Guanxi Work Processes in Chinese Banquets: Varied Styles,Contingency,and Failure in Pay-Back for Favours

Abstract

This study analyses “guanxi work processes” – strategic practices that initiate, develop, shape, suspend, manipulate, and terminate guanxi relations – in Chinese banquets. It explores (i) different styles of carrying out guanxi work in banquets and (ii) guanxi work processes that are relatively unsuccessful and/or shaped by contingencies. Both are research gaps in the relevant scholarship. Adopting a robust processual perspective, this study analyses guanxi work processes that occur during the banquet without losing sight of those that occur pre-banquet and post-banquet. It identifies two styles of doing banquet guanxi work: one that emphasises pseudo-kin declaration and ritualistic toasting, and another that stresses non-reciprocal relational gifting and associated social skills. It also clarifies how guanxi work can fail in various ways and yield unexpected guanxi outcomes. Our data were collected from participant observation of fifteen banquets and multiple rounds of interviews with thirty-nine participants of these banquets.

Introduction

The scholarship on guanxi has substantially grown in the past few decades. Many subfields of guanxi studies emerged; one of them specialises in various details and dynamics of guanxi in banquets. This study contributes to advancing this subfield in three ways. The first way is to sharpen the conceptual tools for analysing guanxi in banquets. It utilises the term “guanxi work” to investigate various possible strategic practices of initiating, developing, shaping, suspending, manipulating, and terminating guanxi relations in banquets. It also conceptualises guanxi in banquets as long-term processes.

The second way that this study contributes to the subfield is to identify different styles of carrying out guanxi work in banquets. As we will show, although current studies usefully uncover a variety of guanxi practices in banquets, surprisingly few of them attempt to identify patterned variations of guanxi processes in banquets. The small number of studies that do so base their typology on guanxi outcomes instead of guanxi practices or processes (Bian, 2017; Lin, 2001). Unlike them, this study attends to differences in guanxi work processes found across different banquet cases. It identifies two styles of banquet guanxi work, each of which entails different guanxi work processes. Due to the small number of cases in our dataset, we certainly do not think that these two styles comprehensively represent all possible styles of banquet guanxi work. They serve as a starting point to facilitate further research on different banquet guanxi work styles.

The third way that this study contributes to the subfield is to extend its investigative focus to banquet guanxi work that does not achieve complete success and/or yield some unexpected outcomes. As our subsequent review of current studies will show, such banquet guanxi work is theoretically intriguing yet completely neglected. This neglect is obstructing the exploration of the fallibility and contingencies of guanxi work processes.

Guanxi Work Processes in Chinese Banquets

“[G]uanxi refers to the personal cultivation of a relationship with another in which favors are exchanged and through which ensuing obligations are incurred” (Barbalet, 2023: 75). As Chinese-styled relations, guanxi receives much scholarly attention due to its relevance for business, politics, society, and international communication. In guanxi studies, various large-scale subfields were formed to investigate topics such as guanxi's decline and persistence in contemporary China, the coexistence of instrumental and non-instrumental elements in guanxi, practical ways to build guanxi for business, and guanxi's positive and negative social impacts (Bian, 2018; Bruckermann and Feuchtwang, 2016; Li and Bian, 2024; Nolan and Rowley, 2020).

Banquets are one of the most important micro-level social settings in which guanxi is initiated, produced, shaped, suspended, manipulated, and terminated. A research subfield on guanxi in banquets is necessary, and it was indeed founded. However, this subfield has remained small and underdeveloped. Its English-language scholarship consists of two dozen dedicated papers and several books that offer extensive relevant discussions (Bian and He, 2022; Farquhar, 2002; Kipnis, 1997; Oxfeld, 2017, 2019; Ruan et al., 2024; Yan, 1996; Yang, 1994). This study adopts two conceptual tools to assist the empirical analysis of guanxi in banquets. The first tool is to conceptualise guanxi in a robustly processual way.

Conceptualising Guanxi in Banquets in a Processual way

This study adopts a processual perspective to analyse guanxi in banquets. The processual perspective is an integral part of relational social science (Vandenberghe, 2018). It is utilised in a broad variety of research fields (Langley and Tsoukas, 2016; Salehi, 2023). The majority of existing studies of banquet guanxi already analyse such guanxi as processes. For example, most anthropological studies of guanxi in banquets fully recognise the processual nature of guanxi and analyse the details of these processes (Kipnis, 1997; Yang, 1994).

However, an important stream of sociological studies of banquet guanxi is the exception. These are studies that emphasise the network conception and analysis of guanxi (Li and Bian, 2024). Network analyses are not necessarily non-processual, but network analyses of banquet guanxi tend to be non-processual. Theorists have warned against interpreting guanxi as mere networks. For example, Barbalet (2021: 168) thinks that “the construction of guanxi connections requires mutual long-term monitoring and surveillance as well as personal disclosure and shared activities of various sorts, as described by ethnographic accounts.” Nonetheless, the kind of long-term methods recommended by Barbalet is seldom employed by sociological studies of banquet guanxi (Lo and Otis, 2003). Following the example of anthropological studies of banquet guanxi and paying heed to Barbalet's (2021) recommendation, we contextualise guanxi in banquets within its pre-banquet processes and post-banquet processes. Our data collection and analyses take into account guanxi processes that occur during a banquet as well as those that occur pre-banquet and post-banquet.

Conceptualising Guanxi in Banquets as Guanxi Work

The second tool that this study adopts to assist the empirical analysis of guanxi in banquets is “guanxi work.” This term has been used in a small number of studies (e.g. Bian and Ang, 1997; Cologna, 2005; Grainger, 2005; Schneider and Wang, 2000; Wang, 2013). In these studies, it is used to metaphorically refer to efforts to build guanxi. Only one of these studies discusses guanxi work in the context of banquets (Farquhar, 2002: 151). In all except two of these studies, the term is casually used without any explanation, as if its meaning is self-evident. A study describes guanxi work as “including building, maintaining, and utilizing guanxi” (Robins and Kashima, 2008: 9). Another study briefly explains that guanxi work is a subtype of identity work (Tian, 2023: 109).

We theoretically clarify the concept of guanxi work. The word “work” in guanxi work borrows its meanings from influential concepts such as “emotion work” and “identity work” (Hochschild, 1979; Winkler, 2018). Emotion work and identity work were coined to highlight individuals’ strategic efforts to manage their emotions and identity. These two concepts have become widely used in social science scholarship. They yielded a family of terms such as aesthetic work and bodily work, which extend the idea of strategic personal management to different realms (Mears, 2014; Zelizer, 2000). The term guanxi work may be understood as yet another member of this family. It extends the idea of strategic management to guanxi, and it highlights individuals’ strategic efforts in managing their guanxi.

The major utility of the term guanxi work is that it serves as a comprehensive shorthand that takes into account all kinds of different actions, such as initiating, developing, shaping, suspending, manipulating, and terminating guanxi. Current studies of guanxi practices in banquets use terms such as “cultivating guanxi” and “building guanxi” (Yang, 1994). A negative, though unintended, consequence of such terms is to discourage the investigation of guanxi practices that do not primarily aim to cultivate guanxi. Careful observers of banquet guanxi processes, including Andrew Kipnis (2002: 22), would prefer a broader view that recognises that “practices of guanxi production” in banquets consist of “all of those social actions […] by which people created, manipulated, and at times eradicated human relationships.”

Fallibility and Contingencies of Guanxi Work in Banquets

Guanxi work is inherently fallible and contingent. This is because the “strength of any guanxi connection is always and necessarily a work in progress, never final, and always capable of being increased as well as decreased through the activities of those involved” (Barbalet, 2021: 168). While this argument is not very new or contested, few studies analyse the fallibility and contingency of guanxi work in banquets. In our view, fallibility and contingency inevitably arise when banquet guanxi work processes deviate from the course set by and/or expected by this banquet's primary organisers.

A reason that fallibility and contingency are neglected is that most banquets do not last for more than a few hours, which encourages the implicit presumption that few contingencies can transpire in such a short-term process. Moreover, some scholars treat guanxi phenomena in banquets as synchronic instead of processual (e.g. Bian, 2017). Cases of banquets with failed or unexpected guanxi outcomes are simply excluded from such studies’ samples.

It is entirely reasonable for past studies to prioritise investigating successful cases rather than failed and contingent cases. Yet, the more the field grows, the more urgent and meaningful it becomes to extend its investigative focus to relatively unsuccessful and contingent cases. This study contributes to making this extension.

Two Styles of Doing Banquet Guanxi Work and Their Respective Patterns of Guanxi Work Processes

A useful way of studying banquet guanxi work is to identify patterns of variation in guanxi work processes. However, most studies of banquet guanxi do not focus on variations. They identify and analyse various components of guanxi work in banquets, such as toasting rituals, sentiment, face (面子, mianzi), pseudo-kin declarations, opportunities for corruption, gifting, etc. These analyses are certainly invaluable, but the investigation of different guanxi work processes should not be sidelined.

A significant minority of studies on banquet guanxi work focus on its variations across different cases. An example is the network sociology-based demarcation between symmetric and asymmetric banquet guanxi (Bian, 2017; Lin, 2001). (Asymmetric banquet guanxi work aims to generate a creditor–debtor relationship between the favour-giver and the favour-receiver, whereas symmetric guanxi work aims to terminate guanxi.) Kipnis (2002) uncovers different guanxi work practices in banquets and emphasises the degree of such variability. Stafford (2006) highlights the distinction between top-down-driven and bottom-up-driven guanxi work in banquets. Bian and He (2022) classify different guanxi work based on their magnitude of instrumentality. While we applaud these studies’ attention to variations, we note that they base their typologies on variations of guanxi outcomes or motivations rather than variations of processes.

This study identifies two variations of guanxi work processes. These variations are caused by different styles of doing guanxi work. The first pattern features a style of guanxi work that stresses the declaration of pseudo-kin or friendship, ritualistic toasting and drinking, and top-down driven emotion work. The second pattern emphasises relational gifting, reciprocal care obligations generated via unreciprocated gifting, and social skills that generate emotional energy in bottom-up ways. We identified the two patterns and styles through grounded theory methods and multiple rounds of hermeneutical interpretations, as the Methods section will explain. The rest of this section discusses the theoretical background of these two styles and patterns.

The first pattern should look very familiar to scholars of guanxi. Its major guanxi work practices, most notably ritualistic toasting and the declaration of pseudo-kin, are seen as the signature characteristics of Chinese banquets. According to Wang et al.'s (2025) typology of toasting rituals, the first pattern's toasting resembles a “planned economy” mode of toasting instead of a “market economy” one. Emotional energy in this style of guanxi work is mainly generated by means of ritualistic toasting and declarations of pseudo-kin or friendship (Collins, 2014). Such declarations may be interpreted as “obfuscatory relational work” instead of truthfully expressive relational work (Rossman, 2014; Ruan et al., 2024: 305). Its emotion work may be interpreted as comparatively contrived (Hochschild, 1979).

The second pattern of guanxi work processes features a different style of doing guanxi work. Its guanxi work is primarily based on gifting, though some banquets of this pattern may include less ritualistic toasting and subtle expressions of friendship as secondary practices. This kind of gifting is conventionally known in anthropological studies as “dyadic gifting” between a favour-giver and a favour-receiver. Such gifting aims to generate a creditor–debtor relationship between the two actors. Adopting the new terminology of recent gifting research, one may also describe such gifting as “relational gifting” that aims to generate “reciprocal care obligations” (Weinberger et al., 2025: 15, 17). Reciprocal care obligations simply refer to the obligations generated in the creditor–debtor relationship.

Highly ritualistic toasting is avoided in banquets that adopt the second style of guanxi work. Less ritualistic toasting still occurs. Wang et al. (2025) describe such toasting as a “market economy” mode of toasting. In such toasting, actions are more courteous, more spontaneous, and more heavily driven from a bottom-up direction by participants’ feelings (Mao et al., 2021; Wang et al., 2025). More and better social skills are required to carry out the second style of guanxi work. “Social skills” is a technical sociological concept that refers to “the ability to engage others in collective […] reproduction of local social orders” (Fligstein, 2001: 105–106). In the specific context of non-reciprocal relational gifting in banquets, relevant social skills include ample preplanning, meticulousness in gift selection, subtlety in the forming of reciprocal care obligations, and efforts devoted to the ritual of gift refusal (Barbalet, 2025; Weinberger et al., 2025). Emotional energy in this pattern of guanxi work is generated through subtle social skills of gifting rather than high-profile declarations of pseudo-kin.

Methods

This study's data were collected from participant observation (PO) of fifteen banquets and interviews with thirty-nine participants of these banquets. Research ethics approval was granted by the Research Ethics Committee of our university. We interviewed two or more major participants from each banquet, most often the host and the main guest. The PO fieldwork was independently carried out by the two authors in two Chinese cities, Chongqing and Shenzhen, respectively, between December 2022 and February 2023. The first author conducted PO at seven banquets in Chongqing; the second author conducted PO at eight banquets in Shenzhen.

We had some sampling criteria for selecting these fifteen banquets. Firstly, they must be at least moderately driven by instrumental social exchange. We avoid banquets that are entirely driven by already intimate networks, such as ones that cater to the bonding between extended family members. This criterion lets us exclude cases of social eating in which non-instrumental interactions clearly dominate. (For a larger research project, the authors also conducted PO on cases in which non-instrumental interactions dominate. Those cases are not analysed in this paper.) Secondly, we only selected small-scale banquets. All banquet cases featured only one banquet table – the number of participants ranged between six and thirteen, including the participant observer. We excluded large life-cycle ritual banquets, which are common in rural China and quite well-researched by anthropological studies of guanxi. Thirdly, we only selected banquets occasioned by one favour-receiver or a small group of favour-receivers aiming to pay back a favour-giver.

Fourthly, we aimed to collect data from a range of banquets composed of participants from different socioeconomic backgrounds. Five of our banquet cases catered to wealthy individuals. The case with the highest expenditure per head was well over 1000 yuan (with an exchange rate of around one US dollar to seven yuan in early 2023). The amount included around 500 yuan for food and around 600 yuan for alcohol, which was most often red wine and Chinese white wine (白酒, baijiu). The alcohol was usually paid for and brought to the banquet by the host or a guest. Five of our banquet cases catered to middle-class individuals. The expenditure per head was over 400 yuan. The amount consisted of around 200 yuan for food and around 200 yuan for alcohol, which was usually beer and red wine. Five of our banquet cases catered to working-class persons; the expenditure per head was around 200 yuan. The amount included around 100 yuan for food and around 100 yuan for beer. Nine of the fifteen banquet cases took place in segregated rooms within commercial restaurants. Three cases occurred in restaurants without segregated rooms. Three other cases took place at private homes; the expenditure was lower in these cases.

The authors obtained access to these banquets through either (i) the host of the banquet, (ii) a main guest of the banquet, or (iii) a friend of the host of the banquet. The two authors have cultivated extensive social networks in the respective cities through relatives and friends. The two authors proactively searched for hosts of banquets in their social networks since mid-2022, which was when plans for this research project began. Through several months of preparation, the authors arranged to join a dozen banquets as minor guests. Many banquets in China are organised in the weeks around the Chinese Spring Festival, which usually occurs in February. The two authors also joined a dozen banquets between July 2022 and November 2022 as a pilot round of PO. The two authors carefully avoided intrusion in the paying-back processes and interaction rituals of the banquets. The first author did so by remaining silent most of the time and playing the role of a young guest who showed up for the learning experience and the food. The second author talked minimally and was introduced as a friend of the host (or the main guest) who happened to be travelling in the city for a few days. Our research purpose was revealed to the hosts, and permission was sought long before the fifteen banquets took place.

We followed similar procedures for conducting the PO at each banquet. We conducted pre-banquet informal interviews with our contact person for the banquet, who was usually the host or the main guest. We did this to ensure that we visited a banquet that would satisfy our sampling criteria and to gain a basic understanding of the social exchange involved. We also conducted informal in-depth interviews after the banquet. In a few cases, the major favour-receiver or favour-giver was unavailable for an interview. For these cases, we instead interviewed others who were informed about the instrumental exchange. We conducted these post-banquet interviews immediately after the banquet or within a week through online means. During the banquet, we adopted grounded theory methods to observe guanxi relations in an open-ended way. Namely, we observed various details of social relations and interactions without interpreting and analysing them with any theory.

In 2024, we conducted a round of follow-up interviews with the host or the main guests of eight of the banquets on which we collected data. These interviews’ objective was to learn about the evolution of guanxi between the favour-giver and favour-receiver(s) a year after the banquet. The most important question was whether their post-banquet guanxi relations maintained the direction and momentum achieved at the end of the banquet.

Our coding and analysis of the data can be described as grounded theory methods with an emphasis on hermeneutical interpretation (Age, 2011; Bryant and Charmaz, 2010). With grounded theory methods, we remained entirely theoretically open at the beginning of data collection; we strived to let new theoretical knowledge emerge during a gradual process of interpretative analysis (Charmaz, 2008). In the first round of hermeneutical interpretation, we tried to make sense of the emotional, moral, and instrumental elements in the banquets. We also began to notice fallibility and contingency in banquet guanxi. In the second round, we explored and compared how various theoretical concepts fit the data. We gradually settled on the relational–processual frame. By building on such interpretations and findings, we entered the third and final round of our hermeneutical interpretation. The data on failed and contingent cases were especially invaluable; they compelled us to adopt a robust processual perspective. The processual perspective helped sharpen our formulation of the research object, which, in turn, resulted in our adoption of the concept of guanxi work. Finally, our interpretation of data with the “guanxi work process” frame enabled us to pinpoint two styles of doing banquet guanxi work.

Our dataset contains fifteen banquet cases; this study presents detailed analyses of only five of them and brief analyses of another two of them. The similarity between some of our cases makes full coverage repetitive and unnecessary. Three banquet cases among those not discussed demonstrated the first style of doing banquet guanxi work. They featured successful guanxi work processes that emphasise declarations of pseudo-kin and ritualistic toasting. (The first section of our analysis discusses such guanxi work processes.) Two cases among those not discussed demonstrated the second style of doing banquet guanxi work. They featured successful guanxi work processes that emphasise gifting and associated social skills (The second section of our analysis discusses such guanxi work processes.) Two cases among those not discussed are banquet guanxi work processes shaped by contingency. (The third and fourth sections of our analysis discuss such guanxi work processes.) One case among those not discussed demonstrated guanxi work processes that failed. (The fifth section of our analysis discusses such guanxi work processes.)

Our analysis bracketed participants who did not play the favour-receiver role or the favour-giver role in the banquet. In fourteen of our fifteen cases, these third-party participants affected guanxi work processes in negligible ways. Like what current studies find, we observed that they fueled emotional energy at ritualistic toasts and acted as witnesses of pseudo-kin declarations. In one of our cases, we indeed observed third-party participants playing a non-negligible role in shaping the guanxi outcome. We believe that third-party participants’ actions constitute a good topic for future research on guanxi work in banquets. However, they deserve dedicated analyses and separate theorisation that cannot be achieved within the confines of this study.

Successful Guanxi Work Processes That Emphasise Declarations of Pseudo-kin and Ritualistic Toasting

This section examines a case that features successful guanxi work processes and the first style of doing guanxi work. Ao, a middle-class businessman, and his wife invited Ben to a banquet as the main guest, accompanied by other participants. Ben, a wealthy and powerful man, introduced a very effective private tutor to Ao's eleven-year-old daughter. A few months after learning from this tutor, Ao's daughter significantly improved her grades. (The context is that after extracurricular academic tutoring became heavily regulated in China in July 2021, an underground market for private tutoring was created. Informal relations became increasingly important for the securing of tutoring services.) The Ao couple organised a banquet at a high-end restaurant to thank Ben. Besides paying the bill for the banquet, Ao did not arrange to give any tangible gifts to Ben. The entire meal cost over 1,500 yuan; the alcohol was partly paid for by some guests.

During the banquet, Ao made a toast to Ben. Ao stood up with a glass of wine and faced Ben, who was sitting next to him. At this moment, all participants of the banquet stopped their conversation and attended the toasting ritual. Ao (2023) gave a toast speech.

I am very happy today at this gathering. I am making a toast to my elder brother (哥, ge) Ben for his tremendous help to me, my daughter, and her wonderful academic performance! Ben, you have introduced a great tutor to us!

After hearing this, Ben (2023) also stood up and asked Ao to sit down.

Don’t do this. It was nothing but a tiny favor. I treat you as my brother; your daughter is my niece. I’m glad that she has improved her academic performance.

Ao showed exaggerated facial expressions and the body language of joy. Then, Ao (2023) made a second toast to Ben and said,

I am so glad to hear that you treat my daughter as your niece. I will also treat your son as my nephew. I now toast to you and your family! If there is anything I can do for my nephew, I will do it.

Ben (2023) accepted Ao's toast and then made a reciprocal toast, saying “To you and your family!” After that, Ben and Ao drank up the wine in their glasses.

Interactions between Ao and Ben during the toasting illustrated their mutual efforts to form a pseudo-kin relationship. Ao (2023) initiated this by saying, “I am making a toast to my elder brother Ben.” This sentence hinted at pseudo-kinship relations because calling people an “elder brother” is quite common in contemporary China; the term originally signifies kinship but currently has varied meanings. Ben (2023) picked up the hint and affirmed that he accepted this pseudo-kinship relation with Ao by explicitly saying, “I treat you as my brother.” Ao (2023) further affirmed the pseudo-kinship by saying, “I will also treat your son as my nephew.” Ao was fully conscious of this ritual; Ao (2023) explained it in the post-banquet interview.

When people start to call each other brothers at a banquet, it represents the promise of long-term relations in which they will keep exchanging favors with one another as if they are real family members.

Besides establishing a pseudo-kin tie, the toasting ritual generated emotional energy. Ao (2023) kept saying phrases such as “I’m very happy today” and “I am so glad.” Ben (2023) reciprocated by saying, “I’m glad that she has improved academically.” All guests smiled and applauded the forging of the new tie. Collective drinking – which is a key part of the toasting ritual – further helped generate emotions through body language and corporeal experiences. Although this toasting ritual successfully generated emotional energy, it was relatively top-down oriented and contrived. It represented what Wang et al. (2025) call the “planned economy mode” of toasting, in which a regimented script guided participants’ actions. To some extent, it resembled a collective performance more than organic sociality and heartfelt conviviality, though it was not as contrived and top-down driven as what Mao et al. (2021) call “coercive toasting.”

The favour-giver was also rewarded with face (面子, mianzi) and status. It was created through the public witnessing and acknowledgement of the favour-giver as the senior partner in newly initiated guanxi. These guanxi work processes enhanced Ben's social status and position in relevant social networks. The emotional energy generated through toasting rituals reinforced Ben's image as a kind and powerful favour-giver. Secondarily, the gaining of such face and status served as a small symbolic pay-back for the favour-giver.

It is difficult to measure how much the pseudo-kin declaration and toasting were truthfully driven by Ao's sentiments. The less truthful they are, the more the guanxi work processes of this banquet may be interpreted as obfuscatory relational work. In obfuscatory relational work, the favour-giver and favour-receiver try to dress up morally ambiguous transactions between themselves as sentimental and/or moral undertakings (Rossman, 2014). Because the guanxi between Ao and Ben did not involve obviously immoral transactions such as corruption or criminal activity, this case did not fully qualify as obfuscatory relational work. Nonetheless, their transaction still involved slight moral ambiguity because private tutorship was banned by the government. The less truthful and the more contrived a banquet's guanxi work practices are, the more this banquet can be interpreted as instrumentally driven. As reflected in this banquet's guanxi work processes, Ao and Ben did not even try to pretend that instrumental exchange was unintended.

Successful Guanxi Work Processes That Emphasise Gifting and Associated Social Skills

This section examines a case that features successful guanxi work processes and the second style of doing guanxi work. Chen was a high-level manager of a large private enterprise. She organised a banquet at a high-end restaurant and invited some of her subordinates, including Wu, to celebrate the Lunar New Year. At the banquet, the participants initiated small talks on various everyday life topics. Wu said that she and her husband were looking for an interior designer for their new apartment. Chen offered help by introducing them to one at a prestigious design studio and getting them a 60 per cent discount. At the end of the banquet, Wu insisted on paying the bill. Chen and others would not let her do this. Wu (2023) responded: “I hardly had the chance to treat you folks to a meal before. So do me a favor; please give me the chance to pay this time.”

After the banquet, Wu contacted the designer introduced by Chen and contracted him to work on Wu's apartment. Three weeks later, Wu invited Chen to a banquet at a high-end restaurant. Wu also invited some of their common colleagues. During the banquet, she brought out a small package of tea and gave it to Chen. Chen declined. If the price of the tea was very high, it would have functioned as a gift that consummated the guanxi relations between Wu and Chen. Chen (2023) said to Wu, “You don’t have to do that, my friend; we have very good guanxi.” Wu did not explicitly announce that the tea was a gift given in return for the favour received. It remained unclear: it could be a gift of equal worth that would terminate guanxi, or it could be a small gift to facilitate the further development of guanxi. After Chen declined it, Wu (2023) reframed her gifting in the following way.

I remember that a long time ago, you said that you love pu’er (普洱, a Chinese tea) tea. I bought some for myself and tried it. However, I could not fall asleep after drinking it. So, I stopped. Now, do me a favor. Take it and finish the tea. It is quite expensive stuff; I don’t want to waste it.

After Chen heard this, she accepted the gift. In the post-banquet interview, Chen (2023) explained that “Wu was trying to do me a favor […]. She (2023) cleverly framed her gift-giving as favor-seeking. I had no choice but to accept it.” In the post-banquet interview, Wu explained that she reframed her gift-giving as favour-seeking to emphasise the non-instrumentality of their guanxi. Firstly, she minimised the monetary value of the gift by downplaying it as a leftover portion of what she bought for herself. Secondly, she framed this pay-back in the name of asking for a favour.

Incidentally, favour-receivers’ framing of pay-back in the name of asking for a favour is quite a common practice in contemporary China. For instance, in one of our banquet cases, a favour-receiver named Xiaomeng (2022) gave a basket of fruits to the favour-giver and said:

These fruits were freshly brought by my mom from my home village. But the amount was too large – Please help me consume them. Otherwise, my mom will scold me.

These details of banquet participants’ efforts document some social skills associated with non-reciprocal relational gifting. In banquets of the second pattern, central practices of guanxi work are based on these social skills. In banquets of the first pattern, social skills are less crucial due to the first pattern's reliance on regimented scripts and rituals. The first social skill observed in these cases was what Barbalet (2025) calls “the ritual of gift refusal.” This ritual helps the favour-receiver and the favour-giver to affirm to each other and other guests the sentimentality and non-instrumentality of their guanxi. We consider this a relatively regimented part of the second pattern of guanxi work processes. However, the social skills required to perform it are still quite demanding and subtle.

The second social skill observed in these cases was meticulousness in gift selection. Wu (2023) subtly revealed this when she said: “Remember that a long time ago, you said that you love pu’er tea.” She hinted that she had paid long-term attention to Chen's taste and well-being. Wu (2023) made known her considerable expenditure when she said that the gift was “quite expensive stuff.” Meanwhile, the gift's (i.e. pu’er tea) monetary value was still much lower than the savings that Wu earned through Chen's connection to the design studio. Hence, the non-reciprocal nature of the gifting was kept, which in turn encouraged the forming of reciprocal care obligations (or creditor–debtor relations in conventional theoretical terms).

The third social skill observed in these cases concerns the forming of reciprocal care obligations without relying on the highly ritualistic (i.e. planned economy mode of) toasting and formal declarations of friendship or pseudo-kin. In banquets of the first pattern, reciprocal care obligations are explicitly and unequivocally declared. Such declarations serve as a signpost for how to orient guanxi in the future. The ritualistic toasting confirms the role of third-party witnesses and guarantors of the newly formed obligations. An outburst of emotional energy is generated with the rituals and the presence of third-party guests; this emotional outburst helps establish the obligations.

In contrast, banquets of the second pattern – including the cases of Wu and Xiaomeng – do not feature pseudo-kin declarations and ritualistic toasting. Instead, we found uncontrived declarations of friendship, spontaneous and bottom-up driven toasting (i.e. in the market economy mode), and gradual generation of emotional energy during the banquet and post-banquet. For example, when Chen (2023) told Wu, “My friend, we have very good guanxi,” friendship is adroitly announced within a context of sociality and conviviality. Emotional energy is not generated in the form of a short affective burst triggered by formulaic rituals.

Based on post-banquet interviews, we found that the second pattern is preferred by some informants for its truthful sentiments. They expressed reservations about pseudo-kin because it looks rather “insincere,” “provincial” (土, tu), or “cheap” (廉價, lianjia). Wu thought that building pseudo-kin ties with Chen was inappropriate. Other informants are “too embarrassed” to rely on pseudo-kin ties for forging care obligations. They believed that because pseudo-kin should involve a substantial investment of sentiments and moral commitment, one should not lightly declare it. They thought that solid obligations should be based on sustained efforts of gifting in the medium-term or long-term period. In short, the second pattern looked less like obfuscatory relational work than the first pattern.

Guanxi Work Processes That are Moderately Shaped by Contingencies

This section analyses a banquet case with guanxi work processes that are moderately shaped by contingencies. Dai owns a food supply business that delivers food to restaurants in Chongqing. He was forty-five years old and solidly belonged to the Chinese middle class. When his delivery system was temporarily disrupted, he sought help from a truck driver, Xia, to deliver the goods. Xia was forty-seven years old and belonged to the lower-middle class. Xia did Dai a favour by delivering the goods. A few days after the delivery, Dai offered Xia a meal at a working-class eatery and invited some common friends and workmates. At the banquet, Dai (2023) took out two cartons of high-end cigarettes and said to Xia, “You have done a favor for me, and I truly appreciate it. Please accept the cigarette as a gift.” After Xia received the gift, Dai (2023) made a toast to Xia: “I dedicate this toast, the cigarettes, and this banquet to you! Thank you again for doing me a favor!”

Xia helped Dai by doing deliveries beyond his job responsibilities. To pay back Xia, Dai paid the bill for the banquet and gifted Xia two cartons of cigarettes. The amount paid by Dai was similar to the amount spent by Xia on giving Dai the favour of delivery. Dai strategically planned his gifting this way. Dai (2023) said this in the post-banquet interview.

I calculated the price of fuel that Xia had to pay for doing my deliveries. It was not a negligible amount. Therefore, I bought cigarettes of similar value to return the favor.

An additional guanxi work practice that occurred at the banquet was Dai's declaration of gratitude through a ritualistic toast. The toast stressed that the instrumental exchange between him and Xia had been consummated. Dai's gifting and toasting blocked the evolution of instrumental exchange into care obligations and sentiment-based relations. Dai (2023) reflected on his guanxi work choice.

There's no need for us to deepen our relationship. So, I invited him to the banquet and sent him a gift. I don’t want to owe him anything. That's why I brought others to the meal and announced during the toast that I had already paid for my debt.

Xia (2023) perfectly understood the practical meaning of this choice.

I became aware at that moment [of toasting] that Dai didn't want to build a deeper relationship with me. I don’t think we will have future contact except when Dai needs a similar favor. He looked grateful to me during the banquet. But it's only a part of the social etiquette; it's not personal.

This case can enrich our understanding of contingencies in guanxi work processes and outcomes. Based on our contextual knowledge of Dai and Xia, we thought that they could have potentially developed long-term guanxi. Dai could have taken up the role of being a long-term favour-giver, and Xia could have been enrolled as a junior member of Dai's social network. Xia admitted in the post-banquet interview that he wished to be enrolled in Dai's guanxi network. Like Liang's banquet case, the strategic incongruence between the favour-receiver and the favour-giver constitutes a major contingency that affects the guanxi outcome. Dai's guanxi objective succeeded when the guanxi between Dai and Xia did not manage to develop. Xia's guanxi objective failed.

A more detailed and adequate interpretation of this banquet's guanxi outcome can be developed. There are at least three potential routes for guanxi to develop post-banquet: termination, weakening, and temporary suspension. The first is Dai's objective: the termination of all subsequent relations between Dai and Xia. This is the conventional understanding of symmetric guanxi phenomena in banquets (Bian and He, 2022; Lin, 2001; Yang, 1994). The second route is the sustenance of a weak tie that will erode through time. For example, as Xia said in the interview, he might inform his working-class acquaintances about this banquet to show off his guanxi to a higher-status person. The third route will materialise if the social exchange between Dai and Xia eventually develops into stronger guanxi relations under suitable circumstances in the future. This could happen because Dai might seek additional help from Xia in the future. These three potential routes are not mutually exclusive – they coexist because guanxi development is processual and contingent. Guanxi work processes that do not aim to develop guanxi would still open up these three routes of post-banquet evolution.

Guanxi Work Processes and Outcomes That are Considerably Shaped by Contingency

This section examines a case with guanxi work processes and outcomes that are considerably shaped by contingency. Deng and her husband were the favour-receivers. They organised a banquet for Zhang, the favour-giver. Both were solidly middle-class. Zhang had given the Deng couple the favour of taking Deng's disabled mother to the hospital every month for seven consecutive months. Because Deng and her husband worked full-time and hospital appointments were during work hours on weekdays, they needed someone to help bring Deng's mother to the hospital and take care of her during these hospital visits. Zhang, a retiree and a neighbour of Deng's mother, already knew Deng's mother and volunteered to help.

The banquet took place at an upper-mid-range restaurant. The Deng couple, Zhang, Deng's son, another neighbour of Deng's mother, and one of the authors, was present. Guanxi work practices in this banquet were limited. There was one round of spontaneous toasting and some words of gratitude expressed by Deng, but no gifting, declaration of friendship, or other rituals. The bill, amounting to 1500 yuan, was paid by Deng. A small amount of emotional energy was generated, but the general social atmosphere was underwhelming throughout the banquet. This was unsurprising given that the Deng couple and Zhang did not aim to further deepen their guanxi.

We highlight some information that can be learned from this case. The first is that this case puts into relief the wide variety of objectives of different banquet guanxi work processes. This case's favour-receiver, Deng (2022), did not aim to terminate guanxi, but wanted to “keep guanxi with Zhang constant.” Deng still needed Zhang's help with her mother's hospital visits after the banquet. Meanwhile, Zhang was similarly uninterested in deepening guanxi with Deng, but was still willing to help Deng's mother. He was being a nice neighbour in offering such help. Besides developing guanxi and limiting the development of guanxi, keeping guanxi constant is a possible objective of banquet guanxi work.

The second piece of information concerns the different patterns of variations of different guanxi work processes. We previously identified two patterns in guanxi work processes that aim to deepen guanxi. However, based on what we observed in banquet cases that did not try to deepen guanxi, we could not identify any patterned variations. Deng paid the bill and informally said “thank you,” but gifting and ritualistic toasting did not occur. Despite this, guanxi processes and outcomes during the banquet did not go astray. The strategic aim of the favour-receiver was still achieved. In yet another banquet case of ours, a favour-receiver who aimed to terminate guanxi gave a gift but skipped the expected practice of ritualistic toasting. Despite this, the strategic aim of the favour-receiver was still achieved. Based on these observations, we preliminarily think that guanxi work processes that do not aim to deepen guanxi are less demanding, less regimented, and, hence, less likely to develop into specific patterns. Even without meticulous design and execution, such guanxi work can proceed and achieve the expected outcome.

The third piece of information revealed by this banquet is the compulsory and involuntary nature of banquets that are mainly organised for guanxi work. We noticed Zhang's modest resistance to spending the time and social efforts to attend this banquet. Deng had already invited Zhang to a banquet after Zhang helped Deng's mother twice. Zhang postponed it as long as possible and eventually accepted the invitation a few months after the first invitation. Zhang (2022) did not mind receiving gifts and words of gratitude from Deng, but he “preferred not receiving them if he had to attend a banquet” for them. After consecutive rounds of invitations from Deng, Zhang felt that he “might appear impolite if he kept refusing.” Neither was Deng eager about the banquet. Deng (2022) did it “out of the social expectation to show gratitude” and she “feared that she would look indecent for relegating family caring responsibilities to a non-kin person.” To keep her decent image, she felt compelled to keep a degree of guanxi with Zhang and seriously thank Zhang. In short, both the favour-receiver and the favour-giver preferred not to perform banquet guanxi work but were still socially pressured to do it. In this light, this banquet can be described as unwanted by both participants. They felt that it was involuntary and laborious rather than enjoyable and/or instrumentally productive. Current studies have already found that some Chinese rural individuals are socially pressured to attend large-scale rural banquets (Kipnis, 1997; Oxfeld, 2017; Stafford, 2006; Yan, 1996). The case of Deng and Zhang shows how people are socially pressured to attend small-scale middle-class urban banquets.

The fourth piece of information concerns contingency. We interpret this case as an example of a banquet that is considerably shaped by contingency. Contingency was involved in whether this banquet could take place. Neither party was eager to have this banquet. The perceptions of social norms and etiquette by Deng and Zhang eventually led to the banquet's realisation. However, other social actors in similar situations of social exchange, or even the same actors in the future, may well choose not to organise a banquet. In our interpretation, the banquet guanxi outcome (i.e. keeping guanxi constant) was also quite contingent. If Zhang and/or Deng had invested more emotions into the banquet guanxi work, or if they happened to find some common personal interests during the banquet, the banquet outcome could well be a modest deepening of guanxi. Random contingent factors, such as whether the banquet participants are in a good mood on the banquet day, also seem to matter.

Guanxi Work Processes That Fail: Fallibility During the Banquet and After the Banquet

This section analyses two cases in which guanxi work processes failed. One of them failed during the banquet, and the other failed after the banquet. The first banquet involved Mou and Liang, who worked in the same private enterprise. Mou is a middle-level manager, a veteran at the company, and a local resident of the city. Liang, a young and non-local person, has recently entered the company. Liang received numerous minor favours from Mou at work, including information and institutional support. Liang set up a banquet at a mid-range restaurant; he invited Mou and several other co-workers. He utilised both gifts and rituals to carry out his pay-back. He aimed for the outcome of developing deeper guanxi relations with Mou.

During the banquet, Liang presented Mou with a gift of local speciality wine from Liang's hometown. Although this gift cost over a 1,000 yuan, it did not exceed the value of the help received from Mou. Hence, this gifting was an attempt to yield creditor–debtor relations and reciprocal care obligations. However, Mou (2023) politely but firmly rejected it.

Even if I don’t accept this gift, we are still great co-workers. If I accept this gift, how would others view me? […] They would blame me for trying to collect protection fees from new workers.

Later on, Liang performed ritualistic toasting in the planned economy mode and declared that Mou was a great and noble person. Liang also motivated other banquet participants to affirm his praise. This would have potentially increased Mou's reputation and status among co-workers. But again, Mou tactfully rejected this symbolic pay-back. Mou (2023) said:

I dare not accept this grand toast. I don’t deserve it; I’m not a very noble person. I have simply done what a co-worker ought to do. This toast should not be only for me. Let's instead toast to all of us, colleagues who strive to make our company better.

Immediately after this, Liang made the third attempt at this banquet to develop guanxi. Liang (2023) thanked Mou for helping him in the past few months and said, “Mou has treated me as an elder brother would.” Mou (2023) negatively responded to Liang's initiation of a pseudo-kin tie.

Please don’t say that. I helped you in a similar way that I helped other new colleagues. Treating me as a helpful colleague would be fine. I truly appreciate your sentiments.

The fallibility of banquet guanxi processes is vividly illustrated in this banquet case. The favour-receiver's guanxi objective has failed. The favour-giver rejected the favour-receiver's declaration of pseudo-kin, forewent the non-reciprocal gift, undermined the ritualistic toasting, and refused to develop care obligations. Although Liang and Mou did not turn into enemies after the banquet, their guanxi at the end of the banquet may be interpreted as suspended and slightly weakened.

In the post-banquet interview, Mou (2023), the favour-giver, told us that he was “somewhat surprised” by the guanxi work at this banquet; Mou “never thought that Liang would so ‘brazenly push the envelope’ (得寸進尺, decun jinchi).” Mou meant that Liang was demanding too much when Liang tried to turn the co-worker relations between Liang and Mou into a pseudo-brotherhood. Mou thought that the banquet should instead be a simple expression of gratitude. Liang (2023), the favour-receiver, also told us that he found the favour-giver's reactions to be “entirely unexpected.” Although he was not absolutely confident that Mou would accept his offer for deeper guanxi, he thought it had a realistic chance of working out.

While Liang's banquet case illustrates fallibility and contingency during the banquet, this section's second banquet case illustrates post-banquet fallibility and contingency. The main actors of this banquet were Wen and Guo. Wen, the favour-receiver at this banquet, made a pseudo-kin declaration. Guo, a wealthy favour-giver, enthusiastically accepted it. A ritualistic toast was smoothly performed with several third-party guests. At the end of this banquet, the guanxi work appeared to us to be entirely successful. However, subsequently, we found through a follow-up interview with Guo that he refused to reply to Wen even though Wen sought favours from him half a year after the banquet. It was because two months after the banquet, Guo decided to leave China and migrate to the US due to China's deteriorating economy. Guo (2023) urgently cultivated new relations and networks, realising that his “old guanxi ties in China would soon become useless” for his family and future business plans.

The banquet case of Guo and Wen supplies two pieces of theoretically relevant information. The first is that it illustrates that an accurate assessment of banquet guanxi work outcome needs to take into consideration post-banquet guanxi development. This is one of the reasons we and theorists such as Barbalet (2021: 168) recommend a long-term processual conception of banquet guanxi. Successful guanxi work during the banquet certainly helps shape the guanxi outcome, but post-banquet action and contexts also contribute much to the eventual guanxi outcome. With the benefit of hindsight, we may say that although the banquet guanxi work of Wen and Guo did not fail during the banquet, it was not sufficiently successful in the sense that their banquet guanxi work did not manage to forge a lasting post-banquet guanxi relation.

The second piece of information concerns the obfuscatory nature of the first style of doing banquet guanxi work. Guo and Wen vowed to become brothers at the banquet, but turned into strangers two months later simply because the national business environment deteriorated. This change was not even triggered by any quarrels or conflicts of interest. This chain of events demonstrates how emotionally empty and contrived the pseudo-kin declarations at banquets are. Based on our interviews, we understand that individuals who are familiar with the contemporary Chinese banquet culture already know that such pseudo-kin declarations are often insincere.

Conclusion

The previous analysis yielded empirical findings and theoretical observations that would advance the research on guanxi in Chinese banquets. It showed the advantages of using the concept of guanxi work to collectively frame various guanxi styles, objectives, and practices. It illustrated the utility of a robust processual conception of banquet guanxi work. It showed that banquet guanxi work processes could be shaped – to different extents in different cases – by contingent factors and contexts. It also helped identify two styles of banquet guanxi work, each of which entails different guanxi work processes. Our analysis of several different banquet cases revealed that banquet guanxi work processes could greatly vary in their degrees of success.

Because this study focuses on uncovering the fallibility and contingency of banquet guanxi work processes, it has not contributed much to the substantial theorisation of such fallibility and contingency. We expect that such theorisation constitutes a promising research direction for future studies of Chinese banquets. For example, scholars may examine and categorise various contingent factors that often play a role in shaping banquet guanxi processes. Or they may analyse how banquet participants tactically navigate guanxi work processes to avoid pending failure. In our future research, we wish to investigate how micro-power conflicts between banquet participants operate as a source of contingency and fallibility.

We also recommend that future research devote more attention to different styles and patterns of banquet guanxi work. This study focuses on identifying two styles of doing guanxi work. Future studies can tackle the intriguing theoretical questions that follow from our discovery. For example, are there social or cultural reasons for banquet organisers to prefer one of the two styles over the other? As we reflected on our data, we vaguely felt that the banquet organisers’ level of education, degree of internationalisation, and gender might affect this preference. However, we hesitated to make any conclusions due to our small dataset.