Managing emotions in ongoing service relationships: new intrinsic and extrinsic emotion regulation strategies from aged care

Abstract Managing emotions at work can be challenging for employees in service roles. Most emotional labor research has been conducted in settings of brief service interactions with strangers. Yet, many traditional family roles, such as aged and child care are now outsourced to personal service roles that have ongoing relationships with clients. The continuing nature of those relationships may result in employees using previously unknown strategies to regulate emotions. To develop an understanding of intrinsic and extrinsic emotion regulation (ER) strategies in service relationships, we interviewed 42 aged care employees in China. Using template analysis we found three new employee-client relationship-based intrinsic ER strategies (i.e. family appreciation, filial reappraisal and deliberate distancing), a self-based strategy (i.e. self-deterioration anticipation) and three client-based strategies (i.e. client deterioration anticipation, vulnerability justification and meaning identification) in managing one’s own emotions in service relationships. When managing others’ emotions in ongoing service settings, a new relationship-based extrinsic ER strategy (i.e. communal bonding) along with situation-based and client-based extrinsic ER strategies were identified. Employees can regulate their own and others’ emotions by crafting their relationships with clients. This study suggests that service relationship is a critical dimension of client interactions and should be considered in workplace ER research.


Introduction
Managing emotions in the workplace is challenging for many people.In service contexts, employees are often required to smile or display Culture; discrete emotions; eldercare; emotion regulation; emotional labor; filial piety; service relationships; workplace emotions particular emotions while serving customers, called emotional labor (Hochschild, 1983).Some service roles, such as child and aged care workers, school teachers and family physicians require employees to repeatedly interact with the same clients and develop ongoing service relationships (Gutek et al., 2002).Managing emotions (i.e.emotion regulation [ER]) in long-term service relationships requires not just smiling but deeper emotional involvement with clients and their families (Clark et al., 2001).For example, aged care employees must manage their own and others' emotions such as stress, worry, frustration and grief in a range of difficult situations particularly during the COVID-19 pandemic (Bassal et al., 2016;Brydon et al., 2022).
However, the study of ER in ongoing service relationships is still at an early stage (Chi & Chen, 2019), with service interactions' temporal relationships rarely investigated (Grandey & Diamond, 2010).The unique ER strategies used by employees in service relationships or whether service relationships enrich or limit the choices of ER strategies has not been established.The purpose of this study is to extend workplace ER theories by understanding ER strategies used in ongoing service relationships.Our main research question is how employees regulate their own and others' emotions in service relationships.To address this, we selected the residential aged care context because eldercare services involve high emotional demands and a range of employee-client relationships.Such service relationships are a result of repeated employeeclient interactions and shaped by cultural norms and values, for example filial piety in a collectivist society.Filial piety, referring to love, respect and obedience toward older family members, plays an important role in informal caregiving (Lai, 2010) and can extend to the whole community including formal aged care settings (Bedford & Yeh, 2019).Understanding how aged care employees use ER strategies is critical to enhance employee well-being and service quality (Clarke & Hill, 2012) and to attract and retain a qualified and sustainable workforce (Radford et al., 2015), especially when a global pandemic increases the salience of employee mental health challenge, long-term staffing shortages and high staff turnover in aged care (Brydon et al., 2022;Davidson & Szanton, 2020).
We first review theories and frameworks relating to workplace intrinsic and extrinsic ER strategies and employee-client relationships in Chinese aged care.We then report results of 42 employee interviews on the use of ER strategies.We briefly report how employees used process-based ER strategies and then report new relationship-based, self-based and client-based intrinsic strategies identified in this ongoing service context.Last, situation-based, relationship-based and client-based extrinsic ER strategies are reported.
The study makes four major contributions.First, we extend workplace ER theories by identifying new ER strategies in service relationships.
Second, the findings of new ER strategies enrich our understanding of workplace relationships and provide insights into job design and job crafting in service relationships.Third, as prior studies have been mainly conducted in Western settings, the Chinese workforce focus brings a cultural perspective to employee-client relationships and ER strategies.Last, these findings provide practical insights for organizations and managers in aged care and broader service sectors to understand employees' emotional lives to enhance employee well-being and retention.

Intrinsic and extrinsic emotion regulation strategies in the workplace
ER is important in the workplace because effective ER strategies can help individuals experience positive emotions and well-being (Côté et al., 2008), whereas ineffective ER strategies can lead to emotional exhaustion (Brotheridge & Grandey, 2002), reduced job satisfaction and decreased service performance (Hülsheger & Schewe, 2011).Managing emotions involves employees regulating their own and others' emotions, such as clients, co-workers and supervisors.ER can be intrinsic or extrinsic depending on whose emotions are regulated: The process of regulating one's own emotions is intrinsic ER, while deliberately regulating or controlling another person's emotions is extrinsic ER (Troth et al., 2018).
In an example where a client complains about a service and vents their anger toward an employee who feels frustrated, the employee's process of managing their own frustration is intrinsic ER, and the employee's process of managing the client's anger is extrinsic ER.There are a variety of established ER models and strategies as shown in Table 1.

Intrinsic ER strategies
In prototypical customer service contexts, employees are often involved in emotional labor, i.e. providing service while displaying required emotions, with three emotional labor strategies: surface acting, deep acting and genuine emotional display (Grandey & Melloy, 2017).Emotional labor strategies focus on whether the experience and expression of emotions are congruent.These strategies are used in short-term client encounters, such as by flight attendants (Hochschild, 1983), call center staff (Grandey et al., 2004), taxi drivers (Hülsheger et al., 2015) and salespeople (Wessel & Steiner, 2015).Meta-analysis shows that surface acting impairs well-being, job satisfaction and performance, but deep acting enhances emotional performance and customer satisfaction (Hülsheger & Schewe, 2011).Gross's (1998) widely used ER process model describes five strategies: situation selection, situation modification, attentional deployment, cognitive change and response modulation.The first four antecedent-focused strategies are used before an emotion is generated, while the last response-focused strategy is used after emotions are generated.Antecedent-focused strategies may increase positive experience and expression, relationship closeness and well-being, while response-focused strategies tend to decrease positive experience and relationship closeness and increase negative experience and depressive symptoms (Gross & John, 2003).Garnefski et al. (2001) categorized nine cognitive strategies into adaptive and maladaptive strategies depending on outcomes.People using adaptive ER strategies experience less negative emotions, depression and anxiety than those using maladaptive strategies (Garnefski et al., 2001).

Extrinsic ER strategies
Regulating another person's emotion involves the regulator and target in an iterative process in which one's actions shape the other's responses (Reeck et al., 2016).Although Gross's (1998) process-based ER model was developed for intrinsic ER, researchers also apply this model to extrinsic ER, such as regulating others' threat (Williams, 2007) and negative emotions (Little et al., 2012), using the same five strategies.Extrinsic ER strategies can also be categorized by potential emotional outcomes as affect-improving strategies (i.e.positive engagement and acceptance) or affect-worsening strategies (i.e.negative engagement, rejection; Niven et al., 2009).These strategies were developed with student samples (Niven et al., 2009) and validated in general working and unemployed population (Niven et al., 2011).Compared with intrinsic ER, the frameworks of extrinsic ER were developed later and need more investigation.Both intrinsic and extrinsic ER strategies have focused on the steps in a process or outcomes of regulation but rarely consider the elements related to interpersonal processes, such as interpersonal relationships.

Emotion regulation in aged care employee-client relationships
Employees in aged care experience a range of emotions, such as frustration, worry, anger and grief (Bassal et al., 2016).They need to manage their own emotions in a range of difficult and stressful situations, including clients' physical and cognitive deterioration, frequent aggression, misunderstanding and disrespect from clients or their families (Gillespie et al., 2011).Physical and psychological aggression from clients often leads aged care employees to feel emotionally exhausted (Evers et al., 2002).Aged care employees also need to manage others' emotions, particularly their clients' emotions, such as loneliness and depression (Liu et al., 2016) and emotional crises -such as feeling angry, sad and hopeless -of older adults with dementia (Aminzadeh et al., 2007).

ER in service encounters versus relationships
The intrinsic and extrinsic ER strategies reviewed above were mainly identified and studied in short-term service interactions.However, temporal relationship is a critical dimension of client interactions and should be considered in workplace ER research (Grandey & Diamond, 2010).Gutek et al. (2002) propose that service interactions may be delivered on a continuum with service encounter at one end and service relationship at the other end.Service relationships occur when a client has repeated contact with the same service provider and both parties expect repeated future contact.
Employees' emotional experience and expression differ in service encounters versus relationships.Regarding emotional experience, employees in service encounters tend to have limited feelings for clients, whereas those in service relationships tend to generate actual feelings for clients (Gutek et al., 2002).Healthcare employees in high patient turnover settings tend to set boundaries to create emotional distance from patients and their families (Hayward & Tuckey, 2011).Regarding emotional expression, employees in service encounters tend to display emotions they do not actually feel (i.e.surface acting) more frequently than employees in service relationships (Wessel & Steiner, 2015).
Motives underlying interpersonal relationships may impact the choice of strategies to regulate emotions (Tamir, 2016).Service encounter employees may have less intrinsic motivation than service relationships employees (Grandey & Diamond, 2010) and therefore may use more surface acting.People with a stronger communal motivation (i.e.caring for another's welfare) are more likely to respond to others' emotional needs and express their own authentic emotions (Clark et al., 2001).Communal motivation is evident in relationships between friends, family members and romantic partners involving prolonged and intimate interactions (Clark et al., 2017) and may apply in ongoing service relationships.

Aged care employee-client relationships
Residential aged care is a typical service relationship context because employees and elderly residents have ongoing interactions and expect to see each other frequently.Most people in the workplace are initially connected with each other through formal roles and conform to exchange norms and instrumental goals, but can later form voluntary and informal workplace friendships characterized by communal norms and socioemotional goals (Pillemer & Rothbard, 2018).Rapport in employee-client relationships benefits service performance such as increased client satisfaction and loyalty (Gremler & Gwinner, 2000).Care service researchers identified the element of relational richness (i.e. the extent to which the carer has an individualized relationship with the client) in the concept of empathetic care (Lamberton et al., 2015).Compared with hospital nursing, residential aged care employees perform intimate personal and physical daily care for residents over not just days or weeks but months and years.
Although the employee-client relationship is integral to care service quality, there is a dilemma for employees between providing person-centered care and maintaining professional boundaries (Jones & Moyle, 2016).Prolonged employee-client relationships can bring social rewards, such as recognition and support (Grandey & Diamond, 2010) and enhance employees' emotional involvement but also increase emotional demands (Clark et al., 2001).Studying ER in service relationships can provide insights to support organizations and employees in their emotional and boundary management.

Filial piety in care relationships
Cultural values and norms also impact employee-client relationships.In Confucian culture countries, such as China and Korea, filial piety, as a central value referring to physical care, love, service, respect and obedience toward parents in child-parent relationships, plays an important role in informal caregiving (Lai, 2010).A close but different concept from filial piety is respect, which can be manifested in linguistic, care and spatial respect for elderly (Sung & Dunkle, 2009).Unlike respect, filial piety is not regarded as a choice but rather an unconditional obligation to provide physical, emotional and financial support for older people in the family (Bedford & Yeh, 2019).
In a collectivist society, filial piety can extend from the family to the whole community of elders and help protect people from caregiving burden by perceiving more gains and values of providing care in family settings (Lai, 2010).Filial piety toward family members can be perceived as rewarding parents in return for their care and sacrifice when adult children were young (Yiu et al., 2021).However, this mechanism does not apply to aged care employees and their clients in formal service settings.Given the generational differences between carers and their clients and the transactional nature of aged care services, there may be ambiguity or new definitions of filial piety toward clients.That said, whether filial piety makes a difference in shaping relationships between care workers and older clients remains to be explored.Filial culture provides a rich ecological context to examine ER in ongoing service relationships.
Given the above, an investigation of how aged care employees manage their own and others' emotions at work may reveal new ER strategies based on the characteristics and underlying motives of service relationships.This study aims to investigate what ER strategies employees use to manage their own and others' emotions in ongoing service relationships.Identifying new ER strategies can guide employees and organizations on effective ER strategies and shaping appropriate service relationships in ongoing, person-centered service.Understanding the nuanced employee-client relationships is helpful for organizations to design jobs involving repeated service interactions and for employees to craft the relational and emotional job components to enhance well-being and deliver high-quality services.

Methods
Consistent with extending theory, this study adopts a qualitative approach to investigate ER strategies in service relationships.Although experience sampling method is used to investigate event-based ER (Hülsheger et al., 2015), in-depth interviewing is most appropriate to explore unique ER strategies in this service relationship context.

Data collection
Aged care facilities in China are operated by the government, non-profit organizations or private investors.Employees in ten facilities from a range of public, private and non-profit providers in metropolitan cities (i.e.Beijing, Guangzhou and Qingdao) and regional towns in Shandong and Guangdong participated in the study.Ethics approval was granted by the University Human Research Ethics Committee.
In-depth interviews with 42 aged care employees lasted about an hour, were audio-recorded, conducted and transcribed verbatim in Chinese then translated into English for analysis.The 38 females and four males aged from 25 to 58 years had one to 27 years aged care tenure.Interviewees were three general managers, eight care managers, 19 care team leaders, 10 carers and two nurses.Other than one casual worker, the remainder was full-time employees 1 .Interviewees were educated to year 9 (40.5%),year 12 (16.7%),college diploma (4.8%) or Bachelor degree (21.4%) level.
The semi-structured interviews included questions about how interviewees regulate their own and others' emotions in the workplace and the outcomes.To collect intrinsic ER incidents, interviewees were asked to describe an event or situation where they managed their own emotions.To collect extrinsic ER incidents, interviewees were asked to describe a situation where they managed the emotions of others in their workplace (i.e.clients, colleagues or supervisors).To understand the ER process, they were specifically asked about discrete regulated emotions and ER strategies.Probe questions about emotions most likely to occur in daily work were asked if interviewees had difficulty recalling any ER incidents, for example 'Please share an event or situation at work when you felt sad, angry, compassionate, happy or proud.'Detailed interview questions are in Appendix A.

Data analysis
Interview data were coded using a template analysis approach (King, 2004) with Gioia et al.'s (2013) methodology.Template analysis is a hybrid of inductive and deductive analysis and appropriate for this study as we start with ER theories and remain open to emerging themes (King, 2004).There were two steps in this template analysis.

Step 1: creating the initial template
The first step is to create an initial template based on the research question and theories shaping the study (King, 2004).In the initial template we created the following ten intrinsic ER strategy codes: five process-based ER strategies (i.e.situation selection, situation modification, attentional deployment, cognitive change and expressive suppression), three emotional labor strategies (i.e.surface acting, deep acting and genuine emotional display), two adaptive strategies (i.e.acceptance and refocusing on planning) and four maladaptive strategies (i.e.self-blame, blaming others, rumination and catastrophizing).The other three adaptive strategies (i.e.putting into perspective, positive refocusing and positive reappraisal) were not included because they are cognitive change strategies.We created nine extrinsic ER strategy codes: five process-based ER strategies corresponding to intrinsic ER and four interpersonal affect regulation strategies (i.e.positive engagement, acceptance, negative engagement and rejection).The initial template included code definitions (see Table 1).
Step 2: coding and revising the template Once the initial template was constructed, the raw data were imported into NVivo version 11 (QSR International Pty Ltd, 2015) for coding.The researchers examined each transcript and marked the relevant text with one or more codes from the initial template.Following Gioia et al. 's (2013) methodology, we conducted first and second-order analyses.
In the first-order analysis, we extracted key quotes and kept faithfully to those terms as first-order concepts, e.g.'go back to the office to take a rest or talk to other staff ' and 'clients treat me as family' .In the second-order analysis, existing ER strategies listed in the initial template were labeled as second-order concepts, such as 'situation selection' , 'cognitive change' and 'expressive suppression' .Then, we checked if the first-order concepts could be categorized into those ER strategies.For example, 'go back to the office to take a rest or talk to other staff ' was categorized into 'situation selection' .Unused codes from the initial template, such as self-blame, blaming others, rumination and catastrophizing strategies were deleted.
We checked similarities among all first-order concepts, grouped them into second-order concepts and labeled them.For example, the common theme of reappraising situations by taking filial responsibility was identified from the first-order concepts of 'thinking of clients as my own parents' , 'fulfilling the filial responsibility for society' , 'acting as clients' children' , so we labeled 'filial reappraisal' as a new second-order intrinsic ER strategy.Other second-order concepts, such as 'family appreciation' and 'deliberate distancing' were created.These three strategies were related to the relationship element, so we aggregated these second-order concepts into the 'relationship-based intrinsic ER strategies' theme.
Although some first-order concepts, such as 'take the client's perspective' and 'understand the causes of elderly's behaviors' were aligned with existing ER strategy codes (e.g.cognitive change), they were re-aggregated into new second-order ER strategies, such as 'vulnerability justification' and 'meaning identification' .The codebook was also independently coded by a management PhD student.Coding discrepancies were discussed and resolved.

Results
Key findings are now presented.Seven intrinsic ER strategies under relationship-based, self-based and client-based strategies themes are presented before the extrinsic ER findings.To protect identities, role and codes (e.g.'Carer 1') are used.'Care Mgr' and 'CTL' refer to Care Manager and Care Team Leader, respectively.

Intrinsic emotion regulation in service relationships
Both known and new intrinsic ER strategies to manage workplace emotions were reported.First, we discuss how the already known process-based strategies are used, then how strategies new and unique in this setting are used.Figure 1 shows the newly identified first-order concepts of intrinsic ER, second-order intrinsic ER strategies and aggregate themes.

Process-based intrinsic ER strategies
Using the ER strategies template, we found that interviewees reported frequent use of process-based intrinsic ER strategies, such as situation selection, attentional deployment, cognitive change and expressive suppression, and surface acting to manage their own emotional experience and expression.Table 2 shows example quotes for the process-based intrinsic ER strategies used by interviewees.
It is worth noting interviewees widely reported using expressive suppression.Carers described how they get used to suppressing their emotions when working with residents: The elderly shouted at me.I can't shout back.[I] just control those emotions and keep them to myself.I have got used to suppressing these emotions.(Carer 1) Interviewees reported a variety of reasons to use suppression, such as complying with employers' emotional display requirements, maintaining interpersonal harmony or saving their energy from ineffective strategies:

When you and your residents have a conflict, you cannot get angry with them. In a word, you cannot show [negative] emotions at work. (CTL20)
No matter how awful his [the client's] words are, you have to calm down.Because shouting at him will not keep things in harmony or solve the problem.( CTL3)

Theme 1: relationship-based intrinsic ER strategies
In addition to the above process-based ER strategies, new strategies were identified.When asked how they managed their own emotions at work, many participants spontaneously addressed a common theme -how they perceive or reframe their relationships with clients -to help manage their own emotions.The three employee-client relationship-based strategies, namely family appreciation, filial reappraisal and deliberate distancing, are described below.

Strategy 1: family appreciation
The family appreciation strategy refers to changing emotional experience by perceiving recognition and gratitude from clients.Such recognition and gratitude are often expressed in the form of being treated as family members by clients: [He] may vent his anger to me sometimes.Since I have been caring for him for so long he treats me as his family.That's why I don't take his venting personally.(Carer 8) The first-order codes 'clients treat me as family' and 'clients want me to be their daughter' emerged, showing that aged care employees perceived appreciation from the elderly clients after repeated interactions, and therefore were aggregated as the second-order code 'family appreciation' .Being treated as clients' own children, indicating appreciation and care from clients, enable carers feel more positive emotions such as happiness: Although these elderly are not my biological family, after spending so long with them, they really treat me as their family.That makes me feel happy here.(CTL27) This gentleman treated us as his own children and even would like to share his snacks with us.He laughed with us, so I felt happy too.(Carer 3) The family appreciation strategy is rooted in recognition, appreciation and reciprocal care from elderly clients.Treating carers as their children is a high level of recognition and appreciation, which may bring resources for carers to feel less negative and more positive when interacting with these clients.

Strategy 2: filial reappraisal
A unique intrinsic ER strategy found in this Chinese aged care context was characterized by appraising situations from a filial perspective.Some carers reported regulating their own emotions by thinking of the filial piety value and perceiving clients as their own parents.First-order codes 'thinking of clients as my own parents' , 'respecting elderly as traditional Chinese virtue' , 'fulfilling the filial responsibility for society' , 'acting as clients' children' were created and aggregated as the 'filial reappraisal' strategy.Filial reappraisal refers to re-interpreting situations by taking filial responsibility for the client to change one's own emotional experience and expression.Carers explained they would not get angry or hurt if their parents show the client's behaviors.Perceptions of relationships will impact how much responsibility people feel in responding to others' needs in those relationships.When employees re-interpret their relationships with clients -perceiving residents as their own parents, they experience less negative emotions: When I imagined encountering such situations in my own family, if my dad or mum lost their temper, would I take it personally?No, impossible.Then, I wouldn't take those emotions personally here too.(CTL3)

When elderly clients make us cry, we can only perceive them as our own parents [to regulate our own emotions]. (CTL2)
Filial reappraisal is an effective intrinsic ER strategy among these carers because filial piety is a critical Chinese value.Employees learn filial piety in their own family, endorse and internalize this value, and then apply it to their workplace interpersonal relationships: Respecting old people… is a key virtue in Chinese culture.(CLT20) My parents fulfill their filial responsibility for my great-grandmas.I am doing the same for my clients here.(Carer 3) We are fulfilling filial piety on behalf of all children.(CTL2) In contrast to family appreciation, filial reappraisal only involves internal cognitive processes and does not require appreciation or social rewards from clients.Employees see clients as parents, while clients may still see employees only as care workers, not as family members: We care for elderly clients very carefully, but we are different from their own children in their eyes (CTL2).
To tell the truth, for some clients, no matter how well we treat them, they still exclude us and consider us as outsiders.(CTL27) Filial reappraisal is neither suppressing external emotional expression nor re-directing one's attention to positive aspects of the situation.It can be perceived as a contextualized cognitive change strategy in this ongoing employee-client relationship context.

Strategy 3: deliberate distancing
If the first two strategies show a tendency to approach the client, then the third strategy involves the tendency to avoid the client and distance that relationship to avoid emotions.The codes 'avoiding talking with the client' , 'swapping units to avoid emotional connection' and 'keeping a distance with the client' were created and aggregated as the deliberate distancing theme, referring to being less emotionally involved by deliberately avoiding interactions or long-term relationships.Examples include a team leader describing how she likes to get to know clients first but deliberately avoids later relationships and interactions.Another swapped units to avoid deep connections so they feel less sad when elderly clients pass away:

I called the client's name, and he thought I was making trouble with him on purpose. Now we don't talk, and it is beneficial for his emotions and mine. (CTL29)
We swap units with other colleagues.I look after the elderly in this unit and will move to look after another unit in a few months.If you work in the same place with the same small group of residents, you would feel too sad to accept it when they pass away suddenly.(CTL 27) Distinct from the expressive suppression strategy, deliberate distancing changes the relational context and subsequent emotions rather than emotional expression.It has also been used as a management intervention to support employees.One manager agreed to a carer's request to avoid interactions with a client: The carer said, "I looked after him so much, but he said I didn't.I felt wronged and hurt.I will apologize to the resident, but I can't work with him anymore."There are only two people on duty, and I have to organize the other colleague to work with her to look after the resident so she can keep a distance from the resident.(Care Mgr 6)

Theme 2: self-based intrinsic ER strategy
In addition to relationship-based ER strategies, one self-based ER strategy was identified.Interviewees reported how they considered their own future to help regulate their emotions in ongoing service interactions.

Strategy 4: self-deterioration anticipation
The emerging codes 'will be in a similar condition' and 'may be in a worse situation than the client' reflected a common theme of anticipating self-deteriorating at clients' age and were aggregated as the self-deterioration anticipation strategy.This process involves social comparison as carers compare their own future condition with their current clients' conditions.Carers believe they would be in a similar or worse situation than their current elderly clients and consequently can better understand clients' current situations and reduce their negative emotions with current client interactions:

When the client gets angry, I would think that I may be in a worse situation than my current clients. Thinking in this way helps me not get frustrated. (Carer 5)
When I am old, I may not behave as well as them.They have been tortured by sicknesses such as dementia.When I am older I may suffer even worse than this.Then the feeling of being hurt is healed.(CTL26) This strategy is different from attentional deployment because it directs the focus to something negative rather than positive.It is not a cognitive change strategy because the regulator is not reappraising the current situation but comparing it with a future self-related situation.Anticipating self in similar or worse situations reminds carers they may need to be looked after by others and will expect good care when they are in a similar condition: Everyone will become old, so the situation for them today is the situation for us tomorrow.I think about when I become old someday.I will be in a similar situation or even worse than them [clients] and need to be looked after.When I shape my thinking this way, I can deal with my emotions.(Carer 3)

Theme 3: client-based intrinsic ER strategies
Beyond relationship-and self-based strategies, three client-based intrinsic ER strategies were identified: client deterioration anticipation, vulnerability justification and meaning identification.

Strategy 5: client deterioration anticipation
The codes 'thinking about a worse future condition of the client' and 'thinking about the client passing away' emerged from the interview data.By thinking about a worse future situation for their clients, such as becoming much sicker or dying, carers reported relief about their current concerns or annoyances:

Time for the elderly is limited. They appear well and energetic now but may pass away tomorrow. Thinking of this, I am relieved about my earlier annoyance. (CTL2) When I think the elderly clients would pass away without their family's company, I treat them better and feel more relieved about their annoying troubles. (CTL26)
The client deterioration anticipation strategy focuses on clients' future situations, not employees' own future situations.Self-deterioration anticipation involves social comparison, while client deterioration anticipation may involve the mechanism of compassion.

Strategy 6: vulnerability justification
Interviewees reported managing emotions by thinking about clients' vulnerability and attributing their behaviors to that vulnerability.The codes 'take the client's perspective' and 'understand the causes of elderly's behaviors' were aggregated as 'vulnerability justification' .In this aged care context, some clients with dementia or other health-related sickness or reduced cognitive ability are irritable or aggressive and not able to control their behaviors.Employees have been educated about the causes of clients' aggressive behavior, which reflect clients' vulnerability, and then use these causes to justify clients' behavior to prevent or stop themselves from experiencing negative emotions: He is sick.He does not realize you are doing something good for him.Dementia makes him confused.( CTL3) We take the client's perspective to understand and explain the situation because his cognitive ability has declined with age.(CTL26) This strategy can also be perceived as a contextualized cognitive change strategy in ongoing asymmetrical relationships where elderly clients are in a more vulnerable and less powerful position than employees.

Strategy 7: meaning identification
Interviewees also reported how they identified and used the meanings of taking care of elderly clients to change their emotional status.The codes 'learn something from the elderly' and 'clients' genuinely satisfied, happy faces bring me happiness' emerged and disclosed the meanings employees identified to help regulate their emotions, for example: I can learn a lot from the elderly.Even if they are sick, I can see many good qualities in them, such as gratitude.These good qualities make me feel moved.(CTL3)

I feel happy because of my clients. I treat the client with dementia sincerely, and her genuinely satisfied and happy face brings me happiness (CTL5).
The meaning identification strategy is different from family appreciation, which is being treated by clients as their children, bringing social rewards and positive resources to employees.The meaning identification strategy does not require explicit client recognition and appreciation but employee proactivity in seeking meaning in their aged care work.

Extrinsic emotion regulation in service relationships
Interviewees were also asked how they managed others' emotions at work.They reported using process-based extrinsic ER strategies including situation selection, situation modification, attentional deployment and cognitive change (Table 2).Beyond those already known, we identified two new strategies of communal bonding and emotional expression.These strategies have been re-aggregated into the three dimensions of situation-, relationship-and client-based extrinsic ER strategies.Figure 2 shows the first-order concepts, second-order extrinsic ER strategies and aggregate dimensions in extrinsic ER.

Theme 1: situation-based extrinsic ER strategies
When regulating others' emotions, especially elderly clients, carers have the autonomy to select or change situations for clients.Interviewees reported using situation-based strategies including situation selection and situation modification to manage their clients' emotions.

Strategy 1: situation selection
In extrinsic ER, the regulator changes the target's emotions.In this context, aged care employees are the regulator and elderly clients are the target who may be exposed to emotion-eliciting situations.Aged care employees are responsible for elderly clients and can select a less emotional-eliciting situation on behalf of the client.Interviewees reported examples of regulating their elderly clients' emotions by avoiding interactions with an emotional client or other clients' family visits: We moved the shouting client to another place in case he influenced other clients.(Care Mgr 11) Especially when other clients' children visit the facility during the festivals, some elderly clients would think about why their children haven't come.To prevent them feeling emotional about that, I would have a walk with the client to avoid other clients' family visits.(CTL2)

Strategy 2: situation modification
Carers have a range of options to modify the situation for their clients.A team leader reported how her team modified the situation that elicited the resident's emotions to help the resident feel less emotional: We immediately contacted the client's family and told them the client was emotional and missing their family.We asked the family to come to see the client.( CTL26) The client felt angry suddenly, and then we asked what he needed or if he missed home and family.To calm him down, we asked his nephew to come to visit him, or we purchased what he needed for him.(CTL21) We aggregate situation selection and situation modification as situation-based extrinsic strategies as they focus on intervening the situation of another person.These strategies are often used in aged care because clients are highly dependent on carers.

Theme 2: relationship-based extrinsic ER strategy
In addition to situation-based strategies, we found aged care employees regulated clients' emotions by building communal bonds with clients.Communal bonding, often seen among family and intimate relationships, has emerged in this service relationship context as a new relationship-based extrinsic ER strategy.

Strategy 3: communal bonding
When regulating residents' emotions, interviewees shared examples of building communal bonding with clients.The codes 'perceive ourselves as the children of the elderly clients' , 'let me know their concerns and emotions' , 'the client perceived me as his family member' and 'celebrate birthday with the elderly' emerged.These codes suggest employees manage others' emotions by building communal bonds with clients and were aggregated as 'communal bonding' .Communal bonding enables people to share their emotions and respond to another person' s emotions.For example, a team leader referred to how perceiving herself as a child of her elderly clients helped to regulate the clients' emotions: We perceive ourselves as their children, so they let me know their concerns and emotions.(CTL22) Building family-like relationships enable carers to establish communal bonds with clients to change clients' behaviors and emotional experience: After I knew him more… he perceived me as his family member.Then I directed him to do a task, watching TV or singing, and he would be absorbed and calm (CTL3).
A care team leader also referred to her experience of changing clients' emotional status by actions that follow communal norms, such as celebrating elderly clients' birthdays: Many clients said their children rarely celebrated birthdays for them at home.They felt very excited and even had tears when we celebrated birthdays for them.(CTL2)

Theme 3: client-based extrinsic ER strategies
Beyond situation-and relationship-based extrinsic ER strategies, we identified two client-based extrinsic ER strategies: emotional expression and attentional deployment.

Strategy 4: emotional expression
Unlike in intrinsic ER, interviewees did not report using response modulation to regulate the client's emotions.On the contrary, they would allow the client to express their emotions first and re-approach the client later:

Allow him to vent his [angry] emotions first and later ask him what happened. (Carer 2)
Sometimes, you have no way to deal with their emotions.Just leave them there.(CTL27) The codes 'allow the client to vent emotions' and 'leave the emotional elderly clients here' were aggregated as the new emotional expression strategy.Carers do not constrain the emotions elderly clients experience or express but leave clients to process their emotions.Emotional expression gives autonomy of managing emotions back to clients.Unlike communal bonding, the emotional expression strategy seems passive or avoidant and could sometimes be perceived as neglect of clients' emotional needs.

Strategy 5: attentional deployment
Interviewees also reported how they directed clients' attention to positive aspects and events to help them feel more positive and less negative.The codes 'remind the elderly to think about positive aspects in his life' and 'talk about happy events with the elderly' were categorized into extrinsic attentional deployment strategy.This carer described how she directed a clients' attention to positive aspects: I talked to this client, saying that you have a very happy family.Your children and grandchildren are very filial to you.They always buy great food for you.You have properties and pension after retirement.You are the happiest resident in our facility!After hearing these, the elderly client cheered up.(Carer 1)

Discussion
This study investigated ER strategies in an ongoing service relationship context to understand interpersonal and cultural influences on ER processes.Beyond existing ER strategies, we found new relationship-, self-and client-based intrinsic ER strategies that aged care employees use to manage their own emotions.We also identified situation-, relationship-and client-based extrinsic ER strategies that aged care employees use to manage others' emotions.

Relationship-, self-, and client-based intrinsic ER strategies
In this ongoing service relationship context, we conceptualized and identified new intrinsic ER strategies from three aspects, employee-client relationships, employee self and clients.The previous intrinsic ER strategies in Gross's (1998) model, emotional labor strategies (Brotheridge & Grandey, 2002) and cognitive strategies (Garnefski et al., 2001) focus on situations and employees' internal attentional and cognitive processes.The new strategies identified in this aged care context have focused on the relational context.
Three relationship-based intrinsic ER strategies suggest aged care employees regulate their own emotions by shaping interpersonal relationships with clients.These relationship-based intrinsic strategies supplement workplace ER theories by providing a different lens to conceptualize workplace relationships.Beyond relational strength, employee-client relationships can be measured by the extent to which employees and clients perceive service responsibilities (Clark et al., 2001).In service relationships, some employees tend to use the deliberate distancing strategy to manage emotions, similar to the emotional boundaries strategy, i.e. controlling emotional distance with a patient or their family, used by hospital health professionals in service encounter settings (Hayward & Tuckey, 2011).Deliberate distancing also aligns with the detached type of client interactions, in which employees typically show low empathy with and high detachment from clients (Lampert & Unterrainer, 2017).
The unique relationship-based intrinsic ER strategies identified in service relationships are family appreciation and filial reappraisal.Repeated interactions and intergenerational identities enable employees and clients to establish family-like relationships beyond traditional, transactional employee-client relationships (Clark et al., 2017).From a resource-based perspective (Hobfoll et al., 2018), regulating emotions is a resource-depleting process.The family appreciation strategy works because employees receive social rewards, such as care or recognition from elderly clients.Employeeclient relationships can provide positive resources for employees as other well-known social exchange relationships in work settings, such as leader-member exchange (Cropanzano & Mitchell, 2005).
Unlike family appreciation, the filial reappraisal strategy is not driven by social rewards but internalized cultural values.A culture with generalized filial piety addresses the children's obligation of respecting, obeying and taking care of elders in one's own family (Lai, 2010).Respecting and taking care of older people may be part of observable filial piety behaviors, but the core of filial piety is feeling obligated to meet older people's physical, emotional and financial needs (Bedford & Yeh, 2019).The filial reappraisal strategy supports the idea that a felt obligation to take care of older people has been extended from one's own family to a broader community, including the workplace.
Of particular interest is why employees can manage their emotions if they wear a hat of clients' children.In Confucian family systems, adult children desire to reward their parents in return for parents' earlier care and even sacrifices (Yiu et al., 2021).Adult children feel obligated to care for and accept their parents' behaviors and emotions because the way children treat their elders is a measure of their moral worth (Bedford & Yeh, 2019).In professional aged care settings, employees are likely to perceive clients as elderly in their own family because these clients are at a similar age to their own parents or grandparents and they interact with these elderly clients on a daily basis.Employees use filial piety to rationalize clients' aggressions and mistreatments so they feel less negative emotions.
Self-based intrinsic ER strategy also addresses the role of service relationships in ER.Self-deterioration anticipation may be explained via a reversed mechanism of pay-it-forward reciprocity, referring to helping an unrelated third party after being helped (Nowak & Sigmund, 2005).When comparing their future self with the client, employees have not been offered any help first but anticipate that they will be cared for and supported by other carers when they are in a similar or worse situation as their current clients.Therefore, they choose to care for their current elderly clients and regulate their own emotions in client interactions to get payback from others in the future.A new label of pay-inadvance reciprocity may be more precise than pay-it-forward reciprocity to explain this strategy.Pay-in-advance reciprocity triggers employees' empathy and prevents them from or relieves their negative emotions.
Client-based intrinsic ER strategies suggest the unique role of ongoing service relationships with vulnerable clients.Ongoing service relationships enable employees to think about the future situation of the client, i.e. client deterioration anticipation.Thinking about a worse future situation for the client, such as more suffering, may trigger employees' compassion (Dutton et al., 2014) and help regulate their own emotions in current interactions.Vulnerability justification shows that attributing others' behavior to their vulnerability can help to regulate one's own emotions.This strategy may be applied in a range of service relationships when clients are vulnerable groups, such as elderly clients, children and patients.
Relationship-, self-and client-based strategies also extend our understanding of cognitive ER strategies.Strategies, such as filial reappraisal, client deterioration anticipation and vulnerability justification demonstrate that employees in ongoing service relationships can regulate their emotions by other cognitive processes, such as anticipation, justification and identifying the meanings relevant to clients beyond Gross's (1998) cognitive change and Garnefski et al.'s (2001) cognitive strategies.

Situation-, relationship-and client-based extrinsic ER strategies
Compared with intrinsic ER strategies, our understanding of extrinsic ER strategies has been at an early stage.Existing extrinsic ER strategies have focused on the steps in a process (Little et al., 2012;Williams, 2007) or outcomes (Niven et al., 2009) of regulation but rarely consider relational elements.Aggregating the existing and new strategies into situation-, relationship-and client-based strategies can provide a new integrated framework, rather than a replication of intrinsic ER models, to understand extrinsic ER.
In extrinsic ER, the relationship-based strategy, i.e. communal bonding, highlights the role of emotional responsiveness in ongoing service relationships.Employees use communal bonding so clients build trust in employees and would like to express their emotions.This aligns with social psychology research: People in communal relationships, evident in family members or romantic partners, express more emotions than in exchange relationships because communal relationships make individuals feel safe and drop some self-protective mechanisms to communicate their needs to others (Clark et al., 2017).Unlike intrinsic family appreciation, extrinsic communal bonding is not for gaining social rewards from the client.Rather, by building communal relationships, employees have a good understanding of their clients' emotional needs and habits, similar to people in communal relationships who tend to keep track of others' needs to respond to others' emotional needs (Clark et al., 2017).
Interestingly, situation-and client-based extrinsic ER strategies require different levels of employee and client involvement.Situation-based extrinsic ER requires a high level of employee involvement as employees select a more desirable situation or avoid a distressing situation for their clients.Client-based extrinsic ER strategies require a high level of client involvement.The emotional expression strategy provides sufficient autonomy for clients to express and regulate their emotions.However, in aged care settings, providing too much autonomy for clients, especially for those who are highly dependent on carers, may be risky and lead to neglect of elderly clients' emotional needs.

Process-based emotion regulation strategies in the aged care context
Although expressive suppression and surface acting have been shown to impair one's own well-being in service encounters (Hülsheger & Schewe, 2011), employees in this study reported frequent use of these strategies.A previous study with aged care employees shows that expressive suppression does not harm one's well-being; on the contrary, it decreases emotional exhaustion and increases personal accomplishment (Guan & Jepsen, 2020).The current qualitative study may provide justification for the positive effects of suppression in service relationships, such as complying with the employer's emotional display requirements and maintaining interpersonal harmony.From a resource-based perspective (Hobfoll et al., 2018), people using suppression can protect, conserve or even gain resources in this context.
Employees in this study reported using situation selection, situation modification and cognitive change to manage others' emotions.Extrinsic response modulation is not used as widely as intrinsic response modulation (i.e.expressive suppression).Requiring the target to keep emotions to themselves accumulates the target's negative feelings, triggers negative emotional dynamics between the regulator and target and impairs the regulator's well-being (Côté, 2005;Guan & Jepsen, 2020).Allowing the target to genuinely express emotions may be the first step of extrinsic ER and re-approaching the target's emotions later, especially given there are often no better alternative strategies.

Theoretical implications
These findings offer several potential contributions to the development of theory.First, this study extends ER theories by taking the temporal nature of service relationships into account.ER studies in service encounters show relative power and closeness between the employee and the target (Diefendorff et al., 2010) and relational strength and service personalization in client interactions (Wang & Groth, 2014) can influence the process and outcomes of ER.The results of this study provide a qualitative understanding of how service relationships play a role in ER.These new intrinsic and extrinsic ER strategies may provide another angle for enhancing employee well-being (Côté et al., 2008), service performance (Clarke & Hill, 2012) and talent attraction and retention in sectors with intensive workforce demands similar to aged care (Radford et al., 2015).Although these strategies have been identified in a service relationship context with a filial culture, they may also be applied to general service relationships and even broader service settings because service interactions are becoming more and more personalized (Wang & Groth, 2014).
Second, relationship-based ER strategies extend our understanding of workplace relationships, particularly employee-client relationships.Employee-client relationships are often based on exchange norms, namely the idea of paying for a service, and sometimes these relationships can form workplace friendships, which are often based on communal norms (Pillemer & Rothbard, 2018).The strategies of family appreciation, filial reappraisal and communal bonding enrich our understanding of workplace relationships.Employees and clients may develop parent-children style relationships after repeated and intimate interactions.While prior work has acknowledged the role of individualized client relationships in caregiving quality (Lamberton et al., 2015), this study may provide insights for organizations to design jobs (Grandey & Diamond, 2010) and to craft relational elements in jobs to make work more meaningful (Berg et al., 2013).

Practical implications
This study provides implications for practice in three respects: training on effective workplace ER strategies, job design with service interactions and employee well-being and retention.First, the newly identified ER strategies provide managers and employees more options for managing their own and others' emotions.Beyond the well-known ER strategies, this study provides a toolkit of intrinsic and extrinsic strategies from an employeeclient relationship perspective.Employees in service relationships need the ability to manage others' emotions, especially those who are vulnerable and dependent.Organizations and managers could encourage employees to use a range of effective ER strategies and improve employee well-being through training or other intervention programs (Ashkanasy et al., 2017).
Second, the relationship-based ER strategies findings provide insights for job design and crafting.Relationship-based ER strategies provide various options for organizations to re-design service interactions.Management may prefer convenient assignment, i.e. the cost-effective assigning of employees to work with clients according to employee availability (Gutek et al., 2002).However, organizations in service settings, particularly those committed to person-centered service provision, may consistently assign the same clients to employees to better build service relationships and understand clients' needs.Individual employees also have the flexibility to craft their relations with others including clients to make work more meaningful (Berg et al., 2013), such as fulfilling their own filial piety values.
Last, service relationships can facilitate employee well-being and retention.The aged care sector is faced with a substantial workforce shortage (Davidson & Szanton, 2020) and employee emotional burnout (Brydon et al., 2022).The relationship-based ER strategies show that service relationships can help employees experience less negative and more positive emotions, enhancing employee well-being.Filial and communal relationships often facilitate trust between employees and clients.Such interpersonal relationships are valuable for retaining employees beyond the classical retention factors, such as perceived supervisor support and job embeddedness (Radford et al., 2015).

Limitations and future research
These findings are limited by the research method and sample.Even careful interviewing can limit explanations of causal relationships based on retrospective verbal descriptions and narratives of critical incidents.Emotions can fluctuate on a within-person basis in a single day and people use ER strategies to deal with specific emotional events and situations (Grandey & Melloy, 2017).Future researchers may use an experience sampling method (Fisher & To, 2012) to capture the real-time ER strategies employees use.
Although there are fewer newly emerged strategies and supporting examples in extrinsic ER compared with intrinsic ER, the extrinsic ER strategies quotes provide critical evidence of how some existing strategies have been contextualized in service relationships.The Chinese aged care sector is still at a developing stage, and training in extrinsic ER may have been neglected for these employees.Interviewees may not be aware of extrinsic strategies they have used due to limited understanding and education in extrinsic emotional management.Further studies may investigate extrinsic ER strategies in a more mature service relationship context to enrich our understanding of extrinsic ER.
Another limitation is generalizability due to the sample from one country.Although the research context provides richness in employeeclient relationships and cultural contexts, the role of service relationships may confound with cultural values.Besides interpersonal-level factors, national culture has an impact on workplace ER and its consequences (Grandey & Melloy, 2017).Further studies are needed to examine whether the new ER strategies apply in other cultural contexts.
Last, relationship-based ER strategies involve both employee and client expectations.However, collecting self-report data only from employees without asking clients also risks an incomplete understanding of the relationships and interpersonal aspects of ER.Employees may use one set of strategies with some clients and another set of strategies with others.Future studies could use dyadic design to extend our understanding of these relationships and relationship-based ER strategies (Tse & Ashkanasy, 2015), perhaps with the actor-partner interdependence research model with employee and client in a single dyad or the one-with-many model with employees who form multiple client relationships.
The qualitative results in this study provide a theoretical framework for future researchers to develop a validated measure of relationship-, self-and client-based intrinsic and situation-, relationship-and client-based extrinsic ER strategies.Future studies may further explore the effects of these new strategies on employee and client well-being and employees' task and contextual performance in both short-and long-term timeframes.Researchers could further investigate the relationship between these new ER strategies and discrete emotions as well as age, gender and education differences in using these new strategies.

Conclusion
In service relationships, in addition to process-based ER, emotional labor and cognitive strategies, employees use relationship-, self-and client-based ER strategies to manage their own emotions and use situation-, relationship-and client-based ER strategies to manage others' emotions at work.Employees can regulate their own and others' emotions by shaping their service relationships with clients, such as family appreciation, filial reappraisal, deliberate distancing and communal bonding.This study provides qualitative evidence on the critical role of service relationships in ER.The continuing nature of service relationships results in employees using previously unknown strategies to regulate emotions.This study, as an early exploration of intrinsic and extrinsic ER strategies in ongoing service relationships, enriches workplace ER theories.The results provide insights for research and practice on improving employee emotional well-being, designing and crafting workplace relationships, and attracting and retaining a healthy workforce in service settings.

Figure 1 .
Figure 1.Interview data structure of intrinsic er strategies in service relationships.

Figure 2 .
Figure 2. Interview data structure of extrinsic er strategies in service relationships.

Table 1 .
established intrinsic and extrinsic er strategies.