Inﬂuence of Guanxi on Job Satisfaction in China: Xinren as a Mediator

. Research has shown that Chinese guanxi (private relationships / networks based on Chinese culture, which allow to obtain social resources informally and privately) are related to job searching and work-related outcomes such as job match and job satisfaction. However, few studies have examined the correlation between the sub-dimensions of guanxi and job satisfaction. With a sample of 267 respondents from di ﬀ erent Chinese companies, this study divided guanxi research into two stages (the renqing & ganqing guanxi used for job-searching processes and the supervisor-subordinate guanxi used for on-the-job processes) and examined the mediating e ﬀ ect of xinren on guanxi and job satisfaction by using structural equation modeling and bootstrapping analysis. The results showed that guanxi was positively related to job satisfaction, and xinren partly mediated these correlations. Our ﬁndings extend previous research on the e ﬀ ects of guanxi on job outcomes and may be used as an exploratory guide to manage employee satisfaction.


Introduction
With the rapid development of economy and digital technology, employment patterns and working modes in China have changed accordingly. Using a social network is an effective job search approach in Western countries, and employers and workers can get accurate information and the best match through referral hires [1]. In China, a social network is also used for job searches; this is called guanxi, which literally means "relation" or "relationship" [2]. Different authors from various academic fields have given different definitions of guanxi. Individuals exist in a guanxi network, which connects relatives, friends, and all other social relations [3,4]. "Guanxi" refers to private relationships or networks based on Chinese culture, which allow to obtain social resources informally and privately [5]. In practice, trustworthiness is also important because the exchange between the jobseeker and the employer can be informal or even illegal, and in workplaces, the supervisor and subordinate (S-s) are also linked together and influence one another via both work-related and non-work-related ties [6].
Although the value of guanxi in understanding how Chinese society operates has been widely recognized, how it is related to employees' job satisfaction remains unclear. Recent studies on the topic of using guanxi in the job-searching process have focused on the effects

Background Theory
This study uses social capital theory to illustrate the influence of guanxi on job satisfaction.
Social Capital Theory. The main idea of social capital theory is that people have access to tangible and intangible resources at the individual, group, and organizational levels through social interaction and contact with others. An important point of this theory is that social capital resources are embedded in the social networks of interconnected individuals, groups, or nations and can be obtained through social networks [10]. The concept of social capital first appeared in a study on the importance of building strong families and local communities. Nahapiet and Ghoshal [11] defined social capital as the sum of actual and potential resources from a relationship network (individual or social unit is an integral part of it) that an individual or social unit can obtain. Since then, the concept of social capital has been widely studied in many fields. Studies have shown that having social capital is beneficial and profitable to individuals, groups, organizations, and nations. Fulkerson and Thompson [12] conducted a meta-analysis of the theory and defined the six dimensions of social capital: (1) community values; (2) collective action, social structure, and benefit realization; (3) trust, reciprocity, and cooperation; (4) individual and group relationship resources; (5) citizen participation and spontaneous assistance; and (6) society relationships and networks. Based on these six dimensions, Fulkerson and Thompson [12] created two primary and relative categories of social capital: resource social capital and normative social capital. Resource social capital mainly refers to resources, networks, and social relationships that are shared with each other. Normative social capital includes norms, trust, reciprocity, civic engagement, and the values of friends, families, and communities. As mentioned above, this paper studies the impact of renqing & ganqing guanxi and supervisor-relationship on job satisfaction. The study of social capital at the micro level, as is embedded in this paper, focuses on three aspects: 1) the relationship characteristics of individual actors and the influence of their own social status on social capital; 2) the overall structural characteristics of the social network in which the actors are concerned; 3) the influence of interaction between networks and constraints on the ability to acquire individual social resources.

Job-Searching Process: Guanxi (Renqing & Ganqing), Xinren (General Trust), and Job Satisfaction
In Chinese, the word "guanxi" has many different meanings. Because of its special cultural connotation, it cannot be translated into "relationship" in English academic circles, and the translation is directly replaced by the Chinese pinyin "guanxi". Based on the GRX scale of guanxi, ganqing reflects the tenor of a social relationship between two people or two organizations as well as an emotional attachment that exists among parties of a network. It describes the degree of emotional understanding, connections, and the sharing of feelings of happiness and fears alike [13], which is always accompanied by some material responsibility (material obligation). "Ganqing" denotes the affective side of guanxi, while "renqing" is another Chinese idea related to it, which emphasizes the human obligation required through actions such as the exchange of favors that nurture social exchange. Human feelings (renqing) can be used as a resource to be presented to another as a gift in the form of money, goods, or service. It is "a set of social norms by which one has to abide in order to get along well with other people in Chinese society". Different from ganqing, renqing denotes the conative part of guanxi. Some research found that guanxi may lead to better job-worker match [3], and searching for a job through guanxi may affect final job outcomes, including job satisfaction. Guanxi can channel job information that accelerates the job-searching process [1], enabling job seekers to pre-assess the quality of their match to the job openings. From the employers'side, guanxi serves as a screening device in the recruitment process; it not only enables employers to have an informational advantage over job seekers, but is also an efficient and less expensive method for identifying satisfied job candidates. Moreover, impartiality in the hiring process can create a more equitable environment in an organization, which reduces job turnover intentions among employees. Thus, this may imply that employees who use guanxi to find a job may have higher job satisfaction. Based on the preceding literature reviews, it is hypothesized that Hypothesis 1. Guanxi (renqing & ganqing) is positively associated with employees' job satisfaction.
Xinren (general trust) refers to trust at the interpersonal level, which represents the interpersonal trust of one's integrity and kindness to stick to and deliver one's promises. Xinren is therefore essential to guanxi as a cognitive prerequisite to any engagement of renqing exchange, and it is established based on evaluations of the other party's trustworthiness as a combination of both. Yet, different from trust in the West, xinren can only be established between relational boundary personnel at the interpersonal level rather than between two organizations [14]. In the relationship exchange, people obtain not only resources but also a certain degree of trust. In weak relationships, the role of moral trust is not significant because Confucian concepts like "love has hierarchy" and the "difference pattern" weaken people's morality with the alienation of relationships. In strong relationships, people's trust is often based on Confucian morality and emotion, and it is not related to their own interests. The job satisfaction brought by job hunting and based on a strong relationship is usually higher; this may imply that employees who rely on people to find a job may have higher job satisfaction, and xinren may have some mediation effect on guanxi and job satisfaction. Thus, this paper proposes the following assumptions: Hypothesis 2. Xinren (general trust) will mediate the relationship between guanxi (renqing & ganqing) and employees' job satisfaction.
Hypothesis 2b. Xinren (general trust) is positively associated with employees' job satisfaction.
To summarize the research hypotheses developed on the job-searching process, figure 1 shows the relationship between the related variables. Supervisor-subordinate guanxi (S-s guanxi) is a key factor used to decode the structure in various organizations [15]. Drawing on social exchange theory, interactions between supervisors and subordinates involve the exchange of resources, friendliness, and trust. Previous researchers have shown that S-s guanxi is related to a wide range of subordinates'work-related outcomes, such as organizational commitment, job satisfaction, turnover intention, organizational citizenship behavior (OCB), task performance, career development, and intention to stay [16,17]. In China, employees always respect authorities; therefore, good relationships with supervisors ensure that employees make things happen at work [18]. Through effective networking with supervisors, employees become well connected and centrally located in their organizations. In general, they can benefit from access to information and to financial or material resources and gain visibility, legitimacy, or sponsorship within organizations [19]. That is, supervisors will provide information on what is happening in the organization, provide access to power structures, and give emotional support to the employees. Therefore, employees' social capital is created through those interpersonal interactions. The present study demonstrates how the effect of S-s guanxi on employees' job satisfaction is influenced by their personal and social resources, i.e., personal power and network resources [9]. Based on the preceding review, Hypothesis 3 states that Hypothesis 3. Guanxi (supervisor-subordinate) is positively associated with employees' job satisfaction.
Xinren (trust in supervisor) refers to subordinates'positive expectations regarding the actions and intentions of their supervisor. Subordinates who receive benefits from having a good relationship with supervisors may have greater trust in their supervisors, and S-s guanxi is positively correlated with trust in supervisors [20]. Regarded as social capital, network resources and personal power are embedded within and derived from employees' relationship ties, which not only provide psychological support via intimacy and trust but also help employees facilitate the exchange of valued resources. Further, employees can utilize those valued resources-including materials, job-related information, and supervisory trust-to achieve favorable outcomes. Researchers have found that organizational resources and information are contextual factors enhancing employees' motivation and performance [19]. In China, social capital is a key factor for work attitudes and outcomes; employees with a high level of social capital are more likely to feel powerful and comfortable and thus be more satisfied with their jobs [19]. In such a high relationship-oriented society, employees with more network resources find it easier to meet their job requirements and execute their duties. Accordingly, employees' job attitudes may depend on their ability to acquire organizational resources. As a key indicator of quality of work life, job satisfaction is related to employees' mental and physical health, job performance, and overall life satisfaction [21]. Based on this literature, it is supposed that xinren (trust in supervisor) would have some mediation effect on the relationship between S-s guanxi and job satisfaction. Therefore, this paper proposes the following other assumptions: Hypothesis 4. Xinren (trust in supervisor) will mediate the relationship between guanxi (supervisor-subordinate) and employees' job satisfaction.
Hypothesis 4b. Xinren (trust in supervisor) is positively associated with employees' job satisfaction.
To have an intuitive understanding of these hypotheses developed for the on-the-job process, the relationships between related variables are shown as figure 2.

Research Context
Responding to the calls made by Barnes et al. [22], we chose China as the study context based on several important considerations. Firstly, guanxi is more prevalent in China than in other countries; secondly, the sheer size and growth of the Chinese economy attracts many Western companies. Hence, to investigate how guanxi may be employed in effectively managing job satisfaction, this paper focuses on investigating the renqing & ganqing dimension of guanxi during the job-searching process as well as supervisor-subordinate guanxi during on-the-job working. We built two models to testify our research hypothesis.

Participants
Questionnaires were distributed and collected via a professional online survey website (Wenjuanxing) in August 2018. We adopted a stratified sampling strategy to select participants for this study. The majority of the sample was in Sichuan-a famous province in the southwest of China. The approach included an online survey and offline interview. On the one hand, we could use social platforms to send an online questionnaire to the participants; on the other hand, we simultaneously did face-to-face interviews to increase efficiency and help participants who could not get through online platforms. Participants were informed that the survey was anonymous, that all data were confidential and would only be used for research purposes and that they could quit at any time. After finishing the questionnaire, the participants had the chance to win a small gift as a reward through an online raffle. We would provide a brief report to those who wished to know the research results. The target sample size was 300, and after deleting questionnaires with incomplete responses, we obtained 267 valid responses with data collected from both the state sector and non-state sector. Of the respondents, 104 were working in the state sector (26% for the government, 36.5% for a public institution, and 37.5% for a state-owned business); 163 respondents were working in the non-state sector (49.1% for a private enterprise, 16% for a foreign-owned enterprise, and 34.9% for a joint venture or others). The respondents were mainly from the information technology and internet, education and training, manufacturing, finance and accounting, and real estate industries. table 1 shows more details about the participants.

Procedure
During the survey practice process, three steps were taken to improve the accuracy of the stratified sample. First, we distinguished prominent demographic and categorical characteristics related to the behavior being studied and stratified them by gender and company type. Second, we determined the overall proportion at each level: male and female respondents each accounted for 50% (150 respondents each); the state sector accounted for 40% (120 respondents); and the non-state sector accounted for 60% (180 people). Third, we drew an independent, simple random sample from each layer to complete the entire sampling survey.

Measures
We developed the measures of all constructs based on a review of the existing literature. Participants rated all items on a 5-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree).
Guanxi (Renqing & Ganqing). The scale for measuring guanxi is adopted from Yen et al. [22], and it contains six items with Cronbach's α = 0.878. Three question items for renqing guanxi reflect the exchange of help and the giving and receiving of favors and contained (1) "I feel a sense of obligation to them for doing them a favor"; (2) "The practice of 'give and take' of favor is a key part of the relationship between them and me"; (3) "I am happy to do a favor when they request one". Cronbach's α was 0.766 in this study. Other three question items for ganqing guanxi are (1) "The others and I are able to talk openly as friends"; (2) "I would consider whether their feelings would be hurt before I made an important decision"; (3) "I would try my best to help out the others when they are in need because they are a friend of mine". Cronbach's α was 0.872 in this study.
Xinren (General Trust). We employed three reversed scale items to measure xinren, which evaluates the degree to which the respondents perceive others as trustworthy [22]. Question items were (1) "They are normally only concerned about themselves" (reverse coded); (2) "They do not seem concerned with others'needs" (reverse coded); (3) "Generally, they are not trustworthy" (reverse coded). Cronbach's α was 0.605 in this study.
Guanxi (Supervisor-Subordinate). This study used the six-item scale developed by Law et al. [23] to measure S-s guanxi in a Chinese context. Question items were (1)  Xinren (Trust in Supervisor). This study used the Chinese version of 6-item scale to measure trust in/loyalty to one's leader, which was translated by Jang et al. [24] and found to have good reliability. Question items were (1) "My supervisor would never try to gain an advantage by deceiving employees' '; (2) "I have complete faith in the integrity of my supervisor"; (3) "I feel a strong loyalty to my supervisor"; (4) "I have a high degree of mutual trust with my supervisor". Cronbach's α was 0.908 in this study.
Job Satisfaction. Hackman and Oldham's [25] 3-item measure was used to assess participants'job satisfaction. Illustrative items were (1) "Generally speaking, I am satisfied with my present job"; (2) "I am generally satisfied with the results of my work"; (3) "Most of my colleagues appreciate their work". Cronbach's α was 0.868 in this study.

Results
In this paper, based on the survey data and M-plus software, structural equation modeling and bootstrapping analysis were used to test the above hypothesis and models, with the aim to explore the essential relationship between guanxi and employees' job satisfaction.

Job-Searching Process
Model 1 and Model 2 supported the research hypothesis. The results of basic descriptive statistics regarding the respondents are shown below (see table 1): In this investigation, some channels were used during the job-searching process. For those respondents who used guanxi during their job-seeking process, social media (such as We Chat, QQ, Facebook, websites, email, etc.) were the most frequently used channel, as shown in figure 3.
As for the role of guanxi in supporting job searching, the results are shown in figure 4. According to our investigation, providing the respondents with recruitment information is its most important role.
Model 1. The relationship between guanxi (renqing & ganqing), xinren (general trust), and job satisfaction during the job-searching process was tested by Model 1, and the results are shown below (see table 2):   Table 2 presents the means, standard deviations, and correlations for guanxi, xinren, and job satisfaction. As we can see from the results, guanxi is positively correlated with job satisfaction (B=0.46,p < 0.001) and xinren (B=0.57,p < 0.001). Xinren is also positively correlated with job satisfaction (B=0.39,p < 0.001). Hypotheses H1, H2a, and H2b were preliminarily supported. Table 3 presents the confirmatory factor analysis (CFA) measurement results of Model 1. CFA showed that the CFA fitting result of the single-factor model (χ 2 /d f = 12.74,RMSEA=0.21,CFI=0.67,TLI= 0.57) is not ideal at all, which indicates that the common method variance between variables does not have a serious impact. Through different combinations of the three variables and CFA testing, it can be found that none of the fitting indexes meet the adaptation standard, while the fitting index of the three-factor model (χ 2 /d f = 4.93,RMSEA=0.08,CFI=0.91,TLI= 0.90) is obviously better than that of the other four models. Therefore, the three-factor model is considered to be more applicable.
The results of structural equation modeling and bootstrapping analysis are shown in figure 5 and table 4. As can be seen in table 4, after the gender, company type, marital status, age, level of education, length of work, and yearly income were controlled, the total effects of Note: Single-factor model: combine all the items. Note: GX = guanxi (renqing & ganqing),XR = xinren,JS = job satisfaction,CI = confidence interval; ***p < 0.001,**p < 0.01,*p < 0.05. guanxi on job satisfaction were significant; thus, hypothesis H1 was supported. In addition, the indirect effects of guanxi on job satisfaction via xinren were significant, and the conditional direct effects of guanxi on job satisfaction were also significant. This means that xinren partly mediated the effect of guanxi on job satisfaction; therefore, hypotheses H2a and H2b, as well as H2, were supported. Note: N=267; ***p < 0.001,**p < 0.01,*p < 0.05. Note: Single-factor model: combine all the items.

On-the-job Process
Model 2. The relationship between guanxi (supervisor-subordinate), xinren (trust in supervisor) and job satisfaction was tested by Model 2, and the results are shown below (see table  5): Table 5 presents the means, standard deviations, and correlations for S-s guanxi, xinren, and job satisfaction. As we can see from the results, S-s guanxi is positively correlated with job satisfaction (B=0.44,p < 0.001), S-s guanxi is positively correlated with xinren (B=0.53,p < 0.001), and xinren is also positively correlated with Job satisfaction (B=0.56,p < 0.001). Table 6 presents the CFA measurement results of Model 2. CFA showed that the fitting result of the single-factor model (χ 2 /d f = 13.13,RMSEA = 0.21,CFI = 0.65,TLI = 0.58) is not ideal, which indicates that the common method variance between variables does not have a serious impact. Through different combinations of the three variables and CFA testing, it can be found that none of the fitting indexes meet the adaptation standard, while the fitting index of the three-factor model (χ 2 /d f = 3.34,RMSEA = 0.08, CFI = 0.94, TLI = 0.92) is obviously better than that of the other four models. Therefore, the three-factor model is considered to be more applicable. Figure 6 and table 7 show the results of structural equation modeling analysis. As can be seen in table 7, after gender, company type, marital status, age, level of education, length of work, and yearly income were controlled for, the total effects of supervisor-subordinate guanxi on job satisfaction were significant; thus, hypothesis H3 was supported. In addition, the indirect effects of supervisor-subordinate guanxi on job satisfaction via trust in supervisor were significant, and the conditional direct effects of supervisor-subordinate guanxi on job satisfaction were also significant. This means that the trust in supervisor partly mediated the effect of supervisor-subordinate guanxi on job satisfaction; therefore, hypotheses H4a and H4b, as well as H4, were supported. Note: SSG = guanxi (supervisor-subordinate), XR = xinren(trust in supervisor),JS = job satisfaction,CI = confidence interval; ***p < 0.001,**p < 0.01,*p < 0.05.

Discussion
According to the research results, all the hypotheses were supported. Guanxi (renqing & ganqing) is positively associated with employees' job satisfaction, and xinren (general trust) will mediate the relationship between guanxi (renqing & ganqing) and employees' job satisfaction. Guanxi (supervisor-subordinate) is also positively associated with employees' job satisfaction, and xinren (trust in supervisor) will mediate the relationship between guanxi (supervisor-subordinate) and employees' job satisfaction.

Implications for Theory and Practice
Using the M-plus software for analysis, the results reveal that all eight proposed hypotheses are supported; furthermore, two proposed hypotheses are contingent upon the moderating effect of general trust, while two hypotheses are moderated by trust in supervisor. Therefore, by theoretically proposing and empirically examining the guanxi-job satisfaction models in relation to xinren, this paper yields several theoretical implications. First, the extant literature as well as guanxi studies tend to treat ganqing, renqing, and xinren collectively as sub-dimensions of guanxi [13,14]. As the findings empirically reveal that both renqing & ganqing guanxi and supervisor-subordinate guanxi significantly improve job satisfaction, we argue that guanxi comprises different types that need to be studied in their own right. In particular, by illuminating how xinren (general trust or trust in supervisor) may be effectively employed to mediate job satisfaction, this paper challenges the previous assumption that treats guanxi as one combined and integrated notion, ignoring its dimensionality, thus leading to muddled empirical findings. Therefore, by responding to the call of further investigate the impact of individual guanxi dimensions on business relationships, this paper extends previous understanding on guanxi and its composition.
Secondly, the results confirm xinren (general trust) as an important moderator that influences the effectiveness of renqing & ganqing guanxi on improving job satisfaction during the job-searching process and xinren (trust in supervisor) as an important moderator that influences the effectiveness of supervisor-subordinate guanxi on improving job satisfaction during the on-the-job process. Specifically, the findings reveal that xinren significantly strengthens guanxi's impact on improving job satisfaction. Overall, the specifics provided in this study greatly contribute to the existing guanxi literature, showing that guanxi does matter during the job-seeking process and job outcomes-especially in the engagement of renqing & ganqing.
Finally, the empirical findings reveal that guanxi has a different influence on job satisfaction for the state and non-state sectors. The results show a significant difference between the job-seeking process and after entering the sector. This competing view of short and long business relationships in studying guanxi and conflict is the first of its kind and thus a contribution of this study.

Limitations and Future Research
Although this research has attempted to empirically validate the proposed model, some caution should be exercised when interpreting the results. First, while the data were solely collected from general Chinese respondents, the findings fail to consider the views from other countries and also to focus on one special industry, which may have skewed the results to some degree. Especially with the sharing economy and digital economy developing prosperously, this study could not obtain deeper data to analyze the difference with traditional industries. Future research is encouraged to investigate guanxi under the new economy.
Second, ganqing, renqing, and xinren are discussed as the key sub-dimensions of Chinese guanxi, but guanxi has more than three dimensions, and this study merely discussed the traditional GRX scale. Future studies may want to further explore and engage with different guanxi behavior norms-such as face-at the personal level.
Third, the scale of xinren (general trust) and job satisfaction used in this study merely contain general questions and thus may not be suitable for use in a specific context. Although it had acceptable reliability in this study, scale development in these two dimensions is recommended for future research.
In addition, this study investigated the sole mediating effect of xinren in examining the relationship between guanxi and employees' job satisfaction. Future research should consider examining the potential mediating effects of other variables on this relationship.

Conclusion
Although the number of studies on renqing & ganqing guanxi or supervisor-subordinate guanxi has increased in the last decade, few researchers have investigated the effect of guanxi on job satisfaction, with xinren as the mediator. The findings of this paper are similar to those of Jie Li et al., [9]-that is, guanxi was observed to be positively correlated with employees' job satisfaction-and thus extend the extant related research. It was also found that S-s guanxi was positively correlated with trust in supervisor, which is in accord with the findings reported in previous studies and indicates that S-s guanxi plays a significant role in the formation and development of subordinates' trust in their supervisor in a Chinese context. Meanwhile, renqing & ganqing guanxi were found to be positively correlated with xinren (general trust) and affect job satisfaction. Further, xinren (general trust or trust in supervisor) partly mediated the effect of guanxi on employees' job satisfaction, which shows that guanxi influences employees' job satisfaction not only directly but also indirectly. This finding extended previous research on the effects of guanxi on job satisfaction and may be used as an exploratory guide to manage employee satisfaction in different companies.