Article Reviews

A SOCIAL INTERACTION MODEL OF THE EFFECTS OF EMOTION REGULATION ON

WORK STRAIN

STÉPHANE CÔTÉ University of Toronto

Current models of how emotion regulation impacts strain focus on intraindividual processes that operate within the mind and body of the person regulating the emotion. This article presents a social interaction model of how emotion regulation impacts strain based on interpersonal processes. In this model, explanations of how emotion regulation impacts strain are based on the receiver’s response to the sender’s emotion regulation and display, the form of emotion regulation, and the emotion being regu- lated.

Organization members often exert effort to publicly display certain emotions and hide oth- ers during interpersonal interactions. For exam- ple, service agents amplify their displays of en- thusiasm to deliver quality service (Pugh, 2001). Bill collectors amplify their displays of anger to garner payments (Sutton, 1991). Negotiators sup- press their displays of happiness to claim value (Thompson, Nadler, & Kim, 1999). Criminal inter- rogators use the “good-cop, bad-cop” technique in which they amplify their displays of compas- sion, on the one hand, and anger, on the other hand, to elicit confessions (Rafaeli & Sutton, 1991).

Despite the prevalence of emotion regulation in the workplace, researchers have only recently begun to explore its consequences. One critical question researchers have explored is whether emotion regulation impacts workers’ strain and, if so, how? Strain represents a set of adverse psychological, physiological, and behavioral re- actions to work stressors, including anxiety, low commitment, elevated heart rate, and absentee- ism (Jex & Beehr, 1991; Karasek, 1979). Strain is costly to employees and organizations; thus, ad- vancing knowledge of factors that influence

strain is an important goal for organizational scientists (Sauter & Hurrell, 1999).

Recent research indicates that emotion regu- lation is an important predictor of strain (Ash- forth & Humphrey, 1993; Grandey & Brauburger, 2002; Pugh, 2002). Our understanding of the mechanisms by which emotion regulation and strain are associated, however, is limited. A re- view of the literature suggests that theory devel- opment is needed, because the current research encompasses a limited range of mechanisms. In particular, current research focuses on intraper- sonal mechanisms that operate inside the mind and body of the person regulating the emotion. I contend that consideration of interpersonal mechanisms that operate between individuals enriches our understanding and produces more comprehensive inquiries of the consequences of emotion regulation for strain.

My goal in this article is to advance theoreti- cal understanding of how emotion regulation influences strain. Precisely, I propose a social interaction model that incorporates the social dynamics of emotion in the explanation of how emotion regulation impacts strain. The social interaction model offers a novel view of how and why emotion regulation impacts strain. A com- mon assumption in the literature is that employ- ees’ emotion regulation benefits organizations but harms employees. As we will see, in the social interaction model, employees’ emotion regulation does not necessarily increase their strain. Instead, factors pertaining to the social dynamics of emotion determine when emotion regulation increases and decreases strain and when it does not influence strain.

I thank Jennifer Berdahl, Kristina Dahlin, Martin Evans, James Gross, Tim Rowley, Alan Saks, Uli Schimmack, Brian Silverman, Glen Whyte, Jia Lin Xie, editor Arthur Brief, and the three anonymous reviewers for their advice and insight- ful and constructive comments on earlier drafts of this arti- cle. This work was supported by a research grant from the Social Sciences and Humanities Research Council of Can- ada.

� Academy of Management Review 2005, Vol. 30, No. 3, 509–530.

509

THE EMOTION PROCESS

According to a componential model, emotions are affective experiences that unfold over short periods of time (Frijda, 1986; Lazarus, 1991; Lev- enson, 1994a; Scherer, 1984). Emotions typically begin with appraisals or evaluations of person- ally relevant events along such dimensions as causation (i.e., “Is the event caused by circum- stances, another person or other persons, or one- self?”; Lazarus, 1991; Roseman, 2001). The pat- tern of appraisals determines what emotion is experienced. For example, events evaluated as unpleasant and caused by another person elicit anger (Roseman, 2001).

Several types of responses to the eliciting event are activated when an emotion occurs. Action tendencies and cognitive tendencies are dispositions to engage in certain behaviors and to process information in certain ways, respec- tively (Fredrickson, 1998; Frijda, 1986). Subjective internal experiences are the conscious experi- ences that people verbally report (Frijda, 1986). Facial and bodily expressions involve muscle movements that produce public displays (Ek- man, 2003). Physiological changes such as heart rate, blood pressure, and skin conductance also co-occur with emotion (Levenson, 1994a). For ex- ample, the threat of being laid off may elicit fear that is manifested in action (e.g., working harder), cognition (e.g., focusing attention on du- ties), subjective experience (e.g., thinking “I am worried”), facial and bodily expression (e.g., raising of the eyebrows), and physiological change (e.g., accelerating heart rate).

Consequences of the experience and expres- sion of emotion are both intrapersonal and in- terpersonal (Frijda & Mesquita, 1994; Keltner & Kring, 1998; Morris & Keltner, 2000). Intraper- sonal consequences concern the effects of emo- tions on individuals’ own behavior, whereas in- terpersonal consequences concern the effects of emotions on the behavior of other persons. Ex- tant models of emotion regulation and strain focus on intrapersonal consequences. In con- trast, the model presented here focuses on inter- personal consequences.

EMOTION REGULATION

Emotion regulation includes all of the efforts to increase, maintain, or decrease one or more components of an emotion (Gross, 1999). These

components can occur relatively early during the unfolding of an emotion, such as when a person changes his or her appraisal of an event, or relatively late during the unfolding of an emotion, such as when a person changes his or her facial display of the emotion.

Theorists have distinguished between two forms of emotion regulation that differ in their timing during the unfolding of an emotion: (1) deep acting—antecedent-focused emotion regu- lation—and (2) surface acting—response-fo- cused emotion regulation (Grandey, 2000; Gross, 1998; Hochschild, 1983). Deep acting concerns the manipulation of components of emotion before the emotion is fully under way. As such, deep acting changes both the internal experience and the public display of emotion. For example, in- dividuals can appraise a threat as harmless to avoid both the internal experience and the pub- lic display of fear. Surface acting concerns the manipulation of components of emotion once the emotion is fully under way. As such, surface acting changes the public display but not the internal experience of emotion. For example, an- gry employees can reduce their public display of anger while leaving their internal experience of anger intact.

Laboratory and field studies provide converg- ing evidence that deep and surface acting both modify public displays (albeit in different ways, as we will see) and that deep acting, but not surface acting, modifies the internal experience of emotion (e.g., Gross, 1998; Gross & John, 2003). Deep and surface acting are differentially asso- ciated with strain (e.g., Grandey, 2003), and, thus, an adequate model of how emotion regu- lation impacts strain must accord an important role to the form of emotion regulation.

Theorists have also distinguished between two directions of emotion regulation: (1) emotion amplification and (2) emotion suppression (Hochschild, 1983; Levenson, 1994b). Emotion am- plification consists of initiating or enhancing public displays of emotion. For example, sales clerks may exert effort to enhance their public displays of enthusiasm. Emotion suppression consists of reducing or eliminating public dis- plays of emotion. For instance, negotiators may exert effort to reduce their public displays of happiness (Thompson et al., 1999). It is important to note that emotion suppression generally does not entirely hide the public display of emotion; laboratory studies of several emotions indicate

510 JulyAcademy of Management Review

that the suppression of emotion typically re- duces, but does not entirely hide, emotion dis- plays (see, for example, Gross & Levenson, 1993, concerning disgust, and Harris, 2001, concerning embarrassment). Emotion amplification and suppression exhibit different associations with strain (e.g., Côté & Morgan, 2002). An adequate model of how emotion regulation impacts strain must therefore accord an important role to the direction of emotion regulation.

Combining the form (deep versus surface act- ing) and the direction (amplification versus sup- pression) of emotion regulation results in a 2 � 2 conceptualization of emotion regulation. Em- ployees can amplify an emotion through deep acting by emitting behaviors that initiate or en- hance the internal experience and, in turn, the public display of that emotion. They can also amplify an emotion through surface acting by emitting behaviors that initiate or enhance the public display of that emotion when that emo- tion is not experienced or is experienced at low levels internally. Employees can suppress an emotion through deep acting by emitting behav- iors that reduce or eliminate the internal expe- rience and, in turn, the public display of that emotion. They can also suppress an emotion through surface acting by emitting behaviors that reduce or eliminate the public display of that emotion when that emotion is internally experienced.

I focus on emotion regulation instead of emo- tional labor because emotion regulation repre- sents a broader and more pervasive set of be- haviors. Most definitions of emotional labor propose that workers manage their public dis- plays of emotion to comply with display rules of the organization (e.g., Brotheridge & Lee, 2002; Glomb & Tews, 2004; Grandey, 2000; Morris & Feldman, 1996). For example, Hochschild origi- nally defined emotional labor as “the manage- ment of feeling to create a publicly observable facial and bodily display; emotional labor is sold for a wage and therefore has exchange value” (1983: 7). Totterdell and Holman, however, note that “not all emotion regulation at work is done in the service of emotional labor” (2003: 71).

Emotional labor excludes many instances of emotion regulation on the job that might impact strain. For example, imagine a supervisor who feels threatened during a meeting because one of her subordinates may reveal information that will cast doubt on one of her decisions. To in-

timidate her subordinate, the supervisor ampli- fies her public display of anger when they make eye contact during the meeting. In this example, the supervisor regulates emotion for personal reasons and not to comply with display rules of the organization; in fact, her emotion regulation may harm the organization by preventing the sharing of information.

To maximize breadth, the model I develop here focuses on emotion regulation.

EXISTING RESEARCH ON EMOTION REGULATION AND STRAIN

Data pertaining to emotion regulation and strain are beginning to accumulate. These data paint a complex picture and reveal important apparent contradictions (Barsade, Brief, & Spa- taro, 2003; Zapf, 2002).

Three findings have emerged from the pub- lished studies. First, there are no consistent dif- ferences in self-reported strain between employ- ees in “people work” jobs, defined as jobs that require workers to often regulate their own emo- tions so as to produce emotional states in others during interactions with the public (Brotheridge & Grandey, 2002; Hochschild, 1983), and employ- ees in other jobs. The available studies report twenty-nine comparisons of the strain levels of employees in people work jobs and employees in other jobs (Brotheridge & Grandey, 2002; Erickson & Wharton, 1997; Schaubroeck & Jones, 2000; Wharton, 1993). Twenty-two comparisons are not significant, three show that employees in people work jobs experience more strain than employees in other jobs, and four show the op- posite association. Second, the perception of de- mands to amplify pleasant emotions and the actual amplification of pleasant emotions are associated with low strain (Adelmann, 1995; Brotheridge & Grandey, 2002; Brotheridge & Lee, 2003; Côté & Morgan, 2002; Diefendorff & Rich- ard, 2003; Zapf, Vogt, Seifert, Mertini, & Isic, 1999). Third, the perception of demands to sup- press unpleasant emotions and the actual sup- pression of unpleasant emotions are associated with high strain (Brotheridge & Grandey, 2002; Brotheridge & Lee, 2003; Côté & Morgan, 2002; Diefendorff & Richard, 2003; Tews & Glomb, 2003; Zapf et al., 1999).

An important theoretical challenge concerns reconciling these seemingly contradictory find- ings. For instance, how can the suppression of

2005 511Côté

unpleasant emotions be related to high strain and the amplification of pleasant emotions be related to low strain? Initial conceptualizations of emotion regulation and strain centered on four intrapersonal mechanisms that concern what happens inside the mind and body of in- dividuals who regulate their own emotions: (1) emotional dissonance, (2) facial feedback, (3) personal control, and (4) effort.

Emotional Dissonance

Emotional dissonance is a state of discrep- ancy between a public display and a subjective experience of emotion that is, in theory, psycho- logically taxing and strain enhancing (Hochs- child, 1983; Morris & Feldman, 1996; Rafaeli & Sutton, 1987). Once an emotion is under way, a person who amplifies or suppresses his or her public display of emotion through surface act- ing creates dissonance between that display and his or her internal experience (Grandey, 2000). In contrast, deep acting does not produce dissonance, because the displayed emotion matches the internally experienced emotion. Consistent with these arguments, in field stud- ies emotional dissonance is correlated with sur- face acting but not with deep acting (Glomb & Tews, 2004; Holman, Chissick, & Totterdell, 2002). The emotional dissonance model thus predicts that surface acting heightens strain, but deep acting is not systematically related to strain.

The finding that there are no consistent differ- ences between the strain experienced in people work jobs and other jobs challenges the emo- tional dissonance model to the extent that em- ployees in people work jobs frequently engage in surface acting (Brotheridge & Grandey, 2002; Hochschild, 1983). The finding that the amplifi- cation of pleasant emotions is associated with low strain also challenges the emotional disso- nance model because, according to the model, strain can either increase as dissonance in- creases or remain stable when there is no dis- sonance. That is, the emotional dissonance model does not explain how emotion regulation might decrease strain.

Laboratory research provides mixed evidence for the emotional dissonance model. In recent studies, participants who suppressed displays of emotion through surface acting while watch- ing film clips exhibited several physiological signs of strain, but also decreased heart rate

(Gross, 1998; Gross & Levenson, 1993, 1997; Har- ris, 2001). In earlier studies, however, the sup- pression of emotion through surface acting low- ered physiological signs of strain (Lanzetta, Cartwright-Smith, & Kleck, 1976; Zuckerman, Klorman, Larrance, & Spiegel, 1981).

Field studies also reveal mixed findings. Self- reports of emotional dissonance are related to burnout, anxiety, depression, and job dissatis- faction (Abraham, 1999; Glomb & Tews, 2004; Holman et al., 2002; Kruml & Geddes, 2000; Mor- ris & Feldman, 1997; Zapf et al., 1999), but also job involvement (Kruml & Geddes, 2000) and per- sonal accomplishment (Zapf et al., 1999). Self- report measures of emotional dissonance are limited, because respondents calculate differ- ences between displayed and experienced emo- tions in the process of generating a response (Zerbe, 2000). Difference scores have serious lim- itations (Edwards, 1994)—for example, differ- ence scores conceal the relative contributions of the internal experience and the display of emo- tion to any association with strain. To overcome these limitations, in three studies researchers measured internal experiences and public dis- plays of emotion separately (Glomb, Miner, & Tews, 2002; Totterdell & Holman, 2001; Zerbe, 2000). None of Zerbe’s (2000) and Totterdell and Holman’s (2001) and only one of Glomb et al.’s (2002) analyses supported the emotional disso- nance model.

Facial Feedback

The facial feedback model posits that the fa- cial display of an emotion enhances or initiates the internal experience of that emotion (Tomkins, 1962). For example, amplifying public displays of happiness increases the internal ex- perience of happiness. Assuming that an in- crease in the valence of emotion decreases strain, this model predicts that the amplification of displays of pleasant emotions lowers strain and the amplification of displays of unpleasant emotions increases strain. Deep and surface act- ing should have similar effects on strain be- cause they both change the display of emotion and, hence, trigger facial feedback processes.

Laboratory studies support the facial feed- back model (e.g., Duclos et al., 1989; Ekman, Lev- enson, & Friesen, 1983; Soussignan, 2002). Even so, the model has difficulty explaining the find- ings that levels of strain are not consistently

512 JulyAcademy of Management Review

different across people work jobs and other jobs and that the suppression of unpleasant emo- tions is associated with high strain. The effect of facial feedback on strain at work has not been explicitly examined in the field, but studies in which researchers used the facial feedback model as a basis for predictions reveal mixed findings. Adelmann (1995) found that emotion regulation, measured with a scale focusing on the amplification of pleasant emotions through surface and deep acting, is associated with high satisfaction with growth opportunities. Diefen- dorff and Richard (2003) found that perceived demands for positive emotion displays predict job satisfaction. Schaubroeck and Jones (2000), however, found that perceived demands for positive emotion displays are related to poor physical health for people who have low orga- nizational identification, job involvement, or emotional adaptability.

Personal Control

Personal control is the belief that one can act to maximize pleasant outcomes and minimize unpleasant ones (Ganster & Fusilier, 1989). Workers experience relatively high personal control when they regulate emotion for personal reasons, such as to intimidate an opponent dur- ing a negotiation, but relatively little personal control when they regulate emotion for outside reasons, such as to conform to display rules enforced by the organization through mystery shoppers and performance evaluations (Hochs- child, 1983; Rafaeli & Sutton, 1987, 1989). The personal control approach predicts that emotion regulation that is outside one’s control, such as emotion regulation performed to comply with display rules, should lead to strain (Grandey & Brauburger, 2002; Pugliesi, 1999). Emotion regu- lation under one’s control, however, such as emotion regulation performed for personal rea- sons, should not lead to strain.

Schaubroeck and Jones’s (2000) findings that perceived demands to amplify public displays of pleasant emotions and to suppress public dis- plays of unpleasant emotions are associated with poor physical health for some workers sup- port the personal control model. These findings, however, differ from Diefendorff and Richard’s (2003) findings that perceptions of demands to display pleasant emotions are associated with low strain. The personal control model also has

difficulty accounting for the finding of no con- sistent differences in the strain levels of employ- ees in people work jobs and those in other jobs. Results of a study that measured control at work also challenge the personal control model. Spe- cifically, employees who had low control over their emotion regulation did not experience more strain than employees who had high con- trol (Pugliesi, 1999).

Effort

Emotion regulation involves efforts to change one or more components of emotion. Effort ex- erted to regulate emotion may increase strain, because this effort uses up resources that can- not be regained (Brotheridge & Lee, 2002; Kruml & Geddes, 2000; Pugh, 2002). Surface acting pre- sumably uses up more resources than deep act- ing, because in surface acting the emotion is already under way (Kanfer & Kantrowitz, 2002). Thus, in this approach, surface acting should be more strongly related to strain than deep acting.

The findings that people work jobs do not dif- fer consistently from other jobs in reported strain and that the amplification of pleasant emotions is associated with low strain chal- lenge the effort model. Some of the laboratory findings summarized in the section on emo- tional dissonance (Gross, 1998; Gross & Leven- son, 1993, 1997; Harris, 2001) support the effort model. Recall, however, that heart rate de- creased as a result of emotion regulation and that other studies showed that emotion regula- tion reduces strain (Lanzetta et al., 1976; Zucker- man et al., 1981). Also challenging the effort model are findings linking self-reported effort to regulate emotion to high personal accomplish- ment, low depersonalization, and high job in- volvement (Kruml & Geddes, 2000).

Limitations of Intrapersonal Models of Emotion Regulation and Strain

The research on the intrapersonal manifesta- tions of emotion has enhanced our understand- ing of how emotion regulation impacts strain. Notwithstanding these important contributions, intraindividual models are less-than-compre- hensive explanations. A first limitation stems from their lack of ability to account for the cur- rently available findings. The research indi- cates that emotion regulation is sometimes pos-

2005 513Côté

itively, sometimes negatively, and sometimes not associated with strain (Barsade et al., 2003). An adequate model of how emotion regulation impacts strain needs to predict when emotion regulation increases and decreases strain and when emotion regulation does not influence strain.

A second limitation is that, other than the fa- cial feedback model, extant models do not dis- tinguish between the amplification and the sup- pression of emotion, even though results tend to differ for them.

A third limitation stems from the lack of atten- tion to research documenting the social or interpersonal functions of emotions. Emotion theorists have argued that emotions have im- portant intrapersonal and interpersonal func- tions (Frijda & Mesquita, 1994; Keltner & Haidt, 1999; Morris & Keltner, 2000). The neglect of the interpersonal functions in existing explanations of how emotion regulation impacts strain may represent a serious omission. Employees regu- late emotion more often when with coworkers and customers than when alone (Glomb et al., 2002), and interaction partners’ characteristics and behaviors may play an important role in the explanation of how workers’ emotion regulation impacts their own strain. In particular, extant models neglect interaction partners’ responses to the authenticity of the emotion displays that follow emotion regulation. Recent evidence that interaction partners respond differently to au- thentic and inauthentic displays of emotion (e.g., Butler et al., 2003) suggests that an ade- quate model must incorporate the authenticity of emotion displays. Extant models also neglect characteristics of interaction partners such as their role, gender, ethnic background, and abil- ities that may play a role in how emotion regu- lation impacts strain.

A final limitation of extant models is the im- plicit assumption that the effects of emotion reg- ulation on strain are the same for each discrete emotion (or for each discrete pleasant emotion and each discrete unpleasant emotion). Evi- dence that discrete emotions arise in response to separate events and communicate different goals, intentions, and attitudes to others (Ek- man, 2003; Frijda, 1986; Lazarus, 1991) challenges that assumption. Theories of emotion regulation and strain may need to accord an important role to the discrete emotion regulated.

Taken together, these limitations suggest that an adequate model of how emotion regulation impacts strain must expand conceptual bound- aries. I propose a social interaction model of the effects of emotion regulation on strain that ad- dresses the limitations of existing models by according important roles to the behavior of in- teraction partners during interpersonal encoun- ters, the form of emotion regulation (i.e., deep versus surface acting), and the discrete emotion being regulated.

A SOCIAL INTERACTION MODEL OF EMOTION REGULATION AND STRAIN

The social interaction model is based on so- cial functional accounts of emotion positing that senders’ public displays of emotion communi- cate rich and important information to receivers during interpersonal encounters (Darwin, 1965; Ekman, 2003; Frijda & Mesquita, 1994; Keltner & Haidt, 1999; Izard, 1971; Plutchik, 1980). Accord- ing to these accounts, the design of the face and the facial muscles evolved to communicate so- cial signals (Darwin, 1965; Öhman, 2002). For example, a sales clerk’s displays of enthusiasm signal his or her intentions to be agreeable to customers. Existing evidence indicates that in- ternal experiences of emotion are reliably ac- companied by public displays (Cacioppo, Petty, Losch, & Kim, 1986; Matsumoto, 1987) and that public displays of emotion reveal information about the goals, intentions, and attitudes of or- ganization members (Sutton, 1991; Van Kleef, De Dreu, & Manstead, 2004). Social functional ac- counts of emotion posit that senders’ displays of emotion provide powerful signals to receivers during interpersonal interactions.

The social interaction model of emotion regu- lation and strain incorporates the social func- tions of emotion by positioning senders and re- ceivers of emotion displays in a feedback loop. Rafaeli and Sutton (1987, 1989) argued that send- ers’ emotion regulation and ensuing displays impact receivers’ responses, which, in turn, im- pact senders’ subsequent emotion regulation and displays. The social interaction model ex- tends Rafaeli and Sutton’s model to the domain of strain and describes factors from the feed- back loop that explain how and why senders’ emotion regulation impacts their own strain.

In the first part of the feedback loop, senders’ emotion regulation influences their own public

514 JulyAcademy of Management Review

displays of emotion. Field research demon- strates that emotion regulation is associated with variability in displays of emotion (Grandey, 2003; Totterdell & Holman, 2003). Laboratory re- search demonstrates that displays of emotion change as a consequence of emotion regulation (Gross, 1998; Harris, 2001). The nature of the emo- tion display depends on the form of emotion regulation (deep or surface acting) and the dis- crete emotion that is regulated. For this reason, the form of emotion regulation and the discrete emotion being regulated play fundamental roles in determining strain in the social interac- tion model.

In the second part of the feedback loop, send- ers’ displays of emotions are affective events (Weiss & Cropanzano, 1996) that are appraised by and that produce responses from receivers. Emotion displays communicate senders’ goals, intentions, and attitudes to receivers, and, ac- cordingly, individuals are more expressive when they are with others than when they are alone (Glomb et al., 2002; Hess, Banse, & Kappas, 1995). Furthermore, people strategically use dis- plays of emotion as “control moves” to achieve personal goals (Clark, Pataki, & Carver, 1996; Frijda & Mesquita, 1994; Rafaeli & Sutton, 1987). For example, people amplify their displays of happiness and suppress their displays of anger to get others to like them (Clark et al., 1996). Theorists have suggested that negotiators stra- tegically use expressions of emotion as tactical gambits to claim value (Barry, Fulmer, & Van Kleef, 2004; Morris & Keltner, 2000; Thompson et al., 1999).

Research also reveals that people are hard- wired to pick up emotional signals from others and that they rely on those signals to guide their own behavior (Elfenbein, Marsh, & Ambady, 2002; Keltner & Kring, 1998; Öhman, 2002). For example, people judge individuals who display happiness as likable and trustworthy (Clark et al., 1996) but, in contrast, judge individuals who display sadness as needy, dependent on others, and incompetent (Clark et al., 1996; Tiedens, 2001). As a result, people offer assistance to sad people (Eisenberg, 2000). Negotiators use oppo- nents’ displays of emotion to infer the location of their bargaining limits, and, based on these in- ferences, negotiators with angry opponents make larger concessions and negotiators with happy opponents make smaller concessions

than negotiators with neutral opponents (Van Kleef et al., 2004).

In the third part of the feedback loop, receiv- ers’ responses influence senders’ strain. For ex- ample, a customer who responds favorably to a salesperson’s display of enthusiasm may act in a friendly manner that may, in turn, lower the salesperson’s strain. There is considerable evi- dence that the behavior of interaction partners is associated with a worker’s strain (Cordes & Dougherty, 1993; Viswesvaran, Sanchez, & Fisher, 1999). This research has identified theo- retical mechanisms by which interaction part- ners’ actions influence a worker’s strain.

More is known about the impact of social sup- port and conflictual behavior on strain than other interpersonal behaviors. The support pro- vided by interaction partners has a direct effect on a worker’s strain by eliciting pleasant emo- tions in the worker, enhancing the worker’s self- worth, and providing understanding of the work- er’s feelings or situation (Beehr, Jex, Stacy, & Murray, 2000; Daniels & Guppy, 1994; Zellars & Perrewé, 2001). The support provided by interac- tion partners also has an indirect effect on the worker’s strain by buffering the impact of work stressors (Cohen & Wills, 1985; Dormann & Zapf, 1999). That is, work stressors augment the strain of a worker who receives high social support less than a worker who receives little social support by helping the worker redefine the po- tential harm in stressors and increasing the worker’s perceptions that others will provide re- sources (Cohen & Wills, 1985; Cordes & Dough- erty, 1993). These processes take place in the social interaction model. That is, when senders regulate emotion, they influence receivers’ pro- vision of social support that, in turn, directly and indirectly impacts senders’ strain.

With respect to conflictual behavior, Spector and Jex’s (1998) meta-analysis indicates that in- terpersonal conflict is associated with several indicators of strain, including anxiety and inten- tions to quit. At an autonomic level, conflict ac- celerates heart rate (Kamarck, Annunziato, & Amateau, 1995). The conflictual behavior of in- teraction partners is also conceptually associ- ated with a worker’s strain because it under- mines the worker’s need for communal exchange and similarity with others and harms the worker’s relationships with authority figures who control resources (Frone, 2000). These pro- cesses occur in the social interaction model.

2005 515Côté

That is, when senders regulate emotion, they influence receivers’ emission of conflictual be- havior that, in turn, impacts senders’ strain.

By expanding the role of receivers beyond customers and adding a link between receivers’ behavior and senders’ strain, the social interac- tion model extends existing models of emotion regulation in the workplace (e.g., Grandey, 2000; Pugh, 2002) stating that senders’ emotion regu- lation impacts receivers’ behavior. According to the social interaction model, receivers’ re- sponses play a critical mediational role in the explanation of how senders’ emotion regulation impacts senders’ strain.

Proposition 1: Receivers’ responses mediate the association between senders’ emotion regulation and send- ers’ strain.

Form of Emotion Regulation

When employees engage in deep acting, they modify both their internal experience and their public display of emotion. When employees en- gage in surface acting, they modify their public display but not their internal experience of emo- tion. In the social interaction model, the form of emotion regulation plays a critical role in deter- mining strain, because it has consequences for the authenticity of the emotion display. Authen- tic displays differ from inauthentic displays of emotion in several ways. Authentic and inau- thentic displays of emotion are induced by dif- ferent patterns of neural activity and involve different combinations of muscles (Ekman, 2003; Rinn, 1984). Authentic displays of emotions tend to be more symmetrical and of more consistent duration than inauthentic displays (Frank, Ek- man, & Friesen, 1993; Rinn, 1984).

More is known about the authentic and inau- thentic displays of happiness than other emo- tions. Research on displays of happiness has revealed important insights. Authentic displays of happiness (Duchenne smiles) involve contrac- tions of the orbicularis oculi muscles around the eyes and the zygomatic muscles around the mouth (Ekman, Davidson, & Friesen, 1990). In contrast, inauthentic displays of happiness (non-Duchenne smiles) only involve contrac- tions of the zygomatic muscles. Authentic dis- plays of happiness— but not inauthentic dis- plays—are related to self-reports of happiness,

and the two types of displays are related to different patterns of neural activity (Ekman et al., 1990). Untrained observers can discern be- tween authentic and inauthentic displays of happiness (Frank et al., 1993; Grandey, Fisk, Mattila, & Sideman, 2002). In future basic re- search, scholars will explore differences be- tween the authentic and inauthentic displays of other emotions (Keltner, Ekman, Gonzaga, & Beer, 2003).

There is a correspondence between the form of emotion regulation employed and the authen- ticity of the display of emotion. Deep acting in- fluences both the internal experience and the public display of emotion. Thus, in deep acting, displays of emotion are authentic because they match internal experiences. Surface acting in- fluences the public display of emotion but not the internal experience of emotion. Thus, in sur- face acting, displays of emotion are inauthentic because they do not match internal experiences.

Empirical evidence supports these proposi- tions. Self-reports of emotion amplification and suppression through surface acting are corre- lated with self-reports of inauthenticity (Brother- idge & Lee, 2002; Gross & John, 2003). Self-reports of emotion suppression through surface acting are also strongly correlated with peer-reports of surface acting, suggesting that receivers detect senders’ inauthentic emotion displays (Gross & John, 2003). Emotion amplification through deep acting is positively—and emotion amplification through surface acting negatively—related to coworker ratings of affective delivery, defined as sincerity, enthusiasm, warmth, courtesy, and friendliness during service encounters (Grandey, 2003). Finally, when watching videoclips of service providers and customers, observers rate ser- vice providers who amplify happiness through deep acting as more authentic than service pro- viders who amplify happiness through surface acting (Grandey et al., 2002).

There is theoretical evidence that receivers generally respond adversely to senders’ inau- thentic displays of emotion and that this effect generalizes across separate discrete emotions. Receivers may interpret senders’ inauthentic displays as calculated attempts to control them (Rafaeli & Sutton, 1989). For example, a debtor might believe that a bill collector who hides displays of experienced anger during a heated exchange is managing the anger to convince the debtor to make a payment. Inauthentic displays

516 JulyAcademy of Management Review

may reveal a lack of effort or interest from send- ers in developing a close interaction or relation- ship with receivers (Grandey et al., 2002). Such displays may reveal that senders are not provid- ing individual attention to receivers (Grandey, 2003) or that senders do not trust receivers (Col- lins & Miller, 1994), and they may reveal dishon- esty—an indicator of low service quality (Para- suraman, Zeithaml, & Berry, 1985). For example, a customer may perceive a salesperson’s inau- thentic display of enthusiasm as dishonest and, as a result, be dissatisfied with the service.

These theoretical arguments are supported by converging empirical evidence from laboratory and field studies. Organizational researchers have found that receivers often respond ad- versely to inauthenticity. For instance, negotia- tors are more likely to reject an offer if they believe their opponent is being deceptive than if they believe their opponent is being honest, even when the offer is exactly the same (Boles, Croson, & Murnighan, 2000).

Similarly, laboratory studies have shown that receivers respond adversely to inauthentic dis- plays of emotions that result from either ampli- fication or suppression through surface acting. In two laboratory studies, researchers found that interaction partners of individuals who sup- pressed emotion through surface acting re- ported liking their partner less, felt less rapport with their partner, and exhibited larger in- creases in blood pressure than interaction part- ners of individuals who suppressed emotion through deep acting or who did not regulate emotion (Butler et al., 2003). In other laboratory studies, individuals experienced less pleasant emotions and rated the target less favorably when they viewed pictures or videoclips of in- authentic displays of happiness than authentic displays of happiness or neutral faces (Frank et al., 1993; Grandey et al., 2002; Surakka & Hiet- anen, 1998). Finally, in one study, individuals who read descriptions of people who displayed emotion that conformed to gender-related dis- play rules believed these displays were less sincere and less appropriate than individuals who read descriptions of people who displayed emotion that departed from display rules (Hut- son-Comeaux & Kelly, 2002).

Field studies have shown that the more indi- viduals amplify or suppress emotion through surface acting (and, hence, the more inauthenti- cally they display emotion), the worse receivers

respond to them. Emotion amplification through surface acting is negatively related to coworker ratings of customer service agents’ affective de- livery (Grandey, 2003) and self-reports of per- ceived social support from the supervisor (Hol- man et al., 2002). Emotion suppression through surface acting is negatively related to self- and peer-ratings of the quality of one’s relationships with others, presumably because peers can de- tect the inauthentic display of emotion (Gross & John, 2003: Studies 4 and 5). The jury is still out concerning receivers’ responses to inauthentic displays of some emotions, such as embarrass- ment and fear. Even so, the laboratory and field evidence accumulated so far indicates that in- authentic displays of emotion and emotion am- plification and suppression through surface act- ing evoke adverse responses in receivers.

The preceding discussion implies that explicit consideration of the form of emotion regulation will enable researchers to distinguish between types of emotion regulation that increase and decrease strain. On the one hand, deep acting is not systematically associated with strain; in- stead, this association depends on the discrete emotion being regulated. On the other hand, surface acting increases senders’ strain be- cause it elicits adverse responses in receivers.

Proposition 2: As senders’ amplifica- tion or suppression of emotion through surface acting increases, receivers’ re- sponses will become more unfavor- able and, in turn, senders’ strain will increase.

Discrete Emotion

Separate discrete emotions signal different intentions, goals, and attitudes to other persons (Darwin, 1965; Ekman, 2003; Frijda, 1986; Lazarus & Cohen-Charash, 2001). For this reason, receiv- ers’ responses depend on the emotion that is regulated through deep acting in the social in- teraction model. The focus on discrete emotions in the social interaction model is consistent with a long history of emotion research (e.g., Darwin, 1965; Izard, 1971; Frijda, 1986) and recent calls to study the role of discrete emotions instead of general positive and negative affective states in the organizational sciences (Barsade et al., 2003; Brief & Weiss, 2002; Lazarus & Cohen-Charash, 2001).

2005 517Côté

Authentic displays of some emotions, such as happiness and embarrassment, generally elicit a favorable response from receivers. Happiness, for instance, communicates an intention to affil- iate and be friendly and agreeable (Izard, 1993; Keltner & Kring, 1998). Authentic displays of happiness produce a favorable stance toward senders and beget agreeable responses (Clark et al., 1996; Keltner & Bonnano, 1997). For exam- ple, salespeople who display authentic happi- ness elicit more favorable customer reactions than their counterparts (Tsai, 2001; Tsai & Huang, 2002).

Embarrassment involves the unpleasant eval- uation of one’s own transgressions (Keltner & Buswell, 1997; Tangney, 1999). Receivers experi- ence sympathy when a person expresses embar- rassment authentically (Keltner, Young, & Buswell, 1997; Semin & Manstead, 1982). Authen- tic displays of embarrassment redress relation- ships after transgressions through apology and appeasement, and, as such, embarrassment pro- tects and enhances social relationships (Keltner & Buswell, 1997).

These findings suggest that unregulated dis- plays of happiness and embarrassment reduce senders’ strain by producing favorable re- sponses in receivers. In the social interaction model, the more senders amplify happiness or embarrassment through deep acting, the more favorable receivers’ responses will be and, in turn, the lower the senders’ strain will be. Con- versely, the more senders suppress happiness or embarrassment through deep acting, the less favorable receivers’ responses will be and, in turn, the higher the senders’ strain will be.

Authentic displays of other emotions, such as anger and contempt, generally elicit an adverse response from receivers. Anger is a response to receivers’ perceived misdeeds (Averill, 1983) that intimidates and persuades receivers to comply with senders (Averill, 1983; Tiedens, 2001). Nego- tiators, for example, display anger in attempts to claim value (Allred, Mallozi, Matsui, & Raia, 1997), and, as a result, opponents of negotiators who display authentic anger place low de- mands and concede high value (Van Kleef et al., 2004). Those individuals who display authentic anger are judged less favorably than individu- als who display no emotion (Clark et al., 1996; Glomb & Hulin, 1997). Anger encourages conflict and aggression in the observers (Allred et al., 1997; Averill, 1983). In the spiral of incivility

(Andersson & Pearson, 1999), receivers recipro- cate senders’ displays of anger, and the escala- tion of anger can result in insults and threats of physical attack.

Contempt signals to others than one is supe- rior to them (Ekman, 2003). People displaying authentic contempt communicate to others that they will not accommodate them. Research on marital interactions indicates that partners of individuals who display authentic contempt ex- hibit negative responses. In one study, the wives of men who displayed authentic contempt had more negative attitudes about the relationship, experienced more unpleasant emotions, and were more likely to become ill than their coun- terparts (Gottman, Levenson, & Wooding, 2001). This research suggests that individuals also re- spond negatively to others’ authentic displays of contempt in the workplace.

Taken together, these findings suggest that unregulated displays of anger and contempt heighten senders’ strain by producing unfavor- able responses in receivers. In the social inter- action model, the more senders amplify anger or contempt through deep acting, the more unfa- vorable receivers’ responses will be and, in turn, the higher the senders’ strain will be. Con- versely, the more senders suppress anger or contempt through deep acting, the less unfavor- able receivers’ responses will be and, in turn, the lower the senders’ strain will be.

The preceding arguments suggest that the failure to consider the discrete emotion being regulated has prevented researchers from pre- dicting when emotion regulation increases and decreases strain. Investigating the regulation of emotion broadly or comparing the regulation of pleasant and unpleasant emotions likely ob- scures important processes. To more clearly un- derstand and predict associations between emotion regulation and strain, researchers must incorporate the discrete emotion being regu- lated into the design of future studies.

Proposition 3: The effects of deep act- ing on strain depend on the discrete emotion that is regulated.

Proposition 3a: As senders’ amplifica- tion of some emotions (e.g., happiness) through deep acting increases, receiv- ers’ responses will become more favor- able and, in turn, senders’ strain will decrease. Conversely, as senders’ sup-

518 JulyAcademy of Management Review

pression of these emotions through deep acting increases, receivers’ re- sponses will become less favorable and, in turn, senders’ strain will in- crease.

Proposition 3b: As senders’ amplifica- tion of other emotions (e.g., anger) through deep acting increases, receiv- ers’ responses will become more unfa- vorable and, in turn, senders’ strain will increase. Conversely, as senders’ suppression of these emotions through deep acting increases, receivers’ re- sponses will become less unfavorable and, in turn, senders’ strain will de- crease.

Moderators of the Influence of Emotion Regulation on Strain

So far, I have argued that receivers’ responses to senders’ emotion displays, the form of emo- tion regulation, and the discrete emotion being regulated play critical roles in the explanation of how emotion regulation impacts strain and in the prediction of when emotion regulation in- creases and decreases strain. The degree to which emotion regulation impacts strain may also vary according to two broad moderators pertaining to social interaction: (1) receivers’ ac- curacy at decoding senders’ displays and (2) the strength of senders’ reactions to receivers’ re- sponses.

Receivers’ accuracy at decoding senders’ emotion displays. It is generally advantageous for receivers to accurately detect the informative signals provided by senders’ displays of emo- tion (Buck, 1984; Salovey & Mayer, 1990). Indeed, evolutionary scientists have argued that the ability to decode others’ displays of emotion has evolved over time by providing significant adaptive advantages (Cosmides & Tooby, 2000; Öhman, 2002). Receivers use the information provided by senders’ displays of emotion to guide their own behavior (Elfenbein et al., 2002). For this reason, receivers focus on senders’ emo- tion displays to a greater extent than the context when inferring senders’ mental states (Naka- mura, Buck, & Kenny, 1990; Watson, 1972).

In the social interaction model, receivers’ ac- curacy at decoding senders’ emotion displays has important consequences for the strength of

receivers’ responses and, in turn, senders’ strain. Receivers’ responses to senders’ emotion displays vary in strength, with strong receiver responses impacting senders’ strain to a greater extent than weak receiver responses. For exam- ple, senders should experience less strain when receivers provide substantial social support than when receivers provide little social sup- port. Receivers who accurately decode senders’ displays of emotion produce relatively strong responses because they accurately detect send- ers’ signals of their goals, intentions, and atti- tudes and use those signals as guides to their own actions (Elfenbein et al., 2002).

In contrast, receivers who do not accurately decode senders’ displays of emotion produce relatively weak responses because they miss senders’ signals of their goals, intentions, and attitudes and, hence, have little accurate infor- mation to guide their own actions. For instance, receivers who do not accurately detect senders’ embarrassment signaling the senders’ intention to remedy past transgressions will not engage in forgiving actions that will reduce senders’ strain.

It is also possible that receivers will incor- rectly perceive that senders display another emotion such as fear. Although evidence exists that people commit such errors (Izard, 1971), lit- tle knowledge exists about when and why ob- servers perceive one emotion when another one is actually displayed. Thus, I leave the study of the consequences of receivers’ erroneous per- ceptions of emotions that senders do not regu- late and display for future research, and the following discussion pertains to inaccuracies whereby receivers fail to detect that senders regulate and display a specific emotion and do not incorrectly perceive another emotion.

Several lines of research converge to suggest that receivers’ accuracy at decoding senders’ emotion displays influences receivers’ re- sponses to senders’ emotion displays. Research on negotiation indicates that factors that im- pede decoding accuracy dampen the behavior of opponents. For instance, opponents separated by a barrier escalate competition in response to displays of anger to a lesser extent than face-to- face opponents, presumably because the barrier impedes the decoding of opponents’ emotion displays (Carnevale, Pruitt, & Seilheimer, 1981). Research on empathy also suggests a link be- tween decoding accuracy and behavior (Eisen-

2005 519Côté

berg, 2000). For example, receivers who accu- rately decode the emotion displays of a fearful person in need are more willing to help that person than their counterparts (Batson et al., 1997).

Associations between individual differences in skills at decoding others’ displays of emotion and indicators of social adjustment (Izard, 1971; Rosenthal, Hall, DiMatteo, Rogers, & Archer, 1979) imply that skilled individuals engage in more adaptive behavior than their counterparts in response to others’ displays of emotion. Al- though the existing research covers several emotions, the jury is still out concerning the consequence for behavior of accurately decod- ing displays of some emotions such as shame and interest. Even so, the available evidence suggests that receivers emit strong responses to senders’ displays of emotion when they accu- rately decode these displays. To the extent that receivers’ responses predict senders’ strain in the social interaction model, receivers’ accuracy at decoding senders’ emotion displays should moderate associations between emotion regula- tion and strain. At one extreme, when receivers decode senders’ emotion displays perfectly, as- sociations between emotion regulation and strain should be very strong. At the other ex- treme, when receivers fail to decode senders’ emotion displays, there should be no associa- tion between emotion regulation and strain.

Several specific factors may determine receiv- ers’ accuracy at decoding senders’ emotion dis- plays. A candidate individual difference charac- teristic is emotion recognition ability—the ability to accurately decode others’ emotion dis- plays (Rosenthal et al., 1979; Salovey & Mayer, 1990). Women generally outperform men on tests of emotion recognition ability (Hall, 1978); thus, gender may also determine the accuracy of re- ceivers’ decoding of senders’ emotion displays.

A candidate dyadic characteristic is ethnic group similarity. Despite substantial cross- cultural similarities (Ekman, 2003), there are subtle differences in facial expressions of dis- crete emotions between cultures (Marsh, Elfen- bein, & Ambady, 2003). As a result, members of an ethnic group recognize displays of emotion of persons from their own ethnic group more accu- rately than displays of emotions of persons from other ethnic groups (Elfenbein & Ambady, 2002).

Another candidate dyadic characteristic is relative status. People with high status possess

resources and can often act at will, without se- rious consequences (Keltner, Gruenfeld, & Anderson, 2003). People with low status, in con- trast, are sensitive to the constraints of others because of impinging threats and lack of re- sources (Snodgrass, 1992); consequently, they carefully attend to the actions of others. Thus, low-status receivers may pay more attention to and, in turn, decode the emotion displays of high-status senders more accurately than vice versa.

The preceding discussion indicates that re- search on emotion regulation and strain would benefit from consideration of factors that deter- mine receivers’ accuracy at decoding senders’ emotion displays. In particular, these factors may help researchers better predict when emo- tion regulation has strong and weak effects on strain and when emotion regulation is not re- lated to strain.

Proposition 4: Receivers’ accuracy at decoding senders’ emotion displays moderates associations between emo- tion regulation and strain. Associa- tions between emotion regulation and strain become stronger as receivers’ accuracy at decoding senders’ emo- tion displays increases.

Senders’ reactions to receivers’ responses. In the last part of the feedback loop of the social interaction model, receivers’ responses to send- ers’ emotion displays impact senders’ strain. The link between receivers’ responses and send- ers’ strain may be moderated by factors pertain- ing to senders’ reactions to receivers’ responses. Strong sender reactions to receivers’ responses should impact senders’ strain to a greater extent than weak sender reactions to receivers’ re- sponses. Results of meta-analysis reveal vari- ability in the association between social sup- port and strain, with factors such as the source of support (e.g., coworkers versus supervisors) moderating this association (Viswesvaran et al., 1999). This indicates that senders’ reactions to receivers’ responses may vary according to ad- ditional factors.

Several specific factors may determine send- ers’ reactions to receivers’ responses. Two can- didate individual-difference characteristics are neuroticism and self-monitoring. Individuals high in neuroticism react more negatively to unpleasant events and less positively to pleas-

520 JulyAcademy of Management Review

ant events than emotionally stable individuals (Côté & Moskowitz, 1998; Larsen & Ketelaar, 1991). Thus, senders high in neuroticism may react more strongly to receivers’ unfavorable re- sponses and less strongly to receivers’ favorable responses than emotionally stable senders.

High self-monitors tune in to social cues to a greater extent than low self-monitors (Snyder, 1974). High self-monitor senders may thus pay more attention and, hence, react more strongly to receivers’ responses than low self-monitor senders. That is, favorable receiver reactions may reduce strain, and unfavorable receiver re- actions may increase strain, to a greater ex- tent for high self-monitors than for low self- monitors.

A candidate dyadic characteristic that may determine senders’ reactions to receivers’ re- sponses is status. As stated above, power influ- ences how much attention individuals devote to others’ social signals. A low-status receiver may pay more attention and react more strongly to a high-status receiver’s response than vice versa.

The preceding arguments suggest that factors that determine senders’ reactions to receivers’ responses may play a key role in the explana- tion of how emotion regulation impacts strain. Researchers may better predict when emotion regulation has strong and weak effects on strain, and when emotion regulation is not re- lated to strain, by considering these factors.

Proposition 5: The strength of senders’ reactions to receivers’ responses mod- erates associations between emotion regulation and strain. Associations between emotion regulation and strain become stronger as the strength of senders’ reactions to receivers’ re- sponses increases.

The Social Interaction Model vis-à-vis Existing Theories and Findings

The social interaction model differs from ex- isting models in a fundamental way— by explic- itly incorporating the social functions of emotion in the explanation of how emotion regulation impacts strain. The research reviewed above re- veals a critical role of factors pertaining to the social dynamics of emotion in the explanation of how emotion regulation impacts strain. Based on this research, I contend that explicit consid- eration of these factors would enhance the re- search in this field. Ignoring the role of interin- dividual processes prevents investigators from researching and predicting when emotion regu- lation increases and decreases strain and when emotion regulation is not related to strain.

Figure 1 summarizes the core propositions from the social interaction model. It also dem- onstrates that the conclusion of the social inter- action model about the effects of emotion regu-

FIGURE 1 A Decision-Tree Summary of the Core Propositions of the Social Interaction Model

2005 521Côté

lation on strain differs fundamentally from extant models, in that the social interaction model posits that emotion regulation can in- crease or decrease strain. Whether emotion reg- ulation increases or decreases strain depends on receiver’s responses to senders’ emotion dis- plays, the form of emotion regulation, and the discrete emotion being regulated.

Even so, none of the propositions from the social interaction model preclude intraindi- vidual processes from operating when workers regulate emotion. Therefore, it seems important to explore whether mechanisms proposed by the social interaction model may coexist with mech- anisms proposed by intraindividual models, and how support for the propositions of the so- cial interaction model would affect the status of the intraindividual models.

Both the social interaction model and the emo- tional dissonance model propose that surface acting increases strain. The two models, how- ever, provide different explanations for this pro- posed effect. The social interaction model pro- poses that the effect of surface acting depends on receivers’ responses, but the emotional dis- sonance model does not accord a role to receiv- ers’ responses. Inconsistencies in past results concerning emotional dissonance and strain may be resolved by incorporating the role of receivers’ responses and receivers’ accuracy at decoding senders’ displays of emotion. It is pos- sible that receivers did not have the opportunity to accurately decode senders’ emotion displays in studies where researchers failed to find an association between emotional dissonance and strain (e.g., Zerbe, 2000). Research participants may not have interacted with others often enough to evoke responses that would have im- pacted their strain.

Evidence that receivers’ responses fully medi- ate associations between surface acting and strain would suggest that surface acting leads to strain principally because inauthentic dis- plays of emotion evoke adverse responses from receivers. That is, evidence for full mediation would suggest that interpersonal mechanisms prevail over dissonance to explain why surface acting is associated with strain. Evidence for partial mediation would suggest that interper- sonal mechanisms and dissonance may both explain why surface acting is associated with strain.

The social interaction model proposes a dif- ferent interpretation of the role of effort in emo- tion regulation than the effort model. Contrary to the effort model, some types of effortful emotion regulation—for example, the amplification of emotions such as happiness and embarrass- ment through deep acting—lower strain in the social interaction model. The social interaction model provides a potential explanation of in- consistent findings concerning the effort model. For example, effort may be related to high per- sonal accomplishment, low depersonalization, and high job involvement, as Kruml and Geddes (2000) found, if happiness is regulated through deep acting. Support for the propositions of the social interaction model would suggest that al- though effort may elicit strain, the receiver’s re- sponse to the sender’s effortful emotion regula- tion is a far more potent determinant of the sender’s strain.

The social interaction model and the facial feedback model share some predictions. For ex- ample, both models predict that the amplifica- tion of happiness through deep acting de- creases strain. Unlike the facial feedback model, however, the social interaction model ac- cords a critical role to receivers’ responses to senders’ emotion regulation. The two models also make some competing predictions. For in- stance, the social interaction model proposes that the amplification of happiness through sur- face acting increases strain, but the facial feed- back model proposes the opposite effect, or no effect (Soussignan, 2002). In the framework of the social interaction model, interpersonal feedback (i.e., receivers’ responses to senders’ emotion regulation and emotion displays) is far more potent than intraindividual feedback (i.e., the influence of senders’ facial muscle movements on senders’ strain), and the interpersonal feed- back dominates if the two processes are in op- position. Support for the propositions of the so- cial interaction model would suggest that, compared to interindividual feedback, intraindi- vidual feedback is not a key factor in the asso- ciation between emotion regulation and strain.

The social interaction model also has common and competing predictions with the personal control model. The personal control model pro- poses that the perception of display rules and the emotion regulation to comply with display rules increase strain. In contrast, the social in- teraction model proposes that the regulation of

522 JulyAcademy of Management Review

some emotions, including happiness and em- barrassment, through deep acting reduces strain, even if it follows display rules of the organization. That is, receivers’ responses to senders’ emotion displays are far more potent forces on senders’ strain than senders’ lack of control. Recent findings that perceived demands to amplify pleasant emotions are associated with low strain (Diefendorff & Richard, 2003) fa- vor the social interaction model. Additional sup- port for the social interaction model’s proposi- tions would indicate that personal control is not a key factor in the explanation of how emotion regulation impacts strain.

The social interaction model also potentially resolves conflicts between the seemingly incon- sistent findings from the literature summarized at the outset of this article by proposing a frame- work that accounts for these inconsistencies. Re- call that the existing research indicates that there are no consistent differences in reported strain between employees in people work jobs and employees in other jobs, that the perception of demands to amplify pleasant emotions and the actual amplification of pleasant emotions are associated with low strain, and that the per- ception of demands to suppress unpleasant emotions and the actual suppression of un- pleasant emotions are associated with high strain.

First, people work jobs may not be systemati- cally associated with more or less strain than other jobs because the effects of emotion regu- lation depend on many factors that determine when it increases and decreases strain. A sim- ple comparison of strain levels between people work jobs and other jobs obscures this complex- ity. Second, links between perceived demands to amplify and the actual amplification of pleas- ant emotions and low strain may occur because authentic displays of happiness provoke pleas- ant responses in receivers. Finally, links be- tween perceived demands to suppress and the actual suppression of unpleasant emotions and high strain occur because suppressing displays of some unpleasant emotions (e.g., embarrass- ment) through deep acting reduces or eliminates receivers’ favorable responses.

Unfortunately, the frequent neglect of the form of emotion regulation and the discrete emotion being regulated in past research precludes de- finitive conclusions. Workers often suppress dis- plays of anger, and future research will deter-

mine whether the suppression of different discrete emotions through deep acting has dif- ferential effects on strain, as proposed by the social interaction model. We cannot currently be certain that the social interaction model pro- vides the best explanation for the findings ac- cumulated so far, because the omission of key variables in past research impedes the interpre- tation of the findings. Even so, it is possible that past findings have been incorrectly attributed to intrapersonal mechanisms, and future research is needed to explicitly test the merits of an in- terpersonal approach.

Limitations of the Social Interaction Model

In this article I have focused on the presenta- tion of how and why the form of emotion regu- lation, the discrete emotion being regulated, and the receiver’s response play critical roles in the explanation of how emotion regulation im- pacts strain in the social interaction model. I also have discussed two broad moderators of the association between emotion regulation and strain: (1) receivers’ accuracy at decoding send- ers’ emotion displays and (2) senders’ reactions to receivers’ responses. I leave the identification of the role of every discrete emotion for future research. For example, I do not explore the role of pride, shame, and guilt. I also leave the iden- tification of every possible determinant of the proposed moderators for future research. For ex- ample, I do not investigate all possible contex- tual factors that potentially influence receivers’ accuracy at decoding senders’ emotion displays. Factors at several levels of analysis may deter- mine the proposed moderators. Although a com- plete analysis of the determinants of the pro- posed moderators is beyond the scope of this article, investigating those possible determi- nants represents an important direction for ex- tending the current model.

The social interaction model predicts the di- rection of effects of emotion regulation on strain. That is, the model predicts whether a specific type of emotion regulation increases or de- creases strain. It does not compare the potency of different types of emotion regulation. For ex- ample, it does not indicate whether the amplifi- cation of happiness or the suppression of anger through deep acting reduces strain the most. Future research should explore this issue.

2005 523Côté

Testing the Social Interaction Model

The social interaction model is more complex and comprehensive than intraindividual models and affords many new research opportunities. An important research agenda includes testing when emotion regulation increases and de- creases strain and when emotion regulation is not related to strain, and, more generally, deter- mining whether the social interaction model il- luminates how emotion regulation impacts strain. The traditional advantages of field re- search (i.e., realism and external validity) and laboratory research (i.e., precision and internal validity) are relevant to strain research (McGrath, 1970). Thus, field studies should be used in conjunction with laboratory studies to test the propositions from the social interaction model.

According to the social interaction model, it is critical to examine the consequences of deep and surface acting separately. Ignoring the form of emotion regulation obscures important and meaningful processes by which emotion regula- tion impacts strain. Brotheridge and Lee (2003) have created a scale that assesses deep and surface acting. Variations of their scale could be used to assess the deep and surface acting of separate discrete emotions. For example, the item “I make an effort to actually feel the emo- tions I need to display to others” could be changed to “I make an effort to actually feel the happiness I need to display to others.” Items may also have to be modified to be relevant to a broader set of interaction partners than just cus- tomers. Glomb and Tews (2004) also have cre- ated a scale to assess the amplification and the suppression of several discrete emotions through surface acting.

I have argued that the discrete emotion being regulated has often been ignored to the detri- ment of an accurate understanding of how emo- tion regulation impacts strain. Conclusions about the regulation of one particular emotion do not necessarily apply to the regulation of other emotions. Separate items from Izard’s (1977) Differential Emotion Scale could be used to measure the regulation of separate emotions (see also Tangney, Miller, Flicker, & Barlow, 1996, for items pertaining to more emotions, and Larsen & Fredrickson, 1999, for a thorough dis- cussion of measurement issues in emotions re- search).

The social interaction model requires aug- menting measures or manipulations of senders’ emotion regulation and measures of senders’ strain with measures of receivers’ responses. For example, Brotheridge and Lee (2002) gath- ered self-reports of the social support provided by coworkers and supervisors and the extent to which relationships with clients were reward- ing. Gross and John (2003: Study 4) gathered peer-reports of relationship closeness. Labora- tory research on the social interaction model requires setting up interactions between indi- viduals. For example, Butler and her associates (2003) studied research participants in pairs. One participant, unbeknownst to her interaction partner, suppressed emotion through surface acting, suppressed emotion through deep act- ing, or expressed emotion naturally. Measures of receivers’ responses included liking of send- ers and blood pressure. Researchers could use variations of this method to study the effects of manipulated sender emotion regulation on re- ceivers’ responses and senders’ strain and test for mediation. Evidence of mediation would po- tentially support the social interaction model and challenge models that do not accord a role to receivers’ responses.

Proposed moderators of the association be- tween emotion regulation and strain should also be assessed. Receivers’ accuracy at decod- ing senders’ emotion displays can be measured, for example, with performance tests of emotion recognition ability, such as the Diagnostic Anal- ysis of Nonverbal Accuracy test (Nowicki & Duke, 2001) and the Mayer-Salovey-Caruso Emo- tional Intelligence Test (Mayer, Salovey, & Ca- ruso, 2002). Because these measures are perfor- mance tests and not self-reports, concerns with common method variance are reduced. Factors that may determine senders’ reactions to receiv- ers’ responses, such as neuroticism and self- monitoring, can be assessed by using estab- lished measures.

A final critical issue concerns the measure- ment of strain. A triangulation strategy may be used in which self-reports of psychological as- pects of strain are augmented with behavioral measures, such as absenteeism, and physiolog- ical measures, such as blood pressure (see Hur- rell, Nelson, & Simmons, 1998, for a thorough discussion of the measurement of work strain).

524 JulyAcademy of Management Review

Implications for Practice

In almost any job, workers regulate their dis- plays of emotion. Although there are exceptions (e.g., Ashforth & Humphrey, 1993; Rafaeli & Sut- ton, 1987), research on emotion regulation and strain at work has predominantly focused on the pernicious effects of emotion regulation. A com- mon assumption is that display rules and emo- tion regulation benefit the organization but harm the worker. The social interaction model offers a different message. Emotion regulation is neither inherently good nor bad for workers. Re- search on the social interaction model may identify ways in which workers can de- crease— or avoid increasing—their strain. The model shifts the focus of strain-reducing inter- ventions from the individual to interpersonal in- teractions. For example, a sales clerk angry with a customer might worsen workplace relation- ships if he or she directs anger at coworkers. Expressing anger toward coworkers might re- duce emotional dissonance or the amount of ef- fort, but strain may nonetheless increase if so- cial relationships worsen because of the outburst (Andersson & Pearson, 1999).

Managers are often responsible for creating and enforcing display rules (Grandey & Brau- burger, 2002; Rafaeli & Sutton, 1989). Research on the social interaction model may inform manag- ers about which rules increase and decrease the strain of subordinates. According to the model, workers should avoid surface acting. Although this recommendation is not new (Grandey, 2000; Hochschild, 1983), the social interaction model’s explanation of why surface acting may be det- rimental is novel. Managers may train workers to engage in deep acting as opposed to surface acting. For instance, one possible deep acting technique concerns cognitive reappraisal—the reinterpretation of events optimistically (Gross, 1998; Lazarus, 1991). In one study researchers found that training teachers in cognitive reap- praisal techniques resulted in improvements in mood over time (Totterdell & Parkinson, 1999). Also, according to the social interaction model, the consequences of deep acting depend on the emotion that is regulated. For example, the sup- pression of anger through deep acting may be beneficial to workers, but this may not be the case for the suppression of embarrassment through deep acting.

Managers might also train workers to under- stand that their strain varies in part as a conse- quence of their own behavior. Through the feed- back loop, workers “create their own reality.” Disgruntled car salespeople, for example, may be dissatisfied because they elicit adverse re- sponses from customers. Raising employees’ awareness of the connection between their own emotion regulation and their own strain repre- sents a possible avenue for intervention. Also, managers play a key role in the social interac- tion model in producing responses to workers’ emotion displays. As previously argued, high- status receivers pay less attention to low-status senders’ displays of emotion than vice versa. Managers should understand that ignoring their subordinates’ emotion displays may impact their subordinates’ strain, for better or worse.

Conclusion

The prevalence of workers’ emotion regula- tion and the impact of strain on workers and their organization underscore the need to under- stand the mechanisms by which they are re- lated. The fledgling literature in this field cur- rently centers on intraindividual processes that happen inside a person’s mind and body and generally tests whether emotion regulation in- creases strain. The social interaction model sug- gests an alternate route by which to proceed with theory building and future research. In this model, the social functions of emotions play a critical role in how emotion regulation impacts strain. Emotion regulation is not inherently good or bad for strain. Instead, how emotion regula- tion impacts strain depends on multiple factors pertaining to the social dynamics of emotion.

REFERENCES

Abraham, R. 1999. The impact of emotional dissonance on organizational commitment and intention to turnover. Journal of Psychology, 133: 441– 455.

Adelmann, P. K. 1995. Emotional labor as a potential source of job stress. In S. L. Sauter & L. R. Murphy (Eds.), Orga- nizational factors for job stress: 371–381. Washington, DC: American Psychological Association.

Allred, K. G., Mallozi, J., Matsui, F., & Raia, C. P. 1997. The influence of anger and compassion on negotiation per- formance. Organizational Behavior and Human Deci- sion Processes, 70: 175–187.

Andersson, L. M., & Pearson, C. M. 1999. Tit for tat? The

2005 525Côté

spiraling effect of incivility in the workplace. Academy of Management Review, 24: 452– 471.

Ashforth, B. E., & Humphrey, R. H. 1993. Emotional labor in service roles: The influence of identity. Academy of Management Review, 18: 88 –115.

Averill, J. R. 1983. Studies on anger and aggression: Implica- tions for theories of emotion. American Psychologist, 38: 1145–1160.

Barry, B., Fulmer, I. S., & Van Kleef, G. A. 2004. I laughed, I cried, I settled: The role of emotion in negotiation. In M. Gelfand & J. Brett (Eds.), Culture and negotiation: Integrative approaches to theory and research: 71–94. Palo Alto, CA: Stanford University Press.

Barsade, S. G., Brief, A. P., & Spataro, S. E. 2003. The affective revolution in organizational behavior: The emergence of a paradigm. In J. Greenberg (Ed.), Organizational behav- ior: The state of the science (2nd ed.): 3–52. Hillsdale, NJ: Lawrence Erlbaum Associates.

Batson, C. D., Sager, K., Garst, E., Kang, M., Rubchinsky, K., & Dawson, K. 1997. Is empathy-induced helping due to self-other merging? Journal of Personality and Social Psychology, 73: 495–509.

Beehr, T. A., Jex, S. M., Stacy, B. A., & Murray, M. A. 2000. Work stressors and coworker support as predictors of individ- ual strain and job performance. Journal of Organiza- tional Behavior, 21: 391– 405.

Boles, T. L., Croson, R. T. A., & Murnighan, J. K. 2000. Decep- tion and retribution in repeated ultimatum bargaining. Organizational Behavior and Human Decision Pro- cesses, 83: 235–259.

Brief, A. P., & Weiss, H. M. 2002. Organizational behavior: Affect in the workplace. Annual Review of Psychology, 53: 279 –307.

Brotheridge, C. M., & Grandey, A. A. 2002. Emotional labor and burnout: Comparing two perspectives of “people work.” Journal of Vocational Behavior, 60: 17–39.

Brotheridge, C. M., & Lee, R. T. 2002. Testing a conservation of resources model of the dynamics of emotional labor. Journal of Occupational Health Psychology, 7: 57– 67.

Brotheridge, C. M., & Lee, R. T. 2003. Development and vali- dation of an emotional labor scale. Journal of Occupa- tional and Organizational Psychology, 76: 365–379.

Buck, R. 1984. The communication of emotion. New York: Guilford Press.

Butler, E. A., Egloff, B., Wilhelm, F. W., Smith, N. C., Erickson, E. A., & Gross, J. J. 2003. The social consequences of expressive suppression. Emotion, 3: 48 – 67.

Cacioppo, J. T., Petty, R. P., Losch, M. E., & Kim, H. S. 1986. Electromyographic activity over facial muscle regions can differentiate the valence and intensity of affective reactions. Journal of Personality and Social Psychology, 50: 260 –268.

Carnevale, P. J., Pruitt, D. G., & Seilheimer, S. D. 1981. Look- ing and competing: Accountability and visual access in integrative bargaining. Journal of Personality and So- cial Psychology, 40: 111–120.

Clark, M. S., Pataki, S. P., & Carver, V. 1996. Some thoughts

and findings on self-presentation of emotions in rela- tionships. In G. J. O. Fletcher & J. Fitness (Eds.), Knowl- edge structures in close relationships: 247–274. Mahwah, NJ: Lawrence Erlbaum Associates.

Cohen, S., & Wills, T. A. 1985. Stress, social support, and the buffering hypothesis. Psychological Bulletin, 98: 310 – 357.

Collins, N. L., & Miller, L. C. 1994. Self-disclosure and liking: A meta-analytic review. Psychological Bulletin, 116: 457– 475.

Cordes, C. L., & Dougherty, T. W. 1993. A review and an integration of research on job burnout. Academy of Man- agement Review, 18: 621– 656.

Cosmides, L., & Tooby, J. 2000. Evolutionary psychology and the emotions. In M. Lewis & J. M. Haviland-Jones (Eds.), Handbook of emotions (2nd ed.): 91–115. New York: Guil- ford Press.

Côté, S., & Morgan, L. M. 2002. A longitudinal analysis of the association between emotion regulation, job satisfac- tion, and intentions to quit. Journal of Organizational Behavior, 23: 947–962.

Côté, S., & Moskowitz, D. S. 1998. On the dynamic covariation between interpersonal behavior and affect: Prediction from neuroticism, extraversion, and agreeableness. Journal of Personality and Social Psychology, 75: 1032– 1046.

Daniels, K., & Guppy, A. 1994. Occupational stress, social support, job control, and psychological well-being. Hu- man Relations, 47: 1523–1544.

Darwin, C. 1965. (First published in 1872.) The expression of the emotions in man and animals. Chicago: University of Chicago Press.

Diefendorff, J. M., & Richard, E. M. 2003. Antecedents and consequences of emotional display rule perceptions. Journal of Applied Psychology, 88: 284 –294.

Dormann, C., & Zapf, D. 1999. Social support, social stressors at work, and depressive symptoms: Testing for main and moderating effects with structural equations in a three- wave longitudinal study. Journal of Applied Psychology, 84: 874 – 884.

Duclos, S. E., Laird, J. D., Schneider, E., Sexter, M., Stern, L., & Van Lighten, O. 1989. Emotion-specific effects of facial expressions and postures on emotional experience. Journal of Personal and Social Psychology, 57: 100 –108.

Edwards, J. R. 1994. The study of congruence in organization- al behavior research: Critique and a proposed alterna- tive. Organizational Behavior and Human Decision Pro- cesses, 58: 51–100.

Eisenberg, N. 2000. Emotion, regulation, and moral develop- ment. Annual Review of Psychology, 51: 665– 697.

Ekman, P. 2003. Emotions revealed. New York: Henry Holt.

Ekman, P., Davidson, R. J., & Friesen, W. V. 1990. The Duch- enne smile: Emotional expression and brain physiology II. Journal of Personality and Social Psychology, 58: 342– 353.

Ekman, P., Levenson, R. W., & Friesen, W. V. 1983. Autonomic

526 JulyAcademy of Management Review

nervous system activity distinguishes among emotions. Science, 221: 1208 –1210.

Elfenbein, H. A., & Ambady, N. 2002. On the universality and cultural specificity of emotion recognition: A meta- analysis. Psychological Bulletin, 128: 203–235.

Elfenbein, H. A., Marsh, A., & Ambady, N. 2002. Emotional intelligence and the recognition of emotion from the face. In L. Feldman Barrett & P. Salovey (Eds.), The wis- dom of feelings: Processes underlying emotional intelli- gence: 37–59. New York: Guilford Press.

Erickson, R. J., & Wharton, A. S. 1997. Inauthenticity and depression: Assessing the consequences of interactive service work. Work and Occupations, 24: 188 –213.

Frank, M. G., Ekman, P., & Friesen, W. V. 1993. Behavioral markers and recognizability of the smile of enjoyment. Journal of Personality and Social Psychology, 64: 83–93.

Fredrickson, B. L. 1998. What good are positive emotions? Review of General Psychology, 3: 300 –319.

Frijda, N. H. 1986. The emotions. Cambridge: Cambridge University Press.

Frijda, N. H., & Mesquita, B. 1994. The social roles and func- tions of emotions. In S. Kitayama & H. R. Markus (Eds.), Emotion and culture: Empirical studies of mutual influ- ence: 51– 87. Washington, DC: American Psychological Association.

Frone, M. R. 2000. Interpersonal conflict at work and psycho- logical outcomes: Testing a model among young work- ers. Journal of Occupational Health Psychology, 5: 246 – 255.

Ganster, D. C., & Fusilier, M. R. 1989. Control in the work- place. In C. L. Cooper & I. T. Robertson (Eds.), Interna- tional review of industrial and organizational psychol- ogy: 235–280. New York: Wiley.

Glomb, T. M., & Hulin, C. L. 1997. Anger and gender effects in observed supervisor-subordinate dyadic interactions. Organizational Behavior and Human Decision Pro- cesses, 72: 281–307.

Glomb, T. M., Miner, A. G., & Tews, M. J. 2002. An experience sampling analysis of emotional dissonance at work. Pa- per presented at the annual meeting of the Society for Industrial and Organizational Psychology, Toronto.

Glomb, T. M., & Tews, M. J. 2004. Emotional labor: A concep- tualization and scale development. Journal of Voca- tional Psychology, 64: 1–23.

Gottman, J., Levenson, R., & Wooding, E. 2001. Facial expres- sions during marital conflict. Journal of Family Commu- nication, 1: 37–57.

Grandey, A. A. 2000. Emotion regulation in the workplace: A new way to conceptualize emotional labor. Journal of Occupational Health Psychology, 5: 95–110.

Grandey, A. A. 2003. When “the show must go on”: Surface and deep acting as determinants of emotional exhaus- tion and peer-rated service delivery. Academy of Man- agement Journal, 46: 86 –96.

Grandey, A. A., & Brauburger, A. L. 2002. The emotion regu- lation behind the customer service smile. In R. G. Lord,

R. J. Klimoski, & R. Kanfer (Eds.), Emotions in the work- place: 260 –294. San Francisco: Jossey-Bass.

Grandey, A. A., Fisk, G., Mattila, A., & Sideman, L. 2002. Is that smile for real? Reaction to inauthenticity in service settings. Paper presented at the annual meeting of the Academy of Management, Denver.

Gross, J. J. 1998. Antecedent- and response-focused emotion regulation: Divergent consequences for experience, ex- pression, and physiology. Journal of Personality and So- cial Psychology, 74: 224 –237.

Gross, J. J. 1999. Emotion and emotion regulation. In L. A. Pervin & O. P. John (Eds.), Handbook of personality: The- ory and research (2nd ed.): 525–552. New York: Guilford Press.

Gross, J. J., & John, O. P. 2003. Individual differences in two emotion regulation processes: Implications for affect, relationships, and well-being. Journal of Personality and Social Psychology, 85: 348 –362.

Gross, J. J., & Levenson, R. W. 1993. Emotional suppression: Physiology, self-report, and expressive behavior. Jour- nal of Personality and Social Psychology, 64: 970 –986.

Gross, J. J., & Levenson, R. W. 1997. Hiding emotions: The acute effects of inhibiting negative and positive emo- tions. Journal of Abnormal Psychology, 106: 95–103.

Hall, J. A. 1978. Gender effects in decoding nonverbal cues. Psychological Bulletin, 85: 845– 857.

Harris, C. R. 2001. Cardiovascular responses of embarrass- ment and effects of emotional suppression in a social setting. Journal of Personality and Social Psychology, 81: 886 – 897.

Hess, U., Banse, R., & Kappas, A. 1995. The intensity of facial expression is determined by underlying affective state and social situation. Journal of Personality and Social Psychology, 69: 280 –288.

Hochschild, A. R. 1983. The managed heart. Berkeley, CA: University of California Press.

Holman, D., Chissick, C., & Totterdell, P. 2002. The effects of performance monitoring on emotional labor and well- being in call centers. Motivation and Emotion, 26: 57– 81.

Hurrell, J. J., Jr., Nelson, D. L., & Simmons, B. L. 1998. Measur- ing job stressors and strains: Where we have been, where we are, and where we need to go. Journal of Occupational Health Psychology, 34: 368 –389.

Hutson-Comeaux, S. L., & Kelly, J. R. 2002. Gender stereo- types of emotional reactions: How we judge an emotion as valid. Sex Roles, 47: 1–10.

Izard, C. E. 1971. The face of emotion. New York: Appleton- Century-Crofts.

Izard, C. E. 1977. Human emotions. New York: Plenum Press.

Izard, C. E. 1993. Organizational and motivational functions of discrete emotions. In M. Lewis & J. M. Haviland (Eds.), Handbook of emotions: 631– 642. New York: Guilford Press.

Jex, S. M., & Beehr, T. A. 1991. Emerging theoretical and methodological issues in the study of work-related

2005 527Côté

stress. Research in Personnel and Human Resources Management, 9: 311–365.

Kamarck, T. W., Annunziato, B., & Amateau, L. M. 1995. Affil- iation moderates the effects of social threat on stress- related cardiovascular responses: Boundary conditions for a laboratory model of social support. Psychosomatic Medicine, 57: 183–194.

Kanfer, R., & Kantrowitz, T. M. 2002. Emotion regulation: Command and control of emotion in work life. In R. G. Lord, R. J. Klimoski, & R. Kanfer (Eds.), Emotions in the workplace: 433– 472. San Francisco: Jossey-Bass.

Karasek, R. A. 1979. Job demands, job decision latitude, and mental strain: Implications for job redesign. Administra- tive Science Quarterly, 24: 285–308.

Keltner, D., & Bonanno, G. A. 1997. A study of laughter and dissociation: The distinct correlates of laughter and smiling during bereavement. Journal of Personality and Social Psychology, 73: 687–702.

Keltner, D., & Buswell, B. N. 1997. Embarrassment: Its distinct form and appeasement functions. Psychological Bulle- tin, 122: 250 –270.

Keltner, D., Ekman, P., Gonzaga, G. C., & Beer, J. 2003. Facial expressions of emotion. In R. J. Davidson, K. R. Scherer, & H. H. Goldsmith (Eds.), Handbook of affective sciences: 415– 429. New York: Oxford University Press.

Keltner, D., Gruenfeld, D. H., & Anderson, C. 2003. Power, approach, and inhibition. Psychological Review, 110: 265–284.

Keltner, D., & Haidt, J. 1999. The social functions of emotions at multiple levels of analysis. Cognition and Emotion, 13: 505–522.

Keltner, D., & Kring, A. 1998. Emotion, social function, and psychopathology. Review of General Psychology, 2: 320 – 342.

Keltner, D., Young, R. C., & Buswell, B. N. 1997. Appeasement in human emotion, social practice, and personality. Ag- gressive Behavior, 23: 359 –374.

Kruml, S. M., & Geddes, D. 2000. Catching fire without burn- ing out: Is there an ideal way to perform emotional labor? In N. M. Ashkanasy, C. E. J. Härtel, & W. J. Zerbe (Eds.), Emotions in the workplace: 177–188. Westport, CT: Quorum.

Lanzetta, J. T., Cartwright-Smith, J., & Kleck, R. E. 1976. Effects of nonverbal dissimulation on emotional experience and autonomic arousal. Journal of Personality and So- cial Psychology, 33: 354 –370.

Larsen, R. J., & Fredrickson, B. L. 1999. Measurement issues in emotion research. In D. Kahneman, E. Diener, & N. Schwarz (Eds.), Well-being: Foundations of hedonic psychology: 40 – 60. New York: Russell Sage Foundation.

Larsen, R. J., & Ketelaar, T. 1991. Personality and susceptibil- ity to positive and negative emotional states. Journal of Personality and Social Psychology, 61: 132–140.

Lazarus, R. S. 1991. Emotion and adaptation. New York: Ox- ford University Press.

Lazarus, R. S., & Cohen-Charash, Y. 2001. Discrete emotions

in organization life. In R. Payne & C. Cooper (Eds.), Emotions at work: 45– 81. Chichester, UK: Wiley.

Levenson, R. W. 1994a. Human emotions: A functional view. In P. Ekman & R. J. Davidson (Eds.), The nature of emo- tion: 123–126. New York: Oxford University Press.

Levenson, R. W. 1994b. Emotional control: Variations and consequences. In P. Ekman & R. J. Davidson (Eds.), The nature of emotion: 273–279. New York: Oxford University Press.

Marsh, A. A., Elfenbein, H. A., & Ambady, N. 2003. Nonverbal “accents”: Cultural differences in facial expressions of emotion. Psychological Science, 14: 373–376.

Matsumoto, D. 1987. The role of facial response in the expe- rience of emotion: More methodological problems and a meta-analysis. Journal of Personality and Social Psy- chology, 52: 769 –774.

Mayer, J. D., Salovey, P., & Caruso, D. 2002. Mayer-Salovey- Caruso Emotional Intelligence Test (MSCEIT): User’s manual. Toronto: Multi-Health Systems.

McGrath, J. E. 1970. Major methodological issues. In J. E. McGrath (Ed.), Social and psychological factors in stress: 41–57. New York: Holt, Rinehart & Winston.

Morris, J. A., & Feldman, D. C. 1996. The dimensions, ante- cedents, and consequences of emotional labor. Acad- emy of Management Review, 21: 986 –1000.

Morris, J. A., & Feldman, D. C. 1997. Managing emotions in the workplace. Journal of Managerial Issues, 9: 257–274.

Morris, M. W., & Keltner, D. 2000. How emotions work: The social functions of emotional expressions in negotia- tions. Research in Organizational Behavior, 22: 1–50.

Nakamura, M., Buck, R., & Kenny, D. A. 1990. Relative contri- butions of expressive behavior and contextual informa- tion to the judgment of the emotional state of another. Journal of Personality and Social Psychology, 59: 1032– 1039.

Nowicki, S., Jr., & Duke, M. P. 2001. Nonverbal receptivity: The Diagnostic Analysis of Nonverbal Accuracy (DANVA). In J. A. Hall & F. J. Bernieri (Eds.), Interpersonal sensitivity: Theory and measurement: 183–198. Mahwah, NJ: Law- rence Erlbaum Associates.

Öhman, A. 2002. Automaticity and the amygdala: Noncon- scious responses to emotional faces. Current Directions in Psychological Science, 11: 62– 66.

Parasuraman, A., Zeithaml, V. A., & Berry, L. L. 1985. A con- ceptual model of service quality and its implications for future research. Journal of Marketing, 49: 41–50.

Plutchik, R. 1980. Emotion: A psychoevolutionary synthesis. New York: Harper & Row.

Pugh, S. D. 2001. Service with a smile: Emotional contagion in the service encounter. Academy of Management Jour- nal, 44: 1018 –1027.

Pugh, S. D. 2002. Emotional regulation in individuals and dyads: Causes, costs, and consequences. In R. G. Lord, R. J. Klimoski, & R. Kanfer (Eds.), Emotions in the work- place: 147–182. San Francisco: Jossey-Bass.

Pugliesi, K. 1999. The consequences of emotional labor: Ef-

528 JulyAcademy of Management Review

fects on work stress, job satisfaction, and well-being. Motivation and Emotion, 23: 125–154.

Rafaeli, A., & Sutton, R. I. 1987. Expression of emotion as part of the work role. Academy of Management Review, 12: 23–37.

Rafaeli, A., & Sutton, R. I. 1989. The expression of emotion in organizational life. Research in Organizational Behav- ior, 11: 1– 42.

Rafaeli, A., & Sutton, R. I. 1991. Emotional contrast strategies as means of social influence: Lessons from criminal interrogators and bill collectors. Academy of Manage- ment Journal, 34: 749 –775.

Rinn, W. E. 1984. The neuropsychology of facial expression: A review of the neurological and psychological mecha- nisms for producing facial expressions. Psychological Bulletin, 95: 52–77.

Roseman, I. J. 2001. A model of appraisal in the emotion system: Integrating theory, research, and applications. In K. Scherer, A. Schorr, & T. Johnstone (Eds.), Appraisal processes in emotion: Theory, methods, and research: 68 –91. New York: Oxford University Press.

Rosenthal, R., Hall, J. A., DiMatteo, M. R., Rogers, P. L., & Archer, D. 1979. Sensitivity to nonverbal communication: The PONS test. Baltimore, MD: Johns Hopkins University.

Salovey, P., & Mayer, J. D. 1990. Emotional intelligence. Imag- ination, Cognition, and Personality, 9: 185–211.

Sauter, S. L., & Hurrell, J. J., Jr. 1999. Occupational health psychology: Origins, content, and directions. Profes- sional Psychology: Research and Practice, 30: 117–122.

Schaubroeck, J., & Jones, J. R. 2000. Antecedents of workplace emotional labor dimensions and moderators of their ef- fects on physical symptoms. Journal of Organizational Behavior, 21: 163–183.

Scherer, K. R. 1984. On the nature and functions of emotion: A component process approach. In K. R. Scherer & P. Ekman (Eds.), Approaches to emotion: 293–317. Hills- dale, NJ: Lawrence Erlbaum Associates.

Semin, G. R., & Manstead, A. S. 1982. The social implications of embarrassment displays and restitution behaviour. European Journal of Social Psychology, 12: 367–377.

Snodgrass, S. E. 1992. Further effects of role versus gender on interpersonal sensitivity. Journal of Personality and So- cial Psychology, 62: 154 –158.

Snyder, M. 1974. Self-monitoring of expressive behavior. Journal of Personality and Social Psychology, 30: 526 – 537.

Soussignan, R. 2002. Duchenne smile, emotional experience, and autonomic reactivity: A test of the facial feedback hypothesis. Emotion, 2: 52–74.

Spector, P. E., & Jex, S. M. 1998. Development of four self- report measures of job stressors and strain: Interper- sonal Conflict at Work Scale, Organizational Con- straints Scale, Quantitative Workload Inventory, and Physical Symptoms Inventory. Journal of Occupational Health Psychology, 3: 356 –367.

Surakka, V., & Hietanen, J. K. 1998. Facial and emotional

reactions to Duchenne and non Duchenne smiles. Inter- national Journal of Psychophysiology, 29: 23–33.

Sutton, R. I. 1991. Maintaining norms about expressed emo- tions: The case of bill collectors. Administrative Science Quarterly, 36: 245–268.

Tangney, J. P. 1999. The self-conscious emotions: Shame, guilt, embarrassment and pride. In T. Dagleish & M. Power (Eds.), Handbook of cognition and emotion: 541–568. Chichester, UK: Wiley.

Tangney, J. P., Miller, R. S., Flicker, L., & Barlow, D. H. 1996. Are shame, guilt, and embarrassment distinct emo- tions? Journal of Personality and Social Psychology, 70: 1256 –1269.

Tews, M. J., & Glomb, T. M. 2003. Emotional labor and the Five-Factor Model of Personality. Paper presented at the annual meeting of the Society for Industrial and Orga- nizational Psychology, Orlando, FL.

Thompson, L. L., Nadler, J., & Kim, P. H. 1999. Some like it hot: The case for the emotional negotiator. In L. Thompson, J. Levine, & D. Messick (Eds.), Shared cognition in orga- nizations: The management of knowledge: 139 –161. Hills- dale, NJ: Lawrence Erlbaum Associates.

Tiedens, L. Z. 2001. Anger and advancement versus sadness and subjugation: The effect of negative emotion expres- sions on social status conferral. Journal of Personality and Social Psychology, 80: 86 –94.

Tomkins, S. S. 1962. Affect, imagery, and consciousness. Vol- ume 1: The positive affects. New York: Springer.

Totterdell, P., & Holman, D. 2001. Just trying to keep my customers satisfied: A diary study of emotional disso- nance in a call centre. Paper presented at the European Congress of Psychology conference, London.

Totterdell, P., & Holman, D. 2003. Emotion regulation in cus- tomer service roles: Testing a model of emotional labor. Journal of Occupational Health Psychology, 8: 55–73.

Totterdell, P., & Parkinson, B. 1999. Use and effectiveness of self-regulation strategies for improving mood in a group of trainee teachers. Journal of Occupational Health Psy- chology, 4: 219 –232.

Tsai, W.-C. 2001. Determinants and consequences of em- ployee displayed positive emotions. Journal of Manage- ment, 27: 497–512.

Tsai, W.-C., & Huang, Y.-M. 2002. Mechanisms linking em- ployee affective delivery and customer behavioral in- tentions. Journal of Applied Psychology, 87: 1001–1008.

Van Kleef, G. A., De Dreu, C. K. W., & Manstead, A. S. R. 2004. The interpersonal effects of anger and happiness in negotiations. Journal of Personality and Social Psychol- ogy, 86: 57–76.

Viswesvaran, C., Sanchez, J. I., & Fischer, J. 1999. The role of social support in the process of work stress: A meta- analysis. Journal of Vocational Behavior, 54: 314 –334.

Watson, S. G. 1972. Judgment of emotion from facial and contextual cue combinations. Journal of Personality and Social Psychology, 24: 334 –342.

Weiss, H. M., & Cropanzano, R. 1996. Affective events theory: A theoretical discussion of the causes and conse-

2005 529Côté

quences of affective experiences at work. Research in Organizational Behavior, 18: 1–74.

Wharton, A. S. 1993. The affective consequences of service work. Work and Occupations, 20: 205–232.

Zapf, D. 2002. Emotion work and psychological well-being: A review of the literature and some conceptual consider- ations. Human Resource Management Review, 12: 237– 268.

Zapf, D., Vogt, C., Seifert, C., Mertini, H., & Isic, A. 1999. Emotion work as a source of stress: The concept and development of an instrument. European Journal of Work and Organizational Psychology, 8: 371– 400.

Zellars, K. L., & Perrewé, P. L. 2001. Affective personality and the content of emotional social support: Coping in orga- nizations. Journal of Applied Psychology, 86: 459 – 467.

Zerbe, W. J. 2000. Emotional dissonance and employee well- being. In N. M. Ashkanasy, C. E. J. Härtel, & W. J. Zerbe (Eds.), Emotions in the workplace: Research, theory, and practice: 189 –214. Westport, CT: Quorum.

Zuckerman, M., Klorman, R., Larrance, D. T., & Spiegel, N. H. 1981. Facial, autonomic, and subjective components of emotion: The facial feedback hypothesis versus the ex- ternalizer-internalizer distinction. Journal of Personality and Social Psychology, 41: 929 –944.

Stéphane Côté is assistant professor of organizational behavior and human resource management at the University of Toronto. He earned his Ph.D. from the University of Michigan. His research focuses on the effects of emotional intelligence and emotion regulation on job performance, leadership effectiveness, and work strain.

530 JulyAcademy of Management Review