Introduction
As both corporate executives and policymakers increasingly recognize environmental concerns, investing in environmental activities has become a critical component in firms’ strategic priorities across the globe (Aguinis & Glavas, Reference Aguinis and Glavas2012; Gast, Gundolf, & Cesinger, Reference Gast, Gundolf and Cesinger2017; Wang, Wijen, & Heugens, Reference Wang, Wijen and Heugens2018; Whiteman, Walker, & Perego, Reference Whiteman, Walker and Perego2013). According to the UN Environment Programme (2021), global investment in nature-based solutions amounted to $133 billion in 2020, approximately 0.10% of global gross domestic product. Strategic investments in environmental activities provide critical support for firms to develop and maintain their environmental commitment (Hart & Dowell, Reference Hart and Dowell2011; Porter & Van der Linde, Reference Porter, Van der Linde and Wubben1995). For example, with increasing environmental investments, firms can promote strategic actions that will enhance environmental sustainability and general social welfare by developing green innovations and incorporating better health and safety standards (Branzei, Ursacki-Bryant, Vertinsky, & Zhang, Reference Branzei, Ursacki-Bryant, Vertinsky and Zhang2004; Zhang, Song, & Zhang, Reference Zhang, Song and Zhang2023).
That said, firms vary significantly in their environmental investments, especially in emerging markets such as China, where incentives for firms to maintain high environmental commitment are often unclear (Wang et al., Reference Wang, Wijen and Heugens2018). For example, in 2020, while 28 Chinese firms were ranked in the Carbon Clean 200 (As You Sow, 2020) to recognize their notable efforts in environmental activities, 705 other Chinese firms were accused of various environmental violations in the same year, incurring nearly ¥300 million in administrative fines.Footnote 1 To understand the heterogeneities among firms in their environmental investments, prior research has examined how factors such as regulatory pressures, normative expectations, and mimetic behaviors affect an organization's environmental practices (De Villiers, Naiker, & Van Staden, Reference De Villiers, Naiker and Van Staden2011; Dixon-Fowler, Ellstrand, & Johnson, Reference Dixon-Fowler, Ellstrand and Johnson2017; García Martín & Herrero, Reference García Martín and Herrero2020; Walls, Berrone, & Phan, Reference Walls, Berrone and Phan2012; Zhu, Xu, Wang, & Zhang, Reference Zhu, Xu, Wang and Zhang2022). Taking an institutional view, a growing body of research focuses on how institutional factors influence firm environmental practices (Bansal, Reference Bansal2005; Jennings & Zandbergen, Reference Jennings and Zandbergen1995; Yuan, Qin, Zhong, & Nicoleta-Claudia, Reference Yuan, Qin, Zhong and Nicoleta-Claudia2023). For example, Bansal (Reference Bansal2005) demonstrates that the institutional environment, such as media pressure and industry norms, influences firm environmental decisions. Focusing on other institutional drivers, Hanim Mohamad Zailani, Eltayeb, Hsu, and Choon Tan (Reference Hanim Mohamad Zailani, Eltayeb, Hsu and Choon Tan2012) find that government regulations and incentives play significant roles in motivating firms to adopt eco-designs that influence environmental performance. While these insights are fruitful, previous studies have primarily focused on firms’ external institutional factors (Berrone, Fosfuri, Gelabert, & Gomez-Mejia, Reference Berrone, Fosfuri, Gelabert and Gomez-Mejia2013; Yin & Jamali, Reference Yin and Jamali2016), overlooking the influence of firms’ internal institutional environment.
An internal institutional environment refers to the set of conditions, factors, and components within the firm that influence its operations and decision-making (Parkhe, Reference Parkhe2003; Swaminathan & Wade, Reference Swaminathan, Wade, Augier and Teece2016). Recent research suggests that, as one of the important manifestations of firms’ internal institutional environment, the institutional logic within organizations drives important strategic decisions (Cobb, Wry, & Zhao, Reference Cobb, Wry and Zhao2016; Greve & Zhang, Reference Greve and Zhang2017; Zhou, Gao, & Zhao, Reference Zhou, Gao and Zhao2017). Institutional logic refers to the ‘socially constructed historical patterns of cultural symbols and material practices, assumptions, values and beliefs by which individuals and organizations provide meaning to their daily activity’ (Thornton & Ocasio, Reference Thornton and Ocasio1999: 804). Institutional logic can shape how firms define goals and thus make decisions (Greenwood, Raynard, Kodeih, Micelotta, & Lounsbury, Reference Greenwood, Raynard, Kodeih, Micelotta and Lounsbury2011; Pache & Santos, Reference Pache and Santos2010; Thornton, Ocasio, & Lounsbury, Reference Thornton, Ocasio and Lounsbury2012). However, while the influence of institutional logic on firm strategic decisions is well recognized, research has provided limited insights into how this logic affects strategic decisions related to firms’ environmental investments.
Moreover, previous studies on institutional logic and firms’ strategic decisions have mainly focused on the influence of a singular dominant logic (Chung & Luo, Reference Chung and Luo2008; Cobb et al., Reference Cobb, Wry and Zhao2016; Raynard & Greenwood, Reference Raynard and Greenwood2023); yet firms usually face multiple logics simultaneously (Pache & Santos, Reference Pache and Santos2013). Thus, significant goal conflicts could occur among coexisting but incongruent logics, which might result in institutional complexity (Greenwood et al., Reference Greenwood, Raynard, Kodeih, Micelotta and Lounsbury2011). Prior research has yielded mixed findings on the influence of institutional complexity on firm behavior. For example, while some studies argue that institutional complexity threatens firm survival (Tracey, Phillips, & Jarvis, Reference Tracey, Phillips and Jarvis2011), others suggest that firms can exploit the advantages of different logics (Andersson, Reference Andersson2022; Yan, Almandoz, & Ferraro, Reference Yan, Almandoz and Ferraro2021). These research gaps result in limited insights into how institutional complexity resulting from multiple incongruent institutional logics affects firms’ environmental investments.
To address these gaps, this study draws on China's mixed-ownership reform of state-owned enterprises (SOEs) as a research context and investigates how the institutional complexity arising from the reform affects SOEs’ environmental investments. Specifically, traditional SOEs are guided by a ‘state logic’ that emphasizes political and societal goals (Greve & Zhang, Reference Greve and Zhang2017; Pan, Cheng, & Gao, Reference Pan, Cheng and Gao2022). This state logic typically puts a higher value on carrying out government objectives and improving social welfare than on maximizing profitability (Greve & Zhang, Reference Greve and Zhang2017). Conversely, private owners are motivated by a ‘financial logic’, which prioritizes financial returns (Battilana & Dorado, Reference Battilana and Dorado2010; Thornton, Reference Thornton2002). With the two institutional logics having incongruent strategic priorities, a mixed-ownership reform brings institutional complexity to SOEs, which provides an ideal context for studying our research questions.
Departing from previous studies that mainly focus on one dominant logic, we suggest that the institutional complexity resulting from SOE privatization can influence firms’ environmental investments in divergent ways, depending on the relative dominance of each institutional logic. On the one hand, the introduction of financial logic could play an enabling role by enhancing firms’ resource availability for environmental investments because of improved performance (Pan et al., Reference Pan, Cheng and Gao2022). On the other hand, a focus on financial returns of financial logic might also constrain SOEs’ motivations for environmental spending, as most environmental investments do not generate immediate financial rewards (Orsato, Reference Orsato2006). Thus, the effect of privatization on environmental investments may hinge on the relative dominance between state and financial logic. As the dominance of each logic is determined by the level of privatization, we theorize an inverted U-shaped relationship between privatization and environmental investments, with environmental investments increasing initially to a point and then declining at higher levels of privatization.
Furthermore, we suggest that the relative dominance of each institutional logic may shift as the perceived relative importance of a particular institutional logic changes in an organization (Greenwood & Hinings, Reference Greenwood and Hinings1996; Lounsbury, Reference Lounsbury2001). Previous studies indicate that two mechanisms can significantly reshape the perceived relative importance of different institutional logics within an organization (Besharov & Smith, Reference Besharov and Smith2014; Pache & Santos, Reference Pache and Santos2010). Internally, organizational members (e.g., CEOs) can influence an organization's primary institutional logic through their personal beliefs and preferences (Camelo-Ordaz, Hernández-Lara, & Valle-Cabrera, Reference Camelo-Ordaz, Hernández-Lara and Valle-Cabrera2005; Hambrick & Mason, Reference Hambrick and Mason1984). Externally, actors in an organization's field environment (e.g., industry rivals and local government) can also shape an organization's preferred institutional logic by generating pressures on the firm's operations. Accordingly, we examine the moderating effects of two sets of factors that can reshape the relative importance of state and financial logic in an SOE. First, to capture the influence of organizational members, we examine how CEOs’ background characteristics, including their political and financial backgrounds, influence their perceived relative importance of state and financial logic, moderating the relationship between SOE privatization and environmental investments. Second, investigating the influence of external actors, we explore the moderating effects of industry competition and regional government–market relationship quality (GMRQ). We propose that they influence the prevailing institutional logic through market competition and the intervention of the local government, thus influencing privatized SOEs’ perceived importance of state or financial logic in making decisions relating to environmental investments.
To test our ideas, we use multisource panel data from Chinese listed SOEs with mixed ownership from 2013 to 2020 and find that the level of privatization has an inverted U-shaped relationship to an SOE's environmental investments. Moreover, our findings reveal that CEOs’ background characteristics and external environmental context moderate this curvilinear relationship.
This study offers several important insights. First, it investigates the influence of an important but understudied internal institutional factor – namely, institutional logic – on firms’ environmental investments. While previous studies have focused on the influence of external institutional factors (Berrone et al., Reference Berrone, Fosfuri, Gelabert and Gomez-Mejia2013; Yin & Jamali, Reference Yin and Jamali2016), few have paid attention to the influence of firms’ internal institutional environment. Our findings thus offer a new perspective on how a firm's institutional environment affects its environmental strategies.
Second, this study explores a unique condition in which institutional complexity results from incongruent institutional logics in the process of SOE privatization and reveals an inverted U-shaped relationship between privatization and environmental investments. Previous research has yet to reach an agreement on the effect of institutional complexity on firm strategic decisions (e.g., Andersson, Reference Andersson2022; Reay & Hinings, Reference Reay and Hinings2005; Siwale, Kimmitt, & Amankwah-Amoah, Reference Siwale, Kimmitt and Amankwah-Amoah2021). Thus, our study joins this conversation by demonstrating that the influence of institutional complexity resulting from SOE privatization on firm strategic behaviors (e.g., environmental investments) is not linear but changes as the relative dominance of each institutional logic changes.
Third, we provide a comprehensive understanding of a firm's decisions relating to environmental investments by exploring the influences from both internal organizational members (i.e., CEOs’ political and financial backgrounds) and the pressures from external actors in an SOE's field environment (i.e., industry competition and regional GMRQ). We find that both these factors can reshape the relative importance of state or financial logic within SOEs, affecting how privatization affects SOE environmental investments.
Theoretical Background
Firm Heterogeneity in Environmental Investments: An Institutional View
While both firm executives and policymakers increasingly recognize concerns about environmental sustainability, firms vary significantly in their level of environmental investments, especially when the incentives to maintain high environmental commitment are unclear. For example, the costs of being environment-friendly for Chinese firms, especially in heavily polluting industries, are often very high as they require significant investment in new equipment, production redesign, and cross-functional coordination (Wang et al., Reference Wang, Wijen and Heugens2018). At the same time, consumers’ awareness of environmental protection and their appreciation of corporate environmental efforts are still relatively low in China (Duanmu, Bu, & Pittman, Reference Duanmu, Bu and Pittman2018).
A growing body of research relies on the institutional view to understand firms’ environmental practices. According to this view, firms exist within institutional environments that have a substantial impact on them (Oliver, Reference Oliver1991). Specifically, the formal regulations, informal social norms, and cultural beliefs constituting these environments greatly shape firm behaviors and strategic decision-making (Kraatz & Block, Reference Kraatz, Block, Greenwood, Oliver, Suddaby and Sahlin-Andersson2008). Drawing on this view, previous studies have provided substantial insights into how the institutional environment affects a firm's environmental practices. For example, Berrone et al. (Reference Berrone, Fosfuri, Gelabert and Gomez-Mejia2013) argue that higher external institutional pressures related to environmental concerns encourage companies, particularly those with notable pollution levels, to pursue environmental innovations. However, a review of the literature reveals that most studies primarily emphasize the role of firms’ external institutional environments in driving their environmental practices, thereby paying relatively scant attention to internal institutional factors. The lack of insight into the influence of the internal institutional environment results in an incomplete understanding of firms’ environmental practices.
To address this gap, we aim to examine the influence of an important internal institutional factor – namely, institutional logic – on firms’ environmental investments. An institutional logic is a socially constructed, coherent, and integrated set of assumptions, values, beliefs, and rules that help individuals and organizations imbue their daily activities with meaning (Thornton & Ocasio, Reference Thornton and Ocasio1999). These factors provide actors with ‘organizing principles’ prescribing legitimate ends and ‘the means by which those ends are achieved’ (Friedland, Reference Friedland, Powell and DiMaggio1991: 248). As such, research has shown that institutional logic can significantly influence important firm strategic decisions (Greenwood et al., Reference Greenwood, Raynard, Kodeih, Micelotta and Lounsbury2011; Pache & Santos, Reference Pache and Santos2010; Thornton, Reference Thornton2002). For example, Chung and Luo (Reference Chung and Luo2008) suggest that institutional logic can shape beliefs about legitimate and efficient corporate strategies, influencing firms’ restructuring process. Given the importance of institutional logic in driving firm strategic decisions, we expect institutional logic to influence organizations’ decisions on environmental investments. However, departing from previous studies focusing mainly on one dominant logic, we explore a unique condition in which a firm is facing multiple coexistent but incongruent logics, which often leads to institutional complexity (Greenwood et al., Reference Greenwood, Raynard, Kodeih, Micelotta and Lounsbury2011; Pache & Santos, Reference Pache and Santos2010). Thus, we investigate how such institutional complexity shapes a firm's environmental strategies, affecting its investments in environmental activities.
SOE Privatization in China and Institutional Complexity
SOEs, which often receive significant resource support from the government, act as agents of the state to carry out the state's policies and regulations (Shleifer, Reference Shleifer1998), particularly in addressing environmental issues (Lin, Lu, Zhang, & Zheng, Reference Lin, Lu, Zhang and Zheng2020). However, as Zhou et al. (Reference Zhou, Gao and Zhao2017) note, SOEs frequently grapple with inefficiencies in resource utilization, constraining their capacity to allocate resources to environmental initiatives (Wang, Liu, & Zhang, Reference Wang, Liu and Zhang2022). To address the low-efficiency problem in SOEs, China has implemented a series of reforms since the 1980s (Bai, Lu, & Tao, Reference Bai, Lu and Tao2009). In 2013, the government launched the mixed-ownership reform of SOEs. The reform involves introducing partial private capital to SOEs (i.e., privatization) to leverage the market expertise and heightened efficiency associated with private ownership (Pan et al., Reference Pan, Cheng and Gao2022; Radić, Ravasi, & Munir, Reference Radić, Ravasi and Munir2021). Despite partial privatization helping improve efficiency in SOEs, it also results in the coexistence of incongruent institutional logics within organizations.
On the one hand, SOEs inherently follow a state logic stemming from state ownership. State logic focuses on ‘rationalization and the regulation of human activity by legal and bureaucratic hierarchies’ (Friedland, Reference Friedland, Powell and DiMaggio1991: 248). It drives firms to serve national interests and prioritize political and societal objectives (Park, Li, & Tse, Reference Park, Li and Tse2006). Because SOEs are the main vehicles for implementing the government's policies and regulations (Shleifer, Reference Shleifer1998), their strategic decisions are motivated by the government's goals, such as promoting environmental practices. Following state logic, firms will prioritize investments in green technology, emissions reductions, and sustainability to serve the public good, even though these investments may not directly contribute to immediate financial returns.
On the other hand, the infusion of private capital introduces a financial logic focused primarily on ‘accumulation and the commodification of human activity’ (Friedland, Reference Friedland, Powell and DiMaggio1991: 248). Financial logic demands that organizations prioritize shareholder value, profit, and cost-efficiencies. Following the financial logic, firms strive for financial returns and pursue maximum shareholder value (Friedman, Reference Friedman1970; Greve & Zhang, Reference Greve and Zhang2017; Jensen, Reference Jensen2002). Through this lens, environmental investments are often considered an unnecessary cost burden that eats into profits without clear financial rewards. This is especially true for Chinese firms in heavily polluting industries because these firms need to spend significantly on environmental activities to improve the environmental condition. Yet the financial rewards of such activities are often unclear, and the opportunity cost of not following the regulations is relatively low. As the state and financial logic present different strategic priorities, the privatization process in SOEs will thus likely produce institutional complexity.
Hypotheses Development
SOE Privatization and Firm Environmental Investments
As the institutional view suggests, a firm's institutional environment can influence its environmental investments (Oliver, Reference Oliver1991) because this environment often determines the motivations and available resources for the firm's environmental activities (Pinder, Reference Pinder2014). We propose that the financial logic introduced in the process of privatization likely affects both resource availability and the motivation for environmental activities, but in different ways.
First, financial logic may play an enabling role by increasing the available resources to support environmental investments. Prior research suggests that, through privatization, SOEs can improve their firm performance by benefiting from improved resource efficiency and transferrable knowledge provided by financial logic, such as financial expertise and market intelligence (Bai et al., Reference Bai, Lu and Tao2009; Cao, Qian, & Weingast, Reference Cao, Qian and Weingast1999; Pan et al., Reference Pan, Cheng and Gao2022). As better performance usually provides a firm with more slack resources (Daniel, Lohrke, Fornaciari, & Turner, Reference Daniel, Lohrke, Fornaciari and Turner2004), SOEs can thus allocate more expenditures to support their environmental activities. Second, financial logic may also play a constraining role by reducing SOEs’ motivation to invest in environmental activities. In contrast with state logic, which aims to carry out state policies and focuses on social welfare, financial logic focuses more on maximizing financial returns (Greve & Zhang, Reference Greve and Zhang2017; Jensen, Reference Jensen2002). Strategies that do not yield short-term financial returns may not be prioritized under this logic. Given the limited immediate financial returns of environmental investments, particularly for polluting firms, which need to spend considerably to improve environmental conditions, the introduction of a financial logic through privatization could dampen SOEs’ environmental activities.
Because privatized SOEs face both state and financial logics and these two logics present different strategic priorities, we suggest that the relative dominance of these two logics, which shifts as the level of privatization changes, will influence firms’ strategic decision-making. When the level of privatization is low to moderate, privatized SOEs are still primarily owned by the state. Under this condition, state logic is the relatively dominant logic within the organization, and the influence of financial logic on the firm's strategic decision-making is limited. Under this condition, financial logic mainly plays the enabling role by helping SOEs improve performance, resulting in more resources for environmental investments. As mentioned previously, state logic emphasizes improving social welfare, offering SOEs a strong motivation to engage in environmental behaviors. With more resources brought by financial logic, privatized SOEs are likely to invest more in environmental activities. As a result, the level of environmental investments will increase as the level of privatization increases.
When the level of privatization increases further (e.g., moderate to high levels), the financial logic becomes the relatively dominant logic within organizations. Following the financial logic, firms will focus more on maximizing financial returns than carrying out government policies and improving social welfare. Although the financial logic can still help SOEs improve performance, this marginal benefit will become incremental with the further increase of privatization. Under this condition, the constraining role of financial logic is more salient, which is likely to reduce these SOEs’ motivations for environmental investment. Therefore, the environmental investment level will decrease as privatization further increases. Overall, we propose the following:
Hypothesis 1 (H1): SOE privatization has an inverted U-shaped relationship to firm environmental investments.
Boundary Conditions: The Perceived Relative Importance of Institutional Logic
Organizations may vary in their responses to incongruent institutional logics because the broader regulatory, social, and cultural environments they face can influence their interpretations of different institutional logics (Greenwood & Hinings, Reference Greenwood and Hinings1996; Lounsbury, Reference Lounsbury2001). When they interpret that an institutional logic is more important for the firm to maintain its competitive advantage, the perceived importance of such institutional logic will be enhanced. Previous studies suggest that two mechanisms can significantly influence such interpretations, reshaping the perceived relative importance of a particular institutional logic (Besharov & Smith, Reference Besharov and Smith2014; Pache & Santos, Reference Pache and Santos2010).
First, internal organizational members, such as executives, can influence an organization's understanding of organizational goals and the selection of appropriate means to achieve these goals by selectively drawing on, interpreting, and enacting an institutional logic they adhere to (Scott, Reference Scott2001; Thornton, Reference Thornton2002). As key strategic decision-makers, CEOs play a critical role in shaping firm behaviors through their personal beliefs and preferences (Camelo-Ordaz et al., Reference Camelo-Ordaz, Hernández-Lara and Valle-Cabrera2005; Hambrick & Mason, Reference Hambrick and Mason1984). Accordingly, we expect that CEOs’ characteristics that may influence their preference for the state of financial logic may reshape the perceived relative importance of that logic. Following this logic, we examine the moderating effects of CEOs’ political and financial backgrounds on the relationship proposed in H1.
Second, actors outside an organization (i.e., the field environment) can also influence the perceived relative importance of a particular institutional logic by exerting compliance pressures on the organization (Pache & Santos, Reference Pache and Santos2010). Two common actors that are highly relevant in an SOE's decisions about environmental investments might be competitive rivals and regulatory bodies, such as local governments (Pache & Santos, Reference Pache and Santos2010). While competition from rivals might increase the perceived importance of financial logic, local government intervention might influence the perceived importance of state logic. SOEs typically need to conform in their operations to the prevailing institutional logic within these outside actors to secure survival and access to key resources (DiMaggio & Powell, Reference DiMaggio and Powell1983; Oliver, Reference Oliver1991). Following this logic, we examine the moderating effects of industry competition and regional GMRQ on the relationship between SOE privatization and firm environmental investments.
Moderating Effects of CEO Background Characteristics
CEO's political background
CEOs in many SOEs have a political background (Lin et al., Reference Lin, Lu, Zhang and Zheng2020; Zhang, Zhou, & Tian, Reference Zhang, Zhou and Tian2022). We identify CEOs as having a political background if they had or currently have working experience in one of the four key government or political organizations: the government, the Communist Party of China, the National People's Congress, and the Chinese People's Political Consultative Conference. CEOs with political backgrounds tend to have more political experiences and/or close ties with the government than those without such backgrounds (Chen, Sun, Tang, & Wu, Reference Chen, Sun, Tang and Wu2011). Moreover, CEOs with political backgrounds are often embedded in the political system (Cao, Lemmon, Pan, Qian, & Tian, Reference Cao, Lemmon, Pan, Qian and Tian2019). As a result, they are more likely to follow state logic and be responsive to the government's policies when making important decisions (Li & Lu, Reference Li and Lu2020). Wang and Wang (Reference Wang and Wang2013) show that CEOs with political backgrounds are more likely to undertake the government's excess employment goals, which leads to an increase in redundant employees in firms. Building on this reasoning, we expect a CEO's political background to influence an SOE's environmental investments by increasing the perceived relative importance of state logic when private capital is introduced into the firm.
Specifically, when SOE privatization increases from low to moderate levels, the CEO's political background can dampen financial logic's enabling role (i.e., improved resource efficiency). As discussed previously, the financial logic introduced in the process of privatization may improve firms’ resource efficiency by changing the incentive systems, organizational structures, and behaviors. As important strategic decision-makers, CEOs will promote practices, norms, and values that they like to prioritize (Oliver, Reference Oliver1991). Compared with other CEOs, those with political backgrounds are more rooted in state logic and are more attentive to strategic priorities that follow state logic. They, thus, might be less receptive to financial logic when the level of privatization is low. Under this condition, the enabling role of privatization in improving resource efficiency may be weakened by CEOs with political backgrounds.
When SOE privatization changes from moderate to high levels, the CEO's political background can also dampen the constraining role (i.e., reduced motivation) arising from dominant financial logic. Despite financial logic intensifying pressures to maximize profits and reduce environmental motivations in firms’ strategic decisions, CEOs rooted in state logic tend to absorb state environmental goals as a personal values-based leadership vision and persist in state-driven environmental policies amid privatization pressure. Under this condition, the role of privatization in reducing firms’ motivation for environmental investments might also be weakened by CEOs’ political backgrounds. Thus, we propose the following:
Hypothesis 2 (H2): The inverted U-shaped relationship between SOE privatization and environmental investments is weakened by CEOs with political backgrounds, such that the shape of the relationship is flatter for SOEs whose CEOs have political backgrounds than those without.
CEOs’ financial background
CEOs’ financial background reflects their working and educational experience in financial industries. In general, CEOs with financial backgrounds have a better understanding of financial knowledge and are usually more cautious about financial returns when making important strategic decisions (Oradi, Asiaei, & Rezaee, Reference Oradi, Asiaei and Rezaee2020). As a result, they are more steeped in financial logic and tend to favor strategies that may help a firm improve its financial conditions (Custódio & Metzger, Reference Custódio and Metzger2014). For example, Jiang, Zhu, and Huang (Reference Jiang, Zhu and Huang2013) find that CEOs with financial backgrounds have more accounting knowledge and can take advantage of their expertise in making financial decisions, thus improving earning quality. As such, we expect a CEO's financial background to influence how privatization affects an SOE's environmental investments by increasing the perceived relative importance of financial logic in this process.
When SOE privatization increases from low to moderate levels, the financial logic introduced by privatization mainly plays an enabling role in providing more resources to support firms’ environmental investments. As CEOs with financial backgrounds are more cognizant of financial logic, they will actively embrace the changes brought about by private capital and strengthen the influence of financial logic, helping their SOEs efficiently accumulate more resources for environmental investments. Under this condition, the enabling role of the financial logic in supporting an SOE's environmental investments will be strengthened.
As SOE privatization changes from moderate to high levels, the financial logic becomes the dominating logic within SOEs, and it mainly takes on a constraining role (i.e., reduced motivation) in a firm's environmental investments. CEOs with financial backgrounds are more attentive to financial logic and thus are likely to be more reluctant to invest in environmental initiatives, which often require substantial up-front costs without immediate payoffs, than CEOs without such backgrounds. Under this condition, the constraining role of financial logic in influencing a firm's environmental decisions will be enhanced. Consequently, the negative effect of privatization on environmental investments will be amplified. Overall, we expect CEOs’ financial backgrounds to strengthen the inverted U-shaped relationship between SOE privatization and environmental investments by enhancing financial logic's influence during the privatization process. Thus:
Hypothesis 3 (H3): The inverted U-shaped relationship between SOE privatization and environmental investments is strengthened by CEOs with financial backgrounds, such that the shape of the relationship is steeper for SOEs whose CEOs have financial backgrounds than those without.
Moderating Effects of External Environmental Context
Industry competition
Industry competition captures the level of competition a firm faces in its industry (Porter & Advantage, Reference Porter and Advantage1985). In highly competitive industries, SOEs are under higher pressure to compete with their private counterparts. SOEs in these industries will thus have stronger incentives to improve their resource efficiency through privatization to survive than those in low-competitive industries. In addition, when industry competition is higher, the intervention from the government also tends to be lower because the competition is mainly driven by market needs, and SOEs need to prioritize their strategies to meet those needs instead of fulfilling the government's goals (Ramaswamy, Reference Ramaswamy2001; Zhou et al., Reference Zhou, Gao and Zhao2017). As a result, financial logic is more prevailing in high-competitive industries than in low-competitive industries.
As mentioned previously, when SOE privatization increases from low to moderate levels, privatization mainly plays a positive enabling role in an SOE's environmental investments by enhancing its resource efficiency. A high level of industry competition may further strengthen this positive effect. Specifically, SOEs in highly competitive industries face higher compliance pressures to emphasize profitability by optimizing their incentive systems, structures, and behaviors (Radić et al., Reference Radić, Ravasi and Munir2021). Under this condition, privatized SOEs are more likely to increase their reliance on financial logic to improve resource efficiency, helping them survive intense competition (Johnson, Smith, & Codling, Reference Johnson, Smith and Codling2000). However, while the emphasis on financial logic is higher, the state logic is still the relatively dominant force in driving firm strategic priorities under low levels of privatization. Thus, the firm will continue to invest more in environmental activities with increased resource efficiency brought about by privatization.
When SOE privatization changes from moderate to high levels, financial logic becomes more prevalent in the organization and increasingly influences strategic decisions. In industries where market competition is more intense, the financial logic is more prevailing, and the government's intervention becomes less influential. Privatized SOEs in these high-competitive industries will rely more on meeting market needs than on government goals. Under this condition, SOEs will be more inclined to follow financial logic and reduce environmental investments to maintain their market competitiveness and profitability in a highly competitive market (Duanmu et al., Reference Duanmu, Bu and Pittman2018). As such, we expect industry competition to strengthen the inverted U-shaped relationship between SOE privatization and environmental investments by enhancing the relative importance of financial logic during the privatization process. Thus:
Hypothesis 4 (H4): The inverted U-shaped relationship between SOE privatization and firm environmental investments is strengthened by industry competition, such that the shape of the relationship is steeper for SOEs in industries with higher market competition.
Regional government–market relationship quality
GMRQ reflects regional institutional development and captures the intervention of the government in local economic activities and business practices (Wang, Fan, & Hu, Reference Wang, Fan and Hu2019). In regions with poor government–market relationships, the government often intervenes in business activities and treats SOEs as essential political instruments by imposing social and political obligations on them (Hoskisson, Eden, Lau, & Wright, Reference Hoskisson, Eden, Lau and Wright2000; Shen, Zhou, & Zhang, Reference Shen, Zhou and Zhang2022), thus strengthening the state logic within privatized SOEs. According to the National Bureau of Statistics of China (2014), in Zhejiang Province, which has a good government–market relationship, economic resources dominated by the government account for only 12.84% of the province's gross domestic product, much lower than that in Gansu Province (37.17%), where the relationship between government and market is worse. This indicates that the government in Zhejiang Province has less control over resource allocation and intervenes less in the market competition than that in Gansu Province.
Moreover, in regions where the GMRQ is lower, SOEs tend to rely more on support from the government than in other regions where the relationship quality is higher (Gao & Hafsi, Reference Gao and Hafsi2015; Zhou et al., Reference Zhou, Gao and Zhao2017). As a result, they often lack market-based skills, experience, and capabilities (Shen et al., Reference Shen, Zhou and Zhang2022). By contrast, in regions with better government–market relationships, the government usually promotes free market competition. Thus, it intervenes less in business activities than in regions with low GMRQ. Consequently, SOEs in those regions tend to face a lower level of social and political obligations, mitigating the influence of state logic in their strategic decisions.
When SOE privatization increases from low to moderate levels, state logic is the relative dominant logic within the organization, and privatization mainly plays an enabling role by enhancing SOEs’ resource efficiency. Compared with regions with lower GMRQ, SOEs in regions with higher GMRQ not only face a lower level of compliance pressures from the government but also have more market-oriented combabilities (Tang, Reference Tang2019), enabling them to improve resource efficiency with the help of financial logic. Under this condition, the enabling role of privatization in enhancing SOEs’ environmental investments will be strengthened.
When SOE privatization changes from moderate to high levels, the relative dominant logic in strategic decisions gradually shifts from state logic to financial logic, and privatization mainly acts negatively by reducing firms’ motivation for environmental investments (i.e., a constraining role). In regions with better government–market relationships, government intervention in privatized SOEs is limited, and financial logic becomes more prevailing than in regions with poor government–market relationships (Shen et al., Reference Shen, Zhou and Zhang2022). Under this condition, privatized SOEs in these regions are also more likely to rely on financial logic when developing their environmental strategies. Consequently, SOEs may further reduce their environmental investments as privatization increases. Overall, we expect regional GMRQ to strengthen the inverted U-shaped relationship between SOE privatization and environmental investments by reducing the relative importance of state logic during the privatization process. Therefore, we posit that:
Hypothesis 5 (H5): The inverted U-shaped relationship between SOE privatization and firm environmental investments is strengthened by regional government–market relationship quality, such that the shape of the relationship is steeper for SOEs located in regions with better relationships.
Methods
Data and Sample
For our sample, we focused on SOEs listed on the Shenzhen or Shanghai stock exchange between 2013 and 2020 (inclusive) that with mixed ownership during this period. We excluded firms with incomplete information and those labeled as special treatment or particular transfer, which indicates that a firm is in dire financial condition, may be delisted (Peng, Wei, & Yang, Reference Peng, Wei and Yang2011), and thus may be incentivized to manipulate its reports to avoid the designation (Firth, Rui, & Wu, Reference Firth, Rui and Wu2011). We chose heavily polluting industriesFootnote 2 for two reasons. First, these industries offer an ideal context to test the influence of institutional complexity on environmental investments. Specifically, as China is increasingly concerned about environmental issues, the government often targets heavily polluting industries to implement its environmental protection policies. As a result, the Chinese government is expending increasing efforts to help firms in these industries, especially SOEs, reduce their negative impact on environmental issues. Therefore, the institutional complexity arising from multiple logics for privatized SOEs is relatively more common in heavily polluting industries than in other industries.
Second, listed companies in heavily polluting industries are required to regularly disclose environmental information on pollutant emissions, environmental compliance, environmental management, and the like, according to the ‘Guidelines for Environmental Information Disclosure of Listed Companies’ issued by the Ministry of Ecology and Environment in September 2010. By contrast, listed companies in nonheavy-polluting industries are not required to disclose their annual environmental reports but are encouraged to do so. As discussed subsequently, we obtain the environmental investment data by analyzing the environmental information disclosed in firms’ annual reports. Therefore, focusing on heavily polluting industries enables us to obtain more accurate data.
We collected data from multiple sources. First, we collected data on SOE privatization and firms’ basic information from the China Stock Market and Accounting Research database. Second, we manually coded data on environmental investments from annual reports of listed firms and the website of China Information (www.cninfo.com.cn). Third, we collected data on government–market relationships from the National Economic Research Institute, a prestigious nongovernmental research center in China. Prior research on Chinese firms has also commonly used these databases (e.g., Li & Qian, Reference Li and Qian2013; Sun, Hu, & Hillman, Reference Sun, Hu and Hillman2016; Zhang, Liang, Li, & Wang, Reference Zhang, Liang, Li and Wang2022). The final sample consists of 2,718 firm-year observations of 447 companies.Footnote 3
Measures
Dependent variable
The dependent variable is environmental investments (EnvInv). To measure this variable, we relied on expenditures reported in ‘management fee’ and ‘construction in progress’ on annual reports, which reflect firms’ endeavors in environmental protection (Zhang, Yu, & Kong, Reference Zhang, Yu and Kong2019). Following Guo and Wei (Reference Guo and Wei2023) and Kong, Liu, Wang, and Zhu (Reference Kong, Liu, Wang and Zhu2024), we first manually identified these expenditures using a set of environmental protection-related keywords, such as desulfurization projects, sewage treatment, waste gas, and greening projects. We constructed the list of environmental protection-related keywords (see Appendix) from recommendations from previous studies (e.g., Guo & Wei, Reference Guo and Wei2023; Xie, Reference Xie2020; Zhang et al., Reference Zhang, Yu and Kong2019). To further confirm the managerial relevance of these keywords, we randomly mixed these environmental protection-related keywords with other firm cost items to form a survey questionnaire and surveyed senior managers from similar firms to those included in our sample. In the survey, respondents rated the degree to which they believed the items listed in the survey were related to a firm's environmental protection activities. The results showed a high overlap (96%) between the selected environmental protection-related keywords in the survey and the full keywords listFootnote 4, indicating that the selected keywords are appropriate. We then aggregate all related expenditures as the firm's environmental investments. Considering the potential influence of firm size, we use total assets to scale environmental investments and then multiply it by 100 to increase the readability of the regression coefficient.
Independent variable
The independent variable SOE privatization (SOEpri) reflects the degree of SOEs’ mixed-ownership reform. We calculated SOE privatization as the total shareholding ratio of nonstate shareholders among the top 10 shareholders (Pan et al., Reference Pan, Cheng and Gao2022).
Moderating variables
The first moderating variable is CEO political background (CEO_PB). As previous studies suggest (Jia & Zhang, Reference Jia and Zhang2013; Pache & Santos, Reference Pache and Santos2010), CEOs with political backgrounds often hold political positions or have governmental experience. As members of government agencies, they are generally socialized into and trained in a specific institutional logic consistent with the government's ideals. Therefore, CEOs with political backgrounds are likely inclined to follow and prefer the state logic in their decisions. Following previous studies (Du, Bai, & Chen, Reference Du, Bai and Chen2019; Zheng, Singh, & Mitchell, Reference Zheng, Singh and Mitchell2015), we use CEOs’ past political experiences to identify their political background. Because CEOs are highly unlikely to obtain high-level political connections through education, we focus on their past working-related political experience in the four key government and political organizations mentioned previously. Consistent with prior research (Li & Lu, Reference Li and Lu2020; Zheng et al., Reference Zheng, Singh and Mitchell2015), we measure CEO_PB as a dummy variable, coded as 1 if a CEO has currently or previously worked in one of the four government agencies and 0 otherwise.
The second moderating variable is CEO financial background (CEO_FB). CEOs with financial backgrounds often have working or educational experience in finance, which means they are more adept in financial situations than other CEOs (Oradi et al., Reference Oradi, Asiaei and Rezaee2020). Therefore, they are likely more steeped in financial logic and more cautious about financial returns in strategic decisions (Custódio & Metzger, Reference Custódio and Metzger2014). We code CEO_FB as a dummy variable that takes the value 1 if a CEO has financial experience and 0 otherwise. Following previous studies (Custódio & Metzger, Reference Custódio and Metzger2014; Kalelkar & Khan, Reference Kalelkar and Khan2016; Liu, Zhou, Chan, & Chen, Reference Liu, Zhou, Chan and Chen2020), we classify CEOs as having financial experience if they have worked in banking, investment, or auditing firms; were employees of a stock exchange; or have an academic degree in accounting, economics, or finance.
We measured industry competition (IndCom), our third moderating variable, using the Herfindahl–Hirschman index (HHI), a widely accepted measurement of industry competition in economics (e.g., Flammer, Reference Flammer2015; Zou, Zeng, Lin, & Xie, Reference Zou, Zeng, Lin and Xie2015). Specifically, we calculated HHI as the sum of the squared market share of each firm competing in an industry based on the total sales of that industry. Considering that the HHI is a negative index and that a higher HHI indicates less competition, we reversely code IndCom using 1 – HHI to increase the readability of the regression coefficient (Pan, Chen, Sinha, & Dong, Reference Pan, Chen, Sinha and Dong2020; Zhou et al., Reference Zhou, Gao and Zhao2017).
Following previous studies (Shen et al., Reference Shen, Zhou and Zhang2022; Tan, Tian, Zhang, & Zhao, Reference Tan, Tian, Zhang and Zhao2020), we measured GMRQ, our fourth moderating variable, using the provincial government–market relationship index developed by the National Economic Research Institute of China (Wang et al., Reference Wang, Fan and Hu2019). The GMRQ index measures the relationship between the local government and the market in a province, reflecting the degree of government intervention in business activities (Gao & Hafsi, Reference Gao and Hafsi2015; Shen et al., Reference Shen, Zhou and Zhang2022; Yang et al., Reference Yang, Zhang, Luo, Tang, Gao, Wan and Zhang2019). It is an aggregate concept comprising three indicators: (1) the proportion of economic resources allocated by the market, (2) government intervention in business-related administration procedures (reversed), and (3) the size of the government relative to the provincial population (reversed). A larger value of GMRQ indicates a better government–market relationship and that the local government barely intervenes in the market.
Control variables
In line with previous studies investigating environmental issues among Chinese firms (e.g., Wei & Zhou, Reference Wei and Zhou2020; Zeng, Cheng, Jin, & Zhou, Reference Zeng, Cheng, Jin and Zhou2022), we added multiple control variables that may be associated with an SOE's environmental investments. These include firm size (Size), measured as the natural logarithm of total assets; leverage (Lev), measured as year-end total liabilities divided by year-end total assets; return on assets (ROA), calculated as net profit divided by total assets; board size (Board), measured as the natural logarithm of the number of board directors; board independence (Indep), measured as the number of independent directors relative to the total number of directors; CEO duality (Dual), coded as 1 if the chairman and the CEO are the same person and 0 otherwise; equity balance degree (Balance), measured as the sum of shares owned by the second to fifth major shareholders divided by the total shares owned by the largest shareholderFootnote 5; and management share (Mshare), measured as the number of shares held by management divided by the total shares outstanding.
Estimation Methods
We employed panel ordinary least squares models with firm, year, industry, and province fixed effects to test our hypotheses. Because we focus on privatized SOEs in heavily polluting industries, sample selection bias may potentially threaten our results. To address this issue, we employed a two-stage Heckman (Reference Heckman1979) estimation procedure in our examinations. In the first stage, we use a probit regression to predict the likelihood of firms being in heavily polluting industries using all privatized SOEs in multiple industries from 2013 to 2020. In the second stage, we include the inverse Mills ratio calculated on the basis of the results obtained from the first stage as an additional control variable to capture potential selection bias. To alleviate the potential threat of reverse causality, we lag all explanatory and control variables by one year in our models.
Results
Table 1 reports the descriptive statistics and correlations of the key variables. The descriptive analysis shows that, on average, SOEpri is 0.207, indicating that the degree of privatization of the SOEs in our sample is 20.7%. We calculated values of the variance inflation factors to check for multicollinearity in the analyses. The values ranged from 1.02 to 1.73, much lower than the threshold value of 10 (Cohen, Cohen, West, & Aiken, Reference Cohen, Cohen, West and Aiken2013). Thus, our results are unlikely to suffer from significant multicollinearity problems.
Table 1. Descriptive statistics and correlations matrix of variables
Notes: N = 2,718; Correlations with an absolute value greater than 0.02 are significant at the 10% level.
Hypotheses Testing Results
Table 2 reports the results of the regression analysis testing our hypotheses. Specifically, Model 1 presents the results of the first-stage selection model, and Models 2–8 report the results of the second-stage regressions.
Table 2. Results of two-stage Heckman regression analysis
Notes: N = 2,718; t-statistics are in parentheses. ***p < 0.01, **p < 0.05, *p < 0.1.
H1 predicts that SOE privatization has an inverse U-shaped relationship to environmental investments. As Model 3 shows, the coefficient for the squared term of SOEpri is negative and significant (b = –1.875, p < 0.05), suggesting an inverted U shape. Following the approach suggested by Haans, Pieters, and He (Reference Haans, Pieters and He2016), we used the utest command in Stata to formally test the U-shaped relationship. The results show that the turning point of the curvilinear relationship occurs when the SOE privatization value is 0.372, with the 95% Fieller (Reference Fieller1954) confidence interval being [0.084, 0.858]. A joint test of the positive slope at lower values of SOE privatization (b = 1.329, p < 0.05) and a negative slope at higher values of SOE privatization (b = –2.180, p < 0.05) is significant. This evidence suggests that the inverted U-shaped relationship between SOE privatization and environmental investments is statistically significant. Furthermore, the results indicate that SOEs whose privatization level is less than 37.0% will invest more in environmental activities with an increase in SOE privatization. By contrast, SOEs with a privatization level greater than 37.2% will engage less in environmental investments as the SOE privatization increases. To better interpret this effect, we graphed Model 3 in Figure 1. Overall, H1 is supported.
Figure 1. Effect of SOE privatization on environmental investments
H2 predicts that CEO political background (CEO_PB) will weaken the inverted U-shaped relationship between SOE privatization and environmental investments. As the results in Model 4 in Table 2 show, the coefficient of the interaction between CEO political background and the squared term of SOE privatization (CEO_PB × SOEpri2) is positive and significant (b = 7.760, p < 0.01), suggesting a flatter curve of the curvilinear relationship. This result is consistent with that in Model 8 (b = 8.078, p < 0.01), which is the complete model including all variables. Therefore, H2 is supported. We plot the results of Model 4 in Figure 2. As the figure illustrates, firms with a higher level of CEO political background show a flatter inverted U shape, indicating a weaker main effect. At the curve's apex, the effect of SOE privatization on environmental investments decreases by 3.9% (3.63 vs 3.49) for firms whose CEOs do not have (vs have) a political background.
Figure 2. The moderating effect of CEO political background
H3 predicts that CEO financial background (CEO_FB) will strengthen the inverted U-shaped relationship between SOE privatization and environmental investments. As the results in Model 5 in Table 2 show, the coefficient of the interaction between CEO financial background and the squared term of SOE privatization (CEO_FB × SOEpri2) is negative and statistically significant (b = –5.747, p < 0.01), suggesting a steepened curve of the curvilinear relationship (Haans et al., Reference Haans, Pieters and He2016). This result is also consistent with that in Model 8 (b = –5.606, p < 0.01). Therefore, H3 is supported. We plot the results of Model 5 in Figure 3. The figure illustrates that firms whose CEOs have a financial background are associated with a steeper inverted U shape, indicating a stronger main effect. At the curve's apex, the effect of SOE privatization on environmental investments increases by approximately 32.1% (2.9 vs 3.83) for firms whose CEOs do not have (vs have) a financial background.
Figure 3. The moderating effect of CEO financial background
Model 6 in Table 2 tests the moderating effect of industry competition (IndCom) on the relationship between SOE privatization and environmental investments. As expected, the coefficient of the interaction term between industry competition and the squared term of SOE privatization (IndCom × SOEpri2) is negative and significant (b = –9.595, p < 0.01), which is similar to that in Model 8 (b = –8.478, p < 0.01). These results suggest a steeper curve of the curvilinear relationship. Therefore, H4 is also supported. Figure 4 plots how different levels of industry competition affect the main relationship. As the figure illustrates, firms with a higher level of industry competition are associated with a steeper inverted U shape, indicating a stronger main effect. At the curve's apex, the effect of SOE privatization on environmental investments increases by approximately 23.2% (1.94 vs 2.39) when comparing firms facing a lower level of industry competition (one standard deviation below the mean) with those facing a higher level of industry competition (one standard deviation above the mean).
Figure 4. The moderating effect of industry competition
Model 7 in Table 2 tests the moderating effect of GMRQ on the relationship between SOE privatization and environmental investments. As expected, the coefficient of the interaction term between government–market relationship quality and the squared term of SOE privatization (GMRQ × SOEpri2) is negative and significant (b = –1.119, p < 0.01). The coefficient of the interaction term in Model 8 is also negative and significant (b = –1.089, p < 0.01). Therefore, H5 is supported. Figure 5 plots how different levels of GMRQ affect the main relationship. As the figure shows, the shape of the curve becomes steeper as the level increases. Specifically, the results indicate that, at the curve's apex, the effect of SOE privatization on environmental investments increases by approximately 19.5% (2.31 vs 2.76) when comparing firms in provinces with lower GMRQ with those in provinces with a higher quality.
Figure 5. The moderating effect of government–market relationship quality
Robustness Checks
Endogeneity concerns
We adopted the control function approach as a robustness test to check the potential threat due to endogeneity problems. This approach includes the first-stage regression's residuals as an additional predictor in the second-stage regression. The residuals substitute for unobserved confounders that could potentially influence the endogenous regressor, thus alleviating endogeneity concerns (Pavićević & Keil, Reference Pavićević and Keil2021; Petrin & Train, Reference Petrin and Train2010; Wooldridge, Reference Wooldridge2015).
In the first stage, we predicted the key endogenous variable, SOE privatization, using all exogenous predictors from the main model along with an instrumental variable – the average value of SOE privatization at the provincial level. Theoretically, the average value of SOE privatization in the province, which indicates the general trend in this province, should be strongly associated with a focal firm's SOE privatization. However, a province's privatization trend does not necessarily affect an individual firm's environmental investments (Wooldridge, Reference Wooldridge2015). After estimating the first-stage model using the instrument, we obtained the residuals and used them as an additional control variable to predict environmental investments in the second-stage regression. Table 3 reports the results of the second-stage analyses. In Model 2, the coefficient of the squared term of SOEpri is negative and significant (b = –1.863, p < 0.05), in further support of H1. In addition, the results related to the moderating effects are consistent with those in Table 2. Given this evidence, endogeneity is unlikely to be an issue in this study.
Table 3. Robustness test using control function approach
Notes: N = 2,718; t-statistics are in parentheses. ***p < 0.01, **p < 0.05, *p < 0.1.
Reverse causality concerns
Another issue that can potentially threaten our results is reversed causality. Although we lagged all predicting variables in our regressions to mitigate this problem, we adopted panel Granger-causality tests to further assess the causal relationships between the dependent and independent variables (Abrigo & Love, Reference Abrigo and Love2016; Love & Zicchino, Reference Love and Zicchino2006). The results reject the null hypothesis that the independent variable (SOEpri2) does not lead to the dependent variable (EnvInv) (p < 0.05), suggesting that SOEpri2 can indeed cause the change of EnvInv. However, the results fail to reject the null hypothesis that the dependent variable (EnvInv) does not cause the change of the independent variable (SOEpri2) (p = 0.904). In summary, the results of the Granger-causality tests suggest that reversed causality is not an issue in our study.
Sensitivity to sample selection
Our sample comprises listed privatized SOEs in heavily polluting industries, which may lead to sample selection bias in our results. To mitigate the potential threat, we first adopted a Heckman two-stage estimation procedure for correction, as the previous empirical results showed that selection bias may not be a threat. To further address selection bias, we performed a robustness check using a new sample with all Chinese listed firms in multiple industries and reran the panel ordinary least squares regressions with firm, year, and industry fixed effects. As Table 4 shows, the results with the new sample are similar to those of our main regressions, confirming the robustness of our results.
Table 4. Robustness test using an alternative sample
Notes: N = 8,941; t-statistics are in parentheses. ***p < 0.01, **p < 0.05, *p < 0.1, ap = 0.107.
In addition, we adopted a propensity score matching (PSM) procedure to examine the robustness of our results (Rubin, Reference Rubin2006). Specifically, we used the PSM procedure with a caliper of 0.15 to construct a new group in which firms in other industries are comparable to firms in heavily polluting industries by matching them on all exogenous predictors from the main model. After matching, we conducted a balance test to confirm the effectiveness of the matching and found no significant difference in the matching variables between the two groups. We then used the matched sample to retest our hypotheses. We found broadly consistent support for our hypotheses, as shown in Table 5.
Table 5. Robustness test using PSM
Notes: N = 8,892; t-statistics are in parentheses. ***p < 0.01, **p < 0.05, *p < 0.1.
Discussion
While environmental concerns are increasingly driving firms’ strategic decisions, insights into firms’ heterogeneous environmental investments are limited (De Villiers et al., Reference De Villiers, Naiker and Van Staden2011; Dixon-Fowler et al., Reference Dixon-Fowler, Ellstrand and Johnson2017; Zhu et al., Reference Zhu, Xu, Wang and Zhang2022). To close this gap, we focus on a unique condition in which firms face institutional complexity resulting from multiple but incongruent institutional logics when making decisions about environmental investments. In contrast with studies that have mainly focused on a single institutional logic (Chung & Luo, Reference Chung and Luo2008; Cobb et al., Reference Cobb, Wry and Zhao2016), we suggest that multiple incongruent institutional logics result in institutional complexity for firms, which in turn affects their environmental investments, depending on the relative dominance of each institutional logic.
Using China's mixed-ownership reform as a research context, we hypothesize and find an inverted U-shaped relationship between SOE privatization and environmental investments. The effect of institutional complexity brought about by privatization is nonlinear because the financial logic introduced by the privatization process plays both enabling and constraining roles in SOEs’ strategic decisions about environmental investments. The extent to which the enabling or the constraining role has a more significant impact on firm decisions depends on the relative dominance of each logic. As our findings show, at low to moderate privatization levels, financial logic plays an enabling role by bringing more available resources to facilitate environmental activities that are aligned with the strategic priorities of state logic. However, at higher privatization levels, when financial logic becomes the dominant logic, firms’ motivation in environmental investments is constrained by the lack of immediate financial payoffs. In particular, we find that the positive role of privatization in facilitating environmental investments reaches a peak point when the level of privatization is around 37%. This finding provides new insights into the mixed findings on the influence of institutional complexity on firm behavior (Andersson, Reference Andersson2022; Raynard & Greenwood, Reference Raynard and Greenwood2023; Reay & Hinings, Reference Reay and Hinings2005; Siwale et al., Reference Siwale, Kimmitt and Amankwah-Amoah2021).
Moreover, our findings reveal that factors that reshape the relative importance of different institutional logics further moderate the curvilinear relationship between SOE privatization and environmental investments. These factors may come from both organizational members and external actors in an SOE's field environment. Regarding organizational members, CEOs play critical roles in important strategic decisions. As a result, the beliefs and preferences resulting from their professional backgrounds will influence their perceived relative importance of a state or financial logic. Regarding external actors in the field environment, the compliance pressures from the market competition and intervention of the local government may also influence the perceived importance of each institutional logic within SOEs. Thus, the industry competition and regional GMRQ will likely influence the perceived importance of each institutional logic. As the perceived relative importance of a particular institutional logic changes, the impact of institutional complexity on an SOE's decision about environmental investments will be further reshaped.
Theoretical Contributions
This study makes several contributions to the literature. First, it contributes to research exploring how institutional factors affect firms’ environmental strategies. While research has examined the role of external institutional pressures, such as regulations, normative expectations, and cultural-cognitive frameworks (e.g., Berrone et al., Reference Berrone, Fosfuri, Gelabert and Gomez-Mejia2013; Yin & Jamali, Reference Yin and Jamali2016), few studies have investigated the impact of firms’ internal institutional factors on their environmental decision-making. Our study extends this line of research by delving into the impact of institutional logic – a critical but underexplored internal institutional factor – on firms’ environmental strategies. By revealing how divergent institutional logics shape a firm's environmental investment decisions, this study sheds new light on the mechanisms through which its institutional context shapes its approach to environmental initiatives.
Second, this study contributes to research on institutional complexity by exploring a unique condition in which institutional complexity results from the SOE privatization process and mapping an inverted U-shaped correlation between institutional complexity induced by different levels of SOE privatization and environmental investments. Scholarship has yet to reach a consensus on the influence of institutional complexity on strategic firm behaviors and decisions, with empirical findings remaining mixed and inconclusive (e.g., Andersson, Reference Andersson2022; Reay & Hinings, Reference Reay and Hinings2005; Siwale et al., Reference Siwale, Kimmitt and Amankwah-Amoah2021). Our study joins this debate by demonstrating that the impact of institutional complexity arising from multiple, potentially incongruent institutional logics on organizational behaviors, such as environmental investments, is not linear but varies depending on the relative dominance of each logic. This nuanced finding provides new insights to help better understand the complex influence of institutional complexity on firm behaviors.
Third, the study also contributes to the literature on strategic decision-making in the context of institutional complexity by exploring the influences of both organizational members and external actors in an organization's field environment on reshaping firms’ strategic environmental investments. Previous studies have provided theoretical anticipations for the influences of these two sources on a firm's important strategic decisions when faced with incongruent institutional logics (Besharov & Smith, Reference Besharov and Smith2014; Pache & Santos, Reference Pache and Santos2010). Yet, few studies have provided comprehensive empirical evidence to support those propositions. Our findings provide new insights by showing that both CEO background characteristics (i.e., CEO political and financial background) and pressures from the field environment (i.e., industry competition and regional GMRQ) can reshape the relative importance of a state or financial logic within SOEs, thereby further influencing how privatization levels ultimately affect firms’ environmental investments. These findings highlight the important interplay between organizational and external field actors in strategic decision-making under institutional complexity.
Managerial Implications
Our findings have several important implications for SOEs and policymakers. First, SOEs should be aware that private capital can play dual roles. More importantly, we inform managers at SOEs that privatization can, under certain circumstances, result in an adverse impact on SOEs’ long-term-focused strategic behaviors. Specifically, while privatization can help SOEs improve resource efficiency, it may lead to an excessive focus on profit maximization and constrain their investment in environmental activities. Therefore, while taking advantage of the benefits of privatization, SOEs should also be aware of the risks associated with excessive privatization. As our findings suggest, a moderate level of privatization is an optimal solution for SOEs.
Second, SOEs should realize that both CEO characteristics (e.g., CEO political and financial background) and external environmental context (e.g., industry competition and regional GMRQ) can reshape the influence of privatization on a firm's investment decision on environmental activities. Thus, during the process of privatization, SOEs should carefully examine their CEOs’ personal characteristics and external institutional environments to make proper decisions about their environmental activities. For example, privatized SOEs whose CEOs have a financial background, those that face strong industry competition, and those located in regions with higher GMRQ should pay more attention to their environment investments in the process of privatization, as these investments are more sensitive to the level of privatization.
Finally, our research also offers important insights for policymakers. Policymakers in China have long attempted to address the inefficiencies of SOEs. Our findings suggest that partial privatization of SOEs can significantly boost their efficiency and also improve their profitability. Therefore, policymakers should be aware that privatization has its own limitations in driving SOEs to make proper organizational changes. Large-scale reforms may not always produce more favorable outcomes. Instead, policymakers should consider both SOEs’ developmental needs and their local market environments when making decisions about their ownership reforms.
Limitations and Future Research Directions
This study has several limitations that offer directions for future research. First, this study only examines institutional complexity through the context of SOEs’ privatization. Previous studies suggest that other types of institutional complexity also exist within organizations. For example, Reay and Hinings (Reference Reay and Hinings2009) suggest that the market logic followed by the commercial sector and the professional logic followed by the technical sector also create conflicts within firms, resulting in institutional complexity. We thus encourage future research to investigate how other types of institutional complexity may affect firms’ strategic behavior.
Second, the research context may limit the generalizability of our findings. While China is a leading emerging market, its government is particularly strong in influencing firms’ activities (Zheng & Huang, Reference Zheng and Huang2018). However, institutional environments in emerging economies are constantly changing. Therefore, we encourage future research to improve the generalizability of our findings by testing our research ideas in other emerging economies.
Finally, we did not test the potential influence of heterogeneity of private ownership. During the mixed-ownership reform, SOEs may bring in various types of private capital, which may present considerable heterogeneity in their objectives, interests, and resources in influencing their strategies and behaviors (Ferreira & Matos, Reference Ferreira and Matos2008). Future research could thus benefit from further investigating this potential interaction effect.
Data availability statement
The data and code that support the findings of this article are openly available in the Open Science Framework at https://osf.io/ethr3/
Acknowledgments
The second author, Yu Chang, acknowledges the support provided by the Key Program of the National Social Science Fund of China under Grant 22AGL005.
Appendix I List of environmental protection-related keywords
Huiying Li ([email protected]) is a PhD candidate in management at the School of Management, Northwestern Polytechnical University. Her research interests include innovation, strategic management, and corporate environmental behavior. Her research has been published in journals, such as IEEE transactions on Engineering Management, Bulletin of Economic Research, and Academy of Management Proceedings, among others.
Yu Chang ([email protected]) is a professor and chair of the marketing department at Northwestern Polytechnical University. Her research focuses on innovation management and marketing strategy. She leads three National Social Science Foundation projects and has published in leading strategy and marketing journals, such as Journal of Business Research, Industrial Marketing Management, and Journal of Business & Industrial Marketing, among others.
Xinchun Wang ([email protected]) is an associate professor of marketing at the John Chambers College of Business and Economics, West Virginia University. His research interests center on issues relating to firm innovation problems and marketing-related problems in entrepreneurial firms. His research has been published in leading strategy and marketing journals, including Journal of Product Innovation Management, Journal of Business Research, and Industrial Marketing Management, among others.
Nan Zhang ([email protected]) is a postdoctoral researcher at the School of Management, Northwestern Polytechnical University. His research interests include corporate nonmarket behavior, corporate governance, and corporate innovation. His research has been published in IEEE transactions on Engineering Management, Bulletin of Economic Research, Chinese Management Studies, and Journal of Cleaner Production, among others.
