Guanxi refers to carefully developed and maintained informal particularistic dyadic ties based on reciprocal obligation and mutual affection (Bedford & Hwang, 2013). Guanxi relationships have a critical impact on a wide variety of organizational and professional outcomes (see Luo, Huang, & Wang, 2012 for a meta-analysis of the guanxi-performance link). Similarly, interpersonal trust is a central relationship requirement that impacts organizational and professional outcomes (see Lascaux, 2020; Paluri & Mishal, 2020 for systematic reviews). Trust is also the most important construct influencing development of business relationships in China (Leung et al., 2005).

Given their common relevance to workplace outcomes, it is not surprising that many studies investigating guanxi in an organizational context also address interpersonal trust (Bian, 1997, finding employment; Chou et al., 2006, team members' effectiveness; Yan, 2018, voice behavior). What is more surprising is the lack of consensus on the relationship between these two important concepts. For example, scholars have posited trust both as an outcome of workplace guanxi relationships (e.g., Song et al., 2012; Wong et al., 2003), and as an antecedent (e.g., Shou et al., 2011). They have also cast trust as a dimension of guanxi (e.g., Barnes et al., 2011), equated trust and guanxi (e.g., Burt et al., 2018), and considered them to be independent variables (e.g., Huang et al., 2011).

This problem is not new with respect to investigations related to trust. Such research is troubled by a “lack of clear differentiation among factors that contribute to trust, trust itself, and trust outcomes” (Mayer et al., 1995, p. 1995). Two interrelated factors may have contributed to this problem: culture and conceptualization.

Scholars have noted an over-reliance on Western conceptions of interpersonal trust in studies of non-Western societies (Kriz & Keating, 2010). Despite encompassing some universal elements of trust, these models may emphasize the values of individualistic cultures (Wasti et al., 2011). However, the issue of culture’s role in such research goes deeper than differences in the understanding of, and perhaps operation of, trust. It relates to the second factor, conceptualization.

Trust development tends to be described as a passive process of assessing the trustee’s characteristics and observing the trustee’s behavior over time (Child & Möllering, 2003; Lewicki & Bunker, 1996). It is a dispositional approach that relies on “stable within party traits” (p. 715) of the trustee such as ability or integrity, and individual traits of the trustor, such as propensity to trust and risk tolerance (Mayer et al., 1995). This perspective does not acknowledge the actor’s ability to take action to create or develop trust; it is not dynamic. The guanxi literature has been similarly criticized for excessive focus on guanxi’s pragmatic utility for business outcomes (the benefits that having guanxi brings) rather than the dynamics of interpersonal exchange and the process of guanxi development (e.g., Mao et al., 2012). The static way guanxi and trust tend to be conceptualized in the management literature may have obscured the relation between them leading to the variety of proposed relationships.

Emphasis on a static conceptualization is a function of culture. The static trait perspective common in research related to guanxi and trust is grounded in an understanding of metaphysical identity reflecting the philosophy of early Greece. In this perspective, objective reality is mainly considered in terms of entities and their external relationships. These relationships are not seen as impacting the entity’s essential nature—analogous to the change in one billiard ball when it is hit by another (Birch & Cobb, 1981). Because individuals are understood to be self-contained and self-sufficient the focus is on identifying their fixed attributes, and each individual is seen as an autonomous unit of action that is independent from social connections (Bedford & Hwang, 2003).

In contrast, scientific investigation from a process-relational perspective focuses on interactions between the entity and its environment, rather than directly on the entity in isolation (Bedford & Yeh, 2020). This ontological perspective allows for interaction to change the nature of the entity (Birch & Cobb, 1981). In other words, entities cannot be understood without attending to their interconnectedness. This scientific perspective is aligned with the human-centered philosophy of Confucianism in which people are defined as interactive beings (King, 1991).

The goal of this paper is to 1) identify how culture and conceptualization have played a role in research on guanxi and trust and 2) propose a dynamic process model integrating the development of guanxi and indigenous forms of trust. In the following, I first provide an overview of the way in which trust has been examined in relation to guanxi in the workplace and the Western norms embedded in that conceptualization. I then review several models of guanxi in order to introduce the concept of renqing (favor), and demonstrate the problem with conceptualizing workplace guanxi processes in terms of a Western model of trust. I introduce sociological conceptualizations of trust that align with mianzi (face/reputation), xinyong (social credit), and xinren (trust), and then integrate these discussions to propose a dynamic process model of guanxi and trust development.

Guanxi and Interpersonal Trust in the Workplace

Researchers examining the role of guanxi in an organizational context often talk about trust using only the English word trust, and they operationalize trust using a Western measure (e.g., Chiu et al., 2018; Chou et al., 2006). The measure used tends to reflect one of two models in which the trustor’s perceptions determine the emergence of trust. The first approach entails identifying a set of cognitive assessments that lead to trust or a determination of trustworthiness: for example, competence, honesty (openness), consistency (reliability), loyalty, and benevolence (concern) (Mishra, 1996). Many researchers investigating organizational outcomes related to guanxi select a subset of these traits to represent and measure a particular aspect of workplace trust.

This conceptualization of interpersonal trust is based on the assumption that work relationships emphasize competence and instrumentality in accomplishing tasks (aspects of cognitive trust). It also reflects the Protestant ethic of keeping emotional and instrumental matters separate, so that professional relationships are distinct from personal relationships, which have an affective component (Sanchez-Burks, 2002). In other words, for Westerners, workplace decisions should be impersonal; mixing emotion with business seems unprofessional (Yum, 1988). The ideal is to keep them separate.

McAllister (1995) extended the cognitive approach. Drawing on research on trust and interpersonal relationships outside the workplace, he recognized that emotions and self-disclosure (elements of affect-based trust) are also important in workplace relationships. However, as with the earlier approach, he cast the trustor as “reacting to trust stimuli” (Child & Möllering, 2003, p. 70). In this perspective, the external context supplies the signals for cognitive-based trust in the form of social similarity, organizational context (including credentials and reputation), and the success of past interactions. This, cognitive trust is a necessary prerequisite for investing in development of affective trust, which is grounded in the perception of the internal motives of the trustee as demonstrated by the trustee’s behavior that meets legitimate needs and demonstrates care without being self-serving.

Culture and Workplace Trust

Some researchers who applied McAllister’s model in organizational studies in Chinese societies found significant results with respect to affect-based trust, but not cognitive trust; that is, cognitive trust did not appear to be a prerequisite of affect-based trust. Researchers who specifically investigated Chinese and Western cultural differences with respect to McAllister’s framework concluded that although trust everywhere may have cognitive and affective components, in Chinese societies affective trust is likely to be more important than (Wasti et al., 2011) or more intertwined with (Chua et al., 2009) cognitive trust than in Western societies.

Hwang’s (1987) face and favor model provides insight into why affect-based trust may be of greater significance for people in Chinese societies. Hwang argued that interpersonal interaction in Chinese societies is guided by Confucian social and ethical norms; the particular norms governing interaction depend on the main type of ties constituting the relationship: expressive (mainly affective and longer term), instrumental (goal-oriented and transient), or mixed ties (Hwang, 1987). This last type–a mixture of expressive and instrumental ties–is the one that prevails in the workplace (Luo, 2011). Relations with strangers are largely instrumental, meaning that they are transactional in nature and operate according to the rule of equity. A relationship transforms to mixed ties with the cultivation of expressive elements. Emotional attachment changes the principle guiding interaction from the rule of equity to the rule of renqing, which carries a moral obligation to return favors (Hwang, 1987). Thus, “Chinese people favor using familiar ties in a business context because strong personal trust can be built up through renqing fa ze–the rules of favor exchange” (Luo, 2011, p. 331).

The term renqing has three simultaneous meanings (Hwang, 1987). First, similar to the noun favor in English, it refers to a resource or an action offered as a favor. The other two aspects specify the ethical requirements for adhering to the rule of renqing: demonstration that one knows and acts according to the social norms for proper behavior (particularly with regard to reciprocity in favors), and sensitivity to the needs of those with whom one is close (offering personalized favors and flexibility when needed). The concept of renqing is bound up with trust. Following the rule of renqing is a kind of test of trustworthiness in that people who do not adhere to this norm (such as by not returning a favor) are seen as untrustworthy and/or immoral as their actions could cause others to lose face (Yau et al., 2000). Reciprocation of favors is morally binding and thus corresponds to perception of a person’s commitment and credibility, which are aspects of trust (Lee & Ellis, 2000).

Expressive ties are a central requirement of workplace guanxi because they transform the rules guiding interpersonal interaction to emphasize renqing—reciprocal obligation. Expressive ties are developed through creation of ganqing (affection), which requires ongoing social interaction and mutual assistance and makes guanxi “more dependable and valuable” (Kiong & Kee, 1998, p. 80). Moreover, unlike in the Western context, in a Chinese context there is not necessarily a clear dichotomy between personal and professional relations (Sanchez-Burks & Lee, 2007). Chinese people expect workplace relationships to encompass affection (Wang et al., 2008), and it is considered normal for professional contacts to ask a personal favor and vice-versa (Bedford & Hwang, 2013).

A Process Model of Guanxi with a Static Model of Trust

Chen and Chen (2004) proposed a process model of guanxi development encompassing renqing and ganqing. They asserted that Chinese interpersonal trust corresponds well to the dispositional model of trust, and organized their presentation of guanxi around McAllister’s (1995) framework. They defined guanxi as a product of mutual trust and feeling, each of which in turn comprises elements of both cognitive and affective trust (see Fig. 1). They represented trust with the Chinese character xin, and specified that xin encompasses perceived ability and sincerity, both of which have been depicted in the Western literature as aspects of cognitive trust (see Mayer et al., 1995 for a review). Chen and Chen (2004) pointed out that the normal translation for the English word trust is xinren, which refers to belief in a person’s ability, reliability, and usability, which are all aspects of cognitive trust. They defined sincerity as manifesting through “being reliable by following social norms” (p. 314).

Fig. 1
figure 1

Process model of guanxi using a dispositional model of trust adapted from Chen and Chen (2004)

Full size image

Chen and Chen (2004) represented the second component of guanxi, feeling, with the Chinese word qing. Establishing a sense of obligation and affective attachment (ganqing) provides the foundation for feeling. Obligation refers to the exchange aspect of renqing in the sense of a “long exchange history” or “wide scope of exchanges” (p. 314).

Chen and Chen (2004) suggested that ability-based trust is more related to the exchange aspect of renqing, and sincerity-based trust is more related to ganqing. In other words, they conceptualized guanxi in terms of two distinct concepts, xin and qing, which respectively relate to cognitive trust and affective trust in a general sense, and each of which is in turn composed of both cognitive and affective trust. Although they did not explicitly state it, in their model cognitive and affective aspects of trust are thoroughly intertwined.

In fact, the relationships among these concepts may be even more complex than Chen and Chen (2004) indicated because there is an inconsistency in their conceptualization of sincerity in terms of reliability in following social norms. They claimed that sincerity relates more to ganqing than to renqing. However, their definition of sincerity aligns with the second meaning of renqing, which implies that renqing plays a role in both the xin and the qing aspects of guanxi.

The inconsistency in conceptualization illustrates the complexity and perhaps futility of dividing Chinese workplace trust development into neat cognitive and affective categories. Although there is evidence that these two forms of trust are relevant to Chinese societies (Wasti et al., 2011), this dichotomy may not be the best way to understand trust development in context of guanxi (Bedford, 2019).

Two additional points regarding Chen and Chen’s (2004) model make it incomplete as a representation of guanxi operation. First, the dispositional model of trust they relied on contains the inherent assumption that trust is low when parties first meet, and then builds as cognitive trust develops through observation and assessment of the target in the given conditions until it reaches a point in the relationship when it is worth affective involvement. This approach does not address the situation in which two strangers can have a relatively high level of initial trust, which is the case in some forms of business-related guanxi (see Bedford, 2011; Fan, 2002). In other words, this model addresses only one kind of guanxi—the kind that builds slowly over time. Second, Chen and Chen’s model does not address other important indigenous Chinese forms of trust that are relevant to the workplace, nor does it address the role of face (mianzi), which is also connected to the operation of guanxi (Hwang, 1987). To address these points, in the next section I review another process model of guanxi that includes two forms of guanxi development in the workplace and also incorporates additional indigenous concepts.

A Process Model of Guanxi

Bedford (2011) reviewed and critiqued several guanxi process models (i.e., Buttery & Wong, 1999; Chen et al., 2004; Fan, 2002; Wong et al., 2007), integrated them, and addressed their shortcomings to develop a dynamic process model of workplace guanxi operation. The resulting model is different from the others in that it 1) is constructed specifically with reference to workplace practices, 2) encompasses guanxi development and maintenance for two types of workplace guanxi, 3) captures the hierarchical nature of relationships, and 4) is constructed on the basis of indigenous concepts.

Bedford’s (2011) model includes a temporal dimension with four stages of guanxi development. The quality of guanxi is different in each stage. In the initial stage, probing, each side identifies the strengths and weaknesses of the other in order to decide whether to invest in building the relationship. In the proving stage, both sides have the intention to form a guanxi channel, but they are still outsiders to one another. The expectation of future exchanges and mutual benefits keeps the process going. Constant social exchanges gradually establish a relationship that can be drawn upon when one needs help in the new friend stage, but it is still important to maintain the relationship through proving activities. The final step is the old friend stage, when the working guanxi channel is fully established and the partner is considered an insider (see Fig. 2).

Fig. 2
figure 2

Process model of guanxi adapted from Bedford (2011)

Full size image

Bedford’s (2011) model differentiates two types of work-related guanxi: working guanxi and backdoor guanxi. Working guanxi entails slow and steady development through the exchange of small favors and affection into a stronger relationship that constitutes a basis for asking for larger favors. It encompasses all four stages of development. Backdoor guanxi makes use of an intermediary (who has working guanxi with both parties) to establish a new guanxi channel without need to expend the effort required to initiate a new relationship. One can skip the preliminary time-intensive stages of guanxi building and begin the relationship at the new friend stage. These two categories of guanxi are operational modes that do not necessarily have a discrete boundary; they may be seen not only as different ways of building guanxi, but also as different ways of using guanxi.

A key insight of Bedford’s (2011) model is to discriminate between two types of face important in the development of guanxi: lian (moral character) and mianzi (status/power and reputation). Development of working guanxi corresponds to growing trust in the target’s lian, which indicates that one recognizes that the actor has accurately represented their character in terms of ability and care and their willingness to adhere to social norms in terms of face, flexibility, and favors. In contrast, backdoor guanxi is powered by the magnetism of mianzi, which reflects one’s reputation, professional credentials and achievements, as well as the caliber of the organization at which one works. Mianzi indexes a person’s ability to satisfy requirements or provide reliable results. The basis for initiating a backdoor relationship is the target’s status and power (mianzi). Because of the intermediary, frequent interaction is not necessary for backdoor guanxi to develop as there is little need to test the target’s moral character if the intermediary is trusted. However, a channel initiated through an intermediary can endure beyond the initial exchange and develop with the continued exchange of reciprocal favors into a relationship of old friends with working guanxi.

Although Bedford’s (2011) model captures the dynamics of different types of guanxi, it is quite general with respect to trust. It does not indicate the process of trust development or how this process differs between the two types of guanxi. It also lacks any discussion of the relation of cognitive and affective aspects of trust to guanxi, and neglects an important aspect of trust in a Chinese business context: xinyong (a kind of social credit score). In the next section, I review a sociological process model of interpersonal trust and discuss how it relates to various Chinese concepts in order to propose an updated process model that integrates the operation of guanxi and trust.

A Process Model of Trust

Instead of focusing on individual differences in readiness to trust and contextual factors that impact this readiness (trust as a psychological state within the individual), sociologists tend to adopt a process perspective that conceptualizes trust as existing both within and between institutions and individuals (see Lewicki & Bunker, 1996), with institution-based trust providing a foundation for interpersonal trust (Durkheim, 1964). Institution-based trust is defined as trust based on the institutional context or third-party structures grounded in reputation systems such as guarantees and recommendations from third parties (Zucker, 1986). It is of particular relevance when direct personal experience is lacking, and thus critical in a business context. Institution-based trust requires no investment or increased vulnerability because it rests on the given context.

Calculus-Based Trust, Xinyong, and Mianzi

In this vein, Lewicki and Bunker (1996) proposed the notion of calculus-based trust, which addresses consistency of behavior. They indicated that individuals continually calculate whether the benefits of having their words and deeds accord outweigh the consequences of defecting. Although there may be benefits to defecting, the largest deterrent is concern for one’s reputation in the wider business network. In the early stages of the relationship, this deterrent is the stronger force, because a single inconsistency can harm not just one’s relationship with the partner, but also one’s standing in the wider network. Inconsistency not only discourages development of trust, it signals a reason for distrust.

This description of calculus-based trust is quite similar to Kiong and Kee’s (1998) discussion of xinyong. Xinyong refers to the usefulness of trust; it is a kind of social credit rating grounded in the community assessment of an individual’s trustworthiness (Leung et al., 2005). It is often used to mean credit in a financial sense (e.g., credit card is translated as xinyong ka), which underscores the instrumentality conveyed by the term. Xinyong goes beyond the simple notion of reliability or predictability to ethicality. That is, if a person is perceived to have good xinyong, that person is seen as ethical (Leung, Chan, Lai, & Ngai, 2011). Early in a relationship, individuals need to prove their xinyong. However, xinyong does not exist only between individuals; it is grounded in social relations. People who break their word lose their xinyong, with the implication that there may be social sanctions. That is, if an actor defects with one guanxi partner, others are likely to react as well. Just as Lewicki and Bunker (1996) proposed with calculus-based trust, fear of damage to xinyong may support a person in remaining trustworthy.

Whereas one’s own reputation provides motivation to remain trustworthy, the other person’s reputation might be a reason to decide to trust them. Sociologist Zucker (1986) argued that an individual’s reputation, as indicated by attributes such as professional association membership, professional certification, academic credentials, job title, or even the caliber of organization at which they work, supports institution-based (or calculus-based) trust. A person’s mianzi (face/reputation) is influenced by these same attributes. Having high mianzi is similar to having strong symbolic capital (prestige as a credit that attracts other forms of capital; Bordieu, 1986). As indicated in Bedford’s (2011) model, high mianzi is a magnet for guanxi. It draws people in without need to expend much effort to create or sustain relationships. In contrast, having low mianzi requires continuous work to develop guanxi.

Xinyong and mianzi clearly relate to aspects of calculus-based trust proposed by sociologists. Xinyong denotes the social consensus on an individual’s credibility and integrity and functions like a social credit score. Mianzi signifies external attributes of an individual such as their academic degrees, where they work, and with whom they are friends. Calculus-based trust is a form of cognitive trust (Lewicki & Bunker, 1996). An empirical study of Chinese employees indirectly supports a connection between cognitive trust, mianzi, and xinyong. Chua et al. (2009) found that the extent of a person’s embeddedness in the social network increased cognitive trust of that person. If it is the case that together mianzi and xinyong (which Chua et al. did not mention) represent the degree of a person’s social embeddedness, then it would be reasonable to conclude they correspond to the level of cognitive trust.

Knowledge-Based Trust and Identification-Based Trust

Lewicki and Bunker (1996) proposed three phases of trust development, of which calculus-based trust is the initial stage. The next two stages are knowledge-, and identification-based trust. Lewicki and Bunker envisioned them as a staircase, with each building on the prior type. The three types of trust play different roles over the course of relationship development, and each entails different processes of relationship building and testing.

The second stage entails building knowledge-based trust on the foundation of calculus-based trust as the relationship progresses. Knowledge-based trust is grounded in the predictability that comes from knowing the partner well enough to foretell what they will do. It requires time to develop and sufficient frequency of interaction to gain a feeling of understanding. Once knowledge-based trust develops, trust is not necessarily damaged by inconsistent behavior so long as one believes one understands the reason behind the partner’s inconsistency. Knowledge-based trust is a cognitive assessment; action is not part of it. The individual observes and notes past interactions as a way to assess the partner’s consistency. Action is the way to move into the next phase of identification-based trust, particularly with actions that demonstrate true caring.

Identification-based trust represents feelings of personal attachment and confidence that the partner will always protect one’s interests so that no monitoring is necessary. Lewicki and Bunker described identification-based trust as an emotional connection that results in harmonization of interests. Whereas cognitive aspects of trust are responsible for maintaining calculus-based trust and developing knowledge-based trust, affective aspects of trust are most important for identification-based trust.

Renqing

Following the rules of renqing is a requirement for building workplace guanxi. I propose that enacting renqing simultaneously supports the development of both knowledge-based trust and identification-based trust. Understanding how it does this requires examination of the expectations surrounding reciprocity.

Knowledge-based trust requires time and personal experience to get to know the partner in order to attain a certain level of understanding and predictability. Although Lewicki and Bunker (1996) did not directly discuss reciprocity, it likely plays a role in the development of knowledge-based trust because most societies have social norms supporting a responsibility to return favors received (Cialdini, 1988; Coyle-Shapiro & Conway, 2005). Social exchanges require belief in the goodwill of the partner to reciprocate, and the observed predictability in exchanges over the long term fosters trust (Blau, 1964). Westerners tend to see this reciprocity as individual-based. Workplace exchanges between Westerners tend to adhere to the equity rule: people anticipate being pretty much even after each round of exchanges (Yum, 1988). Obligation ends with repayment. Short-term symmetrical reciprocity is the expectation (English, 1979). Knowing whether someone is likely to reciprocate adds to the knowledge base on predictability and can increase cognitive trust as Lewicki and Bunker suggested.

Expectations surrounding the Chinese reciprocity norm of renqing are different from those surrounding the Western reciprocity norm in three ways. First, reciprocity is not individual-based, but institution-based. That is, reciprocity is part of the social fabric of Confucian ethics. It is a moral requirement for the self. Failing to repay a favor when it is needed not only violates the relational expectation of the partner, it also violates a wider social norm grounded in Confucian ethics, which not only may cause both parties to lose face, it may also correspond to a reduction in xinyong for the defector (Wong et al., 2007).

Second, reciprocity is an asymmetrical requirement in that the obligation is to return a favor larger than that received. This need to repay a larger favor binds the two individuals together—if the balance is ever made even, there is nothing to drive continuation of the relationship. As exchanges increase in size, so does the individuals’ commitment to each other, which supports a long-term perspective. Given these benefits, individuals may ask favors when they are not necessary so as to promote the development of the relationship (Bedford & Hwang, 2011).

Third, renqing requires sensitivity to the needs of one’s partner. Favors should be personally tailored so as to convey concern for the partner’s best interests and demonstrate knowledge of the partner’s personal situation. Both of these attributes are antecedents of affective trust (McAllister, 1995), and they match the requirements for development of identification-based trust (Lewicki & Bunker, 1996). Moreover, the domain for demonstrating sensitivity to the partner’s needs is broad since there is no expectation of a clear boundary between personal and professional aspects of the relationship (Sanchez-Burks & Lee, 2007). A personal friend may ask a professional favor and vice versa.

In sum, behavior that accords with the rules of renqing bridges the requirements for building knowledge- and identification based trust; it encompasses them both. In the next section, I review the literature on xinren (trust), and propose that xinren develops as a consequence of practicing renqing, and that it embodies both the cognitive aspect of knowledge-based trust and the affective aspect of identification-based trust.

Xinren

As mentioned in the introduction, most workplace studies that encompass both guanxi and trust use only the English word trust and rely on items adapted from Western measures. Of the few studies that do include the term xinren, most conceptualize it by citing previous studies that applied Western descriptions of workplace trust to a Chinese context. The result is that xinren is mainly defined as “akin to the trust construct from Western relationship quality literature” in that it focuses on reliance and confidence between the two parties (Chu et al., 2019, p. 476), and as such is considered a cognitive component of guanxi (e.g., Berger et al., 2015). The most common measure of xinren seems to be that proposed by Yen et al. (2011), who used items reduced from Doney and Cannon’s (1997) seven-item measure of interpersonal trust to represent xinren. Doney and Cannon construed workplace trust as based solely on cognitive processes.

The two studies that specifically investigated the contents of xinren provide a different perspective. Kriz and Keating (2010) empirically explored the conceptualization of xinren with a qualitative study. Kriz and Keating’s analysis identified honesty and sincerity as the most important contents of xinren. Respondents described these elements as a product of continued emotional exchanges that lead to liking (such as providing help and support when needed, and working together with a sense of cooperation). Kriz and Keating pointed out that it is not just the assistance itself or even the confidence that assistance will be provided when needed that builds xinren; affective attachment is also a requirement. They concluded that xinren is driven by reciprocal help and emotional bonding, and, as their respondents stated, having xinren implies you have guanxi. Xinren is a person-to-person construct that does not vary between social and business settings. In other words, xinren transcends context. It is trust in the character of the person as a whole.

Migge et al. (2020) conducted a prototype analysis of xinren to identify its unique features. They found that xinren’s attributes include both cognitive and affective aspects, including honest, reliable, understanding, helpful, friend, happiness, love and joy. They pointed out that honesty, reliability, and cooperation are more closely aligned with the Western conceptualization of workplace trust, while friend, love, joy, and happiness may be more emic features. They also found a negative feature representing distrust embedded in the central conceptualization of xinren: liar. They suggested that this feature extends the spectrum of xinren beyond low trust to encompass distrust, and may represent the existence of a baseline of distrust with which Chinese people tend to approach outsiders.

The two studies just reviewed confirmed that xinren encompasses cognitive and affective aspects of trust that are established over time through interaction. Acting according to renqing simultaneously supports both cognitive and affective aspects of trust, thus it is reasonable to expect following the rules of renqing is a way to develop xinren. As the level of xinren grows guanxi matures.

A Process Model of Trust and Guanxi

Figure 3 displays a model that integrates guanxi and trust development processes. It is grounded in a distinction between rapid trust, which is based on cognitive factors, and process trust, which requires interaction over time to develop and entails a synergy of cognitive and affective aspects. Use of the rapid and process labels to understand trust development in the context of the workplace supports critical aspects of guanxi relationships. Rapid trust is based on information; interpersonal interaction is not a requirement. It is anchored in shared characteristics, guanxi bases, and one’s social network. Rapid trust allows for swift decision-making and flexibility. It highlights the rich sources of information that are relevant to a person’s decision to trust an outsider, and can also encompass the instrumental motivation for initiating guanxi. In contrast, process trust is experiential and takes time to develop. It is personal. It embodies the relational aspect of Chinese identity in that it emphasizes the value that can be created via interpersonal interaction. Process trust cannot exist without action. It is built through activities that accord with the rule of renqing.

Fig. 3
figure 3

A dynamic process model of guanxi and trust

Full size image

These two forms of trust are applied to build two types of guanxi: working guanxi and backdoor guanxi. A person becomes interested in initiating guanxi based on information that signals the instrumental benefits of such a relationship. If the information is sufficient to support rapid trust, a person may progress the relationship by acting in accordance with the rules of renqing. In this proving stage, both sides have the intention to develop guanxi, but they are still outsiders to one another, and so calculus-based aspects of trust play a role in allowing the expectation of future exchanges and mutual benefits to drive the renqing process. Once the reciprocal exchange of favours is initiated, it supports not only the instrumental aspect of the relationship, but also the affective aspect because being attentive to the guanxi partner’s needs and ensuring that the partner’s face is protected breeds affection (ganqing) in the relationship (Kipnis, 1997). This affection entails a degree of emotional understanding and feelings of connectedness and reduced boundaries. Addition of affection to instrumental ties supports development of process trust and transforms the relationship to mixed ties (Luo, 2011).

The instrumental benefits signaled by mianzi and the information available to support rapid trust can also drive an interest in establishing backdoor guanxi. However, instead of personally doing the work to build process trust and guanxi with the target, one relies on a person with whom one already has working guanxi to facilitate backdoor guanxi with the new target. This approach is similar to a cognitive trust process described by Doney and Cannon (1997): transference. They described how one person may use a third party’s trust of another as a basis for defining that other as trustworthy; cognitive trust can be transferred from one source to another. The implication is that, backdoor guanxi is only indirectly supported by affection and process trust by way of the intermediary.

Discussion

Unlike the mainstream research approach to investigation of trust and guanxi, a dynamic process approach highlights the ways in which these two concepts overlap. Both guanxi and trust are relational concepts; neither has meaning for an individual alone. Both are the result of interpersonal interaction. The more time and effort put into developing them, the more they facilitate interaction. The proposed model anchors guanxi development in indigenous trust processes, and in so doing enhances understanding of both guanxi and trust. Specifically, at least in a Chinese context, cognitive and affective elements of trust may be intertwined to the extent that it is not feasible to use this dichotomy to understand trust development. Rather than making a dividing line between these aspects, the proposed model distinguishes between trust that forms quickly (rapid trust), and trust that requires time to build (process trust).

Although a few studies have proposed models of guanxi development specific to the workplace, none have addressed the role of rapid trust or process trust in guanxi development. Rapid trust provides a foundation for developing guanxi and supports the instrumental motivation driving initiation of the relationship. Elements related to this aspect of trust can be considered antecedents to guanxi, although they are also relevant to the guanxi development process, especially in the early stages. Process trust encompasses both cognitive and affective elements, and is an outcome of and motivation for performing the norms of renqing.

Another novel contribution of the model is that it incorporates two indigenous concepts related to trust and posits a different role for each. Although studies have discussed one or the other in relation to guanxi, no previous study has proposed a model of guanxi operation that includes both. Identifying them both together is important because various studies have conceptualized xinyong (e.g., Leung et al., 2005) and xinren (e.g., Barnes et al., 2011) as equivalent to personal trust. Kriz and Keating (2010) suggested that some researchers had confounded these constructs with each other or with general business trust. The proposed model clearly differentiates them: xinyong is grounded in the social network. Xinren evolves between two particular individuals.

Focusing on Chinese indigenous forms of trust highlights new research questions related to these constructs. For example, Migge et al. (2020) indicated that xinren may encompass features related to distrust. They found liar to be a central aspect of xinren and cheat and betray to be peripheral features. They noted that having high trust and low distrust is the optimal configuration for ensuring partners have shared values and lack harmful motives. This distinction is an interesting avenue for future research in exploring the relation of xinyong to xinren. Xinyong captures the notion of reliability and ability to fulfill instrumental expectations. Perhaps having high xinyong is an indicator for low distrust that contributes to this aspect of xinren.

Another area that may be important for future research is the relation of Confucian ethics to the operation of guanxi, xinren and xinyong. The proposed model suggests that the production of xinren is solidly grounded in the norms of Confucian ethics, both directly–trust (xin) is one of the five core concepts of Confucian ethics–and indirectly through renqing, which is grounded in the Confucian ethics of yi (righteousness) and li (propriety). But, what about backdoor guanxi relationships? Other scholars have noted an increasing number of studies presenting the argument that guanxi is now widely used for instrumental purposes, and that this instrumental form of guanxi is dominating social exchange in China (see Wang & Murphy, 2010). At least in the early stages, there is no basis for direct ganqing or xinren in backdoor guanxi. Do such relationships tend to progress into working guanxi so that the instrumental benefits are balanced with affective concerns? Or, do they tend to degrade into bribery or corruption as the instrumental benefits increasingly accrue and outweigh any affective component of the relationship? Moreover, it is important to consider the broader question of the centrality of Confucian ethics in modern Chinese societies and how this may relate to guanxi practices. For example, around the same time that the Chinese authorities were attempting to eliminate traditional Confucian values (that they saw as contaminated with Western values) with the Cultural Revolution, Taiwan’s government was promoting a Chinese cultural renaissance movement. Might there then be a greater affinity for Confucian ethics in modern Taiwan, and if so, would this have implications for the operation of guanxi?

Implications for Other Societies

The proposed model provides a tool for investigation in Chinese societies. It also may have implications for people from other societies. Given the growth of China’s economy and the increasingly international nature of the workplace, understanding guanxi and trust development can be a source of competitive advantage for non-Chinese organizations and individuals. The model provides practical insight for foreign professionals working with Chinese colleagues or business associates in trying to understand these complex concepts. For example, the model highlights the powerful social resources that Chinese people have for quickly ascertaining a basis for rapid trust. While people from individualistic societies may have similar resources available, they may not be as explicit, and the tendency to rely on individual assessments instead of reports from others may mean that they are perceived as less important. Recognizing the role of xinyong and mianzi as well as the importance of practicing renqing to develop affection and xinren may help Westerners to develop relationships with their Chinese colleagues and business associates.

Beyond providing insight into Chinese practices, since the model focuses on the speed with which trust is formed, it may be a better fit than the cognitive-affective dichotomy in understanding trust development in other cultures, especially those that emphasize relational considerations. Moreover, it facilitates new perspectives on the ways in which cognitive and affective aspects of trust may be blended or interdependent in any culture. For example, Lewicki and Bunker (1996) asserted that calculus-based trust focuses solely on cognitive aspects. Yet, there is evidence that affective aspects play a role in rapidly-developed forms of trust in the interpretation of social similarity or organizational context. People may have immediate affective reactions to other groups or situations (Blair & Banaji, 1996) that contribute to the calculation of whether to defect or trust. Thus, some aspects of calculus-based (rapid) trust may be affective.

Another example of research that would extend insight from Chinese notions of trust relates to social embeddedness. The model indicates that important aspects of workplace trust are socially embedded: xinyong and mianzi are formed through collective assessment. Organizational researchers have tended tend to overlook socially embedded influences on behavior (Jeffries & Becker, 2008), but the social embeddedness of workplace trust dynamics may be of growing research interest given the proliferation of social media. Even in China this is a dynamic question as officials are implementing a new social credit score system integrating virtual and interpersonal relationships to assess trust and create Citizen Scores (Botsman, 2017). It remains to be seen what impact the socially embedded information available through social media will have on the development of trust in the organizational context.

Considerations and Conclusions

There are some constraints to keep in mind when considering the utility of the proposed model. First, the model is only intended to apply in the context of development of workplace guanxi relations. Trust development outside that context may proceed differently and is beyond the scope of this article. For example, this article focused on mixed tie guanxi, and I pointed out that backdoor guanxi is likely to move into unethical territory if instrumental ties remain overly central in the relationship. It is also possible that working guanxi can eventually attain a quasifamilial status if expressive ties become the dominant consideration. Such guanxi can become cronyism when one partner feels morally obliged to favor the other against the interests of the company (Ip, 2009).

Second, the analysis identified ways in which a model of trust focusing on a cognitive-affective dichotomy might not be optimal for Chinese culture. This discussion was not aimed at highlighting cultural differences—it was solely intended to identify how well such a model suits the context of guanxi development in a Chinese society. However, by highlighting a different process for building trust, indigenous models such as the one presented in this paper may have implications beyond Chinese culture. Context-specific research can produce novel insights for the local context, and at the same time have the potential to lead to global knowledge (Tsui, 2004). For example, it is possible that in individualistic societies, it is also the case that role-related behavior promotes affective trust, but that this aspect has been neglected due to the tendency to emphasize individual ability. Researchers have already highlighted affective trust development in the workplace (Williams, 2007) and interpersonal relations in business relationships (Lee & Dawes, 2005) as areas in need of research attention.

The importance of guanxi for personal and organizational success is well-established, but the processes that contribute to its development and its relationship with interpersonal trust, are less well understood. I proposed a preliminary theoretical process model for investigating the development and interaction of these key interpersonal relationship concepts. It specifies that different types of trust, xinyong and xinren, play different roles in guanxi development, and that these different types of trust reflect fundamentally different aspects of establishing trust in another person. Given that trust is also an important construct influencing the development of business relationships, it is hoped that the proposed model will stimulate a richer understanding of trust in the context of guanxi development in future research.