with investee firms to promote technological innovation and market development jointly through in-depth integration and resource synergies rather than focusing only on short-term financial returns (Dushnitsky et al., 2021; Gong et al., 2025). Recent studies have also highlighted the significance of corporate governance factors in CVC decision making. For instance, Brinette et al. (2024) found that the educational background of directors in French firms significantly impacts their propensity to engage in CVC activities, with directors from elite institutions being more likely to drive CVC investments. Moreover, CVC plays a multifaceted role in the economy by facilitating technology transfer and innovation cooperation, thereby enhancing investors’ technological innovation capacity (Chemmanur et al., 2014; Schückes et al., 2024). It also helps investors capture emerging market opportunities (Maula et al., 2012) and diversify business risks (Dushnitsky & Lenox, 2006). Additionally, CVC improves the transparency and efficiency of internal governance in entrepreneurial firms by affecting their governance structure (Schildt et al., 2005). CVC often targets high-growth technology and innovation-driven firms, directly influencing their ability to offer competitive compensation, thereby impacting the labor income share. However, the role of CVC in shaping labor income dynamics remains underexplored, raising critical questions about its potential to address—or exacerbate—the declining labor income share. Furthermore, with the rapid development of CVC in China and its critical role in driving industrial upgrading, understanding its impact on labor income distribution is both timely and essential. While this study focuses on the Chinese context, the findings have broader implications for economies worldwide, particularly those experiencing similar challenges in labor income distribution and innovation-driven growth. This study aims to address this gap by examining the impact of CVC on the labor income share of science and innovation enterprises, using data from China’s Science and Technology Innovation Board (STIB) and Growth Enterprise Market (GEM) from 2010 to 2022.

CVC’s influence extends beyond direct financial contributions. By restructuring economic activities, integrating resources, and accelerating the conversion of technological innovation into productivity, CVC has become a key driver of economic transformation (Chemmanur et al., 2014). While CVC-backed firms often benefit from improved employment prospects and diversified income sources, scholars have also noted potential downsides. Excessive CVC shareholding may bias firms toward short-term decision making, complicate management practices, and foster labor-substituting technological innovations, potentially reducing labor income shares (Katila et al., 2008). Consequently, whether and how CVC affects the labor income share, particularly in entrepreneurial firms, is a question of critical importance. In China, where the government actively supports CVC through policies such as the “Opinions on Further Promoting the High-Quality Development of Innovation and Entrepreneurship”, this research aligns closely with national policy priorities.

In exploring how CVC impacts labor income share, the role of media cannot be overlooked. Media serves not only as a channel for information dissemination but also as a key force in shaping public opinion and corporate reputation. For firms receiving CVC investments, increased external attention on their operations, management practices, and social responsibility becomes inevitable. Positive media coverage can enhance a firm’s public image, improve employee satisfaction and loyalty, and promote harmonious development (Wu et al., 2022). Conversely, negative reports may amplify public scrutiny of corporate practices, affecting employee morale and performance. This dual role raises a critical question: does media coverage influence labor income distribution strategies in CVC-backed firms, and if so, how?

This study contributes to the existing literature in three key ways: (1) New perspectives on labor income share in technology-driven enterprises: Previous studies have largely focused on state-owned or large enterprises, with limited attention to the labor income share in science and technology firms. This study uses a CVC lens to explore how industrial capital can enhance labor income shares in innovation-driven businesses. (2) Insights into the financial effects of CVC: While prior research primarily examines CVC’s impact on firm performance, this study investigates its role in shaping internal income distribution, particularly in technology-intensive firms. (3) Mechanisms and policy implications: By analyzing the fundamental mechanisms through which CVC impacts labor income shares—such as human capital quality, corporate governance, and media influence—this study provides actionable insights for fostering a mutually beneficial ecosystem between industrial capital and entrepreneurial firms. These insights are not only relevant to China but also offer valuable lessons for economies globally, where balancing innovation-led growth with equitable income distribution remains a critical policy challenge.

The remainder of this paper is organized as follows. Section 2 reviews the relevant literature and develops the research hypotheses. Section 3 describes the sample selection, data sources, and variable definitions. Section 4 presents the empirical results and conducts robustness tests. Section 5 provides a mechanism analysis to explore the channels through which CVC affects the labor income share. Section 6 discusses the moderating effects of media attention and the impact of CVC on the pay gap. Finally, Section 7 concludes the study and offers policy implications and recommendations.

2. Literature Review and Research Hypotheses

2.1. Literature Review

2.1.1. A Study of the Factors Influencing the Labor Income Share

In addition to representing workers’ advantages in economic activity, the labor income share has a high correlation with social income distribution equality. In recent years, the percentage of worker income has been declining globally, largely due to shifts in the global economic structure and the influence of technological advancements. This topic has attracted considerable attention and in-depth analysis from academic researchers.

Technological advancements, particularly those in automation and artificial intelligence, have had a significant effect on labor income shares (Charalampidis, 2020). As a consequence of the maintenance of wage growth rates below those of labor productivity, increased automation has resulted in a decline in the labor income share. However, the creation of new technologies can partially offset this negative effect, as new tasks increase the demand for labor (Qiao et al., 2024), thus helping to maintain or increase the labor income share. A further crucial factor affecting the proportion of income derived from work is the transformation of the industrial structure. The transfer of inputs of production between various sectors increases with technical advancement and capital deepening (S. Zhang et al., 2022), which impacts the shift in the overall labor income share. As the industrial sector requires greater capital input than the agricultural sector does, labor from the latter is drawn into the former during the process of industrialization. This results in a reduction in the overall labor income share (Ibarra & Ros, 2019). Conversely, as workers transitioned from the industrial sector to the relatively labor-intensive service sector in the postindustrial era, the overall labor income share experienced a resurgence. It is crucial to examine the impact of financialization. The expansion of financial markets and the increased importance of financial capital have changed the pattern of distribution of wages and profits. Nonfinancial corporations’ higher dividends and interest payments are the reasons for the decrease in the labor income share (Dunhaupt, 2017). The negative trend was further exacerbated by the short-term focus of corporate management and the pursuit of maximizing shareholder profit during financialization, which reduced workers’ negotiating leverage. In addition, the decline in government activities and the rise of the financial sector have changed the sectoral composition of the economy, leading to a decline in the overall labor income share. In addition to the main research factors mentioned above, other studies have explored the impact of tax incentives (B. Li et al., 2021), globalization and labor’s bargaining power (Karabarbounis & Neiman, 2013) on the share of income.

2.1.2. A Study of the Economic Consequences of CVC

Existing studies have explored the effects of CVC on investors in multiple dimensions of the economic consequences for investors. In terms of innovation effectiveness, CVC facilitates technology transfer and innovation cooperation and enhances investors’ technological innovation capacity (Chemmanur et al., 2014; Schückes et al., 2024). CVC can help investors capture emerging market opportunities (Maula et al., 2012) and diversify business risks (Dushnitsky & Lenox, 2006). As the information disclosure system is enhanced, the capital market’s response to CVC illustrates that it enhances the transparency of investment projects, reinforces market confidence in the firms that are invested (Mohamed & Schwienbacher, 2016; Kang et al., 2017), and optimizes the visibility and reputation of investors in the capital market (Kim et al., 2017). CVC facilitates the construction of ecological value networks by investors and strengthens interfirm interactions and cooperation (Meng et al., 2023; Cai et al., 2022; Keil et al., 2010).

The impact of CVC on entrepreneurial firms is reflected mainly in its empowerment and governance effects. The empowerment effect is manifested in the fact that CVC provides entrepreneurial firms with financial support and management experience, which enhances their strength (Hallen et al., 2013), helps them expand their market channels and increases their market valuation (Basu et al., 2011). However, CVC may trigger moral hazards and conflicts of interest, affecting the autonomous decision making of startups and the development of independent brands (Katila et al., 2008; Qiao et al., 2024). In addition, CVC improves the transparency and efficiency of internal governance by affecting the entrepreneurial firm’s governance structure (Schildt et al., 2005). Similarly, Ed-Dafali et al. (2022) highlighted the significance of the governance relationship between capital investors and entrepreneurs, indicating that interaction frequency and trust can enhance the performance of VC-backed firms. This suggests that the governance role of CVC is not only limited to internal structural adjustments but also extends to the interactive dynamics between investors and entrepreneurs, thereby providing a more comprehensive understanding of how CVC influences entrepreneurial firms. However, excessive investor intervention can lead to governance conflicts and internal frictions that affect the normal operation and development of the firm (Dushnitsky & Lenox, 2006).

A survey of the literature mentioned above indicates that although earlier studies have examined the variables that could affect the labor income share from different angles, more research is needed to understand the connection between CVC—a significant source of direct financing—and income distribution. Specifically, prior research has predominantly focused on the macro-level impacts of CVC on innovation and market dynamics, yet its micro-level influence on labor income distribution within firms remains underexplored. The majority of existing studies have emphasized the positive roles of CVC in fostering technological advancements and enhancing corporate governance, while insufficient attention has been devoted to the potential negative consequences, such as how excessive CVC involvement might lead to labor-substituting technological innovations or exacerbate income inequality within firms. Moreover, the nonlinear relationship between CVC and labor income share, as well as the moderating effects of contextual factors like media attention, have not been adequately addressed in the literature. To bridge these gaps, this study delves into the intricate relationship between CVC and labor income share, aiming to uncover the underlying mechanisms and provide a more nuanced understanding of CVC’s dual role in shaping income distribution in the context of science and innovation enterprises. By incorporating variables such as the proportion of R&D personnel, the degree of digitalization, and the level of corporate governance as mediators, and examining the moderating role of media attention, this research seeks to offer novel insights and contribute to a more comprehensive theoretical framework in this domain.

2.2. Theoretical Analysis and Research Hypotheses

2.2.1. Impact of CVC Holdings on the Labor Income Share

The labor income share, which is the labor pay component of the total factor compensation, is a key indicator for evaluating a company’s income distribution. According to the resource-based view theory, the competitive advantage of an enterprise comes from its unique resources and capabilities. CVC, as a strategic investment behavior of enterprises, may provide investees with more targeted input resources, drawing upon the parent company’s resources and expertise. If CVC owns more equity in the investee, the interests of the parent company and the investee will converge, and it is more likely to provide technical and resource support for the investee and reduce the technical uncertainty of the investee’s R&D investment (Dushnitsky et al., 2021). Considering a firm’s ability to pay, the parent or affiliated firms behind CVC usually have higher productivity, and they are also able to provide higher levels of wage compensation. In addition, the entry of CVC increases competition in the market, and whether it is to enhance the firm’s competitiveness and incentivize employees or if the firm is competing for talent (Decreuse & Maarek, 2015), CVC investors have incentives to increase compensation.

While the investee may be able to leverage the parent company’s complementary assets to create more value, this does not necessarily guarantee that the investee will derive more value from them. Resource dependence theory suggests that organizations depend on other organizations in order to obtain key resources, and that this dependence affects the organization’s behavior and decision making. Owing to the investee’s high dependence on the parent company’s resources, the two parties may be in an unequal position and, as the CVC’s holdings increase, the parent company tends to engage in potentially opportunistic behaviors, such as stealing the investee’s core technology, if the investee is unable to protect its interests through environmental conditions or deal structures (Benson & Ziedonis, 2010). Second, investors generally have no incentive to nurture the independent commercialization capabilities of investee firms but rather use complementary assets to directly support the activities of investee firms. CVC may place more emphasis on linking the technological innovation outputs of investee firms to the products of their parent firms than on solving the problem from the perspective of the investee firms themselves (Tian & Ye, 2018), thus reducing the incentives of investee firms concerning business innovation activities. While hurting the interests of the investee firms, this may also influence the firm’s labor income share. Additionally, CVC investment involvement encourages capital-biased technical advancement, which causes capital to continually replace labor and, eventually, reduces the labor income share (Y. S. Gao et al., 2024). Therefore, the following hypotheses are proposed:

H1. 

The labor income share of investment firms and CVC holdings have a substantial inverted U-shaped relationship.

2.2.2. Examining How the Labor Income Share Is Impacted by CVC Shareholding

According to the human capital theory, the human capital of an enterprise, especially high-quality R&D personnel, is a key factor in creating value and increasing its share of labor income. CVC, as an important source of funding for enterprises, can help them attract and retain highly skilled personnel, especially R&D personnel, and thus enhance their human capital. CVC supports enterprises’ innovation through the provision of cutting-edge information on technological research and development, professional advice, and the dispatch of R&D personnel. personnel, etc., to support enterprise innovation and increase the demand for high-quality labor and salary expenses. However, when CVC ownership reaches a certain level, firms may tend to pursue short-term financial performance and shareholder returns, reducing investment in high-risk, long-cycle R&D projects in favor of investments that are stable and can achieve returns in the short term. This strategic shift prompts firms to tightly control their cost structure, including limiting the growth of R&D personnel compensation, which negatively affects the share of labor income.

CVC can provide investee enterprises with cutting-edge information on technologies and products and provide professional advice to investee enterprises, assist enterprises in developing and managing innovative projects, and dispatch R&D personnel to enterprises (Braune et al., 2019). This process increases the demand for labor and expends more labor compensation. The process of transforming high-tech achievements promoted by CVC leads to a gradual increase in the demand for high-tech R&D personnel, leading to a sharp rise in high-tech labor costs. Compared with traditional fixed investment, enterprise innovation relies heavily on high-quality talent. With the increase in enterprise innovation ability, competition, demand, and reliance on high-quality labor will greatly increase (Ye et al., 2022), which leads enterprises to take the initiative to increase monetary compensation to attract and retain high-quality labor. Furthermore, the dearth of highly skilled laborers will afford the company’s workforce greater bargaining power, leading to an increase in wages and, ultimately, an expansion of the labor income share.

It can be theorized that an increase in the percentage of employees engaged in research and development (R&D) roles at high remuneration levels may encourage a corresponding rise in the proportion of labor income. This might act as a proxy for how much emphasis a business places on innovation. However, it is important to note that this effect may be subject to variation when the CVC shareholding is high. Specifically, when the proportion of CVC ownership reaches a certain level, a firm’s business strategy may be tilted toward short-term financial performance and immediate shareholder returns (Hill & Birkinshaw, 2006), and accordingly, it will reduce the resources tilted toward high-risk, long-cycle R&D projects and shift to a preference for investments that are highly stable and with short-term realizable returns. This strategic shift has prompted firms to tightly control their cost structure, which inevitably involves scaling back funding for R&D personnel. In the context of high CVC shareholding, firms may limit the growth of R&D personnel compensation in pursuit of short-term financial health. While an increase in the proportion of employees engaged in research and development (R&D) may be necessary to enhance the long-term core competitiveness of businesses, this could lead to an unintended consequence of a reduction in the proportion of labor income.

H2. 

CVC holdings have an inverted U-shaped effect on the labor income share by affecting the proportion of R&D personnel.

According to techno-economic paradigm theory, technological progress is the core driving force for enterprise change and economic growth. Technological innovation not only changes the production mode of enterprises, but also reshapes their organizational structure and competitive advantages, which in turn affects the pattern of income distribution. CVC accelerates this process of technological change by supporting the digital transformation of investee enterprises, which improves their productivity and innovation capacity, thus affecting the share of labor income. CVC plays a key role, which not only provides the necessary financial support, but also effectively reduces the risk of transformation and accelerates the application and marketization of digital technologies. However, when the shareholding of CVC is too high, it may lead to excessive centralization and rigidity in decision making, ignoring the actual needs of enterprises and market dynamics, and affecting the formulation and implementation of digital strategies. As a result, the relationship between CVC shareholding and the share of labor income of an enterprise may exhibit an inverted U-shaped characteristic, and this relationship is mediated by the level of digitization of the enterprise.

Unlike general venture capital, CVC focuses more on strategic orientation, industrial resource integration, business synergy, and long-term cooperation and thus can work more closely with investee firms. Although digital transformation has the potential to increase company revenues and productivity (Hasan et al., 2018; Sezer et al., 2021), numerous businesses are confronted with a lack of clarity regarding their transformation strategies, substantial technology R&D expenditures, and a dearth of skilled professionals. CVC not only provides the necessary financial support but also, more importantly, can effectively reduce transformation risks, and accelerate the application and marketization of digital technology. CVC investors use their strong information and capital advantages to help digital transformation enterprises overcome financial constraints, including providing financial support for infrastructure construction, such as digital equipment and platforms, supporting enterprises in adopting new technologies, and developing better business partners. At the same time, CVC investors can also play a supervisory role, encouraging enterprises to coordinate and integrate resources promptly and improve their risk-resistant ability to cope with the risks of long cycles and high uncertainty caused by digital transformation. Conversely, CVC investors pay attention to the effectiveness of the implementation of the enterprise’s digital transformation, including the scope, goals, and outcomes, and consider whether the transformation will lead to an increase in enterprise value (Gurbaxani & Dunkle, 2019), whereas CVC investors expect enterprises to optimize their management teams, deploy leaders with digital thinking and technology backgrounds, and establish streamlined digital transformation teams to lead top-down digital transformation initiatives. The implementation of digital technology also facilitates enhancements to a company’s R&D capabilities and human capital structure, which subsequently enhances the efficiency of technological innovation and, ultimately, increases revenue through increased production efficiency. As digital technology continues to advance, the distribution of labor resources can be expected to decline further, accompanied by a reduction in production and operating costs. As a result of digitalization and intelligence, labor is likely to benefit from increased profits (Sun, 2023), allowing businesses to increase their labor income share.

Firms backed by CVC investors can mimic the digital business model of the originating firm, take an outsourcing approach to digital transformation, etc. (Furr & Shipilov, 2019) While this reduces the cost of transformation and the risk of failure, it also does not allow them to truly take ownership of the digital transformation, which may achieve only superficial digital mimicry rather than deep, endogenous digital transformation. As CVC investors increase their shareholding, employees and customers may become anxious and resistant. Excessive control may lead to overly centralized and rigid decision making, ignoring the actual needs of the enterprise and market dynamics and affecting the formulation and execution of a digital strategy that suits its development. CVC’s strategic orientation may lead the investee company to reduce its investment in digital transformation and instead invest its resources in projects that can generate economic benefits more quickly. In the long run, this could prevent companies from improving efficiency and innovation through digital transformation, affecting labor productivity and revenue share. Therefore, this paper proposes research Hypothesis H3:

H3. 

CVC holdings have an inverted U-shaped effect on the labor income share by affecting the level of digitization.

According to the resource dependence theory and stakeholder theory, the effectiveness of corporate governance is affected by a variety of factors. When CVC has a low shareholding, it, as an external investor, has an incentive to improve corporate governance and optimize the governance structure through supervision and resource investment. This helps firms attract and retain highly skilled employees and improve the quality of human capital, thereby increasing the labor income share. However, as CVCs’ shareholding increases, their strategic objectives may be more inclined to pursue short-term interests and market competitiveness, leading to excessive intervention in internal governance. This excessive intervention may cause corporate decisions to deviate from long-term development strategies and weaken the protection of employees’ interests, which in turn affects the labor income share. At the same time, CVCs’ resource investment and governance participation may gradually shift from promoting corporate development to protecting their own return on investment, reducing their investment in R&D and employee welfare.

CVC can effectively diversify investment risks by constructing a portfolio that contains multiple projects. This diversified investment strategy reduces the impact of the failure of a single project on the entire enterprise, reduces losses due to information asymmetry or forecasting errors at the macro level, and essentially reduces the information processing costs associated with decision making (Ramaligegowda et al., 2021). By reducing information costs, CVC can identify and respond to potential problems more efficiently, which in turn leads to a more effective regulatory mechanism, improves firms’ tolerance for failures in R&D activities, and promotes firms’ innovation (K. Gao et al., 2019). CVC investors are empowered to require that an investee company promptly address any misconduct and safeguard the rights and interests of its employees in instances where the actions of the managers in question are deemed to violate those rights and interests. Such action may include the rectification of harmed rights and interests, the provision of compensation, the enhancement of the working environment, or the modification of the management system. Moreover, CVC investors can monitor the implementation of corrective measures and require management to improve governance in the interest of companies’ long-term development. According to the capital–skill complementarity hypothesis, corporate governance improvements driven by CVC investors often require the support of many senior managers, and firms rely more on human capital and are willing to attract and hire highly skilled employees through greater compensation.

Despite the positive role of CVC investment in promoting corporate governance, however, there are potential negative effects on the other side of its shareholding. The purpose of CVC shareholding is not only to improve the governance environment of the firms in which it holds a stake but also, more importantly, to utilize its advantage in accessing information to seek benefits in the market. Therefore, from the perspective of governance fraud, as CVC shareholding increases, its supervision of investee firms may be weakened by the pursuit of self-interests, leading to a reduction in governance efficiency (W. A. Li & Li, 2008). The short-sightedness of CVC investors may encourage or require investee firms to reduce their costs, which may lead to slow wage growth, cuts in benefits, or deterioration in working conditions (Du & Ma, 2021), which directly affects the share of firms’ labor income. An increase in CVC shareholding, particularly within the principal–agent framework, provides it with greater influence, which in turn makes management more likely to prioritize short-term CVC interests over long-term development (Falato et al., 2022), thereby further undermining the labor income share. Therefore, when the shareholding ratio is low, CVC shareholding plays the role of “monitoring governance”, whereas when the shareholding ratio exceeds a certain percentage, CVC shareholding has the effect of “governance fraud”. Therefore, this paper proposes Hypothesis H4.

H4. 

CVC holdings have an inverted U-shaped effect on the labor income share by affecting the extent of corporate governance.

2.2.3. Moderating Effect of Media Attention to CVC Holdings on Labor Income Shares in Science-Based Firms

According to external governance theory, in addition to the mandatory regulations, the media serves as a pivotal external instrument for corporate governance within the broader framework of corporate governance. As a vital conduit for corporate communication, it can facilitate the expeditious dissemination of information regarding governance practices and organizational capabilities to the public while concurrently attenuating the informational disparity between businesses and their stakeholders (Saxton & Anker, 2013). Concurrently, stock prices, managerial compensation and turnover, and strategic shifts are all influenced by media scrutiny (Dyck et al., 2008, 2010). An increase in media coverage increases the cost of corporate infractions and encourages managers to engage in self-monitoring and disciplinary actions. This, in turn, enhances the efficacy and performance of corporate governance (Joe et al., 2009). When a firm receives CVC investment, it is closely scrutinized by the government, the capital market, and the media (S. L. Hu et al., 2022), which increases the complexity and risk of corporate reputation management. Businesses that receive considerable media attention are inclined to establish industry benchmarks and have their compensation structures emulated by other businesses. In this case, to maintain competitiveness and attract talent, firms need to maintain a stable share of labor income, thus creating a positive competitive dynamic in the industry. Moreover, media attention intensifies public scrutiny, prompting CVC shareholding firms, especially those in the lower and higher shareholding ranges, to maintain stable socially responsible behaviors to preserve their image and respond to expectations in the face of high exposure (Xiong & Luo, 2021). An important part of this is to improve the treatment of employees (remuneration, working conditions, training, career planning, etc.). In this context, the inverted U-shaped curve will become flatter, which shows that the increase in media attention will weaken the impact of CVC shareholding. Therefore, we propose Hypothesis H5.

H5. 

The inverted U-shaped association is moderated by external media attention, indicating that media attention serves to mitigate the impact of CVC shareholding.

2.2.4. Analysis of the Labor Income Share of CVC Shareholding-Enabled Science and Technology Enterprises Considering the Corporate Pay Gap

Taken from agency theory (Jensen & Meckling, 1979), scholars’ explanations of the intrafirm pay gap are categorized into the optimal pay contract hypothesis and the management power hypothesis. The optimal pay contract hypothesis argues that a performance-related pay system makes agents earn higher income only when they achieve or exceed preset goals, which motivates agents to work harder to satisfy the principal’s goal of corporate profit maximization. Contrary to the hypothesis of the optimal pay contract, the managerial power hypothesis (Bebchuk et al., 2002) argues that when top managers can influence the formulation of the pay contract, it is possible that they may be more driven to safeguard their own interests than the organization’s collective interests. Consequently, they may make decisions that ultimately benefit them more in the long term than the company does, which subsequently contributes to an expansion of the pay gap. Compared with companies that do not receive CVC investment, CVC participation may increase the demand for jobs such as platform construction and business model innovation, and these additional jobs require more knowledge-based and skilled workers. With the increase in highly skilled and educated talent, the overall value of the corporate labor force increases while reducing the demand for low-skilled workers, which gives highly skilled personnel greater bargaining power and narrows the pay gap between ordinary employees and management. Owing to the principal-agent problem, an increase in CVC shareholding leads to a greater resource base for the firm, thereby enhancing management’s control over these resources. Management exerts a stronger influence on the design of the compensation system and therefore has an incentive to develop a compensation system that works in favor of it on the basis of information asymmetry and positional advantages (Robinson & Sensoy, 2013). In this way, with the total compensation fixed, the increase in the level of CVC participation leads to an increase in management power, thus reducing the level of employee pay (S. Li & Wang, 2023), resulting in further widening of the pay gap within the firm. In summary, CVC is associated with a narrowing of the pay gap. While when the CVC shareholding exceeds a certain percentage, the management power keeps increasing and the pay gap between executives and ordinary employees tends to increase. Therefore, we propose Hypothesis H6:

H6. 

There is a U-shaped relationship between CVC shareholding and corporate pay gap.

The research framework of this paper is shown in Figure 1.

3. Sample Selection and Model Construction

3.1. Sample Selection and Data Sources

Since 2010, there has been a growing need to examine the impact of CVC on the labor income share of businesses, particularly startups. This is due to several factors, including shifts in the global economic landscape, the rapid advancement of technology, and growing societal concerns about the equity of income distribution. Such research helps us better understand the income distribution mechanism in modern economies and provides an important reference point for promoting healthier and sustainable economic development. The STIB and the GEM mainly serve science and technology innovative enterprises, which are often characterized by high growth and risk. They are more dependent on external financing, especially the support of venture capital. Choosing listed companies on these two types of boards as samples can more accurately reflect the direct impact of venture capital on the labor income share of Chinese enterprises. Choosing the period of 2010–2022 ensures access to more complete and reliable data records, including but not limited to annual reports, financial reports, and public records of VC transactions of listed companies in China, which helps to conduct in-depth data analysis and empirical research on the research topic.

Accordingly, the research sample for this paper comprises China’s GEM- and KIC-listed firms from 2010 to 2022. Financial and insurance listed companies, as well as ST, *ST, and PT companies, insolvent companies and companies with crucial variables missing, have been excluded, yielding 7624 observations. The majority of the data were obtained from the CSMAR database, whereas the CVC data were gathered from CVsource and other sources, as well as online information query platforms. To limit the impact of extreme values, we reduce the upper and lower 1% of all continuous variables.

3.2. Definition of the Main Variables

3.2.1. Explained Variable: Firm Labor Income Share (LS)

Referring to the research of X. Y. Jiang and Lin (2022), the definition index of the share of corporate labor income is as follows: LS = (cash paid to and for employees + employee compensation payable at the end of the period-employee compensation payable at the beginning of the period)/total operating income.

3.2.2. Explanatory Variable: The Proportion of Shares of Investee Firms Held by CVC (Ownership)

Owing to the absence of direct CVC shareholding data in the database, this paper adopts a manual collection method for CVC shareholding information on China’s GEM and KEC. To determine whether a firm has CVC participation, this paper draws comprehensively on L. Wang and Qi (2019), and given the availability of data, according to different forms of listed large firms’ participation in CVC investment, it is defined according to the following conditions: (1) there are nonfinancial listed firms among the top ten shareholders; (2) nonfinancial listed companies participate in or hold VC institutions. For the judgment of VC institutions, we draw on the practice of G. Zhang and Shao (2022), which is based on the China Entrepreneurial Venture Capital Development Report and CVsource and other databases, to determine whether the top ten shareholders are VC institutions and then through the network information query platform to determine whether the VC institution is a nonfinancial listed company participating in or holding.

Typically, the percentage of equity holdings is a good indicator of the strength of management and control relationships between firms. To consider the degree of CVC participation, this paper uses the CVC shareholding ratio to measure CVC participation in listed firms. The aggregate CVC shareholding ratio is defined as the percentage of shares of China’s GEM- and KTC-listed companies that the CVC’s parent company owns in relation to the total share capital.

3.2.3. Control Variables

This paper selects two main levels of control variables: the level of basic corporate characteristics, including the gearing ratio (Lev), return on assets (Roa), and firm age (FirmAge), and the level of shareholding and governance characteristics, including the nature of property rights (SOE), the number of directors (Board), the proportion of independent directors (Indep), and the combination of two positions (Dual). See

Table 1. Definitions of control variables.

Variable Names	Notation	Details
The nature of property rights	SOE	If the registration type is “state-owned”, then take 1; otherwise 0
Gearing ratio	Lev	Total liabilities/total assets at the end of the year
Return on assets	Roa	Net profit/total assets
The number of directors	Board	The number of directors on the board takes the natural logarithm
The proportion of independent directors	Indep	The ratio of independent directors to board members
The combination of two positions	Dual	If the chairperson of the board and the general manager are one person, then take 1; otherwise 0
Firm age	FirmAge	To calculate the natural logarithm, subtract the current year from the opening year plus one

for details of the main control variable selection and definitions.

3.3. Model Construction

To test the impact of CVC on the labor income share of China’s GEM and KTC-listed firms, this paper constructs Model (1):

$${Ls}_{it}={\alpha }_0+{\alpha }_1{Ownership}_{it}{+\alpha }_2{Ownership}_{it}^2+{\alpha }_j{\sum }_j{Controls}_{it}+\sum Industry+\sum Province+\sum year+{\epsilon }_{it}$$

(1)

In the interim, we employ the ratio of R&D personnel (Rdpersonratio), the degree of digitization (Digital), and the level of corporate governance (Manacost) as mediating variables, as elucidated by the study of T. Jiang (2022), to elucidate the mechanism underlying the ratio of CVC shareholding and the labor income share of investee firms. The following mediation effect model was constructed by incorporating fixed effects and control variables from the baseline regression model.

$$\begin{gathered} {Mediator}_{it}={\beta }_0+{\beta }_1{Ownership}_{it}+{\beta }_2{Ownership}_{it}^2+{\beta }_j{\sum }_j{Controls}_{it}+\sum Industry+\sum Province + \\ \sum year+{\epsilon }_{it} \end{gathered}$$

(2)

$$\begin{gathered} {LS}_{it}={\gamma }_0+{\gamma }_1{Ownership}_{it}+{\gamma }_2{Ownership}_{it}^2+{\gamma }_3{Mediator}_{it}+{\gamma }_j{\sum }_j{Controls}_{it}+\sum Industry+\sum Province + \\ \sum year+{\epsilon }_{it} \end{gathered}$$

(3)

Among them, Mediator represents the mediating variables, including the ratio of R&D personnel (Rdpersonratio), the degree of digitization (Digital), and the level of corporate governance (Manacost). If the coefficients β₁ and β₂ are significant, we will further test the mediation effect. If γ₁, γ₂, and γ₃ are all significant, this indicates a partial mediation effect of R&D personnel proportion, digitalization degree, and corporate governance level. If γ₁ and γ₂ are not significant, but γ₃ is significant, this suggests a full mediation effect of the mediating variable on the impact of CVC shareholding on labor income share.

Finally, to investigate the effect of media attention on CVC shareholding and the labor income share and to test whether Hypothesis H5 is valid, the following model is established, drawing on Haans et al. (2016) and Zhu and Zhou (2018):

$$\begin{gathered} {Ls}_{it}={\alpha }_0+{\alpha }_1{Ownership}_{it}+{\alpha }_2{Ownership}_{it}^2+{\alpha }_3{Media}_{it}+{\alpha }_4{Ownership}_{it}\times {Media}_{it}+{\alpha }_5{Ownership}_{it}^2\times {Media}_{it} + \\ {\alpha }_j{\sum }_j{Controls}_{it}+\sum Industry+\sum Province+\sum year+{\epsilon }_{it} \end{gathered}$$

(4)

The curvature K of the vertex of the quadratic function can be calculated by taking the second-order derivative of LS in Equation (4):

$$K={Ls}_{it}^{″}=2{\alpha }_2+2{\alpha }_5{Media}_{it}$$

(5)

For the inverted U-shaped curve, K < 0, and the smaller K is, the steeper the curve is, and the larger the curve is, the smoother it is. The partial derivation of ${Media}_{it}$ in Equation (5) yields:

$${\displaystyle \frac{\partial K}{\partial {Media}_{it}}}=2{\alpha }_5$$

(6)

As illustrated by Equation (6), the regression coefficient ${\alpha }_5$ of ${Ownership}_{it}^2\ast {Media}_{it}$ represents the point at which the influence of media attention on the inverted U-shaped relationship is most pronounced. It is anticipated that ${\alpha }_5$ will be significantly positive, indicating that there will be a notable increase in media attention as the Media Index increases, K will decrease, and the inverted U-shaped curve will become flatter. In other words, the inverse relationship between the labor income share and CVC shareholding will be weakened by the influence of media attention.

The first-order derivative of Ownership in Equation (4), which is equal to 0, can be used to calculate the inflection point of the curve.

$${Ownership}_{it}^{\ast }=-{\displaystyle \frac{{\alpha }_1{+\alpha }_4{Media}_{it}}{{2(\alpha }_2{+\alpha }_5{Media}_{it})}}$$

(7)

The point of inflection in Equation (7), ${Ownership}_{it}^{\ast }$, is obtained as a partial derivative of ${Media}_{it}$:

$${\displaystyle \frac{\partial {Ownership}_{it}^{\ast }}{\partial {Media}_{it}}}=-{\displaystyle \frac{{\alpha }_1{\alpha }_5-{\alpha }_2{\alpha }_4}{{2\left({\alpha }_2{+\alpha }_5{Media}_{it}\right)}^2}}$$

(8)

If the partial derivative of ${Media}_{it}$ is less than 0, the larger the moderator variable ${Media}_{it}$ is, the smaller the value of ${Ownership}_{it}^{\ast }$ is, and the inflection point of the inverted U-shaped curve will shift to the left. As a result of increased media attention, enterprises’ willingness to accept CVC investment increases. This is an effort to present a favorable company image, which increases the CVC shareholding ratio’s ideal condition. Consequently, the curve’s inflection point is anticipated to move to the right.

4. Empirical Results and Analysis

4.1. Descriptive Statistical Analysis

This paper’s analysis of the impact of CVC is based on the descriptive statistics of the main variables, which are presented in

Table 2. Results of the descriptive statistics of the variables.

Varname	Obs	Mean	SD	Min	P25	Median	P75	Max
LS	7624	0.188	0.122	0.0257	0.107	0.161	0.233	0.699
Ownership	7624	0.0141	0.0440	0	0	0	0	0.274
SOE	7624	0.0725	0.259	0	0	0	0	1
Lev	7624	0.325	0.182	0.0426	0.178	0.300	0.453	0.815
Roa	7624	0.0422	0.0783	−0.302	0.0181	0.0484	0.0802	0.235
Board	7624	2.056	0.190	1.609	1.946	2.079	2.197	2.485
Indep	7624	0.383	0.0534	0.333	0.333	0.375	0.429	0.571
Dual	7624	0.440	0.496	0	0	0	1	1
FirmAge	7624	2.809	0.318	1.792	2.639	2.833	3.045	3.434

. The mean value of firms’ labor income share is 0.188, with a standard deviation of 0.122. The maximum value is 0.699, whereas the minimum value is 0.0257. These findings indicate the existence of notable disparities in the labor income share across firms. The mean and median of each major continuous variable are generally not very different, and the standard deviation is also small, with an overall symmetrical distribution and no outliers.

4.2. Analysis of the Empirical Results

Considering the individual effects, the Hausman tests all rejected the original hypothesis, so the fixed effects model was used. According to the study of Haans et al. (2016), to verify the existence of an inverted U-shaped relationship between CVC shareholding (Ownership) and the share of corporate labor income (LS), three conditions need to be satisfied simultaneously: (1) Ownership² regression coefficient ${\alpha }_2$ is significantly negative; (2) the slope of the curve (i.e., the first-order derivative) is significantly positive when Ownership takes the minimum value, and the slope of the curve is significantly negative when Ownership takes the maximum value; and (3) the inflection point of the curve is within the range of Ownership values. Combining the above conditions, we carry out the following test: as shown in

Table 3. Impact of CVC shareholding and firm’s labor income share.

	(1)	(2)
VARIABLES	LS	LS
Ownership	0.279 ***	0.297 ***
	(3.50)	(3.86)
Ownership2	−1.219 ***	−1.167 ***
	(−3.42)	(−3.39)
SOE		0.001
		(0.18)
Lev		−0.131 ***
		(−18.26)
Roa		−0.346 ***
		(−21.76)
Board		0.017 **
		(2.11)
Indep		0.122 ***
		(4.25)
Dual		0.009 ***
		(3.67)
FirmAge		0.022 ***
		(5.92)
Constant	0.158 ***	0.049
	(8.66)	(1.47)
Industry	Yes	Yes
Province	Yes	Yes
Year	Yes	Yes
Observations	7624	7624
R-squared	0.302	0.360

Note: *** and ** denote significance at the 1% and 5% levels, respectively; the t values in parentheses are the same as those below; Ownership² represents the square term of ownership.

, in the main regression equation, the regression coefficient of Ownership is positive, and the regression coefficient of Ownership² is significantly negative, which satisfies the first condition. Next, taking the first-order derivative of the equation reflecting the relationship (see Equation (1)), the slope of the curve can be obtained as ${\alpha }_1$ + 2${\alpha }_2$ Ownership. The second condition is satisfied when the slope of the curve is positive (0.297) when Ownership takes the minimum value of 0 and negative (−0.343) when Ownership takes the greatest value of 0.274. To determine the location of the inflection point, the second-order derivative was further solved and made equal to zero, using the inflection point formula for quadratic functions −${\alpha }_1$/(2${\alpha }_2$), which, after substituting the specific coefficient value, was calculated to be 0.127. The third requisite is satisfied, as the inflection value in question falls within the designated range of ownership values. In conclusion, the results demonstrate a robust inverted U-shaped relationship between an organization’s labor income share and its CVC shareholding, thereby confirming Hypothesis H1.

5. Robustness Tests and Heterogeneity Analysis

5.1. Discussion of Endogeneity Issues

5.1.1. PSM Propensity Matching Method

This study employs propensity score matching (PSM) to mitigate the influence of endogeneity issues on the findings. The control variables discussed are employed as matching variables, the CVC shareholding ratio is utilized as the matching criterion, and companies situated on the right side of the inflection point of the inverted U-shaped curve are designated the treatment group, in accordance with the methodology of H. Hu et al. (2020). To identify a control group with comparable traits to those of the treatment group, the propensity scores of the samples in the treatment group were then computed using a logit model (using radius matching). Finally, the baseline model was re-estimated using the matched data. The findings show that matched firms with low CVC participation and those with strong CVC participation do not differ significantly ex ante. Column (1) of

Table 4. Test results of the endogeneity problem.

	(1)	(2)	(3)	(4)
VARIABLES	LS	Ownership	Ownership2	LS
Ownership	0.257 *			1.930 ***
	(1.80)			(5.575)
Ownership2	−1.150 **			−9.157 ***
	(−1.99)			(−5.317)
Ownership _IV		0.962 ***
		(16.822)
Ownership2 _IV			0.337 ***
			(5.165)
SOE	−0.011	0.028 ***	0.007 ***	0.007
	(−1.29)	(8.403)	(8.079)	(1.227)
Lev	−0.117 ***	0.007 ***	0.002 ***	−0.125 ***
	(−5.52)	(2.654)	(2.951)	(−16.375)
Roa	−0.365 ***	0.005	0.001	−0.340 ***
	(−6.72)	(0.935)	(0.702)	(−15.452)
Board	−0.015	0.008 ***	0.001 ***	0.012
	(−0.42)	(3.462)	(2.789)	(1.465)
Indep	−0.032	−0.044 ***	−0.007 ***	0.154 ***
	(−0.29)	(−5.199)	(−3.679)	(5.027)
Dual	−0.002	0.002 ***	0.000	0.008 ***
	(−0.21)	(2.841)	(0.855)	(3.037)
FirmAge	0.024 **	−0.005 ***	−0.001 **	0.025 ***
	(2.21)	(−3.678)	(−2.471)	(6.884)
Industry	Yes	Yes	Yes	Yes
Province	Yes	Yes	Yes	Yes
Year	Yes	Yes	Yes	Yes
Observations	972	7624	7624	7624
R-squared	0.407	0.295	0.257	0.295

Note: ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively; Ownership² represents the square term of ownership.

presents the regression findings for the sample following the PSM procedure. The findings of this research are examined, and the results demonstrate that CVC still has a significant inverted U-shaped effect on firms’ labor income shares, even after controlling for the possibility of selectivity bias.

5.1.2. Instrumental Variables Approach

To address the endogeneity issue introduced by reverse causation, the baseline regression model is re-estimated using the instrumental variable technique, as previously described in the literature. In this study, the instrumental variables for IV-2SLS estimation are the average CVC ownership of businesses in the same city (Ownership_IV) and its squared term. CVC investment has significant geographic agglomeration and spillover effects (R. Zhang & Zhang, 2022). The average CVC level of enterprises in the same city is positively correlated with the likelihood of an enterprise making a CVC investment, satisfying the condition of instrumental variable correlation. Conversely, the average CVC shareholding ratio of enterprises in the same city does not directly affect the labor income share of invested enterprises, which meets the condition of instrumental variable homogeneity. Columns (2)–(4) of Table 4 present the IV-2SLS estimation results. The original hypothesis of weak instrumental variables is rejected on the basis of the F statistic (160.555), which is greater than the critical value of the Stock–Yogo weak instrumental variable identification F test at the 10% significance level. Furthermore, there is no issue with overidentification, and the number of instrumental variables is equal to that of endogenous variables. Consequently, the coefficient of the major term in Column (4) of Table 4 is significantly positive, whereas that of the secondary term is notably negative. This illustrates the reasonableness and dependability of the instrumental variables employed in this study. The relationship displays a notable inverted U-shaped correlation, as evidenced by the regression outcomes presented above. This further substantiates the soundness of the paper’s fundamental conclusions after accounting for the influence of endogeneity.

5.2. Other Robustness Tests

5.2.1. Lagged Core Variables

Considering the inverted U-shaped effect of CVC shareholding on the labor income share of China’s KIC and GEM portfolio companies with lags, this paper lags the core explanatory variables by one period and two periods, respectively.

Table 5. Other robustness test results.

	(1)	(2)	(3)	(4)
VARIABLES	LS	LS	Labor	LnLabor
Ownership			0.280 ***	1.673 ***
			(3.70)	(3.39)
Ownership2			−1.099 ***	−5.762 ***
			(−3.25)	(−2.61)
L.Ownership	0.245 ***
	(2.79)
L.Ownership2	−0.870 **
	(−2.21)
L2.Ownership		0.248 **
		(2.49)
L2.Ownership2		−0.879 **
		(−1.97)
Constant	0.050	0.045	0.051	−2.541 ***
	(1.34)	(1.08)	(1.57)	(−11.87)
Controls	Yes	Yes	Yes	Yes
Industry	Yes	Yes	Yes	Yes
Province	Yes	Yes	Yes	Yes
Year	Yes	Yes	Yes	Yes
Observations	6076	4896	7624	7624
R-squared	0.354	0.351	0.362	0.383

Note: *** and ** denote significance at the 1% and 5% levels, respectively; Ownership² represents the square term of ownership.

presents the regression results. According to the findings of the benchmark regression analysis, the inverted U-shaped relationship remains largely unaltered when the lagged terms of the explanatory variables are incorporated into Columns (1) and (2).

5.2.2. Replacement of Explanatory Variables

This study considers the validity of indicator selection, considers the literature (X. Wang & Huang, 2017; Shi et al., 2019), and uses the “cash paid to and for employees” in the cash flow statement divided by the “total operating income” in the income statement “to calculate the labor income share” to determine Labor: LnLabor = Ln[Labor/(1 − Labor)], before another regression is conducted. The findings are presented in Columns (3) and (4) of Table 5. After the newly defined labor income share index is adopted, the inverted U-shaped relationship between the CVC shareholding ratio and portfolio companies’ labor income share is still significant, which further confirms the initial hypothesis.

5.3. Heterogeneity Analysis

5.3.1. Regional Heterogeneity

To investigate the impact of CVC ownership in regions with different degrees of marketization, the paper divides the sample firms into two major regions, Eastern and Midwestern, according to their place of incorporation in China for the subgroup analysis. The results of the grouped regression are presented in

Table 6. Results of heterogeneity analysis.

	(1)	(2)	(3)	(4)
	Midwest	East	Low Financial Constraints	High Financial Constraints
VARIABLES	LS	LS	LS	LS
Ownership	0.349 *	0.294 ***	0.278 ***	0.349 *
	(0.181)	(0.0855)	(0.0856)	(0.186)
Ownership2	−1.139	−1.225 ***	−1.109 ***	−1.197
	(0.815)	(0.382)	(0.378)	(0.896)
SOE	−0.00683	0.00303	0.00255	0.00383
	(0.0111)	(0.00538)	(0.00539)	(0.0111)
Lev	−0.138 ***	−0.131 ***	−0.138 ***	−0.152 ***
	(0.0157)	(0.00818)	(0.00815)	(0.0181)
Roa	−0.383 ***	−0.337 ***	−0.382 ***	−0.229 ***
	(0.0351)	(0.0180)	(0.0186)	(0.0312)
Board	0.0260	0.0149	0.0124	0.0375 **
	(0.0194)	(0.00911)	(0.00947)	(0.0169)
Indep	0.127 *	0.122 ***	0.122 ***	0.115 *
	(0.0659)	(0.0321)	(0.0323)	(0.0637)
Dual	0.0128 **	0.00774 ***	0.00620 **	0.0208 ***
	(0.00541)	(0.00269)	(0.00272)	(0.00518)
FirmAge	0.0165 *	0.0232 ***	0.0262 ***	0.00761
	(0.00853)	(0.00421)	(0.00415)	(0.00980)
Constant	0.0574	0.0461	0.0522	0.0528
	(0.0746)	(0.0379)	(0.0374)	(0.0816)
Industry	Yes	Yes	Yes	Yes
Province	Yes	Yes	Yes	Yes
Year	Yes	Yes	Yes	Yes
Observations	1579	6045	6071	1553
R-squared	0.369	0.364	0.365	0.389

Note: ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively; Ownership² represents the square term of ownership.

, Columns (1) and (2). The findings in Column (2) indicate the presence of an inverted U-shaped relationship and the regression coefficient of the quadratic term of CVC shareholding (Ownership²) in the Eastern region. The coefficient is −1.225 and is statistically significant. In contrast, research conducted in the Midwest region indicates that while the primary term of CVC shareholding is significantly and positively correlated with its labor income share, the quadratic term is not statistically significant. This may be because the Midwest region faces a scarcity of top talent, and the increase in CVC ownership encourages firms to adopt more attractive compensation and career development strategies to maintain and attract talent. Moreover, China’s coordinated regional development policies, which include tax incentives, financial subsidies, and low-interest loans, also prove advantageous for the central and western regions. Consequently, an increase in CVC ownership enables businesses to utilize external resources in a more efficacious manner, facilitates accelerated development, and provides incentives for businesses to increase their labor income share.

5.3.2. Heterogeneity of Financing Constraints

This study examines the impact of CVC holdings on the labor income share in relation to the degree of financing constraints. Specifically, the WW index from 2010 to 2022 is used to assess the financing constraint status of each firm, and the sample is categorized into a higher-financing-constraint group and a lower-financing-constraint group on the basis of the mean of the scores. In Table 6, Column (3) shows the regression results for the lower-financing-constraints group, in which the inverted U-shaped relationship is still significant. This is mainly because firms with lower financing constraints usually have stronger access to capital. The resources and strategic direction provided by CVC significantly increase business effectiveness and technical innovation during the initial growth phase of CVC ownership, which in turn increases labor productivity and labor income share. However, as CVC ownership continues to increase, the marginal utility of additional resources diminishes, and excessive external intervention may disrupt the normal operating rhythm of the firm, leading to increased management complexity and decreased decision-making efficiency and ultimately reducing the share of labor income. Conversely, the primary term of CVC ownership has a substantial positive correlation with the labor income share of the investee firm. However, the coefficient of the quadratic term (Ownership²) is not statistically significant in the high-financing-constraint group. The increased involvement and expanded involvement of CVC is of particular importance for businesses that are constrained by significant financial limitations. The involvement of CVC can facilitate the enhancement of enterprise management standards and competitiveness in these domains, which often exhibit deficiencies in their benefit systems, market strategies, and management structures. Through the design and implementation of effective incentive mechanisms, CVC not only improves employee satisfaction and loyalty but also directly and positively impacts labor income share.

6. Mechanism Analysis

The purpose of this section is to empirically test hypotheses H2, H3, and H4, and explore the potential mediating mechanisms of the relationship between CVC shareholding and firms’ labor income shares.

6.1. Mediating Effect of the R&D Personnel Ratio

This study calculates the R&D personnel ratio (Rdpersonratio) by dividing the total number of R&D personnel by the total number of employees. The regression results between the Rdpersonratio in the investee company and the proportion of CVC ownership are examined in Column (1) of

Table 7. Regression results of the mechanism of action.

	(1)	(2)	(3)	(4)	(5)	(6)
VARIABLES	Rdpersonratio	LS	Digital	LS	Manacost	LS
Ownership	0.496 ***	0.221 ***	3.627 ***	0.261 ***	0.084 ***	0.388 ***
	(4.31)	(2.91)	(4.07)	(3.42)	(4.34)	(5.23)
Ownership2	−2.108 ***	−0.862 **	−10.800 ***	−1.062 ***	−0.332 ***	−1.529 ***
	(−4.10)	(−2.55)	(−2.71)	(−3.11)	(−3.85)	(−4.62)
Rdpersonratio		0.048 ***
		(5.32)
Digital				0.010 ***
				(9.82)
Manacost						1.088 ***
						(24.63)
SOE	0.014 *	−0.005	0.190 ***	−0.001	0.004 ***	0.005
	(1.94)	(−1.01)	(3.42)	(−0.21)	(3.39)	(1.15)
Lev	−0.004	−0.147 ***	0.017	−0.132 ***	0.005 ***	−0.126 ***
	(−0.37)	(−20.22)	(0.21)	(−18.40)	(2.85)	(−18.17)
Roa	−0.030	−0.337 ***	−0.037	−0.346 ***	0.032 ***	−0.311 ***
	(−1.27)	(−21.56)	(−0.20)	(−21.88)	(8.03)	(−20.26)
Board	−0.018	0.025 ***	−0.105	0.018 **	−0.002	0.015 **
	(−1.46)	(3.15)	(−1.11)	(2.25)	(−0.81)	(1.96)
Indep	0.086 **	0.122 ***	−0.937 ***	0.131 ***	−0.014 *	0.107 ***
	(2.01)	(4.36)	(−2.83)	(4.60)	(−1.89)	(3.88)
Dual	0.012 ***	0.007 ***	0.101 ***	0.008 ***	0.002 ***	0.011 ***
	(3.24)	(3.06)	(3.64)	(3.28)	(3.65)	(4.85)
FirmAge	0.100 ***	−0.013 ***	0.747 ***	0.015 ***	0.011 ***	0.035 ***
	(17.93)	(−3.14)	(17.21)	(3.93)	(12.14)	(9.51)
Constant	−0.185 ***	0.065 **	1.526 ***	0.034	−0.070 ***	−0.027
	(−3.71)	(1.97)	(3.95)	(1.03)	(−8.34)	(−0.84)
Industry	Yes	Yes	Yes	Yes	Yes	Yes
Province	Yes	Yes	Yes	Yes	Yes	Yes
Year	Yes	Yes	Yes	Yes	Yes	Yes
Observations	7624	7624	7624	7624	7624	7624
R-squared	0.268	0.370	0.255	0.368	0.191	0.407
Sobel Z	−2.064 **		−1.811 *		−2.111 **

Note: ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively; Ownership² represents the square term of ownership.

. The data indicate that the coefficient of the squared term of the proportion of CVC shareholding is −2.108, which is significantly negative. The results indicate that the Rdpersonratio in invested businesses and CVC shareholdings exhibit a notable inverted U-shaped relationship. Specifically, moderate CVC shareholding is associated with an increase in the Rdpersonratio in businesses, whereas excessive CVC participation is linked to a gradual decline in this figure. Further analysis in column (2) of Table 7 shows that the coefficients of CVC shareholding and its squared term remain significant at the 1% and 5% levels, respectively, after the introduction of the proportion of R&D personnel in investee firms as a mediating variable, while the coefficients of the proportion of R&D personnel are also significant at the 1% level. In addition, the Sobel Z-test is statistically significant at the 5% level, further confirming that the proportion of R&D personnel in investee firms plays a partial mediating role in the effect of CVC shareholding on firms’ labor income shares. In conclusion, the veracity of Hypothesis H2 is confirmed.

6.2. Mediating Effect of the Degree of Digitization

To measure the degree of firm digitization, this study employs the methodology proposed by Yuan et al. (2021), which quantifies the extent of microfirm digitization (Digital) by dividing the total number of words pertaining to firm digitization by the length of the annual report’s management discussion and analysis (MD&A) section and subsequently multiplying the result by 100. A company is deemed to have a more pronounced digital orientation if its Digital indicator exhibits a higher value. The regression results, with the firm’s level of digitalization (Digital) serving as the explanatory variable, are presented in Column (3) of Table 7. At the 1% level of significance, the coefficients of CVC shareholding and its squared term remain statistically significant. This finding indicates that the degree of digitization of investee enterprises displays an inverted U-shaped trajectory, initially increasing and subsequently declining as CVC institutions expand their holdings. Column (4) of Table 7 presents the regression results after including Digital as a mediating variable in the baseline model and finds that the coefficients are still significant at the 1% level and the corresponding Sobel Z-test is statistically significant at the 10% level. Thus, digitalization has a partial mediating effect in the inverted “U” shaped relationship between CVC ownership and firms’ labor income share. Hypothesis H3 is confirmed.

6.3. Mediating Effect of Corporate Governance Level

In this research, the percentage of corporate overhead to total assets is employed as a proxy variable for the corporate governance level. This approach is grounded in a study by S. Li (2007). Lower values imply higher organizational management efficiency within the firm and better corporate governance. To facilitate the interpretation of the empirical results, the indicator reflecting the level of corporate governance (Manacost) is obtained by taking the opposite of the aforementioned results. The results with the level of corporate governance (Manacost) as the explanatory variable are presented in Column (5) of Table 7. The findings indicate that both the coefficient of CVC shareholding and its squared term are statistically significant. This finding suggests that as CVC firms expand their interest, the corporate governance level of investee enterprises tends to increase and then decreases in an inverted U-shaped manner. Column (6) of Table 7 shows the regression results after adding the mediating variable of corporate governance level (Manacost) to the baseline regression model. It can be seen that the regression coefficients of the coefficients of the CVC shareholding ratio and its squared term and the corporate governance level are significant at the 1% level and the corresponding Sobel Z-test is statistically significant at the 5% level, which indicates that there is a significant indirect effect of the corporate governance level. Consequently, the level of corporate governance has a partial mediating effect in the inverted “U” shaped relationship between CVC ownership and firms’ labor income shares. Hypothesis H4 passes the test.

Through the above analysis, the empirical results in this section support hypotheses H2, H3, and H4, that is, CVC shareholding shows a significant inverted U-shape relationship with the share of firms’ labor income by affecting the proportion of R&D personnel, the degree of digitalization of firms, and the level of corporate governance.

7. Further Discussion

This section further explores the moderating effect of media attention on the inverted U-shaped relationship between CVC holdings and the share of firms’ labor income (H5), as well as the U-shaped relationship between CVC holdings and the pay gap within firms (H6).

7.1. Moderating Effects

7.1.1. Effect of Media Attention on the Inverted U-Shaped Curve

This study uses the frequency of online media coverage of listed businesses annually as the media attention indicator (Media). To facilitate further analysis of the data, the amount of online media coverage is logarithmized in this paper.

The degree of flatness or steepness of the quadratic function curve can be judged by calculating the curvature K obtained from the second-order derivative at the vertex of the quadratic function. According to Equation (4), the inverted U-shaped curve of CVC shareholding versus labor income share has a negative curvature K at the apex, indicating that the initial shape of the curve is downward concave. As the absolute value of K decreases (i.e., the negative value increases), the shape of the curve tends to flatten. To analyze the specific effect of media attention, this paper examines the trend in curvature K by calculating the partial derivative of curvature K for ${Media}_{it}$ (see Equation (6)). The effect of Media on the curvature K of the inverted U-shaped curve mainly depends on regression coefficients of the interaction term between Ownership² and Media (Ownership² × Media).

A regression analysis of the data presented in

Table 8. Regression results for further analysis.

	(1)	(2)
VARIABLES	LS	GAP
Ownership	1.221 ***	−4.078 **
	(3.21)	(−2.24)
Ownership2	−5.606 ***	34.168 ***
	(−3.21)	(4.20)
Media	−0.000
	(−0.16)
Ownership × Media	−0.188 **
	(−2.48)
Ownership2 × Media	0.899 ***
	(2.60)
SOE	0.001	−0.182
	(0.23)	(−1.60)
Lev	−0.131 ***	1.627 ***
	(−17.98)	(9.56)
Roa	−0.344 ***	6.430 ***
	(−21.37)	(17.12)
Board	0.017 **	1.430 ***
	(2.12)	(7.39)
Indep	0.122 ***	0.289
	(4.25)	(0.43)
Dual	0.009 ***	−0.235 ***
	(3.76)	(−4.14)
FirmAge	0.022 ***	−0.005
	(5.71)	(−0.05)
Constant	0.051	0.716
	(1.50)	(0.91)
Industry	Yes	Yes
Province	Yes	Yes
Year	Yes	Yes
Observations	7624	7624
R-squared	0.360	0.184

Note: *** and ** denote significance at the 1% and 5% levels, respectively; Ownership² represents the square term of ownership.

indicates that the coefficient of Ownership² × Media is significantly positive. This implies that as media attention increases, the shape of the inverted U-shaped curve becomes flatter, with the absolute value of the curvature K at the apex of the curve decreasing. This suggests that the inverted U-shaped relationship is weakened by increased media attention. To verify the effect of media attention, values are categorized based on quartiles: low media attention is defined as the first quartile (Q1, 4.33) and high media attention as the third quartile (Q3, 5.41). The magnitude of the vertex curvature, K, is then incorporated into Equation (5). The result shows that in the regression equation, when ${Media}_{it}$ takes the 1/4 quartile, K is −3.43; when ${Media}_{it}$ takes the 3/4 quartile, K is −1.18. This result shows that, as the Media value increases, despite K remaining negative, its absolute value diminishes. This directly confirms that heightened media attention leads to a flattening of the inverted “U” shaped curve.

7.1.2. Impact of Media Attention on the Inflection Point

The inflection point of a quadratic function curve is determined by the value of the independent variable at which the first derivative equals zero. The influence of a moderating variable, such as media attention, on this inflection point is assessed through the partial derivative of the inflection point with respect to the moderating variable. More specifically, based on Equation (8), when the denominator is positive, the sign of the numerator dictates the direction in which media attention (Media) affects the location of the curve’s inflection point. The sign of the numerator determines the direction in which media attention (Media) affects the location of the curve’s inflection point. The fact that the numerator ${\alpha }_1{\alpha }_5-{\alpha }_2{\alpha }_4$ is greater than 0 indicates that the inflection point of the inverted “U” curve of CVC ownership versus firm’s labor income share shifts to the right as media attention increases. To specifically quantify the above effects, the corresponding inflection point ${Ownership}_{it}^{\ast }$ is calculated by taking the 1/4 quartile (4.33) and the 3/4 quartile (5.41) of ${Media}_{it}$, respectively, into Equation (7). The results show that when ${Media}_{it}$ is in the 1/4th quartile, the inflection point ${Ownership}_{it}^{\ast }$ is 0.116, and when Media_it is in the 3/4th quartile, the inflection point ${Ownership}_{it}^{\ast }$ shifts to 0.137, which confirms that the inflection point of the curve of CVC ownership versus firms’ share of labor income shifts to the right when media attention is high. This suggests that an increase in media attention may result not only in startups becoming more reliant on CVC but also in the optimal level of CVC ownership being raised, which would cause the inflection point of the inverted U-shaped curve to shift to the right.

7.2. Impact of CVC Shareholding on the Pay Gap

In conjunction with the studies on the pay gap, this study calculates the internal pay gap (GAP) by dividing the average salary of executives by the average salary of employees. A greater GAP indicates a more pronounced disparity in remuneration between executives and regular employees in publicly traded companies. CVC shareholding and the internal pay gap have a strong positive U-shaped relationship, according to the regression results in Column (2) of Table 8. In other words, when the CVC ownership ratio is low, the change in the internal wage gap within the company aligns with the optimal compensation contract hypothesis, leading to a gradual narrowing of the wage gap between executives and regular employees. However, once the CVC ownership ratio surpasses a certain threshold, the influence of management significantly increases, and the shift in the compensation structure begins to conform to the predictions of the management power theory, resulting in an increasing trend in the wage gap between executives and employees.

The empirical results in this section further confirm the validity of hypotheses H5 and H6 that media attention does moderate the inverted U-shaped relationship between CVC holdings and firms’ share of labor income, and that the relationship between CVC holdings and the intra-firm pay gap exhibits a U-shaped relationship.

8. Conclusions and Implications of the Study

8.1. Conclusion and Contributions

This study investigates the mechanisms through which CVC affects the labor income share of enterprises, using data from China’s Science and Technology Innovation Board (STIB) and Growth Enterprise Market (GEM) from 2010 to 2022. The results reveal a significant inverted U-shaped relationship between CVC shareholding and the labor income share of invested firms. By enhancing human capital quality, promoting digital transformation, and improving corporate governance, CVC can increase the labor income share up to a certain point, beyond which its effects diminish or reverse. These findings align with global trends, highlighting the dual role of technological advancements and capital participation in shaping income distribution. For instance, in advanced economies like the United States and Europe, similar patterns emerge where CVC fosters innovation but may also exacerbate income inequality, underscoring the universal relevance of this study’s insights.

This paper makes three key contributions: (1) It bridges a critical gap in the literature by analyzing the nonlinear relationship between CVC shareholding and labor income distribution, a topic of growing importance globally as economies grapple with rising income inequality. (2) The study identifies mediating mechanisms—R&D personnel ratio, digital transformation, and corporate governance—that provide actionable insights for policymakers and corporate managers, both in China and internationally, to address challenges in balancing innovation with income equity. (3) By highlighting significant regional disparities and the role of financial constraints, this research offers practical policy recommendations adaptable to diverse economic contexts, emphasizing the importance of tailored strategies to harness CVC’s potential benefits while mitigating its potential downsides.

In conclusion, this study contributes to a novel understanding of how CVC influences labor income distribution in science and innovation enterprises. It advances academic research by offering a robust framework to analyze the dual role of CVC and provides practical guidance for firms and policymakers to achieve sustainable development goals. By optimizing CVC ownership structures and emphasizing long-term strategies, firms can simultaneously drive economic growth and foster social equity, contributing to a more inclusive global economy.

8.2. Recommendations

Regarding the findings, the following recommendations are put forth in this paper:

(1)

Optimize CVC shareholding ratios to balance innovation and equity. Firms should conduct data-driven analysis to determine the optimal CVC shareholding ratio, maximizing the labor income share without sacrificing innovation efficiency. This recommendation is applicable globally, particularly in advanced economies like the United States and Europe, where CVC plays a pivotal role in fostering technological innovation. Policymakers in these regions can use these findings to encourage firms to adopt more balanced strategies that align short-term financial goals with long-term social objectives.

(2)

Strengthen talent management and governance structures. Firms should prioritize the design of flexible incentive systems that link employee compensation to performance, especially for R&D personnel. Equity-based incentives can further align the interests of employees and firms, fostering a virtuous cycle of innovation and equitable growth. Globally, policymakers should promote skill enhancement programs to bridge the digital divide and ensure that the benefits of technological transformation are widely distributed. This is particularly critical in both advanced markets, where automation is displacing low-skilled labor, and in emerging markets, where skill gaps hinder equitable growth.

(3)

Tailor strategies to regional disparities and financial constraints. Policymakers in emerging economies should encourage CVC participation through targeted subsidies, tax incentives, and financing channels to reduce regional inequalities and promote labor income share growth. Conversely, in developed economies, regulators should focus on mitigating the potential downsides of excessive capital concentration, such as wage stagnation and widening pay gaps, by implementing stricter disclosure requirements and fostering greater accountability.

(4)

Leverage media as a governance tool and establish globally competitive remuneration systems. Firms should actively engage with media to enhance transparency and align corporate strategies with public expectations. Media attention has been shown to moderate the effects of CVC ownership on labor income share, making it a powerful tool for ensuring accountability. Internationally, businesses in regions with strong media scrutiny, such as the United States and Europe, can integrate corporate social responsibility (CSR) initiatives to maintain public trust. To establish globally competitive remuneration systems, firms should adopt international best practices, including performance-based pay schemes, employee equity incentives, and transparent reporting frameworks. Policymakers could incentivize firms to maintain wage growth aligned with productivity increases, ensuring that innovation-driven growth benefits employees at all levels. In emerging markets with weaker labor protections, governments should strengthen oversight mechanisms and require CVC-backed firms to disclose pay structures, fostering greater accountability and social fairness.

In summary, these recommendations provide actionable strategies for firms and policymakers worldwide to leverage the benefits of CVC while minimizing its potential drawbacks. By adopting a balanced approach tailored to regional and economic contexts, firms can drive sustainable growth, enhance labor income shares, and contribute to a more equitable global economy.