You are on page 1of 44

Accountability and Emotional Labor 1 Running head: ACCOUNTABILITY AND EMOTIONAL LABOR

Power in the Process: Accountability and Emotional Labor in a Recruitment Negotiation Sharmin Spencer University of Illinois at Urbana-Champaign Peter J. Carnevale New York University

Accountability and Emotional Labor 2 Abstract Emotional labor is an individuals control over their emotions to achieve an organizational objective. This study used a laboratory method to examine the combined impact of two forms of accountability on emotional labor: outcome and process accountability (J. S. Lerner & P. E. Tetlock, 1999). We expected process accountability to heighten emotional labor and we expected outcome accountability to moderate the impact of process accountability. The data supported both expectations. In a simulated recruitment negotiation, process accountability encouraged emotional labor; it led interviewers to act more positive toward an unpleasant job recruit, but only in the absence of outcome accountability. Outcome accountability nullified the impact of process accountability. The findings comport with previous research on process and outcome accountability in the decision-making and negotiation literatures, and suggest that accountability may be an important form of organizational norm-enforcement that can affect the occurrence of emotional labor.

Accountability and Emotional Labor 3 The Effects of Process and Outcome Accountability on Emotional Labor Darwin (1872) argued that human emotions have evolutionary roots, and that specific facial expressions reflect the same emotion across species. A century later, Ekman (1973) found evidence of a complex interplay of innate and environmental influences on human emotional expression: environmental forces determine when or the frequency with which a behavior (such as a smile) is produced; genes determine the emotion behind a particular facial expression. There is today broad agreement that human facial expressions convey sentiments that can be interpreted by an outside observer, and that facial expression are sometimes but not always honest reflections of private thoughts and feelings. So pervasive are emotions both within individuals as well as cross-culturally they often are hard to ignore even in the most professional of settings. The question is, if emotions are innate and universally recognized, as Darwin argued, then how easily can we learn to fake or suppress emotions when the need arises in a professional context? Certainly, the power of emotions as a communicating device has long been recognized and used as a means toward an end in a myriad of settings. But how easily can we learn to monitor our facial feelings so as to express a message that is consistent with the spoken word? This is the question employers are now asking their employees to consider. In a time when jobs might be at stake, or incentives on the horizon, this is a question many employees are motivated to answer. In recent years, especially in the service industry, increasing emphasis is being placed on the regulation and expression of specific emotions as a tool to lure and tempt customers into buying a product. Indeed, many organizations today require that their

Accountability and Emotional Labor 4 employees display particular emotions while suppressing others (Hochschild, 1983). The idea behind this principle is that quality interactions between customers and employees foster customer satisfaction, which in turn benefit and ensure the success of an organization geared toward making profits (Morris & Feldman, 1996). Hochschild (1983) was the first to recognize the potential pitfalls of a job requirement that invades something so fundamentally private as ones emotions. Following an examination of expressed emotion in flight attendants, she introduced the term emotional labor a phrase used to describe the management of feeling to create a publicly observable facial and bodily display. This vague description could mean anything from the expression of emotions not genuinely felt to the suppression of emotions genuinely felt, but was used to reflect the artificial manufacturing of an organizationally sanctioned emotion. In an effort to clarify the concept, Morris and Feldman (1996) described emotional labor as the effort, planning, and control needed to express organizationally desired emotion during interpersonal transactions. Because such activities are laborintensive, and are tailored to the ever-changing demands of customer-related transactions, they are collectively referred to as emotional labor. Morris and Feldman argued that there are three essential components to this construct: frequency of interaction, duration of interaction, and emotional dissonance. Within this framework, these authors identified and found support for three antecedents of emotional labor routineness of task, job autonomy, and power of role receiver. According to Morris and Feldmans (1997) conceptualization of emotional labor, frequency, duration, and emotional dissonance are independent of each other. However,

Accountability and Emotional Labor 5 one could argue that it is emotional dissonance the discrepancy between actual and required emotional expression that is fundamental to the concept of emotional labor (see Mann, 1999a). The other dimensions remain offshoots of this underlying construct because they simply refer to its features rather than act as separate descriptors of the concept. Additionally, it can be argued that the other characteristics serve to magnify the influence of emotional dissonance. Frequency of emotional display, for instance, is classified as a symptom of emotional labor precisely because it could serve as an indicator of the number of times there is a conflict between real and manifested emotion. At the same time, however, repeated practice with this requirement could in fact reduce the emotional labor component. Whether or not frequency of emotional dissonance serves to accentuate or eliminate emotional labor is perhaps an empirical question. But ultimately it is the internal incongruity between what is expected and what is truly felt, regardless of frequency, that we believe lies at the heart of emotional labor. Display rules, or norms about how to interact with customers, are considered the leading causes of emotional display employees express at work (Rafaeli & Sutton, 1990). Calling the construct Display of Positive Emotion, or good cheer, Rafaeli and Sutton (1990) argued that ground rules for how employees should emote with their customers stem from two sources the organizational setting, and the target person. By making structured observations of cashiers in a busy Israeli store, and recording the amount of positive emotions they exhibited, they were able to compile a composite of behavioral indicators that promised to measure Display of Positive Emotion. A factor analysis of the construct revealed a two-factor solution: mechanics of displayed positive emotion, and interactive display of positive emotion. Behavioral indicators of the first factor included

Accountability and Emotional Labor 6 smiling, making eye contact, and thanking the customers, while indicators of the second factor included pleasantness and attentiveness. As predicted, busyness of the store was negatively correlated with display of positive emotions, while customer demand (defined as the degree of complexity involved in cashier-customer interactions) was positively correlated with display of positive emotion. Rafaeli and Suttons (1990) work provides researchers a useful mechanism with which to measure the expression of positive emotions. They primarily investigated the display of positive emotions and its antecedents. In the present study, we focus on both the display of positive emotions and the suppression of negative emotions, as well as on genuinely felt emotion. We conceptualize emotional labor as performance-relevant behavior that is consistent with organizationally determined display rules (Tews & Glomb, in revision). Following Tews and Glomb (in revision), we define emotional labor as behavior that can be reflected along two dimensions: the act of positive emotional display displaying organizationally mandated positive expressions that might or might not belie ones true sentiments, and the act of negative emotional suppression suppressing genuine feelings of negativity that go against ground rules of appropriate emotional expression. This behavioral definition allows the construct to be measured quantitatively and comprehensively. A caveat, however, is in order. With any latent trait, the finding of a measure with construct validity is a difficult task; but with emotional labor, the difficulty is heightened. This is because outward manifestations that purport to measure the construct might be reflective of emotional labor, but they might also reflect genuinely felt emotion. For example, a waitress who smiles at her customers might do so because she is genuinely

Accountability and Emotional Labor 7 happy to greet them, or because she is expending effort to conceal her dissatisfaction at having received a small tip by the previous customer, and hopes now to ensure that the insult is not repeated. Mann (1999a) would probably argue that only if the latter case is true, could her behavior (smiling) be described as emotional labor. This is because there is a distinct clash between required public expression, and genuine private feeling. Thus, an act governed by emotion is only considered emotional labor when the outward emotions contradict inwardly felt emotions. As such, then, what might appear as a display of genuinely positive emotions, might well in fact be a carefully crafted faade put on by the assiduous employee simply doing their job. The discovery, so to speak, of emotional labor, might only be recognized after the employee has later admitted in a self-report to the artificial nature of his/her role, and to the difficulty involved in emoting that way. On the other hand, whether or not self-reports of emotional labor indicate its presence or absence, there are often instances when people are able to retain their composure and display positive emotions even in the face of publicly recognized unpleasantness. As such, then, despite their outward positive display to the casual observer, such individuals behavior must certainly be described as emotional labor because of the difficulty that must be involved in masking what the outside observer might consider to be the targets true emotions. Indeed, when it comes to workplace interactions, and from an employers vantage, a premium is placed on what researchers refer to as emotional labor. Therefore, in such situations, the critical factor of emotional labor might not be self-reports of genuine feelings, so much as how good employees are at expressing positive emotions and suppressing negative emotions in universally acclaimed instances of unpleasantness.

Accountability and Emotional Labor 8 Because investigation on emotional labor is still in its infancy, little is known with certainty as to its effect on performance. Researchers have found that emotional labor could lead to burnout, exhaustion, and other physical symptoms of stress (Grandey, 2000), and that its impact on job satisfaction is mixed (see Morris & Feldman, 1996), but for the most part, the assumption is made that its influence on job performance, at least in the short run, is positive. From an employers perspective, emotional labor is beneficial to any profit-making organization, and as such is consistent with a high level of performance. This is because when the effort is made to exhibit emotions that are in keeping with the dictates of ones organization, presumably, that is tantamount to good performance. Accountability Accountability, a norm-enforcing device, operates to make individuals answerable to those in authority by having them behave according to a certain standard expected of their position (Schlenker, 1986). In anticipation of being evaluated and rewarded by a superior, accountable individuals feel pressured to adhere to a pre-ordained standard, and must justify their beliefs and actions should they deviate from the norm (Gelfand & Realo, 1999). Within the context of a negotiation, accountability produces high aspirations and contentious behavior that can lead performance to suffer (Ben-Yoav & Pruitt, 1984; Kramer 1995). High accountability situations evoke competitive behaviors from representatives, which make them less open to their opponents offers, and which subsequently hinder them from achieving optimal outcomes. Low accountability, on the other hand, results in more cooperation between negotiating opponents that serves to increase outcomes that benefited both parties (Carnevale & Pruitt, 1992).

Accountability and Emotional Labor 9 Lerner and Tetlock (1999) cite studies that distinguish between outcome accountability and process accountability within the context of judgment and decisionmaking. Outcome accountability refers to the extent to which constituents feel accountable for the outcome or end result of a work-related activity. For instance, constituents might be accountable to their supervisors for the outcomes obtained during decision-making. Process accountability, on the other hand, refers to the extent to which constituents feel accountable for the process with which they make decisions, that is, the extent to which they evaluate alternative decision-making strategies, rather than for the value of their outcomes. The irony is that outcome accountable employees make poorer decisions than process accountable employees (Lerner & Tetlock, 1999). When the focus is on outcome accountability, more detrimental judgments and decisions result presumably due to a more rigid way of thinking, the stress of having to achieve a given result, and a prior commitment to a given line of action (Simonson & Staw, 1992). It is as if constituents are aware of what they are supposed to achieve, but are ignorant of the means by which to achieve it. However, when process accountability is the focus, greater accuracy in judgments and decisions is achieved due to an increased emphasis on openness to alternative perspectives that help in attaining optimal outcomes. Accountability Effects on Emotional Labor We suggest that outcome and process accountability well fit into a paradigm of workplace emotions. Specifically, outcome accountability refers to the extent to which constituents are responsible for attaining a certain goal. For example, employees are accountable for the number of products they sell to customers. Process accountability refers to the extent to which constituents are responsible for the method they choose to

Accountability and Emotional Labor 10 implement. For example, employees are accountable for the quality of an interaction they might have with a customer, which presumably might be reflective of an organizations overall image. In such a situation, then, employees might feel accountable for how they present themselves to their clients, rather than how many products they are able to sell. Although research has not specifically examined the impact of accountability on emotional labor, indirect evidence citing a positive relationship between the two has been found. Hochschilds (1983) qualitative study on flight attendants found that those employees dealing directly with passengers and hence those more accountable for upholding their airlines public reputation were found to exhibit higher levels of emotional labor than those employees who did not have to deal with passengers. Additionally, Rafaeli and Suttons quantitative studies (1987, 1990) examined front line workers in stores cashiers and clerks who came face-to-face with customers, and who were hence in a position that made them accountable for the protection of their organizations public image. They found that a positive relationship existed between customer demand defined as the level of complexity in transactions between cashiers and customers and the display of good cheer. Finally, Morris and Feldmans (1997) finding on the negative relation between job autonomy and emotional labor also seems to indicate that employees with less job latitude are also those who are more accountable to a superior, and hence pressured to exert emotional labor. It stands to reason that being accountable for how one presents oneself may have an impact on the level of emotional labor expended. If emotional labor is viewed as a vehicle through which to enhance performance in a way that benefits the organization, then accountability, a construct that encourages

Accountability and Emotional Labor 11 emotional labor, should be seen as an asset to the organization, as it indirectly affects performance through emotional labor. Because outcome accountability has a deleterious effect on decision-making effectiveness, whereas process accountability has a positive effect, it may be that process accountability, and not so much outcome accountability, has the positive effect on emotional labor. This implies that process accountability might induce superior customerrelated performance over outcome accountability. Hypotheses The main purpose of this research is to examine whether specific types of accountability serve as antecedents to emotional labor. Specifically, we seek to understand whether accountability at the outcome or process levels affect the amount of emotion management or emotional labor of employees. We expected process accountability in particular to have a facilitating effect on emotional labor, more than outcome accountability. But we expected that the two forms of accountability would interact. Process-facilitating effect. Several studies report that process accountability improved the accuracy of decisions made within a decision-making paradigm (SiegalJacobs & Yates, 1996; Mano, 1992; see Lerner & Tetlock, 1999; Tetlock, 1985, 1992). In a field study examining purchasing managers, Doney and Armstrong (1996) also found that process accountability increased the effort managers devoted to analyzing competing products, while outcome accountability had no such effect. Given the facilitating effect that process accountability has on attaining optimal decisions, and in encouraging greater managerial diligence, we expected that process accountability would have a similar,

Accountability and Emotional Labor 12 positive, impact on emotional labor. Specifically, we expected that process accountability would accentuate cooperativeness in an employee whose task was to display courtesy, politeness, and a positive demeanor in an effort to achieve an organizational objective, namely to recruit an unpleasant person to the organization. An argument could be made that process accountability affects genuinely felt emotions rather than simply the display of positive emotions. However, we expected that process accountability would drive the outward display of positive emotions rather than privately experienced emotion, because most individuals would, in all likelihood, at least internally react negatively to someone who was clearly unpleasant. This internal negativity would manifest itself in self-reports of true attitudes, which would stand in contrast to observed ratings of positive and negative emotions. The process facilitating effect of process accountability formed the basis of our first hypothesis: Hypothesis 1: Process accountability will foster high levels of emotional labor. Individuals under process accountability will outwardly express more positive emotions, and suppress more negative emotions, than individuals who do not experience process accountability. Outcome-rigidity effect. The evidence on outcome accountability appears mixed, with the bulk of the evidence suggesting it may have little or no effect on emotional labor. In the negotiation literature, there is evidence that outcome accountability fosters higher aspirations and contentious behavior, which can in some cases produce a rigid approach to negotiation and poor outcomes (Carnevale, 1985; Carnevale & Pruitt, 1992; but see OConnor, 1997); there is evidence that it can do this especially among people socialized in an individualistic culture (Gelfand & Realo, 1999). In a decision context,

Accountability and Emotional Labor 13 Simonson and Staw (1992) found that outcome accountability resulted in poorer judgments and reduced accuracy than process accountability. This was attributed to a narrower attention capacity associated with outcome, but not process, accountability. This is consistent with Janis and Mann's (1977) conflict theory: the high uncertainty about decision success coupled with high decision consequence produces stress. Stress can produce rigidity and poor performance. Consistent with this, Siegel-Jacobs and Yates (1996) found that outcome accountability produced greater stress and greater judgmental inconsistency. Taken together, these studies suggest that outcome accountability may interfere with a persons efforts at emotional labor. Given the debilitating effect that outcome accountability can have on negotiation and decision-making, we expected it to have a similarly flagging effect on emotional labor. A stubborn focus on the outcome to the exclusion of the means with which to achieve that outcome, may lead to stilted emotional expression, that is, to an inability to show positive affect and an inability to refrain from showing negative affect. Hypothesis 2: Outcome accountability will moderate the effect of process accountability such that the effects of process accountability will only obtain when there is no outcome accountability. In other words, outcome accountability will lessen the positive impact of process accountability on emotional labor. Method Participants and Design 85 participants (44 males, and 41 females), who were drawn from the Introductory Psychology subject pool, participated in this experiment for which they

Accountability and Emotional Labor 14 received partial course credit. All were between the ages of 18-22. Out of them, 88% were White, 4% African-American, 4% Hispanic and 2% were Asian American. Outcome accountability (present, not present), and process accountability (present, not present), were manipulated in a 2 x 2 factorial design, which contained approximately 20 participants per cell. Procedure Participants were randomly assigned to one of the four experimental conditions (outcome, process, combined outcome and process, and a no accountability control group). Two participants were run in each session of the experiment, both of whom were assigned to the same condition. This was followed by the next pair of participants who were randomly assigned to another condition, and so on, for each session run. Table 1 lists the sequential steps involved in the experiment. Upon arrival, each participant was greeted, presented with a folder of instructions that included an informed consent form, and directed to their own little room in which they were asked to read the contents of their folder. The first item was a cover sheet that gave a brief synopsis of the experiment. It described the experiment as a simulated recruiting interview of the sort that typically takes place in a job interview, and about how interviews might be affected by different communication technologies. The cover story given to each participant informed them that they were employees of a student organization called UI Computing Assist, which was in the process of recruiting a new worker, and that they (the participants) were in charge of the interview process. Their job was to try to recruit the hired job candidate in the upcoming interview by selling the organization of which they were a member, to the person they were about

Accountability and Emotional Labor 15 to interview. The cover story added that the interview would be with another subject (in reality a confederate), who was given instructions on playing the role of a hired job candidate, and who was currently in a remote location within the building. Participants were also informed that the interaction would take place using a video conferencing system. The participant would view the other over a TV monitor, and the experiment was an attempt to investigate how this sort of technological medium affected the interview process. They were also told that while they would not be able to view the interviewee while they (the participants) spoke, they would be able to see the interviewee over the TV monitor when she responded. This unusual situation was explained as a technological form that pertained to the purported purpose of the study to investigate technology on the interview process. Participants were also informed that their interview would be video taped for examination at a later date. In fact, the reason for the TV-video hookup was to provide experimental control. Our concern was to ensure that every participant received the same stimulus. Thus we decided to record the confederate before hand rather than have them pose live as an interviewee. This ensured that we could standardize the stimulus presented to each subject, and hence provided greater control over the interview proceedings. The confederate on the tape was trained to exhibit negative emotional expression, use a sharp tone of voice, and be skeptical of any claims made, so as to encourage emotional labor in the participant. The interviews were recorded to later code for verbal, non-verbal or paralinguistic signs of emotional labor in the participants. Interview. In each experimental session, one of the participants was directed by an experimental aide and called to a small interview room in which there was a table

Accountability and Emotional Labor 16 and chair, a TV monitor, and a digital camera. A script for the participants to follow was taped to the table (see Appendix). There was also a stopwatch set at two minutes that was used to time the participant each time they spoke. Before asking the participant to press the start button on the stopwatch, the experimenter orally reviewed the instructions so as to make sure they had been fully understood. The participant then pressed the start button while the experimenter pressed the record button on the video camera. These cues started the interview process in which the participant tried to recruit the interviewee by using the information written on the script. The first phase of the interview ended when the stopwatch beeped after two minutes (see Appendix A for the interview script used by participants). At this point, an experimental aide in an outside control room surreptitiously pressed the start button on a VCR so the tape of the confederate posing as the unpleasant interviewee was played. This enabled the participant in the interview room to view the recorded confederate whom they believed was live and responding to their opening pitch for the first time. After the confederate had finished speaking, the experimental aide stopped the tape from playing, and the participant spoke in response to what they had just heard, for another two minutes. The video camera recorded all three phases the participant speaking first, followed by the confederate tape, followed by the participants response of the interview. Upon concluding the interview, the participant was led back to their little room and was asked to complete the experimental measures included in their folder (see Appendix B for excerpts of the confederates script). It was only after the participant was safely back in their room that the second participant seated in their own separate room was called upon for their turn to interview.

Accountability and Emotional Labor 17 The same exercise was repeated with the second participant as what took place with the first. Once they had completed the interview, they too were led back to their room, and asked to complete the experimental measures inside their folder. After each participant had turned in their responses, they were debriefed, thanked for their participation, and allowed to leave. Manipulation of accountability. Accountability was manipulated in the instructions participants read in each condition. In the outcome accountability condition, participants were informed that there is a chance that you can earn real money for participating in this study. The person who you interview in the recruiting interview will decide to either join or not join the organization. If they say yes, that they will join the organization, then you will receive a ticket that will be entered in a lottery. One in every 10 tickets will be randomly drawn from this lottery, and will be converted into a cash prize of $20 for each winner. In the process accountability condition, participants were told (in the information folder), there is a chance that you can earn real money for participating in this study. After you have completed the interview, the videotape of the interview will be submitted for review by a senior graduate student in human resources who works in our lab, and who is an expert on interview processes. This expert will write an evaluation of your intentions and behaviors during the interview. This expert will also decide, based on the interview tape, if you did a good enough job to earn a prize. If they judge that your interview performance does deserve a reward, then you will receive a ticket that will be entered in a lottery. One in every 10 tickets will be randomly drawn from this lottery, and will be converted into a cash prize of $20 for each winner. Participants in this condition

Accountability and Emotional Labor 18 were also informed that the person evaluating their tapes would not know the outcome of the interview. That is, the senior graduate student would not know whether the interviewee decided to accept or reject her job offer. In the condition that combined both outcome accountability and process accountability, participants were given instructions that combined those given to participants within the outcome and process conditions. In the no accountability condition, an effort was made to ensure that participants did not feel answerable to a superior. This was accomplished in the opening statement that emphasized to participants that they were temporary workers of the student organization, and were substituting for another employee who had taken the day off. Specifically, participants were told that you are temporary, which means that you are only in this organization for the day, and you already have another, more permanent, job lined up in another organization. This organization, in which you are temporary, is in the process of hiring a new worker, and the person who is in charge of the interview process is sick, and you are substituting for them for the day. Because you are only a substitute, you will never have to see any members of UI Computing Assist after today. They were also told that, there is a chance that you can earn real money for participating in this study. For being in this study, you will receive a ticket that will be entered in a lottery. One in every 10 tickets will be randomly drawn from this lottery, and will be converted into a cash prize of $20 for each winner. Our lab assistant may eventually view the tape of your interview for purposes of data coding regarding communication technology, but it will not be evaluated in any way in terms of your interview behavior or the outcome of the interview.

Accountability and Emotional Labor 19 In each condition, these instructions were reinforced with a reminder inserted after the instructions on the interview process, and the script on what to say during it. Experimental measures. The manipulation check for process accountability was assessed with 4 inter-related items that were averaged (Cronbachs = .74). Items included: I felt I had to justify my behavior during the interview, I felt pressured to maintain a pleasant conversation, I felt responsible for the quality of the interview, and I felt my organizations image rested on my shoulders. For outcome accountability, the manipulation check was measured by 4 inter-related items which were also averaged (Cronbachs = .79). Items included: I felt pressured to ensure that the interviewee was hired, I felt responsible for the outcome of the interview, I was concerned about recruiting the interviewee, and I felt responsible about whether the interviewee was hired or not. Anchors ranged from 1 (definitely did not feel this way) to 7 (definitely felt this way). In order to assess emotional labor, we felt it was important to develop a measure that, from the participants perspective, tapped into the difficulty involved in generating emotions and suppressing emotions that were consistent with display rules. Self-reported emotional labor was measured through responses to 9 inter-related items. This scale measured the extent to which participants felt they engaged in emotional labor during the course of their interviews, and in particular, the level of ease or difficulty associated with managing their emotions. Two subscales that measured the display of positive emotions, and the suppression of negative emotions composed the larger scale. The items measuring the display of positive emotions included: I tried to display enthusiasm, I tried to act cheerful and sociable, I tried to act excited, I tried to be interested and

Accountability and Emotional Labor 20 attentive to the other person, and I tried hard to do a good interview. (Cronbachs = .71). The suppression of negative emotions was measured by responses to the following 4 items: I tried to suppress how upset or distressed I felt, I tried to suppress the anger and contempt that I felt, I tried to pretend I was not upset or distressed, and I tried to pretend I was not angry or feeling contempt. (Cronbachs = .87). Participants indicated their responses with anchors that ranged from 1 (very easy) to 7 (very difficult). For each subscale, the emotional labor items were averaged for an overall score, with higher mean scores indicating higher levels of displayed positive emotion, and suppressed negative emotion. Participants were also asked their true attitudes (i.e., their genuine feelings) toward their interviewee with 5 inter-related items that were averaged. Items included: indicate how competent you thought the person interviewed was, indicate your overall impression of the person interviewed, indicate how much you would like to work with the person interviewed, indicate how likeable you felt the person interviewed was, and indicate how much you want the person interviewed to join your organization. (Cronbachs = .69). These items sought to understand the participants true feelings and attitudes toward the person they interviewed. Finally, participants were also asked the extent to which they were convinced they were speaking to a live subject during their interview. Only 14% of the participants were completely suspicious of the claim that the applicant whom they interviewed was live. These participants were evenly distributed across the four accountability conditions. A one-way analysis of variance measuring the effect suspicion had on observed and self-

Accountability and Emotional Labor 21 reported emotional labor indicated that participants did not differ from each other in the extent of emotional labor they engaged in as a function of their level of suspicion. Observer measures. Recall that each interview was video taped for subsequent viewing and evaluation. Two independent raters who were blind to the experimental conditions, watched the videotapes of the participants, and rated their emotional labor using a scale that measured the degree to which the raters perceived emotion management in the participants during their interviews. This scale was divided into two subscales that measured the expression of positive emotions, and the suppression of negative emotions. That is, this 7 point scale measured the extent to which the raters felt participants faked or suppressed their emotions during their interview. 9 inter-related items composed the subscale measuring the expression of positive emotions. Items included: how much did the participant smile?, how much eye-contact did the participant make?, did the participant praise the applicant? did the participant speak for the entire two minutes?, did the participant use a calm tone of voice?, did the participant show enthusiasm?, was the participant attentive?, did the participant express hope that the applicant would join?, and did the participant apologize for any insults taken by the applicant? (Cronbachs = .78). 5 inter-related items composed the subscale measuring the suppression of negative emotions. Items included: did the participant suppress their anger?, did the participant suppress their contempt?, did the participant suppress their sarcasm?, did the participant suppress a sharp tone of voice? did the participant hide irritation? (Cronbachs = .97). Anchors ranged from 1 (not at all) to 7 (a lot). Furthermore, the raters also made an overall assessment of the amount of

Accountability and Emotional Labor 22 emotional labor they felt participants engaged in during their interview. Table 2 presents descriptive statistics and correlations for all variables. Results Manipulation Checks Two 2 (process: present, absent) x 2 (outcome: present, absent) analyses of variance were conducted on the process and outcome accountability scales respectively to determine if participants were aware of the type of accountability they had been manipulated by. There was a main effect for process accountability, F(1, 84) = 8.77, p< . 01, (Mprocess = 5.3; Mno process = 4.6) indicating that process accountable individuals felt more accountable for the process of their interview than individuals not accountable at all. While differences between outcome accountability, and no accountability were not significant, they were in the expected direction on the outcome accountability index, (Moutcome = 4.9; Mno outcome = 4.5), that is, individuals accountable at the outcome level felt more accountable than individuals not accountable at all. Observer Analysis Based on the videotapes made of the participants, independent observers made ratings on the display of positive emotions, the suppression of negative emotions, and overall emotional labor, as exhibited by the participants. A 2 (process: present, absent) x 2 (outcome: present, absent) analysis of variance examining observed differences in the display of positive emotions across the accountability conditions, was conducted. In support of Hypothesis 1., there was a main effect for process accountability. Further examination using Tukeys post hoc comparisons found that process accountability participants displayed significantly more positive emotions than no accountability

Accountability and Emotional Labor 23 participants, F(1,84) = 5.80, p< .02, 2 = .06. As Figure 1 illustrates, those accountable at the process level displayed significantly more positive emotions than no accountability individuals. There was no main effect of outcome accountability. In support of Hypothesis 2., the process by outcome accountability interaction was significant, F(1, 84) = 14.27, p< .001, 2 = .15. Individuals on whom process and outcome accountability were both emphasized displayed fewer outward positive emotions than process accountability individuals (Figure 1). Figures 2 and 3 show that individuals accountable at both the process and outcome levels similarly suppressed fewer negative emotions than process accountability participants, F (1, 84) = 5.15, p< .04, 2 = .05, and from an overall standpoint were also judged by independent raters as exhibiting less emotional labor than process accountability individuals, F (1, 84) = 7.40, p< .009, 2 = .08. Outcome accountability attenuated the impact of process accountability on emotional labor. No other main effects or interactions were significant. Self Report Data After conducting their interviews, participants responded to questions that assessed their display of positive emotions, suppression of negative emotions, overall level of emotional labor, and their true attitudes toward the unpleasant person they interviewed. Results based on self-reported measures of emotional labor were not significantly different across the various accountability conditions. Participants accountable at the process, outcome, combined process and outcome levels, and no accountability individuals, did not significantly differ from each other on self-reported

Accountability and Emotional Labor 24 positive emotions, the suppression of negative emotions, and the overall level of emotional labor. The nonsignificant self-report findings of emotional labor could imply that the participants did not feel as if the interview taxed their emotional selves. However, it is more likely that for the most part they felt that their interview was indeed emotionally laborious, as evidenced by the two-way analysis of variance on the expression of true attitudes that was not significant. In addition to the analyses conducted on faked positive emotions, and suppressed negative emotion, a 2 (process: present, absent) x 2 (outcome: present, absent) analysis of variance was also conducted on the measure of true attitudes to see if participants differed in their attitudes and feelings toward the person they interviewed. Ratings were consistently low across the four accountability conditions, and were not significantly different from each other (Mprocess = 2.4, Mno process = 2.5; Moutcome = 2.6, Mno outcome = 2.7), thus indicating that most participants found the interviewee unlikable and unpleasant. In fact, the low mean levels of positive genuine feelings and attitudes that were reported by most participants suggests that at all levels of accountability, participants felt that the person whom they interviewed was clearly negative, and unlikable. From a self-reported standpoint, then, emotional labor was certainly engaged in, just not significantly more or less so by the four groups of participants. Discussion The purpose of this research was to examine whether two different types of normenforcing mechanisms process accountability and outcome accountability affect emotional labor. Based on the decision-making and negotiation literatures that document

Accountability and Emotional Labor 25 the efficacy of process accountability in achieving optimal outcomes, we expected that process accountability would principally drive emotion management. We expected that outcome accountability would moderate the impact of process accountability and would lessen the impact of process accountability on emotional labor. The data supported both expectations. We defined emotional labor as two separate, though related, components; the expression or display of positive emotions, and the suppression of negative emotions. We used both observer ratings and self-reports to determine the level of emotional labor participants engaged in. Within the framework of a simulated organization, participants served as employees and conducted interviews with a prospective employee (a confederate trained to project negative affect) whom they were given to understand had to decide about whether or not to accept her job offer. Through the use of manipulations and incentives in a controlled setting, we isolated the mechanisms that we believed drove the expression of emotional labor. Process accountability participants were made to feel answerable to their supervisor for the means with which they conducted their interviews, and were told that their entry into a lottery would be contingent upon the quality of the interview and not on the outcome of the interview. Outcome accountability participants were informed that their entry into the lottery would be based on the end result of the interview, that is, whether or not the applicant accepted her job offer. Comparisons between the accountability conditions on the three observed measures of positive emotional expression, negative emotional suppression, and overall emotional labor showed findings that were consistent with our hypotheses. In keeping with the process facilitating effect, the principle type of accountability underlying

Accountability and Emotional Labor 26 emotion management remained process accountability. As predicted, we found that when participants were made to recognize that they would be held accountable for the quality of their professional interactions, that is, they would be evaluated by their supervisor on the basis of how well they maintained their social decorum, they made the extra effort to put up a good show, and keep up their positive faade, more so than when they were not made accountable at all. This particular finding was obtained through the use of observer ratings, one that suggests that from an outward perspective, emotion management appeared more pronounced in those individuals who had been tutored with the instruction to maintain an outward calm than in those who were not given this instruction. Process accountability clearly affected the display of positive affect as opposed to actual affect in participants. It could be argued that participants accountable at the process level became more invested and involved in their jobs and that alone made them feel more positively about their interview task, and so when it came time to interact with the confederate, these accountable participants were able to do so more positively and cheerfully than unaccountable participants. However, comparisons between the four accountability groups on the measure of self-reported true attitudes indicated that there were no significant differences between the four conditions; in general, regardless of the condition they were in, participants felt negatively about the person they interviewed as evidenced by the low mean ratings they provided on their attitudes toward her. These individuals were evenly distributed across the four accountability conditions. Such findings stood in contrast to those obtained through the use of observer ratings. Independent raters who watched the videotapes made of the interviews consistently gave

Accountability and Emotional Labor 27 higher ratings to process accountable participants than to unaccountable participants on the measures of the display of positive emotions, and overall emotional labor. This clearly suggests that process accountability affects the display of, rather than the actual, positive emotions, and as such it serves as an underlying mechanism driving emotional labor. We also expected an interaction between process accountability and outcome accountability. We expected that outcome accountability would dilute the effect of process accountability on emotional labor. Consistent with the outcome rigidity effect, outcome accountability was in fact the debilitating force that undermined the effect of process accountability a finding observed when it interacted with process accountability. Participants accountable at the outcome as well as process levels (both accountability) were rated as expressing fewer positive emotions and suppressing fewer negative emotions than process accountability participants. The findings suggest that the outcome rigidity effect is one that exerts no tangible impact on workers emotional labor at best, and at worst, impairs the benefit accrued from process accountability when it is made to combine with it. It is important to note that the differential effects of process and outcome accountability can not be explained by a difference in motivation. In all conditions of the experiment, there was a promise of a reward, and the reward was held constant. There is no reason to believe that the participants in the 4 conditions of the experiment had any less expected value for the reward than in any other condition, or that the intrinsic engagement with the task varied across conditions with varying effort (cf. Lerner & Tetlock, 1999).

Accountability and Emotional Labor 28 The results of this study fit with conclusions from the decision and negotiation literatures. Process accountability asks that individuals focus on the means toward an end; emotional labor suggests what that means might be. Process accountability encourages flexibility in considering alternative lines of action; emotional labor emphasizes public equanimity in workplace incumbents. It is not a big inferential leap to suggest that the two constructs work hand-in-hand in promoting positive emotions, and in suppressing negative emotions. Moreover, when one considers the detrimental effects of outcome accountability on decision-making, it comes as no further surprise that this construct should not be an antecedent of emotional labor. Indeed, based on the results of this study, it appears that outcome accountability serves to hamper the performance of emotional labor as it might promote rigidity in behavior, pressure in having to obtain a certain outcome, commitment to a prior course of action, and consequently, the antithesis of what employers hope for their employees to attain. From a theoretical viewpoint, the findings of this study support previous findings on process accountability as an antecedent underlying several outcomes. Within the context of emotional labor, process accountability emerged as the sole and arguably the most central factor driving emotional expression and suppression. Perhaps the greatest contribution that this study provides to the literature on emotional labor lies in its identification of a construct process accountability that was found to act as an antecedent to emotional labor. One of the strengths of this study lay in the controlled nature of the study. By conducting an experiment on process and outcome accountability on emotional labor, we were able to pinpoint with certainty the mechanism driving emotional expression in

Accountability and Emotional Labor 29 participants. Similarly, the use of multiple measures of evaluation, observer ratings as well as self-reports, provided two distinct ways of assessing emotional labor. From a practical (organizational) standpoint, observer reports of emotional labor might be considered the more important of the two as employers seek to cultivate outwardly appropriate emotional expression in their employees. Visible manifestations of emotion work are, after all, what customers, interviewers, and other observers see and file away for future reference. However, if emotional labor leads to burnout and exhaustion as research demonstrates it has been known to do (Grandey, 2000), then organizations might consider taking measures to ameliorate the amount of emotional labor employees perform so as to reduce the negative impact that it might have in the long run. More research needs to be done on which of the two measures of emotional labor self-reports or observer reports leads to burnout and/or exhaustion. This research added to the body of literature on accountability and emotional labor by testing their possible relationship in a controlled setting. To our knowledge, these two areas had not been examined together, nor had the causal nature of their relationship been empirically tested. Indeed, we viewed accountability as a neglected predictor of emotional labor as previous research had overlooked its potential impact on emotional labor. By examining their relationship between the two constructs within the precincts of the laboratory, we hoped to shed light on a possible antecedent of emotional labor, and hence, add to the literature in these important areas. Research is never a flawless enterprise, and laboratory research is subject to oft heard criticisms (see Sears, 1986). One of the shortcomings of this experiment was that several participants remained unconvinced that they were interviewing a live person. This

Accountability and Emotional Labor 30 knowledge might have reduced the emotional burden of behaving in an emotionally appropriate way. Even so, tests conducted on the differences between suspicious and unsuspicious participants showed that they did not differ from each other in the amount of emotional labor they engaged in. Secondly, the manipulation check on outcome accountability indicated that it was not found to be statistically significantly different from no accountability. It could be that this lack of significance might have contributed toward the absence of significant differences between outcome accountability participants and no accountability individuals. However, the finding of a significant interaction between outcome and process accountability on the degree of emotional labor observed between both accountability and process accountability participants seems to suggest that outcome accountability did serve to lower the amount of emotion work participants manifested. As with any experiment conducted within the lab, the greatest strength of the study is also its greatest limitation. Given that we used college students with minimal working experience as our participants, it could be argued that they did not serve as a representative sample from which to draw generalizable conclusions. However, as evidenced from their self-reports of genuine feelings, it was clear that the experiment was successful in capturing the same internal negative sentiments that employees experience during the trials of their real-time jobs, and as such conjured up some type of emotional labor reflective of that which is engaged in at work. Moreover, participants in this experiment were given an incentive (prize money) in the same way that employees in work settings are entitled to incentives, thus providing them with yet another reason to engage in emotional labor.

Accountability and Emotional Labor 31 Directions for future research One of the goals in the emotional labor literature is to create a comprehensive model that outlines the antecedents and consequences of emotional labor. More research needs to be done on outcome accountability, and in particular, on the mechanisms that cause it to have a detrimental effect on emotional labor. In the negotiation literature, for example, Ben-Yoav & Pruitt (1984) found that expectation of cooperation in future interactions interacted positively with accountability on negotiation outcomes. Perhaps the same dynamic holds with regard to emotional labor, and future research might consider manipulating both accountability and expectation of cooperation in order to see their interacting effect on emotional labor. Emotional labor is fundamentally a construct that has been identified and known to exist in the workplace. Therefore, research would stand to gain from adopting other practices of obtaining data (i.e., field research) and surveying employees in real organization. If a model of emotional labor is to be drawn, a number of questions on the consequence side remain unanswered: is emotional labor tantamount to good performance? Does emotional labor lead to performance deficits in unrelated work activities? Can it spill over from interactions with customers to coworkers and supervisors as well? Does emotional labor fade with time as employees become more rehearsed with it? How do past episodes with unpleasant individuals influence behavior toward other individuals? More lab and field research needs to be done to identify the missing pieces needed to create a clearer picture. This research expands upon work done in the emotional labor domain by identifying antecedents that encourage and impede the performance of emotional labor.

Accountability and Emotional Labor 32 The double-edged sword of accountability was highlighted in the mixed effects accountability had on participants emotion management depending on whether they were exposed to process accountability or to outcome accountability. By understanding how multiple forms of accountability impact emotional labor we will be in a better position to develop theories of emotional labor and better able to apply our findings to the practical world of organizational productivity.

Accountability and Emotional Labor 33 References Ben-Yoav, O., Pruitt, D. G. (1984). Accountability to constituents: A Two-Edged Sword. Organizational Behavior and Human Performance, 34, 283-295. Carnevale, P.J. (1985). Accountability of group representatives and intergroup relations. In E.J. Lawler (Ed.), Advances in group processes: Theory and research, Vol. 2. Greenwich, Connecticut: JAI Press. Carnevale, P. J., & Pruitt, D. G. (1992). Negotiation and mediation. Annual Review of Psychology, 43, 531-582. Darwin, C. (1872). The expression of the emotions in man and animals. London: John Murray. Doney, P. M., & Armstrong, G. M. (1996). Effects of accountability on symbolic information search and information analysis by organizational buyers. Journal of Academy Marketing Science, 24, 57-65. Ekman, P. (1973). Darwin and facial expression. New York, NY: Academic Press. Gelfand, M. J., & Realo, A. (1999). Individualism-collectivism and accountability in intergroup negotiations. Journal of Applied Psychology, 84, 721-736. Grandey, A. A. (2000). Emotion regulation in the workplace: A new way to conceptualize emotional labor. Journal of Occupational Health Psychology, 5, 95-110. Hochschild, A. (1983). The managed heart. Berkeley, CA: University of California Press.

Accountability and Emotional Labor 34 Kramer, R. M. (1995). In dubious battle: Heightened accountability, dysphoric cognition, and self-defeating bargaining behavior. In R. Kramer & D. Messick (Eds.), Negotiation as a social process, pp. 95-120. Beverly-Hills: Sage. Lerner, J. S., & Tetlock, P. E. (1999). Accounting for the effects of accountability. Psychological Bulletin, 125, 255-275. Mann, S. (1999a). Emotion at work: To what extent are we expressing, suppressing, or faking it? European Journal of Work and Organizational Psychology, 8, 347-369. Mano, H. (1992). Judgments under distress. Organizational Behavior and Human Decision Processes, 52, 216-245. Morris, J. A. & Feldman, D. C. (1996). The dimensions, antecedents and consequences of emotional labor. Academy of Management Review, 21, 986-1010. Morris, J. A., & Feldman, D. C. (1997). Managing emotions in the workplace. Journal of Managerial Issues, 9, 257-274. OConnor, K. M. (1997). Groups and solos in context: The effects of accountability on team negotiation. Organizational Behavior and Human Decision Processes, 72, 384-407. 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. (1990). Busy stores and demanding customers: How do they affect the display of positive emotion? Academy of Management Journal, 33, 623-637.

Accountability and Emotional Labor 35 Schlenker, B. R. (1986). Personal accountability: Challenges and impediments in the quest for excellence. San Diego, CA: Naval Personnel Research and Development Center. Sears, D. O. (1986). College sophomores in the laboratory: Influences of a narrow data base on social psychologys view of human nature. Journal of Personality and Social Psychology, 51, 515-530. Siegal-Jacobs, K., & Yates, J. F. (1996). Effects of procedural and outcome accountability on judgment quality. Organizational Behavior and Human Decision Processes, 65, 1-17. Simonson, I., & Staw, B. M. (1992). Deescalation strategies: A comparison of techniques for reducing commitment to losing courses of action. Journal of Applied Psychology, 77, 419-426. Tetlock, P. E. (1985). Accountability: The neglected social context of judgment and choice. In B. Staw & L. Cummings (Eds.), Research in organizational behavior (Vol. 7, pp. 297-332). Greenwich, CT: JAI Press. Tetlock, P.E. (1992). The impact of accountability on judgment and choice: Toward a social contingency model. In M.P. Zanna (Ed.), Advances in experimental social psychology, pp. 331-376. (Vol. 25). New York: Academic Press. Tews, M. J., & Glomb, T. M. (in revision). Expressing and suppressing: A conceptualization and operationalization of emotional labor. Journal of Management.

Accountability and Emotional Labor 36 Appendix A Interview Script used by Participants to Recruit the Interviewee In the following interview, please bring up the following points. You may refer to this sheet while you are speaking, but try not to read directly from it.

1. Talk about UI Computing Assists excellent computer facility lots of computers, with high speed Internet access. We have the fastest Internet speed in the Midwest!

2. Friendly environment with an excellent support system. We train you to be friendly and efficient. Our students never leave the facility dissatisfied. 3. Has ongoing workshops and training for new and existing employees to improve their skills. We teach you that theres always room for self-improvement. Even a die-hard computer geek has to keep up with the pace of technology. 4. Students who come to the organization for questions are intelligent and motivated. Students never leave the facility dissatisfied. We address their every problem!

5. This is the highest paying job on campus for students. You will be one of the richest students on campus!

6. Lots of benefits can access the labs facilities any time of the day or night. Employees can receive upgrades for personal use at discounted prices. Youll be living in the lap of Technological luxury!

Accountability and Emotional Labor 37 Appendix B Excerpts from the Confederates Tape Yeah, well, Im not really sure I can buy all the stuff that youre trying to tell me. You say you have the fastest internet speed in the Midwest? Thats a pretty lofty claim! I have a better computer system in my own apartment! You say Ill be one of the richest students on campus? Im sure I can get a better job working somewhere else because I am one of the best workers. Then you call me a computer geek! Are you trying to insult me to get me to take your job? I dont think thats the best way to go.

Accountability and Emotional Labor 38 Table 1 Sequence of Steps used in the Experiment Step Two participants arrived Given folder of materials Description Each participant was directed to their own room. Participants read that they were employees of a simulated organization, UI Computing Assist. They were told to interview and recruit another subject who had been offered a job with them. They were also informed that interview will take place over a TV 1st participant directed to interview room. 2nd participant waited in their room. Experimenters role Interview monitor, and that it would be video taped. The first participant sat at a table in front of a TV monitor in the interview room. The table was equipped with a timer, and a taped script for the participant to follow during the interview. Meanwhile, the second participant waited alone in their room until called. The experimenter orally reviewed the instructions again, and then pressed the record button on the camera. After starting the timer, the participant spoke for two minutes until the timer went off. Then, in a control room, an experimental aide pressed the start button on a VCR so that the interviewees (confederates) taped image appeared on the TV in the interview room. The participant viewed the interviewee until she stopped Step Interview (table continues) Description speaking, at which point the aide pressed the stop button. The participant responded to what was said for another two minutes. The recording was stopped when the timer went off. The first

Accountability and Emotional Labor 39 participant went back to their room. The experimental aide 2 participant
nd

rewound the confederates tape. The second participant was called to the interview room after the first one left, and followed the same sequence of steps in the

Post interview

interview room as was used by the first participant. Back in their room, each participant filled out a questionnaire that asked about the interview, and left after completing it.

Accountability and Emotional Labor 40 Table 2 Means, Standard Deviations, and Correlations of all Measures Subscale M Self report 1. Display positive emotion 2. Suppress negative emotion 3. True attitudes 4. Level of suspicion Observed 5. Display positive emotion 6. Suppress negative emotion 4. 4 5. 1.0 1.4 5 1.5 2 -.0 9 -.0 5 -.1 0 .12 -.1 4 -.1 6 .07 .07 .15 -.0 6 -.0 2 .04 -.55 .58 .54 -.94 .58 .94 -3. 9 3. 4 2. 7 3. 7 1.0 1 1.3 3 .84 2.1 1 -.36 .04 -.0 7 .36 --.1 6 -.1 6 .04 -.1 6 -.31 -.0 7 -.1 6 .31 --.0 9 .12 .07 -.0 6 -.0 5 -.1 4 .07 -.0 2 -.1 0 -.1 6 .15 .04 SD 1 2 Correlation 3 4 5 6 7

6 7. Emotional labor (composite) 5. 8

Accountability and Emotional Labor 41 Figure Captions Figure 1. Display of positive emotions as a function of process accountability and outcome accountability. Figure 2. Suppression of negative emotions as a function of process accountability and outcome accountability. Figure 3. Emotional labor (composite index) as a function of process accountability and outcome accountability.

Accountability and Emotional Labor 42 Figure 1.

5.5

Display of Positive Emotions

4.5

No Outcome Accountability Outcome Accountability

3.5

3 No Process Accountability Process Accountability

Accountability and Emotional Labor 43 Figure 2.

6.5

6 Suppression of Negative Emotions

5.5 No Outcome Accountability Outcome Accountability

4.5

3.5

3 No Process Accountability Process Accountability

Accountability and Emotional Labor 44 Figure 3.

6.5

6 Overall Emotional Labor No Outcome Accountability Outcome Accountability

5.5

4.5

3.5

3 No Process Accountability Process Accountability

You might also like