INTRODUCTION
Ownership structure is a key characteristic of corporate governance and holds important implications for organizational strategy and productivity (Hill & Snell, Reference Hill and Snell1989). State ownership – organization's property interest is vested in the state or a public body representing the state – is still an important phenomenon in the world economy (La Porta, Lopez-De-Silanes, & Shleifer, Reference La Porta, Lopez-De-Silanes and Shleifer1999). State-Owned Enterprises (SOEs) contribute approximately 10% of the worlds’ GDP (Bruton, Peng, Ahlstrom, Stan, & Xu, Reference Bruton, Peng, Ahlstrom, Stan and Xu2015) and still exist in countries including the United States, Germany, France, Italy, China, Brazil (Pargendler, Reference Pargendler2012), Canada, the United Kingdom, New Zealand, Finland, Sweden, Hungary, Norway, the Philippines, Malaysia, Singapore, Egypt, Serbia, Turkey, Bhutan, Chile, Ghana, Kenya, India, Pakistan, Peru, South Africa, Zambia, Korea, and many other countries (World Bank, 2014). Therefore, Peng, Bruton, Stan, and Huang (Reference Peng, Bruton, Stan and Huang2016) suggest that state ownership is one of the mainstream organizational forms and holds important implications for organizational theories. Existing research on state ownership has focused on its implications for organizational performance (Le & O'Brien, Reference Le and O'Brien2010; Ramaswamy, Reference Ramaswamy2001) and found that state ownership is on average associated with decreased labor productivity and corporate performance (Chen, Reference Chen2001; Xu & Wang, Reference Xu and Wang1999). However, privatization reform, which aims to reduce state ownership, has only achieved mixed results (Dharwadkar, George, & Brandes, Reference Dharwadkar, George and Brandes2000). Because employees are an important stakeholder of organizations and driver of organizational performance (Aguilera & Jackson, Reference Aguilera and Jackson2010), how state ownership and its reduction influence employee behavior may provide a clue about the transition effect. However, very few studies have paid attention to their experiences under state ownership (Fiss, Reference Fiss, Greenwood, Oliver, Suddaby and Sahlin2008). Therefore, more research is needed on the micro foundation of ownership reform (Cuervo & Villalonga, Reference Cuervo and Villalonga2000; Parker, Reference Parker1995). A few micro studies found that employees in SOEs prefer egalitarian distribution (He, Chen, & Zhang, Reference He, Chen and Zhang2004) and perform extra-role behaviors that align with SOEs’ goals (Farh, Zhong, & Organ, Reference Farh, Zhong and Organ2004). However, it is unclear what motives drive SOE employees to engage in extra-role behaviors. To fill this gap, this study focuses on how organizations with different ownership structure motivate employee behavior that is important for organizational performance – extra-role behavior.
In this study, I propose that SOEs rely on a distinctive mechanism to motivate employees’ extra-role behavior. SOEs foster a strong collective identity among employees and encourage them to treat their enterprises as families. This objective is achieved through widespread participation in organizational governance. As the reform emphasizing market efficiency and organizational productivity proceeds, privatized companies reinforce employees’ concern about their individual outcomes and utilize equitable allocation of outcomes to motivate employees. Thus, employees in SOEs and privatized companies will hold different relational models about their relationships with their organizations and pay attention to organizational practice that is consistent with their relational model. In particular, I suggest that the communitarian and egalitarian relational model in SOEs accentuates positive reactions to procedural justice, whereas the market and exchange relational model in privatized companies accentuates positive reactions to distributive justice. To test this argument, the present research draws upon the variability in ownership structure of organizations during China's transition to explore how ownership structure moderates the effects of distributive justice and procedural justice on employees’ extra-role behavior.
The current work contributes to existing theory and research in a number of ways. First, I bridge micro and macro research on China's economic reform (Naughton, Reference Naughton1996; Ramamurti, Reference Ramamurti2000), developing a theory regarding how ownership structure influences employee relational model and reactions to organizational practices. This is key to understanding the micro-foundations of institutional reform. I suggest that ownership structure is not associated with different levels of employee motivation (Burawoy & Lukacs, Reference Burawoy and Lukacs1985), but rather associated with different predictors of employee motivation. Thus, whether these differences are acknowledged and addressed properly during the transition may help explain the mixed results of transition effects in macro research.
Second, by exploring how ownership structure serves as a key contextual factor altering the well-known relationship between organizational justice and extra-role behavior, I offer new insights into the boundary conditions of justice effects. Because different ownership structures rely on different approaches to motivate employees, it will shape which mechanism of justice is operative – social exchange or social identity. Specifically, distributive justice, by emphasizing equitable allocation of outcomes, is especially important in soliciting employee extra-role behavior in privatized organizations. Procedural justice, by verifying individuals’ collective identity, plays an important role in state-owned organizations. Thus, I show how different justice dimensions matter in different contexts.
Finally, this study holds important implications for designing reform measures in transition economies. Previous reform efforts are based on the assumption that SOE managers and employees lack incentives and motivation, and the main reform measure is to privatize SOEs. This study uncovers an unrecognized predictor of employee extra-role behavior in SOEs. I suggest that SOEs do not simply lack motivation but motivate employees in different ways than privatized organizations. Recognizing this new motive can generate more creative ways to reform and manage SOEs, given the increasing presence of state ownership around the globe (The Economist, 2012).
THEORETICAL BACKGROUND AND HYPOTHESES
In this article, I propose that ownership structure is associated with the relational model that individuals hold regarding their relationships with their organizations and their primary concerns in organizations. Ownership structure and the role of labor are two important dimensions in corporate governance (Aguilera & Jackson, Reference Aguilera and Jackson2010). Because ownership structure defines the social relations among organizational actors (Fiss, Reference Fiss, Greenwood, Oliver, Suddaby and Sahlin2008), it will influence how employees construe their relationships with their organizations. In addition, ownership structure shapes organizational goals (Fiss, Reference Fiss, Greenwood, Oliver, Suddaby and Sahlin2008), which will influence the primary concerns among employees. For instance, a state owner focuses more on organizational solidarity and treats workers as ‘master of enterprises’ (Chiu, Reference Chiu2006), whereas private owners may focus more on organizational productivity and treat labor as a means to achieve organizational productivity. Thus, their relative weights in the ownership structure of organizations will influence how employees construe their relationships with their organizations – as enterprise master or exchange partners – and whether they are primarily concerned with social identity or exchange outcomes. The relational schemas that employees hold and their primary concerns will direct their attention to different organizational practices (Thornton, Ocasio, & Lounsbury, Reference Thornton, Ocasio and Lounsbury2012), with implications for which form of justice is important for guiding employee extra-role behavior. I selected China as the empirical setting to examine this proposal because its transition from planned economy to market economy generates great diversity in the ownership structure of organizations.
In the context of China, I suggest that state ownership and privatization reform activate different relational models people apply to their organizations. These relational models are the cognitive schemas that people use to process complex information and guide their actions. These cognitive schemas are constructed in a specific institutional environment and mediate the impact of institutions on human behaviors (Seo & Creed, Reference Seo and Creed2002). Therefore, only one or a few relational models are salient in a specific context, and economic development generally moves the relational model from communal sharing to market pricing (Fiske, Reference Fiske1992). Applying this theory to employee-organization relationships in China, I argue that the communal sharing model is dominant in SOEs, and the market pricing model is dominant in privatized companies. The following sections will develop these arguments based on previous research.
State Ownership in China
State ownership is an important characteristic of China's pre-reform socialist economy. China adopted the Soviet model shortly after its revolution (Jackson, Reference Jackson1992), leading SOEs to play a central role in the composition of its planned economy (Groves, Yongmiao, McMillan, & Naughton, Reference Groves, Yongmiao, McMillan and Naughton1994). State ownership is associated with obligations and privileges that have evolved historically. In particular, SOEs have alternative goals beyond financial performance, such as maintaining political stability, increasing employment, and providing public facilities (Walder, Reference Walder1989; Zif, Reference Zif1981). In conjunction with these obligations, they have greater access to government loans and purchases and face softer budget constraints (Bai & Wang, Reference Bai and Wang1998; Dong & Putterman, Reference Dong and Putterman2003). The incentive system of SOEs has been historically characterized by high social benefits and low cash wages (Walder, Reference Walder1983), both of which are allocated equally within classes of workers (Giacobbe-Miller, Miller, & Zhang, Reference Giacobbe-Miller, Miller and Zhang1997). SOEs also provide employment security and used to grant all employees lifelong employment until retirement (Naughton, Reference Naughton1996). SOE employees have more secure employment than employees of non-state-owned organizations (Gong & Chang, Reference Gong and Chang2008).
SOEs’ equal treatment of workers in their need satisfaction foster a communal sharing schema (Fiske, Reference Fiske1992) in which those within a group are not differentiated and taken care of by the group. Based on interviews with 30 employees and a survey of 500 employees of two SOEs in northeast China, Liu (Reference Liu2003) found that SOEs emphasize group solidarity and treat employees as families. This family culture is further intensified by the socialist ideology, which regards workers as the ‘leading class’ and ‘master of socialist society’ (Wang & Greenwood, Reference Wang and Greenwood2015). This ideology cultivated a collective identity of SOE employees as the ‘master of enterprises’ (zhurenweng) (Chiu, Reference Chiu2006). This identity is manifested in their reactions to unemployment. Due to their sense of class status and entitlement, laid-off workers prevalently experienced loss of face and a sense of betrayal (Mok, Wong, & Lee, Reference Mok, Wong and Lee2002; Wang & Greenwood, Reference Wang and Greenwood2015), and state workers resorted to this identity to defend their rights and resist SOEs’ downsizing effort, as illustrated in interviews with eight steel SOEs throughout China in 1997 (Hassard, Morris, Sheehan, & Yuxin, Reference Hassard, Morris, Sheehan and Yuxin2006). SOEs' family culture and collective identity have been acknowledged by private companies during their acquisition of SOEs, as documented in a case study (Xing & Liu, Reference Xing and Liu2016). The above studies unanimously show that SOEs activate a communal sharing relational model among SOE employees, as reflected in their culture and collective identity as ‘master of enterprises’.
A very important institutional embodiment of that identity is the widespread participation opportunities of SOE employees in the governance of SOEs. SOEs institutionalize widespread employee participation in organizational affairs through daily production meetings, yearly workers’ congress, various management committees, and incentive-suggestion systems (Tang, Reference Tang1993; Walder, Reference Walder1981). Although these participation opportunities are limited by central planning and party control over leadership selection (Walder, Reference Walder1981), they still have a symbolic function of enhancing the communal sharing model in SOEs. Based on case studies of six enterprises in Shanghai from 1997 to 1998, Benson and Zhu (Reference Benson and Zhu1999) found that SOEs are characterized by traditional management systems including teamwork, information sharing, and harmonious work conditions. In these organizations, unions and work congress participate in important organizational decision making, including redundancy decisions, organizational restructuring, developing training, and welfare and housing. In addition, employee creativity is recognized as an asset during ownership transformation, and employees can receive shares or form new companies based on their adopted ideas (Benson & Zhu, Reference Benson and Zhu1999). Participation opportunity has become so deeply grounded in SOE employees’ relational model that they rely on it to evaluate new management practices. For instance, a study of 194 employees from four SOEs in northeast China in 2008 showed that procedural justice, especially being able to participate in the performance appraisal process, is very important for employees’ perception of the system (Tsai & Wang, Reference Tsai and Wang2013). Therefore, procedural justice, especially the opportunity to participate in organizational decision making, plays an important role in upholding SOE employees’ collective identity and sustaining their motivation.
Gradualist Reform in China
China's economy has experienced an unprecedented change as it transitions from a centrally-planned socialist economy to a market-oriented capitalist economy (Guthrie, Reference Guthrie1999). However, the capitalist economy in China is centrally-managed capitalism, where the state plays an important role in many aspects of the economy (Lin, Reference Lin2011). Under this background, ownership reform in China adopts a gradualist approach, with SOEs coexisting with organizations of hybrid ownership (Nee, Reference Nee1992). For example, SOEs are partially privatized via public listing on stock exchanges, building joint ventures with local or foreign private firms, or transferring property rights to private holders (Walder, Reference Walder1995). Analogous to other gradual reforms in China, this privatization reform is only a partial one in the sense that the state remains as a dominant shareholder in many privatized firms. Typically, public listing of SOEs allows the state to retain between 40% and 50% of the company's shares. Between 20% and 30% of the shares are designated for institutional shares, and the remaining 30% are designated for public consumption as free-floating shares (Guthrie, Reference Guthrie1999; Xu & Wang, Reference Xu and Wang1999). In joint ventures, foreign parent companies often control half or more of the shares of joint ventures, and state-owned parent companies hold the other half or less (Guthrie, Reference Guthrie1999). These privatized companies constitute an important sector in China's economy (Walder, Reference Walder2011). The privatization reform abolished privileges provided by the government, tightened budget constraints, and increased market pressure in privatized companies (Zahra, Ireland, Gutierrez, & Hitt, Reference Zahra, Ireland, Gutierrez and Hitt2000). As a result, privatized companies place a higher priority on efficiency and productivity as organizational goals than SOEs do (He et al., Reference He, Chen and Zhang2004).
Along with change in ownership structure, management practices of privatized companies are different from SOEs as well. The capitalist market exchanges human activities based on prices generated from market competition (Friedland & Alford, Reference Friedland, Alford, Powell and DiMaggio1991). Under this logic, privatized companies allocate rewards based on individual performance and contribution to the organization (Giacobbe-Miller, Miller, Zhang, & Victorov, Reference Giacobbe-Miller, Miller, Zhang and Victorov2003), provide opportunities for career advancement to motivate their employees (Gong & Chang, Reference Gong and Chang2008), and use meritocracy as the basis for promotion (Zhao & Zhou, Reference Zhao and Zhou2004). That is, the incentive system of privatized organizations follows the rule of equity (Chen, Meindl, & Hui, Reference Chen, Meindl and Hui1998). Although SOEs are also increasingly adopting these labor practices, such as bonus payment and piece-rate wages (Groves et al., Reference Groves, Yongmiao, McMillan and Naughton1994; Keister, Reference Keister2002), their movement towards a modern human resource management system is constrained by government involvement and union strength (Benson & Zhu, Reference Benson and Zhu1999). According to a survey of 600 Chinese companies in 2003, SOEs adopt less strategic human resource management practices than foreign-invested enterprises and private-owned enterprises (Ngo, Lau, & Foley, Reference Ngo, Lau and Foley2008).
With the change in ownership structure and management practices of privatized companies, the relationship between employees and their organizations change as well. As privatized companies reward individuals according to their contribution, employees will adopt the market pricing model – in which they view their relationship with organizations as exchanges and pay attention to the ratio between their output and input (Guthrie, Reference Guthrie2002). Accordingly, they will expect that every effort that they contribute to the organization is fairly rewarded. Indeed, compared to SOE employees, employees of public firms and joint ventures have a stronger preference for equitable allocation based on individual contribution and a lower preference for equal allocation of outcomes (Choi & Chen, Reference Choi and Chen2007; He et al., Reference He, Chen and Zhang2004). At the same time, the introduction of private owners and multinational companies downplays the status of workers in corporate governance (Hassard, Morris, & Sheehan, Reference Hassard, Morris and Sheehan2002). Indeed, the majority of workers perceive stricter management control after the reform (Chiu, Reference Chiu2006), and privatized organizations, such as private enterprises and joint ventures, have lower employee participation than SOEs (Chiu, Reference Chiu2002). As a result, employees of privatized companies may be less likely to regard themselves as ‘master’ of their organizations than SOE employees. The different relational models in SOEs and privatized companies can also be evidenced by employees’ organizational commitment. SOE employees had higher continuance commitment than employees of private enterprise (Chiu, Reference Chiu2002) and foreign-invested enterprises (Wang, Reference Wang2004). In contrast, employees of foreign-invested enterprises perceive higher value congruence with organizations than SOE employees, due to their common interest in the exchange relationship.
Overall, previous research has found a significant difference between SOEs and privatized companies in organizational practices and employee models. SOEs create a family culture and activate a communal sharing model among employees. Accordingly, SOE employees participate widely in organizational activities and develop the collective identity as ‘master of enterprises’. In contrast, privatized companies emphasize productivity and foster a market pricing schema. As a result, employees regard themselves as exchange partners of their organizations and expect fair treatment for their contribution. In the following sections, I explicate that the different relational models in SOEs and privatized companies will influence how employees react to organizational justice.
Organizational Justice and Extra-Role Behavior
An important indicator of employees’ contribution to their organizations is extra-role behavior (Van Dyne, Cummings, & Parks, Reference Van Dyne, Cummings and Parks1995). Extra-role behaviors are those that lie outside of formal role requirements, are directed toward benefitting the organization, and are not explicitly rewarded (Van Dyne et al., Reference Van Dyne, Cummings and Parks1995).[Footnote 1] Extra-role behaviors reflect employees’ engagement with the organization and have been associated with important organizational outcomes, such as sales, efficiency, quality, and customer satisfaction (Organ, Podsakoff, & MacKenzie, Reference Organ, Podsakoff and MacKenzie2006; Podsakoff, Whiting, Podsakoff, & Blume, Reference Podsakoff, Whiting, Podsakoff and Blume2009). Indeed, it may be extra-role behaviors that partly justify the existence of organizations within markets by lowering transaction costs and increasing coordination at little tangible cost to the firm (Coase, Reference Coase1937; Williamson, Reference Williamson1981). Previous research has found that the ownership structure of organizations is related to different forms of extra-role behaviors that individuals perform (Farh et al., Reference Farh, Zhong and Organ2004). For instance, because SOEs incorporate social welfare as an indicator of organizational performance, SOE employees engage more in behaviors that are community-oriented, such as participating in social welfare, than non-SOE employees. On the other hand, because SOEs place a lower emphasis on organizational efficiency than non-SOEs, SOE employees engage less in extra-role behaviors that enhance organizational efficiency – such as taking initiatives and saving and protecting organizational resources – than non-SOE employees. In the current study, I take a different perspective and focus on extra-role behaviors that have been widely regarded as important to all kinds of organizations. I examine how ownership structure influences predictors of these extra-role behaviors that are essential for organizational effectiveness.
I argue that ownership reform changes people's models of their relationships with organizations, and the different relational models of employees in SOEs and privatized companies will lead them to pay attention to different organizational practices. Perceived justice is an important way through which organizational practices increase employees’ extra-role behaviors. Organizational justice includes distributive justice – the allocation of outcomes according to individuals’ performance and contribution – and procedural justice – the fair process of organizational decision making, such as allowing individuals to participate (Colquitt, Reference Colquitt2001). Both distributive justice and procedural justice robustly predict employees’ motivation and extra-role behaviors (Cohen-Charash & Spector, Reference Cohen-Charash and Spector2001). However, individuals don't take all forms of justice into consideration when making a decision about their extra-role behaviors (Lind, Reference Lind, Greenberg and Russell2001). Instead, they rely on the most salient form of justice in their environment to make an overall evaluation of their organization and rely on that evaluation to guide their extra-role behaviors (Lind, Kray, & Thompson, Reference Lind, Kray and Thompson2001). Based on relational models theory (Fiske, Reference Fiske1992), I propose that which form of justice becomes salient in a context will depend on the relational model in that context because relational model guides individuals’ information processing. Information that is consistent with the model is attended to and becomes salient, whereas information that is inconsistent with the model is ignored or downplayed (Fiske, Reference Fiske1992; Sanchez-Burks, Nisbett, & Ybarra, Reference Sanchez-Burks, Nisbett and Ybarra2000). As a result, the salience of distributive justice and procedural justice in a specific context depends on the dominant relational model in that context.
Based on cross-cultural research, I suggest that distributive justice is especially salient under the market pricing model, whereas procedural justice is especially salient under the communal sharing model. When people hold the market pricing model, they are primarily concerned with the ratio of their output to their input, which is exactly the definition of distributive justice (Adams, Reference Adams1965). Social exchange theory suggests that individuals strive to maximize the resources they receive in social exchanges, and distributive justice makes sure that individuals’ inputs into social exchange are fairly rewarded (Colquitt et al., Reference Colquitt, Scott, Rodell, Long, Zapata, Conlon and Wesson2013; Organ, Reference Organ1990). Because distributive justice is fair distribution of outcomes, people pay attention to distributive justice when their instrumental concern is highlighted. For instance, distributive justice is especially important for countries high in materialism, such as China and Korea (Kim & Leung, Reference Kim and Leung2007), or when the productivity goal is highlighted (Chen et al., Reference Chen, Meindl and Hui1998). In contrast, when the relational concern is highlighted, people attend to procedural justice to make sense of their relational status. Because procedural justice, such as whether people can participate in the decision making process, carries expressive value of how people are treated in their groups and helps individuals address their relational concern (Tyler, Reference Tyler1989, Reference Tyler1994). According to the relational model of justice, procedural justice signals to individuals that they are valued members of their groups and plays an important role in enhancing employee cooperation in organizations (Tyler & Blader, Reference Tyler and Blader2000). Procedural justice leads individuals to interpret their interactions with organizations as social relationships, rather than economic transactions (Tyler, Degoey, & Smith, Reference Tyler, Degoey and Smith1996; Tyler & Lind, Reference Tyler and Lind1992), and such interpretation reinforces the communal sharing model. Empirical research also shows that procedural justice is especially important when people care about their relationships (Kwong & Leung, Reference Kwong and Leung2002). For instance, procedural justice is especially important for people with interdependent self-construal, i.e., those who define themselves according to their relationships (Brockner, De Cremer, van den Bos, & Chen, Reference Brockner, De Cremer, van den Bos and Chen2005). When their highlighted relational concerns are addressed by procedural justice, they don't pay attention to distributive justice any more (Brockner, Chen, Mannix, Leung, & Skarlicki, Reference Brockner, Chen, Mannix, Leung and Skarlicki2000; Kwong & Leung, Reference Kwong and Leung2002). All of these studies suggest that procedural justice is especially salient when people are primarily concerned with relationships, whereas distributive justice plays a large role when people are primarily concerned with outcomes.
The Moderation Effect of Ownership Structure
In this section, I argue that the effects of organizational justice will depend on the ownership structure of organizations. As stated above, the ownership structure of organizations is associated with the dominant relational models employees hold, and the relational models will direct people's attention to the justice practice that is consistent with the dominant model. Specifically, the communal sharing model in SOEs lead SOE employees to pay attention to procedural justice to address their relational concern, whereas the market pricing model in privatized companies will lead employees to resort to distributive justice to address their instrumental concern. Thus, the ownership structure of organizations will moderate the impact of organizational justice on individual behaviors.
I propose that the communal sharing model in SOEs reinforces the importance of procedural justice in verifying employees’ collective identity. Scholars have drawn upon social identity theory to explain why procedural justice elicits extra-role behaviors – described in the group engagement model (Tyler & Blader, Reference Tyler and Blader2003). This model suggests that organization members have a need to belong to the organization and therefore identify with it (Lind & Tyler, Reference Lind and Tyler1988; Tyler, Reference Tyler1989), and procedural justice satisfies that need because it enhances members’ sense of pride and respect (Tyler & Blader, Reference Tyler and Blader2000). I argue that this identity mechanism of justice is especially salient in highly state-owned organizations, where employees hold the communal sharing model and value their organizational membership. Because procedural justice enhances their pride about their organizational membership and makes them feel respected and honored in SOEs, it leads them to invest their social identities in their organizations and engage in extra-role behaviors to express that identity. Thus, employees in state-owned organizations will display a stronger positive reaction to procedural justice via extra-role behaviors than those in organizations with lower levels of state-ownership.
Hypothesis 1: State-ownership will moderate the positive relationship between procedural justice and employee extra-role behaviors, such that the relationship will be stronger for employees of organizations with higher levels of state ownership.
In privatized companies, where employees are more concerned about their economic outcomes, I propose that distributive justice will be a more important driver of individual extra-role behaviors than in SOEs. According to social exchange theory, both justice and extra-role behavior are social resources used in the exchange between organizations and employees, and their exchange follows the rule of reciprocity (Colquitt et al., Reference Colquitt, Scott, Rodell, Long, Zapata, Conlon and Wesson2013). This is consistent with the market pricing model, which matches input to output proportionally (Fiske, Reference Fiske1992). As employees of privatized organizations view their relationship with organizations as exchanges, they will match their extra-role behaviors to the outcomes received from their organizations. When these employees perceive the outcomes they receive from organizations to be fair, they will reciprocate by engaging in extra-role behavior. In addition, distributive justice may lead employees to increase extra-role behavior to exchange for outcomes they value. Although extra-role behaviors are not prescribed in role-definitions, employees deem these behaviors instrumental to increase their performance evaluation and promotion opportunities (Hui, Lam, & Law, Reference Hui, Lam and Law2000; MacKenzie, Podsakoff, & Fetter, Reference MacKenzie, Podsakoff and Fetter1991). When people perceive their companies to be fair and believe these behaviors are reciprocated by their organizations, they are more likely to engage in extra-role behaviors (Podsakoff, MacKenzie, Paine, & Bachrach, Reference Podsakoff, MacKenzie, Paine and Bachrach2000). Taken together, employees of privatized organizations will pay more attention to distributive justice to make a decision about whether to engage in extra-role behaviors than SOE employees. Therefore, I hypothesize that:
Hypothesis 2: State-ownership will moderate the positive relationship between distributive justice and employee extra-role behaviors, such that the relationship will be stronger for employees of organizations with lower levels of state ownership.
In summary, I suggest that the reason of why employees engage in extra-role behaviors varies with the ownership structure of organizations. SOEs activate the communal sharing model and relational concern among employees, who pay attention to procedural justice to determine whether to engage in extra-role behaviors. In contrast, privatized organizations reinforce the market pricing model and instrumental concern among employees, who resort to distributive justice to determine whether to engage in extra-role behavior. To examine these hypotheses, I first conducted a qualitative study to examine the relational models dominant in different kinds of organizations. After that, I conducted a field survey with employees from organizations with different degrees of state ownership to directly test my hypotheses.
QUALITATIVE STUDY
Sample and Procedure
Because Fiske (Reference Fiske1992)’s relational models are targeted to interpersonal relationships, existing measures of relational modles – such as Haslam and Fiske (Reference Haslam and Fiske1999) – are not suitable to characterize employee-organization relationships in this study. Therefore, I conducted a qualitative study to investigate the different relational models under different ownership structures. This is consistent with the qualitative methodology employed in research of relational models (Fiske, Reference Fiske1991). In order to understand employees’ relational models in organizations with different ownership structures, I selected four firms under the same group company, including one state-owned firm (Case 1), two public firms (Case 2 and 3), and one joint venture (Case 4). I interviewed six human resource managers from these four firms.
The interviews were conducted at the managers’ offices. The interviews were designed in a semi-structured approach. First, I asked about the history of the firm and the career history and responsibilities of the manager. Second, I asked about ownership structure, organizational goals, and requested a chart of organizational structure. Third, I asked how they carry out the functions of human resource management, including recruiting, training, performance evaluation, compensation, and career management. Finally, I asked how employees construe their relationships with their company. I took notes of all interviews and recorded and transcribed the interviews for which approval was granted. The duration of interviews varied between one to two hours.
Results
The management practices of the four cases are summarized in Table 1. Because Case 2 and Case 3 were both public firms, and their practices were very similar, I combined them into one category. As shown in Table 1, the cases represented organizations with various degrees of state ownership. In terms of organizational goals, employee satisfaction was one of organizational goals in SOEs, but not in public firms or joint ventures, indicating the special employee status in SOEs. According to company law in China, limited companies with two or more state-owned investors need to have a union representative on the board to represent the interest of workers, as reflected in Cases 2 and 3. These characteristics indicate that in SOEs and public firms, employee satisfaction is an ends rather than a means to achieve organizational profit. This is different from the schema in highly privatized organizations, such as the joint venture. Their organizational goals focus on personnel development, which treats employees as human resources for achieving organizational goals. Therefore, different organizational goals regarding employees reflect different relational models.
Table 1. Organizational information of state-owned enterprises, public firms, and joint ventures. Summary of case organizational information
In SOEs, employees rely on their organizations to satisfy their needs, reflecting the communal sharing schema (Fiske, Reference Fiske1992). An SOE manager commented about SOE employees’ loyalty (Case 1):
The workers of SOEs are very loyal. They think that they belong to the firm even after death, and the firm will send them a wreath (hua quan). Our employees have a very high happiness index, and this index is even higher after they retire. If employees are hospitalized, the firm will visit them. I buy employees casualty insurance. If they get cancer, I give them 100,000 Yuan. I visit them during holidays. If an employee dies, I cover a series of services. Overall, employee loyalty is highest in SOEs, less in public firms, and even less in joint ventures.
Another SOE manager commented about the organizational culture of the SOE and how it compares with the market pricing schema in joint ventures (Case 1):
With the development of the times, employees of SOEs are not as committed as in the past. If the firm is performing well, people have hope. SOEs value people and give employees many opportunities, such as rotation. Although the foreign companies pay well, people have to work very hard. My friends working there admire my job.
SOEs and public firms hold annual worker congress meetings, which institutionalize employees’ participation in organizational decision making. During these meetings, top managers debrief employee representatives, who evaluate managers’ performance, express the concerns of employees, and vote on important organizational policies. A public firm manager described the procedures for employee participation (Case 2):
I have a series of democratic management procedures, including employee representative meetings, evaluation of top management teams, publicity of party and administrative policies and affairs. If employees have some big problems, they may even go to the top managers.
The public firms and the joint venture inherited the incentive-suggestion system (helihua jianyi) from former SOEs, in which employees provide suggestions for improving the work process and receive recognition or bonuses (in case 4) for their valuable suggestions. Yet in the joint venture, employees’ suggestions become a resource of exchange, because the division of labor is clear – managers are responsible for making decisions, and employees are to execute decisions. Their communication is mostly carried out in the informal way, in which employees directly express their concerns to their supervisors.
The incentive systems are different between the three kinds of organizations I studied, reflecting different underlying relational models. In the SOE, factory managers are evaluated via financial indicators, but there is no formal performance evaluation or performance-based bonus for employees. The equality in outcomes is aligned with the solidarity goal and communal sharing schema in SOEs (Chen et al., Reference Chen, Meindl and Hui1998). The joint venture evaluates and rewards individual performance consistent with the market pricing model. Their incentive system emphasizes not only internal equity but also external competitiveness, i.e., it ensures that employees are compensated for the value they create for their organizations. A joint venture manager, who was dispatched to the joint venture from an SOE in 2009, described the incentive system of joint ventures this way (Case 4):
Our system is objective, scientific, and based on data and evidence. It's not like SOE, where bonuses often come as surprises. The incentive system gave priority to performance in evaluation, reward, and promotion. The organization and managers and employees at each level set goals at the beginning of every year. With reference to the goals, the evaluation of employees is carried out every year, combining self-evaluation and supervisor evaluation. The compensation is based on the result of evaluation, adjusting up to 20% above or below the salary band for each level. The salary levels are adjusted annually for market competitiveness by comparing with market salary data.
The public firms were experimenting with different evaluation and incentive systems, such as Key-Performance Indicators (KPI) and balanced score card and experiencing some tension during the transition. A public firm manager described the challenges that she encountered in enforcing the new incentive systems (Case 3). Because the firm used to be SOE, employees still held the communal sharing model and react negatively to the management practices that contradict this model.
I have established all the institutions relying on economic measures. However, when the (firm) performance is not good, I can't enforce these institutions, because I don't have so much money. SOE employees are not as qualified (su zhi) as joint venture employees, and their attitudes are not good. When you evaluate their performance, they think that you are going to deduct their wages. Therefore, I can't do performance evaluation right now. The rules can't rule the mass (fa bu ze zhong). I will hold KPI trainings this year.
The interviews and case studies corroborated my argument that highly state-owned organizations activate a communal sharing model whereas highly privatized organizations reinforce a market pricing model. The evidence provided support for my theoretical argument that different relational models underlie different ownership structures. The following quantitative study further tests my hypotheses regarding how ownership structure moderates employees’ reactions to different justice practices.
QUANTITATIVE STUDY
Utilizing the ownership diversity of organizations in China, this study examines the moderating effect of state ownership on individual behaviors in a variety of firms located in Shanghai, China. The quantitative study was conducted with Shanghai State-owned Assets Supervision and Administration Commission (SASAC) from 2009 to 2010. In 2008, Shanghai SASAC supervised 40 group companies, which were composed of 3923 enterprises; the total revenue of these enterprises was equivalent to 101.49 billion US dollars, and their total profit was 1.66 billion dollars. Among the organizations under the supervision of SASAC, I selected three types of organizations with different degrees of state ownership: (1) entirely state owned, which have the most state ownership, (2) publicly-traded, which typically have a middle-level of state ownership, and (3) international joint ventures, which typically have the lowest level of state ownership. This unique design has three advantages. First, examining and comparing organizations under the same city's governance reduces the potential influence from other contextual factors because the same state office standardizes the governance of local enterprises under its supervision (Naughton, Reference Naughton2005). Second, because all these firms used to be SOEs before the reform, the cross-sectional sample can provide a clue about the transition effect. Third, because the state directs and controls the transition process, the mobility and transfer of personnel between organizations is kept at a minimum. Thus, the observed relationships are more likely to be driven by change in ownership structure than personal selection.
Sample and Procedure
Among the 40 group companies supervised by Shanghai SASAC, four agreed to participate in my survey study. These companies covered a wide range of industries, including food, commercial, chemical, and automobile. I varied ownership structure within each industry and selected 12 firms, including four state-owned firms, three public firms, and five joint ventures. I requested 50 to 100 respondents from each firm, summing up to 800 respondents from the 12 firms altogether. In order to create a random sample of each firm, I requested that managers of each firm select respondents from employee rosters randomly based on the sample size that they agreed on. Finally, 721 participants returned the questionnaires, resulting in a response rate of 90%. The distribution of the sample among the three types of organizations was 282 from SOEs, 230 from public firms, and 209 from joint ventures. The hierarchical composition of the sample was 42% general employees, 28% supervisors, 21% middle managers, and 9% top managers. The respondents had an average age of 39 years and an average tenure of 13 years. In addition, 39% were females, 56% were Chinese Communist Party members, and 78.2% of the respondents had college or higher education.
The questionnaire was translated from English to Chinese by the author and back translated to English by a research assistant, following the procedure suggested by Brislin (Reference Brislin1980). In addition, the wording of the questions was discussed with a local manager to ensure that employees can understand it. Questionnaires enclosed in envelopes were distributed to employees at their workplaces, and they were informed that the survey was only for research purposes and assured of the anonymity of their responses. They were required to seal the envelopes after completing the survey and drop the envelopes at a central location. After the survey, the companies returned the envelopes to the author.
Measures
Independent variables
I obtained the annual statistics of all the firms from the government office and calculated the degree of state ownership by the proportion of state-owned equity in the total equity of each organization, following the example of previous research (Le & Buck, Reference Le and Buck2009; Le & O'Brien, Reference Le and O'Brien2010). The information of state-owned equity was readily available in the government report (SASAC, 2008), and the total equity was calculated by subtracting total liabilities from total assets. The distribution of state ownership in the sample is presented in Figure 1.
Figure 1. Histogram of state ownership in 12 organizations.
Following the example of previous literature (Colquitt, Reference Colquitt2001), distributive justice was measured by the extent to which the outcomes are allocated based on individual contribution. I selected a scale widely used in previous studies (Blader & Tyler, Reference Blader and Tyler2009; Tyler & Blader, Reference Tyler and Blader2000, Reference Tyler and Blader2003). This scale contained five items, including ‘The resources I receive are linked to how well I do my job’ and ‘In general, resources are fairly allocated among employees at my organization’ (α = 0.91). Responses were made with a five-point scale (1-highly disagree, 5-highly agree).
One of the most important manifestations of procedural justice is organizational members’ participation in organizational decision making (Bies & Shapiro, Reference Bies and Shapiro1988), and the function of participation in conveying the relational value of procedural justice has been found to be cross-cultural (Lind, Tyler, & Huo, Reference Lind, Tyler and Huo1997). Therefore, I measured procedural justice with a scale asking about employees’ participation in organizational decision making (Hage & Aiken, Reference Hage and Aiken1969). The scale has four questions, such as ‘How frequently can you participate in the decision on the adoption of new programs?’ and ‘How frequently can you participate in decisions on the adoption of new policies?’ (1-never to 5-always, α = 0.90).
Dependent variable
Since my objective is to examine extra-role behavior that is important for all types of organizations, I adopted the widely-used measure of extra-role behavior that represents individuals’ engagement with organizations (Blader & Tyler, Reference Blader and Tyler2009; Tyler & Blader, Reference Tyler and Blader2000, Reference Tyler and Blader2003). With a five-point scale ranging from 1-‘never’ to 5-‘always’, respondents rated the frequency of how often they engage in six behaviors, such as volunteering to do things that are not required in order to help the organization; putting extra effort into doing their jobs well, beyond what is normally expected; working extra hours even when they would not receive credit for doing so; and helping others with work related problems (α = 0.91).
Control variables
I controlled for demographic variables including gender, education, tenure (how many years they had worked in the organization), and position in the hierarchy, which have been found to be related to extra-role behaviors in previous research (Morrison, Reference Morrison1994; Podsakoff et al., Reference Podsakoff, MacKenzie, Paine and Bachrach2000). Gender was a dummy variable (1= Female, 0 = male), whereas the other variables were treated as continous variables. Previous research suggests that another dimention of organizational justice – the treatment individuals receive from their leaders, i.e., interactional justice – also influences their extra role behavior (Colquitt, Reference Colquitt2001). Therefore, this study controlled for interactional justice, which was measured with a four-item scale used in previous research (Blader & Tyler, Reference Blader and Tyler2003a, Reference Blader and Tyler2003b). Respondents indicated to what extent their supervisors consider their views, respect their rights, care about employees’ well-being, and give them an explanation for the decisions made when there is a disagreement (1-highly disagree, 5-highly agree; α = 0.92). In order to rule out the alternative argument that the effect of state ownership is due to individual differences on value of groups and relationships, I also included psychological collectivism and interdependent self-construal, which have been found to either enhance extra-role behaviors (Moorman & Blakely, Reference Moorman and Blakely1995) or moderate the effect of procedural justice (Brockner et al., Reference Brockner, De Cremer, van den Bos and Chen2005). Psychological collectivism is individuals’ value of groups and was measured by selecting five items with the highest loadings on each dimension of the psychological collectivism scale (Jackson, Colquitt, Wesson, & Zapata-Phelan, Reference Jackson, Colquitt, Wesson and Zapata-Phelan2006). The selected items were: ‘I preferred to work in groups rather than working alone’; ‘I felt comfortable trusting group members to handle their tasks’; ‘I followed the norms of groups’; ‘I was concerned about the needs of groups’; and ‘Group goals were more important to me than my personal goals’ (1-highly disagree, 5-highly agree; α = 0.85). I measured interdependent self-construal with four items with the highest loadings in the relational-interdependent self-construal scale (Cross, Bacon, & Morris, Reference Cross, Bacon and Morris2000): ‘my close relationships are an important reflection of who I am’; ‘when I feel very close to someone, it often feels to me like that person is an important part of who I am’; ‘I think one of the most important parts of who I am can be captured by looking at my close friends and understanding who they are’; and ‘when I think of myself, I often think of my close friends or family also’ (1-highly disagree, 5-highly agree; α = 0.71).
Results
Firstly, I tested whether missing data was a serious problem in the sample. Missing value analysis showed that the majority of missing values occurred in the demographic variables (the number of observations for other variables ranged from 716 to 721). I divided the sample into two subsamples: one without any missing values of gender, tenure, education, and position (N = 464), and the other with at least one missing value (N = 255). The two subsamples did not have a significant difference on extra-role behaviors (t (458) = 1.63, p = 0.10). Therefore, missing data did not cause a serious concern. After that, I also examined whether SOEs and privatized organizations differ in age or tenure because of the different histories of these organizations. ANOVA results showed that SOEs, public firms, and joint ventures did not have a significant difference on employee age (F (2, 646) = 1.00, p = 0.37) or tenure (F (2, 596) = 2.65, p = 0.07).
Secondly, I conducted confirmatory factor analysis to examine whether the measurement of variables was valid. The confirmatory factor analysis showed that the six-factor model, using all the items of the measured variables without parceling, fit very well with the data (χ 2 (390) = 1579.68, p < 0.001, CFI = 0.92, TLI = 0.91, RMSEA = 0.06). In order to examine whether a substantial common method variance was present, I conducted the one-factor test as suggested in previous studies (Podsakoff, MacKenzie, Jeong-Yeon, & Podsakoff, Reference Podsakoff, MacKenzie, Jeong-Yeon and Podsakoff2003). The model that loaded all the items on a common method factor did not fit the data well (χ 2 (405) = 7234.70, p < 0.001, CFI = 0.52, TLI = 0.48, RMSEA = 0.15), which suggested that the common method cannot account for the relationship between variables. I also compared the measurement model with several alternative models. For instance, the six-factor model fit better than a five-factor model that combined distributive justice and procedural justice (Δχ2 (5) = 1207.72, p < 0.001) and a four-factor model that combined distributive justice, procedural justice, and interactional justice (Δχ2 (9) = 2172.27, p < 0.001). These comparisons suggest that distributive justice, procedural justice, and interactional justice measured in this study represent distinctive aspects of organizational justice. I also compared the measurement model with a five-factor model that combined psychological collectivism and interdependent self-construal, and the latter fit the data significantly worse (Δχ 2 (5) = 219.49, p < 0.001), indicating that collectivism and interdependent self-construal represent distinctive aspects of cultural values. The means, standard deviations, and correlations among all the variables are presented in Table 2.
Table 2. Correlations and descriptive statisticsa
Notes: a Entries on the diagonal are Cronbach's alphas. Significance levels are based on two-tailed tests. For position, 1 = Employee, 2= Supervisor, 3 = Middle manager, 4 = Top manager. For education, 1 = Middle school, 2 = High school, 3 = College, 4 = Bachelor, 5 = Master or higher.
* p < 0.05; ** p < 0.01
Thirdly, I used the software HLM7 to test my hypotheses. Considering the nested nature of my data, I constructed a three-level hierarchical linear model to control for industry and firm effects. The results are presented in Table 3. In the intercept-only model, both firm effects and industry effects were significant, which indicated the necessity of controlling for these effects. In Model 1, I entered the demographic variables into the model. Both position and tenure had significant and positive effects on extra-role behaviors. To test my hypotheses that state ownership accentuates the effect of procedural justice (H1) and attenuates the effect of distributive justice (H2) on extra role behavior, I centered procedural justice and distributive justice around their group means and centered state-ownership around its grand mean, following the suggestion of previous research (Aguinis, Gottfredson, & Culpepper, Reference Aguinis, Gottfredson and Culpepper2013).[Footnote 2] In Model 2, I entered state ownership, distributive justice, and procedural justice into the model. Consistent with previous research, both distributive justice and procedural justice had significant and positive effects on extra-role behaviors. To test the moderation effect of state ownership, I entered these interaction terms in Model 3. The interaction effects significantly improved model fit (χ 2 (2) = 11.74, p = 0.003) and explained 3% of individual-level variance and 5% of industry-level variance according to the procedure suggested by Hox (Reference Hox2010). According to Cohen's standard, the effect sizes were above the low level (0.02) and under the medium level (0.15). Besides the variances explained which indicate explanatory power, Aguinis et al. (Reference Aguinis, Gottfredson and Culpepper2013) also recommended reporting predictive power as indicated by the coefficients of moderation effect. The interaction effect between state ownership and procedural justice (γ = 0.30) was significant and positive, which supported H1. This effect has achieved the medium standard of effect size (Cohen, Reference Cohen1988). I did a simple slope analysis at one standard deviation above, at, and below the mean level of state ownership (Preacher, Curran, & Bauer, Reference Preacher, Curran and Bauer2006) and plotted the simple slopes in Figure 2. Simple slope analysis showed that the effect of procedural justice on extra-role behaviors was positive at high (simple slope = 0.21 s.e. = 0.05, t = 4.19, p < 0.001) and medium levels (simple slope = 0.14, s.e. = 0.04, t = 3.94, p = 0.001) of state ownership, but the effect became non-significant at low level of state ownership (simple slope = 0.07, s.e. = 0.05, t = 1.58, p = 0.115).
Table 3. Hierarchical linear models of extra-role behavior
Notes: ISC=Interdependent Self-Construal. DJ= Distributive justice. PJ= Procedural justice (participation). For position, 1 = Employee, 2= Supervisor, 3 = Middle manager, 4 = Top manager. For education, 1 = Middle school, 2 = High school, 3 = College, 4 = Bachelor, 5 = Master or higher.
Figure 2. The interaction effect of procedural justice and state ownership on extra-role behavior.
In Model 3 of Table 3, the interaction effect between state ownership and distributive justice was significant and negative; supporting H2 that state ownership attenuates the positive effect of distributive justice on extra-role behaviors. The effect size of the coefficient (γ = -0.69) was large according to Cohen's standard (above 0.50) (Cohen, Reference Cohen1988). The results of simple slope analysis are presented in Figure 3. Distributive justice had a positive effect on extra-role behaviors among organizations with medium (simple slope = 0.13, s.e. = 0.05, t = 2.83, p = 0.005) or low levels (simple slope = 0.29, s.e. = 0.06, t = 4.54, p < 0.001) of state ownership. At high level of state ownership, distributive justice did not have a significant effect on extra-role behaviors (simple slope = -0.03, s.e. = 0.07, t = -0.43, p = 0.66).
Figure 3. The interaction effect of distributive justice and state ownership on extra-role behavior.
Finally, I conducted robustness checks to test whether the moderation effects of state ownership are driven by individual differences on collectivism and interdependent self-construal. To achieve that objective, I first tested whether state ownership was associated with cultural values. I regressed cultural values on state ownership in hierarchical linear models. After controlling for demographic variables, state ownership did not have a significant effect on psychological collectivism (b = -0.05, s.e. = 0.16, p = 0.76) or interdependent self-construal (b = 0.00, s.e. = 0.11, p = 0.99). Therefore, state ownership was not systematically related with individual difference in cultural values. Next, I entered cultural values and their interaction effects with distributive justice and procedural justice in the model. In Model 4 of Table 3, both psychological collectivism and interdependent self-construal had a significant and positive effect on extra role behavior. Controlling for these effects and their moderating effects, the hypothesized interaction effects remained significant. It indicates that the moderation effects of state ownership were not due to individual differences on collectivism, interdependent self-construal, or their interaction effects with justice. I also tested whether the hypothesized interaction effects were robust after controlling for the effect of interactional justice. Interactional justice had a positive effect on extra-role behavior, and this effect was not moderated by state ownership (b = -0.35, s.e. = 0.23, p = 0.12). After controlling for the main effect of interactional justice, the hypothesized moderation effects remained significant.[Footnote 3] Therefore, the results were robust to individual difference in cultural values and interactional justice individuals receive.
DISCUSSION
Drawing on the institutional diversity in the reform context of China, this study demonstrates that ownership structure moderates the effects of justice on extra-role behavior. Specifically, distributive justice is positively associated with extra role behavior in privatized organizations but has no effect (or even negative effect after controlling for individual values and interactional justice) in highly state-owned organizations. In contrast, participation in decision making – a key component of procedural justice – is positively associated with extra-role behavior in organizations with high and medium levels of state ownership, but its effect became non-significant in highly privatized companies. The moderation effect of state ownership is not reducible to individual difference in cultural values. Therefore, the findings support my proposal that employees pay attention to the form of justice that is consistent with their relational schema to guide their extra role behavior.
The findings also indicate that the effect of state ownership on employee behavior depends on perceived justice. Specifically, employees of privatized organizations demonstrate less extra-role behavior than SOE employees when distributive justice is low, whereas SOE employees engage in less extra-role behavior than employees of privatized organizations when they perceive low procedural justice or high distributive justice. In other words, people react negatively when their dominant relational schema is not verified, but people in different organizations react differently to practices that are inconsistent with their schemas, depending on whether such practice is irrelevant or contradictory to their schemas. Since participation is irrelevant to privatized companies’ market pricing schema, it is ignored by their employees, and their extra-role behavior is not reduced. In contrast, distributive justice is not only inconsistent with SOEs’ communal sharing schema, it contradicts that schema by undermining the equal status of employees. As a result, employees of SOEs engage in less extra-role behavior than privatized companies when distributive justice is high.
Theoretical Contributions
This study makes important theoretical contributions to multiple literatures. First, this study contributes to research on the micro-foundations of corporate governance. Research on corporate governance has called for more attention to the role of labor and research on the transition economies (Aguilera & Jackson, Reference Aguilera and Jackson2010; Fiss, Reference Fiss, Greenwood, Oliver, Suddaby and Sahlin2008). Many countries’ SOEs have employee representatives on board (World Bank, 2014), and even more organizations have various forms of worker participation, such as employee stock ownership plans (Doucouliagos, Reference Doucouliagos1995). This study suggests that these organizations should pay attention to the unique relational schema triggered by these institutions. Increasing labor representation in corporate governance can change the nature of how employees relate to their organizations. Employees will regard themselves as owners of organizations and attend to participation opportunities to verify their owner identity. Meanwhile, they will be less sensitive to the outcomes they receive. Therefore, even minor and symbolic change in corporate governance can have far-reaching implications for employee motivation.
In addition, I offer a theory of the micro foundations of institutional transition. I find that the relationship between perceived justice and individual behavior depends upon ownership structure and, likewise, the relationship between ownership structure and individual behavior depends upon justice perceptions. Thus, the favored variables of micro and macro scholars do not work independently – they combine interactively to influence individual behaviors. I found that ownership structure is not associated with different levels of employee motivation (Burawoy & Lukacs, Reference Burawoy and Lukacs1985), but rather associated with different predictors of employee motivation. This finding helps explain the mixed results on the transition effect. Since the transition effect depends on how people perceive organizational practices, the method of transition may greatly influence its effectiveness. For instance, previous research found that non-state controlled firms are more likely to enhance post-transformation performance than state-controlled companies (Li, Xia, Long, & Tan, Reference Li, Xia, Long and Tan2012). One explanation is that removing state control can change the relational schema of employees and facilitate their adaptation to new organizational practices after transition.
Furthermore, this study discovers a boundary condition of the well-established effects of organizational justice. Previous research has proposed at least two reasons for why justice is important – instrumental reason represented by social exchange theory and relational reason represented by social identity theory. The current study suggests that justice is important for different reasons under different organizational contexts, and distributive justice and procedural justice are differentially positioned to serve people's instrumental or relational needs. Specifically, when organizational contexts foster a communal sharing schema, the function of procedural justice in satisfying individuals’ relational needs and verifying their social identity is especially important. In these contexts, the well-established positive effect of distributive justice becomes non-significant or even reversed. In contrast, when organizational contexts emphasize productivity and efficiency, distributive justice plays an important role in fulfilling people's instrumental motive. Therefore, this study highlights the importance of organizational context and its underlying relational model as factors that shape why people care about justice and which aspect of justice people care about.
Finally, this study contributes to cross-cultural research by highlighting the importance of organizational context and relational model in activating cultural knowledge and guiding individual behaviors. Cross-cultural research has investigated how national differences in cultural values influence individuals’ reactions to justice (Kirkman, Chen, Farh, Chen, & Lowe, Reference Kirkman, Chen, Farh, Chen and Lowe2009). However, studies relying on different cultural values generate conflicting predictions regarding how people within the same culture react to justice. For instance, since Chinese are high on both materialism and interdependent self-construal, both distributive justice and procedural justice should be important for them (Brockner et al., Reference Brockner, Chen, Mannix, Leung and Skarlicki2000; Brockner et al., Reference Brockner, De Cremer, van den Bos and Chen2005; Kim & Leung, Reference Kim and Leung2007). At the same time, because Chinese people are high in power distance and traditional values, justice – especially procedural justice – should be less important for them (Brockner et al., Reference Brockner, Ackerman, Greenberg, Gelfand, Francesco, Chen, Leung, Bierbrauer, Gomez, Kirkman and Shapiro2001; Farh, Earley, & Lin, Reference Farh, Earley and Lin1997; Kim & Leung, Reference Kim and Leung2007). These contradictions are not specific to China. Given that power distance and collectivism are highly correlated with each other (Hofstede, Reference Hofstede2001), it would be challenging to predict whether justice is more or less important for a particular culture. Fiske (Reference Fiske1992) argued that all cultures share the four relational schemas in social relationships. Indeed, the correlations between collectivism values and relational models are generally small (Realo, Kästik, & Allik, Reference Realo, Kästik and Allik2004; Vodosek, Reference Vodosek2009). It is the specific context within a culture that determines which relational model is salient. As found in this study, organizational context provides a better prediction about organizational members’ reaction to justice than cultural values.
Implications for Management Practices
First, this study generates important implications for applying justice practices. This study suggests that distributive justice is especially suitable for satisfying instrumental need, whereas procedural justice is especially important for satisfying relational need. Thus, organizations may emphasize different aspects of justice under different organizational agenda. If organizations aim to achieve high productivity, they should emphasize distributive justice and use practices such as performance evaluations and pay for performance. If organizations want to solicit employee identification, they should emphasize procedural justice and establish institutions for employees to participate in organizational decision making.
In addition, this study generates new insights for transition economies such as China by emphasizing the perspective of employees in reform design. The institutions of state ownership cultivated a communal sharing schema among SOE employees. As a consequence, they do not react positively to distributive practices such as pay for performance because it contradicts their identity as enterprise masters. Therefore, managers and policymakers should recognize this relational schema while changing the incentive system of SOEs. Just as the manager in Case 3 did, managers should transform employees’ relational schemas through training and communication before implementing practice change. At the same time, this sense of ownership can be regarded as a unique legacy of SOEs and be leveraged to their transformation. Reform methods such as profit sharing and employee stock ownership can protect employees’ collective identity and reinforce its motivating force. In addition, the reform process should also preserve SOE employees’ social identity by inviting them to participate in the design and implementation of the reform. If the reform simply privatizes SOEs and discharges workers without recognizing their relational schema, the reform will encounter unforeseen resistance and lose the motivational legacy of SOEs.
Limitations and Future Research
Despite the important contributions, this study has some limitations. First, although the degree of state ownership was measured with archival data, the measurement of justice and extra-role behaviors may be subject to common-method bias. However, ‘in the absence of true effects, it is extremely unlikely for common-method variance to generate significant cross-level interactions. In fact, if a true cross-level interaction exists, common-method variance tends to lower the likelihood of its identification’ (Lai, Li, & Leung, Reference Lai, Li and Leung2013: 243). Therefore, the findings of cross-level interactions argue against a common-method variance explanation. Furthermore, common-method variance cannot explain the existence of non-significant simple slopes. If common method bias was driving the results, it would have caused all the relationships between justice and extra-role behaviors to be significant. Nonetheless, future research should measure behaviors with another source and corroborate the results.
Second, although I have conducted qualitative studies to investigate the relational schemas under different ownership structures, I did not measure relational models in the quantitative study. The primary reason for this limitation is because the existing scales are targeted to interpersonal relationships and not suitable for the current study. This limitation makes it difficult to rule out alternative explanations. For instance, one might argue that it's the authority ranking relational schema in SOEs that constrains participation opportunities in SOEs, which further makes participation especially scarce and important for SOE employees. This alternative explanation is not consistent with previous research and theory. Previous research found non-significant difference in hierarchical organizational culture between SOEs and privatized companies, such as foreign-controlled enterprises (Ralston, Terpstra-Tong, Terpstra, Wang, & Egri, Reference Ralston, Terpstra-Tong, Terpstra, Wang and Egri2006). Relational models theory (Fiske, Reference Fiske1992) predicts that people attend to information consistent with their relational schema and behave accordingly. Under the authority ranking schema, employees will not expect participation opportunities but instead simply comply with their authorities. Thus, they will not pay attention to participation opportunities to guide their extra-role behavior. This alternative explanation does not fit well with the data either. My qualitative studies indeed found the widespread participation opportunities enjoyed by SOE employees, and my quantitative study didn't find a negative correlation between state ownership and participation. Therefore, it is not very likely that this explanation can explain the findings. Future research should develop scales of relational schemas for employee-organization relationships and directly test the moderating effects of relational schema.
Another limitation is that the cross-sectional study could not make causal argument about the observed relationships. However, the unique design of this study reduces the likelihood of reverse causality. In the sample design, all the organizations used to be SOEs, and the majority of the participants used to be SOE employees. Because the reform in Shanghai was carried forward with close state control, the labor transfer between organizations during transition was controlled by the state and kept at a minimum level. Thus, the different relationships between justice and extra-role behaviors observed in different organizations are more attributable to ownership change than individual self-selection. Future research can employ longitudinal study to track individual behaviors during the reform and corroborate my findings.
CONCLUSION
Drawing on the institutional diversity during ownership reform in China, this study shows that the ownership structure of organizations influences why people care about justice and which aspect of justice people value. In state-owned organizations, the socialist legacy activates a communal sharing schema among employees and renders participation an important factor for verifying their social identity. As the privatization reform emphasizes productivity and efficiency, employees of privatized organizations construe their relationships with their organizations as social exchanges and pay close attention to equitable allocation of outcomes. Therefore, policymakers and managers should acknowledge the different schemas activated by different ownership structures and design reform methods and management practices creatively.
