An Ethnographic Toolkit for Studying the Networking Pathways of Hard-to-Reach Populations: The Case of Cosmetic Surgery Consumers in South Korea

Abstract

This article develops a novel ethnographic toolkit for examining the networking pathways that hard-to-reach populations use to socially survive. The toolkit consists of two sampling strategies (snowball and purposive sampling) and three data collection practices (role shuttling, site shuttling, and autoethnography). This article illustrates the applications of the toolkit in an ethnography of South Korean cosmetic surgery clinics and digital forums from 2018 to 2019 by uncovering the role that furtive networks play in facilitating cosmetic surgery consumption. Longitudinal in nature, the toolkit excels in examining the network’s dynamism, informal hierarchy, and the meaning-making and networking pathways that allow members of a hard-to-reach population like cosmetic surgery consumers in South Korea to participate in stigmatized practices. In the hard-to-reach population of surgery enthusiasts, I find that surgery is purchased by consumers through persuasive reconstructions of the meanings of success, body, and self by an elusive network of clinicians, who are introduced by an ever-changing roster of past cosmetic surgery consumers perceived to be high-status.

Introduction

This article examines the networking pathways that cosmetic surgery consumers as a hard-to-reach population use to obtain cosmetic surgery as a stigmatized practice. In demonstrating how local attitudes toward cosmetic surgery are shaped and sustained by networks, this article outlines an ethnographic toolkit for examining the networking pathways that surgery consumers in particular and hard-to-reach populations in general leverage to preserve their communities and consumption of deviant interests.

Hard-to-reach populations are defined as any population that is associated with socially sanctioned or illegal identities and behaviors (Atkinson & Flint, 2001). The sanctioned quality of these identities and behaviors thus forces them out of view of institutional membership lists and census records. Simultaneously, people belonging to sanctioned identities and behaviors labor to hide their identities, making them difficult to identify.

We need not look very far into the recent past to find rife examples of hard-to-reach populations. At the height of the HIV epidemic during the 1980s to 1990s, having HIV was a brand that forced patients to hide the fact, even if it meant neglecting self-care (Chenard, 2007). The previous U.S. administration under Donald Trump intensified a widely publicized crackdown on illegal immigrants, forcing migrants into fringe jobs and hiding their migrant status from official institutions (Alamillo et al., 2019). Drug users and prostitutes strive to continue their practices outside the purview of institutions and the public eye, even as it comes at the cost of losing access to healthcare and police services as a result (Goffman, 2014).

It is politically and academically important to understand these populations. Hard-to-reach populations have typically been marginalized people pushed unto the fringes of society. They are significant because their status reveals fundamental questions about inequality that continue to preoccupy influential agendas in sociology, anthropology, and cultural studies. Understanding participation in sanctioned behaviors is essential to unpacking the ways a hard-to-reach population in particular and marginalized identities at large endure and survive.

Yet, our methodological understanding of how and where hard-to-reach populations network has not kept pace. This article sheds light on this picture by adopting an ethnographic approach to examine networking pathways of hard-to-reach populations, exemplified in an ethnography of cosmetic surgery consumers in South Korea from 2018 to 2019. Rather than addressing the question of who purchases cosmetic surgery, this article focuses on why and how individuals seek cosmetic surgery with a focus on the influence of their ties (Gage, 2013; Wellman et al., 2001). I argue that an ethnographic approach is thus well-suited to understanding networking pathways by tapping into the nuances of marginalized lived experiences, the agency responsible for the negotiation of an identity, and the relational lines of action that eventually build into a network structure.

In what follows, I first review the conceptualization of hard-to-reach populations, the qualities of their network structures (informal hierarchies and dynamism), and the methodological challenges they pose. I then articulate a novel ethnographic toolkit as a data collection tool for networking pathways of hard-to-reach populations. This toolkit consists of two sampling strategies (snowball and purposive sampling) and three data collection practices (role shuttling, site shuttling, autoethnography. The following section demonstrates the application of the ethnographic toolkit in studying cosmetic surgery in South Korea through the stages of entering the field, collecting and analyzing data, and theorizing and identifying networking pathways.

Hard-to-Reach Populations: An Overview

Networks and Networking Pathways

This section reviews a loosely bound literature on hard-to-reach populations and infer the methodological lessons they yield for the study of their networks, most of which are constituted by members associated with identities or behaviors that are sanctioned or deemed deviant, if not taboo, in mainstream society.

For hard-to-reach populations, social networks are indispensable to community-building, creating new social ties and exchanging support and information crucial for such populations to subsist—while keeping their identities a secret (see Mo & Coulson, 2008). The general avoidance that comes out of repression makes these individuals difficult to identify and their lived experiences elusive to understanding. A strong tradition of public health and sociology research shows how this has been the case for stigmatized individuals such as drug users (Griffiths et al., 1993), HIV-positive individuals (Solomon et al., 2018), and illegal immigrants (Larchanché, 2012).

Networking pathways, or the movement of individuals through chains of ties to procure illicit goods or participate in deviant activities, are essential to the social survival of hard-to-reach populations. Networking pathways can also be conceptualized as a variant of the “network capital” that Wellman and Frank (2001) theorize as the “form of ‘social capital’ that makes resources available through interpersonal ties” (p.234). Following Ryan and Mulholland (2014), this conceptual angle allows us to capture in our observations a wider range of network ties; ties that are not constrained by a “rigid dichotomy of bridging versus bonding… a continuum of relationships that are spatially and temporally dynamic” (p.151).

Social embeddedness is what ultimately comes to shape—and so cannot be extracted from—questions about meaning-making. As an influential strand of social network scholarship holds, the attitudes a person holds are not entirely determined by personal attributes, but by the relational qualities of the social structure of such networks (Rivera et al., 2010). Much research shows, for instance, that the attitudes a person holds toward any given issue do not originate purely from personal biases or experiences, so much as they are shaped by the interactions with people they interact most with (Erickson, 1988) or must interact most with per constraints of a shared organizational or social setting (Rivera et al., 2010). That is, in general, greater interaction anticipates stronger cohesion in values and sense of group belongingness (Lawler et al., 2008), an association found to be true in workplace settings (Granovetter, 1973, p.162) and general geographic proximity (Bidart & Lavenu, 2005). In seeking to understand hard-to-reach populations, it thus becomes especially important to examine how attitudes and interpersonal processes are actually shaped by structured relations.

Methodological Challenges of Studying Hard-to-Reach Populations

The challenge at hand in attempting to understand the network structure of hard-to-reach populations is not how to analyze their networks, but how to collect data on their networks in the first place. Two methodological challenges loom large for studying these networks.

(1) Informal hierarchy. Hard-to-reach populations are hidden from the oversight of formal institutions. For fear of persecution, their stigmatized practices and behaviors are not documented in institutional membership lists or census surveys. This makes identifying and accessing them a particular challenge. They do not have a formal organization to tap into to conveniently access a comprehensive list of members and directly contact each one as we might with a systematic, quantitative study (Erickson & Nosanchuk, 1983).

But the absence of a formal structure does not mean they have no structure at all. Social structure, to paraphrase Giddens (1984, p.25), appears in any form of social life where people meet and interact. Hard-to-reach populations form new networks to sustain their practices and support one another in the shadows. In the process, they adopt an unofficial, informal hierarchy to order their behaviors. We may, thus, define an informal hierarchy as a set of social relationships of dominance and subordination that emerges with continued interaction over time. Unlike formal institutions, what bulwarks the infrastructure of this hierarchy is the consent of the many that gradually shift into relationships of dominance.

We observe direct and indirect evidence of an informal hierarchy in hard-to-reach populations. A healthy body of organizational network literature suggests that, in general, the decline of formal hierarchy is correlated with the rise of informal hierarchy (Diefenbach & Sillince, 2011). Following Erickson’s (1981) influential review of the most prominent types of secret societies, hard-to-reach populations are often not simply an assorted collection of one-time transactions but are typically characterized by a “persistent pattern of relationships” (p.189). Like in William Golding’s (1987) notorious Lord of the Flies, people within a network give and/or receive orders and interact with specific people in patterns that shuffle them over time into different, unequal positions (see also Warr, 2005). It means that some people are connected to a larger proportion of the network, and others less. It may also mean that specific members act as brokers who facilitate interactions or initiate individuals into the network.

These characteristics complicate the matter of accessing a hard-to-reach population. For a researcher to gain access to their network, s/he requires an introduction—and endorsement—by an existing member and intimate knowledge on how to navigate the process of entering. This grows particularly important as many informal hierarchies may co-exist in different social spaces within the same ecology or field for hard-to-reach populations (Liu & Emirbayer, 2016). Researchers must earn participants’ trust to integrate themselves into the sites where these informal hierarchies are located—and to understand the significance assigned to different positions and social structures once they are in (Collins, 1986).

(2) Dynamism. Virtually all networks and populations are dynamic. Old members depart and new members enter. Changes in membership are easily traceable if associated with formal institutions. One need only track updates in a comprehensive, official list of member names (Erickson & Nosanchuk, 1983).

Hard-to-reach populations, however, are a different story. The perpetual flux of their networks, with actors shuffling in and out and ties between them waxing and waning, is easily a “black-box” for researchers. Krebs (2002) importantly advances the argument by showing the fuzziness and dynamism of their network boundaries—how a member may serve multiple roles and how roles may shift to different members over time.

Without complete membership lists to work with, researchers must trace and examine the dynamism of hard-to-reach populations in different ways. One approach could be longitudinal data collection. It may be better served for teasing out the regularities of the networking choices members make in regular, everyday life. Indeed, how members order their social lives to successfully participate in stigmatized practices while protecting their identities and social well-being may require data collection beyond a single, fixed point in time (Monge & Contractor, 2003). Longitudinal data collection may also capture moments where these regularities are interrupted and identify situations that threaten members’ successful participation in their stigmatized practice or the concealment of their identities (Valente, 2007).

Another approach for tracing the dynamism of hard-to-reach populations could be relying on word-of-mouth from well-connected members in the network. A strong connection with members of the network could prove useful in understanding the context behind dynamism. What motivates certain members to join or leave? Like Elijah Anderson (2000) argues, having a direct line to members of hard-to-reach populations captures configurations in the social structure and lends predictive power for how and where network changes might occur in the future.

Developing an Ethnographic Toolkit

How do we collect data about the network structure of hard-to-reach populations in the first place? An ethnographic toolkit accounts for challenges with securing participant trust, defines the positionality of the researcher in a way that accomplishes this, and uncovers how ties change.

Ethnography lays out different typologies of social networks in human relationships, finding unique relationships that are potentially crucial to network analysis but overlooked for their deep embeddedness in micro-level interactions. Constituting a longitudinal data collection, ethnography powerfully unearths the overlooked patterned dynamics of everyday life and draws attention to interpretive accounts which deepen – and resist – existing theories of the social world (Desmond, 2012). Its grounded, proximal observation of events as they occur imparts important insights on several network characteristics: (i) it explores how actors react to stress or an event or change that shakes up the usual procedure of things; (ii) it illustrates the context in which ties are strengthened, weakened, formed or broken, which also lends to (iii) scoping the networking pathways and choices that mediate interactions among members of a network.

Thus, ethnography is an ideal tool to circumvent the methodological challenges stated in the previous section. To further sharpen its potential for navigating the dynamic nature of networks of hard-to-reach populations, I further propose a new toolkit for network ethnography consisting of two sampling and three data collection practices.

Two Modes of Sampling

The ethnographic toolkit’s sampling methods draw inspiration from the constructivist grounded theory practice of theoretical sampling. Beginning from the recognition of the fact that analysis, coding, and memo writing begin at the same time as data collection, this consists of choosing a sample for the purpose of theory construction (Clarke, 2005): intentionally recruiting subjects who both identify with your phenomenon of study, and whose perspectives are as diverse as possible – to capture the range of dimensions that structure experiences within the boundaries of a single site.

For sampling, I argue that snowball and purposive sampling should be concurrently conducted. (1) Snowball sampling. Snowball sampling consists of asking respondents for referrals to social ties among their personal networks who might be interested in participating in a researcher’s project. Snowball sampling offers practical advantages for accessing members of hard-to-reach populations, who are difficult to identify or initiate contact with.

When a researcher is referred to members of a hard-to-reach population by an existing member, s/he becomes associated with a degree of trustworthiness (Atkinson & Flint, 2001). Indeed, first impressions are important for fostering lasting relationships, especially for hard-to-reach populations wary of outsiders (Contreras, 2019). Their marginalized status makes them vulnerable to persecution by authorities and stigmatization by social ties. Being introduced by an existing member of the hard-to-reach population lends credence to the researcher’s background and vindicates his/her intentions, being seen as possessed of values that insiders share.

Recent work from urban studies richly documents the proficiency of snowball sampling for studying hard-to-reach populations (Contreras, 2019), distinguishing itself as a way to access marginalized hard-to-reach populations and examine in-depth results in a reduced amount of time. Here, snowball sampling prompts participants to assign their own meanings to identities, more accurately capturing their lived experiences. It also allows researchers to only include in their research participants who are motivated, engaged, and interested in participating. Finally, like Erickson and Nosanchuk (1983) suggest, snowball sampling can be altered to verify its own data sources when members of the population check information provided by others.

(2) Purposive sampling, which consists of intentionally selecting specific subjects for study (respondents, field sites, etc.) based on the study purpose. Purposive sampling is motivated by the assumption that each subject will yield thick, contextual data about flows of action, meaning, and people. It is thus particularly well-suited for qualitative research on hard-to-reach populations characterized by stigmatized practices because it focuses on a few socially significant sites that, because of their rarity, do not lend well to broad sampling, but which instead lend well to exploring their internal depth.

I assert that purposive sampling should be conducted with key sites and actors – those who are deemed important to the perpetuation of a network of hard-to-reach populations. The importance of a site should be determined by how well it performs as hubs of exchange among the network, determined by information that a researcher “scavenges” from academic, media, grey literature, and archival sources pertinent to the hard-to-reach population (Holliday & Elfving-Hwang, 2012). I emphasize that the sites to be sampled for data collection are not simply in-person, but online as well, given the continuity of individual and collective behaviors across the online-offline threshold (boyd, 2014). By capturing the context in which these exchanges happen, one can thus infer how the structure, formation, and content of ties inform the relational embeddedness that constitutes and ultimately perpetuates the illegitimate practice in question.

Concurrently conducting snowball and purposive sampling allows researchers to discover the important actors in a given network. This evokes what might be called a qualitative conceptualization of network centrality by targeting and uncovering who the most important actors are in the network structure that exists within the population (Bonacich, 2007). These actors can be nominally important as key informants for researchers, giving them access to further parts of the hard-to-reach population network and scoping out new networking pathways. These actors can also be relationally important to the facilitation of social exchanges in the population itself.

Three Modes of Data Collection

For data collection, I propose role shuttling, site shuttling, and autoethnography be concurrently conducted within the field. (i) Role shuttling is the practice of changing different roles in the field within the sites selected. I propose a systematic attempt to alternate the levels of engagement that a researcher partakes in. A common thread in ethnographic research is the recognition of the fluidity of the boundaries that ethnographers cross when they become part of a field site (Burawoy, 1998). As time goes by, ethnographers are often called to play different roles within a local social structure, particularly as they become more familiar with participants and gain new access to unforeseen parts of participants’ lives.

I build on this idea by suggesting that researchers consciously shift their level of engagement across different activities they engage in, such as from a complete participant to a complete observer or vice versa (Au, 2017; Adler & Adler, 1994). This is important because varying the degrees of our engagement in the field enables us to interact with a wider range of actors, through which we can ascertain who connects with who, under what conditions, and how they do so. Ultimately, we gain better visualization of, and even access to, the different layers of the informal hierarchy of a hidden network and the networking pathways that sustain it.

(ii) Site shuttling. Similar to role shuttling, site shuttling consists of conducting fieldwork and observing in different sites among the sites selected. This spreads out a researcher’s exposure to explore the hubs of exchange among a hard-to-reach population in both offline and online contexts. This practice also delineates the social and symbolic boundaries of a network: how many sites are involved in the practice of interest among a hard-to-reach population? What facilitates entry into these sites? We thus glimpse how members of a hard-to-reach population participate in the sanctioned/stigmatized practice.

Site shuttling and role shuttling complement one another to generate variation from within. Uncovering different actors and different perspectives help to document details with significant depth and attention to their immediate (i.e., local) and higher level (i.e., governmental) contexts.

(iii) Autoethnography consists of systematically recording reflections about one’s feelings, reactions, and experiences on a routine basis. The idea is that the temporal process of integration you experience from observer to participant renders your own experience a source of valuable data, moving beyond the fringes of the community to stepping across social and symbolic boundaries that bar outsiders. The transition lays bare (changes in) experiences with which to not only triangulate accounts provided by other participants in the field, but also to document shifts in your own presuppositions as well as to make inferences about the repertoires of values and schemas that order actions in the community (Au, 2017; Swidler, 1986). Autoethnography is also an invitation to account for positionality, most of all as it comes to bear on how one navigates the social and symbolic boundaries that impose social distances between researchers and a hard-to-reach population.

The Relational Bases of Cosmetic Surgery Consumption

Cosmetic surgery consumption among non-celebrity consumers has surged over the past decade (Featherstone, 2010; Gimlin, 2007). Growth ranging from to 200% to 500% has been recorded in South Korea, Italy, Britain, Germany, Brazil, among other nations (International Society of Aesthetic Plastic Surgery, 2017). My study focuses on South Korea, which exhibits the fastest growth with 20 procedures occurring per 1000 people (Baer, 2015) and extremely gendered consumption where over 30% of women have had surgery (Holliday & Elfving-Hwang, 2012).

Yet despite its popularity and even valorization in select social spaces in South Korea, cosmetic surgery does not enjoy widespread cultural acceptance (Holliday et al., 2017). Although the boundaries of this population are not closed—there is no limit to how many people can be a part of it—there remains a collective identity with status-meanings toward which members socialize their attitudes about cosmetic surgery and fellow consumers. These characteristics strongly suggest qualities of boundedness and hiddenness about cosmetic surgery consumption that make consumers a hard-to-reach population.

On the one hand, cosmetic surgery is rising in popularity among non-celebrity consumers. On the other hand, cosmetic surgery is met with cultural intolerance. In light of this paradox, the substantive research question my study set out to address was: why do consumers desire cosmetic surgery?

I applied the ethnographic toolkit in a year-long ethnography from 2018 to 2019 in seven top cosmetic surgery clinics and digital ethnography of online blogs and forums concerning cosmetic surgery based in Seoul, South Korea. I examine the applications of the toolkit for studying the networking pathways of hard-to-reach populations and how they retain participation in deviant activities despite stigma. All participants’ names have been anonymized.

Part I: Entering the Field

I began my fieldwork selecting the typologies and locations of my sites widely, but purposively. Typologically, I decided to conduct digital ethnographies of online networking sites and ethnographies of physical clinics. Before entering the field, these methodological choices were motivated by the theoretical characteristics of my population of study.

Gleaning from digital sociology on East Asian networks that hearken to the significance of online networking sites for social life (Au, 2021), it was determined that bounded digital spaces of interaction are a particularly important anchor for topic discussion and group memberships in South Korea. This is evinced by faster flows of information among members of local social networks and with broader reach to a larger proportion of the network (Barbalet, 2016). Coupled with the fact that individuals within these networks have a higher probability of knowing each other, this type of network works against individuals interested in behaviors that might be shunned by strong (particularly kinship) ties ¹ . Discussions about such issues are thus pushed into online spaces, where consumers of stigmatized practices can discuss their concerns anonymously and, just as important, build ties with fellow members otherwise difficult to identify.

Keeping this in mind, I determined that online blogs were an important part of the social fabric of whatever structural cohesion existed in the hard-to-reach population that cosmetic surgery consumers represented. I turned to NAVER and DAUM, the two most prominent sites for hosting blogs in South Korea, and identified several prominent forums and blogs. There, interested consumers congregated within moderated forums to discuss the benefits, risks, concerns, reviews, opinions, and experiences of different cosmetic surgery clinics, practices, and trends.

Physically, I decided to conduct fieldwork in clinics. Clinics represent the lynchpin site of interaction surrounding cosmetic surgery. Constituting where consumers go for consultations and, in most cases, the surgeries themselves, the clinic is a site of rich dynamical interactions between different classes of actors: surgeons, responsible for addressing the technical issues of surgery and conducting the surgery itself; clinicians, responsible for mediating between surgeons and consumers, facilitating consumer consultations pre-surgery and following up with them post-surgery, and essentially “selling” the procedures to consumers; and consumers, the interested parties and members of the hard-to-reach population purchasing the practice and integrating its results into their lives.

Location-wise, I decided on Gangnam-gu. I drew upon government reports and grey literature to identify this as a prime location for cosmetic surgery. According to the most recent Korean Gallup poll report, 36.7% of cosmetic surgeries occurred in Gangnam-gu, a single district of Seoul (Korean Gallup, 2015). My experiences touring through different districts corroborated this observation. In Gangnam-gu, a district most saturated with tourists and towering skyscrapers that housed a gamut of global brands, I could identify roughly five to six (and sometimes more) cosmetic surgery clinics in every couple of buildings I passed by.

Thus, I selected the clinical sites for my fieldwork from within Gangnam-gu. I began my site-selection by selecting the most high-profile, large clinics (which occupied anywhere from four to 17 floors of a building, each one a site for specialist types of procedures ² ), then several small clinics (which typically occupied one or two floors of a building), which were taken to represent the largest congregations of consumers as well as the broadest scope for potential consumers. Seven clinics were selected. Many clinics responded positively to the prospect of learning more about their own actual and potential consumers.

In both typologies of sites, online and offline, I also conducted snowball sampling to recruit individuals for later interviews, as well as to build my own network as a would-be participant navigating these networks. Individuals I became connected with were always made aware of my intent and my line of work as a researcher. Clinicians and consumers alike were motivated to spread the word on my project to other potentially interested others because of their interest in the practice of cosmetic surgery itself and in the results of the study – why, for instance, people were interested in surgery and how they found their way to a clinic.

Part II: Data Collection and Analysis in the Field

Role Shuttling and Site Shuttling

Physically, I site shuttled between different clinics by attending each one once a week. Each different day of the week, I attended a different clinic, repeated over the span of a year for seven clinics. Having this breadth allowed me to cross-compare themes that emerged in one clinic versus another to build more explanatory theoretical categories. It also allowed me to experiment with different participation roles, across which I role shuttled between being an engaged consumer and a passive observer at different clinics and in online discussions. At some clinics and forums, I spoke more often with consumers who were new initiates; those were interested at best and had not yet consumed. At others, I spoke more often with those who were recurrent consumers. At others still, I constrained myself by interacting more with clinicians instead of consumers. The diversity of perspectives allowed me to observe patterns in who spoke to who and the social junctures when these interactions occurred.

As a passive observer, I documented the dialogical exchanges between clinicians and consumers, as well as between fellow consumers. Here, I could excavate substantive and structural themes that patterned the interactions. Substantively, for instance, I discovered that clinicians often use a variety of discursive strategies, offering bonuses, discounts, and advantages to consumers under the guise of “luck,” to market the sale of a procedure to consumers. In an exchange between Anthony [SK023], an interested consumer, and Raina [SK032], a senior consultant at one of the largest cosmetic surgery organizations in South Korea, discourses of “luck” emerged in persuading consumers to purchase:

“Raina: [This package of] zygoma reduction and jawline surgery really suits you. It is about 12 million WON. It will make your face more even and symmetrical, which is what you want, right?

Anthony: I see… it does look enticing.

Raina: It’s perfect. What holds you back?

Anthony: Well… the cost is a bit high. I don’t know if I can afford it.

Raina: Hm… well, since you are so interested in us, I think there’s a chance to make something work. You are lucky, very lucky, that you spoke with me today. I can make it so you can have the same deal for 9 million won.”

The invocation of “luck” discursively singled out the consumer. Used as a rhetorical frame for a price change, “luck” constituted a strategy to dissuade doubt and establish a sense of connection and trust with other consumers [SK028, SK029]. As Anthony later recounted after the consultation, “it made me feel special. You know, even if I know it’s probably a sales strategy, I still can’t help that it made me feel unique and cared for.” He consequently chose to purchase the cosmetic surgery package with Raina. This discursive strategy featured strongly in consultations with many other consumers as well.

Additionally, among consumers, one of the most powerful and universal themes in their accounts of cosmetic surgery was the desire for unofficial advantages in different social spaces outside cosmetic surgery. Kim-sook [SK007], a saleswoman at a mid-sized (and which I later corroborated after directly interacting with discussants) dispenser in South Korea, recounted how she obtained more workplace benefits by superiors for appearing “beautiful” after surgically modifying her appearance: “My boss paid more attention to me. He gave me more leeway in choosing which branch [in Seoul] I wanted to work at.” Others corroborated this account of workplace advantages and broad social advantages [SK009, SK035]. All participants spoke favorably about how surgical modifications yielded greater interest from their social ties, who responded positively to their modifications, praised their newfound beauty, and invited them to more outings and events.

Structurally, I discovered that clinical interactions were organized to institute a two-way flow of power between consumers and medical professionals. On the one hand, consumers exerted influence over the clinical consultation by leveraging their position as clients and the capital they were to invest, akin to how customers do the same with service-oriented professions like seeking a manicure at a salon (Johnson, 1972, p.65). On the other hand, clinicians exerted influence over the consultation by framing the body and face as a canvas upon which standards of bodily appearance are projected—and persons sanctioned when these standards are not met—and clinical judgments of appearance in terms of the need for cosmetic surgery as a guiding hand of professional expertise (ibid, p.51).

Continually observing online forums, I discovered that this tension was a powerful motivating force for consumers outside the walls of the clinic. Online, consumers often discussed what they felt was “wrong” with “themselves” in terms of their bodies and faces, reporting not only their own self-diagnoses but also (in almost every single case) the corroboration they received from clinicians after a consultation(s). As Kim-Sook elaborates, “when I spoke to the consultant [at the clinic] about my desire to enlarge my eyes, she told me that it was true. But she also said that my double eyelids were too thick. I did not even notice it before, but then I started seeing what she saw.” For Kim-Sook as well as others [SK028, SK042, SK051], the self-diagnosed imperfections in their appearance were given credence by clinicians like Raina and her colleagues [SK038, SK039], who often capitalized on these insecurities to depict even more imperfections. “It mattered because it meant I was imperfect,” Kim-Sook went on, “your appearance is like your business card. It is your first encounter, what others see before they even get to your actual business card, your experience, your education, and so on.”

Captured is an important eclipse between the body and self, wherein conceptualizations of the self are subsumed into that of bodily appearances ³ . By conceiving their selves in terms of the territory upon which the clinician laid claims of expertise (the body), consumers relented the jurisdiction of bodily evaluation—and thus, the decision to obtain cosmetic surgery—to clinicians under the desire to correct a perceived imbalance between their imaginations of their external and inner selves.

The same dialogues (between ties in online forums and in-person) that gave rise to this interpretation were also what informed the networking pathways consumers used to locate a clinic. Consumers chose a clinic by soliciting advice from others online and offline. Online, they would post questions about a given procedure, often within a statement of their appearance goals, which others would respond to and begin a dialogue about, sometimes continuing beyond the post itself (that is, the poster and responder would transition into a private conversation). Subsequent conversations with some of these individuals uncovered how they solicited advice to select a clinic from others in offline contexts as well, asking friends they first knew in-person.

The importance of networking as a method for facilitating clinic selection became key to ramping up my access to these networks. Becoming an engaged consumer in clinics, I went beyond mere observation to actively generate dialogue about my own body as a canvas for cosmetic surgery. I participated in consultations with clinicians, soliciting recommendations and discussing possible procedures – and even appearances – that my body could become.

Through these fluid transitions between passive observer and engaged consumer, I was able to ascertain trust more quickly from both clinicians and consumers in different sites. Viewing my identity as one of them, they trusted me not only as someone who would do no harm but as someone who shared an interest enough in their practice to give opinions to and solicit advice from, to engage with by means of social support in building a genuine relationship with. Thus, these connections allowed me to protract my study into the long-term and catalyze my snowball sampling for consumers by receiving referrals for contacts for conversations, which expanded the number of individuals I worked with and provided access to different parts of the hard-to-reach population.

As Duneier (2002) suggests, qualitative researchers in general and ethnographers in particular are not simply instruments of their craft, but humanized agents (p.1560); the same impulse to empathize with my participants also helped fuel my adoption of their desires. The interpretive processes I asked consumers to report about (what they felt about their own bodies, their consultations, etc.) and the cross-comparative questions I asked (how they felt about others’ bodies, how they felt different categories of actors figured into their decisions) helped me identify how (and why) types of ties mattered to evaluating cosmetic surgery and bodies. This kind of abductive analysis built robust theoretical inferences about the whole of the hard-to-reach population on what role networks played in motivating and facilitating consumption (Timmermans & Tavory, 2012).

Autoethnography

Autoethnography enmeshed well with the longitudinal approach of the ethnographic toolkit, capturing changes in my own theorizations—and personal feelings—over time. I documented my feelings, experiences, presuppositions, and hunches regularly through the year after every consultation with clinicians and consumers. In so doing, I recorded the interpretive changes that I was experiencing, adding nuance, triangulating, or holding accountable the interpretations offered by participants and the theoretical categories I was trying to build.

My curiosity in cosmetic surgery was not ingenuine. It began from a place of intellectual curiosity, and, without attempting to sway any interested consumer, no ethical barriers were breached ⁴ . But as I deepened my engagement, with every passing clinical consultation and conversation about cosmetic surgery, my curiosity began to change into agreement, and later, desire.

At first, I asked why others were interested in cosmetic surgery. Participants, clinicians, and consumers alike repeatedly spoke to a wealth of social benefits, operationalized in more ties, stronger ties, more resilient ties, a greater tendency to form ties to begin with, all of which extended into the likelihood of forming different types of ties altogether (i.e., romantic ties, workplace ties). Several months into the fieldwork, I realized that their discursive depictions of social capital in connection to cosmetic surgery revolved around the theme of contingency: just as one felt rewarded with social capital after obtaining cosmetic surgery, socialization to this fact made social capital interpretively contingent on having surgery. It hearkened to a basic, cliché instinct about networking in social life: better-looking people are better liked ⁵ .

According to another recurrent female consumer, Tracy [SK002], who I spoke with in an informal interview at a cosmetic surgery clinic, “having a great social life depends on being beautiful… this is true for friends [and] for romance. People were more motivated to speak with me, and I didn’t even realize it till I got surgery. It’s scary to imagine what would’ve happened now if I didn’t get surgery.” For Seok-Min [SK003], another woman consumer at the same clinic, this fear was a lived reality. According to her, “I haven’t been very successful at dating. Men see me, and they don’t see me again after a couple dates. They end up seeing other girls, who I know have had cosmetic surgery. They’re so much prettier. Getting cosmetic surgery myself might even the odds. It might improve my chances with men and reverse my losing streak [laughs].”

As these informal interviews suggest, there was an unwritten appearance bias that governed how one is received in social settings that include friendship circles and romantic dating. Observing other social ties obtain surgical modifications and subsequently succeed in these social settings, consumers were socialized to associate success with cosmetic surgery. By the same token, they also learned to infer that failure to obtain cosmetic surgery constituted deprivation of the same social rewards.

As such, the desire for cosmetic surgery was bound up in an instrumental desire to improve one’s own life by satisfying informal demands for appearance imposed by others. As I continued to participate in my observatory and engaged roles, reflexively engaging with this theme throughout, the question turned itself on me: would I improve my own life after doing so? I became convinced that it would. If others valorized surgically modified bodies, then regardless of how I felt, my value in the eyes of others would improve. Another question had surfaced: what was I potentially losing in failing to be “beautiful”?

For the purposes of my research, I had moved to Seoul. This is often an underappreciated, but paramount insight. As urban ethnographies of the underclass have demonstrated, we learn the traits, habits, orientations that are key to social survival not as a researcher in a field site, but as an integrated part of the larger social fabric in which our field sites themselves are embedded (Goffman, 2014). Thus, the contingency of social capital on cosmetic surgery began for me as a nominal insight (an observation of themes in the local networking cultures I was being exposed to while socially distanced as a researcher observer). These insights then became expectations of real experiences that I was socializing myself toward (an account of personal losses and gains in my own social life).

This was a classic case of the researcher going native, being so absorbed into their field site that they came to internalize norms and values of locals—and it was my autoethnography that helped me realize this as it happened to me and convert it into a source of data itself to visualize how networks informed consumers’ interpretations of surgery. The apparent shift in my own interpretations of cosmetic surgery detected by my autoethnography provided me with an insider glimpse into the affective power over evaluation that the discursive narratives I was being exposed to wield. That is, my first-hand experiences, feelings, and reflections became a valuable source of data that sharpened the analytical precision of the theoretical inferences I aimed to make and, just as important, how the structure and content of networking itself ultimately combined to inform the desire for cosmetic surgery.

Part III: Theorizing and Identifying Networking Pathways

The emergent analysis from the previous section, comparing cases within and across themes in cosmetic surgery consumption, helped to push toward higher-level theorizations that make sense of its local popularity. Finding cross-code commonalities helped formulate higher-level abstractions, allowing abductive theorizing on the thematic categories of motivations and actions for the sake of cosmetic surgery (Timmermans & Tavory, 2012). I cast inductively made observations of a field site against concepts from the literature to determine “the overall sense of a situation” (Hammersley, 2004, p.27) or theories that better “fit” the data (Burawoy, 1998). This ensured reliability by continually examining and re-examining separate reports until data saturation was reached in unearthing the social embeddedness that linked desire and relational action.

Indeed, in seeking to understand why consumers sought cosmetic surgery, I discovered the answer in how they went about seeking it out. The contextual proximity and intimate trust from participants accessed by the ethnographic toolkit allowed me to develop important insights into how the desire for cosmetic surgery itself was socially embedded and, correspondingly, how network remained dynamic through a value-based social ordering. Just as one felt rewarded with social capital after obtaining cosmetic surgery, socialization to this fact made social capital interpretively contingent on having surgery, an evaluative process enforced by key actors (past cosmetic surgery consumers) in interested consumers’ networking pathways and eventual pathways to surgery. The results suggest that networking itself is not only the method by which consumers seek out cosmetic surgery, but a consequence (in terms of perceived social capital gains) and therefore cause of its motivation (socialization toward these gains, enforced by interactions with key actors).

Thus, the ethnographic toolkit was key to parsing out the role that networks played by accounting for the (i) informal hierarchy and (ii) dynamism that defined this population. Concurrently, the ethnographic toolkit unearthed the networking pathways and pathways that ensured the population’s social survival and continuity. The ethnographic toolkit gave me proximity to the context that networking choices occurred and as they did, which was integral to uncovering how (and why) types of ties mattered to evaluating cosmetic surgery and bodies.

(i) Informal hierarchy. The hierarchy that characterized the hard-to-reach population of cosmetic surgery consumers was not institutionalized per se ⁶ , but diffusely stratified into status orderings of classes of actors. Surgically modified bodies became an important value ⁷ for participants in a way that rendered cosmetic surgery consumers representative of a high-status in-group with pronounced abilities in maintaining, forming, and strengthening ties as they pleased (see also Lamont & Molnar, 2002). These fell into two conceptual classes of actors: past cosmetic surgery consumers (high-status) and non-cosmetic surgery consumers (low-status).

People interested in cosmetic surgery resorted to a common networking strategy to begin and navigate the market. Interested consumers (low-status) selectively turned to past cosmetic surgery consumers (high-status) from amongst online and offline friends and acquaintances in seeking out information on the practice. Like studies of patients seeking out alternative medical practitioners (Wellman et al., 2001), ties evidently mattered for cosmetic surgery consumption by transferring information needed to identify the right clinics and connect consumers to them. The present study advances the argument by showing how the very interactions within this networking strategy, the way in which networks were a method of facilitating cosmetic surgery, also became a core part of the motivation for cosmetic surgery. Hearkening to social embeddedness, the structure of the relationship interweaved with the content of interaction within network pathways that anticipated action (cosmetic surgery).

Interested consumers felt they had surpassed the social boundaries into a higher status during their interactions with surgically modified alters, feeling accepted and encouraged by the latter toward the practice. It was here that many consumers like Anthony, Kim-Sook, Tracy, and Seok-Min moved from indecision to action in deciding to obtain cosmetic surgery, prompted by encouragement from surgically modified alters who endorsed the practice by sharing insights on their enhanced social capital post-surgery and corroborated the imagined benefits of surgery for a more satisfied and engaged social life. Within this interaction process, participants reported [SK51, SK101], cosmetic surgery conceptually evolved from a practice for self-satisfaction alone to one about the relationships between themselves, their peers, and high-status alters, which surgery would help improve. Concurrently, the interaction imbued cosmetic surgery with the promise of elevating individuals’ identities past the symbolic boundaries of this high-status in-group of surgically modified individuals to count amongst their midst, not only in company (socially interacting with surgically modified alters), but in identity (being equal with surgically modified alters by undergoing the same rituals that afforded them their status).

(ii) Dynamism. A longitudinal perspective allowed me to observe changes in membership in the population and identify explanations for change. Some members of the population left, no longer interacting with others in the online and offline communities about cosmetic surgery, because they had given up their interest in it. Others in the population changed their roles over time. Finally, consumers like Tracy and Seok-Min who ended up purchasing surgery would come to dispense advice of their own to newcomers [SK028, SK029]. This subtle change in roles was a form of upward mobility that revealed how new network pathways were born: experts who interested consumers consulted in seeking cosmetic surgery only came from within the existing population. This meant that the hard-to-reach population was more insulated and its informal hierarchy more structured than expected, because those who held the highest status (past cosmetic surgery consumers) and who were key to facilitating the survival of the population (guiding non-consumers) only came from within.

Amidst the constant shuffle in the population from non-cosmetic surgery consumers to past cosmetic surgery consumers (a steady flow of members who went ahead to consume surgery), I discovered that the networking pathway stayed the same. People who wanted to consume cosmetic surgery did so by initiating contact with someone who had already undergone surgery, even if who they sought out changed over time.

In the cosmetic surgery market, only one type of actor did not change: clinicians. After interested consumers had contacted and discussed the subject at length with a past cosmetic surgery consumer, they finally went on to see a clinician. Clinicians were the very last juncture in the network pathway to obtaining cosmetic surgery, but because they were professionally affiliated with the surgery clinics themselves and were not active participants in online forums, they were neutral in the status game of the informal hierarchy.

Almost universally, people interested in cosmetic surgery acted on this interest by networking with specific types of ties. They sought out past consumers, whether seeking them from among personal ties or from among members of online forums [SK008, SK009, SK028, SK035]. When they finally came to clinicians, consumers appeared more receptive of the advice and recommendations that clinicians gave, as witnessed earlier.

The complacency that consumers show when they reach clinicians enhances our qualitative understanding of the influence that a member of a network has over the actions and interpretations of others in the network (Bonacich, 2007). Through clinicians, I observed that although they had no stake in the informal hierarchy that ordered the hard-to-reach population, they appeared to exert powerful influence over consumers’ decisions. The opinions of clinicians like Raina’s about what procedures a consumer needed were sought after and well-heeded [SK038, SK039]. This was not simply because of their professional identity, but because consumers had developed preconceptions about their expertise and craft before reaching them. Clinicians were referred to by past consumers based on which and how many other clients they had consulted for. Those with a higher number of consultations, experience consulting dozens of clients every day, embodied a kind of network centrality, working to their advantage. A clinician believed to be well-networked gained superior influence over consumer decisions. The accumulation of this type of influence, in turn, sensitizes us to the underrecognized importance of actor locations in a network pathway for influence. As the terminal node in many network pathways of interested consumers (the very last person a consumer saw), clinicians had already been talked about at length and evaluated by past consumers—before interested consumers walked in the door. Having this status beforehand gave clinicians an important edge in legitimating their expertise and influencing consumers’ decisions and interpretations during eventual consultations.

Discussion

This article has outlined an ethnographic toolkit that contributes to the qualitative study of hard-to-reach populations by articulating how to collect data and examine the networking pathways through which they subsist, using the case of cosmetic surgery consumers in South Korea. Parts I and II focused on the methodological details of my project and how the ethnographic toolkit worked in practice to collect data. Part III showcased the implications and significance of my approach for the content of the social exchanges, combining nominal insights (meta-level feelings about cosmetic surgery as a practice) and personal experiences (of cosmetic surgery consultation and consumption) to generate a picture of social embeddedness that originates the motivation for cosmetic surgery consumption.

In the present study, interested consumers (low-status) leaned heavily on others online and in reality who had already undergone cosmetic surgery in the past to assortatively access clinics through elusive networking pathways. Afterward, gatekeeping clinicians lulled consumers into discursive jousts that valorized surgery, aggrandized the novelty of their positionality (e.g., how “lucky” consumers were to receive the chance of surgery), and reconstructed expectations of surgery as a matter of obtaining social and romantic capital – which consumers were suddenly at risk of losing should they back out at that point. This typology of networking pathway gained its power from being elusive, rooted in an informal hierarchy of dominance and subordination that emerges with continued interaction over time (the allocation of high-status to past consumers, the influence consigned to clinicians, and desire for status rife among low-status non-consumers), and in dynamism (following surgery, formerly low-status non-consumers apotheosized into high-status consumers who then took up the mantle of inducting new others into the practice). These observations were triangulated and verified by my personal experiences in the field. I understood not only the factual credence of these networking pathways, but felt their emotional sway, powerfully capturing the gravity vested in the shifting meanings of surgery and the ability of the hard-to-reach population to survive.

I conclude with a discussion about the social and ethical boundaries encountered in this project. The ethnographic toolkit can be applied to any ethnography of hard-to-reach populations. Though I spent a considerable amount of time in the field, as with conventional ethnographies, what is key is not the duration of the ethnography but how successful one is in scoping out the networks that sustain hard-to-reach populations, which may require greater or lesser amounts of time depending on the subject. The social boundaries of hard-to-reach populations are closed in theory, but never in practice so long as they accept new members. The ethnographic toolkit, in systematically unpacking the network structures that siphon members of hard-to-reach populations in and out of different social spaces, aims to expedite this research process. Reflective of this fact, I did not typically interview surgery participants more than once, especially in the field.

On this note, clinics wanted to take part in my research project to learn more about their potential customers. It must be noted that the data was shared with clinics, but also with customers. This ensured the findings were accurate and offered transparency for both parties, so that neither felt like they had an edge over the other. Furthermore, I was never covert in the research process, always bringing to light my identity and aims as a researcher when engaging in any conversation with any participant. My interest in surgery for myself, on that note, was not feigned, but originated in a genuine place of curiosity. Though it did bring me closer to the participants, this was evidence of “going native” or the process of my identity becoming embedded in the field and its dynamics, which proved to be an important source of understanding of the emotional valences that participants experience in the solicitation of surgery.

This, ultimately, was a methodological limitation of the project and the toolkit. Though autoethnography may help the researcher detect and recognize the psychological effect of “going native” when it happens, it may not prevent it altogether. Future research should investigate how this might be accomplished, perhaps through reliance on third-parties or research teams that provide the researcher with sources to triangulate their theorizations and methodological notes with.