A MULTILEVEL FIELD INVESTIGATION OF EMOTIONAL LABOR,AFFECT,WORK WITHDRAWAL,AND GENDER
BRENT A.SCOTT
Michigan State University
CHRISTOPHER M.BARNES
United States Military Academy at West Point
Using experience-sampling methodology,we examined within-individual relation-ships among emotional labor,negative and positive affective states,and work with-drawal,as well as the moderating role of gender.Fifty-eight bus drivers completed two daily surveys over a two-week period,producing415matched surveys.Results of hierarchical linear models revealed that affective states worned when employees engaged in surface acting but improved when they engaged in deep acting.Surface acting was positively associated with work withdrawal,and state negative affect mediated this relationship.Results also revealed moderating effects of gender:the within-individual relationships were stronger for females than for males.
As countries such as the United States have in-creasingly shifted to rvice-oriented economies (Wharton,1993),the emotions and moods(or more generally,the affect)displayed by members of or-ganizations has become increasingly important (Ashforth&Humphrey,1993;Grandey,2000,2003; Hochschild,1979,1983;Morris&Feldman,1996). Generally speaking,customer rvice employees are expected to conform to integrative display rules (e Ekman&Frien,1969),which stipulate either explicitly or implicitly that they should express positive emotions and suppress negative emotions (e.g.,Brotheridge&Grandey,2002;Hochschild, 1983;VanMaanen&Kunda,1989).However,em-ployees naturally experience a variety of emotions throughout their workdays.As a result,they often must manage their affective displays in order to conform to display rules.
Hochschild(1979,1983)referred to the lf-man-aging of affective displays as emotional labor and distinguished between two primary forms:surface acting and deep acting(e also Grandey,2000). Surface acting involves attempting to modify affec-tive displays without changing underlying feelings. In contrast,deep acting involves attempting to modify actual feelings to match required displays. Thus,with surface acting,an employee manipu-lates or“fakes”the appropriate emotional display; with deep acting,the employee tries to genuinely feel the desired emotion.
account numberCote(2005)noted that a common assumption is that emotional labor,whether it occurs via surface act
ing or deep acting,is good for organizations yet bad for employees.With the accumulation of re-arch on emotional labor,this assumption has been challenged,as differences between surface acting and deep acting,particularly with regards to their effects on organizational and employee criteria,have be-come apparent.For example,rearch has shown that surface acting is negatively associated,but deep act-ing is positively associated,with effective emotional displays during customer interactions(Grandey, 2003).In addition,rearch has revealed that surface acting,but not deep acting,is associated with per-sonal costs such as emotional exhaustion,physical complaints,and burnout(for a meta-analysis,e Bono and Vey[2005]).Thus,the effects of emotional labor appear to depend not only on the type of labor ,surface or deep acting),but also on the criteria to which it is related.
Although knowledge about emotional labor is accruing,much remains unknown.For instance, despite the dynamic interplay between felt and dis-played emotions in theories of emotional labor (e.g.,Grandey,2000;Gross,1998),to our knowl-edge,no study has examined how affective states actually experienced by employees change as a re-sult of engaging in surface acting or deep acting.In addition,little is known about the effects of emo-tional labor on employees beyond outcomes such as strain and burnout,even though theoretical models specify links between emotion
al labor and work withdrawal(Grandey,2000),which refers to such behavior as taking longer breaks than permit-ted,spending work time on personal matters,or putting less effort into one’s ,Lehman& Simpson,1992).Little is also known about the role of individual differences in the context of emo-tional labor,even though Grandey(2000)noted that individual differences need to be taken into ac-
Academy of Management Journal
2011,Vol.54,No.1,116–136.
116
comebackCopyright of the Academy of Management,all rights rerved.Contents may not be copied,emailed,posted to a listrv,or otherwi transmitted without the copyright holder’s express written permission.Urs may print,download or email articles for individual u only.
count to more fully understand emotional labor.Indeed,Bono and Vey (2005)called specifically for future rearch to examine individual differences as moderators of the relationships between emo-tional labor and work outcomes.
Finally,the vast majority of existing rearch has been cross-ctional,focusing on static differences in surface acting and deep acting between individ-uals at a single point in time.Yet,as Beal,Trouga-kos,Weiss,and Green (2006)noted,emotions,and,conquently,emotional labor,are dynamic,and thus rearch should examine emotional labor within individuals over time.This is an important point,becau,typically,in cross-ctional studies,within-individual variation is implicitly treated as error variance.The fluctuations,however,may be meaningful and related to important work crite-ria.To date,a few studies have taken a within-individual approach to the study of emotional la-bor,linking daily engagement in surface acting to lower well-being in the form of emotional exhaus-tion,negative affect,and job dissatisfaction (Judge,Woolf,&Hurst,2009;Totterdell &Holman,2003)and deep acting to the display of positive emotions and lf-rated job performance (Totterdell &Hol-man,2003).Indeed,Judge et al.(2009)found that 39.3and 31.7percent of the variance in surface acting and deep acting,respectively,was within individuals.The noteworthy studies suggest that within-individual variation in emotional labor is systematic and important to consider.According to Kozlowski and Klein (2000),a mean-ingful understanding of workplace phenomena re-quires integrative approaches that traver multiple levels.Thus,if rearchers are to more fully under-stand emotional labor,theory and rearch must be extended not only to take into account emotional labor process occurring within individuals,but also to consider the ways in which between-individual diffe
rences influence the within-person process.With this in mind,we aimed in the current study to extend theory and rearch on emotional labor by taking a multilevel approach and examining (1)the effects of emotional labor on affective states and work withdrawal within individuals over time and (2)the moderating role of gender on the within-individual relationships.Using Grandey’s (2000)model of emo-tional labor as an overarching theoretical framework,we hypothesized that when employees engage in sur-face acting,they experience an increa in negative affect and a decrea in positive affect and,con-quently,are more likely to report withdrawing from work.In contrast,we hypothesized that when em-ployees engage in deep acting,they experience a de-crea in negative affect and an increa in positive affect and,conquently,are less likely to report withdrawing from work.Moreover,in accordance with Grandey’s (2000)theorizing that emotional labor relationships may vary by gender,we hypothesized that the within-individual relationships are stron-ger for females than males.Figure 1prents a model of our hypothesized relationships.
FIGURE 1
Hypothesized Multilevel Model of the Relationships among Emotional Labor,Affect,
Work Withdrawal,and
Gender
2011117
Scott and Barnes
Overall,our multilevel approach responds si-multaneously to calls for rearch on emotional labor process within individuals(Bono&Vey, 2005;Gosrand&Diefendorff,2005)and calls for rearch on the extent to which individual differ-ences influence tho process(Bono&Vey, 2005).In doing so,we advance theory by demon-strating the importance of taking into account the dynamic,within-individual aspect of emotional labor and by explicating how individual differ-ences emphasized in extant theories of emotional labor,such as gender(Grandey,2000),may,in a top-down manner,influence the strength of with-in-individual emotional labor relationships.The examination of gender in concert with intraindi-vidual emotional labor process links between-and within-individual levels of analysis and therefore extends the range of existing theory.In addition,we advance practice by providing em-ployees and managers with guidance on how to facilitate the cultivation of desired emotional and work outcomes on day-to-day basis and by illumi-nating how men and women may best benefit from emotional labor.Theories of affect have tended to focus on external events as elicitors of emotion (e.g.,Weiss&Cropanzano,1996).Here,we con-sider a more internal ,employee choice of emotional labor strategy)whereby em-ployees may alter their own experienced states and subquent perceptions about their actions at work. Having provided an overview of our model,we prent the theoretical logic for each hypothesis in the ctions below.slow life
HYPOTHESES
Emotional Labor and Affect
Despite being recruited,lected,socialized,and rewarded for conforming to display rules(Rafaeli& Sutton,1987),employees may not actually experi-ence the affect that display rules prescribe.Emo-tion-eliciting events may occur at work(Weiss& Cropanzano,1996),or they may occur at home and emotions may spill over into work(Judge&Ilies, 2004),or individuals may experience diffu mood states that are unconnected to specific events(Wat-son,2000).Regardless of what precipitates affect, there are days and times when employees’feelings are at odds with the display requirements of their organization.As noted above,employees can at-tempt to clo the gap between what they feel and what they are suppod to feel either by faking appropriate affective displays(surface acting)or by actively modifying underlying affective states to match display requirements(deep acting)(Hoch-schild,1979,1983).
Integrating surface acting and deep acting with Gross’s(1998)taxonomy of emotion regulation, Grandey’s(2000)model of emotional labor posi-tions surface acting as a form of respon-focud emotion regulation,whereby affective states are manipulated after they have commenced.In con-trast,deep acting is a form of antecedent-focud emotion regulation,whereby affective states are manipulated before they have commenced.Ac-cording to Grandey(2000),with surface acting,in-divid
uals may attempt to suppress unwanted feel-ings by simply faking appropriate displays.With deep acting,individuals may deploy their attention elwhere by conjuring thoughts to elicit desired affective states,or they may change their cognitive perspective by reappraising their situation.The end result of such efforts is that surface acting does nothing to change underlying emotions,whereas deep acting actually creates emotions that are com-mensurate with display rules.Conquently,emo-tional dissonance,which is a n of tension that occurs when experienced and displayed affect di-verge(Hochschild,1983),is exacerbated by surface acting but diminished by deep acting(Grandey, 2003).
From an affective display standpoint,both sur-face acting and deep acting lead to the same result: namely,the expression of affective states that con-form to display rules,which are typically“integra-tive”in customer rvice ttings,meaning that positive emotions such as enthusiasm,interest,and cheer are encouraged,and negative emotions such as anger,hostility,and distress are discouraged (Beal et al.,2006;Hochschild,1983).However,the affective states actually experienced in the two forms of emotional labor differ.When individuals surface act,they do nothing to change their under-lying affective states,and indeed,tho affective states may actually worn as a result of this emo-tional labor strategy.Rearch has shown that at-tempting to suppress negative emotions may have t
wifi direct
he ironic effect of causing individuals to think about the eliciting situation even more,which ul-timately intensifies tho negative feelings (Wegner,1994).In contrast,deep acting by defini-tion produces a change in felt affect.It follows that, to the extent that display rules emphasize positive affective displays,surface acting throughout a given day should increa negative affect and de-crea positive affect.In contrast,deep acting throughout a given day should decrea negative affect and increa positive affect.
118February
Academy of Management Journal
Hypothesis1a.Within individuals,daily sur-face acting is associated with an increa in negative affect.
Hypothesis1b.Within individuals,daily sur-face acting is associated with a decrea in positive affect.
Hypothesis2a.Within individuals,daily deep acting is associated with a decrea in negative affect.
Hypothesis2b.Within individuals,daily deep acting is associated with an increa in positive affect.
Emotional Labor,Affect,and Work Withdrawal According to Grandey’s(2000)model,work withdrawal is an outcome of emotional labor. Drawing on our above arguments regarding the re-lationships between emotional labor and experi-enced affective states,we propo that employees are more likely to withdraw from work when they surface act throughout a day but are less likely to withdraw from work when they deep act through-out the day.To the extent that surface acting pro-duces a change in affect for the ,in-cread negative affect and decread positive affect),surface acting should be positively associ-ated with work withdrawal.In contrast,to the ex-tent that deep acting produces a change in affect for the ,decread negative affect and in-cread positive affect),deep acting should be neg-atively associated with work withdrawal.Below, we draw from the emotions literature to support the propositions.
A fundamental tenet of theories of emotion is that individuals’behaviors are influenced by how they feel.Specifically,affective experiences are ac-companied by action tendencies,which are impul-sive,automatic urges to achieve a particular goal (Frijda,1994,2007).By taking“control prece-dence”relative to other concerns(Frijda,1994, 2007),affective states energize and prioritize be-haviors(Elfenbein,2007).
As Elfenbein(2007)noted,the action tendencies of negative affective states address current prob-lem
s to improve the situation,which is accom-plished through avoidance(as oppod to ap-proach)in either the short or long term(Cacioppo, Gardner,&Bernston,1999;Fitness,2000).Applied to a work context,the notion of action tendencies implies that employees should report higher levels of work withdrawal on days in which they experi-ence negative affective states.Rearch on related outcomes such as counterproductive work behav-iors supports this idea,showing that such out-comes are positively associated with negative affect (for a review,e Elfenbein[2007]).Although un-desired by their organization,work withdrawal may be perceived as constructive by employees insofar as it rves as a form of coping,a way to calm negative affective states to return to a desired baline state(Grandey&Brauburger,2002).
In contrast,positive affective states prompt ap-proach rather than avoidance behavior by stimulat-ing creativity and arch as well as by expanding individuals’action repertoires(Cacioppo et al., 1999;Fredrickson,2001).Given that individuals experiencing positive affective states engage their environments,it follows that,in work contexts, employees should report lower levels of work with-drawal on days when they experience positive af-fective states.Here again,rearch indicating that related outcomes such as counterproductive work behaviors are negatively associated with positive affect provides some support for this idea(Elfen-bein,2007).
In view of the above,we propo that employees will report higher levels of work withdrawal on days when they experience high levels of negative affect and lower levels of work withdrawal on days when they experience high levels of positive affect. Furthermore,we propo that“state negative”and “state positive”affect transmit the effects of daily surface acting and deep acting on reports of work withdrawal.1According to Frijda(1994,2007),af-fective states are proximal drivers of action.Thus, daily surface acting should be positively associated with reports of work withdrawal becau this emo-tional labor strategy is associated with worning affect.In contrast,daily deep acting should be neg-atively associated with reports of work withdrawal becau this emotional labor strategy is associated with improved affect.Some indirect empirical ev-idence supports this notion,as Judge et al.(2009) found that state negative affect mediated the within-individual relationships between surface acting and both emotional exhaustion and job dissatisfaction.
We expect this mediation to be partial rather than full becau reasons in addition to affect may ex-plain associations between emotional labor and work withdrawal.For example,surface acting re-quires high levels of lf-regulation and thus de-1Here,our concern is with affective“states,”which refer to more transient,temporary feelings,as oppod to affective“traits,”which refer to more enduring individ-ual differences in the proclivity to experience particular ,Elfenbein,2007).when while 区别
2011119
Scott and Barnes
pletes lf-regulatory resources(Baumeister,Mu-raven,&Tice,2000;Beal et al.,2006;Richards& Gross,1999).Conquently,employees may utilize work withdrawal as a form of rest,allowing them to return to lf-regulatory demands with renewed strength(Muraven&Baumeister,2000). Hypothesis3.Within individuals,daily surface acting is positively related to perceptions of work withdrawal.ai
Hypothesis4.Within individuals,daily deep acting is negatively related to perceptions of work withdrawal.法国的文化
Hypothesis5a.Within individuals,negative af-fect partially mediates the positive relation-ship between daily surface acting and percep-tions of work withdrawal.
Hypothesis5b.Within individuals,positive affect partially mediates the positive relation-ship between daily surface acting and percep-tions of work withdrawal.
Hypothesis6a.Within individuals,negative affect partially mediates the negative relation-ship between
daily deep acting and percep-tions of work withdrawal.
Hypothesis6b.Within individuals,positive affect partially mediates the negative relation-ship between daily deep acting and percep-tions of work withdrawal.justin tv
爱词霸沙龙
Gender as a Boundary Condition on Emotional Labor Effects
A large body of work has been devoted to under-standing gender differences in the experience and expression of emotion(for a review,e Brody and Hall[2008]),making gender a natural individual difference to examine in concert with emotional labor(Grandey,2000;Morris&Feldman,1996). Drawing from this literature,we propo that gen-der moderates the within-individual relationships among emotional labor,affect,and work with-drawal in such a way that the relationships are stronger for females than males.
男英文名A relatively consistent finding in the emotions literature is that women are both expected to and do show greater emotional intensity and emotional expressiveness than men,and such differences hold for both positive and negative emotions (Brody&Hall,2008).The root of such differences may lie in role development,whereby females are socialized to be more emotionally expressive and men are socialized to be more emotionally re-strained(Eagly,1987;Grossman&Wood,1993;Kring&Gordon,199
8).Socialization pressures also influence the types of emotions that females and males are expected to express.Positive,relation-ship-facilitating emotions such as warmth and cheer are considered more role-appropriate for women than men,whereas negative,distancing emotions such as anger and hostility are viewed as more role-appropriate for men than women(Brody, 1999;Brody&Hall,2008;Simpson&Stroh,2004). According to poststructuralist feminist theorists (e.g.,Weedon,1987;e also Mumby&Putnam, 1992),the expectations carry over into work-places and are perpetuated by organizational norms and practices that construct different emotional roles for women and men.
The gender differences should have implica-tions for the within-individual relationships hy-pothesized above.Beginning with surface acting, when women attempt to mask or fake an emotion, they should experience greater emotional disso-nance becau their actions are at odds with their tendency to display and express what they are ac-tually feeling.Men,however,should experience less emotional dissonance,becau they are more accustomed to hiding emotions from others and faking affective states(Ashmore,Del Boca,& Wohlers,1986;Fabes&Martin,1991).Indeed, Kruml and Geddes(1998)found that women were more likely to experience emotional dissonance than men.Moreover,when display requirements call for the expression of positive emotion,the u of surfa
ce acting should be especially detrimental to women,becau the persistence of an underlying negative affective state conflicts with gender role requirements to display positive emotion.Taken together,the findings suggest that the adver effects of daily surface acting on both experienced affective ,incread negative affect,de-cread positive affect)and reports of work with-drawal should be greater for women than men.On this point,in a between-individual study,Johnson and Spector(2007)found that surface acting was more strongly associated with decread well-being (e.g.,emotional exhaustion,job dissatisfaction)for females than males.
In contrast to surface acting,deep acting should be more beneficial for females than males.In a customer rvice context,the cultivation of posi-tive affective states via deep acting should not only be more role-appropriate for women than for men, but also,to the extent that women experience such states more intenly(Brody&Hall,2008),deep acting should yield a greater gain for women than for men.In addition,rearch has shown that women are more adept than men at producing“au-thentic”smiles(Merton,1997),implying that
120February
Academy of Management Journal
