Qualitative Sociology

, Volume 34, Issue 1, pp 101–119

Sympathy Work: Identity and Emotion Management Among Victim-Advocates and Counselors


    • Department of SociologyFurman University

DOI: 10.1007/s11133-010-9177-6

Cite this article as:
Kolb, K.H. Qual Sociol (2011) 34: 101. doi:10.1007/s11133-010-9177-6


Advocates and counselors who assist victims of domestic violence and sexual assault often claim a moral identity that depends, in part, on their ability to sympathize with their clients. However, when their clients behave in ways that staff members perceive as “difficult” (i.e., lie, return to their abusers, break rules, express anger at those trying to help them, or fail to show up for appointments), feelings of sympathy can begin to wane—creating a moral identity dilemma. Data collected from participant observation and in-depth interviews outline four generic processes that advocates and counselors engage in to overcome this dilemma. These findings highlight the interdependent nature of identity and emotion management and contribute to previous scholarship on how those who claim a moral identity negotiate feelings that run counter to their identity code.


IdentityEmotionsDomestic violenceSexual assaultVictim-advocacy


Christina sighed, leaned back in her chair, and began to describe the behavior of one client she characterized as particularly “difficult.” Initially, she had helped the client regain custody of her daughter—only to watch her relapse into illegal drug use and lose custody again. Later, Christina admitted her back into the shelter on the condition that she meet with a counselor and a social worker—appointments the client made, but did not keep. Christina knew that healing from domestic violence took time and was willing to forgive the client, but her patience for this client was beginning to wane. Finally, after learning that the client was using drugs in the shelter, Christina told her that if she did it again, she would be asked to leave:

It can be frustrating....[S]he came in, we did the screening, and we took her back in. And we said, “okay, we’re going to start over.” And, she may have stayed a week....She started breaking all the rules in the shelter. She ended up leaving and she never...she didn’t even get her things, she just left.

Looking back, Christina remembered that she used to feel differently about her work and her clients. Indulging in a bit of nostalgia, she said “I want to feel that feeling again.”

Advocates and counselors who assist victims of domestic violence (DV) and sexual assault (SA) consider themselves good people who help others in vulnerable situations. However, their identity is dependent, in part, on their ability to sympathize with their clients. When their clients behave in ways that staff members define as “difficult”—lie, return to their abusers,1 break rules, express anger at those trying to help them, or fail to show up for appointments—they break norms regarding how victims should behave (Holstein and Miller 1990) and make it more difficult for staff members to see them as innocent and blameless (Clark 1987, 1997; Dunn 2001, 2002; Loseke 1992). As a result, staff members may find themselves unable to feel and display sympathy for the same people they have chosen to help, thus creating an identity dilemma.

Advocates and counselors in DV and SA agencies can point towards their work with victims as evidence of their “moral identity” (a social role—such as mother or pastor—that defines someone as a good person) (Kleinman 1996, p. 5; see also Katz 1975). Like social workers (Loseke and Cahill 1986), their job indicates that they are caring people. Additionally, the troubling emotions they experience—anger, fear, loss, and pain (Campbell 2002)—serve as additional evidence of their capacity to care. However, in some cases, it is the lack of a particular emotion (sympathy) that can be the most troubling of all. Their ability to feel and display sympathy for victims sets them apart as good people. Without sympathy, advocates and counselors have trouble adhering to their “identity code”—“a set of rules or enabling conventions for signifying [an] identity” (Schwalbe and Mason-Schrock 1996, p. 125). In this paper, I outline this moral identity dilemma and describe four ways that advocates and counselors are able to overcome it.

Intersection of Emotion and Identity Work

For advocates and counselors, sympathy indicates not only an understanding of how their clients feel (empathy), but an additional recognition that their clients’ “hurt or anguish or worry” (Clark 1997, p. 44) is undeserved and unfair. As a result, “empathy alone is necessary but not sufficient for sympathy” (emphasis in original) (Clark 1997, p. 35). Feeling sympathy (and displaying those feelings through words and gestures) confirms their moral stance relative to abuse: they believe it is painful and wrong. Consequently, an inability to display sympathy puts their claims to a moral identity in jeopardy and potentially means losing access to the “psychic and material rewards” (Schwalbe and Mason-Schrock 1996, p. 126) that accompany being recognized as a caring and compassionate person. Because victim advocacy and counseling typically yields meager earnings and public prestige, a moral identity offers significant symbolic compensation for selfless work.

The emotion and identity work of advocates and counselors has consequences for their clients, too. If clients cannot elicit sympathy from staff members at DV and SA agencies, they are even less likely to do so among the wider public. Advocates and counselors operate as “moral entrepreneurs” (Becker 1963, pp. 147–163) who signify their clients’ worthiness for help. Their displays of sympathy serve as evidence that their clients’ victimization is “no fault of [their] own” (Loseke 2000, p. 48) and that—despite their “difficult” behavior—they are “innocent and therefore blameless” (Dunn 2001, p. 288; see also Leisenring 2006). Clients who cannot adhere to the behavioral expectations for victims (Best 1997; Dunn and Powell-Williams 2007) not only threaten their access to sympathy (and services) in DV and SA agencies (Loseke 1992), they also risk having their claims denied in the criminal justice system (Dietz and Martin 2007; Dunn 2001).

Advocates and counselors want to sympathize with their clients and see themselves as caring and compassionate people. However, claiming a moral identity does not guarantee that others will accept those claims (Goffman 1959). Consequently, we can expect advocates and counselors to engage in “identity work” (Snow and Anderson 1987) to project an image of themselves to others, gauge how others interpret their identity, and tailor their responses accordingly. The purpose of such “work” is to “establish, change, or lay claim to meanings as particular kinds of people” (Schwalbe 1996, p. 105). Yet, identity work is not a singular event or an individual project. Schwalbe and Mason-Shrock argue that identity work is primarily a group process, or “something that people do together” (1996, p. 114) on a continuing basis. Thus, if the identity code of advocates and counselors entails offering sympathy to their clients, we can expect them to work together to remind one another how they “should” be feeling—especially if they find their sympathy reserves lacking. I argue that this process includes both identity and emotion management.

Typically, identity dilemmas are analyzed from the perspective of contradictory behavior that appears, at least on the surface, to be inauthentic; or, in Erickson’s terms, displays a lack of commitment to “one’s self-values” (1995, p. 127). As a result, when a person behaves in ways that run counter to their values, we can expect them either to alter their behavior or offer “accounts” (Scott and Lyman 1968) to explain any perceived inconsistencies. Yet, feeling the “wrong” way can pose a potentially greater threat to one’s identity claims than contradictory behavior. Inappropriate emotional displays may not be obvious to outsiders, but insiders know different. Those who share an identity are especially attuned to subtle inconsistencies within the group. As a result, group members may “police” one another’s emotions “to ensure the integrity and survival” of their shared identity (Schwalbe and Mason-Schrock 1996, p. 127). Hochschild defines this process as aligning one’s emotions with the prevailing subcultural “feeling rules” (1983, pp. 118–119). These rules outline which emotions “real” members should feel under particular circumstances. Smith and Kleinman, for example, in their study of socialization in medical school, highlighted the strategies students used to manage away “unsettling reactions to patients and procedures” (1989, p. 57) in order to feel like professionals.

Previous scholarship has outlined the ways that organizations who hire and train workers based on their claims to a moral identity sometimes ask more of them than they are able to accomplish. In Copp’s (1998) study of a sheltered workshop for people with developmental disabilities, she found that instructors were often unable to infuse their clients with positive feelings given the structural conditions of their job (boring, repetitious work beset by long periods of “down time”). In Kleinman’s (1996) study of a holistic health center, the volunteers for the organization saw their sacrifices for the organization (taking extra shifts, sometimes without pay) as indicators of their moral identity; yet, they became angry and frustrated at their lack of influence compared to the paid practitioners. In both these studies, we can identify these predicaments as rooted in organizational arrangements that tax members’ ability to feel “good” about their work and themselves. They responded to these problems in patterned ways. Copp’s (1998) instructors peacefully increased their control over their clients—with limited success; and Kleinman’s (1996) volunteers individualized their frustrations rather than question unequal arrangements in the organization. Both solutions managed—but did not completely solve—their problem. One reason for this was that their options were constrained by their desire to maintain their claims to a moral identity.

Although frustration at coworkers or daily tasks can make it difficult to accomplish a moral identity, a lack of sympathy for clients can present an even more challenging scenario. Joffe found that abortion counselors, despite their initial “sympathetic pose toward client groups” (Joffe 1978, p. 113), became angered by cynical clients and repeat aborters. The emergence of such unexpected emotions shows how “clients hold considerable leverage in determining whether such work can be experienced by counselors as morally suspect or ‘heroic’” (1978, p. 119). Wolkomir and Powers’ (2007) study, also of abortion clinic staff members, described how difficult patients evoked undesired feelings that challenged their perceptions of themselves. Although “easy” patients were a source of satisfaction, “hostile,” “ambivalent,” and “hard” patients (2007, pp. 162–166) made it difficult for staff to “feel like good helpers” (2007, p. 165). As a result, staff responded by “detaching” emotionally and “building boundaries” around the most troubling aspects of their interactions (2007, pp. 162–164). Their experiences show that “[h]elping others can be a risky business” (2007, p. 155), especially when it comes to affirming a moral identity in the absence of sympathy.

This study builds upon past research that has outlined the ways that those who claim a moral identity overcome their particular identity dilemmas. Wolkomir (2001, 2006) showed how gay and ex-gay Christian men were able to learn how to “feel like a good Christian and to manage all of the negative emotions that arose during the learning process” (2006, p. 119) of accepting their new “selves.” Group leaders engaged these men in “emotion work” (Hochschild 1979, 1983) “to teach the men how and what to feel” (Wolkomir 2006, p. 120) until they believed their emotional experiences were consistent with those of authentic Christians. Yet, what if these men had been unable to evoke the feelings necessary to affirm their identity? Holden’s (1997) study of homeless shelter volunteers partly answers this question. When required to enforce arbitrary policies—work they interpreted as inconsistent with their moral identity as “egalitarians”—the volunteers responded by re-fashioning their clients’ identities as immature and in need of structure. This strategy allowed them to justify their strict enforcement of shelter rules (such as requiring “lights out” at 11 p.m. and confiscating food in bedrooms) and still see themselves as selfless and kind people. Yet, Holden’s study does not focus on how the volunteers elicited the emotions needed to manage this identity dilemma.

My paper attempts to extend the analysis of these studies to outline four generic processes whereby those who claim a moral identity can affirm their identity when confronted with an absence of sympathy. As expected, generating additional sympathy for victims is relatively easy, even when they behave in “difficult” ways. The participants in this study did this in two ways: one, by offering accounts to deflect blame away from their clients; and two, by reconstructing their clients’ “victim-biographies” to make up for missing information that could justify their “difficult” behavior. When these strategies failed, the advocates and counselors struggled to reconcile their lack of sympathy with their moral identity. They responded by—one—enforcing a micro-hierarchical relationship between themselves and their clients as well as by—two—calling attention to their clients’ inappropriate emotional displays. These two refusal strategies allowed them to frame their denial of sympathy as in the best interests of either the client, the reputation of the agency, or the DV and SA movement as a whole. By engaging in these four processes, the advocates and counselors in this study were able to point to either their generation or refusal of sympathy as in indicator of their moral identity.

Methods and Setting

The data in this paper are part of a larger ethnographic project. To learn about work inside a DV and SA agency, I first trained as a volunteer at a nearby agency for 35 hours during nights and weekends over a period of 4 weeks (in addition to two, half-day “shadow” shifts). After training, I answered the crisis hotline and met with walk-in clients once a week for 3 months.

Through the contacts I established during my volunteer work, I secured an informational interview with a similar, although much larger, agency. I requested permission to study staff members while they interacted with each other and clients. All staff members responded positively to my request. They were eager to share their experiences with someone interested in their agency (partly because they found it difficult to find people outside the agency willing to talk to them about their work). They granted me nearly unlimited access and allowed me to take part in client sessions, observe staff meetings, and ask questions as they worked. I also participated in a number of activities: over a period of 14 months at the agency, I answered phones, made photocopies, ran errands, and on a few occasions I accompanied clients in court to seek signatures from a judge for their protective orders.

Stopping Abuse in Family Environments (SAFE)2 is an agency that helps victims of DV and SA. At the time, June 2005 to August 2006, the agency had an annual budget of nearly one million dollars (with 80% from municipal, state, and federal grants; and 20% from private donations) and served a local population of roughly 60,000 people within a 700 square mile area. Highly regarded by neighboring social service organizations, the agency won a regional award for outstanding service the year this study was conducted. At SAFE, most clients were white, working-class women, although I observed SAFE advocates assist African American and Latina women, as well as those from middle- and upper-class backgrounds. Men were welcome to seek services, but these instances were rare. Also, cases that involved DV outnumbered those of SA.

The agency employed between 25 and 30 people who offered a variety of client services: a crisis hotline, court advocacy, counseling, support groups, and an emergency shelter. Off-site services also included an educational intervention program for abusers, community awareness projects, and a youth educational program. The data for this study come from the advocates and counselors in the main office and the shelter (all of whom I refer to when I use the label “staff member”).

I visited SAFE (and their satellite offices) an average of once a week, for 14 months. As a participant and observer, I was present in the main office, shelter, or court about 7 hours per visit. Staff members introduced me to clients as someone who was researching the agency and shadowing them in the office. I took notes throughout the day (which I expanded and typed immediately afterwards). I also conducted 14 interviews, averaging between 75 and 90 minutes each, using a semi-structured interview guide (Lofland et al. 2006). Interview questions probed staff member’s understanding of their jobs, their clients, and the emotions they encountered in the course of their work. I digitally recorded and transcribed each interview.

All questions and observations during fieldwork were guided by an inductive, grounded theory approach (Charmaz 2006). I was constantly “looking and listening...participating and asking” (Lofland et al. 2006, p. 18) in order to identify themes as they emerged. To do this, I used a “theoretical sampling” method (Corbin and Strauss 2007, p. 143), which allowed me to follow up on cases involving clients labeled “difficult” and seek out opportunities to ask staff members how these cases made them feel. Operating from a symbolic interactionist perspective (Blumer 1969; Mead 1934), I paid close attention to the meanings staff members ascribed to their actions, including how those meanings were subject to refinement and contestation during interactions. Once it became clear that staff members were able to generate and refuse sympathy to some clients yet still claim a moral identity, I began constructing “memos” (Charmaz 2006) to analyze these data early in the project.

The Emotional Subculture at SAFE

The advocates and counselors at SAFE knew that signing on with the agency meant joining a unique “emotional subculture” (see Clark 1997, pp. 185–193; Scott 2005) that outlined the “rules about the feelings workers should and should not, feel and display” (Martin et al. 2007, p. 46). Yet, policy alone cannot dictate how much staff members will sympathize with their clients in everyday interactions, as Clark explains:

Many occupations have their own rules for creating sympathy margins. The rules may be imposed by management, yet workers also shape and modify the rules informally as they interact with clients or customers and each other (1997, p. 185).

As a result, the feeling rules at SAFE were shaped by forces inside and outside the agency. Below, I outline three primary influences: essentialist notions regarding women’s emotionality, DV and SA agency policies, and dominant cultural assumptions regarding victims’ claims to sympathy. These were not the only factors, by any means; but the influences of these three factors were the most visible when observing staff members’ emotional displays in practice.

All of the advocates and counselors at SAFE were women. As a result, an absence of sympathy could jeopardize not only their moral identity, but their claims to femininity as well. Women are typically cast as “emotional experts” (Martin et al. 2007, p. 46) whose job it is—paid or unpaid—to manage the feelings of others (Hochschild 1983). These norms are so pervasive that Shields argues that “emotionality” is the “master stereotype” for all women (2002, p. 45). Although sociologists of emotion disagree with this essentialist logic, the women of SAFE did not. They embraced the folk belief that “women are ‘by nature’ more sensitive, understanding, and compassionate than men” (Clark 1997, p. 51). They routinely referred to these qualities as an asset in this line of work—meaning that an inability to sympathize with clients was especially troubling for them on two levels: as staff members and as women.

On paper, the feeling rules at SAFE—what to feel, under which conditions, and to what extent (Hochschild 1983)—were set by organizational policy: a victim deserves sympathy, no matter what, and “difficult” behavior is a product of abuse, thus “[h]er unreasonable behavior is not her fault” (Loseke 1992, p. 22). Historically, DV and SA agencies have played a critical role in constructing this image of abused women (Martin 2005). These agencies have painted an image of victims as “pure, innocent, blameless, and free of problems (before the abuse)” (Lamb 1999, p. 108) whose victimhood is “morally unambiguous” (Best 1997, p. 13). Davies, Lyon and Monti-Catania (1998) argue that this depiction of victimhood was essential to persuade community leaders to accept and support the establishment of battered women’s shelters and rape crisis centers. Kelly, the SAFE co-director in charge of hiring, designed her hiring and training policies around this constructed image of victimhood. She trained staff members to forgive clients their failures on the basis that their behavior was a product of being deprived of “power and control” over their lives. This rationale, common among agencies like SAFE who adopt a “Duluth” model of victim services (Pence and Paymar 1993), emphasizes patience and compassion until the root cause of “difficult” behavior—abuse—ceases. In practice, this meant screening out applicants in the hiring process Kelly believed were unwilling to adopt a feminist analysis of abuse as a social, rather than individualized phenomenon. To gauge whether job candidates were capable of a feminist understanding of DV and SA, she looked for particular characteristics in the interviewing process: “I’m looking to see if they have the passion for [this line of work], if they come from a really committed, dedicated place, where they know this is hard work....And that they have compassion for people, in general...[and] are non-judgmental.” To stay at SAFE, new hires had to convince Kelly that they were, in her words, “warm and compassionate and understanding” people. There were practical reasons for this. Kelly believed that offering clients a continual flow of sympathy kept them coming back—even when they made mistakes. At the same time, Kelly also trained staff members that too much sympathy in the face of defiant or overly “difficult” behavior could prove counterproductive and non-“empowering.” Given these conflicting instructions, it is unsurprising that SAFE staff members were occasionally unsure of how to respond to “difficult” clients.

Decisions regarding sympathy in the SAFE office were also shaped by dominant cultural portrayals of victims and how much sympathy they deserve. Unlike DV and SA agency policies, the wider public “still holds the victim responsible for preventing abuse or leaving the relationship” (Berns 2004, p. 28). Outside of agencies like SAFE, victims are still characterized by “low self-esteem, passiveness, and weakness” (Berns 2004, p. 27). Lamb found that lawyers, judges, and police officers expect victims to be “polite, composed, deferential” (1999, p. 118), and Dunn (2002) and Konradi (1996) found that victims who defy these norms risk having their legal claims dismissed. SAFE staff members often critiqued these popular notions, but their interactions with clients revealed that they were not entirely immune to them. Clark sums up these dominant cultural feeling rules when she states that even those most eager to sympathize with others still have limits; they are “are nice—up to a point” (Clark 1997, p. 183). They may be quicker to offer “excuses and justifications” (Clark 1997, p. 183) for others’ “difficult” behavior, but they still set boundaries and expectations for them. Although SAFE staff members were more forgiving of “difficult” behavior than most, their patience was not infinite.

Given the competing influences on the emotional subculture at SAFE, it is understandable that the advocates and counselors occasionally had trouble displaying unlimited amounts of sympathy for each and every client. In the cases of particularly “difficult” clients, they confronted multiple and sometimes contradictory sets of feeling rules—as women, as SAFE staff members, and as constituents of the prevailing dominant culture.

The Moral Identity Dilemma of Advocates and Counselors

“Difficult” behavior created a problem for the advocates and counselors of SAFE. They faced an emotional dilemma—i.e., experiencing “‘forbidden’ emotions that they are unable to manage” (Martin et al. 2007, p. 46)—which had consequences for their identity. On the one hand, they were called upon to display sympathy unconditionally. On the other hand, they feared that unlimited gifts of sympathy, no matter how “difficult” their clients behaved, could do more harm than good for their clients, the reputation of the agency, and the DV and SA movement as a whole. When clients exceeded their limits on “difficult” behavior, SAFE staff members were left to create their own strategies to manage this moral identity dilemma.

Complicating this task, policies at agencies like SAFE generally warn against “tough love” strategies (such as scolding or reprimanding) on the basis that those actions mimic the callous or infantilizing behavior of abusers (Martin 2005). Many DV and SA agencies nation-wide (including the one in this study) avoid intervening or telling their clients what to do. Borrowing a concept from “Duluth model” intervention programs for abusers, staff members in this study defined abuse as the act of denying another person “power and control” over their life (Pence and Paymar 1993). Under this rationale, staff members believed that demanding their clients obey demands (no matter how well intentioned) was counter-productive, harmful, and non-“empowering.” As a result, it was frowned upon at SAFE to tell clients how to behave.

Being on friendly, egalitarian terms with their clients helped the advocates and counselors derive satisfaction and meaning from their work. They wanted to see their advocacy and counseling as more than just a “job.” Heather, an advocate, was outspoken in this regards: “I know what I do and I love what I do. And it’s my calling in life.” Melissa, a counselor, echoed her sentiments: “I never have to worry about what I’m doing for the world. You know, on my death bed, I don’t have to think back and say, ‘what did I do?’” At SAFE, their moral identity meant they were willing to make sacrifices for their clients. For example, when I mentioned to Heather that a local defense lawyer had subpoenaed confidential files at another DV and SA agency, she responded that she would defy such a court order (if given the opportunity). Jesse, overhearing the conversation, nodded in affirmation. When I joked with Heather that I would not write down her answer in case my fieldnotes were ever subpoenaed, she responded proudly: “You can write it down. I don’t care. I’ll go to jail for it.” Following Heather’s lead, Jesse announced “me too.” Staff members positioned themselves as different—more caring, more compassionate—than outsiders. As a result, generating sympathy for “difficult” clients was the first, and most obvious, solution when moral identity dilemmas arose.

Generating Sympathy

In most cases, staff members had little difficulty sympathizing with clients. Knocking on the office door or calling the crisis hotline almost always entitled clients to a provisional victim status, guaranteeing them enough sympathy to withstand most of their “difficult” behavior. As Holstein and Miller (1990) explain, being a victim typically shields one from blame for most, if not all, of one’s mistakes. This was typically the case at SAFE: staff members did not expect all clients to be perfect. Over 14 months, in private interviews and discussions, they identified few clients as “ideal.” Many clients behaved in ways that the staff members described (in private) as unhealthy and counterproductive, yet staff members were usually able to see past small missteps.

Clients whose “difficult” behavior persisted, however, taxed staff members’ capacity for sympathy. In response, the advocates and counselors sought out ways to generate additional sympathy for them. Of course, generating particular emotions was not their explicit goal. Rather, they saw keeping the flow of sympathy open as part of their job. They believed that listening to clients and acknowledging their point of view helped create positive working relationships, thus making it more likely their clients would return to seek more services if needed. When analyzed from the perspective of identity and emotion management, however, we can recognize that this process also served a latent function: generating sympathy for clients helped the advocates and counselors affirm their moral identity.

The following processes of sympathy generation are representative of staff members’ interactions with “difficult” clients. Below, I present these strategies (deflecting blame and reconstructing clients’ victim-biographies) as distinct categories, but there was considerable overlap between them; in some cases, staff members employed both strategies at once.

Deflecting Blame Elsewhere

Of all the strategies to generate sympathy for their clients, deflecting blame was the easiest and most readily available. As long as staff members could identify an external cause for their clients’ behavior, they could help their clients accomplish an “innocent and blameless” identity (Dunn 2001) worthy of additional sympathy.

Abusers were typically the first target for blame. Yet, this became more difficult when clients returned to their abusers. Doing so placed staff members in an uncomfortable position. On the one hand, they saw their job as “empowering” their clients and supporting their decisions—not telling them what to do. On the other hand, staff members worried that some clients might not recognize the potential risks involved. These findings are consistent with what Dunn and Powell-William found: “despite training specifically designed to recognize battered women’s agency and to honor all the choices they make (including “staying”), victim advocates have considerable difficulty doing this” (2007, pp. 977–978). One of Cathleen’s clients, for example, developed a pattern of leaving her abuser and then returning to him a few weeks later. Cathleen (an advocate) became frustrated. After expressing her dismay to the other advocates in the office that day (Heather and Jesse), Cathleen still managed to deflect blame away from her client and towards the abuser. Cathleen explained to her office mates: “[S]he keeps coming back [to SAFE]. And I just get frustrated. Not so much at her, but really it is with him...[that] he has that level of control.” Cathleen was able to generate sympathy for the client by placing the blame “with him.” Her coworkers in the office that day also helped Cathleen in this regard by reminding her that abusers often exert financial control over their victims and cut off avenues of support from their families, making it harder for them to escape (and stay away). This client was poor, like most SAFE clients. Thus, a financial explanation was easy for Cathleen to interpret and act upon. Collectively, by focusing the blame on the abuser, they framed this client’s behavior as typical of a victim whose abuser denies her “power and control” over her life—making her eligible for more sympathy.

When they had exhausted the sympathy generated by blaming abusers, SAFE staff members looked elsewhere for potential causes of their clients’ “difficult” behavior. In one case, Heather found an unlikely source: her client’s mother. The client had accused Heather, in court, of forcing her to file for a protective order against her will—a serious offense. Even as Heather was being scolded by the judge in open court, she did not blame the client or express frustration. Afterwards, Heather explained how thinking about the mother’s influence on her client allowed her to “forgive and forget” the incident in court.

Heather’s client stretched the advocates’ abilities to generate sympathy to the limit because of the egregiousness of the “difficult” behavior. Public accusations like this harmed the advocates’ reputation in court. However, redirecting blame onto the mother in this case was relatively easy given that Heather’s explanation fit seamlessly with the dominant cultural image of victims of DV and SA as compliant and impressionable. Given the client’s history of abuse and subsequent vulnerability, Heather explained that she was in no position to object to her mother’s pressure. As a result, she believed it was reasonable—in some ways preferable—that the client blame her rather than her own mother. Heather explained that she had expected early on that the mother might “push” the client to file for a protective order before she was emotionally prepared to confront her abuser in court. Deflecting blame away from the client enabled Heather to generate sympathy for her client, and take pride in her ability to withstand the judge’s criticism.

There’s a point when victims are ready [to file for a protective order] and when they’re not...and she’s not ready yet....That’s okay, it is part of the job…it takes the heat off [the client] and it makes it easier, especially with abusers, to say that we made them do it, we made them take out the charges....It’s frustrating [smiles], but that’s what an advocate does.

By not putting up a fight in court, Heather believed she had provided a valuable service.

Word of the case spread quickly in the SAFE office. Other staff members worried that the client’s accusation might have upset Heather. One by one, they poked their heads into the advocates’ office and asked Heather how she was feeling. When other staff members began to express frustration at the client, Heather reminded them of her earlier suspicions that the client’s mother did not understand the value of waiting until her daughter was ready to initiate legal procedures. Throughout the rest of the office, other staff members consistently expressed their admiration at Heather’s ability to take “heat” for her client. Ironically, being berated by a judge enhanced Heather’s status in the office that day. Given the opportunity to contradict her client in public, she refused. Her public display of sympathy (stoically accepting blame) bolstered her claims to a moral identity.

Reconstructing Clients’ Victim-Biographies

Although deflecting blame from clients for their “difficult” behavior was usually sufficient to generate sympathy for them, some cases called for additional measures. Often, staff members lacked sufficient information to understand why their clients behaved the way they did. In response, they wove a generic narrative of victimhood between the bits and pieces of information they did have. The end result of this biographical work (Holstein and Gubrium 2000) was the reconstruction of what I call a “victim-biography.” They did this by “selecting and highlighting the defining aspects” (Holstein and Gubrium 2000, p. 157) of their clients’ past to reformulate a story line consistent with an “ideology of victimization” (Best 1997, p. 10) that depicted their clients—in spite of their recent behavior—as innocent and blameless.

This biographical work most often occurred when clients failed to show up for appointments or court hearings. Their absence evoked feelings inconsistent with staff member’s moral identity. For example, Andrea, a counseling intern who spent 6 months at SAFE as part of her graduate degree program, became discouraged when her client continued to skip her counseling sessions: “I don’t take it personally, but it is kind of frustrating.... The reason I’m doing this is to help people, and [the client] is not allowing me to help [her].” Soon after, however, Andrea was able to transform her frustration into sympathy. She did this by framing her client’s behavior as typical of all clients:

I think it is just the nature of the clientele...they are in crisis, and they are dealing with their immediate concerns. A lot of these women have had different experiences in life [than me]...abuse and putdowns....They don’t believe that anything good is going to happen [to them].

Without a precise explanation for her client’s behavior, she inserted a generic one. This provided a context for Andrea to act upon her feelings, allowing her to generate sympathy—and affirm her moral identity in the process.

The more “difficult” the behavior, the more extensive the repair work to the client’s victim-biography. In one case, Jesse (an advocate) began to lose her ability to forgive a client’s outbursts over the phone. She took a call on the crisis hotline, but put down the receiver without a word a few minutes later. She looked confused and frustrated. The client had hung up on her. Shaking her head, Jesse told me that Meg and Kelly had warned her that this was a “draining client.” I asked Jesse to explain what this meant to her:

I mean, the client first says that she needs to go to a doctor’s appointment because of an eye problem, and later she says that she has chest pains. I’ve taken first aid classes before and I know that chest pains are more serious than an eye infection, so I said she should call 911 if she has chest pains, and she hung up on me.

It became clear that Jesse’s patience was waning with this client. In the weeks before the hang-up, the client had been calling repeatedly to request transportation to and from doctors’ appointments. Jesse, and the other advocates, began to complain that it was becoming more difficult to “feel sorry” for this client because she was so demanding. A week after the hang up, Jesse was given a message to call the client; she let out a deep breath and exclaimed—with a mix of humor and frustration—“Someone save me from [this client]!” She did not return the call. From Jesse’s standpoint, her sympathy reserves for this client were near depletion.

To generate additional sympathy, Jesse and her co-workers began to reconstruct the client’s victim-biography. Specifically, the advocates went back through the client’s file and re-interpreted her outbursts alongside a generic narrative of victims’ resilience and persistence. The client had a compelling history of abuse. Her eye infection, for example, was a result of being hit in the face while being raped by two men. Her transportation issues were caused by her poverty and her rural address far from public transportation. She had asked her landlord for help when she felt afraid, but he responded with a threat to evict her if she kept calling the police. Jesse still characterized her as “draining,” yet this label began to take the form of a symbolic badge of honor. Rather than diminish her claims to sympathy, her hang-ups and outbursts were now seen as evidence of her perseverance—thus offering an explanation for why she kept asking staff members for more and more help. In the end, Jesse was able to interpret the client’s behavior as “difficult,” but understandable—and, in some ways, admirable.

In the end, Jesse called the shelter and asked them to admit the client. Because she had continued to help the client (albeit with less patience), Jesse was able to take some satisfaction from her work with the client: “well, at least there is some resolution here...at least I’ve passed her on to someone else; so yes, I do feel better.” After reconstructing her client’s victim-biography, Jesse was then able to continue the flow of sympathy for her client. Jesse’s renewed feelings—especially after facing such “difficult” behavior—served as evidence to others that she was still a caring and compassionate person.

Refusing Sympathy

Staff members believed that their ability to sympathize with their clients set them apart as good and moral people. When clients behaved in ways that put their claims to sympathy in doubt, generating sympathy was their first response. However, as “sympathy entrepreneurs,” they also “create moral boundaries distinguishing the sympathy-worthy from those undeserving of sympathy” (Clark 1997, p. 125). This meant that SAFE staff members sometimes had to decide who was “undeserving” of sympathy: this posed the greatest threat to their moral identity.

Refusing sympathy was neither an easy choice for staff members nor one that they took lightly. Yet, they saw unlimited gifts of sympathy in the face of overly “difficult” behavior as problematic. From their perspective, putting limits on how much they were willing to forgive preserved the value of their displays of sympathy. As a result, the advocates and counselors at SAFE justified their lack of sympathy on the basis of their clients’ failure to comply with the implicit micro-hierarchy between service provider and recipient or their clients’ inability to display the emotions expected of victims. They believed their refusals of sympathy in these cases were in their clients’ best interests. They saw continued gifts of sympathy in the face of persistent “difficult” behavior as a disservice, and ultimately non-“empowering.” Thus, they affirmed their moral identity not in spite of their refusal of sympathy, but rather because of it.

Enforcing Micro-Hierarchy

Although the advocates and counselors cast themselves as equals to their clients, Clark (1987, 1997) has found that sympathy typically flows downward in hierarchical arrangements. As Best (1997) argues, becoming a “victim” is a downward move in status. Consequently, accepting sympathy usually obligates the recipient to reciprocate with displays of gratitude and deference (Clark 1987, p. 300). On the surface, this prerequisite for emotional exchange would appear counter-intuitive to the advocates and counselors’ moral identity. In trainings and hypothetical discussions, they claimed that their gifts of sympathy were unconditional, regardless of how their clients’ behaved. In practice, as long as clients’ “difficult” behavior stayed within limits, the expectation that clients conform to an implicit micro-hierarchy went ignored (for similar findings, see Loseke [1992, p. 130]). Yet, the existence of such a micro-hierarchy—even if unacknowledged—was still there. Breaking the agency’s rules came to be seen as a lack of respect for the administrators of those rules. In response, the advocates and counselors began setting limits on how much they were willing to tolerate from clients. When clients refused those terms, staff members no longer saw themselves as obligated “to feel sorry” for them.

The advocates and counselors at SAFE were hesitant to treat their clients as subordinates. As Holden (1997) found, telling clients what to do runs counter to an egalitarian identity. No one joined SAFE to become “rule enforcers” (Loseke 1992, p. 123). As a result, staff members were careful to avoid appearing too quick to judge or disdainful of their clients. Instead, they proceeded cautiously before declaring a client’s behavior to be out of bounds. Before they would make such an accusation, they often went to their supervisor, Kelly, or other staff members for confirmation of their doubts. I never observed anyone at SAFE unilaterally label a client as disrespectful or undeserving of sympathy. Staff members always sought a second, third, and sometimes fourth opinion to confirm their suspicions that their client was unwilling to abide by staff members’ expectations.

The more a client had suffered in the past, the more the advocates and counselors were willing to wait before enforcing a micro-hierarchy. One SAFE client, for example, became well known among DV and SA agencies in the region for lobbying state legislators to enhance the penalty for non-fatal strangulation (making it a felony crime). The client, a poor white woman, had been brutally assaulted by her abuser, and her bruised-laden photos served as compelling evidence in support of the new law. Staff members often referred to the client as the “poster child” of the new statute. Afterwards, SAFE advocates secured a scholarship for her to attend a community college. Within months, however, the client relapsed into illegal drug use and moved in with a new abusive partner, a man whose past SAFE staff members knew well and had warned her about. Later, the client stopped attending classes, leaving her scholarship unused. In light of her past injuries and suffering, staff members initially labeled her decisions as unintentional “mistakes”; once they began to add up, though, staff members began to set clear limits, thus bringing the micro-hierarchy to light.

When staff members believed there was no end to the client’s infractions, they began to make explicit their expectations and the consequences for violating them. For example, when, a sheriff’s deputy warned Meg that the client was selling drugs again, Meg simply shrugged her shoulders. She explained that she had warned this client time and time again. Typically, the advocates and counselors do not issue ultimatums to the victims they are trying to help, but this client had exceeded their limits. Soon after, staff members began to suspect she was using her suffering to get money from SAFE. The next time the client asked for rent assistance, Jesse doubted her motives and asked Cathleen, “Is she still in crisis?” To which Cathleen replied, “Not really, but she pretends like she is for money.” Weeks later, the client was caught selling drugs and sent to jail. When a sheriff’s deputy told Meg and Lisa “she’s in jail now, and she deserves to be in jail,” they both nodded quietly in agreement. Staff members did not reach out to the client in jail (as they typically did with other clients).

A new drug conviction, a new relationship with a known abuser, and using SAFE money for drugs proved too much for staff members. In response to her persistent “difficult” behavior, they enforced the unspoken micro-hierarchy between staff member and client; subsequently, as the client defied their limits, they saw their refusal of sympathy as justified. Staff members now saw the client as manipulative (“she pretends...for money”) and framed her behavior as not only as counter-productive, but also as disrespectful. Although speaking of a client in this way could call their moral identity into question, they saw unlimited gifts of sympathy as potentially more dangerous. From their perspective, continuing to offer sympathy to this client risked their reputation with local law enforcement which could hinder their ability to help other clients.

Enforcing the micro-hierarchy between staff members and clients also helped preserve the value of sympathy within the organization. Even if staff members saw their clients as blameless for their original victimization, they still held them to account for their willingness to work on their own behalf afterwards (i.e., take their medication, show up for appointments, display effort). SAFE staff members believed that setting limits on how much difficult behavior they were willing to tolerate was in the interest of all clients (and, by extension, all victims of DV and SA). They worried that if clients could receive unlimited gifts of sympathy no matter how badly they behaved, then what symbolic gestures would be left to offer to clients who were not as “difficult” in the face of adversity?

Similar to the experience of Holden’s (1997) egalitarian volunteers, dictating the terms of acceptable behavior was difficult for SAFE staff members. Even though they framed their relationships with clients as cooperative, they became frustrated when clients were not willing to do their “fair share.” Jen, a counselor, expected effort on the part of her clients, even if it was minimal. When dealing with a client who had stopped taking her medications, she said: “If you’re not taking your meds, we are really wasting our time here.” For another client who kept missing appointments, Jen eventually stopped trying to get the client to reschedule:

I kept putting my hand out there farther and farther and she wasn’t meeting me halfway. Instead of it being fifty-fifty, it was more like sixty-forty or seventy-thirty. At this point, I just had to back off if she [was not] putting in the effort.

On the surface, “fifty-fifty” does not indicate a micro-hierarchy. Yet, it was Jen who set the terms of their relationship, not the client. In Jen’s mind, there was a good reason for this: letting a client dictate the conditions of their therapeutic relationship was unhealthy. Although she framed this arrangement with the rhetoric of “empowerment” at SAFE—or helping the client to help herself—it was Jen who decided how the client should behave. Rather than hurt her identity claims, such a refusal of sympathy bolstered Jen’s claims to a moral identity: she “empowered” her clients; she did not coddle them.

Identifying Clients’ Inappropriate Emotional Displays

Doubting a client’s claim to victimhood put the advocates and counselors at SAFE in a difficult predicament. They saw themselves as the last line of defense for victims of DV and SA—willing to believe their clients when no one else would. They understood that claims to victimhood were hard to verify. SAFE staff members defined abuse as denying someone “power and control” over their life. As a result, physical scars or bruises were not necessarily required because instances of emotional and verbal abuse leave no marks. However, in those cases, staff members did look to their clients’ emotional displays to authenticate their claims to victimhood. Despite their eagerness to believe their clients, staff members did require some evidence (albeit small) that their clients were, in fact, “real” victims.

Hochschild (1983) conceptualizes feeling rules as subcultural norms outlining the type, direction, and intensity of emotions people should feel under particular conditions. Above, I have outlined what emotional responses SAFE staff members expected of themselves. Yet, clients have their own set of feeling rules, too—that is, if they want to persuade others that they are victims. At SAFE, when clients claimed recent abuse and anticipated more, staff members expected some measure of fear (even if they were flexible in regards to its degree and consistency). They also expected anger, as long as it was primarily directed toward their abusers—not staff members. Staff members believed that continuing to offer sympathy when they could not confirm their clients’ victimization could harm not only their reputation as advocates and counselors, but also that of the agency and the DV and SA movement as a whole. Thus, counter-intuitively, refusing sympathy in these cases served as an indicator of their moral identity.

SAFE staff members still believed they were willing to excuse their clients’ breaches of feeling-rules more than outsiders. As Dunn (2001) and Leisenring (2006) have found, victims often express emotions that run counter to an “ideal” version of victimhood. The advocates and counselors understood this and took it into account when evaluating their clients’ emotional displays. In one case, the advocates and counselors were confused by a Latina client’s emotional display. She had accused her partner of recently assaulting her, but, after her hearing in court, Meg began to doubt her: “The thing that makes her different from most of our clients is that she just isn’t afraid of him.... I basically told her she wasn’t eligible for a [protective order].” After hearing a description of her behavior, Cathleen (also an advocate) confirmed Meg’s analysis: “Is she really fearful of him? If so, why is she trying to contact him?” After observing their client repeatedly violate the feeling rules of victimhood, they began to question her identity claims.

Despite their skepticism, Meg and Cathleen referred the client to the shelter. Again, they were hesitant to cut off the flow of sympathy without ample evidence to support their decision. They did not expect all clients to express fear at all times. As Dietz and Martin (2007) have found, women of color are less likely to express fear in response to abuse. However, after a few days in the shelter, the client began to display anger and aggression towards the other residents. Shelter staff members were not entirely surprised by this, either. They were aware—as Garfield (2005) found—that abused women from marginalized groups often express anger in response to abuse instead of fear. Yet, when the client gave her abuser the shelter’s phone number—one of the worst violations of shelter rules—the shelter staff members were left scrambling for ways to explain her behavior. Days later, staff members began to discuss whether they should remove her from the shelter.

Weeks later, however, new information surfaced and staff members began to employ their sympathy generating strategies. The client, who was 6 months pregnant, became delusional and threatened suicide. Staff members called 911 and the client spent 2 weeks in a psychiatric ward. When she returned to the shelter, Rebecca explained that hospital psychiatrists had prescribed her medication: “It was good for her to get a diagnosis.” Quickly, the perception of the client changed; staff members framed her inappropriate emotional displays as a product of a mental disorder. She became a victim again. They readmitted her into the shelter.

Weeks later, however, she began to express a new set of feelings inconsistent with victimhood. She voluntarily left the shelter and reunited with her former abuser. Typically, staff members could understand clients’ decisions to return to their abusers as evidence of coerced dependence or a lack of other options. Yet, this explanation soon fell short. The client appeared happy and was now engaged to marry her former abuser. In the advocates’ office, when Cathleen and Jesse heard the news, they both shook their heads in disappointment. They appeared unwilling to engage in another round of sympathy generation for the client. Jesse joked about how she might approach the client and try to stop the wedding: “No! Take [the ring] off. Throw it away!” Joking at a client’s expense was rare at SAFE. However, in this case, their initial skepticism about the case, combined with her disruptive behavior in the shelter (even if the product of a mental disorder), proved to be too much. When staff members no longer considered her to be an authentic victim (a status that entitled her to sympathy), the client was no longer shielded from criticism (or sarcasm). Instead of working to improve her situation, staff members believed the client was doing the opposite and that additional sympathy might only make matters worse. Thus, her inappropriate feelings (lack of fear, anger towards other shelter residents, and affection for her former-abuser) became the justification for their refusal of sympathy. As a result, they believed that saving their energy (and sympathy) was in the best interest of this client and all their clients—helping them affirm their moral identity.


I analyze the generation and refusal of sympathy at SAFE to show how identity and emotion management are interdependent processes. For those who claim a moral identity, whether they be advocates and counselors, “good” Christians (Wolkomir 2006), or homeless shelter volunteers (Holden 1997), knowing how and when to offer sympathy is a key part of their “identity code” (Schwalbe and Mason-Schrock 1996). I do not claim that these findings are generalizable to all staff members in agencies that assist victims of DV and SA. Instead, I outline four “generic processes” (Schwalbe et al. 2000) that explain how people who claim a moral identity, under specific circumstances, are able to point to their decisions to generate or refuse sympathy as indicators of that identity. These processes offer a few answers for how individuals and groups address and negotiate an emotional subculture with feeling rules that sometimes conflict with their identity code. Future research may identify more. An increased understanding of these generic processes can help establish “analytic generalizability” (Becker 1990, p. 240) regarding how and why particular groups use emotions to accomplish their identities.

The symbiotic relationship between SAFE staff members and their clients also offers a useful context for studying the intersection between identity and emotion management. Staff members needed victims to show themselves and others that they are caring and compassionate people, and victims need advocates and counselors to authenticate their claims of suffering to outsiders. As a consequence, a breach of feeling rules from one party could affect the identity claims of the other. Therefore, highlighting identities which must coordinate with specific counterparts—teachers and students, doctors and patients, police officers and criminals—may offer a useful platform for future research on how identity-pairs respond when one provokes the “wrong” emotions in the other.

Identity dilemmas offer a unique opportunity to conduct this type of research. The moments when identity claims are called into question—either by the subject or their audience—require more than just altering behavior. “Feeling” the wrong way poses a subtle, but more challenging obstacle to identity affirmation. Yet this hurdle also offers opportunities. Interactions with “difficult” clients served as focal points to show that their work was more than a job, it was a vocation. Staff members believed that anyone could offer sympathy to “ideal” clients (even if such clients were rare). To them, feeling and displaying sympathy to a client who accused them of lying in open court, as Heather experienced, was a truer test of their virtuosity. Yet, not all moral identity dilemmas can be solved by simply generating more sympathy. Although we expect people to align their feelings with the “emotional subculture” (see Clark 1997, pp. 185–193) as Wolkomir found (2006), how can we explain instances when participants refuse to feel as they believe they typically “should”? How can willful violations of feeling rules also be pathways for identity affirmation? Identity dilemmas provide valuable data to answer these questions.

One potential explanation for why SAFE staff members believed it was okay to refuse sympathy is that group leaders grant others permission to feel the “wrong” way. It is true that Kelly, the co-director, held considerable influence over the feeling rules of the SAFE office (through selective hiring, training, and management procedures), yet this answer falls short in two related ways. First, organizational policies are not enough to dictate feeling rules. Managers can exert only so much control when it comes to how employees should manage their emotions (Clark 1997; Copp 1998; Garot 2004; Leidner 1999). Second, group members uphold, amend, and modify their feeling rules to preserve the general meaning and salience of their shared identity. At SAFE, answers to questions on how to respond emotionally to “difficult” clients were constantly subject to negotiation and modification among staff members. As a result, decisions to generate or refuse sympathy were typically made after lengthy consultations with others. With each encounter with “difficult” clients, staff members remade the meaning of their moral identity by adding exceptions and loopholes to their feeling rules as long as they could believe that they had the interests of their clients at heart—even if this meant cutting off the flow of sympathy to some of them.

To analyze decisions of whether to elicit or suppress particular emotions, researchers must focus their attention on the local logic of those being studied. SAFE staff members believed that gifts of sympathy—even when difficult to evoke—increased the likelihood that clients would seek services (if needed) in the future, even when they made mistakes. From the perspective of the advocates and counselors, expressing frustration or disappointment with clients might turn them away, increase their shame, or prevent them from seeking services in the future. Refusing sympathy for clients also had its own logic. Being discerning about their gifts of sympathy preserved their value. Setting limits on how much sympathy they were willing to offer and when, helped preserve the reputation of the organization as a whole. Refusing sympathy in some cases also helped staff members maintain their claims to outsiders of being honest brokers of information when they acted on behalf of their clients. They were aware that critics, particularly in the criminal justice system, were eager to paint them as zealous advocates with little regard for the truth. By showing that they were able to disagree with clients—sometimes even dispute their claims—they believed they were better able to help all victims of DV and SA (not just SAFE’s clients). Ironically, under this rationale, disbelieving some clients served the greater good by making it more likely that other claims of abuse would be given thoughtful consideration.

Those who work at agencies that assist victims of DV and SA will not be surprised by the existence of “difficult” clients and staff members’ identity dilemmas. This problem does not have a simple solution. One tactic worth consideration for those in this line of work is to follow Lamb’s advice (1999) and reconceptualize the meaning of “victim” to more accurately reflect the behavior and experiences of actual clients. DV and SA agency policies that construct an image of “pure victims” (Davies et al. 1998, p. 17) do so for a reason: they are competing against unsympathetic images in the prevailing dominant culture that depict their clients as passive—or worse, willing—participants in their own abuse. Yet, the empirical reality of clients who seek services at DV and SA agencies rarely aligns with the implicit image contained within the training manuals and handouts at agencies like SAFE. Reconstructing the image of DV and SA victims is one possible solution, albeit a difficult task for overworked advocates and counselors struggling to keep up with the crisis hotline and walk-in clients. A more practical solution would be to redouble efforts to prepare staff members for moral identity dilemmas and develop new strategies on how to manage them. I have outlined four generic strategies that emerged organically, but these need not be the only ones. In either case, understanding these dilemmas as social problems—rather than individual ones—is the first step.


I use the label “abuser” rather than “batterer” because this term allows for a more comprehensive depiction of people who control, oppress, and hurt others. I also use the term “victim,” although there is a debate among DV and SA agencies about whether to use “victim” or “survivor.” Staff used “survivor” on occasion, but usually to refer to a former client who had succeeded in achieving their goals. Staff also used “survivor” to move away from the stigma associated with “victim” (Dunn 2005).


All names of organizations, places, and people are pseudonyms.



The author thanks Sherryl Kleinman, Allison Hurst, Michelle Wolkomir, Jennifer Dunn, and the anonymous reviewers for their thoughtful suggestions and comments on previous drafts of this paper.

Copyright information

© Springer Science+Business Media, LLC 2010