The cooperation and reciprocity in mutual aid are crucial for the survival of species, contribute to the evolution of societies, and serve as fundamental organizing principles of human communities (38). Mutual aid refers to the formation of self-organized associations in which individuals work together to address shared health or social issues through reciprocal support (39). These organizations typically arise within community settings and consist of neighborhood members (40). While conventional understandings of mutual aid emphasize interpersonal reciprocity between individuals, recent scholarship highlights a crucial shift toward community-level cooperation (5, 29, 41). In crisis contexts such as the COVID-19 pandemic, mutual aid increasingly manifests as collective community action—where residents jointly mobilize resources and organize self-help activities to address shared risks and disruptions, transforming individual reciprocity into community-based resilience (5).

Community mutual aid can be understood as an organizational form within a community where residents provide assistance through essential provisions, such as food, and emotional support while coordinating to address shared challenges (42). This includes supporting marginalized groups and delivering community care during crises (7), the latter being the focus of this study. Disaster risk is defined as the potential for loss of life, injury, or damage to assets within a system, society, or community over a specific period (43). Community members experienced unprecedented feelings of fear, uncertainty, and confusion amid disaster risks, which in turn motivated mutual aid actions (44). Previous studies have highlighted the importance of community mutual aid in crisis response, including disasters such as Hurricanes Katrina, Sandy, Maria (40), as well as the 2005 London bombing (45). It supports local residents and strengthens community activities at the micro level (9).

During the COVID-19 pandemic, community mutual aid became a central form of community response and a means of strengthening social connections (29). In the United Kingdom, for example, more than 4,000 mutual aid groups were established during the early stages of the pandemic (46). These groups offered various types of support, including the procurement and delivery of essentials, prescription collection, meal provision, pet care, informational sharing, emotional support through helplines, and food bank management (47). By combining information exchange, mutual care, and the mobilization of local resources, these community initiatives collectively addressed residents’ everyday challenges (5). It is noteworthy that early mutual aid was often linked to anarchist perspectives (32); however, during the pandemic, such community-based mutual aid evolved beyond that context and displayed plural and pragmatic orientations—some were entirely grassroots, others partly supported by local governments or embedded within market infrastructures (2). This evolution illustrates that mutual aid has transcended ideological boundaries, becoming a flexible framework for public participation that integrates civic, state, and market logics (48).

Community mutual aid during the pandemic displayed several defining characteristics. First, spontaneity and grassroots participation were key features; when community members faced shared threats, mutual aid groups typically emerged from the bottom up rather than through administrative directives (40). Second, mutual aid groups often maintained flat organizational structures, operating flexibly through neighborhood networks and digital platforms such as WeChat and Facebook (30, 33). Third, mutual aid acted as a resource integrator, mobilizing internal community resources, such as food, medicine, labor, skills, and information, to achieve internal redistribution and relative self-sufficiency (5, 10). Fourth, mutual aid networks provided not only material assistance but also comprehensive care systems through emotional support and information sharing (29, 42, 49). Fifth, reciprocity and collaboration formed its operational logic, emphasizing “solidarity over charity” (50). Finally, emotional motivation was a key sustaining factor. Participants experienced joy through helping others, which reinforced community identity and sustained network operations (28, 29).

Social capital theory provides a crucial theoretical foundation for understanding community-based mutual aid. Coleman conceptualizes social capital as resources embedded within individuals and collectives that influence behavior through particular social structures, thereby generating benefits in specific contexts. Social capital is inherently productive, serving explicit instrumental purposes—individuals can deliberately mobilize it to achieve desired goals (24). Building on Coleman’s insights, Putnam defines social capital as features of social organization—such as networks, norms, and trust—that facilitate coordinated action and cooperation for mutual benefit (25). As a driver of civic engagement, social capital encourages participation in voluntary organizations, which in turn fosters norms of reciprocity and interpersonal trust, ultimately sustaining collective well-being (51).

Banks extends this framework to the domain of mutual aid, arguing that in contexts characterized by individualism and weakened social ties, mutual aid groups maintain participation primarily through their internal social capital (26). Drawing upon Coleman and Putnam’s perspectives, he distinguishes two interrelated dimensions of social capital. The first, relational social capital, is characterized by networks grounded in shared norms and trust that enable individuals to overcome collective action dilemmas such as the “tragedy of the commons” and to pursue shared interests. The second, structural social capital, concerns the configuration and density of social networks through which individuals gain influence and access to resources via exchanges. Together, these two dimensions form the foundation of community mutual aid.

Empirical research further demonstrates that mutual aid practices during the COVID-19 pandemic operated through pre-existing structural networks within communities and relied heavily on relational social capital—particularly trust and reciprocity (30). Simultaneously, the mutual aid process contributes to the reproduction and reinforcement of these forms of social capital: members expand their personal support networks through assistance exchanges (52), while the mechanism of “healing through helping” strengthens trust and reciprocity within the community (26).

CGB is a novel retailing model that combines online e-commerce with offline community retailing, involves many stakeholders (e.g., group leaders, consumers, and retailers) (53) and is supported by CGB platforms (e.g., WeChat, Meituan, and Pinduoduo) (54). Notably, Meituan and Pinduoduo are central to China’s CGB ecosystem, operating through sub-brands such as Meituan Youxuan and Duoduo Maicai. These platforms provide group leaders with standardized product catalogs, logistics support, and integrated payment systems (55). By leveraging technology and capital, they offer advantages in supply chain integration and service standardization, thereby becoming deeply embedded in the everyday consumption practices of local communities (18).

However, compared to these large-scale platforms, WeChat plays a more fundamental and ubiquitous role in the CGB ecosystem (56). On the one hand, most CGB platforms offer mini-program interfaces within WeChat, enabling users to browse products, place orders, and complete payments without the need to install additional apps. On the other hand, in areas with limited or no platform involvement, many CGB activities are organized entirely through WeChat groups, where group leaders manually manage product listings, order collection, and delivery coordination (53).

This self-organized operational form within the WeChat ecosystem reflects a higher degree of grassroots participation, flexibility, and social embeddedness (57). Research shows that CGB in China has effectively leveraged social tools such as WeChat group chats and mini-programs to cultivate strong neighborhood ties, thereby accelerating the growth of the community-based digital economy (58).

To further illustrate the operational mechanism of CGB within WeChat-based networks, Figure 1 presents a typical workflow from supply chain establishment to after-sales service management. The process involves multiple actors—including digital platforms, group leaders, and community members—who collaboratively coordinate procurement, distribution, and communication through WeChat groups or mini-program tools (37).

As a promising retail platform model, the sales characteristics of CGB are closely related to consumer psychology (36). Consumers are driven by factors such as perceived value, including convenience and time savings (49); perceived trust in the quality of services offered by the group leader (59); online platforms (36); and good distribution models (60). Importantly, Zhang and Tsai find that collectivist consumer psychology significantly influences purchase decisions in CGB (61). Studies have also scrutinized the impact of digital platforms and highlighted how their features (e.g., convenience, usefulness, and ease of use) influence consumer attitudes and purchase intentions (53, 54, 62). The dual role of group leaders, bridging community ties and commercial transactions, crucially affects perceived trust and consumer behavior (37, 59).

During the COVID-19 pandemic, CGB surged in popularity, particularly in cities such as Wuhan and Shanghai, where it was deployed to address emergency needs and ensure community welfare (63). This process typically involves establishing online platforms, setting community norms, appointing group leaders, and optimizing logistical procedures (19). Stakeholders, including group leaders, residents, estate managers, and volunteers, collaborate to create supply chains, organize food distribution, and enhance community resilience (63). Participation motives are driven by both individual interests and communal benefits, thus fostering community cohesion and increasing local governance engagement (19).

Social media groups are integral components of platforms such as Facebook, Flickr, and WeChat (64), which form virtual communities where individuals share interests and networks linked with family members, friends, or colleagues (65). These groups typically consist of group owners (in charge of group management) and members (who join through open acceptance, authorization, or invitation) (66).

During crises, social media groups provide three types of social assistance. Instrumental support includes concrete aid and services that help restore communal life in affected areas by mobilizing resources and services within the group (67). Second, informational support enables disaster-affected individuals to share information, ask questions, and solicit advice, thus facilitating knowledge exchange and situational awareness (67). Finally, emotional support allows members to express empathy, affection, confidence, and compassion (68). For example, during the Camp Fire disaster, individuals in a Facebook group shared uplifting quotes and focused on mental health, recovery, personal growth, and appreciation (68). During the COVID-19 pandemic, these social media communities broke through time and space restrictions and enabled individuals to chat online (69). Community managers (e.g., neighborhood committees and grid administrations) have specific responsibilities in each community. They disseminate information related to community-based preventive measures to these groups (70). It is important to note that during this period, the circulation of Mis/Dis/harmful Information (MDHI), as emphasized in the RCRC framework, posed significant challenges. Therefore, community managers not only provided preventive information but also played a role in countering misinformation and maintaining trust within the community (71). Additionally, these groups serve as platforms for residents to seek help, acquire necessities, engage in public affairs, and alleviate loneliness caused by isolation (72). Research has indicated that participation in these groups can enhance residents’ trust and attachment to their community, encouraging greater involvement in community activities (69). In summary, residents’ WeChat groups became essential tools for community outreach during the pandemic, connecting community networks, maintaining operations, and shaping community dynamics (72).

As COVID-19 rapidly spread worldwide, many countries adopted lockdown measures in accordance with the WHO guidelines to curb virus transmission (75). In April 2022, the government of Jinan, China, implemented community-based lockdown measures in response to the local COVID-19 outbreak. In this context, two notable cases of CGB emerged in Community W. Community W is a pseudonym used to protect participants’ privacy. It is also the residence of the first author, which facilitated immersive observation and direct interaction with community members, enabling the collection of rich, contextually informed data. To address potential biases associated with insider research, we implemented rigorous qualitative procedures, including systematic data collection, coding, and participant validation, to ensure the credibility and reliability of our findings.

On April 24, 2022, a resident (with the No. L01, Table 1) created a WeChat group called the FeiFei Group for CGB, which included 498 members, all residing in Community W. Five days later, another resident (No. L02) established a similar WeChat group known as the AA Group, which comprised 410 members from the same neighborhood. Both “FeiFei” and “AA” are pseudonyms used to represent these two CGB WeChat groups.

These two cases were purposefully selected for the study for the following reasons: First, they represent predominant CGB forms emerging in Chinese cities during COVID-19. Model 1, which is characterized as “small and scattered” and initiated by residents (57), is exemplified by the Feifei Group. This model leverages group leaders’ personal networks to establish independent supply chains involving suburban producers and small wholesalers, primarily distributing fresh produce. Despite flexibility and cost advantages, this model faces significant coordination challenges. In contrast, the AA Group represents Model 2, which operates through commercial platforms such as Meituan (76). This model ensures a stable supply of diverse goods, including daily necessities, through centralized operations, but it may limit community decision-making autonomy. These contrasting organizational logics provide a theoretical basis for analyzing the adaptability and limitations of CGB models in crisis contexts.

Second, while large platforms such as Meituan and Pinduoduo constitute major actors in China’s CGB ecosystem, our study focuses on CGB practices led by local group leaders and coordinated through WeChat social networks (19). This focus allows us to investigate the micro-level interaction mechanisms, mobilization strategies, and the reproduction of community engagement logics within the group buying process, rather than the centralized and commercialized operations typical of platform economies. Both selected cases exemplify this form of grassroots, WeChat-based CGB, characterized by spontaneity, flexibility, and close-knit community engagement, which directly aligns with the objectives of our study.

Third, the selection of Feifei Group and AA Group as case studies aligns with the abductive logic adopted in this study. Abductive reasoning emphasizes starting from observed phenomena and inferring the most plausible causes or mechanisms to generate new theoretical insights (77). Field observations in Community W revealed that, despite their independent operational models and organizational structures, these two groups were relatively large, representative, and exhibited a notable similarity in mutual aid patterns. This unexpected phenomenon serves as the starting point for abduction: Why did CGB spontaneously display a consistent logic of mutual aid during the pandemic? Guided by social capital theory, we conducted a comparative analysis of organizational structures, decision-making processes, and member interactions, focusing on the common factors that facilitated mutual aid behaviors. Through a systematic examination of similarities and differences, we abductively inferred the core mechanisms driving mutual aid practices in CGB and constructed a practice-rooted, generalizable theoretical framework.

After the formation of the two WeChat groups, we promptly joined both and obtained permission from the group leaders to issue a recruitment announcement. This announcement outlined two criteria for participation: residing in the neighborhood and involvement in at least one CGB activity. Interested individuals were offered a modest compensation as an incentive for participation, but not contingent on providing any specific responses. To minimize the risk of participants joining solely for the compensation, we verified eligibility by confirming prior participation in CGB activities and residence within the community.

As shown in Table 1, 12 participants were voluntarily recruited, including 2 group leaders and 10 members (7 women and 5 men), aged between 22 and 52 years (M = 37.92, SD = 8.35). Additionally, 70% of the participants engaged in 10 or more group purchases. To ensure anonymity, informed consent was obtained from all participants, who were also informed of the research purpose.

From April 24 to November 30, 2022, ethnographic data were collected through immersive participant observation within a CGB WeChat group. Before commencing ethnographic observation, informed consent was obtained from both the group leaders and group members. The group leaders provided verbal consent, and no members expressed objections to the study. All chat records, images, and interaction data were collected solely for academic research, with all personal identifiers removed to ensure participants’ confidentiality. The researcher actively participated in daily interactions, discussions, and group-purchasing activities, documenting digital communication including text, images, videos, voice messages, emojis, and hyperlinks. To enhance reflexivity and minimize researcher bias, reflective field notes were maintained throughout the study, recording personal perceptions and potential influences on data interpretation.

The two CGB groups studied were established during the pandemic; thus, direct ethnographic observation was only possible during this period. After May 15, 2022, when strict community lockdowns in Jinan were largely lifted, the context shifted to the post-pandemic stage. To reconstruct a time-bound comparison across stages, pre-pandemic practices were inferred from participants’ retrospective accounts, supplemented with literature and media reports illustrating general CGB operations before the crisis. Post-pandemic observations and archival data (e.g., WeChat group histories) captured the persistence or decline of mutual aid practices and the reversion to commercial functions. This triangulated approach provided a comprehensive understanding of CGB mechanisms and social dynamics across the pre-pandemic, pandemic, and post-pandemic stages, highlighting their functional evolution.

Therein, we conducted semi-structured interviews with the participants. The interviews lasted between 30 and 60 min, with an average of 48 min. They were conducted both online and offline, in compliance with social distancing guidelines. Online interviews were conducted using WeChat’s voice communication features, whereas offline interviews were primarily conducted during the pickup of CGB merchandise. Following the interview outline, we posed a series of semi-structured questions, as shown in Appendix 1, covering topics such as buyers’ motives, experiences, tools, and perspectives on CGB.

Since all interviews were conducted in Chinese while the manuscript is written in English, we provide an explanation of the translation process. Data analysis was conducted based on the original Chinese transcripts to avoid semantic bias (78). During the writing stage, the first author translated the relevant excerpts into English, and the corresponding author performed back-translation (79). Both authors are bilingual. When encountering expressions with multiple possible translations or culture-specific meanings, we referred to relevant literature, discussed until consensus was reached, and developed a consistent glossary (e.g., translating “团长” as “group leader” and “接龙” as “group note”). To ensure that the translations accurately reflected participants’ intended meanings, we also conducted member checking with selected participants (80).

We adopted an abductive analytical perspective to code and interpret the interview data (77). This approach integrates the systematic, theory-informed procedures of grounded theory—open coding, axial coding, and selective coding—with abductive reasoning as its core logic (81). Abductive reasoning allows iterative movement between empirical data and theoretical frameworks, continuously comparing, refining, and optimizing interpretations to identify the most plausible explanations for observed phenomena (77). This approach is particularly suitable for our study, as it enables the use of social capital theory as a sensitizing concept rather than a predetermined framework, providing heuristic guidance without constraining theoretical development (82). All interview recordings were fully transcribed and imported into NVivo 14 software for a systematic three-stage coding process.

The coding process followed a systematic three-stage procedure. During open coding, we read transcripts line by line, identifying and labeling initial concepts while writing analytical memos to capture emerging insights. For example, the statement, “I think it’s because we are all neighbors, and we trust information that comes from our neighbors (M08 group member),” was coded as “neighborhood relations.” In the subsequent axial coding stage, relationships among concepts were examined and organized hierarchically into sub-categories and main categories. For instance, “neighborhood relationship” was grouped under the sub-category “networks of acquaintance,” which, guided by social capital theory, was further categorized under “community networks.” During this stage, we employed abductive theory matching, iteratively comparing data with theory to explore underlying mechanisms. During the selective coding phase, drawing on the conceptualization of social networks within social capital theory, “community networks” were identified as part of the core category “the causes of community mutual aid,” reflecting their role in facilitating collective assistance. Other main categories were integrated into two additional core dimensions: “the process of community mutual aid” and “the outcomes of community mutual aid.” Together, these core categories formed a theoretical framework explaining the mechanisms of community mutual aid in CGB.

Throughout the analysis, we maintained theoretical openness. For example, the statement, “I was able and felt obligated to help my neighbors during a period of panic when the supermarkets were empty (L01 group leader),” was initially coded as “responsibility” and later refined into the sub-category “sense of rescue.” Although this psychological motivation extends beyond conventional social capital theory, it was incorporated into the broader category “psychological motivations,” alongside reciprocity, reflecting underlying emotional and moral drivers of community mutual aid.

Two researchers conducted independent coding and met regularly to discuss discrepancies, which were resolved through constant reference to the original data (83). All participants were invited to review the research findings. Most confirmed the accuracy of the coding and agreed with our interpretations. For divergent opinions, we revisited the original transcripts and refined coding categories or their relationships where necessary, while retaining these differing perspectives to capture the diversity of experiences and deepen our understanding of mutual aid mechanisms (84).

The COVID-19 pandemic posed a significant risk to individuals’ health and well-being (75). To curb the spread of COVID-19, governments implemented emergency lockdown measures according to the WHO guidelines. Some negative impacts on communities and individuals occurred, such as the temporary disruption of market activities and cessation of express delivery services. Consequently, residents faced multiple disaster risks, such as the scarcity of essential resources and physical isolation.

Uncertainty is an inherent aspect of disaster risk that cannot be ignored (85). Previous research indicated that uncertainty surrounding the COVID-19 pandemic may lead to persistent and intense fear (86). Additionally, the lockdown intensified feelings of loneliness due to heightened uncertainty (87). Our findings showed that residents experienced significant fear and loneliness during the lockdown period.

Residents form acquaintance networks on WeChat groups based on their existing relationships. These networks also serve as hubs of social support, where residents can offer assistance to one another. CGB is often facilitated through WeChat groups, which can include anywhere from 3 to 500 members (88). Membership in these WeChat groups requires invitations from current members and thus relies on personal connections between new and existing members. This network of acquaintances on WeChat plays a crucial role in facilitating CGB.

This network is characterized by a lack of clear hierarchical structures and a significant degree of equality among its members. Group leaders, who are often the most influential individuals within the organization, act more as initiators than managers, and do not hold prominent positions within the community network. Group members voluntarily follow rules such as refraining from spreading rumors or mistreating others, rather than being forced into compliance. This structure emphasizes unity and equal status among members, which is consistent with other models of community mutual aid.

In China, the social structure is sometimes described as an “acquaintance society” where interpersonal connections are primarily local with minimal movement among social groups. However, modernization has weakened these acquaintance relationships in urban communities (89). During a shared crisis, the emotions and capacity to gather resources among neighbors within a social network of acquaintances are revived. The Chinese proverb “A close neighbor is preferable to a distant cousin” underscores the reliability of neighborhood support during emergencies.

The network structure of social relationships significantly affects the functional content of social support (90). In CGB WeChat groups, the network of acquaintances among residents functions as a hub for social support, thereby facilitating community mutual aid. Four primary forms of social assistance emerged within this network.

The primary and crucial element is instrumental support, in which the group leader provides supplies to the residents, and the residents also provide supplies to each other.

Second, residents served as information nodes, gathering firsthand information about the pandemic from other networks and sending it to the AA and FeiFei groups. Thus, community networks provide informational support.

Third, the community network creates a secure environment for residents to express their emotions, which helps mitigate feelings of isolation and anxiety.

Finally, trust within a community network is crucial beyond typical forms of social support. Trust in group leaders forms the foundation for instrumental, informational, and emotional support, especially in commercial activities such as CGB (49).

Two main aspects of how residents participate in CGB exist: first, they engage in CGB through the utilization of low-cost technologies, such as mini-programs and group notes within WeChat groups, and second, they negotiate their emotions and interests during the group purchasing process.

Mini-programs are crucial technological tools for facilitating CGB. As components of WeChat, mini-programs can be utilized without the need to download or install them, which provides convenient accessibility by allowing users to open them effortlessly through swipes or searches (91).

The AA Group employs a mini-program called the “Vertical Order King.” The mini-program’s homepage displays essential details, such as the leader’s ID (L02), the offered products, their prices, the associated images, and pickup date. Users can easily add products to their shopping cart by clicking the “+” button and complete the checkout process by entering the recipient’s name, phone number, WeChat ID, pick-up point, and other necessary information.

A mini program enhances the shopping experience by offering various features. From the leader’s perspective, this includes product presentations, price labeling, and a user interface for completing purchases. From a consumer’s perspective, it provides a range of products to choose from, a one-click purchase button, and a streamlined checkout screen. These features streamline the purchasing experience and increase the likelihood of purchasing.

The FeiFei Group uses the “Group Note” feature to monitor statistical data of the group leader and members’ participation in group buying (Figure 3). The group leader disseminates product information by creating notes within the WeChat group. Users can engage by selecting the “#Group Note” button in the group chat, which allows them to track the sales data for various goods visually.

Integrating mini-programs that simulate the shopping experience or group notes that provide a clear view of collective purchasing, coupled with effective technology distribution, establishes a cost-effective mobilization method within the CGB. Moreover, within CGB, group leaders and members negotiate their respective interests. During the COVID-19 lockdown, group leaders often demonstrated altruistic intentions by sacrificing a portion of their earnings to assist others within the community.

L02, a mother of two, initiated a CGB business to leverage its flexibility and manage her time efficiently. Her original intention was to ensure the project sustainability by offering inexpensive prices. However, as the lockdown progressed, she aimed to utilize her resources to help her neighbors overcome challenges regardless of her earnings.

In turn, purchasers often perceive buying products from their neighbors as a means of supporting them. Mutual assistance reflects a community’s collective efforts to manage its interests and emotions.

This negotiation process enables the establishment of a consensus and subtle management of the interests and emotions of buyers and sellers in the CGB. For example, buyers commonly express satisfaction with purchased items through images and written remarks. To incentivize the group members to contribute valuable ideas, as a symbol of gratitude, the group leader distributed digital red packets to the WeChat group.

In CGB, residents derive psychological motivation from two main factors: a perceived sense of rescue and a strengthened sense of community. Upon receiving notifications about the lockdown, residents of Community W experienced a decrease in the availability of goods in nearby superstores and marketplaces. L01 organized a collective purchase of cucumbers from her suburban garden and sold them at a significantly lower price. When asked about her motivation for initiating CGB, she stated:

Group members who engaged in CGB perceived it as an effective approach to address the scarcity issue when “eggs and vegetables had been depleted and were not accessible.” (M10 group member). Group members understood that the group leaders would profit from these purchases, indicating a reciprocal relationship.

CGB has emerged as a pivotal catalyst in fostering a sense of community among neighbors. The detection of a COVID-19 case in a nearby community would trigger a lot of information (news, screenshots, audio clips, and videos) across CGB WeChat groups. While some residents initially felt alarmed by the sudden influx of information and confirmed cases, others took steps to alleviate the tensions. This period marked a significant shift, in which residents felt a profound connection to their community, emphasizing a sense of inclusion and membership.

For critical prevention moments, community officers allow some residents to leave their homes. These individuals subsequently organized transportation teams to facilitate the delivery of collectively purchased goods to residents’ homes. These acts of mutual aid not only fulfilled practical needs, but also instilled a deep sense of accomplishment among participants, further strengthening community bonds.

When CGB evolved into a mutual aid practice, it significantly enhanced the community’s resilience during the lockdown. First, it alleviated community challenges. On the one hand, the material shortages were mitigated within the community, particularly by ensuring the provision of food and other essential needs for residents. Organized efforts to source and distribute goods mean that, even in times of crisis, the basic needs of the community are met, thereby reducing stress and uncertainty. As one participant noted,

On the other hand, WeChat groups used for organizing these purchases provided a platform for residents to interact, reducing feelings of loneliness and panic. These virtual interactions became a crucial lifeline for emotional support, allowing residents to share experiences, offer help, and build a sense of togetherness despite physical isolation. As reflected in one participant’s words,

Second, the CGB fostered a spirit of mutual aid among residents. In tense and high-pressure environments, residents supported each other by strengthening community solidarity and cohesion. This spirit of mutual support involved not only sharing resources but also building trust and mutual respect among neighbors.

Initiating CGB on WeChat involved collaboration with neighborhood committees and community grid managers, indicating government involvement in supporting these activities during the pandemic (L02 group leader). Government officials and property managers were involved in ensuring smooth operation of the CGB. Their involvement ensured that logistical challenges were effectively managed, and communication channels between residents and authorities were open. This demonstrated that the community mutual aid initiative received support from both internal and external sources, thereby developing a strong spirit of mutual aid among residents.

Inequality in the community mutual aid system posed by the CGB also has several negative effects. First, the supply of materials was uneven. Group leaders made efforts to mobilize resources to meet residents’ needs for food and other daily necessities. However, a notable shortage of critical anti-epidemic materials, such as masks and alcohol, occurred temporarily. This scarcity resulted in prohibitively high prices of anti-epidemic materials purchased by community groups.

Second, inequality existed in terms of digital access. Not all residents of Community W joined WeChat groups for CGB, with some continuing to rely on other retail channels. Our survey found that this group mainly included older adults with limited access to digital tools, along with renters who lacked the social connections needed to be invited into WeChat groups. These vulnerable groups are excluded from the CGB-based mutual aid system.

This study demonstrates that, amid the COVID-19 pandemic, CGB evolved from a pragmatic procurement mechanism into a grassroots model of mutual aid, fully embodying the core features of community mutual aid.

First, the emergence of CGB was neither state-mandated nor market-driven; it represented a form of social self-rescue rooted in residents’ immediate needs and collective survival instincts. Motivated by shared risk perceptions, ordinary community residents spontaneously organized procurement, distribution, and last-mile delivery. This bottom-up mobilization bypassed formal institutions (19), relying instead on informal trust networks and reciprocal norms characteristic of social capital (24, 25). These dynamics illustrate how embedded social capital can be activated to sustain community vitality during crises (92).

Second, CGB adopted a flat, decentralized structure that sharply contrasted with the hierarchical logic of bureaucratic and commercial systems (26). Group leaders acted as coordinators rather than commanders, while members flexibly assumed roles such as information gatherers and volunteers. Digital technologies reduced coordination costs and enhanced agency and equality, illustrating how digital empowerment fosters community resilience amid disruption (35).

Third, CGB WeChat groups functioned as central nodes for resource aggregation and redistribution. Group leaders bridged fragmented market supply and localized demand through collective purchasing and optimized logistics (22). This transformation of individual consumption into collective resource management exemplifies the resource mobilization function of social capital (93), demonstrating how community networks compensate for institutional gaps to maintain social continuity.

Fourth, the dynamics of CGB were grounded in trust and reciprocity. Repeated interactions between organizers and members cultivated enduring obligations and shared norms, which constitute the core mechanisms for the reproduction of social capital. These informal reciprocal norms operated as efficient systems of social regulation (94), promoting cooperation and reducing uncertainty. Over time, they reinforced collective efficacy and facilitated the accumulation of both relational and structural social capital.

Fifth, CGB platforms gradually evolved beyond their initial function of material exchange, becoming multifunctional spaces for social support. Communication channels originally designed for logistical coordination transformed into venues for information sharing and emotional sustenance, consistent with prior findings on digital mutual aid (95). This hybrid online–offline interaction strengthened neighborhood solidarity, alleviated isolation, and fostered a sense of community under conditions of physical distancing (96).

Finally, the enduring vitality of CGB stemmed from its emotional foundations. Amid uncertainty, empathy, shared hardship, and a collective sense of solidarity emerged as intrinsic drivers of participation (97). What began as pragmatic risk mitigation gradually became a ritual of collective identity imbued with emotional resonance, such as sending red envelopes (98). Emotional energy transformed economic transactions into symbolic acts of mutual aid, illustrating how affective ties sustain the reproduction of social capital during crises.

Having established that CGB functioned as a form of community mutual aid during the pandemic, this section examines the underlying conditions that enabled this transformation. Understanding why CGB evolved into mutual aid helps clarify the broader circumstances under which mutual aid–based community engagement emerges.

From a social capital perspective, this transformation was not accidental but driven by the activation of social capital within disaster contexts. Disaster sociology emphasizes that disasters are not merely physical events but structural shocks at the societal level, revealing and amplifying vulnerabilities and inequalities within social systems (7, 8). During the COVID-19 pandemic, for instance, lockdown measures led to shortages of essential goods, exposing weaknesses in food supply chains and intensifying feelings of isolation and anxiety (9). Existing research further suggests that such perceptions of disaster risk can activate social capital, thereby encouraging mutual aid at the community level (10). Disaster risk can therefore be understood as an external catalyst that transformed CGB from an economic exchange network into a mutual aid practice, as members mobilized shared resources under common risk perceptions to establish responsive systems of collective support.

The intrinsic catalyst for this transformation lies in the community’s pre-existing social capital, encompassing established neighborhood networks, shared norms, and trust in local organizers. During the crisis, these elements were reactivated and reintegrated into community interactions, serving as internal driving forces behind the transformation (30). Amid lockdowns and resource shortages, residents utilized digital platforms such as WeChat groups to rebuild social connections and form online mutual aid networks. Although these networks operated through digital media, their core participants were primarily drawn from existing neighborhood ties, reflecting the characteristics of a society of acquaintances. These networks exhibited a horizontal, closed, and stable social structure that embodies structural social capital, wherein residents mobilize and coordinate social resources through established relational networks.

Meanwhile, neighborhood norms were reactivated and gradually institutionalized through ongoing processes of resource coordination and responsibility allocation, giving rise to informal institutions that sustained the mutual aid order. As the Chinese proverb goes, “A good neighbor is better than a distant relative,” highlighting that neighborhood-based mutual aid norms, rooted in acquaintance networks, exhibit strong emotional cohesion and resource mobilization capacity during crises (99). Trust, as another critical form of social capital—particularly institutional trust in community organizers—constituted a key mechanism enabling CGB to maintain operations and fulfill its mutual aid functions during crises. This aligns with the concept of relational social capital (26), which emphasizes social networks grounded in shared norms and mutual trust.

This section explores how CGB transformed into community mutual aid during the pandemic. We compare the evolution of this model before, during, and after the pandemic. As shown in Table 3, it initially functioned as a market-oriented procurement mechanism, shifted to mutual aid practices during the pandemic, and later emerged as a hybrid form combining commercial efficiency with community cohesion. This trajectory reflects functional adaptation rather than organizational restructuring. The shift from a commercial to a mutual aid model was primarily enabled by the reproduction of structural and relational social capital, with digital participation pathways and psychological motivations playing pivotal roles in this process.

First, at the technological level, digital participation pathways facilitated the reproduction of structural social capital. Previous studies suggest that mutual aid practices increasingly depend on digital platforms to achieve social objectives, as the hybrid digital–physical spaces they generate constitute a vital foundation for forming and reproducing of social capital (100). Consistent with this, digital platforms became central arenas for regenerating social capital within CGB. Through these platforms, individuals were organized, resources mobilized, and social bonds maintained (33–35).

WeChat, for instance, served as the primary medium for CGB. It operated not merely as a technological tool but as a form of social infrastructure embedded in everyday life. As Joyce notes, digital infrastructures—comprising networks, code, applications, and devices—perform structural functions in social mobilization and resource coordination (101). During lockdowns, WeChat’s architecture and functionalities—such as group chats, payment systems, and mini-programs—significantly strengthened community cohesion. Its open application programming interfaces (APIs) further enabled participation from external organizations and users (102), embedding additional forms of social capital, including governmental and commercial resources, into community networks. However, digital platforms are not neutral conduits of mutual aid. As state, market, and public interests intersect within platform spaces, the reproduction of social capital also exposes institutional tensions. Platform algorithms and governance mechanisms can generate power asymmetries, constrain community autonomy, and potentially undermine the equity and sustainability of mutual aid efforts (88). Thus, digital platforms function simultaneously as the infrastructure for reproducing structural social capital and as potential sources of its fragility.

Second, psychological motivations fostered the reproduction of relational social capital. Consistent with prior research (1, 6, 10, 28, 33), residents’ participation in mutual aid during crises stems from a reciprocal mindset—a core element of relational social capital. This sense of reciprocity cultivates empathy and emotional support among members while guiding participants to balance individual and collective interests.

In the context of CGB, participants pursue personal needs—such as securing daily necessities—while simultaneously upholding collective interests by adhering to group norms, engaging in mutual assistance, and participating in volunteer services. This alignment of self-interest with collective well-being encourages residents to internalize community-oriented values, thereby deepening reciprocity and trust among members and mitigating collective action dilemmas such as the tragedy of the commons (26).

Consequently, although community mutual aid networks retain the form of material exchange, their operational logic during the pandemic transcended market rationality. This finding aligns with prior research emphasizing that mutual aid embodies non-commodified social relations (5). Furthermore, the psychological motivations sustaining participation extend beyond reciprocity to include a sense of responsibility, membership, and accomplishment. Viewed through the lens of mutual aid as healing (26), these psychological dimensions enhance both the stability and the emotional resilience of mutual aid networks.

This finding calls for a deeper engagement with Banks’ conceptualization of structural and relational social capital in the context of digitalization and crisis (26). In traditional community settings, these two dimensions of social capital often operated hierarchically: structural capital facilitated coordination, while relational capital sustained commitment (26, 103). However, the results of this study reveal that within the CGB model, the boundary between these dimensions has become increasingly fluid. WeChat-based infrastructures, while primarily serving structural functions by organizing the flow of goods and information, simultaneously cultivate relational ties through everyday interactions. Neighbors, for instance, build mutual trust and reciprocity through digital gestures such as sharing information, embedding emotional exchanges within structural coordination and thereby generating what may be termed hybrid social capital. This hybridization extends Banks’ framework by demonstrating that in digital crisis mobilization, structural and relational social capital are not merely complementary but mutually constitutive. Structural connectivity afforded by digital tools enhances relational trust, while affective reciprocity, in turn, reinforces network efficiency. Thus, the CGB model exemplifies how social capital has evolved into a digital form of community resilience within emerging sociotechnical and crisis contexts (104).

The transformation of CGB into community mutual aid produced dual effects: it strengthened community resilience while simultaneously amplifying social inequalities.

Consistent with previous studies (7, 28, 29), our findings indicate that CGB, as a form of community mutual aid, significantly enhanced community resilience—not only by alleviating short-term hardships but also by reinforcing the spirit of solidarity among residents. The social capital cultivated during the crisis facilitated collective mobilization (26), and with support from external factors such as government agencies, this spirit of mutual aid further expanded. Although initially temporary, these effects may prove enduring. Just as community mutual aid during the COVID-19 pandemic relied on preexisting resources, future forms of community mutual aid may have activated the spirit of mutual assistance developed during this period.

However, the study also reveals a “dark side” of social capital: its potential to reproduce social inequality (105). Although CGB embodies the defining features of community mutual aid—operating as a bottom-up, informal support mechanism that complements top-down government aid (6, 7, 28)—it does not benefit all residents equally. While some research suggests that mutual aid can mitigate inequality through the accumulation of social capital (2, 9), our findings indicate that, in certain contexts, these very mechanisms may unintentionally deepen disparities. Such inequalities are particularly pronounced among older adults. Consistent with prior research (106), the digital divide—an enduring form of structural inequality—is especially evident in mutual aid activities that rely heavily on digital platforms such as WeChat groups and mini-programs. As one interviewee noted, many older adults residents were unable to place orders independently and had to depend on government supplies (M02 group members).

This suggests that, while digital platforms have strengthened social cohesion and enhanced the efficiency of mutual aid within certain groups, they have also exacerbated intergenerational disparities in access to information, participation, and resources (107). Moreover, although informal assistance systems like CGB can partially relieve pressure on formal aid mechanisms, they may inadvertently weaken these systems’ responsiveness to vulnerable populations. Formal institutions may assume that all individuals benefit equally from community-based mutual aid, thereby overlooking structural challenges rooted in unequal distributions of social capital (108). These findings highlight the need for future research on community mutual aid to examine inclusivity and redistribution mechanisms within social capital, ensuring that mutual aid practices strengthen, rather than reproduce social inequality.

This study highlights the substantial potential of resident-led mutual aid initiatives in responding to future public health emergencies. Community mutual aid can effectively mitigate the collective action problem and promote reciprocal benefit, as neighborhood reciprocity plays a vital role in enhancing cohesion and fostering resilience during crises (109). Nevertheless, inequalities in supply distribution and participation barriers—particularly among vulnerable groups such as the older adults remain a pressing concern. To improve the effectiveness of mutual aid and reduce its potential drawbacks, governments and communities are encouraged to institutionalize community mutual aid programs, establish transparent distribution protocols, and provide digital training or offline support to ensure equitable participation. Furthermore, cultivating a sustained culture of mutual aid among residents can contribute to resilient community governance in China (19, 69). Policymakers should therefore strengthen these efforts by providing institutional and technological support for community-based initiatives.

Additionally, this study underscores the essential role of digital platforms in coordinating participants, mobilizing resources, and providing social support during crises. Neighborhood digital platforms, particularly WeChat groups, serve as key infrastructures for sustaining mutual aid networks and community cohesion (110). Community managers can leverage the functional affordances of these platforms to foster engagement and enhance resilience, while authorities should focus on empowering digital tools and ensuring equitable access—especially in marginalized communities—to maximize the benefits of digital inclusion for grassroots governance.

This study had some limitations. First, during the pandemic, a variety of community-based mutual aid approaches emerged across different regions. Although the CGB model captures shared themes, it is insufficient to comprehensively represent the diverse practices of mutual aid and solidarity observed in communities worldwide. Second, we employed an interpretive case study approach to examine the community mutual aid process based on the CGB model. This methodological choice resulted in a relatively small sample size and primarily descriptive data, limiting the generalizability of the findings. Third, this study acknowledges a key limitation in its underrepresentation of older adults and the lack of systematic data concerning this demographic. Therefore, future research should focus on two specific areas: (1) evaluating the applicability of the CGB model across different sociocultural contexts, (2) assessing the post-pandemic impact of CGB and its contribution to strengthening community resilience and (3) expanding the sample by including focused in-depth interviews and customized questionnaires specifically designed for older participants.